APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1976, p. 562-568Copyright © 1976 American Society for Microbiology

Vol. 31, No. 4Printed in U.S.A.

Ultrastructure of Rumen Bacterial Attachment to ForageCell WallsDANNY E. AKIN

Field Crop Utilization and Marketing Research Laboratory, Richard B. Russell Agricultural ResearchCenter, Agricultural Research Service, Athens, Georgia 30604

Received for publication 31 December 1975

The degradation of forage cell walls by rumen bacteria was investigated withcritical-point drying/scanning electron microscopy and ruthenium red staining/transmission electron microscopy. Differences were observed in the manner ofattachment of different morphological types ofrumen bacteria to plant cell wallsduring degradation. Cocci, constituting about 22% of the attached bacteria,appeared to be attached to degraded plant walls via capsule-like substancesaveraging 58 nm in width (range, 21 to 84 nm). Many bacilli appeared to adhereto forage substrates without distinct capsule-like material, although unattachedbacteria with capsules were observed occasionally. Certain bacilli appeared to beattached to degraded tissue via small amounts of extracellular material, butothers apparently had no extracellular material. Bacilli with a distinct morphol-ogy due to an irregularly folded, electron-dense outer layer or layers (about 15nm thick) and without fibrous extracellular material constituted about 37% ofthe attached bacteria and were observed to adhere so closely to degraded plantwalls that the bacterial shape conformed to the shape of the degraded zone. Inthe rumen ecosystem, bacteria appeared to adhere to plant substrates duringdegradation by capsule-like material and by small amounts of extracellularmaterial, as well as by other means not observable by electron microscopy.

Capsular and slime materials (21) have beenshown to be involved in the association of bacte-ria with inert substances (13) and to cellulosicsubstrates (4, 19). The technique of rutheniumred staining (15) has been useful to show thefibrillar nature of capsule-like substances (13,18, 20). Recently, the technique of critical-pointdrying (3) of bacteria, used to study extracellu-lar material, showed a fibrous nature to thecapsule-like polymer surrounding bacteria (5).

Investigations of the rumen microfloral deg-radation of forage tissue have revealed thepresence of capsule-like material adjoininglarge cocci to plant walls (1, 2). Additionally,bacteria from pure cultures of Ruminococcusalbus were shown to attach to cellulose by fi-brillar structures external to the bacterial cellwall (19). Research from this laboratory indi-cated that differences exist in the mode of at-tachment of rumen bacteria to forage tissue (2).The objectives of this work were to (i) investi-

gate further the ultrastructural nature of ru-men bacterial attachment to forage tissue, and(ii) delineate the manner of attachment of dif-ferent morphological types of rumen bacteria toforage tissue, by means of critical-point drying/scanning electron microscopy and rutheniumred staining/transmission electron microscopy.

MATERIALS ANI) METHOI)SInoculum. For in vitro studies, rumen digesta

obtained from a cannulated steer was strainedthrough cheesecloth into a previously warmed vac-uum bottle for transportation to the laboratory. Therumen fluid was strained again through cheeseclothand mixed 1:2 with McDougall carbonate buffer (17)previously bubbled with CO2 at 39 C.

For in vivo studies, rumen bacteria that wereassociated with partially degraded leaf blades wereremoved from the fistulated steer and studied insitu.

Substrate. For in vivo studies, substrates con-sisted of leaf blades of Coastal bermudagrass (Cyno-don dactylon I L.l Pers.) frozen immediately afterharvest at 4 weeks of age and maintained at -30 C.Leaf blades were cut into sections 2 to 5 mm longwith a new, cleaned razor blade to reduce dragacross the tissue cross section. These leaf sectionswere then inoculated with the rumen bacteria-buffer inoculum at 39 C with continuous CO2 bub-bling.

For in vivo studies, leaf blades from the partiallydigested hay that was predominately Coastal ber-mudagrass were examined.

