The present invention relates to a transgenic expression construct for plants, comprising an anti-apoptotic gene of the BAG family of plant genes and a promoter overexpressing said gene in a plant or part thereof. Respective transgenic plants exhibit at least one yield gain phenotype when compared to a non-transgenic plant. The present invention further relates to a method for producing a transgenic plant or part thereof, comprising introducing the transgenic expression construct of the invention into said plant or part thereof, and respective transgenic plants or parts thereof as produced.

Background of the invention

Yield in crops and in particular in agricultural crops is a complex multifactorial trait reflecting the interaction of the environment with all growth and development processes. Positive selection over long periods of time has led to a yield increase as a result of genomic changes that impact yield component traits like seed weight and seed number and yield related traits (for example biomass, plant architecture, adaptation and stress resistance).

By 2050, the world population is expected to reach nearly 10 billion. To sustainably provide sufficient food for this increasing population, crop productivity needs to increase by more than 40% compared to the current level. This is a challenge because there is limited potential for future expansion of arable lands whilst at the same time climate predictions suggest that a larger portion of the globe will be subject to erratic environmental conditions.

To date, strategies to increase yield in crops have utilized three major approaches, namely conventional breeding, marker assisted selection and genetic engineering. Of these, genetic engineering promises a great potential since it allows the introduction of selected genes without affecting the desirable characteristics of an elite genotype.

Genetic engineering has been used to improve the resistance of crops against environmental stresses like drought or salinity. Some success has been achieved to reduce the losses caused by these environmental stress factors, which contribute to a major part to crop yield losses worldwide.

While improving the tolerance against environmental stresses may help to reduce the crop yield losses, the solutions found today in this regard do not normally lead to a yield increase compared to the wild type in the absence of such external stress factors.

Programmed cell death (PCD) is essential for all multicellular organisms. The decision of whether a given cell should live or die is the result of a competition between anti-apoptopic and pro-apoptopic (pro-death) proteins and signal transduction pathways.

Compared to the vast molecular knowledge base available for PCD (frequently referred to as apoptosis) in animal or mammalian cells, much less data are available for plant PCD process. It remains undisputed that genetic programs leading to cellular suicide do exist in plants, being involved in resistance to disease and developmental processes. With respect to the latter, PCD pathways are known to play a role, for instance, in the male and female gametophyte development, during incompatible pollen-pistil interactions, during seed development and germination, seed coat and endosperm development, the formation of tracheary elements, aerenchyma formation and lateral and adventitious root emergence, root cap development, leaf abscission processes and senescence.

Plants often respond to pathogen attack by executing PCD programs such as the hypersensitive (HR) response. For instance, plant resistance to biotrophic pathogens are generally a result of the pathogen triggering cell death. Hemi-biotrophs and necrotrophs take over the plant HR response, thus reprogramming plant pathways, tricking the plant to undergo PCD that provides a nutritional benenfit entirely for the pathogen.

Programmed cell death in plants appears distinct from animal apoptosis; plant homologs of mammalian regulators of apoptosis are scarce using standard search algorithms. For example no sequence specific caspase homologs can be found in plant genomes; yet enzyme activities have been demonstrated. Further, there are plant proteases that mediate similar phenomena, with a similar conceptual framework but are not true caspases. Importantly, the ectopic expression of several animal anti-apoptotic genes in transgenic plants is known to confer protection from pathogens and other stresses as described by (Dickman et al, 2001; Lincoln et al., 2002). It is therefore conceivable that the divergent evolution of PCD molecular constituents has led to mechanisms in plants that may be functionally similar to animal genes and pathways independent of primary sequence. However, these cannot be easily deciphered in the absence of searchable sufficiently homologous genes.

Arabidopsis thaliana (At) AtBAG4, Hsp70 and p35 are anti-apoptotic genes that have been reported to confer tolerance to salinity and drought stresses in transgenic tobacco.

The mammalian BAG (Bcl-2-associated athanogene) family of proteins (7 homologs in humans) promote cell survival acting synergistically with Bcl-2. Various additional functions have been ascribed to BAG proteins including apoptosis, tumorigenesis, neuronal differentiation, stress response and cell cycle regulation. BAG proteins are characterized by a common C-terminally localized BAG domain mediating the interaction with Hsp70/Hsc70 molecular chaperones. Additional domains mediate the interaction with the proteasome system, the localization within cells and the interaction with transcription factors (Kabbage and Dickman, 2008).

Heat shock proteins comprise a set of powerful chaperones that are expressed in response to a variety of physiological and environmental stresses and are localized in a variety of subcellular compartments. By protein-protein interaction Hsp70 proteins can facilitate anti- apoptotic Bcl-2 proteins to inhibit apoptosis pathways at distinct key points (Joly et al, 2010). The broad-spectrum activity of Hsp70s requires the recruitment of co-chaperones and other chaperone systems (Brodsky and Bracher, 2007). Overexpression of Hsp70 in tobacco conferred tolerance to drought stress (Cho and Hong, 2006).

The function of p35, a Baculovirus anti-apoptotic protein has been extensively studied in different organisms (Clem et al, 1991; Clem and Miller, 1994; Bump et al, 1995; Xue and Horvitz, 1995; Bertin et al, 1996). Expression of p35 in tobacco, tomato and passion fruit significantly improved tolerance to abiotic and biotic stresses (Lincoln et al, 2002; Wang et al, 2009). Transgenic tobacco expressing p35 displayed broad-spectrum tolerance to a range of abiotic stresses including salinity, upon further investigation this tolerance was attributed to the ability to sequester ROS (Wang et al, 2009).

A moderate low- level expression of the AtBAG-4 gene was shown to confer protection to plants from a variety of abiotic stresses, by inhibiting programmed cell death, while knockout plants were very susceptible to various stresses including drought tolerance, increased salt stress, increased UV-B tolerance, cold treatment and oxidant exposure. In contrast to AtBAGl,5, 7, AtBAG6 T-DNA insertion mutants are extremely susceptible to the necro trophic fungus B. cinerea to which it is normally resistant, suggesting that AtBAG6 has a role in basal resistance limiting necrotrophic pathogen colonization (Kabbage and Dickman, 2008).

Hoang et al. (2015) describe the development of salinity tolerance in rice by constitutive over- expression of genses involved in the regulation of programed cell death. The authors show that the exogenous expression of anti-apoptopic genes from diverse origins, including the AtBAG4 gene of Arabidopsis improves salinity tolerance in rice at the wholeplant level. However, this is related to the avoidance of losses, not to a net yield gain that surpasses standard yield potentials.

In fact, until today no way to improve yield in crops or plants in the absence of environmental stress factors and leading to an improvement in the overall harvest yield compared to the wild-type under the same condition has not been found and still remains a need to be solved.

It is therefore an object of the present invention to provide a process for the increase of yield in crops, in partcilular in the absence of environmental stresses, leading to a yield exceeding the yield in wild-type plants. Other objects of the present invention will become apparent to the person of skill when reading the following specification, examples, claims and figures of the present invention.

According to a first aspect of the present invention, this object is solved by a transgenic expression construct for plants, comprising an anti-apoptotic gene of the BAG family of plant genes comprising at least one BAG domain, or a plant homolog of said BAG gene, and a promoter overexpressing said gene in a plant or part thereof, when compared to the expression the wild-type promoter of said anti-apoptotic gene. Preferred embodiments of the present invention are set forth in the dependent claims and the detailed specification hereinafter.

It was surprisingly found that when an anti-apoptotic gene of the BAG family of plant genes was overexpressed in a plant or part thereof, a substantial increase in yield could be achieved. Their overexpression seems to delay PCD in (e.g.) endosperm cells and has additional effects leading to increased yield potential in monocot (e.g. Oryza sativa) and dicot (Q.g.Nicotiana tabacum, Arabidopsis thaliana) plants, upon overexpression. This, for example, affects the grain size for assimilates, but also inflorescence and panicle architecture leading to increased seed numbers. In addition, constitutive expression leads to larger leaves and thicker stems and thus to an increase in overall bio mass.

