turkeys has been described in detail elsewhere (Harrison and Hansen, 1950a).

Transcription

1 LACTOBACILLI FROM TURKEYS ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN Department of Bacteriology and the Livestock Sanitary Service Laboratory, University of Maryland, College Park, Maryland Received for publication July 15, 1950 Lactobacilli have been reported as being a predominant bacterial type in the intestinal tract of various warm-blooded animals. For instance, Crecelius and Rettger (1943) found that lactobacilli are the most prevalent bacteria in the intestinal tract of guinea pigs. Shapiro and Sarles (1949) reported that lactobacilli are the most numerous bacteria in most areas of the intestinal tract of chickens, and Harrison and Hansen (1950a) that anaerobic lactobacilli (the Lactobacillus bifidus type) comprise the largest part of the flora of the cecal feces of turkeys. As early as the turn of the century it was known that lactobacilli of this type predominate in the stools of breast-fed infants (Tissier, 1900). Weiss and Rettger (1934) conclude from their experiments that lactobacilli may constitute from 90 to 95 per cent of the fecal flora of breast-fed infants. The grampositive, anaerobic rods isolated from the stools of normal adults by Eggerth (1935) in our opinion were probably lactobacilli. It has been definitely demonstrated by Orla-Jensen et al. (1936) that both Lactobacillus bifidus and Lactobacillus acidophilus may be isolated from the stools of humans, the former being the predominant species among the noncoliform types. These workers also pointed out some important physiological differences between the two species, which not infrequently have been confused in the past. The obligate anaerobe, Lactobacillus bifidus, was discovered to be the most prevalent species of bacteria in the feces of turkeys (Harrison and Hansen, 1950a), but in addition a flora of facultative anaerobic lactobacilli was found. These organisms, although less numerous, were of particular interest because they differed in some respects from species in the literature. It is the purpose of this paper to describe the little known avian facultative anaerobic lactobacilli and to discuss their taxonomy. METHODS The technique employed for the isolation of bacteria from the cecal feces of turkeys has been described in detail elsewhere (Harrison and Hansen, 1950a). Pour plates of diluted fecal samples were prepared using various selective media. Numerous colonies, which subsequent tests revealed to be members of the genus Lactobacillus, were picked from these pour plates. When transplanted to agar slants, these bacteria produced a thin surface growth. Incubation of the slants under a reduced oxygen tension or under complete anaerobiosis, however, allowed development of a heavier growth on the agar. The cultures were found to be gram-positive, nonsporeforming rods that varied somewhat in size and in tendency toward the formation of chains. Preliminary tests indicated these rods to be 543

2 544 ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN [VOL. 60 catalase-negative and incapable of reducing nitrate. Some strains were found to produce considerable amounts of gas as a result of carbohydrate fermentation. The cultures were screened with respect to fermentation reactions by incubation for a week at about 37 C in a medium recommended to us by Tittsler and Rogosa (1948) to which we added about 1.5 per cent agar. The medium used has the following percentage composition: dehydrated liver, 0.5; tryptone, 0.5; tryptose, 0.5; yeast extract, 0.5; bromeresol purple, about ; and agar, 1.5. The ph was adjusted to about 6.9, and the carbohydrates were sterilized by filtration and added aseptically to the autoclaved base. The changes produced in litmus milk were also recorded. By means of these tests, cultures considered representative of the group were selected for lyophilization. Later these cultures were revived, restreaked, subcultured, and studied en masse by the methods to be described in order to ensure correct identification. Nineteen representative strains of the facultative anaerobic lactobacili were grown in flasks in