Organization and expression of leghaemoglobin genes

10 downloads 0 Views 820KB Size Report
J., John, C. K., Royarao, J. C. & Lim, B., eds.), pp. 253-281, University of ... Vema, D. P. S., Ball, S., GuCrin, C. & Wanamaker, L. (1979) Biochemistry 18,476-483.
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

Organization and expression of leghaemoglobin genes N. BRISSON,A. POMBO-GENTILE, AND 6). P.S. VERMA' Deprtment of Biology, McCill University, Montreal, P.Q.,Canada H3A 1BI Received July 13, 1981 Bpisson, N.,Pombo-Gentile, A. & Vema, D. P. S. (1982) Organization and expression of leghaemoglobin genes. Can. J. Biochem. 68,272-278 kghaemoglobin genes in soybean (Clycine m a ) are present as a moderately reiterated family of sequences. Since there sue identical restriction site patterns of these sequences in DNA isolated from leaf, root, or nodule tissue, the data suggest that no major changes in the organization or methylation of leghaemoglobin genes occur during their induction. Cloned soybean leghaemoglobin-cDNA cross hybridized with RNA fiom root nodules of kidney bean (Phaseolus vulgaris), and to a lesser extent, of pea (Pisurn sativurn) indicating sequence homology in the leghaernoglobin genes of these species. Hybridizationto the genomic DNA restriction fragments of two other species, Glycine soja and Vicia faba, also indicated interspecies sequence homologies. Several restriction fragments appear to be common to all the species examined. The induction of these genes occurs following infection of the plant by Rhizsbiurn m d is independent of the appearance of nitrogenase activity in the nodules. The level of expression is, however, influenced by various mutations in Whizobiurn that result in the development of ineffective (nornitrogen fixing) nodules. Brisson, N., Pombo-Gentile, A. & Vema, D. P. S. (1982) Organization and expression of leghaemoglobin genes. Can. J. Bischem. 60,272-278 hs gi?nes de la kghkmoglobine dans la Rve soya (Clycine mar) sont prbsents sous fome d'une famille de s6quences madCr6ment rk@t&s. C o m e les profils des sites de restriction de ces skquences sont identiques dans le DNA isolk des feuilles, des raines ou du tissu nodulaire, les donnees sugg2rent qu'il ne se produit aucun chmgement rnajeur dans lvorganisationou la mkthylation des gbnes de la legh6moglobine durant leur induction. La cDNA clon6, correspondant h la leghkmoglobine de la five soya, s'kybride bien avec le RNA des nodules de la racine du haricot (Bhaseolus vulgaris) et, A un de@ moindre, des p i s (Pisum sativurn); il y a donc homologie de skquences dans les gbnes de la legh6moglsbine de ces e s w e s . L'hybridation avec les fragments de restriction du DNA gCnsmique de deux autres esphces, Glycine so@ et Viciafaba indique aussi une homologie skquentielle en&e ces es@ces. k profil gknkral des g&nesde la leghkmoglobine semble plutbt similaire d'une esp&ce A I'autre. I1 y a induction de ces g%nessuite h l'infection de la plante par Rhizobium et elle est indkpendante de l'appaadtisn de I'activig nitrogknasique dans les nodules. Cependant, le taux d'expression est influence par diverses mufations chez Whizobiurn, lesquelles mhnent le dCveloppement de nodules inefficients (incapables de fixer I'azote). [Traduit par le journal]

Introduction A myoglobin-like haernoprotein, leghaemoglobin, present in nitrogen-fixing root nodules of leguminous plants diverged from animal globins more than $08 million yeas ago (1). Major amino acid changes that occurred during evolution (2) resulted in the alteration of immunological and some other physicochemicd properties of this molecule (2, 3). However, it appears to have retained its basic molecular structure to perform the function of oxygen binding. The presence of this unique molecule in legume root nodules developed because the association of Rhizobiurn sp. is obligatory for the process of symbiotic nitrogen fixation. The activity of the enzyme system responsible for nitrogen fixation has been found to be proportional to the amount of leghaemoglobin present particularly during the early stages of the nodule development (4, 5). Leghaemoglobin is B ~ u atoRw~ ~h correspondence should be addressed.

