APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 1990, p. 3081-30870099-2240/90/103081-07$02.00/0Copyright C) 1990, American Society for Microbiology

Vol. 56, No. 10

Effects of Physicochemical Factors on the Adhesion to CelluloseAvicel of the Ruminal Bacteria Ruminococcus flavefaciens and

Fibrobacter succinogenes subsp. succinogenesVALERIE ROGER,1 GE1RARD FONTY,l.2* SYLVIE KOMISARCZUK-BONY,3t AND PHILIPPE GOUET1

Laboratoire de Microbiologie, INRA, CR de Clermont-Ferrand-Theix, 63122 Ceyrat,1 Laboratoire de BiologieCompar&e des Protistes, CNRS URA 138, Universite Blaise Pascal, Clermont II, 63170 Aubiere,2

and Station de Recherches de Nutrition, INRA, 78350 Jouy-en-Josas, France

Received 23 March 1990/Accepted 14 July 1990

Ruminococcus flavefaciens adhered instantly to cellulose, while Fibrobacter succinogenes had the highestpercentage of adherent cells after about 25 min of contact between bacteria and cellulose. Adhesion of R.flavefaciens was unaffected by high concentrations of sugars (5%), temperature, pH, oxygen, metabolicinhibitors, and lack of Na+. In contrast, the attachment was affected by the removal of divalent cations (Mg2+and Ca2+), the presence of cellulose derivatives (methylcellulose and hydroxyethylcellulose), and cystine.Adhesion of F. succinogenes was sensitive to low and high temperatures, high concentrations of glucose andcellobiose (5%), hydroxyethylcellulose (0.1%), redox potential, pH, lack of monovalent cations, and thepresence of an inhibitor of membrane ATPases or lasalocid and monensin. Cells of F. succinogenes heated at100°C no longer were adherent. On the other hand, adhesion was insensitive to the lack of divalent cations(Mg2+ and Ca2+), the presence of 2,4-dinitrophenol, tetrachlorosalicylanilide, or inhibitors of the electrontransfer chains. Adhesion of F. succinogenes seems to be related to the metabolic functions of the cell. Externalproteins and/or cellulases themselves might play a part in the attachment process. Several mechanisms are

probably involved in the adhesion of R. flavefaciens, the main one being the interaction between the largeglycocalyx and the divalent cations Ca2' and Mg2+. Hydrophobic bonds and enzymes may also be involved.

In the rumen, fibrolytic microorganisms rapidly colonizeplant particles after their ingestion (2, 19). Adhesion ofcellulolytic bacteria brings the cell into close contact with itsspecific substrate and concentrates hydrolytic enzymes oncellulose. Several qualitative electron microscopic studieshave been done on adhesion of the three dominant ruminalcellulolytic bacteria to cellulose and plant cell walls (3, 4, 8,9, 11, 13, 17, 19, 22). The influence of some physicochemicalfactors on the adhesion of these bacteria has been investi-gated (14, 20, 21, 24, 27, 32). The adherence process of thesespecies, however, is still unknown. Ruminococcus albus andRuminococcusflavefaciens possess a large glycocalyx that ismainly glycoproteic (19). The integrity of the cell coat wouldseem to be essential for their adhesion since the action ofperiodate or pepsin prevents bacterial adhesion (18). Oncontact with the substrate, the coat is often thicker than onthe free surface of the bacterium, as the result of a stretchingof the extracellular material (1). The glycocalyx of Fibrobac-ter succinogenes, formerly Bacteroides succinogenes (23),unlike that of R. flavefaciens and R. albus, is very thin (19).The bacterium seems to make up for the small quantity ofthis extracellular material by the flexibility of the cell wall,which enables it to conform to the topography of thesubstrate. F. succinogenes also has fine filaments that linkthe cells and might be involved in the adhesion to cellulose(15). In old cultures, small vesicles deriving from the bacte-rial outer membrane have been observed in close contactwith cellulose (13). These vesicles have been shown to havecellulase and xylanase activities, and it has been suggested

* Corresponding author.t Present address: Laboratoire de la Chaire d'Alimentation,

Ecole Nationale Vdterinaire de Lyon, 69280 Marcy l'Etoile, France.

but not proved that they are involved in bacterial adhesion(11).The aim of this work was to compare the effects of some

abiotic factors on the adhesion of F. succinogenes and R.flavefaciens to cellulose.

MATERIALS AND METHODSBacterial strains used. We studied the effects of all factors

on R. flavefaciens 007 and F. succinogenes subsp. succino-genes S85 (ATCC 19169), kindly provided by C. S. Stewart(Rowett Research Institute, Aberdeen, United Kingdom)and M. P. Bryant (University of Illinois, Urbana-Cham-paign), respectively.We also investigated the effects of some of these factors

(see Results) on the adhesion of the following newly isolatedruminal cellulolytic strains provided by K. J. Cheng (Re-search Station, Lethbridge, Alberta, Canada): R. flave-faciens 083 and F. succinogenes LRS095 isolated from asteer and R. flavefaciens 131 and F. siuccinogenes LRS128isolated from a buffalo.

