JOURNAL OF BACTERIOLOGY, Oct 1971, p. 535-544 Copyright O 1971 American Society for Microbiology

Vol. 108, No. I Printed in U.S.A.

Electron Microscopy of Tissue Culture Cells Infected with Brucella abortus BETTY A. HATTEN, SHYI YI HUANG, M. L. SCHULZE, AND S. EDWARD SULKIN Department of Microbiology, The University of Texas (Southwestern), Medical School at Dallas, Dallas, Texas 75235

Received for publication 15 July 1971

Thin sections of hamster kidney tissue cultures were examined by electron mi7-day period after infection with Brucella abortus 3183. Numerous bacteria and structures resembling L-forms were present both intracellularly and extracellularly after the first 24 hr of infection. Most intracellular microorganisms were enclosed by a cytoplasmic membrane, but in a few instances no limiting membrane was detected. After 4 to 7 days, fewer microorganisms were present, and most normal-appearing bacteria were intracellular, particularly in antibiotictreated cultures. Structures typical of Brucella L-forms were extracellular at the latter time intervals. Several structures were observed in cells from infected cultures whose relationship to the infecting organisms is not known. These consisted of various membranous structures within cytoplasmic vacuoles, myelin-like structures surrounding occasional intracellular organisms, and small bodies present within vacuoles and extracellularly. The latter structures observed throughout the experimental period appeared to occur more frequently as the duration of the infection increased. croscopy over a

Intracellular production of Brucella L-forms has been suggested by light microscopy studies of hamster kidney cells infected with Brucella abortus 3183 (10, 12). A direct relationship of presumably altered forms to the bacterial strain used to infect the tissue cultures was established by specific immunofluorescent staining. Recovery of L-forms from infected tissue culture cells provided additional indirect evidence that such forms might be present, existing perhaps for several days after elimination of bacteria from cell cultures by addition of antibiotics to the maintenance medium (11, 12). Nevertheless, it is possible that ilitracellular fluorescence observed in tissue cultures infected with B. abortus was caused by accumulations of antigens and not by microorganisms, per se, and also that L-forms growing in subcultures transferred from infected tissue cultures to artificial media were induced by recovery procedures. Because of the significant function that L-forms might have in perpetuating Brucella infections, direct evidence regarding their presence or absence in infected tissue culture cells was imperative. Consequently, an electron microscope examination was made of infected tissue cultures containing cell types representative of those present after establishment of naturally occurring infections. Results of a preliminary study have been reported (12), and

our conclusions are drawn from more extensive studies of infected hamster kidney cell cultures. All microorganisms which did not appear to be typical bacteria are designated as altered forms. Organisms resembling those previously observed in penicillin-treated, artificial-medium cultures (9) are specifically referred to as L-forms in some instances.

535

MATERIALS AND METHODS Microorganisms. Stock cultures of B. abortus 3183 were maintained on Trypticase Soy Agar (BBL) slants at 4 C. Hamster kidney cell cultures were infected with bacterial organisms grown in Trypticase Soy Broth (BBL) under CO2 tension at 37 C by a procedure previously described (10). Hamster-kidney cell cultures. Hamster kidney cell cultures were grown and maintained at pH 7.4 in Hanks 0.5% lactalbumin hydrolysate medium with 5.0% calf serum added. Cells in monolayers grown in Leighton tubes were infected by 5-hr exposure to 1.6 x 109 bacteria suspended in Hanks balanced salt solution. Sufficient numbers of cells were used so that organisms could be found in thin sections throughout the 7-day experimental period. Antibiotics were removed from the tissue culture medium several days before the culture was infected, so that bacterial growth would not be inhibited nor L-forms induced by residual antibiotics in the cells. After infection, the monolayers were maintained in medium without antibiotics, in medium containing 5.0 or 10.0 gg of penicillin/ml, or in medium

A-RNW '11.,i

;

IW

#I

qe

t5~~~~~~~~~~~~.~ tINfrF F.Y-%it

FIG. 1. Thin sections of uninfected hamster kidney monolayers showing some cellular structures of the two predominant cell types. a, Nuclear region (N), mitochondria (M), lipid bodies (L), cisternae (C), ribosomes (R), and pinocytotic processes (P) as they appeared in one cell type. b, Nuclear region (N), mitochondria (M), lipid bodies (L), lysosomes (Ly), granular (G) and agranular (A) reticuli, ribosomal aggregates (R), and fibrillar structure of another cell type (F). 536

%k

40

.1-

40

I

'I";,

..

