Proc. Natl. Acad. Sci. USAVol. 82, pp. 7999-8003, December 1985Biochemistry

Synthesis of pathogenesis-related proteins in tobacco is regulatedat the level of mRNA accumulation and occurs onmembrane-bound polysomes

(transcriptional control/in vitro coupled translation-processing/stress protein)

JOHN P. CARR, DAVID C. DIXON, AND DANIEL F. KLESSIGDepartment of Cellular, Viral and Molecular Biology, School of Medicine, University of Utah, Salt Lake City, UT 84132

Communicated by Barbara McClintock, August 9, 1985

ABSTRACT The pathogenesis-related (PR) proteins oftobacco plants are induced in response to a variety of patho-genic and chemical agents. Although the function of theseproteins is unknown, they are associated with resistance tomultiplication and/or spread of tobacco mosaic virus. Wereport that functional mRNAs encoding PR proteins arepresent only when synthesis of these proteins has been induced,suggesting that their synthesis is controlled in part at the levelof mRNA accumulation. In addition PR proteins appear to besynthesized and processed in a manner analogous to proteinsdestined for the endoplasmic reticulum since (i) the in vitrotranslation products synthesized in the wheat-germ cell-freesystem are slightly larger than the in vivo products, (ii)translation of PR mRNAs in the rabbit reticulocyte lysatesystem is blocked unless that system is supplemented with dogpancreas microsomes, and (iii) mRNAs for PR proteins areassociated predominantly with membrane-bound polysomes invivo. This pathway of synthesis and posttranslational modifi-cation suggests possible sites of action of these proteins.

Infection of tobacco plants (Nicotiana tabacum L.) withtobacco mosaic virus (TMV) results in one of two distinctresponses characteristic ofthe particular combination of hostcultivar and virus strain. Many cultivars undergo a systemicinfection in which TMV spreads from the original point ofentry to many parts of the host. Such systemic infectionsoften result in extensive damage to the plant, especially in theyoungest leaves, which may become deformed and display amosaic pattern of green and yellow areas. In contrast to thesystemic infection, certain tobacco cultivars can restrict viralmultiplication to a small zone of cells around the infectionsite. In this localized infection, necrotic lesions form at thesites of virus entry. When these localizing (resistant) tobaccocultivars undergo a second infection with TMV, the localizingresponse is enhanced, resulting in the appearance of fewerand smaller lesions. This enhancement of virus localization,acquired resistance, is not confined to the tissue adjacent tothe primary lesions but is found throughout the plant (1, 2).Acquired resistance to TMV can also be induced by otherlesion-forming viruses, bacteria, and fungi (3), as well as bycertain chemicals such as acetylsalicylic acid (3, 4).

Virus localization and expression of acquired resistanceare accompanied by the appearance of abundant amounts ofpathogenesis-related (PR) proteins. Like acquired resistance,the appearance of these proteins is not restricted to theinfected areas but is seen throughout the plant (5-9). Thesecorrelations suggest that accumulation ofPR proteins may bepart of a generalized resistance response to many types ofplant pathogen (3, 10).

PR proteins can be regarded as a class of stress proteinsinduced in response to the stress of infection. However, theydiffer from the widely studied stress proteins, the heat shockproteins, in a number of respects: (i) their synthesis isinhibited at elevated temperatures; (ii) they begin to accu-mulate days rather than minutes after the inducing stimulus;(iii) their appearance is not transient; and (iv) unlike someheat shock proteins, which are highly conserved and serologi-cally related across a wide spectrum oforganisms (11-13), PRproteins from different plant groups do not usually react tothe same antisera (8).However, PR proteins from Nicotiana and other plant

groups do have several common properties. These aremolecular masses between 10 and 20 kDa, solubility in lowpH buffers, high proportions of acidic and aromatic aminoacid residues, and resistance to endogenous plant proteases(8). The best-characterized PR proteins are the PRi group,which are synthesized in the TMV-localizing (resistant)tobacco cultivar Xanthi-nc. PRs la, lb, and lc are chargeisomers having the same molecular mass (=16 kDa). They aresimilar in amino acid composition (9), and all three react withantiserum raised against PRla (14-16). Biochemical, genetic,and serological evidence implies that each member ofthe PRifamily of proteins is encoded by a separate gene (14-16).Although considerable information has been obtained re-