induced several days prior to the appearance of nitrogenase activity (6-8) and represents 20-30% of the totd cytoplasmic proteins in mature nodules. Located in the cytoplasm of the infected host cell (8), it transports sufficient O2to support high respiratory rates and ATB production in the bacteroids while at the same time maintains an O2 tension below that which would be damaging to the nitrogenase enzyme complex (5, 10). Nodules of most legumes appear to contain more than one species of leghaemoglobins. This mdti~licitywithin a species and the close relationship m o n g leghaemoglobins of related legumes suggests that genes for this group of proteins are well conserved in evolution. That various leghaemoglobins are distinct gene products is evident from the protein sequence data (1 1- 13) as well as from the in vitro translation of nodule mRNA which produces polypeptides corresponding to distinct leghaemoglobin species (6). Whether the different leghaemoglobins have distinct

0008-401$/$2/030272-07W1 .m/o 61982 National Research Council of Canah/ConseiX national de recherches du Canada

BRISSON ET AL.

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

functions in root nodules is not known but they appear to be induced at different times during root nodule development and exhibit different rates of turnover (6). The symbiotic nature of the biosynthesis of leghaemoglobin (haem is made by the Rhizobium while the apoprotein is synthesized by the plant) has recently been supported by genetic evidence showing that infection with Rhizobim mutants unable to synthesize haem always results in ineffective (nonnitrogen fixing) nodules lacking hnctional leghaemoglobin (14). However, ineffective nodules contain significant amounts of leghaemoglobin A which can be translated in vitrs into an apoprotein (7) suggesting that induction of leghaemoglobin is independent of the ability of nodules to fix nitrogen. Using purified complementary DNA to leghaemoglobin W A we estimated the number of leghaemoglobin genes in soybean. The saturation hybridization kinetics with DNA either from uninfected plants or from nodule tissue yielded a value of about 40 copies per haploid genome (15). This observation was recently strengthened by direct demonstration of the complexity of the leghaemoglobin gene family in soybean genome using a molecular clone (16). We present here evidence that the leghaemoglobin genes are sufficiently conserved to allow cross hybridization of the leghaemoglobin cDNA clone with not only other s p i e s but also with other genera. Its molecular organization appears to be similar in each species. There is no apparent correlation between the extent of methylation in genomic DNA and the induction of the leghaemoglobin genes.

273

was obtained from W. Newcomb, Queen's University and 19. phaseoli from T. Hall, University of Wisconsin. Late logphase cultures were harvested and the bacterial pellets were suspended in 5% sucrose for inoculation of seeds.

DNA extraction a d digestion DNA was isolated from soybean nodules and uninf~ted tissues (root, leaves, and embryonic axes), kidney bean, pea, and wild soybeans (Glycine javonica and G. soja) seedlings using the cesium chloride - ethidiurn bromide banding procedure (18). For nodule DNA only the upper band from cesium chloride gradient was collected; the lower bmd consists of bacteroid DNA of higher density. Plasmid DNA was prepared by the cleared lysate method (19) and purified by two cycles of isspycnic ethidium bromide - cesium chloride centrifugation. Phage DNA containing leghaemoglobin sequences (Ch4 GmLbl 1 and Ch4 G d M , see Ref. 16) was prepared from large-scale lysates according to Blattner et al. (20). DNA digestions were carried out in conditions furnished by the suppliers. Following electrophoresis in 1.2% agarose gels (horizontal submerged), the DNA was transferred to nitrocellulose paper (BA85, Schleicher and Schuell) (21) and hybridized to radioactively labelled DNA probes.

