microorganisms in the rumen of cattle fed a constant ration
TRANSCRIPT
MICROORGANISMS IN THE RUMEN OF CATTLE FED A CONSTANT RATION1
Abstract Direct microscopic and c ~ ~ l t u r e coi~nts were made on samples of rumen contents
obtained by stomach tube from each of 25 cows a t monthly intervals during 3 months. The ration \\.as timothy hay plus various amounts of cottonseecl 111ea1, rolled barley, and salt. No correlation between different rations and differences in the microbiota c o ~ ~ l d be detected. Slreplococczls boois was identified by its rapid growth in the feed nledii~nl and sho\\red counts ranging between 2 X 10j and 1 .4X108. Most animals showed a S. bovis count between 1 and 20 million with an average of about 10 million per milliliter. The characteristics of some rumen strains are described. The cellulose-digesting bacteria were referable to four chief groups and the numbers of each were recorded. The cellulolytic cocci varied in alrnost every character studied. Two types were distinguished as representing the greatest divergence. R ~ ~ ~ n i ~ z o c o c c l ~ s albz~s n. sp. differs from R. juvefacims in being Gram-negative, forming little or no yellow pigment, and producing no succinic acid. A cellulose-digesting sporeformer, Clostridiz~nt Ioclzlzeadii n. sp., was found in many of the sanlples. I t was estre~nely active in digesting cellulose, exceeding in this respect the species of anaerobic ccll~~lose digesters previoirsly isolated. Spores were fornled in abun- dance but rapidly disintegrated and many strains were lost before subculture. Closlridil~nz l o ~ z g i s p o r ~ ~ ~ ~ z 11. sp. was encoi~~ltered cluring the investigation and is dcscribcd.
" In dell giinstigsterl Falle~l wurde schon innerhalb eilles Tages (bei 38°C.) Icraftige Zersetzung (of cellulose) erreicht. Die Resultate stimmen also hier mit jenem iiberein, \vie sie speziell im Verdauungslcanal der Wiederkauer beobachtet \vurden, wie den11 auch das Anhaufungsvei-fahren selbst sic11 den dort gegebenen Verhlilt~lisse~l moglichst eng anschlieszt." (Lohnis and Lochhead 1913 (12)).
Dr. Grant Lochhead was not only early concernecl with methods for demonstrating cellulolytic bacteria, he also appreciated the uniquelless of the rumen habitat ancl its pre-eminent fitness for rapid cligestion of cellulose. For this reason and in recog~lition of Dr. Lochhead's lifelong interest in microorga~lis~ns as they esist in nature, the author talces great pleasure in contributing i l l his honor an account of some of the characteristics of certain ~nicroorganisrns from the rumen.
During the summer of 1951, the Departlnent of Animal Husbandry a t the State College of Washington tested the effects of socli~~m chloride used to limit the feed collsumed (4). During 4 months 25 pregnant range cows were fed eight different combiuations of chopped timothy hay, cottonseed meal, rolled barley, ancl salt. They were available for a s t ~ l d y of the rumen microorganisn~s.
'Manuscript received October 4, 1956. Contribution from the Department of Bacteriology, University of California, Davis,
California, U.S.X. This investigation was supported in part by a research grant, E-899, from the National Institutes of Health. This work \\,as completed a t The State College of \Vashington, I'ullman, \Vashingto~~, U.S.A.
Can. J. Mlcroblol. 3 (1957)
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
290 CANADIAN JOURNAL O F MICROBIOLOGY. VOL. 3. 1957
Methods
The rations fed are shown in Table I. Rumen contents were sampled a t four different periods, June 15, July 5
and 6, August 3 and 4, and September 7 and 8. Each sample was removed through a stomach tube and, after i t was mixed, 10 ml. was added to 90 ml. of salt solution in equilibrium with 5yo carbon dioxide in a screw-capped bottle. The dilution fluid had the following composition: NaCl O.6yO, I<H?P04 O.lyo, NIgS04 O.Olr r /o , CCa12 0.01%. After sterilization, 0 . 06y0 Na2C03 was added and the solution equilibrated with 95% nitrogen - 5y0 carbon dioxide. As soon a s the samples were returned to the laboratory, air was again displaced with the Na-C02 mixture and the diluted materials were left undisturbed until just prior to inoculation when each sample was shaken vigorously for 30 seconds on a shalter having 600 oscillations per minute with an excursion from center of 8 mm. Further dilutions were made in the same diluent and inoculated into rumen fluid cellulose agar and feed extract agar series. The inoculated tubes were rolled as they solidified and were incubated a t 39' C.
The rumen fluid cellulose agar medium c o ~ ~ t a i n e d 1yo finely divided filter paper cellulose, wet ground in a pebble mill, and Soyo rumen liquid obtained by squeezing rumen contents through cotton baclted by cheesecloth. Ingre- dients of the medium in percentages were: I<H4POI 0.02, K3HPOJ 0.03, MgSOI 0.01, CaC12 0.01, NaCl 0 .1 , (NHi)?SOI 0 . 1 , and resazurin 0.0001. Cysteine (0. 02y0) and sodium bicarbonate (0.5%) were added after sterili- zation and the mixture equilibrated with an atmosphere of carbon dioxide.
The feed extract medium mas composed of 50 ml. of hay extract (4 g. timothy hay boiled 5 minutes in 100 ml. t ap water and filtered), 25 ml. cottonseed meal extract (2 g. boiled 5 minutes in 100 ml. water), and 25 ml. barley extract (2 g. boiled 5 minutes in 100 ml. water). Other constituents were the sallle as for the rumen fluid cellulose agar except that the (NH1)?SOI, NaC1, cellulose, and rumen fluid were omitted.
TABLE I
Amount of tirnothy hay fed I<incl of concentrate and Amount of Ration (as fraction of total a r n o ~ ~ n t (as fraction of NaC1
Xo. digestible nutrients) total digestible nutrients) (Ib. per day)
1 2 /3 Cottonseed meal 1/3 2 2 /3 Cottonseed meal 1/3 3 2/3 Iiolled barley 1/4
Cotto~~seed ~neal 1/12 4 2 /3 Rolled barley 1/4
Cottonseed meal 1/12 5 3/4 Cottonseed meal 1/4 6 3 /4 Cottonseed meal 1/4 7 8/9 Cottonseed meal 1/9 8 8/9 Cottonseed meal 1/9
Ad libitum
0.2 Ad libitum
0 . 1 ~ i c l l i b i t~~n l
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUNGATE: RUMEN MICROORGANISMS 291
011 June 15 all animals were sampled and cultured on the same day. Sampling required about 2 hours and inoculations were not completed until 10 hours after the last sampling. During this time the anaerobic conditions kept the bacteria viable ancl the lower temperature and lack of added substrate reduced changes in numbers and kinds of organisms due to growth and death. I n the June 15 series the order of inoculation and the culture counts were compared to detect any correlation, but none was found. For the remaining samplings, half of the animals were examined on one day and the other half on the next. This reduced the time between sampling and completion of culturing to 4 to 6 hours.