garding the appearance and properties of PR proteins (8),little is known concerning regulation of the genes encodingthese proteins. In this paper, we report a study of themechanism(s) governing synthesis of this group of stressproteins. Our results suggest that expression of the genesencoding PR proteins is regulated in part at the level ofmRNA accumulation and that synthesis of PR proteinsoccurs on membrane-bound polysomes.

MATERIALS AND METHODSPlant Material and Induction of PR Protein Synthesis.

Nicotiana tabacum cv. Xanthi-nc plants were grown fromseed in a sand/peat compost in a Conviron growth chamber(17) and used for experiments when 4-6 wk old. To induce PRprotein synthesis, batches of 24 Xanthi-nc plants weretransferred to trays containing approximately 4000 cm3 ofcompost and watered daily with 500 ml of 830 .uM acetylsal-icylic acid adjusted to pH 6.5 for 6 days (4).

Extraction of RNA from Leaves and Crude PolysomalPreparations. Total RNA was extracted (17) from leaves ofacetylsalicylic acid-treated and untreated Xanthi-nc plants.Poly(A)+ RNA was obtained by oligo(dT)-cellulose chroma-tography (18). Crude preparations of free and membrane-bound polysomes were made using a variation of the proce-

Abbreviations: PR, pathogenesis-related; TMV, tobacco mosaicvirus; DPM, dog pancreas microsomes; SRP, signal recognitionparticle.

7999

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 82 (1985)

dure ofJackson and Larkins (19): leaves (5 g) were pulverizedin liquid N2 and homogenized in 50 ml of ice-cold buffer A[200 mM Tris HCl, pH 9.0/400 mM KCl/35 mM MgCl2/25mM EGTA/25% (wt/vol) sucrose/1% (vol/vol) 2-mercapto-ethanol/0.005% cycloheximide], and the homogenate wascentrifuged at 15,000 rpm for 20 min at 40C in a Beckman JA20rotor. The supernatant (containing free polysomes) wasplaced on ice, and the pellet (containing membrane-boundpolysomes) was washed with two 25-ml portions of buffer A.Membrane-bound polysomes were released by suspendingthe pellet in buffer A containing 1% (vol/vol) Triton X-100.Polysomes (free and released membrane-bound) were con-centrated by ultracentrifugation through 60% (wt/vol) su-crose cushions (19). RNA was obtained from polysomalpellets by phenol extraction and ethanol precipitation (17).

In Vitro Synthesis and Processing of Proteins. RNA wastranslated in nuclease-treated cell-free rabbit reticulocyte(Green Hectares, Oregon, WI) or wheat germ (BethesdaResearch Laboratories) (20, 21) lysates. In some experimentsreticulocyte lysates were supplemented with dog pancreasmicrosomes (DPM) or a blank solution according to theinstructions of the manufacturer (Amersham).

In Vivo Labeling of Proteins. Leaves were excised from thestem under water and fed via the petiole with 200 A.l of H20containing 100 ,uCi of [35S]methionine (1 Ci = 37 GBq;Amersham) for 18 h. Soluble leaf proteins were extractedwith 50 mM Tris*HCl pH 8.0/1 mM EDTA/0.1% 2-mercap-toethanol (1 ml of buffer/g of leaf) and stored at -20°C.

Analysis of Radiolabeled Polypeptides. Immunoprecipita-tions (17) were carried out using 1-5 Al of anti-PRla serum(22) or 50 ug of anti-pea ribulose-1,5-bisphosphate carbox-ylase small subunit IgG. Preimmune rabbit serum was usedin control immunoprecipitations. Antigen-antibody compe-tition experiments were carried out in immunoprecipitationreaction mixtures containing 20 ,ul of in vitro translationmixture, 1 ,ul ofanti-PRla serum, and 0-4 ,ug ofpurified PRlaprotein. Immunoprecipitated polypeptides were analyzed byNaDodSO4/polyacrylamide gel electrophoresis (23) and so-dium salicylate fluorography (24).

