HeLa Cell Infection by Yersinia enterocolitica - Infection and Immunity

INFECTION AND IMMUNITY, Apr. 1981, p. 48-55 0019-9567/81/040048-08$02.00/0

Vol. 32, No. 1

HeLa Cell Infection by Yersinia enterocolitica: Evidence for Lack of Intracellular Multiplication and Development of a New Procedure for Quantitative Expression of Infectivity J. A. DEVENISH AND D. A. SCHIEMANN* Environmental Bacteriology Laboratory, Ontario Ministry of Health, Toronto, Ontario M5W 1R5, Canada

The in vitro invasive properties of bacteria have frequently been studied by the use of HeLa cell cultures in chamber slides, using microscopic examination to enumerate intracellular bacteria. When this system was used to examine invasive properties of Yersinia enterocolitica, it resulted in rapid internalization of high numbers of bacteria during the infection phase which prevented subsequent discrimination of intracellular multiplication. A modified procedure was developed which standardized the ratio of bacteria to HeLa cells (i.e., multiplicity), the time for the infection phase, and the addition of specific antiserum with gentamicin for restricting bacterial uptake during the intracellular growth phase. Studies with this modified chamber slide system found that strains of human isolates of Y. enterocolitica (serotypes 0:3, 0:8, 0:5,27, and 0:6,30) exhibited different degrees of cell infection but did not multiply intracellularly. A second test system was developed that used roller tubes and viable cell counts for enumeration of intracellular bacteria. This roller tube system confirmed that internalized bacteria did not multiply inside HeLa cells over a 24-h period. The roller tube system with viable cell counts for enumeration is a simplified technique for quantitative comparison of in vitro infectivity of HeLa cells by Y. enterocolitica. The pathogenesis of invasive bacteria such as shigellae begins with penetration of intestinal epithelial cells (13, 22) followed by intracellular multiplication (5, 6, 35). In vitro tissue culture invasiveness has been used to evaluate the pathogenic potential of various bacteria, including Mycobacterium spp. (29), Shigella flexneri (4,6, 8, 13, 22, 23), and, more recently, Escherichia coli (16), Salmonella typhimurium (7, 11, 12), and Yersinia pseudotuberculosis (27). Studies on the interaction of Y. enterocolitica with in vitro cell cultures have concluded that common human serotypes of this organism are also invasive (14, 15, 17, 25, 32, 33). Techniques for studying in vitro invasiveness by bacteria have been designed either to observe infection ability only or to determine that there has been intracellular multiplication after infection. To distinguish between intracellular multiplication and penetration or infection only of in vitro cell cultures, one needs a methodology

repeated washings of the cell monolayer. Intracellular bacteria are then confirmed either by microscopic examination of stained monolayers or by viable counts on cell lysates. Reported studies of in vitro cell invasiveness with Y. enterocolitica have sometimes used a two-phase procedure with chamber slides and microscopic verification of intracellular bacteria; these studies concluded that there had been intracellular multiplication (14, 19). We undertook an investigation of this two-phase method for the purpose of standardizing the test so that intracellular multiplication could be irrefutably distinguished from cell infection only. These studies led to the development of a roller tube method which permitted quantitative expression of the relative in vitro HeLa cell infection capability of Y. enterocolitica. These investigations were completed with human isolates of Y. enterocolitica representing serotypes 0:3, common in Canada (31) and many other countries (21, 37), that will separate the two processes into distinct 0:8, the predominant human type isolated in the phases. The infection phase must limit the num- United States (1, 36), and 0:5,27 and 0:6,30, also ber of intracellular bacteria to less than the isolated from humans in Canada but with a maximum which can be accommodated by the much lower frequency than 0:3 (31). animal cell without disruption, and the intracelMATERIALS AND METHODS lular growth phase (IGP) must include a provision for "blocking" continual infection by extraBacterial strains. All cultures of Y. enterocolitica cellular bacteria. This blocking is usually accom- were originally isolated from humans and were proplished by the addition of an antibiotic or by vided by S. Toma, Canadian National Reference Ser48