After 24 hours' incubation the feed agar tubes were examined for colonies of Streptococcus bovis, identified by their large size and the microscopic appearance of the cells. The tubes were further incubated and a total colony count made after 10 days. The cellulolytic colonies in the rumen cellulose agar were examined and counted after 20 days of incubation.
For direct counts of bacteria, 1/10 ml. of the 10X dilution was mixed ~v i t h 9 . 9 ml. distilled water, and 0.01 ml. of this 1000X dilution spread evenly over 2 sq. cm. of a clean slide which had just previously been covered \vith a loopful of a 0.1% solution of sodiu~n lauryl sulphate to facilitate even spreading. The smears were dried, fixed, and Gram stained. The bacteria were counted in a total of 15-20 microscope fields, evenly spaced across the smear on the vertical or horizontal mid-line. A greater concentration of organisms was always noted a t the edge of the smear so four of the fields chosen were a t the encls of the horizontal and vertical bisects. For the June 5th samplings no direct counts of bacteria were obtained.
Foi- direct counts of protozoa, 1 ml. of the 10X dilution was added to 1 1111. O F 40y0 ethanol, 8% acetic acid, 0.05% methyl green. The samples mere left in this preservative until all had been collectecl. Counts were made using a ~llolcl counting chamber.
Results of the Various Counts
The bacterial counts are shown in Table 11. The direct microscopic counts ranged from a low of 8 billion per milliliter to a high of 47 billion. 'The total average count uras 20 billion ancl the average for each of the three samplings did not differ significantly from this a t the 5% level.
The total culture counts wit11 the feed evtract agar medium showed great variation. The averages for the different samplings differ significantly a t the 1% level when tested by analysis of variance (16). The counts on June 15 for animals 44, 52EL0, and 59 \Xiere as high as many of those reported by other investigators for alfalfa hay rations. But the cultures from the other animals showed much smaller ~lumbers. This variation may have been due to the shorter period (2 weeks) on the ration prior to sampling.
The feed extract medium apparently satisfied the culture requirements of only a small fraction of the bacteria in the rumen. The average total counts were onl). 0 . 1 to ly0 of the direct count. Higher counts could u~ldoubtedly
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
NUYBERS OF BACTERIA AS ESTIJIATISI) BY SEIrER.41, METHODS
C o u n t s in nill lions except direct c o t ~ n t in b i l l i o n s
June 15 samples --
July 5, 6 san~ples .-\ugusL 3. 4 samples September 7. 8 samples
Cell~iio- Cellulo- Cellulo- Cellulo- Cow Ration Total S. lytic Total .S. Istic Total S. lytic Total S. ls t ic Xo. No. cul ture bouis bacteria Direct cul ture borlis bacteria Direct culturc buois bacteria Direct culture bouis bacteria
Average 450 4 . 8 3 . 0
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
1-1UNG.-ITE: RUMEN MICROORGANISMS 293
have been obtained by using the improved techniques of Bryant and Burliey (1) and Malii (14). These include addition of rumen fluid to the medium, mixing the sample in a Waring blendor, and adding clrsteine to the dilution water. Even with improvements in culture techniques it is questionable whether the culture count \vould have constituted more than a few per cent of the direct count. According to Maki (14) the culture count for animals on a high fiber ration (alfalfa hay) is a much smaller fraction of the direct count than for animals on a low fiber ration. Tilnothy hay would be expected to contain a higher fiber content than alfalfa and might show an even greater discrepancj~.
The numbers of S. bovis detected with the feed extract medium fell below 200,000 per milliliter in only 2 out of 93 samples. The largest count was 140 million per milliliter. The averages for each sa~npling period do not differ significantly a t the 20% level when tested by analysis oi variance. The culture counts for S. bovis are more consistent than either the total or the cellulolytic colony counts. They are about the salne as the counts previously obtained for normal sheep (10) but are sornemhat larger than the counts of Wilson (18).
The cou~lts of cellulolytic bacteria were slightly lower than the cou~lts for S. bovis. This probably does not indicate fewer numbers, since the cells of S. bovis are single or in twos whereas the cellulolytic bacteria accumulate on particles of plant material. Particularly since they were not dispersed \\-it11 a Waring blendor, the cellulolytic cell clumps would be relatively larger and less numerous. With the exception of the September sainples the average values agree fairly well.
The September samples were immediately diluted to 1000X (rather than l o x ) and kept in this dilutioil until just before inoculation. 111 the 10X dilutio~l the organisms were sufficieiltly coilcentrated that traces of oxygen or other toxic factors \\-ere absorbed without damage during the storage period. In the 1000X dilution this muti~ally protective effect was much less. The cells of S. bovis ancl others included in the total culture count are lnore resistant than the cellulose digesters and were less injured by this change in procedure. In inany of the September sanlples the total culture count ancl the S. bovis count mere the same.
In counting the protozoa, Diplodiniunz was identified ancl counted accurately. The average nuinber was 6000 per nlilliliter of rumen contents, with a maxiinum of 27,000. The numbers did not vary with the ration. -. I he factor for the direct couilt was 800. In inany cases the number of Isotricha ancl Dasytricha per milliliter was below this value, so accurate average numbers were not obtained. The mavi~num count for I s o t ~ i c h a was 10,400 pel- milliliter and for Dasytricha 8800. If those animals showing fewer than 800 per milliliter mere arbitrarily assu~ned to contain a t least 400 per milliliter, the average couilts for Isotricha and Dasytricha were 1200 per ~nilliliter each. The Entodiniz~nz were considerably more nulilerous than any of the other protozoa but were not accurately counted.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
294 CANADIAN JOURNAL O F MICROBIOLOGY. VOL. 3. 1957
Characteristics of S. bovis
The colonies of S. bovis were white, light orange, or orange in color. Each of 48 colonies was suspended in a small quantity of sterile t ap water and from this a loopful was inoculated into aerobic tubes of agar medium containing an inorganic solution plus ruinen fluid and 0.5yo starch. No reducing materials were added and cotton-plugged tubes were used. Incubation was a t 45" C. No growth occurred in any of the tubes. I t seemed possible tha t conditions in these cultures were too aerobic, so 18 of the same tubes were autoclaved, reinoculated, the air in the tubes replaced with nitrogen, and the tubes rubber stoppered. These tubes showed excellent growth the next morning. T h e results indicate tha t on initial isolation the strains were inhibited by oxygen.
The original colors were not maintained during subsequeilt transfers. Some of the white strains showed orange colonies and vice versa. T h e same medium was used in all tubes. Since in the original tubes both white and orange colonies had been found in the same regions, the color variatioils could not be accounted for by differences in the medium.
