Backoround of the Invention

Transcriptional and translational fusions to reporter genes whose products can be easily assayed offer powerful approaches to studying gene structure, expression, regulation, gene product assembly, trans¬ port and compartmentalization (Rosenberg e_£ __LI• , Science 222 :734-739, 1983; Bonnerot <__i al . , Proc. Nat. Acad. Sci. USA jϋ:6795-6799. 1987; Finnegan _=£ al.. The Plant Cell 1:757-764, 1989). Some of these gene fusions have been applied in more wide-ranging studies such as discerning cell lineage during development and purifying gene products (Germino ≤i ϋl., Proc. Nat. Acad. Sci. USA 80:6848-6852. 1983; Silhavy and Beckwith, Microbiol. Rev. 49:398-418, 1985; Scholtissek and Grosse, QsΑS. £2:55-64, 1988). Pioneered with £__. coli β-galactosidase (lacZ) , gene fusions have been adapted to other organisms includ¬ ing yeast, animals and plants. While transcriptional fusions might be construed as straightforward, translational fusions require reporters which can function despite covalent addition of extraneous polypeptides to their amino- or carboxy-terminus.

Some genetic markers impart a selectable phenotype such as antibiotic resistance while other markers specify an enzymatic reporter activity for which there is a colorimetric or luminescence assay, making them suitable in combination for selection and screening. Single genetic markers which provide both of these features would be particularly useful.

S mmary pf the Inven ion

This invention pertains to genetic markers which have a biochemically assayable (reporter) activity and confer a conditionally selectable growth advantage. The genetic markers are fused genes comprising a first structural gene which encodes the biochemically assayable product and a second, different structural gene which encodes a product whose activity confers the selectable growth advantage in a transformed cell. The translational product of the fused gene is a single contiguous polypeptide (fusion protein) which exhibits the activities of both gene products. In a preferred embodiment, the first structural gene encodes an enzyme that acts on a chromogenic substrate, such as the enzyme β-glucuronidase (GUS) , and the second structural gene encodes an enzyme which confers antibiotic resistance to a transformed cell, such as the enzyme neomycin phosphotransferase (NPT-II) which confers resistance to several aminoglycoside antibiotics.

The genetic markers of this invention can be adapted for use in prokaryotic and eukaryotic cells, including yeast, animal and plant cells. They provide for powerful positive genetic selection and facile, sensitive biochemical and histochemical detection in transformed cells. They facilitate genetic selection of transformed cells and permit subsequent spatial localization and quantitative estimation of gene activity. In addition, the markers can be used to probe for and recover novel genes and genetic regulatory elements such as promoters.

Brief Description of the Drawings

Figure 1 is a schematic representation of a fusion between the carboxy-terminus of GUS and the amino-terminus of NPT-II.

Figure 2 shows the comigration of GUS and NPT-II activities in extracts of DH5α cells transformed with a plasmid carrying the as.: :npt-II fused gene on molecular sieve chromatography.

Figure 3 shows immunoblot analysis of _£__. coli- produced GUS and GUS::NPT-II fusion proteins.

Figure 4 shows the elution profile of GUS and NPT-II activity in extracts of tobacco plants trans¬ formed with a plasmid carrying the qus: :npt-II fusion gene on gel permeation chromatography.

Figure 5 shows the immunodetection of GUS and NPT-II polypeptides in leaf extracts from the transgenic tobacco plants.

Figure 6 shows gel electrophoretic analysis of GUS activity in tobacco plants.

Figure 7 is a schematic diagram of a probe designed for insertional tagging of plant promoters through T-DNA mediated transformation.

Figure 8 is a histogram detailing tissue and/or developmental specific β-GUS expression.

Detailed Description of the Invention

The genetic markers of this invention are bifunctional fused genes which provide a reporter activity and a trait which can be selected for. The fused genes comprise a first structural gene which encodes a product whose activity can be assayed bio¬ chemically fused to a second structural gene which encodes a product whose activity confers a growth advantage that is conditionally selectable.

The reporter gene preferably encodes an enzyme which acts on a chromogenic or luminogenic sub¬ strate. In general, the enzyme must have low enough activity in any prospective host cell so that the activity associated with the genetic marker can be distinguished from any endogenous (background) activity of the cell. Suitable enzymes may include GUS, chloramphenicol acetyltransferase (CAT), nopaline synthase (NOS) , β-galactosidase (LAC) and luciferase (for example, firefly or bacterial luxAB luciferase) .

The second gene of the fusion confers conditional growth advantage. This provides for positive genetic selection of cells transformed with the marker. The selectable trait can be antibiotic resistance, such as resistance to an aminoglycoside antibiotic (e.g. neomycin, kanamycin, hygromycin or gentamycin) . Alternatively, genes which confer resistance to other types of antibiotics, such as CAT which confers resistance to chloramphenicol, can be used. In addition, the gene can be a modified or mutant gene. Some examples are a modified actin gene which confers resistance to bleomycin, the str gene which encodes an altered bacterial rRNA resulting in streptomycin resistance or the mdr gene which encodes a mutant membrane protein that confers multiple drug resistance.

A preferred embodiment of the marker is a qus: :npt-II fusion gene. This marker is especially useful in plants. The qus gene (E___ coli uidA) has all of the relevant features of a reporter gene for plants as well as other systems that lack significant endogenous GUS activity (Jefferson, Plant ol. Biol. Reo. 1:387-405, 1987). It meets the criteria of

sensitive and simple assayability — a low background of endogenous activity in plants, lack of toxicity in transgenic plants (Jefferson e_£ _a - , EMBO J. £:3901- 3907, 1987) and the availability of convenient substrates such as the indigogenic substrate 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (XGluc) . GUS has been used as a biochemical and histochemical marker, alone or as a transcriptional reporter in studying developmental patterns of gene expression.

The npt-II gene provides resistance to the aminoglycoside antibiotics neomycin, kanamycin and geniticin. It is a widely used marker in several taxonomically diverse representatives of bacteria, fungi, plants and animals. Its broad host range, its ability to provide for positive genetic selection and the availability of a radiochemical assay for NPT-II contribute to the widespread use of the gene. The kanamycin-resistance (Km ^R ) phenotype conferred by the npt-II gene is the most prevalent selection marker in plant genetic transformation (Klee e_t ____!• _> Ann. Rev. Plant Physiol. M:467-486, 1987) where it has also been used as a reporter in transcriptional and translational fusions (Teeri e_t a_l. , EMBO J. £:1755- 1760, 1986; Vaeck ££ &!., ]5__Lt_JL_____ 223. :33-37, 1987). However, the use of the gene as a reporter has been limited because of the lack of an assay for NPT-II activity other than the radiochemical method (Radke e_t aJL . , Theor. APPI. Genet. 25:685-694, 1988) or the less commonly used method of immunoscreening with polyclonal antibodies (Vaeck e_t ja_l. , Nature £2£:33-37, 1987).

The component genes for constructing the markers of this invention can be obtained from standard sources or synthesized. The first and

second structural genes can be fused together either directly or through an intergenic nucleotide sequence. The order of fusion can vary depending on the capability of a particular gene product to accept fusions at the carboxy- or amino-terminus. This capability can be determined empirically. A 5' location of the gene encoding the biochemically assayable enzyme, in relation to the gene conferring the conditionally selectable growth advantage, increases the likelihood that transformed cells selected for the growth advantage will also exhibit the reporter activity.

