:‘53 p115“. . ‘w ‘u ...»~.;..v..u‘\7.-..-. -~.s-:\ . ~ .. > t ,5.- 3‘ ‘ .- ': nu," I ‘fi‘fi‘uz‘t'r I _‘.~- 4, L. 3: 'e -' mung a. Lgfllj- h... 4‘”: ‘5‘ " L31“ ‘ EN”? "3. ,V n. .325 5 #22 ,.~ ,0 a"; . .1.‘ x,{ :54 k. , S ;d . :’ “an. I.‘ '"t‘ ,...1 .- famm .5. . .353}. '. . .-_. N- . . ....;,.._.....-,’. , .z,,...‘.v.... . ,..;. ..-... ...,,-. ) I“ . u yarn... ... 'Inr-o , . ,. .mt- ;" ' .,,. ,.... n. n. . .mu ... - . um” r .. ”.w-u. . ' mu.“ 1 .. .. ”nu...” .“u- .n‘. .‘ ... . .. u ,,. ....A.‘ v . ,.....,....-u v uy.. I. u. O‘INDM a. - hummom u .- , ' I r‘. ‘A . . v .. .. 'H'Q‘W. . , . A. mm“.-. — .... ' . . H. .. .wur;"¢,A1 o .. . . v Hull! .‘J. . . ., “Mug“... . r . ,—f a .Clwn - . - . ‘,‘;","';.', ‘1}. 1,, .T .. .- 4‘ 4-;- l. ...t...,.'n ~r— .... . , ... .. .. . .‘ . 271.2223, ‘* . :1"). , - ., - «WV: ".3... ‘ . '3'.“ 7M\368 ANTS'TA EU NIEV HSILTY lllllllll lllllllllllllllll 300898 7772 This is to certify that the dissertation entitled THE ROLE OF PLANT-DEFENSE RESPONSES IN THE DETERMINATION OF HOST SPECIFICITY FOR THE RHIZOBIUM-LEGUME SYMBIOSIS presented by Janet Lynn Salzwedel has been accepted towards fulfillment of the requirements for Ph.D. Botany and Plant Pathology degree in 74446 Dear Major professor/ Date May 22, 1991 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to remove We checkout from your record. TO AVOID FINES return on or before ode due. ll DATE DUE DATE DUE DATE DUE IS. L——- ——l fil—jl—W MSU lo An Affirmative AotlorVEquel Opportunlty Institution CWMa-pj THE ROLE OF PIANTvflfiEENSE RESPONSES IN THE DETERMINATION'OF’HDST SPECIFICTTY FOR.THE RHIZOBIUMFLEGUME SYMBIOSIS By Janet.1ynn.Salzwedel A.DISSERT%EIGN Eitnfixxed.to Nfirmdgan.State‘University in.partial fulfillment.of the requirements fer the.degree of DOCTORAOF PHILOSOPHY Department of Botany and Plant.Pathology 1991 mar ‘IHEIDIEOFPIAN'I‘WERESIONSES IN THE WOT OF HOST SPECIFICITY RR 1m: H'IIZOBIUM-IESUME SYMBI$IS By JanetlynnSalzwedel Our objective was to test whether rhizobia elicited a plant defense response which could affect the determination of host specificity in the Rhizdaium-legume symbiosis. Heterologous Rhizobium legmninosannu biovars elicited increased specific activity of salt-elutable peroxidase fremthesurfaceofpeaandcloverroots. 'Ihecell-freesupernatantof 3. legmnirnsarum bv. y_i£i_._a_§ also elicited increased peroxidase activity from clover roots. 'Ihe excreted elicitor of peroxidase activity was flavone-dependent, heat-stable, and ethanol-soluble. Permddase activity was localized to the site of attempted penetration in clover roothairs. Pemidaseactivityincloverroothairsbegantoirrzrease 6 hrs after heterologous inoculation. Inoculation with homologous bv. trifolii suppressed peroxidase activity for 12 hrs. The transient sugaression of activity could be mimicked by treatmart with purified EPS from bv. trifolii. 'Ihe Sym plasmid-cured strain, _m_®_::'m§ mutant, flung mutant, and hybrid recombinant bv. trifolii strain cartaining the host specific nodulation (hgi) genes from bv. lic_ia£ each elicited lessperoxidaseactivityfrompearootsthandidthewildtypebv. trifolii. We conclude that 1151; genes interact with others present on the Sym plasmid to contribute to host-specificity through the modification of an elicitor which increases root hair peroxidase activitywhichinturnmayalterthestructureoftheroothairwall at the site of incipient penetration. MicropHelectrodes wereusedtomeasuretheresponseof single white clover root hairs to inoculation with rhizobia or treatment with purified LPS. Cells of homologous bv. trifolii or their LPS iniuced a rapid (30 min) increase in pH at the surface of individual root hairs from pH 6.5 to 6.7. Heterologous rhizobia did not elicit this response. The flung and 9939mm; mutants of bv. trifolii also failed to elicit this response. We conclude that since homologous (compatible) rhizobia elicit this increase in pH, this respome is ftmdamentally differentflianflrefuptakebymstcellsmdergoingahypersersitive response elicited by incompatible pathogens. 'Ihis Rhizobimn-indwed neutralization at the root hair surface may enhance successful infection by overcoming the acid-inhibition of early infection events. WW Cutofthechaosweimposeorder. Suchisthegenesisofthis dissertation. I am grateful to those who have helped me along the way. My guidance committee, Drs. Ken Nadler, Ray Hammerschmidt, Robert Hardin-ski, andmost of allFrankDazzowasmost patient withme. Frank DazzoisthemostemberantscientistIhaveeverlmownarrithoughI sometimesresisted,hetaughtmetokeepanopenmind. ManyhavepassedthroughDr. Dazzo's labandallhavecontributed tomy learningexperience. I wishtothankArrireaSquartini forhelp with molecular genetics, J ing-wen Chen for discussions, and Guy Orgambide for help with chemistry. I am grateful to Dr. N. S. Subba Rao for his kind help in reviewing this dissertation. In particular, I wish to thank Drs. Rawle and Saleela Hollingsworth who camseled me in science arri were also my friends. IamgratefultoDr.BobHausingeranithemembersofhislabfor the kind use of their spectrophotometer and helpful enzyme discussions. I thankthetmdergraduateworkerswhohelpedmewithmyresearch, Laura Amenzeller and Ann Aggarwaal. IaminiebtedtoDr.TiTienfortheuseofhisequipmentarrito Dr. Peter Vassilev for participating in the root hair pi-I experiments and forteadringmehowtomakemicropflelectrodes. Partial financial support from the USDA National Needs in Bio- technology Program is gratefully acknowledged. Ofmyfriends, IwishtothankBrianPalik forallthathetaught me. Emiragement from Kent MCOJe, Jorge Santa-Domingo, and Steve Spatz was muchappreciated. SpecialthanksgotoJimBretzwhodriedmytears duringthedarkdays. Mostofall, Iowemysuccesstotheloveandsugnortofmy parents, brother, andsister. Icouldnothavedoneitwithoutthem. iv 1788130me3 PAGE LIST OF TABLES... ..................................... vii LIST OF FIGURES........... ................... ............viii LITERATURE REVIEW ........................................ 1 Introduction ..... .......... 1 Rhizobium-legume host specificity................... 1 Mechanisms of plant defense against microorganisms“ 5 Plant responses to Rhizobium .............. .. ...... .. 8 Peroxidase review... ......... 15 CHAPTER 1. NE gene dependence for the elicitation ofperoxidaseactivityfromcloverroothairsandpea roots by Rhizobium and their cell-free supernatants. ..... 21 Abstract...... .............. ..... . .............. 21 Introductim............... ....................... .. 23 Materials and Methods.... ................ . ..... 26 Bacteria and cell-free bacterial washings..." 26 Isolation of EPS 27 Seed sterilization................ ............ . 27 Plant growth and inoculation................... 27 Treatment of clover roots with isolated EPS.... 28 Removal of roots......... ..... ...... 28 Clover growth and inoculation for studies with root hairs.......... ...... . ............ 28 PeI'oxidase localization in clover root hairs... 29 V PAGE Isolation of peroxidase........................ 29 Peroxidase assay............................... 30 Separation of isozymes by electrqinresisu... 30 Staining gels for peroxidase activity. ...... 31 Heat stability and ethanol fractionation of cell-free bacterial washings................ 31 Infection thread and nodule initiation bioassays ...... ...... . 32 Results ...... . ....... 34 GIAPIER 2. Rhizobium leguminosarum bv. trifolii and its purified LPS stimulates a rapid neutralization at the surface of clover root hairs 67 Abstract 67 Introduction ...................... .. 68 Materials and methods....................... ...... .. 71 Bacterial cultures...... ..... ...... .. 71 LPS isolation.......................... ...... .. 71 Preparation of pH microelectrodes....... ...... . 72 Measurement of root hair pH.... ................ 73 ReSIflts................................ ............ . 74 DiMimOCOOOOOOOOOOOOOOO00.0.0.0... ........ .00... 77 LIST OF REFERENCESOOOOOOOO0.0.0.0..........OOOOOOOOOOOOOO 82 vi LISTOF‘EBIFS PAGE Rhizobium legmninosarumstrainsused in this study 33 Effect of purified EPS from bv. trifolii AN0843 on specific activity of peroxidase eluted from clover roots. 44 Effect of culture supernatant from broth-grown B. leguminosarum bv. viciae on number of infected root hairs and nodule initiations formed by ANU843 on white clover cultivars 54 vii LIST OF FIGURES FIcJRE PAGE m I 1 Specific activity of peroxidase eluted from clover arripearootsZ4hrafterinoculatimwitth10 cells/ml suspension of wild type rhizobia. Each barrepresentsthemeanofatleastBexperiments +/- SE. Nitrogen-free Fahraeus medium (-NF) was used as the control ..... ............... 35 2 A. Native polyacrylamide gel stained for peroxidase activity with 3-amino-9-ethyl carbazole. mroxidase was isolated from clover roots 24 hr after imculation with rhizobia. Ehchlanewasloadedwichugofprotein. Treatments included: lane 1 horseradish peroxidase, lane 2 —NF control, lane 3 ANUB43, lane 4 R1300. B. Native polyacrylamide gel stained for peroxi- dase activity with 3-amino-9-ethyl carbazole. kroxidase was isolated from pea roots 24 hr after inoculation. Treatments included: lane 1 -NF control, lane 2 ANU843, lane 3 R1300. C. Native gel made with a 3-10 % gradient of acrylamide to illustrate the homogeneity of the band at the interface of 7.5 % gels..... ..................... 36 3 Whitecloverroothairsstainedforigsitu peroxidase activity with DAB + H233, pH 5.5. A. uninocilated. B. 1 d post-' ation with AN0843. C. 5 d post-inoculation with ANU843. D. 5 d post-inoculation with R1300. Arrows indicate areas of enhanced staining......................... 39 4 Specific activity of peroxidase from isolated clover root hairs at 0, 6, 12, or 24 hr after inocilation with wild type rhizobia. Data points represent the mean of 2 experiments +/- SE....... ............ 43 5 Specific activity of peroxidase eluted from clover ard pea roots 24 hr after inoculation with the pSym-cired bv. trifolii strain or recombinants containing cloned fragments from pSym in the pSym-cured backgromd. -NF = - control, wt = wild type, pSym" = ANUB45, viii FIGJRE 10 PAGE 032 = 845pRI'032 containing 14 kb _no_d gene region,A910 = 84511210324910 containing 8kb nodJICBAD. Bars represent the mean of 3 experiments +/- SE ..... . .............. .. .......... . 46 Specific activity of peroxidase eluted from cloverarripearoots24hrafterinoculationwith suspensions of wild type (wt) rhizobia or their respective nodE: rm; and nodL: :TnS mutants. Bars represent th—e mean of 3 —experiments +/- SE.. ................. 47 Specific activity of peroxidase eluted from cloverandpearoot24hrafterinocilationwith wild type or hybrid recombinant strains carrying heterologoush_sn_genes. wt=wildtype, pKI'K85= plasmid carrying bv. viciae nodDFEIMN genes, pRT290 = plasmid carrying bv. trifolii nodrERIMN genes. BarsrepreseuuttluemeanofBeuzperimerxts+/-SE. ..... 49 Specific activity of peroxidase eluted from cloverroots 24 hraftertreatmentwithstandard- ized cell-free bacterial washings (002612; 0.08- 0.12) from bacteria grown on plates wi without 4 uM flavone. Data are from one experiment .......... 50 Specific activity of peroxidase eluted from clo- ver root 24 hr after addition of autoclaved bac- terial washing (0024 = 0.18) from R1300 grown with4qu1avone. 4gataarefromone acperiment ........... 52 Specific activity of peroxidase eluted from clover roots 24 hr after treatment with the ethanol-insoluble pellet and ethanol—soluble supernatant fractions of bacterial washings from R1300 grown with 4 uM naringenin. Data are from one experiment .......... . ........ ........ 53 we 11-! at the surface of single white clover root hairs inocu- lated with wild type rhizobia. Bars = SE 75 pH at the surface of single white clover root hairs inocu- lated with Tn§ mutants derived from Rt843. Bars = SE-...... 75 pH at the surface of single white clover root hairs treated With purified LPS. Bars = SE.......OOOOOOOOOOOOOOOOO ........ 76 II'I'ERA'IUREREVIEN Introductim 'Ihe mutually beneficial symbiosis between soil bacteria of the genus Rhizobium and plants in the legume family is agriculturally important world wide. The fixation of atmospheric N2 into NH3 by the bacteria residing inside root nodules eliminates the need for costly nitrogenous fertilizers. 'Ihe symbiosis is characterized by highly evolved and restricted host specificity with only Rhizobium legmninosarmn bv. Mableto infectpeasarrivetdu, 3. leguminosanum bv. trifolii able to infect clovers, and g. meliloti able to infect alfalfa. 'nuesequenceofeventsleadingtosuccessful nodulation forthe above Rhizobium-legume combinations are well known. Microscopy has revealedthatrhizobiaareattractedtohostroots, attachtotheroot hairtip, induucemarkedcurling ofthetip, andpenetratetheroothair within the confines of the advancing infection thread (Nutman, 1959: Vincent, 1980). Eventually, the bacteria are released from the infection thread into the proliferating cortical cells and are surrounded by a plant derived peribacteroid membrane. 'Ihe bacteria then differentiate into bacteroids and a nitrogen fixing nodule emerges. mizduium-legmne host quecificity. The genes governing host specificity in the fast-growing rhizobia have been well characterized (Horvath gt_ a1" 1986; Debelle and Sharma, 1 2 1986; Djordjevic g; g]_.., 1986; Martinez e_tgl_., 1990). For g. leguminosarum bv. trifolii, bv. ligigg, and B. meliloti, successful symbiosis culminating in a nitrogen-fixing nodule filled with bacteroids deperrisonthepresenceofonetoseverallargebacterialplasmids, appropriately known as symbiotic plasmids (pSym) . In bv. trifolii strainANU843, there isoneSymplasmidonwhichgenesessential for white clover nodulation are found (Djordjevic gt_ g1” 1983). To date, thenfigenes inANU843 thathavebeennamedarethecommon_n0_d_genes W; the host-specific nodulation (hsn) nodFERIMN'I‘ genes; and two genes that affect infection thread development, _no_dIJ_. "Oommcm _n_og genes" aredefinedasgeneshavinghomologous Msequencesamongthe rhizobia, whose loss by mutaticm _or deletion renders the bacteria unable to infect theirownoranyhost, arriwhiduwhenmutatedordeleted, can be replaced by the equivalent gene from another Rhizobium species to restore nodulation ability on the original host. To be designated an h_sn_gene, either (1) thegenemust alterthebacterialhostrangewhen thatgenefunctionislost (e.g. 'mginsertioninflextendsbv. trifolii host range to nodulate peas), or (2) the gene must be transferred to a recipient strain of a different Rhizobium species in order to allow that recipient to efficiently modulate the homologous host of the donor Rhizobium (eg. bv. trifolii nodFERIMN genes transferred to bv. Egg enable the latter to efficiently nodulate white clover, Djordjevic gt. _a_1_., 1986). Thus, manipulation of the bacterial h_sn_ genes affects the determination of host specificity. The precise function ofthe_h_sn_genesisstillunknown. Fauchergt_gl_. (1988), however, have reported that the bacterial production of a host specific extracellular factor which deforms root hairs is dependent on 3 fling. meliloti. 'n'ueauthorsbelievethatthecommonggenes encode a non-specific hair deformation factor and it is thew gene product which then modifies the factor to produce the host-specific form of the molecule known as nodle (Ierouge gt_ 11,, 1990). Root exudates contain various carbohydrates, carboxylic acids, phendic compounds, and amino acids (Klein g3; _a_1_., 1990). Within this milieu, the class of compounds called flavones are a pivotal plant signal which stimulates expression of the fast-growing Rhizobium nodulation genes. Flavones are synthesized from one branch of the phenylpropanoid pathway with other branches giving rise to lignin precursors, anthocyanins, and pterocarpan phytoalexirs. ‘Ihe greatest stimulation of E gene expression depends upon the specific flavone and the particular Rhizobium biovars or species being examined. 'Ihe most stimulatory flavones are 7,4'-dihydroxyflavone (DiF) for bv. trifolii, naringenin for bv. m and luteolin for g. meliloti (Redmond g; Q" 1986; Peters g 91., 1986). Very low concentrations (10-9-10.6 M) of these compounds will induce Rhizobium nod genes. The pattern of hydroxylation specifies whether the flavone will be stimulatory, inhibitory, or have no effect on the particular strain being examined. Some host specificity is dictated by the specific flavone to which a particular Rhizobium species responds, and the basis of this flavone specificity lies in specific domains of the protein product of the pSym regulatory gene, @ (Spaink gt 11:, 1987). Those genes shown by transcriptionalgugfiusimstobeeuquressedaftereuquocuretoflavorns haveaconservedupstreamregimofmmwnasa'gbox". new protein binds to this E box region of WA (Fisher $31., 1988, Kondorosi gt 11., 1988). 4 In additiau to release of _ngg gene-inducing flavones, the plant contribution to host-specificity is thought to involve specific recognitimarriattadumentofbacteriatothehostroothairs. Atthe root hair surface, an initial reversible attachment occurs that is not symbiont-specific (Dazzo g g" 1984b) and may involve a (EH-dependent bacterial adhesin (Smit g g" 1986). Within an hour, this step is followed by a symbiont-specific, lectin-mediated step of bacterial aggregation (Dazzo g g" 1984b). In soybean, pea, and clover, lectirs which specifically bind the EPS or LPS of the homologous symbiont have been observed (Bohlool atfl Schmidt, 1974: Dazzo ard Hubbell, 1975: vander Schaal g g" 1983). In white clover, immunofluorescence localization has shown that the lectin, trifoliin A, is present at root hair tips (Dazzo and Brill, 1977). The most compelling evidence that a lectin is a host-encoded determinant of symbiont specificity has come through genetic manipulation of the host plant. Diaz 5 11: (1989) showed that transgenic white clover containing the gene for pea lectin could be infected with 3. leguminosarum bv. v_ic_i_ag_. 