Strains S85 and LRS095 were grown on Ml culturemedium derived from that of Bryant et al. (7), modified by S.Komisarczuk-Bony and G. Fonty in our laboratory (Table 1)(unpublished data), and strains 007 and 083 were grown onthe medium of Scott and Dehority (31). Since R. flavefaciens131 and F. succinogenes LRS128 do not grow on syntheticor semisynthetic medium, they were grown on mediumRGMC (20), modified in our laboratory to contain thefollowing (per 100 ml): mineral solution 1, 7.5 ml; mineralsolution 2, 7.5 ml; resazurin (0.1%), 0.1 ml; NaHCO3, 0.4 g;clarified ruminal fluid, 30 ml; cellobiose, 0.2 g; cysteinehydrochloride, 0.05 g; distilled water, 85 ml. Mineral solu-tions 1 and 2 were those described previously by Bryant andBurkey (5).

3081

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

TABLE 1. Composition of medium Ml

Substance or parameter Amt

Mineral Ml solution"................ 75 mlVitamin solutionb................ 10 mlTrace mineral solution ................ 10 mlFatty acid solution"................ 3 mlHemin solutione................ 1 mlResazurin solutiont ................ 1 mlCellobiose................ 3 gNaHCO3 ................ 4 gCysteine hydrochloride ................ 0.6 gpHI ................ 7Gas phase................ 100% CO2

Mineral Ml solution consists of the following (grams per liter of distilledwater): KH2PO4, 6; (NH4)2S04, 12; NaCI, 12; KCI, 6; MgSO4- 7H20, 1.2;CaCl,- 2H20, 0.6.

Vitamin solution, consists of the following (milligrams per liter of distilledwater): biotin, 50; para-aminobenzoic acid, 100; pyridoxamine, 1.000.

'Trace mineral solution consists of the following (milligrams per liter ofdistilled water): FeSO4 7H20, 1,000; MnSO4* H20, 150; CoCI2 6H20, 150.

d Fatty acid solution consists of the following (milliliters): acetic acid, 17;propionic acid, 6; butyric acid, 4; isobutyric acid, 1; N-valeric acid, 1;isovaleric acid, 1; 2-methylbutyric acid, 1.

e Hemin (0.1 g) is dissolved in 10 ml of ethanol and then brought to 1,000 mlwith 0.05 M NaOH.fResazurin (0.1 g) is dissolved in 100 ml of distilled water.g Final pH 7 is adjusted with 3 M NaOH.

All strains were cultured under 100% CO2 in Hungatetubes (16 by 25 mm; Bellco Glass, Inc., Vineland, N.J.)containing 9 ml of medium and sealed with butyl stoppers.Cellobiose was the sole energy source in the media.

Determination of percentage of bacteria attached to cellu-lose. The effects of the different factors were investigatedwhen bacterial growth was at the end of the exponentialphase. At this time, the optical density was usually between0.6 and 0.7 for R. flavefaciens and 1.2 and 1.3 for F.succinogenes. The bacterial cultures were centrifuged at3,500 x g for 10 min. The supernatant was decanted, and thepellet was suspended in 9 ml of medium of the same

composition containing the test factor and without cellobioseand resazurin, which may have modified optical densityreadings.The percentage of bacteria adhering to cellulose was

calculated by the method of Minato and Suto (20), whichconsists of measuring the optical density of the culturebefore and after the addition of cellulose. Optical densitywas read at 600 nm. We studied adhesion to microcrystallinecellulose Avicel (microcrystalline cellulose for chromatogra-phy; Macherey Nagel, Duren, Federal Republic of Ger-many) added at a concentration of 0.2%. At least fourreplicates were made for each factor.

Effects of physicochemical factors on adhesion of bacteria tocellulose Avicel. (i) Contact time between bacteria and cellu-lose. After the addition of cellulose Avicel, the followingcontact times between bacteria and cellulose were studied:1-min hand shaking (S); S + 4 min of incubation at 38°C; S +14 min of incubation at 38°C; S + 29 min of incubation at

38°C; S + 59 min of incubation at 38°C. The cultures were

hand shaken every 5 min to resuspend the cellulose.(ii) Influence of temperature. Adhesion was measured at 4,

13, 20, 30, 38, 42, 43, 44, 46, 48, and 52°C. The cultures of R.flavefaciens 007 were left for 10 min at the desired temper-

ature before cellulose was added. They were then shaken for1 min and immediately centrifuged at 500 x g. The culturesof F. succinogenes S85 were left at the desired temperaturefor 24 min after the addition of the cellulose and 1 min ofshaking. A bacterial culture of each species was also heated

at 100°C for 10 min. Its adhesion was then compared withthat of an untreated control culture.

(iii) Effects of carbohydrates. The effects of cellobiose,glucose, mannose, xylose, maltose, and soluble starch werestudied at concentrations of 0.1, 0.2, 0.5, 1, and 5%. Toeliminate any traces of cellobiose, which had been used asenergy source during culture, the bacteria were centrifugedat 3,500 x g for 10 min, washed in a medium of the samecomposition but without cellobiose or resazurin, centrifugedagain at the same speed, and then placed in the same mediumcontaining the test carbohydrate. A sugar-free control wasalso prepared.The percentage of adherent cells was also determined after

subculturing F. succinogenes S85 one, five, or ten times into4% glucose or cellobiose. The percentage of adhesion inthese different subcultures was compared with that of acontrol containing the same sugar at a concentration of0.3%.

