Rhizobium Leguminosarum Biovar Trifolii PEDRO F

Home , Polygalacturonase

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1992, p. 1816-1822 Vol. 58, No. 6 0099-2240/92/061816-07$02.00/0 Copyright X) 1992, American Society for Microbiology Cell-Associated Pectinolytic and Cellulolytic Enzymes in Rhizobium leguminosarum Biovar Trifolii PEDRO F. MATEOS,1 JOSE I. JIMENEZ-ZURDO,1 JIN CHEN,' ANDREA S. SQUARTINI,1 SHERIDAN K. HAACK,' EUSTOQUIO MARTINEZ-MOLINA,2 DAVID H. HUBBELL,3 AND FRANK B. DAZZO1* Department ofMicrobiology and Center for Microbial Ecology, Michigan State University, East Lansing, Michigan 488241; Departamento de Microbiologia, Facultad de Farnacia, Universidad de Salamanca, 37007 Salamanca, Spain2; and Department of Soil Science, University of Florida, Gainesville, Flonida 3261 3 Received 19 November 1991/Accepted 20 March 1992

The involvement of Rhizobium enzymes that degrade plant cell wall polymers has long been an unresolved question about the infection process in root nodule symbiosis. Here we report the production of enzymes from Rhizobium kguminosarum bv. trifolii that degrade carboxymethyl cellulose and polypectate model substrates with sensitive methods that reliably detect the enzyme activities: a double-layer plate assay, quantitation of reducing sugars with a bicinchoninate reagent, and activity gel electrophoresis-isoelectric focusing. Both enzyme activities were (i) produced commonly by diverse wild-type strains, (ii) cell bound with at least some of the activity associated with the cell envelope, and (iii) not changed appreciably by growth in the presence of the model substrates or a flavone that activates expression of nodulation (nod) genes on the resident symbiotic plasmid (pSym). Equivalent levels of carboxymethyl cellulase activity were found in wild-type strain ANU843 and its pSym-cured derivative, ANU845, consistent with previous results of Morales et al. (V. Morales, E. Martinez-Molina, and D. Hubbell, Plant Soil 80:407-415, 1984). However, polygalacturonase activity was lower in ANU845 and was not restored to wild-type levels in the recombinant derivative of pSym- ANU845 containing the common and host-specific nod genes within a 14-kb HindlIl DNA fragment of pSym from ANU843 cloned on plasmid pRtO32. Activity gel electrophoresis resolved three carboxymethyl cellulase isozymes of approximately 102, 56, and 33 kDa in cell extracts from ANU843. Isoelectric focusing activity gels revealed one ANU843 polygalacturonase isozyme with a pI of approximately 7.2. These studies show that R. leguminosarum bv. trifolii produces multiple enzymes that cleave glycosidic bonds in plant cell walls and that are cell bound.

An important event in the infection of white clover roots (5, 14, 26, 31, 35, 40), crossing of root cortical cell walls by by Rhizobium leguminosarum biovar trifolii leading to de- rhizobia in the advancing infection thread (3, 6), and release velopment of the root nodule symbiosis is the passage of the of rhizobia from infection threads into the nodule cell bacteria across the root hair wall (5, 10, 14, 24, 26, 35). This cytoplasm (6, 41). The strongest evidence for involvement of rigid assemblage of plant polysaccharides and glycoproteins wall hydrolysis in the R. leguminosarum bv. trifolii-white constitutes a barrier to host specificity (1). Several hypoth- clover infection process was obtained by Callaham and eses have been proposed to explain how this event occurs. Torrey (5), who showed a localized degradation of the root Nutman (27) proposed that rhizobia redirect growth of the hair wall coincident with deposition of a new wall layer, root hair wall from the tip to the localized site of infection. above the site of degradation, which is continuous with the This would cause invagination rather than penetration of the root hair wall. Recently, Baker et al. found many root hair wall, forming the tubular structure of the infection (2) that thread. Ljunggren and Fahraeus (20) proposed that homolo- cells of R. leguminosarum bv. trifolii attached to the root surface of white clover produce pit erosions in gous Rhizobium strains specifically induce the host plant to epidermal produce polygalacturonases, which soften the root hair wall walls that follow the contour of the bacterium, suggesting at the site of infection and thus allow the bacteria to that wall-degrading enzymes are associated with the bacte- penetrate between microfibrils to the cell membrane and rial cell surface itself and/or locally induced in the plant by initiate an infection thread. The ability ofRhizobium meliloti components of the bacterial surface. to induce polygalacturonase production by alfalfa roots is Rhizobial infection of legumes is a delicately balanced