SomeProperties of the Pili of Corynebacterium renale · diffusion method of Ouchterlony, and extraction of whole cell antigen by 1% sodium deoxycholate were described in aprecedingreport

JOURNAL OF BACTERIOLOGY, Mar. 1970, p. 1063-1069Copyright X 1970 American Society for Microbiology

Vol. 101, No. 3Printed in U.S.A.

Some Properties of the Pili of Corynebacteriumrenale

RYO YANAGAWA AND KOICHI OTSUKIDepartmenit ofHygienie antd Microbiology, Faculty of Veterinary Medicine, Hokkaido University, Sapporo, Japan

Received for publication 18 August 1969

Some properties of the pili of the gram-positive bacteria Corynebacterium renalewere described. A relationship was found between the morphological features of piliand the types of C. renale. Strains of types I and III usually possessed a small numberof pili, whereas those of type II possessed numerous pili. Thick and long bundles ofpili characteristic of C. renale were frequently observable in type II strains. Piliationof C. renale was stable under various cultural conditions. No ability to agglutinatered blood cells was demonstrated by piliated strains of C. renale. Pili were isolatedfrom the cells of C. renale and studied serologically by immunodiffusion. The pili ofa type II strain were serologically identical with the pili of another type II strain butnot with those of the strains belonging to types I and III. The pili were serologicallydistinct from the cell wall. The pili were broken into short pieces by boiling, but theirantigenicity was increased after boiling.

Filamentous appendages of bacteria, differentfrom flagella, were designated as fimbriae (11)or pili (1). Duguid et al. (11) reported that knownhemagglutinating ability of Escherichia coli wasdue to pili. Pili were also found in other genera,Shigella (9), Salmonella, Klebsiella, Proteus, andChromobacterium (10). Brinton and co-workersextended the study of pili electrophoretically(4), genetically (2, 3, 5, 6), biochemically (14),morphologically, and chemically (2). Thus, theproperties of bacterial pili have been considerablyclarified. But the species of bacteria which havebeen known to possess pili were all gram-negative(5, 12).In gram-positive bacteria, existence of pili in

Corynebacterium renale was recently reported (16).Since the presence of pili had not been known be-fore in gram-positive bacteria, it is of interestto study the properties of C. renale pili and tocompare them with those of gram-negative bacte-ria. The present report deals with some propertiesof C. renale pili.

MATERIALS AND METHODS

Microorganisms. A total of 63 strains of C. renalewere used. Of these, 55 were isolated in Japan andclassified serologically and biochemically into types1, Ht, and III (15). Four strains were given by J. E.Phillips of the Royal (Dick) School of VeterinaryStudies, Summerhall, Edinburgh; three strains wereprovided by the American Type Culture Collection,Rockville, Md. One strain isolated in Uruguay andobtained from T. Hiramune of the Hokkaido Branch

Laboratory, National Institute of Animal Health,Sapporo, Japan, was identified in our laboratory astype I. Four strains sent us from A. Vallee, InstitutPasteur, Paris, were identified as type III. Eight strains,strains 1, 2, 3, and 9 (type I), strains 35 and 45(type It), and strains 42 and 43 (type III), were usedas representatives of each type for the studies ofpiliation under various cultural conditions and fortests for agglutination. Nutrient agar and broth wereused as culture media for C. renale. Usually 1- or 2-day-old cultures were tested.

Detection of pili by electron microscopy. Pili wereexamined electron microscopically. Bacteria grown innutrient broth were fixed by adding osmium tetroxideto the broth culture to give a final concentration of0.5%. After fixing overnight at 4 C, the bacterialcells were sedimented by centrifugation and washedtwice with distilled water. Bacteria grown on agarmedium were suspended in distilled water, and,after fixing in final 0.5% osmium tetroxide overnightat 4 C, the cells were sedimented and washed like-wise. The washed bacteria were mounted on thecollodion grids and, after reducing the excess amountof the material by absorption with filter paper, wereshadowed with palladium. Later, the procedure offixation was frequently omitted because no markeddifference was observed between the pili of fixed andnonfixed bacteria. In such cases, the bacterial cellsgrown on agar medium were suspended in distilledwater and shadowed likewise. Preparations wereexamined in a JEM 7 electron microscope (JapanElectrical Optics Laboratory Co.).

