Methods and means to modify fiber strength in fiber-producing plants

Field of the invention

[1] This invention relates to the field of agriculture, more specifically to the use of molecular biology techniques to alter fiber-producing plants, particularly cotton plants, and/or accelerate breeding of such fiber-producing plants. Methods and means are provided to alter fiber qualities, such as increasing fiber strength. Methods are also provided to identify molecular markers associated with fiber strength in a population of cotton varieties and related progenitor plants.

Background of the invention

[2] Cotton provides much of the high quality fiber for the textile industry. The modification of cotton fiber characteristics to better suit the requirements of the industry is a major effort in breeding by either classical methods or by genetically altering the genome of cotton plants.

[3] About 90% of cotton grown worldwide is Gossypium hirsutum L., whereas Gossypium barbadense accounts for about 8%. As in most flowering plants, cotton genomes are thought to have incurred one or more polyploidization events and to have evolved by the joining of divergent genomes in a common nucleus. The cotton commerce is dominated by improved forms of two "AD" allotetraploid species, Gossypium hirsutum L. and Gossypium barbadense L (both 2n=4x=52). Allotetraploid cottons are thought to have formed about 1-2 million years ago, in the New World, by hybridization between a maternal Old World "A" genome taxon resembling Gossypium herbaceum (2n=2x=26) and paternal New World "D" genome taxon resembling Gossypium raimondii or Gossypium gossypioides (both 2n=2x=26). Wild A genome diploid and AD allotetraploid Gossypium taxa produce spinnable fibers. One A genome diploid species, Gossypium arboreum (2n=2x=26), remains intensively bred and cultivated in Asia. Its close relative and possible Gossypium progenitor, the A genome diploid species G. herbaceum, also produces spinnable fiber. Although the seeds of D genome diploids are pubescent, none produce spinnable fibers. No taxa from the other recognized diploid Gossypium genomes

(B, C, E, F, G and K) have been domesticated. Intense directional selection by humans has consistently produced AD allotetraploid cottons that have superior yield and/or quality characteristics compared to the A genome diploid cultivars. Selective breeding of G. hirsutum (AADD; "Upland" cotton) has emphasized maximum yield, whereas G. barbadense (AADD; "Sea Island", "Pima", or "Egyptian" cotton) is prized for its fibers of superior length, strength, and fineness (Jiang et al., 1998, Proc Natl Acad Sci U S A. 95(8): 4419-4424).

[4] A cotton fiber is a single cell that initiates from the epidermis of the outer integument of the ovules, at or just prior to anthesis. Thereafter, the fibers elongate rapidly for about 3 weeks before they switch to intensive secondary cell wall cellulose synthesis. Fiber cells interconnect only to the underlying seed coat at their basal ends and influx of solute, water and other molecules occurs through either plasmodesmata or plasma membrane. Ruan et al. 2001 (Plant Cell 13: 47-63) demonstrated a transient closure of plasmodesmata during fiber elongation. Ruan et al. 2004 (Plant Physiology 136: 4104-4113) compared the duration of plasmodesmata closure among different cotton genotypes differing in fiber length and found a positive correlation between the duration of the plasmodesmata closure and fiber length. Furthermore, microscopic evidence was presented showing callose deposition and degradation at the fiber base, correlating with the timing of plasmodesmata closure and reopening. Expression of a endo-l,3-beta- glucanase gene in the fibers, allowing to degrade callose, correlated with the reopening of the plasmodesmata at the fiber base.

[5] W02005/017157 describes methods and means for modulating fiber length in fiber producing plants such as cotton by altering the fiber elongation phase. The fiber elongation phase may be increased or decreased by interfering with callose deposition in plasmodesmata at the base of the fiber cells.

[6] WO2008/083969 (claiming priority of European patent application EP 07000550) discloses isolated DNA molecules comprising a nucleotide sequence encoding cotton endo-l,3-beta-glucanases and fiber cell preferential promoter or promoter regions, as well

as methods for modifying the length of a fiber of a cotton plant using these sequences or promoters. WO2008/083969 also describes that the timing of expression of the A and D subgenome specific alleles of the fiber specific endo-l,3-beta-glucanase gene in Gossypium hirsutum is different. Whereas the onset of the expression of the D subgenome specific allele correlates with the end of the rapid elongation phase (about 14 to 17 days post-anthesis, hereinafter abbreviated "DPA"), onset of the expression of the A subgenome specific allele is delayed until the beginning of the late fiber maturation phase (about 35-40 DPA) depending on growth conditions.

[7] One fiber characteristic that is of special interest for the cotton industry is fiber strength. There is not only a high correlation between fiber strength and yarn strength, but also cotton with high fiber strength is more likely to withstand breakage during the manufacturing process.

[8] Fiber strength is, among many other textile properties of cotton fibers (e.g., fiber wall thickness or maturity, dyeability, extensibility ...), described to be directly dependent on the amount and properties (e.g., degree of polymerization, crystallite size, and microfibril orientation) of cellulose (Ramey, 1986, In: Mauney J.R. and Stewart J. McD. (eds.) Cotton Physiology. The Cotton Foundation, Memphis, TN, pp. 351-360; Triplett, 1993, In: Cellulosics: Pulp, Fibre, and Environmental Aspects. Ellis Horwood, Chichester, UK, pp. 135-140; Hsieh, 1999, In: Basra A.S. (ed.) Cotton Fibers: Developmental Biology, Quality Improvement, and Textile Processing. The Haworth Press, New York, pp. 137-166). Advances in the past decade, particularly using the model plant Arabidopsis (Arioli et al., 1998, Science 279(5351): 717-720), have led to a great increase in the knowledge of the proteins involved in cellulose synthesis. Despite this, there is still much to learn about cellulose synthesis, especially about how it is regulated at both transcriptional and post-transcriptional levels (Taylor, 2008, New Phytologist 178 (2) , 239-252).

[9] Typical primary fiber cell walls in G. hirsutum, which are about 0.5 μM thick and contain 20-25% cellulose along with pectin, xyloglucan, and protein (Meinert and

Delmer 1977, Plant Physiol 59:1088-1097), are synthesized during fiber elongation (Haigler, 2007, In: R.M. Brown, Jr. and LM. Saxena (eds.), Cellulose: Molecular and Structural Biology, 147-168, Springer.). Primary wall deposition proceeds alone until 14-17 DPA, then a transition phase with concurrent primary and secondary wall deposition occurs between 15-24 DPA (representing deposition of the "winding layer"), followed by predominantly secondary wall synthesis until at least 40 DPA. The first period of wall thickening (12-16 DPA) is accomplished by continued synthesis in the same proportions of primary wall components (Meinert and Delmer, 1977, supra), an observation that is consistent with increasing wall birefringence while the cellulose microfibrils remain transversely oriented (Seagull, 1986, Can J Bot 64:1373-1381). The secondary wall finally attains a thickness of 3-6 μM around the whole circumference of the fiber, becoming thinner only at the fiber tip. In G. barbadense, there is an overlap between primary and secondary wall deposition within each fiber rather than in the fiber population because the overlapping period is greatly prolonged, and 90% of secondary wall deposition is complete before elongation ceases (DeLanghe, 1986, In: Mauney J.R. and Stewart J. McD. (eds.) cotton Physiology. The Cotton Foundation, Memphis, TN, pp. 325-350). It is thought that elongation continues exclusively at the fiber tip as secondary wall is deposited over most of the cell surface.

[10] Maltby et al. (1979, Plant Physiol. 63, 1158-1164) describe that developing fibers of Gossypium hirsutum transiently synthesize 1,3-beta-D-glucan (callose) at the onset of secondary wall deposition followed by massive synthesis of cellulose. Meier et al. (1981, Nature 289: 821-822) describe that callose may be a probable intermediate in biosynthesis of cellulose of cotton fibers. DeLanghe (1986, supra) describes that callose may be required in cotton fiber secondary walls to provide a space for the crystallization and final orientation of cellulose microfibrils in the exoplasmic zone in the absence of typical matrix molecules.

[11] The inventions described hereinafter in the different embodiments, examples, figures and claims provide improved methods and means for modulating fiber strength. More specifically, the present invention describes how to increase fiber strength and at

the same time maintain a high fiber yield in plants. In particular, the invention describes how to increase fiber strength in cotton species selected for high yield, such as Gossypium hirsutum, by introgression of fiber strength determining genes from other cotton species selected for high fiber strength, such as Gossypium barbadense. Methods are also provided to identify molecular markers associated with fiber strength in a population of cotton varieties and related progenitor plants. The inventions described hereinafter also provide novel nucleic acid molecules encoding fiber-specific Gossypium glucanase proteins (GLUC J.I) and the proteins as such.

Summary of the invention

[12] The inventors identified a quantitative trait locus for fiber strength on chromosome A05 of Gossypium and found that Gossypium barbadense comprises an allele of this fiber strength locus that is superior to the allele of this QTL from Gossypium hirsutum, i.e. the presence of the Gossypium barbadense fiber strength allele in a Gossypium plant results in a higher fiber strength as compared to the fiber strength of a Gossypium plant comprising the Gossypium hirsutum fiber strength allele.

[13] Thus, in a first aspect, the present invention provides a non-naturally occurring Gossypium plant, and parts and progeny thereof, comprising at least one superior allele of a fiber strength locus on chromosome A05.

[14] In one embodiment, the plant is a plant from an A genome diploid Gossypium species, such as Gossypium herbaceum or Gossypium arboreum, or an AD genome allotetraploid Gossypium species, such as Gossypium hirsutum and Gossypium barbadense, and the superior fiber strength allele is derived from a different A or AD genome Gossypium species.

[15] In another embodiment, the plant is a Gossypium hirsutum, a Gossypium herbaceum or a Gossypium arboreum plant, preferably a Gossypium hirsutum plant, and the superior fiber strength allele is derived from Gossypium barbadense.

[16] In one aspect, the Gossypium barbadense fiber strength allele is located on chromosome A05 of Gossypium barbadense between AFLP marker P5M50-M126.7 and SSR marker CIR280. In another aspect, between AFLP marker P5M50-M 126.7 and SSR marker BNL3992. In still another aspect, between AFLP marker P5M50-M 126.7 and SSR marker CIR401c. In yet another aspect, is the LOD peak of the Gossypium barbadense fiber strength allele located between SSR marker NAU861 or the GLUC 1.1 gene and SSR marker CER401c. In a further aspect, is the LOD peak of the Gossypium barbadense fiber strength allele located at about O to 5 cM, more specifically at about 4.008 cM, from SSR marker NAU861 or the GLUC 1.1 gene, hi still a further aspect, is the LOD peak of the Gossypium barbadense fiber strength allele is located at about 0 to 12 cM, more specifically at about 10 cM, especially at about 10.52 cM, from SSR marker CIR401C

[17] In another aspect, the Gossypium barbadense fiber strength allele comprises at least one Gossypium barbadense ortholog of a nucleotide sequence comprised in the genomic DNA sequence spanning the Gossypium hirsutum GLUC 1.1 A gene represented in SEQ ID NO: 53.

[18] hi still another aspect, the Gossypium barbadense fiber strength allele comprises a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein. In one aspect, the Gossypium barbadense GLUC 1.1 gene is characterised by the presence of a T nucleotide at a nucleotide position corresponding to nucleotide position 712 of SEQ ID NO: 5. In a further aspect, the Gossypium barbadense GLUC 1.1 gene is located at about 0 to 5 cM, more specifically at about 4 cM, from the LOD peak of the Gossypium barbadense fiber strength allele, hi yet a further aspect, the Gossypium barbadense GLUC 1.1 gene is located at about 0 to 2 cM, at about 0 to 1 cM, more specifically at about 0.008 cM of the NAU861 marker.

[19] In yet another embodiment, the plant is a Gossypium hirsutum, Gossypium barbadense, a Gossypium herbaceum or a Gossypium arboreum plant, preferably a Gossypium hirsutum plant, and the superior fiber strength allele is derived from

Gossypium darwinii. In one aspect, the Gossypium darwinii fiber strength allele comprises a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein. In another aspect, the Gossypium darwinii GLUCl.1 gene is characterised by the presence of a T nucleotide at a nucleotide position corresponding to nucleotide position 761 of SEQ ID NO: 56.

[20] In still another embodiment, the plant is a Gossypium hirsutum, Gossypium barbadense or a Gossypium herbaceum plant, preferably a Gossypium hirsutum plant, and the superior fiber strength allele is derived from Gossypium arboreum. In one aspect, the Gossypium arboreum fiber strength allele comprises a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein. In another aspect, the Gossypium arboreum GLUC 1.1 gene is characterised by the abscence of a C nucleotide at a nucleotide position corresponding to the nucleotide position between position 327 and 328 of SEQ ID NO: 21.

[21] In a further embodiment, the callose content of the fibers is increased in the plant compared to the callose content of the fibers of an equivalent Gossypium plant that does not comprise the at least one superior allele of the fiber strength locus.

[22] In yet a further embodiment, the strength of the fibers is increased in the plant compared to the strength of the fibers of an equivalent Gossypium plant that does not comprise the at least one superior allele of the fiber strength locus. In one aspect, the strength of the fibers is on average between about 5% and about 10%, preferably about 7.5%, higher. In another aspect, the strength of the fibers is on average between about 1.6 g/tex and about 3.3 g/tex, preferably about 2.5 g/tex, higher. In still another aspect, the strength of the fibers is on average between about 34.6 g/tex and about 36.3 g/tex, preferably about 35.5 g/tex.

[23] In another embodiment, the plant is a Gossypium hirsutum plant homozygous for the Gossypium barbadense fiber strength allele.

[24] In still another embodiment, the invention provides a fiber obtainable from the plant of any one of paragraphs 13 to 23.

[25] In a further embodiment, the invention provides a method of identifying a Gossypium barbadense allele of a fiber strength locus on chromosome A05 in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, comprising the step of determining the presence of a Gossypium barbadense allele of a marker linked to the fiber strength locus in the genomic DNA of the plant selected from the group consisting of: AFLP marker P5M50-M126.7, SSR marker CIR280, SSR marker BNL3992, SSR marker CIR401c, SSR marker NAU861, a polymorphic site in an ortholog of a nucleotide sequence comprised in the genomic DNA sequence spanning a Gossypium hirsutum GLUC 1.1 A gene represented in SEQ ID NO: 53 of the plant; and a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene of the plant, such as SNP marker GLUC1.1A-SNP2 located at a nucleotide position corresponding to nucleotide position 418 to 428 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP3 located at a nucleotide position corresponding to nucleotide position 573 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP5 located at a nucleotide position corresponding to nucleotide position 712 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP6 located at a nucleotide position corresponding to nucleotide position 864 in SEQ ID NO: 5 or SNP marker GLUC1.1A-SNP8 located at a nucleotide position corresponding to nucleotide position 832 in SEQ ID NO: 5.

[26] In a particular aspect, the Gossypium barbadense allele of AFLP marker P5M50- M 126.7 is detected by amplification of a DNA fragment of about 126.7 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 43 and 44, respectively; the Gossypium barbadense allele of SSR marker CIR280 is detected by amplification of a DNA fragment of about 205 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 51 and 52, respectively; the Gossypium barbadense allele of SSR marker BNL3992 is detected by amplification of a DNA fragment of about 140 bp to about 145 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 49 and 50, respectively; the Gossypium barbadense allele of SSR marker CIR401c is

detected by amplification of a DNA fragment of about 245 to about 250 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 47 and 48, respectively; the Gossypium barbadense allele of SSR marker NAU861 is detected by amplification of a DNA fragment of about 215 bp to about 220 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 45 and 46, respectively; the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP2 is detected by detecting a CTCATCAAA nucleotide sequence at a position corresponding to the position of SNP marker GLUC1.1A-SNP2 or by amplification of a DNA fragment of about 143 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively; the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP3 is detected by detecting a C nucleotide at a position corresponding to the position of SNP marker GLUC 1.1 A- SNP3; the Gossypium barbadense allele of SNP marker GLUCl.1A-SNP5 is detected by detecting a T nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP5; the Gossypium barbadense allele of SNP marker GLUCl.1A-SNP6 is detected by detecting an A nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP6; the Gossypium barbadense allele of SNP marker GLUCl. IA- SNP8 is detected by detecting a C nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP8.

[27] In a further embodiment, the invention provides a method of identifying a Gossypium darwinii allele of a fiber strength locus on chromosome A05 in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, comprising the step of determining the presence of a Gossypium darwinii specific polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUC 1.1 A gene of SEQ ID NO: 56, such as SNP marker GLUC1.1A-SNP2 located at a nucleotide position corresponding to nucleotide position 476 to 477 in SEQ ID NO: 56, SNP marker GLUCl.1A-SNP3 located at a nucleotide position corresponding to nucleotide position 622 in SEQ ID NO: 56, SNP marker GLUC1.1A-SNP5 located at a nucleotide position corresponding to nucleotide position 761 in SEQ ID NO: 56, SNP marker GLUC1.1A-SNP6 located at a nucleotide position corresponding to nucleotide position 913 in SEQ ID NO: 56 or SNP

marker GLUC1.1A-SNP8 located at a nucleotide position corresponding to nucleotide position 881 in SEQ ID NO: 56.

[28] In a particular aspect, the Gossypium darwinii allele of SNP marker GLUC 1.1 A- SNP2 is detected by detecting a CTCATCAAA nucleotide sequence at a position corresponding to the position of SNP marker GLUC1.1A-SNP2 or by amplification of a DNA fragment of about 143 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively; the Gossypium darwinii allele of SNP marker GLUCl.1A-SNP3 is detected by detecting a C nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP3; the Gossypium darwinii allele of SNP marker GLUC1.1A-SNP5 is detected by detecting a T nucleotide at a position corresponding to the position of SNP marker GLUCl.1A-SNP5; the Gossypium darwinii allele of SNP marker GLUC1.1A-SNP6 is detected by detecting an A nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP6, and the Gossypium darwinii allele of SNP marker GLUCl.1A-SNP8 is detected by detecting a G nucleotide at a position corresponding to the position of SNP marker GLUCl.1A-SNP8.

[29] In a further embodiment, the invention provides a method of identifying a Gossypium arboreum allele of a fiber strength locus on chromosome A05 in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, comprising the step of determining the presence of a Gossypium arboreum specific polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUC 1.1 A gene of SEQ ED NO: 21, such as SNP marker GLUC1.1A-SNP7 located at a nucleotide position corresponding to a nucleotide position between nucleotide position 327 and 328 in SEQ ID NO: 21. In a particular aspect, the Gossypium arboreum allele of SNP marker GLUC1.1A-SNP7 is detected by detecting the absence of a C nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP7.

[30] In a further embodiment, the invention provides a method of distinguishing a Gossypium barbadense allele of a fiber strength locus on chromosome A05 from a

Gossypium hirsutum allele of the fiber strength locus in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, comprising the step of determining the presence of Gossypium barbadense alleles and/or Gossypium hirsutum alleles of markers linked to the fiber strength locus in the genomic DNA of the plant selected from the group consisting of: AFLP marker P5M50-M126.7, SSR marker CER280, SSR marker BNL3992, SSR marker CIR401, SSR marker NAU861; a polymorphic site in an ortholog of a nucleotide sequence comprised in the genomic DNA sequence spanning the Gossypium hirsutum GLUC 1.1 A gene represented in SEQ ID NO: 53 of the plant; and a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant, such as SNP marker GLUC1.1A-SNP2 located at a nucleotide position corresponding to nucleotide position 418 to 428 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP3 located at a nucleotide position corresponding to nucleotide position 573 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP5 located at a nucleotide position corresponding to nucleotide position 712 in SEQ ID NO: 5, SNP marker GLUC 1.1 A- SNP6 located at a nucleotide position corresponding to nucleotide position 864 in SEQ ID NO: 5 or SNP marker GLUCl.1A-SNP8 located at a nucleotide position corresponding to nucleotide position 832 in SEQ ID NO: 5.

[31] In a particular aspect, the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of AFLP marker P5M50-M 126.7 by amplification of, respectively, no DNA fragment and a DNA fragment of about 126.7 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 43 and 44, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SSR marker CIR280 by amplification of, respectively, no DNA fragment and a DNA fragment of about 205 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 51 and 52, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SSR marker BNL3992 by amplification of, respectively, two DNA fragments, one of about 160 bp to about 165 bp and one of about 85 bp to about 90 bp, and a DNA fragment of about 140 bp to about 145 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 49 and 50, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SSR

marker CIR401 by amplification of, respectively, a DNA fragment of about 255 bp (CIR401b) and a DNA fragment of about 245 bp to about 250 bp (CIR401c) with at least two primers comprising at their extreme 3' end SEQ ID NO: 47 and 48, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SSR marker NAU861 by amplification of, respectively, a DNA fragment of about 205 bp to about 210 bp and a DNA fragment of about 215 bp to about 220 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 45 and 46, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP2 by detecting, respectively, no nucleotide or a CTCATCAAA nucleotide sequence at a position corresponding to the position of SNP marker GLUC1.1A-SNP2, or by amplification of, respectively, a DNA fragment of about 134 bp and a DNA fragment of about 143 bp with at least two primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP3 by detecting, respectively, a G or a C nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP3; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP5 by detecting, respectively, a C or a T nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP5; the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP6 by detecting, respectively, a G or an A nucleotide at a position corresponding to the position of SNP marker GLUC1.1A-SNP6; and the Gossypium hirsutum allele is distinguished from the Gossypium barbadense allele of SNP marker GLUC1.1A-SNP8 by detecting, respectively, a G or a C nucleotide at a position corresponding to the position of SNP marker GLUCl.1A-SNP8.

[32] In another embodiment, the invention provides a method for generating and/or selecting a non-naturally occurring Gossypium plant, and parts and progeny thereof, comprising at least one superior allele of a fiber strength locus on chromosome A05, wherein the superior fiber strength allele is derived from Gossypium barbadense, comprising the steps of crossing a plant from an A genome diploid Gossypium species,

such as Gossypium herbaceum or Gossypium arboreum, or an AD genome allotetraploid Gossypium species, such as Gossypium hirsutum, with a Gossypium barbadense plant, and identifying the Gossypium barbadense fiber strength allele according to paragraph 25 or 26.

[33] In another embodiment, the invention provides a method for generating and/or selecting a non-naturally occurring Gossypium plant, and parts and progeny thereof, comprising at least one superior allele of a fiber strength locus on chromosome A05, wherein the superior fiber strength allele is derived from Gossypium darwinii, comprising the steps of crossing a plant from an A genome diploid Gossypium species, such as Gossypium herbaceum or Gossypium arboreum, or an AD genome allotetraploid Gossypium species, such as Gossypium hirsutum or Gossypium barbadense, with a Gossypium darwinii plant, and identifying the Gossypium darwinii fiber strength allele according to paragraph 27 or 28.

[34] In another embodiment, the invention provides a method for generating and/or selecting a non-naturally occurring Gossypium plant, and parts and progeny thereof, comprising at least one superior allele of a fiber strength locus on chromosome A05, wherein the superior fiber strength allele is derived from Gossypium arboreum, comprising the steps of crossing a plant from an A genome diploid Gossypium species, such as Gossypium herbaceum, or an AD genome allotetraploid Gossypium species, such as Gossypium hirsutum or Gossypium barbadense, with a Gossypium arboreum plant, and identifying the Gossypium arboreum fiber strength allele according to paragraph 29.

[35] hi still another embodiment, the invention provides a method for altering the callose content of a fiber in a Gossypium plant, particularly increasing the callose content of a fiber, comprising the steps of: introgressing a superior allele of the fiber strength locus on chromosome A05 in the Gossypium plant according to any one of paragraph 32 to 34, and selecting a plant with an altered callose content in its fibers, in particular an increased callose content.

[36] In yet another embodiment, the invention provides a method for altering the properties of a fiber in a Gossypium plant, particularly increasing the strength of a fiber, comprising the steps of: introgressing a superior allele of the fiber strength locus on chromosome A05 in the Gossypium plant according to any one of paragraph 32 to 34, selecting a plant with an altered fiber strength, in particular an increased fiber strength.

[37] In a further embodiment, the invention provides a kit for of identifying a Gossypium barbadense allele of a fiber strength locus on chromosome A05 or for distinguishing a Gossypium barbadense allele of a fiber strength locus on chromosome A05 from a Gossypium hirsutum allele of the fiber strength locus in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, comprising primers and/or probes for determining the presence of Gossypium barbadense alleles and/or Gossypium hirsutum alleles of markers linked to the fiber strength locus in the genomic DNA of the plant selected from the group consisting of: AFLP marker P5M50-M 126.7, SSR marker CIR280, SSR marker BNL3992, SSR marker CIR401, SSR marker NAU861 , a polymorphic site in an ortholog of a nucleotide sequence comprised in the genomic DNA sequence spanning the Gossypium hirsutum GLUC 1.1 A gene represented in SEQ ID NO: 53, and a polymorphic site in a nucleotide sequence of the GLUC 1.1 A gene in the genomic DNA of the plant, such as SNP marker GLUC1.1A-SNP2 located at a nucleotide position corresponding to nucleotide position 418 to 428 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP3 located at a nucleotide position corresponding to nucleotide position 573 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP5 located at a nucleotide position corresponding to nucleotide position 712 in SEQ ID NO: 5, SNP marker GLUC1.1A-SNP6 located at a nucleotide position corresponding to nucleotide position 864 in SEQ ID NO: 5 or SNP marker GLUC1.1A-SNP8 located at a nucleotide position corresponding to nucleotide position 832 in SEQ ID NO: 5.

[38] In one aspect, the kit comprises at least two primers and/or probes selected from the group consisting of: primers comprising at their extreme 3' end SEQ ID NO: 43 and 44, respectively; primers comprising at their extreme 3' end SEQ ID NO: 51 and 52, respectively; primers comprising at their extreme 3' end SEQ ID NO: 49 and 50,

respectively; primers comprising at their extreme 3' end SEQ ID NO: 47 and 48, respectively; primers comprising at their extreme 3' end SEQ ED NO: 45 and 46, respectively; primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively.

[39] The inventors have further found that the properties of fibers in cotton plants can be controlled by controlling the number of endo-l,3-beta-glucanase genes/alleles that are "functionally expressed", i.e. that result in functional (biologically active) endo-l,3-beta- glucanase protein (GLUC), in fibers during the secondary cell wall synthesis phase and the maturation phase, herein commonly referred to as fiber strength building phase, of fiber development. By abolishing the functional expression of a number of endo-l,3-beta- glucanase genes/alleles that are functionally expressed in fibers during the fiber strength building phase, in particular during the maturation phase, of fiber development, such as the A-subgenome specific endo-l,3-beta-glucanase gene in G. hirsutum, while maintaining the functional expression of a number of such endo-l,3-beta-glucanase genes/alleles, such as the D-subgenome specific endo-l,3-beta-glucanase gene in G. hirsutum, it is believed that the degradation of callose can be decreased to a level allowing a higher fiber strength, while maintaining a level of callose degradation sufficient to obtain an industrially relevant fiber length.

[40] Thus, in another aspect, the present invention provides a non-naturally occurring fiber-producing plant, and parts and progeny thereof, characterized in that the functional expression of at least one allele of at least one fiber-specific GLUC gene that is functionally expressed during the fiber strength building phase, in particular the fiber maturation phase, of fiber development is abolished. Such plants, and parts and progeny thereof, can be used for obtaining plants with modified callose content and/or modified fiber properties, in particular for obtaining fiber-producing plants with increased callose content in the fibers and/or increased fiber strength that preferably maintain an industrially relevant fiber length. As used herein, "plant part" includes anything derived from a plant of the invention, including plant parts such as cells, tissues, organs, seeds, fibers, seed fats or oils.

[41] In one embodiment, the GLUC gene is a GLUC 1.1 gene encoding a GLUC protein that has at least 90% sequence identity to SEQ ID NO: 4.

[42] In another embodiment, the plant is a Gossypium plant, wherein the GLUC gene is a GLUC 1.1 A gene encoding a GLUC protein that has at least 97% sequence identity to SEQ ID NO: 4 or a GLUC 1.1 D gene encoding a GLUC protein that has at least 97% sequence identity to SEQ ID NO: 10, preferably the GLUC 1.1 A gene.

[43] In still another embodiment, the plant is a Gossypium hirsutum plant.

[44] hi a further embodiment, the amount of functional GLUC protein is significantly reduced in fibers during the fiber strength building phase, in particular the fiber maturation phase, of fiber development in the plant compared to the amount of functional GLUC protein produced in fibers during the fiber strength building phase, in particular the fiber maturation phase, of fiber development in a plant in which the functional expression of the at least one GLUC allele is not abolished.

[45] hi still a further embodiment, the callose content is significantly increased in fibers of the plant compared to the callose content in fibers in a plant in which the functional expression of the at least one GLUC allele is not abolished.

[46] In yet a further embodiment, the strength of the fibers is significantly increased compared to the strength of the fibers in a plant in which the functional expression of the at least one GLUC allele is not abolished. In one aspect, the strength of the fibers is on average between about 5% and about 10%, preferably about 7.5%, higher. In another aspect, the strength of the fibers is on average between about 1.6 g/tex and about 3.3 g/tex, preferably about 2.5 g/tex, higher. In still another aspect, the strength of the fibers is on average between about 34.6 g/tex and about 36.3 g/tex.

[47] In still a further embodiment, the plant is a Gossypium hirsutum plant characterized in that the functional expression of at least two alleles of at least one fiber- specific GLUC gene is abolished.

[48] In another embodiment, the present invention provides a fiber obtainable from the fiber-producing plant of any one of paragraphs 40 to 47.

[49] In a further embodiment, the present invention provides a nucleic acid molecule encoding a non-functional GLUC 1.1 protein having an amino acid sequence wherein at least one amino acid residue similar to the active site residues or to the glycosylation site residues of the GLUC 1.1 protein of SEQ ID NO: 4 is lacking or is substituted for a non- similar amino acid residue. In one aspect, the active site residues of the GLUC 1.1 protein of SEQ ID NO: 4 are selected from the group consisting of Tyr48, Glu249, Trp252, and Glu308, and wherein the glycosylation site residue of the GLUC 1.1 protein of SEQ ID NO: 4 is Asn202. In another aspect, the non- functional GLUC 1.1 protein comprises an amino acid sequence at least 90% identical to the amino acid sequence of SEQ ID NO: 6, SEQ ID NO: 18, SEQ ID NO: 57 or SEQ ID NO: 22. In another aspect, the nucleic acid molecule comprises a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 3 from nucleotide 101 to 1078, wherein at least one nucleic acid residue is deleted, inserted or substituted. In yet another aspect, the nucleic acid molecule comprises a nucleotide sequence at least 92% identical to the nucleic acid sequence of SEQ ID NO: 54 from nucleotide 50 to 589. In still a further aspect, the nucleic acid molecule comprises the nucleic acid sequence of SEQ ID NO: 54 from nucleotide 50 to 589. In still another aspect, the nucleic acid molecule comprises a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 1 from nucleotide 2410 to 3499, wherein at least one nucleic acid residue is deleted, inserted or substituted. In yet another aspect, the nucleic acid molecule comprises a nucleotide sequence at least 92% identical to the nucleic acid sequence of SEQ ID NO: 5 from nucleotide 63 to 711, SEQ ID NO: 17 from nucleotide 2 to 472, SEQ ID NO: 56 from nucleotide 112 to 760 or SEQ ID NO: 21 from nucleotide 27 to 372. In still a further aspect, the nucleic acid molecule comprises the nucleic acid sequence of SEQ ID NO: 5 from nucleotide 63 to 711, SEQ

ID NO: 17 from nucleotide 2 to 472, SEQ ID NO: 56 from nucleotide 112 to 760, or SEQ ID NO: 21 from nucleotide 27 to 372.