VOL. 32, 1981

HeLa INFECTION BY Y. ENTEROCOLITICA

vice for Yersina, Toronto, Ontario. The cultures represented two strains of serotype 0:3 (E546 and E549), two of serotype 0:8 (E543 and E557), two of serotype 0:5,27 (E544 and E555), and one of serotype 0:6,30 (E545). Strain M42-43 of Shigella flexneri 2a (identified as Sh228) was provided by S. Formal, Walter Reed Army Medical Center, Washington, D.C. The invasive capability of this strain was confirmed in our laboratory by a positive Sereny test. Bacterial cultures were prepared for testing by streaking Trypticase soy-yeast extract (0.6%; BBL Microbiology Systems) agar plates from stock slants of the same medium. Y. enterocolitica was incubated at either 32°C for 24 h or 22°C for 48 h for slide cultures and at 22°C only for suspension cultures. S. flexneri was incubated at 35°C for 18 to 20 h. At least 10 colonies were touched with a loop, and a light inoculum made in 5 ml of Trypticase soy-yeast extract broth. Broth cultures of Y. enterocolitica were incubated at 22°C for 18 to 22 h, and cultures of S. flexneri were incubated at 35°C for 18 to 20 h in roller tubes. The broth cultures were centrifuged (3,200 rpm for 20 min), and the recovered cells were washed twice with Earle balanced salt solution. A diluted suspension of the bacteria was prepared in Earle balanced salt solution containing 0.2% bovine albumin. The density of this suspension was determined by counting with a Petroff-Hausser chamber and then adjusted to provide the desired multiplicity (i.e., ratio of number of bacteria to number of HeLa cells). Microscopic counts were confirmed by colony counts of viable cells and used as the basis for calculating reported bacterial multiplicities. Slide cultures. HeLa cell cultures were maintained in Eagle minimal essential medium (MEM) with 10% fetal bovine serum (FBS), 50 IU of penicillin G per ml, and 50 .ug of streptomycin per ml, with incubation at 35°C under 5% CO2. Culture medium was removed and replaced with fresh medium 24 h before harvesting. Cells were recovered from culture flasks by removing the medium, washing once with Dulbecco phosphate-buffered saline, and then adding 5 ml of Ca- and Mg-free phosphate-buffered saline containing 0.25% trypsin. The density of the recovered HeLa cell suspension was determined by microscopic counting in a Fuchs-Rosenthal chamber. The concentration was adjusted to 6.4 x 104 cells per ml in antibiotic-free MEM-FBS, and 1.0 ml was added to each of the chambers on a Lab-Tek tissue culture four-chamber slide (Miles Laboratories, Inc.). Slides were incubated at 35°C under 5% CO2 for 24 h. Based on three different cell suspensions and counts on 29 chambers from 8 prepared slides, we determined that there were on the average 1.4 generations of the HeLa cells during 24 h of incubation. Consequently, the actual number of HeLa cells at the time of infection was 1.5 x 105 per chamber, and this value was used as the basis for calculating bacterial multiplicities. HeLa cell monolayers were prepared for infection by removing the medium and washing once with Earle balanced salt solution. The bacterial suspension (0.2 ml) in Earle balanced salt solution containing 0.2% bovine albumin was added to each chamber on the slide, which was then incubated at 35°C under 5% CO2. After the infection period, the medium containing

49

extracellular bacteria was removed and the infected monolayer was gently washed once with Earle balanced salt solution. One milliliter of MEM containing gentamicin (50 jg/ml) and 4% specific antiserum was added to each chamber, and the slide was reincubated. After the IGP, the medium was removed and the monolayer was washed three times with phosphatebuffered sline. The monolayer was fixed with methanol and stained with May-Griinwald and Giemsa stains for microscopic examination. Infection was quantitatively determined by counting the number of intracellular bacteria present in 100 HeLa cells selected at random from the stained monolayer. Suspension cultures. HeLa cells were recovered from culture flasks as previously described and suspended in MEM with Earle salts (GIBCO Laboratories). The density was adjusted to 2.5 x 105 cells per ml, and 2.0 ml of this suspension plus 0.05 ml of additional MEM with Earle salts was placed in tissue culture tubes (Coming Glass Works: polystyrene, 16 by 125 mm). Duplicate tubes were inoculated with 0.25 ml of a bacterial suspension prepared as previously described and containing 2.5 x 106 cells of Y. enterocolitica (multiplicity of 5) or 5 x 107 cells of S. flexneri (multiplicity of 100) based on microscopic counts. Multiplicities of Y. enterocolitica based on viable counts ranged from 4.4 to 6.8. The multiplicity of S. flexneri based on viable counts was equal to that derived from the microscopic count for the data reported herein. After an infection period at 35°C in a roller apparatus (30 min for Y. enterocolitica and 3 h for S. flexneri), two tubes were removed, and the infected cells and extracellular bacteria were recovered by centrifugation at 4°C (6,000 rpm for 10 min). The remaining tubes received 0.1 ml of undiluted specific antiserum and 0.1 ml of gentamicin (2.5 mg/ml) and were reincubated in the roller tube apparatus. The concentration of gentamicin was increased in suspension cultures from 50 Ag/ml used with slide cultures to 100 yig/ ml. This was done to further ensure that there would be no bacterial uptake after the infection period and that viable counts would represent intracellular bacteria only. The supernatant from centrifuged suspensions was removed, and 10 ml of 1.0% Tergitol 7 in phosphate-buffered saline (pH 7.2) warmed to 44.5°C was added to each tube. The tubes were placed in a side-arm shaker for 25 min at 35°C to allow for cell disruption. Two-milliliter samples from each tube were pooled, and viable counts were determined with triplicate spread plates for each dilution in an appropriate series. The first count at time zero represented the total intracellular and extracellular bacteria present after the infection period. Counts at various times during the IGP were determined in the same manner. All counts were expressed as the number of bacteria per milliliter of medium. Although counts were highly reproducible and relatively stable during the IGP for strains of serotypes 0:8 and 0:3, some variability was observed with strains of serotypes 0:5,27 (E544) and 0:6,30 (E545). To eliminate the possible influence of separate tubes at each time point on this variability and to provide a sound basis for interpreting whatever changes were occurring during the IGP, a single cell suspension was prepared in a spinner flask. To further reduce the possibility