(iv) Effects of cellulose derivatives. The effects of methyl-cellulose (MC) (2% viscosity, 400 cP; Sigma Chemical Co.,St. Louis, Mo.), carboxymethyl cellulose (CMC), and hy-droxyethylcellulose (E. Merck AG, Darmstadt, Federal Re-public of Germany) were studied at a concentration of 0.1%.The bacterial cultures were treated as above; the cells weresuspended after a second centrifugation in a sugar-freemedium of the same composition supplemented with MC,CMC, or hydroxyethylcellulose at the required concentra-tion. Controls without cellulose derivatives were prepared atthe same time.

(v) Effect of divalent cations Mg2' and Ca2+. F. succino-genes S85 and R. flavefaciens 007 were grown on Ml culturemedium (Table 1). For R. flavefaciens, an amino acidsolution containing isoleucine, leucine, and valine (100 mg ofeach per 100 ml of distilled water) was added at a concen-tration of 1% to satisfy its nutritional requirements (6). Aftercentrifugation, the pellet was resuspended in 9 ml of Mlmedium containing no resazurin or cellobiose and lackingCa2+, Mg2+, or both. The mineral composition of themedium was as follows (grams per 100 ml): KH2PO4, 0.6;(NH4)2SO4, 1.2; NaCl, 1.2; KCl, 0.6. In addition,MgSO4- 7H20 was added at a concentration of 0.12 g/100 mlto the medium lacking Ca2+ and CaCl2 2H20 was added ata concentration of 0.06 g/100 ml to the medium lackingMg2 . The bacterial cells were centrifuged two more timesand resuspended after each centrifugation in their respectivemedia. At the same time, controls containing Ca2+ or Mg2+were prepared.

(vi) Effect of pH. Adhesion of R. flavefaciens 007 and F.succinogenes S85 was studied in a pH range of 3.5 to 9 and4.5 to 7, respectively, at intervals of 0.5 unit. After centrif-ugation, R. flavefac iens and F. succinogenes were sus-pended in the medium of Scott and Dehority (31) and in Mlmedium, respectively, containing no cellobiose, no res-azurin, and no buffer (NaHCO3). The medium of Scott andDehority (31) was adjusted to the required pH with 3 MNaOH or with 1 M HCI. In Ml medium, a double concen-tration of NaCl and 8 ml of 3 M NaOH per liter were addedto compensate for the sodium lost by removal of the buffer.Then the required pH was adjusted with 3 M KOH or with 1M HCl. This was done since upon total sodium deprivationF. succinogenes did not adhere at all (see Results). Experi-ments were done under nitrogen to avoid the variations inpH caused by the dissolution of CO2 in the medium.

(vii) Effect of oxygen. After the addition of cellulose, thebacterial culture was left in contact with air for 5, 15, 30, or60 min.

3082 ROGER ET AL.

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ADHESION OF RUMINAL CELLULOLYTIC BACTERIA TO CELLULOSE 3083

(viii) Effect of redox potential. Adhesion of R. flavefaciens007 and F. succinogenes S85 was studied in media contain-ing a reducing agent under 100% CO2, without a reducingagent under 100% CO2, containing a reducing agent andprepared in aerobiosis, or without a reducing agent andprepared in aerobiosis.Three different reducing agents (cysteine hydrochloride,

dithiothreitol, and sodium sulfide) were used at a concentra-tion of 0.1% for R. flavefaciens 007. With F. succinogenesS85, cysteine hydrochloride was the only reducing agentstudied. Both species were incubated at 38°C for 25 min afterthe addition of cellulose. The redox potential was measuredwith an rH meter (CG 819; Schott Gerate, Hofheim, FederalRepublic of Germany); the electrode was immersed in thebacterial culture for 20 s.

(ix) Effect of sodium and gas phase. R. flavefaciens 007 wasgrown on Ml medium containing amino acids as detailedabove. Two media were tested: medium Ml and medium Ml- Na from which all sodium sources had been removed andthe pH adjusted to 7 with 3 M KOH. Both were tested undera gas phase of 100% N2 or 100% CO2.Adhesion of F. succinogenes S85 was studied on the

following media: medium Ml - Na from which all sodiumsources had been removed; medium Ml + NaHCO3 inwhich the sodium was added at a concentration of 50 mM inthe. form of NaHCO3; medium Ml + Na2CO3 in which thesodium was added at a concentration of 50 mM in the formof Na2CO3; medium Ml + NaCl in which the sodium wasadded at a concentration of 50 mM in the form of NaCl;medium Ml + KHCO3 in which no sodium ions were addedto the medium (the concentration of KHCO3 was calculatedso that the supply of HCO3- ions was the same as that in themedium Ml + NaHCO3); medium Ml + LiCl in which nosodium was added to the medium and lithium was added ata concentration of 50 mM.

All the media were prepared under 100% N2. The mediumwith LiCl and those without sodium were also preparedunder 100% CO2. When NaCl was removed from the mineralsolution, it was replaced by the same concentration of KCI.In all media, the pH was adjusted to 7 with 3 M KOH.We also studied the adhesion of the bacteria on medium

Ml under 100% N2 or 100% CO2. The latter was considereda control.