RESULTSPR Protein Synthesis Is Regulated at the Level of mRNA

Accumulation. Acetylsalicylic acid treatment of Xanthi-ncplants induced synthesis ofthe PR1 (16 kDa) proteins (Fig. 1).These proteins were not detected in untreated plants (Fig. 1).To establish whether mRNAs encoding the PR1 proteins

were present in untreated plants, RNA was prepared fromacetylsalicylic acid-treated and untreated plants and used toprogram two different cell-free translation systems. Reticu-locyte lysates programmed with poly(A)+ RI4A from eitheracetylsalicylic acid-treated or untreated leaves generated thesame immunoprecipitable polypeptides of =40 and 14 kDabut neither synthesized a 16-kDa protein (Fig. 1; see Fig. 3).Wheat germ extracts produced this same set of im-munoprecipitable polypeptides with both RNAs (Fig. 1).However, when programmed with poly(A)+ RNA fromtreated (but not untreated) plants, the wheat germ systemsynthesized a polypeptide similar in size to PR1 made in vivo.This polypeptide was also synthesized in wheat germ extractsprogrammed with RNA obtained from TMV-infected Xanthi-nc leaves (data not shown).Antigen-antibody competition experiments were conduct-

ed to determine the relationship of these in vitro products togenuine PRla. Addition ofpurified PRla had little or no effecton immunoprecipitation of the 40- and 14-kDa in vitrotranslation products (Fig. 2, see Fig. 3B), suggesting thatthese products are not related to PRla. These polypeptideswere probably immunoprecipitated by antibodies in the

kDa M68- %43- ,

In vivo

-ASA+ASAP 1 P I

Reticulocyte Wheat germAn~~~~~~~--ASA +ASA-ASA + ASAP I P I P I P I

25- _

18.4- s14.3-a

4...

FIG. 1. Comparison of PR proteins synthesized in vivo and invitro. In vivo-labeled proteins and poly(A)+ RNAs were extractedfrom leaves of acetylsalicylic acid (ASA)-treated or from untreatedXanthi-nc tobacco plants. The poly(A)+ RNA preparations wereused to program reticulocyte and wheat germ cell-free translationsystems. 35S-labeled in vivo- and in vitro-synthesized polypeptideswere immunoprecipitated with immune serum raised against PRia(lanes I) or with preimmune serum (lanes P), the immunoprecipitateswere resolved by electrophoresis on a NaDodSO4/15% polyacryl-amide gel, and the gel was fluorographed. The arrow marks theposition of PR1. Marker proteins (lane M) were bovine serumalbumin (68 kDa), ovalbumin (43 kDa), a-chymotrypsinogen (25.7kDa), P3lactoglobulin (18.4 kDa), and lysozyme (14.3 kDa).

anti-PRla serum produced against contaminants in the orig-inal antigen preparation.

In contrast, immunoprecipitation of the 16-kDa translationproduct was quantitatively inhibited by the addition ofauthentic PRla (Fig. 2). Hence the 16-kDa product isserologically related to PRla. It is assumed that the 16-kDaPR1 polypeptide synthesized in vitro and detected by immu-noprecipitation with anti-PRla serum consists of in vitro-synthesized forms of all three members of the PR1 familysince polyclonal antisera raised against PRla also react withPRs lb and lc (14-16).Thus, it appears that functional mRNAs encoding the

16-kDa PR1 are present in acetylsalicylic acid-treated orTMV-infected plants but not in control plants. These resultssuggest that induction ofPR1 protein synthesis is regulated inpart at the level of mRNA accumulation.The Block to PR1 Protein Synthesis in Reticulocyte Lysates

Is Released by DPM. The inability of the rabbit reticulocytecell-free system to synthesize detectable PR1 suggests twopossibilities. Either reticulocyte lysates contain a factor,