In the context of the present invention, the anti-apoptotic gene of the BAG family of plant genes is characterized in that it comprises at least one BAG domain as described, for example, above. BAG domains are known to functionally mediate the interaction with molecular chaperones, in particular Hsp70/Hsc70. Thus, an "anti-apoptotic gene of the BAG family" shall mean a gene, preferably a plant gene, the product of which a) promotes cell survival (i.e. is anti-apoptotic), and b) comprises a (at least one) BAG domain binding to chaperones, for example Hsp70 and/or 90. The BAG domain can be N- or C-terminally localized. The definition also includes protein complexes, wherein the proteins as involved provide at least one of these "BAG family" functions when present in said complex. Optionally, the BAG- gene can also comprise at least one Ubiquitin- like- domain (see Doukhanina et al., 2006b).

Preferred is the transgenic expression construct according to the present invention, wherein said anti-apoptotic gene is selected from a gene from Arabidopsis thaliana (At)- or Oryza sativa (Os)- or homologs thereof. Seven BAG homologs were identified in Arabidopsis, four of which have domain organization similar to animal members. The remaining three members contain a predicted calmodulin-binding motif near the BAG domain, a feature unique to plant Bcl-2-associated athanogene (BAG) proteins (Doukhanina et al, 2006).

The following Table 1 shows preferred BAG proteins for use in the present invention.

Name/SEQ ID NCBI accession no mass length Localization

NO.

AtBAGl/1 NP 200019 36.4 326 Cytoplasm

AtBAG2/2 BAB10172 25.5 232 Microbody

AtBAG3/3 NP 196339 34.2 303 Cytoplasm

AtBAG4/4 AAC 14405 29.2 269 Cytoplasm

AtBAG5/5 AAC 17606 24.5 215 Mitochondrion

AtBAG6/6 AAC62882 116.8 1043 Nucleus AtBAG7/7 NP_201045 51.6 446 Nucleus

OsBAG4/8 NP 001044700.1 262

Other BAG genes that are useful in the present invention are found, for example, in Raphanus raphanistrum, Brassica napus, Brassica oleracea, Raphanus sativus, Brassica rapa, Triticum aestivum, Gossypium hirsutum, Vitis vinifera, Quercus robur, Nicotiana tabacum, Solanum tuberosum, Solanum lycopersicum, Solanum melongena, and Coffea canephora.

A "homolog" of the BAG gene according to the present invention shall mean a gene that encodes for a protein that exhibits at least about 60%, preferably at least about 70%, and more preferred at least about 80% or at least about 90% and even more preferred at least about 95% sequence identity to a B AG-protein and preferably to the/a B AG-domain thereof as described herein, and in particular as above. Of course, this protein also has to be functional with respect to anti-apoptosis and BAG domain, as explained above.

In the context of the present invention, the term "about" shall generally mean a variability of +/- 10%) of the value as indicated.

In the context of the present invention, a gene is overexpressed by a promoter, particularly in a plant or part thereof, if as a result of said expression an altered spatial distribution and/or an increased quantity of the gene product is found, compared to the expression of the wild-type (regular) promoter for this gene in, for example, a non-transgenic plant or part thereof. Preferred levels of overexpression lead to a more than 2-fold, more than 5-fold, more than 10- fold or even more than 20-fold higher expression of the gene, when compared to the expression of the wild-type (regular) promoter for this gene. Overexpression can be measured using well-known methdos in the art, such as mRNA detection (e.g. using rtPCR), promoter activity assays, and Western blotting (immunoblot) analysis.

In the context of the present invention, it is important to distinguish yield potential and potential yield. Yield potential is defined as the yield of a cultivar when grown in environments to which it is adapted, with nutrients and water non-limiting and with pests, diseases, weeds, lodging, and other stresses effectively controlled. As such, it is distinguished from potential yield, which is defined as the maximum yield which could be reached by a crop in given environments, as determined, for example, by simulation models with plausible physiological and agronomic assumptions (see, for example, Evans, L. T., and R. A. Fischer. 1999. Yield Potential: Its Definition, Measurement, and Significance. Crop Sci. 39: 1544-1551).

Hoang et al. (2015) describe that exogenous expression of anti-apoptotic genes from diverse origins, AtBAG4 (Arabidopsis) , Hsp70 (Citrus tristeza virus) and p35 (Baculovirus), significantly improved salinity tolerance in rice at the whole plant level. They demonstrated traits associated with stress-tolerant transgenic events including, improved photosynthesis, membrane integrity, ion and ROS maintenance systems, growth rate, and yield components. Anti- apoptotic genes are said to facilitate maintenance of metabolic activity at the whole plant level to create favorable conditions for cellular survival under salt stress. AtBAG4 from A. thaliana, Hsp70 from Citrus tristeza virus and p35 from Baculovirus driven by the maize polyubiqui- tin-1 (Ubi-1) promoter were cloned into the binary vector pCAMBIA1301.

Hoang et al. (2015) disclose a yield gain only in the well-known context of protection from salinity stress. Under regular conditions, no improved effect was found when expression the gene (see suppl. data, in particular Figs. 3 and 4 at seedling and reproductive stage, respectively).

For the transgenic expression construct of the present invention, any suitable ubiquitous or tissue-specific plant promoter can be used, as long as the promoter causes an overexpression of the gene. Respective promoters are known to the person of skill, and described in the state of the art (e.g. Dutt et al., 2014). Preferred is a transgenic expression construct according to the present invention, wherein said promoter overexpressing said gene is selected from ubiquitous and tissue-specific plant promoters, such as BSV, CaMV 35S, TA29, actin, FBP7, CMPS, Lhca3.St.l, Mannopin synthase, RolD, Uepl, Ubiquitin and NapA, Gtl (synonymously termed GluA-2), 2S, CuMFTl, Dc3, HaG3-A, LeB4, LegA, Phas, Psl, Str, and USP. For instance, to express a gene only in grains and not in leaves, a seed-specific promoter can be used. The codons of the gene can be optimized for the organism in order to overcome codon usage bias.

Yet another aspect of the present invention then relates to a method for producing a transgenic plant or part thereof, comprising introducing the transgenic expression construct according to the present invention as described above into said plant or part thereof. Methods for making transgenic plants or parts thereof are known to the person of skill, and may involve, for example, transfer with Agrobacterium tumefaciens, the "gene gun" technology, electroporation or microinjection.

Consequently, another aspect of the present invention relates to a transgenic plant or part thereof, produced according to the method according to the present invention. Said transgenic plant or part thereof according to the present invention, preferably overexpresses an anti- apoptotic gene of the BAG family of plant genes comprising at least one BAG domain as described above, or a homo log of said BAG gene, such as, for example AtBAG4 or OsBAG4 as described above. Most preferred are transgenic plants comprising Gt-1 promoter-driven At- BAG4 or OsBAG4.

In general, the transgenic plant or part thereof according to the present invention can be any monocot or dicot plant, and in particular is a plant used in agriculture, e.g. a crop plant. Thus, preferably the transgenic plant is selected from Arabidopsis, and crop plants, such as wheat, rice, maize, barley, rye, sorghum, soy, sugar cane, pumpkin, tobacco, coffee, wine, broccoli, cauliflower, cassava, potato, and cotton.

Further preferred is the transgenic plant or part thereof according to the present invention, wherein said overexpression is non-ubiquitous (i.e. not in every part or tissue of the plant, or in substantially all parts or tissues of the plant), such as, for example, tissue-specific, e.g. specific for meristematic tissues, fast growing tissues-specific, endosperm-specific, leaf-, root-, stem-, flower-, and/or fruit-specific, or specific for certain plant cell types.

It was surprisingly found that the transgenic plants or parts thereof according to the present invention exhibit at least one yield gain phenotype, when compared to a non-transgenic plant, preferably a plant of the same variety or species without the transgene, or even without any transgene. A yield gain phenotype in the context of the present invention shall mean an increase of the mass per plant, when compared to a non-transgenic plant, also leading to an increase of mass, or grains or seeds, generated from a unit of land expressed as kilograms per hectare. According to the invention, sub-aspects of the yield gain can be selected from larger and/or longer ears or panicles, a tissue specific yield gain, a yield gain under no stress conditions, larger seeds with increased weight, seeds having an increased number of endosperm cells, and larger seed hulls having an increased number of cells. It was furthermore found that, when employing the present invention, substantial yield gains, e.g. as defined above, could be achieved in the glasshouse. In general, yield gains of a few percentages already result in quite substantial gains. A yield gain according to the present invention is selected from more than about 5% increase in seed mass, more than about 5% increased seed weight per plant, more than about 5% increased seed number, and more than about 5% increase in overall plant biomass such as, for example, more than about 5% of bio- mass in rice or tobacco plants in the field. In some instances, a yield gain can be about 5%- 10% increase in seed mass per plant, and about 5%-15%, preferably about 10%>, increase in overall plant biomass, or even more. As can be seen in the examples below, a grain yield gain per plant of up to 60% (rice) or a yield in overall biomass (dry mass) of up to 43 % (tobacco) as found in the glasshouse can be achieved.