order that sufficient lactic acid could be collected for the identification of these organisms. Each flask contained the following: tomato juice (filtered), ml; trypticase (Baltimore Biological Laboratories), 5 g; yeast extract (BBL), 5 g; glucose, 25 g; and sufficient water to make 500 ml. (The ph was adjusted to 6.9 before the addition of all the water so that the final volume would not deviate from 500 ml.) Each flask received 16 g of calcium carbonate powder and was stoppered, steriized at 15 pounds for 15 minutes, and weighed to the nearest gram. Each was inoculated with a loopful of culture and incubated at 37 C for 2 weeks. After the water lost through evaporation had been replaced, an aliquot was removed for the determination of residual glucose. The method of Bertrand described by Kertesz (1930) was employed after the protein had first been precipitated with copper sulfate and the calcium salts with sodium fluoride at neutrality. Then another aliquot was poured from any remaining calcium carbonate and acidified to a ph of about 1.0 with concentrated sulfuric acid. Steam distillation followed, and the total volatile acid (calculated as acetic acid) was determined by titrating an aliquot of the distillate. The steam-distilled culture was then extracted for 4 days with diethyl ether in a liquid extraction apparatus. Water was added to the ether extract, the ether evaporated, and an aliquot titrated with base from which the total nonvolatile acid (calculated as lactic acid) was determined. The remaining solution of acid was neutralized by adding zinc carbonate, decolorized by boiling with norit, and filtered; and the zinc lactate was crystallized and, in most instances, collected in three fractions. Finally, the percentage of water of crystallization and the optical rotation of each fraction were determined. In addition, 18 representative cultures were grown in various substrates for 2 weeks at 37 C. Fermentation was detected in this instance through the use of the ph meter, no indicator being employed in the medium. This served to verify the reactions obtained earlier in the fermentation screening tests and provided information of a quantitative nature. The basal medium used had the following percentage composition: yeast extract, 2.2; K2HPO4, 0.2; and MgSO -7H20, Sufficient substrate was added to produce a concentration of 2 per cent substrate.

3 1950] LACTOBACILLI FROM TURKEYS 545 All the substrates were sterilized with the base after the ph was adjusted to 6.9 except xylose, arabinose, maltose, melibiose, and cellobiose, which were sterilized by filtration and added aseptically to the autoclaved base. The temperature requirements of our cultures were determined by incubation in broth at various temperatures (20, 26, 30, 36.5, 44.5, 45.5, and 49.5 C) for a period of 2 weeks, the time required for the appearance of growth (turbidity) being noted, also the ph of the culture at the end of the incubation period. The medium used was the same as that employed as the base in the fermentation studies described above; to this base was added 0.5 per cent glucose. An additional set was inoculated, the surface of the broth was covered with sterile vaseline, and the tubes were incubated for 2 weeks at 37 C to test again for the production of gas. The inability of our cultures to reduce nitrate was confirmed by incubation for 2 weeks at 37 C in the same basal medium, to which had been added 1 per cent glucose and 0.1 per cent potassium nitrate. Agar slant cultures were incubated aerobically at 37 C for 28 hours, after which time a smear was prepared and stained with aqueous crystal violet for microscopic observation. The slants were then reincubated for an additional 20 hours and then flooded with 3 per cent hydrogen peroxide to test for catalase, cultures being considered catalase-negative if no bubbles of gas appeared. Again the medium employed was the same as that used as the base in the fermentation studies above except that 0.25 per cent glucose and 1.5 per cent agar were additional ingredients. At first litmus was used as an indicator in skim milk to indicate changes in acidity, but its use was discontinued because the litmus at times stimulated acid production by our strains of lactobacilli. Also, the amount of litmus affected the speed at which milk curdled. In the final experiments only plain skim milk (adjusted to a ph of 7.0 with sodium hydroxide and autoclaved at 12 pounds for 12 minutes) was employed. The percentage of acid (calculated as lactic acid) and the final ph produced in this substrate were determined after incubation at 37 C for 2 weeks. A defined medium employed by Rogosa et at. (1947) as an aid in the identification of lactobacilli has been used by us. Its composition is tabulated below. Casein hydrolyzate... 