Eighteen strains of S. bovis were isolated fronz the cattle a t Pullman, three from cattle a t I thaca, and six from sheep a t Ithaca. Their characteristics are shown in Table 111. Many of the sugars listed in Bergey's Manual a s always fermented were not tested.
AH strains were Gram-positive cocci. In fresh mounts they were often in pairs and then sometimes slightly lanceolate. Action on litmus milk was variable: 16 of the Pullman and two of the Ithaca strains from cattle formed acid with no reduction or coag~~la t ion , and two Pullman strains formed acid with slight coagulation but no reduction. T h e six sheep and one cattle strain from Ithaca showed acid, coagulation, and reduc t io~~ . Most isolates caused no visible change in blood agar, b u t a few gave a very slight green color.
All strains grew readily in feed extract medium. I t seemed probable t h a t the starch in the barley was the fermentable substrate, yet the culture counts did not correlate with the amount of barley in the ration (Table IV). Rations 3 and 4 with the barley did not show sigilificantly higher counts. T h e possibility tha t S. bovis might utilize other constituents was tested by incor- porating them separately into anaerobic agar slants and iiloculatiilg with 11 of the Pullma11 strains. All showed growth and acid production within 24 hours on the hay extract and cottonseed meal extract, respectively, and 10 showed growth and acid production on the barley extract. Growth was not a s profuse on the latter medium and in lloile of the tubes was growth as abundant a s on the combined extracts. Before inoculation, only the barley extract medium gave a positive starch test. T h e results indicate tha t constituents of the ration other than starch can support growth.
Very high couilts of S. bovis were observed in previous experiments (10) in which grain was suddenly administered to hay-fed sheep. In order t o test whether continuous consumption of a high grain ration would cause high counts of S. bovis, rumen contents of a youilg bull on a fattening ration of
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUNGATE: RUMEN MICROORGANISMS
TABLE 111
CH~~RACTERISTICS OF STRAINS OF S. bovix ISOLATED FROM THE RUMEN OF CATTLE AXD SHEEP
Strains from cattle Strains from sheep
Pullman, Ithaca, I thaca, 18 strains 3 strains 6 strains
Fermentation characteristics
Glucose Salicin Raffinose Lactose Cellobiose Starch I n ~ ~ l i n Esculin Trehalose L-Arablnose Mannitol D-Arabinose Sorbitol Glycerol D - X ~ ~ O S ~
Other characteristics Hydrolysis of Na hippirate Growth a t 45" C. Survi\w 60' C. for 30 minutes Growth in 2% NaCl Growth in 6.570 NaCl Gelatin liquefaction Catalase Indol Capsule formation Urease
TABLE IV
AVERAGE CULTURE COUXTS OF S. bouix ON DIFFERENT RATIONS, MILLIONS/ML.
Rations (Table I )
Sampling 1 2 3 4 5 6 7 8
K~Y- August September
Average 11.6 10.0 16.7 4.1 7.7 13 5 .3 11.3
$ grain - 5 hay were inoculated into dilution series of wheat extract agar (filtrate from 1 g. ground wheat boiled in water and filtered through cotton). A S. bovis colony count of 5 million per gram rumen contents was obtained. Starch-digesting rods of various sorts were much more numerous. Repetition of the test gave a count of 7 million S. bovis. A contiilued high grain ration does not necessarily support ullusually high numbers of S. bovis; other amylolytic organisms may outnumber it. A partial explanation for this
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
296 CANADIAN JOURNAL O F MICROBIOLOGY. VOL. 3. 1957
may be inability of S. bovis to attach to starch grains. Cells inoculated into liquid media containing raw starch sho~ved no tendency to aggregate on the grains, in contrast to Bacteroides amylophi l z~s (6) and the coccus of van der Wath (17).
The gro~vth of S. bovis on hay suggested its presence in the hay before the latter was ingested. A sample of good quality grass hay and one of alfalfa hap were inoculated into a dilution series of rumen liquid glucose agar. No colonies of S. bonis developed, sho~ving that numbers were less than 500 per gram of air-dried hay.
Since S. bovis could gronT on the ingredients of the feed without adcled rumen fluid, its ability to produce nutrients for the cellulolytic cocci requiring rumen fluid was tested. Four grams of poor quality hay and 1 g. of dairy feed concentrate* were boilccl ~ i ~ i t h 140 ml. of inorganic culture solution and filtered. The filtrate was clivided into two equal portions which were sterilized and cooled. One was buffered by adcling 0 . 3y0 sodium carbonate and bubbling with carbon dioxide and was inoculated; the other served as a control. After incubation a t 38" C. the inoculated culture sho\ved typical development. A t 48 hours the flasks were reinoved and to each were added 1 .2 g. of agar and 30 ml. of a suspension of finely divided cellulose, and the contents were resterilized. Experimental and control tubes were brought to the saine pH and inoculated ~ i ~ i t h four strains of cellulolptic cocci isolated froin the rumen. The strains were also inoculated into the usual rumen fluid cellulose agar medium (9).
The growth of S. bovis exertecl no favorable influence on the development of any of the cocci tested. Coccal strains 43-1 and 46-1 grew only in the ruinen fluid medium. Strains 61 and 69 sho\iIed about the same development in the three different meclia.
The actively ainylolytic coccus described by van der Wath (17) possessed many of the characteristics of S. bovis. In addition to attackiilg starch, lactose, inulin, raffinose, arabinose, and glucose, it survived a temperature of 60" C. for 30 minutes. Sorbitol and mannitol were not fermented. I t differed from S. bovis in failing to ferment salicin and in being strongly iodo- philic.
A strain of S. bovis which had been carried for some time in the laboratory was inoculated into both liquid and agar media containing starch and the resulting cells tested for iodophily. None was found, even though the test was made 6 to 8 hours after inoculation ~vhen gronjth was rapid. A culture freshly isolated froin a sheep (Pullman) was similarly testecl but no ioclophily found. Of 25 strains of rumen streptococci testecl by Hobson and iVIann (7) only seven sholi~ed iodophilp and these only \\illen the cells were old. The coccus of van der Wath, although siinilar in many respects to the tested strains of S. bovis, is apparently not identical with them. Whether it repre- sents a different species or a variant cannot be determined until additional
"Forty parts corn gluten feed, 54 hominy and cornmeal, 38 distillers' dricd grains, 36 brewers' dricd grains, 8 linseed oil meal, and 20 molasses.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
TIUSGATE: RUMEN MICROORGANISMS 297
strains of amylolytic cocci have beell isolated from the rurnen of animals in widely separated geographic localities and their characteristics described.
The streptococci isolated from sheep by MacPherson (13) are quite similar to the S. bovis strains except that glycerol was occasionally fermented and mannitol was not. The aerobic plating technique ~ ~ s e d would select against organisms as anaerobic as those of the present study and may explain the fermentation differences.