linked to a plasmid carried by the bacteria (28). The third process. If wall-degrading enzymes are involved, their pro- model (5, 10, 16) proposes that wall-degrading enzymes duction would have to be restricted to account for slow, produce a localized degradation that completely traverses localized penetration without destruction of the root hair and the root hair wall, allowing direct penetration by the bacte- subsequent abortion of the infection process (10, 16). McCoy ria. (24), in considering the role of hydrolytic enzymes in active Electron microscopic studies support the hypothesis that penetration of plant cell walls by certain pathogenic micro- hydrolytic enzymes are involved in various steps in the organisms, was the first to investigate the possible involve- infection process, including entry of rhizobia into root hairs ment of these enzymes in the infection of legumes by Rhizobium species. She found no evidence for these enzymes from rhizobia. However, sensitive procedures to * Corresponding author. detect minute amounts of cell wall-degrading enzymes were 1816 VOL. 58, 1992 CELLULASES AND PECTINASE IN R. LEGUMINOSARUM 1817 not available. More recently, several studies have detected B-INOS medium was overlaid with 5 ml of 0.2% (wt/vol) pectinolytic (17, 22, 23, 29, 30, 36), cellulolytic (25, 36), and substrate and 0.5% agarose in B-INOS medium. To detect in hemicellulolytic (21, 22) enzyme activities from pure cul- vitro activity, B-INOS medium was replaced with water in tures of rhizobia. In general, the activities of these rhizobial the bottom layer and 50 mM potassium phosphate-citric acid enzymes are very low and at the limit of sensitivity of buffer (PCA; pH 5.2) in the thin overlay; the same concen- conventional reducing sugar assays; this has hampered re- trations of agarose were maintained. Plates were inoculated search progress in this area. Nucleotide sequences homolo- by transferring cells from colonies with sterile toothpicks or gous to pectate lyase genes from Erwinia spp. do not by adding 10 p.l of cell extract containing 15 to 30 p.g of hybridize to DNA from R. tnifolii (13); however, a prelimi- protein; plates were incubated at 30°C. After the bacteria nary report of hybridization to DNA from Rhizobium fredii were grown for 4 days, colonies were washed off with water has been made (39). A cloned 2.5-kb fragment of DNA and discarded. To detect carboxymethyl cellulase (CM- encoding polygalacturonase from Agrobacterium tume- cellulase) activity, plates were flooded with 0.1% Congo red faciens biovar 3 did not hybridize to DNA from R. legumi- solution for 30 min and then rinsed several times with 1 M nosarum bv. viciae or Bradyrhizobium japonicum (33). NaCl. Areas of degradation of polygalacturonate were de- In this study, we used assays with improved sensitivity tected by adding filtered 0.05% ruthenium red solution to and reliability to verify the production of R. leguminosarum plates for 20 min and then rinsing with water. This procedure bv. trifolii enzymes that degrade polygalacturonate and revealed clear areas of hydrolysis. carboxymethyl cellulose (CMC) as model substrates of plant Preparation of culture supernatants and cell extracts. Broth cell wall polymers. We also used R. leguminosarum bv. cultures in the early stationary phase (9 x 108 cells per ml) trifolii wild-type strain ANU843 to evaluate (i) the propor- were centrifuged at 3,500 x g for 1 h at 4°C. Ammonium tion of extracellular activity to cell-associated activity, (ii) sulfate was added to the culture supernatant to 100% satu- methods for releasing these enzymes from cells, and (iii) ration at 4°C with constant stirring. Precipitated proteins in whether production of these enzymes is affected by the this extracellular fraction were centrifuged at 13,000 x g for resident symbiotic plasmid (pSym) (especially its 14-kb nod 15 min at 4°C, redissolved, and dialyzed for 4 h with three region), the flavone that enhances nod expression, or the changes of 300 ml of glycine-urea buffer (25) (1% glycine, 1 substrates themselves. M urea, 156 mM potassium phosphate, 22 mM citric acid (Portions of this work were presented at the 12th North [pH 7.4]) at 4°C. The cell pellet was suspended in 1.5 ml of American Symbiotic Nitrogen Fixation Conference [2].) PCA buffer on ice, sonicated in five cycles of 1-min bursts at 32 W with a microprobe (Fisher model 300 sonic dismem- MATERLILS AND METHODS brator), and centrifuged at 11,000 x g for 15 min at 4°C. Alternatively, the cells were washed with 30 mM Tris-HCI Bacterial strains and growth conditions. The strains of R. (pH 8.1), centrifuged at 11,000 x g for 15 min at 4°C, leguminosarum bv. trifolii used in this study were as follows: suspended in 