In more detail, preferred transgenic constructs, such as AtBag4 overexpression in rice using the Gtl promoter (endosperm- and meristematic- and fast growing tissues-specific promoter), showed 10%) longer seeds (due to increased endosperm cell number, longitudinal), and 10%> higher mass. The additional yield parameters such as increased panicle length accommodating an increased number of seeds per plant contribute substantially to the strongly increased grain yield noted in Figure 17. Similar results, albeit at lower levels of increase were found upon OsBag4 overexpression in rice using the Gtl promoter (endosperm- and meristematic- and fast growing tissues-specific promoter). AtBag4 overexpression in tobacco using the ubiquitous 35S promoter showed about 18% increased seed weight caused by larger seeds and in the greenhouse a faster growth and the development of much larger leaves and thicker stems resulting in the dry mass increase noted above. AtBag4 overexpression in Arabidopsis thaliana using the ubiquitous 35 S promoter showed about 25% increased seed weight. These constructs thus serve as proof of concept for the overall invention, and in particular for the alternative successful embodiments (which could be combined as well) of using endosperm- and meristematic- and fast growing tissues-specific promoter or a ubiquitous promoter.

Yet another aspect of the present invention then relates to a method for identifying a yield gain plant or part thereof, comprising identifying the overexpression and/or a modification of an anti-apoptotic gene as described herein selected from genes of the BAG family of plant genes comprising at least one BAG domain, or a plant homolog of said BAG gene in said plant or part thereof, compared to a plant not overexpressing said anti-apoptotic gene and/or comprising said modification, and identifying said yield gain plant or part thereof based on said overexpression. Preferred is the method according to the present invention which further comprises identifying the underlying marker for said overexpression yield gain and/or identifying said modification, and preferably further isolating said marker.

This aspect of the present invention allows to select for and identify yet unknown favorable alleles (mutants) of genes of the BAG family of plant genes comprising at least one BAG domain, or a plant homolog of said BAG gene(s) that lead to a yield gain phenotype as described herein. The yield gain preferably is the result of an overexpression of said gene, but can also be based on functionally improved modifications, such as mutations or gene duplications of said BAG gene(s). The plants used for this method can either be obtained from spontaneously mutated plants, but the method can also comprise a mutagenesis step (for example using gene editing, such as, for example introduction of mutations by Crispr-Cas or TALEN), and then a screening for respective plants and/or markers. Consequently, another aspect of the present invention relates to a plant or part thereof as identified according to the method according to the present invention.

As above, a yield gain phenotype in the context of the present invention shall mean an increase of the mass of a plant or parts thereof, when compared to a non-transgenic plant, including but not restricted to mass of grains or seeds, leaves, stems, roots, tubers and fruit generated from a unit of land expressed as kilograms per hectare. According to the invention, sub-aspects of the yield gain can be selected from larger and/or longer ears or panicles, a tissue specific yield gain, a yield gain under no stress conditions, larger seeds with increased weight, seeds having an increased number of endosperm cells, and larger seed hulls having an increased number of cells. As above, a yield gain according to the present invention is selected from more than about 5% increase in seed mass, more than about 5% increased seed weight per plant, more than about 5% increased seed number, and more than about 5% increase in overall plant biomass such as, for example, more than about 5% of biomass in rice or tobacco plants in the field. In some instances, a yield gain can be about 5%-10% increase in seed mass per plant, and about 5%-15%, preferably about 10%, increase in overall plant biomass, or even more.

The inventors studied PCD-related processes occurring in the developing rice grain and during panicle development as a pathway that has the potential of increasing yield, when down- regulated. They showed that a player of canonical PCD, the anti-apoptotic AtBAG4 gene, delays PCD in endosperm cells and has additional effects leading to increased yield potential in monocot (Oryza sativa) and dicot (Nicotiana tabacum, Arabidopsis thaliana) plants, upon overexpression. This affects the grain sink size for assimilates, but also inflorescence and panicle architecture leading to increased seed numbers.

Further advantages, properties, and features of the present invention become apparent by the following description of preferred examples of the present invention with reference to the enclosed Figures, nevertheless, without being limited thereto. For the purposes of the invention, all references as cited herein are incorporated by reference in their entireties.

Figure 1 shows that rice seed length correlates with seed weight. Independent T2 events were preselected an analyzed in the T3 seed generation. Black circles, control; green triangles, HvSOD events; dark blue squares AtBAG4 events; light blue squares, OsBAG4 events. Individual data points represent mean values (n=100).

Figure 2 shows Southern blots of preselected events. Transformation events in the T2 generation were preselected by segregation analysis for single integration and EcoRI-digested DNA of the resulting leaf material was probed. A, OsBAG4 events; B, AtBAG4 events. Lanes 1,2 and 8,9 represent the linearized transformation vector and a 0.8 kb PCR amplicon, respectively, as positive controls. All other lanes represent independent transformation events, among which there are the singly integrated AtBAG4 events s #2, #3 and #4 (lanes 11, 12, 13, respectively). Two singly integrated OsBAG4 events are shown in A (lanes 4, 5); lane 7 and 16 represent the control, and the former also hybridizing with the internal OsBAG 4 gene.

Figure 3 shows seed length and width distribution in selected rice events. A, length distribution of control seeds and of the T5 generation of AtBAG4 and HvSOD overexpressing seeds. The event numbers are indicated. (n= > 1300 for every data set composed of randomly selected grains of 14 siblings randomly planted in the greenhouse. Events #2,3,4 AtBAG; event #30, HvSOD. The data are based on a 3D analysis using a Microtrac Dynamic Image Analyzer instrument. The curves were fitted with the Gauss equation R ² = 0.94-0.97; width R ² = 0.98 - 0.99. Figure 4 shows seed metrics of Control and 35S::AtBAG4 overexpressing tobacco seeds. Seeds from 6 siblings of the control (black line) and 5 plants of 35S::AtBAG4 overexpressing events (green line) were measured using Microtrac Dynamic Image Analyzer instrument. The curves were fitted with the Gauss equation. A, seed diameter (R ² = 0.98 control; R ² = 0.99 AtBAG4 events); B, seed volume (R ² = 0.98 control and AtBAG4 events), n = > 1100 seeds (mixed from the different control and overexpressing plants).

Figure 5 shows that larger seeds are caused by an increased endosperm cell number. T3 seeds of homozygous events were sectioned and morphological parameters determined along the axes indicated. A, Longitudinal sections. 1, Longitudinal cell number; 2, longitudinal cell diameter; 3, lateral cell number; lateral cell diameter. B, Cross sections; dorsal (d), ventral (v) and lateral (1) axes were defined to account for the anisodimetric shape. 1, Dorsal cell number; 2, dorsal cell diameter; 3, ventral cell number; 4, ventral cell diameter; 5, lateral cell number; 6, lateral cell diameter. Measurements were made using the ImageJ software; n = > 6 per event.

Figure 6 shows that transgenic events have a lower proportion of TUNEL-positive nuclei. Isolated rice endosperm nuclei derived from spikelets, individually tagged at the time of flowering were double stained with DAPI (A) and with TUNEL (B) allowing to assess the proportion of apoptotic nuclei. The numeric values determined are shown in C. black line, untrans- formed control; blue line, event #30 expressing HvSOD; green line, event # 1 expressing At- BAG4; orange lines events #9 and #10 expressing OsBAG4. In parallel aliquots, the number of nuclei was determined based on Feulgen staining. (D), number of nuclei determined independently after Feulgen staining. Each data point represents the average (± SE) of 11 replicates of 4 seeds pools. DAF, days after flowering; bar represents 10 μιη.