The genes can be fused by several different methods. One method is based upon the assumption that if stable fusions are possible between any two genes of interest, they will form and persist in vivo. A construct is prepared in which the coding region of the second gene is placed out of frame relative to the coding region of the first gene such that selection for function of the second gene will favor appropriate mutations for its expression.

As described in greater detail in the Exemplification below, this in vivo approach was taken to construct qus: :npt-II fusions. The procedure is applicable to the fusion of other genes. Gene fusions were prepared in which the npt-II gene was placed out of frame with respect to an upstream qus gene. To prevent secondary trans¬ lational initiation of the npt-II coding region, its initiation codon was removed. The fused, gene was placed under the control of the tac promoter. The plasmid carrying this construct was phenotypically GUS ⁺ and Km ^s as the initial construct. !__. coli cells were transformed with the plasmid and spontaneous

bacterial mutants which had acquired kanamycin resistance were selected. Analysis of the nucleotide sequence of the fused gene at the junction region revealed that a single nucleotide had been deleted close to the qus translational stop codon. This deletion brought the npt-II coding region into frame with the ous coding region but placed the original qus stop codon out of frame. Biochemical analysis showed the fusion protein was bifunctional having the enzymatic activities of the two individual genes.

In an alternative approach to constructing the gene fusions of this invention, the coding regions of the two genes can be fused together in proper frame and placed directly under the control of an appro¬ priate promoter. To create these fusions, of course, the translational stop codon associated with the 5' structural gene must be inactivated so that transla¬ tion will carry through to the coding region of the 3' structural gene. This can be accomplished by deleting the codon, changing it or by placing it out of frame by making an upstream nucleotide insertion or deletion, in conjunction with an appropriate downstream deletion ^" or insertion which brings the coding region of the 3' structural gene into frame with the coding region of the first structural gene. In addition, the coding region of the 3' gene should be modified to remove or deactivate the initiation codon and any surrounding ribosomal binding sites to prevent secondary initiations. Positive gene fusions can be verified as described in the first approach.

In some cases, the fusion protein may be deleterious to cells. When this is the case, expression of the fusion gene can be regulated. For example, in 3 _. coli. a lad (repressor) gene can be

used to regulate the expression of the fusion gene. Differential screening in the presence and absence of the tac-promoter inducer isopropylthiogalactoside can be used to assess deleterious expression.

The genetic markers of this invention have a wide spectrum of utility. They are useful in pro- karyotic and eukaryotic systems including yeast, other fungi and animal cells — generally any system for which gene-transfer protocols have been devised and where combinations of assayable and selectable marker genes can be usefully employed. The genetic markers of this invention simplify construction of nucleic acid vectors. The markers can be used in standard animal, plant and bacterial nucleic acid vectors. To activate the marker, it is operatively linked to a transcriptional promoter (and optionally an enhancer) which is active in the biological host of interest.

As described, the fused gene qus: :npt-JI is particularly useful as a marker in plants. It can be used in standard plant cell vectors such as Ti plasmids of &__ tumefaciens in connection with a suitable transcriptional promoter active in plants. The marker gene allows selection at the whole plant level or callus stage. For instance, placed in combination with a constitutive promoter, the marker gene enables the selection of kana ycin resistant seedlings in progeny. In addition, the qus: :npt-II marker allows direct kanamycin selection of trans¬ formed plant cells at the callus stage and subsequent enzymatic or histochemical assessment of the GUS activity in regenerated plants. The GUS activity of the marker can be employed to conveniently enzymati- cally correlate the expression of the marker with a

kanamycin resistant phenotype. This is especially useful in recalcitrant species where the efficiency of genetic selection using aminoglycoside antibiotics . is poor.

The genetic markers of this invention can also be used as molecular probes for genetic regulatory elements. They offer a powerful approach for the identification, characterization and recovery of these elements. As described below, the probes can insertionally tag active genetic regulatory elements such as transcriptional promoters in genomic or other DNA. These may include tissue-specific and developmentally specific regulatory elements.

In general, a probe for insertional tagging of a transcriptional promoter comprises a mobile genetic element, such as a plasmid, virus or a transposon, containing a genetic marker of this invention, without its own transcriptional promoter, located proximate to an insertion junction of the element. The location of the marker next to an insertion junction allows it to be activated by a transcrip¬ tional promoter situated adjacent to the other side of the junction. Thus, insertion of the probe into a cellular genome will occasionally result in the genetic activation of the marker gene by its juxtaposition to a chromosomal regulatory sequence. The active regulatory element so identified can then be isolated by any of several recombinant DNA approaches.

Figure 7 shows a preferred embodiment of a probe designed for insertional tagging of plant transcrip¬ tional promoters through T-DNA mediated transform¬ ation. The probe contains the qus: :npt-II fused gene located between the right border (RB) and left border

(LB) of the T-DNA. Preferably, the marker is located proximate to, and transcriptionally downstream from, the Ti right border sequence. A transcriptional terminator such as the transcriptional terminator sequence from the 3' region of the Ti nopaline synthetase gene (NOS-T) is placed 3* to the marker gene. A translational enhancer, such as the AMV enhancer shown, may be included between the border and the marker. Optionally, a sequence containing translational termination signals in all three reading frames (.poly-TGA) can be located between the right border junction and the marker gene. This enables the exclusion of further peptide fusions upon insertion of the marker gene into a chromosomal or other gene. This arrangement serves to further specifiy the selection of transcriptional fusions exclusively, for purposes of identification of new regulatory sequences.

Upon integration of the probe into plant cell¬ ular DNA, activation of the flus.: :j_t_≥£.-_LI gene can be directly selected for on the basis of resistance to an aminoglycoside antibiotic such as kanamycin. Expression characteristics of the tagged regulatory sequence can be conveniently followed by assessment of GUS expression. In this way, the qus: :npt-II gene can greatly facilitate the selection and character¬ ization of plant genes which exhibit various expres¬ sion patterns, the evaluation of regulatory elements in vivo and the determination of their activity through development.

The regulatory sequence identified by the insertional tagging procedure can be recovered by any of several approaches. Total genomic DNA can be isolated from transformed cells or tissues, cleaved

with restriction enzymes and then submitted to Southern analysis using the qus: :not-II coding region or a portion thereof as a hybridization probe to identify fragments containing the marker gene. In this way, useful restriction enzyme sites upstream of the insertion junction of the probe can be approxi¬ mately localized. Genomic DNA can then be cleaved with the cognate restriction endonuclease to generate restriction fragments which include the region upstream of the junction containing the regulatory sequence and some portion of the gus: :npt-II sequence. The restriction fragment can then be subcloned and identified using an appropriate hybridization probe for the unique sequence associated with the retained portion of gus: :npt-Il .

Alternatively, a polymerase chain reaction (PCR) can be used to amplify the sequence. When necessary, in the event the chosen restriction site leaves a 3 ' recessed end, the restriction fragment can be blunt- ended (using a DNA polymerase in the presence of all four deoxynucleotides) and 'tailed ^* with oligo(dA) at its 3' terminus using terminal deoxynucleotidyl- transferase. The PCR can then be performed with two primers: oligo(dT) and a us-specific oligonucleo- tide complementary to the non-coding strand of 5' region of the gus gene. The amplified PCR product (by virtue of the specificity and orientation of the gus-specific primer) is the chromosomal region between the insertion junction and the site of oligo(dA) tail at the restriction enzyme site.

In yet another method, the restriction fragment can be circularized by self-ligation and then amplified by an "inverse" PCR. For this, diverging

oligonucleotide primers are used. One is complemen¬ tary to the non-coding strand in the 5* region of the qus: :npt-II coding region; the other is complementary to the coding strand in the 3' region of the qus: :n_p£.-IX coding region of the NOS polyadenylation sequence.

