The roles of intrahepatic V?14+ NK1.1+ T cells for liver injury induced bySalmonella infection in...

The Roles of Intrahepatic Va141 NK1.11 T Cells for Liver InjuryInduced by Salmonella Infection in Mice

MASATOSHI ISHIGAMI,1,2 HITOSHI NISHIMURA,1 YOSHIKAZU NAIKI,1 KENTARO YOSHIOKA,2 TETSU KAWANO,3 YUJIRO TANAKA,3

MASARU TANIGUCHI,3 SHINICHI KAKUMU,4 AND YASUNOBU YOSHIKAI1

To investigate the roles of intrahepatic T cells in liverinjury after Salmonella infection, we examined serum ala-nine transaminase (ALT), histopathology, and bacterialnumbers in liver after infection with Salmonella choleraesuisstrain 31N-1 in mice genetically lacking TCRab1, CD41,CD81, or NK1.11T cells with C57BL/6 background. Incontrol (1/1) mice, serum ALT reached a peak level by day7 after an intraperitoneal inoculation of 2 3 106 CFUSalmonella choleraesuis 31N-1. In TCR-b2/2 mice, liverinjury, as assessed by serum ALT level and histologicalexamination, was significantly suppressed on day 7 afterSalmonella infection but the numbers of bacteria in liver didnot differ from those in normal mice, suggesting that ab Tcells are responsible for liver injury induced by Salmonellainfection. To further determine which subsets in ab T cellsare important for the liver injury, we compared serum ALTlevel in mice genetically lacking CD4, CD8, b2-microglobu-lin (b2m, IAb, or Ja281 after Salmonella infection. InCD42/2 mice, serum ALT was significantly lower in compari-son with control mice, but there was no difference in serumALT levels in CD82/2 and IAb2/2 mice from that in controlmice. Notably, serum ALT levels and pathological lesions inliver were significantly decreased in b2m2/2 or Ja2812/2

mice, which lacked in NK1.11 T cells bearing TCR Va14-Ja281 specific for b2m-associated CD1d, following Salmo-nella infection. Taken together, it is suggested that ab Tcells bearing NK1.1 and CD4 may be main effector cells forliver injury after Salmonella infection. (HEPATOLOGY

1999;29:1799-1808.)

Salmonella serovar are facultative parasites that cause avarious spectrum of diseases ranging from self-limited gastro-intestinal infections to systemic infections with high mortal-ity named typhoid fever in human.1,2 Th1 type immuneresponse producing IFN-g is important for eliminating suchintracellular bacteria.3-6 The protective role of ab T cellsagainst bacterial growth of Salmonella in liver was proven inthe experiments used anti-TCR ab mAb for depletion ofthese cells.7 The study with gene knockout mice lacking ab Tcells revealed that ab T cells are important to eliminate thesebacteria in the late phase of infection rather than in the earlyphase until 6 days after intramural infection with S. dublin.8

Depletion experiments of CD41 and CD81 T cells with mAbsduring S. typhimurium infections showed that CD41 T cellsplayed important roles for the late phase elimination ofbacteria and CD81 T cells had only marginal effects.9 We havepreviously reported that unique T cells expressing intermedi-ate intensity of TCR ab and high intensity of LFA-1 in liverare important mediators to control Salmonella infection,especially in early phase of infection.7 The unique liver T cellsare now known to correspond to NK1.11 T cells in regard totheir restricted TCR ab repertoire and restriction of MHCclass I-like Ag.10-12 NK1.11 T cells are found in thymus,13

liver,14 and bone marrow,15 where they account for popula-tions of CD42CD82 and CD41CD82 TCRab cells. A largenumber of NK1.11 T cells express an invariant TCR encodedby the Va14 and Ja281 gene segments10,11 and are selected bythe nonpolymorphic MHC class I-like surface protein, CD1d.12

NK1.11 T cells in liver are recently found to be an essentialtarget of IL-12 for their cytotoxity by an NK-like effectormechanism.16 Therefore, it is possible that NK1.11 T cells inliver may contribute to protection against Salmonella infec-tion via cytotoxity against infected hepatocytes.