Figure 7 shows vitality stain of seed cross sections. Seeds derived from individually tagged spikelets were harvested at 16 DAF and stained with FDA (upper panel) and JC-1 (lower panel). Pictures were taken under standardized photographic conditions. Seeds expressing At- BAG4 and OsBAG4 showed consistently a more intense fluorescence compared to control seeds (WT).

Figure 8 shows that spikelets and seed grains are equivalently increased in size. Examples of T3 seeds are shown (A) as paddy, (B) after removal of the seed coat and (C ) spikelets (before fertilization). An alignment of ten corresponding spikelets is shown in (D), showing a difference of one spikelet (= 10%). The transgenic events used are indicated.

Figure 9 shows cryo-SEM of rice spikelets. Spikelets were taken at random from the middle part s of panicles and photographed. Arrow points to the 5 ^th discernible row used for cell counting. A, HvSOD #22; B, OsBAG4#33; C, Control.

Figure 10 is based on cryo-SEM and reveals increased cell numbers in the hull. Spikelets were taken at random from the middle parts of panicles and the number and size of cells determined in the 5 ^th row of the lemma, as shown in the inset (arrow), in Figure 9. 1, Longitudinal cell number; 2, seed length; 3. cell length; 4 cell width. Data represent the mean (± SE) of three spikelets per event.

Figure 11 shows Gtl promoter activity in developing spikelets. The 1.85 kb version of the Gtl promoter used to express HvSOD and BAG4 was used to express GUS. A discrete spatial pattern was observed in the basal growth zones of the palea (1) and lemna (2, 9) and in the outgrowth zone of the awn (3). Activity was also sometimes noted in the ovary (6). A, dissected spikelet; B, outgrowing awn; D, overview of a pre-emergent panicle. 5, stamen; 7, lod- icules 8, rachilla. Bar refers to D and represents 5 mm.

Figure 12 shows Gtl promoter activity in developing panicles. The 1.85 kb version of the Gtl promoter used to express HvSOD and BAG4 was used to express GUS. A, six developmental stages of pre-emergent panicles. The spotted spatial expression profile shown in Figure 11 is observed from stage 3-5. When grown to almost full size (stage 6), spikelet hulls show the uniformous GUS stain, shown in B. Bars represents 5 mm (A) and 10 mm (B).

Figure 13 shows Gtl : reporter gene expression in dividing cells. A, GUS signal obtained in regenerating calli. B, YFP expression in root tips of regenerating plantlets.

Figure 14 shows Gtl::YFV reporter gene expression in inflorescence primordia. Developing inflorescences are ordered according to size to reflect progressing developmental stages. Bar represents 200 μιη (a-c) and 500 μιη (d). The cellular expression in primordia substructures leading to panicle side branches and spikelets can be resolved by confocal microscopy (e, f; bar represents 50 μηι. The nuclar localization is caused by a nuclear localization sequence trans lationally fused to GFP.

Figure 15 shows Gtl : reporter gene expression in vascular tissues, (a) GUS-stained, stem cross section at the 2 ^nd internode showing signals in the (b) large vascular bundles and (c), epidermal vascular tissues, (d) YFP-signals showing vascular tissue staining in the (d) collar and (e) mature leaf.

Figure 16 shows Gtl : reporter gene expression in rice seeds. Longitudinal (a) and transversal (c) section of a mature rice seed at 25 DAF; (b) longitudinal section of a late milky-stage seed at 10 DAF.

Figure 17 shows that several parameters contribute to increased yield per plant. Radar plot, in which the parameter changes given in Table 3 are displayed on a per cent scale (Control = 100%). All parameters displayed deviating from equilateral triangle contribute to the increased yield (gram seeds per plant; black line).

Figure 18 shows that transgenic lines show longer panicle accommodating a higher number of seeds. 1, Transgenic line overexpressing AtBAG4 (event #2); 2, control.

Figure 19 shows that _^ 4t L4 G4-overexpressing plants exhibit increased leaf size in young (A) and adult (B) plants. The general leaf shape remains unchanged.

Figure 20 shows that _^ 4t L4 G4-overexpressing plants have thicker stems. A, aspect of stems at the same reproduction scale taken at the 11 ^th leaf position. B, macroscopic view of cross section (at the 11 ^th leaf position). The white double arrow indicates the diameter and representing 16 mm (wild-type) and 22 mm (AtBAG4 overexpressor); a, pith (parenchyma); b, cortex (chlorenchyma). C, microscopic view of stem sections. Bar represents 1 mm.

Figure 21 shows that cell proliferation is promoted by. A, young leaves of the AtBAG4- overexpressing plants are darker green and show a strongly proliferating phenotype characterized by the increased call mass between the leaf veins. This leads to a strongly "spinach-like" appearance which disappears when leaves grow larger. B, protoplasts were generated from in- vitro grown wild-type and _^ 4t ^4 G4-overexpressing tobacco plants Their concentration was adjusted to 2 x 10 ⁴ cells ml ^"1 PNT medium and allowed to grow under a dim light/dark cycle at 27°C for seven days. The AtBAG4 overexpressing cells show much faster cell proliferation.

Figure 22 shows the increased longevity of _^ 4t L4 G4-overexpressing tobacco plants. A, wild- type plants show senescence during growth, starting at the bottom at the plant, while overexpressing plants (possessing larger leaves) do not. B, delayed senescence is caused by a loss of ethylene sensing. Leaves (with no signs of senescence) at the indicated defined middle positions of plants of the same age were wrapped on the plant into plastic and exposed for 1.5 days to ethylene (generated from 2 ml of a 50 mM Ethephon solution). While wild-type leaves responded strongly, no signs of senescence were apparent in the AtBAG4- overexpressing leaves.

SEQ ID No. 1 to SEQ ID No. 8 show amino acid sequences of preferred i^4G-genes (see table 1).

Examples

Abbreviations: CTAB, Hexadecyl trimethyl-ammonium bromide; DAF, days after flowering; DAPI, 4 ^' ,6-Diamidino-2-phenylindole; FDA, Fluorescein diacetate; JC-1,5,5 ^' ,6,6 ^' - tetrachloro-l, ,3,3 ^' -tetraethylbenzimidazol-carbocyanine iodide; TUNEL, Terminal deox- ynucleotidyl transferase-mediated dUTP nick end labeling.

Material and methods

Chemicals used

Phusion™ High-Fidelity DNA Polymerase was a product of Finnzymes. Nylon membranes (positively charged), DIG High Prime DNA Labeling and Detection Starter Kit II were from Roche. The following chemicals were purchased from Sigma- Aldrich: JC-1, DAPI, FDA, chemicals for Gibson cloning, cellulose from Trichoderma viride, alpha-amylase from Bacillus licheniformis, Feulgen reagent and toluidine blue.

DNA constructs

HvSOD (Database Acc. No. AK248365.1) lacking its predicted 40-amino-acid transit peptide (ChloroP 1.1 software), individually or in the combination with a 5 '-fusion transit peptide of Pea ribulose- 1.5-bisphospatc carboxylase/oxygenase small subunit (SSU), were synthesized (GenScript), respectively. Both were equipped with a 3 ^' - BamRl and a 5 ^' HmdIII site. The excised fragments were ligated into an accordingly digested pCambial380 -derived binary vector harboring the hptll selectable marker gene and the 1.8 kb rice endosperm- specific glu- telin Gtl (syn. GluA-2) promoter (Okita et al, 1989; Wu et al, 1998). This yielded pC1380- Gtl-SSU-SOD and pC1380-Gtl-SOD.

AtBAG4 was amplified using Phusion™ High-Fidelity DNA Polymerase and suitable primers. The purified, Small Sa/I-digested PCR product was inserted into the same binary vector. The OsBAG4 fragment was obtained by BamRl digestion from the original vector (pYC30- OsBAG4) and ligated into the BamRl digested and dephosphorated backbone of the same binary vector. The resulting vectors, pC1380-Gtl-AtBAG4 and pC1380-Gtl-OsBAG4, were used for Agrobacterium transformation.