1 ml of 20% sucrose containing 30 mM Tris- wild-type ANU843 (34), its pSym-cured derivative, ANU845 HCl (pH 8.1), and incubated in an ice bath for 40 min with (37), a recombinant strain containing the 14-kb HindIII DNA 100 p.l of 1-mg/ml lysozyme (Sigma Chemical Co., St. Louis, fragment encoding the entire nod region of pSym cloned on Mo.) in 100 mM EDTA at pH 7.3 to release proteins plasmid pRtO32 and introduced into the pSym-cured deriva- associated with the cell envelope (18). The lysozyme-EDTA- tive (strain ANU845pRtO32A6 with a Tn5 insertion in the treated suspension was centrifuged at 11,000 x g for 15 min vector) (38), derivatives of ANU843 containing a Tn5 inser- at 4°C, and the supernatant was dialyzed (Spectrapor mem- tion in pSym nodD (ANU851) (11, 37) or nodC (ANU277) brane tubing; molecular weight cutoff, 12,000 to 14,000) (12) (B. Rolfe, Australian National University), and a col- against PCA. The presence of cytoplasmic constituents in lection of strains (Fl through F9) isolated from nodules of these cell fractions was examined by measuring NADP- red clover (Tnfolium pratense) grown in soils of various pHs dependent glucose-6-phosphate dehydrogenase (E.C. near Salamanca, Spain (E. Martinez-Molina, University of 1.1.1.49) (43), a cytoplasmic enzyme marker in R. legumi- Salamanca). nosarum bv. trifolii (15). All enzyme samples were stored at Bacteria were cultured in BIII defined medium (8) with 40C. 0.5% (wt/vol) inositol (hereafter called B-INOS) instead of Quantitative enzyme assays. Substrates for CM-cellulase mannitol as the carbon source (25). Broth cultures were and polygalacturonase were the sodium salts of CMC (me- grown in 75 ml of B-INOS medium in 300-ml flasks shaken at dium viscosity) and polygalacturonic acid (both from Sig- 150 rpm at 28°C. Inocula were prepared by suspending cells ma), respectively, at 0.2% (wt/vol) in PCA buffer. Sodium from 5-day-old plate cultures into sterile B-INOS medium, polygalacturonate was dialyzed against 1 liter of PCA buffer centrifuging aseptically at 4,000 x g for 15 min, and resus- with three changes in 3 days. The reaction mixture contained pending in B-INOS medium to an initial population density 0.4 ml of substrate, 0.4 ml of enzyme solution, and 0.8 ml of of 107 cells per ml. In induction experiments, cells were PCA buffer in an Eppendorf tube. After the reagents were grown in B-INOS broth supplemented with 2 F.M 4',7- mixed quickly very well, an 0.8-ml aliquot (considered t = 0 dihydroxyflavone (DHF), 0.2% (wt/vol) CMC, or 0.2% so- h) was frozen at -80°C. The rest of the sample was incu- dium polygalacturonate. Plate cultures were grown on bated at 30°C for 5 h. B-INOS medium containing 4 p.M DHF and solidified with Product formation was measured by using the sensitive 2% purified agar (U.S. Biochemical Corp., Cleveland, Ohio). 2,2'-bicinchoninate method of Waffenschmidt and Jaenicke Kanamycin sulfate (30 ,ug/ml) was included in culture me- (42). This assay measures reducing sugars in the nanomole dium to retain the recombinant plasmid in strain ANU845 range without interference from proteins. Samples of 200 ,ul pRtO32A6. Cultures were routinely checked for purity by were mixed with 0.8 ml of water and 1 ml of 2,2'-bicinchon- streaking on BIII and Trypticase soy agar plates. inate reagent (42) in tubes, which were tightly capped and Double-layer plate assay. A modification of the Saleh- heated at 100°C for 15 min in a heating block. Samples were Rastin et al. (36) plate assay was used. To detect in vivo then cooled to room temperature for 20 min, and their optical activity, a bottom layer containing 15 ml of 0.7% agarose in density was measured at 540 nm; the blank contained 1 ml of 1818 MATEOS ET AL. APPL. ENVIRON. MICROBIOL. water and 1 ml of 2,2'-bicinchoninate reagent with no sugar. plate assay in sonicated cell extracts of R. leguminosarum Standard curves for reducing sugar were prepared with bv. trifolii ANU843 grown on B-INOS plates (Fig. 1A). Both glucose and galacturonic acid in the range of 0 to 55 nmol for enzyme activities were detected within 4 h of incubation and CM-cellulase and polygalacturonase activities, respectively. were destroyed by boiling. No activity was found in this One unit of enzyme activity is defined as the amount extract when assayed on plates formulated to detect pectate releasing 1 nmol of reducing sugar equivalent per min at lyase activity (100 mM Tris-HCl [pH 8.5], 1.5 mM CaCl2 [7]) 30°C. The amount of protein in samples was measured by the (data not shown). CM-cellulase activity could also be de- dye-binding method of Bradford (4) with bovine serum tected directly underneath colonies growing on plates con- albumin (Sigma) as standard. taining B-INOS agar plus 0.2% CMC. This