(1 1 4

-41h 4=61 -k

FIG. 2. Antigen-antibody competition experiment to determinerelationships of in vitro translation products to PRla. Equal amountsof radioactive translation products synthesized in wheat germ ex-tracts programmed with RNA from leaves of acetylsalicylic acid-treated Xanthi-nc tobacco were immunoprecipitated with 1 Al ofanti-PRla serum in the presence of0, 1, or 4 Ag ofpurified PRla. Theimmunoprecipitated 3"S-labeled polypeptides were analyzed byNaDodSO4/15% polyacrylamide gel electrophoresis and fluorogra-phy. The arrow marks the position of PR1, while the asterisksindicate the positions of the 14- and 40-kDa polypeptides.

8000 Biochemistry: Carr et al.

Proc. Natl. Acad. Sci. USA 82 (1985) 8001

absent from wheat germ extracts, that inhibits synthesis ofPR1 or wheat germ extracts contain a factor, absent fromreticulocyte lysates, that promotes PR1 synthesis. Theformer possibility was investigated because rabbit reticulo-cyte lysates, but not wheat germ extracts, contain levels ofsignal recognition particles (SRPs) (25) that block elongationof nascent polypeptides possessing the N-terminal signalsequence necessary for cotranslational translocation into theendoplasmic reticulum (25-27). If the PR1 proteins areinitially synthesized as precursors with signal sequences,then arrest of polypeptide elongation by SRPs may explainthe lack of PR1 synthesis in reticulocyte lysates.To test this hypothesis, reticulocyte lysates were supple-

mented with DPM. The membrane vesicles of DPM containSRP receptors (docking proteins) that release the SRP-mediated elongation arrest and allow translation to continue,coupled to transport of the nascent polypeptide into thevesicles (27, 28). In the presence ofDPM, reticulocyte lysatesprogrammed with RNA from acetylsalicylic acid-treatedXanthi-nc leaves synthesized a polypeptide of approximately16 kDa (Fig. 3A). The addition ofpurified PRla quantitativelyinhibited the immunoprecipitation of this polypeptide, show-ing that it is related to the PR1 family of proteins (Fig. 3B).This in vitro-synthesized PR1 was not produced by reticulo-cyte lysates programmed with RNA from untreated Xanthi-nc leaves even in the presence of DPM (Fig. 3), againsuggesting that PR protein induction is regulated at the levelof mRNA accumulation.

Since the 16-kDa polypeptide is not synthesized in retic-ulocyte lysates in the absence of DPM, the results of aprevious study using unsupplemented reticulocyte lysatessuggested that the 14-kDa polypeptide immunoprecipitatedby anti-PRla serum might be related to PRla and implied thatPR1 mRNAs may be constitutive (22). However, the resultsof antigen-antibody competition experiments in the presentstudy exclude this possibility and confirm the results ob-tained using the wheat germ system (Figs. 1 and 2).PR1 Proteins Are Synthesized as Larger Precursors. If PR

proteins are synthesized as precursors that undergo cleavageduring transport into the endoplasmic reticulum, then PR1

A-ASA ±ASA

kDa BM -M

68-L r

43-

25- _

18.4-

14.3-

B+ASA -ASA

kDa6

68 -

43- a____

25 -

protein made in the wheat germ system should be larger thanthe processed product found in vivo. In the experimentsdescribed above, the polypeptides synthesized in vitro or invivo could not be distinguished on the basis of size using aNaDodSO4/15% polyacrylamide gel system. However, whenthe products were fractionated on NaDodSO4/20%o poly-acrylamide gels the PR1 synthesized in vivo migrated faster,suggesting a difference in size of -1 kDa (Fig. 4).PR Protein mRNAs Are Associated with Membrane-Bound