To determine the tissue-specificity of the 1.85 kb Gtl promoter used, the vectors pCA1380- Gtl-NLS-3vYFP and pCAl 380-Gtl-NLS-GUS were constructed using the same binary vector. The source vectors containing NLS-3vYFP (pMU006) and NLS-GUS (pNK186) were kindly provided by T. Laux (University of Freiburg). The NLS-3vYFP fragment was obtained by BamRl/ BgUl digestion and ligated into the BamRl digested and dephosphorated backbone of the same binary vector. The NLS-GUS fragment was obtained by PCR amplification using pNK186 as template and suitable primers. The PCR product was mixed with BamRl I Spel digested backbone fragment and the ligation was done by Gibson cloning (Gibson et al, 2009). All vectors were verified by sequencing.

Rice transformation and plant propagation

Embryogenic calli derived from immature embryos of an experimental Golden Rice 2 (GR 2 (Paine et al., 2005; Ye et al., 2000) event in an Oryza sativa var. japonica (cv Kaybonnet) background (hereafter termed "control") were used for the co-cultivation with Agrobacteria, hygromycin selection and plant regeneration as described (Ilg et al, 2010). After 3-4 rounds of hygromycin selection, resistant calli were subjected to regeneration. Plantlets were transferred to rooting medium for 2 weeks, and then to Yoshida solution (Yoshida et al, 1976) for 2 weeks at 28 °C under 16h/8h light-dark cycle, and transplanted to soil with a spacing of 15x 10 cm for greenhouse growth under a 16h/8h light/dark cycle at 31°C/26 °C. A number of 45-70 independent events were generated for each construct. Preliminary selection for single integration was done by germinating the segregating Tl seed generation in the presence of hygromycin (80 mg Γ ¹ ). Events with a 3: 1 germination ratio were carried into the T2 seed generation, which were subjected to the same selection. Those showing no segregation were assumed to be homozygous. Southern-blot analysis was done to verify the copy number of the insertion. Genomic DNA (20 μg) were digested with EcoRI, a unique restriction site localized adjacent to the 5' end of the Gtl promoter. The probes hybridizing with OsBAG4 or AtBAG4 were labeled, and the blotting carried out according to the manufacturer ^' s instructions (DIG High Prime DNA Labeling and Detection Starter Kit II, Roche). The corresponding linearized binary vector plasmids (100 pg/lane) and cDNA fragments generated by PCR or restriction digestion (20 pg/lane) were used as positive controls.

Azygous seeds were isolated from the segregating Tl plant generation using a leaf assay for hygromycin sensitivity, essentially according to (Roy et al, 2012). An optimized concentration of 120 μg ml ^"1 dissolved in water was applied to 2-3 cm leaf segments of 1-month-old transgenic and control plants and kept in Petri dishes at RT under which condition bleaching of azygous plants became apparent after 5 d. The bleached segments were tracked back to the corresponding plants.

2L4G -expressing Arabidopsis thaliana and Nicotiana tabacum plants

The transformation events used expressing AtBAG4 (Acc. At3g51780; NM 115037,7) under constitutive cauliflower mosaic virus 35S (35S CaMV) promoter control (Doukhanina et al, 2006). Tobacco plants (cv Glurk) were transformed with the vector pEarlygatel04-AtBag4, resulting in the expression of an N-terminal translational fusion YFP with AtBAG4. The data presented relate to a random mix of four integration events.

Detection of Programed Cell Death

Flowers located at the 3rd to 6th position from the apex of the upper 3 primary branches on the panicle were individually tagged at their flowering date. Freshly harvested seeds at 10, 13, 16, 19 and 22 days after flowering (DAF) from at least 5 plants for each time point were manually dehulled. For the isolation of nuclei, the embryos were excised and the endosperms cut into small pieces, followed by fixation in acetic/ethanol (25 : 75) over night and then re- hydrated with an ethanol/water graded series (75%, 50%, 25%, 0% ethanol) . The endosperm samples were kept in a water bath at 60°C for 10 min and, after rinsing with water, digested over night with 0,5% (w/w) cellulase (from Trichoderma viride (Sigma) and 1250 U ml ^"1 of Type VII a-Amylase (from Bacillus licheniformis, Sigma) at 35°C in 1 ml/ 4 grains of Mcllvaine's citrate-phosphate buffer at pH 5.0. The samples were macerated by gently press- ing the pellet with a glass rod. After washing with IX PBS and centrifugation, the pellets were weighted and resuspended inl X PBS to reach a final concentration of 0.1 g ml ^"1 . A 200 μΐ aliquot was used for nuclei staining using the Feulgen method as described (Singh and Jenner, 1982). Another 200 μΐ aliquot was used first for TUNEL staining using the In Situ Cell Death Detection Kit POD (Roche Diagnostics) according to the manufacturer ^' s instruction, followed by centrifugation and resuspension of the pellet in PBS and treatment with a 0.6 μΜ DAPI solution. This double staining was used to estimate the ratio of viable and apop- totic nuclei. The number of nuclei was determined using a counting chamber with an Axi- oskop 40 (Zeiss) microscope equipped with corresponding florescence filters. The isolation of nuclei proved to be necessary because TUNEL-detection is not possible in rice endosperm sections (Kobayashi et al, 2013).

The viability of endosperm cells was assessed in 16 DAF seeds with the viability stains fluorescein diacetate (FDA, Sigma- Aldrich) and 5,5 ',6,6'-tetrachloro-l,10,3,30-tetraethyl- benzimidazolo carbocyanine iodide (JC-1, Sigma- Aldrich) as described by (Kobayashi et al, 2013). Fresh seeds were dehulled with forceps, and cut into halves submerged in H ₂ 0 with a razor blade. Sections were treated with 2 mg Γ ¹ FDA in PBS at 25°C for 25 min in the dark, or 2 mg Γ ¹ JC-1 at 37°C for 25 min in the dark, followed by 2 x washing with PBS at 37°C. Sections of samples were observed and photographed with a fluorescence microscope (Axio- vert 135 TV, Zeiss).

Seed morphology assessment

For the study of cells in the outer parenchyma layer of the hull, young spike lets section just before flowering were fixed with 2.5% glutaraldehyde in 100 mM potassium phosphate buffer, pH 7.0 and embedded in Epon. Cross-sections of 500 nm were made in the middle part using an Ultra Microtome (Ultracut E; Reichert-Jung) equipped with glass knife and stained with 0.25% toluidine blue. Endosperm cell number was analyzed using cross and longitudinal razor blade hand sections of freshly harvested 22 DAF seeds. Sections were stained with 0.25%) toluidine blue and then photographed. Cells were counted using the ImageJ program.

The surface of young spikelets collected just before flowering was visualized by Cryo Scanning Electron microscopy (Cryo-SEM). The samples were mounted on the cryo holder with Tissue Tek O.C.T "glue", plunge frozen in nitrogen slush, transferred under vacuum into the preparation chamber of the cryo system (Gatan), sputtered with a 20 nm layer of gold and finally transferred into the SEM cold stage. The samples were kept stable at -150°C under high vacuum. Pictures were taken at 5kV acceleration voltage. Images were processed using Adobe Photoshop CS3 software. The hull length was estimated with Image J and the cell number in 5th row from the middle side of lemma was counted.

Yield potential characteristics

Twenty four siblings of each event were planted in 57 x 17 x 10 cm trays (4 siblings per tray) which were randomly placed into the greenhouse and grown under the greenhouse conditions given above. Seeds of homozygous events were first germinated in Yoshida solution for 14 days (Yoshida et al, 1976), and transplanted to soil. The following traits for each plant were recorded at the mature stage: panicle number, seed number per plant, seed setting rate (% of filled grains rel. total spikelet number per panicle), 1000-grain weight and the grain yield in gram per plant. The distribution of seed dimensions was analyzed, after being dried at 37°C for 3 days using a Microtrac Dynamic Image Analyzer (DIA) and the data evaluated (n =200- 2000) with the GraphpadPrism program.

Promoter characterization

Cellular localization of the histochemical GUS signals and YFP fluorescence was investigated in the TO generation of transgenic plants. Hand cuttings and inflorescences at progressing developmental stages were collected and photographed. YFP expression in vascular tissues was observed with samples from leaf collar, sheath, cross sections at the 2nd internode, the peduncles of panicle primary branches and the spikelet rachilla. Images were captured with an Axioskop 40 (Zeiss) microscope equipped with the suitable filters for YFP detection.