The liver is one of the target organs of Salmonella.17 In theliver, the bacteria are known to invade hepatocytes.4,18 Recentreports show that S. typhi induced acute hepatitis is seen in 27patients who were well-documented cases of typhoid feverwith clinical aspect of acute hepatitis in humans.19 In themurine model for Con A-induced hepatitis, the Fas and Fas-Lsystem plays a central role in liver injury.20 In addition, IFN-gand TNF-a productions are known to be involved in liverinjury.21,22 In case of liver injury induced by bacterial infec-tions, CD81 cytotoxic T cells are reported to be responsiblefor liver damage after a lethal infection with Listeria monocy-togenes,23 whereas excessive TNF-a may contribute to liverdamage after endotoxin-induced liver injury after infection ofGram-negative bacteria.24 Salmonella species (spp.), Gram-

Abbreviations: 2/2, knockout; TCR, T cell receptor; b2m, b2-microglobulin; NK,natural killer; LFA-1, leucocyte function antigen-1; MHC, major histocompatibilitycomplex; Con A, concanavalin A; TNF-a, tumor necrosis factor-a; spp., species; FcR,Fc-receptor; mAb, monoclonal antibody; IL-12, interleukin-12; IL-4, interleukin-4;CTL, cytotoxic T lymphocyte; FasL, Fas ligand; IFN-g, interferon-g; ALT, alaninetransaminase; CFU, colony forming unit.

From the 1Laboratory of Host Defense and Germfree Life, Research Institute forDisease Mechanism and Control; 2Third Department of Internal Medicine, NagoyaUniversity School of Medicine, Nagoya; 3The CREST, Japan Science and TechnologyCorporation (JST), and Division of Molecular Immunology, Center for BiomedicalScience, Chiba University School of Medicine, Chiba; and the 4First Department ofInternal Medicine, Aichi Medical School, Aichi, Japan.

Received August 10, 1998; accepted March 29, 1999.Address reprint requests to: Yasunobu Yoshikai, Laboratory of Host Defense and

Germfree Life, Research Institute for Disease Mechanism and Control, NagoyaUniversity School of Medicine, 65 Tsusumai-cho, Showa-ku, Nagoya, 466-8550, Japan.E-mail: [email protected]; fax: 81-52-744-2449.

Copyright r 1999 by the American Association for the Study of Liver Diseases.0270-9139/99/2906-0025$3.00/0

1799

negative bacteria having LPS in cell walls, are intracellularbacteria that preferentially induce cell-mediated immu-nity.3,9,25-28 But within our knowledge, the direct relationshipbetween Salmonella spp. and liver injury was not still proven.

In the present study, we estimated the liver injury andbacterial growth in liver during the infection with an intracel-lular bacterium, S. choleraesuis avirulent strain 31N-1 curedof 50-kb plasmid.29 Analyses of the role of various T cellsubsets were performed by using various genes-knockoutmice, and NK1.11 T cells were found to have an importantrole for the formation of liver injury.

MATERIALS AND METHODS

Mice. TCRb, CD4, and CD8-knockout mice generated withhomologous recombination in pluripotent embryonic stem cells andbred homozygosity have been described.30-35 Ja281-knockout micelacking only Va141 NK1.11T cells have also been described.36

b2-microglobulin (b2m) and MHC class II IAb-knockout mice wereobtained from Taconic (Germantown, NY). Age- and sex-machedC57BL/6 mice were purchased from Japan SLC (Hamamatsu, Japan)and were used as the normal control in each experiments. All micewere used at 6 to 10 weeks of age and maintained under pathogen-free conditions in the facilities of Nagoya University. All experimentswere done along the criteria outlined in the Guide for the Care andUse of Laboratory Animals prepared by the National Academy ofScience.

Microorganisms. S. choleraesuis strain 31N-1 (2 3 106 CFU)29

were inoculated intraperitoneally in all experiments.Antibodies. Phycoerythtin (PE)-conjugated anti-TCR ab, anti-gd,

anti-NK1.1, and anti-CD8 and FITC-conjugated anti-CD3 andanti-CD4 mAbs were purchased from PharMingen (San Diego, CA).