Piliation in various cultural conditions. Piliation ofC. renale was studied under the following culturalconditions: incubation at 20, 37, and 40 C; incuba-tion for 5 days in a tube in an atmosphere of low

1063

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

YANAGAWA AND OTSUKI

oxygen and elevated carbon dioxide tension (13);incubation in a desiccator in which a candle wasburned to extinction; cultivation for 3 days in boiled25% calf serum with 0.5% glucose, mannose, mal-tose, galactose, sucrose, or arabinose; and cultivationon agar media containing various concentrations ofantibiotics, chloramphenicol (1.25 to 5.0 ,ug/ml),penicillin (0.2 to 0.5 unit/ml), streptomycin (0.625 to5.0 ,g/ml), and mitomycin C (0.08 to 0.32 lAg/ml).The concentrations of the antibiotics were the rangeof the maximum concentrations which permitted theeight strains to grow. Cultivation of strain 35 onagar medium with 10% homologous hyperimmuneserum was also tested. The hyperimmune rabbitserum used was that prepared for immunodiffusion.

Agglutination. Cow, pig, dog, cat, sheep, guineapig, chicken, and human red blood cells were used toexamine whether these red blood cells could be ag-glutinated by the cells of C. renale. Media and pHrange used for the hemagglutination test were citratebuffer (0.15 M citric acid, 0.15 M sodium citrate; pH5.0, 5.5, 6.0, 6.5, 7.0), 0.065 M phosphate buffer(pH 7.0, 7.2, 7.4, 7.6, 7.8, 8.0), Veronal buffer (0.15 Msodium barbital, 0.15 N HCl, pH 7.0, 7.4, 8.0), boricacid borax buffer (0.15 M boric acid, 0.15 M borax;pH 7.5, 8.0), and tris(hydroxymethyl)aminomethane(Tris) buffer (0.15 M Tris, 0.15 N HCl; pH 7.0, 7.4,8.0). The red blood cells were washed three timesand suspended in the above buffer solutions to obtainfinal 0.3% (v/v) concentration. HeLa cells cultivatedfor 3 days, dispersed by trypsinization, washed bycentrifugation, and resuspended in Hanks balancedsalt solution (8 X 105 cells per ml) were also usedfor detecting the possibility of agglutination by thecells of C. renale. In addition, bacterial cells of otherspecies, particularly those of gram-positive bacteriasuch as Staphylococcus aureus, Streptococcus dys-galactiae, Listeria monocytogenes, and Bacillussubtilis, and Proteus vulgaris, which is gram-negative,were used; cells were obtained from 2-day-old cultureson nutrient agar. These bacterial cells, as well as cellsof C. renale, were washed twice and resuspended inthe buffer solutions to obtain a turbidity equivalentto an optical density of 0.5 at 450 nm.

Agglutination tests were carried out by mixing0.5-ml amounts of cell suspensions of C. renale with0.5-ml amounts of the suspensions of red bloodcells, HeLa cells, or other bacterial cells in test tubes.These tubes were then incubated, respectively, for 2hr at 37 C, overnight at room temperature, or over-night at 4 C. Then the tubes were examined foragglutination.

Isolation of pili. About 10 g (wet weight) of C.renale grown on agar medium was collected, washedthree times with distilled water, resuspended in dis-tilled water, and agitated in a high-speed mixer for Smin, keeping the bacterial suspension cool (2). Aftercentrifugation at 7,000 X g for 30 min, the super-natant fluid was again centrifuged. The cell-freesupernatant fluid thus obtained was sedimented at40,000 X g for 60 min. The sediment containing pillwas suspended in 1 ml of distilled water, which wasused for electron microscopy, and used as pili anti-gen.

Immunodiffusion. The serum of a rabbit whichhad been immunized against the whole cells of C.renale strain 35, the piliated type II strain, was used.Antigens used were the pili antigens obtained fromfour strains, 9, 35, 42, and 46, and the whole cellantigen extracted from strain 35. The methods ofpreparation of the immune rabbit serum, the agar-gel-diffusion method of Ouchterlony, and extraction ofwhole cell antigen by 1% sodium deoxycholate weredescribed in a preceding report (15). Pili samples wereboiled for 20 min, and pili samples treated with 0.25%trypsin (EC 3.4.4.4; Difco) in 0.065 M phosphatebuffer (pH 7.8) and 0.25% Streptomyces protease(Prozyme: EC class 3.4.4; Kyowa Hakko KogyoCo., Tokyo) in the same buffer (pH 7.2) at 37 C for 1hr were also used.