50

DEVENISH AND SCHIEMANN

that the variability was due to a minimal infection, or that there was some threshold number of internalized bacteria required for either multiplication or a stable state, the multiplicity in this system was increased from 5 to 10. Because strain E545 (0:6,30) was weakly infective compared with other strains, the infection phase in the spinner flask was increased from 30 to 60 min. The initial volume in the spinner flask was 95 ml, permitting withdrawal of 2.0-ml portions at periodic intervals for viable counts. Cells were recovered as previously described except that cell disruption was accomplished by sonication of the Tergitol suspension. This technique improved cell lysis with no loss of viable bacteria in a much shorter time. This spinner flask system was used to describe changes in number of intracellular bacteria during the IGP for strains E544 (0:5,27) and E545 (0:6,30).

INFECT. IMMUN.

TABLE 1. Influence of multiplicity on HeLa cell infection rates (serotype)

1-6 39 46 36 36 51 61 8 8 5 53 29 29 20 45 27 10 8 33 45 51 28 57 44 35 22

0

E543 (0:8)

16.0 8.0 4.0 15.7 7.9 3.1 15.5 7.8 3.1 17.0 8.5 4.2 1.7 16.5 8.2 4.1 1.7 17.6 8.8 4.4 1.8 16.4 8.2 4.1 1.6

E644 (0:5,27)

E545 (0:6,30) E646 (0:3)

RESULTS Multiplicity. Initial studies of HeLa cell invasiveness by Y. enterocolitica were completed by the method described by Mehlman et al. (19) and Mehlman and Aulisio (18). We found that this procedure gave a very heavy monolayer of HeLa cells and that the expected multiplicity of 10 to 20, based on turbidimetric adjustment of the bacterial suspension and an infection period of 90 min, resulted in large numbers of intracellular bacteria with invasive strains of Y. enterocolitica. Consequently, discrimination of intracellular multiplication by microscopic counts was not possible. Further studies undertook to standardize the number of HeLa cells per chamber and the conditions of time and multiplicity in order to achieve a maximum number of infected cells with a minimum number of bacteria per cell. Holding the time constant at 30 min and varying the multiplicity gave the results shown in Table 1 and Fig. 1 and 2. Strains of serotypes 0:8 (E543 and E557) and 0:5,27 (E544 and E555) were highly infective, followed by strains of serotype 0:3 (E546 and E549). The strain of serotype 0:6,30 (E545) was only weakly infective. The results indicated that a multiplicity of about 10 was best with a 30-min time period for achieving low numbers of intemalized bacteria that would not confuse further distinction of intracellular multiplication. Bacterial uptake during IGP. To be certain that increases in number ofintracellular bacteria are actually due to multiplication, it is necessary to prevent continual uptake of extracellular bacteria after the infection phase. This is accomplished by removal of the infection phase medium containing part but not all of the extracellular bacteria and addition of new medium with an antibiotic for destruction of remaining extracellular bacteria. The test organism must, of

% of HeLa cells with given no. of bacteria/cell

Multiplicity

Culture no.

