Microbial Pathogenesis 1990 ; 9 : 47-53

Effect of motility and chemotaxis on the invasion of Salmonella typhimurium into HeLa cells T . Khoramian-Falsafi, l S . Harayama,2 K . Kutsukake3 and J . C . Pechere l 'Department of Microbiology and 2 Department of Medical Biochemistry, University Medical Center, Geneva, Switzerland,• 'Department of Biology, Faculty of Science, University of Tokyo, Hongo Tokyo 113, Japan (Received April 24,1990 ; accepted May 8, 1990)

Khoramian-Falsafi, T . (Department of Microbiology, University Medical Center, Geneva, Switzerland), S . Harayama, K . Kutsukake and J . C . Pechere . Effect of motility and chemotaxis on the invasion of Salmonella typhimurium into HeLa cells . Microbial Pathogenesis 1990 ; 9 : 47-53 . Salmonella typhimurium strain LT2 is able to invade HeLa cells in vitro . The effect of the motility and chemotaxis of the bacteria on cell invasion were examined by two methods : (1) conventional invasion assays where the HeLa cell monolayers were placed horizontally at the bottom of plastic wells and (2) vertical assays where the HeLa cell monolayer attached to one face of plastic bottles was placed vertically . In both assays, the invasion rate of the wild-type strain was higher than that in isogenic non-motile mutants . There was no significant difference between the invasion rate of non-flagellated mutants and that of a flagellated but non-motile mutant . These observations indicated that the motility per se increases the rate of the bacterial invasion by increasing the chance of encounter between Salmonella and the HeLa cells . Smooth-swimming non-chemotactic mutants exhibited 10 times higher invasion rates than the wild-type strain in conventional assays but their invasion rates in vertical assays were approximately equal to that of the wild-type strain . This result indicated that in the conventional assays, the migration of the wild-type bacteria towards the HeLa cells was hampered by their chemotactic responses . Tumbly non-chemotactic mutants exhibited invasion rates intermediate between the wild-type and non-motile strains presumably because of their intermediate net speeds of migration . Key words : Salmonella typhimurium ; invasion ; motility; chemotaxis ; infection .

Introduction Bacteria are exposed to a wide range of environments, in which the motility and chemotaxis towards favorable environments may be important factors to increase their chance of survival . Whether motility and chemotaxis constitute one of the virulence factors has been examined in several pathogenic bacteria . In Vibrio cholerae 01, nonmotile mutants exhibit diminished virulence ." Flagella and motility are important for the invasive virulence of Pseudomonas aeruginosa in burned mice .' In these examples, motility per se seems to be important for the pathogenicity . A different result was obtained with Salmonella typhimurium. Using mutants defective in flagellar synthesis, motility or chemotaxis, Carsiotis et a/. 5 have demonstrated that flagella but not motility are important for the virulence of S . typhimurium . Furthermore, they showed that flagella improved the survival of the pathogen with murine macrophages .' However they reported' in a recent publication that the attenuated virulence of non-flagellate mutants was associated with the loss of a virulence gene mviS adjacent to the f/g 0882-4010/90/070047+07 $03 .00/0