RESULTSMorphological features of pili in relation to C.

renale types. Pili were detected in all 63 strains.No essential differences were found between thefeatures of pili of the fixed and unfixed bacteria.Features of pili were not the same among thethree types of C. renale. Pili of C. renale strainsbelonging to type I were small in number (Fig. 1and 2). Not all of the bacterial population waspiliated in the strains of type I, even after carefulfixation of bacteria with osmium tetroxide.Usually, the pili of type I strains were short. Onthe contrary, all the strains of type II possessednumerous long pili (Fig. 3 and 4). Strains of typeIII were variable; many of them possessed fewpili (Fig. 5), as did strains of type I, but somestrains of type III had many pili (Fig. 6).Thick and long bundles of pili, frequently

having a thickness of more than 0.4 Am and alength of more than 10 Aum, were common inthe strains of type II (Fig. 7) and in three strainsof type III which possessed many pili. Thesethick pili sometimes encircled the bacterial cell(Fig. 8). Such bundles of pili were rare in all thestrains of type I and the remaining strains of typeIII which possessed a small number of pili.

General features of pili in relation to C. renaletypes are summarized in Table 1.Among the eight strains frequently used in

the following biological studies, strains 1, 2, 3,and 9, which belonged to type I, possessed a smallnumber of pili; strains 35 and 45 (type II) andstrains 42 and 43 (type III) possessed numerousp..Piliation of C. renale under various cultural

conditions. The original features of the pili ofthe eight strains were unchanged when thesestrains were serially subcultivated in nutrientbroth or on agar plates.

In nutrient broth, strains of type I possessinga small number of pili and those of type II whichpossessed numerous pili formed surface pellicle.

1064 J. BACTrERIOL.

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PILL OF C. RENALE

FIG. 1. C. renale strain 9 (type 1). X 34,000.FIG. 2. C. renale ATCC 10849 (type I). X 48,000.FIG. 3. C. renale strain 35 (type 11). X 40,000.FIG. 4. C. renale strain 46 (type 11). X 17,000.

VOL. 101, 1970 1065

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

YANAGAWA AND OTSUKI

FIG. 5. C. renale strain 48 (type III). X 34,000.FiG. 6. C. renale strain 42 (type IIl). X 28,000.FIG. 7. C. renale strain 35 (type II). X 26,000.FiG. 8. C. renale strain 35 (type II). X 25,600.

1066 J. BACTERIOL.

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PILI OF C. RENALE

In addition, no pellicle was formed in the culturesof type III strains possessing numerous pili.These findings indicated that piliation did notcorrelate with pellicle formation.

Piliation was not changed when the eightstrains were cultivated in an atmosphere ofdecreased oxygen and elevated carbon dioxideconcentration. Growth at 20, 37, and 40 C didnot cause change in piliation. Growth in themedia with one of six sugars, mannose, glucose,maltose, galactose, sucrose, or arabinose, didnot induce alteration in piliation. Chlorampheni-col, penicillin, streptomycin, and mitomycin Cdid not change piliation. Strain 35 inoculatedon agar medium containing 10% homologoushyperimmune rabbit serum grew rather profuselyand showed no change in the features of pili.

Tests for hemagglutination and agglutinationof other cells by C. renale. It is well known inEnterobacteriaceae that piliated strains aggluti-nate various kinds of cells. The possibility ofagglutination was studied with the eight strainsof the three types of C. renale. The cells usedwere of human, cow, pig, chicken, cat, dog, rab-bit, sheep, and guinea pig red blood cells, HeLacells, and the bacterial cells described above.No agglutination of red blood cells was observed,even in the pH range from 5.0 to 8.0 with thevarious buffer solutions. HeLa cells and bacterialcells were not agglutinated by the cells of C.renale, but were agglutinated by P. vulgaris.Microscopic examination showed that cells ofC. renale did not adhere to these various cells.

Serological properties of C. renale pili. Sero-logical properties of C. renale pili were studiedby immunodiffusion. Prior to the study, it wasnecessary to isolate pili as described above.The mixer treatment, conducted for 5 min,

was effective in removing pili from the cells ofC. renale. Longer treatments, such as 30 min,resulted in breaking the pili into pieces rangingfrom about 20 nm to several times that length.Such long blender treatment also resulted incontamination by the cell wall antigen.The isolated pili of strain 35 obtained by 5 min

of mixer treatment are shown in Fig. 9. Long piliare attached side by side and form thick bundles.Boiling for 10 min broke the long pili into shortpieces (Fig. 10).Immunodiffusion of rabbit antiserum to whole

cells of strain 35 with the homologous whole cellantigen and the pili antigens from the four strainsis shown in Fig. 11. The pili of strain 35 (well 1)gave a precipitin line which was characteristicin that the line formed close to the antigen welland extended from there in the direction oppositeto the serum well. The pili of strain 35 boiledfor 20 min (well 2) formed a similar precipitinline, but the line was stronger and wider. Another