'Ihe barrier of host specificity has also been breached through external manipulation of the host plant. Al-Mallah gt a_1. (1987) succeeded in infecting white clover with Rhizobium loti at a low frequencyaftertherootshadbeentreatedwithamixtureofcellulase and pectolyase in the presence of polyethylene glycol. Similar treatment allowed nodulation of the non-legume Brassica napus by rhizobia (Al-Mallah g g" 1990). It is the early interaction between Rhizobium and the root hair which must determine host-specificity (Li and Hubbell, 1969). Both homologous and heterologcus rhizobia attach to root hairs, but they 5 differ in their pattern of attachment. Homologous rhizobia are heavily concentrated at the root hair tip with additional bacteria polarly attaduedalongthelengthoftheroothair. 'Ihishasbeentermedthe symbiont specific pattern of attachment. 'Ihe heterologous rhizobia also bindtoroothairs, thoughnotinthesymbiontspecificpatternof attachment (Dazzo g 9;" 1984). Both homologous and heterologous rhizobia induce root hair deformations, but it is only the homologous which induces the symbiont specific defamation called a shepherd's crook which leads to infection thread initiation. Heterologous rhizobia haveneverbeenobservedtoformaninfectionthread. 'Ihus, thefirst structuralbarrierencounteredbyrhizobiaistheroothairwall. Mechanisms of plant defense agairst microorganisms. Daringthelifecycleofaplant, itsrootsareexposedtoafull spectrum of rhizosphere—colonizing organisms including saprophytic, beneficial, and pathogenic fungi and bacteria. ‘Ihe first line of defense against colouization by microbes is preformed structures including the cell wall. The wall is a matrix composed of cellulose fibrils, hemicelluloses, pectirs, pectic acid, interwoven with a network of hydroxyprol ins-rich glycoprotein (extensin) (Iamport, 1986; Varner and Lin, 1989). A number of enzymes are also covalently or ionically bound to the wall matrix. In legmne roots, additional proteins such as lectinsandadhesinsarefourrionroothairtips. Furtherprotectionof roots is thought to be provided by a hydrophobic suberin layer. Suberin is a heterogeneous polymer composed of both long chain (C16-C26) fatty acids and alcohols, and aromatic compounds as suggested by the appearance of p-hydroxybenzoic acid, vanillin, and syringaldehyde after nitrobenzene oxidation (Kolattukudy, 1977). While. cutin is the 6 protectivepolymercoveringtheaerialpartsofplants, suberinis tlnugnttobetheprotectivecoveringforrootsandtubers. ‘Ihusfar, analysisoftheskinofrootvegetablessud‘uascarrot, parsnip, and turnip, andtheepidermis ofyoungbeanrootshasshownthatthe characteristic components of suberin are present (Kolattulariy g a" 1975; Sijmons $11., 1985). Plants have also evolved inducible structural arrl Chemical defense mechanisms to resist attack by pathogers. 'Ihe hypersensitive reaction (HR) isoftencitedasahost-specificdefensemeduanism. 'IheHRis irduced in incompatible (rm-host) interactions with pathogenic bacteria. It is characterized by a rapid localized host cell death and ruecrosis oftissuewhich isthoughttolimit furtherpathogen ingress. 'Iheearliesteventobservedinthetmistheuptakeoffi'l'arri concomitant efflux of K+ within 30 min after inoculation of tobacco suspension cell cultures (Atkirson g g" 1985). After several hours there is decompartmentalization of the cells, loss of membrane integrity, a general efflux of ions, and finally cell death. The HR cell death can be delayed by high external pa (Salzwedel _eg gag, 1939) andHRinducedperoxidatim ofthehostcellplasmamembranecanbe inhibitedbyadditiouoftheenzymesuperrncidedismutase (Kemlerand Novacky, 1937). Arotheriniucibledefenserespousewhidlisprevalentamougthe legumes is the production of antimicrobial metabolites known as puytoalexins. In legumes, pterocarpan derivatives of the phenylprquanoid pathway are primarily induced in the tissues being attacked by pathogens (Ingham, 1982). These compounds are all toxic to fungi while a few are also bactericidal; e.g. soybean glyoeollin is .7 toxic to Pseudomonas syringae pv. glminea (Wyman auri Van Etten, 1978). Itisttnughtthattolerameofhostfiuytoaleudrscontribrtestothe virulence of a pathogen, although some nor-pathogens are also tolerant (Smith and Banks, 1986). Virulent isolates of the fungus Nectria haematococca detoxify pisatin via pisatin demethylase while avirulent isolates are unable to do so (Van Etten_e£_a_l_., 1980). 'Ihis system provides the best evidence for the role of tolerance to phytoalexins in the virulence of a pathogen. Many phenolic componuds also have antimicrobial properties. In the resistance resporse of tobacco and tomato, increased concentrations of rhenolies sud) as scopoletin have been measured (Nadolny and Sequeira, 1980; Bashan _e; g" 1987). A number of induced structural modifications of plant tissues are thought to hinder colonization by pathogens. In the epidermal cells of barley coleoptiles, resistance to Erisyphe graminis has been correlated with the rapid formation of oversized cell wall appositiors known as papillae at the sites of attempted fumgal penetratio'u (Aist and Israel, 1986). Papillae are composed of callose and menolic residues as indicated by histoduemical staining and autofluorescence (Aist and Israel, 1986). In cucnnber, non-host resistance and induced resistance to ftmgi were correlated with deposition of a lignin halo at the site of penetration in epidermal cells of hypocotyls (Hammersdlmidt g5 91., 1985; Hammerschmidt and and, 1982). Deposition of suberin is important in the potato tuber wonud response (Kollattukudy and Dean, 1974) and in coating the infected vascular cells of resistant tomato plants inoculated with Verticillium albo-atrum (Street and Ellis, 1986). In melonandcucnnber, fmgalpathogensirrluceanincreaseinthe hydroocyproline-rich cell wall glycoprotein lawn as extensin (EEquerré— 8 mgayé and Iamport, 1979; Hammerschmidt gt $1., 1984). This increase in extensin in melou has been correlated with ethylene-inducible resistance to pathogens Wye _e_t_ 11., 1979). actensin also accumulates inthecell wallsfromcucnnberseedlings subjectedtoheatshock. Such extensin-enriched walls are more resistant to pectolytic enzymes than walls from cortrol plants (Sterner and Hammerschmidt, 1987). Arothertypeofresponsetoinfectiouistheproductionof pathogenesis related (PR) proteins. In tobacco, resistance to W tabacina involves systemic induction of both chitinase and B-l,3-glucanase (Ye g_t_ 11,, 1990). Other PR proteins have been identified as proteinase-inhibitors (Bowles, 1990). Ntnnerous reports link an increase in the activity of certain isozymes of peroxidase with resistance responses (Campa, 1991). 11115 enzyme is capable of catalyzing the cross-linking of the various wall strengthening polymers including lignin, suberin, and extensin (Campa, 1991). A detailed review of peroxidase literature appears later in this section. 'Iheabove inducible resistancemeduanismsarerarelyusedalonebut ratherinconcert. Sinceplantsareconstantlyexposedtoalegimof potentialpathogens, thequestionremainsastohowrhizobiaavoidsudu defense mechanisms in the natural erwimnment. Plant m to Ruizduiun. We already know the phenotype of homologous interactions—a nitrogen fixingnodule. Yet, tofullyunrierstarrithespectnnnofevents which contribute to the determination of host specificity, it is also necessary to study the heterologous interactions. Vance (1983) suggested that the symbiotic interaction was a beneficial plant disease. A plant's molecular recognition of both pathogens and Rhizobium depends 9 upolthebacterialpolysacduaridesresidimouthecellcurface—EPS, cps, and res (Keen and Holliday, 1932). In his review, Vance (1983) addressed the heterologous Rhizobium-legmne interaction in asking: 1. Does pretreatment with heterologous polysaccharide affect infection by homologous bacteria? 2. Are there rhizobial (EPS?) suppressors of plant mechanisms that limit infection? In another review (Djordj evic g; g" 1987a) Rhizobium is described as a refined parasite and the authors suggestthattheroleof lectinsinsuccessful infectionistotitrate out EPS which would stimulate a defense response. There are no outward signs of gross defensive responses or injury to the legume host after inoculation with wild type heterologous rhizobia. Still, this metabolically competent bacterium with a full complement ofgenes andenzymesnecessarytoinfectahost, isunuableto penetrate the heterologous root hair. Rhizobia presumably enconrter phytoalexirs in competition with other rhizosphere organisms which elicit plant defense responses. Rhizobia differ in degree of tolerance to various legume phytoalexins. 'Ihe slow-growing Bradyrhizobium was sensitive to the major phytoalexins in clover, medicarpin and maakiain, with an Effective Dose (E050) = 10—60 Lug/ml. 5. leguminosarum bv. trifolii, bv. yicig, and 3. M were much less sersitive to medicarpin and maakiain with an EDSO > 100 urg/ml (Panlduurst and Biggs, 1980). Pisatin purified from peas increases the generatiou time of R. leguminosarum bv. yigigag'gyitgg, howeverpisatinwasnotdetectable in peanodulesmrtilbacteroidsbegantosenesceatBOdafterinoculatiou (Van Iren _e_t_ g_l_., 1933). Chakraborty and Chakraborty (1939) report that homologous rhizobia did not elicit either of two phytoalexins in pea epicotyls after 5 d. Glyceollin was detected (175 pmole/mg dry weight) 10 but did not accnnulate in either the effective nodules of Glycine max or the ineffective nodules of g. s_ojg 131342434 after inoculation with g. m 113mm (Parniske et a1., 1990). Parniske g g. (1933) found thatsoybeanroothairs isolatedafter infectiouwithgm contain 4 new flavonoid componuds, one identified as glyceollin I, although environmental factors can also lead to differences in the flavonoid pattern. Bothg. Manda. mammimm inducible tolerance to glyceollin (Parniske g; a_l., 1991). 'lhe authors conclude that other rhizosphere organisms probably stimulate phytoalexin production and it is the isoflavone-inducible tolerance to the phytoalexin which gives these rhizobia a competitive advantage in the rhizosphere. Even in interactions between wild type _R. leguminosarum bv. trifolii and clover, the majority of the infection threads that are initiated eventually abort (Nutman, 1959). One possible interpretation of such data is that infection threads abort because the plant limits infection by an invading organism with a defense-like response. Recently, a number of workers claimed to have observed plant- deferseresponsesinlegumeswhentreatedwithcertainmutantsof Rhizobium. Inoculation of Macrcptillium with an EPS Wing mutant of NGR234 resulted in rapid accumulatiou of osmiophillic droplets in the epidermal cells of the host (Djordjevic _e_t_ _a_l., 1988). ‘Ihe authors suggest this is a hyperseusitive response. P'L'uhler e_t_ Q. (1991) reported that autofluorescent (polyphenolic) materials accumulated in thickened cortical cell walls of pseudonodules on alfalfa induced by an EPS'mutantofR. meliloti. ‘IheyproposethatEPS isessentialto suppressadefenseresporsewhiduwouldpreventpenetratimbythe 11 homologous rhizobia. Both of these reports show that plants are capable of responding to Rhizobium mutants defensively. It is not necessarily indicative of the pl'uenomena which contribute to the host specificity of wild type rhizobia in the rhizospuere. What, then, is the plant responsetoaheterologous wildtypeRhizobium andcouldsucharesponse inclufle a defense whidu would exclude the bacteria? The following are possible mwels to explain why rhizobia are unable to infect the root hairs of a heterologous legume host. 1. Heterologous rhizobia do not bind to host moot lectin. Iectins may have multiple functions in interactions with rhizobia, including attachment. Eyen though Diaz gt_ gl_. (1989) overcame the host- specificity barrier by introducing the pea lectin gene into white clover, Al-Mallah _e_t_ g. (1987) overcame the host specificity barrier by removing the tips of clover root hairs with exogenous cell wall degradingenzymes. Regardlessofthebacterialmodeofentryinthis case, thedatahintthatattadumenttotheroothairwall isnotthe sole factor for the determination of host specificity. 'Ihe data do not precluude the possibility that lectin was present on root hair plasmamembranes, however the non-legume Brassica napus could also be infected by Rhizduium with removal of the root hair wall (Al-Mallah g 11., 1990) . 2. 'Iherhizobiadonothavetheenzymestoerodethewallofthe heterologous host nor do the rhizobia specifically elicit the host's own wall degrading enzymes. Martinez-Molina g; g_]._. (1979) studied cellulase and hemicellulase activity in rhizduia and suggested that hydrolytic enzymes may be an additional factor in host specificity. Such enzymes maybeimportantindistinguishingnon-legumespeciessuxduasinthe 12 Gramineae where wall composition differs markedly from the Ieguminosae. Mort's analysis of the root hair walls from a number of plant species showed that within the legume family, cell wall polysaccharide compositions ameared very similar (Mort and Grover, 1988). Even with similar compositions, a thorough analysis of the particular linkages within the wall matrix may yet reveal differences amoug legume genera. Theoretically, rhizobia that produce cellulases and pectinases should be able to penetrate any legume root hair given that wall substrates are similar. In addition to bacterial enzymes, homologous rhizobia and their isolated EPS stimulate the activity of clover root polygalacturonase (Ljundggren and Fahraeus, 1961). It was believed that induction of host polygalacturonase led to softenirg of the root hair wall allowing penetration by the bacteria. 3. The flavones of the host do not effectively induce expression of the heterologous Rhizobium nod genes. Spaink gl_: g_l.. (1987) costructed hybridstrainscontainirgnodABCUfromg. leguminosarumpRLlJIandthe _nggm gene from 3. leguminosarum, & trifolii, or g. meliloti. 'Ihe exudates from Melilotus alba, Pisum sativum, Vicia hirsuuta, and 'I'rifolium repens each stimulated the expression of all cloned nodD constructstoatleast50%oftheiniuctionlevelof1uteolin. Both wild type 3. leguminosarum bv. yiciag arud R. leguminosarum bv. trifolii are stimulated equually by apigenin. There is a difference in stimulation by 7-hydroxyflavone (bv. trifolii is stimulated, bv. yi_c_i_a_e_ is not). 'Ihere may be some host-specificity dictated by the sensitivity of a Rhizobium strain to a particular flavone (Spaink g _al" 1987), however clover elaudates contain a mixture of stimulatory arud inhibitory puenolic componuds (Djordjevic _e_t g1" 1987b). In addition, a number of l3 legume plants produuce 7,4'-dihydroxy flavone (DiF) (Venkataraman, 1981) anithereforeamorecarefulanalysisofflavmesintherhizosphereis required to understand the contribution of flavones to host-specificity. 4. 'IheheterologouthuizobiumEgoirecognizedasasymbioutarrl therefore the plant's casti‘tutive level of defense is not reduced sufficiently to allow penetration by the bacteria. All plants possess preformed deterrents to microbial colonization. Recognition of a beneficial microorganism probably involves biochemical cell-cell communuication. Sudu recognition could occur at the level of attachment, during initial cell wall degradationbybacterialorhostenzymes, in the release of biologically active bacterial polysaccharides which affectroothairs, otherroothaircurlu’rg factorsexcretedbythe bacteria, orsomeasyetunuknownproduuctofbacterialhflgenes necessary to establish the symbiosis. 5. 'nueheterologousmizobium'grecognizedasaninvaderarriplant defenses are induced which exclude the bacteria from penetration. Plant-pathogen interactions which involve gene-forhgene recognition requirearesistancegeneinthehostarrianavirulencegeneinthe bacteria for expression of resistance (Ellingboe, 1981). Rolfe (pers. comm.) suggests that Rhizobium hsn genes act as aviruulence genes since 'In§ mutation in ANU843 mug (presumed loss of function) results in expanded host rarge to include peas. Rolfe g £41988) claim that increased flavcme exudation stimulated by heterologous rhizobia representsadefenseresuonse, althoughthedirectconsequuenceofthis flavoueexudatio'uisnotdisclssed. ItisthoughtthathomologousEPS somehow masks elicitors of plant defense which are present in both homologous and heterologous rhizobia (Verna and Nadler, 1984: Djordjevic l4 _e_t_ a_1., 1987a). 6. ‘Iheheterologousrhizduiadonotproducetheappropriate extracellular "leg" signal. Ierouge e_t_ 11, (1990) have isolated a compon'd called nodle from _R_. meliloti. 'Ihis compond is a sulfated— lipooligosacduaride whose synthesis is deperudent uupon host specificity genes and which is active in root hair deformatioru arud initiation of cortical cell divisiors at nanomolar concentrations. 'Ihe biological activity is restricted to the host plants of R. meliloti. Althoughthesixthhypothesisiscurrentlythemostpooular, more has yetdemoustratedhow__nod_signals alonemightbeaffectingtheroot hair to allow homologous infections to occur. Item must be some mechanism forthebacteriatopenetratetheroothairwall. Noone hypothesishaseverbeenshowntobedefinitive. Mostlikely, a combination of the above hypotheses for Rhizobium-host interactions will ultimately be needed to explain the determinatiou of host specificity. mus, theworkinthisdissertationadiressesouepossible determinant contributingtohost specificityandhasbeenguidedbythe fifth hypothesis: namely that Rhizobium biovars are unable to infect a heterologous host because they are recognized as invaders and elicit plant defense reactions which then exclude the bacteria. Many plant resistance reactions to pathogens have been duaracterized which develop todetectable levelsoverthecourseofdays. 'lheworkpresentedhere examines only the very early pheruomena in the interacticn of rhizobia arriroothairswhidumustplayaroleintheshortwirdowofopportmity wheru host specificity is determined. Several different plant reactions could contribute to host specificity, however this dissertation research focuses on the first structural barrier that rhizobia enconrter—the 15 roothaircellwallauriirducedwallmodifications. Numerousreports havecorrelatedirereasedactivityinplantcellwallperouddasewith resistancetomicroorganisms. 'Iherefore,experimentswereperformedto test how homologoe and heterologous rhizobia affect clover and pea root androothairperoxidase. Inaddition,t1einfluenceofbacterialhost— specificity genes on possible plant responses was examined usirg a collectiou of _no_d gene mutants and recombinants. Muddase Mia! 