E549 (0:3)

E555

(0:5,27) E657 (0:8)

z

2

w 0 M. -I

A1

w M

9 27 62 15 28 35 92 92 95 45 71 71 80 52 72 90 92 18 15 39 71 32 53 65 78

6-10

>10

30 21 2 25 16 4 0 0 0 2 0 0 0 3 1 0 0 17 27 6 1 11 3 0 0

22 6 0 24 5 0 0 0 0 0 0 0 0 0 0 0 0 32 13 4 0 0 0 0 0

100 90 80 70 60 50 40

E 543 0

z

E 546 x E 549 E 557 A

w

0

5

10

15

20

MULTIPLICITY FIG. 1. Influence of multiplicity on HeLa cell infection by four strains of Y. enterocolitica representing serotypes 0:3 (E546, E549) and 0:8 (E543, E547). Infection time was 30 min. Degree of infection was determined by microscopic counts on stained monolayers.

51


VOL. 32, 1981 z

0 U IL

z

w -I '4

-I

I-z w

U

w a.

71 90 80 70 60 50 40 30

E 544 0 E 545 E 555 X

x E543 (0:8)

i

* E544 (0:5,27)

Ul '4

* E549 (0:3)

m 0

20 10

41 cc

0

5

10

15

20

MULTIPLICITY FIG. 2. Influence of multiplicity on HeLa cell infection by three strains of Y. enterocolitica representing serotypes 0:5,27 (E544, E555) and 0:6,30 (E545). Infection time was 30 min. Degree of infection was determined by microscopic counts on stained monolayers.

course, be susceptible to the antibiotic being used. This was confirmed for three of the test strains (E543, E544, and E549) by suspending them in MEM-FBS and making viable counts over time. The number of viable bacteria was reduced by more than 99.9% within the first 15 min of contact (Fig. 3). This reduced the multiplicity to a level where further infection of the HeLa cells would be barely detectable. The general susceptibility of Y. enterocolitica to gentamicin has been reported by others (10, 26). To further evaluate the effectiveness of gentamicin for restricting cell infection, HeLa cell monolayers in chamber slides were charged with a bacterial suspension in MEM at a density which gave a multiplicity of 10. Other chambers were charged with the same suspension containing gentamicin (50 ,g/ml) or with a suspension containing gentamicin (50 Ag/ml) plus 4% specific antiserum (rabbit). The chamber slides were held for 3 h at 35°C under 5% C02 and then stained for microscopic examination. Gentamicin alone was not capable of completely blocking cell infection during the IGP, but the combination of the antibiotic and specific antiserum was a significant improvement (Table 2). The use of antiserum alone had no effect on bacterial infection. The use of FBS instead of specific antiserum did not change the results obtained with gentamicin alone. From these results, we concluded that the system with antiserum was the more reliable, and it was adopted for all further studies.

0

30

60

90

120

TIME (MIN) FIG. 3. Susceptibility of three strains of Y. enterocolitica suspended in MEM-FBS to gentamicin (50 pg/ml) at 35°C.

Intracellular multiplication. Table 3 presents microscopic data relating to changes in number of intracellular bacteria during the IGP. The four strains of Y. enterocolitica examined did not multiply intracellularly. Further observations up to 24 h with strain E543 gave the same result. To support the conclusion that Y. enterocolitica does not multiply inside HeLa cells, a virulent strain of S. flexneri, previously reported to multiply in HeLa cells (8), was examined under essentially the same test conditions. Difficulties with loss of cells from the monolayer during washings and cytotoxicity made the results equivocal. Consequently, another system using a roller tube and viable cell counts was adopted as a more precise tool for studying intracellular multiplication. (i) Roller tube. Figure 4 presents the results from the roller tube system for three strains of Y. enterocolitica and one of S. flexneri. S. flexneri, although weakly infective, multiplied rapidly inside the HeLa cells, causing cell lysis with a decrease in number of viable bacteria as they were exposed to extracellular antibiotic. In marked contrast, the number of intracellular Y. enterocolitica cells remained essentially unchanged over 24 h. The relative difference in infectivity among the three strains of Y. enterocolitica agreed with the previous results obtained with the chamber slide (Fig. 1).

52


INFECT. IMMUN.