"

*

-,

Iti.-

,

'k

-

CV

e

*+

q4

Q0.2 pm

FIG. 2. Thin sections of hamster kidney cells showing intracellular bacteria 24 hr after infection. a, Bacterium (B) which appears to be dividing within a cytoplasmic vacuole of a cell from a monolayer treated with 5.0 1ig of penicillinlml. b, Bacteria (B) surrounded by a thickened cytoplasmic membrane (CM) and myelin-like structures (My) in a cell from a monolayer treated with 5.0 ,ug each ofpenicillin and streptomycin/ml. c, Bacteria (B), possibly an L-form (LF), and a membranous structure (MS) which appear to be lying free in the cytoplasm adjacent to the nuclear membrane (NM) and between two cytoplasmic vacuoles (V) enclosed by granular reticuli (G) in a cell from a monolayer treated with 5.0 Ag each of penicillin and streptomycin/ml. 537

c

K'l

11

PI.1

--l-

4

4b

,V" :

i

-

.

I

.0 I

..Z . .k,ip,o

i.l

'L.I'

L

Le ~~

0.2

FIG. 3. Thin sections of hamster kidney cells showing possible L-forms in cytoplasmic vacuoles 24 hr after infection. a, Several cytoplasmic vacuoles containing variable-appearing forms (VF) in addition to a bacterium (B) in a cell from a monolayer treated with 2.5 Asg each of penicillin and streptomycin/ml. b, Portion of a cell from a monolayer treated with 5.0 Aig of penicillinlml whose cytoplasm was filled with vacuoles containing many 538

VOL. 108, 1971

CELLS INFECTED WITH B. ABORTUS

containing 2.5 or 5.0 /Lg each of penicillin and streptomycin/mi. Uninfected cultures served as controls, showing cellular structures as they normally occurred under the experimental conditions employed. Approximately 95% of untreated tissue cultures infected in this manner yielded positive cultures when transferred to artificial media as previously described (10). Treatment with 5.0 ug of penicillin/ml reduced only slightly the percentage of cultures from which Lforms and bacteria could be recovered. L-forms were recovered from approximately 85, 79, and 67%, and bacteria from 54, 37, and 32% of those cultures exposed for 1, 4, and 7 days, respectively, to 10.0 jig of penicillin/ml, or to 2.5 or 5.0 jg each of penicillin and streptomycin/ ml. Electron microscopy. Tissue cultures were prepared for electron microscopy at 1, 4, and 7 days after infection. These time intervals were selected for this study because organisms present in the hamster kidney cell cultures should be representative of those occurring early in the infection, or after establishment of infection for intermediate and prolonged periods of time. Monolayers were washed several times with physiological saline, scraped from the glass surfaces, and sedimented by slow-speed centrifugation. The pellets were fixed and dehydrated by a glutaraldehyde-osmic acid technique similar to that outlined by Anderson et al. (1). Methods and equipment used for cutting, staining, and examining the specimen were the same as previously described (9).

RESULTS Uninfected hamster kidney cells. Most cells observed in the hamster kidney tissue cultures were of two general types. In one type, the nuclear material was evenly dispersed with some marginal concentrations at the inner surface of the membrane (Fig. la). The cytoplasm contained numerous mitochondria, lipid bodies, cisternae, and ribosomes. Most cells had numerous pinocytotic processes at the surface. In the second, most prevalent type, the nucleus was irregularly shaped and contained aggregates of nuclear material in addition to heavy deposits along the nuclear membrane (Fig. lb). Cytoplasmic regions of the latter cells were filled with many agranular reticula and vacuoles bounded by granular reticula, as well as numerous large mitochondria, some lipoidal and lysosomal bodies, occasional ribosomal aggregates, and fibrous structures. Internal structures of these cells did not vary appreciably in uninfected control cultures during the 7-day experimental period. Infected hamster kidney cells. Intracellular bacterial organisms were generally enclosed in a