Polysomes in Vivo. The above results imply that PR1polypeptides are synthesized in vitro as precursors withsignal peptides that interact with SRPs and suggests that invivo these proteins are destined for transport into and/oracross a membrane. If so, their mRNAs should be associatedwith membrane-bound polysomes (27).To test this hypothesis, RNA extracted from crude prep-

arations of membrane-bound and free polysomes from ace-tylsalicylic acid-treated or untreated plants was translated inwheat germ extracts. The translation products were analyzedby immunoprecipitation with anti-PRla serum and IgG spe-cific for the small subunit of ribulose-1,5-bisphosphate car-boxylase. Since this small subunit (SSU) is synthesized onfree polysomes as a 20-kDa precursor (29-31), it served as asuitable internal control for the separation of free andmembrane-bound polysomes. As expected, functionalmRNA encoding the small subunit precursor (pSSU) wasprimarily associated with free polysomes (Fig. 5), indicatingthat little cross-contamination took place between the freeand membrane-bound polysomes during the fractionationprocedure. In contrast, mRNAs for PR1 proteins wereassociated principally with membrane-bound polysomes(Fig. 5). mRNAs encoding PR1 proteins were also associatedwith membrane-bound polysomes extracted from TMV-infected Xanthi-nc tobacco leaves (data not shown). Theseresults indicate that PR proteins are synthesized on mem-brane-bound polysomes and suggest that in vivo they aretranslocated into the lumen of the endoplasmic reticulum forsubsequent processing and/or transport.

DISCUSSIONThe mechanism(s) governing expression of genes encodingPR proteins has been investigated using an in vitro translationassay. mRNAs encoding the PR1 family of proteins arepresent only in tobacco plants that are synthesizing theseproteins, indicating that their synthesis is regulated in part at

A B C D

4-

18.4-

143 3- __i$mm

FIG. 3. Abolition of the block to PR1 synthesis in reticulocytelysates by DPM. (A) Reticulocyte lysates supplemented with a blanksolution (lanes B) or with DPM (lanes M) were programmed withRNA extracted from leaves of acetylsalicylic acid (ASA)-treated oruntreated Xanthi-nc tobacco. "S-labeled translation products wereanalyzed by immunoprecipitation with anti-PRla serum, NaDod-S04/15% polyacrylamide gel electrophoresis, and fluorography. (B)Antigen-antibody competition experiment. Reticulocyte lysates sup-plemented with DPM were programmed with RNA from leaves ofASA-treated or untreated Xanthi-nc tobacco. Equal amounts of"5S-labeled translation products were immunoprecipitated with 1 ILIof anti-PRla serum. Immunoprecipitation of the '16-kDa productsynthesized in lysates programmed with RNA from ASA-treatedtobacco was inhibited by the addition of 1 or 4 ,ug of purified PRla.Arrows, position of PR1 proteins.

In vitroI.tIn~In vl(l . _! In l'il'

FIG. 4. Size difference between PR1 synthesized in vivo and invitro. 35S-labeled PR1 proteins were immunoprecipitated from invivo-labeled protein extracts or from wheat germ cell-free translationsystems. Lanes: A, PR1 synthesized in vitro; B and C, in vivo-labeledPR1; D, a mixture of in vivo- and in vitro-synthesized PR1. Theimmunoprecipitates were analyzed by NaDodSO4/20%o polyacryl-amide gel electrophoresis and fluorography.

Biochemistry: Carr et al.

Proc. Natl. Acad. Sci. USA 82 (1985)

kDa68-

anti-small subunit anti-PRIa- forAI-

-ASA +ASA -ASA -+ASAT T F M T T FM

43-

25 -w

- pSSU

4_ 4SI6-PRI18.4-

14.3- -

FIG. 5. Association ofPR protein mRNAs with membrane-boundpolysomes. Wheat germ cell-free lysates were programmed with total(lanes T) leaf RNA from acetylsalicylic acid (ASA)-treated oruntreated Xanthi-nc tobacco or with RNA extracted from crudepreparations of free (lanes F) or membrane-bound (lanes M)polysomes from leaves of acetylsalicylic acid-treated Xanthi-nctobacco. Aliquots of translation products containing equivalentamounts of trichloroacetic acid-insoluble radioactivity were im-munoprecipitated using anti-PRla serum or anti-small subunit IgGand analyzed by NaDodSO4/15% polyacrylamide gel electrophoresisand fluorography. pSSU, precursor of the ribulose-1,5-bisphos-phate carboxylase small subunit.

the level of mRNA accumulation. Because the amount offunctional message reflects not only the rate of transcriptionbut also the efficiency of RNA processing and mRNAstability, further studies will be required to determine theprecise regulatory mechanism(s).