method gave Activity stain overlay technique to detect CM-cellulase after positive results for CM-cellulase activity from several native PAGE. Sodium dodecyl sulfate-polyacrylamide gel electro- strains of R. leguminosarum bv. trifolii isolated from nod- phoresis (SDS-PAGE) was performed in 0.75-mm gels in a ules of T. pratense (Fig. 1B). Polygalacturonase activity vertical slab unit (Mini-Protean II; Bio-Rad) by a modifica- could not be detected directly by this colony assay on tion of the method of Laemmli (19). The separating gel B-INOS-polygalacturonate plates, since the medium be- contained 10% acrylamide, 0.33% Bis, and 20 ,ug of bovine neath the colony stained more intensely with ruthenium red serum albumin per ml to enhance renaturation of electro- (data not shown). This is probably due to rhizobial acidic phoresed enzymes (32). The stacking gel contained 4% exopolysaccharides (9) that have diffused from the colony acrylamide/Bis. SDS was excluded from both gels but added into the agar medium and are also stained intensely with at 0.05% to the running buffer. Enzyme samples were mixed ruthenium red. Alternatively, polygalacturonase activity with sample buffer (62 mM Tris-HCl [pH 6.8], 10% glycerol, was detected in sonicated cell extracts from these same R. 0.025% bromophenol blue, and 2% SDS) in a ratio of 5:1 leguminosarum bv. trifolii strains spotted onto double-layer (vol/vol) without heating. In native PAGE, the SDS and plates made with polygalacturonate in PCA buffer (Fig. 1B). bovine serum albumin were eliminated and the separating gel These results indicate that the double-layer plate assay is contained 7% acrylamide-Bis. Samples were electro- sensitive enough to detect the low levels of CM-cellulase and phoresed at 200 V until the tracking dye migrated to the polygalacturonase activities produced by a diversity of wild- bottom of the gel. The proteins in SDS gels were renatured type R. leguminosarum bv. trifolii strains and should have by incubating for 2 h with shaking in three 100-ml changes of utility in studies that require sensitive methods to screen for 10 mM PCA buffer. CMC-agarose overlays were constructed these enzyme activities. by using a 0.4-mm layer of 0.2% CMC and 0.5% agarose in CM-cellulase and polygalacturonase are cell bound in R. PCA buffer on a sheet of GelBond support film as recom- leguminosarum bv. trifolii ANU843. Previous studies of R. mended by the manufacturer (FMC BioProducts, Rockland, leguminosarum bv. trifolii 0403 grown in BIll defined me- Maine). The running gel was placed on the substrate-impreg- dium (15) have shown that cells in the early stationary phase nated overlay and incubated in a moist chamber at 30°C (7 h at a population density of ca. 9 x 108/ml have not begun for native PAGE, 12 h for SDS-PAGE). Then the overlay stationary-phase lysis (detected by assaying the cytoplasmic was immersed in aqueous 0.1% Congo red solution for 30 enzyme marker glucose-6-phosphate dehydrogenase in the min and rinsed in 1 M NaCl. extracellular fraction). The total and specific activities of Activity stain for polygalacturonase in isoelectric focusing polygalacturonase and CM-cellulase were measured by the gels. Agarose gels (10% glycerol, 1% agarose, 5% sorbitol, BCA-reducing sugar method in various fractions of R. legu- 2% Bio-Lyte 3/10 or 6/8 ampholytes from Bio-Rad Labora- minosarum bv. trifolii ANU843 grown in B-INOS broth to a tories) were cast in 0.8-mm trays. Paper electrode wicks density of 9 x 108 cells per ml. Enzyme activities were not were moistened in 100 mM glutamic acid or 500 mM acetic found in the extracellular fraction containing the proteins acid (anode) and 500 mM NaOH (cathode). Enzyme samples precipitated by ammonium sulfate (Table 1). In contrast, were prepared by sonicating cells of strain ANU843 as both enzyme activities were detected repeatedly in extracts described above but in deionized water; then ampholyte of pelleted cells produced by sonication or treatment with solution was added to a 2% (vol/vol) final concentration. lysozyme-EDTA (Table 1). Although the specific activity for Samples (10 [l) were applied to paper wicks placed near the polygalacturonase was lower in the supernatant of sonicated cathode. A mixture of isoelectric focusing standards (Bio- cells (due to more extraneous protein), total polygalactu- Rad) with a pl range of 4.6 to 9.6 was applied next to the ronase activity was higher in this fraction. Lysozyme-EDTA enzyme sample. Isoelectric focusing was performed at 4°C treatment released a larger portion of the polygalacturonase on a Bio-Phoresis Horizontal Electrophoresis cell (Bio-Rad) activity than the CM-cellulase activity from pelleted cells. at a constant 6 W with a maximum of 1.5 