The Kinds of Cellulolytic Bacteria Observed
The cellulolytic colonies were identified by inspection of the colony itself and by phase microscopy of fresh mounts. When several very similar colollies were in the same tube, usually only one was examined microscopically. Most of the cellulose digesters could be classified into one of the following groups: (1) cocci 0.8-1.2 p in diameter, single or diplo, or in chains, forming discrete colonies; (2) small rods similar to Bacteroides succinogenes, formiilg no discrete cololly and moving through the agar as the cellulose dissolved; (3) curved rods, often motile, forming discrete colonies with usually only a narrow zone of cellulose digestion, similar to the butyric rod previously reported (9); (4) large sporeforming rods, non-motile, forming a diffuse irregular colony; and (5) minute cocci 0.3-0.4 p in diameter. The occurrence of these types in the initial rumen fluid cellulose agar cultures is shown in Table V. Table VI summarizes their incidence.
Many of the colonies, which could not be identified by macroscopic and microscopic examination, were subcultured to dilution tubes of rurnen fluid cellulose agar, because they might be new types. In a few instances the subcultures failed to grow, but in most cases as they were freed of accom- panying bacteria they could be assigned to one of the above groups. Many of the unidentified colonies ~vould probably also have fallen into one of these groups but they could not be cultured.
As seen in Table VI, the butyric rods occurred most commonly, followecl by the cocci, sporeformers, Bacteroides succinogenes, and the tiny cocci. The sporeformers \vere detected in most animals, followed by the cocci and the butyric rods. The cocci, the butyric rods, and B . succinogenes had been identified as important digesters of cellulose in the rumen, but the spore- foriners and tiny cocci had not previously been encountered. Only a few coloilies of the tiny cocci were seen and none greul in subculture. Repre- sentatives of the other types were isolated and, together with strains from other sources, were st~ldied in more detail. The results for the sporeformers and the cell~ilolytic cocci are reported.
Sporeformers from the Rumen
The rather large numbers of cell~~lolytic sporeformers (Tables V alicl VI) in the rumen contents of the animals fed timothy hay, and the regularity of their occurreilce illdicatecl an importance in the rumen. Three strains from aliirnals 28, 52EL0, and 66, respectively, were isolated and their character-
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
CAN.4DIAN JOURNAL O F MICROBIOLOGY. VOL. 3, 1957
m u 0 0 U NTJ 5 -W zg 9 1 9 9 9 f?? 1 9
0 0 0 0 0 0 3 0 0 0
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUNGATE: RUMEN MICROORGANISMS 299
'TABLE VI
OCCURRENCE 01; CmLuLoLYric TYPES IS THE COWS FED TIMOTHY HAY
Sporefor~ners Cocci Bacteroides Butyric rods Tiny cocci Unidentilied
Per cent of the celluloly tic colonics
Per cent of the animals in which found
June July August Sept. Total June July August Scpt. Total
istics studied. The remaining strains from the cows fed timothy hay resembled these three morphologically. Additional strains were encountered when a tablet of dried rumen bacteria prepared by a veterinary supply house was examined for cellulolytic bacteria. None of the previously encountered non-sporeforming rods and cocci were found but several colonies of spore- formers appeared in the 600X dilution in rumen fluid cellulose agar. Five of these were pure cultured and designated B1 to B5. During studies in RiIississippi (11) another strain (No. 30) was found when rumen contents from an animal 011 poor grass pasture were diluted into rumen fluid cellulose agar tubes. No sporeformers were encountered in cultures inoculated from the animals on Ladino clover pasture.
The deep colonies of these sporeformers in rumen fluicl cellulose agar were small, irregular in outline, not very opaque, and often with a seemingly empty center. Surface colonies in cellulose agar were almost invisible except for the zone of cellulose digestion. They spread rapidly as a thin layer with fingerlike extensions which could be detected under the dissecting microscope if the light was carefully adjusted. Clear spots of cellulose digestion were usually visible in rumen fluid cellulose agar 24 hours after inoculation and all cellulose in the tube was often digested within a few days. A very considerable gas pressure developed.
The oval to oblong spores formed a t one end of the sporangium. The spor- angia were of the clostridial type (Fig. I ) , spindle-shaped, and distinctly different in shape and size from the vegetative cells, which were non-motile rods. The clostridia occurred in large masses in certain parts of the colonies in rumen cellulose agar. In some preparations they were attached by their tips in rosettes of three to six cells (Fig. 2). The distal tips were in turn attached to other cells, giving a striking three-dimensional network. The appearance suggested that the clostridial forms divided. Spores developed in the clostridia as shown in Fig. 1. I n most of the strains the sporangial wall initially remained attached to the spore but after culture in the laboratory for a long time this characteristic was often absent. Strain 30 never showed the old sporangial wall attached, even when first isolated. This strain differed also in forming small amounts of yellow pigment.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
300 CANADIAN JOURNAL O F MICROBIOLOGY. VOL. 3. 1957
Spores were observed in those parts of the colony just inside the periphery. As the colony enlarged a shell of spores was alnrays found in the clistal portion and those more centrallj- locatecl disintegrated. Also in the long fingerlilte surface growths the spores \\,ere found near the tips but \\,ere allnost entirely absent from the older portions. In many old cultures no spores a t all were found in the areas cleared of cell~~lose. They formed initially and then conlpletely disappeared. This behavior was some\vhat variable. Spores were usually forlned more abundailtly on first isolatioll and there was some indication that the abunclance of spores was influencecl by the particular batch of culture medium. M a n j ~ of the strains n-ere lost because subcult~~res failed to grow.
The heat resistance of the spores was not determined, though strain 30 mas viable after being heated to SO0 C. for 1 minute. During the initial isolatiolls subcultures often failed to grom unless fairly massive illocula were used. Also, the number of colonies developing was usually less than espectecl. Preliminary tests suggested that the organisms might be sensitive to oxygen. A colony was divided in two, and one half subcultured with as little exposure to air as possible. The other half was held for 5 minutes on the encl of the inicrospat~~la used for inoculatiilg and then similarly subcultured. The inoculum protected from air gave many more colollies t h o ~ ~ g h a few developed froin the one exposed. In another experiment, exposure to air for 2 minutes sigilificantly decreasecl the viabilitl-, though not to the same extent. These results suggest that oxygen 1;ills the cells though it is possible instead that their gro\vth was inhibited bl- the sinall amouilt of oxygen addecl wit11 the inoculum. The sporeformer strains were extreinely sensitive to oxygen ancl never sho\ved grourth if resazurin was oxidized. In some cases, even though this indicator reduced, it was suspected tha t failure to grom was clue to too high an osygen tensioll in the medium.