[50] In another embodiment, the present invention provides a non- functional GLUC 1.1 protein encoded by the nucleic acid molecule of paragraph 49.

[51] In still another embodiment, the present invention provides a method for identifying a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein in a plant, preferably a Gossypium plant, such as a Gossypium hirsitum plant, said GLUC 1.1 gene comprising a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 1 from nucleotide 2410 to 3499, comprising the step of identifying a polymorphic site in the nucleotide sequence of the GLUC 1.1 gene in the genomic DNA of the plant that results in the production of a non- functional GLUC 1.1 protein. In one aspect, the present invention provides a method for identifying a GLUC 1.1 gene from Gossypium barbadense or from Gossypium darwinii comprising the step of identifying a T nucleotide at a nucleotide position corresponding to nucleotide position 3050 in SEQ ID NO: 1. hi another aspect, the present invention provides a method for identifying a GLUC 1.1 gene from Gossypium arboreum comprising the step of identifying a deletion of a C nucleotide at a nucleotide position corresponding to nucleotide position 2674, 2675 or 2676 in SEQ ID NO: 1.

[52] In a further embodiment, the present invention provides a method of distinguishing a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein from a GLUC 1.1 gene encoding a functional GLUC 1.1 protein, said GLUC 1.1 genes both comprising a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 1 from nucleotide 2410 to 3499, comprising the step of identifying a polymorphic site in the nucleotide sequences of the GLUC 1.1 genes. In one aspect, the present invention provides a method of distinguishing a GLUCl.1 from Gossypium barbadense, from Gossypium darwinii or from Gossypium arboreum from a GLUC 1.1 gene from Gossypium hirsutum, respectively, comprising the step of identifying a polymorphic site selected from the group consisting of: polymorphic sequence marker GLUC1.1A-SNP2

located between the nucleotide at position 2765 and 2766 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP3 located at nucleotide position 2911 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP5 located at nucleotide position 3050 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP6 located at nucleotide position 3202 in SEQ ID NO: 1, SNP marker GLUC 1.1 A-SNP7 located at nucleotide position 2674, 2675 or 2676 in SEQ ID NO: 1 and SNP marker GLUCl .1 A-SNP8 located at nucleotide position 3170 in SEQ ID NO: 1. In another aspect, polymorphic sequence marker GLUC1.1A-SNP2 from Gossypium barbadense or Gossypium darwinii and from Gossypium hirsutum, respectively, is detected by amplification of a DNA fragment of about 143 bp and about 134 bp, respectively, with primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively. In still another aspect, SNP marker GLUC1.1A-SNP3 from Gossypium barbadense or Gossypium darwinii and from Gossypium hirsutum, respectively, is detected by amplification of a DNA fragment of about 57 bp with primers comprising SEQ ID NO: 41 and 42 and detection of the DNA fragment with fluorescently labeled probes comprising SEQ ID NO: 39 and 40, respectively.

[53] In a further embodiment, the present invention provides a method for generating and/or selecting a non-naturally occurring fiber-producing plant, and parts and progeny thereof, wherein the functional expression of at least one allele of at least one fiber- specific GLUC gene that is functionally expressed during the fiber strength building phase, in particular the fiber maturation phase, of fiber development is abolished, comprising the step of: mutagenizing at least one allele of the GLUC gene, or introgressing at least one allele of a non-functionally expressed ortholog of the GLUC gene or at least one allele of a mutagenized GLUC gene, or introducing a chimeric gene comprises the following operably linked DNA elements: (a) a plant expressible promoter, (b) a transcribed DNA region, which when transcribed yields an inhibitory RNA molecule capable of reducing the expression of the GLUC allele, and (c) a 3' end region comprising transcription termination and polyadenylation signals functioning in cells of the plant. In one aspect, the GLUC gene is a GLUCl.1 gene encoding a GLUC protein that has at least 90% sequence identity to SEQ ID NO: 4. In another aspect, the fiber- producing plant is a Gossypium plant, and the GLUC gene is a GLUC 1.1 A gene encoding

a GLUC protein that has at least 97% sequence identity to SEQ ID NO: 4 or a GLUCl. ID gene encoding a GLUC protein that has at least 97% sequence identity to SEQ ID NO: 9, preferably a GLUCl. IA gene, hi still another aspect, the fiber-producing plant is a Gossypium plant, and the non- functionally expressed ortholog of the GLUC gene is a GLUCl. IA gene which is derived from a Gossypium barbadense, from a Gossypium darwinii or a Gossypium arboreum plant, preferably from a Gossypium barbadense. hi a further aspect, the method further comprises the step of identifying the non-functionally expressed ortholog of the GLUC gene or the mutagenized GLUC gene according to the method of paragraph 51.

[54] In a further embodiment, the present invention provides a method for altering the callose content of a fiber in a fiber-producing plant, particularly increasing the callose content of a fiber, comprising the steps of: generating and/or selecting a non-naturally occurring fiber-producing plant, and parts and progeny thereof, wherein the functional expression of at least one allele of at least one fiber-specific GLUC gene that is functionally expressed during the fiber strength building phase, in particular the fiber maturation phase, of fiber development is abolished, according to the method of paragraph 53, and selecting a plant with an altered callose content in its fibers, in particular an increased callose content.

[55] In a further embodiment, the present invention provides a method for altering the properties of a fiber in a fiber-producing plant, particularly increasing the strength of a fiber, comprising the steps of: generating and/or selecting a non-naturally occurring fiber- producing plant, and parts and progeny thereof, wherein the functional expression of at least one allele of at least one fiber-specific GLUC gene that is functionally expressed during the fiber strength building phase, in particular the fiber maturation phase, of fiber development is abolished, according to the method of paragraph 53, and selecting a plant with an altered fiber strength, in particular an increased fiber strength.

[56] hi another embodiment, the present invention provides a kit for identifying a GLUC 1.1 gene encoding a non-functional GLUC 1.1 protein in a plant, said GLUC 1.1

gene comprising a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 1 from nucleotide 2410 to 3499, comprising primers and/or probes for determining the presence of a polymorphic site in the nucleotide sequence of the GLUC 1.1 gene in the genomic DNA of the plant that results in the production of a nonfunctional GLUC 1.1 protein. In one aspect, the kit comprises primers and/or probes for determining the presence of a T nucleotide at a nucleotide position corresponding to nucleotide position 3050 in SEQ ID NO: 1 or for determining a deletion of a C nucleotide at a nucleotide position corresponding to nucleotide position 2674, 2675 or 2676 in SEQ ID NO: 1.

[57] In still another embodiment, the present invention provides a kit for distinguishing a GLUC 1.1 gene encoding a non- functional GLUC 1.1 protein from a GLUC 1.1 gene encoding a functional GLUC 1.1 protein, said GLUC 1.1 genes both comprising a nucleic acid sequence having at least 92% sequence identity to SEQ ID NO: 1 from nucleotide 2410 to 3499, comprising primers and/or probes for determining the presence of a polymorphic site in the nucleotide sequences of the GLUC 1.1 genes. In one aspect, the present invention provides a kit comprising primers and/or probes for distinghuishing Gossypium barbadense, Gossypium darwinii or Gossypium arboreum specific alleles from Gossypium hirsutum specific alleles of a polymorphic site selected from the group consisting of: polymorphic sequence marker GLUC1.1A-SNP2 located between the nucleotide at position 2765 and 2766 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP3 located at nucleotide position 2911 in SEQ ID NO: 1, SNP marker GLUCl.1A-SNP5 located at nucleotide position 3050 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP6 located at nucleotide position 3202 in SEQ ID NO: 1, SNP marker GLUC1.1A-SNP7 located at nucleotide position 2674, 2675 or 2676 in SEQ ID NO: 1 and SNP marker GLUC1.1A-SNP8 located at nucleotide position 3170 in SEQ ID NO: 1. hi another aspect, the kit comprises at least two primers and/or probes selected from the group consisting of: primers comprising at their extreme 3' end SEQ ID NO: 37 and 38, respectively, to identify polymorphic sequence marker GLUC1.1A-SNP2, primers comprising SEQ ID NO: 41 and 42, respectively, to identify SNP marker GLUC 1.1 A- SNP3, probes comprising SEQ ID NO: 39 and 40, respectively, to identify SNP marker

GLUC1.1A-SNP3, primers comprising SEQ ID NO: 62 and 63, respectively, to identify SNP marker GLUC1.1A-SNP5, and probes comprising SEQ ID NO: 60 and 61, respectively, to identify SNP marker GLUCl.1A-SNP5.

Brief description of the figures

[58] Figure 1: Alignment of genomic and cDNA sequences of A and D subgenome- specific GLUCl.1 genes from Gossypium hirsutum (' GhGLUC l.lA-gDN A ' corresponds to SEQ ID NO: 1 from nucleotide 2246 to 3753, ' GhGLUC LlA-cDN A ' corresponds to SEQ ID NO: 3, ' GhGLUCL 1 D-gDNA ' corresponds to SEQ ID NO: 7 from nucleotide 3206 to 4694, and ' GhGLUC 7.7£>-cDN A' corresponds to SEQ ID NO: 9) and Gossypium barbadense (' GbGLUCl. lA-gDNA ' corresponds to SEQ ID NO: 5 , ' GbGLUC 7,/λ-cDN A' corresponds to SEQ ID NO: 54, 'GbGLUC 1.1 D-gDNA ' corresponds to SEQ ID NO: 11 , and ' GbGLUC l.lD-cOH A' corresponds to SEQ ID NO: 13). The putative TATA box is indicated in bold, the putative start codons and the putative first exons are indicated in bold and in bold with an arrow, respectively, the putative intron and second exon sequences are indicated in regular with an arrow, the putative intron sequences are further indicated between 7', the putative (premature) STOP codons are indicated in italic and underlined.

[59] Figure 2: Alignment of amino acid sequences of A and D subgenome-specific GLUC 1.1 proteins from Gossypium hirsutum ('GhGLUC 1.1 A' corresponds to SEQ ID NO: 2 and 4 and 'GhGLUCLlD' corresponds to SEQ ID NO: 8 and 10) and Gossypium barbadense ('GbGLUCl. IA' corresponds to SEQ ID NO: 6 and 55 and 'GbGLUCLlD' corresponds to SEQ ID NO: 12 and 14). The putative signal peptide is indicated in italic, the putative post-translational splicing site is indicated as '><', the GH 17 signature is indicated in bold. Amino acids that are identical between at least three of the four sequences are highlighted. The dashed line indicates the protein segment that is missing in GbGLUCLlA.

[60] Figure 3: Protein model of GLUCl. IA protein of G. hirsutum (Figure 3a; right) and G. barbadense (Figure 3b; right) based on an X-ray structure of a barley 1,3-1,4-

beta-glucanase (laqO; Figure 3a&b; left). The active site of laqO is located in an open cleft at the bottom of the barrel defined by the C-terminal ends of the parallel intra-barrel beta-strands (Muller et al., 1998, J.Biol.Chem 273 (6): 3438-3446) and is indicated by the amino acids and their position numbers displayed in the upper left part of the protein model of laqO in Figure 3a and b at the left. Active site residues Glu288, Glu232 and Tyr33 in laqO (Figure 3a, left) correspond to Glu308, Glu249 and Tyr48 in GhGLUCl. IA (Figure 3a, right) and are absent in GbGLUCl. IA (Figure 3b, right). The glycosylation site Asnl90 in laqO (Figure 3 A, left) corresponds to Asn 202 in GhGLUCl. IA (Figure 3a, right) and is also absent in GbGLUCl. IA (Figure 3b, right). Figure 3b further shows that the threonine, histidine and glutamine amino acids at position 82, 83 and 84 of GbGLUC 1.1 A (Figure 3b; right) that are not present in GhGLUC 1.1 A (see for example Figure 7) are located in a distant loop which is not part of the active site and not involved in glycosylation.

[61] Figure 4: Box plot indicating the difference in fiber strength (as determined by measuring the breaking force of single fibers; indicated in cN on the Y-axis) between untreated fibers ('untreated') and fibers treated with exogenous glucanase ('treated') derived from Gossypium hirsutum cultivar FM966 grown in a greenhouse in Europe ('FM966 Astene'), in the field in the US ('FM966 Sellers') and in the field in Australia ('FM966 Australia'), from Gossypium hirsutum cultivar Coker 312 grown in a greenhouse in Europe ('Coker 312'), from Gossypium barbadense cultivar PimaS7 grown in a greenhouse in Europe ('PimaS7'), and from Gossypium barbadense cultivar PimaY5 grown in the field in Australia ('PimaY5').

[62] Figure 5: Box plot indicating the difference in callose content (as determined by fluorescence measurements of aniline blue stained fibers; indicated as the ratio of green over blue fluorescence on the Y-axis) between untreated fibers ('untreated') and fibers treated with exogenous glucanase ('treated') derived from Gossypium hirsutum cultivar FM966 grown in a greenhouse in Europe ('FM966 Astene'), in the field in the US ('FM966 Sellers') and in the field in Australia ('FM966 Australia'), from Gossypium hirsutum cultivar Coker 312 grown in a greenhouse in Europe ('Coker 312'), from

Gossypium barbadense cultivar PimaS7 grown in a greenhouse in Europe ('PimaS7'), and from Gossypium barbadense cultivar PimaY5 grown in the field in Australia ('PimaY5').

[63] Figure 6: Alignment of genomic DNA sequences of A and D subgenome-specific GLUCl.1 genes from Gossypium hirsutum ('GhGLUCl. IA gDN A' corresponds to SEQ ID NO: 1 from nucleotide 2348 to 3554 and ' GhGLUC LlD gDN A' corresponds to SEQ ID NO: 7 from nucleotide 3311 to 4496), Gossypium tomentosum ('GtGLUC LlA_gDNA ' corresponds to SEQ ID NO: 15 and ' GtGLUC LlD_gDN A ' corresponds to SEQ ID NO: 25), Gossypium barbadense ('GbGLUCl. IA gDNA ' corresponds to SEQ ID NO: 5 and ' GbGLUC 1.1D gDNA ' corresponds to SEQ ID NO: 11), Gossypium darwinii (' GdGLUC LlA_gDNA' corresponds to SEQ ID NO: 17 and 'GdGLUC LlD gDNA ' corresponds to SEQ ID NO: 27), Gossypium mustelinum, ('GmGLUCl. lA_gDNA ' corresponds to SEQ ID NO: 19 and ' GmGLUC LlD_gDN A' corresponds to SEQ ID NO: 29), Gossypium arboreum (' GaGLUC LlA_gDNA ' corresponds to SEQ ID NO: 21), Gossypium herbaceum ('GheGLUCl.lA gDNA' corresponds to SEQ ID NO: 23), and Gossypium raimondii (' GrGLUC LlD gDN A' corresponds to SEQ ID NO: 31). The positions of primers SE077 and SE078, used to generate the complete coding sequence from start to stop codon, and the positions of primers SE003 and SE002, used to generate partial coding sequences, are underlined. The putative start codons and the putative first exons are indicated in bold and in bold with an arrow, respectively, the putative intron and second exon sequences are indicated in regular with an arrow, the putative intron sequences are further indicated between '/', the putative (premature) STOP codons are indicated in italic and underlined. Five polymorphic sites (4 single nucleotide polymorphisms (SNPs) and one extended indel) that exist between the GLUCl. IA or GLUC 1.1 D sequences of , e.g., G. hirsutum FM966 and G. barbadense Pima S7 or G. darwinii, are indicated with arrows and named 'GLUCl. ID-SNPl' and 'GLUC1.1A-SNP2, 3, 5 and 6'. Allelic variants are indicated as follows: [G. hirsutum allele/ G. barbadense or G. darwinii allele]. One polymorphic site (1 SNP) that exist between the GLUCl. IA sequences of, e.g., G. hirsutum FM966 and G. arboreum is indicated with an arrow and named 'GLUCl.1A-SNP7'. Allelic variants are

indicated as follows: [G. hirsutum allele/ G. arboreum allele]. One polymorphic site (1 SNP) that exist between the GLUC 1.1 A sequences of, e.g., G. barbadense Pima S7 or G. darwinii is indicated with an arrow and named 'GLUC1.1A-SNP8'. Allelic variants are indicated as follows: [G. barbadense allele/ G. darwinii allele].

[64] Figure 7: Alignment of amino acid sequences of A and D subgenome-specifϊc GLUC 1.1 proteins from Gossypium hirsutum (GhG LUC l.lA_prot' corresponds to SEQ ID NO: 2 and 4 and GhGLUCl.1D_ prof corresponds to SEQ ID NO: 8 and 10; full- length sequences), Gossypium tomentosum (GtGLUC l.lA_prot ' corresponds to SEQ ID NO: 16 and GtGLUCl.1D_ prof corresponds to SEQ ID NO: 26; partial sequences), Gossypium barbadense (GbGLUC l.lA_prof corresponds to SEQ ID NO: 6 and 55 and GbGLUC l.lD_prof corresponds to SEQ ID NO: 12 and 14; full-length sequences), Gossypium darwinii (GdGLUC l.lA_prof corresponds to SEQ ID NO: 57 and GdGLUC l.lD_prof corresponds to SEQ ID NO: 59; full-length sequences), Gossypium mustelinum, (GmGLUC l.lA_prof corresponds to SEQ ID NO: 20 and GmGLUC 1.1 D_prof corresponds to SEQ ID NO: 30; partial sequences), Gossypium arboreum (GaG LUC l.lA_prof corresponds to SEQ ID NO: 22; full-length sequence), Gossypium herbaceum (GheGLUCl.lA_prof corresponds to SEQ ID NO: 24; full- length sequence), and Gossypium raimondii (GrGLUCl.1D_ prof corresponds to SEQ ID NO: 32; partial sequences). The putative signal peptide is indicated in italic, the putative post-translational splicing site is indicated as '><', the GHl 7 signature is indicated in bold. Amino acids that differ from the amino acids in the upper sequence, i.e. GhGLUC l.lA_prot, are highlighted.

[65] Figure 8: Expression of GLUCl. IA and GLUCl. ID in G. barbadense. DNA from a cDNA library from (developing) fibers in Gossypium barbadense was extracted and equalized. PCR fragments were amplified using oligonucleotide primers SE002 and SE003 (SEQ ID NO: 35 and 36) and digested with restriction enzyme AIwI. A PCR amplified product for GLUC 1.1 A yields 3 fragments (479bp+118bp+59bp) while for GLUCl. ID it only yields 2 fragments (538bp +118bp). Lane 1 and 12: 1 kb size markers; lanes 2 to 9: GbGLUCl. IA and D expression at 0, 5, 10, 15, 20, 25, 30 and 40 DPA; lane

10: negative (no template; NTC); lane 11: positive control (genomic DNA from Pima S7).

[66] Figure 9: Schematic representation of 165250 bps DNA fragment spanning the GLUCl. IA gene of Gossypium hirsutum (SEQ ID NO: 53). Box: retrotransposon region; *: position of CIR280 homology region; arrow: DNA fragment encoding protein indicated with following abbreviations: SHMT (Serine HydroxyMethylTransferase); GrpE/HSP-70 (GrpE protein/ HSP-70 cofactor); ARFl 7: putative Auxin Response Factor similar to At-ARF 17; eIF-5-1: probable eukaryotic translation Initiation Ffactor 5-1; Avr9: putative Avr9 elicitor response protein; VPS9: similar to Vacuolar Protein Sorting- associated protein VPS9; HAT: putative Histon Acetyl Transferase gene; Glue 1.1 : GLUC 1.1 A encoding region; MEKKl: putative Mitogen-activated protein kinase kinase kinase 1; PIP5K1: PhosphatidylInositol-4-Phosphate 5-Kinase 1.

Detailed embodiments

[67] The current invention is based on the unexpected finding that the presence of the Gossypium barbadense ortholog of a fiber strength locus on chromosome A05, hereinafter called Gossypium barbadense fiber strength allele, in Gossypium hirsutum plants results in an increased strength of the fibers of the Gossypium hirsutum plants compared to the strength of the fibers of Gossypium hirsutum plants comprising the Gossypium hirsutum ortholog of the fiber strength locus.

[68] Thus, in a first aspect, the present invention provides a non-naturally occurring Gossypium plant, and parts and progeny thereof, comprising at least one superior allele of a quantitative trait locus (QTL) for fiber strength located on chromosome A05.

[69] As used herein, the term "non-naturally occurring" or "cultivated" when used in reference to a plant, means a plant with a genome that has been modified by man. A transgenic fiber-producing plant, for example, is a non-naturally occurring fiber- producing plant that contains an exogenous nucleic acid molecule, e.g., a chimeric gene comprising a transcribed region which when transcribed yields a biologically active RNA

molecule capable of reducing the expression of a GLUC gene according to the invention and, therefore, has been genetically modified by man. In addition, a fiber-producing plant that contains, for example, a mutation in an endogenous GLUC gene (e.g. in a regulatory element or in the coding sequence) as a result of an exposure to a mutagenic agent is also considered a non-naturally occurring fiber-producing plant, since it has been genetically modified by man. Furthermore, a fiber-producing plant of a particular species, such as Gossypium hirsutum, that contains, for example, a mutation in an endogenous GLUC gene that in nature does not occur in that particular plant species, as a result of, for example, directed breeding processes, such as marker-assisted breeding and selection or introgression, with another species of that fiber-producing plant, such as Gossypium barbadense, is also considered a non-naturally occurring fiber-producing plant. In contrast, a fiber-producing plant containing only spontaneous or naturally occurring mutations, i.e. a plant that has not been genetically modified by man, is not a "non- naturally occurring plant" as defined herein and, therefore, is not encompassed within the invention. One skilled in the art understands that, while a non-naturally occurring fiber- producing plant typically has a nucleotide sequence that is altered as compared to a naturally occurring fiber-producing plant, a non-naturally occurring fiber-producing plant also can be genetically modified by man without altering its nucleotide sequence, for example, by modifying its methylation pattern.

[70] The term "quantitative trait" refers herein to a trait, such as fiber strength, whose phenotypic characteristics vary in degree and can be attributed to the interactions between two or more genes and their environment.

[71] As used herein, the term "locus" (loci plural) or "site" means a specific place or places on a chromosome where, for example, a gene, a genetic marker or a QTL is found.

[72] A "quantitative trait locus (QTL)" is a stretch of DNA (such as a chromosome arm, a chromosome region, a nucleotide sequence, a gene, and the like) that is closely linked to a gene that underlies the trait in question. "QTL mapping" involves the creation of a map of the genome using genetic or molecular markers, like AFLP, RAPD, RFLP,

SNP, SSR, and the like, visible polymorphisms and allozymes, and determining the degree of association of a specific region on the genome to the inheritance of the trait of interest. As the markers do not necessarily involve genes, QTL mapping results involve the degree of association of a stretch of DNA with a trait rather than pointing directly at the gene responsible for that trait. Different statistical methods are used to ascertain whether the degree of association is significant or not. A molecular marker is said to be "linked" to a gene or locus, if the marker and the gene or locus have a greater association in inheritance than would be expected from independent assortment, i.e. the marker and the locus co-segregate in a segregating population and are located on the same chromosome. "Linkage" refers to the genetic distance of the marker to the locus or gene (or two loci or two markers to each other). The closer the linkage, the smaller the likelihood of a recombination event taking place, which separates the marker from the gene or locus. Genetic distance (map distance) is calculated from recombination frequencies and is expressed in centiMorgans (cM) [Kosambi (1944), Ann. Eugenet. 12:172-175].

[73] "Fiber strength locus" or "strength locus", as used herein, refers to a stretch of DNA on chromosome A05 of Gossypium species that is closely linked to (a) gene(s) that is(are) involved in the regulation of fiber strength. The "fiber strength locus" is a QTL said to be linked to the "(fiber strength) causal gene(s)".

[74] A "fiber", such as a "cotton fiber", as used herein, refers to a seed trichome, more specifically a single cell of a fiber-producing plant, such as cotton, that initiates from the epidermis of the outer integument of the ovules, at or just prior to anthesis. The morphological development of cotton fibers has been well documented (Basra and Malik, 1984, Int Rev of Cytology 89: 65-113; Graves and Stewart, 1988, supra; Ramsey and Berlin, 1976, American Journal of Botany 63 (6): 868-876; Ruan and Chourey, 1998, Plant Physiology 118: 399^06; Ruan et al. 2000, Aust. J. Plant Physiol. 27:795-800; Stewart, 1975, Am. J. Bot. 62, 723-730). Cotton fibers, in particular from Gossypium hirsutum, undergo four overlapping developmental stages: fiber cell initiation, elongation, secondary cell wall biosynthesis, and maturation. Fiber cell initiation is a

rapid process. White fuzzy fibers begin to develop immediately after anthesis and continue up to about 3 days post-anthesis (DPA), which is followed by fiber cell elongation (until about 10 to about 17 DPA). Depending upon growth conditions, secondary cell wall biosynthesis initiates and continues to about 25 to about 40 DPA, followed by a maturation process until about 45 to about 60 DPA. The secondary cell wall synthesis and maturation phase are herein commonly referred to as "fiber strenght building phase". Only about 25 to 30% of the epidermal cells differentiate into the commercially important lint fibers (Kim and Triplett, 2001). The majority of cells does not differentiate into fibers or develop into short fibers or fuzz. During fiber elongation and secondary wall metabolism, the fiber cells elongate rapidly, synthesize secondary wall components, and show dramatic cellular, molecular and physiological changes. Fiber elongation is coupled with rapid cell growth and expansion (Seagull, 1991. In Biosynthesis and biodegradation of cellulose (Haigler, C. H. & Weimer, P. J., eds) pp. 1432163, MarcelDekker, New York) and constant synthesis of a large amount of cell metabolites and cell wall components such as cellulose. About 95% of the dry-weight in mature cotton fibers is cellulose (Pfluger and Zambryski, 2001, Curr Biol 11: R436— R439; Ruan et al., 2001, Plant Cell 13: 47-63). Non-celluloid components are also important to fiber cell development (Hayashi and Delmer, 1988, Carbohydr. Res. 181 : 273-277; Huwyler et al., 1979, Planta 146: 635-642; Meinert and Delmer, 1977, Plant Physiol 59: 1088-1097; Peng et al, 2002, Science 295: 147-150). Compared to other plant cells, cotton fibers do not contain lignin in secondary walls but have large vacuoles that are presumably related to rapid cell growth and expansion (Basra and Malik, 1984, supra; Kim and Triplett, 2001, Plant Physiology 127: 1361-1366; Mauney, 1984, supra; Ruan and Chourey, 1998, supra; Ruan et al., 2000, supra; Van 't Hof, 1999, American Journal of Botany 86: 776-779).

[75] "Fiber strength", as used herein, can be determined by determining the strength of a bundle of fibers, i.e. "fiber bundle strength", or by determining the strength of single fibers. The higher the single fiber strength and the lower the variations of single fiber breaking elongation, the closer the bundle and yarn tensile strength would be to the sum of single fiber strength; ideally, fiber bundle tenacity would equal the total single fiber

breaking tenacity had all fibers within the bundle equal breaking elongation and no slack (Liu et al., February 2005, Textile Res. J).

[76] "Fiber bundle strength", as used herein, refers to a measure that is usually expressed in terms of grams per tex. This commercial High Volume Instruments (HVI) measure of fiber bundle strength ("HVI strength") is also called "tenacity". A tex unit is equal to the weight in grams of 1,000 meters of fiber. Therefore, the strength reported is the force in grams required to break a bundle of fibers one tex unit in size. Measurements of cotton fiber bundle strength can, for example, be made according to USDA standards. A beard of cotton is clamped in two sets of jaws, one eighth inch apart, and the force required to break the fibers is determined. Table 1 can be used as a guide in interpreting fiber strength measurements.

Table 1 : Interpretation of HVI fiber strength measurements

Degree of Strength HVI* Strength (grams per tex)

Very Strong 31 or more

Strong 29 -30

Average 26 - 28

Intermediate 24 - 25

Weak 23 or less

*High Volume Precision Instruments

[77] Alternatively, the strength of fibers can be compared by determining the "single fiber strength" by performing single fiber tensile tests, for example, on a FAVIMAT Robot (Textechno) as described on http://www.textechno.com/ and in the Examples. Briefly, a single fiber is clamped between two fiber clamps with a continuously adjustable gauge length between 5 and 100 mm (set e.g. on 8 mm) and a draw-off clamp speed between 0.1 and 100 mm/min (set e.g. on 4 mm/min), and the force (cN) required to break the fibers ("breaking force") is determined. Average breaking forces of specific cotton varieties can be found in the Examples.

[78] "Chromosome A05", as used herein, refers to chromosome A05 (numbering according to Wang et al., 2006, Theor Appl Genet 113(l):73-80) in an A genome diploid Gossypium plant, such as Gossypium herbaceum or Gossypium arboreum, or in an AD allotetraploid Gossypium plant, such as Gossypium hirsutum, Gossypium barbadense and Gossypium darwinii. In one embodiment, the Gossypium plant is an A genome diploid Gossypium plant comprising 13 A genome chromosome pairs, numbered AOl to Al 3 according to Wang et al. (2006, Theor Appl Genet 113(l):73-80), such as Gossypium herbaceum or Gossypium arboreum. In another embodiment, the Gossypium plant is an AD genome allotetraploid Gossypium plant comprising 13 A genome and 13 D genome chromosome pairs, numbered AOl to Al 3 and DOl to D 13, respectively, according to Wang et al. {supra), such as Gossypium hirsutum, Gossypium barbadense and Gossypium darwinii.

[79] In one embodiment, the non-naturally occurring Gossypium plant is a Gossypium hirsutum, a Gossypium herbaceum or a Gossypium arboreum plant, preferably a Gossypium hirsutum plant, and the superior allele of the fiber strength locus is derived from Gossypium barbadense.

[80] Gossypium barbadense, in particular Gossypium barbadense cv. Pima S7, seeds are publicly available and can be obtained for example from the Cotton Collection (USDA, ARS, Crop Germplasm Research, 2765 F&B Road, College Station, Texas 77845; http://www.ars-grin.gov/).

[81] The term "superior allele" of the fiber strength locus refers herein to an allele of the fiber strength locus the presence of which in the genome of a fiber-producing plant results in a higher fiber strength compared to the fiber strength in such fiber-producing plant not comprising the superior allele (i.e., comprising a non-superior allele).