The regulatory sequence associated with the gus: :npt-II probe can also be isolated by standard recombinant DNA methods. For example, a genomic library of the insertionally tagged cellular DNA (e.g. a genomic library of DNA purified from a regenerated plant) can be constructed. The library can be screened with a hybridization probe derived from the qus: :npt-II coding region or from the transcribed, non-translated "leader" region immediately 5' to the insertion junction. This leader sequence can be determined by RACE-PCR cloning and sequencing of the specif c jams.: :m_£.-XI cDNA associated with the insertional tagging, using cDNA libraries generated from the tissue where the qus: :npt-II marker gene is expressed.

In addition to _. the regulatory sequence, the gene naturally associated with the regulatory sequence can be isolated. The insertionally tagged DNA is cleaved with a convenient restriction enzyme to produce a restriction fragment which includes some portion of the region transcriptionally 'downstream* from the site of chromosomal insertion of the probe (for example the Sad site situated between the 3' ter¬ minus of gus: :npt-ll and the NOS-T polyadenylation sequence) . The fragment can be subcloned using a hybridization probe specific for the carboxy-terminal region of the qus: :npt-II marker as a probe for the gene.

A PCR procedure similar to that described above can also be used to isolate the gene. An oligo(dT) primer is used in combination with a qus: :npt-II or NOS terminator-specific oligonucleotide primer complementary to the coding strand of 3* region of the qus: :npt-II gene or NOS-T sequence. The ampli¬ fied product, by virtue of the specificity and orientation of the qus: :npt-II or NOS terminator- specific primer, will be the chromosomal DNA region between the 3' end of the gas.: :nEi.-I_l and the site of oligo-dA addition at the chosen restriction enzyme site.

The invention is illustrated further by the following exemplification.

Exemplification

The following abbreviations are used: Ap, ampicillin; bp, base pair(s); CaMV, cauliflower mosaic virus; DTT, dithiothreitol; GUS, β-glucuronidase; qus, E. coli gene encoding GUS (uidA) ; kb, kilobase(s) or 1000 bp; Km, kanamycin; NPT-II, neomycin phosphotransferase-II; not-II, gene encoding NPT-II; nt, nucleotide(s) ; oligo, oligode- oxyribonucleotide; PAGE, polyacrylamide gel electro- phoresis; Pollk, Klenow (large) fragment of _=L__ coli DNA polymerase I; ^R , resistance/resistant; SDS, sodium dodecyl sulfate; XGluc, 5-bromo-4-chloro-3- indolyl-β-D-glucuronide; ::, novel joint (fusion); [], denotes plasmid carrier state.

Assembly of ous::npt-II fusions in vivo and in vitro

Plasmid DNAs were prepared in large volume as described (Clewell and Helinski, Biochemistry 1:4428-4440, 1970), or using a minipreparation

procedure (Ish-Horowicz and Burke, Nucl. Acids. Res. 1:2989-2998, 1981). Standard recombinant DNA techniques were used (Maniatis eJt jg .. , Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1982). A 0.8 Kb Hindlll fragment containing the truncated npt-II structural gene from plasmid pABDI (Paszkowski fit al. EMBO Jt 2:2717-2722, 1984) was modified by filling-in with Pollk and by addition of an 8-mer SacI linker. The resulting £a£l fragment was ligated into the corresponding site in pBI221 (Jefferson e_t al. , EMBO J. 1:3901-3907, 1987) 3' to the gus gene. Lastly, the resident plant promoter element was replaced by a Hi lII-_-_aiI__iI fragment containing a 96 nucleotide (nt) bacterial tac promoter cassette (Pharmacia-LKB) to yield pGKl. Approximately 10 ¹¹ cells of DH5α[pGKl] were plated on kanamycih-containing medium and some 10 ³ Km ^R colonies were recovered and pooled. DH5α was re-transformed with plasmid DNA from the pool, and colonies plated to arnpicillin-containing medium. These were screened for Km ^R and some twenty were found to be doubly ampicillin and kanamycin resistant. Colonies were replica-plated and assayed for GUS overexpression on Whatman filter disks impregnated with a bacterial lysis solution (Holmes and Quigley, Anal. Biochem. 114:193-197, 1981) containing the indigogenic GUS substrate 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (XGluc Research Organics Inc.; Jefferson, Plant Mol. Biol. Rep. 1:387-405, 1987). The junction nucleotide sequence (determined by the procedure of Sanger e_£ 21 - . Proc. Natl. Acad. Sci. USA 74:5463-5467, 1977) of the fusion gene is shown_ in Figure 1 within stippled boxes to indicate coding region domains derived from the qus and npt-II structural genes, and

are separated by the 61 nt intergenic region origi¬ nating from the 3' nontranslated sequence of E__. coli uidA. (The nucleotide and amino sequence corresponding to the junction within pGKK14 are provided as Sequence ID No. 1 and 2 respectively. Likewise, the junction nucleotide and amino acid sequence within pGKl are given as Sequence ID No. 3 and 4.) The qus termination codon and npt-II 5' SacI linker modification are underlined. Site-directed mutagenesis was carried out using the procedure of Kunkel, Proc. Nat. Acad. Sci. USA 82:488-492. (1985), with oligόnucleotides synthesized using phospho- ramidite chemistry on a Biosearch model 8700 synthesizer. Oligonucleotide OL30 (Sequence ID No. 5 in the Sequence Listing) is shown above the bold and asterisked 'A' nucleotide at position 1790, which is present in GK1 and deleted in the pBI-403A and pBI- 403B site-directed mutants. The deduced amino acid sequence spanning the qus. intergenic and npt-II coding regions is shown as predicted for the parental GUS (Seq. ID No. 4) and fusion peptides (Seq. ID No. 2).

Molecular sieve chromatography of native proteins from E. coli strains expressing qus. npt-II and qus::nPt-II fusion genes

All procedures were performed at 0°-4°C. For E. coli. mid-log phase cells (20 ml) were washed with an extraction buffer containing protease inhibitors as described (Platt and Yang, Anal. Biochem. 162:529- 535, 1987) and resuspended in 1 ml. Cells were lysed in a pressure cell and the nucleic acids were removed by protamine sulphate precipitation (0.4% final concentration) . 200μl samples of these extracts were

loaded onto a Superose 6 column (Pharmacia LKB) and the proteins eluted with a Tris-maleate/DTT buffer (67 mM Tris-maleate, 42 mM MgCl2. 400 mM NH ₄ C1, 5 mM DTT, pH 7.2). The NPT-II assays measured phosphoryl- ation of kanamycin using a dot blot method to P81 paper essentially as described by Radke e_£ si. , Theor. Appl. Genet. 11:685-694, 1988. Spots were located by autoradiography and activities were quantified by excising radioactive P81 spots and counting in a universal liquid scintillation cocktail (Scintiverse-II; .Fisher Scientific). GUS activity was measured spectrophotometrically using p-nitrophenyl-β-D-glucuronide as substrate, or by histochemical localization as described (Jefferson, Plant Mol. Biol. Rep. 1:387-405, 1987). Panel A of Figure 2 shows a chromatogram for strain DH5α carrying plasmids pGKl and pGS39, expressing gus and npt-II respectively. Panel B shows a chromatogram for strain DH5α carrying pGKK14, which has the fused gus: :npt-II gene. The column was calibrated with size standards (Pharmacia or BioRad) and sizes in kilodaltons are shown by the arrows: 440, ferritin; 232, catalase; 43, ovalbumin; 25, chymotrypsinogen A; 1.3, vitamin B12. Protein concentrations were determined as described (Jefferson e_t si - , Proc. Nat. Acad. Sci. USA 83:8447-8451. 1986). The relative values of 1 in Panel.A correspond to 116,816 cpm (NPT-II assay) and an OD415 of 3.25 (GUS assay, inclusive of a multiplication factor for absorbance) ; Panel B values were 17,424 cpm (NPT assay) and OD415 1.24 (GUS).