The sporeforlners do not require rumen fluid. They grow on inally common laboratory media and show a rapid developinent in an infusion of hay plus concentrate. Their ability to digest alfalfa hay was tested by inoculating into rumen fluid broth in tubes containing 40 mg. air-dried hay. Uilinoculated control tubes incubated in parallel contailled 22.7 mg. dried material. The t~ibes inoculated with strains B4, B5, 30, 66, and 52ELO showed weight decreases of 44, 45, 47, 32, and SO%, respectivelj.. The residues uTere not further analyzed to measure the digestioil of the various constituents of the hay but the decrease in total meight suggests that nlore than one fraction was attaclted (see later section on hay cligestion by the cocci). Strains 66 ancl 52ELO caused decreases of 22 ancl 41% in the residual weight of poor cl~lality grass hay. These results support the view that the sporeformers play a part in fiber digestion in the rumen.
Other characteristics of the sporeformers were as follows: strain 30 gave no visible growth in beef estract - peptone broth or in brain heart illfusion broth; no visible growth in nutrient gelatin but the gelatin coinpletely liquefied in 96 hours; no visible growth in R4R-VP broth, and the methyl
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
FIG. 1. Gram stain oi C. loclrhcc~dii, strain 30, from a 24-hour rumcn fltlid g l~~cosc agar c u l t ~ ~ r c . \'egciati\.c cclls i111d a few larger cells regarded as clostriclia. hlag. 1550 X. Green liltcr.
FIG. 2. Same snlcar as Fig. 1. Selected portion to sIio\\. clostriclia and their arrangement.. FIG. 3. S~nca r of 45-hot~r c ~ ~ l i l ~ l - c of C. lotr~gispor~iirt in rumen fluicl ccllol~iosc agar.
\Tcgciati\.e cclls and clustritlia. Gram stain. Rlag. 1550X.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
FIG. 4. Smcar of 5-day cult~irc of C. / n n g i s ( ~ o r l c ~ ~ ~ in rumcn f l~~ i t l cellulose agar. Vegetative cells and clostriclia containing spores. Spore stain wi th rnalaciiitc grocn and safranin counter stain. ( a ) Ketl liltcr. Spores sho\r as clarl; arcas in ccntcr of clostridia. (6) Green lilter. Vegetative cells d~stinct. Sporcs almost colorless.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUiXG.4TB: RUMES MICROORG.4NISMS 30 1
red and Voges-I'rosltauer tests both negative; acid and gas in I<ligler's iron agar, no H?S; litmus mill; reduced in 24 hours and co~npletely peptonized in 48 hours; gas and acid in thioglycolate agar. Strains B4 and B5 showed similar characteristics except that gelatin mas not liquefied by B5 until after 20 dajys ancl by B4 only partially after 20 days.
Substrates fer~nellted by strains 28, 52EL0, 66, and B1 to B5 are slxown in Table VII. Of the sugars commo~xly at tacl~ed only two were monosaccharides. These tests were made with a medium containing runlell fluid and 0.4% substrate. Copious slime was produced in the tubes containing fennentable sugars. In some, the estracellular slime was so abundant that the medium gelled and the tube could be inverted without flow of the contents. Substrates not fermented were D - X J - ~ O S ~ , L-nrabinose, D-al-abinose, galactose, mailnose, rhaninose, trehalose, lactose, raffinose, inulin, and esculin.
TABLE YI I
SCBSTRATES FERMENTED BY THE SPOREFORMERS
Strain
Substrate 28 52ELO 66 B 1 B 2 B3 I34 B5 30
I.-Xylose GIl~cose Levl~lose Cellobiosc Mal lose Sucl-ose Starch Salicin Cellulose
The fermentation products of strain 52ELO were determined. I t was grown in all-glass sealed flaslts with 1 g. of poor quality grass hay and 0 .5 g. concentrate as substrate. The weight of insoluble substrate digested mas 335 mg. but since soluble substrates were also present this callnot be used as the basis for calculating carbon recovery. I'roducts formed, in mM., were: carbon dioxide 2.55, hydrogen 2.76, acetic acid 0.92, formic acid 0.52, butyric acid 0.87, and ethanol 0.54. S o succinic acid was formed and only a trace of lactic acicl. The large proportions of hydrogen and carboll dioxide account for the collsiderable gas pressures observed in the culture tubes.
The characteristics of these sporeformers differ in several respects from the previously described species of Clostridiurtz and ~lecessitate that they be assigned to a new species. In recognition of Dr. Lochhead's pioneer work in demonstrating cellulose digestion by colonies of aerobic bacteria (12) and in further recognitiolx of his important contributions to our lc~lowledge of bacteria as they exist in nature this new species is lla~ned Clostr idiz~m lochheadii with the follo~ving characteristics.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
302 CANADIAN JOURNAL OF MICROBIOLOGY. VOL. 3, 1957
Gram-positive rods 0.7-1.5 p by 3-7 p. Non-motile. Anaerobic. Form oval to oblong spores 1-1.5 p by 2-3 p. Spindle-shaped clostridia, 1.5-1.7 p by 5-6 p \with tapered ends, are formed prior to sporulation and the spores develop terminally and slightly laterally in them. The clostridia are often arranged in star-shaped aggregates, with the cells joined by the tips and these aggregates in turn joined to form a three-dimensional network. The appearance suggests that the clostridia divide. Spores form in the peripheral part of a colony, degenerate, and are replaced by new ones formed as the colony grows out. Par t of the sporangial wall often remains attached to the spore, e~tencling from one end as a tube. Spores are unstained by simple stains but show the typical reaction to spore stains.
Subsurface colonies are irregular in shape in cellulose agar, solnewhat diifuse, and the center ~ ~ s u a l l y less dense than the periphery. Colonies spread rapidly on the surface of cellulose arrar.
Cellulose is rapidly feFmented and also glucose, cellobiose, maltose, sucrose, starch, and salicin. Levulose often and L-sylose seldom fermented. Not fermented were D-sylose, L-arabinose, D-arabinose, galactose, mannose, rharnnosc, trehalose, lactose, raffinose, inulin, and esculln. Forty to fifty per cent of the insoli~ble materials in alfalfa hay were digested. Copious slime formed when carbohyclrates were attacked.
Growth occurs on a wide variety ol media containing fermentable carbohyclrate, including infusions of various ruminant feeds. Does not require rumen fluicl. No visible growth in beef extract - peptone broth or beef heart infusion broth. No visible growth in nutrient gelatin but gelatin liquefaction may be complete \\;thin 96 hours, or only partial. No visible growth in hIII-VP broth ancl the methyl red and T'oges-Proslcaucr tests both negative. 'lcid and gas in ICligler's iron agar; 110 H2S. Litmus milk recluced in 24 hours and completely peptonized in I S hours. Gas and acid in thioglycolatc agar.
Habitat: the rumrn of cattle.
Another sporeformer was first observed as a very orange colony in the rumen fluicl cellulose agar tube (200,000X dilution) inoculated from animal 52WM'. I t produced much more pigment than any of the other cellulolytic bacteria thus far observed. The vegetative cells were motile rods 1 ,u in diameter by 7-15 ,u long. Some of these transformed into clostriclia 2-3 ,u by 7 ,u (Fig. 3) i11 some of which slightly CLII-ved spores 1 ,I.L in diameter and 3-6 ,u long were seen (Fig. 4). The curvature of some of the long spores appeared to be slightly spiral rather than in a flat plane.