[82] As used herein, the term "allele(s)" means any of one or more alternative forms of a gene or a marker at a particular locus or of a quantitative trait locus (QTL). In a diploid or allotetraploid (amphidiploid) cell of an organism, alleles of a given gene, marker or

QTL are located at a specific location or locus (loci plural) on a chromosome. One allele is present on each chromosome of the pair of homologous chromosomes. As used herein, the term "homologous chromosomes" means chromosomes that contain information for the same biological features and contain the same genes or markers at the same loci and the same quantitative trait loci but possibly different alleles of those genes, markers or quantitative trait loci. Homologous chromosomes are chromosomes that pair during meiosis. "Non-homologous chromosomes", representing all the biological features of an organism, form a set, and the number of sets in a cell is called ploidy. Diploid organisms contain two sets of non-homologous chromosomes, wherein each homologous chromosome is inherited from a different parent. In allotetraploid (amphidiploid) species, like cotton, essentially two sets of diploid genomes exist, whereby the chromosomes of the two genomes are referred to as "homeologous chromosomes" (and similarly, the genes, markers and loci of the two genomes are referred to as homeologous genes, markers or loci). A diploid, or allotetraploid (amphidiploid), plant species may comprise a large number of different alleles at a particular locus.

[83] The term "ortholog" of a gene or protein or QTL refers herein to the homologous gene or protein or QTL found in another species, which has the same function as the gene or protein or QTL, but is (usually) diverged in sequence from the time point on when the species harboring the genes or quantitative trait loci diverged (i.e. the genes or quantitative trait loci evolved from a common ancestor by speciation). Orthologs of, e.g., the Gossypium barbadense GLUC genes or fiber strength locus may thus be identified in other plant species (e.g. Gossypium arboreum, Gossypium darwinii, etc.) based on both sequence comparisons (e.g. based on percentages sequence identity over the entire sequence or over specific domains) and/or functional analysis.

[84] In one embodiment, the superior allele of the fiber strength locus is obtainable from Gossypium barbadense, in particular Gossypium barbadense cv. PimaS7, i.e. the presence of the Gossypium barbadense fiber strength allele in a Gossypium plant, such as a Gossypium hirsutum plant, results in an increased fiber strength compared to the fiber

strength in the Gossypiwn plant, such as the Gossypium hirsutum plant, not comprising the Gossypium barbadense allele, but, for example, the Gossypium hirsutum allele.

[85] In still another embodiment, the Gossypium barbadense fiber strength allele is located on chromosome A05 of Gossypium barbadense between AFLP marker P5M50- Ml 26.7 and SSR marker CIR280. In another embodiment, the Gossypium barbadense fiber strength allele is located on chromosome A05 of Gossypium barbadense between AFLP marker P5M50-M 126.7 and SSR marker BNL3992. In yet another embodiment, the Gossypium barbadense allele is located on chromosome A05 of Gossypium barbadense between AFLP marker P5M50-M126.7 and SSR marker CIR401c. In a further embodiment, the LOD peak of the fiber strenght QTL allele of Gossypium barbadense is located between SSR marker NAU861 or the GLUC 1.1 marker and SSR marker CIR401c, in particular at about O to 5 cM, more specifically at about 4 cM, especially at about 4.008 cM, from SSR marker NAU861 or the GLUC 1.1 marker and at about 0 to 12 cM, more specifically at about 10 cM, especially at about 10.52 cM, from SSR marker CIR40 Ic.

[86] A "(genetic or molecular) marker", as used herein, refers to a polymorphic locus, i.e. a polymorphic nucleotide (a so-called single nucleotide polymorphism or SNP) or a polymorphic DNA sequence at a specific locus. A marker refers to a measurable, genetic characteristic with a fixed position in the genome, which is normally inherited in a Mendelian fashion, and which can be used for mapping of a trait of interest. For example, the fiber strength trait was mapped on chromosome A05 of Gossypium barbadense between, amongst others, markers P5M50-M 126.7 and CIR280, P5M50-M 126.7 and BNL3992, P5M50-M126.7 and CIR401, and linked to markers NAU861, GLUCl.1, and others, as indicated, e.g., in Table 6 in the Examples. Thus, a genetic marker may be a short DNA sequence, such as a sequence surrounding a single base-pair change, i.e. a single nucleotide polymorphism or SNP, or a long DNA sequence, such as microsatellites or Simple Sequence Repeats (SSRs). The nature of the marker is dependent on the molecular analysis used and can be detected at the DNA, RNA or protein level. Genetic mapping can be performed using molecular markers such as, but not limited to, RFLP

(restriction fragment length polymorphisms; Botstein et al. (1980), Am J Hum Genet 32:314-331; Tanksley et al. (1989), Bio/Technology 7:257-263), RAPD [random amplified polymorphic DNA; Williams et al. (1990), NAR 18:6531-6535], AFLP [Amplified Fragment Length Polymorphism; Vos et al. (1995) NAR 23:4407-4414], SSRs or microsatellites [Tautz et al. (1989), NAR 17:6463-6471]. Appropriate primers or probes are dictated by the mapping method used.

[87] The term "AFLP ^® " (AFLP ^® is a registered trademark of KeyGene N.V., Wageningen, The Netherlands), "AFLP analysis" and "AFLP marker" is used according to standard terminology [Vos et al. (1995), NAR 23:4407-4414; EP0534858; http://www.keygene.com/keygene/techs-apps/]. Briefly, AFLP analysis is a DNA fingerprinting technique which detects multiple DNA restriction fragments by means of PCR amplification. The AFLP technology usually comprises the following steps: (i) the restriction of the DNA with two restriction enzymes, preferably a hexa-cutter and a tetra- cutter, such as EcoRI, Pstl and Msel; (ii) the ligation of double-stranded adapters to the ends of the restriction fragments, such as EcoRI, Pstl and Msel adaptors; (iii) the amplification of a subset of the restriction fragments using two primers complementary to the adapter and restriction site sequences, and extended at their 3' ends by one to three "selective" nucleotides, i.e., the selective amplification is achieved by the use of primers that extend into the restriction fragments, amplifying only those fragments in which the primer extensions match the nucleotides flanking the restriction sites. AFLP primers thus have a specific sequence and each AFLP primer has a specific code (the primer codes and their sequences can be found at the Keygene website: http://www.keygene.com/keygene/pdf/PRIMERCO.pdf; herein incorporated by reference); (iv) gel electrophoresis of the amplified restriction fragments on denaturing slab gels or cappilaries; (v) the visualization of the DNA fingerprints by means of autoradiography, phospho-imaging, or other methods. Using this method, sets of restriction fragments may be visualized by PCR without knowledge of nucleotide sequence. An AFLP marker, as used herein, is a DNA fragment of a specific size, which is generated and visualized as a band on a gel by carrying out an AFLP analysis. Each AFLP marker is designated by the primer combination used to amplify it, followed by the

approximate size (in base pairs) of the amplified DNA fragment, e.g. P5M50-M126.7 refers to AFLP primer combination P05 (or Keygene code PI l, which is a Pstl primer with additional nucleotides AA; see Table 2) and M50 (which is a Msel primer with additional nucleotides CAT; see Table 2), the use of which in Gossypium barbadense results in an amplified DNA fragment of 126.7 bp (see Table 2). It is understood that the size of these fragments may vary slightly depending on laboratory conditions and equipment used. Every time reference is made herein to an AFLP marker by referring to a primer combination and the specific size of a fragment, it is to be understood that such size is approximate, and comprises or is intended to include the slight variations observed in different labs. Each AFLP marker represents a certain locus in the genome.

[88] The term "SSR" refers to Simple Sequence Repeats or microsatellite [Tautz et al. (1989), NAR 17:6463-6471]. Short Simple Sequence stretches occur as highly repetitive elements in all eukaryotic genomes. Simple sequence loci usually show extensive length polymorphisms. These simple sequence length polymorphisms (SSLP) can be detected by polymerase chain reaction (PCR) analysis and be used for identity testing, population studies, linkage analysis and genome mapping. "SSR marker", as used herein, refers to markers indicated as CIRx, NAUx and BNLx (wherein x is a number) that are publicly available markers which are used to create genetic maps of different Gossypium species (see Cotton Microsatellite Database at http://www.cottonmarker.org/).

[89] A "(genetic or molecular) marker", such as an AFLP or SSR marker, can be dominant (homozygous and heterozygous individuals are not distinguishable) or co- dominant (distinguishing homozygous and heterozygous individuals, e.g., by band intensity), as exemplified in Table 2 below. A "(genetic or molecular) marker", such as an AFLP or SSR marker, can be linked to a gene or locus in "coupling phase" or in "repulsion phase'. For example, a dominant marker linked in coupling to a gene or locus is present in individuals with the gene or locus and absent in individuals without the gene or locus, while a dominant marker linked in repulsion phase to a gene or locus is absent in individuals with the gene or locus and present in individuals without the gene or locus.

[90] Different alleles of markers can exist in different plant species. "Gossypium barbadense or Gossypium hirsutum alleles of markers linked to the fiber strength locus", as used herein, refers to a form of a marker that is derived from and specific for Gossypium barbadense or Gossypium hirsutum, respectively. Table 2 examplifies how different alleles of different markers can be identified or distinguished: column 1 indicates different marker loci on chromosome A05 of Gossypium barbadense and/or Gossypium hirsutum, column 2 indicates for each marker locus a specific primer pair that can be used to identify the presence or absence of the specific marker locus, column 3 indicates whether a specific marker allele of Gossypium barbadense (in particular cv. Pima S7; indicated as 'Pima') and Gossypium hirsutum (in particular cv. FM966; indicated as 'FM') generates an amplified DNA fragment and, if so, the size of the amplified DNA fragment, column 4 indicates whether the marker indicated in column 1 is a dominant or a codominant marker as defined above.

Table 2: Detection of specific Gossypium barbadense or Gossypium hirsutum alleles of markers on chromosome A05

Marker Primer pair: Amplified fragment Codominant/ locus on (in bp) dominant chromosome from FM from Pima marker

A05

P5M50- P5 5' GACTGCGTACATGCAGAA 3' 126.7 dominant

M126.7 (SEQ ID NO: 43)

M50 5' GATGAGTCCTGAGTAACAT 3'

(SEQ ID NO: 44)

GLUCl. IA- forward 5' TAT CCC TCT CGA TGA GTA CGA C 3' 134 143 codominant

SNP2 (SEQ ID NO: 37) reverse 5'CCC AAT GAT GAT GAA CCT GAA

TTG3'

(SEQ ID NO: 38)

NAU861 forward 5' CCAAAACTTGTCCCATTAGC 3' 205-210 215-220 codominant

(SEQ ID NO: 45) reverse 5' TTCATCTGTTGCCAGATCC 3'

(SEQ ID NO: 46)

CIR401C forward 5' TGGCGACTCCCTTTT 3' 245-250 dominant

(SEQ ID NO: 47) reverse 5' AAAAGATGTTACACACACACAC 3'

(SEQ ID NO: 48)

CIR401b forward 5' TGGCGACTCCCTTTT 3' 255 dominant

(SEQ ID NO: 47) reverse 5' AAAAGATGTTACACACACACAC 3'

(SEQ ID NO: 48)

BNL3992 forward 5' CAGAAGAGGAGGAGGTGGAG 3' 160-165/ 140-145 codominant

(SEQ ID NO: 49) 85-90

reverse 5' TGCCAATGATGG AAAACTC A 3'

(SEQ ID NO: 50)

CIR280 forward 5 ' ACTGCGTTC ATT AC ACC 3 ' 205 dominant

(SEQ ID NO: 51) reverse 5' GCTTCACCCATTCATC 3'

(SEQ ID NO: 52)

[91] As indicated above, the location of the Gossypium barbadense fiber strength allele on chromosome A05 can be determined by linked AFLP and/or SSR markers, such as AFLP marker P5M50-M126.7, and SSR markers BNL3992, CIR401b and NAU861. However, it is understood that these AFLP and SSR markers can be converted into other types of molecular markers. When referring to a specific (molecular or genetic) marker in the present invention, it is understood that the definition encompasses other types of molecular markers used to detect the genetic variation originally identified by the AFLP and SSR markers. For example, if an AFLP marker is converted into another molecular marker using known methods, this other marker is included in the definition. For example, AFLP markers can be converted into sequence-specific markers such as, but not limited to STS (sequenced-tagged-site) or SCAR (sequence-characterized-amplified- region) markers using standard technology as described in Meksem et al. [(2001), MoI Gen Genomics 265(2):207-214], Negi et al. [(2000), TAG 101: 146-152], Barret et al. (1989), TAG 97:828-833], Xu et al. [(2001), Genome 44(l):63-70], Dussel et al. [(2002), TAG 105:1190-1195] or Guo et al. [(2003), TAG 103:1011-1017]. For example, Dussel et al. [(2002), TAG 105:1190-1195] converted AFLP markers linked to resistance into PCR- based sequence tagged site markers such as indel (insertion/deletion) markers and CAPS (cleaved amplified polymorphic sequence) markers.

[92] The conversion of an AFLP marker into an STS marker, for example, generally involves the purification of the DNA fragment from the AFLP gel and the cloning and sequencing of the DNA fragment. Cloning and sequencing of AFLP fragments (bands) can be carried out using known methods [Guo et al. TAG 103:1011-1017]. Based on the marker sequence (internal) locus specific PCR primers can be developed [Paran and Michelmore (1993), TAG 85:985-993], which amplify fragments of different sizes or wherein the PCR product is cleaved with a restriction enzyme after amplification to reveal a polymorphism. As internal PCR primers often do not reveal polymorphisms

related to the EcoRl, Mse\ or Pstl (or other enzymes) restriction site differences, inverse PCR [Haiti and Ochmann (1996), In: Harwood A, editor, Methods in molecular biology vol58: basic DNA and RNA protocols, Humana Press, Totowa NJ pp293-301] or PCR- walking [Negi et al. (2000), TAG 101:146-152 ; Siebert et al, (1995), NAR 23:1087-1088] may be used to identify flanking sequences, which can then be used to generate simple, locus specific, PCR based markers. Primers can easily be designed using computer software programs such as provided by Sci-Ed (Scientific & Educational Software PO Box 72045, Durham, NC 27722-2045 USA). The polymorphism of the STS marker can be detected by gel electrophoresis, or can be detected using fluorometric assays, such as TaqMan® technology (Roche Diagnostics).

[93] In another embodiment, the fiber strenght QTL allele of Gossypium barbadense comprises at least one Gossypium barbadense ortholog of a nucleotide sequence comprised in the genomic DNA sequence spanning the Gossypium hirsutum GLUC 1.1 A gene represented in SEQ ID NO: 53 (see Figure 9 and the sequence listing).

[94] In another embodiment, the fiber strenght QTL allele of Gossypium barbadense comprises at least a GLUC 1.1 gene encoding a non-functional GLUC 1.1 protein as further described below. In one aspect the Gossypium barbadense GLUC 1.1 gene is located at about 0 to 5 cM, more specifically at about 4 cM, from the LOD peak of the fiber strenght QTL allele of Gossypium barbadense. In another aspect the Gossypium barbadense GLUCl.1 gene is located at about 0 to 2 cM, at about 0 to 1 cM, more specifically at about 0.008 cM of the NAU861 marker located in the fiber strenght QTL allele of Gossypium barbadense.

[95] In another embodiment, the non-naturally occurring Gossypium plant is a Gossypium hirsutum, Gossypium barbadense, a Gossypium herbaceum or a Gossypium arboreum plant, preferably a Gossypium hirsutum plant, and wherein the superior fiber strength allele is derived from Gossypium darwinii. In one aspect, the fiber strenght QTL allele of Gossypium darwinii comprises at least a GLUC 1.1 gene as further described below.

[96] In still another embodiment, the non-naturally occurring Gossypium plant is a Gossypium hirsutum, Gossypium barbadense or a Gossypium herbaceum plant, preferably a Gossypium hirsutum plant, and wherein the superior fiber strength allele is derived from Gossypium arboreum. hi one aspect, the fiber strenght QTL allele of Gossypium arboreum comprises at least a GLUCLl gene as further described below.

[97] In a particular embodiment, the callose content of the fibers of the non-naturally occurring Gossypium plant is increased compared to the callose content of the fibers of an equivalent Gossypium plant that does not comprise the at least one superior allele of the fiber strength locus.

[98] "Callose" refers to a plant polysaccharide that comprises glucose residues linked together through beta-l,3-linkages, and is termed a beta-glucan. It is thought to be manufactured at the cell wall by callose synthases and is degraded by beta- 1,3- glucanases. The callose content of fibers can be measured by staining the fibers with aniline blue, a dye specific for 1,3-beta-glucans. Under UV, callose deposits present an intense yellow-green fluorescence. Images are analyzed and the ratio Green/Blue is used as a measure for callose. "Cellulose" is the major structural polysaccharide of higher plant cell walls. Chains of beta-l,4-linked glucosyl residues assemble soon after synthesis to form rigid, chemically resistant microfibrils. Their mechanical properties together with their orientation in the wall influence the relative expansion of cells in different directions and determine many of the final mechanical properties of mature cells and organs.

[99] In a particular embodiment, the strength of the fibers of the non-naturally occurring Gossypium plant is increased compared to the strength of the fibers of an equivalent Gossypium plant that does not comprise the at least one superior allele of the fiber strength locus.

[100] "Increase in fiber strength", as used herein, refers to an average strength of fibers of a specific fiber-producing plant species, such as cotton, which is significantly higher

than the average strength of fibers of that specific plant species normally observed. Fiber strength is largely determined by variety. However, it may be affected by plant nutrient deficiencies and weather.

[101] hi one aspect of this embodiment, the non-naturally occuring Gossypium plant is a Gossypium hirsutum plant which is homozygous for the Gossypium barbadense fiber strength allele. In a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 5% and about 10%, more specifically about 7,5%, higher than the fiber strength of a Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum fiber strength allele. In still a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 1.6 g/tex and about 3.3 g/tex, more specifically about 2.5 g/tex higher than the fiber strength of a Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum fiber strength allele. In yet a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 34.6 g/tex and about 36.3 g/tex, more specifically about 35.5 g/tex, as compared to a fiber strength of on average between about 32.2 g/tex and about 33.8 g/tex, more specifically about 33.0 g/tex of a Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum fiber strength allele.

[102] A "variety" (abbreviated as var.) or "cultivar" (abbreviated as cv.) is used herein in conformity with the UPOV convention and refers to a plant grouping within a single botanical taxon of the lowest known rank, which grouping can be defined by the expression of the characteristics resulting from a given genotype or combination of genotypes, can be distinguished from any other plant grouping by the expression of at least one of the said characteristics and is considered as a unit with regard to its suitability for being propagated unchanged (stable).

[103] As used herein, the term "heterozygous" means a genetic condition existing when two different alleles reside at a specific locus, but are positioned individually on corresponding pairs of homologous chromosomes in the cell. Conversely, as used herein,

the term "homozygous" means a genetic condition existing when two identical alleles reside at a specific locus, but are positioned individually on corresponding pairs of homologous chromosomes in the cell.

[104] A "fiber-producing plant" refers to a plant species that produces fibers as defined above, such as a cotton plant. Of the Gossypium species, the A genome diploid Gossypium species and AD genome allotetraploid Gossypium species are known to produce spinnable fiber. Botanically, there are three principal groups of cotton that are of commercial importance. The first, Gossypium hirsutum (AADD), is native to Mexico and Central America and has been developed for extensive use in the United States, accounting for more than 95 % of U.S. production. This group is known in the United States as American Upland cotton, and their fibers vary in length from about 7/8 to about 1 5/16 inches (about 22 - about 33 mm). Worldwide it accounts for about 90% of the cotton production. A second botanical group, G. barbadense (AADD), which accounts for about 5% of U.S. production and about 8% of the worldwide production, is of early South American origin. With fibers varying in length from about 1 1/4 to about 1 9/16 inches (about 32 - about 40 mm), it is known in the United States as American Pima, but is also commonly referred to as Extra Long Staple (ELS) cotton. A third group, G. herbaceum (AA) and G. arboreum (AA), embraces cotton plants with fibers of shorter length, about 1/2 to about 1 inch (about 13 - about 25 mm), that are native to India and Eastern Asia. None from this group is cultivated in the United States.

[105] "Fiber length", as used herein, refers to the average length of the longer one-half of the fibers (upper half mean length). In the US, it is usually reported in lOOths or 32nds of an inch (see Table 3; 1 inch is 25.4 mm). It is measured, for example, according to United States Department of Agriculture (USDA) standards by passing a "beard" of parallel fibers through a sensing point. The beard is formed when fibers from a sample of cotton are grasped by a clamp, then combed and brushed to straighten and parallel the fibers. Fiber length is largely determined by variety, but the cotton plant's exposure to extreme temperatures, water stress, or nutrient deficiencies may shorten the length. Excessive cleaning and/or drying at the gin may also result in shorter fiber length. Fiber

length affects yarn strength, yarn evenness, and the efficiency of the spinning process. The fineness of the yarn which can be successfully produced from given fibers is also influenced by the length of the fiber.

Source: http://www.cottoninc.com/; 1 inch = 2.54 cm

[106] An "industrially relevant fiber length", as used herein, refers to a length of fibers of a specific cotton species which is on average at least equal to or not significantly smaller than the length of fibers of that specific cotton variety normally observed. For G. hirsutum, an industrially relevant fiber length is reported to vary from about 7/8 to 1 5/16 inches (about 22 - about 33 mm). For G. barbadense, an industrially relevant fiber length is reported to vary from 1 1/4 to 1 9/16 inches (about 32 - about 40 mm). For G. herbaceum (AA) and G. arboreum (AA), an industrially relevant fiber length is reported to vary from 1/2 to 1 inch (about 13 - about 25 mm).

[107] Whenever reference to a "plant" or "plants" according to the invention is made, it is understood that also plant parts (cells, tissues or organs, seeds, fibers, severed parts such as roots, leaves, flowers, pollen, etc.), progeny of the plants which retain the distinguishing characteristics of the parents (especially the fiber properties), such as seed

obtained by selfing or crossing, e.g. hybrid seed (obtained by crossing two inbred parental lines), hybrid plants and plant parts derived there from are encompassed herein, unless otherwise indicated.

[108] The term "fiber strength allele detection assay" refers herein to an assay that indicates (directly or indirectly) the presence or absence of specific alleles of the fiber strength locus of the present invention. In one embodiment it allows one to determine whether a particular fiber strength allele is homozygous or heterozygous at the locus in any individual plant.

[109] hi another aspect of the invention, methods are provided for generating and/or selecting Gossypium plants, and parts and progeny thereof, comprising at least one superior allele of the fiber strength locus.

[110] hi one embodiment, the superior allele of the fiber strength locus is the Gossypium barbadense allele and the method comprises the step of identifying a Gossypium plant that comprises the Gossypium barbadense fiber strength allele based on the presence of Gossypium barbadense alleles of markers linked to the fiber strength locus, such as the markers linked to the Gossypium barbadense fiber strength allele indicated above and in Table 6 and 13.

[I ll] In a particular aspect, the method comprises the step of determining the presence of Gossypium barbadense alleles of markers linked to the fiber strength locus in the genomic DNA of a plant selected from the group consisting of: AFLP marker P5M50- M126.7, SSR marker CIR280, SSR marker BNL3992, SSR marker CIR401c, SSR marker NAU861, a polymorphic site in a genomic DNA sequence of the plant corresponding to a genomic DNA sequence comprised in SEQ ID NO: 53, and a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUC 1.1 A gene of SEQ ID NO: 5, such as the SNP markers indicated as GLUC1.1A-SNP2, 3, 5, 6 and 8 below and in Table 13.

[112] In a further embodiment, the superior allele of the fiber strength locus is the Gossypium danvinii allele and the method comprises the step of identifying a Gossypium plant that comprises the Gossypium danvinii fiber strength allele based on the presence of Gossypium danvinii alleles of markers linked to the fiber strength locus, such as the markers linked to the Gossypium danvinii fiber strength allele indicated above and in Table 13.

[113] In a particular aspect, the method comprises the step of determining the presence of a Gossypium danvinii allele of a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUCl. IA gene of SEQ ID NO: 56, such as the SNP markers indicated as GLUC1.1A-SNP2, 3, 5, 6 and 8 below and in Table 13.

[114] In a further embodiment, the superior allele of the fiber strength locus is the Gossypium arboreum allele and the method comprises the step of identifying a Gossypium plant that comprises the Gossypium arboreum fiber strength allele based on the presence of Gossypium arboreum alleles of markers linked to the fiber strength locus, such as the markers linked to the Gossypium arboreum fiber strength allele indicated above and in Table 13.

[115] In a particular aspect, the method comprises the step of determining the presence of a Gossypium arboreum allele of a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUC 1.1 A gene of SEQ ID NO: 21, such as the SNP marker indicated as GLUC1.1A-SNP7 below and in Table 13.

[116] Markers linked to the fiber strength locus can be used for marker assisted selection (MAS) or map based cloning of the fiber strength locus. MAS involves screening plants for the presence or absence of linked markers. In particular plants are screened for the presence of markers flanking the locus or gene or linked to the locus or

gene. Based on the presence/absence of the marker(s) plants are selected or discarded during the breeding program. MAS can significantly speed up breeding programs and introgression of a particular locus or gene into another genetic background, and can also reduce problems with genotype x environment interactions. MAS is also useful in combining different fiber strength loci in one plant. The presence or absence of a specific fiber strength allele, such as the Gossypium barbadense fiber strength allele, can be inferred from the presence or absence of molecular markers, such as the AFLP and SSR markers indicated above (see for example Table 2) or markers derived from them, linked to the specific allele. For example, Gossypium barbadense plants, in particular Gossypium barbadense cv. Pima S7 plants, may be crossed to Gossypium hirsutum plants and progeny plants from this cross are then screened for the presence of one or more AFLP and/or SSR markers linked to the Gossypium barbadense fiber strength allele, for example, by using the barbadense allele identification protocol.

[117] Breeding procedures such as crossing, selfing, and backcrossing are well known in the art [see Allard RW (1960) Principles of Plant Breeding. John Wiley & Sons, New York, and Fehr WR (1987,) Principles of Cultivar Development, Volume 1, Theory and Techniques, Collier Macmillan Publishers, London. ISBN 0-02-949920-8]. Superior alleles of the fiber strength locus, such as the Gossypium barbadense fiber strength allele, can be transferred into other breeding lines or varieties either by using traditional breeding methods alone or by using additionally MAS. In traditional breeding methods the increased callose content and/or increased fiber strength phenotype is assessed in the field or in controlled environment tests in order to select or discard plants comprising or lacking the superior fiber strength allele. Different crosses can be made to transfer the superior fiber strength allele, such as the Gossypium barbadense fiber strength allele, into lines of other Gossypium species or varieties, such as A genome diploid Gossypium plant lines, such as Gossypium herbaceum or Gossypium arboreum plant lines, or in AD allotetraploid Gossypium plant lines, such as Gossypium hirsutum and Gossypium barbadense plant lines, in particularly in Gossypium barbadense plant lines different from the Pima S7 variety. The breeding program may involve crossing to generate an Fl (first filial generation), followed by several generations of selfing (generating F2, F3,

etc.)- The breeding program may also involve backcrossing (BC) steps, whereby the offspring are backcrossed to one of the parental lines (termed the recurrent parent). Breeders select for agronomically important traits, such as high yield, high fiber quality, disease resistance, etc., and develop thereby elite breeding lines (lines with good agronomic characteristics). In addition, plants are bred to comply with fiber quality standards, such as American Pima or American Upland fiber quality.

[118] The "barbadense or hirsutum allele identification protocol", as used herein, refers to the identification of the Gossypium barbadense and/or Gossypium hirsutum allele of the fiber strength locus comprising the steps of: extracting DNA from plant tissue such as leaf tissue or seeds and carrying out an analysis of linked markers, such as an AFLP and/or SSR analysis for one or more of the linked AFLP and/or SSR markers, using, for example, specific primer pairs to identify the barbadense or hirsutum allele, such as those indicated in Table 2. The barbadense or hirsutum allele identification protocol may be carried out on DNA obtained from individual plants or on DNA obtained from bulks (or pools). In one embodiment kits for detecting the presence of the Gossypium barbadense and/or Gossypium hirsutum fiber strength allele in Gossypium DNA are provided. Such a kit comprises, for example, primers or probes able to detect a DNA marker, such as an AFLP and/or an SSR marker, linked to the Gossypium barbadense and/or Gossypium hirsutum fiber strength allele. The kit may further comprise samples, which can be used as positive or negative controls and additional reagents for AFLP and/or SSR analysis. The samples may be tissue samples or DNA samples. As positive control may, for example, Gossypium barbadense seeds, in particular from cv. Pima S7, be included. As negative controls may, for example, Gossypium hirsutum seeds, in particular from cv. FM966, be included.

[1 19] In a further aspect, methods are provided to distinguish between the presence of superior and non-superior alleles of the fiber strength locus. In one embodiment, methods are provided to distinguish between the presence of the Gossypium barbadense allele and the Gossypium hirsutum allele comprising the step of determining the presence of Gossypium barbadense and/or Gossypium hirsutum alleles of markers linked to the fiber

strength locus, such as the markers linked to the fiber strength locus indicated above, for example, those indicated in Table 2 and Table 13.

[120] Thus, in one embodiment, a method is provided for distinguishing between the presence of the Gossypium barbadense and Gossypium hirsutum fiber strength alleles by determining the presence of Gossypium barbadense and Gossypium hirsutum alleles of markers linked to the fiber strength locus in the genomic DNA of a plant selected from the group consisting of: AFLP marker P5M50-M 126.7, SSR marker CIR280, SSR marker BNL3992, SSR marker CIR401, SSR marker NAU861, a polymorphic site in a genomic DNA sequence of the plant corresponding to a genomic DNA sequence comprised in SEQ ID NO: 53, and a polymorphic site in a nucleotide sequence of a GLUC 1.1 A gene in the genomic DNA of the plant corresponding to the nucleotide sequence of a GLUC 1.1 A gene of SEQ ID NO: 5, such as the SNP markers indicated as GLUC1.1A-SNP2, 3, 5, 6 and 8 below and in Table 13.

[121] According to another aspect of the invention, methods are provided for altering the callose content of a fiber in a Gossypium plant, particularly increasing the callose content of a fiber, comprising the step of introgressing a superior allele of the cotton fiber strength locus on chromosome A05, such as the Gossypium barbadense allele, in the Gossypium plant.

[122] According to yet another aspect of the invention, methods are provided for altering the properties of a fiber in a Gossypium plant, particularly increasing the strength of a fiber, comprising the step of introgressing a superior allele of the cotton fiber strength locus on chromosome A05, such as the Gossypium barbadense allele, in the Gossypium plant.

[123] The current invention is further based on the unexpected finding that the functionality and the timing of expression of the GLUC 1.1 A gene, which was located in the support interval of the strength locus, differ between G. hirsutum and G. barbadense. It was found that, while G. hirsutum plants comprise a GLUC 1.1 A gene which is

functionally expressed during the fiber strength buiding stage of fiber development, more particularly during the fiber maturation phase, G. barbadense plants comprise a GLUC 1.1 A gene which is non- functionally expressed during the fiber strength building phase. The GLUC 1.1 D gene on the other hand is functionally expressed during the entire fiber strength building stage in both Gossypium species. It was further found that addition of exogenous endo-l,3-beta-glucanase to fibers of Gossypium barbadense reduces the callose content and the strength of the fibers. Based on these findings, it is believed that the renown strength of the fibers of G. barbadense might be, at least in part, caused by a higher callose content in the fibers and that this higher callose content might be caused by the abscence of a functionally expressed A subgenome-specific fiber-specific endo-1,3- beta-glucanase gene. It is further believed that by abolishing the functional expression of specific alleles of GLUC genes during the fiber strength building stage in fiber-producing plants while maintaining the functional expression of specific other GLUC genes during the fiiber strength building stage, it is possible to fine tune the amount and/or type of functional GLUC proteins produced during the fiber strength building stage, thus influencing the degradation of callose in the fiber which in turn influences the strength and length of the fiber produced. It is believed that the absolute and relative amount of different GLUC proteins in fibers can thus be tuned in such a way so as to attain a proper balance between fiber length and strength.