© 1990 Academic Press Limited

48

T. Khoramian-Falsafi eta/.

genes and not with the non-flagellate phenotype . Recently, Lockman and Curtiss' have demonstrated that a defect in the motility reduces the ability of S. typhimurium to invade Henle cells in vitro but does not affect its virulence for mice . Here, we have studied the effect of motility and chemotaxis on the invasiveness of S. typhimurium into HeLa cells using a large collection of f/a and the mutants, which were placed in various experimental conditions . Results Control experiments Gentamicin at a concentration of 200 yg/mI in Dulbecco minimal essential medium supplemented with 2% fetal calf serum (FCS) efficiently killed SJW1 103, a derivative of S. typhimurium LT2, and gentamicin-resistant derivatives of SJW1 103 were never isolated . We therefore assumed that the bacteria surviving gentamicin treatment in the invasion assay described in the Materials and methods section represented mostly bacteria internalized into HeLa cells as shown in our previous studies . 9' 10 There was no significant difference in the growth rates in the minimal essential medium containing 2% FCS of S. typhimurium SJW1 103 or its fla, mot and the derivatives used in this study. Microscopic observation showed that treatment with 0 .1 % Triton X-100 completely lysed the HeLa cells . The viability of S. typhimurium, in contrast, was not affected by Triton X-100 at that concentration . Therefore, treatment with Triton X-100 allowed the release of internalized bacteria from HeLa cells, and allowed counting of the internalized bacteria as colonies formed on antibiotic-free plates . The viability of the HeLa cells was estimated by microscopic observation after trypan blue staining : more than 80% of the HeLa cells were viable after 3 h invasion experiments . Invasion of wild-type S . typhimurium Figure 1 demonstrates the time course of appearance of a bacterial population protected from gentamicin killing . In the wild-type SJW1103 strain, this population started to increase linearly following a lag period of about 45 min, and reached around 6 x 10 4 cells per well after 3 h . Microscopic examination of the HeLa cells incubated with Salmonella for 90 min showed that most of the HeLa cells contained only few bacteria, generally less than five . Invasion of non-motile mutants When mutants, KK1015, SJW1 400, SJW1 386 and SJW1 350, defective in motility or flagellar synthesis, were incubated with HeLa cells for up to 90 min, very few bacteria invaded the HeLa cells . The number of internalized bacteria started to increase if the incubation time was extended over a second hour (Fig . 1) . The invasion rates in the mutant strains were significantly lower than those of the wild-type strain especially after short incubation times (0 .5 to 2 h) . All non-motile mutants showed similar invasion rates (non-significant differences) . Invasion of the mutants We used a set of the derivatives of SJW1 103 to examine the effect of the mutations on invasion of Salmonella into HeLa cells . As shown in Fig . 1, ail the the mutants that are 'smooth swimmers' (cheA, chew, cheY and cheR) invaded the HeLa cells efficiently . Surprisingly, this efficiency was about 10 times higher than that of the wild-type strain . The cheB 'tumbly' mutant in contrast was as poor an invader as the non-motile mutants . The cheZ mutant which is classed as tumbly was not a good

49

Motility and invasion in Salmonella 60000

60000

SJ W 1103 (wild-type)

40000

60000

KKIU15 (flo-25 )

40000

40000

A

20000

20000 . 0

2

3

0

20000 n

4

3

60000

60000

60000

40000

40000

40000

20000

20000

20000

3

2

3

0

HB101 (E. COO

0

I

2

3

2

3

a 600000

500000

600000

400000

400000

400000

200000

200000

200000

U

0

G 2

3

0

600000,

60000

60000

400000

40000

40000

200000

20000

20000

0

SJW3100 (cher)

I 2 Time(h)

0

I

SJ W 3037 (cheZ )

1

Fig . 1 . Invasion kinetics in conventional invasion assays . The S. typhimurium strains were incubated with the HeLa cells for the periods indicated in this figure, then incubated with gentamicin for 1 .5 h . The ordinates are cfu per well after the gentamicin treatment . The following numbers of bacteria were employed for the inocula : SJW1103, 3 .3x10 6 /ml ; KK1015, 3.3x10 6/ml, SJW1400, 2.3x10 6/ml ; SJW1386, 4 .0x106/ml ; SJW1350, 3 .0x10 6 /ml ; HB101, 2.7 x10 6 /ml ; SJW3111, 4 .Oxl06 /ml ; SJW3113, 2 .4 x106 /ml ; SJW2905, 3 .3 x 106/ml ; SJW3112, 4.5 x 106/ml ; SJW3100, 4 .0 x 106 /mI and SJW3037, 0 .9 x 10 6 /ml .

invader in the first 2 h of incubation, but the number of invaded bacteria increased dramatically after that time .