Nick translation and hybridization DNA was labelled in vitro by nick translation (22). In a typical reaction, 280ng of DNA was labelled to a specific activity of 1 x 10' to 2 X l O % ~ m / ~byg using a mixture of a-["PIBCTP and dATP (specific activity 2 x 10' to 3 x 10' Ci/mmol(l Ci = 37 GBq), Amersham Corp.). Poly(A) + RNAs from soybean, kidney bean md pea nodules, uninfected roots, and wheat embryos were isolated as described previously (17), electrophoresed on methyl mercury hydroxide gels according to Bailey and Davidson (23) and transfewed to diazotized (DBM) papers (24). Hybridization was carried out at 42°C for 16 h in sealed plastic bags as described previously Materials and methods (16). Autoradiograms were obtained by exposing the X-ray Plant material Soybean (Glycine max L. cv. Prize and cv. Hxcor), Pea film at -70°C for 24-72 h using Dupont intensifying screens. For the filter disc hybridization, poly(A)+ RNA isolated (fisum sativum), and kidney bean (Phaseolus vulgaris) seeds were inoculated with their respective Rhizobium strains (see from different tissues was titrated by %-labelled poly(U) below). Seeds of Glycine soja (obtained from N. Neilson, hybridization (25) and 0.05 to 10 pg of W A was bound to h r d u e University), Glycine jovonica (obtained from ICRIS- 1l-mm discs of DBM paper. Hybridization was at 42'C for AT, Hydrabad, India), and Vicia faba were geminated for 1 48 h with 3 x 106 cpm (specific activity 1.5 x lo6cpm/bg week without inoculation. Plants were grown in a controlled DNA) of a nick-translate8plasmid pLbl containing part of the environment chamber with a photoperiod of 16 h at a day leghaemoglobin sequence (16). Linearity of the assay was temperature of 28°C and a night temperature of 22°C. They checked by using filters with increasing amounts of W A . were watered with nitrogen-free nutrient medium (17). The Results and discussion soybean plants bearing ineffective (unable to fix nitrogen) nodules formed by Rhizobium japreicum mutant strains were DNA methylation and organization of leghaemoglobin genes in soybean kept in a separate growth chamber to prevent possible The interest in methylation of eukaqotic DNA has contarnination by wild-type strains. The nodules were hawested 21 days after inocuktion and stored under liquid nitrogen been stimulated by its possible role in the control of gene until used. activity. The DNA of higher organisms contains Rhizotsium strains Rhizobisrmjaponicsrm strains 61A7Q(wild type, effective), 61A24 (wild type, ineffective) were obtained from Nitragin Co. Milwaukee and strain SM5 was kindly provided by W. Brill, University of Wisconsin. Rhizobim legumimsarurn

5-methyl cytosine (m5C), most often in the sequence C-G (26). Restriction endonucleases whose recognition sequences include the sequence C-G (e.g., HpdH and Hhd), are thus valuable tools for mapping m ' ~in DNA.

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

274

CAN. J. BIWKE :M.VOL. 60, 1982

To find out if any correlation exists between the Tlae similarity of HhaI with HpaII both for the pattern expression of soybean leghaemoglobin genes and their obtained and for the C-G sequence in the recognition site state of methylation, DNA was isolated from various suggests that HhaI sites may also be methylated in the tissues and digested with a series of restriction enzymes. high molecular weight fraction. Thus, it appears that at Figure 1 shows the ethidium bromide staining pattern of least for bulk of the DNA sequences in these plant digests of leaf, embryonic axes, and nodule DNAs from tissues, the total amount of methylation remains almost soybean as displayed by agarose gel electrophoresis. constant. DNA from embryonic axes digested with EcoRI and A recombinant plasmid (pLbl) containing a portion of Hind 111gives a pattern of fragments with a model size of one of the soybean leghaemoglobin coding sequences about 10kb. MboI generates fragments with a modal (16) was used to study the genomic organization of size of about 1.7 kb,while HhaI and HpaII cleave the leghaemoglobin gene family. Following digestion with DNA to a heterogeneous mixture of fragment sizes with several restriction enzymes and electrophoresis through a large fraction remaining almost undigested. The an agarose gel, DNA was blotted onto a nitrocellulose presence of this high molecular weight DNA can be paper which was then hybridized to 32~-labelledpLbl explained if the sites for these enzymes are methylated, as described before (16). Since all soybean leghaemoor if few sites exist. A significant fraction of both nodule globins exhibit extensive m i n o acid sequence homoloand embryo DNA which is not digested by HpaII is gies (11- 13), this probe is expected to cross hybridize reduced in size by MspI (an isoschizomer of HpaII to aU genomic leghaemoglobin sequences. Figure 2 which cleaves C-C-G-G irrespective of methylation) shows the pattern of hybridization of pLbl to Hind 111, suggesting that this DNA does contain HpaII sites, but MspI, and HpaII digested soybean DNA. It is apparent methylation prevents their cleavage by the enzyme. that leghaemoglobin sequences are represented in the However, comparison of the MspI pattern with that of soybean genome as a moderately reiterated family N o 1 shows that the occurrence of the HpaII sites is of genes (see also Ref. 16). Based upon the size much lower than expected for a four base pair sequence. of restriction fragments containing leghaemoglobin sequences, it appears that this family is spread in almost 50 kb of DNA and these genes are not contained within a Hindlll Hpall Mspl Mbol Hhol Eco R I -----simple tandemly repeated array. The fact that the pattern M N E N E N E N E N E L N E L of hybridization is identical whether the DNA was isolated from uninfected tissue or from nodules suggests that there is no apparent amplification or rearrangement in these sequences following infection of the plant by Rhizobium. Although soybean genomic DNA is methylated, hybridization to the leghaemoglobin probe showed no change in the methylation pattern of these genes between DNA from induced and uninduced tissues. In contrast, tissue-specific variations have been demonstrated in methylation patterns of globin and ovalbumin genes (27, 28).