Strain 52M7W was estremely difficult to maintain in subculti~res. Large iilocula often failed to grow and even \vhen gro\vth occurred ollly a few colonies fonned. A piire culture was obtained by piclcing fro111 ruinen fluid cellobiose agar and was maintainecl for about 3 months before it was lost by too long a delay in subculturing. The spores resemble those of C. lochheadii in losing their viabilit), \\;ithi11 a few ~ee l c s .
The carbohydrates fermented nrere cellulose, destrose, galactose, cellobiose, and sucrose. R/Iuch sliine was produced on these sugars. Cellulose digestion was rapid but not as rapid as by C. lochhendii.
The strain grew on nutrient broth or ).east extract i f cellulose was preseilt. Growth with carbon dioside and bicarbonate n-as better than with nitroge~l gas.
Fermentation products froin 1 .14 m3/1. cellulose (as hesosc) were, ill mR4.: hydrogen 0.59, carbon dioxide 1.13, acetic acid 0.51, formic acid 0.75, and ethanol 0.26. No propionic, butyric, lactic, and succinic acids were produced.
Strain 52M7W differed from C. lochheadii in being motile, in producing very long spores, much more yello\v-orange pigment, 110 butyric acid, and in
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUNGATE: RUMEN MICROORGANISMS 303
fermenting galactose. These features necessitate placing it i l l a separate species, Clostridium longisporum n. sp., with characteristics as described above.
The habitat of C. longisporum is unknowil. 111 the dilutioil series i l l which it was originally isolated there were no yellow-orange colonies visible in the next lower dilution so i t may have been a co~~tamina i l t . Once previously an organism similar in color and cellulolytic activity was found as an undoubted contaminant in a rumen cellulose agar series, but was lost before other char- acteristics could be ascertained.
The Cellulolytic Cocci
Seven strains isolated from the cows on the timothy hay rations werc picked because they represeilted the range of differences in colony types. Other strains mere isolated from sheep a t Itllaca. Some of the cllaracteristics are s l~o \v i~ in Table VIII.
The reaction of the rumen cellulolytic cocci to the Gram stain is not uniform. The most thorough study of the Gram reaction was made with strain 69. An 8-hour culture i l l rumen cellobiose broth was subcultured to a tube of similar medium, which was incubated 12 hours, and the cells then Gram stained in parallel with Eschericljia coli and Streptococcus pyogenes. The cells were Grain-negative without exception. The slightlj, more intense color of the cocci as compared with E. coli was just as marlced \\-hen safranin was applied as a simple stain as \\rile11 it was part of the Gram procedure. With the other strains, the Gram reaction was often variable.
Strain S2-23 was observed in the first dilution tube of a rumen fluicl cellulose agar series illoculated with rumen contents from a sheep which had been given excess glucose 10 hours prior to sampling. The culture count of cellulo- l ~ t i c bacteria before the glucose was admillisterecl \\as 10 million per milliliter, most of the coloilies appearing to be the 11011-sporeforming butyric rods. Follo~ring the administration of glucose the rumen coilteilts became quite acid (10) and the animal died 7 days later. Ten hours after the glucose n-as given another sample of rumen contents was reilloved and diluted into a rumen cellulose agar series. These sho\\red 500 cellulolytic colonies per milliliter of rumen contents, a great decrease from the earlier count. Colonies of Gram-positive streptococci were prominent among the cellulolytic colonies from the acid rumen and one of these was isolatecl ill pure cul t~lre as strain S2-23. Out of 24 colonies esaininetl from the pre-glucose series none were of this type, inclicating tha t they were much outnumbered by other cellulose decomposers in the original ruinen contents but were relatively more abundant in the acid ruinen. Smears of the rumen coilteilts before and after adminis- tration of the glucose arere examined and streptococci identical in morphology with the pure culture were observed in both, but were much more numerous ill the normal contents, indicating tha t the streptococci also dimiilishecl in numbers wit11 acid but not as rapidly a s the other forms. Strain S2-23
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
C.4SADIAN JOURNAL O F MICROBIOLOGY. VOL. 3. 1957
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUKGATE: IiUMEN MICROORGANISMS 305
digested cellulose nluclz more rapidly and completelj. than the strains of streptococci previously obtai~led. Unfortuilately, the culture was lost before the iermentatio~l products could be analyzed.
The sugars ierillented by 10 strains of streptococci are shoxvn in Table IX. xone of the tested cocci fermented glucose, L-sylose, D-xylose, D-arabinose, galactose, rhainnose, raffinose, inulin, trehalose, esculin, salicin, and starch. Cellobiose mas the o~lly sugar fermented bj- all strains.
TABLE IS
L-Xrabinose - + + - - - - - - - Cellobiose + + + + + + +
- + - +
- Lactose + + * + - - Mzl tose * - - - - - - i - hIannose + - - - i - - - - Sucrose i - - i -
- - - - +
Fructose + - - - -
The results of the anal!.ses for fermentation products are shown in Table X, together xvith the fermentation products of other cocci reported in the literature. The values for carbon dioxide involved a coilsiderable error owing to use of 10070 carbon dioxide and 0.5% sodium bicarbonate in the medium. The equilibration between gas and liquid phases was not always complete.
According to their fermentation pattern certain of the cocci are closely related, but many show differeilces indicatiilg diverse origins. Types producing much or little of a certain product can be designated, but there are intergradatioils of types, and strain S1-2 produced a t least a small quantity of each product.
The products s h o ~ ~ ~ n in Table X can arise fro111 one fermentatioil pattern, e.g. that of Escherichia coli , and there is no evidence to excl~~cle the possibility that the differences in metabolic products of the cellulolytic cocci are due to variations in a basic pattern coininon to all strains. On this basis they should be included in a single genus.