5 g Glucose g Solution A ml Solution B... 5 ml L-Asparagine mg L-Tryptophan mg L-Cystine... mg DL-Methionine... mg Cysteine... mg Ammonium citrate... 2 g Sodium acetate (anhydrous)... 6 g Adenine; guanine; xanthine; uracil; each mg Riboflavin; thiamine; pantothenate; niacin; each Pg Pyridoxamine ,g Pyridoxal... pg

4 546 ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN [VOL. 60 Pyridoxine... Inositol and choline, each... p-aminobenzoic acid... Biotin pg 10 mg 200 pg 5pg 3 pg Folic acid (synthetic)... Make up to 1 liter with distilled water. Solution A. K2HP04 and KH2PQ4, each 25 g, into distilled water to a volume of 250 ml. Solution B. FeSO4c7H20, 0.5 g; MnS4O2H20, 2.0 g; NaCl, 0.5 g; and MgS9047H20, 10 g. Dissolve in distilled water to a volume of 250 ml. The medium is dispensed in 8-ml amounts in test tubes after the ph is adjusted to 6.7 with sodium hydroxide. It is sterilized in the autoclave for 12 minutes. The 12-minute period begins when the temperature reaches 116 C; the final temperature is 121 C. Our lactobacilli strains were studied in this medium in the following manner: One-tenth ml of an 18-hour, 37 C, yeast extract glucose broth culture was inoculated into a tube of the defined medium, and when good growth (turbidity) appeared, 2 loopfuls were transferred to a fresh tube of the same medium. This process was repeated until the organism in question had passed serially through 5 tubes of the defined medium. Lactobacilli that could grow in the defined medium produced as great a turbidity and lowering of ph in the fifth tube as in the first and second tubes, whereas those that were unable to grow in this medium were not transferable beyond the third, or even the first, tube. When good growth appeared in the fifth tube, 2 loopfuls were transferred to a sixth and final tube of this medium; at the same time 2 loopfuls were transferred to a tube of medium lacking riboflavin to determine whether the organism required added riboflavin for growth. An attempt was then made to maintain the organism in the medium containing no added riboflavin. It was found that none of the cultures capable of growth in the defined medium with riboflavin were able to grow in this same medium in the absence of this vitamin. Incubation was always at 37 C, and continued for 2 weeks unless growth appeared before this time. RESULTS AND DISCUSSION Our cultures of facultative anaerobic lactobacilli have been identified as Lactobacilus plantarum, Lactobacillus acidophilus, and Lactobacillus fermenti. The majority of the cultures of L. plantarum and L. fermenti were isolated from aerobic heart infusion agar plates representing a 10-6 to 10-7 dilution of cecal feces. The cultures of L. acidophilus were recovered from anaerobic tomato juice glucose agar plates representing a dilution of 10-7 or more. The colonies were small and nearly always subsurface. L. acidophilus and L. fermenti will grow on aerobic slants, but produce better surface growth when incubated under a reduced oxygen tension. All cultures were gram-positive and were unable to reduce nitrate or produce catalase. The homofermentative rods (L. plantarum and L. acidophilus) produce little or no gas in the vaseline-sealed broth tubes; L. fermenti, on the other hand, produces considerable gas, the production of which forces the vaseline seal up from the surface of the medium. Shapiro and Sarles (1949) observed in the intestinal tract of baby chicks the presence of certain gas pro-