FIG. 1. Restriction endonuclease digestion of soybean DNA. Nodule, leaf, and embryonic axes DNA (18 g each) were digested with various restriction enzymes, elwtrophoresed on 1.2%agarose gel, and stained with edhidium bromide. Ed, nodule;E, embryonic axes; L,l&, M,molecular weight markers.

General structure of IeghQemoglobin genes in soybean There are at least 10 &OM restriction fragments ranging from 1 to 12kb in length that contain leghaemoglobin sequences (16). A detailed analysis of the two fragments isolated from a soybean genomic DNA library constructed in Charon 4 X vector, revealed (N. Brisson and D. P. S. Verrna, manuscript in preparation) that the 11.5 kb &oRI fragment (Gmll) contains two leghaemoglobin genes separated by about 2 kb of DNA, while Gm4 conbins a truncated gene. Flsuaking regions of these three genes appears to be quite unique. K-loop analysis and DNA sequencing data showed that one gene coding for Lbc3 on Gmll is intempted by three introns and includes potential regulatory sequences (e.g., TATA and CAT boxes, etc.). Another leghaemoglobin gene located on the 7.5 kb EcoRI fragment also

BEPISSON ET AL.

Hpal l Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

Mspl

- - Hindill E N E N E N

M t

S! rn coding regmn

coding regmn

+

3'-noncding regoon

FIG. 3. General structure of leghaemoglobin genes in soybean. The sizes of the introns in Lbc3, a gene located on fragment Grnll, are shorter than those on the 7.5 EcoRI fragment (see Ref. 29).

the intemption of the coding region into four exons. The length of the intervening sequences varies between different leghaemoglobin genes of soybean (cf. Ref. 29). Homologies in the structws of the proteins (32) and genes of globin and leghaemoglobins particularly in the positions of the two introns strongly suggest that these genes evolved from a c o m o n ancestor. Furthermore, the presence of the conserved sites for the two intervening sequences suggests that there may be specific regions in these proteins encoded by exons (33). Recently, G6 (34) proposed that globin proteins are composed of fow distinct units, the middle exon containing two of these units. The presence of the third intervening sequence in leghaemoglobins appear to divide the central exon into two "functionsll" units.