kV until a rate of As expected, the cytoplasmic enzyme marker NADP-depen- 1.0 kV- h was reached. After focusing was complete, the dent glucose-6-phosphate dehydrogenase was present in the lane containing pI standards was excised and stained for cell sonic extract (0.36 U/mg of protein; 1 U reduces 1 p,mol protein with Coomassie blue according to the instructions of of NADP/min at 30°C and pH 7.8 [43]) but was not detected the manufacturer. The remaining gel was rinsed for 15 min in in the fraction of cell-bound proteins released by lysozyme- PCA buffer, placed on the substrate-impregnated overlay (as EDTA. This indicated that cell lysis did not occur during described above with 0.2% sodium polygalacturonate), and lysozyme-EDTA treatment and that portions of the polyga- incubated in a moist chamber at 30°C for 3 h. Then the lacturonase and CM-cellulase activities are associated with overlay was stained with 0.05% ruthenium red for 20 min, the cell envelope of R. leguminosarum bv. trifolii ANU843. rinsed with water, and photographed. These results corroborate an earlier report (25) that the main proportion of the total cellulase activity in broth cultures of RESULTS AND DISCUSSION R. leguminosarum bv. trifolii remains associated with the cells after centrifugation. CM-cellulase and polygalacturonase activities are produced The symbiotic plasmid pSym of R. leguminosarum bv. by R. leguminosarum bv. trifolii. CM-cellulase and polyga- trifolii ANU843 contributes to its polygalacturonase activity lacturonase activities were detected by the double-layer but not to its CM-cellulase activity. Root extracts stimulate VOL. 58, 1992 CELLULASES AND PECTINASE IN R. LEGUMINOSARUM 1819 CM-CELLULOSE Na POLYPECTATE

4h 8h 24h 4h 8h 24h

A UN BOILED

BOILED

FIG. 1. CM-cellulase and polygalacturonase activities associated with R. leguminosarum bv. trifolii and detected by the double-layer plate assay. (A) Sonicated cell extracts (unboiled or boiled) from strain ANU843 were incubated for 4, 8, and 24 h on plates. (B) Cells or sonicated cell extracts from strains Fl (upper left corner) through F9 (lower right corner) were spotted onto plates containing a CMC or sodium polypectate substrate, respectively. the production of cellulases by R. leguminosarum bv. trifolii flavone did not significantly alter the levels of polygalactur- BAL (25). Expression of nod genes on pSym in R. legumi- onase and CM-cellulase activities of R. leguminosarum bv. nosarum bv. trifolii ANU843 is increased significantly by trifolii ANU843, and these enzyme activities remained asso- growth in the presence of DHF, a flavone released from ciated with the cells rather than in the extracellular fraction white clover roots (11). We therefore examined whether (Table 1). growth with DHF affected the levels of CM-cellulase and Morales et al. (25) showed that curing of pSym did not polygalacturonase activities of R. leguminosarum bv. trifolii affect cellulase production in R. leguminosarum bv. trifolii ANU843. Growth in the presence of this nod-inducing 521. We examined the influence of pSym in the CM-cellulase

TABLE 1. CM-cellulase and polygalacturonase activities in R. leguminosarum bv. trifolii wild-type ANU843 and its pSym-cured derivative, ANU845a

2 ,uM DHF in Cell fraction or Polygalacturonase activity (U) CM-cellulase activity (U) StralnB-INOSmedium treatment Total Specific Total Specific ANU843 (wild type) - Extracellular 0 + 0 0 ± 0 0 ± 0 0 ± 0 ANU843 - Sonic extract 5.77 ± 0.35 2.25 ± 0.14 0.98 ± 0.33 0.38 + 0.13 ANU843 - Lysozyme-EDTA 2.75 ± 0.03 17.38 ± 0.18 0.09 ± 0.01 0.57 + 0.06 ANU843 + Extracellular 0 0 0 0 0 0 0 0 ANU843 + Sonic extract 6.17 + 0.87 2.37 + 0.33 0.91 + 0.17 0.35 ± 0.07 ANU843 + Lysozyme-EDTA 2.68 ± 0.06 20.40 ± 0.46 0.06 ± 0.01 0.46 + 0.10 ANU845 (pSym-) + Extracellular 0 ± 0 0 ± 0 0 ± 0 0 ± 0 ANU845 + Sonic extract 1.65 ± 0.20 0.65 ± 0.08 1.26 + 0.35 0.49 ± 0.13 a One unit of enzyme activity is the amount releasing 1 nmol of reducing sugar per min at 30'C. The values reported are the means of triplicate samples derived from equivalent numbers of cells ± the standard errors of the means. 1820 MATEOS ET AL. APPL. ENVIRON. MICROBIOL. CM-CELLULOSE Na POLYPECTATE