Scanning electron microscopy. Leaf blade sec-tions digested in vitro or in vivo were fixed in 4%glutaraldehyde (in 0.1 M cacodylate buffer, pH 7.0 to7.2) for 16 to 24 h and postfixed in 1.5% bufferedosmium tetroxide for about 4 h at 4 C. Samples weredehydrated in a series of ethanol-water washes (25%

562

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

ULTRASTRUCTURE OF RUMEN BACTERIAL ATTACHMENT 563

through 100%; vol/vol) and critical-point dried inliquid CO2 without transitional solvents. Sampleswere then coated for conductive purposes with gold-palladium (60-40) alloy wire and observed with thescanning electron microscope (SEM) at about 15 kV.

Transmission electron microscopy. A pretreat-ment of 0.1 M cacodylate buffer with 0.15% (wt/vol)ruthenium red for 30 min at room temperature wasused for one group of samples of degraded leaves; allsamples were then prepared for the transmissionelectron microscope (TEM) as described previously(2) with the inclusion of 0.15% (wt/vol) rutheniumred in both fixatives according to the formula usedby Cagle et al. (6). Other samples were preparedwithout ruthenium red staining for comparativepurposes.

Sizes of capsules and external layers of the bacte-ria were determined from enlargements of thin sec-tions with a particle-size analyzer. Capsular widthsfor cocci were determined by averaging the threeunattached sides of 23 randomly chosen bacteria.The percentages of morphological types of attachedbacteria were determined from bacteria associatedwith degraded zones in forage walls observed in 15micrographs.

RESULTSMost interpretations were made from in vi-

tro-digested leaf blade sections. However, thesame morphological types of bacteria werefound in both in vitro- and in vivo-digestedspecimens, and the differences in the mode ofattachment by these morphological types weresimilar in both types of digestion.

Ultrastructure shown by the SEM. Manybacteria, particularly the cocci, appeared to beattached by extracellular, fibrous structures toplant walls undergoing degradation (Fig. 1[F]). That many of these bacteria were activelyinvolved with plant tissue degradation was in-dicated by degraded zones surrounding the bac-teria (Fig. 1, inset). However, rumen bacteria,particularly bacilli, without the extracellular,fibrous structures were often observed in de-graded zones of plant walls (Fig. 2 IB]). Theseobservations suggested that some bacilli in-volved with the attachment and degradation ofplant tissues have no extracellular material orinsufficient amounts to be observed with theSEM using critical-point-dried samples.

Ultrastructure shown by the TEM. In thinsections of ruthenium red-stained, degradedleaf sections, numerous bacteria were attachedto the plant wall by a capsule-like material(Fig. 3, 4, 5). Often, the width of the side of thecapsule attached to the plant material waslonger than the unattached side, indicating astretching of the extracellular material (Fig. 4,insets).Many round-appearing bacteria were at-

tached to cell walls. Cocci were distinguishedfrom bacilli that had been sectioned perpendic-

ularly to the long axis of the cell by similarityin morphology and electron denseness to cap-sules and cell walls of known cocci (as revealedby round bacteria dividing by binary fission,Fig. 3 [C]). No bacilli that resembled thesedividing cocci in capsular and cell wall mor-phology were seen in thin sections. Althoughmost cocci-degrading plant walls appeared tobe attached via capsules, the capsules varied insize (Fig. 5 [C]), averaging 58 nm wide andranging from 21 to 84 nm. These encapsulatedcocci constituted about 22% of the attached bac-teria.

Distinct capsules were not observed with ba-cilli that were degrading plant material, andextracellular material was present only insmall amounts (Fig. 4) or absent (Fig. 6 [B])between bacteria and plant walls. On occasion,large capsules were observed surrounding ba-cilli, but these bacteria were not attached toplant walls (Fig. 6, arrow).