5 1950] LACTOBACILLI FROM TURKEYS 547 ducers that were not coliforms; it seems possible, in the light of our experience with turkeys, that these may be heterofermentative lactobacilli. The strains of L. fermenti studied form volatile acid in amounts only slightly greater than the homofermentative lactobacilli. The fermented glucose converted to volatile acid (calculated as acetic acid) by L. plantarum and L. acidophilu was with a single exception found to be less than 3 per cent; the volatile acidity produced by the two L. fermenti strains was somewhat over 4 per cent (table 1). In addition to the production of gas, the heterofermentative nature of these two strains is STRIN NUMKBE * GAS PlO- DUC- TION G G GROWTH IN TEE DEFINED CON- TAINING FLAVIN sm ILE Final ph 5.5 n.t n.t TABLE 1 Characteristics of lactobacilli from turkeys Per cent acid calculated 20 C 26 C as lactic acid 0.24 n.t n.t GROWTH AT 30 to 41 C 44.5 C C IC N ra 0 n.t PEE CE.NT FER- MENTED GLUCOSE CONVERTED TO OPTI- CAL Tm Volatile Nonvolacid atile LACTIC OF cac_ acid ACID lae scalcu. PROacetic lated as DUCED acid lactic acid n.t. n.t n.t I I I I spzazs Lactobacillu plantarum Lactobacillu= acidophilau Lactobacclu..fermenti, no gas; G, gas; n.t., not tested; -, no growth (no turbidity or lowering of ph);, growth (turbidity and loweing of ph); I, inactive lactic acid; D, dextro-lactic acid. I Lactobacilu pntarum var. mobilis. apparent from the fact that they convert less than 35 per cent of fermented glucose to nonvolatile acid (calculated as lactic acid). L. plantarum and L. acidophilus, on the other hand, convert between 70 and 85 per cent of fermented glucose to nonvolatile acid (table 1). The strains of L. fermenti (12-12 and 29-35) produce little or no growth in skim milk; they form inactive lactic acid from the fermentation of glucose; they do not grow in the defined medium (table 1); and they ferment fructose, glucose, mannose, galactose, sucrose, maltose, melibiose, lactose, and raffilose (table 2). They do not ferment glycerol, arabinose, rhamnose, sorbitol, mannitol, inositol, trehalose, cellobiose, melizitose, salicin, dextrin, or starch. Culture 12-12

6 548 ARTHUR P. HARRISON, JR., AND P-. ARNE HANSEN [VOL. 60 eq 0 H9 m ~ x 0. m CnQ 06 fa ZLYKLSgfls ON 001! " co~ 9 40 ;s rl p.9 P. 9 O -.% i P. -lb 40 ;s I S.' C- go co co t_ t- co co co eq C- Iq00t t c o tvs- C C t- t- co co 0to ECIXI 0 o! I" C i 0 C in a' co so t@ t- 46 H0000 q e CO CO ZSL 0 co co to to to go 40Yo co to co to co co cc; zso.azrxx rsuvco 00 t 1 't "t! 1 I C.! '9 I I! 16' C- 't C'- "' IR 1 1 0! W! I,q C- " XESO.IOVI 44^ lfl 4 fe 0000H444400O04600 H I co to t t4 zsoliurxzx ao 400 0! 0!-COC 0!@0!0!@1 CO CO 460 CO I1SOIs soot Ij t. t eq 00 -C -C 00 SOII^ ;2SOIVIHlUL i; u.9 r..9 P4 e0 00 o eq O 0V eq o o 00o 0 Ison t! e l 1 1 el 1 11 CI1 el 1 f It "t 1s t KSOxolI o q eq eq e e oeie ;s10 r.'- 1 4 t t c c w c C In! x b t t tb 11 "t '9 cl cl cl se t9 n 1 el "d ** zso. o o neso liluoi 00 eq 0 e C ORti 'I ti 'It CO t~ ti 'I co 0: 09 * CO co q 11 '9 t eq eq 0o zsom m 'I eq. eq0 000t " I 0 to 10 w co V Vl v e 10 co co v co co co co co -0 la go la o la to 44 &44 44 WI q0 4444eq44C--@00q 90 4C -- eq CO e CO zsoxlolao eq CO 0 CO eq tco O eq 0 ISOLISONI 001 sec o X 4 0e &.NV1cI '- -00 C-. -0 C- -0 C- C- C- -C t h z so a l a..o.a 703I.lf IC la o l to so soravxv 000 q SOAXI b We O f)l# b)lf 6044 ex e q e o.0co 0 X I S.g~ 04 ii 00e co _ _ es 0o 04 o o0

7 1950] LACTOBACILLI FROM TURKEYS 549 ferments xylose but culture does not. Although our cultures are now nonfermenters of dextrin, when originally tested in the yeast extract basal medium they were able to bring about a weak fermentation of this substrate. Incidentally, dextrin is not a well-defined carbohydrate but may vary in composition. A B :, 19* ork. Figure 1. Lactobacilli from turkeys. Twenty-eight-hour, 37 C, yeast extract glucose agar cultures stained with aqueous crystal violet (X 1,050). A: Lactobacillus fermenti, B: Lactobacillits acidophiluis, Microscopic examination of crystal-violet-stained films made from yeast extract glucose agar slants revealed rather long rods (0.7 to 0.8,u by 5 to 15,u) with blunt ends. The cells for the most part were rather straight, and chain formation was absent (figure 1A).