Vertical invasion assay In the conventional invasion assay, as used above, the HeLa cell monolayer was spread horizontally at the bottom of microtiter well plates . Part of the contact between bacteria and the HeLa cells in this assay may be simply due to sedimentation of the bacteria onto the HeLa cell monolayer . To avoid such a gravity effect, we designed experiments in which the HeLa cell monolayer was placed vertically . As shown in Fig . 2, the difference in the efficiency of the invasion between the wild type and non-motile mutants was much more pronounced in this vertical assay . The results in Fig . 2 showed that the invasion by the non-motile mutants observed in the conventional assay (Fig . 1) was in fact mostly due to precipitation of these mutants onto the monolayer of the HeLa cells . We also examined the invasion of chemotactic mutants in the vertical assay . From the results in Table 1, we concluded that (1) the invasion of non-motile mutants was

50

T . Khoramian-Falsafi et al. 2500000

2000000

s o

1500000

u

1000000

0

500000

0

2

3

Fig . 2 . Invasion kinetics in vertical assays . The ordinate is cfu per ml protected from the gentamicin killing . (o) SJW1 103 (wild-type) ; (0) SJW1350 (fliC1001) ; (0) KK1015 (fla-25 ) . The inocula were 3 .8x 108 /ml, 3 .6x 108/ml and 7 .1 x 10' ml for SJW1 104, SJW1 350 and KK1015, respectively .

Table 1 Invasion of the wild-type, non-motile and non-chemotactic Salmonella into the HeLa cells in vertical invasion assays S. typhimurium strain

SJW1 103 (wild type) KK1015 (f/a-25) SJ W 1400 (f/hD2040) SJW1353 (fliC1001) SJ W 1386 (mot) SJW3100 (cheB) SJ W3037 (cheZ) SJW3111 (cheA)

Inoculum/ml

Mean CFU/ml

2 .5 x 10' 5 .5x108 1 .5 x 108 1 .2x108 1 .O X 108 3 .O x 10' 2 .5 x 10' 6 .0x10'

1 .5x 105 3 .5x103 2 .5 x 103 5 .6x103 1 .2x 103 1 .3 x 104 2 .6 x 105 2 .5x105

The HeLa cells were incubated with the bacterial strains for 2 h, then with gentamicin (200,ug/ml) for 1 .5 h as described in vertical invasion assay in the Materials and methods section .

much less efficient than that of the wild-type strain ; (2) the smooth-swimming cheA mutant was as proficient as the wild-type strain for invasion ; (3) invasion of the cheZ mutant was as good as that of the wild-type strain and the cheA mutant; (4) invasion of the tumbly cheB mutant was intermediate between non-motile and wild-type strains .

Discussion If S. typhimurium could continuously infect HeLa cells and propagate within them, the number of intracellular bacteria would increase exponentially . However, the saturation of the number of bacteria per HeLa cell suggested by the infection kinetics indicated that the metabolism of the HeLa cells was altered after the initial invasion of S. typhimurium so that the HeLa cells would not allow further entry and intracellular multiplication of the bacteria . It is possible, for example, that the bacterial invasion reduced the phagocytosis by the host cell, the mechanism of bacterial internalization .

Motility and invasion in Salmonella

51

The possibility of inhibition of the intracellular bacterial proliferation by gentamicin should also be considered . The efficiency of invasion by the wild-type S . typhimurium was much higher than

that of the isogenic non-motile mutants . A comparison of invasion between the mot and non-flagellated mutants demonstrated that motility per se was important for efficient invasion of S . typhimurium into HeLa cells . We did not obtain any indication that the flagellar structures would act as adhesins or that the flagella would stimulate the phagocytic activity of HeLa cells. The present data therefore indicate that bacterial motility probably increases the frequency of bacterial invasion by increasing the chance of collision between the pathogen and the HeLa cells . Studying the phagocytic response of macrophages from C3H/He mice, Tomita and Kanegasaki" concluded that physical impact caused by bacterial motility enhanced the phagocytic activity of macrophages . We do not believe that a threshold of physical impact exists for the invasion of Salmonella into the HeLa cells because simple sedimentation of nonmotile bacteria was enough to trigger the internalization of the bacteria . S . typhimurium is not only motile but also chemotactic . In the wild-type strain, the cells swim by rotating their flagella with an occasional change in their swimming direction by reversing the flagellar rotation . A chemotactic response is manifested by modulating the frequency of reversal of the swimming direction : if the cells swim toward a favorable stimulus, the reversal is repressed ; if they swim toward a nonfavorable stimulus, the reversal is increased . In chemotactic mutants, regulation of the reversal is defective, therefore, these mutants show aberrant swimming behavior : mutants defective in cheA, chew, cheY and cheR swim ahead without any reversal of