'Iheparadigm ofplantperoxidases ishorseradishperoxidase (Meunuier, 1991). This enzyme has a MW of 42,100 and contains 308 amino acids, 1 Fe-protcporhyrin prosthetic group, 2 Ca”, 17% carbohydrate from 8 neutral side chains each with a glycosyl compositio'u of N- acetylglucosaminez, marmose3, fucose, and xylose (unford, 1991; Campa, 1991) . 'Ihe reaction catalyzed by peroxidase follows modified ping-pong kinetics with two 1-electron oxidations of two substrate molecules via activated enzyme intermediates formed by interaction with 11202 as follows: 1:221 1313;. 229.. E-II +Afi2 -> E+'AH+H20 wrere E is enzyme, A is substrate. Peroxidaseisirreversibly inactivatedatpfllll orgreaterduetothe loss of reme, but is typically stable at room temperature and pH 5-10 (Dmford, 1991). Horseradish peroxidase has a pH optimum of 7.0 (Maehly, 1955); however tte pH optimum for manyplantperoxidases is slightly acidic (Fielding and Hall, 1978; liverdeen e_t _a_1_., 1988). Becauuseoftlesuorgredoxpropertiesoftteoxidizedformsof perouddasearritteobservedlogdistareeelectrmtrarsferprocessesof 16 proteins, it is possible for peroxidases to oxidize substrates which are not in the direct vicinity of the active site (Meunier, 1991). 'Ihis explaixswhypermddasescanoxidizealaxgerangeofmuzralard artificial substrates imluding O-methoxyphenol (guaiacol) , O- dianisidine, 3,3'-diaminobenzidine, 3-amim-9-ethylcarbazole, 4—methoxy naphthol, 2,2'-azim-di[3-ethyl-benzothiazoline—(6) sulfonic acid] (ABIS) , pyrogallol, gallic acid, acetosyringcne, 3-hydroxy-flavone, ferulic acid, vanillic acid, syrirgaldehyde, and indole-B-acetic acid (1AA) (Dunford, 1991; Gaspar gt a_1., 1982; Grison and Pilet, 1985). It isthaaghttlntthetruefmnctimofcertainpemxidasesisdictatedby some substrate specificity i_n yiLo. Peroxidasehasbeendetectedinhomogenatesofleavee, stems, and roots from numerous plant species (Gaspar e_t a1., 1982). Both cytoplasmic and wall-bound peroxidases have been isolated (Gaspar gt a_1., 1982). kroxidase may be covalently or icmically bound to cell walls. Treatment with cell wall degrading enzymes was necessary to release covalentlybonmdpermcidase inmaizerootsandpeaepicotyls (Grison and Pilet, 1985; Ridge and Osborne, 1970). Ionically bound wall peroxidaseshavebeenrecoveredfmmtheintercellularfluidfmmvaanm infiltrated cucumber and barley leaves, lupin hypoootyls, the medium from sugaension-afltured peanut cells, the stele, cortex, and tip of pea roots, and the surface of bean roots (Hammersdunidt e; g" 1982: Srivastava and VanHuystee, 1977; Albert g9 a_1., 1986: Ros Barcel6 and Mufioz, 1989; Fielding and Hall, 1973). Tissue prints of cross-sections from untreated pea epicotyls showed permidase activity was localized in vascular bundles. Plants treated with ethylene underwent cessation of elongation, increase in radial 17 growth, andarpearanceofperoxidase activityintheepidermaland cortical cells (Cassab gt_ a_1_., 1988). Pa roots, however, showed histochemical staining for peroxidase in epidermal cells, stele, and cortical cells (Fielding ani Hall, 1978). Density gradient centrifugaticn and electron microscopy have shown that peroxidase activity was most often associated with the central vacuole, rough endoplasmic reticulum, golgi bodies, the plasmalemma, and the cell wall (Gaspar gt gin 1982). Cress root hair walls stained for peroxidase activity with the strongest reaction at the tips (Zaar, 1979). Golgi bodies within the root hairs also stained for peroxidase activity, thus suggesting the subcellular site of post-translational glycosylaticn (Zaar, 1979). Plant peroxidases exist both as anionic (acidic) and mtionic (basic) isozymes. Certain peroxidase isozymes differ in their substrate specificity (mic, 1968: Grison and Pilet, 1985; Gibson and Liu, 1978). In pea tissue homogenates, some isozymes were unique to root, epicotyl, and leaf (Gibson and Liu, 1973). Van Huystee (Srivastava and Van Huystee, 1977), however, has shown that peanut isozymes are artifacts of interaction with phenols. Filtrate from peamrt alspension cell cultures contained five different isozymes of peroxidase based on mobility in a native gel. Treatment of the filtrate with Dowex l-Xl to remove Woliceresultedinthelossofllmtofsisozymebards. mthe other hand, Ros Barcel6 and Mu’r'loz (1989) suggest that the interaction of peroxidase isozymes with phenolic compounds exerts epigenetic control of peroxidase activity. 'Iheir data show that incubation of lupisisoflavone with 2 isozymes purified from lupin cell walls specifically converts the 2 isozymestoathird isozymewhidlgairstheabilitytooxidize 18 scopoletin. 'nleauthorsprcposethattheflavmemayactasan intermediate inthetransfercfahydrogenatomfromscopoletinto peroxidase. In other plant systems, peroxidase isozymes are selectively stimulated in response to particular stimuli. Kay and Basile (1987) showed that the stimulation or supression of certain peroxidase isozymes was correlated with different stages of crganogenesis in tobacco. Womdingofanmberhypocotylsanlofpotatotubersreslltsinthe increase in certain peroxidase isozymes (SNalheim and Robertsen, .1990; Espelie g a1., 1986). In cucurbits challenged with fungi, induced resistance responses are associated with systemic stimulation of particular animic isozymes (Hammersdlmidt g a" 1982; Smith and Hammerschmidt, 1988). In a resistant barley cultivar, resistance to Erysiphe graminis was correlated with an increase in 2 isozymes which didmtappearinwourriedtissueanddidmtincreaseinthenear isogenic susceptible cultivar (Kerby and Somerville, 1989) . 'Iherearereportsthatircreasedperoxidaseactivityintobaccoand tomato is not correlated with resistance to bacterial pathogens, but rather that increased concentration of pl'lenolic compounds is important (Nadclny and Sequeira, 1980; Bashan e_t_ g" 1987). ‘Ihepresencecfpercxidase incell wallsanditspotential catalytic activity suggests a possible function in cell wall synthesis. Extensin monomers are cross-linked via isodityrosine ether linkages and peroxidase catalyzes the formation of these cross-links in m (Everdeen g a_1., 1988). In xylem elements and woody tissues, cell walls are impregnated with lignin which is composed of derivatives of p- coumaryl, califeryl, and sinapyl alcohols covalently bourd in a 19 heterogeneals matrix (Lewis airi Yamamoto, 1990). Again, the polymerization of known components of lignin can be achieved _i_n_ 21%; by peroxidase (campa, 1991). 'Ihus, by virtue of its ability to cross-link substratesingg wallbom'ripermcidascearetholghttoparticipate in wall synthesis with wall-bound malate dd’lydrogenase supplying the necessary 3202 from NADIH (Campa, 1991). Cross-linkingcf suberinmayalsobecatalyzedbypermcidase. A suberin-associated peroxidase has been immunochemically localized to wounded potato ’olber tissues undergoing suberizaticn (Fspelie and Kolattuhxiy, 1985; Espelie _e_t g" 1986). In tomato, resistance to the wilt pathogen Verticillium albo-atrum is correlated with the deposition of a suberin coating in the xylem (Street g g" 1986). ‘lhe cum clone fort-hesuberin-associatedperoxidase frcmpotatowasusedtcprobe NorthernblctscftotalRNAfromnearisogeniclinescftcmato. In tomato suspension cell clltures treated with fungal elicitors, the resistant line showed peroxidase message within 15 min while the susceptible line showed barely detectable hybridizaticn in 3 hr (Mohan and Kolattuhldy, 1990). In addition to cross-linking activity, some peroxidase isozymes can degrade indole acetic acid (1AA) QM Anumberof stuiieshave shown a correlation between IAA catabolism by peroxidase and cessation of exponential plant growth (Gaspar et a1., 1982). cationic peroxidase isozymeshaveahigheroxidc—reductimpotential againstIAAthando anionic isozymes (Gaspar g a1., 1982: Campa, 1991). Such cationic peroxidaseshavebeenforriintobaccorootsarrlcalluscllture, andin isolated vacuoles from tobacco (Campa, 1991). In contrast, peroxidases isolated from cell walls are either strongly or weakly anionic isozymes 20 (Campa, 1991) . Another possible role for peroxidase in plant-microbe interactions is in the production of molecules with antimicrdlial activity. Peroxidase and polyphenol cxidase (PPO) catalyze the oxidation of phends to quinones which have greater antimicrobial activity than the substrates (Gaspar, 1982). Peroxidase also participates in the generation of active oxygen species, such as supercxide and peroxide, which are toxic to microbes (Elstner, 1982: Katsuwon and Andersm, 1989). In summary, plant peroxidases have been studied extensively. Yet because of the diversity of substrates which it can accomodate, direct evidence for the role(s) of peroxidase has eluded researchers. None the less, the strong correlations between plant resistance to microorganisms and peroxidase activity make peroxidase an important target for study in the Rhizobium-legmne symbiosis and therefore is the main focus of this Q'IAPI'ERQ‘IE _NQQGENEWCEHRIHEEIICIMTWOFFEWHMSE ACTIVITYMCLOVERKDI‘HAIISANDPEAKDI’SBYH-IIZOBHJM AND THEIR CELL-FREE W Extract Our objective was to test whether rhizobia elicited a peroxidase- dependent plant defense response which could affect the determination of host specificity in the Rhizobium-legume symbiosis. Heterologous Rhizobium leguminosarum biovars elicited increased specific activity of salt-elutable peroxidase from the surface of pea arfl clover roots. Likewise, the cell-free supernatant of B. leguminosarum bv. M also elicited increased peroxidase activity from clover roots. The excreted elicitor of peroxidase activity was flavours-dependent, heat stable, and ethanol soluble. Treatment of clover seedlings with the heterologous cell-free supernatantdecreasedthenumberof infectedroothairshlt not the number of nodule initiations formed by bv. trifolii. In infected root hairs with shepherd's crook deformations, the stain for peroxidase activity only accumulated at the site of infection thread initiation. Heterologous bv. 1i_cj.§_e_ caused irregular root hair deformations with stain for peroxidase activity accumulating over the entire deformation where the bacteria were attached. In isolated clover roothairs, percxidaseactivitybegantoincreaseéhrsafter heterologous inoculation. In contrast, inoculation with homologous bv. trifolii suppressed peroxidase activity below the level of the 21 22 uninoculated control. 'Ihe suppression was mimicked by treatment with purified EPS from bv. trifolii. After 12 hrs, the specific activity of roothairperoxidaseincreasedtothelevel cfthecontrol. 'Ihusduring the period from 6-12 hrs after inoculation, suppression or elicitation of peroxidase activity may affect the structure of the root hair walls to either facilitate or prevent penetration by the bacteria. 'Ihe salt-elutable peroxidase from pea roots migrated as one weakly acidic isozyme in alkaline native gel electrophoresis which had greater activity after inoculation with the wild type biovar trifolii. Neither the bv. trifolii strain aired of its Sym plasmid nor the strain cmrtainingtleclcredBkbfmgmentcfpSymwhidrcarriesflscommonggd genes elicited an increase in the specific activity of pea root peroxidase. 'Iheclomd14kbregimcontainingthecomma1_no_dgenesard theh_sn_genes restoredsome elicitaticn ofperoxidasebutnottcthe level of the wild type. Its 14 kb regicm is not sufficient to elicit the wild type level of peroxidase and therefore additional regions of pSym must play a role in the elicitation of peroxidase. Single ‘Ih_5_ mutations in _nofl or Q effectively reduced the level of peroxidase activity elicited by these mutants which correlates with the extended host range phenotype for a _noiil mutant of bv. trifolii. 'Ihe wild type elicitation of peroxidase can be overcome by presence of homologous _hg genes inhybrid recombinants. ‘Ihus, _hggenesmay interact withothers present on pSym to control host—specificity throlgh the modification of an elicitor which increases root hair peroxidase activity which in turn may alter the structure of the root hair wall at the site of incipient penetration. 23 Introdlctim mizobium leguminosarum bv. trifolii forms a mutually beneficial symbiosis with Trifclium repers (white clover) as well as with other species of the genus Trifclium. Microscopy has revealed that the bacteriaattachtoroothairs, irrmcemarkedclrlingofroothairs, and thenpenetratetheroothairwall withintheconfinesofanadvancing infection thread (Nlrunan, 1959; Vincent, 1980) . Ultimately, the bacteriaarereleased fromthe infectim‘threadintcthecortical cells whichhavedividedandfcrmedarootnodule. 'Ihercotnodulethen becomes the site of nitrogen fixation. Fad) biovar or species of Rhizobium has a specific host range. For example, 3. legmninosanm bv. trifolii infects and modulates clovers, B. legmninosanlm bv. ligag nodulates peas and vetch, and 3. meliloti modulates alfalfa. ‘Ihe bacterial genes controlling host-specificity have been well characterized in a number of rhizobia (Horvath _e_t a" 1986; Debellé and Sharma, 1986; chrdjevic gt g" 1986; Martinez g g" 1990). Among the B. leguminosarum biovars, the host specific nodulation (E1) genes are mdFEle or mdFEIMN for bv. trifolii and bv. yic_iag_, respectively (Martinez gt gt, 1990; B. Rolfe, pers. comm.). 'Ihesegenesarefoundmthesymbioticplasmid (pSym). 'lhefilenotypecf bv. trifolii mm: with a m; insertion in fl is ch+ Fix- on peas, thus extending the host range of the strain (chrdjevic g g" 1985). Inordertcectendthehostrangeofbv. tigi_ag, however, avectcr containing bv. trifolii nodFERIMN (as in strain RIBOOERtZQO) must be introduced before efficient nodulatim of white clover will occur (chrdjevic gt a" 1986). 24 'Ihe biochemical basis for host-specificity is currently under investigation in a number of laboratories. Among various hypotheses, it hasbeensuggestedthatsuccessful infectiondeperdsontheabilitycf homologous rhizcbia to avoid eliciting a plant defense response (Vance, 1983: Djordjevic g _a_1_., 1987a; Hihler gt at" 1991). Plant defelse responsesmaybestructllralorchemicaldeterrantstoinfectimbya microorganism. Inducible defense responses include hypersersitive host cell death (HR), production of antimicrobial firy‘toalexins, ard the deposition of wall-strengthening polymers such as lignin, suberin, or extelsin at sites of attempted penetration by a pathogen (Misaghi, 1982). Some investigators have reported that certain Rhizobium mutants altered in EPS production are capable of eliciting an HR—like necrotic respase in plant roots (Djordjevic gt gl_., 1988; Pfihler e_t_ a_1., 1991). M suggest that the EPS of homologous rhizobia allows successful infection to occur by masking cell surface determinants of plant deferse. This interpretation carries the tacit assumption that heterologous rhizobia, lacking the proper EPS, induce plant defense responseswhidlcontrihltetoits inabilitytcinfect. Itislikely that the determimtim of host specifcity involves a combination of mechanisms andplant defense maybeoneofthem. Host-specificity is determined (hiring the interaction of Rhizobium withthehostroothairsbltpriortoinfectimthreadfcrmatimaiarrl Hubbell, 1969). Al-Mallah _e_t gl_. (1987) overcame the barrier to host specificity by treating clover roots with a mixture of cellulase and pectolyase. 'Iheseenzymesdegradedthewall atthetipofroothairs and allowed the heterologous g. tgt_i to nodulate white clover. 'nms, 25 ttewallofcloverroothairsappearstobeanimportantcompeentin the determination of host-specificity. ‘Ihe primary plant cell wall is a matrix of cellulose fibrils, hemicelluloses, and pectils interconnected with a network of hydroxyprolire—rich glycoprotein called extesin (Iamport, 1986: Varrer and Lin, 1989). In addition, a mnnber of enzymes are associated with the cell wall including ionically bound peroxidase (Gaspar gt g_l_., 1982). Peroxidase is capable of forming isodityrcsire links in extesin and polymerizing the aromatic constituents of lignin and suberin in 3.1.133 (Everdeel gt g" 1988; Lewis and Yamamoto, 1990; Espelie and Kolattukudy, 1985) . 'ltms, increased peroxidase activity during plant defeserespesesisthoughttoirereasetleamotmtofcross-linksin cell wall polymers which then prevents peetration by a pathogen. Albert E g. (1986) isolated peroxidase from tie surface of bean roots grown in both sterile and non-sterile soil. The activity of peroxidase isolated from non-sterile plants was greater than from sterile plants, suggesting that root colonizing organisms in tl'e m1- sterile soil stimulate greater peroxidase activity. A systemic increase inperoxidaseactivityisapartoftleinducedresistanceresporsein cucmnber inoculated with a fngal pathogen (Hammersdlmidt g g" 1982). 'Be activity of two specific peroxidase isozymes increased in a resistant barley alltivar inoculated with tie pathogen m gtatlinis. 'ne activity of the two isozymes reither increased in wounded plants, nor in us rear isogenic susceptible cultivar after inoculation (Kerby and Sommerville, 1989). Inthepresertstudy, peroxidase activitywasusedasameasureof a plant defese respese during tl'e early interaction between rhizobia 26 and tle roots of pea and clover. We hypothesized that rhizobia influence peroxidase activity, particularly in root hair walls, which thencartributestottesuccessorfailurecf infection. Wecompared peroxidaseactivity from rootsandroothairsafter'inoculatimwith homologous and heterologous rhizobia, as well as strairs with various geetic constructiorstoexploretrercle ofbacterialhflgeles in eliciting a defese response. Materials am nethods Bacteria ard cell-free bacterial awnings. BacteriaweregrownatBOConagarplatescertainingBergersen's modified medium (BIII, mzzo, 1982), amelded with tte appropriate antibiotieswhennecessary. 'nestrairsusedinthisstuiy, their souroe, and relevant characteristics are listed in Table 1. Bacterial inoculum was prepared by suspendirg 5 d-cld bacteria in nitrogen-free Fahraeus medium (-NF, Dazzo, 1982) and adjusted to a desity of Klett 5 (5 x 107 cells/ml) as measured with a Klett-Summerscn calorimeter using the no. 66 red filter. To obtain cell-free bacterial washings, bacteria were grown for 3 d with and without 4 uM 4',7-dihydroxyflavore ([1117) or naringenin (Nar) on BIII agar plates. 