539

cytoplasmic membrane. The appearance of the bacteria was not unusual, and some apparently were capable of cell division (Fig. 2a). Sometimes the enclosing membrane was thickened, and myelin-like structures surrounded the bacteria (Fig. 2b). Vacuoles rarely contained more than six to eight bacteria and frequently had only one to three organisms. In a few instances, no cellular membrane could be detected around the microorganisms. Figure 2c shows several bacteria and another structure, possibly an L-form, which appear to lie within the cytoplasm, close to the nuclear membrane. Other forms, along with bacterial organisms, sometimes were observed in vacuoles (Fig. 3a). Occasionally the cytoplasm was filled with vacuoles containing such structures, which consisted of large forms of varying size and shape, and small round bodies (Fig. 3b). Some larger forms contained internal membranous structures, dense granular material, and fibrous structures (Fig. 3c, d, and e), and in these respects resembled Brucella L-forms induced by penicillin in artificial media (9). Small bodies seen in vacuoles of infected cells also resembled those occurring in Brucella L-form cultures. Additional vacuoles present in infected cells contained a variety of membranous structures and small bodies less readily related to the infecting organisms (Fig. 4a). When these vacuoles, or inclusions, were found near the nucleus, they appeared to cause slight to marked indentations in the nuclear membrane. The cytoplasms of a few cells were almost completely filled with vacuoles which varied considerably in size. Occasionally, the vacuoles were continuous with the extracellular spaces, allowing their contents (large, complex forms, membranous structures, and small bodies) to merge with microorganisms outside the cell (Fig. 4b). Extracellular microorganisms. Both bacterial and altered forms were observed extracellularly in infected tissue cultures. Their occurrence and appearance varied from large clusters of apparently normal bacteria in varying numbers (Fig. 5a) to small groups of organisms in which few typical bacteria were present (Fig. 5b). Membranous structures and other features of some extracellular organisms (Fig. 5c) were typical of Brucelia L-forms induced by penicillin in artificial medium cultures (9). Relationship of cultural conditions to occurrence of various organisms. Microorganisms and

variable forms (VF) and small bodies (SB). c, Small bodies (SB) and a variable form (VF) with internal membranous structures (IM) in a cell from a monolayer treated with 5.0 Ag of penicillin/ml. d, Variable form with membrane-enclosed area containing dense granular material (GM) in a cell from a monolayer treated with 5.0 ,g of penicillin/mi. e, Variable form containing fibrous structures (FS) in a cell from a monolayer treated with 5.0 ug of penicillin/ml.

a

lw I.W.A?

r

11

Ms

4

t

44 .d.0

0.5 pm h

FIG. 4. Thin sections of hamster kidney cells showing a variety of membranous structures and small bodies present in cytoplasmic vacuoles 24 hr after infection. a, Numerous membranous structures (MS) in a cytoplasmic vacuole (V) lying close to the nuclear membrane (NM) in a cell from a monolayer treated with 10.0 Ag of penicillin/ml. b, Many membranous structures (MS), small bodies (SB), and variable forms (VF) in a vacuole that is continuous with the extracellular spaces around a cell from a monolayer not treated with antibiotics. 540