It has been suggested that the appearance of PR proteinsmight be due to breakdown of normal cellular proteins (32).However, the observation that functional PR mRNAs arepresent only in plant tissue in which PR1 proteins can bedetected argues strongly against such an origin for this family ofPR proteins.The finding that PR1 proteins were synthesized in wheat

germ extracts but not in rabbit reticulocyte lysates wassurprising and initially perplexing. However, addition ofDPM to reticulocyte lysates programmed with RNA fromacetylsalicylic acid-treated or TMV-infected Xanthi-nc to-bacco leaves permitted the in vitro synthesis ofPR1 proteins.It is inferred from these results that SRPs present in thereticulocyte lysates are capable of recognizing a putativeN-terminal signal sequence on the PR proteins synthesized invitro and preventing their translational elongation in the sameway as SRPs can arrest in vitro synthesis of animal proteinsdestined for the endoplasmic reticulum. Further evidencesuggesting the existence of a signal sequence on nascent PR1proteins comes from the observation that PR1 produced inwheat germ lysates appears to be larger than PR1 synthesizedin vivo. Moreover, the association of mRNAs encoding PRproteins with membrane-bound polysomes suggests that invivo PR proteins are cotranslationally translocated into amembrane-bound subcellular structure such as the endo-plasmic reticulum (27).

In general, three classes ofproteins are synthesized as highmolecular weight precursors on membrane-bound polysomesand cotranslationally translocated across the endoplasmicreticulum membrane: secreted proteins, intracellular vesicle(e.g., lysosomal) proteins, and membrane-associated pro-teins. Since PR proteins are soluble in the absence ofdetergents, it is unlikely that they are integral membraneproteins or accumulate in any type of insoluble form analo-gous to seed storage proteins, which are also synthesized asprecursors on membrane-bound polysomes (33). However, itis conceivable that they remain in the cell to be transportedvia the endoplasmic reticulum to the Golgi, like certain other

types of stress proteins (34), or to lysosomes. It is alsopossible that PR proteins pass through the plasma membraneto remain in the periplasmic space like extracellular carbonicanhydrase (35) or to be extruded beyond the cell wall likea-amylase (36). The latter is consistent with the preliminaryresults of others (8, 37) which suggest that the PR proteins oftobacco may be extracellular. Since little is known about thefunction(s) ofPR proteins, studies aimed at determining theirlocation in vivo should help to define their role in the responseof plants to the stress of infection.We thank Drs. J. F. Antoniw and R. F. White for purified PRla

and their advice; Dr. N. H. Chua for pea anti-SSU IgG; and Drs.J. 0. Berry, B. J. Nikolau, G. P. Thomas, G. Herrck, and D. Carrolfor helpful discussions. This work was supported by a SearleScholarship from the Chicago Community Trust and an IndividualResearch Award in Plant Biology from the McKnight Foundation toD.F.K. D.F.K. is the recipient of Faculty Research Award 270 fromthe American Cancer Society.

1. Ross, A. F. (1961) Virology 14, 329-339.2. Ross, A. F. (1961) Virology 14, 340-358.3. Gianinazzi, S. (1983) in Plant-Microbe Interactions. Molecular

and Genetic Perspectives, eds. Nester, E. W. & Kosuge, T.(Macmillan, New York), Vol. 1.

4. White, R. F. (1979) Virology 99, 410-412.5. Gianinazzi, S., Vallee, J. C. & Martin, C. (1969) C. R. Acad.

Sci. (Paris), Ser. D 268, 800-802.6. Gianinazzi, S., Martin, C. & Vallee, J. C. (1970) C. R. Acad.