8 550 % XtIio o s S ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN J:1 C). it.:}bus~~~~~~~d.0 } IMi} -?s L 4i v... C S A. _lf : L...-,,. B : s i ( \ e_k. w ell k; _. E _K. as 4X4. i 'zsiw -i.s! b. a as' -i: 1 SPy.:.-*k. }6#Ss :: 9s ft Ax. i e....e si..s'\, k%l We are using the name Lactobacillus fermenti for cultures and although Beijerinck's (1901) original definition is too vague to permit any degree of certainty in determining the identity of the present strains with his "Lactobacillus fermentum." The lack of fermentation of arabinose, which is now Figure 2. [VOL. 60 Lactobacillus plantarum from turkeys. Twenty-eight-hour, 37 C, yeast extract glucose agar cultures stained with aqueous crystal violet (X 1,0o9). A and B: C: D: used as one of the decisive characters for placing strains within the species, was not considered by Beijerinck. However, the later emended and much more detailed description by Smit (1915), also of the Delft laboratory, agrees in the main with the strains here investigated except for the temperature relationships.

9 1950] LACTOBACILLI FROM TURKEYS 551 The bacterium of Smit developed between 18 C and 49 C. A detailed description of this lactobacillus species was given by Orla-Jensen (1919) under the name Betabacterium longum and by Pederson (1938), who considered this to be a synonym of fermentum and who found its maximum to be 48 to 50 C. The specific epithet was first given as fermentum, which was corrected in Bergey's Manual (Breed et al., 1939) to the genitive fermenti. Our strains of this organism have a temperature maximum for growth between 44 and 46 C. They grow faster at 41 C than at 30 C; no growth occurs at 20 C (table 1). These temperature relationships agree with those considered characteristic of Betabacterium longum by Orla-Jensen (1919). We are inclined to disregard the slight discrepancy between the growth ranges of Smit's and our organisms and rather emphasize their common inability to ferment arabinose and their positive fermentation of sucrose and raffinose; these fermentation characters were considered of importance in the heterofermentative group by Orla-Jensen (1919). Lactobacillus plantarum was isolated from 10 of the 12 turkeys sampled. The morphology varies from strain to strain. On yeast extract glucose agar slants, most cultures (typified by 29-34) contain cells of rather uniform size having rounded ends which average 0.8 Iu by 2 to 8,u; they occur singly, in pairs, and in trios, chains being absent (figure 2D). Other cultures (for instance, 7-41) are more pleomorphic with cells of varying length and diameter, averaging 0.6 to 1.0,u by 3 to 15,. In these cultures long filamentous cells may be observed, and there is a tendency toward chain formation (figure 2C). The cells of all the cultures show curving or bending to a greater or lesser degree. In some instances individual cells or a pair of cells curl up into ringlets, typified by culture 24-7 (figures 2A and 2B). The key to the genus Lactobacillus in Bergey's Manual (Breed et al., 1948, p. 350) leaves the impression that L. plantarum produces only inactive lactic acid. However, in the original description of the species by Orla-Jensen (1919) it was stated: "Plantarum most frequently forms pure inactive lactic acid, but can also form dextro-lactic acid; in some few cases, indeed even exclusively dextro-lactic acid." In the present study all L. plantarum strains produce dextrolactic acid in excess of the levo form. Since Lactobacillus casei also produces dextro-lactic acid, much interest has been displayed in finding additional differentiating characters for these species. Fermentation of melibiose is considered characteristic of L. plantarum; Orla-Jensen (1943) states that L. casei never ferments this carbohydrate. Our cultures