TABLE 2. Effect of specific antiserum with gentamicin on infection of HeLa cells by Y. enterocolitica % of HeLa no. (serotype)

8

celis with given

Control 13.5 Gentamicina 46.7 Gentamicin + 90.7 antiserum" E546 Control 29.0 (0:3) Gentamicin 55.3 Gentamicin + 85.7 antiserum E549 Control 29.0 (0:3) Gentamicin 59.3 Gentamicin + 82.0 antiserum E557 Control 36.0 43.7 (0:8) Gentamicin Gentamicin + 80.3 antiserum a Concentration, 50 ug/ml. bConcentration, 4.0%.

6-10

1-5

0

E543 (0:8)

>10

28.5 26.0 32.0 52.3 1.0 0.0 8.7 0.3 0.3 49.5 43.7 14.3

15.0 1.0 0.0

6.5 0.0 0.0

52.0 39.7 17.3

12.0 1.0 0.7

7.0 0.0 0.0

43.5 53.6 19.0

18.5 2.7 0.4

2.0 0.0 0.3

TABLE 3. Relative stability of number of Y. enterocolitica in HeLa cells during the IGP % of HeLa celLs with given no. of bacteria

Cuerotype) E543 (0:8) E544

(0:5,27) E545 (0:6,30) E546 (0:3)

7

no. of bacteria/cell

Condition

Condition

Infection, 30niin IGP, 2 h IGP,3h IGP, 6h Infection, 30 min IGP, 2 h IGP, 3 h IGP, 6h Infection, 30 min IGP, 2 h IGP, 3h IGP,6h Infection, 30 min IGP, 2h IGP, 3h IGP, 6h

per cell

0

1-5 6-10 >10

24 19 19 23 21 31 31 16 89

38 35 42 26 47 38 40 48 11 9 10 9 33 36 36 31

91 90 91 64 60 60 69

15 19 14 19 18 12 14 18 0 0 0 0 2 3 2 0

23 25 25 32 14 19 15 18 0 0 0 0 1 1

2 0

(ii) Antiserum. Recognizing that the roller tube system would be more acceptable if antiserum were not required, one test strain (E557) was examined with and without specific antiserum. The results were nearly identical to those shown in Fig. 4, indicating that the same relative degree of infection was obtained under both conditions and that the stability of intracellular bacteria was not changed. (iii) Antibiotic-free HeLa celis. There have been conflicting reports on the uptake of strep-

Uc1-I 6'I I-

5

0

0

4

-i

3

0 2 4

6 8 10 12 14 16 18 20 22 24

TIME (HOURS)

FIG. 4. Number of viable intracellular bacteria in HeLa cells suspended in roller tubes. Count at zero time represents total intra- and extracellular bacteria after a 30-min infection period. Strains represented are Y. enterocolitica serotypes 0:8 (E543, E557) and 0:3 (E546) and S. flexneri 2a (Sh228).

tomycin by animal cells and the effect of the antibiotic on intracellular bacteria (2, 3, 16, 24, 30). Although all cell cultures were prepared for infection in antibiotic-free medium, there remained a question of whether there could be a carryover of intracellular antibiotic accumulated during normal cell maintenance. To eliminate any possibility that bacteriostatic levels of intracellular antibiotic were the explanation for nonmultiplication of Y. enterocolitica, HeLa cells were cultured in antibiotic-free medium. Five subcultures were made before using the cells for infection by strain E557 (0:8). The results were the same as those shown in Fig. 4 and those of the experiment previously described on the effect of antiserum. (iv) Spinner flask. The results for strains E544 (0:5,27) and E545 (0:6,30) (Fig. 5) were obtained by use of a single cell suspension in a spinner flask instead of individual roller tubes. Strain E545 showed a lower degree of infection, and although there appeared to be an initial small increase in number, there was no further change nor any decrease that should eventually occur because multiplication had induced cell lysis. Strain E544 was highly infective but also showed no intracellular multiplication. Unlike all other test strains, E544 showed a slight decline in number of viable intracellular bacteria over 24 h. Reproducibility of infectivity index by roller tube. The ability to reproduce the infec-


VOL. 32, 1981

53

TABLE 4. Reproducibility of the HeLa cell infectivity index

7

00 TE (

E 544 2

)E 545 -

Culture no. Log bacteria/ml for trial no.: Mean ± stan(serotype) 2 3 4 5 dard deviation 1 4.14 4.01 4.09 4.17 4.26 4.13 ± 0.09 E674

(0:3)

E557

(0:8)

OII

E544

(0:5,27) E661

3 -fs

(0:8) E763

4.47 4.40 4.29 4.15

4.33 ± 0.14

4.72 5.20 5.41

5.11 ± 0.35

4.73 4.61

4.67 ± 0.08

3.31 3.02

3.16 ± 0.20

4.38 4.78

4.58 ± 0.28

3.61 4.20

3.90 ± 0.42

(0:16) E764

O

2 4

6

8 10 12 14 16 IS 20 22 24 TIMWE (HOURS)