Immunoblot analysis of E. coli-produced GUS and GUS::NPT-II fusion proteins

Extracts from £__. coli were denatured and electrophoresed in a 0.1% SDS 7.5% PAGE (SDS-PAGE; Laemmli, Nature 222:680-685, 1970) before electro- blotting to nitrocellulose membranes (Burnette, A. Anal. Biochem. 112:195-203, 1981). Protein estima¬ tions employed the dye-binding method (Bradford, Anal. Biochem. 72:248-254, 1976). The blots were screened for immunoreactive bands with polyclonal antibodies against GUS (Clontech, San Francisco) or NPT-II (5*-3' Inc., Cleveland) using goat anti-rabbit alkaline phosphatase conjugate as secondary antibody (ProMega Biotec, Madison). Figure 3 shows the immunoblots. Lane 1, DH5α[pBI-403A] ; 2, DH5α[pBI-403B] ; 3, DH5α[pGKl] ; 4, DH5α; 5, 10 ng purified GUS. Size markers used were from BRL Life Technologies (not shown) and are indicated in kilodaltons.

Gel permeation chromatography of plant extracts

Gel permeation chromatography was performed on extracts (A) from transgenic plants expression gus and npt-II from two independent promoters within a disarmed T DNA, or (B) from a single promoter as a fused gene product. The bifunctional GUS::NPT-II fusion gene generated in £___. coli was fused to a Hindlll-BamHI CaMV 35S transcript promoter fragment from pBI221 (Jefferson e_t _al., EMBO J. 1:3901-3907, 1987) and cloned into pMON806 resulting in pBI405. pMON806 (the kind gift of Dr. Harry Klee, Monsanto Co., St. Louis, MO.) is an Aqrobacterium tumefaciens binary plasmid which incorporates a methotrexate

resistance marker selectable in plants, but which lacks plant-selectable npt-II or qus genes. This construct was transferred by conjugation to &___ tumefaciens strain MP90 (Koncz and Schell, MPl. Gen- Genet. 204:383-396, 1986), and used in transforming tobacco leaf disc explants as described (Horsch e_£ ___!. _# Science 277:1229-1231, 1985). Plant extracts were prepared from 0.5 g of leaf material from a young plant was ground to a fine powder in liquid nitrogen with 0.125 g of polyvinylpolypyrrolidone and extracted with 50 μl of β-mercaptoethanol and 250 μl of buffer containing 100 mM Tris-HCl/0.1% sodium dodecylsulfate/1 mM phenylmethylsulfonylfluoride/5 rr_>: DTT, pH 7.4. The extract was centrifuged at 20,000 X g for 30 min and 200 μl of the supernatant was loaded on a Superose 6 column. The proteins were eluted and the enzymes were assayed as described (Jefferson, Plan MQl. BiPl, Rep_.1:387-405, 1987; Jefferson e_£ il. ₍ EMBO J. 6:3901-3907, 1987). The chromatograms are shown in Figure 4. Size standards were the same as for Figure 2, except aldolase (158 kDa) was also included. The slight shift in elution volumes is due to the use of independent columns. The relative values of 1 for NPT-II and GUS assays in Panel A correspond to 2,000 cpm and an absorbance of 7.68 respectively. Panel B, 39,255 cpm (NPT-II) and 3.67 (GUS) .

Immunodetection of GUS and NPT polvpeotides in leaf extracts from transoenic tobacco

Plant extracts were prepared as described above. The immunoblot is shown in Figure 5. Lanes 1-4, transgenic plants for the GUS::NPT-II fusion; 5, transgenic for independently expression GUS and NPT-II; 6, 10 ng purified £«. coli GUS.

Activity gel assay of GUS in transgenic tobacco

Tobacco protein extracts were prepared as described above with the addition of lOμg/ml leupeptin (Sigma) to the extraction buffer. Aliquots containing lOOμg of protein were incubated in lx SDS reducing buffer at 37°C for 2 minutes, and resolved in a 0.1% SDS 7.5% PAGE gel (Laemmli, Nature 277:680- 685, 1970). Gels were incubated in 0.5mg XGluc per ml, 0.05% NaN3 as described (Jefferson e_£ si . , Proc. Nat. Acad. Sci. USA 83:8447-8451. 1986). Figure 5 is a gel showing GUS activity. The GUS activity was visible as insoluble indigo precipitates in the gel matrix. Lane 1, plant transgenic for the qus: :npt-II fusion gene in pBI-405; lane 2, plant transgenic for native qus from pRJ221 (Jefferson, Plant Mol. Biol. Rep. 1:387-405, 1987); lane 3, purified GUS.

Assembly of a gene encoding both qus and npt activities

The experimental design for generating fusions between the C-terminus of GUS and the N-terminus of NPT-II is illustrated in Figure 1. While protein fusions to the aminό-terminus of both GUS and NPT-II have been reported, NPT-II has been shown refractory to carboxy-terminal fusions (Reiss e_t s . , EMBO J. 2:3317-3322, 1984). Since there existed no precedent for fusions at the C-terminus of GUS, an in vivo approach was used to generate such fusions on the assumption that if stable fusions in the NH2-GUS: :NPT-COOH configuration were possible, they would form and persist in vivo. The gus coding region expressed from an upstream tac promoter was ligated to a npt-II coding region (Beck e_£ si . , Gene 11:327-336, 1982) that was deleted for the first four

codons, including the initiation codon. This arrangement provided a functional qus gene with a termination codon, a gus-derived 3' non-translated "intergenic" region of 61 nt, followed by the truncated coding region of npt-II out of frame,to the qus gene (see Figure 1 and Sequence Listing). Attempts to maintain this construct pGKl (and its derivatives) in the £___. coli gus deletion mutant strain SO200 (Jochimsen si Si. , Mol, gen. Cenet, 143:85-91. 1975) were unsuccessful due to an unexplained high frequency of plasmid loss under non-selective conditions, and therefore DH5α was used for all subsequent studi.es. The multicopy plasmid pGKl afforded overproduction of GUS in £_ _* . coli strain DH5α but did not confer kanamycin resistance on the host (data not shown) . This indicated that there was no spurious transcription and/or translation of the npt-II coding region culminating in kanamycin resistance in vivo.

Since spontaneous genetic deletions are not uncommon, we anticipated to generate in-frame fusions between GUS and NPT-II in pGKl as a rare but identi¬ fiable Km ^R event among a large number of Km ^s bacter¬ ial cells harboring the plasmid. This approach would also provide a solution in the event of transcrip¬ tional termination of qus at or before the intergenic region in pGKl. A further assumption was that only the stable fusion events or those which precisely cleaved the fusion peptide to its component npt-II domain would persist. To confine our further experiments to those that expressed a plasmid- associated Km ^R phenotype, we isolated plasmid DNA from the pooled Km ^R colonies (10 ³ obtained from 10 ¹¹ cells) and re-transformed DH5α, selecting for the

ampicillin resistance marker associated with the vector component of pGKl. The Ap ^R colonies were replica-plated for Km ^R and GUS ⁺ phenotype as scored by filter assay. The overproduction of GUS was further confirmed by enzymatic assay of crude cell¬ ular extracts of 15 Aρ ^R Km ^R GUS ⁺ clones (data not shown). Two such clones (pGKK7 and pGKK14) were then chosen to investigate the molecular basis of the GUS ⁺ NPT-II+ phenotype.