FIG. 2. Influence of pSym and pSym nod genes on CM-cellulase and polygalacturonase activities in sonicated cell extracts of R. leguminosarum by. trifolii ANUL843 grown on B-INOS plates with 4 pLM DHF. Numbered spots correspond to the following extracts: 1, wild-type strain ANU843; 2, pSym-cured derivative ANU845; 3, recombinant strain ANU845pRtO32A6; 4, nodD::TnS derivative ANU851; 5, nodC::TnS derivative ANU277. and polygalacturonase activities of R. leguminosarum bv. turonase activities of ANU845 and ANU845pRtO32A6 were trifolii ANU843 by comparison to these enzyme activities in markedly reduced (Fig. 2). These results were confirmed by the pSym-cured derivative, strain ANU845. Consistent with using the quantitative BCA-reducing sugar assay, which the results of Morales et al. (25), CM-cellulase activity showed a ca. 3.5-fold reduction in polygalacturonase activity associated with cells grown in broth containing DHF was not in both ANU845 and ANU845pRtO32A6 (Table 2). altered in the pSym-cured strain ANU845 (Table 1). How- These data obtained with R. leguminosarum bv. trifolii ever, the polygalacturonase activity in the sonicated cell ANU843 grown on B-INOS medium indicate that (i) neither extract was ca. 3.5-fold lower in this mutant, indicating a flavone induction of pSym nod genes nor the positive regula- contribution of pSym to the total polygalacturonase activity tory gene, nodD, is required for production of the CM- of wild-type ANU843 (Table 1). cellulase and polygalacturonase activities detected by these The possible influence of the 14-kb nod region of pSym on assays; (ii) genes encoding the CM-cellulase(s) are present on CM-cellulase and polygalacturonase activities of ANU843 either the bacterial chromosome or cryptic plasmids; (iii) was first examined by using the double-layer plate assay. pSym does contribute positively to the total polygalactur- CM-cellulase activity was detected in sonicated cell extracts onase activity; and (iv) the recombinant plasmid pRtO32A6 of wild-type ANU843, pSym-cured ANU845, recombinant containing the cloned 14-kb pSym nod region does not restore strain ANU845pRtO32A6 (containing the pSym nod region of the polygalacturonase activity of the pSym-cured strain to ANU843 on plasmid pRtO32 in the pSym-cured ANU845 that of its ANU843 parent. It is likely that a locus that background), and the nodD::TnS (ANU851) and nodC::Tn5 contributes to wild-type levels of polygalacturonase activity (ANU277) insertion mutant derivatives of ANU843 (Fig. 2). associated with R. leguminosamum bv. trifolii ANU843 re- These results are consistent with those of quantitative assays sides on its pSym outside of the 14-kb HindIII nod region. of CM-cellulase from ANU843 and ANU845 (Table 1). In CM-cellulase and polygalacturonase activities are not af- addition, assay of the same sonicated cell extracts on plates fected in R. leguminosarum bv. trifolii ANU843 by growth in containing sodium polypectate indicated that the polygalac- the presence of substrates. Growth in B-INOS broth supple-

TABLE 2. Contribution of pSym to the polygalacturonase TABLE 3. Polygalacturonase and CM-cellulase activities of activity of sonicated cell extracts from R. leguminosarum sonicated cell extracts of R. leguminosarum bv. trifolii ANU843 bv. trifolii ANU843 and derivativesa grown in the presence of polysaccharide substratesa Strain Polygalacturonase Polygalacturonase CM-cellulase ANU843 Supplement to activity (U) activity (U) (wild type)...... 1.90 +0.03 B-INOS medium ANU845 (pSym-)...... 0.54 ± 0.10 Total Specific Total Specific ANU845pRtO32A6 ...... 0.48 ± 0.07 ANU851 (nodD::TnS)...... 2.00 +0.20 None 2.24 + 0.06 2.25 + 0.06 0.56 ± 0.03 0.38 + 0.02 ANU277 (nodC::TnS) ...... 1.80 ± 0.06 0.2% CMC ND ND 0.62 + 0.05 0.42 ± 0.03 0.2% Polypectate 1.92 + 0.15 1.93 + 0.15 ND ND a Cells were grown on B-INOS plates containing 4 ,uM DHF. Specific activitiy is given as units of enzyme activity per milligram of total protein. a Values reported are the means of triplicate samples derived from equiv- Values are means from triplicate experiments + standard errors of the means. alent numbers of cells ± standard errors of the means. ND, not determined. VOL. 58, 1992 CELLULASES AND PECTINASE IN R. LEGUMINOSARUM 1821 A B In summary, the improved sensitivity of methods used in this work has permitted reliable detection and quantitation of 1 2 1 2 CM-cellulase and polygalacturonase activities associated with R. leguminosarum bv. trifolii, without the handicap of their inherently low activity. These studies are consistent 97.4 with earlier reports that rhizobia produce enzymes that can 66.2- break glycosidic bonds present in plant cell walls and pro- vide evidence that these enzyme activities are cell bound. At 45.0 least a portion of these enzyme activities is released by 31.0- treatment of cells with lysozyme-EDTA to disrupt their cell envelope without lysis of the protoplast. This is the first report of CM-cellulase and polygalacturonase isozymes in R. 21.5- leguminosarum bv. trifolii and the contribution of pSym to cell-associated polygalacturonase activity. Further use of FIG. 3. Activity gel overlay detection of CM-cellulase isozymes these improved methods should make it possible to investi- in sonicated cell extracts of R. leguminosarum bv. trifolii ANU843 gate the role of these polysaccharide-degrading enzymes in after separation by native PAGE (A) and SDS-PAGE (B). Lanes: 1, the 100 ,ug of protein; 2, 150 ,ug of protein. Lane 2 in panel B was infection process of this nitrogen-fixing Rhizobium- overloaded to show the third isozyme at an apparent molecular mass legume symbiosis. of 33 kDa. The molecular masses (in kilodaltons) of protein standards are indicated to the left of panel B. ACKNOWLEDGMENTS