VOL. 108, 1971

CELLS INFECTED WITH B. ABORTUS

other intracellular structures apparently related to the infection of hamster kidney cells with B. abortus were present within 24 hr after infection and are representative of those seen throughout the 7-day experimental period. The occurrence of various forms and structures depended on the duration of the infection, and, to an extent, on the presence or absence of antibiotics in the medium. Table 1 shows data from one series of sections of infected hamster kidney cells cultured for various periods of time under various conditions. Figures 2 and 5 show organisms referred to as bacterial forms; Fig. 3 and 5 contain organisms referred to as altered forms. Other forms (Fig. 3 and 5), which could not be classified as bacteria or L-forms because of either the level at which they were sectioned or their state of degeneration, are termed unidentified forms. Membranous structures and small bodies are not included. The number of intracellular bacterial organisms decreased as the duration of infection increased, and they were rarely found in antibiotic-treated cultures by the 7th day. Differences in the frequencies with which other forms occurred in cultures treated with different combinations or concentrations of antibiotics used in this study probably are not significant. Large intracellular forms that structurally resembled Brucella L-forms were not observed after longer periods of incubation; but membranous structures and small bodies were more common. No variation in number or type was noted among forms and structures occurring in the cell types present in these cultures. Numerous areas containing many extracellular microorganisms were observed 1 day after infection of hamster kidney cell cultures with B. abortus 3183. Most of th-ese organisms appeared to be normal bacteria in cultures where no antibiotics had been added to the medium. But in those tissue culture media that contained antibiotics, numerous altered microorganisms were present (Table 1). The number of extracellular organisms decreased as the duration of infection increased, particularly in antibiotic-treated cultures. The number of altered forms seen in the antibiotic-treated cultures increased until, by the

7th day of infection, typical bacterial organisms were rarely seen. DISCUSSION Electron microscopy studies such as those presented in this paper are by no means quantitative, and their inability to detect specific structures does not exclude the possibility that these structures are present. However, these studies do provide an overall estimate of the relative proportions of different forms present under the var-

541

ious conditions described. The results of this investigation indicate that bacteria and, possibly, L-forms were present intracellularly after short periods of infection, but that probably only bacteria remained after longer periods. Localization of most microorganisms within cytoplasmic vacuoles agrees with similar reports concerning the relationship of other Brucella species and strains of B. abortus to the host cell (13, 18). Failure to observe an increase in number of intracellular bacteria as the duration of infection increased suggests that these organisms may have become dormant or may have been destroyed by cellular enzymes, by antibiotics present in the medium, or by a combination of these. An interplay of these processes may be indicated by: (i) presence of a few intracellular bacteria throughout the experimental period, (ii) occurrence of intracellular structures resembling L-forms, and (iii) presence of a variety of membranous structures and small bodies possibly resulting from partial breakdown of the infecting organisms. Apparent lack of intracellular structures resembling Lforms after longer periods of infection indicates that such structures may occur intracellularly only as transitory or degenerative forms. In contrast, the number of altered forms, some of which appeared to be L-forms, tended to increase extracellularly as the duration of infection increased, particularly under conditions adverse to bacterial survival. Whether similar results could be obtained with other strains of B. abortus is not known. McCune et al. (16) described induction of a sterile state in tubercle bacilli by treatment of infected animals with certain drugs (15, 16). This phenomenon is thought to account for microbial persistence in tuberculosis, as some bacteria could be revived after existing for many months in this state. Observation of fewer bacteria and fewer structures resembling L-forms as the duration of infection increased, if not resulting from a dormant state, must have been caused by destruction of most of these organisms before a large number of replications had occurred. Previous studies, however, have shown that some viable brucellae are present for at least 14 days after infection of hamster kidney cell cultures treated with antibiotics (10, 12). The origin of small bodies present in considerable numbers and extracellularly both intracellularly throughout the 7-day incubation period is unknown. If these small bodies are related to the infecting organisms, their capacity to initiate or perpetuate infection is questionable, as conflicting results have been obtained regarding viability of such structures in L-form cultures of other bacterial species (3, 4, 6, 21 ).

b

T.. +s,

\-

C.

, %

Ww

H

...

.1

.,

s

.x.

-d R.tOF ' ;\

c

^; *

t

@

wft

'#.t w

0.2 zm

FIG. 5. Thin sections of hamster kidney cells showing extracellular organisms present in infected monolayers. a, Altered bacterial forms from a monolayer treated with 5.0 ,ug of penicillinlml that are representative of those present in antibiotic-treated tissue cultures 24 hr after infection. b, Altered bacterial forms from a monolayer treated with 10.0 gg of penicillin/ml that are representative of those present in antibiotic-treated tissue cultures 7 days after infection. c, Altered bacterial forms from a monolayer treated for 4 days with 5.0 gg each ofpenicillin and streptomycinlml that are representative of those in which the varied sizes and shapes, internal fibrous structures (FS), and membranes (M) appear to be typical of Brucella L-forms. 542