Results

In brief, the examples show that the present invention led to larger seeds with increased weight in monocot and dicot seeds. Larger seeds are caused by an increased number of endosperm cells, an increased endosperm cell number caused by delayed programmed cell death, an increased number of (larger) seed grains at longer panicles, and larger seed hulls, caused by increased cell number, which allowed the seeds to grow larger. Constitutive expression in tobacco also led to larger leaves, thicker stems caused by more and larger cells and thus, in increased bio mass.

Larger seeds with increased weight in monocot and dicot seeds Larger rice seeds relative to the control were visually apparent in transgenic events in the T2 rice seed generation of transformation events expressing individually the anti-apoptotic OsBAG4 and AtBAG4 gene. This appeared more pronounced with the latter which led the inventors to concentrate on these events. Interestingly, a similar phenotype was observed with events expressing superoxide dismutase from barley. As shown in Figure 1, increased size correlates with increased seed weight; roughly 10% longer seeds produced about 10 % higher mass.

A more detailed investigation was conducted with the T5 rice seed generation, after confirmation of single integration by Southern blotting (Figure 2) and continued phenotypic selection. This showed that seed sizes of AtBAG4 events were normally distributed all showing increased length relative to the control of 8.7 ± 0.9 mm (Figure 3 A) to 9.4 ± 0.8 mm (event 2; ~ +10%), 9.5 ± 0.7 (event 3; ~+10%), 9.1 ± 0.8 (event 4; ~ +6% ). Seed width showed some small increase not exceeding ~ 2% which may be not significant (Figure 3 B). Only one out of several HvSOD events showing very similar characteristics (comp. Figure 1) was carried through the process, for scientific reasons. This is because of homology to known allergens (www.allergenoline.org) common to SODs excluding this transgene from further developments. A series of 40 transformation events expressing catalase from wheat (Acc. X94352.1) under the control of the same promoter did not show any seed phenotypic changes.

In order to assess whether the increased seed size extends to dicotyledonous plants, the inventors examined Nicotiana tabacum (Doukhanina et al, 2006) and Arabidopsis thaliana plants overexpressing AtBAG4 under constitutive promoter control. Arabidopsis seeds were not amenable to 3D analysis because of their too small size, however, they were increased in weight by 24.5% relative to the control (Table 2). Tobacco seeds showed a 30%> weight gain (Table 2) which corresponds to an average seed volume gain of ca. 40% and an average increase in length of ca. 10% (Figure 4). Furthermore, the _^ 4t ^4 G4-overpressing tobacco lines showed a grain mass per 10 capsules which was 19%> increased while the number of capsules remained approximately constant.

Table 2. Weight increase of Arabidopsis thaliana and Nicotiana tabacum seeds upon 35S::BAG4 expression

Material Seeds mg/1000 mg/1000 Inc. weight

measured (±SE) seeds (%)

Arabidopsis thaliana WT, sibling 1 490 15.3

WT, sibling 2 706 15.0

WT, sibling 3 808 15.7 15.4±0.4 0.0

35S::AtBAG4 #\, sibling 1 778 19.0

35S::AtBAG4 #\, sibling 2 774 19.6

35S::AtBAG4 #\, sibling 3, 720 18.8 19.4±0.4 24.5

Nicotiana tabacum

WT, sibling 1 321 81.3

WT, sibling 2 547 81.2

WT, sibling 3 523 99.4

WT, sibling 4 786 84.3 82.2±0.9

WT, sibling 5 634 74.2

WT, sibling 6 615 76.1 0.0

35S::AtBAG4 #\ sibling 1 526 106.3

35S::AtBAG4 #\ sibling 2 515 104.2

35S::AtBAG4 #\ sibling 3 518 105.4

35S::AtBAG4 #\ sibling 4 500 113.1 107.3±0.4 30.5

Larger seeds are caused by an increased number of endosperm cells

Randomly selected T4 seeds of successful homozygous rice events were harvested and the number of endosperm cells determined in longitudinal and transversal sections. This showed that the number of endosperm cells was increased by maximally 20% in the longitudinal direction (Figure 5 A), thus in comparable range with increased seed weight and size. No significant changes with respect to cell sizes longitudinal and lateral directions were observed. The latter also did not reveal differences in cell number; thus it can be concluded that increased seed size is predominantly determined by an increased cell number along the longitudinal axis. This is corroborated by equivalent measurements carried out with cross sections (Figure 5B), where these metrics showed no significant changes.

Increased endosperm cell number caused by delayed programmed cell death

Using a DAPI/TUNEL double staining with isolated rice endosperm nuclei, the proportion of apoptotic nuclei containing fragmented DNA was estimated. As shown in Figure 6C, the transgenic events expressing the anti-apoptotic AtBAG4, OSBAG4 as well as HvSOD all showed a diminished proportion of TUNEL-positive nuclei over the entire time course. Consequently, the total number of detectable endosperm nuclei was higher in the same transgenics as compared to the untransformed control (Figure 6 D). Viability staining of seed cross sections was used to visualize this process (Figure 7). This higher intensity of fluorescence in the transgenic endosperms indicates the presence of more viable cells. Larger seed hulls, caused by increased cell number, allow seeds to grow larger

The size and shape of the rice grain is determined by the spikelet hull before fertilization, which consists of the palea and lemma acting as a limiting frame within which grain filling occurs. It is long standing knowledge that, when allowed to grow without this restriction, kernels grow larger (T. Takeda, 1990). Moreover, factors affecting spikelet hull size and shape will impact seed grain dimensions, such as OsSPL16 (GW8), encoding a member of the SBP domain family of transcription factors (Wang et al, 2012), GW2, encoding a RING-type ubiquitin E3 ligase (Song et al, 2007), qSW5, with unknown function (Shomura et al, 2008); OsPPKLl-3, encoding a Kelch-like repeat domain protein phosphatase (Zhang et al, 2012); GS3, encoding a regulatory transmembrane protein and GS5, a putative serine protease (Li et al., 2011).

Given this correlation, the larger seed grains observed upon BAG4 and SOD expression should be preceded by the outgrowth of larger spikelets. As shown in Figure 8, this condition is met; the about 10% increased seeds size is accompanied by an equivalently increased spikelet size. To investigate the underlying cause, spikelets (of different events than shown in Figure 8) were investigated by Cryo-SEM, photographed, and the cell number determined as shown in Figure 9. The quantitative analysis across several spikelets gave the result shown in Figure 10. The increase in seed length of 12-14% (2) is mirrored by an increased cell number in the longitudinal direction of 13 - 17% (1). This implies that the cell size in this direction remains unchanged, a condition that is met (4), just like with cell sizes in cross sections that showed no significant increase. Thus, these results with hull cells are essentially the same as those obtained with endosperm cells (see Figure 5).

Gtl promoter leakiness explains increased spikelet size

The increased cell number present in spikelet hulls and the increased hull size are only explicable when strict endosperm specificity of the Gtl -promoter (Okita et al., 1989; Zheng et al, 1993; Wu et al, 1998; Qu et al, 2008) is doubted. The inventors therefore used exactly the same 1.85 kb Gtl promoter version to drive the expression of the reporter genes NLS-GUS and NLS-YFP in the same (cv. Kaybonnet) rice background. The analysis of the resulting TO plants revealed promoter leakiness in the developing spikelet ^' s lemna and palea, before panicle emergence (Figure 11 A) in dividing zones, the leaf basis and the outgrowing awn. It is only when the growth of spikelets is almost completed, shortly before emergence, that this spotted expression profile changes in favor of a uniform spatial expression across the hull (Figure 12).

Preferential expression actively dividing cells is also witnessed by reporter signals in regenerating embryogenic calli and in root tips (Figure 13). Moreover, a signal was obtained in stages of inflorescence development (Figure 14) during which panicle branches and spikelet primor- dia form. At later stages, a strong signal is obtained in the differentiating vascular bundles. In fact, expression in vascular tissues is maintained throughout vegetative growth (Figure 15.

As expected for an endosperm promoter, reporter gene expression was very strong in the endosperm of developing seeds, however, decreased expression is observed at later stages (Figure 16). This starts at the center expending outwards and can most probably be attributed to PCD which shows this pattern of histological spread in the endosperm (see Figure 7; and Kobayashi et al., 2013).