Cross hybridizability ojsoybean cloned leghuemoglobin cDNA to mRNA frOm other legume nodules Poly(A) + RNAs isolated from 3-week-old nodules of kidney bean and pea were electrophoresed on a methyl mercury gel and transferred to diazotized paper for FIG. 2. Identification of restriction enzyme fragments hybridization with labelled pLbl. W A S fro& sbybean containing leghaemoglobin sequences in soybean genomic uninfected roots and from wheat germ were also run and DNA. A Southern blot of the gel in Rg. 1 was hybridized with transferred in parallel. Figure 4 indicates that pLbl did 32~-labelledpLbl (see Materids and methods). Autoradio- not hybridize to uninfected root and wheat germ M A S graph was obtained after exposure of the paper for 3 days. E, but that extensive hybridization occurred with RNA embryonic axes, N, nodule DNA; M, molecular weight from bean, and to a lesser degree, with pea nodules. The markers. cross hybridization with bean RNA was somewhat shows the presence of the intervening sequences at the predicted since the protein sequences of soybean Ebcs same locations (29). Based on these data, a general and kindey bean Lb at this end of the molecule are about structure of soybean leghaemoglobin genes can be 76% homologous. However, hybridization of pEbl with Brawn as shown in Fig. 3, which shows features pea RNA constitutes the first evidence of the relatedness common to other eukaryotic genes. The transcript is of pea and soybean leghaemoglobin sequences. These spliced, capped by 7-methylguanosine (asevidenced by cross hybridization results indicate that the degree of the competitive inhibition of the in vitro translation by homology existing between the leghaemoglobins of 7-rnethylguanosine-5'-monophosphate (30; D. P. S. different leguminous plants could in fact be similar to Vema, unpublished data) and polyadenylated resulting the degree of homology between globins from different animal species (35). Thus soybean leghaemoglobin in the production of a 9S h addition to the two intervening sequences found in clones could be successfully used to probe the genomes d l globin genes (31), there is a thid intervening of other leguminous plants for studying the organization sequence present in leghaernoglobin genes resulting in of these genes.

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

CAN. J. BIOCHEM. VOL. 60, 1982

FIG.4. Cross hybridization of a soybean leghaemoglobin cDNA clone (pLb1) with W A isolated from other legume nodules. (1) Soybean; (2) kidney bean; (3) pea; (4) uninfected soybean root; (5) wheat embryos. Poly/A)+ RNA was electrophoresed on methyl mercury gels and transferred to a diazotized paper which was then hybridized to 32~-labelled pLbl. h o w indicates the position sf 9% leghaemoglobh d W A of soybean.

Osganizatisn sf the leghaemsglobin gene family in the legumes In view of the evolutionary conservation of leghaemoglobin sequences and their specific regulation by Rhizobium, it was of interest to compare the overdl organization of this family of sequences within different legumes. Although pLbl showed sufficient cross hybridizability with WNA from kidney bean and pea nodules, we used mother cBNA clone, pLbl4, representing drnost the complete leghaemoglobin sequence (F. Fuller and B.P. S. Vema, unpublished) to obtain m a imum sequence homology. The DNA isolated from two varieties of Glycine mtw (cv. Brize and cv. Harcor), Phsseslus vulgaris, Glycine ssja, Pisum sativum, and Vicia faba was digested with Hind IIH and following electrophoresis and blotting, was hybridized with nicktranslated pLb14. ~h~~~ is an extensive cross hybridization of p ~ genomic DNA of four species (Fig. 5), andthe overpattern appears to be similar. F*emore9 some fragments containing legbaemoglobin sequences dire C O ~ toOd l ~genornes including Gb'cinejavonica (data not shown). Few minor differences were also observed between two varieties of soybean (cf. lanes 1

FIG.5. Hind III restriction enzyme pattern of leghaernoglobin b ~genes in five legume species. DNA isolated from (1) G1~cinem. cv. Rize; (2) Hmcor; (3) Phe~lusvulgarb; (4) Glysine soja; (5) Vicda faba; (6) Pisum sativurn, digested with Hind IU, and elcctrophoresed on 0.8% a g m e gel. The Southem blot was hybridized 32~-]aklled a leghaemoglobin cDNA clone which contains most of the leghaemoglobin sequence. M, molecular weight markers in kilobase pairs.

IN ET AL.

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

and 2, Fig. 5 ) . Thus the divergence in leghaemoglobins sf different species (changes in amino acid sequence and antibody cross-reactivity) is not well reflected by changes in the positions of restriction enzyme cleavage sites, A detailed analysis of this family of genes is essential to derive any general conclusions related to the evolution of these genes.