VOL. 108, 1971

543

CELLS INFECTED WITH B. ABORTUS

Membranous structures have been demonstrated in association with intracellular infections by other microorganisms (7, 19, 22) and were mentioned in connection with infection of monocytes with B. abortus over a 48-hr incubation period (13). In cells infected with Japanese encephalitis virus, these structures were thought to represent lipids or lipoproteins released by breakdown of cytoplasmic organelles, as such products were known to form lamellar structures under certain conditions (22). In other instances, membranous or myelin-like structures appear to be produced by the cell, possibly in an attempt to isolate the intruding organism and subsequently destroy it (19). The latter response was observed in some cells infected with B. abortus. In other instances, vacuoles containing predominantly membranous structures present after longer periods of infection were of a similar size and occurred with the same frequency as inclusions which stained specifically with fluorescein-conjugated antisera (10, 12). Regardless of origin, the contents present in the vacuoles (other than identifiable microorganisms) probably consist of material that has been partially digested by hydrolytic enzymes released from lysosomes stimulated by the infection. It seems logical to assume that L-forms having little or no cell wall for protection would be more susceptible to lysis by cellular enzymes than would bacterial organisms. In contrast, Lforms usually are less susceptible to antibiotics because they lack certain cell wall constituents and, probably, because they are less active metabolically than bacterial organisms. Although most antibiotics can penetrate cellular membranes, it is unlikely that internal and external concentrations are equilibrated. These factors may account for intracellular localization of most bacteria after 4 to 7 days of infection, whereas organisms resembling L-forms generally were extracellular at these times. Presence of serum in the tissue culture medium could have contributed to extracellular occurrence of altered forms, as it is well known that normal sera promote protoplast formation (17). Any further influence that the presence of antiserum might have on induction and survival of such forms is not known; this could depend on the antigenic relationship of altered forms to the parent bacteria and on the diversity of antigenic modifications among them. Some L-forms and their parent bacteria have been found to be very similar antigenically, and others have little or no antigenic relationship (8, 14, 20). Whether Lforms could perpetuate an infection in the event that reversions could not occur would be contingent on viability of the L-forms and on the pres-

TABLE 1. Occurrence of various types of intracellular

and extracellular organismsa a _

T

'25 E |

24hr

None

4 days

5.0 P 10.0 P 2.5 PS 5.0 PS None OP io.o P 5.0 PS 10.0 P

7 days

E) "

5/90 80/52 0/50 37/30 47/104 26/14 24/46/45/128 40/20 75/12/2/22 100/14 1/70 100/7 9/100 79/7 2/32 100/9

E

0/28 20/20 70/27 30/23 14/43 60/43 16/- 38/13/61 47/19 0/- 25/0/64 0/22 0/66 0/27 0/75 21/18 0/69 0/22

Broth cultures 4 days None -/9175 /58 -/14 -/28 10.0 P -/69 -/1 -/83 -/16 a Data compiled from sections of infected hamster kidney cell cultured under different conditions; bacterial and altered forms shown in Fig. 2-5. bp, micrograms of penicillin per milliliter; PS, micrograms each of penicillin and streptomycin per milliliter. cIntracellular organisms/extracellular organisms.

ence of components associated with virulence of the bacterial organisms. Viability as well as antigenic and toxic properties of Brucella L-forms induced in artificial media under several conditions are being examined. ACKNOWLEDGMENT This investigation was supported by Public Health Service grant A -08913 from the National Institute of Allergy and Infectious Diseases. LITERATURE CITED