Molecular basis of the αus::npt-II fusion and reconstruction of the fusion in vitro

The region surrounding the junction of gus and npt-II in pGKK7 and pGKK14 was sequenced and compared to that in the parental plasmid, pGKl (Figure 1). DNA sequencing was initially limited to approximately 80 nt on both sides of the qus stop codon on the assumption that the TGA (underlined in Figure 1) would have to be removed from the reading frame by a mutation in the triplet or by an upstream frame- shift. A further assumption was that such a frame- shift, if present, would be closer to the original stop codon in order to retain GUS activity. Like¬ wise, any lesions in the npt-II coding region would not likely extend too far into the gene. The results revealed a single base pair deletion near the 3* end of the qus coding region (A:T, bold type and asterisked at position 1790 in Figure 1; nt 2089; Jefferson e_t si . , Proc, flat, Acafl, Sci, VSA 22:8447-8451, 1986). The relevant segments of the reading frames are shown in Figure 1 and indicates this deletion would frameshift the TGA codon of qus and place the npt-II coding region in-frame to qus. Thus, the deletion in pGKK7 and pGKK14 was predicted

to encode a fusion polypeptide of 885 amino acids encompassing the gus, intergenic and npt-II coding region domains.

The in vivo experiments and subsequent partial sequenced analysis of pGKK7 and pGKK14 did not exclude the possibility of additional mutational events elsewhere on pGKl. To determine that the only relevant change was the deletion of the A:T pair at position 1790, and that it alone was sufficient to confer GUS ⁺ and Km ^R phenotypes, the predicted mutational event was reconstructed from pGKl by site-directed mutagenesis with an 18-mer oligo- nucleotide (OL30; Figure 1 and Seq ID No. 5). Of 100 amp ^r colonies analyzed, approximately 70% were Km ^R in subsequent screening and two such randomly chosen clones, pBI-403A and pBI-403B, were resequenced and shown to contain the predicted A:T deletion (data not shown) . These experiments confirmed our interpretation of the in vivo experiments with regard to the genetic basis of the GUS::NPT-II protein fusion, and the functional equivalents of pBI-403A/B with pGKK7 and pGKK14.

Biochemical and immunolooical evidence for bifunctional GUS::NPT-II protein

Crude extracts of DH5α[pGKK14] were fractionated by molecular sieve chromatography. The results shown in Figure 2 indicated co-migration of all of the GUS activity with approximately 40% of the total NPT-II activity in the extract. The remainder of the NPT-II activity was found in a slower-migrating peak. The control extract from a strain expressing GUS from the parental plasmid pGKl, and NPT-II in trans from a second plasmid pGS39 (derived from pGS38; Selvaraj

and Iyer, J. Bacteriol. 158:580-589. 1984) resolved into two distinct peaks containing independently the GUS and NPT-II activities, effectively ruling out the possibility of a spurious post-translational associa¬ tion of NPT-II and GUS subunits into a quaternary complex. The apparent sizes for these classes of proteins were estimated at 270 kDa for the fusion peak and 25 kDa for the NPT-II peak in the same extract, compared with the control values of 140 kDa and 20 kDa for the unfused GUS and NPT-II respec¬ tively. The 25 kDa protein in the fusion extract may result from secondary translation from an in-frame ATG found at position 1768 with its attendant ribosome binding site-like sequence 5 nt upstream (nt 2067; Jefferson ≤±. Si. , Proc. Nat. Acad. Sci. USA 22:8447-8451, 1986). However, since anti-NPT-II polyclonal antibodies failed to identify a discrete NPT-II polypeptide of lower molecular weight in sonicated extracts (see below and Figure 3), this fraction of NPT-II activity is more likely the result of general proteolytic degradation arising from extended handling of samples for column chroma¬ tography. Notwithstanding, this property of the GUS::NPT-II protein in £__. coli does not detract from its usefulness as discussed below. This constituted evidence for a protein fusion and suggested that the native _£__. coli GUS may be active as a homodimer. Further evidence for expression of the fusion was obtained by immunoblot analysis of expressed polypeptides in DH5α carrying the relevant plasmids.

Immunoblot blot analyses of protein extracts from DH5α, DH5α[pGKl] and DH5c_[pGKK14] using poly¬ clonal antibodies against GUS or NPT-II enzymes showed the presence of the fusion peptide exclusively

in pGKK14-containing strains (Figure 3). The estimated size for the fusion polypeptide of 96 kDa corresponds closely to the sum of GUS (68 kDa; Jefferson e_£ _al., Proc, Fa , Acaflt Sci. USA 22:8447-8451, 1986) and NPT-II (25 kDa, Horsch fit al. , Science 227:1229-1231. 1985). These analyses also showed the presence of discrete smaller antigenic peptide fragments, suggesting some degradation of the GUS polypeptide and the fusion derivative. Activity gels with MUG as the substrate (Jefferson, Plant Mθl. B ol. ReP. 2:387-405, 1987) showed GUS activity in bands of higher M _r than the parental GUS, thus confirming the independent gel filtration and immunoblot experiments (data not shown; see Figure 6 for a related experiment) .

Taken together, the DNA sequence analysis, the co-purification on gel filtration columns of GUS and NPT-II activities as a single protein of predicted fusion molecular weight and the Western analysis argue in favor of expression of a bifunctional fusion gene and peptide in the NH ₂ -GUS::NPT-II-COOH configuration.

Expression of the gus;;npt-II bifunctional gene in plants

An objective of this study was to construct a versatile biochemical and genetic marker for simultaneous selection and screening in transgenic plants using the qus and npt-II genes. The utility of the bifunctional protein was tested in plants by transforming Nicotiana tabacu with pBI405 carrying the qus: :npt-II chimeric gene (Figure 4) in combina¬ tion with the 35S transcript promoter from CaMV. Of 32 transgenic plants selected directly for Km ^R , 28

were found to be GUS ⁺ and in those plants tested, both markers were inherited as a single dominant locus through meiosis. While each positive plant exhibited GUS activity well above the control (plants transgenic for npt-II alone), there was variation in the GUS specific activity among individuals. Such variation could also be seen when plants were transformed with a non-fusion φus marker from pBH21 (data not shown; Jefferson fit .Li., EMBO J. 2:3901- 3907, 1987). Similar differences in gene activity among individual transgenic plants have been attributed to chromosomal position effects (Weising e± _al., Ann. Rev. Genet. 22:421-478. 1988).