Expression of leghaemoglobin genes A direct correlation has been established between the effectiveness of legume root nodules in fixing atmospheric nitrogen and the presence of leghaemoglobin (4). However, the appearance of leghaemoglobin apoprotein and A has been detected prior to the nitrogenase activity during the development sf the nodule (6). To determine further the interdependence of the induction of leghaemoglobin and nitrogenase, two strains of Rhizobitlrn were used which f o m nodules that are unable to fix nitrogen (ineffective). One of these strains, SM5, has a mutation in the nitrogenase genes (36) and is thus unable to synthesize functional nitrogenase (37). Figure 6 shows the results obtained when pEbl was hybridized to polysomal poBy(A)+ RNA extracted from wild-type, SM5 or 61A24 (wild type, ineffective) induced nodules, and from uninfected roots. Poly(A)% W A S were bound to diazotized filter discs and hybridized with labelled DNA. A marked reduction in the level of leghaemoglobin m W A sequences is observed in ineffective nodules as compafed with the wild type. The reduction ranges from approximately 5-fold in SM5 to about 126-fold in 61AB a d is consistent with other results obtained by liquid hybridization using a leghaemoglobin cDNA probe (7)* Thus, the leghaemoglobin genes appear to be expressed in nodules irrespective of their effectiveness in fixing nitrogen although at greatly reduced levels (see also Ref. 17). Even at the lowest level, however, as in strain 61A24-induced nodules, these sequences are fairly abundant, representing about 1200 molecules/cell (7) (based upan the estimation of 20% leghaemoglobin A in effective nodules, Ref. 8). These data are consistent with the notion that the induction of the leghemoglobin genes is independent of nitrogen fixation in nodules and is an early manifestation of the infection process. Understanding the regulation of expression of the leghaemoglobin genes is important for the improvement or manipulation of the process of symbiotic nitrogen fixation since these genes appear to be part of the developmental program of the nodule (38) which requires physical invasion of the host by Rhizobium and may thus represent the host response to infection. These sets of genes are not expressed if the plants are not infected by Rhizobium or if ammonia is available for

A

E

277

4000

FIG. 6. Expression of leghaemoglobin genes in soybean nodules and effect of mutation in Rhizobiurn japonicum. Poly(A)+ RNA was isolated from 3-week-OMnodules formed by wild-type (61A76) and ineffective (61A24 and SM5) strains of Rhizobia and FWA was bound to DBM filter discs which were then hybridized to gLbl. The radioactivity associated with each disc was measured and data were nomdized for 1 pg RNA.

growth. The reason for a moderate reiteration in leghaemoglobin genes is not apparent; however, there %reat least four distinct leghaemsglobins (39) which are encoded by different genes. If each sequence occurs at two loci, there may be at least eight functional genes in addition to the possible pseudogenes often present in association with the gene fadlies (3I). Acknowledgements This study was supported by an operating and a strategic grant from the Natllrd Sciences and Engineering Research Council of Canada. N.B. is thankful for a NSERC graduate fellowship. We wish to h a n k S. Auger and C. EWlouri for characterizing one of the leghaemoglobin cDNA clones (pLbl4) from a cDNA library developed by B. Goodchild and F. Fuller for reading the manuscript. 1. Goodman, EM., Moore, G . W. & Matsuda, G . (1975) Nature ( L o d o n ) 253,603-408 2. Hurrell, 9. G . R.,Nicola, N. A. & Leach, S. J, (1979) in Soil Microbiology and Plant Nutrition (Broughton, W .

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by China University of Science and Technology on 06/06/13 For personal use only.

278

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 28.

CAN. J. BIOCHEM. VOL. 68, 1982

J., John, C. K., Royarao, J. C. & Lim, B., eds.), pp. 253-281, University of Malaya Press, Kuala Lumpu Hurrell, J. G. R., Nicola, N. A., Broughton, W. J., Dilworth, M. J., M i n a s h , E. &Leach, S. J. (1976) Eur. J. Biochem. 60, 389-399 Virtanen, A. I., Jorma, J., Lincola, H. & Linnasalmi, A. (1947) Acta Chem. Scand. Ser. B: l,90-111 Bergersen, F. J. & Goodchild, D. J. (1973) Aust. J. Biol. Sci. 26, 741-756 Vema, D. P. S., Ball, S., GuCrin, C. & Wanamaker, L. (1979) Biochemistry 18,476-483 Vema, D. P. S., Haugland, R. Brisson, N., Legocki, R. & Lacroix, L. (1981) Biochim. Biophys. Acta 653, 98- 107 Auger, S. Baulcombe, D. & Vema, D. P. S. (1979) Biochim. Biophys. Acta 563, 496-507 Vema, D. P. S. & Bal, A. K. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 3843-3847 Appleby, C. A., Turner, G. L. &MacNicol, P. K. (1975) Biochim. Biophys. Acta 387,46 1-474 Ellfolk, N. Be Sievers, 6;. (1971) Acta. @hem. Scand. Ser. B: 25, 3335-3534 Nicola, N. A. (1975) Ph.D. thesis, University of Melbourne, Australia Sievers, S. G., Huhtala, M.-L. & Ellfolk, N. (1978) Acta Chem. Scand. Ser. B: 32,380-386 Nalder, D. (1981) Plant Physiol. Suppl. 67, 135 Baulcombe, D. & Verma, D. P. S. (1978) Nucleic Acids Res. 5,4141-4153 Sulivm, D., Brisson, M., Goodchild, B., Vema, D. B. S. & Thomas, D. Y. (1981) Nature (London) 289, 516-518