This work was supported by the Center for Microbial Ecology and mented with 0.2% CMC or polypectate substrates did not the Biotechnology Research Center at Michigan State University (to appreciably change the specific activities of the correspond- F.B.D.) and by Direcci6n Generalde Investigaci6n Cientifica y ing depolymerizing enzymes in the sonicated cell fraction of Tecnica grant PB87-0435-CO3-03 (to E.M.M.). R. leguminosarum bv. trifolii ANU843, indicating no evi- We thank B. Rolfe for providing nod mutant and recombinant dence for substrate induction (Table 3). strains of ANU843 and S. Coleman, G. Orgambide, and S. Philip- Three CM-cellulase isozymes and one polygalacturonase Hollingsworth for helpful comments. isozyme are detected in sonicated extracts ofR. leguminosarum bv. trifolii ANU843. Three CM-cellulase isozymes in the REFERENCES sonicated cell extract of R. trifolii leguminosarum bv. 1. Al-Mallah, M., M. Davey, and E. Cocking. 1987. Enzymatic ANU843 were resolved by native PAGE and detected by the treatment of clover root hairs removes a barrier to Rhizobium- activity stain overlay technique (Fig. 3A). The same sample host specificity. Bio/Technology 5:1319-1322. electrophoresed in the presence of SDS yielded three CM- 2. Baker, D., M. Petersen, M. Robeles, J. Chen, A. Squartini, F. cellulase isozymes, which, by comparison to the low-molec- Dazzo, and D. Hubbell. 1989. Pit erosion of root epidermal cell ular-mass standards of Bio-Rad, behaved like proteins with walls in the Rhizobium-white clover symbiosis. 12th North approximate molecular masses of 102, 56, and 33 kDa under American Symbiotic Nitrogen Fixation Conference, Iowa State the conditions described (Fig. 3B). One polygalacturonase University Press, Ames. isozyme was detected in the same sonicated cell extract of 3. Bakhuizen, R., P. van Spronsen, and J. Kijne. 1988. Cell wall degradation and infection thread initiation in pea root cortical ANU843 by using isoelectric focusing followed by the activ- cells during infection by Rhizobium leguminosarum biovar ity stain overlay technique, first with the 3/10 ampholyte viciae, p. 83-107. In R. Bakhuizen, The plant cytoskeleton in mixture and later with the 6/8 ampholyte mixture. Calibra- the Rhizobium-legume symbiosis. Ph.D. dissertation, Univer- tion of the isoelectric focusing gel made with the 6/8 am- sity of Leiden, The Netherlands. pholyte mixture with Bio-Rad standards indicated that the 4. Bradford, M. M. 1976. A rapid and sensitive method for the major polygalacturonase associated with ANU843 has an quantitation of microgram quantities of protein utilizing the approximate pl of 7.2 (Fig. 4). principle of protein-dye binding. Anal. Biochem. 72:248-254. 5. Callaham, D., and J. Torrey. 1981. The structural basis for infection of root hairs of Trifolium repens by Rhizobium. Can. J. Bot. 59:1647-1664. pi 6. Chalifour, F., and A. Benhamou. 1989. Indirect evidence for cellulase production by Rhizobium in pea root nodules during 7.75 bacteroid differentiation: cytochemical aspects of cellulose breakdown in rhizobial droplets. Can. J. Microbiol. 35:821-829. _7.0 7. Collmer, A., J. Ried, and M. Mount. 1988. Assay methods for pectic enzymes. Methods Enzymol. 161:329-335. 8. Dazzo, F. 1984. Leguminous root nodules, p. 431-446. In R. 1-6.5 Burns and J. Slater (ed.), Experimental microbial ecology. Blackwell Scientific Publishers, Oxford. 9. Dazzo, F., and W. Brill. 1979. Bacterial polysaccharide which binds Rhizobium trifolii to white clover root hairs. J. Bacteriol. -6.0 137:1362-1373. 10. Dazzo, F., and D. Hubbell. 1982. Control of root hair infection, p. 274-310. In W. Broughton (ed.), Nitrogen fixation, vol. 2. FIG. 4. Activity gel overlay detection of the major polygalactu- Rhizobium. Clarendon Press, Oxford. ronase isozyme in sonicated cell extracts of R. leguminosarum bv. 11. Djordjevic, M., J. Redmond, M. Batley, and B. Rolfe. 1987. trifolii ANU843 after separation by isoelectrofocusing with 6/8 Clovers secrete specific phenolic compounds which stimulate or ampholytes. The sample contained 20 pug of total protein. The pls of repress nod gene expression in Rhizobium trifolii. EMBO J. isoelectrofocusing standards are indicated. 6:1173-1179. 1822 MATEOS ET AL. APPL. ENVIRON. MICROBIOL.