Protein extracts from two independent GUS ⁺ Km ^R plants were again subjected to gel filtration (Figure 4) and immunoblot analyses (Figure 5). The protein eluting at about 270 kDa contained both GUS and NPT-II activities, in contrast to the control plants where the proteins were expressed from independent promoters and the size of the unfused GUS was around 140 kDa. Interestingly, the bifunctional protein appeared to chromatograph as a trimer in both £.__ coli and tobacco although, in the absence of sedimentation coefficient data, such chromatographic behavior may merely reflect an exaggerated anisometry of the fusion peptide as a dimer. Alternatively, this difference may be due to the modified carboxy terminus of GUS in the fusion protein — an observation that is reminiscent of the different oligomeric structures of β-galactosidase mutants in E. coli (Fowler and Zabin, Science 154:1027-1029. 1966) . It would be informative to dissect the functional domains of GUS and NPT-II from a practical viewpoint as well as with regard to studying the

structure-function relationships of the two catalytic domains — particularly as it relates to a novel fusion between two proteins of different native quaternary structures. Unlike protein extracts from E. coli expressing the GUS::NPT-II fusion gene under the control of a tac promoter, only 10% of a smaller NPT-II protein was observed in transgenic plants expressing the fusion (see Panel B in Figures 2 and 4). Detection of GUS and NPT-II epitopes in a >97 kDa polypeptide fraction confirmed expression of GUS::NPT-II as a fusion protein in transgenic leaf tissues, and indicated a relative stability of the protein in these plants (Figure 5). Since trans¬ lation of polycistronic rriRNA is limited to viral and organellar messages in plants (Bonneville fit al. , Mosaic V rus Cell 1!:1135-1143, 1989); Angenon fit Si . , Mol ^' ec. Cell. Biol. 9:5676-5684. 1989), we surmise that the smaller NPT-II protein represents a product of the fusion and not a result of reinitia¬ tion within the qus: :npt-II mRNA. Thus, the Km ^R phenotype can be manifested only upon expression of the intact GUS::NPT-II fusion peptide.

Attempts were made to obtain zy ograms using XGluc as the substrate for both unfused or fused GUS after SDS-PAGE of leaf extracts in a 7.5% gel (Figure 6). Incubation of the samples in a loading buffer (Jefferson _f it si. , Proc. Nat. Acad. Sci. USA 22:8447-8451, 1986) at 37°C for 2 min prior to electrophoresis showed one prominent larger activity band for the fusion at >200 kDa position and a smaller one for the unfused GUS (and the purified GUS) at >97 kDa. Prolonged., incubation at 37°C abolished the GUS activity, although a polypeptide of predicted size could be seen under these conditions

as shown in the immunoblots of Figure 5. We surmise that under experimental conditions permissive for retaining the activity, the protein is not dissociated to yield subunits as reflected in the overall slower migration (Figure 6). Thus, the results shown in Figures 4-6 clearly indicate stable expression of the bifunctional protein in plants. Histochemical analysis of transgenic tobacco plants generated using the qus: :npt-II fusion gene for selection showed patterns of expression similar to 35S promoter-driven qus constructs (Jefferson gt al.. 1987). The qus: :npt-II fusion gene has been similarly used to generate transgenic plants of Brassica naous and Arabidoosis thaliana. and fusion- based plasmid constructs have been successfully employed in assays of transient gene expression in electroporated protoplasts of tobacco. White Spruce (Picea qlauca) and maize.

Analysis of Transgenic Plants

Analysis of 1000 transgenic plants arising from the use of the qus: :npt-II fusion gene was performed, and the results graphically summarized as a frequency histogram in Figure 8. This study identified 243 plants which expressed some degree of tissue and/or developmental specificity of β-GUS expression. Figure 8 indicates a 24-precent frequency of tissue-specific phenotypes and is in close agreement with previous results from analysis of a group of 200 transgenic plants (results not shown). All singular tissue-specific phenotypes sought (anther, leaf, ovary, petal, root, seed, stem) were found. In addition, most binary combinations of flower or somatic tissue-specificity were identified, including

[anther+ovary] , [anther+petal] , [root+leaf], [root+stem] and [stem+leaf] . Only one binary flower tissue combination was not identified, [petal+ovary] .

Assembly of g lu*;.npt Bifunctional Fusion gene

The construct is based upon the qus: npt bifunctional fusion maker gene described above.

A near full-length coding region of the firefly (P. pyralis) luciferase gene was defined and amplified from a co ercially available plasmid (pLUC; Promega, Madison WI) using PCR and oligonucleotides BC175 (Sequence ID No. 6) and BC176 (Sequence ID No. 7). This region was substituted for the NcoI-SacI GUS domain of the qus: :npt bifunctional construct, thus generating the lux::npt combination. The complete nucleotide sequence of the chimeric lux::npt gene is given in Sequence ID No. 8, along with its corresponding deduced amino acid sequence in Sequence ID No. 9. This construct was used to transform E. coli cells, initially selecting for the expression of kanomycin resistance from the downstream npt domain. Kanomycin-resistant colonies were cultured and extracts assayed qualitatively for the expression of light-emitting activity using a liquid scintillation counter and a commercial assay kit (Promega, Madison WI). Among 4 colonies, three were positive for luciferase activity.

Equivalents

Those skilled in the art will recognize, or be able to ascertain using no more than routine experi¬ mentation, numerous equivalents to the specific procedures described herein. Such equivalents are considered to be within the scope of this invention and are covered by the following claims.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: Crosby, William L. Datla, Raju S.S. Hammerlindl, Joesph K. Selvaraj, Gopalan

(ii) TITLE OF INVENTION: BIFUNCTIONAL GENETIC MARKERS

(iii) NUMBER OF SEQUENCES: 9

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: LAHIVE & COCKFIELD

(B) STREET: 60 State Street Suite 510

(C) CITY: Boston

(D) STATE: MA

(E) COUNTRY: USA

(F) ZIP: 02109

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: ASCII TEXT

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER: PCT

(B) FILING DATE:

(C) CLASSIFICATION:

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: DeConti, Giulio A.

(B) REGISTRATION NUMBER: 31,503

(C) REFERENCE/DOCKET NUMBER: NPI-002PC

(ix) TELECOMMUNICATION INFORMATION: (A) TELEPHONE: (617) 227-7400

(2) INFORMATION FOR SEQ ID NO:1:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 123 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

SUBSTITUTE SHEET

(ii) MOLECULE TYPE: cDNA

(ix) FEATURE:

(A) NAME/KEY: CDS

(B) LOCATION: 1..123

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:l:

ATG AAC TTC GGT GAA AAA CCG CGC AGG GAG GCA AAC AAT GAA TCA ACA .£ Met Asn Phe Gly Glu Lys Pro Arg Arg Glu Ala Asn Asn Glu Ser Thr

1 5 10 15

ACT CTC CTG GCG CAC CAT CGT CGG CTA CAG CCT CGG GAA TTG CTA CCG r-_ Thr Leu Leu Ala His His Arg Arg Leu Gin Pro Arg Glu Leu Leu Pro 20 25 30

AGC TCG AGC TTG GAT GGA TTG CAC GCA ::_

Ser Ser Ser Leu Asp Gly Leu His Ala

35 40

(2) INFORMATION FOR SEQ ID NO: 2 :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 41 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2 :

Met Asn Phe Gly Glu Lys Pro Arg Arg Glu Ala Asn Asn Glu Ser Thr 1 5 10 15

Thr Leu Leu Ala His His Arg Arg Leu Gin Pro Arg Glu Leu Leu Pre 20 25 30

Ser Ser Ser Leu Asp Gly Leu His Ala 35 40

(2) INFORMATION FOR SEQ ID NO: 3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 124 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

SUBSTITUTE SHEET

(ii) MOLECULE TYPE: cDNA

(ix) FEATURE:

(A) NAME/KEY: CDS

(B) LOCATION: 1..42

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:

ATG AAC TTC GGT GAA AAA CCG CAG CAG GGA GGC AAA CAA TGA ATCAλCλ Met Asn Phe Gly Glu Lys Pro Gin Gin Gly Gly Lys Gin * 1 5 10

ACTCTCCTGG CGCACCATCG TCGGCTACAG CCTCGGGAAT TGCTACCGAG CTCGAGCTTG 109

GATGGATTGC ACGCA i:.