i. Anderson, D. R., H. E. Hopps, M. F. Barile, and B. Bemheim. 1965. Comparison of the ultrastructure of several rickettsiae, ornithosis virus, and mycoplasma in tissue culture. J . Bacteriol. 90:1387-1404. 2. Ayoub, E. M., and J. G. White. 1969. Intraphagocytic degradation of group A streptococci: electron microscopic studies. J. Bacteriol. 98:728-736. 3. Boven, C. P. A. van, H. L. Ensering, and W. Hijmans. 1968. Size determination by the filtration method of the reproductive elements of group A streptococcal L-forms. J. Gen. Microbiol. 52:403-412. 4. Boven, C. P. A. van, H. L. Ensering, and W. Hijmans. 1968. Size determination by phase microscopy of the reproductive elements of group A streptococcal L-forms. J. Gen. Microbiol. 52:413-420. 5. Chapman, G. B., J. H. Hanks, and J. H. Wallace. 1959. An electron microscope study of the disposition and fine structure of Mycobacterium lepraemurium in mouse spleen. J. Bacteriol. 77:205-2141. 6. Cole, R. 1967. Discussion. Mycoplasma and L-forms. Ann. N.Y. Acad. Sci. 143:813-823.

544

HATTEN ET AL.

7. Dales, S. 1965. Penetration of animal viruses into cells. Prog. Med. Virol. 7:1-44. 8. Dienes, L., H. J. Weinberger, and S. Madoff. 1950. Serological reactions of L type cultures isolated from Proteus. Proc. Soc. Exp. Biol. Med. 75:405-412. 9. Hatten, B. A., M. L. Schulze, S. Y. Huang, and S. E. Sulkin. 1969. Ultrastructure of Brucella abortus L-forms induced by penicillin in a liquid and in a semisolid medium. J. Bacteriol. 99:611-618. 10. Hatten, B. A., and S. E. Sulkin. 1966. Intracellular production of Brucella L forms. I. Recovery of L-forms from tissue culture cells infected with Brucella abortus. J. Bacteriol. 91:285-296. 11. Hatten, B. A., and S. E. Sulkin. 1966. Intracellular production of Brucella L forms. II. Induction and survival of Brucella abortus L-forms in tissue culture. J. Bacteriol. 91:14-20. 12. Hatten, B. A., and S. E. Sulkin. 1968. Possible role of Brucella L-forms in the pathogenesis of brucellosis, p. 457-471. In L. B. Guze (ed.), Microbial protoplasts, spheroplasts, and L-forms. The Williams & Wilkins Co., Baltimore. 13. Karlsbad, G., R. W. I. Kessel, S. de Petris, and L. Monaco. 1964. Electron microscope observations of Brucella abortus grown within monocytes in vitro. J. Gen. Microbiol. 35:383-390. 14. Lynn, R. J., and G. J. Haller. 1968. Bacterial L-forms as immunogenic agents, p. 270-278. In L. B. Guze (ed.), Microbial protoplasts, spheroplasts, and L-forms. The

J. BACTERIOL.

Williams & Wilkins Co., Baltimore. 15. McCune, R., F. M. Feldman, H. Lambert, and W. McDermott. 1966. Microbial persistence. 1. The capacity of tubercle bacilli to survive sterilization in mouse tissues. J. Exp. Med. 123:445-468. 16. McCune, R., F. M. Feldman, and W. McDermott. 1966. Microbial persistence. II. Characteristics of the sterile state of tubercle bacilli. J. Exp. Med. 123:469-486. 17. Muschel, L. H., W. F. Carey, and L. S. Baron. 1959. Formation of bacterial protoplasts by serum components. J. Immunol. 82:38. 18. Pearson, G. R., B. A. Freeman, and W. D. Hines. 1963. Thin-section electron micrographs of monocytes infected with Brucella suis. J. Bacteriol. 86:1123-1125. 19. Scanga, F. 1964. Atlas of electron microscopy, p. 38. Elsevier, New York. 20. Weibull, C., W. D. Bickel, W. T. Haskins, K. C. Milner, and E. Ribi. 1967. Chemical, biological, and structural properties of stable Proteus L forms and their parent bacteria. J. Bacteriol. 93:1143-1159. 21. Wyrick, P. B., and H. Gooder. 1971. Growth of streptococcal protoplasts and L-colonies on membrane filters. J. Bacteriol. 105:646-656. 22. Yasuzumi, G., and I. Tsubo. 1965. Analysis of the development of Japanese B encephalitis (JBE) virus. III. Electron microscope studies on inclusion bodies appearing in neurons and microglial cells infected with JBE virus. J. Ultrastruct. Res. 12:317-327.