[ 124] Thus, in a further aspect, the present invention provides a non-naturally occurring fiber-producing plant, and parts and progeny thereof, characterized in that the functional expression of at least one allele of at least one fiber-specific GLUC gene that is functionally expressed during the fiber strength building phase, in particular during the maturation phase of fiber development, is abolished.

[125] The term "gene" means a DNA sequence comprising a region (transcribed region), which is transcribed into an RNA molecule (e.g. into a pre-mRNA, comprising intron sequences, which is then spliced into a mature mRNA, or directly into a mRNA without intron sequences) in a cell, operable linked to regulatory regions (e.g. a promoter). A gene (genomic DNA) may thus comprise several operably linked

sequences, such as a promoter, a 5' leader sequence comprising e.g. sequences involved in translation initiation, a (protein) coding region (with introns) and a 3' non-translated sequence comprising e.g. transcription termination sites. "cDNA sequence" refers to a nucleic acid sequence comprising the 5' untranslated region, the coding region without introns and the 3' untranslated region and a polyA tail. "Endogenous gene" is used to differentiate from a "foreign gene", "transgene" or "chimeric gene", and refers to a gene from a plant of a certain plant genus, species or variety, which has not been introduced into that plant by transformation (i.e. it is not a "transgene"), but which is normally present in plants of that genus, species or variety, or which is introduced in that plant from plants of another plant genus, species or variety, in which it is normally present, by normal breeding techniques or by somatic hybridization, e.g., by protoplast fusion. Similarly, an "endogenous allele" of a gene is not introduced into a plant or plant tissue by plant transformation, but is, for example, generated by plant mutagenesis and/or selection, introgressed from another plant species by, e.g., marker-assisted selection, or obtained by screening natural populations of plants.

[126] "Expression of a gene" or "gene expression" refers to the process wherein a DNA region, which is operably linked to appropriate regulatory regions, particularly a promoter, is transcribed into an RNA molecule. The RNA molecule is then processed further (by post-transcriptional processes) within the cell, e.g. by RNA splicing and translation initiation and translation into an amino acid chain (polypeptide), and translation termination by translation stop codons. The term "functionally expressed" is used herein to indicate that a functional, i.e. biologically active, protein is produced; the term "not functionally expressed" to indicate that a protein with significantly reduced or no functionality (biological activity) is produced or that no or a significantly reduced amount of protein is produced.

[127] The term "fiber specific" or "fiber cell specific", with respect to the expression of a gene, refers to, for practical purposes, the highly specific, expression of a gene in fiber cells of plants, such as cotton plants. In other words, transcript levels of a DNA in tissues

different of fiber cells is either below the detection limit or very low (less than about 0.2 picogram per microgram total RNA).

[128] The term "fiber strenght building phase" commonly refers herein to the secondary cell wall synthesis and maturation phase of fiber development as defined above.

[129] The term "GLUC gene" refers herein to a nucleic acid sequence encoding an endo-l,3-beta-glucanase (GLUC) protein.

[130] The term "nucleic acid sequence" (or nucleic acid molecule) refers to a DNA or RNA molecule in single or double stranded form, particularly a DNA encoding a protein or protein fragment according to the invention. An "endogenous nucleic acid sequence" refers to a nucleic acid sequence within a plant cell, e.g. an endogenous (allele of a) GLUC gene present within the nuclear genome of a plant cell. An "isolated nucleic acid sequence" is used to refer to a nucleic acid sequence that is no longer in its natural environment, for example in vitro or in a recombinant bacterial or plant host cell.

[131] The terms "protein" and "polypeptide" are used interchangeably and refer to molecules consisting of a chain of amino acids, without reference to a specific mode of action, size, 3-dimensional structure or origin. A "fragment" or "portion" of a protein may thus still be referred to as a "protein". An "isolated protein" is used to refer to a protein that is no longer in its natural environment, for example in vitro or in a recombinant bacterial or plant host cell. "Amino acids" are the principal building blocks of proteins and enzymes. They are incorporated into proteins by transfer RNA according to the genetic code while messenger RNA is being decoded by ribosomes. During and after the final assembly of a protein, the amino acid content dictates the spatial and biochemical properties of the protein or enzyme. The amino acid backbone determines the primary sequence of a protein, but the nature of the side chains determines the protein's properties. "Similar amino acids", as used herein, refers to amino acids that have similar amino acid side chains, i.e. amino acids that have polar, non-polar or practically neutral side chains. "Non-similar amino acids", as used herein, refers to amino acids that

have different amino acid side chains, for example an amino acid with a polar side chain is non-similar to an amino acid with a non-polar side chain. Polar side chains usually tend to be present on the surface of a protein where they can interact with the aqueous environment found in cells ("hydrophilic" amino acids). On the other hand, "non-polar" amino acids tend to reside within the center of the protein where they can interact with similar non-polar neighbors ("hydrophobic" amino acids"). Examples of amino acids that have polar side chains are arginine, asparagine, aspartate, cysteine, glutamine, glutamate, histidine, lysine, serine, and threonine (all hydrophilic, except for cysteine which is hydrophobic). Examples of amino acids that have non-polar side chains are alanine, glycine, isoleucine, leucine, methionine, phenylalanine, proline, and tryptophan (all hydrophobic, except for glycine which is neutral).

[132] An "enzyme" is a protein comprising enzymatic activity, such as functional, i.e. biologically active, endo-l,3-beta-glucanase or glucan endo-l,3-beta-D-glucosidase (GLUC) proteins (EC 3.2.1.39). GLUC proteins belong to the glycosyl hydrolase family 17 (GH 17) enzyme grouping and are capable of hydrolyzing 1,3-beta-D-glucosidic linkages in 1,3-beta-D-glucans, including long chain 1,3-beta-D-glucans called callose (see also http://www.cazy.org/fam/GH17.html). The GH17 group is identified by the following amino acid recognition signature: [LIVMKS]-X-[LIVMFYWA](3)-[STAG]-E- [STACVI]-G-[WY]*-P-[STN]-X-[SAGQ], where E, such as Glu249 in GhGLUCl. IA (SEQ ID NO: 2 and 4) and similar or identical amino acids in other GLUC 1.1 proteins (for example as indicated in Figure 7), is an active site residue. The GH 17 recognition signal of GLUC 1.1 enzymes, as described herein, further contains a conserved tryptophan (W) residue at the position indicated with *, such as Trp252 in GhGLUC 1.1 A (SEQ ID NO: 2 and 4) and similar or identical amino acids in other GLUC 1.1 proteins (for example as indicated in Figure 7), which is predicted to be involved in the interaction with the glucan substrate.

[133] In one embodiment, the fiber-specific GLUC gene that is functionally expressed during the fiber strength building phase, is a GLUC 1.1 gene.

[134] The term "GLUC 1.1 gene" refers herein to a nucleic acid sequence encoding a GLUC 1.1 protein. In particular, a "GLUCl.1 gene", as used herein, refers to a GLUC gene encoding a cDNA sequence with at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, 100% sequence identity to SEQ ID NO: 3 or comprises a coding sequence with at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, 100% sequence identity to the nucleotide at position 2410 to the nucleotide at position 3499 of SEQ ID NO: 1.

[135] A "GLUC 1.1 protein", as used herein, refers to a GLUC protein that has at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99%, 1000% sequence identity to SEQ ID NO: 4.

[136] A functional "GLUC 1.1 protein", as used herein, refers to a GLUC 1.1 protein that is capable of hydro lyzing 1,3-beta-D-glucosidic linkages in 1,3-beta-D-glucans, that has at least 90%, at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 4 and that comprises amino acid residues similar to the active site residues of the GLUC 1.1 protein of SEQ ID NO:4. A non-functional "GLUC 1.1 protein", as used herein, refers to a GLUC 1.1 protein that is not capable of hydrolyzing 1,3-beta- D-glucosidic linkages in 1,3-beta-D-glucans. In particular, a non- functional GLUC 1.1 protein lacks one or more amino acid residues similar to the active site residues of the GLUC 1.1 protein of SEQ ID NO:4.

[137] An "active site" or "catalytic site", as used herein, refers to a position on the three-dimensional structure of an enzyme which is involved in substrate binding, such as binding of 1,3-beta-D-glucans to GLUC enzymes, and in the biological activity of the enzyme, such as the hydrolyzation of 1,3-beta-D-glucosidic linkages in 1,3-beta-D- glucans of GLUC enzymes. "Active site (amino acid) residues", as used herein, refer to amino acid residues that are located within the active site of an enzyme and play a crucial role in substrate binding or in enzyme activity. A "glycosylation site", as used herein, refers to a position on the three-dimensional structure of an enzyme which is

glycosylated, i.e. a site to which (branched) oligosaccharides bind which may function in increasing stability, such as thermostability, of the protein. "Glycosylation site (amino acid) residues", as used herein, refer to amino acid residues within the glycosylation site of an enzyme to which (branched) oligosaccharides bind. Predictions of the three- dimensional structure of the endo-l,3-beta-glucanase enzymes as described herein indicate that the active site and the glycosylation site of the barley 1,3-1,4-beta-glucanase (as described by Mύller et al., 1998, J of Biol Chem 273 (6): 3438-3446; called "laqO" in the Protein Data Bank, which is freely available at http://www.rcsb.org/pdb/) are conserved, for example, in the Gossypium hirsutum GLUCl. IA and D, the Gossypium barbadense GLUC 1.1 D and the Gossypium herbaceum GLUC 1.1 A proteins as described herein, while the Gossypium barbadense GLUC 1.1 A protein, the Gossypium darwinii GLUC 1.1 A protein, and the Gossypium arbor eum GLUC 1.1 A protein as described herein lack most conserved amino acids located within these sites these sites (see, e.g., Table 4, Figure 3 and Examples). Active site and glycosylation residues in other GLUC 1.1 proteins can be determined by aligning the amino acid sequences of the different GLUC 1.1 proteins with the GLUC 1.1 proteins of the present invention, such as the amino acid sequence of GhGLUC 1.1 A in SEQ ID NO:4, and identifying identical or similar residues in the other GLUC 1.1 proteins.

Table 4: Amino acid regions and positions of active site residues and glycosylation site residues in GLUC 1.1 A and D proteins of the three principal groups of cotton of commercial interest

exon 2 Active site residue

Tyr33 Tyr48 TyrόO TyrόO TyrόO

Glu232 Glu249 Glu261 Glu261 Trp252 Tφ264 Tφ264

Glu288 Glu3O8 Glu320 Glu320 Glycosylation site residue: Asnl90 Asn202 ND Asn214

-: not present; ND: not determined

[138] The terms "target peptide", "transit peptide" or "signal peptide" refer to amino acid sequences which target a protein to intracellular organelles. The GLUCl.1 proteins as described herein comprise a signal peptide at their N-terminal end, such as the amino acid sequence indicated before the putative post-translational splicing site in Figures 2 and 7. "Mature protein" refers to a protein without the signal peptide, such as the GLUC 1.1 proteins as described herein without the amino acid sequence indicated before the putative post-translational splicing site in Figures 2 and 7. "Precursor protein" or "preproenzyme" refers to the mature protein with its signal peptide.

[139] In another embodiment, the fiber-producing plant is a Gossypium plant. In a particular aspect, the Gossypium GLUC 1.1 allele is a GLUC 1.1 A or D allele.

[140] A "GLUCl. IA gene", as used herein, refers to a GLUCl.1 gene located on the A subgenome of a Gossypium diploid or allotetraploid species ("GLUC 1.1 A locus") and encoding a GLUC 1.1 A protein. In particular, a GLUC 1.1 A gene encodes a cDNA sequence with at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 3 or comprises a coding sequence with at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to the nucleotide at position 2410 to the nucleotide at position 3499 of SEQ ID NO: 1. Similarly, a "GLUCl. ID gene", as used herein, refers to

a GLUCl.1 gene located on the D subgenome of a Gossypium diploid or allotetraploid species(" GLUC 1.1 D locus") and encoding a GLUCl. ID protein. In particular, a GLUCl. ID gene encodes a cDNA sequence with at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 9 or comprises a coding sequence with at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to the nucleotide at position 3337 to the nucleotide at position 4444 of SEQ ID NO: 7.

[141] A "GLUC LlA protein", as used herein, refers to a GLUC 1.1 protein encoded by a GLUC 1.1 gene located on the A subgenome of a Gossypium diploid or allotetraploid species and having at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 4. Similarly, a "GLUCl. ID protein", as used herein, refers to a GLUC protein encoded by a GLUC 1.1 gene located on the D subgenome of a Gossypium diploid or allotetraploid species and having at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 10.

[142] In another embodiment the fiber-producing plant is a Gossypium hirsutum plant. In a particular aspect, the Gossypium hirsutum GLUC 1.1 allele is a GhGLUC 1.1 A or a GhGLUCl. ID allele, preferably a GhGLUCl. IA allele.

[143] As described in WO2008/083969, the GLUCl. IA and GLUCl. ID genes of Gossypium hirsutum can be distinguished by the presence of a cleaved amplified polymorphic sequence (CAPS) marker using an AIwI restriction enzyme recognition site present in the nucleotide sequence of GhGLUCl. IA that is absent in the nucleotide sequence of GhGLUCl. ID and by their timing of expression: whereas the GhGLUCLlD is expressed during the entire fiber strength building phase (from about 14 to 17 DPA on depending on growth conditions), onset of GhGLUCl. IA is delayed until the beginning of the late fiber maturation phase (about 30-40 DPA depending on growth conditions). The GLUCl. IA and GLUCl. ID genes of Gossypium barbadense can also be distinguished by the presence of the CAPS marker using the AIwI restriction enzyme recognition site present in the nucleotide sequence of GbGLUC 1.1 A that is absent in the nucleotide sequence of GbGLUCl. ID. Both genes are however expressed during the

entire fiber strength building phase (from about 14 to 17 DPA on depending on growth conditions). The level of expression of GbGLUCLlA is however much lower than the level of expression of GbGLUClJD.

[144] In one embodiment, the functional expression of the at least one GLUC allele is abolished by mutagenesis.

[145] "Mutagenesis", as used herein, refers to the process in which plant cells (e.g., Gossypium seeds or other parts, such as pollen, etc.) are subjected to a technique which induces mutations in the DNA of the cells, such as contact with a mutagenic agent, such as a chemical substance (such as ethylmethylsulfonate (EMS), ethylnitrosourea (ENU), etc.) or ionizing radiation (neutrons (such as in fast neutron mutagenesis, etc.), alpha rays, gamma rays (such as that supplied by a Cobalt 60 source), X-rays, UV-radiation, etc.), or a combination of two or more of these. Thus, the desired mutagenesis of one or more GLUC alleles may be accomplished by use of chemical means such as by contact of one or more plant tissues with ethylmethylsulfonate (EMS), ethylnitrosourea, etc., by the use of physical means such as x-ray, etc, or by gamma radiation, such as that supplied by a Cobalt 60 source. While mutations created by irradiation are often large deletions or other gross lesions such as translocations or complex rearrangements, mutations created by chemical mutagens are often more discrete lesions such as point mutations. For example, EMS alkylates guanine bases, which results in base mispairing: an alkylated guanine will pair with a thymine base, resulting primarily in G/C to A/T transitions. Following mutagenesis, Gossypium plants are regenerated from the treated cells using known techniques. For instance, the resulting Gossypium seeds may be planted in accordance with conventional growing procedures and following self-pollination seed is formed on the plants. Additional seed that is formed as a result of such self-pollination in the present or a subsequent generation may be harvested and screened for the presence of mutant GLUC alleles. Several techniques are known to screen for specific mutant alleles, e.g., Deleteagene™ (Delete-a-gene; Li et al, 2001, Plant J 27: 235-242) uses polymerase chain reaction (PCR) assays to screen for deletion mutants generated by fast neutron mutagenesis, TILLING (targeted induced local lesions in genomes; McCallum et al,

2000, Nat Biotechnol 18:455-457) identifies EMS-induced point mutations, etc. Additional techniques to screen for the presence of specific mutant GLUC alleles are described in the Examples below.

[146] "Wild type" (also written "wildtype" or "wild-type"), as used herein, refers to a typical form of a plant or a gene as it most commonly occurs in nature. A "wild type plant" refers to a plant with the most common phenotype of such plant in the natural population. A "wild type allele" refers to an allele of a gene required to produce the wild- type phenotype. By contrast, a "mutant plant" refers to a plant with a different rare phenotype of such plant in the natural population or produced by human intervention, e.g. by mutagenesis, and a "mutant allele" refers to an allele of a gene required to produce the mutant phenotype.

[147] As used herein, the term "wild type GLUC (e.g. wild type GLUCl. IA or GLUCl. ID), means a naturally occurring GLUC allele found within plants, in particular Gossypium plants, which encodes a functional GLUC protein (e.g. a functional GLUCl. IA or GLUCl. ID, respectively). In contrast, the term "mutant GLUC (e.g. mutant GLUC 1.1 A or GLUCl. ID), as used herein, refers to a GLUC allele, which does not encode a functional GLUC protein, i.e. a GLUC allele encoding a non-functional GLUC protein (e.g. a non- functional GLUC 1.1 A or GLUC 1.1 D, respectively), which, as used herein, refers to a GLUC protein having no biological activity or a significantly reduced biological activity as compared to the corresponding wild-type functional GLUC protein, or encoding no GLUC protein or a significantly reduced amount of GLUC protein. Such a "mutant GLUC allele" is a GLUC allele, which comprises one or more mutations in its nucleic acid sequence, whereby the mutation(s) preferably result in a significantly reduced (absolute or relative) amount of functional GLUC protein in the cell in vivo. As used herein, a "full knock-out GLUCl. IA allele" is a mutant GLUCl. IA allele the presence of which in homozygous state in the plant (e.g. a Gossypium hirsutum plant with two full knock-out GLUCl. IA alleles and two wild-type GLUCl. ID alleles) results in an increase of fiber strength in that plant. Mutant alleles of the GLUC protein- encoding nucleic acid sequences are designated as "glue" (e.g. glue J. Ia or glucl.ld,

respectively) herein. Mutant alleles can be either "natural mutant" alleles, which are mutant alleles found in nature (e.g. produced spontaneously without human application of mutagens), such as the Gossypium barbadense GLUCl. IA allele, the Gossypium darwinii GLUCl. IA allele, and the Gossypium arboreum GLUC 1.1 A allele, or "induced mutant" alleles, which are induced by human intervention, e.g. by mutagenesis.

[148] Thus in one aspect of the embodiment, GLUC mutant plants are provided herein, whereby the mutant alleles are selected from the GLUC 1.1 A and/or GLUCLlD genes. Thus in a particular aspect, the genotype of these GLUC mutant plants can be described as: GLUCl. IA/ glucl.la; GLUCl. ID/ glucl.ld; GLUCl. IA/ glucl.la, GLUCl. ID/ GLUCl. ID; or GLUCl. IA/ GLUCl. IA, GLUCl. ID/ glucl.ld.

[149] In a further aspect of the embodiment, homozygous GLUC mutant plants or plant parts are provided, whereby the mutant alleles are selected from the GLUC 1.1 A and GLUCl. ID genes. Thus in a particular aspect, homozygous GLUC mutant plants are provided herein, wherein the genotype of the plant can be described as: glucl.la / glucl.la; glucl.ld / glucl.ld; glucl.la/glucl.la, GLUCl.1D/GLUC1. ID or GLUC 1.1 A/ GLUCl. IA, glucl.ld /glucl.ld.

[150] hi a further aspect of the invention the homozygous GLUC mutant plants or plant parts comprise a. further mutant allele, wherein the mutant plants or plant parts are heterozygous for the additional mutant GLUC allele. Thus in a further particular aspect, homozygous GLUC mutant plants comprising one further mutant GLUC allele are provided herein, wherein the genotype of the plant can be described as: GLUCl.1 -A/ glucl.1-a, glue J.1-d /glucl.1-d or glucl. la/glucl. Ia, GLUCl. ID/ glucl.ld.

[151] In another embodiment, the functional expression of the at least one GLUC allele is abolished by introgression of a non-functionally expressed orthologous GLUC allele or of a mutagenized allele of the GLUC gene.

[152] In one aspect of this embodiment, the non- functionally expressed orthologous GLUC allele can be isolated from specific cotton species, for example from Gossypium barbadense, darwinii or arboreum.

[153] In yet another embodiment, the functional expression of the at least one allele of the GLUC gene is abolished by introduction of a chimeric gene comprises the following operably linked DNA elements:

(a) a plant expressible promoter,

(b) a transcribed DNA region, which when transcribed yields an inhibitory RNA molecule capable of reducing the expression of the GLUC allele, and

(c) a 3' end region comprising transcription termination and polyadenylation signals functioning in cells of the plant.

[154] Several methods are available in the art to produce an inhibitory or a silencing RNA molecule, i.e. an RNA molecule which when expressed reduces the expression of a particular gene or group of genes, including the so-called "sense" or "antisense" RNA technologies.

[155] Thus in one embodiment, the inhibitory RNA molecule encoding chimeric gene is based on the so-called antisense technology. In other words, the coding region of the chimeric gene comprises a nucleotide sequence of at least 19 or 20 consecutive nucleotides of the complement of the nucleotide sequence of the GLUC allele. Such a chimeric gene may be constructed by operably linking a DNA fragment comprising at least 19 or 20 nucleotides from the GLUC allele, isolated or identified as described elsewhere in this application, in inverse orientation to a plant expressible promoter and 3' end formation region involved in transcription termination and polyadenylation.

[156] In another embodiment, the inhibitory RNA molecule encoding chimeric gene is based on the so-called co-suppression technology. In other words, the coding region of the chimeric gene comprises a nucleotide sequence of at least 19 or 20 consecutive nucleotides of the nucleotide sequence of the GLUC allele. Such a chimeric gene may be

constructed by operably linking a DNA fragment comprising at least 19 or 20 nucleotides from the GLUC allele, in direct orientation to a plant expressible promoter and 3' end formation region involved in transcription termination and polyadenylation.

[157] The efficiency of the above mentioned chimeric genes in reducing the expression of the GLUC allele may be further enhanced by the inclusion of a DNA element which results in the expression of aberrant, unpolyadenylated inhibitory RNA molecules or results in the retention of the inhibitory RNA molecules in the nucleus of the cells. One such DNA element suitable for that purpose is a DNA region encoding a self-splicing ribozyme, as described in WO 00/01133 (incorporated by reference). Another such DNA element suitable for that purpose is a DNA region encoding an RNA nuclear localization or retention signal, as described in WO03/076619 (incorporated by reference).

[158] A convenient and very efficient way of downregulating the expression of a gene of interest uses so-called double-stranded RNA (dsRNA) or interfering RNA (RNAi), as described e.g. in WO99/53050 (incorporated by reference). In this technology, an RNA molecule is introduced into a plant cell, whereby the RNA molecule is capable of forming a double stranded RNA region over at least about 19 to about 21 nucleotides, and whereby one of the strands of this double stranded RNA region is about identical in nucleotide sequence to the target gene ("sense region"), whereas the other strand is about identical in nucleotide sequence to the complement of the target gene or of the sense region ("antisense region"). It is expected that for silencing of the target gene expression, the nucleotide sequence of the 19 consecutive nucleotide sequences may have one mismatch, or the sense and antisense region may differ in one nucleotide. To achieve the construction of such RNA molecules or the encoding chimeric genes, use can be made of the vector as described in WO 02/059294.

[159] Thus, in one aspect of the embodiment, the chimeric gene comprises the following operably linked DNA elements:

(a) a plant expressible promoter, preferably a plant expressible promoter which controls transcription preferentially in the fiber cells;

(b) a transcribed DNA region, which when transcribed yields a double-stranded RNA molecule capable of reducing the expression of the GLUC allele and the RNA molecule comprising a first and second RNA region wherein i) the first RNA region comprises a nucleotide sequence of at least 19 consecutive nucleotides having at least about 94% sequence identity to the nucleotide sequence of the GLUC allele; ii) the second RNA region comprises a nucleotide sequence complementary to the at least 19 consecutive nucleotides of the first RNA region; iii) the first and second RNA region are capable of base-pairing to form a double stranded RNA molecule between at least the 19 consecutive nucleotides of the first and second region; and

(c) a 3' end region comprising transcription termination and polyadenylation signals functioning in cells of the plant.

[160] The length of the first or second RNA region (sense or antisense region) may vary from about 19 nucleotides (nt) up to a length equaling the length (in nucleotides) of the GLUC allele. The total length of the sense or antisense nucleotide sequence may thus be at least about 25 nt, or at least about 50 nt, or at least about 100 nt, or at least about 150 nt, or at least about 200 nt, or at least about 500 nt. It is expected that there is no upper limit to the total length of the sense or the antisense nucleotide sequence. However for practical reasons (such as e.g. stability of the chimeric genes) it is expected that the length of the sense or antisense nucleotide sequence should not exceed 5000 nt, particularly should not exceed 2500 nt and could be limited to about 1000 nt.

[161] It will be appreciated that the longer the total length of the sense or antisense region, the less stringent the requirements for sequence identity between these regions and the corresponding sequence in the GLUC allele or its complement. Preferably, the nucleic acid of interest should have a sequence identity of at least about 75% with the corresponding target sequence, particularly at least about 80 %, more particularly at least about 85%, quite particularly about 90%, especially about 95%, more especially about 100%, quite especially be identical to the corresponding part of the target sequence or its

complement. However, it is preferred that the nucleic acid of interest always includes a sequence of about 19 consecutive nucleotides, particularly about 25 nt, more particularly about 50 nt, especially about 100 nt, quite especially about 150 nt with 100% sequence identity to the corresponding part of the target nucleic acid. Preferably, for calculating the sequence identity and designing the corresponding sense or antisense sequence, the number of gaps should be minimized, particularly for the shorter sense sequences.

[162] For the purpose of this invention, the "sequence identity" of two related nucleotide or amino acid sequences, expressed as a percentage, refers to the number of positions in the two optimally aligned sequences which have identical residues (xlOO) divided by the number of positions compared. A gap, i.e., a position in an alignment where a residue is present in one sequence but not in the other, is regarded as a position with non-identical residues. The "optimal alignment" of two sequences is found by aligning the two sequences over the entire length according to the Needleman and Wunsch global alignment algorithm (Needleman and Wunsch, 1970, J MoI Biol 48(3):443-53) in The European Molecular Biology Open Software Suite (EMBOSS, Rice et al, 2000, Trends in Genetics 16(6): 276 — 277; see e.g. http://www.ebi.ac.uk/emboss/align/index.html) using default settings (gap opening penalty = 10 (for nucleotides) / 10 (for proteins) and gap extension penalty = 0.5 (for nucleotides) / 0.5 (for proteins)). For nucleotides the default scoring matrix used is EDNAFULL and for proteins the default scoring matrix is EBLOSUM62.

[163] "Substantially identical", "essentially similar", or "corresponding to", as used herein, refers to sequences, which, when optimally aligned as defined above, share at least a certain minimal percentage of sequence identity (as defined further below). "(A nucleotide or a nucleotide sequence) at a position corresponding to a position of (a nucleotide or a nucleotide sequence in a specific nucleotide sequence)", as used herein, refers to (nucleotides or nucleotide sequences) of two essentially similar sequences, which are aligned with each other in an optimal alignment of the two essentially similar sequences.

[164] dsRNA encoding chimeric genes according to the invention may comprise an intron, such as a heterologous intron, located e.g. in the spacer sequence between the sense and antisense RNA regions in accordance with the disclosure of WO 99/53050 (incorporated herein by reference).

[165] It is preferred for the current invention that the target specific gene sequence included in the antisense, sense or double stranded RNA molecule comprises at least one nucleotide, and preferably more which are specific for the specific GLUC allele whose expression is to be downregulated. Such specific nucleotides are indicated at least in figure 6 by the gray boxes.

[166] In a preferred embodiment, the inhibitory RNA molecule is specifically adapted to downregulate the A-subgenomic allele of the GLUC 1.1 gene. In another preferred embodiment, the biologically active RNA is specifically adapted to downregulate the D subgenome-specific allele of the GLUCl.1 gene.

[167] The use of synthetic micro-RNA's to downregulate expression of a particular gene in a plant cell, provides for very high sequence specificity of the target gene, and thus allows conveniently to discriminate between closely related alleles as target genes the expression of which is to be downregulated.

[168] Thus, in another embodiment of the invention, the inhibitory RNA or silencing RNA or biologically active RNA molecule may be a microRNA molecule, designed, synthesized and/or modulated to target and cause the cleavage of specific subgenomic alleles, preferably the A subgenomic allele of the GLUC 1.1 gene in a fiber producing plant, such as a cotton plant. Various methods have been described to generate and use miRNAs for a specific target gene (including but not limited to Schwab et al. (2006, Plant Cell, 18(5):1121-1 133), WO2006/044322, WO2005/047505, EP 06009836, incorporated by reference). Usually, an existing miRNA scaffold is modified in the target gene recognizing portion so that the generated miRNA now guides the RISC complex to cleave the RNA molecules transcribed from the target nucleic acid. miRNA scaffolds

could be modified or synthesized such that the miRNA now comprises 21 consecutive nucleotides of one of the subgenomic alleles of the fiber selective β-1,3 endoglucanase encoding nucleotide sequence, such as the sequences represented in the Sequence listing of WO2008/083969, and allowing mismatches according to the herein below described rules.

[169] Thus, in one embodiment, the invention provides a chimeric gene comprising the following operably linked DNA regions:

(a) a plant expressible promoter;

(b) a DNA region which upon introduction and transcription in a plant cell is processed into a miRNA, whereby the miRNA is capable of recognizing and guiding the cleavage of the mRNA of a GLUC allele of the plant but not another GLUC allele, such as the mRNA of the A subgenome specific GLUC allele but not the D subgenome specific GLUC allele; andoptionally,

(c) a 3' DNA region involved in transcription termination and polyadenylation.

[170] The mentioned DNA region processed into a miRNA may comprise a nucleotide sequence which is essentially complementary to a nucleotide sequence of at least 21 consecutive nucleotides of a GLUC allele, provided that one or more of following mismatches are allowed: a mismatch between the nucleotide at the 5' end of the miRNA and the corresponding nucleotide sequence in the RNA molecule; a mismatch between any one of the nucleotides in position 1 to position 9 of the miRNA and the corresponding nucleotide sequence in the RNA molecule; three mismatches between any one of the nucleotides in position 12 to position 21 of the miRNA and the corresponding nucleotide sequence in the RNA molecule provided that there are no more than two consecutive mismatches.

[171] As used herein, a "miRNA" is an RNA molecule of about 20 to 22 nucleotides in length which can be loaded into a RISC complex and direct the cleavage of another RNA molecule, wherein the other RNA molecule comprises a nucleotide sequence essentially complementary to the nucleotide sequence of the miRNA molecule whereby one or more

of the following mismatches may occur: a mismatch between the nucleotide at the 5' end of said miRNA and the corresponding nucleotide sequence in the target RNA molecule; a mismatch between any one of the nucleotides in position 1 to position 9 of said miRNA and the corresponding nucleotide sequence in the target RNA molecule; three mismatches between any one of the nucleotides in position 12 to position 21 of said miRNA and the corresponding nucleotide sequence in the target RNA molecule provided that there are no more than two consecutive mismatches, no mismatch is allowed at positions 10 and 11 of the miRNA (all miRNA positions are indicated starting from the 5' end of the miRNA molecule).