21. Southern, E. M. (1975) J. Mol. Biol. 98, 503-5 17 22. Rigby, P. W., Dieckman, M., Rodes, C. & Berg, P. J. (1977) J. Mol. Biol. 113, 237-251 23. Bailey, J. & Davidson, N. (1976) Anal. Bioehem. 70, 295-314 24. Alwine, J. C., Kemp, D. J. & Stark, G. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5350-5354 25. Bishop, J. O., Rssbash, M. &Evans, D. (1974) J. Mol. Biol. 85, 75-86 26. Grippo, P., Iaccarino, M., Parisi, E. & Scarano, E. (1968) J. Mol. Biol. 36, 195-208 27. Waalwijk, C. Be Flavell, R. A. (1978) Nucleic Acids Res. 5,3231-3236 28. Mmdel, J. L. & Chambon, P. (1979) Nucleic Acids Res. 7,2081-2104 29. Jenson, E. O., Paludan, K., Hyldig-Nielsen, J. J. Jorgensen, P. & Marcker, K. A. (198 1) Nature ('London) 291,667-679 30. Hickey, E. D., Weber, L. A. & Baglioni, C. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 19-23 31. Efstratiadis, A., Posdcony, J. W., Maniatis, T., Lawn, R. M., O'Connell, C., Spritz, R. A., DeRiel, J. K., Forget, B. G., Weissman, S. M., Slightom, J. L., Blechl, A. E., Smithies, O . , Baralle, F. E., Shoulders, C. C. Be Prsudfoot, N. J. (1980) Cell 21,653-668 32. Hunt, T. L., Hurst-Calderone, S. & Dayhoff. M. 0. (1978) Atlas of Protein Sequence and Structure, vol. 5, suppl. 3, pp. 229-25 1, National Biomedical Research Foundation, Washington, DC 33. Gilbert, W. (1978) Nature ('London)271, 501 34. G6, M. (1981) Nature ('London)291, 90-92 35. Zimmer, E. A., Maretin, S. L., Beverley, S. M., Kan, Y. W. Be Wilson, A. C. (1980) Proc. Natl. Acad. Sci. Verma,D.P.S.,Nash,D.&Shulman,H.(1974)Natuae U.S.A. 77, 2158-2162 (London) 251,74-77 Bendich, A. J., Anderson, R. S. & Ward, B. L. (1980) 36. Hauglmd, R. & Verma, D. P. S. (1981) J. Mol. App. Gen,, in press in Genome Organization and Expression in Plants (Leaver, C. J., ed.), pp. 31-33, Plenum Press, New York 37. Maier, R. J. & Brill, W. J. (1976) J. Bacterlol 127, 763-769 Clewell, D. P. & Helinski, D. P. (1970) Biochemistry 9, 38. Verma, D. P. S. (1981) in Molecular Biology of Plant 428-4440 Development (Smith, H. & Grierson, D., eds.), pp. Blattner, F. R., Williams, B. G., Blechl, A. E., 437-466, Blackwell Scientific Publications, Oxford Thompson, K. Q., Faber, H. E., Furlong, L. Grunwdd, D. J., Kiefer, D. O., Moore, D. Q., Schumm, J. W., 39. Appleby, C. A., Nicolas, N. A,, Hurrell, J. G. R. & Leach, S. J. (1975) Biochemistry 14,4444-4450 Sheldon, E. L. & Smithies, 0. (1977) Science (Washington, B . C . ) 196, 161-169