The first generic desigilation for a member of this group was R u m i n o c o c c z ~ s , with R. j7avefaciens as the type species (15). In order to accommodate all the strains of cell~~lolytic cocci in the genus it is necessary to redefine it as follows: R z ~ m i n o c o c c z ~ s (15)-Spherical cells, somewhat elongated prior to division; single, in twos or in chaills, never in tetrads or cubical packets. Non-motile, 11011-sporeforming. Gram-positive or -negative or -variable. Anaerobic. Ferment carbohydrate to form acetic acid, a t least traces of hydrogen, and various combinations of ethanol, formic acid, lactic acid, and succinic acid. Type species Ruminococcz~s~uvefuciens Sijpesteijn.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
TABLE X
" Substrate Products formed, mM. llesose fermented
,+ - ?:
.4mount L.
fermented, Acetic Formic Lactic Succinic Carbon CI
Strain Source Kind mM. I I? CO? Ethanol acid acid acid acid recovery, OJo 0 - s
61-1 Cow Ceilrllosc 0.83 0.037 R~r?ni?zococc~rs
fla~cfaciens S Sheep Cellulose 1.6 Sijpesteijn Ccllobiose 2.05
(15) 52 Maki (14) 53 Maki (14) S1-y F-6 I-Iall (5)
69 Colorless coccus U
'Iungate (9) 56-2 Yclloxv coccus h
I-Iungate (9) 66-1
Cow Coxv Sheep Rabbit
Sheep Cow
Cow Stcer
Cow
Steer Cow
Cellobiose Cellobiose Cellulose
Cellulose Cellobiose Cellulose
Cellulose Cellulose Cellobiose Cellulose
0.5 0.193 0.094 0.675 70 F
None 0.65 0.64 Trace 0.51 66 - 0 None 0.73 0.58 0 .5 67
0.26 0.4 Tracc None
0.213 0.43 Kone 0.07 0.10 0.61 1.022
1.03 0.69 0.378 0.175 0.31 0.27 1.06 None None None
None
0.223 0.486 0 .6 Trace 75 w Trace -0.19 0.384 None 26 - 0.102 0.029 0.147 0.378 None 26 'O
0.14 0.38 0.141 0.403 None 42 (n -2
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
I-IU?iG.-\TI?: RUMEN MICROORGANISMS 307
On the assumption that strain 61-1 of Table I X is identical with Sijpesteijn's strain S (the only one for which fermentation products were determined), production of hydrogen should be added to the characteristics of R. JZavefaciens. The yellow color is not always exhibited, though usually present under certain conditions. T o this species may be assigned also strains S1-y, F-6, S1-2, and probably 52 and 53 of Table X.
Strains 69, 56-2, 46-1, 43-1, and U (9) are considered to belong to a new species, Rzcminococcus albzls, which cl~aracteristically produces none or little of the yellow pigment. T h e colonies are usually white. Deep colonies are initially lens-shaped, but soon become multiple lens-shaped and often spread out in a complex arrangement. Cells ~lsually single or in twos, often sliglltly elongated prior to division, 0 . 8 to 2 . 0 ,u in diameter, Gram-negative to Gram- variable. Capsule often formed. Ferments cellulose ancl cellobiose, bu t usually does not ferment glucose and other sugars. Nutrient recluirements are not met b y usual bacteriological culture media, bu t can be met by rumen flaicl, extracts of feces, or other media containing metabolic products of other microorga~~isms. Hydrogen, carbon dioxide, ethanol, acetic acicl, lormic acid, and lactic acicl are produced in various combi~lations and proportio~ls, b u t no succi~lic acid. Usually produces more hydrogen and carbon dioxide tllan does R . ,flavefaciens. Usually not iodophilic. Strain 69 is the type strain. Habi ta t : I-umen of cattle and sheep.
The yellow coccus strain A is co~~siclered a variant of R. JaveJaciens, proclucing less than the usual amount of succinic acid. Strain 66-1 does not easily fit either o i these species. The low carbon recovery suggests tha t not all the iern~entation products were identified.
T h e number of cellulolytic cocci in the rumen is one measure of their importance. Bryant and Burltey (2) found tha t they comprised 5 to 10y0 of the total c u l t ~ ~ r e count. A seco~lcl measure is the degree to which they call cligest the i~lsoluble co~ls t i tue~l ts of roughages. I11 measuring this ability it is important that growth of the test organisnl be limitecl only by the amount of aclclecl roughage. With a low concentration this requirenlent is fulfilled ancl the concentration of metabolic products remains below the inhibitory level.
Air-dried grouncl alfalfa hay (40 mg.) was weighed into each of numerous culture tubes containing 10 1111. of 30% rumen fluid broth. These were inocu- latecl with the test c u l t ~ ~ r e s and i~lcubatecl 2 weeks a t 39O C. The supernatant fluid was removecl, the sediment suspended in a little distilled water and transferred to a weighing tube. T h e plant fiber particles soon settled and the supernate with its still-suspended smaller particles (including the bacteria) was again removecl. This washing process was repeated until the supernate \\ras clear. The sediment was dried and weighed. Uninoculated tubes similarly treated showed the dry weight of sedimented fiber particles in the alfalfa to be 22.7 mg. T h e percentages of the insoluble materials digested by the tested strains were: 21 for 61-1, 35 for 43-1, 33 for 66-1, and 45 for 69.
The utilizatio~l of various fractions of hay b y other coccal strains was tested. T h e experiment was similar to the above except tha t after determination of
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
308 C.4NADIAN JOURNAL OF IvIICROBIOLOGY. VOL. 3. 19Si
the residual dry weight the residue was s~~ccessively treated with benzene- alcohol, hot water a t 8S0 C. for 24 hours, 2% (w/v) sulphuric acid a t room temperature for 24 hours, and 70% (w/v) su lph~~r ic acid a t room temperature for 24 hours. After each treatment the fiber was washed, dried, and weighed. The amount of each fraction digested was calculated and expressed as a percentage of the total weight of that fraction. Tlle results arc s11owll in Table XI for t\vo different feeds, alfalfa hay and a poor quality grass ha)..
Betnreen 44 and 59y0 of tlle materials i l l the alfalfa hay were digested and 19 to 29% of the grass hay. The action on the benzene-alcohol fraction \\;as about the same for the two hay samples. Tlle amount of this fraction was small. The hot-water-soluble fraction in the alfalfa was digested to a greater extent tllan in the grass. Fractions 3, 4, and 5, regarded as consisting chiefly of l~emicellulose, cellulose, and lignin, respectively, sho\ved more digestion in the alfalfa than in the grass. The digestion of the "lignin" fraction in the alfalfa was unexpectedly high and was of significant magnitude even with the grass hay.
All of the fractions of the alfalfa llay were attaclced to about the salne extent. With the grass hay, the attacl; on the "cellulose" fraction was usually greater than that on any other except tlle benzene-alcohol. Although the method for estimating these different fractions was crude, the results indicate that susceptibility to acid l~g~drolysis is not closely correlated \vith digestibilit~; by the bacteria.
Discussion
During the past 10 years various cocci and non-sporeforn~i~lg rods of tlle rumen have repeatedly been identified as cellulose digesters. The present study confirms their importance and provides additional information on their characteristics. The difference in succinic acid production by strain S of Sijpesteijn (15) and the )-ello\\7 cocci (9) is apparently not due to differences in the culture conditions. The bacteria themselves differ. Also otller variations in the characteristics of individual strains show that the rumen celluloiytic cocci are not a hoinoge~leous group. Study of additional strains ma>- disclose better defined and more restricted species than R. jZavefac%e7rs and R. albz~s but present information is insufficient to warrant further splitting. Cellulose digestion is not regarded as a distinctive characteristic of the rumen cocci. So much variation has been observed in the rate of cellulose digestion by pure c u l t ~ ~ r e s that it seems highly probable that non-cellulolytic strains referable to R. jlavefacie7~s or R. albz~s exist in the rumen. , . I he true relative importance of Bacteroides szLcci7zoge?zes in the rumell is
probably not disclosed by the present experiments. Many strains apparently do not grow readily in cellulose agar (3), though cellulose is rapidly digested in broth tubes. I~~clications of sporadic cellulose digestion by B . succinoge~zes were observed when it was first isolated (8) and repeatedly since that time. The factors concerned have not been identified, but they almost certainly affected the agar clilutio~l experiments of the present study and the counts of B. sz~ccitzoge7zes do not reflect its relative importance.