[172] A miRNA is processed from a "pre-miRNA" molecule by proteins, such as DCL proteins, present in any plant cell and loaded onto a RISC complex where it can guide the cleavage of the target RNA molecules.

[173] As used herein, a "pre-miRNA" molecule is an RNA molecule of about 100 to about 200 nucleotides, preferably about 100 to about 130 nucleotides which can adopt a secondary structure comprising a double stranded RNA stem and a single stranded RNA loop and further comprising the nucleotide sequence of the miRNA (and its complement sequence) in the double stranded RNA stem. Preferably, the miRNA and its complement are located about 10 to about 20 nucleotides from the free ends of the miRNA double stranded RNA stem. The length and sequence of the single stranded loop region are not critical and may vary considerably, e.g. between 30 and 50 nt in length. Preferably, the difference in free energy between unpaired and paired RNA structure is between -20 and -60 kcal/mole, particularly around -40 kcal/mole. The complementarity between the miRNA and the miRNA* need not be perfect and about 1 to 3 bulges of unpaired nucleotides can be tolerated.The secondary structure adopted by an RNA molecule can be predicted by computer algorithms conventional in the art such as mFOLD. The particular strand of the double stranded RNA stem from the pre-miRNA which is released by DCL activity and loaded onto the RISC complex is determined by the degree of complementarity at the 5' end, whereby the strand which at its 5' end is the least involved in hydrogen bounding between the nucleotides of the different strands of the cleaved

dsRNA stem is loaded onto the RISC complex and will determine the sequence specificity of the target RNA molecule degradation. However, if empirically the miRNA molecule from a particular synthetic pre-miRNA molecule is not functional (because the "wrong" strand is loaded on the RISC complex, it will be immediately evident that this problem can be solved by exchanging the position of the miRNA molecule and its complement on the respective strands of the dsRNA stem of the pre-miRNA molecule. As is known in the art, binding between A and U involving two hydrogen bounds, or G and U involving two hydrogen bounds is less strong that between G and C involving three hydrogen bounds.

[174] Naturally occurring miRNA molecules may be comprised within their naturally occurring pre-miRNA molecules but they can also be introduced into existing pre- miRNA molecule scaffolds by exchanging the nucleotide sequence of the miRNA molecule normally processed from such existing pre-miRNA molecule for the nucleotide sequence of another miRNA of interest. The scaffold of the pre-miRNA can also be completely synthetic. Likewise, synthetic miRNA molecules may be comprised within, and processed from, existing pre-miRNA molecule scaffolds or synthetic pre-miRNA scaffolds.

[175] The pre-miRNA molecules (and consequently also the miRNA molecules) can be conveniently introduced into a plant cell by providing the plant cells with a gene comprising a plant-expressible promoter operably linked to a DNA region, which when transcribed yields the pre-miRNA molecule. The plant expressible promoter may be the promoter naturally associated with the pre-miRNA molecule or it may be a heterologous promoter.

[176] Suitable miRNA and pre microRNA molecules for the specific downregulation of the expression of the GhGLUC 1.1 A gene are set forth in the sequence listing entries SEQ ID NO: 13, 14, 17, 18 and 19 of WO2008/083969.

[177] Suitable miRNA and pre microRNA molecules for the specific downregulation of the expression of the GhGLUC 1.1 D gene are set forth in the sequence listing entries SEQ ID NO: 15, 16, 20 and 21 of WO2008/083969.

[178] As used herein, the term "plant-expressible promoter" means a DNA sequence which is capable of controlling (initiating) transcription in a plant cell. This includes any promoter of plant origin, but also any promoter of non-plant origin which is capable of directing transcription in a plant cell, i.e., certain promoters of viral or bacterial origin such as the CaMV35S, the subterranean clover virus promoter No. 4 or No. 7, or T-DNA gene promoters and the like.

[179] A plant-expressible promoter that controls initiation and maintenance of transcription preferentially in fiber cells is a promoter that drives transcription of the operably linked DNA region to a higher level in fiber cells and the underlying epidermis cells than in other cells or tissues of the plant. Such promoters include the promoter from cotton from a fiber-specific β- tubulin gene (as described in WO0210377), the promoter from cotton from a fiber-specific actin gene(as described in WO0210413), the promoter from a fiber specific lipid transfer protein gene from cotton (as described in US5792933), a promoter from an expansin gene from cotton (WO9830698) or a promoter from a chitinase gene in cotton (US2003106097) or the promoters of the fiber specific genes described in US6259003 or US6166294. Fiber selective promoters as described herein may also be used.

[180] The invention also encompasses the chimeric genes herein described, as well as plants, seeds, tissues comprising these chimeric genes, and fibers produced from such plants.

[181] Methods to transform plants are well known in the art and are of minor relevance for the current invention. Methods to transform cotton plants are also well known in the art. Agrobacteri wm-mediated transformation of cotton has been described e.g. in US

patent 5,004,863 or in US patent 6,483,013 and cotton transformation by particle bombardment is reported e.g. in WO 92/15675.

[182] The chimeric genes according to the invention may be introduced into plants in a stable manner or in a transient manner using methods well known in the art. The chimeric genes may be introduced into plants, or may be generated inside the plant cell as described e.g. in EP 1339859.

[183] The chimeric genes may be introduced by transformation in cotton plants from which embryogenic callus can be derived, such as Coker 312, Coker310, Coker 5Acala SJ-5, GSC25110, FIBERMAX 819 , Siokra 1-3, T25, GSA75, Acala SJ2, Acala SJ4, Acala SJ5, Acala SJ-Cl, Acala B1644, Acala B1654-26, Acala B1654-43, Acala B3991, Acala GC356, Acala GC510, Acala GAMl, Acala Cl, Acala Royale, Acala Maxxa, Acala Prema, Acala B638, Acala B 1810, Acala B2724, Acala B4894, Acala B5002, non Acala "picker" Siokra, "stripper" variety FC2017, Coker 315, STONEVILLE 506, STONEVILLE 825, DP50, DP61, DP90, DP77, DESl 19, McN235, HBX87, HBX191, HBX107, FC 3027, CHEMBRED Al, CHEMBRED A2, CHEMBRED A3, CHEMBRED A4, CHEMBRED Bl, CHEMBRED B2, CHEMBRED B3, CHEMBRED Cl, CHEMBRED C2, CHEMBRED C3, CHEMBRED C4, PAYMASTER 145, HS26, HS46, SICALA, PIMA S6 ORO BLANCO PIMA, FIBERMAX FM5013, FIBERMAX FM5015, FIBERMAX FM5017, FIBERMAX FM989, FIBERMAX FM832, FIBERMAX FM966, FIBERMAX FM958, FIBERMAX FM989, FIBERMAX FM958, FIBERMAX FM832, FIBERMAX FM991, FIBERMAX FM819, FIBERMAX FM800, FIBERMAX FM960, FIBERMAX FM966, FIBERMAX FM981, FIBERMAX FM5035, FIBERMAX FM5044, FIBERMAX FM5045, FIBERMAX FM5013, FIBERMAX FM5015, FIBERMAX FM5017 or FIBERMAX FM5024 and plants with genotypes derived thereof.

[184] "Cotton" as used herein includes Gossypium hirsutum, Gossypium barbadense, Gossypium arboreum and Gossypium herbaceum. "Cotton progenitor plants" include

Gossypium arboreum, Gossypium herbaceum, Gossypium raimondii, Gossypium longicalyx and Gossypium kirkii.

[185] The methods and means of the current invention may also be employed for other plant species such as hemp, jute, flax and woody plants, including but not limited to Pinus spp., Populus spp., Picea spp., Eucalyptus spp. etc.

[186] The obtained transformed plant can be used in a conventional breeding scheme to produce more transformed plants with the same characteristics or to introduce the chimeric gene according to the invention in other varieties of the same or related plant species, or in hybrid plants. Seeds obtained from the transformed plants contain the chimeric genes of the invention as a stable genomic insert and are also encompassed by the invention.

[187] hi one embodiment, the amount of functional GLUC protein is significantly reduced in fibers of the fiber-producing plant during the fiber strength building phase of fiber development compared to the amount of functional GLUC protein produced during the fiber strength building phase in a plant in which the functional expression of the at least one GLUC allele is not abolished.

[188] A "significantly reduced amount of functional GLUC protein" (e.g. functional GLUC 1.1 A or GLUCl. ID protein) refers to a reduction in the amount of a functional GLUC protein produced by the cell comprising a mutant GLUC allele by at least 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95% or 100% (i.e. no functional GLUC protein is produced by the cell) as compared to the amount of the functional GLUC protein produced by the cell not comprising the mutant GLUC allele. This definition encompasses the production of a "non-functional" GLUC protein (e.g. truncated GLUC protein) having no biological activity in vivo, the reduction in the absolute amount of the functional GLUC protein (e.g. no functional GLUC protein being made due to the mutation in the GLUC gene), and/or the production of a GLUC protein with significantly reduced biological activity compared to the activity of a functional wild type GLUC

protein (such as a GLUC protein in which one or more amino acid residues that are crucial for the biological activity of the encoded GLUC protein, as exemplified above and below, are substituted for another amino acid residue). The term "mutant GLUC protein", as used herein, refers to a GLUC protein encoded by a mutant GLUC nucleic acid sequence ("glue allele") whereby the mutation results in a significantly reduced and/or no GLUC activity in vivo, compared to the activity of the GLUC protein encoded by a non- mutant, wild type GLUC sequence ("GLUC allele").

[189] In yet a further embodiment, the fibers of the non-naturally occurring fiber- producing plant have a higher callose content compared to the callose content of the fibers of an equivalent fiber-producing plant wherein the expression of the at least one GLUC allele is not abolished.

[190] In a particular aspect of this embodiment, the strength of the fibers of the non- naturally occurring fiber-producing plant is increased compared to the strength of the fibers of an equivalent fiber-producing plant wherein the expression of the at least one GLUC allele is not abolished.

[191] In one aspect of this embodiment, the non-naturally occuring Gossypium plant is a Gossypium hirsutum plant which is homozygous for the Gossypium barbadense GLUC 1.1 A allele. In a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 5% and about 10%, more specifically about 7,5%, higher than the fiber strength of a Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum GLUCl. IA allele. In still a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 1.6 g/tex and about 3.3 g/tex, more specifically about 2.5 g/tex higher than the fiber strength of a Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum GLUCl. IA allele. In yet a further aspect of this embodiment, the strength of the fibers of the Gossypium plant is on average between about 34.6 g/tex and about 36.3 g/tex, more specifically about 35.5 g/tex, as compared to a fiber strength of on average between about 32.2 g/tex and about 33.8 g/tex, more specifically about 33.0 g/tex of a

Gossypium hirsutum plant which is homozygous for the Gossypium hirsutum GLUC 1.1 A allele.

[192] Further provided herein are nucleic acid sequences of wild type and mutant GLUC 1.1 genes/alleles from Gossypium species, as well as the wild type and mutant GLUC 1.1 proteins. Also provided are methods of generating and combining mutant and wild type GLUC 1.1 alleles in Gossypium plants, as well as Gossypium plants and plant parts comprising specific combinations of wild type and mutant GLUC 1.1 alleles in their genome, whereby these plants produce fibers with altered fiber strength and whereby the plants preferably grow normally and have a normal phenotype. The use of these plants for transferring mutant GLUC 1.1 alleles to other plants is also an embodiment of the invention, as are the plant products of any of the plants described. In addition kits and methods for marker assisted selection (MAS) for combining or detecting GLUC genes and/or alleles are provided. Each of the embodiments of the invention is described in detail herein below.

[193] Provided are both wild type (GLUCl.1) nucleic acid sequences, encoding functional GLUCl.1 proteins, and mutant (glucl.l) nucleic acid sequences (comprising one or more mutations, preferably mutations which result in a significantly reduced biological activity of the encoded GLUC 1.1 protein or in no GLUC 1.1 protein being produced) of GLUC 1.1 genes from Gossypium species, especially from Gossypium hirsutum and Gossypium barbadense, but also from other Gossypium species. For example, Gossypium species comprising an A and/or a D genome may comprise different alleles of GLUC 1.1 A or GLUCl. ID genes which can be identified and combined in a single plant according to the invention. In addition, mutagenesis methods can be used to generate mutations in wild type GLUCl. IA or GLUCl. ID alleles, thereby generating mutant alleles for use according to the invention. Because specific GLUC 1.1 alleles are preferably combined in a Gossypium plant by crossing and selection, in one embodiment the GLUC 1.1 and/or glucl.l nucleic acid sequences are provided within a Gossypium plant (i.e. endogenously).

[194] However, isolated GLUCl.1 and glucl.J nucleic acid sequences (e.g. isolated from the plant by cloning or made synthetically by DNA synthesis), as well as variants thereof and fragments of any of these are also provided herein, as these can be used to determine which sequence is present endogenously in a plant or plant part, whether the sequence encodes a functional protein or a protein with significantly reduced or no functionality (e.g. by expression in a recombinant host cell and enzyme assays) and for selection and transfer of specific alleles from one Gossypium plant into another, in order to generate a plant having the desired combination of functional and mutant alleles.

[195] Nucleic acid sequences of GLUCl. IA and/or GLUC 1.1 D have been isolated from Gossypium hirsutum, from Gossypium barbadense, from Gossypium tomentosum, from Gossypium darwinii, from Gossypium mustelinum, from Gossypium arboreum, from Gossypium herbaceum, and from Gossypium raimondii as depicted in the sequence listing. The wild type GLUC 1.1 A sequences of Gossypium hirsutum, tomentosum, mustelinum and herbaceum and wild type GLUC 1.1 D sequences of Gossypium hirsutum, tomentosum, barbadense, darwinii, mustelinum and raimondii are depicted, while the mutant glucl.la and/or glue 1. Id sequences of these sequences, and of sequences essentially similar to these, are described herein below and in the Examples, with reference to the wild type GLUCl. IA and GLUCl. ID sequences. Further, the mutant GLUCl. IA sequences of Gossypium barbadense, darwinii and arboreum are depicted, while the alternative mutant glucl.la sequences of these sequences, and of sequences essentially similar to these, are described herein below and in the Examples. The genomic GLUC 1.1 A and D protein-encoding DNA, and corresponding pre-mRNA, comprises 2 exons (numbered exons 1 and 2 starting from the 5 'end) interrupted by 1 intron. In the cDNA and corresponding processed mRNA (i.e. the spliced RNA), introns are removed and exons are joined, as depicted in the sequence listing and Figures 1 and 6. Exon sequences are more conserved evolutionarily and are therefore less variable than intron sequences.

[196] "GLUC 1.1 A nucleic acid sequences" or "GLUC 1.1 A variant nucleic acid sequences" according to the invention are nucleic acid sequences encoding an amino acid

sequence having at least 95%, at least 96%, at least 97%, at least 98%, at least 99% or 100% sequence identity to SEQ ID NO: 4 or nucleic acid sequences encoding a cDNA sequence with at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 3 or comprises a coding sequence with at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to the nucleotide at position 2410 to the nucleotide at position 3499 of SEQ ID NO: 1. These nucleic acid sequences may also be referred to as being "essentially similar" or "essentially identical" or "corresponding to" the GLUC 1.1 A sequences provided in the sequence listing.

[197] "GLUC 1.1 D nucleic acid sequences" or "GLUC 1.1 D variant nucleic acid sequences" according to the invention are nucleic acid sequences encoding an amino acid sequence having at least 95%, at least 96%, at least 97%, at least 98%, at least 99% or 100% sequence identity to SEQ ID NO: 10 or nucleic acid sequences encoding a cDNA sequence with at least 97%, at least 98%, at least 99% sequence identity to SEQ ID NO: 3 or comprises a coding sequence with at least 95%, at least 96%, at least 97%, at least 98%, at least 99% sequence identity to the nucleotide at position 3337 to the nucleotide at position 4444 of SEQ ID NO: 7. These nucleic acid sequences may also be referred to as being "essentially similar" or "essentially identical" or "corresponding to" the GLUC 1.1 A sequences provided in the sequence listing.

[198] Thus, the invention provides both nucleic acid sequences encoding wild type, functional GLUC 1.1 A and GLUC 1.1 D proteins, including variants and fragments thereof (as defined further below), as well as mutant nucleic acid sequences of any of these, whereby the mutation in the nucleic acid sequence preferably results in one or more amino acids being inserted, deleted or substituted in comparison to the wild type protein. Preferably the mutation(s) in the nucleic acid sequence result in one or more amino acid changes (i.e. in relation to the wild type amino acid sequence one or more amino acids are inserted, deleted and/or substituted) whereby the biological activity of the GLUC 1.1 protein is significantly reduced. A significant reduction in biological activity of the mutant GLUC 1.1 protein, refers to a reduction in enzymatic activity by at least 30%, at

least 40%, 50% or more, at least 90% or 100% (no biological activity) compared to the activity of the wild type protein.

[199] Both endogenous and isolated nucleic acid sequences are provided herein. Also provided are fragments of the GLUC 1.1 sequences and GLUCl.1 variant nucleic acid sequences defined above, for use as primers or probes and as components of kits according to another aspect of the invention (see further below). A "fragment" of a GLUCl.1 or glucl.l nucleic acid sequence or variant thereof (as defined) may be of various lengths, such as at least 10, 12, 15, 18, 20, 50, 100, 200, 500, 1000 contiguous nucleotides of the GLUCl.1 or glucl.l sequence (or of the variant sequence).

[200] Nucleic acid sequences of GLUC 1.1 A and/or GLUC 1.1 D have been isolated from Gossypium hirsutum, from Gossypium barbadense, from Gossypium tomentosum, from Gossypium darwinii, from Gossypium mustelinum, from Gossypium arboreum, from Gossypium herbaceum, and from Gossypium raimondii as depicted in the sequence listing. The wild type GLUC 1.1 A sequences of Gossypium hirsutum, tomentosum, mustelinum and herbaceum and wild type GLUCl. ID sequences of Gossypium hirsutum, tomentosum, barbadense, darwinii, mustelinum and raimondii are depicted, while the mutant glucl.l a and/or glucl.ld sequences of these sequences, and of sequences essentially similar to these, are described herein below and in the Examples, with reference to the wild type GLUCl. IA and GLUCl. ID sequences. Further, the mutant GLUCl. IA sequences of Gossypium barbadense, darwinii and arboreum are depicted, while the alternative mutant glucl.l a sequences of these sequences, and of sequences essentially similar to these, are described herein below and in the Examples. The genomic GLUC 1.1 A and D protein-encoding DNA, and corresponding pre-mRNA, comprises 2 exons (numbered exons 1 and 2 starting from the 5 'end) interrupted by 1 intron. In the cDNA and corresponding processed mRNA (i.e. the spliced RNA), introns are removed and exons are joined, as depicted in the sequence listing and Figures 1 and 6. Exon sequences are more conserved evolutionarily and are therefore less variable than intron sequences.

[201] The nucleic acid sequences of GLUCl. IA and/or GLUCl. ID from Gossypium hirsutism, from Gossypium barbadense, from Gossypium tomentosum, from Gossypium darwinii, from Gossypium mustelinum, from Gossypium arboreum, from Gossypium herbaceum, and from Gossypium raimondii depicted in the sequence listing encode wild type, functional GLUC 1.1 proteins from these Gossypium species. Further, the mutant GLUCl. IA sequences of Gossypium barbadense, darwinii and arboreum depicted in the sequence listing encode wild type, non-functional GLUC 1.1 proteins from these Gossypium species. Thus, these sequences are endogenous to the Gossypium species from which they were isolated. Other Gossypium species, varieties, breeding lines or wild accessions may be screened for other GLUCl. IA and GLUC 1.1 D alleles, encoding the same GLUC 1.1 A and GLUC 1.1 D proteins or variants thereof. For example, nucleic acid hybridization techniques (e.g. Southern blot, using for example stringent hybridization conditions) or PCR-based techniques may be used to identify GLUCl.1 alleles endogenous to other Gossypium plants. To screen such plants or plant tissues for the presence of GLUC 1.1 alleles, the GLUC 1.1 nucleic acid sequences provided in the sequence listing, or variants or fragments of any of these, may be used. For example whole sequences or fragments may be used as probes or primers. For example specific or degenerate primers may be used to amplify nucleic acid sequences encoding GLUC 1.1 proteins from the genomic DNA of the plant or plant tissue. These GLUCl.1 nucleic acid sequences may be isolated and sequenced using standard molecular biology techniques. Bioinformatics analysis may then be used to characterize the allele(s), for example in order to determine which GLUC 1.1 allele the sequence corresponds to and which GLUC 1.1 protein or protein variant is encoded by the sequence.

[202] Whether a nucleic acid sequence encodes a functional GLUC 1.1 protein can be analyzed by recombinant DNA techniques as known in the art, e.g. expressing the nucleic acid molecule in a host cell (e.g. a bacterium, such as E. colϊ) and analyzing the endo-1,3- beta-glucanase activity of the resulting protein or cells.

[203] In addition, it is understood that GLUC 1.1 nucleic acid sequences and variants thereof (or fragments of any of these) may be identified in silico, by screening nucleic

acid databases for essentially similar sequences. Likewise, a nucleic acid sequence may be synthesized chemically. Fragments of nucleic acid molecules according to the invention are also provided, which are described further below. Fragments include nucleic acid sequences encoding only the mature protein, or smaller fragments comprising all or part of the exon and/or intron sequences, etc.

[204] Nucleic acid sequences comprising one or more nucleotide deletions, insertions or substitutions relative to the wild type nucleic acid sequences are another embodiment of the invention, as are fragments of such mutant nucleic acid molecules. Such mutant nucleic acid sequences (referred to as glue 1.1 sequences) can be generated and/or identified using various known methods, as described further below. Again, such nucleic acid molecules are provided both in endogenous form and in isolated form. In one embodiment, the mutation(s) result in one or more changes (deletions, insertions and/or substitutions) in the amino acid sequence of the encoded GLUC 1.1 protein (i.e. it is not a "silent mutation"). In another embodiment, the mutation(s) in the nucleic acid sequence result in a significantly reduced or completely abolished biological activity of the encoded GLUC 1.1 protein relative to the wild type protein.

[205] The nucleic acid molecules may, thus, comprise one or more mutations, such as:

(a) a "missense mutation", which is a change in the nucleic acid sequence that results in the substitution of an amino acid for another amino acid;

(b) a "nonsense mutation" or "STOP codon mutation", which is a change in the nucleic acid sequence that results in the introduction of a premature STOP codon and thus the termination of translation (resulting in a truncated protein); plant genes contain the translation stop codons "TGA" (UGA in RNA), "TAA" (UAA in RNA) and "TAG" (UAG in RNA); thus any nucleotide substitution, insertion, deletion which results in one of these codons to be in the mature mRNA being translated (in the reading frame) will terminate translation.

(c) an "insertion mutation" of one or more amino acids, due to one or more codons having been added in the coding sequence of the nucleic acid;

(d) a "deletion mutation" of one or more amino acids, due to one or more codons having been deleted in the coding sequence of the nucleic acid;

(e) a "frameshift mutation", resulting in the nucleic acid sequence being translated in a different frame downstream of the mutation. A frameshift mutation can have various causes, such as the insertion, deletion or duplication of one or more nucleotides, but also mutations which affect pre-mRNA splicing (splice site mutations) can result in frameshifts;

(f) a "splice site mutation", which alters or abolishes the correct splicing of the pre- mRNA sequence, resulting in a protein of different amino acid sequence than the wild type. For example, one or more exons may be skipped during RNA splicing, resulting in a protein lacking the amino acids encoded by the skipped exons. Alternatively, the reading frame may be altered through incorrect splicing, or one or more introns may be retained, or alternate splice donors or acceptors may be generated, or splicing may be initiated at an alternate position (e.g. within an intron), or alternate polyadenylation signals may be generated. Correct pre-mRNA splicing is a complex process, which can be affected by various mutations in the nucleotide sequence of the GLUC 1.1 -encoding gene. In higher eukaryotes, such as plants, the major spliceosome splices introns containing GU at the 5' splice site (donor site) and AG at the 3' splice site (acceptor site). This GU-AG rule (or GT-AG rule; see Lewin, Genes VI, Oxford University Press 1998, pp885-920, ISBN 0198577788) is followed in about 99% of splice sites of nuclear eukaryotic genes, while introns containing other dinucleotides at the 5' and 3' splice site, such as GC-AG and AU-AC account for only about 1% and 0.1% respectively.

[206] As already mentioned, it is desired that the mutation(s) in the nucleic acid sequence preferably result in a mutant protein comprising significantly reduced or no enzymatic activity in vivo. Basically, any mutation which results in a protein comprising at least one amino acid insertion, deletion and/or substitution relative to the wild type protein can lead to significantly reduced or no enzymatic activity. It is, however, understood that mutations in certain parts of the protein are more likely to result in a reduced function of the mutant GLUC 1.1 protein, such as mutations leading to truncated

proteins, whereby significant portions of the functional domains, such as the catalytic domain, are lacking.

[207] The functional GLUC 1.1 proteins of Gossypium described herein are about 325 - 337 amino acids in length and comprise a number of structural and functional domains. These include the following: An N-terminal plastid target peptide of about 14-26 amino acids followed by what constitutes the mature GLUC 1.1 protein. The mature GLUC 1.1 protein comprises active site and glycosylation amino acid residues as indicated in Table 4 above.

[208] Thus in one embodiment, nucleic acid sequences comprising one or more of any of the types of mutations described above are provided. In another embodiment, glue 1.1 sequences comprising one or more deletion mutations, one or more stop codon (nonsense) mutations and/or one or more splice site mutations are provided. Any of the above mutant nucleic acid sequences are provided per se (in isolated form), as are plants and plant parts comprising such sequences endogenously.

[209] A deletion mutation in a GLUC 1.1 allele, as used herein, is a mutation in a GLUCl.1 allele whereby at least 1, at least 2, 3, 4, 5, 10, 20, 30, 50, 100, 200, 500, 1000 or more bases are deleted from the corresponding wild type GLUC 1.1 allele, and whereby the deletion results in the mutant GLUC 1.1 allele being transcribed and translated into a mutant protein which has significantly reduced or no activity in vivo. A deletion may lead to a frame-shift and/or it may introduce a premature stop codon, or may lead to one amino acid or more amino acids (e.g. large parts) of coding sequence being removed, etc. The exact underlying molecular basis by which the deletion results in a mutant protein having significantly reduced biological activity is not important. Also provided herein are plants and plant parts in which specific GLUCl.1 alleles are completely deleted, i.e. plants and plant parts lacking one or more GLUCl.1 alleles.

[210] A nonsense mutation in a GLUC 1.1 allele, as used herein, is a mutation in a GLUC 1.1 allele whereby one or more translation stop codons are introduced into the

coding DNA and the corresponding mRNA sequence of the corresponding wild type GLUCl.1 allele. Translation stop codons are TGA (UGA in the mRNA), TAA (UAA) and TAG (UAG). Thus, any mutation (deletion, insertion or substitution) which leads to the generation of an in-frame stop codon in the coding sequence (exon sequence) will result in termination of translation and truncation of the amino acid chain, hi one embodiment, a mutant GLUCl.1 allele comprising a nonsense mutation is a GLUCl.1 allele wherein an in- frame stop codon is introduced in the GLUC 1.1 codon sequence by a single nucleotide substitution, such as the mutation of CAG to TAG, TGG to TAG, TGG to TGA, or CGA to TGA. In another embodiment, a mutant GLUCl.1 allele comprising a nonsense mutation is a GLUC 1.1 allele wherein an in- frame stop codon is introduced in the GLUCl.1 codon sequence by double nucleotide substitutions, such as the mutation of CAG to TAA, TGG to TAA, CGG to TAG or TGA, CGA to TAA. In yet another embodiment, a mutant GLUCl.1 allele comprising a nonsense mutation is a GLUCl.1 allele wherein an in-frame stop codon is introduced in the GLUC 1.1 codon sequence by triple nucleotide substitutions, such as the mutation of CGG to TAA. The truncated protein lacks the amino acids encoded by the coding DNA downstream of the mutation (i.e. the C-terminal part of the GLUCl.1 protein) and maintains the amino acids encoded by the coding DNA upstream of the mutation (i.e. the N-terminal part of the GLUC 1.1 protein). In one embodiment, the nonsense mutation is present anywhere in front of the second conserved GIu residue, the Tip residue, the first GIu residue, and/or the Tyr residue of the active site, so that at least the conserved GIu residue, the Trp residue, the first GIu residue, and/or the Tyr residue is lacking, resulting in significantly reduced activity of the truncated protein. The more truncated the mutant protein is in comparison to the wild type protein, the more likely it is that it will lack any enzymatic activity. Thus in another embodiment, a mutant GLUC 1.1 allele comprising a nonsense mutation which result in a truncated protein lacking the second conserved GIu, a truncated protein lacking the second conserved GIu residue and the Trp residue, a truncated protein lacking the second conserved GIu residue, the Trp residue and the first GIu residue, a truncated protein lacking the second conserved GIu residue, the Trp residue, the first GIu residue and the Tyr residue, or a truncated protein with even less amino acids in length are

provided. In yet another embodiment, the nonsense mutation results in one or more exons not being translated into protein, such as exon 1, exon 2 or exons 1 and 2.

[211] A splice site mutation in a GLUC 1.1 allele, as used herein, is a mutation in a GLUC 1.1 allele whereby a mutation in the corresponding wild type functional GLUC 1.1 allele results in aberrant splicing of the pre-mRNA thereby resulting in a mutant protein having significantly reduced or no activity. The mutation may be in the consensus splice site sequence. For example, Table 5 describes consensus sequences, which - if mutated - are likely to affect correct splicing. The GT-AG splice sites commonly have other conserved nucleotides, such as 2 highly conserved nucleotides on the 5 'end of the intron (in the exon), often being 5'-AG-3'. On the 3'-side of the GT dinucleotide (thus in the intron) high conservation can be found for a tetranucleotide 5'-AAGT-3'. This means that 8 nucleotides can be identified as highly conserved at the donor site.

Table 5: Consensus splice site sequences

^λ depicts the splice site; R = A or G; Y = C or T; N = A, C, G or T (but often G); n multiple nucleotides; in bold = consensus dinucleotides in the intron sequence. Pu purine base; Py = pyrimidine base.

[212] Splice site structure and consensus sequences are described in the art and computer programs for identifying exons and splice site sequences, such as NetPLAntgene, BDGP or Genio, est2genome, FgeneSH, and the like, are available. Comparison of the genomic sequence or pre-mRNA sequence with the translated protein can be used to determine or verify splice sites and aberrant splicing.