TABLE 1. Features of pili in relationi toC. renale types

Type of Pili numerous, and Pili not numerous,C. renale thick and long pili and thick and long

frequent pili infrequent

I 0 30aII 19 0III 3 11

a Number of strains examined.

line was produced when deoxycholate extract ofwhole cells of strain 35 (well 3) was examined.This line, considered to be the reaction of cellwall antigen, appeared between the antigen andserum wells and was different from the reactionof the pili. The same figure shows that the piliof strain 46, another strain of type II (well 4),gave a precipitin line common to strain 35.However, the pili from other strains of differenttypes of C. renale such as I (well 5) and III (well6) gave no reactions.

Effects of two proteolytic enzymes, trypsinand Prozyme, on pili were also examined. Piliwere considerably degraded morphologicallyby these enzymes. However, the enzyme-treatedpili still formed a line similar to that of theoriginal pili.

DISCUSSION

Cells of C. renale, causative agent of bovinepyelonephritis and cystitis, are unique amonggram-positive bacteria in that they possess pili.Pili have been known as fine fibrillar appendagesof certain bacteria, but hitherto were knownonly in gram-negative species (5, 12).

Morphological features of pili were not thesame among three types of C. renale. C. renalewas classified into three types serologically andbiochemically (15). The relation of features ofpili to C. renale types is interesting. The bundlesof pili, which most frequently appeared in thestrains of type II, are considered to be charac-teristic to C. renale. Such bundles of pili havenot been described in gram-negative bacteriaThe diameter of an individual pilus of C. renale

was, regardless of type, about 2.5 to 3 nm (16),the diameter similar to that of type III pili of thegram-negative bacteria described by Brinton (2).Thicker pili of C. renale were the bundles of thefine pili.

Pili of C. renale were present when the bacteriawere grown in an atmosphere of low 02 andelevated CO2 concentrations or at 20, 37, and40 C. The presence of sugars including mannose,which inhibits piliation of some gram-negativebacteria (9), antibiotics, and homologous anti-bacterial serum did not affect piliation of C.

1067VOL. 101, 1970

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

YANAGAWA AND OTSUKI

FIG. 9. Pili of C. renale strain 35. X 40,000.FIG. 10. Pili of C. renale strain 35. Boiledfor 10 min. X 40,000.

FiG. 11. Immunodiffusion of rabbit antiserum towhole C. renale strain 35 (type II; well 7) with thefollowing antigens: well 1, pili of strain 35; well 2,boiled pili of strain 35; well 3, deoxycholate extract ofwhole cells of strain 35; well 4, pili of strain 46 (typeII); well 5, boiled pili of strain 9 (type I); and well 6,boiled pili of strain 42 (type IlI). The photograph was

taken at 15th day.

renale. Pili were present as long as C. renalegrew. Thus, piliation of C. renale is stable.

Pili of C. renale differ from the pili of manygram-negative bacteria in that they do not adhereto red blood cells. In many gram-negative bac-teria, piliated strains can easily be recognized by asimple hemagglutination test, for they adhere tored blood cells and cause hemagglutination (7-9,11). No hemagglutination was caused by anystrain of C. renale.Although adhesive properties were not clearly

shown in this report, pili of C. renale, particularlythose of type II strains, showed a tendency toattach side by side and to form thick bundles.The adhesive property of C. renale pili needsfurther investigation.The three types of C. renale possess cell wall

and pili antigens, and the cell wall and piliantigens are peculiar to each type.

Serological studies of type II pili showed thatantigenic specificity of pili was different fromthat of cell walls. Also antigenicity of pili isspecific of each type of C. renale. C. renale strainswere classified into three types, primarily basedon the antigens extracted with deoxycholate fromcell wall (15).

1068 J. BACTERIOL.

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

VOL. 101, 1970 PILI OF C.

The peculiar position of the precipitin lineformed by pili antigen might be attributed to thelong filamentous shape of isolated pili whichdiffuse slowly in gels. After boiling, the pilifraction gave a wider and stronger precipitinline. On examination with an electron micro-scope, such boiled pili were not intact filamentsbut were broken into segments, which diffusedmore rapidly in agar gel. From the fact that C.renale pili, broken morphologically by boiling orby the action of proteolytic enzymes, were stillserologically active, the pilus is thought to consistof small antigenic subunits assembled into afilament.