(x) Effect of metabolic inhibitors. We tested monensin andlasalocid (ionophore antibiotics) at a concentration of 0.02mM, 2,4-dinitrophenol and tetrachlorosalicylanilide (proton-ophores) at concentrations of 1.6 and 0.02 mM, respectively,N,N'-dicyclohexylcarbodiimide (DCCD) (inhibitor of mem-brane ATPases) at a concentration of 0.02 mM, antimycin Aand hydroxyquinoline-N-oxide at a concentration of 0.1mM, and sodium azide at a concentration of 40 mM (thelatter three are all inhibitors of electron transfer chains).These inhibitors were added to the bacterial culture and

left in contact with the bacteria for 10 min before the additionof cellulose. Sodium azide was dissolved in distilled water,hydroxyquinoline-N-oxide was dissolved in dimethyl sulfox-ide, and the other inhibitors were dissolved in ethanol.Control cultures containing ethanol or dimethyl sulfoxide atthe same concentrations as in the test tubes were made tocheck the eventual effects of these compounds. The finalconcentration of dimethyl sulfoxide or ethanol in the bacte-rial culture was less than 1 or 3%, respectively.

RESULTSF. succinogenes had the highest percentage of adherent

cells at the end of the exponential growth phase, while R.

8

Ea 40

* 20

1,6

Time (hours)

go

8

17 22Time (hours)

FIG. 1. Effect of the growth phase on the adhesion of F. succi-nogenes S85 (a) and R. flavefaciens 007 (b). Symbols: A, growth(A6w); *, adherent cells (%).

flavefaciens adhered in the early exponential phase. There-after, the adhesion of F. succinogenes fell abruptly, whereasthe high percentage of adherent ruminococci was maintaineduntil the late stationary phase (Fig. la and b).

R. flavefaciens 007 and 083 adhered rapidly to celluloseAvicel; after only 1 min of contact, 80 to 90% of the bacterialcells were adherent. For the study of other factors, thisspecies was maintained in contact with cellulose for 5 min.In contrast, the greatest number of adhering cells of F.succinogenes S85 was reached between 15 and 30 min ofcontact with cellulose; after 1 min of contact, only 30% ofthe cells adhered (Table 2). The effects of other factors onthis species were therefore studied with cultures harvestedat the end of the exponential growth phase and after 25 minof contact with cellulose.Adhesion of both species was maximal at 38°C. Adhesion

of F. succinogenes varied according to the temperature

TABLE 2. Effect of contact time between cellulose and bacteriaon percentage of cells adhering to cellulose Avicel

% of adherent cellsaContact time

(min) R. flavefaciens R. flavefaciens F. succinogenes007 083 S85

1 74 4 71 3 36 15 81 3 82 1 57 4

15 83 2 75 8 78 530 82 7 82 2 83 360 80 6 89 2 83 4

a Average of four replicates + standard deviation.

VOL. 56, 1990

11--k

I

Z.

I

9

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

3084 ROGER ET AL.

TABLE 3. Effect of temperature on adhesion of F. succinogenesS85 and R. flavefaciens 007 to cellulose Avicel

Temp % of adherent cellsa(OC) R. flavefaciens F. succinogenes

4 99±7 71±313 98 5 81± 220 98 7 84 ±+530 100 96 ± 342 100 98± 343 NDb 85 ± 644 ND 77 346 ND 83 648 ND 39 452 93 ± 3 ND' Percentage of adherent cells compared with a control tube at 38°C

(average of four replicates ± standard deviation).b ND, Not determined.

(Table 3), while that of R. flavefaciens was unaffected. Thepercentage of adhering cells of F. succinogenes decreasedmarkedly at 48°C. After exposure of the bacterial cuitures to100°C, the adhesion of F. succinogenes S85 and LRS095 wascompletely suppressed, while there was a decrease of only15% in attachment of R. flavefaciens.At low concentrations (0.1 to 1%), carbohydrates had no

effect on the adhesion of the two species. At a concentrationof 5%, the adhesion ofR. flavefaciens was slightly depressedwhile that of F. succinogenes was partly affected by xyloseand mannose and more by cellobiose and glucose (Table 4).There was a considerable decrease in the adhesion of F.succinogenes after it had been transferred several times oncellobiose or glucose at a high concentration (4%). Thenumber of adherent cells of F. succinogenes decreased by60, 67, and 80% after 1, 5, and 10 transfers, respectively, ona high concentration (4%) of cellobiose. On 4% glucose,whatever the number of transfers, there was only a 35%decrease. Hydroxyethylcellulose partly inhibited the adhe-sion of F. succinogenes and greatly inhibited that of R.flavefaciens, while MC only inhibited the adhesion of R.flavefaciens and CMC had no effect on either species (Table4).Only the attachment of R. flavefaciens was affected by a

lack of divalent cations (Table 4). Deprivation of either Ca2"or Mg2" had little effect, but the lack of both cationsinhibited adhesion to a greater extent (Table 4). The numberof adherent cells of F. succinogenes increased regularly frompH 4.5 to 6.0 (30% of adherent cells at pH 4.5, 80% at pH6.0), remained stable between pH 6.0 and 7.0, and fellabruptly at pH 7.5 (25% of adherent bacteria). The adhesionof R. flavefaciens was stable at pHs between 3.5 and 7.5,decreased at pH 8.0 (40% of adherent cells), and remainedthe same at pH 8.5 and 9.0. After 5 min of exposure to air,80% inhibition of adherent cells of F. succinogenes S85,LRS095, and LRS128 was observed. After 1 h, adhesion wascompletely inhibited. Adhesion of R. flavefaciens 007 was

little affected by the presence of oxygen, but strains 083 and131 were slightly inhibited: 25% after 5 min and 35% after 1h of exposure.The percentage of adherent bacteria of F. succinogenes

decreased with oxidation of the medium compared with thecontrol (medium with cysteine and under 100% CO2. -250mV). It decreased by 20% in medium without cysteine under100% CO2 (-150 mV), by 80% in medium with cysteineunder aerobic conditions (-50 mV), and by 90% in mediumwithout cysteine under aerobic conditions (+100 mV). Ad-