'Ihe bacteria were scraped off the platesandsusperiedinliqrid-NFmedimn,gertlystekenfcr1hr, and the) celtrifuged to pellet cells at 10,780 x g for 30 min. This bacterial wash fluid was sterilized by sequentially passirg tie supernatant through 0.8 um, 0.45 um, arr! 0.2 um Millipore filters. 'Ite filtrates were diluted with sterile -NF medium to an 0D245 nm which was standardized among treatments for each experiment. 27 IsolatimofEPs. StrainANUB43wasgrownataocforSdonBIIIagarplateswithout flavore. Cells weresusperledinpesrhateblfferedsalire (PBS), stirred for 30 min at4C, andthen centrifuged at 16,270ngcrlhr. 'nesupematantwasrenovedandceeeltratedtcca.20mltmdervacmlm onarotaryevaporator (40 C). Two volumes of cold 99%ethanol were added and tie mixture allowed to precipitate while gently stirring for 24hrat4c. ‘Bemixb.1rewastlence1trifugedat20,190ngorlhr at4C. 'neslpernatantwasremovedandtleEPSpelletdriedmrder vacuumatroomtemperaturefor30min. 'lhepelletwasredissolvedin ca. 100 ml of water, rapidly stirred at 4 C until tl'e solutim was homogeeous, and then placed in dialysis tubing (12,000-14,000 MWC). 'neEPSsolutimwasdialyzedagainstwaterfchdatllC. Finally, t1esolutimwasceee¢ratedtoce10mlontherotaryevaportatcn andthellyquilizedfcrstorage. Seedsterilizatim Seedsofbothpeacv.LittleMarvelarriclovercv.Dutd1Whitewere usedunlessotlerwiselisted. Seedsweresurfacesterilizedbyshaking seedsin70%ethanolfor4minfollowedby3x10minwasheswith1/lo strength commercial bleach, and finally several washes with sterile water. Plantgrowthardirmflntim Forsmdiesmperoxidasefromcloverroots,slrfacesterilized clover seeds were embedded in blocks of -NF medium solidified with 1 % mifiedagaranisuspefledmstainlesssteelwiremeshsupportsover 50mlofliglid-NFmedi1min9ancoveredglassdistes(mzzo, 1982). CloverseedsweregeminatedardseedlingsweregrownforBdina 28 growth chamber with a day/night regime of 14 hr light/10 hr dark, 23 C day/20 C night, and 70 % relative humidity. For inoculation or treatmert of roots in these wire mesh assemblies, tte original plant growthmedium ineadldishwasremovedardrellacedwitheitherSOmlof bacterial inocllum suspelded in fresh -NF or 50 m1 of sterile bacterial wash fluid for 24 hr. For studies m pea roots, surface sterilized seeds were germinated in tie dark in sterile water for 2 d, then placed on -NF agar plates (3 seedlings per plate) and incubated vertically in the growth diamber for 2d. marootsweretlenimculatedbyapplyinngrcpscfaKletts bacterial suspesion to each root. 'Ite plates were then covered, inclbated flat atroomtenperature for 1hr, andthenincubated vertically in us growth chamber for 24 hr. heathen: of clover roots with isolate! EPS». the EPS from strain ANUB43 was dissolved in sterile -NF medium at concentrations of 5, 50, and 500 ug/ml. EPS solutions were added to clover roots in wire mesh assemblies and incubated for 24 hr. m1 of roots. Cloverandpearootswere removedafter24hrscf inclbaticnwith bacteria, bacterial wash fluid, or purified EPS. Clover roots were obtairedby immersingthewiremesh inadishof liquidnitrogel. 'Ihe frozenrootsweretlenbrokencffintcachilledbeaker. marootswere cut from cotyledons with a razor blade. Isolated roots were stored at -20 C. Clover grown: an! inacflatim for studies with root hairs. Seedsweregerminated for2doninverted-NFagarplates. Seedlings were inoculated by gently shaking plants in a suspension of 29 bacteria (KlettS) for30min. Formicroscopyofroothairs, seedlings were then incubated vertically on —NF agar plates for 1-5 d. For isolation of root hairs, inoculated seedlings were rinsed with sterile water, separated on moist filter paper, covered, and then ireubated for 0, 6, 12, or24hrintlegrowthchamber. Seedlingswerethenplacedin glass vials and immersed in liquid nitrogen. 'Ite frozen vials were shaken vigorously according to tie method of Gerhold gt Q. (1985) to fracturetletteroothairs. Fragmentscfrootswerepouredcut, the vials were thawed, and 1.5 ml of 1 M NaCl in water was added to each vial. ‘Ihe irside walls of the vials were rirsed with the NaCl solution tosuspeidtheroothairs, ardtlenroothairsuspesioaswerepooled. Intact seedlings were placed on glass microscope slides, a coverslip added, and the substrate mixture (12 mg 3,3'-diamindaenzidile in 3 ml 60 mM Na-K-phcsphate buffer, pH 5.5, with 50 ul H202, Zaar 1979) injected under the coverslip. ‘Ite seedlings were rinsed with -NF medium afterSmin of incubation atroomtemperature. Roothairsweretten observedmaZeissmotomicroscope usingtrequartz-halogenlampfcr brightfield illuminatial. Photos were taken with Kodak Plus-X black and white film and Kodak Enctachrome color slide film. Isolatim of permfidme. To isolate peroxidases from root surfaces or root hairs, plant tissue was placed in 1 M NaCl and subjected to a sonic bath (Cole—Parmer UltrasonicCleaner) fcr20minat4C. 'misprocedureelutedproteils from surfaceswithminimaldamagetocells. 'nedebriswaspelletedby centrifugation at 5000 x g and the supernatant trarsferred to dialysis tubing (12,000—14,000 MWC) and desalted in water for 2 d at 4 C. The 30 proteincontentof samples wasmeasuredbytteBradforddyebinding assay using BSA as a standard. Peroxidase my. For kinetic studies of peroxidase activity, tle total salt-eluted protein solutim was assayed for peroxidase activity by adding 1-100 ul ofsampletolmlcfreactionmixtureinacuvette. 'nereaction mixturewas freshlymadeeachdaywith 50ml of 10mMNa-plcsphate buffer, pH 6.0, 125 ul guaiacol (Sigma), and 350 ul H202 (Hammerschmidt g g" 1982). 'Be increase in 00470 due to the formation of tetra- guaiacol was measured for 3 min on a Gilfcrd Respesetm UV/VIS scaming spectrophotometer. 'ne Gilfcrd kireties software package was used to calculate tte initial rate of tle reaction as 00470/min. We: of isozymes by electrqhoresis. Native polyacrylamide gel electrophoresis was performed to separate acidicperoxidase isozymes. Alkalirenmninggelsweremadelmmthick with 7.5 % acrylamide in 0.15 M Tris-Hm buffer at pH 9.3. 'Be stacking gels were 2.5 % acrylamide in 0.02 M Tris-phosphoric acid buffer at pH 6.7. 'Be lower electrode buffer was 0.1 M Tris-HCl, 11-! 8.8, arri tte upper electrode buffer was 0.04 M Tris-glycine, pH 9.6 (Biochemical Handbook). Samplescmtainingz ugofproteinweremixedwithSXsample buffer (0.5 M 'I'ris-HCl, pH 6.8, 0.05 % w/v bromphenol blue, and 10 95 glycerol v/v). Samples were run through tie staddng gel at 10 mA, and thelnmatacestantcurrentof 24mALn'rtiltlebrcmpierelbluehad migrated 3/4 of tte length of tie gel. 'ne electrophoresis apparatus was cooled by runnirg tap water. 3 1 Staining wls for pendfhse activity. Perciddaseisozymeswerevisualizedbysoakingttegelina solution containing 190 ml of 50 mM Na-acetate buffer, pH 5.5, 40 mg of 3-amino-9-ethyl carbazole (Sigma) dissolved in 10 ml of N,N-dimethyl formamide, and 66ul ofH202. 'nereactionwasstoppedafterZOminby rinsing gels with water. Gels were stored in a sole of 50 % methanol, 5 % acetic acid v/v in water. Beatstabilityarrletharelfractjrnatimofcell-freebacterial washings. ‘Ihe bacterial wash fluid from bv. M12 strain 300 which elicited clover peroxidase (00245 0.18) was divided into two aliquots. Ore aliquot was autoclaved for 20 min and then certrifuged to remove any precipitates. The original and the autoclaved aliquot were each added to clover roots in wire mesh assemblies (50 ml per assembly) and incubated for 24 hrs. Roots were isolated and peroxidase activity was assayed as described above. To fractionate the bacterial wash fluid, .2 volumes of cold 95 % ethanol were added to precipitate ethanol-insoluble components. The mixture was certrifuged at 10,000 x g for 40 min at 4 C. 'De ethanol supernatant was removed, evaporatedtodrynessurdervaclmm, andthe residue redissolved in a volume of sterile -NF medium equal to the original sample volume. 'Ihe ethanol-ilsoluble pellet (primarily polysaccharides) was also redissolved in tie same volume of -NF medium. Fifty ml each of the ethanol-soluble and tie ethanol-insoluble fractions were incubatedwithcloverrootsfor24 hr. ‘Iherootswerethen isolated and tie peroxidase activity was assayed. 32 Infectim threal arr! tulle initiatim bioassays. 'Be cell-free supernatants from broth-grown bacteria were used in infectim thread and nodule initiation bioassays. Bacteria were grown at 30Cinshakenflasks (175RPM) containingBIIImedium for2—3d. 'nemediumwassupplemertedwithz uMIiiFornaringenin forANUB43 and R1300, respectively. 'Ite bacteria were pelleted by centrifugation at 16,000 x g for 30 min and tie supernatants were filter-sterilized as previously described. Supernatants were diluted with BIII medium to an 00245 of 0.625-0.633. Seeds of the clover cultivars Ditch White, Iadim, louisiana S-1, and Nolin & laBorde '83 were germinated in the darkoninverted-NFagarplates forld, andthenseedlingswere transferredtofresh-NFagarplatesandiJeubatedvertically forldin ttegrowthdlamber. Seedlingrootswerettentreatedbyaddingwulof bacterial supernatant and covering with a sterile 18 mm2 coverslip positieed with tie lower edge just below tle root tip. After incubating for 4 hr at room temperature, tte coverslip was lifted and 10 ul of ANU843 inoculum was applied without rirsing and the coverslip replaced. 'ne final inoculum density was 105 cells/seedling. Seedlings were incubated vertically for 4 d, then mormted on microscope slides and stained with 0.01 % methylee blue dissolved in -NF medium. Recess stainwasrirsedawaywith-NFmedimnandttenumberofinfectim threadscountedusingphasecontrastmicroscopy. later, tieseedlings were submerged in 33 % bleadlandclearedundervaclum for15min. 'Ihe seedlings were rilsed with water, restaired with nethylee blue, and the rumberofnodule initiatiorsintterootcortexwerecounted. Root legthwasmeasuredfromtl'eroottipmarkmadeatttetimeof ireculation to tte root tip at tre time of staining. :33 Table 1. Ahizobiua leguainosaruw strains used in this study. nod henot e strain no. description clover 253 source/ref. bv. trifolii A80843 wild type (wt) 4 - Rolfe/Djordjevic 1983 A80845 pSya- deriv. of 843 - - Aolfe/Djordjevic 1983 AH0291 843, pSyI ggd§;:1n5, Arr delayed + + lolfe/Djordjevic 1985 A80251 843, pSyn gg§£::1n5, KIr + - Aolfe/ieinaan 1988 ANUD32 845p11032 (14kb HindIII iragaent nodJICBADlKRLHn froa 843 pSya, + - Aolie/Scbofield 1984 cloned into p11240), Cbr and Djordjevic 1985 AN0910 845p110324910 (8 kb fraglent - - Rolfe/Innes 1985 nodJICBAD in pxtz30), Cbr, Kar bv. viciae 81300 wild type - + Rolfe/Brewin 1980 5039 81fr deriv. of wt 248 with Hijffelaan/unpub. pahlzzins in uonsyabiotic gene, (Ir - +vetch 881601 5039, pALIJI nodllzlns (=p11601), [Ir + delayed Hijffelaan/Hijffelaan +vetch 1985 and Salzwedel unpub. 1003 Rifr deriv. of wt 1001 - + Squartini [Ph.D. Dissertation hybrids A80290 81300p81290 (9 kb Banal frag. nodlslhll frol pSya + + Rolfe/Diordjevic 1986 of A80843 in pltZ30), III 843-85 th43plix85 (8.4 kb [pol fragrant (8ac+ vetch) Squartini and Salzwedel /unpub. fro: 811001 in 911230), Sar iAbhreviations and concentrations for antibiotics: la, kanaaycin 30 ug/al, Cb, carbenicillin 75 ug/al, SI, streptoaycin 250 ug/Il, Aif. rifalpicin 20 ug/al. hAddress of Sources: Barry Rolfe, Plant Aicrobe Interactions Groups, Research School of Biological Sciences, Australian, national University, Canberra, 1.0.1. 2601, Australia. Carel iijfflelan, Dept. of Plant Molecular Biology, Leiden University, lonnensteeg 3, 2311 VJ Leiden, The letharlands. ” Andrea Squartini, Bipartilento di Biotecnologie Agrarie, Universita di Padova, via Gradenigo 6, 35131 Padova, Italy. 34 Insults Heterologous rhizobia elicited an increase in the specific activity ofperoxidaseeluted fromtherootsofpeaandclover (Fig. 1). The activity from pea roots inoculated with bv. trifolii was 2-fold greater thanfromrootsttnseinoibatedwithhomologoasbv.v_ici_a§orthe-NF control for 24 hrs. Similarly, the activity from clover roots incubated with bv. we strairs 5039 and 1003 was significantly greater than from the -NF control, but not significantly greater than the homologous combination after 24 hrs. The peroxidase activity from clover roots inoculated with strain R1300 was not significantly greater than either the -NF control or the homologous W843. Native polyacrylamide gel electmxnhoresis shown in Fig. 2 revealed 4 acidic peroxidase isozymes from clover roots. One sharp band of activity was extremely mobile, 2 broader bands had moderate mobility, anflthe4flnbandwassharparflappearedjustattheinterfaceofthe stackingardthe7.5%nnminggel. 'Ihis4thbandstainedmore intensely inthe lane containingprotein fromcloverrootsinooilated withheterologous R1300 comparedtotheotheerarflswhidnarpeared equivalent to the intensity for samples from the homologos combination. In contrast, pea roots yielded only one sharp band of activity with low mobilityjustpasttheinterfacebetweenthestaddngardmfinggels. 'Ihe intensity of this isozyme was increased after inoculation with the heterologous ANU843. 35 2000 1 E Worm bv. trifolii E 1800'? EX! 3W bv. rich: 2 : // 5 1600-: f/ A g _ < 0’ 1400-“ __ / Q g : / E E 1200-: § LaJ E . Q_ 1000- (I) \O 2 ‘ / LL, r; 800-: 1 y , / T m . _ E 8 600-: § i V 2 ’w / O 400: M V E 3 / / CL 200: / / / o: / / . , / . . _, -~r so 300 5039 1003 -NF 343 300 5039 1003 BACTERIAL INOCULUM WHITE CLOVER PEA Fig. 1. Specific activity of peroxidase el from clover and pea roots 24 hr after inoculation with 5 x 10 cells/ml suspension of wild type rhizobia. Each bar represents the mean of at least 3 experiments +/- SE. Nitrogen-free Fahraeus medium (-NF) was used as the control. 36 Fig. 2. A. Native polyacrylamide gel stained for peroxidase activity with 3-amino—9-ethyl oarbazole. Peroxidase was isolated from clover roots 24 hr after inoculation with rhizobia. Each lane was loaded with 2 ug of protein. Treatments inclined: lane 1 horseradishperoxidase, lanez -NFcountrol, 1ane3ANUB43, lane 4 R1300. B. Native polyacrylam ide gel stained for peroxidase activity with 3-amino-9-ethyl carbazole. Peroxidase was isolated from pea roots 24 hr after inoculation. Treatments included: lane 1 -NF control, lane 2 ANU843, lane 3 R1300. C. Nativegel madewitha 3-10 %gradientofacrylamideto illustrate the homogeneity of the band at the interface of 7.5 % native gels. 37 38 The location of peroxidase activity in clover root hairs was shown by the deposition of the insoluble brown product of DAB oxidation (Fig. 3). Uninoonlated clover root hairs had an even distribution of golden- brownstainovertheentireroothair (Fig. 3A). Cloverroothairs formed characteristic shepherd's crooks after inoculation with homologous ANU843 (Fig. 38,6). In these root hairs, stain for peroxidase activity only accumulated at the point of infection thread initiationbothldandeafterinoculation. 'Iheinfectionthread itself did not accumulate any more stain than the background. In contrast, after inoculation with heterologois R1300, clover root hairs show dark deposits over the entire irregular deformation (Fig. 3D) . 39 Fig. 3. White clover root hairs stained for _ig situ pemidase activity with DAB + H202, 113 5.5. A. uninoculated B. 1 d post-inoculation with ANU843 C. S d post-inoculation with ANU843 D. 5 d post-inoculation with R1300 Arrows inndionte areas of enhanced staining. 38 41 42 Cloverroothairswere isolatedat 0, 6, 12, and24hrsafter inoculation to determine when peroxidase activity began to increase. 'Ihe specific activity of peroxidase elicited by R1300 began to increase 6hrsafterincoilationardwasgreaterthaneitherthe-NFcortmlor the homologous ANUB43 at 12 and 24 hrs (Fig. 4). Surprisingly, the homologous ANU843 suppressed the specific activity of peroxidase from roothairscomparedtothe-NFcontrol at0and12hrafterinoo11ation. ('Ihose samples designated time 0 were harvested immediately after addition of inconlum, however preparation of the tissue took aproximately 15-30 min). At 6 hr after inoculation, specific activities for the -NF control and the ANU843 treatment were not significantly different. 'Ihe specific activity began to increase 12 hrs after inoculation with ANU843, although not significantly greater than the -NF control. PEROXIDASE SPECIFIC ACTIVITY 43 175 H —NF control H Rt843 150 H R1300 ’6 125 E \ .E 100 E \o r\ 75 d. O O 50 V \ 25 0' r I I I I l I I I Y I l I I I I I l I T U 0 6 12 18 TIME AFTER INOCULATION (hr) Fig. 4. Specific activity of peroxidase from isolated cloverroothairsato, 6, 12, or24hrafterinoculation with wild type rhizobia. Data points represent the mean of 2 experiments +/-SE. 24 44 'IhepurifiedEPS frouu homologousANU843 alsosurpressedthe specific activity of peroxidase from clover roots (with root hairs intact) after 24 hrs of incubation (Table 2). Treatment with EPS at 500 ug/ml yielded aproximately 2-fold lower specific activity of root peroddasecomparedtothe-NFcontrol. TreatmentwithEPSatSorso ug/ml resulted in peroxidase activity similar to the -NF control. Table 2. Effect of purified EPS from bv. trifolii ANU843 on specific activity of peroxidase eluted from clover roots. treatmennt specific activity OD min/mg protein 7 -NF control fl/ 794 EPS 5 ug/ml 857 EPS 50 ug/ml 767 EPS 500 ug/ml 416 aSeedlings were incubated with EPS solutions for 24 hr. bData are from one experiment. The heterologous elicitation of pea root peroxidase was dependent upon the presence of pSym in bv. trifolii. 