12. Djordjevic, M., P. Schofield, and B. Rolfe. 1985. Tn5 mutagen- 29. Plazinski, J., and B. Rolfe. 1985. Analysis of the pectolytic esis of Rhizobium tnfolii host specific nodulation genes results activity of Rhizobium and Azospirillum strains isolated from in mutants with altered host range ability. Mol. Gen. Genet. Trifolium repens. J. Plant Physiol. 120:181-187. 200:463-471. 30. Prasuna, P., and S. Ali. 1987. Detection and characterization of 13. Fennington, G., and T. Hughes. 1990. Erwinia pel gene homol- two thermally reactive pectinases in cultures ofRhizobium. Ind. ogy survey in selected bacteria. Plant Soil 125:285-287. J. Exp. Biol. 25:632-633. 14. Higashi, S., and M. Abe. 1980. Scanning electron microscopy of 31. Ridge, R., and B. Rolfe. 1985. Rhizobium species degradation of Rhizobium trifolii infection sites on root hairs of white clover. legume root hair cell wall at the site of infection thread origin. Appl. Environ. Microbiol. 40:1094-1099. Appl. Environ. Microbiol. 50:717-720. 15. Hrabak, E., M. Urbano, and F. Dazzo. 1981. Growth-phase 32. Ried, J., and A. Collmer. 1985. Activity stain for rapid charac- dependent immunodeterminants of Rhizobium tnjfolii lipopoly- terization of pectic enzymes in isoelectric focusing and sodium saccharide which bind the white clover lectin, trifoliin A. J. dodecyl sulfate-polyacrylamide gels. Appl. Environ. Microbiol. Bacteriol. 148:697-711. 50:615-622. 16. Hubbell, D. 1981. Legume infection by Rhizobium: a conceptual 33. Rodriguez-Palenzuela, P., T. Burr, and A. Collmer. 1991. Poly- approach. BioScience 31:832-837. galacturonase is a virulence factor in Agrobacterium tume- 17. Hubbell, D., V. Morales, and M. Umali-Garcia. 1978. Pectolytic faciens biovar 3. J. Bacteriol. 173:6547-6552. enzymes in Rhizobium. Appl. Environ. Microbiol. 35:210-224. 34. Rolfe, B., P. Gresshoff, and J. Shine. 1980. Rapid screening for 18. Ito, K., T. Sato, and T. Yura. 1977. Synthesis and assembly of symbiotic mutants of Rhizobium and white clover. Plant Sci. the membrane proteins in E. coli. Cell 11:551-559. Lett. 19:227-284. 19. Laemmli, U. K. 1970. Cleavage of structural proteins during the 35. Sahlman, K., and G. Fahraeus. 1963. An electron microscope assembly of the head of bacteriophage T4. Nature (London) study of root hair infection by Rhizobium. J. Gen. Microbiol. 227:680-685. 33:425-427. 20. Ljunggren, H., and G. Fahraeus. 1961. The role of polygalactu- 36. Saleh-Rastin, N., M. Petersen, S. Coleman, and D. Hubbell. ronase in root-hair invasion by nodule bacteria. J. Gen. Micro- 1991. Rapid plate assay for hydrolytic enzymes ofRhizobium, p. biol. 26:521-528. 188. In D. Keister and P. Gregon (ed.), The rhizosphere and 21. Lopez, M., and E. Signer. 1987. Degradative enzymes in Rhizo- plant growth. Kleuvier Academic Publishers, Dordrecht, The bium meliloti, p. 185-187. In D. Verma and N. Brisson (ed.), Netherlands. Molecular genetics of plant-microbe interactions. Martinus Ni- 37. Schofield, P., M. Djordjevic, B. Rolfe, J. Shine, and J. Watson. jhoff Publishers, Dordrecht, The Netherlands. 1983. A molecular linkage map of nitrogenase and nodulation 22. Martinez-Molina, E., V. Morales, and D. Hubbell. 1979. Hydro- genes in Rhizobium tnfolii. Mol. Gen. Genet. 192:459-465. lytic enzyme production of Rhizobium. Appl. Environ. Micro- 38. Schofield, P., R. Ridge, B. Rolfe, J. Shine, and J. Watson. 1984. biol. 38:1186-1188. Host-specific nodulation is encoded on a 14 kb fragment in 23. Martinez-Molina, E., and J. Olivares. 1982. A note on evidence Rhizobium tnfolii. Plant Mol. Biol. 3:3-11. for involvement of pectolytic enzymes in the infection process 39. Shantharam, S., H. Dae Yun, D. Swanlund, and A. Atherly. of Medicago sativa by Rhizobium meliloti. J. Appl. Bacteriol. 1988. Identification and cloning of two pectate lyase genes from 52:453-455. Rhizobium fredii USDA193, p. 202-203. In R. Palacios and D. 24. McCoy, E. 1932. Infection by Bact. radicicola in relation to the Verma (ed.), Molecular genetics of plant-microbe interactions. microchemistry of the host's cell walls. Proc. R. Soc. London American Phytopathological Society Press, St. Paul, Minn. Ser. B 110:514-533. 40. Turgeon, B., and W. Bauer. 1985. Ultrastructure of infection- 25. Morales, V., E. Martinez-Molina, and D. Hubbell. 1984. Cellu- thread development during the infection of soybean by Rhizo- lase production by Rhizobium. Plant Soil 80:407-415. bium japonicum. Planta 163:328-349. 26. Napoli, C., and D. Hubbell. 1975. Ultrastructure of Rhizobium- 41. Verma, D., V. Zogbi, and A. Bal. 1978. A cooperative action of induced infection threads in clover root hairs. Appl. Microbiol. plant and Rhizobium to dissolve the host cell wall during 30:1003-1009. development of root nodule symbiosis. Plant Sci. Lett. 13:137- 27. Nutman, P. S. 1956. The influence of the legume in root nodule 142. symbiosis. A comparative study of host determinants and 42. Waffenschmidt, S., and L. Jaenicke. 1987. Assay of reducing functions. Biol. Rev. Cambridge Phil. Soc. 31:109-151. sugars in the nanomole range with 2,2'-bicinchoninate. Anal. 28. Palomares, A., E. Montoya, and J. Olivares. 1978. Induction of Biochem. 165:337-340. polygalacturonase production in legume roots as a consequence 43. Worthington, C. C. 1988. Glucose-6-phosphate dehydrogenase. of extrachromosomal DNA carried by Rhizobium meliloti. Mi- Worthington enzyme manual, p. 159-161. Worthington Bio- crobios 21:33-39. chemical Corporation, Freehold, N.J.