TABLE 4. Effect of sugars, cellulose derivatives, and lack ofdivalent cations on adhesion of F. succinogenes S85 and

R. flavefaciens 007 to cellulose Avicel

% of adherent cellsaFactor studied

R. flavefaciens F. succinogenes

Glucose (5%) 82 ± 4 64 ± 2Cellobiose (5%) 82 ± 6 54 ± 2Mannose (5%) 89 ± 1 71 ± 8Xylose (5%) 90 ± 1 77 ± 6Maltose (5%) 93 ± 2 100Soluble starch (5%) 90 ± 5 100

CMC (0.1%) 100 100HEC (0.1%) 21 3 70 ± 3MC (0.1%) 46 ± 4 100

Minus Ca2+ 91 ± 4 100Minus Mg2+ 87 ± 4 100Minus Ca2' and Mg2" 77 ± 5 100

a Percentage of adherent cells compared with a control culture containingCa2" and Mg2+ and without carbohydrates (average of four replicates +standard deviation).

hesion of R. flavefaciens was not greatly affected by redoxpotential (Table 5). The presence of cysteine or Na2S underaerobic conditions almost completely inhibited adhesion ofthe ruminococci (Table 5).Under a nitrogen atmosphere and in the absence of sodium

ions, F. succinogenes did not adhere to cellulose (20% ofadherent bacteria compared with control values). An atmo-sphere of CO2 or the presence of Na+ was required for theadhesion of this species. In its effect, Li' could substitutefor Na+ (Table 6). The adhesion of the bacterium was alsostrongly inhibited by DCCD and by certain ionophore anti-biotics (Table 7). At the same concentration, the effect oflasalocid was greater than that of monensin. The otherprotonophores and inhibitors of the electron transfer chainsdid not significantly alter the percentage of adherent bacte-ria.Adhesion of R. flavefaciens was little affected by the

absence of sodium ions in the medium (90% adherent cellscompared with the control) (Table 6) or by the action of thedifferent metabolic inhibitors tested (Table 7).

DISCUSSION

From the various quantitative techniques used for thedetermination of bacterial adhesion, we chose that describedby Minato and Suto (20) because it is simple, rapid, repro-ducible, and allows the study of many factors. While it istrue that with this method nonadherent bacterial cells maybe entrapped within cellulose during centrifugation (27), thisshould not be considered a major drawback. If large num-bers of cells were to be entrapped during centrifugation,strong inhibitory factors would never be found; yet weobserved that adhesion of R. flavefaciens in the presence ofcysteine in aerobic medium and that of F. succinogenes inthe presence of several factors (Na+ deprivation under anatmosphere of N2, presence of oxygen) were almost com-pletely inhibited. Furthermore, in each assay we comparedour results with those of controls lacking the factor the effectof which was to be studied.

R. flavefaciens adheres in less than 1 min to cellulose,which suggests that the adherence process involves a struc-ture that is an integral part of the bacterial cell. The adhesion

APPL. ENVIRON. MICROBIOL.

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ADHESION OF RUMINAL CELLULOLYTIC BACTERIA TO CELLULOSE 3085

TABLE 5. Effect of redox potential on adhesion of R. flavefaciens 007 to cellulose Avicel

Gas phase Reducing % of adherent cells with Eh (mV)aagent Cysteine hydrochloride Dithiothreitol Na2S

100% C02 Used 90 ± 7 (-210 ± 88) 94 + 4 (-290 + 23) 94 + 2 (-190 ± 24)Omitted 86 + 11 (-80 ± 38) 88 ± 8 (-180 ± 32) 95 ± 2 (-60 ± 46)

100% air Used 13 ± 10 (+40 ± 78) 60 ± 15 (-220 ± 45) 9 ± 9 (+120 ± 18)Omitted 58 ± 18 (+140 ± 59) 56 ± 15 (+210 ± 17) 57 ± 7 (+200 ± 24)

a Average of four replicates ± standard deviation.

of this bacterium, as shown for R. albus (25), occurs withinS min after the addition of the substrate. R. flavefaciens andR. albus appear therefore to be able to colonize plant celltissues as soon as they arrive in the rumen and possiblycolonize plant tissues faster than F. succinogenes. In an

environment in which there is strong competition from othermicroorganisms, this may represent an ecological advan-tage. Our findings are in line with those of Akin (3), whoobserved by electron microscopy a larger number of F.succinogenes adherent to plant cell walls during the laterstages of cellulolysis.The sudden drop in the percentage of adherent cells of F.

succinogenes at low and high temperatures is reminiscent ofan enzymatic type of bacterium-cellulose linkage. This isfurther suggested by the fact that adhesion reached itsmaximum at the end of the exponential phase, i.e., whenenzyme production is at its maximum.The carbohydrates studied, all constitutive elements of the

polyosides of the plant cell walls, had an inhibitory effect onadhesion only at a high concentration (5%), which indicatesthat they have a weak specificity for possible receptors of thebacterial wall. The inhibitory effect of glucose and cellobioseon F. succinogenes at high concentrations may be due to aretroactive inhibitory effect of these sugars on enzymaticactivity and strongly suggests that the cellulases of F.succinogenes are involved in the adherence process. After10 transfers on these sugars at 4%, we observed that thisspecies was still able to degrade filter paper. Furthermore,Taylor et al. (33) have recently shown that these sugars canrepress certain cellulase genes of the bacterium. As for R.