Bacilli with a distinctive morphology ap-peared to be attached without external fibrousmaterial to plant walls (Fig. 5, 7, 8 [B]). Thesebacteria had an electron-dense, irregularlyfolded outer layer (or layers) about 15 nm thick,which was separated from the plasma mem-brane by an electron-transparent layer (Fig. 7,8). The electron-dense layer was present inspecimens not treated with ruthenium red, butfixed with OsO and stained with uranyl ace-tate and lead citrate; however, the layer usu-ally appeared to be more intensely stained inruthenium red-stained specimens. These bacilli(with dimensions of 1 to 2 by 0.3 to 0.5 Am forunattached cells) were closely associated withdegraded regions, and the bacterial shape wasmodified to conform with the shape of the de-graded zones (Fig. 5 IB], 8). This type of mi-crobe is numerous in degraded plant wall re-gions and constituted about 37% of the bacteriaattached to plant cell walls. Diffuse, fibrillarmaterial external to the irregularly foldedouter layer of an unattached bacterium wasfound in one sample of in vivo-degraded tissue(Fig. 7, inset), but usually the extracellularcoat material was negligible or perhaps con-tained in the thin electron-dense layer. Exami-nation at high magnification revealed a gram-negative-type cell wall morphology (Fig. 8, in-set) with, in one case, a small layer, not denselystained, external to the outer membrane (Fig.8, arrow). Capsule-like material was not appar-ent between bacterial and plant cell walls.

I)ISCUSSIONHungate (12) reported that the cellulolytic

ruminococci possessed "abundant" capsules.Akin et al. (2) reported that large cocci from

VOL. 31, 1976

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

FIG. 1. Scanning electron micrograph of critical-point-dried leaf digested in vitro with rumen bacteria.Bacteria appear to be attached to plant cell walls via fibrous structures (F) typical of critical-point-driedextracellular material. A coccus (inset) with fibrous structures (arrow) adjoining bacterial and plant wall liesin a degraded zone. x8,000; inset, x13,500.

FIG. 2. Scanning electron micrograph of critical-point-dried leaf digested in vitro with rumen bacteria.Some bacilli (B) that lack extracellular material are within depressions of the plant wall. Two cocci appearto be joined by extracellular material (arrow). x15,000.

564

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

ULTRASTRUCTURE OF RUMEN BACTERIAL ATTACHMENT 565

mixed bacterial populations of the rumen wereattached to plant structures by a capsule-likematerial. Additionally, Patterson et al. (19),using critical-point drying and the TEM,showed that extracellular fibrils from cells of apure culture of R. albus were involved withattachment of the bacteria to cellulose. Leath-erwood (14), using the SEM for cultures grownon cellulose, reported the presence of "tube-like" appendages on R. albus.Specimens observed in this study by critical-

point drying/SEM and ruthenium red staining/TEM revealed a fibrous nature to the capsule-like material that was apparently involved inthe adhesion of cocci to plant walls during deg-radation. The average capsule width of 58 nmmay be an underestimation due to shrinkageduring TEM preparation of the highly hydrated(21) capsule.However, many bacilli lacking capsule-like,

fibrous material in the critical-point-driedpreparations appeared to be within depressionsin the plant wall. That many of the bacillilacked or possessed small amounts of extracel-lular material was confirmed by the TEM.Cheng and Costerton (7), using ruthenium redstaining, described 10 different morphologicaltypes of external layers in bacteria (not at-tached to a substrate) from a mixed rumenmicroflora. Although some layers were small,these authors (7) reported that all bacteria ex-amined had at least some external coat mate-rial. Conversely, Akin et al. (2) reported that inthin sections not stained with ruthenium red,some bacilli were attached to plant walls with-

out apparent extracellular material; this wasconfirmed by the observations reported herein.Apparently, an adhesive layer not detectedwith the methodology used herein or other at-tractive forces (16) enabled certain rumen ba-cilli to adhere to the plant substrate duringdegradation.

Bacilli with a distinct morphology due to anelectron-dense, irregularly folded outer layer(or layers) were seen closely attached to plantwalls undergoing degradation (Fig. 8). Thesebacteria possessed a gram-negative-type cellwall structure (10) without external, fibrouscoat material in most cases. Costerton et al. (9)showed that Bacteroides succinogenes, a preva-lent cellulolytic species in the rumen (12), inpure culture possessed an irregular outer mem-brane with small and irregular amounts of coatmaterial similar to the bacilli mentioned above.These authors (9) concluded that the irregularfolding, observed with two other bacteria, withfoldings of different extents among species, wasan artifact of fixation or embedding based onfreeze-etch data. The bacilli reported herein tohave the irregular layer(s) appeared to be of acertain dimension (1 to 2 by 0.3 to 0.5 Am),whereas bacteria of other dimensions andshapes lacked this morphology. The shapes ofthese bacteria appeared at times to be modifiedby close contact with substrate or other bacte-ria. The irregular shape of the bacilli in contactwith plant cell walls is an unexplained phe-nomenon; it is not known if the thinness of thepeptidoglycan layer, which has been reportedfor B. succinogenes (10), could be responsible. It

FIG. 3. Thin section ofruthenium red-stained leafdegraded in vitro by rumen bacteria. Cocci (as shown bycells dividing by binary fission IC]) are attached to the plant wall by capsule-like material (arrows). x20,OOO.