ferment the following: glycerol (weakly), mannitol, fructose, glucose, mannose, galactose, sucrose, maltose, melibiose, lactose, raffinose, and dextrin; most ferment sorbitol, trehalose, and salicin; a few attack rhamnose, inositol, and cellobiose; but xylose, arabinose, melizitose, and starch are fermented only slightly or not at all. These reactions agree with those considered characteristic of L. plantarum by Orla-Jensen (1919, 1943) except for the occasional lack of salicin fermentation. All the strains studied by Orla-Jensen and by Tittsler et al. (1947) fermented salicin. In addition, the final ph produced in certain disaccharides can sometimes help differentiate between these two species, since it has been shown (Orla-

10 552 ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN [VOL. 60 Jensen, 1919) that lactose fermentation by L. casei usually results in a greater degree of acidity than the fermentation of maltose and sucrose, whereas in the case of L. plantarum lactose is generally fermented to no greater degree than these other two disaccharides. The final ph produced in maltose and sucrose by our cultures is almost invariably slightly lower than that produced in lactose (table 2). In addition to the characters just pointed out it may be observed that acid production in skim milk by L. casei is greater than that produced by L. plantarum, the former usually for g from 0.9 to 1.5 per cent acid (calculated as lactic acid) in this substrate (Orla-Jensen, 1919; Tittsler et al., 1947). However, there is admittedly too often an overlapping of the two ranges to give this test much practical value. Our cultures produce a comparatively low degree of acidity, between 0.10 and 0.61 per cent (table 1), in skim milk. The temperature limits allowing growth of our cultures differ from those considered characteristic for this species. Orla-Jensen (1919) states that as a rule the maximum temperature for growth of L. plantarum is 37.5 to 40 C, although a few of his strains grew at 45 C. The minimum temperature supporting growth was in most cases about 10 C. Tittsler et al. (1947) observed that this species usually grows at 16 C but not at 45 C. All of our cultures grow rapidly at 44.5 C and even fairly well at 45.5 C, but none show growth (turbidity or lowering of ph) at 20 C, even after incubation at this temperature for 2 weeks. Growth (turbidity) develops more rapidly at 41 C than at 30 C, and acid production is greater at 36.5 C than at 30 C. These facts indicate a higher optimum temperature for growth than is usually considered characteristic for this species, but this is not surprising when one considers the source of our cultures. Turkeys have an average body (rectal) temperature of 41.4 C (Marsden and Martin, 1939), and the mere establishment of the lactobacilli therein would indicate ability to grow at a temperature higher than that which will allow growth of most of Orla-Jensen's strains. Whether this organism is ingested by turkeys with grain and mash and adapts itself to the higher temperature of the bird's body is a matter of conjecture. Although our strains differed slightly from those studied by Orla-Jensen, they have nevertheless been considered to be L. plantarum. Incidentally, Pederson (1929) isolated strains of this species from spoiled tomato products which were able to grow at 45 C, and which grew most rapidly at a temperature of 30 to 37 C. The avian L. plantarum strains studied here do not grow in the defined medium unless riboflavin is added. Strain is of particular interest because it shows active motility in young cultures. The flagellation of this organism has already been discussed (Harrison and Hansen, 1950b). It is apparent from the data recorded in the accompanying tables that this one characteristic (motility) is its single differentiating feature. We feel that this organism is truly a LactobaciUo and have labeled it L. plantarum var. mobilis. Johnson and Pollard (1940) isolated a motile, gram-positive, nonsporeforming diplobacillus (from the tissues of a turkey) which they considered to be a lactobacillus and which, although incompletely described, they named