(2) INFORMATION FOR SEQ ID NO:4:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 13 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4:

Met Asn Phe Gly Glu Lys Pro Gin Gin Gly Gly Lys Gin

1 5 10

(2) INFORMATION FOR SEQ ID NO: ^' 5:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 18 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5: GAAAAACCGC GCAGGGAG 1_

SUBSTITUTE SHEET

(2) INFORMATION FOR SEQ ID NO:6:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: CCGGATCCAA ATGGAAGACG 2_

(2) INFORMATION FOR SEQ ID NO:7:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 26 base pairs

(B) TYPE: nucleic acid .

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: GGAGCTCGGC AATTTGGACT TTCCGC 26

(2) INFORMATION FOR SEQ ID NO:8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 2448 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(ix) FEATURE:

(A) NAME/KEY: CDS

(B) LOCATION: 1..2448

(ix) FEATURE:

(A) NAME/KEY: misc_feature

(B) LOCATION: 1652..1659

(D) OTHER INFORMATION: /function-- "Sad site"

SUBSTITUTE SHEET

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 8:

ATG GAA GAC GCC AAA AAC ATA AAG AAA GGC CCG GCG CCA TTC TAT CC7 4- Met Glu Asp Ala Lys Asn lie Lys Lys Gly Pro Ala Pro Phe Tyr Pre 1 5 10 15

CTA GAG GAT GGA ACC GCT GGA GAG CAA CTG CAT AAG GCT ATG AAG AGλ - ^{' ■'} Leu Glu Asp Gly Thr Ala Gly Glu Gin Leu His Lys Ala Met Lys Arc 20 25 30

TAC GCC CTG GTT CCT GGA ACA ATT GCT TTT ACA GAT GCA CAT ATC GAG 144 Tyr Ala Leu Val Pro Gly Thr lie Ala Phe Thr Asp Ala His lie Glu 35 40 45

GTG AAC ATC ACG TAC GCG GAA TAC TTC GAA ATG TCC GTT CGG TTG GCA 192 Val Asn lie Thr Tyr Ala Glu Tyr Phe Glu Met Ser Val Arg Leu Ala 50 55 60

GAA GCT ATG AAA CGA TAT GGG CTG AAT ACA AAT CAC AGA ATC GTC GTλ 240 Glu Ala Met Lys Arg Tyr Gly Leu Asn Thr Asn His Arg lie Val Val

65 70 75 80

TGC AGT GAA AAC TCT CTT CAA TTC TTT ATG CCG GTG TTG GGC GCG TTA 288 Cys Ser Glu Asn Ser Leu Gin Phe Phe Met Pro Val Leu Gly Ala Leu 85 90 95

TTT ATC GGA GTT GCA GTT GCG CCC GCG AAC GAC ATT TAT AAT GAA CGT 33c Phe lie Gly Val Ala Val Ala Pro Ala Asn Asp lie Tyr Asn Glu Arg 100 105 110

GAA TTG CTC AAC AGT ATG AAC ATT TCG CAG CCT ACC GTA GTG TTT GTT 384 Glu Leu Leu Asn Ser Met Asn lie Ser Gin Pro Thr Val Val Phe Val 115 120 125

TCC AAA AAG GGG TTG CAA AAA ATT TTG AAC GTG CAA AAA AAA TTA CCA 432

Ser Lys Lys Gly Leu Gin Lys lie Leu Asn Val Gin Lys Lys Leu Pre 130 135 140

ATA ATC CAG AAA ATT ATT ATC ATG GAT TCT AAA ACG GAT TAC CAG GGA 4EC lie lie Gin Lys lie lie lie Met Asp Ser Lys Thr Asp Tyr Gin Gly 145 150 155 160

TTT CAG TCG ATG TAC ACG TTC GTC ACA TCT CAT CTA CCT CCC GGT TTT £2- Phe Gin Ser Met Tyr Thr Phe Val Thr Ser His Leu Pro Pro Gly Phe 165 170 175

AAT GAA TAC GAT TTT GTA CCA GAG TCC TTT GAT CGT GAC AAA ACA A I _ ^" Asn Glu Tyr Asp Phe Val Pro Glu Ser Phe Asp Arg Asp Lys Thr He 180 185 190

SUBSTITUTE SHEET

£24

72C

76 £

eie

960

1008

1104

SUBSTITUTE SHEET

.29'

1392

144'

148E

158-

:eε:

SUBSTITUTE SHEET

CAG GGG CGC CCG GTT CTT TTT GTC AAG ACC GAC CTG TCC GGT GCC CTG 1824 Gin Gly Arg Pro Val Leu Phe Val Lys Thr Asp Leu Ser Gly Ala Leu 595 600 605

AAT GAA CTG CAG GAC GAG GCA GCG CGG CTA TCG TGG CTG GCC ACG ACG 1872 Asn Glu Leu Gin Asp Glu Ala Ala Arg Leu Ser Trp Leu Ala Thr Thr 610 615 620

GGC GTT CCT TGC GCA GCT GTG CTC GAC GTT GTC ACT GAA GCG GGA AGG 192C Gly Val Pro Cys Ala Ala Val Leu Asp Val Val Thr Glu Ala Gly Arg 625 630 635 640

GAC TGG CTG CTA TTG GGC GAA GTG CCG GGG CAG GAT CTC CTG TCA TCT 1968 Asp Trp Leu Leu Leu Gly Glu Val Pro Gly Gin Asp Leu Leu Ser Ser 645 650 655

CAC CTT GCT CCT GCC GAG AAA GTA TCC ATC ATG GCT GAT GCA ATG CGG 2016 His Leu Ala Pro Ala Glu Lys Val Ser He Met Ala Asp Ala Met Arg 660 665 670

CGG CTG CAT ACG CTT GAT CCG GCT ACC TGC CCA TTC GAC CAC CAA GCG *2064 Arg Leu His Thr Leu Asp Pro Ala Thr Cys Pro Phe Asp His Gin Ala 675 680 685

AAA CAT CGC ATC GAG CGA GCA CGT ACT CGG ATG GAA GCC GGT CTT GTC 2112 Lys His Arg He Glu Arg Ala Arg Thr Arg Met Glu Ala Gly Leu Val 690 695 700

GAT CAG GAT GAT CTG GAC GAA GAG CAT CAG GGG CTC GCG CCA GCC GAA 2160 Asp Gin Asp Asp Leu Asp Glu Glu His Gin Gly Leu Ala Pro Ala Glu 705 710 715 720

CTG TTC GCC AGG CTC AAG GCG CGC ATG CCC GAC GGC GAG GAT CTC GTC 22C8 Leu Phe Ala Arg Leu Lys Ala Arg Met Pro Asp Gly Glu Asp Leu Val 725 730 735

GTG ACC CAT GGC GAT GCC TGC TTG CCG AAT ATC ATG GTG GAA AAT GGC 2256 Val Thr His Gly Asp Ala Cys Leu Pro Asn ^' He Met Val Glu Asn Gly 740 745 750

CGC TTT TCT GGA TTC ATC GAC TGT GGC CGG CTG GGT GTG GCG GAC CGC 2304 Arg Phe Ser Gly Phe He Asp Cys Gly Arg Leu Gly Val Ala Asp Arg 755 760 765