FIG. 4. Enlargment of Fig. 3. A bacillus (B) appears to adhere to the plant wall by a small amount ofextracellular material, whereas the coccus (C) is attached via a structure resembling capsule-like material.Stretching of the extracellular material is indicated in a coccus shown by the TEM (lower inset) and in abacillus shown by the SEM (upper inset) (arrows). x40,OOO; lower inset, x25,500; upper inset, x12,000.

FIG. 5. Thin section of ruthenium red-stained leaf degraded in vitro by rumen bacteria. Cocci (C) andbacilli (B) with an irregular shape appear to be attached to the plant wall. The extracellular material aroundthe cocci varies in width. The bacilli possess an electron-dense outer layer but lack capsule-like material; theirshape is altered when in contact with plant material (arrows). x18,000.

FIG. 6. Thin section of ruthenium red-stained leaf degraded in vitro by rumen bacteria. A bacillus (B) inclose proximity to an epidermal cell lacks any evidence ofextracellular material but is degrading the cell wallmaterial. The inset shows a bacillus with a diffuse fibrous capsule (arrow) from an in vivo-degraded section;this microbe is not attached to plant material. x46,000; inset, x28,900.

FIG. 7. Thin section of ruthenium red-stained leaf degraded in vitro by rumen bacteria. Bacteria (B) withelectron-dense, irregularly folded outer layers are attached closely to plant material; the bacteria appear topossess a pliable cell wall as shown by the modification in cell wall close to other bacteria (arrow). Noapparent fibrous extracellular material is visible. The inset shows a morphologically similar, but unattachedbacterium from the in vivo degradation of leaf tissue showing a diffuse fibrous capsule-like material (inset,arrow). x32,000; inset, x28,900.

FIG. 8. Thin section of ruthenium red-stained leaf degraded in vitro by rumen bacteria. The bacteriumwith irregular folding of the outer layer is attached without apparent fibrous extracellular material, althougha thin layer is seen on one side (arrow). The attached side of the bacterium is very irregular, with portions ofthe bacillus (B) extending into the plant wall (P). The inset reveals the double membrane (arrows) of the cellwall structure characteristic ofgram-negative bacteria. x80,OOO; inset, x 150,000.

VOL. 31, 1976

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

FIGS. 3-5.

566

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

ULTRASTRUCTURE OF RUMEN BACTERIAL ATTACHMENT 567

t

i 46.\ .'--;7A~~~~.l : 8AI- .t :

FIGS. 6-8.

VOL. 31, 1976

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

is doubtful that these irregular shapes are arti-facts ofTEM preparation since, at times, exten-sions of the bacterial cells seem to be within thedegraded zones of rigid plant walls.

Factors other than capsules or slime havebeen reported to be responsible for bacterialattachment to substances (8, 16). Certain of thebacilli described herein appeared to have athin, irregular, electron-dense layer that mightbe responsible for attachment like the "primaryacid polysaccharide" reported for initial adhe-sion of a marine bacterium to solid surfaces(11). Other bacilli appeared to be in close prox-imity to plant walls undergoing degradationwithout noticeable extracellular material de-tectable by the TEM. However, that the fixa-tion, dehydration, and embedding steps couldhave obscured particular extracellular sub-stances of these bacteria cannot be completelyruled out. The observations reported herein re-veal differences in the manner of attachment ofvarious morphological types of rumen bacteriato forage material during degradation.

ACKNOWLEDGMENTSI thank Billy D. Nelson, Southeast Louisiana Experi-

ment Station, Franklinton, La., for the forage samples andHenry E. Amos, Field Crops Laboratory, Athens, Ga., forassistance in collecting and preparing the rumen fluid inoc-ulum.