FIG. 5. Number of viable intracellular bacteria in HeLa cells suspended in a spinner flask. Strains represented are Y. enterocolitica serotypes 0:5,27 (E544) and 0:6,30 (E545).

tivity index was determined by repeating the procedure with the same strain of Y. enterocolitica on different days. The results (Table 4) indicated a high reproducibility by the roller tube system. It is quite possible that a low infectivity index, as was demonstrated by strain E545, for example, does not truly represent intracellular bacteria but only extracellular survivors not destroyed by gentamicin. For this reason, all strains showing a high infectivity index by the roller tube system must be examined for antibiotic susceptibility. The fraction of bacteria that are extracellular and may survive exposure to gentamicin at 100 Ag/ml would not significantly change a high infectivity index that is expressed as a logarithm of a viable count.

(nontypable) E766 (0:1,2,3)

with the conclusions of other investigators who used slightly different techniques and criteria for judging invasiveness (27, 32, 33). The strain of serotype 0:6,30 included in our studies was only weakly infective. Lee et al. (14) have reported that a strain of this serotype was invasive for HeLa cells by their procedure. The infectivity which we observed correlates well with our recent report that guinea pig conjunctivitis is produced only by strains of serotypes 0:3, 0:8, and

0:5,27 (28). Although Y. enterocolitica is infective for HeLa celLs, our studies show that it, unlike S. flexneri, does not multiply intracellularly. Experiments with HeLa cells cultured in antibioticfree medium and the elimination of antiserum from the test system verified that this inability was not the result of either of these conditions. It is also unlikely that the absence of multiplication resulted from penetration of gentamicin, since S. flexneri, which was susceptible to this antibiotic, multiplied rapidly under the same conditions. Vaudaux and Waldvogel (34) found DISCUSSION that gentamicin does not readily penetrate epiThe word "invasive" has been used to describe thelial cells. Furthermore, if there was antibiotic the ability of certain bacteria to infect and mul- penetration, it is not likely that the number of tiply inside cultured animal cells. Studies with intemal bacteria would remain stable over the Y. enterocolitica have associated invasive prop- period of 24 h, as was the case. The stability of erties with the pathogenic potential of these this number of internal bacteria also verifies the organisms (14, 33). Microscopic enumeration of effectiveness of the antibiotic in preventing furintracellular bacteria and viable cell counts of ther uptake of extracellular bacteria after the disrupted HeLa cells reported here demonstrate infection phase. that strains of Y. enterocolitica isolated from The inability to multiply intracellularly does humans have the ability to infect but not to not seem related to virulence, since strains E546 multiply inside cultured HeLa cells. Thus the (0:3), E557 (0:8), and E549 (0:3) were previterm "infective" rather than "invasive" would ously shown to be capable of producing guinea be more appropriate for describing this ability pig conjunctivitis (28). Strains E557 and E549 of Y. enterocolitica. also bear the V and W virulence antigens, acThree of the four serotypes selected for study cording to an indirect calcium dependency test (0:3; 0:8; and 0:5,27) were infective for HeLa (28). The absence of intracellular multiplication cells, although to different degrees. This agrees in HeLa cells may not be highly significant to

54


interpretation of pathogenic capabilities. It may be that infection of HeLa cells by Y. enterocolitica represents only a measure of the bacterial cell's ability to physically attach, and that ingestion is then an endocytic process initiated by the HeLa cells, as was suggested by Hale et al. (9) for Shigella and Henle 407 cells. The infection by Y. enterocolitica of HeLa cells is, however, a far more rapid process than it is for Shigella. The use of specific antiserum with gentamicin in the IGP significantly reduced bacterial infection of the HeLa cells. The conditions for these experiments were more extreme than actual test conditions in that the number of extracellular bacteria was considerably higher. This number is reduced in the test procedure by physical removal with medium, and it is likely that the antibiotic then reduces the multiplicity quickly to a level where further cell infection is negligible. The results obtained with the roller tube showed that the presence of antiserum was not necessary to control continual infection during the IGP. The disagreement with the effect of antiserum in the chamber slide may be that the increased numbers observed microscopically with gentamicin alone were actually due to nonviable bacteria which would not be represented in viable counts on cell lysates from roller tubes. Pedersen et al. (25) have stated that ultraviolet light- and Formalin-treated cells of Y. enterocolitica do not lose their ability to penetrate HeLa cells. In any event, the use of specific antiserum with antibiotic in the roller tube technique is not considered essential. The Lab-Tek tissue culture chamber slide is a convenient tool for microscopic observation of infection, although the preparation of stained slides is somewhat tedious. The slide was not, however, useful for recovery of HeLa cells for completing viable bacteria counts, particularly because HeLa cells were easily removed during washing, especially if there had been any cytotoxic effects. There also seemed to be a consistent, unexplainable pattern of cell loss or lack of adherence under all conditions, with the highest numbers of HeLa cells remaining in the center and along the edges of the chamber. The roller tube offers an alternative system which is simpler, allows stricter control of multiplicity, and provides more reliable results through viable counts of intracellular bacteria. These viable counts provide a quantitative index for comparing the relative infection capability of Y. enterocolitica for HeLa cells. ACKNOWLEDGMENTS This work was supported in part by research grant PR-742 from the Ontario Ministry of Health.