the swimming direction ; mutants defective in cheB and cheZ change continuously their swimming direction except when they receive strong chemotactic stimuli which causes smooth swimming behavior ." The rates of invasion of the smooth-swimming chemotactic mutants (cheA, cheW, cheY and cheR) was higher than the wild-type in the conventional assay, but approximately equal to the wild-type in the vertical assay . These results indicate that the chemotactic ability of Salmonella was not necessary for efficient invasion into HeLa cells but rather interfered with invasion under some experimental condition . The reason why the smooth chemotactic mutants invaded more efficiently than the wild-type strain into the HeLa cells in the conventional assay may be that chemotaxis toward oxygen or the geotaxis away from gravity in the wildtype bacteria inhibited its migration toward the HeLa cells on the bottom of the wells . The results with the tumbly mutants (cheB and cheZ) demonstrated that aberrant chemotactic behavior could reduce the infection of HeLa cells . The tumbly mutants continuously change their swimming direction because of which the mean speed of their spatial migration may be much smaller than that of the wild-type strain . The invasion rate of the cheZ mutant strongly increased when the bacteria were incubated

with the HeLa cells for 2 h or more . We assume that the cheZ mutant started to swim smoothly 2 h after incubation with HeLa cells by adapting to the new environment . Unfortunately, it was not possible to confirm this hypothesis experimentally because microscopic observation of the chemotactic behavior of the bacteria swimming above the monolayer of the HeLa cells was difficult . The present results suggest precautions to be taken for the utilization of chemotactic mutants for determining the role of chemotaxis in some biological process : it is advisable to use both smooth and tumbly mutants in such experiments .

We inoculated the bacteria at a concentration of about 3 x 10 6 cells/ml . The bacterial suspension was placed on a monolayer of confluent HeLa cells on a surface of 2 cm 2 . The mean swimming speed of the motile bacteria is about 20 µm/s, and one sixth of the bacteria are statistically moving toward the HeLa cells . Therefore, the net migration

52

T. Khoramian-Falsafi eta[

of bacteria toward the HeLa cells is of 3 ym/s . Therefore, about 10 5 cells (3 x 10 6 cells/cm 3 x 2 cm 2 x O .0003 cm/s x 60 s) would collide with the HeLa cells every minute . Experimentally 2x10 5 bacteria invaded the HeLa cells during 30 min (data from smooth the mutants) . Therefore, one invasion event would occur in every 15 [(2 x 10 5 ) : (105 x 30)] collisions . Although this is a rough estimation, it still indicates that a significant fraction of bacteria which interacted with the HeLa cells invaded . In the natural oral route infection of Salmonella, the site of infection is in the gut . Once the initial invasion has provoked some tissue injury, the site of the bacterial infection will be surrounded by inflammatory cells, and the pathogen will have to cope with them in order to establish a systemic disease . Therefore, several factors including efficiency of colonization and invasion in epithelial cells and the ability to survive and multiply in host macrophages would determine overall pathogenicity . The data presented in this paper demonstrate that the bacterial motility could contribute to the invasion process . The in vivo data using BALB/c mice as a model animal of salmonellosis, however, indicated that the bacterial motility was not important for the virulence . $ The precision of virulence tests by the challenge through the oral route of infection was however not very high . Thus, like with other pathogenic bacteria, the motility could well be an auxiliary factor contributing to the pathogenicity of Salmonella .