LITERATURE CITED

1. Akin, D. E., and H. E. Amos. 1975. Rumen bacterialdegradation of forage cell walls investigated by elec-tron microscopy. Appl. Microbiol. 29:692-701.

2. Akin, D. E., D. Burdick, and G. E. Michaels. 1974.Rumen bacterial interrelationships with plant tissueduring degradation revealed by transmission electronmicroscopy. Appl. Microbiol. 27:1149-1156.

3. Anderson, T. F. 1950-51. Techniques for the preserva-tion of three-dimensional structure in preparing spec-imens for the electron microscope. Trans. N.Y. Acad.Sci. 13:130-134.

4. Berg, B., B. V. Hofstein, and G. Pettersson. 1972.Electron microscopic observations on the degradationof cellulose fibres by Celluibrio fulvus and Sporocyto-phaga myxococcoides. J. Appl. Bacteriol. 35:215-219.

5. Cagle, G. D. 1974. Critical point drying: rapid method

for the determination of bacterial extracellular poly-mer and surface structures. Appl. Microbiol. 28:312-316.

6. Cagle, G. D., R. M. Pfister, and G. R. Vela. 1972.Improved staining of extracellular polymer for elec-tron microscopy: examination of Azotobacter, Zoog-loea, Leuconostoc, and Bacillus. Appl. Microbiol.24:477-487.

7. Cheng, K. J., and J. W. Costerton. 1975. Ultrastructureof cell envelopes of bacteria of the bovine rumen.Appl. Microbiol. 29:841-849.

8. Corpe, W. A. 1970. Attachment of marine bacteria tosolid surfaces, p. 73-87. In R. S. Manly (ed.), Adhe-sion in biological systems. Academic Press Inc., NewYork.

9. Costerton, J. W., H. N. Damgaard, and K. J. Cheng.1974. Cell envelope morphology of rumen bacteria. J.Bacteriol. 118:1132-1143.

10. Costerton, J. W., J. M. Ingram, and K. J. Cheng. 1974.Structure and function of the cell envelope of gram-negative bacteria. Bacteriol. Rev. 38:87-110.

11. Fletcher, M., and G. D. Floodgate. 1973. An electron-microscopic demonstration of an acidic polysaccha-ride involved in the adhesion of a marine bacteriumto solid surfaces. J. Gen. Microbiol. 74:325-334.

12. Hungate, R. E. 1966. The rumen bacteria, p. 8-90. In R.E. Hungate (ed.), The rumen and its microbes. Aca-demic Press Inc., New York.

13. Jones, H. C., I. L. Roth, and W. M. Sanders III. 1969.Electron microscopic study of a slime layer. J. Bacte-riol. 99:316-325.

14. Leatherwood, J. M. 1973. Cellulose degradation byRuminococcus. Fed. Proc. 32:1814-1818.

15. Luft, J. H. 1971. Ruthenium red and violet. I. Chemis-try, purification, methods of use for electron micros-copy and mechanism of action. Anat. Rec. 171:347-368.

16. Marshall, K. C., R. Stout, and R. Mitchell. 1971. Mech-anism of the initial events in the sorption of marinebacteria to surfaces. J. Gen. Microbiol. 68:337-348.

17. McDougall, E. I. 1948. Studies on ruminant saliva. I.The composition and output of sheep's saliva. Bio-chem J. 43:99-109.

18. Pate, J. L., and E. J. Ordal. 1967. The fine structure ofChondrococcus columnaris. III. The surface layers ofChondrococcus columnaris. J. Cell Biol. 35:37-51.

19. Patterson, H., R. Irvin, J. W. Costerton, and K. J.Cheng. 1975. Ultrastructure and adhesion propertiesof Ruminococcus albus. J. Bacteriol. 122:278-287.

20. Springer, E. L., and I. L. Roth. 1973. The ultrastruc-ture of the capsules of Diplococcus pneumoniae andKlebsiella pneumoniae stained with ruthenium red.J. Gen. Microbiol. 74:21-31.

21. Wilkinson, J. F. 1958. The extracellular polysaccha-rides of bacteria. Bacterial. Rev. 22:46-73.

568 AKIN

on Decem

ber 20, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from