'Ihe pSym-cured strain (ANUB45) did not elicit an increase in peroxidase activity from pea roots, butrathersuppressedtheactivitybelowthelevelfromthe homologous combination (Fig. 5). Strain 845pRI'032 (containing the entire 14 kb {Q region from ANUB43) elicited slightly more peroxidase activitythanstrainANU84S, althoughnotasmudnasdidthe heterologous wild type ANU843 (Fig. 5). Strain 845pRt032A910 (8kb fragmentcontainingcommon_rc_dgeneshntnctthe_h§n_genes) didnot elicit an increase in specific activity from pea roots with the level similar to that found with strain ANU845. For clover roots, ANU845, 845110032, 8451111032A910, and ANU843, all elicited SpeCific activities 45 similar to the -NF control. The qnecific activity of peroxidase from pea roots inoculated with the bv. trifolii rch: :1an mutant (strain ANUZ9‘7) was aproodnnately 2- fold less than the activity elicited by the heterologous wild type W43 (Fig. 6). Similarly, the peroxidase activity from clover roots innoonlated with the bv. viciae nodEstng mutant (strain RBIéOl) was less than the activity elicited by the heterologous wild type 5039. On pea, the bv. viciae nch: :1an mutant elicited peroxidase activity similar to that of the homologous wild type 5039. On clover, however, the bv. trifolii nodE: :'m_s_ mutant elicited slightly greater specific activity of peroxidase than did ANUB43. PEROXIDASE SPECIFIC ACTIVITY 46 20003 E bv. 101.0111 genetic background 1800- [A 1600- j! ’6 - E 1400 ? E 1200-; / E 1000-3 ? O I I; 800-; I T é . .8, 60°"; Ti // 9 wn/éé/ éfg/T : / * 2001////// //// o: / / Z - s 431' w? pSJm- 032 K10 ' 43$ «'1 pSym- 032 1910 BACTERIAL INOCULUM WHITE CLOVER PEA Fig. 5. Specific activity of peroxidase eluted from clover and pearootsZ4hrafterinconlationwiththepSym-curedbv. trifolii soain or recombinants containing cloned fragmennts frompSym inthepSym-curedbackground. -NF=control, wt= wild type, pann- = ANU845, 032 = scsprrroaz containing 14 kb pic: gerne region, A910 = 845pRI‘032A910 containing 8 kb nodIICBAD. Barsrepresentthemeanof3experiments+/-SE. PEROXIDASE SPECIFIC ACTIVITY 47 20004 :CIBJeumirmnmmbvdrifelfi 18001mmmmM.fimn 16005 ///’ ,4 ,r" 0‘ 1400-3 / E 1200-3 IT ? £3 10004: T é T ? rx 800-3 5?, 6005 :j;:’/// 1- :::: ’ ‘ 3 E // / T 4001 ./”//’/:::: y//’ 200% ::::’////,// :::j’///V,// 0‘ . . . . . . . . . . ,, BACTERIAL INOCULUM WHITE CLOVER PEA Fig. 6. Specific activity of peroxidase eluted from clover and pea roots 24 hr after inoculation with suspensions of wild type (wt) rhizobia or their respective nch::Tn§ annd nchn:Tn§mutants. Barsrepresentthemeancheaqneriments +/- SE. 48 Rather effects of the Rhizobium hsrn genes were tested using hybrid recombinants. Specific activity of peroxidase from pea roots after inconlation with 8431316165 (pKI'KBS contains nodDFEIMN genes from R11003) was significantly less than that elicited by the heterologous wild type 11011843 (Fig. '7). Ms specific activity was also lower than that elicited by the homologous R1300. The specific activity elicited by R1300pRt290 (containingtherchERIMNgenesfromANUBfl) onpeaswasnot significantly different than the activity elicited by R1300. For clover roots, 843pK1‘K85, R1300pRt290, ANUB43, and R1300 all eliCited Similar peroxidase activity. The cell-free bacterial washing from heterologous R1300 grown with naringenin elicited increased peroxidase activity in clover roots while thebacterialwashingfromR1300grownwithouttheflavonedidnct (Fig. 8). IIhe specific activity after treatment with the bacterial washing from ANU843 grown with [HF was similar to treatmennt with washing from ANUB43 without flavone, R1300 without flavone, and the -NF control. PEROXIDASE SPECIFIC ACTIVITY 2000‘ 49 1800; 1600; 1400§ 1200§ 1000; 800% 600% 400% 200% (OD47O/min/mg) d 0 E bv. trifolii genetic background Em bv. grime genetic bockgroun I 1 ? ——1 \\\\\\\\\\\\\ i 1‘ \\\\Pd-1 85 in parieo WHITE CLOVER wipKTKBSwtpR‘bQO BACTERIAL INOCULUM PEA Fig. 7. Specific activity of peroxidase eluted from clover andpearoots 24hrafterinoculationwithwildtypeor hybrid recombinant strains carrying heterologous _hsLn_ genes. Wt = wild type, plums = plasmid carrying bv. 112113 nodDFELMN gennes, W90 = plasmid carrying bv. trifolii nodFERIMNgenes. Barsrepresentthemeanof3experiments +/- SE. PEROXIDASE SPECIFIC ACTIVITY 50 1200 i E cell-free wash fluid from bv. trifolii 1100: a cell—free wash fluid from bv 11:19.9. Q 2 El controls x 1000-; >< a = 52 \ - >< r: 800: >< E 700-: ’ ii \ j / 1C3 600: / g a- : >< >< o . K >< / >< 8 500: / >< / >< 400-: / E3 / 300: / >< / § 200: - . V s :2 s / s x -NF -NF Rt843 Rl300 Rt843 RI300 control +Nar +DHF +Nar TREATMENT Fig. 8. Specific activity of peroxidase eluted from clover roots 24 hr after treatment with standardized cell-free bacterialwashings( 45=0.08-0.12) frcmbacteriagrownon plates withandwi 4qulavone. Dataarefromone experiment. 5 1 The cell-free wash fluid from R1300 grown with [HF also elicited an increase in peroxidase activity from clover roots (Fig. 9). Autoclaved wash fluid retained its ability to elicit peroxidase activity. After fractionation of the wash fluid, the ethannol fraction elicited greater peroxidase activity than the -NF control (Fig. 10). The fraction containing the ethanol-insoluble pellet also elicited some clover root peroxidase activity compared to -NF, althogh it was less than the ethanol-soluble fraction. PEROXIDASE SPECIFIC ACTIVITY 52 1600. 1500§ 1400§ 1300§ 1200§ 1100§ 1000§ 900% 800% 700% 600% (09470/ min#108) 500 $1? Rfioo RI300 R1300 R1300+Nar control -f|avone +DHF +Nar autoclaved TR EATM ENT Fig. 9. Specific activity of peroxidase eluted from clover root524hrafteradditionofautoclavedbacterialwashing (00245“; 0.18) frole300grownwith 4 uM flavone. Dataare oneexperiment. PEROXIDASE SPECIFIC ACTIVITY 53 (OD47O/min/mg) \1 .8 600-l 5003 400 - ’ethanol-ir‘nsoluble r ethanoILsoluble control pellet supernatant TREATMENT Fig. 10. Specific activity of peroxidase eluted from clover roots 24 hr after treatmernt with the ethanol-insoluble pellet annd ethanol-soluble supernatant fractios of bacterial washings froleBOOgrownwith4uMnarirgenin. mtaarefromone experiment. 54 calmnesupernatantfrombroth-grownbv.y_igi_agredncedthemmber of infectedroothairsbutnotthernnnberofncduleinitiatiosformed by bv. trifolii on white clover cultivars (Table 3). Root length was nnotaffected. SupernatantfromANU843grownwithDIE‘increasedthe rnunnberof infectedroothairsandnoduleinitiatioscomparedtothe BIII control for olltivar Dntch White. Table 3. Effect of culture supernatant from broth-grown R. leguminosarum bv. viciae on number of infected root hairsja-nd nodule initiatios formed by AN0843 on white clover cultivars. Inf 1 Noilpl Root lent/pl (m) g BIII R1300 an; BIII R1300 _s_up _B_I__II R1_3___00 s_up 1 22.2 14.8 1.8 10.1 0.7 10.3 1.4 $0.4 10. 7 :1. 2 13.7 :1. 7 2 6.7 12.1 0.9 10.4 0.9 $0.4 1.4 10.4 8. 7 11.2 9.6 10.8 3 8.0 14.2 2.4 10.9 0.3 10.2 0.1 10.1 7. 8 :1. 4 7.2 10. 6 4 4.4 1:1.7 1.3 $0.4 0.0 t0.0 0.3 10.2 5.3 to. 8 4.8 10.6 aCultivar 1 often White, 2 Regal Iadino, 3 Iouisiana s—1, 4 Nolin a. LaBorde '83. bForty ul of filter-sterilized cultmre supernatant from R1300 grown with 2 uM naringennin ( 45 = 0.625) was incubated for 4 hr followed by inoculation with 10 cells/seedling of ANU843 without rinsing. Plants were innonbated for4d. Dataarethemeansperseedling+/-SE, n=9. C"I‘reatmennt of cultivar Dutch White with supernatant from ANU843 grown with 2 um um having anOD2 = 0.633 followed by inoculation with ANU843 resulted in 30.4 +/-3.% infected root hairs per plant and 4.1 +/- 0.7 Noi per plant. Discussion Anumberof factorsarelikelytocontrihntetothedetermination of host-specificity in the Rhizobimnn-legmne symbiosis. Although peroxidase activity is required for the normal growth annd maturation of plant cell walls, localized increases in peroxidase activity contribute todefenseagainst invadingmicroorganisms. Wehypothesizedthatsudna planntdefenseresposehinderedpenetrationofroothairsby 55 heterologous rhizdnia. Cytochemical localization of peroxidase activity inclorerroothairsprovidedtheinitialsupportforthishypothesis. 'meextensivestainingofirregularcloverroothairdeformatios in heterologous combination indicates inncreased peroxidase activity in resposetoattemptedponetrationbythebacteria. Sudnenzymeactivity colldmodifythestrnctureoftheroothairwallarritlmsmakeitmore diffionlt for the heterologous rhizobia to penetrate In contrast, the localization of peroxidase activity in clover root hairs infected by homologous bv. trifolii was limited to the site of infection thread initiation both 1 annd 5 days after inoculation. It is possible that thisperoridaseactivityrepresentsrqnairactivityoftlcmothair wall after the bacteria have successfully penetrated. 'n'ne liglnt golden stainevenlydistrihntedovertlcnmimonlatedrcothairsrepresentsttc peroxidasepresenntfornormalgrowth. 'nnemBsubstrateforperoddaselocelizationcrossestheroothair wallsincebothflnewallandtlccytoplasmwerestaincdinplasmclyzed activity associated with the infection thread. Therefore, the infection threadiscompatiblewiththeroothaircytoplasmandocebeyondthe initial barrier atthecell wall, doesnotelicitadefenserespose. Anassessmentofabortedinfectionthreadswasnotincludedinthis study. Peroddasefromisolatedcloverroothairswasassayedto quantitate the activity elicited by rhizdoia. The immediate suppression of peroxidase activity from isolated clover root hairs after addition of homologous inconlum suggests that some preformed factor is present in tlcinccllmnwhidnsnppresseetlcmrmallevelofperocidaseactivityin 56 roothairs. 'Ihedifferencebetweenthe-NFandANUBB treatnnenntswas notsignificant at 6hrs andmaybeduetostatistical variation forthe two replicates. The specific activity began to increase up to control levels between 12 and 24 hr after inoculation with ANU843. 'Ihis correlates with the enhanced deposition of stain at the site of infection thread initiation since bacterial attachment, onrling of the roothair, andinitiationof infectionthreadscantakeplaceduringthe first 24 hr. To test the possibility that homologous EPS was the preformed factor suppressing peroxidase activity, clover seedlings were inonbated with three concenntratios of EPS from ANU843. Inncubation with 500 ug/ml EPSsuppressedtheactivityofperoxidasefrcmcloverroots (roothairs still inntact), suggesting that homologous EPS may block peroxidase activity. The homologous rhizobia aggregate at root hair tips annd may provide a localized concenntration of EPS that is effective in supressing root hair peroxidase activity. To mimic that effect, a large exogenous cocentrationcfEPSisrequiredbecausetheEPSisspreadontoverthe entireroot epidermis ratherthanconcentratedattheroothairtips. Inacflition, enzymes releasedfromcloverrootsarecapableofdegrading bacterialpolysacdnaridesanflmayhavereducedtheacmal cocenntration of EPS within the 24 hr innoubation (Dazzo g 31;, 1982). A lower cocenntration of EPS applied to clover roots may be effective earlier than24 hrsaftertreatment, althonghwedidrctdirectlytestthis possibility. By pretreating clover seedlings with oligosaccharides from honologonsEPSanriCPS, thenumberof infectionthreadsfcrmedbybv. trifolii strain 0403 onwhite clover is increasedandhasbeentermed infection-related biologin activity (Abe gt _a_1_., -1984) . The 57 suppressionofroothairperoxidasebyEPSmaybeapartofthis biological activity. Incontrast, bv. Moellselicitedanincreaseinthe perocidase activity in heterologous combination with isolated clover roothairs. 'IheincreasebeganbetweenGandDhrsafterinoculation while the rise in activity elicited by homologous ANU843 began 12 hrs afterinoculation. 'nnisperiodfrom6t012hrsafterinnoculationmay bethecriticeltinnewhensuccessor failureof initialpennetrationis determined. ‘nneeffectivenessofadefenseresponseoftendependson its rapid initiation annd development (Misaghi, 1982). ‘Ihe peroxidase responsefcr isolatedcloverroothairs ismorerapidanndmoreintense for the heterologous than for the homologous combination. ‘nnus, our model for the determination of host specificity includes the transient suppression of peroxidase activity in root hairs by homologous rhizobia and their EPS. During this suppression, the root hairwall maycontinnuetogrow. Withoutthenormal level ofperoxidase activity, however, fewer cross-links annog polymers would be formed during extension of the wall. Sudn a wall would be more susceptible to penetration by the homologous strainn. After penetration, highly localized peroxidase activity could facilitate repair of the wall at the site of bacterial penetration annd infection thread initiation. At the site of attempted penetration by heterologous rhizobia, a rapid increase in peroxidase activity would contribute to increased cross-linking of wall polymerssuchasextensinandsuberin. 'Iheroothairwallswould thenbemore resistanttopennetrationbythebacteriaandmayactually exclude the heterologous rhizobia. The localization of activity and the time course for elicitation in clover root hairs provide evidence to 58 support our plant-defense hypothesis. Elution ofwallproteinns fromisolatedcloverroothairs unavoidably includes cytoplasmic proteins as well. Therefore we determined wlnether the activity of peroxidase from the surface of inntact roots (which includesroothairs) was affectedbyrhizcubiasimilarto alterations observed in root hairs. The heterologous combinnation of AN0843 on pea significantly increases the specific activity of peroxidase within 24 hr in a mannnner similar to an incompatible pathogenn. Similarly, the heterologous combination of R1300 on clover resulted in a slight increase in peroxidase specific activity from clover roots, altlnghtheincreasewasnotasgreatasforpea. Forperoxidase eluted fronthe surface ofbothcloveranndpearoots, homologous rhizobia failed to suppress the specific activity. This may reflect the costitutive baclcgound of peroxidase presennt in the walls of all the epidermal cells of the root in which a slight suppression in root hair peroxidase is nnot detected. Heterologous elicitation of peroxidase activity, however, isadominnannteffectanndthusisdetectedin peroxidase eluted from roots and root hairs. 'nnespecificactivityofanenzymecouldbeincreasedduetol. an increase in the activity of existing ernzyme (e.g. increase supply of nneesssarycofactor), 2. gmsynthesisofanewisozyme, or3. a decrease in the total protein cocenntration (with total activity of peroxidase remaining constant). Total activity for peroxidase eluted from clover roots increased after heterologous inoculation (data not sham). 'nnerefare, peroxidase isozymes from the surface of roots were separatedbyelectrquhoresistodistinguishbetweenthefirsttwo possibilities. Initially, peronridase isozymes from unninnoculated clover 59 roots were squarated by isoelectric focusing followed by staining with 3-anninno-9-ethyl carbazole. Five well-resolved bands of peroxidase activity were apparent (data nnot shown). All but one isozyme migrated to a pI < 7.0 annd therefore routine separation of isozymes was performed using alkaline native gels to detect the acidic peroxidases. Four acidic isozymes from clover roots were separated in alkaline native gels and detected with the activity stainn. No new isozymes were observed after innoculation with heterologous rhizobia compared to the -NF control. Inoculation with heterologous bv. v_ic_i§ did ennhance the activity of only 1 of the 4 isozymes. Its low mobility suggests that the native enzyme was either very weakly acidic, had a high MW (possibly complexed with other root surface components), or both. In pea roots, only one isozyme was apparent in native gels. It also had low mobility and migrated just slightly farther than the low mobility clover isozyme. Itwculdbeinterestingtoanalyzetherootisozymesofotlerlegumesto determine whethertheperoxidaseisozyme increasedbyheterologous rhizobia was conserved amog the legumes. The magnitude of peroxidase activity elicited by bv. 31% on clover was substantially lower than that elicited by bv. trifolii on pea. 'Ihis may be due to the method of determining specific activity since the spectrophotometric assay detects percnxidase activity in a mixtnrreofproteinsfromtherootsurface. Onlylofthe4isozymes from cloverroots was increasedafterheterologous inoculationwhilea single isozyme was eluted from pea roots. Thus, we believe that the activity of the constitutive peroxidase isozymes from clover roots obscuresthe increase intheactivity of 1 isozymeafterheterologous inoculation. In comparison, the dnannges in the single isozyme from pea 60 rootsareeasilydetectedbythespectropnotometric assayofperoxidase activity. We reasoed that if the elicitation of peroxidase is an important determinant of host specificity, thenn elicitation should be controlled by the host-specific nodulation Q_ns_n) genes present on the Sym plasmid. 'Ihishypothesis wastestedby ineculatingrootswithstrainscontaining cloed fragmennts ofpSym inapSym-curedbackground. We finrithat deletion of the entire Sym plasmid from ANU843 (strain ANUB45) eliminates the heterologous elicitation of peroxidase on pea roots. 'Ihus, somebacterial factorgovern'edbygenesontleSymplasmidmustbe resposible for the heterologous elicitation of peroxidase activity. Surprisingly, the specific activity after innoculation with ANU845 is evennlessthanthe-NFcontrol. WespeculatethatthelackofpSym affects both the bacterial factors which elicit greater peroxidase activity annd which lead to snppression of peroxidase activity. 'nne14kb_ng_d_generegionofpSym fromAN0843hadbeenc1oed (Sdnofield g1; _a_1_., 1984) annd we determined its influence on the elicitation of pea root peroxidase using the strain 845pR1032. The ability to elicit peroxidase activity was partially restored by the presenceofthe 14 kb fragment comparedtoAN0845, butwasnotrestored towildtypeANU843 levelsonpea. 'nnissnggeststhatprodnctsoftle 14 kb 2g gene region contribute to peroxidase elicitation, but that othergenesontheSym plasmid, ortheinteractionbetweenntle_no_d_genss with others on the Sym plasmid, have a more profound effect on us elicitation of peroxidase. Strain 845pRI'0324910 contains the common goesdeJmBChntladcsttehgggenesncd—Lmlbflardelicitsless peroxidase activity than does 845pRI'032, with the level similar to the 61 elicited by pSym' ANUB45. This indicates that the common E genes aloe are innsnfficient to elicit the level of peroxidase activity observed after inoculation of pea with 845pRI032. 'Ihe _hg genes thennselves orthe interactionbetweentheflandcommonLoggenes may controlttecontributionoftheflgeneregionintheheterologous elicitation of peroridase activity. In clover, the resposes to 845pRI‘0324910, 845pRI‘032, ANU845, and ANUB43 are all similar. This may reflect the insensitivity of the spectrophotometric assay for a mixture of peroxidase isozymes from clover roots. 