Materials and methods Bacteria/strains. SJW1 103 is a Of/iB (OH2) derivative of S . typhimurium LT2 . All the flagellar and non-chemotactic mutants were constructed on this genetic background using P22-assisted transduction . The KK1015 strain carries the fla-25 mutation . Our studies have demonstrated that the f/a-25 allele is a deletion mutation and defective in fIgABCDE, 13 (see lino et al." for gene symbols) . Therefore, we now call this mutation Lflg(A-E) . A mutant of S . typhimurium carrying the same mutant allele has been used by Carsiotis et al." to demonstrate that the mviS gene is important for virulence . The SJW1350 strain carries the fliC1001 mutation in the structural gene of one of two flagellins . Since the other flagellin gene f/iB was defective in the parent strain, the SJW1 350 strain does not synthesize any flagellin although it synthesizes the basal body and hook structure . The SJW1 353 strain is a flgK2004 derivative of SJW1 103 . The hook protein of this strain is defective, and the cell only synthesizes the basal body without hook. The SJW1 400 strain carries the flhD2040 mutation . This gene encodes a central controller for the flagellar regulation, which most likely is the flagellar gene-specific sigma factor . We also used SJW1 386, a motAS61 derivative of SJW1 103 in these experiments . This bacterial strain produces morphologically intact flagella, but, the cell is non-motile because its flagella do not rotate. The strains SJW3111, 3113, 3112, 3100, 2905 and 3037 were cheA, chew, cheR, cheY, cheB and cheZ derivatives of SJW1103 . Escherichia coli H131 01 (F - ara hsdS pro recA lac gal xy/mt/supErpsL thi) was used as a negative control . Conventional invasion assay . The bacteria were grown in L-broth 15 at 37°C overnight . Next morning, the culture was diluted into 100 volumes of the same medium and cultivated for a further 3 h . At that stage, the bacterial motility was checked microscopically and the bacterial cells resuspended in Dulbecco minimal essential medium supplemented with 2% FCS (Gibco) to a concentration of about 3 x 106 cells/ml . The bacterial suspension (1 ml) was used to infect a monolayer of cultured HeLa cells in 24-well plates (Nunc) and the plates were incubated for 30-180 min at 37°C under 5% CO 2 . The bacteria that were unable to associate tightly with the HeLa cells were removed by washing with Hanks balanced salt solution (HBSS), and the cell monolayer incubated at the same temperature in the fresh medium containing 200 pg/ml of gentamicin (Seromed) for 90 min . The medium was then removed and the HeLa cells were washed twice with minimal essential medium . The second wash was kept and the number of the extracellular bacteria in it was estimated by viable counting on L-broth plates . The viable bacteria were of the order of 0-30 cells/well in most of the experiments . The infected HeLa cells were mixed with 0 .5 ml 0 .1 M phosphate buffered saline (pH 7 .6) containing 0 .1% Triton X-100 to release intracellular bacteria . Gentamicin protected bacteria were then appropriately

Motility and invasion in Salmonella

53

diluted in phosphate buffered saline and plated onto L-broth to determine the number of intracellular cfu .

Vertical invasion assay . HeLa cells were cultivated in plastic bottles (Falcon, 50 ml) to obtain a monolayer of the cells . 45 ml of bacterial suspension in Dulbecco minimal essential medium supplemented by 2% FCS (approximately 2 .5x107 to 5x10 8 bacteria/ml) was added to the bottles and they were incubated for 0 .5 to 3 h at 37°C under 5% CO 2 . After the incubation, the HeLa cell monolayers were washed with HBSS and 50 ml of the same medium containing gentamicin (200 µg/ml) was added and incubated further for 1 .5 h at 37°C under 5% CO 2 . After the treatment, the HeLa cells were lysed by 0 .1 % Triton X-1 00 and the number of bacteria in the lysates was determined by viable counting as described above .