Strains of type II, which possessed the mostnumerous pili, were frequently isolated fromapparently healthy cattle, whereas strains oftypes I and III, which possessed rather few pili,were frequently isolated from diseased cattle (15).Strains of type II are not very pathogenic but areparasitic in the urinary tract. Possibly these cellsattach themselves to the cells of the urinary tractso that they are not easily ejected outside theanimals' body. Presumably, C. renale pili arehelpful for this attachment. But this is only a

speculation at present. Connected with thisspeculation, it is interesting that saprophyticKlebsiella are piliated but pathogenic Klebsiellaare not piliated (7).At present, it is almost impossible for us to

give an adequate idea of the functions of C. renalepili. Except for the function of F pili (5), the func-tions of pili in gram-negative bacteria are said toremain still to be discovered (8).

ACKNOWLEDGMENTS

We thank Y. Mifune for electron microscopy and YumikoFukagawa for technical assistance.

LITERATURE CITED

1. Brinton, C. C., Jr. 1959. Non-flagellar appendages of bacteria.Nature (London) 183:782-786.

RENALE 1069

2. Brinton, C. C., Jr. 1965. The structure, function, synthesisand genetic control of bacterial pili and a molecular modelfor DNA and RNA transport in gram-negative bacteria.Trans. N.Y. Acad. Sci. Ser. II 27:1003-1054.

3. Brinton, C. C., Jr., and L. S. Baron. 1960. Transfer of piliationfrom Escherichia coli to Salmo,zella typhosa by geneticrecombination. Biochim. Biophys. Acta 42:298-311.

4. Brinton, C. C., Jr., A. Buzzell, and M. A. Lauffer. 1954.Electrophoresis and phage susceptibility studies on a fila-ment-producing variant of the E. coli B bacterium. Biochim.Biophys. Acta 15:533-542.

5. Brinton, C. C., Jr., P. Gemski, Jr., and J. Carnahan. 1964.A new type of bacterial pilus genetically controlled by thefertility factor of E. coli K 12 and its role in chromosometransfer. Proc. Nat. Acad. Sci. U.S.A. 52:776-783.

6. Brinton, C. C., Jr., P. Gemski, Jr., S. Falkow, and L. S. Baron.1961. Location of the piliation factor on the chromosome ofEscherichia coli. Biochem. Biophys. Res. Commun. 5:293-298.

7. Duguid, J. P. 1959. Fimbriae and adhesive properties inKlebsiella strains. J. Gen. Microbiol. 21:271-286.

8. Duguid, J. P., E. S. Anderson, and I. Campbell. 1966. Fimbriaeand adhesive properties in Salmonella. J. Pathol. Bacteriol.92:107-138.

9. Duguid, J. P., and R. R. Gillies. 1957. Fimbriae and adhesiveproperties in dysentery bacilli. J. Pathol. Bacteriol. 74:397-411.

10. Duguid, J. P., and R. R. Gillies. 1958. Fimbriae and hemag-glutinating activity in Salmonella, Klebsiella, Proteus andChromobacterium. J. Pathol. Bacteriol. 75:519-520.

11. Duguid, J. P., I. W. Smith, G. Dempster, and P. N. Edmunds.1955. Non-flagellar filamentous appendages ("Fimbriae")and hemagglutinating activity in Bacterium coli. J. Pathol.Bacteriol. 70:335-348.

12. Reynolds, P. E. 1968. The bacterial cell: major structures, p.73-135. In J. Mandelstam and K. McQuillen (ed.), Bio-chemistry of bacterial growth. Blackwell Scientific Publica-tions, Oxford and Edinburgh.

13. Shoetensack, M. 1930. A simple and practical method for theisolation of anaerobic microbes. Kitasato Arch. Exp. Med.7:310-313.

14. Wohlhieter, J. A., C. C. Brinton, Jr., and L. S. Baron. 1962.Utilization of carbohydrates and metabolic intermediates bypiliated and nonpiliated bacteria. J. Bacteriol. 84:416-421.

15. Yanagawa, R., H. Basri, and K. Otsuki. 1967. Three types ofCorynebacterium renale classified by precipitin reactions ingels. Jap. J. Vet. Res. 15:111-120.

16. Yanagawa, R., K. Otsuki, and T. Tokui. 1968. Electronmicroscopy of fine structure of Corynebacterium renalewith special reference to pili. Jap. J. Vet. Res. 16:31-38.

on October 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from