[213] Any mutation (insertion, deletion and/or substitution of one or more nucleotides) which alters pre-mRNA splicing and thereby leads to a protein with significantly reduced biological activity is encompassed herein. In one embodiment, a mutant GLUCl.1 allele comprising a splice site mutation is a GLUC 1.1 allele wherein altered splicing is caused by the introduction in the GLUC 1.1 transcribed DNA region of one or more nucleotide substitution(s) of the consensus dinucleotides depicted in bold above. For example, ^A GU may for example be mutated to ^A AU in the donor splice site and/or AG ^A may be mutated to AA ^λ in the acceptor splice site sequence. In another embodiment, a mutant GLUC 1.1 allele comprising a splice site mutation is a GLUCl.1 allele wherein altered splicing is caused by the introduction in the GLUCl.1 transcribed DNA region of one or more nucleotide substitution(s)in the conserved nucleotides in the exon sequences.

[214] Further provided are both functional GLUC 1.1 amino acid sequences and nonfunctional GLUC 1.1 amino acid sequences (comprising one or more mutations, preferably mutations which result in a significantly reduced or no biological activity of the GLUC 1.1 protein) from Gossypium species, especially from Gossypium hirsutum and Gossypium barbadense, but also from other Gossypium species, such as those indicated below, hi addition, mutagenesis methods can be used to generate mutations in wild type functional GLUC 1.1 alleles, thereby generating mutant non-functional GLUC 1.1 alleles which can encode further non- functional GLUC 1.1 proteins, hi one embodiment the functional and/or non-functional GLUC 1.1 amino acid sequences are provided within a Gossypium plant (i.e. endogenously). However, isolated GLUC 1.1 amino acid sequences (e.g. isolated from the plant or made synthetically), as well as variants thereof and fragments of any of these are also provided herein.

[215] Amino acid sequences of GLUCl. IA and GLUC 1.1 D proteins have been determined from Gossypium hirsutum, from Gossypium barbadense, from Gossypium tomentosum, from Gossypium darwinii, from Gossypium mustilinum, from Gossypium arboreum, from Gossypium herbaceum, and from Gossypium raimondii as depicted in the sequence listing and Figures 2 and 7. The wild type functional GLUC 1.1 A sequences of Gossypium hirsutum, tomentosum, mustilinum and herbaceum and wild type functional GLUC 1.1 D sequences of Gossypium hirsutum, tomentosum, barbadense, darwinii, mustilinum and raimondii are depicted, while mutant non-functional GLUC 1.1 A sequences of these, and of sequences essentially similar to these, are described herein below, with reference to the wild type functional GLUC 1.1 A and GLUC 1.1 D sequences. Further, the wild type non- functional GLUC 1.1 A sequences of Gossypium barbadense, darwinii and arboreum are depicted, while alternative (mutant) non-functional GLUC 1.1 A sequences of these sequences, and of sequences essentially similar to these, are described herein below and in the Examples.

[216] As described above, the functional GLUC 1.1 proteins of Gossypium described herein are about 325-337 amino acids in length and comprise a number of structural and functional domains. The sequences of the N-terminal part of the GLUC 1.1 proteins are less conserved evolutionarily than the sequences of the mature GLUC 1.1 proteins. The sequences of the mature GLUC 1.1 proteins are therefore less variable than the sequences of the precursor proteins.

[217] "GLUC 1.1 A amino acid sequences" or "GLUC 1.1 A variant amino acid sequences" according to the invention are amino acid sequences having at least 95%, at least 96%, at least 97%, at least 98%, at least 99% or at least 100% sequence identity to SEQ ID NO: 4. These amino acid sequences may also be referred to as being "essentially similar" or "essentially identical" or "corresponding to" the GLUC 1.1 A sequences provided in the sequence listing.

[218] "GLUC 1.1 D amino acid sequences" or "GLUC 1.1 D variant amino acid sequences" according to the invention are amino acid sequences having at least 95%, at

least 96%, at least 97%, at least 98%, at least 99% or at least 100% sequence identity to SEQ ID NO: 10. These amino acid sequences may also be referred to as being "essentially similar" or "essentially identical" or "corresponding to" the GLUC 1.1 D sequences provided in the sequence listing.

[219] Thus, the invention provides both amino acid sequences of wild type functional and non- functional GLUC 1.1 A and GLUC 1.1 D proteins, including variants and fragments thereof (as defined further below), as well as mutant non-functional amino acid sequences of any of these, whereby the mutation in the amino acid sequence preferably results in a significant reduction in the biological activity of the GLUC 1.1 protein. A significant reduction in biological activity of the (wild type or mutant) non-functional GLUC 1.1 protein, refers to a reduction in enzymatic activity (i.e. in endo-l,3-beta- glucanase activity) by at least 30%, at least 40%, 50% or more, at least 90% or 100% (no biological activity) compared to the activity of the functional protein.

[220] Both endogenous and isolated amino acid sequences are provided herein. A "fragment" of a GLUC 1.1 amino acid sequence or variant thereof (as defined) may be of various lengths, such as at least 10, 12, 15, 18, 20, 50, 100, 200, 400 contiguous amino acids of the GLUC 1.1 sequence (or of the variant sequence).

[221] The amino acid sequences depicted in the sequence listing are wild type GLUC 1.1 proteins from Gossypium species. Thus, these sequences are endogenous to the Gossypium plants from which they were isolated. Other Gossypium species, varieties, breeding lines or wild accessions may be screened for other (functional or nonfunctional) GLUC 1.1 proteins with the same amino acid sequences or variants thereof, as described above.

[222] In addition, it is understood that GLUC 1.1 amino acid sequences and variants thereof (or fragments of any of these) may be identified in silico, by screening amino acid databases for essentially similar sequences. Fragments of amino acid molecules according to the invention are also provided. Fragments include amino acid sequences of

the mature protein, or smaller fragments comprising all or part of the amino acid sequences, etc.

[223] Amino acid sequences comprising one or more amino acid deletions, insertions or substitutions relative to the wild type (functional or non-functional) amino acid sequences are another embodiment of the invention, as are fragments of such mutant amino acid molecules. Such mutant amino acid sequences can be generated and/or identified using various known methods, as described above. Again, such amino acid molecules are provided both in endogenous form and in isolated form.

[224] In one embodiment, the mutation(s) in the amino acid sequence result in a significantly reduced or completely abolished biological activity of the GLUC 1.1 protein relative to the wild type protein. As described above, basically, any mutation which results in a protein comprising at least one amino acid insertion, deletion and/or substitution relative to the wild type protein can lead to significantly reduced (or no) enzymatic activity. It is, however, understood that mutations in certain parts of the protein are more likely to result in a reduced function of the mutant GLUC 1.1 protein, such as mutations leading to truncated proteins, whereby significant portions of the functional domains, such as the active site or glycosylation site (see above), are lacking or mutations whereby conserved amino acid residues which have a catalytic function or which are involved in substrate specificity are substituted.

[225] Thus in one embodiment, mutant GLUC 1.1 proteins are provided comprising one or more deletion or insertion mutations, whereby the deletion(s) or insertion(s) result(s) in a mutant protein which has significantly reduced or no activity in vivo. Such mutant GLUC 1.1 proteins are GLUC 1.1 proteins wherein at least 1, at least 2, 3, 4, 5, 10, 20, 30, 50, 100, 200, 300, 400 or more amino acids are deleted or inserted as compared to the wild type GLUC 1.1 protein, whereby the deletion(s) or insertion(s) result(s) in a mutant protein which has significantly reduced or no activity in vivo.

[226] In another embodiment, mutant GLUC 1.1 proteins are provided which are truncated whereby the truncation results in a mutant protein which has significantly reduced or no activity in vivo. Such truncated GLUC 1.1 proteins are GLUCl.1 proteins which lack functional domains, such as active site residues and/or glycosylation site residues, in the C-terminal part of the corresponding wild type (mature) GLUC 1.1 protein and which maintain the N-terminal part of the corresponding wild type (mature) GLUC 1.1 protein. Thus in one embodiment, a truncated GLUC 1.1 protein comprising the N-terminal part of the corresponding wild type (mature) GLUC 1.1 protein up to but not including the conserved second GIu residue (as described above) is provided. The more truncated the mutant protein is in comparison to the wild type protein, the more likely it is that it will lack any enzymatic activity. Thus in another embodiment, a truncated GLUC 1.1 protein comprising the N-terminal part of the corresponding wild type (mature) GLUC 1.1 protein up to but not including the conserved Tip and/or the first GIu residue (as described above) is provided. In yet another embodiment, a truncated GLUC 1.1 protein comprising the N-terminal part of the corresponding wild type (mature) GLUC 1.1 protein up to but not including the conserved Tyr residue (as described above), or lacking even more amino acids, is provided.

[227] In yet another embodiment, mutant GLUC 1.1 proteins are provided comprising one or more substitution mutations, whereby the substitution(s) result(s) in a mutant protein which has significantly reduced or no activity in vivo. Such mutant GLUC 1.1 proteins are GLUC 1.1 proteins whereby conserved amino acid residues which have a catalytic function or which are involved in substrate binding or specificity (for example, those described above) are substituted. Thus in one embodiment, a mutant GLUC 1.1 protein comprising a substitution of a conserved amino acid residue which has a catalytic function, such as the conserved first or second GIu, Trp, and/or Tyr residues, is provided, hi another embodiment, a mutant GLUC 1.1 protein comprising a substitution of a conserved amino acid residue involved in glycosylation, such as the conserved Asn residue, is provided.

[228] In another aspect of the invention, methods are provided for generating mutant glucl.l alleles (for example induced by mutagenesis) and/or identifying mutant glucl.l alleles using a range of methods, which are conventional in the art, for example using PCR based methods to amplify part or all of the glucl.l genomic or cDNA.

[229] The term "mutagenesis", as used herein, refers to the process in which plant cells (e.g., a plurality of Gossypium seeds or other parts, such as pollen) are subjected to a technique which induces mutations in the DNA of the cells, such as contact with a mutagenic agent, such as a chemical substance (such as ethylmethylsulfonate (EMS), ethylnitrosourea (ENU), etc.) or ionizing radiation (neutrons (such as in fast neutron mutagenesis, etc.), alpha rays, gamma rays (such as that supplied by a Cobalt 60 source), X-rays, UV-radiation, etc.), or a combination of two or more of these. Thus, the desired mutagenesis of one or more GLUCl.1 alleles may be accomplished by use of chemical means such as by contact of one or more plant tissues with ethylmethylsulfonate (EMS), ethylnitrosourea, etc., by the use of physical means such as x-ray, etc, or by gamma radiation, such as that supplied by a Cobalt 60 source.

[230] Following mutagenesis, Gossypium plants are grown from the treated seeds, or regenerated from the treated cells using known techniques. For instance, the resulting Gossypium seeds may be planted in accordance with conventional growing procedures and following self-pollination seed is formed on the plants. Additional seed which is formed as a result of such self-pollination in the present or a subsequent generation may be harvested and screened for the presence of mutant GLUC 1.1 alleles, using techniques which are conventional in the art, for example polymerase chain reaction (PCR) based techniques (amplification of the glucl.l alleles) or hybridization based techniques, e.g. Southern blot analysis, and/or direct sequencing of glucl.l alleles. To screen for the presence of point mutations (so called Single Nucleotide Polymorphisms or SNPs) in mutant GLUC 1.1 alleles, SNP detection methods conventional in the art can be used, for example oligoligation-based techniques, single base extension-based techniques or techniques based on differences in restriction sites, such as TILLING.

[231] As described above, mutagenization (spontaneous as well as induced) of a specific wild-type (functional or non- functional) GLUC 1.1 allele results in the presence of one or more deleted, inserted, or substituted nucleotides (hereinafter called "mutation region") in the resulting mutant GLUC 1.1 allele. The mutant GLUC 1.1 allele can thus be characterized by the location and the configuration of the one or more deleted, inserted, or substituted nucleotides in the wild type GLUC 1.1 allele. The site in the wild type GLUC 1.1 allele where the one or more nucleotides have been inserted, deleted, or substituted, respectively, is also referred to as the "mutation region". A "5' or 3' flanking region or sequence" as used herein refers to a DNA region or sequence in the mutant (or the corresponding wild type) GLUC 1.1 allele of at least 20 bp, preferably at least 50 bp, at least 750 bp, at least 1500 bp, and up to 5000 bp of DNA different from the DNA containing the one or more deleted, inserted, or substituted nucleotides, preferably DNA from the mutant (or the corresponding wild type) GLUC 1.1 allele which is located either immediately upstream of and contiguous with (5' flanking region or sequence") or immediately downstream of and contiguous with (3' flanking region or sequence") the mutation region in the mutant GLUCl.1 allele (or in the corresponding wild type GLUC 1.1 allele).

[232] The tools developed to identify a specific mutant GLUC 1.1 allele or the plant or plant material comprising a specific mutant GLUC 1.1 allele, or products which comprise plant material comprising a specific mutant GLUC 1.1 allele are based on the specific genomic characteristics of the specific mutant GLUC 1.1 allele as compared to the genomic characteristics of the corresponding wild type GLUC 1.1 allele, such as, a specific restriction map of the genomic region comprising the mutation region, molecular markers or the sequence of the flanking and/or mutation regions.

[233] Once a specific mutant GLUC 1.1 allele has been sequenced, primers and probes can be developed which specifically recognize a sequence within the 5' flanking, 3' flanking and/or mutation regions of the mutant GLUC 1.1 allele in the nucleic acid (DNA or RNA) of a sample by way of a molecular biological technique. For instance a PCR method can be developed to identify the mutant GLUC 1.1 allele in biological samples

(such as samples of plants, plant material or products comprising plant material). Such a PCR is based on at least two specific "primers": one recognizing a sequence within the 5' or 3' flanking region of the mutant GLUC 1.1 allele and the other recognizing a sequence within the 3' or 5' flanking region of the mutant GLUC 1.1 allele, respectively; or one recognizing a sequence within the 5' or 3' flanking region of the mutant GLUC 1.1 allele and the other recognizing a sequence within the mutation region of the mutant GLUC 1.1 allele; or one recognizing a sequence within the 5' or 3' flanking region of the mutant GLUC 1.1 allele and the other recognizing a sequence spanning the joining region between the 3' or 5' flanking region and the mutation region of the specific mutant GLUC 1.1 allele (as described further below), respectively.

[234] The primers preferably have a sequence of between 15 and 35 nucleotides which under optimized PCR conditions "specifically recognize" a sequence within the 5' or 3' flanking region, a sequence within the mutation region, or a sequence spanning the joining region between the 3' or 5' flanking and mutation regions of the specific mutant GLUC 1.1 allele, so that a specific fragment ("mutant GLUC 1.1 specific fragment" or discriminating amplicon) is amplified from a nucleic acid sample comprising the specific mutant GLUC 1.1 allele. This means that only the targeted mutant GLUC 1.1 allele, and no other sequence in the plant genome, is amplified under optimized PCR conditions.

[235] PCR primers suitable for the invention may be the following: oligonucleotides ranging in length from 17 nt to about 200 nt, comprising a nucleotide sequence of at least 17 consecutive nucleotides, preferably 20 consecutive nucleotides selected from the 5' flanking sequence of a specific mutant GLUC 1.1 allele (i.e., for example, the sequence 5' flanking the one or more nucleotides deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention, such as the sequence 5' flanking the deletion, non-sense or splice site mutations described above or the sequence 5' flanking the potential STOP codon or splice site mutations indicated above) at their 3' end (primers recognizing 5' flanking sequences); or oligonucleotides ranging in length from 17 nt to about 200 nt, comprising a nucleotide sequence of at least 17 consecutive nucleotides, preferably 20 consecutive

nucleotides, selected from the 3' flanking sequence of a specific mutant GLUCl.1 allele (i.e., for example, the complement of the sequence 3' flanking the one or more nucleotides deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention, such as the complement of the sequence 3' flanking the deletion, non-sense or splice site mutations described above or the complement of the sequence 3' flanking the potential STOP codon or splice site mutations indicated above) at their 3' end (primers recognizing 3' flanking sequences); or oligonucleotides ranging in length from 17 nt to about 200 nt, comprising a nucleotide sequence of at least 17 consecutive nucleotides, preferably 20 nucleotides selected from the sequence of the mutation region of a specific mutant GLUC 1.1 allele (i.e., for example, the sequence of nucleotides inserted or substituted in the GLUC 1.1 genes of the invention, or the complement thereof) at their 3' end (primers recognizing mutation sequences).

[236] The primers may of course be longer than the mentioned 17 consecutive nucleotides, and may e.g. be 20, 21, 30, 35, 50, 75, 100, 150, 200 nt long or even longer. The primers may entirely consist of nucleotide sequence selected from the mentioned nucleotide sequences of flanking and mutation sequences. However, the nucleotide sequence of the primers at their 5' end (i.e. outside of the 3 '-located 17 consecutive nucleotides) is less critical. Thus, the 5' sequence of the primers may consist of a nucleotide sequence selected from the flanking or mutation sequences, as appropriate, but may contain several (e.g. 1, 2, 5, 10) mismatches. The 5' sequence of the primers may even entirely consist of a nucleotide sequence unrelated to the flanking or mutation sequences, such as e.g. a nucleotide sequence representing restriction enzyme recognition sites. Such unrelated sequences or flanking DNA sequences with mismatches should preferably be not longer than 100, more preferably not longer than 50 or even 25 nucleotides.

[237] Moreover, suitable primers may comprise or consist of a nucleotide sequence at their 3' end spanning the joining region between flanking and mutation sequences (i.e., for example, the joining region between a sequence 5' flanking one or more nucleotides

deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention and the sequence of the one or more nucleotides inserted or substituted or the sequence 3' flanking the one or more nucleotides deleted, such as the joining region between a sequence 5' flanking deletion, non-sense or splice site mutations in the GLUC 1.1 genes of the invention described above and the sequence of the non-sense or splice site mutations or the sequence 3' flanking the deletion mutation, or the joining region between a sequence 5' flanking a potential STOP codon or splice site mutation as indicated above and the sequence of the potential STOP codon or splice site mutation), provided the mentioned 3 '-located nucleotides are not derived exclusively from either the mutation region or flanking regions.

[238] It will also be immediately clear to the skilled artisan that properly selected PCR primer pairs should also not comprise sequences complementary to each other.

[239] For the purpose of the invention, the "complement of a nucleotide sequence represented in SEQ ID NO: X" is the nucleotide sequence which can be derived from the represented nucleotide sequence by replacing the nucleotides through their complementary nucleotide according to ChargafPs rules (AOT; GOC) and reading the sequence in the 5' to 3' direction, i.e in opposite direction of the represented nucleotide sequence.

[240] Examples of primers suitable to identify specific mutant GLUC 1.1 alleles are described in the Examples.

[241] As used herein, "the nucleotide sequence of SEQ ID No. Z from position X to position Y" indicates the nucleotide sequence including both nucleotide endpoints.

[242] Preferably, the amplified fragment has a length of between 50 and 1000 nucleotides, such as a length between 50 and 500 nucleotides, or a length between 100 and 350 nucleotides. The specific primers may have a sequence which is between 80 and 100% identical to a sequence within the 5' or 3' flanking region, a sequence within the

mutation region, or a sequence spanning the joining region between the 3' or 5' flanking and mutation regions of the specific mutant GLUC 1.1 allele, provided the mismatches still allow specific identification of the specific mutant GLUC 1.1 allele with these primers under optimized PCR conditions. The range of allowable mismatches however, can easily be determined experimentally and are known to a person skilled in the art.

[243] Detection and/or identification of a "mutant GLUC 1.1 specific fragment" can occur in various ways, e.g., via size estimation after gel or capillary electrophoresis or via fluorescence-based detection methods. The mutant GLUC 1.1 specific fragments may also be directly sequenced. Other sequence specific methods for detection of amplified DNA fragments are also known in the art.

[244] Standard PCR protocols are described in the art, such as in 'PCR Applications Manual" (Roche Molecular Biochemicals, 2nd Edition, 1999) and other references. The optimal conditions for the PCR, including the sequence of the specific primers, is specified in a "PCR identification protocol" for each specific mutant GLUC 1.1 allele. It is however understood that a number of parameters in the PCR identification protocol may need to be adjusted to specific laboratory conditions, and may be modified slightly to obtain similar results. For instance, use of a different method for preparation of DNA may require adjustment of, for instance, the amount of primers, polymerase, MgCl ₂ concentration or annealing conditions used. Similarly, the selection of other primers may dictate other optimal conditions for the PCR identification protocol. These adjustments will however be apparent to a person skilled in the art, and are furthermore detailed in current PCR application manuals such as the one cited above.

[245] Examples of PCR identification protocols to identify specific mutant GLUC 1.1 alleles are described in the Examples.

[246] Alternatively, specific primers can be used to amplify a mutant GLUC 1.1 specific fragment that can be used as a "specific probe" for identifying a specific mutant GLUC 1.1 allele in biological samples. Contacting nucleic acid of a biological sample,

with the probe, under conditions which allow hybridization of the probe with its corresponding fragment in the nucleic acid, results in the formation of a nucleic acid/probe hybrid. The formation of this hybrid can be detected (e.g. labeling of the nucleic acid or probe), whereby the formation of this hybrid indicates the presence of the specific mutant GLUC 1.1 allele. Such identification methods based on hybridization with a specific probe (either on a solid phase carrier or in solution) have been described in the art. The specific probe is preferably a sequence which, under optimized conditions, hybridizes specifically to a region within the 5' or 3' flanking region and/or within the mutation region of the specific mutant GLUC 1.1 allele (hereinafter referred to as "GLUC 1.1 mutation specific region"). Preferably, the specific probe comprises a sequence of between 20 and 1000 bp, 50 and 600 bp, between 100 to 500 bp, between 150 to 350bp, which is at least 80%, preferably between 80 and 85%, more preferably between 85 and 90%, especially preferably between 90 and 95%, most preferably between 95% and 100% identical (or complementary) to the nucleotide sequence of a specific region. Preferably, the specific probe will comprise a sequence of about 15 to about 100 contiguous nucleotides identical (or complementary) to a specific region of the specific mutant GLUC 1.1 allele.

[247] Specific probes suitable for the invention may be the following: oligonucleotides ranging in length from 20 nt to about 1000 nt, comprising a nucleotide sequence of at least 20 consecutive nucleotides selected from the 5' flanking sequence of a specific mutant GLUC 1.1 allele (i.e., for example, the sequence 5' flanking the one or more nucleotides deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention, such as the sequence 5' flanking the deletion, non-sense or splice site mutations described above or the sequence 5' flanking the potential STOP codon or splice site mutations indicated above), or a sequence having at least 80% sequence identity therewith (probes recognizing 5' flanking sequences); or oligonucleotides ranging in length from 20 nt to about 1000 nt, comprising a nucleotide sequence of at least 20 consecutive nucleotides selected from the 3' flanking sequence of a specific mutant GLUC 1.1 allele (i.e., for example, the

sequence 3' flanking the one or more nucleotides deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention, such as the sequence 3' flanking the deletion, non-sense or splice site mutations described above or the sequence 3' flanking the potential STOP codon or splice site mutations indicated above), or a sequence having at least 80% sequence identity therewith (probes recognizing 3' flanking sequences); or oligonucleotides ranging in length from 20 nt to about 1000 nt, comprising a nucleotide sequence of at least 20 consecutive nucleotides selected from the mutation sequence of a specific mutant GLUC 1.1 allele (i.e., for example, the sequence of nucleotides inserted or substituted in the GLUC 1.1 genes of the invention, or the complement thereof), or a sequence having at least 80% sequence identity therewith (probes recognizing mutation sequences).

[248] The probes may entirely consist of nucleotide sequence selected from the mentioned nucleotide sequences of flanking and mutation sequences. However, the nucleotide sequence of the probes at their 5' or 3' ends is less critical. Thus, the 5' or 3' sequences of the probes may consist of a nucleotide sequence selected from the flanking or mutation sequences, as appropriate, but may consist of a nucleotide sequence unrelated to the flanking or mutation sequences. Such unrelated sequences should preferably be not longer than 50, more preferably not longer than 25 or even not longer than20 or 15 nucleotides.

[249] Moreover, suitable probes may comprise or consist of a nucleotide sequence spanning the joining region between flanking and mutation sequences (i.e., for example, the joining region between a sequence 5' flanking one or more nucleotides deleted, inserted or substituted in the mutant GLUC 1.1 alleles of the invention and the sequence of the one or more nucleotides inserted or substituted or the sequence 3 ' flanking the one or more nucleotides deleted, such as the joining region between a sequence 5' flanking deletion, non-sense or splice site mutations in the GLUC 1.1 genes of the invention described above and the sequence of the non-sense or splice site mutations or the sequence 3' flanking the deletion mutation, or the joining region between a sequence 5'

flanking a potential STOP codon or splice site mutation indicated above and the sequence of the potential STOP codon or splice site mutation), provided the mentioned nucleotide sequence is not derived exclusively from either the mutation region or flanking regions.

[250] Examples of specific probes suitable to identify specific mutant GLUC 1.1 alleles are described in the Examples.

[251] Detection and/or identification of a "mutant GLUC 1.1 specific region" hybridizing to a specific probe can occur in various ways, e.g., via size estimation after gel electrophoresis or via fluorescence-based detection methods. Other sequence specific methods for detection of a "mutant GLUC 1.1 specific region" hybridizing to a specific probe are also known in the art.

[252] Alternatively, plants or plant parts comprising one or more mutant glucl.l alleles can be generated and identified using other methods, such as the "Delete-a-gene™" method which uses PCR to screen for deletion mutants generated by fast neutron mutagenesis (reviewed by Li and Zhang, 2002, Funct Integr Genomics 2:254-258), by the TILLING (Targeting Induced Local Lesions IN Genomes) method which identifies EMS- induced point mutations using denaturing high-performance liquid chromatography (DHPLC) to detect base pair changes by heteroduplex analysis (McCallum et al., 2000, Nat Biotech 18:455, and McCallum et al. 2000, Plant Physiol. 123, 439-442), etc. As mentioned, TILLING uses high-throughput screening for mutations (e.g. using CeI 1 cleavage of mutant- wildtype DNA heteroduplexes and detection using a sequencing gel system). Thus, the use of TILLING to identify plants, seeds and tissues comprising one or more mutant glucl.l alleles in one or more tissues and methods for generating and identifying such plants is encompassed herein. Thus in one embodiment, the method according to the invention comprises the steps of mutagenizing plant seeds (e.g. EMS mutagenesis), pooling of plant individuals or DNA, PCR amplification of a region of interest, heteroduplex formation and high-throughput detection, identification of the mutant plant, sequencing of the mutant PCR product. It is understood that other mutagenesis and selection methods may equally be used to generate such mutant plants.

[253] Instead of inducing mutations in GLUCl.1 alleles, natural (spontaneous) mutant alleles may be identified by methods known in the art. For example, ECOTILLING may be used (Henikoff et al. 2004, Plant Physiology 135(2):630-6) to screen a plurality of plants or plant parts for the presence of natural mutant glue J.I alleles. As for the mutagenesis techniques above, preferably Gossypium species are screened which comprise an A and/or a D genome, so that the identified glue 1.1 allele can subsequently be introduced into other Gossypium species, such as Gossypium hirsutum, by crossing (inter- or intraspecific crosses) and selection. In ECOTILLING natural polymorphisms in breeding lines or related species are screened for by the TILLING methodology described above, in which individual or pools of plants are used for PCR amplification of the glucl.l target, heteroduplex formation and high-throughput analysis. This can be followed up by selecting individual plants having a required mutation that can be used subsequently in a breeding program to incorporate the desired mutant allele.

[254] The identified mutant alleles can then be sequenced and the sequence can be compared to the wild type allele to identify the mutation(s). Optionally functionality can be tested by expression in a homologous or heterologous host and testing the mutant GLUC 1.1 protein for functionality in an enzyme assay. Using this approach a plurality of mutant glucl.l alleles (and Gossypium plants comprising one or more of these) can be identified. The desired mutant alleles can then be combined with the desired wild type alleles by crossing and selection methods as described further below. Finally a single plant comprising the desired number of mutant glucl.l and the desired number of wild type GLUCl.1 alleles is generated.

[255] Oligonucleotides suitable as PCR primers or specific probes for detection of a specific mutant GLUC 1.1 allele can also be used to develop methods to determine the zygosity status of the specific mutant GLUC 1.1 allele.

[256] To determine the zygosity status of a specific mutant GLUC 1.1 allele, a PCR- based assay can be developed to determine the presence of a mutant and/or corresponding wild type GLUC 1.1 specific allele:

[257] To determine the zygosity status of a specific mutant GLUC 1.1 allele, two primers specifically recognizing the wild-type GLUC 1.1 allele can be designed in such a way that they are directed towards each other and have the mutation region located in between the primers. These primers may be primers specifically recognizing the 5' and 3' flanking sequences, respectively. This set of primers allows simultaneous diagnostic PCR amplification of the mutant, as well as of the corresponding wild type GLUC 1.1 allele.

[258] Alternatively, to determine the zygosity status of a specific mutant GLUC 1.1 allele, two primers specifically recognizing the wild-type GLUC 1.1 allele can be designed in such a way that they are directed towards each other and that one of them specifically recognizes the mutation region. These primers may be primers specifically recognizing the sequence of the 5' or 3' flanking region and the mutation region of the wild type GLUC 1.1 allele, respectively. This set of primers, together with a third primer which specifically recognizes the sequence of the mutation region in the mutant GLUCl.1 allele, allow simultaneous diagnostic PCR amplification of the mutant GLUCl.1 gene, as well as of the wild type GLUC 1.1 gene.

[259] Alternatively, to determine the zygosity status of a specific mutant GLUC 1.1 allele, two primers specifically recognizing the wild-type GLUC 1.1 allele can be designed in such a way that they are directed towards each other and that one of them specifically recognizes the joining region between the 5' or 3' flanking region and the mutation region. These primers may be primers specifically recognizing the 5' or 3' flanking sequence and the joining region between the mutation region and the 3' or 5' flanking region of the wild type GLUC 1.1 allele, respectively. This set of primers, together with a third primer which specifically recognizes the joining region between the mutation region and the 3' or 5' flanking region of the mutant GLUC 1.1 allele,

respectively, allow simultaneous diagnostic PCR amplification of the mutant GLUC 1.1 gene, as well as of the wild type GLUC 1.1 gene.

[260] Alternatively, the zygosity status of a specific mutant GLUC 1.1 allele can be determined by using alternative primer sets which specifically recognize mutant and wild type GLUC 1.1 alleles.

[261] If the plant is homozygous for the mutant GLUC 1.1 gene or the corresponding wild type GLUC 1.1 gene, the diagnostic PCR assays described above will give rise to a single PCR product typical, preferably typical in length, for either the mutant or wild type GLUC 1.1 allele. If the plant is hemizygous for the mutant GLUC 1.1 allele, two specific PCR products will appear, reflecting both the amplification of the mutant and the wild type GLUC 1.1 allele.

[262] Identification of the wild type and mutant GLUC 1.1 specific PCR products can occur e.g. by size estimation after gel or capillary electrophoresis (e.g. for mutant GLUCl.1 alleles comprising a number of inserted or deleted nucleotides which results in a size difference between the fragments amplified from the wild type and the mutant GLUC 1.1 allele, such that said fragments can be visibly separated on a gel); by evaluating the presence or absence of the two different fragments after gel or capillary electrophoresis, whereby the diagnostic PCR amplification of the mutant GLUC 1.1 allele can, optionally, be performed separately from the diagnostic PCR amplification of the wild type GLUC 1.1 allele; by direct sequencing of the amplified fragments; or by fluorescence-based detection methods.