Microscopic observation of invaded bacteria . The monolayers of HeLa cells were grown on Lab-Tek chamber glass slides (Nunc, 4-chambers) and were infected with 1 ml of bacterial suspension as described above . After incubation with gentamicin, cells were washed, then fixed in 2% glutaraldehyde or 9-aminoacrydine (Sigma) and stained with Giemsa for microscopic observation .

We thank Shigeru Yamaguchi for bacterial strains, Elisabeth Buhiman for technical assistance and Frangoise Rey for secretarial work . This work was supported by the Fonds National Suisse de la Recherche Scientifique to JCP (Grant 31-28007 .89) .

References 1 . Freter R, Allweiss B, O'Brien PCM, Halstead SA, Macsai MS . Role of chemotaxis in the association of motile bacteria with intestinal mucosa : in vitro studies . Infect Immun 1981 ; 34 : 241-9 . 2 . Freter R, O'Brien PCM, Macsai MS. Role of chemotaxis in the association of motile bacteria with intestinal mucosa : in vivo studies, Infect Immun 1981 ; 343 : 234-40 . 3 . Guentzle MN, Berry LN . Motility as virulence factor for Vibrio cholerae . Infect Immun 1975 ; 11 : 890-7 . 4 . Drake D, Montie TC . Flagella, motility and invasive virulence of Pseudomonas aeruginosa. J Gen Microbiol 1988 ; 134 : 43-52 . 5 . Carsiotis M, Weinstein DL, Karch H, Holder IA, O'Brien AD . Flagella of Salmonella typhimurium are a virulence factor in infected C57BL/6J mice . Infect Immun 1984; 46 : 814-18 . 6 . Weinstein DL, Carsiotis M, Lessner CR, O'Brien AD . Flagella help Salmonella typhimurium survive with murine macrophages . Infect Immun 1984 ; 46: 819-25, 7 . Carsiotis M, Stocker BA, Weinstein DL, O'Brien AD . A Salmonella typhimurium virulence gene linked to fig. Infect Immun 1989; 57 : 3276-80 . 8 . Lockman HA, Curtiss III R . Salmonella typhimurium mutants lacking flagella or motility remain virulent in BALB/c mice . Infect Immun 1990; 58 :137-43 . 9 . De Melo M, Pechere JC . Effect of mucin on Campylobacter jejuni association and invasion on Hep-2 cells . Microbial Pathogenesis 1988 ; 5 : 71-6 . 10 . De Melo M, Gabbiani G, Pech6re JC . Cellular events and intracellular survival of Campylobacter jejuni during infection of Hep-2 cells . Infect Immun 1989 ; 57 : 2214-22 . 11 . Tomita T, Kanegasaki S . Enhanced phagocytic response of macrophages to bacteria by physical impact caused by bacterial motility or centrifugation . Infect Immun 1982 ; 38 : 865-70 . 12 . Macnab RM . Motility and chemotaxis . In : Neidhardt FC, Ingraham JL, Low KB, Magasanik B, Schaechter M, Umberger HE, eds . Escherichia coli and Salmonella typhimurium : cellular and molecular biology . Washington, D .C. : American Society for Microbiology, 1987 ; 732-59 . 13 . Kutsukake K, lino T, Komeda Y, Yamaguchi S . Functional homology of f/a genes between Salmonella typhimurium and Escherichia coli . Mol Gen Genet 1980 ; 178: 59-67 . 14 . lino T, Komeda Y, Kutsukake K et al. New unified nomenclature for the flagellar genes of Escherichia coli and Salmonella typhimurium. Microbial Rev 1988 ; 52 : 533-5 . 15 . Miller JH . Experiments in molecular genetics . Cold Spring Harbor : Cold Spring Harbor Laboratory, 1972 .

Effect of motility and chemotaxis on the invasion of Salmonella typhimurium into HeLa cells.

Salmonella typhimurium strain LT2 is able to invade HeLa cells in vitro. The effect of the motility and chemotaxis of the bacteria on cell invasion we...
455KB Sizes 0 Downloads 0 Views