Increased number of (larger) seed grains at longer panicles

During early generations of transgenic BAG4 and HvSOD events, the inventors noted that the yield, in terms of gram seeds per plant, exceeded - by far at times - the ca. 10% yield increase that is explicable by increased seed size and weight. This led the inventors to conduct more systematic studies with two selected rice events in the T3 and T4 plant generation (yielding T4 and T5 seeds, respectively) at different times of the year. T4 and T5 seeds were harvested in April and December, respectively, and therefore differed quantitatively in several parameters determined (Table 3). As shown in Figure 17, all parameters investigated, with the exception of the seed setting rate (= filled/total seeds) and seed diameter, contributed to yield increase. Their relative contribution can be variable and have different weights. Particularly, an increase in panicle size and its secondary branches, however small in percentage, can lead to the accommodation of significantly more seeds. A typical appearance is shown in Figure 18. The reason for these changes in panicle size and the concomitant formation of more spikelets probably resides in the expression of BAG4 in the flower primordia (Figure 14).

Table 3. Components contributing to increased rice grain yield (g/plant).

T3 plants (producing T4 seeds) and T4 plants (producing T5 seeds) were randomly grown in the greenhouse and harvested in April and December 2014, respectively. Numbers represent the mean (±SD); n= 20 plants (T3) and 24 plants (T4). Seed parameters (n > 500) were meas- ured manually in the T3 generation and automatically with a high speed camera (Microtrac Dynamic Image Analyzer) in the T4 plant generation allowing n > 1500.

Panicle Main Seed Inc. Seed Inc. Seed g/1000 Inc. Mature Inc. mature g/plant Inc. Transgen

Nr./plant Panicle length seed width seed setting grain g/1000 seeds/ seeds./plant g/plant copy No.

length [mm] length [mm] width rate [96] grains plant [96] [96]

[cm] [%] [%] [96]

0

Control 3.3±0.8 22.3±3.2 8.5±0.7 0.0 2.0±0.1 0.0 78.2±8.6 16.1±0.9 437±78 0 6.7±1.8 0

T4 8.1±2.2

3.3±0.8 25.2±3.1 9.0±0.6 5.7 2.0±0.1 0.0 75.2±5.8 17.3±0.3 514±70 17.7 21.0 1

AtBAG4#2 (ns)

T4 10.9±3.0

3.7±0.7 25.1+3.0 9.0±0.7 5.7 2.1+0.1 0.0 79.8±5.5 16.9±1.0 656±205 50.1 63.0 1

AtBAG4M p<0.01)

4.4±1.2 19.5±1.4 8.7±0.9 0.0 2.3±0.04 0.0 91.4 17.0±0.7 0.0 548±129 0.00 9.3±2.2 0.0 0

Control

T5 12.5±2.3

4.8±1.4 20.2±1.9 9.5±0.9 9.3 2.3±0.02 0.0 92.9 18.8±0.7 10.2 668±130 22.0

AtBAG4#2 P<0.01 ³⁴ - ^{2 1} T5 11.8±3.3 _nn

4.4±1.2 22.0±1.7 9.2±0.9 6.2 2.3±0.01 0.0 92.1 18.2±0.5 6.9 648±172 18.4 _/ ,- _> 26.90 1

AtBAGa p<0.05

Increased organ size and biomass of TV. tabacum plants constitutively expressing AtBAG4

Transgenic tobacco plants expressing 35S: :AtBAG4-GFP (four events) showed pronounced growth characteristics, in addition to the larger seeds noted above. Young plantlets already showed strongly increased leaf size (Figure 19A), a feature that was maintained during growth to maturity (Figure 19B). Leaf areas were determined at the given positions on the plant (Table 4) and showed to be increased by ca. 48%. The morphological appearance of these larger leaves was unchanged. The maximal length/maximal width ratio remained largely unaffected (1.33 in the wild-type, 1.36 in the transgenics, at average; n= 6, each) and so was the number of main leaf veins (16.7 in the wild-type, 17.0 in the transgenics, at average; n= 7, each, at position #15 and #16 ).

Table 4. Leaf areas of WT and _^ 4f2L4G¥-overexpressing plants (n=6, each). Leaf areas were determined at the given leaf positions on the plant (numbered from the bottom).

Wild-type (n=6) _^ 4flL4G -overexpressors

Leaf # 15 Leaf # 16 Average Leaf # 15 Leaf # 16 Average(±SE) [cm ² ] [cm ² ] (±SE) [cm ² ] [cm ² ] (#15, #16)

(#15, #16)

287.7 273.6 500.0 454.7

206.9 408.2 384.0 452.1

319.6 298.7 379.0 408.7

326.0 329.2 578.1 452.3

371.5 374.0 529.8 360.5

236.2 367.1 316.6±31.4 600.0 508.1 467.3±60.4

All transgenic plants also showed a pronounced increase in stem thickness (19.4 ± 1.8 mm, diameter; n = 7) compared to the wild-type (13.7 ± 2.2 mm, diameter; n=7), thus an average increase of ca. 42%, measured at the 15 ^th leaf position of flowering plants (Figure 20A). Cross sections of the stems (Figure 20B) show that this increase is mainly due to increased growth of the pith parenchyma (a) and the cortex (b). Measurements made with microscopic thin sections as shown in Figure 20C suggest that the increase is caused by both, an increased cell number as well as an increase in cell diameter.

At a given age of the plants (shortly before flowering), the leaves were somewhat lower in number in the AtBAG4 overexpressing plants. The latter showed 30.1±1.6 (n=7) leaves while 39±1.6 (n=7) leaves were found in the wild-type (ca. -23%). To cope with this ambiguity potentially compensating the increased leaf size, the biomass produced was determined with mature plants (possessing seed capsules) after drying to constant weight at 70°C (n = 7; WT and AtBAG4 overexpressors, each), separated into roots and above-ground parts (seed capsules excluded). This revealed pronounced increases in biomass of the green plant material growing above ground of ca. +43%> (Table 5), while the root mass increase of ca. + 8%> may not be significant.

Table 5. Dry mass of wild-type and _^ 4f2L4G -overexpressing plants (n=6, each).

Wild-type _^ 4f2L4G -overexpressors

Dry mass Dry mass Dry mass Dry mass

Above ground [g] Root [g] Above ground [g] Root [g]

59.8 5.8 91.0 7.0

43.1 4.9 91.8 6.7

73.3 6.5 87.0 6.2

78.0 8.5 86.8 5.1

58.6 4.2 89.3 6.4

60.1 4.4

Av: 62.1±12.3 Av: 5.7±1.6 Av: 89.1±2.6 Av: 6.2±0.8 Taken together, At-BAG4 - overexpressing plants show significantly larger leaves (+40%), significantly thicker stems (+ 42%). They have a reduced leaf number (ca. -23%) at a specific age which does not compensate the increased biomass produced of ca. + 43%. This is accompanied by the production of heavier and larger seeds (see above).

Increased cell proliferation and increased longevity is at the basis of the invention

Based on the data presented, our current understanding is that AtBAG4, directly or indirectly, acts as a positive regulator of cell proliferation (the sum of cell division and cell expansion) under standard growth conditions, this forming the basis of the here-presented invention. This is a property that is novel in the field of Programmed Cell Death Research implying that PCD in plants constitutes a continuous process involved in cell homeostasis and morphogenesis. Even in very young leaves (where no PCD is expected), increased cell proliferation is witnessed in the AtBAG4 overexpressing lines (Figure 21A,B) which are greener and have a curly shape signaling rapid cell proliferation which does not cope with epidermal expansion and the outgrowth of leaf veins. This situation becomes balanced when leaves develop further. To investigate this issue out of the plant context with its changing (phyto-)hormonal environment, protoplasts were generated according to standard methods and grown in a liquid medium. This showed that the _^ 4t ^4 G4-overexpressing cells grew very significantly faster as compared to the wild-type cells (Figure 21C). This indicates that cell proliferation, as provoked by AtBAG4 is a plant-independent cellular process. Cell proliferation as the basis is also witnessed, for instance, in rice grains (more cells), in tobacco stems (more and larger cells). The larger rice panicles accommodating more seeds is attributed to AtBAG4 - expression in the flower primordia (comp. Figure 14) caused by the specificity of the Gtl- promoter, this leading to enhanced spikelet formation.