INFECT. IMMUN. LITERATURE CITED 1. Bissett, M. L. 1976. Yersinia enterocolitica isolates from humans in California. J. Clin. Microbiol. 4:137-144. 2. Bonventre, P. F., R. Hayes, and J. Imhoff. 1967. Autoradiographic evidence for the impermeability of mouse peritoneal macrophages to tritiated streptomycin. J. Bacteriol. 93:445-450. 3. Bonventre, P. F., and J. G. Imhoff. 1970. Uptake of 3Hdihydrostreptomycin by macrophages in culture. Infect. Immun. 2:89-95. 4. Calabi, 0. 1970. In-vitro interaction of Shigella flexneri with leukocytes and HeLa cells. J. Infect. Dis. 122:1-9. 5. Formal, S. B., E. H. LaBrec, T. H. Kent, and S. Falkow. 1965. Abortive intestinal infection with an Escherichia coli-Shigella flexneri hybrid strain. J. Bacteriol. 89:1374-1382. 6. Gerber, D. F., and H. M. S. Watkins. 1961. Growth of shigellae in monolayer tissue.cultures. J. Bacteriol. 82: 815-822. 7. Giannella, R. A., 0. Washington, P. Gemski, and S. B. Formal. 1973. Invasion of HeLa cells by Salmonella typhimurium: a model for study of invasivness of Salmonella. J. Infect. Dis. 128:69-75. 8. Hale, T. L, and P. F. Bonventre. 1979. Shigella infection of Henle epithelial cells: role of the bacterium. Infect. Immun. 24:879-886. 9. Hale, T. L, R. E. Morris, and P. F. Bonventre. 1979. Shigella infection of Henle intestinal epithelial cells: role of the host cell. Infect. Immun. 24:887-894. 10. Hammberberg, S., S. Sorger, and M. I. Marks. 1977. Antimicrobial susceptibilities of Yersinia enterocolitica biotype 4, serotype 0:3. Antimicrob. Agents Chemother. 11:566-568. 11. Kiblstrdm, E. 1977. Infection of HeLa cells with Salnonella typhimnurium 395 MS and MR1O bacteria. Infect. Inmun. 17:290-295. 12. Kihlstrom, E., and IL Edebo. 1976. Association of viable and inactivated Salmonella typhimurium 395 MS and MR10 with HeLa cells. Infect. Immun. 14:851-857. 13. LaBrec, E. H., H. Schneider, T. J. Magnani, and S. B. Formal. 1964. Epithelial cell penetration as an essential step in the pathogenesis of bacillary dysentery. J. Bacteriol. 88:1503-1518. 14. Lee, W. H., P. P. McGrath, P. H. Carter, and E. L Eide. 1977. The ability of some Yersinia enterocolitica strains to invade HeLa cells. Can. J. Microbiol. 23: 1714-1722. 15. Miki, M., P. Gronroos, and T. Vesikari. 1978. In vitro invasiveness of Yersinia enterocolitica isolated from children with diarrhea. J. Infect. Dis. 138:677-680. 16. Mandell, G. L. 1973. Interaction of intraleukocytic bacteria and antibiotics. J. Clin. Invest. 52:1673-1679. 17. Maruyama, T., T. Une, and H. Zen-Yoji. 1979. Observations on the correlation between pathogenicity and serovars of Yersinia enterocolitica by the assay applying cell culture system and experimental mouse infection. Contrib. Microbiol. Immunol. 5:317-323. 18. Mehlman, I. J., and C. C. G. Aulisio. 1978. Yersinia enterocolitica, Yersinia pseudotuberculosis, and related bacteria, p. XVHI-l-XVIU-12. In Bacteriological analytical manual. Association of Official Analytical Chemists, Washington, D.C. 19. Mehlman, L. J., C. C. G. Aulisio, and A. C. Sanders. 1978. Problems in the recovery and identification of Yersinia Erom food. J. Assoc. Off. Anal. Chem. 61:761771. 20. Mehlman, I. J., E. L Eide, A. C. Sanders, M. Fishbein, and C. C. G. Aulisio. 1977. Methodology for recognition of invasive potential of Escherichia coli. J. Assoc. Off. Anal. Chem. 60:546-562. 21. Mollaret, H. H. 1971. L'infection humaine i "yersinia