'Ihemutationofasinglegene, _nog, canextendthehostrangefor bothbv. trifolii andbv. yiclag. Ifourhypothesis iscorrectand heterologous elicitation of peroxidase activity hinders penetration by rhizobia, thentlosemutantswithextendedhostrangeshouldnologer elicit increased peroxidase activity. Therefore, the @31an mutannts of botln ANU843 and 5039 were tested for their ability to elicit peroxidase activity. Strain 5039_no_c!:;::Tn§ on clover and Bug-Quinn; on peas each elicited significantly less peroxidase activity than the correspodingheterologcuswildtypestrains. ‘Ihesedataarecosistent with our hypothesis and correlate with the nodulation ptnennotype of these mutants in that 503%:81‘115 forms small white, presumably Fix" nodules on clover (Salzwedel, unpublished) and Djordjevic et a1. (1985) report thatan 843E31h§mutanthasexterdeditslostrangetopeas (Nod+Fix-) . Therefore, we believe that the pivotal host range gene, mtg, functionsasanavirulenoegereinwhidntlelossoffunction leadsto increasedl'ost-range (virulence). 'Iheflgereproductmay affect the production of the bacterial factor which elicits peroxidase activity in the heterologous host which in turn affects the efficiency 62 of successful nodulation. A'Inngmutationintheflgeneofbv. yi_ci£e_doesnotaffectthe level of peroxidase elicited on pea; however, 843_no_c§::'m_5_ on clover elicits more peroxidase activity than ANU843. This increased elicitation of peroxidase may contribute to the delayed nodulation phnennotype for 843m: :‘Ih§ on clover. The level of peroxidase activity elicited by the 843_no@_::Tnn§ mutant and the 845pRI'032 strain on pea are similar. Since the 14 kb fragmennt in pRI'032 contains Egbert lacks the rest of pSym outside the 14 kb region, and the 843_no_dE_::Tn§ mutant containstheentireSymplasmidwithonlytheflgeneinter-rupted, the implication is that the interaction of _nofl with some as yet unncharacterized region of pSym is essential to the expression of the elicitor of peroxidase activity. 'BeflgeneistranscribedinthenodFERLoperonofpSyminbv. trifolii. We tested whether a mutation in Log; had a polar effect on theexpression ofgenes locateddownstream inthneoperonbydetermining the effect of an 84351131115 mutant on the elicitation of pea root peroxidase. 'Beflmutantelicitedareducedlevel ofpearoot peroxidase similar to the level elicited by the _rogz; mutant. 'Ihis indicatesthateithertle'mginsertioninflhasapolareffecton theexpressionofflwhichthen isthnekeygennecontrollingthe elicitation of peroxidase activity, or bothn fl and Q are needed for the elicitation of peroxidase activity. 'Ihehostrangeofmnizobiumcanalsobeextendedbyintrodncing homologous_hsn_genes inntoahneterologous wildtypebackground. Strain R1300pRI'290 is able to rodulate white clover (Djordjevic g a_l_., 1986) and the reciprocal strain 843pm<85 becomes hac+ on vetch (Squartini, 63 unpublished). Strain Rt843pK1‘K85 containing the bv. viciae nodDFEUdN genes on a multicopy plasmid elicits substantially less per'ooridase activityonpeasthandoestheparentalwildtypeANUMBsggestingthe presence of the bv. viciae hsnn genes modifies the expression of the _ elicitor encoded by the AN0843 genes. Strain Rt843pKI'K85, however, did not elicit greater peroxidase activity than the wild type bv. 15% strains on white clover. 'Ihis suggests either that heterologous _hg genesdonotoodeforanelicitorthemselvesorthatthewildtypebv. trifolii genes are dominannt. The recombinant R1300pRt290, and wild type bv. 2% R1300 elicited similar levels of peroxidase activity in clover roots. 'Ihis again may reflect the insensitivity of the spectrophotometric assay to detect changes in a single peroxidase isozyme within a mixture of isozymes from clover roots. no determine whether the elicitor of peroxidase activity in clover roots was an excreted metabolite, the cell-free washings from plate- grown R1300 were tested for elicitation of peroxidase activity. Such washingsfmmthehetemlogcusbv.yi_qi_a_eoncloverelicitedanincrease inperoxidase fromtherootsurfacewithin 24hrjustasdidthelive bacteria. However, onlythnewashingfromheterologcusRl300grownin thepresenoeofaflavone (IHFornarinngenin) will elicitanincreasein perocidase activity (Fig. 8). Therefore, the bacterial factor must be an ometed metabolite whose production, modification, or export is dependentonaflavone—induciblegene. 'Ihelanown_no_d_genesareinduced by flavoe, but the 14 kb _rngq gene region in 845pRI'032 is not sufficiennt to elicit wild type levels of peroxidase. ‘Ihe flavoe-induction of the elicitor from R1300 maydependonh_srn_genes inpart: howeversomeother genes, perhapsalsoinducedbyflavoe, maybefoundoutsidethefl 64 gene region and may have a large impact on the elicitation of peroxidase. 'Ihe dominant flavone from the clover rhizosphere, man, also induces the heterologous R1300 elicitor and thnus it is feasible that the elicitor is produced in the interaction of field-grown clover and soil— borne bv. Vic—iae cells. This elicitor from R13 00 is heat-stable since eliciting activity is retained after autoclaving (Fig. 9) . Both the ethanol-soluble and ethanol-insoluble fractios of the cell-free washn fluid elicited increased peroxidase activity with the greatest elicitation by the ethanol-soluble fraction. The ethanol-insoluble precipitate (primarily EPS) may have some innherennt eliciting activity, however there may also be a small ethanol-soluble molecule caught in the polysacchuaride matrix to acoounnt for the eliciting activity of that fraction. Bioassays were performed to test whether the cell-free washing which elicited peroxidase could also influence infection of root hairs. Treatment of white clover with sterile cell-free washing from homologous ANU843 followed by inoculation with ANU843 cells increased the rnumber of infectedrcothairscounparedtothecontrol. 'nnisisconsistentwith previous studies in which curling factors and biologically active polysaccharides were found in homologous culture supernatant (Fauchner gt_ g" 1988; Abe e_t_ 31., 1984). In contrast, the washning from the heterologous R1300 decreasedthe numberof roothairs infectedbyANU843 on clover (Table 3) . The number of nodule initiatios formed, however, was not affected by heterologous culture supernatant. We believe that thneexcretedelicitor frole300 acts atthne stageofrcothair penetration and that those few bacteria which overcome the peroxidase resposeelicitedbythesupernatantareuninpaired intheinnitiationof 65 nodules. Growth of ANU843 in broth-culture was not affected by addition of the culture supernatant from R1300 (data not shown). Root length was notaffectedbythetreatmont indicatingthesupernatantdidnotexert negative effects on plannts. Thus, the correlation betweon elicitation ofperoxidaseandthnedecrease innnumberof infectionsaftertreatmont with heterologous culture supernatant supports the hypothesis that the heterologous biovar elicits a plannt defense respose which is a determinant of host specificity. To coclude that Rhizobium elicits a plannt defense response, the following minimum criteria must be met: 1. the response is elicited to a greater extort by the incompatible than the compatible microorganism. 2. 'Iheresposeoccursintissuewherethemicroorganismislccated. 3. 'Iheresposemustoccurrapidlyatatimepriortoingressbythe microbe. 4. 'Iheresponsemustcontributetotheonclusionofthe microbe. The results presented here meet these criteria. We therefore propose that heterologous rhizobia produce an excreted metabolite whose synthesis or modification is depodont on flavone-inducible gones present on the Sym plasmid and whichn elicits increased peroxidase activityonthnerootsofheterologous legumes. We furtherproposethnat this rapid increase in peroxidase activity acts on cell wall polymers to increasetheintegrityoftheroothairwallandtherebyhinderits poetration by the bacteria. In contrast, the transiont suppression of root hair peroxidase activity by homologous rhizobia contributes to successful infection by delaying the normal polymerization of wall polymers and thnus facilitates poetration by the bacteria. We coclude that the increased peroxidase activity elicited by heterologous rhizobia is a plant defose response which contributes to expression of host 6 6 specificity during the infection process in the Rhizobium-legume syflniosis. GIAPI'ER'IWO HHZOBIIM WIN BV. 'I‘RIFOLII AND ITS HJRIFIED IPSWARAPIDWZATIONAT'IHEWOF CLOVER m HAIRS Alstract Inoculation of white clover seedlings with homologous Rhizobium leguminosarum bv. trifolii results in the rapid (30 min) iuncrease in the piofthemedinnnfronG.5toG.7 atthesurfaceof individualroothairs as measured by pH microelectrodes. Inoculation with bv. m or 3. meliloti results in a only a slight rise in pH above that of the uninoculated control. Mutant derivatives of bv. trifolii strain ANUB43 with a 'Inn_5_ insertion in the regulatory gene _@ or the host specificity gone Log; fail to stimulate the neutralization response The 19:15::an mutant is slightly impaired in its ability to stimulate alkalinization compared to the wild type strain. Purified LPS from bv. trifolii strain ANUB43 annd strain 0403 also stimulates the neutralization at the surface of root hairs while the IPS from 3. meliloti 1021728 does not. Enacreted LPSmayserveasabacterial signaltowhichtheplantrespodsand facilitates development of the symbiosis. 'Ihis neutralization response may conferanadvantagetohomolcgcnus rhizobiaintherhizosphereby providing a localized 11-! environment which allows infection to proceed past the acid-sensitive step. Neutralization at the surface of root hairs may represent an uptake of h-I‘". Since homologous (compatible) rhizobia are stimulatory, this response is fundamentally different than 67 68 thehfuptakecoservedduringthehypersesitiveresposeelicitedby inoompatible pathoges. Wm 'Ihemovementofiosacrcsscellplasmamembranesisessential for numerous activities in plant annd animal cells. 'Ihese fluxes are mediated by integral membrane proteins such as H+-A‘I'Pases, symports, antiports, carrier proteins, and voltage-gated ion channels (Hedrich and Schroeder, 1989). In roots, net ATP-dependent H+ efflux drives the uptake of sugars, amino acids, and catios (Marschner, 1986). With the application of patch-clamp technology to plant cell membranes, single ion channel measurements have been possible. In particular, chnannels forCl'andCaH'transporthavebeenstudied, aswellasseveraltypes of plasma membrane channels for K+ transport (Hedrich and Schroeder, 1989). Studies on cell elogation in barley and maize led to the acid- growth hypothesis in which insole-acetic acid (1AA) treatment is believed to induce H+ extrusion which acidifies and looses the cell wall to allow turgor-driven cell elo'gation (Cleland, 1976) . 'ne most recent studies, however, demostrate that the low [it (3.0-3.5) required to achnieve loosening of the cell wall is not biologically feasible with IAAapplicztionandstate thattheacid-growthhypothesishasyettobe confirmed (Sdopfer, 1989). Enrenthnoughfi+ fluxacrossmembranesisneoessarytomaintain homeostasis for actively metabolizing cells, there are predictios that picouldserveasthesecoomessegerofexternalstimuliinplant tissue. Annalogstoanimalcellseoorimessengers (e.g. cAMP) havebeen sought but not founnd in plant tissues (Felle, 1989). Although the 69 regulation of enzyme activity depends on appropriate cytoplasmic pi, and normal metabolism is costame affecting internal pH, Felle (1989) proposesthatp-Ifunctiosasasecondmessegerbyinteractingwith vacuolar Ca'H' and thnus is perceived diffently than normal metabolic changes in pH. 7 'Ihe plant-defose respose known as the hypersensitive reaction (HR) is irdnoed by inoompatible plant pathogenic bacteria. An early eventinsuspension-culturedtobaccocellsduringtnnenunisanetu+ uptake anud concomitannt K" efflux (Atkinson g 11,, 1985). 'Ihis phenomenon is significantly reduced by ATPase inhibitors and apparently requires an electrochemical gradient across the plasmalemma (Atkinson annd Baker, 1989). Neither the precise mechnanism of the H+/n<" exchange northerelatioshipbetweenmuptakeandhmcelldeatharelcownt Ion fluxes which geerate electric currents play a role in orienting plant growth during development (Jaffe annd Nuccitelli, 1977; Weisenseel 3 an, 1979). Miller g 21; (1988) describe thnese cellular currentsasa3-partsystencosistingofaninternalcurrentloop,an enternalcurrentloop, andtheinterfaceofthetwoatthe plasmamenbrane. Ion replacement studies with barley roots and root hairshaveshmnthnatsohecternnalcurrentsarecarriedbyh-f" (Weiseseel _et_: a1., 1979). currents have also been studied in tobacco, maize, and white clover roots using a non-invasive vibrating microelectrode. Foreachtypeofrootsystem, thedirectionofcurrent flow is inward formeristematicandelogatiungregiosand forthetips of actively growing root hairs, whereas mature epidermal cells (post- elongation region) have an outward current (Miller g al_., 1986, 1988, Miller, 1989). 7O Inthesoil environment, sudnenternualcurrentsarethoghtto affect the interaction of the plant root withn the rhizcsfiere. The sites of inwardcurrent flow intobaccoroots, includinggrowirgrcot tips, sites of energing laterals, and wounnd sites, are precisely the same areas to which Phytophthora parasition zoospores are attracted (Miller gt Q" 1988). Miller g g. (1988) suggest that the zoospores are electrotactic, as well as chemotactic, thnus allowing the organism to distinguish between living annd dead host cells. 'Ihe surface of rhizobia carries a not negative charge at pH '7 (Marshall, 1967). It has been suggestedthatcurrents inwhitecloverroothairscouldplayarolein attracting the bacteria to the proper sites for infection (Miller _e_t a_1., 1986). Plasmamembrane potential in cortical cells and root hairs is affected by treatment with rhizobia or their supernatants. firsek gt 31.- (1986) measured chnanges in transmembrane potential for soybean cortical cells after inoculation with homologous annd heterologous rhizobia. After 1 d, bothn homologous Bradyrhizobium japonicum and the heterologos Rhizcoium meliloti reduced the transmembrane potential (depolarized) , with B. meliloti eliciting the greatest reduction. Neither heat-killed rhizobianorPseudomonas fluorescensdecreasedthemembranepotential. 'Ihey coclude that host cells do respond physiologically to heterologous inoculation annd suggest that living rhizobia increase membrane permeability in root cells, thnnus giving rise to the altered potential difference similartothatseenincellsundergoingHR. Loggtal. (1991) report that the transmembrane potential of single alfalfa root hairsmaybemeasured. Roothairsimpaledwithanelectrodeandthen enposed to the flavoe—induced supernatant of wild type 3. meliloti 7 1 show a rapid depolarization withnin 1-2 min followed by recovery and slight hyperpolarization in 10—15 minn. Snupernnatants from a man; derivative did not induce the depolarization. Purified bacterial factor NodRm-l whidn induce alfalfa root hnair deformation also stimulated the rapid depolarization. Previously, ecternalpI-Iarcundrootswasmeasuredinthehulk medium or by pH sesitive dye in agar plates. Actively growing white clwerrootheirsgeerateedogeouselectrielcurrentswhidnare carried by H” and which mighnt influence rhizosphere microoganisms. ‘Ihe objective of this study was to determine the effect of inoculation with homologous and heterologous rhizobia on the In" cooentration at the surface of individual clover root hnairs. Materials and methods Bacterial cultures. 'Ihe wild type Rhizobium strains _13. leguminosarum bv. trifolii AN0843 and B. lfluminosarum bv. gigging 300 were supplied by Barry Rolfe, Australian National University. 3 meliloti strain 1021 was provided by Jean Dénnarié, ERA-(NB Toulouse, France. In; mutants derived from ANU843 (from Barry Rolfe) included strain ANUBSl fining, ANUZ58 flung, annd AMJZSI gyms; Bacteria were grown for 4 d at 30 C on BIII agar plate (Dazzo, 1982) for wild type or BIII agar containing 100 ng/ml kannamycin for'Innfimutants. Bacterial inoculumwaspreparedby suspeding the bacteria in nitrogen-free Fahracus medium (-NF, Dazzo, 1982) to a density of klett 100 (ca. 109 cells/m1). LPS isolation. RuizobiaweregrowninshakenflaskscontainningBIIIbrothatwc. _R_. leguminosarum bv. trifolii strains ANUB43 and 0403 (Rothaunsted 72 Experimental Station, I-hrpeden, U.K.) and BL meliloti strain 102F28 (Joseph Burton, 'Ihe Nitragin Co., Milwaukee, WI) were eachn grown to a turbidity of 90 klett unnits (early stationary phnase, ca. 9 x 108 cells/m1) as measured with a Klett-Summerson colorimeter. Cells were pelleted by low speed centrifugation, reuspeded in 0.5 M NaCl, and thenstirredrapidly forlhtoremoveonpsularpolysaccharide. Following centrifugation at 10000 x g, the IPS was extracted from the cell pellet with hot phenol/water and the aqueous portion dialyzed against water. Rurifietion of IPS was achnieved by a combination of ion ecchangednromatographyonAGle, INaseandRNasetreatment, ultracenntrifugation, gel filtration thnrough Bio-led ALSM, and then through Sepharose 4B in EDrA-triethylamine to yield an IPS peak with a costant ratio of total carbohydrate, 2-keto-3-deonyocbu1onic acid, and heptose (Carlson gt _a_l_. 1976; Hrabak g g. 1981). as peak fractions were pooled, dialyzed against deionized water, and lyophilized. IPS was dissolved in -NF medium and f ilter-sterilized before assay. prqaaratim of pH microelectrode. Liquid membrane microelectrode were prepared daily. Glass micropipette with tip diameters of 0.8-1.5 nm were made from Pyrex microcapillarie (Corning no. 7740, Corning, NY) using a PP-83 Narishnige micrquipette puller. Capillarie were thnen cleared for 15 hrs in nitric acid, rinsed with double distilled water and distilled methanol, dried under a stream of argon to evaporate the solvent, further dried at 200 C for 16 hrs, and stored over silica gel (Ammannu gt_ 21- , 1981) . Prior to filling, micropipette were silanized with N,N-Diethyl trimethnylsilylamine and dried again at 200 c for «no miun. 