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUNGATE: RUMEN MICROORG.4NISMS
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
310 CANADI4N JOURNAL OF MICROBIOLOGY. VOL. 3. 1957
On a number of occasions the low dilution tubes of cellulose agar series inoculated with rumen contents have shown an estl-emely rapid and complete clisappearance of all the cellulose. 111 the nest lower dilutions, zones of cellulose cligestio~l cleveloped only after much longer incubation, inclicating that the i~loculum contained o~ l ly a few colony-producing ~111its of the rapid cellulose digester. Of all tile isolated cell~ilolytic rum-n bacteria, C. 1oci~iz.eadii is tile only olle which shows such rapid grolvth from a few initial cells. 111 cleep agar colonies its zone of cell~ilose cligestion increases o~ilj. slightly Inore ral>iclly than with most other cellulolytic bacteria, but ill the thin agar of roll tubes the cells soon reach the surface and the11 spread vastly more rapiclly than other forins thus far encou~lterecl. I f the mpicl ancl complete cellulose cligestio~i be ascribed to C. locl~headi i , this orgaiiisin is fairly widespread in the bovine rumell though often in small numbers.
When a colony of C. lochlzeadii in cellulose agar is subcul t~~red , the n ~ ~ m b e r of colonies developing is much fewer than \vheil other strains of ceII~ilose digesters, shelving equal clearing of cellulose, are transferred. The ilumbers demonstrable by culturiilg are relatively less than the cellulose digestion or, conversely, the cellulose cligestion by this species is greater than the numbers indicate. If this observation with agar cultures applies also to the rumen, C. lochkead~ii is of considerable importance with certain ratioils.
The minute cocci seen in areas of cellulose digestion cluriilg the present experime~its suggest that they represent a11 additional cel l~i lol~~tic type. Ho~vever, the capacity to digest cellulose sl~ould not be inferred ~iiitil proved with pure cultures. I t will be iilteresti~lg to see whether these cocci prove to be cellulolytic and also whether additioilal l;i~lcls of cellulose-digesti~ig bacteria of the rumen will be discovered. Whe~lever ruminants are fed a disti~lctive type of roughage it would seem profitable to tletermine the liinds of cellulolytic bacteria concerned.
References 1. BRYANT, M . P. and BGRKEY, L. A. C u l t ~ ~ r a l riiethods ant1 some characteristics of sollie
of the more numerous groLips of bacteria in the bovine rumen. J . Dairy Sci. 36, 205-217 (1953) .
2. BRYANT, M . P. arid BURKEY. L. A. Numbers and some predominant groups of bacteria in the rumen of cows fed dilferent rations. J . Dairy Sci. 36, 218-224 (1953) .
3 . BRYANT, M . 1'. and DOETSCH, R. N. A study of actively cellulolytic rodshaped bacteria of the bovine rumen. J. Dairy Sci. 37, 1176-1183 (1954) .
4 . GALGAN, M. W . and SCHNEIDISR, B. H. Self-feeding salt protein mixtures to range cattle. State Coll. Wash. Azr. Espt . Sta. Circ. 166 (1951) .
5 . HALL, E. R. Investigations on the microbiology of cellulose utilization in dornestic rabbits. J. Gen. WIicrobiol. 7, 350-35': (1952) .
6. Ht\nr~iix, L. J . and HUNGATE, R. E:. Culture and physiology of a starch-digesting Imcterit~rn (Bncleroides alnylophilz~s n. sp.) from the bovine rumen. J. Bacteriol. 72, 548-554 (1956) .
7 . Honsox, 1'. N . and WIAXN, S. 0. Sorile factors alfecting the formation of iodophilic polysaccharide in group D streptococci from the rumen. J . Gen. Microbial. 13, 420-435 (1955) .
8. HUNGATE, R. E. Studies on cellulose fermentation 111. The culture and isolation of cellulose-decomposing bacteria from the rumen of cattle. J. Bacteriol. 53, 631-645 (1947) . . .
9. HUNGATB, R. E. The anaerobic rnesophilic cellulolytic bacteria. Bacteriol. Revs. 14, 1-49 (1950) .
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.
HUKG.-\TE: RUMEN MICROORGANISMS 311
10. HUXGATE, R. E., DOUGHERTY, R. \V., BRYANT, &I. P., and CELLO, R. R'l. Microbiological and physiological changes associated with acute indigestion in sheep. Cornell Vet. 42, 423-449 (1952).
11. HUNGATE, R. E., FLETCHER, D. W., DOUGHERTY, R. \V., and B.IRRENTINE, B. F. Microbial activity in the bovine runlen; its measureme~~t and relation to bloat. Appl. AIicrobiol. 3, 161-173 (1955).
12. L ~ H N I S , F. and LOCHHEAD, G. Uber Zellulose-Zersetzu~lg. Zentr. Baliteriol. Para- sitenl:., Abt. 11, 37, 490-492 (1913).
13. MACPIIERSOS, RI. J. Isolatio~l and identihcation of a~llylolytic streptococci from the rumen of sheep. J . Pathol. Bacteriol. 66, 95-102 (1953).
14. 1 1 , L . Characteristics of rumen bacteria isolated from cattle fed high and low roughage rations. Thesis. University of Wisconsin. 1955.
15. SIJPESTEIJN, A. I<. 011 RZLIILZ'~IOCOCCILS jEavcjacic~ts, a cellulose-dec0111posiq bacterium from the rumen of sheep and cattle. J . Gen. Microbial. 5, 869-879 (19jl) .
16. S T E A R M . ~ , R. L. Statistical concepts in microbiology. Bacteriol. Revs. 19, 160-215 (1955).
17. \Tax DER \\IATH, J . G. Studies 011 the alimentary tract of merino sheep in South .Africa. XI. Digestion and synthesis of starch by rumi~lal bacteria. Onderstepoort J. Vet. Sci. .qninlal Ind. 23, 367-383 (1948).
18. \\'ILSOY, S. M. Some carbohydrate-ferme~ltitig orgallis~ns isolated from the rumen of the sheep. J . Gen. hIicrobiol. 9, i-ii (1953).
Can
. J. M
icro
biol
. Dow
nloa
ded
from
ww
w.n
rcre
sear
chpr
ess.
com
by
San
Fran
cisc
o (U
CSF
) on
11/
18/1
4Fo
r pe
rson
al u
se o
nly.