VOL. 32, 1981

22.

23.

24.

25.

26.

27.

28.


enterocolitica" en 1970, a la lumiere de 642 cas recents. Pathol. Biol. 19:189-205. Ogawa, H., A. Nakamura, and R. Nakaya. 1968. Cinemicrographic study of tissue cell cultures infected with Shigella flexneri. Jpn. J. Med. Sci. Biol. 21:259273. Ogawa, H., A. Nakamura, R. Nakaya, K. Mise, S. Honjo, T. Taksaka, T. Fujiwara, and K. Imaizumi. 1967. Virulence and epithelial cell invasiveness of dysentery bacilli. Jpn. J. Med. Sci. Biol. 20:315-328. Osada, Y., M. Nakajo, T. Une, H. Ogawa, and Y. Oshima. 1972. Application of cell culture in studying antibacterial activity of rifampicin to ShigeUa and enteropathogenic Escherichia coli. Jpn. J. Microbiol. 16: 525-533. Pedersen, K. B., S. Winblad, and V. Bitsch. 1979. Studies on the interaction between different 0-serotypes of Yersinia enterocolitica and HeLa cells. Acta Pathol. Microbiol. Scand. 87:141-145. Raevuori, M., S. M. Harvey, M. J. Pickett, and W. J. Martin. 1978. Yersinia enterocolitica: in vitro antimicrobial susceptibility. Antimicrob. Agents Chemother. 13:888890. Richardson, M., and T. K. Harkness. 1970. Intracellular Pasteurella pseudotuberculosis: multiplication in cultured spleen and kidney cells. Infect. Immun. 2:631639. Schiemann, D. A., and J. A. Devenish. 1980. Virulence of Yersinia enterocolitica determined by lethality in Mongolian gerbils and by the Sereny test. Infect. Immun. 29:500-506.

55

29. Shepard, C. C. 1957. Growth characteristics of tubercle bacilli and certain other Mycobacteria in HeLa cells. J. Exp. Med. 105:39-47. 30. Showacre, J. L, H E. Hopps, H. G. du Buy, and J. E. Smadel. 1961. Effect of antibiotics on intracellular Salmonella typhosa. I. Demonstration by phase microscopy ofprompt inhibition of intracellular multiplication. J. Immunol. 87:153-161. 31. Toma, S., L Lafleur, and V. R. Deidrick. 1979. Canadian experience with Yersinia enterocolitica (19661977). Contrib. Microbiol. Immunol. 5:144-149. 32. Une, T. 1977. Studies on the pathogenicity of Yersinia enterocolitica. fi. Interaction with cultured cells in vitro. Microbiol. Immunol. 21:365-377. 33. Une, T. H. Zen-Yoji, T. Maruyama, and Y. Yanagawa. 1977. Correlation between epithelial cell infectivity in vitro and 0-antigen groups of Yersinia enterocolitica. Microbiol. Immunol. 21:727-729. 34. Vaudaux, P., and F. A. WaldvogeL 1979. Gentamicin antibacterial activity in the presence of human polymorphonuclear leukocytes. Antimicrob. Agents Chemother. 16:743-749. 35. Watkins, H. M. S. 1960. Some attributes of virulence in Shigella. Ann. N.Y. Acad. Sci. 88:1167-1186. 36. Weaver, R. E., and J. G. Jordan. 1973. Recent human isolates of Yersinia enterocolitica in the United States. Contrib. Microbiol. Immunol. 2:120-125. 37. Zen-Yoji, H., and T. Maruyama. 1972. The first successful isolations and identification of Yersinia enterocolitica from human cases in Japan. Jpn. J. Microbiol. 16:493-500.