TABLE 6. Effect of sodium and gas phase on adhesion ofF. succinogenes S85 and R. flavefaciens 007 to cellulose Avicel

Gasa Ion supple- cells'Gasphasementb R. flavefaciens F. succinogenes

100% N2 Minus Na 91 5 19 7NaHCO3 NDd 84 6Na2CO3 ND 86 3NaCl ND 57 4KHCO3 ND 10 6Nae 89 5 84 ± 2LiC1f ND 48 6

100% C02 Minus Na 89 + 4 81 8LiC1f ND 91 4

a Media prepared under either 100% N2 or 100% CO2 gas phase.b Na+, Li', or HC03- supplied at 50 mM concentration.' Percentage of adherent cells compared with a control culture (average of

four replicates standard deviation).d ND, Not determined.eAll the sodium sources of medium Ml are present (NaCI, NaHCO3, and

NaOH).f No sodium added.

albus (24), soluble starch did not inhibit adhesion to cellu-lose, even at high concentrations, which suggests the spec-ificity of the binding with the substrate. Moreover, it hasbeen observed elsewhere (G. Fonty, I. Guerry, and P.Gouet, First Congress of the French Society for Microbiol-ogy, Toulouse, 1986) that R. flavefaciens adheres weakly ornot at all to several noncellulolytic substrates (Sephadex,Amberlite resin, starch granules).

Unlike Minato and Suto (20), Kudo et al. (16), andRasmussen et al. (27), we observed no inhibitory effect ofCMC. This controversial observation might be explained bythe fact that the inhibitory activity of CMC on cellulases wastransitory, as this compound was very inhibitory in immedi-ate assays but a portion of this inhibition was relieved bypreincubating CMC with the enzyme source (26). Under ourexperimental conditions, F. succinogenes was in contactwith cellulose Avicel for 25 min, and this might explain whyour results were different. They might also be due to thebacterial strains and the cellulose source used. Morris (24)recently observed that the adhesion of R. albus differedaccording to strains, while Wood et al. (35) have shown thatthe cellulases of this bacterium varied in size depending onthe energy source. However, our results on the effect of MCon R. flavefaciens are in agreement with those of Rasmussenet al. (27). We also observed an eluting effect of MC (V.Roger and G. Fonty, unpublished data) against this bacte-rium. Rasmussen et al. (26) and White et al. (34) haverecently shown that MC at a low concentration has an

TABLE 7. Effect of metabolic inhibitors on adhesion ofF. succinogenes S85 and R. flavefaciens 007 to cellulose Avicel

% of adherent cellsbInhibitor (mM)'

R. flavefaciens F. succinogenes

Metal ionophoresLasalocid (0.02) 82 ± 14 6 ± 4Monensin (0.02) 85 ± 6 70 ± 8

Proton ionophores2,4-dinitrophenol (1.6) 100 94 ± 3Tetrachlorosalicylanilide (0.02) 100 95 ± 4

Membrane ATPase inhibitor 94 ± 1 10 ± 6(DCCD [0.2])

Electron transport inhibitorsAntimycin A (0.1) 87 ± 9 88 ± 10Hydroxyquinoline-N-oxide (0.1) 90 ± 4 90 ± 4NaN3 (40) 100 100

a All factors are dissolved in ethanol except NaN3 and hydroxyquinoline-N-oxide, which were dissolved, respectively, in distilled water and dimethylsulfoxide.

b Percentage of adherent cells of a control culture (average of four repli-cates ± standard deviation).

VOL. 56, 1990

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

inhibitory effect on an exo-,B-1,4-glucanase (called exo A)and has surfactant properties at a high concentration. Theseobservations and our own results suggest that this enzymeacts in the adherence process and that hydrophobic interac-tions are involved in the adherence of R. flavefaciens.However, these statements have to be demonstrated.

Unlike F. succinogenes, R. flavefaciens requires the pres-ence of Ca2" and Mg2" ions for its adhesion. These cationsmight bind the large glycocalyx observed by electron micros-copy (8, 19) and cellulose. However, the fact that in theirabsence adhesion decreased by only 20 to 30% indicates thationic interactions are not the sole mechanism of adhesion ofthis bacterial species. The greatest number of adherent cellsof F. succinogenes observed at pHs between 5.5 and 7.0 isconsistent with an enzymatic-type adhesion process since atthose pHs the different cellulases of F. succinogenes exhib-ited the greatest activity. However, pH could have alsoaffected electrostatic interactions between the cell surfaceand cellulose. In agreement with Rasmussen et al. (27), we

observed that the adhesion of R. flavefaciens, like that of R.albus (24), was not affected by low pHs.The attachment of R. flavefaciens would seem to be