'Ihe n+- 73 selective neutral carrier was prepared by mixing double distilled tri-n- dodecylamine as the Hts-elective ligand (finnal cocentration 10 wt%) with sodium tetraphenylborate (0.7 wt%) in o—nitrophenyl cctyl ether, andthen incubated atroomtemperature for18hrsunder100%CD2 (Ammannnn e_t 11,, 1981). Micrqnipette were filled with the H+-selective carrier by capillary action to a column height of 600 um from the tip with the remainder of the pipette back-filled with a buffer containing 40 mM mzpo4, 15 mM NaCl, and 23 mM NaCl-I (pH 7.0). Microelectrode were inserted intoaholderwithaAg/AgCl wireimmerseddirectlyintothe microelectrode buffer solution, whichn in turn was connected to a high- impedance 614 Keithly electrometer and one of the chnannels of an 8800 Began Total Clamp System with specific adjustments for ion-selective electrode measurements. The Ag/AgCl pellet of the referece electrode was in contact with the seedling bathing solution by an agar bridge. During measurements, all electrode and highn impedance compoents were located inside a Faraday cage. Measurement of root hair 11!. Surface sterilized seeds of white clover cv. Dutch Whnite were germinated in the dark on -NF agar plate at 23 C day/20 C nighnt. For experiments, single 3-d old seedlings were anchored to small 5.5 cm plastic petri dishne with a drop of molten agarose, and then submerged in 2 ml of -NF medium, [11 6.5. Seedlings were allowed to equilibrate for 5 min prior to inoculation or addition of LPS treatnnent solutios. Bacteria were inoculated by adding 100 ul of the klett 100 bacterial suspesion to the seedling bathing solution. ‘Ihe purified LPS from Rm102F28, ANU843, and 0403 was dissolved in -NF medium and applied to 74 seedling bathning solution at a final cooentration of 5 ug/seedling. Microelectrode were positioned roughnly perpendicular to and midway alog the surface of single root hnairs with a micromanipulatcr and invertedmicrcscope. Changeinelectricalpotentialdueto fluctuatios in}!+ cocentration were recorded every 5 min foruptoGO minandlaterconvertedtopfieachdaybymeaslrementofelectrode respose to pH standards. lbsults 'IhepHatthesurfaceofroothairsonmninoculatedseedlingsdid not vary significanntly from the pH of the originnal bathning solution over 60 min (Fig. 1). Addition of heterologous rhizobia (R1300, Rm1021) raised the pi! slightly, whereas the homologous bv. trifolii AN0843 stimulated a signnificannt rise in 11-1 and neutralization of the medium at the surface of root hnairs beginnnning 30 min after inoculation. A 60 min incubation of AN0843 in -NF medium lacking a seedling did not affect the pH of the medium (data not shown). Mutants with Tn; insertios in either ER (a positive regulatory gee) or _no_d_L (a host-specific nodulation gee) did not stimulate a pH chnange (Fig. 2). 'Ihe mum; mutant was slightly impaired its ability to elicit the neutralization as compared to the homologous wild type. 'Ihepl-IatthesurfaceofroothairsrosewithinZOminafter treatment with purified IPS from both homologous Rhizobium strains (ANUB43 and 0403) (Fig. 3.). The heterologous Run102F28 LPS did not elicit this neutralization. pH pH 75 6.75 e—o -NF control n - 2 H Rt843 n - 5 5.70 H Rl300 n - 6 H Rm1021 n - 1 6.65 6.60 bacteria added " 6.55 l, .. /-/ \' 6.50 ,--:— .l‘/ 6.45 of}:Eu'szbz'ssbnrsunbissbs'ssbss TIME (min) Fig. 1. pH at the surface of single white clover root hairs inoculated with wild type rhizobia. Bars = SE. 6.75 * o—o 84311951an5 n - 3 a—a 843mm!) n - 3 5.70 H 843mmuTn5 n - 2 H 843 wild type n - 5 6.65 g‘ "— 6.60 l. -- bacteria I .. 655 added i . . A ’ .. 5.50 99’ r 6.45 I r l l I I T I l I T T I O 5101520253035404550556065 TIME (min) Fig. 2. pH at the surface of single white clover root hairs inoculated with Tn5 mutants derived from Rt843. Bars = SE. pH 76 6.65 HRtO403LPSn-4 -o—aRt843LPSn-2 ‘HRmFZBLPSn-3 6.60- I LPS I added 6.55- 1 I I I I I I I I I I I I 0 5 10 15 20 25 30 35 4O 45 5O 55 60 65 TIME (min) Fig. 3. pH at the surface of single white clover root hairs treated with purified LPS. Bars = SE. 77 Discussion Bulk measurements of the medium around roots shows acidification in several days, especially after inoculation with Azospirilluum (Bashan, 1990). By measuring pH at the surface of single root hairs, the microenvironment at the site importannt to infection were annalyzed. Our results using non-invasive electrode and wild type bacteria are more representative of what must occur during competition in the rhizosphere. Previous studiehnavemappedcurrentpaths of singleroothairsand showed thnattherearesite of outwardandinwardfluxattheroot surface (base of root heirs) and the root hair tip, repectively (Miller _e_t g" 1986). In our study, the electrode was positioned midway betweenthepoinnts ofoutwardan‘dinmardflmtinordertomeasurethe net chnange in H+ cooentration. Nitrogen—free Fahraeus medium was used in experiments to maintain theosmoticbalanoewithninroothairs. 'IhismediumcontainlemM phosphate and thus provide some buffering which might affect the magunitudeofnfldnangeobserved. Wethereforebelievethnatour measurementsrepresenttheminimumphdnangepossibleandany differences could be e'hnanced in a nnatural environment. The inncrease in medium pl after treatment with homologous Rhizobium crourifieleScanbeinterpretedasanuptakeobeytheroothair. Alternatively, the external dnange in pH may be an effect of the efflux ofcatios. 'n'ednargeimbalanoecausedbysudnaneffluxinthe aqueous solution could easily be compesated by the association/dissociation of water molecule reulting in a net change in nil (Good, 1986). Using a K+-specific microelectrode, we also observed a 35%increaseinentennalK’atthesurfaceofroot-hairs 20minafter 78 inoculation with strain AN0843 (data not shown). 'Ihis efflux of Id coupled withn the rise in run (low nn” concentration) is suggetive of a hi+/I<+antiportsudnasthosereportedinannmberofsystems(aehland Raschke, 1987; Sze, 1985). InSuu_nap' mgmm, thecytoplasmicpiis7.3 (Bertland Felle, 1985). If pi of the cytoplasm in clover root hairs is similar, the external medium (pi 6.5) would surely a higher cooentration of H+ ontsideoftheroothairthnanwaspreentinthecytoplasm. Itis possible that the Rhizobium-induced neutralization of the external medium wouldnotrequireeergyinpntsinoethefi+gradient favors uptakeratherthanefflux. 'nerequirement forATPtopumpI-f‘out acrossthemembraneofrootcellsiswellcharacterized (Marsduner, 1986). Nodulation by homologous rhnizobia can be blocked by inoculating roots in a medium with a pi of 5.0 or les. Muunnns (1968) identified an early acid-sesitive step in the infection proces whichn occurred prior to infection thnread formation. Subsequent work showed that acidic pi innhnibited the activity of host polygalacturonase which is a possible factor in the softening of root hair walls duuring homologous innfection (Mums, 1969; Ljundgreun and Fahraeus, 1961). Dazzo and Hubhell (1975) found that components of the bv. trifolii CPS likely to be recognized by the host plant lectin are very acid labile. Bacteria treated with buufferatpi5.00rlesno1ongerboundtheantibodymadeagainstthe cross-reactive clover antigen. Growth of plannts at pi <5.0 decreased thenetadgee—inducingactivityofrooteadatefromsubterrranean clover but not whnite clover (Rolfe _e_t; gl_., 1988). Item may be additional infection-related events sesitive to pi including the 79 innteraction of pectic carbonyls or cocentration of Ca‘H' which would affect cell wall structure. 'Ihus the neutralization at the surface of white clover root hairs may provide a localized pi environment favorable to hnonnologous infection and alleviate the potential acid-inhibition of Ms Rhizobium-induced respose in root huairs is significant because it is a rapid biovarbspecific pheomeonn which is affected by mutatios in bacterial _no_d gee Strain ANU851 _n_o_d_D::'I‘n_5_ is mutated in its regulatorygeegivingaNod'pheotypeonanyhost. StrainANUBSl is also unable to induce the wild type neutralization respose, suggetingthattheactionofsomegeeundertheregulationofflis esenntialtostimulatetherespose. ‘IheoperonsnodFERLandLoiihj repreent the 8 kb of the ANU843 symbiotic plasmid deignated host specific noduulation (bin) gees. A ‘1an insertion in _nogg results in expanded host range while retaining some nodulation ability on white clover (Djordjevic gt _a_1_., 1985). The phenotype for a g mutant is delayed nodnulation on white clover. In addition, inoculation of the £1.10 muutant on white clover severely impairs cytoplasmic streaming in developing root hairs (Dazzo and Appenzeller, unpublished observation). Althogh a limited number of mutannts were teted, we find that mutation of nag only slightly diminishes the stimulation of the neutralization, howeveramginmabolishetherepose. ‘Ihismayrepresenta function for _no_d_L whichn contribute to nodulation efficiecy. Rhizobium LPS may be part of the complex chnenical signaling which must take place to form a sooesful symbiosis. Rhizobium leguminosarum bv. trifolii LPS is encreted in the clover rhnizospere and pretreatment of clover roots with purified LPS at low concenntration ehnanoe the 80 number of infection threads formed by wild type _13. leguminosarum bv. trifolii (Ihzzo gt Q" 1991). Flavones released from legume roots signaltheRhizobiumtoexpresnecesarygenes. ‘IheLPScouldbea specie-specific signal from the bacteria which elicits neutralization at the root hair surface which in turn provide a favorable environment for infection. Whether the neutralization of the external medium results in cytoplasmic or cell wall acidification which then cause directchemical effectsorserveasafurtherinternnalsignalcannotbe determined from these measurements. 'Ihe site of interaction between IPS and clover root hairs may be the cell wall or more likely the cell membrane. In any event, it is significannt that homologous rhnizobial LPS can stimulate the clover root hair reponse while heterologous LPS doe not. 'Iheuptakeofli'I'andconcomitant efflux ofK+hnasbeenobservedfor plant tissue culture cells undergoing the hypersesitive reaction (HR) in respose to incompatible pathoges (Atkinson gt _a_1_., 1985). Based on electron micrograpns, Djordjevic gt fl. (1988) have suggeted that a mutannt of strain NGR234 cause a hypersensitive-like respose on Macrogtillium. Our results, however, indicate that the wild type Rhizobium strain equivalent to an incompatible pathogen, the heterologous biovar, doe not elicit the 11* uptake respose Instead, it is the homologous rhizobial strain (analogous to the compatible pathogen) whidn elicits the respose. 'Ihus the plannt's respose to wild typerhizobia istheoppositeofwhathasbeenobservedforthel—IR. Therefore the Rhizobium-stimulated change in Ii+ cocentration doe not indicateareistanceresposebutmore likelyisacontributingfactor in succesful symbiosis. These results do not rule out the possibility 8 1 that other plannt defose-like repose contribute to the determination of host specificity. 2. 3. 7. 8. 10. LIST OF REFERENCES Abe, M., J.E. Sherwood, R.I. Hollingsworth, and F.B. Dazzo, 1984. Stimulation of clover root hair infection by lectin-binding oligosacchnarides from the capsular and extracellular polysaccharides of Rhizobium trifolii. J. Bacteriol. 160: 517-520. Aist, J.R. and hi.W. Israel, 1986. Auutofluuorescent and ultraviolet- absorbing compoennts in cell walls and papillae of barley coleoptile ardtheirrelatioshiptodiseaseresistance. en. J. Bot. 64: 266-272. Albert, F. and A.J. Anderson, 1987. The effect of Pseudomonas M colonization on root surface peroxidase. Plannt Phnysiol. 85: 537-541. Albert, F.G., I...W. Bennett, and A.J. Anderson, 1986. Peroxidase associated with the root surface of Phaseolus vulgaris. Can. J. Al-Mallah, M.I(., M.R. Davey, and E.C. Cooking, 1987. Enzymatic treatment of clover root hairs remove a barrier to Rhizobium-host specificity. Biotechnnology 5: 1319-1322. Al-Mallah, M.K., M.R. Davey, and E.C. Cooking, 1990. Nodulation of oilseed rape (Brassica napus) by rhizobia. J. Exp. Bot. 41: 1567-1572. Ammann, D., F. Lanter, R.A. Steiner, P. Schulthes, Y. Shijo, and W. Simon, 1981. Neutral carrier based hydrogen ion selective micro- electrode for extra- and intracellular stdie. Anal. Chem. 53: 2267-2269. Atkinson, nut. and CJ. Baker, 1989. Role of the plasmalemma n+- A'I'PaseinPscudomonassyringae-inducedKVhFednangein suspesion-cultured tobacco cells. Plannt Physiol. 91: 298-303. Atkinson, M.M., J.S. Huang, and J.A. Knapp, 1985. The hypersesitive reaction of tobacco to Pseudomonas syringae pv. pisi. Plant Buysiol. 79: 843-847. Bashnan, Y., 1990. Short exposure to Azospirillum brasilese Cd inoculationenhancedprotoneffluxof intactwheatroots. Can. J. Microbiol. 36: 419-425. 82 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 83 Bashan, Y., Y. Okon, and Y. Henis, 1987. Peroxidase, polyphenol- ocidase, andpenolsinrelationtoresistanceagainstPseudo- monas syringae pv. tomato in tomato plants. Can. J. Bot. 65: 366-372. Behnl, R. and K. Raschke,1987. Close coupling between extrusion of H+ and uptake of K+ by barley roots. Planta 172: 531-538. Bertl, A. and H. Felle, 1985. Cytoplasmic pi of root hair cells of Sinnapis alba recorded by a phi-sensitive micro-electrode. Doe fusicoccin stimulate the proton pump by cytoplasmic acidification? J. Dtp. Bot. 36: 1142-1149. Bohnlool, B.B. and R.I.. Schmidt, 1974. Lectins: a possible basis for specificity in the Rhizobium-legume root nodule symbiosis. Sciece 185: 269-271. Bowle, D.J., 1990. Defese-related proteins in higher plants. Ann. Rev. Biochem. 59: 873-907. Brewin, N.J., J.E. Beringer, A.W.B. Johnston, 1980. Plasmid mediated transfer of host-range specificity between two strains of Rhnizobium _1_gguminosarum. J. Gen. Microbiol. 128: 413-420. Campa, A., 1991. Biological role of plant peroxidase: known and potential functios. g: Peroxidase in Chemistry and Biology, Vol. II, J. Everse, K.E. Everse, and M.B. Grisham, eds., CRC Press, Boca Raton, FL, pp. 25-50. Cassab, G.I., J.J. Lin, L.S. Lin, and J.E. Varner, 1988. Ethylene effect on extesin and peroxidase distribution in the subapical region of pea epicotyls. Plant Physiol. 88: 522-524. Chakraborty, U. and B.N. Gnakraborty, 1989. Innteraction of Rhizobium leggminosarum and Fuusariuun solani f. sp. pisi on pea affecting disease development and pnytoalexin production. Can. J. Bot. 67: 1698-1701. Cleland, B.B., 1976. Kinetics of hormoe-induced Inf-excretion. Plant Physiol. 58: 210-213. Dazzo, F.B. and W.J. Brill, 1977. Receptor site on clover and alfalfa roots for Rhnizobium. Appl. Environ. Microbiol. 33: 132- 136. Dazzo, F.B., 1982. Ieguminous root noduule. g: Duperimental Microbial Ecology, R.G. Buurns and J.L. Slater, eds., Blackwell Scientific Publicatios, Oxford, UK, pp. 431-446. Dazzo, F.B. and D.Ii. Iiubhell, 1975. Cross-reactive antiges and lectin as determinannts of symbiotic specificity in the Rhizobium- clover associatiou. Anal. Microbiol. 30: 1017-1033. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 84 Dazzo, F.B., 6.1.. Truchet, J.E. Sherwood, ELM. Hrabak, arnd A.E. Gardiol, 1982. Alteration of tle trifoliin A-binfling capsule of Rhizobium trifolii 0403 by enzymes released from clover roots. Appl. Environ. Microbiol. 44: 478-490. Dazzo, F.B., 6.1.. Truchet, J.E. Sherwood, B.N. Hrabak, M. Abe, and S.B. Panlcratz, 1984. Specific phases of root leir attachment in tte Rhizdnium trifolii clover symbiosis. Appl. Environ. Microbiol. 48: 1140-1150. Dazzo, F.B., 6.1.. Truchet, R.I. Hollingsworth, E.M. Hrabak, H.S. Pankratz, S. Philip-Hollingsworth, J.L. Salzwedel, K. Chapman, L. Amenzeller, A. Squartini, D. Gerhold, and G. Orgambide, 1991. Rhizobium lipopolysaccharide modulates infection thread develop- ment in white clover root hairs. J. Bacteriol., accepted with r'ev151os. Debellé, F. and S.B. Sharma, 1986. Nucleotide seqnenoe of Rhizobium meliloti R622011 genes involved in host specificity of nodulation. Nucl. Acids Res. 14: 7453-7472. Dewick, P.M., 1975. Pterocarpan biosynthesis: chalcone and isoflavoe preomsors of demethylhomopterocarpin and maackiain in Trifolium pratense. Bnytodemistry 14: 979-982. Diaz, C.L., L.S. Melchers, P.J.J. Hooykaas, B.J.J. Lugtenberg, and J.W. Kijre, 1989. Root lectin as a determinant of host-plant specificity in the Rhizobium-legume symbiosis. Nature 338: 579- 581. Djordjevic, M.A., W. Zurkowski, J. Shine, and B.G. Rolfe, 1983. Sym plasmid transfer to various symbiotic mutants of mizobium trifolii, _13. leguminosarum, arnd B. meliloti. J. Bacteriol. 156: 1035-1045. Djordjevic, M.A., P.R. Schofield, and B.G. Rolfe, 1985. ‘I‘nS mutagenesis of Rhizobium trifolii host-specific nodulation genes result in mutants with altered host-range ability. Mol. Gen. Genet. 200: 463-471. Djordjevic, M.A., R.W. Innes, C.A. Wij ffelman, P.R. Schofield, and B.G. Rolfe, 1986. Nodulation of specific legumes is controlled by several distinnct loci in Rhizobium trifolii. Plant Mol. Biol. 6: 389-401. Djordjevic, M.A., D.W. Gabriel, and B.G. Rolfe, 1987. Rhizobium tte refined parasite of legumes. Ann. Rev. Phytopat'hol. 25: 145-168. Djordjevic, M.A., J.W. Redmond, M. Batley, and B.G. Rolfe, 1987. Clovers secrete specific plenolic compounds which eitner stimu- late or repress nod gere expression in Rhizobium trifolii. Elm Journal 6: 1173-1179. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 85 Djordjevic, S.P., R.W. Ridge, H. Chen, J.W. Redmond, H. Batley, and BG. Rolfe, 1988. Induction of pathogenic-like resmonses in tie legume Macrcptilium atropirpnrenm by a transposon-irduced mutant of tie fast-growing, broad-host range Rhizobinmn strain @234. J. Bacteriol. 170: 1848-1857. nmford, B.B., 1991. Horseradish peroxidase: stmcture and kinetic prcperties E Peroxidases in demistry and Biology, Vol. II, J. Everse, K.E. Everse, and M.B. Grisham, eds., CRC Press, Boca Raton, FL, pp. 1-24. Ellirgboe, A.H., 1981. Changing concepts in host-pathogen genetics. Ann. Rev. Phytcpath. 19: 125-143. Elstner, E.F., 1982. Oxygen activation arnd oxygen toxicity. Ann. Rev. Plant Physiol. 33: 73-96. Ehdo, T., 1968. Irndoleacetate oxidase activity of horseradish and otter plant peroxidase isozymes. Plant & Cell Physiol. 9: 333- 341. érsek, T., A. Novacky, and s. G. Pueppke,1986. Compatible and incompatible rhizobia alter membrane potentials of soybean root cells. Plant Physiol. 82: 1115-1118. Espelie, K.E., V.R. Franceschi, arnd P.E. Kolattukudy, 1986. Immunocytodemionl localization and time course of appearance of an aninonic peroxidase associated with suberization in wound- healing potato tuber tissue. Plant Physiol. 81: 487-492. Espelie, KB. and P.R. Kolattnkudy, 1985. Purification and characterization of an abscisic acid-inducible anionic peroxidase associated with suberization in potato (Solanum tuberosum) . Arch. Biochem. arnd Biophys. 