TAT CAG GAC ATA GCG TTG GCT ACC CGT GAT ATT GCT GAA -GAG CTT GGC 2352 Tyr Gin Asp He Ala Leu Ala Thr Arg Asp He Ala Glu Glu Leu Gly 770 ^■*• 775 780

GGC GAA TGG GCT GAC CGC TTC CTC GTG CTT TAC GGT ATC GCC GCT CCC 240C Gly Glu Trp Ala Asp Arg Phe Leu Val Leu Tyr Gly He Ala Ala Pre 785 790 795 800

SUBSTITUTE SHEET

GAT TCG CAG CGC ATC GCC TTC TAT CGC CTT CTT GAC GAG TTC TTC TA . _ 1 i Asp Ser Gin Arg He Ala Phe Tyr Arg Leu Leu Asp Glu Phe Phe 805 810 815

(2) INFORMATION FOR SEQ ID NO: 9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 815 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:9:

Met Glu Asp Ala Lys Asn He Lys Lys Gly Pro Ala Pro Phe Tyr Pro

1 5 10 15

Leu Glu Asp Gly Thr Ala Gly Glu Gin Leu His Lys Ala Met Lys Arg 20 25 30

Tyr Ala Leu Val Pro Gly Thr He Ala Phe Thr Asp Ala His He Glu 35 40 45

Val Asn He Thr Tyr Ala Glu Tyr Phe Glu Met Ser Val Arg Leu Ala 50 55 60

Glu Ala Met Lys Arg Tyr Gly Leu Asn Thr Asn His Arg He Val Val 65 70 75 80

Cys Ser Glu Asn Ser Leu Gin Phe Phe Met Pro Val Leu Gly Ala Leu 85 ' 90 95

Phe He Gly Val Ala Val Ala Pro Ala Asn Asp He Tyr Asn Glu Arg 100 105 110

Glu Leu Leu Asn Ser Met Asn He Ser Gin Pro Thr Val Val Phe Val 115 120 125

Ser Lys Lys Gly Leu Gin Lys He Leu Asn Val Gin Lys Lys Leu Pro 130 135 140

He He Gin Lys He He He Met Asp Ser Lys Thr Asp Tyr Gin Gly 145 150 155 160

Phe Gin Ser Met Tyr Thr Phe Val Thr Ser His Leu Pro Pro Gly Phe 165 170 175

SUBSTITUTE SHEET

Asn Glu Tyr Asp Phe Val Pro Glu Ser Phe Asp Arg Asp Lys Thr He 180 185 190

Ala Leu He Met Asn Ser Ser Gly Ser Thr Gly Leu Pro Lys Gly Val 195 200 205

Ala Leu Pro His Arg Thr Ala Cys Val Arg Phe Ser His Ala Arg Asp 210 215 220

Pro He Phe Gly Asn Gin He He Pro Asp Thr Ala He Leu Ser Val 225 230 235 240

Val Pro Phe His His Gly Phe Gly Met Phe Thr Thr Leu Gly Tyr Leu 245 250 255

He Cys Gly Phe Arg Val Val Leu Met Tyr Arg Phe Glu Glu Glu Leu 260 265 270

Phe Leu Arg Ser Leu Gin Asp Tyr Lys He Gin Ser Ala Leu Leu Val 275 280 285

Pro Thr Leu Phe Ser Phe Phe Ala Lys Ser Thr Leu He Asp Lys Tyr 290 295 300

Asp Leu Ser Asn Leu His Glu He Ala Ser Gly Gly Ala Pro Leu Ser 305 310 315 320

Lys Glu Val Gly Glu Ala Val Ala Lys Arg Phe His Leu Pro Gly He 325 330 335

Arg Gin Gly Tyr Gly Leu Thr Glu Thr Thr Ser Ala He Leu He Thr 340 345 350

Pro Glu Gly Asp Asp Lys Pro Gly Ala Val Gly Lys Val Val Pro Phe 355 360 365

Phe Glu Ala Lys Val Val Asp Leu Asp Thr Gly Lys Thr Leu Gly Val 370 375 380

Asn Gin Arg Gly Glu Leu Cys Val Arg Gly Pro Met He Met Ser Gly 385 390 395 400

Tyr Val Asn Asn Pro Glu Ala Thr Asn Ala Leu He Asp Lys Asp Gly 405 410 415

Trp Leu His Ser Gly Asp He Ala Tyr Trp Asp Glu Asp Glu His Phe 420 425 430

Phe He Val Asp Arg Leu Lys Ser Leu He Lys Tyr Lys Gly Tyr Glr. 435 440 445

SUBSTITUTE SHEET

Val Ala Pro Ala Glu Leu Glu Ser He Leu Leu Gin His Pro Asn He 450 455 460

Phe Asp Ala Gly Val Ala Gly Leu Pro Asp Asp Asp Ala Gly Glu Leu 465 470 475 48C

Pro Ala Ala Val Val Val Leu Glu His Gly Lys Thr Met Thr Glu Lys 485 490 495

Glu He Val Asp Tyr Val Ala Ser Gin Val Thr Thr Ala Lys Lys Leu 500 505 510

Arg Gly Gly Val Val Phe Val Asp Glu Val Pro Lys Gly Leu Thr Gly 515 520 525

Lys Leu Asp Ala Arg Lys He Arg Glu He Leu He Lys Ala Lys Lys 530 535 540

Gly Gly Lys Ser Lys Leu Pro Ser Ser Ser Leu Asp Gly Leu His Ala

545 550 555 560

Gly Ser Pro Ala Ala Trp Val Glu Arg Leu Phe Gly Tyr Asp Trp Ala 565 570 575

Gin Gin Thr He Gly Cys Ser Asp Ala Ala Val Phe Arg Leu Ser Ala 580 585 590

Gin Gly Arg Pro Val Leu Phe Val Lys Thr Asp Leu Ser Gly Ala Leu 595 600 605

Asn Glu Leu Gin Asp Glu Ala Ala Arg Leu Ser Trp Leu Ala Thr Thr 610 615 620

Gly Val Pro Cys Ala Ala Val Leu Asp Val Val Thr Glu Ala Gly Arg 625 630 635 640

Asp Trp Leu Leu Leu Gly Glu Val Pro Gly Gin Asp Leu Leu Ser Ser 645 ^' 650 655

His Leu Ala Pro Ala Glu Lys Val Ser He Met Ala Asp Ala Met Arg 660 665 670

Arg Leu His Thr Leu Asp Pro Ala Thr Cys Pro Phe Asp His Gin Ala 675 680 685

Lys His Arg He Glu Arg Ala Arg Thr Arg Met Glu Ala Gly Leu Val 690 695 700

SUBSTITUTE SHEET

Asp Gin Asp Asp Leu Asp Glu Glu His Gin Gly Leu Ala Pro Ala Glu

705 710 715 720

Leu Phe Ala Arg Leu Lys Ala Arg Met Pro Asp Gly Glu Asp Leu Val 725 730 735

Val Thr His Gly Asp Ala Cys Leu Pro Asn He Met Val Glu Asn Gly 740 745 750

Arg Phe Ser Gly Phe He Asp Cys Gly Arg Leu Gly Val Ala Asp Arg 755 760 765

Tyr Gin Asp He Ala Leu Ala Thr Arg Asp He Ala Glu Glu Leu Gly 770 775 780

Gly Glu Trp Ala Asp Arg Phe Leu Val Leu Tyr Gly He Ala Ala Pro

785 790- 795 800

Asp Ser Gin Arg He Ala Phe Tyr Arg Leu Leu Asp Glu Phe Phe 805 810 815

SUBSTITUTE SHEET