unaffected by cell metabolism, as evidenced by the highadhesion percentage obtained in the presence of metabolicinhibitors and in a culture medium heated at 100°C. On thecontrary, adhesion of F. succinogenes seems to require theintact metabolic functions of the bacteria. Indeed, the adhe-sion decreases in the absence of Na, which is necessary forgrowth and glucose uptake (12, 29), and is almost completelyinhibited by oxygen. This finding was confirmed by theobservation that heat-treated cultures (100°C for 10 min) donot adhere and is another argument in favor of the partici-pation of proteins in adhesion. The effect of redox potentialsupports this hypothesis; the more the medium is oxidized,and thus incompatible with the survival of the bacterium, thefewer bacteria there are adhering to cellulose. DCCD, whichalkylates the carboxyl groups, might act by inhibitingATPases or altering a protein involved in the adheringprocess. However, Gong and Forsberg (14) recently demon-strated that the carboxyl groups of a protein do not play an

important role in adhesion. Thus, DCCD seems to inhibitadhesion via an inhibition of ATPases. The decrease in thedegree of adhesion of F. succinogenes in the presence ofmonensin and lasalocid might be due to modifications ofionic gradients. The proton motive force of the bacterium,which might be an energetic source for different cellularfunctions, is reduced in the presence of monensin (10, 28,30). Our results suggest that only live cells of F. succino-genes are able to adhere to cellulose. However, this hypoth-esis is not in agreement with results of Gong and Forsberg(14), who observed that Formalin-treated cells still adhere.The absence of effects of protonophores and inhibitors ofelectron transfer chains, used at concentrations that inhibitthe growth of F. succinogenes (12), remains unexplained.

Contrary to what might be expected for a strictly anaero-

bic bacterium, attachment of R. flavefaciens remains high inthe presence of oxygen or in a medium prepared underaerobic conditions in the absence of reducing agent. The

almost total inhibition of adhesion in the presence of cyste-

ine in a medium prepared in aerobiosis might be due to the

fact that under these conditions, cysteine is transformed intocystine, a highly hydrophobic amino acid, which couldtherefore compete with the hydrophobic interactions thatpossibly exist between the bacterium and its specific sub-strate. The absence of an effect of dithiothreitol in aerobiosiscan be explained by the low redox potential still reached.

The effect of a reducing agent might also be explained by thepresence of the SH groups.The adherence process seems to be different for R. flave-

faciens and F. succinogenes. The adhesion of the latter isaffected by low temperatures, the presence of oxygen, thelack of sodium, and the inhibition of its membrane ATPases.These results suggest that proteins and bacterial cellulasesthemselves play a part in the adhesion of F. succinogenes.The proteins responsible for adhesion should now be iso-lated and purified. Several mechanisms appear to be in-volved in the adhesion of R.flavefaciens, the main one beingthe interaction between the bacterial cell and the divalentcations Ca2" and Mg2+. Hydrophobic interactions and en-zymes may also be involved.

LITERATURE CITED

1. Akin, D. E. 1976. Ultrastructure of rumen bacterial attachmentto forage cell walls. Appl. Environ. Microbiol. 31:562-568.

2. Akin, D. E. 1979. Microscopic evaluation of forage digestion byrumen microorganisms-a review. J. Anim. Sci. 48:701-710.

3. Akin, D. E. 1980. Evaluation by electron microscopy andanaerobic culture of types of rumen bacteria associated withdigestion of forage cell walls. Appl. Environ. Microbiol. 39:242-252.

4. Akin, D. E., and F. E. Barton II. 1983. Rumen microbialattachment and degradation of plant cell walls. Fed. Proc.42:114-121.

5. Bryant, M. P., and L. A. Burkey. 1953. Cultural methods andsome characteristics of some of the more numerous groups ofbacteria in the bovine rumen. J. Dairy Sci. 36:205-217.

6. Bryant, M. P., and I. M. Robinson. 1962. Some nutritionalcharacteristics of predominant cultural ruminal bacteria. J.Bacteriol. 84:605-614.

7. Bryant, M. P., I. M. Robinson, and H. Chu. 1959. Observationson the nutrition of Bacteroides succinogenes-a ruminal bacte-rium. J. Dairy Sci. 42:1831-1847.

8. Cheng, K. J., C. S. Stewart, D. Dinsdale, and J. W. Costerton.1983/1984. Electron microscopy of bacteria involved in thedigestion of plant cell walls. Anim. Feed Sci. Technol. 10:93-120.

9. Dinsdale, D., E. J. Morris, and J. S. D. Bacon. 1978. Electronmicroscopy of the microbial populations present and theirmodes of attack on various cellulosic substrates undergoingdigestion in the sheep rumen. Appl. Environ. Microbiol. 36:160-168.

10. Forano, E. 1988. Effets de la monensine sur les gradientstransmembranaires de H', Na+ et K' chez Bacteroides succi-nogenes, bacterie cellulolytique du rumen. Reprod. Nutr. Dev.28(Suppl. 1):81-82.

11. Forsberg, C. W., T. J. Beveridge, and A. Hellstrom. 1981.Cellulase and xylanase release from Bacteroides succinogenesand its importance in the rumen environment. Appl. Environ.Microbiol. 42:886-896.

12. Franklund, C. V., and T. L. Glass. 1987. Glucose uptake by thecellulolytic ruminal anaerobe Bacteroides succinogenes. J. Bac-teriol. 169:500-506.

13. Gaudet, G., and B. Gaillard. 1987. Vesicle formation andcellulose degradation in Bacteroides succinogenes cultures:ultrastructural aspects. Arch. Microbiol. 148:150-154.

14. Gong, J., and C. W. Forsberg. 1989. Factors affecting adhesionof Fibrobacter succinogenes subsp. succinogenes S85 and ad-herence-defective mutants to cellulose. Appl. Environ. Micro-biol. 55:3039-3044.