11 1950] LACTOBACILLI FROM TURKEYS 5,53 Lactobacillus meleagridis. These workers did not give data sufficient to prove it to be a member of this genus, however. Lactobacillus acidophilus was found to be the most numerous of the facultative anaerobic lactobacilli isolated, usually being encountered on plates representing dilutions higher than would yield the other two species. The strains isolated produce only low acidity in milk, the two representative cultures (12-35 and 2042) tabulated forming, respectively, 0.52 and 0.34 per cent acid (calculated as lactic acid). Inactive lactic acid is formed from the fermentation of glucose, and these strains do not grow in the defined medium with or without riboflavin. When grown on yeast extract glucose agar slants, these rods average 0.6 to 0.8 j, by 3 to 7,u, although some longer filamentous cells may also be seen. The rods are sometimes curved, and they usually occur singly or attached in twos and threes; no real chain formation occurs, however (figure 1B). Both strains grow between 26 C and 41 C, growth being more rapid in the high range. Strain grows at 44.5 C, whereas does not; neither grows at 45.5 C. This agrees with the findings of Curran et al. (1933), who observed that, although the maximum temperature for growth is usually 43 C, some strains will grow at 46 C, and even a few at 48 C. No growth (turbidity or lowering of ph) occurs at 20 C, and in this respect our cultures behave like those first isolated by Moro (1900a,b). In general, the more complete description given by Orla-Jensen et al. (1936) has been used for the identification of our strains. Orla-Jensen and co-workers stated that L. acidophilus may vary considerably in fermentation characters, not only according to the basic medium but from one experiment to another. Consequently fermentation characters should preferably be observed under varied conditions, and too great dependence on the significance of the behavior in a single carbohydrate must be warned against, though the whole "spectrum" of fermentation characters usually is helpful in placing the strain taxonomically. Our cultures ferment fructose, glucose, mannose, galactose, maltose, sucrose, and lactose; and glycerol, xylose, arabinose, rhamnose, sorbitol, mannitol, inositol, melizitose, and starch are not attacked (table 2). These reactions agree in the main with those considered characteristic of this species by Orla-Jensen et al. (1936) and by Tittsler et al. (1947). However, according to Orla-Jensen et al., cellobiose is not fermented by this species, but Tittsler et al. found it to be attacked, as did we. Fermentation of trehalose, melibiose, and salicin is variable according to Orla-Jensen et at. Our cultures ferment salicin but not trehalose. Although when last tested our cultures did not ferment melibiose, when first tested in the yeast extract basal medium both fermented this disaccharide, and one (20-42) was able to bring about a weak fermentation of starch. The ability to ferment melibiose and starch seems therefore to have been lost. It is interesting in this connection that Shapiro et al. (1949) divided into four types, on the basis of fermentation reactions, the unidentified facultative anaerobic lactobacilli they isolated from the intestinal tract of chickens, and there seem to be no discrepancies between the fermentation characters of the lactobacilli of their type III and our L. acidophilus cultures.