Sci. (Paris), Ser. D 270, 2383-2386.7. Van Loon, L. C. & Van Kammen, A. (1970) Virology 40,

199-211.8. Van Loon, L. C. (1985) Plant Mol. Biol. 4, 111-116.9. Antoniw, J. F., Ritter, C. E., Pierpoint, W. S. & Van Loon,

L. C. (1980) J. Gen. Virol. 47, 79-87.10. Kuc, J. (1982) BioScience 32, 854-860.11. Schlesinger, M. J., Ashburner, M. & Tissieres, A. (1982) Heat

Shockfrom Bacteria to Man (Cold Spring Harbor Laboratory,Cold Spring Harbor, NY).

12. Neidhardt, F. C., Van Bogelen, R. A. & Vaughn, V. (1984)Annu. Rev. Genet. 18, 295-329.

13. Kelley, P. M. & Schlesinger, M. J. (1982) Mol. Cell. Biol. 2,267-274.

14. Ahl, P., Antoniw, J. F., White, R. F. & Gianinazzi, S. (1985)Plant Mol. Biol. 4, 31-37.

15. Antoniw, J. F., White, R. F., Barbara, D. J., Jones, P. &Longley, A. (1985) Plant Mol. Biol. 4, 55-60.

16. Matsuoka, M. & Ohashi, Y. (1984) J. Gen. Virol. 65,2209-2215.

17. Berry, J. O., Nikolau, B. J., Carr, J. P. & Klessig, D. F.(1985) Mol. Cell. Biol. 5, 2238-2246.

18. Aviv, H. & Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69,1408-1412.

19. Jackson, A. 0. & Larkins, B. A. (1976) Plant Physiol. 57,5-10.

20. Roberts, B. E. & Paterson, B. M. (1973) Proc. NatI. Acad.Sci. USA 70, 2330-2334.

21. Pelham, H. R. B. & Jackson, R. J. (1976) Eur. J. Biochem. 67,247-262.

22. Carr, J. P., Antoniw, J. F., White, R. F. & Wilson, T. M. A.(1982) Biochem. Soc. Trans. 10, 353-354.

23. Laemmli, U. K. (1970) Nature (London) 227, 680-685.24. Chamberlain, J. P. (1979) Anal. Biochem. 98, 132-135.25. Meyer, D. I., Krause, E. & Dobberstein, B. (1982) Nature

(London) 297, 647-650.26. Walter, P. & Blobel, G. (1981) J. Cell Biol. 91, 557-561.27. Walter, P., Gilmore, R. & Blobel, G. (1984) Cell 38, 5-8.28. Gilmore, R. & Blobel, G. (1983) Cell 35, 677-685.29. Dobberstein, B., Blobel, G. & Chua, N.-H. (1977) Proc. NatI.

Acad. Sci. USA 74, 1082-1085.30. Chua, N.-H. & Schmidt, G. W. (1978) Proc. NatI. Acad. Sci.

USA 75, 6110-6114.31. Highfield, P. E. & Ellis, R. J. (1978) Nature (London) 271,

420-424.32. Coutts, R. H. A. & Wagih, E. E. (1981) Plant Sci. Lett. 21,

51-54.33. Higgins, T. J. V. (1984) Annu. Rev. Plant Physiol. 35, 191-221.

8002 Biochemistry: Carr et al.

Biochemistry: Carr et al.

34. Welch, W. J., Garrels, J. I., Thomas, G. P., Lin, L. J.-C. &Feramisco, J. R. (1983) J. Biol. Chem. 258, 7102-7111.

35. Coleman, J. R. & Grossman, A. R. (1984) Proc. Nati. Acad.Sci. USA 81, 6049-6053.

Proc. Nati. Acad. Sci. USA 82 (1985) 8003

36. Okita, T. W., DeCalvera, R. & Rappaport, L. (1979) PlantPhysiol. 63, 195-200.

37. Antoniw, J. F. & White, R. F. (1983) Neth. J. Plant Pathol.89, 255-264.