15. Groleau, D., and C. W. Forsberg. 1981. Cellulolytic activity ofthe rumen bacterium Bacteroides succinogenes. Can. J. Micro-biol. 27:517-530.

16. Kudo, H., K. J. Cheng, and J. W. Costerton. 1987. Electronmicroscopic study of the methylcellulose-mediated detachmentof cellulolytic rumen bacteria from cellulose fibers. Can. J.Microbiol. 33:267-272.

17. Lamed, R., J. Naimark, E. Morgenstern, and E. A. Bayer. 1987.

3086 ROGER ET AL.

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ADHESION OF RUMINAL CELLULOLYTIC BACTERIA TO CELLULOSE 3087

Specialized cell surface structures in cellulolytic bacteria. J.Bacteriol. 169:3792-3800.

18. Latham, M. J. 1980. Adhesion of rumen bacteria to plant cellwalls, p. 339-350. In R. C. W. Berkeley, J. M. Lynch, J.Melling, P. R. Rutter, and B. Vincent (ed.), Microbial adhesionto surfaces. Ellis Horwood Limited Publishers, Chichester,United Kingdom.

19. Latham, M. J., B. E. Brooker, G. L. Pettipher, and P. J. Harris.1978. Ruminococcus flavefaciens cell coat and adhesion tocotton cellulose and to cell walls in leaves of perennial ryegrass(Lolium perenne). AppI. Environ. Microbiol. 35:156-165.

20. Minato, H., and T. Suto. 1978. Technique for fractionation ofbacteria in rumen microbial ecosystem. It. Attachment ofbacteria isolated from bovine rumen to cellulose powder in vitroand elution of bacteria attached to it. J. Gen. Appl. Microbiol.16:1-16.

21. Minato, H., and T. Suto. 1981. Technique for fractionation ofbacteria in rumen microbial ecosystem. IV. Attachment ofrumen bacteria to cellulose powder and elution of bacteriaattached to it. J. Gen. Appl. Microbiol. 27:21-31.

22. Miron, J., M. T. Yokoyama, and R. Lamed. 1989. Bacterial cellsurface structures involved in lucerne cell wall degradation bypure cultures of cellulolytic rumen bacteria. Appl. Microbiol.Biotechnol. 32:218-222.

23. Montgomery, L., B. Flesher, and D. Stahl. 1988. Transfer ofBacteroides succinogenes (Hungate) to Fibrobaccter gen. nov.as Fibrobacter succinogenes comb. nov. and description ofFibrobacter intestinalis sp. nov. Int. J. Syst. Bacteriol. 38:430-435.

24. Morris, E. J. 1988. Characteristics of the adhesion of R. albuis tocellulose. FEMS Microbiol. Lett. 51:113-118.

25. Morris, E. J., and 0. J. Cole. 1987. Relationship betweencellulolytic activity and adhesion to cellulose in R. albus. J.

Gen. Microbiol. 133:1023-1032.26. Rasmussen, M. A., R. B. Hespell, B. A. White, and R. J.

Bathast. 1988. Inhibitory effects of methylcellulose on cellulosedegradation by Ruminococcus flav'efaciens. Appl. Environ.Microbiol. 54:890-897.

27. Rasmussen, M. A., B. A. White, and R. B. Hespell. 1989.Improved assay for quantitating adherence of ruminal bacteriato cellulose. Appl. Environ. Microbiol. 55:2089-2091.

28. Russell, J. B. 1987. A proposed mechanism of monensin actionin inhibiting ruminal bacterial growth: effects on ion flux andprotonmotive force. J. Anim. Sci. 64:1519-1525.

29. Russell, J. B. 1987. Effect of extracellular pH on growth andproton motive force of Bacteroides suc cinogenes, a cellulolyticruminal bacterium. Appl. Environ. Microbiol. 53:2379-2383.

30. Russell, J. B., and H. J. Strobel. 1989. Effect of ionophores onruminal fermentation. Appl. Environ. Microbiol. 55:1-6.

31. Scott, H. W., and B. A. Dehority. 1965. Vitamin requirements ofseveral cellulolytic bacteria. J. Bacteriol. 89:1169-1175.

32. Stack, R. J., and R. E. Hungate. 1984. Effect of 3-phenylpropanoic acid on capsule and cellulases of Rumninococcusalbus 8. Appl. Environ. Microbiol. 48:218-223.

33. Taylor, K. A., B. Crosby, M. McGavin, C. W. Forsberg, andD. Y. Thomas. 1987. Characteristics of the endoglucanase en-coded by a cel gene from Bacteroides succinogenes expressedin Escheric/hia coli. Appl. Environ. Microbiol. 53:41-46.

34. White, B. A., M. A. Rasmussen, and R. M. Gardner. 1988.Methylcellulose inhibition of exo-,-1,4-glucanase A from Rumi-nococcusflavefaciens FD1. Appl. Environ. Microbiol. 54:1634-1636.

35. Wood, T. M., C. A. Wilson, and C. S. Stewart. 1982. Prepara-tion of the cellulase from the cellulolytic anaerobic rumenbacterium Rumninococcus albus and its release from the bacte-rial cell wall. Biochem. J. 205:129-137.

VOL. 56, 1990

on February 10, 2020 by guest

http://aem.asm

.org/D

ownloaded from