240: 539-545. Esquerré-Tugayé, M.T., C. Iafitte, D. Mazau, A. Toppan, and A. me, 1979. Cell anrfaces in plant-microorganism interactios. II. EVidence for tte aconnulation of hydrocyprolinne—rich glyco- proteirs intlecell wall ofdiseasedplantsasadefensemedn- anism. Plant Physiol. 64: 320-326. Esquerré-Tngayé, M.T. and D.T.A. Lamport,1979. Cell surfaces in plant-microorganism interactions. I. A structural investigation of cell wall hydrocyproline-rich glycoproteins whidn accumulate in fungus-infected plants. Plant Physiol. 64: 314-319. Everdeen, D.S.,S. Kiefer, J.J. Willard, E.P. Muldoon, PM. Dey, X.B. Li, and D.T.A. Iamport, 1988. mzymic cross-linkage of monomeric extersin precursors _inn vitro. Plant Physiol. 87: 616- 621. Everse, J., KB. Everse, and 11.3. Grisham, eds., 1991. Peroxidases in chemistry and biology, Vol. II. CRC Press, Inc., Boca Inton, FL, 264 pp. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 86 Faucher, C. F.Mai11et, J. Vasse, C. Rosenberg, A.A.N. van Brussel, G. ‘I‘rudet, annd J. Dénarié,1988. mizobium meliloti host range nnodl-I gone determines production of an alfalfa-specific extracel- 1u1ar signal. J. Bacteriol. 170: 5489-5499. Felle, H., 1989. pH as a second messenger in plants. in: Second messengers in plant growth annd development, W.F. Boss annd D.J. Morre, eds., Alan R. Liss, Inc., NY, pp. 145-166. Fielding, J. I. and J.L. Hall, 1978. A biodemical arnd cytochemical study of peroridase activity in roots of Pisum sativum. I. A comparison of DAB—peroxidase annd guaiacol-peroxidase with particular emfinasis on tre properties of cell wall activity. J. Exp. Bot. 111: 969-981. Fielding, J.L. annd J.L. Hall, 1978. A biodemical and cytochemical study of peroxidase activity in roots of Pisum sativum. II. Distribution of ennzymes in relation to root developmennt. J. Ecp. Bot. 111: 983-991. Gaspar, T., C. Penel, T. Thorpe, and H. Greppin, 1982. Peroxidases 1970-1980. Universite de Geneva-Centre de Botanique, Geneve, Suisse. 324 pp. Gerhold, D.L., F.B. Dazzo, annd P.M. Gresshoff, 1985. Selective removal of seedling root hairs for studies of tie mizobium- legume symbiosis. J. Microbiol. Meth. 4: 95-102. Gibson, D.M. annd E.H. Liu, 1978. Substrate specificities of ' isozymesinttedevelopingpeaseedling. Ann.Bot. 42: 1075-1083. Good, N.E., 1986. Confessions of a habitual skeptic. Ann. Rev. Plannt Physiol. 37: 1-22. Grison, R. and RE. Pilet,1985. Maize root peroxidases: relatioship with polyptenol oxidases. mytod'emistry 24: 2519- 2521. Gustine, D.L., R.'I‘. Sherwood, B.G. Moyer, and F.L. Lukezic, 1990. Metabolites from Psenriomonas conggata elicit pnytoalexin biosynthesis in white clover. Phytcpatlnology 80: 1427-1432. Hammersdnmidt, R. annd J. 10.16, 1982. Lignifioation as a mechanism for induced systemic resistance in cucumber. Physiol. Plant Pathol. 20: 61-71. Hammerschmidt, R., A.M. Bonnen, G.C. Bergstrom, and KK. Baker, 1985. Association of widermal lignification with nnonhost re- sistance of cucurbits to fungi. Can. J. Bot. 63: 2393-2398. Hammerschmidt, R., D.T.A. Iamport, annd E.P. Muldoon, 1984. Cell wall hydroxyproline enhancement and lignin deposition as an early event in tte resistance of cucumber to Cladosporium cucumerinum. 60. 61. 62. 63. 64. 65. 66. 67. 69. 70. 71. 87 Physiol. Plant Pathol. 24: 43-47. Hammerschmidt, R., E.M. Nuckles, annd J. 10.16, 1982. Association of enhanced peroxidase activity with induced systemic resistance of cuonmber to Colletotridmm lagenarium. Bnysiol. Plant Pathol. 20: 73-82. Hedrich, R. and J.I. Schroeder, 1989. The physiology of ion dnannelsandelectrogenicpumps inhigherplants. Ann. Rev. Plant Physiol. 40: 539-569. Higgins, V.J. annd B.G. Smith, 1972. Separation and identification of two pterocarpanoid phytoalexins produced by red clover leaves. Phytopathology 62: 235-238. Horvath, B., E. Kondorosi, M. John, J. Schmidt, I. Torok, z. (yiirgypal, I. Barabas, U. Wieneke, J. Sclell, and A. Kondorosi, 1986. Organization, structure and symbiotic function of Rhizobium meliloti nodulation gennes determining host specificity for alfalfa. Cell 46: 335-343. Hrabak, E., M. Urbano, and F. Dazzo, 1981. Growth-[hasedependennt immunodeterminants of mizobium trifolii lipopolysacdnaride which binnd trifoliin A, a white clover lectin. J. Bacteriol. 148: 697-711. Ingham, J.L., 1978. Isoflavonoid and stilbene phytoalexins of tie gennus Trifolium, Biochemical Systematics annd Ecology 6: 217-223. Ingham, J.L., 1982. Phytoalexins from the Leguminnosae E: Phytoalexins, J.A. Bailey arnd J.W. Mansfield, eds., John Wiley and Sons, NY, 334 pp. Innes, R.W., P.L. Kuempel, J. Plazinski, H. Canter-Cremers, B.G. Rolfe, annd M.A. Djordjevic, 1985. Plant factors induce expres- sion of nodulation and host-range genes in mizobium trifolii. Molec. Gen. Genet. 201: 426-432. Jaffe, LF. annd R. Nuccitelli, 1977. Electrical controls of develcpment. Ann. Rev. of Biophysics and Bioengineering 6: 445- 476. Katsuwon, J. annd A.J. Annderson, 1989. Respose of plant- colonizinng psandomonads to hydrogen peroxide. Aml. Environ. Micrdaiol. 55: 2985-2989. Kay, LE. annd K.V. Basile, 1987. Specific peroxidase isoenzymes are correlated with organogenesis. Plant mysiol. 84: 99-105. Keen, N.T. annd M.J. Holliday, 1982. .12.: Phytopathogennic prokaryotes, Vol. 2., M.S. Mount and G.H. Lacy, eds., Academic press, NY, pp. 179-220. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 88 Kerby, K. annd S. Somerville, 1989. Enhancement of specific inntercellular peroxidases following inoculation of barley with Erysiple minis f. sp. hordei. Physiol. and Molec. Plant Pathol. 35: 323-337. Klein, D.A., J.L. Salzwedel, annd F.B. Dazzo, 1990. Microbial colonization of plannt roots. Q: Biotechnology of plant-microbe interactios, J.P. Nakas and C. Hagedorn, eds., Mozraw Hill Publishing Co, NY, pp. 189-225. Kolattukndy, P.E., 1977. Biosynthesis and degradation of lipid polymers. Q: Lipids annd lipid polymers in higher plants, M. Tevini and H.K. Lichtenthaler, eds., pp. 271-292. Kolattukudy, P.E. and B.B. Dean, 1974. Structure, gas chromatographic measurement, and funnction of suberin synthesized by potato tuber tissue slices. Plant Physiol. 54: 116-121. Kolattukudy, P.E., K. Kronman, and A.J. Poulose, 1975. Determination of structure annd composition of suberin from roots of carrot, parsnip, rutabaga, turnip, red beet, annd sweet potato by combined gas-liquid chromatography and mass spectrometry. Plant Physiol. 55: 567-573. Kondorosi, E., J. Guyris, J. Schmidt, M. John, E. Duda, J. Schell, and A. Kondorosi, 1988. Positive arnd negative control of nodulation genes in mizobium meliloti strain 41. in; Molecular plant-microbe interactions, D.P.S. Verma annd R. Palacios, eds., APS Press, St. Paul, MN, pp. 73-78. Iannport, D.‘I‘.A., 1986. ‘Ite primary cell wall: a new model. Q: Cellulose: structure, modification, annd hydrolysis, R.A. Young and R.M. Rowell, eds., John Wiley & Sons, NY, pp. 77-90. Lerouge, P., P. Roche, C. Faucher, F. Maillet, G. Truchet, J. Promé, annd J. Dénarié, 1990. Symbiotic host-specificity of mizobium meliloti is determined by a sulpnated annd acylated glucosamine oligosaccharide signal. Nature 344: 781-784. Lewis, N.G. and E. Yamamoto, 1990. Ligninn: occurrennce, biogennesis and biodegradation. Ann. Rev. Plannt Hnysiol Plant Mol. Biol. 41: 455-496. Li, D. annd D.H. Huflell, 1969. Infection thread formation as a basis of nodulation specificity in mizobium—strawberry clover associatios. can. J. Microbiol. 15: 1133-1136. Ljunggren, H. and G. Fahraeus, 1961. 'Ite role of polygalacturonase in root-hair invasion by nodule bacteria. J. Gen. Microbiol. 26: 521-528. Long, S.R., R.F. Fisher, J. Ogawa, J. Swanson, D.W. Ehrhardt, R.M. Atkinson, and J.S. Schwedock, 1991. Rhizobium meliloti nodula- tion gene regulation annd molecular signals. _i_n: Advances in 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 89 molecular genetics of plant-microbe innteractios, H. Hennnecke and D.P.S. Verma, eds., Kluwer Academic Publishers, Dordrednt, 'ne Netlerlands, pp. 127-133. Lucy, M.C., P.S. Matthews, and H.D. VanEtten, 1988. Metabolic detoxification of tie phytoalexins maackian annd medicarpin by Nectria haematococca field isolates: relatioship to virulennce on dnickpea.‘ Physiol. annd Molec. Plant Pathol. 33: 187-199. Maehly, A.C., 1955. Plant peroxidase. ;1n_n: Methods in enzymology, vol. 2, S.P. Colowick and N.0. Kaplan, eds., Academic Press, NY, p. 807. Marschnner, H., 1986. Mineral Nutrition in Higher Plants, Academic Press, London, 674 pp. Marshall, K.C., 1967. Electrophoretic properties of fast- annd slow-growing species of mizobium. Aust. J. Biol. Sci. 20: 429- 438. Martinez, E., D. Romero, and R. Palacios, 1990. Its mizobium genome. Critical Rev. in Plant Sciennces 9: 59-93. Martinez-Molina, B., R.M. Morales, annd D.H. Hubbell, 1979. Hydrolytic enzyme production by mizobium. Appl. Environ. Micrdaiol. 38: 1186-1188. Menmier, B., 1991. N- and o-Demethylatios catalyzed by peroxidases. _ig: Peroddases in Gemistny and Biology, Vol. II, J. Everse, K.E. Everse, and M.B. Grisham, eds., CRC Press, Boca Raton, FL, pp.201-218. Miller, A.L., 1989. Ion onrrennts and growth regulators in plannt root development. Biol. Bill. 176(8): 65-70. Miller, A.L., J.A. Raven, J.I. Sprent, and M.H. Weisenseel, 1986. Endogenmsioncurrentstraversegrowingrootsandroothairsof Trifolium m. Plant Cell Environ. 9: 79-83. Miller, A.L., E. Shand, annd N.A.R. Gow, 1988. Ion currents associated with root tips, emerging laterals annd indnced wound sites in Nicotiana tabacum: spatial relatioship proposed betweenn resulting electrical fields annd phytophthoran zoospore infection. Plannt Cell Environ. 11: 21-25. Misaghi, I.J., 1982. Physiology annd Biodemistry of Plant-Pathogen Interactios. Plenum Press, NY, pp. 287. Mohan, R. annd P.E. Kolattukudy, 1990. Differential activation of expression of a suberization-associated anionic peroxidase gene in near-isogenic resistant and susceptible tomato linnes by elici- tors of Verticiglium albo-atrum. Plant Physiol. 921: 276-280. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 90 Mort, A.J. and P.B. Grover Jr., 1988. Characterization of root hair cell walls as potential barriers to the infection of plants by rhizobia. Plant mysiol. 86: 638-641. Mums, D.N. , 1968. Nodulation of Medicago sativa in solution culture. I. Acid-sensitive steps. Plannt and Soil 28: 129-146. Munnns, D.N., 1969. Enzymatic breakdown of pectin and acid- inhibition of tie infection of Medimg roots by mizobium. Plant and Soil 30: 117-120. Nadolny, I. and L. Sequeira, 1980. Increase in perotidase activity isnotdirectlyinnvolvedininducedresistanceintobacco. Physiol. Plant Pathol. 16: 1-8. Nutman, P.S., 1959. Some observations on root-hair infection by nodule bacteria. J. Exp. Bot. 10: 250-263. Pankhurst, C.E., and D.R. Biggs, 1980. Sensitivity of mizobium to selected isoflavooids. can. J. Microbiol. 26: 542-545. Parniske, M., B. Ahlborn, and D. Werner, 1991. Isoflavooid- indncible resistance to tie pnytoalexin glyceollin in soybean rhizobia. J. Bacteriol. 173: 3432-3439. Parnniske, M., G. Pausch, and D. Werner, 1988. Changes in flavooid pattern of root hairs of Glycine max in respose to symbiotic infection with Bradyrhizobiunnn japonicum, (Abstr.) _ig: Nitrogenn Fixation: Hundred Years After, H. Botle, F.J. deBruijn, W.E. Newton, eds., Gustav Fischer, Stuttgart, p. 466. Parniske, M., C. Zimmermann, P.B. Cregan, and D. Werner, 1990. Hypersensitive reaction of nodule cells in tte GlEine sp/Bradyrhizobium japonicum-symbiosis occurs at the genotype- specific level. Botanica Acta 103: 143-148. Peters, N.K., J.W. Frost, and S.R. Long, 1986. A plant flavone, luteolin, induces expression of mizobium meliloti nodulation gennes. Sciennce 233: 977-980. Pflhler, A., W. Arnold, A. Buenndia-Claveria, D. Kapp, M. Keller, K. Niehaus, J. Quandt, A. Roxlau, and W.M. Weng, 1991. The role of tie mizobium meliloti exopolysaccharides EPS I and EPS II in do infection process of alfalfa nodules, _ig: Advances in Molecular Genetics of Plant-Microbe Interactions, Vol. 1, H. Henneke and D.P.S. Verma, eds., Kluwer Academic Rnblisners, Dordrecht, Netherlands, pp. 189-194. Redmond, J.W., M. Batley, M.A. Djordjevic, R.W. Innes, P.L. Kuempel, and B.G. Rolfe, 1986. Flavones induce expression of nodulation gennes in Rhizobium. Nature 323: 632-635. Ridge, R. and DJ. Osborne, 1970. Hydroxyproline and peroxidases in cell walls of Pisum sativunnn: regulation by ethylene. J. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 91 Exp. Bot. 21: 843-856. Rolfe, B.G., M. Batley, J.W. Redmond, A.E. Richardson, R.J. Simpson, B.J. Bassam, C.L. Sargent, J.J. Weinman, M.A. Djordjevic, and F.B. mzzo, 1988. Phenolic compounds secreted by legumes. i_n_: Nitrogen Fixation: Hundred Years After, H. Bothe, F.J. de Bruijn, and W.E. Newton, eds., Gustav Fischer, Stuttgart, pp. 405-409. Rolfe, B.G. and RM. Gresshoff, 1988. Genetic annalysis of legnmne nodule initiation. Ann. Rev. Plannt Physiol. Plannt Mol. Biol. 39: 297-319. Ros Baroelc’n, A. and R. Mufioz, 1989. Enigenetic control of a cell wall scopoletin peroxidase by lupisisoflavoe in I._upinnum. Phytochem. 28: 1331-1333. Salzwedel, J.L., M.E. Daub, and J.S. Huang, 1989. Effects of singletocygenquendnersandwontlebacteriallyindncoi hypersensitive reaction in tobacco suspension cell cultures. Plant Physiol. 90: 25-28. Schofield, P.R., R.W. Ridge, B.G. Rolfe, J. Shine, and J.M. Watson, 1984. Host-specific nodulation is encoded on a 14 kb INA fragment in Rhizobium trifolii. Plant Molec. Biol. 3: 3- 11. Schopfer, P., 1989. nil-dependence of extension growth in Avena - coleoptiles and its implioatios for tie mechanism of auxin action. Plant Physiol. 90: 202-207. Sijmons, P.C., P.E. Kolattukudy, and H.F. Bienfait, 1985. Iron deficiency decreases suberization in bean roots through a decrease in suberin-specific peroxidase activity. Plannt Physiol. 78: 115-120. Smith, DA. and S.W. Bannks, 1986. Biosynthesis, elicitation and biological activity of isoflavooid phytoalexins. Phytoclem. 25: 979-995. Smith, J.A. and R. Hammerschmidt, 1988. Comparative stndy of acidic peroxidases associated with indnced resistance in cucum- ber, muskmelon, and watermelon. Physiol. and Molec. Plant Pathol. 33: Spaink, H.F., C.A. Wijffelman, E. Pees, R.J.H. 0kker, and B.J.J. Irgtonberg, 1987. Rhizobium nodulation gene nodD as a determinant of host specificity. Nature 328: 337-340. Srivastava, O.P. and RB. Van Huystee, 1977. Innteractios amog onenolics and peroxidase isozymes. Bot. Gaz. 138: 457-464. Sterner, B.A. and R. Hammerschmidt, 1987. Association of teat stock induced resistannce to disease with increased accumulation 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 92 of insoluble extensin and ethylene synthesis. Physiol. and Molec. Plant Patlol. 31: 453-461. Street, P.F.S., J. Robb, and B.B. Ellis, 1986. Secretion of vasonlar coatingcomponentsbyxylemparenonymacells of tomatoes infected with Verticillium albo-atrum. Protoplasma 132: l-ll. Svalheim, O. and B. Robertson, 1990. Induction of peroxidases in oncumber hypcotyls and fungal infection. Physiol. Plannt. 78: 261-267. Sze, H., 1985. H+-translooatinng A'I'Pases: advannces using membrane vesicles. Annn. Rev. Plant Physiol. 36: 175-208. Vance, O.P., 1983. mizobium infection and nodulation: a beneficial plant disease? Annn. Rev. Microbiol. 37: 399-424. van der Schaal, I., J. Kijne, C. Diaz, and F. van Iren, 1983. Pea lectin binding by mizobium. _iln: Iectins, Vol. III, T.C. Bog- Hansen and G.A. Spengler, eds., Walter de Gruyter and Co., Berlin, pp. 531-338. VanEtten, H.D., P.S. Matthews, R.J. Tegtmeier, M.F. Dietert, and J.I. Stein, 1980. 'Be association of pisatin tolerannce and demethylation with virulenoe on pea in Nectria haematococca. Physiol. Plant Pathol. 16: 257-268. van Iren, F., M. van der Knapp, J. van den Heuvel, and J.W. Kijne, 1983. mytoalexins and nodulation in Pisum sativum (Abstr.) ' for Cogress on International Nitrogen Fixation, Holland. p. 433. Varner, J.E. and LS. Lin, 1989. Plant cell wall architecture. Cell 56: 231-239. Venkataraman, K., 1981. Q: The flavonoids, J.B. Harborne, T.J. Mabry, and H. Mabry, eds., Chapman & Hall, London, pp. 267-295. Verma, D.P.S. and K. Nadler, 1984. Iegume-mizobium-symbiosis: host's poinnt of view, Q: Gennes innvolved in microbe—plant innteractios, D.P.S. Verma and Th. Hohn, eds., Springer-Verlag, Wien, Austria, pp. 57-93. Vincent, J.M., 1980. Factors controlling the legume-mizobium symbiosis. _i_n: Nitrogen Fixation, Vol. II, W.E. Newton and W.H. Orme-Johnson, eds., University Park Press, Baltimore, pp. 103-129. Weinman, J.J., M.A. Djordjevic, C.L. Sargent, F.B. Dazzo, and B.G. Rolfe, 1988. A molecular analysis of the host range gennes of mizobium trifolii. g: Molecular Genetics of Plant-Microbe Interactions, R. Palacios and D.P.S. Verma, eds., APS Press, St. 133. 134. 135. 136. 137. 138. 93 Paul, MN, pp. 33-34. Weisenseel, M.H., A. Dorn, and LP. Jaffe, 1979. Natural H+ onrrentstraversegrowingrootsandroothairsofbarley (Hordeum vulgare 1.). Plant Physiol. 64: 512-518. Wij ffelman, C.A., E. Pees, A.A.N. van Brussel, R.J.H. Okker, and B.J.J. ILgtenberg, 1985. Genetic and functional annalysis of the nodulation region of tie mizobium leguminosarum Wm plasmid pRLlJI. Arch. Microbiol. 143: 225-232. Wymann, J.G. and H.D. Van Etten, 1978. Antibacterial activity of selected isoflavooids. Phytopathol. 68: 583-589. Ye, x.s., S.Q. Pan, and J. Kué, 1990. Association of pathogenesis-related proteins and activities of peroxidase, ,8-1,3-glucanase and chitinase with systemic induced resistanoe to blne mold of tobacco but not to systennnic tobacco mosaic virus. Hnysiol. Mol. Plant Pathol. 36: 523-531. Zaar, H., 1979. Peroxidase activity in root hairs of cress @idium sativum L). cytochemical localization and radio- active labelling of wall bound peroxidase. Protcplasma 99: 263-274. Zaat, S.A.J., A.A.N. van Brussel, T. Tak, E. Pees, and B.J.J. Irgtenberg, 1987. Flavooids indnce mizobium leguminosarum to produce nodnABC gene-related factors that cause thick, short rootsandroothairresposesoncommonvetch. J. Bacteriol. 169: 3388-3391. HICHIGAN STATE UNIV. LIBRARIES J“VIWINWINWINIIIIIIIWINHIHIIIHIIIUWI 31293008987772