[263] Examples of primers suitable to determine the zygosity of specific mutant GLUC 1.1 alleles are described in the Examples.

[264] Alternatively, to determine the zygosity status of a specific mutant GLUC 1.1 allele, a hybridization-based assay can be developed to determine the presence of a mutant and/or corresponding wild type GLUC 1.1 specific allele:

[265] To determine the zygosity status of a specific mutant GLUC 1.1 allele, two specific probes recognizing the wild-type GLUC 1.1 allele can be designed in such a way that each probe specifically recognizes a sequence within the GLUC 1.1 wild type allele and that the mutation region is located in between the sequences recognized by the probes. These probes may be probes specifically recognizing the 5' and 3' flanking sequences, respectively. The use of one or, preferably, both of these probes allows simultaneous diagnostic hybridization of the mutant, as well as of the corresponding wild type GLUC 1.1 allele.

[266] Alternatively, to determine the zygosity status of a specific mutant GLUC 1.1 allele, two specific probes recognizing the wild-type GLUC 1.1 allele can be designed in such a way that one of them specifically recognizes a sequence within the GLUC 1.1 wild type allele upstream or downstream of the mutation region, preferably upstream of the mutation region, and that one of them specifically recognizes the mutation region. These probes may be probes specifically recognizing the sequence of the 5' or 3' flanking region, preferably the 5' flanking region, and the mutation region of the wild type GLUC 1.1 allele, respectively. The use of one or, preferably, both of these probes, optionally, together with a third probe which specifically recognizes the sequence of the mutation region in the mutant GLUC 1.1 allele, allow diagnostic hybridization of the mutant and of the wild type GLUC 1.1 gene.

[267] Alternatively, to determine the zygosity status of a specific mutant GLUC 1.1 allele, a specific probe recognizing the wild-type GLUC 1.1 allele can be designed in such a way that the probe specifically recognizes the joining region between the 5' or 3' flanking region, preferably the 5' flanking region, and the mutation region of the wild type GLUC 1.1 allele. This probe, optionally, together with a second probe which specifically recognizes the joining region between the 5' or 3' flanking region, preferably the 5' flanking region, and the mutation region of the mutant GLUCl.1 allele, allows diagnostic hybridization of the mutant and of the wild type GLUC 1.1 gene.

[268] Alternatively, the zygosity status of a specific mutant GLUCl.1 allele can be determined by using alternative sets of probes which specifically recognize mutant and wild type GLUC 1.1 alleles.

[269] If the plant is homozygous for the mutant GLUC 1.1 gene or the corresponding wild type GLUC 1.1 gene, the diagnostic hybridization assays described above will give rise to a single specific hybridization product, such as one or more hybridizing DNA (restriction) fragments, typical, preferably typical in length, for either the mutant or wild type GLUC 1.1 allele. If the plant is hemizygous for the mutant GLUC 1.1 allele, two specific hybridization products will appear, reflecting both the hybridization of the mutant and the wild type GLUC 1.1 allele.

[270] Identification of the wild type and mutant GLUC 1.1 specific hybridization products can occur e.g. by size estimation after gel or capillary electrophoresis (e.g. for mutant GLUC 1.1 alleles comprising a number of inserted or deleted nucleotides which results in a size difference between the hybridizing DNA (restriction) fragments from the wild type and the mutant GLUC 1.1 allele, such that said fragments can be visibly separated on a gel); by evaluating the presence or absence of the two different specific hybridization products after gel or capillary electrophoresis , whereby the diagnostic hybridization of the mutant GLUC 1.1 allele can, optionally, be performed separately from the diagnostic hybridization of the wild type GLUC 1.1 allele; by direct sequencing of the hybridizing DNA (restriction) fragments; or by fluorescence-based detection methods.

[271] Examples of probes suitable to determine the zygosity of specific mutant GLUC 1.1 alleles are described in the Examples.

[272] Furthermore, detection methods specific for a specific mutant GLUC 1.1 allele which differ from PCR- or hybridization-based amplification methods can also be developed using the specific mutant GLUCl.1 allele specific sequence information provided herein. Such alternative detection methods include linear signal amplification

detection methods based on invasive cleavage of particular nucleic acid structures, also known as InvaderTM technology, (as described e.g. in US patent 5,985,557 "Invasive Cleavage of Nucleic Acids", 6,001,567 "Detection of Nucleic Acid sequences by Invader Directed Cleavage, incorporated herein by reference), RT-PCR-based detection methods, such as Taqman, or other detection methods, such as SNPlex.

[273] hi another aspect of the invention, kits are provided. A "kit" as used herein refers to a set of reagents for the purpose of performing the methods of the invention, more particularly, the identification of a specific mutant GLUC 1.1 allele in biological samples or the determination of the zygosity status of plant material comprising a specific mutant GLUC 1.1 allele. More particularly, a preferred embodiment of the kit of the invention comprises at least two specific primers, as described above, for identification of a specific mutant GLUC 1.1 allele, or at least two or three specific primers for the determination of the zygosity status. Optionally, the kit can further comprise any other reagent described herein in the PCR identification protocol. Alternatively, according to another embodiment of this invention, the kit can comprise at least one specific probe, which specifically hybridizes with nucleic acid of biological samples to identify the presence of a specific mutant GLUC 1.1 allele therein, as described above, for identification of a specific mutant GLUC 1.1 allele, or at least two or three specific probes for the determination of the zygosity status. Optionally, the kit can further comprise any other reagent (such as but not limited to hybridizing buffer, label) for identification of a specific mutant GLUC 1.1 allele in biological samples, using the specific probe.

[274] The kit of the invention can be used, and its components can be specifically adjusted, for purposes of quality control (e.g., purity of seed lots), detection of the presence or absence of a specific mutant GLUC 1.1 allele in plant material or material comprising or derived from plant material, such as but not limited to cotton seeds, raw cotton, cotton bales, yarn, fabric, apparel, etc.

[275] The term "primer" as used herein encompasses any nucleic acid that is capable of priming the synthesis of a nascent nucleic acid in a template-dependent process, such as

PCR. Typically, primers are oligonucleotides from 10 to 30 nucleotides, but longer sequences can be employed. Primers may be provided in double-stranded form, though the single-stranded form is preferred. Probes can be used as primers, but are designed to bind to the target DNA or RNA and need not be used in an amplification process.

[276] The term "recognizing" as used herein when referring to specific primers, refers to the fact that the specific primers specifically hybridize to a nucleic acid sequence in a specific mutant GLUC 1.1 allele under the conditions set forth in the method (such as the conditions of the PCR identification protocol), whereby the specificity is determined by the presence of positive and negative controls.

[277] The term "hybridizing" as used herein when referring to specific probes, refers to the fact that the probe binds to a specific region in the nucleic acid sequence of a specific mutant GLUC 1.1 allele under standard stringency conditions. Standard stringency conditions as used herein refers to the conditions for hybridization described herein or to the conventional hybridizing conditions as described by Sambrook et al., 1989 (Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbour Laboratory Press, NY) which for instance can comprise the following steps: 1) immobilizing plant genomic DNA fragments or BAC library DNA on a filter, 2) prehybridizing the filter for 1 to 2 hours at 65°C in 6 X SSC, 5 X Denhardt's reagent, 0.5% SDS and 20 μg/ml denaturated carrier DNA, 3) adding the hybridization probe which has been labeled, 4) incubating for 16 to 24 hours, 5) washing the filter once for 30 min. at 68°C in 6X SSC, 0.1 %SDS, 6) washing the filter three times (two times for 30 min. in 30ml and once for 10 min in 500ml) at 68°C in 2 X SSC, 0.1 %SDS, and 7) exposing the filter for 4 to 48 hours to X-ray film at -70°C.

[278] As used in herein, a "biological sample" is a sample of a plant, plant material or product comprising plant material. The term "plant" is intended to encompass Gossyphim plant tissues, at any stage of maturity, as well as any cells, tissues, or organs taken from or derived from any such plant, including without limitation, any fibers, seeds, leaves, stems, flowers, roots, single cells, gametes, cell cultures, tissue cultures or protoplasts.

"Plant material", as used herein refers to material which is obtained or derived from a plant. Products comprising plant material relate to food, feed or other products, such as raw cotton, cotton bales, yarn, fabric, apparel, etc., which are produced using plant material or can be contaminated by plant material. It is understood that, in the context of the present invention, such biological samples are tested for the presence of nucleic acids specific for a specific mutant GLUC 1.1 allele, implying the presence of nucleic acids in the samples. Thus the methods referred to herein for identifying a specific mutant GLUC 1.1 allele in biological samples, relate to the identification in biological samples of nucleic acids which comprise the specific mutant GLUC 1.1 allele.

[279] The present invention also relates to the transfer of one or more specific mutant GLUC 1.1 allele(s) in one Gossypium plant to another Gossypium plant, to the combination of specific GLUCl.1 alleles in one plant, to the plants comprising one or more specific mutant GLUC 1.1 allele(s), the progeny obtained from these plants and to the plant cells, or plant material derived from these plants.

[280] Thus, in one embodiment of the invention a method for transferring a non- functionally expressed GLUCl.1 allele from one Gossypium plant to another Gossypium plant is provided comprising the steps of:

(a) crossing a Gossypium plant comprising a non-functionally expressed GLUC 1.1 allele, as described above, with a second Gossypium plant,

(b) collecting Fl hybrid seeds from the cross,

(c) optionally, backcrossing the Fl plants, derived from the Fl seeds, for one or more generations (x), collecting BCx seeds from the crosses, and identifying in every generation BCx plants, derived from the BCx seeds, comprising the non-functionally expressed GLUC 1.1 allele as described above,

(d) selfing the Fl or BCx plants, derived from the Fl or BCx seeds,

(e) collecting Fl Sl or BCx Sl seeds from the selfing,

(f) identifying Fl Sl or BCx Sl plants, derived from the Fl Sl or BCx Sl seeds, comprising the non-functionally expressed GLUC 1.1 allele as described above.

[282] In another embodiment of the invention a method for combining at least two non- functionally expressed GLUC 1.1 alleles in one Gossypium plant is provided comprising the steps of:

(a) transferring a non- functionally expressed GLUC 1.1 allele(s) from one Gossypium plant to another Gossypium plant as described above,

(b) repeating step (a) until the desired number and/or types of non- functionally expressed GLUC 1.1 alleles are combined in the second plant.

[283] In yet another embodiment of the invention, a method is provided for altering the callose content of a fiber in a fiber producing plant, such as Gossypium plants, comprising the steps of:

(a) abolishing the functional expression of at least one allele of at least one fiber specific GLUC gene that is functionally expressed during the fiber strength building phase of fiber development,

(b) identifying a plant, which produces fibers, the callose content of which is increased as compared to the callose content of the fibers of a corresponding plant in which the functional expression of the GLUC gene is not abolished.

[284] hi still another embodiment of the invention, a method is provided for altering the properties of a fiber, particularly increasing the strength of a fiber, in a fiber producing plant, such as a Gossypium plant, comprising the steps of:

(c) abolishing the functional expression of at least one allele of at least one fiber specific GLUC gene that is functionally expressed during the fiber strength building phase of fiber development,

(d) identifying a plant, which produces fibers, the strength of which is increased as compared to the strength of fibers of a corresponding plant in which the functional expression of the GLUC gene is not abolished.

[285] In another aspect of the invention, plant fibers with increased fiber strength are are provided derived from fiber-producing plants according to the invention, especially of Gossypium hirsutum plants as provided herein, but also from other Gossypium species.

For example, Gossypium species wherein the expression of at least one fiber specific GLUC gene that is functionally expressed during the fiber strength building phase of fiber development, such as a GLUC 1.1 A and/or GLUCl. ID gene, can be abolished, for example Gossypium tomentosum, Gossypium mustilinum, Gossypium herbaceum, or Gossypium raimondii.

[286] Also included in the invention is the use of the fibers of this invention, for example, in the production of raw cotton, cotton bales, yarn, fabric, apparel, etc.

[287] Other applications, such as mixing fibers with a specific callose content and/or a specific modified strength according to the invention with other fibers with a lower callose content and/or a lower fiber to increase the average callose content and/or fiber strength in, for example, cotton bales, yarn, fabric, apparel, etc; thus making it more suitable for certain applications, such as but not limited to, the production of biodiesel, stronger textile, etc., are also included in the invention.

[288] It will be clear that whenever nucleotide sequences of RNA molecules are defined by reference to nucleotide sequence of corresponding DNA molecules, the thymine (T) in the nucleotide sequence should be replaced by uracil (U). Whether reference is made to RNA or DNA molecules will be clear from the context of the application.

[289] It is understood that when referring to a word in the singular (e.g. plant or root), the plural is also included herein (e.g. a plurality of plants, a plurality of roots). Thus, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one".

[290] As used herein "comprising" is to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more features, integers, steps or components, or groups

thereof. Thus, e.g., a nucleic acid or protein comprising a sequence of nucleotides or amino acids, may comprise more nucleotides or amino acids than the actually cited ones, i.e., be embedded in a larger nucleic acid or protein. A chimeric gene comprising a DNA region, which is functionally or structurally defined, may comprise additional DNA regions etc. A plant comprising a certain trait may thus comprise additional traits etc.

[291] The following non-limiting Examples describe the identification of a fiber strength locus on chromosome A05 in cotton and the characterization of a GLUC 1.1 gene located in the 1-LOD support interval of the Strengt QTL. Unless stated otherwise in the Examples, all recombinant DNA techniques are carried out according to standard protocols as described in Sambrook et al. (1989) Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor Laboratory Press, NY and in Volumes 1 and 2 of Ausubel et al. (1994) Current Protocols in Molecular Biology, Current Protocols, USA. Standard materials and methods for plant molecular work are described in Plant Molecular Biology Labfax (1993) by R.D.D. Croy, jointly published by BIOS Scientific Publications Ltd (UK) and Blackwell Scientific Publications, UK.

[292] Throughout the description and Examples, reference is made to the following sequences represented in the sequence listing:

SEQ ID NO: 1: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium hirsutum cv. Fiber Max966, A-subgenome specific SEQ ID NO: 2: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 1 SEQ ID NO: 3: amplified cDNA fragment of endo-l,3-beta-glucanase gene from

Gossypium hirsutum cv. Fiber Max966, A-subgenome specific SEQ ID NO: 4: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 3 SEQ ID NO: 5: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium barbadense cv. PimaS7, A-subgenome specific SEQ ID NO: 6: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 5 SEQ ID NO: 7: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium hirsutum cv. Fiber Max966, D-subgenome specific

SEQ ID NO: 8: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 7

SEQ ID NO: 9: amplified cDNA fragment of endo-l,3-beta-glucanase gene from

Gossypium hirsutum cv. Fiber Max966, D-subgenome specific SEQ ID NO: 10: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 9 SEQ ID NO: 11: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium barbadense cv. PimaS7, D-subgenome specific SEQ ID NO: 12: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 11 SEQ ED NO: 13: amplified cDNA fragment of endo-l,3-beta-glucanase gene from

Gossypium barbadense cv. PimaS7, D-subgenome specific SEQ ID NO: 14: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 13 SEQ ID NO: 15: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium tomentosum, A-subgenome specific

SEQ ID NO: 16: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 15 SEQ ID NO: 17: amplified genomic DNA fragment of endo- 1,3 -beta- glucanase gene from Gossypium darwinii, A-subgenome specific

SEQ ID NO: 18: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 17 SEQ ID NO: 19: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium mustelinum, A-subgenome specific

SEQ ID NO: 20: endo- 1 ,3-beta-glucanase protein encoded by SEQ ID NO: 19 SEQ ID NO: 21: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium arboreum, A-subgenome specific

SEQ ID NO: 22: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 21 SEQ ID NO: 23: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium herbaceum, A-subgenome specific

SEQ ID NO: 24: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 23 SEQ ID NO: 25: amplified genomic DNA fragment of endo- 1,3 -beta- glucanase gene from Gossypium tomentosum, D-subgenome specific

SEQ ID NO: 26: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 25 SEQ ID NO: 27: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium darwinii, D-subgenome specific SEQ ID NO: 28: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 27

SEQ ID NO: 29: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium mustelinum, D-subgenome specific

SEQ ID NO: 30: endo- 1 ,3-beta-glucanase protein encoded by SEQ ID NO: 29 SEQ ID NO: 31: amplified genomic DNA fragment of endo-l,3-beta-glucanase gene from Gossypium raimondii, D-subgenome specific

SEQ ID NO: 32: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 31 SEQ ID NO: 33: forward primer SE077 for amplification of endo-l,3-beta-glucanase genomic fragment SEQ ID NO: 34: reverse primer SE078 for amplification of endo-l,3-beta-glucanase genomic fragment SEQ ID NO: 35: forward primer SE002 for amplification of endo-l,3-beta-glucanase genomic fragment SEQ ID NO: 36: reverse primer SE003 for amplification of endo-l,3-beta-glucanase genomic fragment SEQ ID NO: 37: forward primer pUGlucaAf for amplification of endo- 1,3 -beta- glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP2 SEQ ID NO: 38: reverse primer pl.3GlucaAr for amplification of endo- 1,3 -beta- glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP2 SEQ ID NO: 39: probe TM249-GCM1 for detecting the G. barbadense variant of polymorphic site GLUC 1.1 A-SNP3 SEQ ID NO: 40: probe TM249-GCV1 for detecting the G. hirsutum variant of polymorphic site GLUC1.1A-SNP3 SEQ ID NO: 41: forward primer TM249-GCF for amplification of endo- 1,3 -beta- glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP3 SEQ ID NO: 42: reverse primer TM249-GCR for amplification of endo- 1,3 -beta- glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP3

SEQ ID NO: 43: AFLP primer P5 for amplification of genomic DNA fragment corresponding to marker P5M50-M 126.7, in particular for discriminating different variants of marker P5M50-M 126.7 SEQ ID NO: 44: AFLP primer M50 for amplification of genomic DNA fragment corresponding to marker P5M50-M 126.7, in particular for discriminating different variants of marker P5M50-M126.7 SEQ ID NO: 45: forward SSR primer for amplification of genomic DNA fragment corresponding to marker NAU861, in particular for discriminating different variants of marker NAU861 SEQ ID NO: 46: reverse SSR primer for amplification of genomic DNA fragment corresponding to marker NAU861, in particular for discriminating different variants of marker NAU861 SEQ ID NO: 47: forward SSR primer for amplification of genomic DNA fragment corresponding to marker CIR401, in particular for discriminating different variants of marker CIR401 SEQ ID NO: 48: reverse SSR primer for amplification of genomic DNA fragment corresponding to marker CIR401, in particular for discriminating different variants of marker CIR401 SEQ ID NO: 49: forward SSR primer for amplification of genomic DNA fragment corresponding to marker BNL3992, in particular for discriminating different variants of marker BNL3992 SEQ ID NO: 50: reverse SSR primer for amplification of genomic DNA fragment corresponding to marker BNL3992, in particular for discriminating different variants of marker BNL3992 SEQ ID NO: 51: forward SSR primer for amplification of genomic DNA fragment corresponding to marker CIR280, in particular for discriminating different variants of marker CIR280 SEQ ID NO: 52: reverse SSR primer for amplification of genomic DNA fragment corresponding to marker CIR280, in particular for discriminating different variants of marker CIR280

SEQ ID NO: 53: DNA sequence of a 165250 bps DNA fragment spanning the

GLUC 1.1 A gene in G. hirsutum SEQ ID NO: 54: amplified cDNA fragment of endo-l,3-beta-glucanase gene from

Gossypium barbadense cv. PimaS7, A-subgenome specific SEQ ID NO: 55: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 54 SEQ ID NO: 56: amplified genomic DNA fragment of endo- 1,3 -beta- glucanase gene from Gossypium darwinii, A-subgenome specific

SEQ ID NO: 57: endo- 1 ,3-beta-glucanase protein encoded by SEQ ID NO: 56 SEQ ID NO: 58: amplified genomic DNA fragment of endo- 1,3 -beta- glucanase gene from Gossypium darwinii, D-subgenome specific

SEQ ID NO: 59: endo-l,3-beta-glucanase protein encoded by SEQ ID NO: 58 SEQ ID NO: 60: probe for detecting the G. barbadense variant of polymorphic site

GLUC1.1A-SNP5 SEQ ID NO: 61: probe for detecting the G. hirsutum variant of polymorphic site

GLUC1.1A-SNP5 SEQ ID NO: 62: forward primer for amplification of endo-l,3-beta-glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP5 SEQ ID NO: 63: reverse primer for amplification of endo-l,3-beta-glucanase genomic fragment, in particular for discriminating different variants of polymorphic site GLUC1.1A-SNP5 SEQ ID NO: 64: forward primer Gl.l-SGA-F for amplification of endo- 1,3 -beta- glucanase genomic fragment SEQ ID NO: 65: forward primer Gl.l-fl-Fl for amplification of endo- 1,3 -beta- glucanase genomic fragment

EXAMPLES

Example 1: Identification and characterization of a quantitative trait locus (QTL) on cotton chromosome A05 linked to fiber strength

1.1. QTL discovery

[293] Discovery of quantitative trait loci associated with cotton fiber properties was performed according to standard procedures. Briefly, parental cotton plant lines with fiber phenotypes of interest were selected, segregating populations were generated and the impact of the presence of specific chromosomal regions on measurable cotton fiber phenotypes was determined. The parental lines were Gossypium hirsutum cv. FM966 (used as female parent in the initial cross; abbreviated hereinafter as "FM"; particularly known for its high fiber yield, but lower fiber quality compared to Gossypium barbadense varieties) and Gossypium barbadense cv. PimaS7 (used as male parent in the initial cross; abbreviated hereinafter as "Pima"; particularly known for its excellent fiber quality, but lower fiber yield compared to Gossypium hirsutum varieties). Backcross populations with both parental lines were generated and evaluated in the greenhouse as well as in the field.

1.2. Evaluation of plants derived from a first backcross to the Gossypium barbadense Pima S7 parental line ("Pima BClFl population")

[294] A QTL for fiber strength on chromosome A05 was originally detected in a BClFl mapping population [(FM x Pima) x Pima; recurrent parent used as male parent] of 119 individuals. The population was grown under standard growing conditions in a greenhouse. A genome- wide genetic map of about 800 markers was constructed based on amplified fragment length polymorphism PCR (AFLP-PCR or AFLP) marker data and simple sequence repeat (SSR or microsatellite) marker data from the 119 individuals using JoinMap software (map version 8 and 13; Stam, 1993, Plant J 3: 739-744). Fiber strength was measured by High- Volume Instruments (HVI) (United States Department of Agriculture, Agricultural Marketing Service) on samples from 88 of the 119 individual plants. QTL mapping was performed using MapQTL software (Van Ooijen and

Maliepaard, 1996, Plant Genome IV Abstracts, World Wide Web site: http://www.intl- pag.org). Final QTL data are based on the restricted multiple QTL mapping (rMQM; Jansen, 1993, Genetics 135:205-211; Jansen and Stam, 1994, Genetics 136: 1447-1455) analysis.

[295] A clear QTL associated with fiber strength (also referred to as "Strength locus" or "Stren locus") was detected on chromosome A05. The QTL had a sharp LOD (logarithm of the odds) score peak with a maximum value of LOD 4.92 at a position of 98.61 cM from the tip of chromosome A05, with a 1-LOD support interval of 14 cM (from 91.515 cM to 105.61 cM). The 1-LOD QTL support interval was flanked by one AFLP marker, P5M50-M 126.7, at 85.515 cM, and one microsatellite marker, CIR401c, at 109.13 cM. Within the QTL support interval one microsatellite marker NAU861 (94.61 cM) and a GLUC 1.1 gene (94.602 cM) were located at close distance (ca 4 cM) to the position of maximum LOD value (Table 6). Primer pairs used to distinguish between the G. hirsutum and G. barbadense alleles of the markers are indicated in Table 2 above.

Table 6: Estimated position (according to JoinMap version 8 and 13) on chromosome A05 of markers linked to the fiber strength locus in the FM and Pima BClFl population

As indicated above, the GLUCl. IA gene was mapped within the support interval of the Strength locus (LOD of 4.431) using SNP marker GLUC1.1A-SNP2 as indicated in

Table 13 and primers pl.3GlucaAf (SEQ ID NO: 37) and pl.3GlucaAr (SEQ ID NO: 38) as described in Example 6 below. Plants homozygous for the GLUCl. IA allele of Gossypium barbadense Pima S7 (Pima GLUC 1.1 A allele or Gbglucl.lA) had 9.7% higher fiber strength compared to plants heterozygous for Gbglucl.lA (Ho/He ratio of 109.7%). The QTL explained 17.8% of the variation for fiber strength in the population.

1.3. Evaluation of plants derived from a first backcross to the Gossypium hirsutum FM966 parental line ("FM BClFl population")

[296] QTL mapping was also performed in a complementary BClFl population [(FM x Pima) x FM; recurrent parent used as male parent] of 130 individuals. Fiber strength was measured on samples from 94 of the 130 individual plants. The QTL for fiber strength in the region flanked by markers P5M50-M126.7 and CIR401 was not detected in this FM BClFl population (max LOD = 0.42, i.e. below the critical threshold value of LOD = 3). However, technically, plants heterozygous for the GLUC 1.1 A allele of Gossypium barbadense Pima S7 of this population did show about 1 to about 2% higher fiber strength compared to plants homozygous for the GLUCl. IA allele of Gossypium hirsutum FM966 (FM GLUCl. IA allele or GhGLUCl. IA). Together with the data from the Pima BClFl population this suggested that the GLUCl. IA allele of Gossypium barbadense Pima S7 provides superior fiber strength.

1.4. Evaluation of plants derived from a fourth backcross to the Gossypium hirsutum FM966 parental line ("FM BC4F1 population")

[297] With the purpose of improving fiber quality in Gossypium hirsutum, in particular in Gossypium hirsutum cv. FM966, genome fragments of the Gossypium barbadense parental line were backcrossed into the FM BClFl population by single seed descent and without selection during 4 generations (FM BC4F1 population). The Pima region of chromosome A05 carrying the candidate Strength locus was expected to be present in a number of these introgression lines.

[298] A total of 219 FM BC4F1 plants originating from 75 FM BC3F1 plants (average 3 sister plants per line) were grown under standard growing conditions in a greenhouse.

All plants were genotyped for 450 SSR markers and the strength of fibers from all plants was measured by HVI (see above). In the region of the Strength locus, 14 and 23 FM BC4F1 plants were heterozygous for the NAU861 and the GLUC 1.1 A markers, respectively, versus 196 and 194 plants that were homozygous for the NAU861 marker and the GLUCl. IA allele of Gossypium hirsutum FM966.

[299] Table 7 summarizes the impact on fiber strength of the presence of different Pima marker alleles in heterozygous state versus the equivalent FM marker alleles in homozygous state (He/Ho ratio) in FM BClFl and FM BC4F1 populations. Markers indicated as CIRx, NAUx, JESPRx and BNLx are publicly available markers (see Cotton Microsatellite Database at http://www.cottonmarker.org/). Markers indicated as 'Primer combination X and Y-amplified fragment size' are AFLP markers (Vos et ai, 1995, NAR 23:4407-4414).

[300] A similar effect on fiber strength was observed in both the FM BClFl and FM BC4F1 populations for the presence of the Pima GLUCl. IA allele (i.e. plants heterozygous for the Pima GLUC 1.1 A allele showed about 1 to about 2% higher fiber strength compared to plants homozygous for the FM GLUC 1.1 A allele).

Table 7: Estimated position (according to JoinMap version 8 and 13) on chromosome A05 and impact on fiber strength of different allele combinations (He versus Ho FM) for markers linked to the fiber strength locus in FM BClFl and FM BC4F1 populations

1.5. Evaluation of plants derived from the F2 generation of a fourth backcross to the Gossypium hirsutum FM966 parental line ("FM BC4F2 population")

[301] As a next step, QTL validation in FM BC4F2 families was performed under field conditions in summer in Mississippi. FM BC4F2 plants segregate in 3 genetic classes: plants homozygous for FM marker alleles, plants homozygous for Pima marker alleles and plants heterozygous for FM and Pima marker alleles, hi most cases 75-80 plants were genotyped per line and fiber samples from about 50 single plants were analyzed. This allowed testing of the effect of the FM or Pima marker alleles (and predicted linked genes) in heterozygous and homozygous condition.

The field trial included 4 FM BC4F2 families (called lines 6, 10, 20 and 94) segregating for various portions of the region of chromosome A05 carrying the Strength locus from Pima S7. Segregation was tested using 6 markers: BNL0542, BNL3995, CIR139a, NAU861, GLUC 1.1 A, BNL3992.

[302] All BC4F2 plants of line 6 were homozygous for the FM allele of the markers tested. Line 94 produced only 38 FM BC4F2 plants and only 10 of those produced sufficient fiber for single plant analysis. The two remaining lines, lines 10 and 20, produced larger numbers of plants and had good marker segregation. Line 10 contained a segment of chromosome A05 of Pima carrying the Strength locus centered around the GLUC 1.1 gene. The second line, line 20, contained a segment of chromosome A05 of Pima shifted to the lower end of the Strength locus support region.

[303] In line 10 the expectation that plants homozygous for the Pima GLUCl. IA allele produce stronger fibers was confirmed. The fiber strength of plants homozygous for the Pima GLUC 1.1 A allele was on average 2.5 grams per tex higher than the fiber strength of plants homozygous for the FM GLUCl. IA allele (35.5 g/tex versus 33.0 g/tex or 7.5% increase in fiber strength). A similar result was observed for the two markers NAU861 and BNL3992 which are closely linked to GLUC 1.1 A on either side. The differences in fiber strength between homozygous FM plants, homozygous Pima plants and

heterozygous plants were not significant in Anova, but they were significant in paired t- test between homozygous FM plants and the other two classes.

[304] In line 20 the Pima alleles of markers NAU861 and BNL3992 did not provide stronger fiber. This line segregates for a lower section of the region of Pima chromosome A05, in the tail of the QTL support interval. This line also does not contain the Pima allele of the GLUC 1.1 A gene.

[305] The data in Table 8 consolidate the results for line 10 in terms of "Marker Trait Performance" for fiber strength (MTP, calculated as ratio of the difference in average trait performance for two marker classes (HoFM-HoPima) and the average standard deviation for trait performance in both marker classes). It is shown that plants homozygous for the Pima allele of markers NAU861, GLUCl. IA and BNL3992 had stronger fibers than plants homozygous for the FM allele of these markers (negative MTP). However, the difference in performance was smaller than the average standard deviation (MTP value between 0 and -1).

[306] Thus, the field trial data provide evidence in support of the idea that there is a QTL associated with fiber strength on chromosome A05, close to or coinciding with the GLUC 1.1 A gene, with the superior allele coming from Gossypium barbadense PimaS7.

[307] Due to the low number of plants in the FM BC4F2 population it was not possible to fine map the QTL position. In this respect it is noted that the Pima allele of a marker (BNL3992) that was included in the introgressed Pima fragment in line 10, but resided at a position outside the original support interval on the Pima BClFl map also segregated with the enhanced fiber strength derived from PimaS7. This can be explained by the fact that in the original BCl population sufficient recombinations had occurred to place this marker outside the QTL support interval, while in the (smaller) BC4F2 populations it remained linked to the QTL causal gene more frequently.