12 554 ARTHUR P. HARRISON, JR., AND P. ARNE HANSEN [VOL. 60 SUMMARY AND CONCLUSIONS Lactobacillus was found to be the predominating bacterial genus in the cecal feces of turkeys. Lactobacillus bifidus was the most numerous anaerobe and Lactobacillus acidophilus the most numerous facultative anaerobe in the turkeys tested. Lactobacillus plantarum and Lactobacillus fermenti were also encountered, but in smaller numbers than the other two species. Although L. bifidus and L. acidophilus isolated from the ceca of turkeys may be considered typical, some variation from the accepted description of the species was noticed in the physiological characters of the avian strains of L. plantarum: about 50 per cent of the strains did not ferment salicin; none of the strains would grow without added riboflavin; the maximum, optimum, and minimum temperatures were slightly higher, no cultures developing at 20 C; and one strain was motile. However, our L. plantarum strains agreed with the original definition in all major characters, and particularly in fermenting melibiose and in forming inactive lactic acid with a surplus of the dextro form. REFERENCES BEIJERINCK, M. W Sur les ferments lactiques de l'industrie. Arch. n6erland. sci., ser. 2, 6, BREED, R. S., MURRAY, E. G. D., AND HITCHENS, A. P Bergey's manual of determinative bacteriology. 5th ed. Williams & Wilkins Co., Baltimore, Md. BREED, R. S., MURRAY, E. G. D., AND HITCHENS, A. P Bergey's manual of determinative bacteriology. 6th ed. Williams & Wilkins Co., Baltimore, Md. CRECEI2US, H. G., AND RETTGER, L. F Intestinal flora of the guinea pig. J. Bact., 46, CURRAN, H. R., ROGERS, L. A., AND WHITTIER, E The distinguishing characteristics of Lactobacillus acidophilus. J. Bact., 25, EGGERTH, A. H The gram-positive non-spore-bearing anaerobic bacilli of human feces. J. Bact., 30, HARiuSON, A. P., JR., AND HANSEN, P. A. 1950a The bacterial flora of the cecal feces of healthy turkeys. J. Bact., 59, HARRISON, A. P., JR., AND HANSEN, P. A. 1950b A motile Lactobacillus from the cecal feces of turkeys. J. Bact., 59, JOHNSON, E. P., AND POLLARD, M Further observations on an organism in turkeys whose blood serums agglutinate Salmonella pullorum. J. Infectious Diseases, 66, KERTESZ, Z. I Recalculated tables for the determination of reducing sugars by Bertrand's method. The author, Geneva, N. Y. MARSDEN, S. J., AND MARTN, J. H Turkey management. 1st ed. The Interstate, Danville, fli. MORO, E. 1900a Ueber die nach Gram farbbaren Bacillen des Sauglingsstuhles. Wien. klin. Wochschr., 13, MoRo, E. 1900b Ueber den Bacillus acidophilus n. spec. Jahrb. Kinderheilk., 52, ORLA-JENSEN, S The lactic acid bacteria. Mem. acad. roy. sci. Danemark, Sect. sci., 8 ser., 5, ORLA-JENSEN, S The lactic acid bacteria. Erganzungsband. M6m. acad. roy. sci. Danemark, Sect. sci. biol., 2 (3), ORLA-JENSEN, S., ORLA-JENSEN, A. D., AND WINTHER, Bacterium bifidum and Thermobacterium intestinale. Zentr. Bakt. Parasitenk., II, 93,

13 1950] LACTOBACIMLI FROM TURKEYS 555 PEDERSON, C. S The types of organisms found in spoiled tomato products. N. Y. State Agr. Expt. Sta., Tech. Bull PEDERSON, C. S The gas-producing species of the genus Lactobacillus. J. Bact., 35, ROGOSA, M., TITTSLER, R. P., AND GEIB, D. S Correlation of vitamin requirements and cultural and biochemical characteristics of the genus Lactobacillus. J. Bact., 54, SHAPIRO, S. K., RHODES, R. A., AND SARLES, W. B Lactobacilli in the intestinal tract of the chicken. J. Bact., 58, SHAPIRO, S. K., AND SARLES, W. B Microorganisms in the intestinal tract of normal chickens. J. Bact., 58, S[IT, J Studien ilber Lactobacillus fermentum. Z. Garungsphysiol., 5, TISSIER, H Recherches sur la flore intestinale des nourrissons. Tbese pour le doctorat en m6decine, Facult6 de MWdecine de Paris, Paris. TITTSLER, R. P., GzIB, D. S., AND ROGOBA, M Taxonomy of the genus Lactobacillus, with special reference to correlations of differential characteristics. J. Bact., 54, TiTTsIxR, R. P., AND RoGOSA, M Personal communication. WEIss, J. E., AND RETTGER, L. F Lactobacillus bifidus. J. Bact., 28, Downloaded from on January 24, 2019 by guest