Increased longevity of plants and of cells is yet another component leading to increased yield potential. Phenotypically, _^ 4t ^4 G4-overexpressing tobacco plants are also characterized by their delayed senescence (Figure 22A) which can contribute to the yield potential by allowing prolonged assimilate production. This delay is caused by the ethylene insensitivity of the transgenics. When non-senescent leaves (of the same middle positions on the plants) are treated with ethylene, the trangenics showed no response (Fig. 22B,C). The mechanism, by which AtBAG4 can attenuate ethylene sensing, is unknown.

Literature Cited Bertin J, Mendrysa SM, LaCount DJ, Gaur S, Krebs JF, Armstrong RC, Tomaselli KJ, Friesen PD (1996) Apoptotic suppression by baculo virus P35 involves cleavage by and inhibition of a virus-induced CED-3/ICE-like protease. J Virol 70: 6251-6259

Brodsky J.L. and Bracher A. (2007). Nucleotide exchange factors for Hsp70 molecular chaperones in: Networking of Chaperones by Co-Chaperones, ed. G. L. Blatch (New York: Landes Bioscience/Eurekah.com), 1-12.

Bump NJ, Hackett M, Hugunin M, Seshagiri S, Brady K, Chen P, Ferenz C, Franklin S, Ghayur T, Li P (1995) Inhibition of ICE family proteases by baculovirus antiapoptotic protein p35. Science 269: 1885-1888

Cho EK, Hong CB (2006) Over-expression of tobacco NtHSP70-l contributes to drought- stress tolerance in plants. Plant Cell Rep 25: 349-358

Clem RJ, Fechheimer M, Miller LK (1991) Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science 254: 1388-1390

Clem RJ, Miller LK (1994) Control of programmed cell death by the baculovirus genes p35 and iap. Mol Cell Biol 14: 5212-5222

Dickman MB, Park YK, Oltersdorf T, Li W, Clemente T, French R (2001) Abrogation of disease development in plants expressing animal antiapoptotic genes. Proc Natl Acad Sci U S A 98: 6957-6962

Doukhanina E V, Chen S, Zalm E Van Der, Godzik A, Reed J, Dickman MB (2006a)

Identification and Functional Characterization of the BAG Protein Family in Arabidopsis thaliana . J Biol Chem 281: 18793-18801

Dutt M, Dhekney S a, Soriano L, Kandel R, Grosser JW (2014) Temporal and spatial control of gene expression in horticultural crops. Hortic Res 1: 14047

Gibson DG, Young L, Chuang R, Venter JC, Iii CAH, Smith HO (2009) Enzymatic

assembly of DNA molecules up to several hundred kilobases. 6: 12-16

Hoang TML, Moghaddam L, Williams B, Khanna H, Dale J, Mundree SG (2015)

Development of salinity tolerance in rice by constitutive-overexpression of genes involved in the regulation of programmed cell death. Front Plant Sci 6: 1-14

Ilg A, Yu Q, Schaub P, Beyer P, Al-Babili S (2010) Overexpression of the rice carotenoid cleavage dioxygenase 1 gene in Golden Rice endosperm suggests apocarotenoids as substrates in planta. Planta 232: 691-699

Joly AL, Wettstein G, Mignot G, Ghiringhelli F, Garrido C (2010) Dual role of heat shock proteins as regulators of apoptosis and innate immunity. J Innate Immun 2: 238-247

Kabbage M, Dickman MB (2008) The BAG proteins: A ubiquitous family of chaperone regulators. Cell Mol Life Sci 65: 1390-1402

Kobayashi H, Ikeda TM, Nagata K (2013) Spatial and temporal progress of programmed cell death in the developing starchy endosperm of rice. Planta 237: 1393-400 Li Y, Fan C, Xing Y, Jiang Y, Luo L, Sun L, Shao D, Xu C, Li X, Xiao J, et al (2011)

Natural variation in GS5 plays an important role in regulating grain size and yield in rice. Nat Genet 43: 1266-1269

Lincoln JE, Richael C, Overduin B, Smith K, Bostock R, Gilchrist DG (2002) Expression of the antiapoptotic baculovirus p35 gene in tomato blocks programmed cell death and provides broad-spectrum resistance to disease. Proc Natl Acad Sci U S A 99: 15217- 15221

Okita TW, Hwang YS, Hnilo J, Kim WT, Aryan a P, Larson R, Krishnan HB (1989) Structure and expression of the rice glutelin multigene family. J Biol Chem 264: 12573- 12581

Paine JA, Shipton CA, Chaggar S, Howells RM, Kennedy MJ, Vernon G, Wright SY, Hinchliffe E, Adams JL, Silverstone AL, et al (2005) Improving the nutritional value of Golden Rice through increased pro-vitamin A content. Nat Biotechnol 23: 482-487

Qu LQ, Xing YP, Liu WX, Xu XP, Song YR (2008) Expression pattern and activity of six glutelin gene promoters in transgenic rice. J Exp Bot 59: 2417-24

Roy S, Banerjee A, Tarafdar J, Senapati BK (2012) Detection of probable marker- free transgene-positive rice plants resistant to rice tungro disease from backcross progenies of transgenic Pusa Basmati 1. J Genet 91: 213-218

Shomura A, Izawa T, Ebana K, Ebitani T, Kanegae H, Konishi S, Yano M (2008)

Deletion in a gene associated with grain size increased yields during rice domestication. Nat Genet 40: 1023-1028

Singh BK, Jenner CF (1982) A modified method for the determination of cell number in wheat endosperm. Plant Sci Lett 26: 273-278

Song X- J, Huang W, Shi M, Zhu M-Z, Lin H-X (2007) A QTL for rice grain width and weight encodes a previously unknown RING-type E3 ubiquitin ligase. Nat Genet 39: 623-630

Takeda K (1990) Inheritance of grain size and its implications for rice breeding. In: Rice genetics II, Proceedings of the Second International Rice Genetics Symposium 14- 18 May 1990, IRRI, pp 181 - 190. ISBN 971-22-0027-2

Wang S, Wu K, Yuan Q, Liu X, Liu Z, Lin X, Zeng R, Zhu H, Dong G, Qian Q, et al

(2012) Control of grain size, shape and quality by OsSPL16 in rice. Nat Genet 44: 950- 954

Wang Z, Song J, Zhang Y, Yang B, Chen S (2009) Expression of baculovirus anti-apoptotic p35 gene in tobacco enhances tolerance to abiotic stress. Biotechnol Lett 31: 585-589

Wu C, Suzuki A, Washida H, Takaiwa F (1998a) The GCN4 motif in a rice glutelin gene is essential for endosperm-specific gene expression and is activated by Opaque-2 in transgenic rice plants. 14: 673-683 Wu CY, Suzuki A, Washida H, Takaiwa F (1998b) The GCN4 motif in a rice glutelin gene is essential for endosperm-specific gene expression and is activated by opaque-2 in transgenic rice plants. Plant J 14: 673-683

Xue D, Horvitz HR (1995) Inhibition of the Caenorhabditis elegans cell-death protease

CED-3 by a CED-3 cleavage site in baculovirus p35 protein. Nature 377: 248-251

Ye X, Al-Babili S, Kloti A, Zhang J, Lucca P, Beyer P, Potrykus I (2000) Engineering the provitamin A (beta-carotene) biosynthetic pathway into (carotenoid-free) rice

endosperm. Science 287: 303-5

Yoshida S, Forno DA, Cock JH, Gomez KA (1976). Laboratory Manual for Physiological studies of Rice. IRRI, Las Banos, Laguna. pp.83.

Zhang X, Wang J, Huang J, Lan H, Wang C, Yin C, Wu Y, Tang H, Qian Q, Li J, et al

(2012) Rare allele of OsPPKLl associated with grain length causes extra-large grain and a significant yield increase in rice. Proc Natl Acad Sci U S A 109: 21534-9

Zheng Z, Kawagoe Y, Xiao S, Li Z, Okita T, HYoshidau TL, Lin A, Murai N (1993) 5' distal and proximal cis-acting regulator elements are required for developmental control of a rice seed storage protein glutelin gene. Plant J 4: 357-366