Table 8: Estimated position on chromosome A05 and impact on fiber strength (indicated as MTP) of different allele combinations (HH FM versus HH Pima) for markers linked to the Strength locus in FM BC4F2 plant lines

[308] Column 2 lists markers on chromosome A05 linked to the Strength locus. Markers indicated as CIRx, NAUx and BNLx are publicly available markers (see Cotton Microsatellite Database at http://www.cottonmarker.org/). Markers indicated as 'Primer combination X and Y-amplified fragment size' are AFLP markers (Vos et al., 1995, NAR 23:4407-4414). Column 1 indicates their map positions on the genetic map (in cM) of the FM BClFl mapping population constructed using JoinMap software map version 8. Graphical genotypes for the markers are indicated for BC4F1 plants that gave rise to BC4F2 families 6, 10, 20 and 94: a = homozygous FM966, h = heterozygous. Segregation of the 'h' regions in the graphical genotypes was investigated using marker data for markers indicated with *. Average phenotypic performance for fiber strength was compared for groups of plants homozygous for FM966 markers (genotype "HH FM")

and for groups of plants homozygous for Pima markers (genotype "HH Pima"). Marker Trait Performance (MTP) is expressed as ((average phenotype HH FM - average phenotype HH Pima)/0.5 x (SD HH FM + SD HH Pima)). Positive MTP means performance FM is higher than performance Pima. Negative MTP means performance Pima is higher than performance FM. MTP higher than 1 and MTP lower than -1 means delta performance exceeds average standard deviation (SD). Data for fiber strength properties are based on homozygous segregates among 60 plants.

Example 2: Identification and characterization of a glucanase gene linked to the fiber strength locus on cotton chromosome A05

2.1 Characterization of the GLUCl. IA gene localized in the support interval of the Strength locus

[309] As described in Example 1.2, a GLUCl.1 gene was mapped within the support interval of the predicted QTL for fiber strength on chromosome A05, suggesting that the GLUCl. IA candidate gene might be the causal gene for fiber strength. As further described in Example 1 , the superior allele comes from the Pima parental line rather than from the FM parental line.

[310] Based on the GhGLUCl. IA and D nucleotide sequences described in WO2008/083969 (SEQ ID NO: 1 and 7, respectively), 2 primers (forward primer SE077 (SEQ ID NO: 33) en reverse primer SE078 (SEQ ID NO: 34)) were designed to amplify genomic DNA fragments for G. barbadense (reaction mix and PCR conditions as described in Example 4). Two genomic DNA sequences were derived: one for GbGLUCl. IA (SEQ ID NO: 5) and one for GbGLUCLlD (SEQ ID NO: 11).

[311] The 2 primers (forward primer SE077 (SEQ ID NO: 33) en reverse primer SE078 (SEQ ID NO: 34)) were also used to amplify GLUCl. IA and GLUCl. ID cDNA from cDNA libraries from G. hirsutum and G. barbadense (reaction mix and PCR conditions as described in Example 4). cDNA sequences were derived for GhGLUCl. IA (SEQ ID NO: 3), for GhGLUCl. ID (SEQ ID NO: 9), and for GbGLUCl. ID (SEQ ID NO: 13).

Forward primer G 1.1 -SGA-F (SEQ ID NO: 64) en reverse primer SE078 (SEQ ID NO: 34) were used to amplify GLUC 1.1 A cDNA from a cDNA libraries from G. barbadense. The cDNA sequence was derived for GbGLUCl. IA (SEQ ID NO: 54).

[312] Alignment of genomic and cDNA sequences of A and D subgenome-specific GLUC 1.1 genes from Gossypium hirsutum and Gossypium barbadense indicated that the GLUC 1.1 A gene from Gossypium barbadense displayed a c to t nucleotide substitution (at position 712 of SEQ ID NO: 5) that resulted in a putative premature STOP codon (cga to tga) as compared to the GLUC 1.1 A and D genes from Gossypium hirsutum and the GLUC 1.1 D gene from Gossypium barbadense (Figure 1), that is predicted to result in the production of a truncated GLUC 1.1 A protein in Gossypium barbadense (Figure 2). Compared to the Gossypium hirsutum ortholog, the Gossypium barbadense GLUC 1.1 A amino acid sequence lacks the GH 17 signature (Figure 2).

2.2. Characterization of the GLUCl. IA protein from different Gossypium sp.

[313] Protein modeling based on an X-ray structure of a barley 1,3-1,4-beta-glucanase belonging to the GH 17 family of glycosidase hydrolases (laqO in Protein Data Bank) (Figure 3, left), using FUGUE™ and ORCHESTRAR™ technologies from Sybyl7.3, showed that the GLUC 1.1 A protein of G. barbadense (Figure 3b, right) is missing the active site and substrate binding cleft (located within the area indicated by the amino acids and their position numbers, displayed in the upper left part of the protein model of laqO and described in Mϋller et al., 1998, J Biol Chem 273: 3438-3446), which was found to be present in the GLUC 1.1 A and D proteins of G. hirsutum and in the GLUCl. ID protein of G. barbadense (Figure 3a, right). The GLUCl. IA protein of G. barbadense is therefore predicted to be inactive.

2.3. Characterization of the genomic regions spanning the GLUCl.1 alleles from different Gossypium sp.

[314] DNA sequencing of an about 165kb and 136kb region spanning the GLUCl. IA (SEQ ID NO: 53) and GLUCl. ID alleles (not shown), respectively, of Gossypium hirsutum was undertaken using 454 DNA sequencing (454 Life Sciences): Firstly BAC

clones with genomic DNA spanning each GhGLUC 1.1 allele were identified by hybridization using part of the GLUC 1.1 gene as a probe against a FM BAC library. The BAC clones were isolated, confirmed by PCR and grouped into alleles. Selected BAC clones were sequenced to define neighboring genes facilitated by bioinformatics annotation software programs and EST searches (see Figure 9). The BAC sequence data also identified an additional molecular marker (CIR280) located on an adjacent gene (HAT) (see Table 6 and 7 for estimated position on chromosome A05 in the FM BCl population).

Example 3: Analysis of the biological role of glucanase in fiber strength

3.1. Determination of link between inactive GbGLUCl. IA enzyme and fiber strength

[315] To determine if there is a link between the inactive GbGLUC 1.1 A enzyme and fiber strength, the impact of glucanase activity on fiber strength was analyzed by exogenous addition of a 1,3-beta-glucanase enzyme to fibers from G. barbadense (comprising a GLUC 1.1 A predicted to be inactive), as well as fibers from G. hirsutum (comprising a GLUC 1.1 A predicted to be active). It was expected that the strength of the G. barbadense fibers would significantly decrease, if there was indeed a link between the inactive GbGLUC 1.1 A enzyme and fiber strength.

[316] Individual fibers were treated with a beta-l,3-D-glucanase from Helix pomatia (Fluka, 49103). 10 mg of fibers were incubated in 1OmM sodium acetate buffer (pH 5) and 500 μl of glucanase (lmg/ml) was added. They were subjected to infiltration under vacuum for 10 minutes and overnight incubation at 37°C. The strength of individual cotton fibers was measured using a Favimat R device (Textechno) in a single fiber tensile test at 8 mm gauge length and a speed of 4 mm/min. The strength measure is recorded in force (cN). The results were statistically analyzed and are presented in Table 9 and Figure 4.

Table 9: Callose content (as measured by the green/blue fluorescence ratio of aniline blue stained fibers (ratio green/blue)) and strength (as measured by the breaking force (cN)) of untreated fibers (no GLUC) and fibers treated with glucanase (GLUC) from different G. hirsutum and G. barbadense varieties

[317] A pronounced drop in strength was observed for Pima fibers treated with the glucanase and a less pronounced but still noticeable reduction in strength was observed for fibers from various G. hirsutum lines. In this respect, it is important to note that the extent of secondary cell wall formation and cellulose content contribute to fiber strength in G. hirsutum, while the stronger fibers of G. barbadense have a lower cellulose content than those of G. hirsutum. The complementation experiment thus indicated that the presence of the Gbglucl.lA allele within the fiber strength locus contributes to the renowned strength of Pima fibers.

3.2. Determination of link between 1,3-beta-D-glucan content and fiber strength

[318] 1,3-beta-D-glucans, including long chain 1,3-beta-D-glucans called callose, are the substrate for 1,3-beta-glucanase enzymes. Aniline blue is a dye specific for 1,3-beta- glucans. This dye was used to determine if fibers treated with 1,3-beta-glucanase and displaying a reduced fiber strength also displayed a reduced level of the 1,3-beta-glucan substrate in the cotton fiber walls.

[319] A 0.05% solution of aniline blue in 0.067M K ₂ HPO ₄ (pH 9) was used. The fibers were incubated for 15 minutes under vacuum. Under UV, callose deposits present an intense yellow-green fluorescence. Images are analyzed and the ratio Green/Blue is used as a measure for callose. The average value of 3 images was calculated.

[320] As indicated in Table 9 and Figure 5, this staining technique showed that cotton fibers treated with the glucanase had a lower level of 1,3-beta-glucan and that elevated 1,3-beta-glucan levels were linked to enhanced fiber strength.

3.3. Statistical analysis of effect of glucanase treatment on fiber strength and callose content

[321] The effect of the treatment (untreated minus treated) was statistically analyzed. The results are presented in Table 10.

Table 10: Statistical analysis of glucanase treatment (untreated minus treated) on callose content and strength of fibers from different G. hirsutum and G. barbadense varieties

Callose content Fiber strength

(ratio G/B) (Force) difference p-value difference p-value

G. hirsutum cv. FM966 (greenhouse) 0.01 0.882 -0.18 0.618

G. hirsutum cv. FM966 (field US) -0.04 0.634 1.05 0.041*

G. hirsutum cv. FM966 (field AU) 0.01 0.922 1.03 0.003*

G. hirsutum cv. Coker312 (greenhouse) 0.03 0.415 1.41 0.002*

G. barbadense cv. PimaS7 (greenhouse) 0.11 0.278 2.55 0.000*

G. barbadense cv. PimaY5 (field AU) 0.08 0.121 3.07 0.000*

[322] The correlations between the treatment and callose content as well as fiber strength were statistically analyzed. The results are presented in Table 11 for G. hirsutum and in Table 11 for G. barbadense.

Table 11 : Statistical analysis of correlations between glucanase treatment of fibers of G. hirsutum, their callose content and their strength

Glucanase Callose content Fiber strength (Force) treatment (ratio G/B)

Glucanase Correlation 1.00 -0.03 -0.48 treatment Sig. (2-tailed) 0.944 0.233

Callose Correlation -0.03 1.00 0.66 content (ratio Sig. (2-tailed) 0.944 0.075

G/B)

Fiber Correlation -0.48 0.66 1.00 strength Sig. (2-tailed) 0.233 0.075

(Force)

Table 12: Statistical analysis of correlations between glucanase treatment of fibers of G. barbadense, their callose content and their strength

Glucanase Callose content treatment (ratio G/B) Fiber strength (Force)

Glucanase Correlation 1.00 -0.96 -0.99 treatment Sig. (2-tailed) 0.044* 0.013*

Callose Correlation -0.96 1.00 0.90 content (ratio Sig. (2-tailed)

G/B) 0.044* 0.103

Fiber Correlation -0.99 0.90 1.00 strength Sig. (2-tailed)

(Force) 0.013* 0.103

[323] In summary, cotton fibers with a higher 1,3-beta-glucan content displayed higher fiber strength and reduction in 1,3-beta-glucan content by exogenously supplied 1,3-beta- glucanase enzyme significantly reduced fiber strength and callose content in G. barbadense, indicating that 1,3-beta-glucan or callose has a specific role in cotton fiber strength which can be modulated by enzymes such as GLUC 1.1.

Example 4: Identification of GLUCl. IA alleles in different cotton species

[324] GLUCl.1 sequences were isolated from six different Gossypium hirsutum varieties (Guazuncho; DP16; Cooker 312 (C312); Fiber Max 966 (FM966); Acala SJ2; Acala Maxxa), from five different Gossypium barbadense varieties (PimaS7; Tanguis LMW 1737-60; Tanguis CN(C.P.R.)712-60; Sea Island Tipless; VH8), from Gossypium herbacium, Gossypium tomentosum, Gossypium darwinii, Gossypium arboreum, Gossypium raimondii, Gossypium kirkii, Gossypium longicalyx, and Gossypium mustelinum

[325] Based on the GhGLUCl. IA and D nucleotide sequences described in WO2008/083969 (SEQ ID NO: 1 and 7, respectively), primer pairs (forward primer SE077 (SEQ ID NO: 33) and Gl . l-fl-Fl (SEQ ID NO: 65) en reverse primer SE078 (SEQ ID NO: 34) or forward primer SE002 (SEQ ID NO: 35) en reverse primer SE003

(SEQ ID NO: 36)) were designed to amplify full-length or partial, respectively, genomic DNA fragments. The reaction mix used contained: 2μl DNA (200ng/μl genomic DNA), lμl forward primer (10 pM), lμl reverse primer (10 pM), 4 μl 5x High Fidelity buffer, 0.2 μl Phusion enzyme (Finnzymes), 0.4 μl dNTP's (1OmM), 11.4 μl water (MiIIiQ). The PCR protocol used was as follows: 1 min at 98°C; 30 times: 10 sec at 98°C (denaturation), 30 sec at 56°C (annealing), 1 min at 72°C (elongation); 30 sec at 58°C; 10 min at 72°C; 4°C.

[326] GLUCl. IA sequences from all G. barbadense lines tested and from Gossypium darwinii display a single nucleotide substitution (c to t at position 712 of SEQ ID NO: 5 and at position 470 of SEQ ID NO: 17 or at position 761 of SEQ ID NO: 56, respectively; see also GLUC1.1A-SNP5 in Table 13) resulting in a premature stop codon (cga to tga) in their sequences (Figure 6; since the GLUCl.1 sequences from the different Gossypium hirsutwn varieties and the different Gossypium barbadense varieties, respectively, were identical to each other, only the GLUC 1.1 sequences of the FM966 and PimaS7 variety, respectively, were included in the alignment). The GLUCl. IA sequence from G. arboreum displayed a single nucleotide deletion (deletion of c nucleotide between position 327 and 328 of SEQ ID NO: 21) also resulting in a premature stop codon (tga at position 373-375 of SEQ ID NO: 21) further downstream in its sequence (Figure 6). The premature stop codons in the GLUC 1.1 A sequences from G. barbadense, from Gossypium darwinii and from G. arboreum resulted in a predicted truncated GLUCl. IA protein sequence (Figure 7; GLUCl. IA protein of 179 (SEQ ID NO: 6), of 179 (SEQ ID NO: 57), and of 78 (SEQ ID NO: 22) amino acids, respectively), while the GLUCl. IA sequences from all other Gossypium species tested did not display premature stop codons and are predicted to produce a complete GLUC 1.1 protein (Figure 6 and 7).

[327] As indicated above, G. barbadense is commercially recognized for its superior fiber quality, particularly for fiber strength, length and fineness. G. darwinii is the closest relative of G. barbadense and some even consider it as a variety of G. barbadense rather than a separate species. However, G. darwinii produces sparse, non-spinnable, khaki or

brown fiber, usually less than 1.3 cm in length (see e.g. Wendel and Percy, 1990, Bioch. Systematics And Ecology 18 (7/8): 517-528). As the fibers from G. darwinii are not commercially used, little information is available about its commercially relevant fiber qualities, such as fiber strength.

Example 5: Genotyping of GLUCl.1 genes in commercial germplasm

[328] The genotype of GLUCl. IA and GLUCLlD genes was determined in commercially available germplasm by determining the genotype of GLUC1.1A-SNP3, 5 and 6 and GLUC 1.1 D-SNPl (as indicated in Figure 6 and Table 13) in a total of 73 G. hirsutum varieties, one G. barbadense variety, 2 G. arboreum varieties, one G.herbaceum variety, and one G. mustilinum variety using Illumina GoldenGate SNP Genotyping and BeadArray technology as prescribed by the manufacturer. Briefly, a GoldenGate Genotyping assay uses allele-specific extension and ligation for genotype calling using a discriminatory DNA polymerase and ligase (Illumina).

Table 13: Position and genotype of GLUC 1.1 D-SNPl and GLUCl.1A-SNP2, 3, 5, 6, 7 and 8 in GLUCl. ID and A genes, respectively of different Gossypium species {G.h.: G. hirsutum, G.b.: G. barbadense, G.t.\ G. tomentosum; G.d.: G.darwinii; G.m.: G. mustilinum; G.a.\ G. arboreum G.he.: G.herbaceum Gr.: G.raimondiϊ)

[329] The results confirmed that the genotypes of GLUC1.1A-SNP3, 5 and 6 and GLUC 1.1 D-SNPl in the different analysed Gossypium species and varieties were as indicated in Figure 6 and Table 13. In particular, geno typing of GLUC1.1A-SNP5 in the different Gossypium species and varieties indicated that all analysed Gossypium species and varieties different from G. barbadense comprise the cga codon found in GLUC 1.1 A of Gossypium hirsutum instead of the tga stop codon found in glue 1.1 A of Gossypium barbadense Pima S7.

Example 6: Detection of GLUCl.1 allele encoding an inactive GLUCl.1 protein in Gossypium plants and/or transfer of GLUCl.1 allele encoding an inactive GLUCl.1 protein into Gossypium lines comprising a corresponding GLUCl.1 allele encoding an active GLUCl.1 protein

[330] A GLUCl.1 allele encoding an inactive GLUC 1.1 enzyme, such as a Gbglucl.lA allele, Gdglucl.lA allele or Gaglucl.lA allele, is transferred into cotton lines comprising a corresponding GLUCl.1 allele encoding an active GLUC 1.1 enzyme, such as Gossypium hirsutum breeding lines, by the following method:

[331] A plant containing a GLUC 1.1 allele encoding an inactive GLUC 1.1 enzyme, such as a Gossypium barbadense plant, a Gossypium danvinii plant or a Gossypium arbor eum plant containing a GLUC 1.1 A allele encoding an inactive GLUC 1.1 A enzyme,

or a mutagenized Gossypium hirsutum plant containing a mutant GLUC 1.1 allele encoding an inactive GLUC 1.1 enzyme (donor plant), is crossed with a plant containing a corresponding GLUC 1.1 allele encoding an active GLUC 1.1 enzyme, such as a Gossypium hirsutum plant containing a GLUCl. IA allele encoding an active GLUC 1.1 A enzyme (recurrent parent). The following introgression scheme is used (the GLUCl.1 allele encoding an inactive GLUC 1.1 enzyme is abbreviated to glue while the GLUC 1.1 allele encoding an active GLUC 1.1 enzyme is depicted as GLUC):

[332] Initial cross: glue I glue (donor) X GLUCI GL UC (recurrent parent)

[333] Fl plant: GLUC /glue

[334] BCl cross: GLUC / glue (Fl) X GLUCI GL UC (recurrent parent)

[335] BC 1 plants: 50% GLUC /glue and 50% GLUC I GLUC

[336] The 50% GLUC /glue are selected using a specific assay (e.g. PCR, TaqMan™,

Invader™, and the like; see also below) for the glue 1.1 allele.

[337] BC2 cross: GLUC / glue (BCl) X GLUCI GL UC (recurrent parent)

[338] BC2 plants: 50% GLUC /glue and 50% GLUC I GLUC

[339] The 50% GLUC /glue are selected using a specific assay (e.g. PCR, TaqMan™,

Invader™, and the like; see also below) for the glucl.l allele.

[340] Backcrossing is repeated until BC4 to BC5 (e.g. if the donor plant is a Gossypium barbadense plant and the recurrent parent is a Gossypium hirsutum plant) or until BC3

(e.g. if the donor plant and the recurrent parent are Gossypium hirsutum plants)

[341] BC3-5 plants: 50% GLUC /glue and 50% GLUC I GLUC

[342] The 50% GLUC /glue are selected using a specific assay (e.g. PCR, TaqMan™,

Invader™, and the like; see also below) for the glucl.l allele.

[343] To reduce the number of backcrossings (e.g. until BC2 if the donor plant and the recurrent parent are Gossypium hirsutum plants, or until BC3 to BC4 if the donor plant is a Gossypium barbadense plant and the recurrent parent is a Gossypium hirsutum plant), molecular markers can be used in each generation that are specific for the genetic background of the recurrent parent.

[344] BC3-5 Sl cross: GLUC / glue X GLUC /glue

[345] BC3-5 S 1 plants: 25% GLUC / GLUC and 50% GLUC /glue and 25% glue /glue

[346] Plants containing the glucl.l allele are selected using molecular markers for the glue 1.1 allele. Individual BC3-5 Sl plants that are homozygous for the glucl.l allele {glue /glue) are selected using molecular markers for the glucl.l and GLUC 1.1 alleles. These plants are then used for fiber production.

[347] Molecular markers which can be used to detect a specific glucl.l or GLUC 1.1 allele or to discriminate between a specific glucl.l and GLUCl.1 allele are, for example, single nucleotide polymorphisms (SNPs) or polymorphic nucleotide sequences:

[348] As an example, SNPs and polymorphic nucleotide sequences which can be used to discriminate between the Gbglucl.lA or Gdglucl.lA allele and the GhGLUCl. IA allele and between the GbGLUCl. ID or Gdglucl.lD allele and the GhGLUC 1.1 D allele or to detect their presence in DNA samples or plants, are SNPs indicated as GLUC 1.1 A- SNP3, 5 and 6 in Figure 6 and Table 13 and the polymorphic nucleotide sequence indicated as GLUC1.1A-SNP2 in Figure 6 and Table 13 and the SNP indicated as GLUC 1.1 D-SNPl in Figure 6 and Table 13, respectively.

[349] In particular, a SNP which can be used to discriminate between the Gbglucl.lA or Gdglucl.lA allele that comprises a premature tga STOP codon and the corresponding GhGLUCl. IA allele that comprises a cga codon instead, is the SNP indicated as GLUC1.1A-SNP5 in Figure 6 and Table 13.

[350] The genotype of such SNPs and polymorphic nucleotide sequences can be determined, for example, using a PCR assay.

[351] As an example, PCR assays were developed to determine the genotype of the SNP indicated as GLUC 1.1 D-SNPl in Figure 6 and Table 13 and of the polymorphic nucleotide sequence indicated as GLUC1.1A-SNP2 in Figure 6 and Table 13 of plants of the BCl populations described in Example 1 in order to map the GLUCl. ID and A genes of G. hirsutum and barbadense, respectively. More specifically, following PCR assay was developed to discriminate between the Gbglucl.lA allele and the GhGLUC 1.1 A

allele based on the genotype of the SNP indicated as GLUC1.1A-SNP2 in Figure 6 and Table 13: Primers: Forward: 5 ' TAT CCC TCT CGA TGA GTA CGA C 3 '

(pl.3GlucaAf- SEQ ID NO: 37) Reverse: 5'CCC AAT GAT GAT GAA CCT GAA TTG3 '

(pl.3GlucaAr - SEQ ID NO: 38) Amplicon size: 134 bps for G. hirsutum and 143 bps for G. barbadense.

- PCR conditions: 5μl gDNA (20ng/μl) + 15μl PCR mix (PCR mix: 2 μl 10 x Taq PCR buffer, lμl labeled pl.3GlucaAf (100 pmol/μl), 0.2μl pl.3GlucaAr (100 pmol/μl), 0.25μl dNTPs (2OmM), 0.5μl MgCl ₂ (50 mM), 0.2μl Taq polymerase, 10.85 μl MiIiQ)

- Labeling of forward primer: O.lμl 10 x T4 kinase buffer, 0.2μl pl.3GlucaAf(100 pmol/μl), 0.0 lμl T4 kinase, O.lμl P ³³ γ ATP, 0.59μl MiIIiQ = 1 μl; Ih at 37°C and 10 min at 65°C

- PCR profile: 5 min at 95 ⁰ C; 35 times: 45s at 95°C, 45s at 58°C, 1 min at 72°C; 10 min at 72°C.

Gel analysis: PCR fragments are separted on 4.5% denaturing acrylamide gels Overnight exposure of gel to BIOMAX MR films

[352] Alternatively, the genotype of such SNPs can be determined, for example, using Illumina GoldenGate SNP Genotyping as indicated in Example 5 for the SNPs indicated as GLUC1.1A-SNP3, 5 and 6 and GLUCl. ID-SNPl in Figure 6 and Table 13.

[353] Alternatively, the genotype of such SNPs and polymorphic nucleotide sequences can be determined by direct sequencing by standard sequencing techniques known in the art to determine the complete GLUCl.1 nucleotide sequence present in a plant followed by analysis of the obtained sequence, e.g., by alignment with the GLUCl.1 sequences described herein (see, e.g., Figure 6 and 7).

[354] Alternatively, the genotype of such SNPs and polymorphic nucleotide sequences can be determined by a Taqman assay. The TaqMan assay procedure and interpretation of the data are performed as prescribed by the manufacturer (Applied Biosystems). Briefly, a probe specific for a specific variant of a polymorphic site in a GLUC 1.1 gene binds the template DNA if this specific variant is present. The probe has a fluorescent reporter or fluorophore, such as 6-carboxyfluorescein (acronym: FAM) and VIC (a proprietary dye from Applied Biosystems), attached to its 5' end and a quencher (e.g., tetramethylrhodamine, acronym: TAMRA, of dihydrocyclopyrroloindole tripeptide "minor groove binder", acronym: MGB) attached to its 3' end. The close proximity between fluorophore and quencher attached to the probe inhibits fluorescence from the fluorophore. During a PCR with two primers capable of amplifying a DNA fragment comprising the polymorphic site, the 5' to 3' exonuclease activity of the Taq polymerase degrades that proportion of the probe that has annealed to the template as DNA synthesis commences. Degradation of the probe releases the fluorophore from it and breaks the close proximity to the quencher, thus relieving the quenching effect and allowing fluorescence of the fluorophore. Hence, fluorescence detected in the real-time PCR thermal cycler is directly proportional to the fluorophore released and the amount of DNA template present in the PCR. The following discriminating Taqman probes and primers were thus developed to discriminate different variants of GLUC1.1A-SNP3 and GLUC1.1A-SNP5 (see Figure 6 and Table 13):

Table 14a

Probes

GLUCl.lA -SNP3 of

Gbglucl.lA 5' FAM- AACTCGCTCGCCTCA 3' (SEQ ID NO: 39)

GhGLUCLlA 5' VIC-AACTCGCTGGCCTCA 3' (SEQ ID NO: 40)

Forward primer 5' CCTGGTGCCATGAACAACATAATG 3' (SEQ ID NO: 41) reverse primer 5' CGTCGTGCCTAGCCCAAA 3' (SEQ ID NO: 42)

Table 14b

Probes

GLUCl.lA -SNP5 of

Gbglucl.lA 5' FAM- ATCCTGTCAAACCAG 3' (SEQ ID NO: 60) GhGLUCLlA 5' VIC-ATCCTGTCAAACCAG 3' (SEQ ID NO: 61)

Forward primer 5' GCTTTTGGAAGCGATATAACATCGA 3' (SEQ ID NO: 62) reverse primer 5' GGCATAGGCAAAATAAGGGTACACA 3' (SEQ ID NO: 63)

[355] Probes specific for polymorphic sites in the Gbglucl.lA or corresponding GhGLUCLlA target gene, such as the probes specific for GLUCl.1A-SNP3 of Gbglucl.lA and GhGLUCLlA indicated as "5' FAM- AACTCGCTCGCCTCA 3" and "5' VIC-AACTCGCTGGCCTCA 3', respectively, in Table 14a, and forward and reverse primers that are capable of amplifying a fragment comprising the polymorphic site and that can thus be used in combination with them are indicated in Table 14a. Generally, each probe set consists of two probes each specific for one variant of the polymorphic site in the GLUCLl target gene which comprises the variant nucleotide (e.g., the underlined nucleotide in Table 14) or variant nucleotide sequence (e.g. the probe with SEQ ID NO: 39 is specific for GLUC1.1A-SNP3 of Gbglucl.lA and the probe with SEQ ID NO: 40 is specific for GLUCl.1A-SNP3 of GhGLUCLlA) and a set of two primers that are capable of amplifying a fragment comprising the polymorphic site (e.g. the primer with SEQ ID NO: 41 is specific for a nucleotide sequence upstream of GLUC1.1A-SNP3 and the primer with SEQ ID NO: 42 is specific for a nucleotide sequence downstream of GLUC1.1A-SNP3, such that the use of both primers results in the amplification of a DNA fragment comprising GLUC1.1A-SNP3).

[356] Alternatively, the genotype of such SNPs and polymorphic nucleotide sequences can be determined by Invader technology (Third Wave Agbio).

Example 7: Comparison of expression of GLUCLlA and GLUCl. ID during fiber growth and development in Gossypium barbadense and in Gossypium hirsutum

[357] Expression of GLUCLlA and GLUCLlD during fiber growth and development was analyzed for G. barbadense and compared with the expression of GLUCLlA and GLUCLlD during fiber growth and development of G. hirsutum as described in WO2008/083969.

[358] DNA from a cDNA library of G. barbadense created from fiber cells and seed at O and 5 DPA and from fiber cells at 10, 15, 20, 25, 30 and 40 DPA was extracted, the concentration was equalized and a PCR amplification was performed using primers SE002 (SEQ ID NO: 35) and SE003 (SEQ ID NO: 36). The PCR reaction mix used contained: 1 μl template DNA (200ng/μl), 5 μl 5x GreenGoTaq buffer, 0.75 μl SE002 (lOμM), 0.75 μl SE003 (lOμM), 0.5 μl dNTP's (20 mM), 0.25 μl GoTaq polymerase, 16.75 μl MiIIiQ water (total of 25μl). The PCR conditions used were as follows: 5 min at 95°C; 5 times: 1 min at 95°C, 1 min at 58 ⁰ C, 2 min at 72°C; 25 times: 30 s at 92°C, 30 s at 58°C, 1 min at 72°C; 10 min at 72°C, cooldown to 4°C. The expected length of the PCR product is 655 bp. After PCR amplification, the PCR fragment is digested with AIwI digest (3h incubation at 37°C) using 10 μl template; 1 μl AIwI enzyme; 2 μl NEB 4 restriction buffer; 7 μl MQ water. The resulting fragments are analysed on 1.5% TAE gel stained with EtBr. The expected band sizes for the A subgenome allele specific PCR fragment are : 479 bp, 118 bp and 59 bp. The expected band sizes for the D subgenome allele specific PCR fragment are: 538 bp and 118 bp.

[359] Figure 8, lanes 2 to 9, represent GbGLUCJ. IA and D expression at 0, 5, 10, 15, 20, 25, 30 and 40 DPA. Differences in band intensities in Figure 8 correspond to relative differences in expression. A negative (no template; NTC; Figure 8, lane 10) and a positive control (genomic DNA from Pima S7; Figure 8, lane 11) were included. The expression profile of the GhGLUCl. IA and D and GbGLUCl. IA and D genes can be summarized as follows:

[360] Thus while the expression of GLUCl. IA in G. hirsutum starts only at 30 DPA, GLUCl. IA in G. barbadense is expressed from 15 DPA on. However, as indicated above, the GbGLUCl. IA gene is predicted to encode a non- functional GLUC 1.1 A protein.