Liver-Derived T-cell Preparation. Mice were inoculated 2 3 106 CFUof viable S. choleraesuis intraperitoneally in 0.2 mL of PBS. Micewere sacrificed on indicated days after bacterial infection by cuttingtheir cervical arteries under deep anesthesia with ether. The wholeliver was removed after portal vein with PBS, homogenized thor-oughly, and serially diluted and adjusted 10 mL with PBS. Hepato-cytes were removed from these homogenates with 500 3 g and thencell precipitate was prepared at 1,100 3 g centrifuge. For gradientcentrifugation to derive lymphocytes, we prepared 45% and 66.6%Percoll solution and performed gradient centrifugation, and theintermittent band was derived carefully. After vigorous wash withcentrifuge, cell precipitate was used as liver-derived total lympho-cytes.

Bacterial Growth in Liver. Hundred µL out of 10 mL of liverhomogenates were spread on tryptosoya agar (Nissui Pharmaceuti-cal, Tokyo, Japan) plates, and colonies were counted after incuba-tion for 24 hours at 37°C.

Flow Cytometry. Liver-derived lymphocytes were stained with PE-and FITC-conjugated mAbs after blocking FcR by 2.4G2 (anti-FcgRmAb). After incubation with PE- and FITC-conjugated mAbs at 4°Cfor 30 minutes and extensive wash, the stained cells were analyzedby a FACScan flow cytometry (Beckton Dickinson, San Jose, CA).Live lymphocytes were gated by forward and side scatter.

Histological Examination. Liver specimens were fixed in 10% buff-ered formalin for more than 2 days and embedded in paraffin. Fourmicron-thick sections were cut and stained with hematoxylin andeosin.

Serum ALT Levels. Blood samples were collected at the time ofsacrifice. These samples were centrifuged at 5,000 3 g for 10minutes, and aqueous phase was used as serum. Serum ALT levelswere measured by the commercial available kit (IATRON, Tokyo,Japan) according to manufacturer’s instruction.

Cytokine Enzyme-Linked Immunosorbent Assay (ELISA). IFN-g andTNF-a production in the serum were measured by usingthe commercial available kits (IFN-g, Genzyme Diagnostics, Cam-

bridge, MA; TNF-a, Amersham, Buckinghamshire, UK) accordingto manufacturer’s instructions.

Statistical Analysis. All data were analyzed by Student’s t-test. P ,.05 was considered as statistically significant.

RESULTS

Bacterial Growth and Serum ALT Levels in TCRb2/2 Mice AfterSalmonella Infection. To examine the role of ab T cells inprotection and liver injury after infection with Salmonellaspp., we compared bacterial number in liver and serum ALTlevel between TCR b1/1 mice and TCRb2/2 mice with thesame genetic backgrounds as C57BL/6 mice after an intraperi-toneal inoculation with a sublethal dose (2 3 106 CFU) of Scholeraesuis 31N-1. The number of bacteria in liver ofTCRb2/2 mice was much the same as those in TCRb1/1 miceon days 3, 7, and 12 after Salmonella infection (Fig. 1A, Table1) and both groups of mice cleared this bacteria within 30days.37 In control (1/1) mice, serum ALT increased to reacha peak level by day 7 then decreased by day 12 after anintraperitoneal inoculation of 2 3 106 S. choleraesuis 31N-1(Fig. 1B, Table 1). These data indicate that infection withavirulent strain of S. choleraesuis actually elicited moderate

FIG. 1. Kinetics of bacterial growth and serum ALT levels in TCRb-knockout mice. TCRb-knockout and control mice were inoculated ip with2 3 106 CFU of S. choleraesuis 31N-1. Serum and livers were collected ondays 0, 1, 3, 7, and 12 after infection, and (A) bacterial counts and (B) serumALT levels were measured. Day 0 means the data from mice withoutinfection. Each point and vertical bar indicate the mean 6 SD for 5 mice ofeach group. (h), TCRb2/2; (m), control mice. *Significantly different fromthe values for control mice; P , .005.

1800 ISHIGAMI ET AL. HEPATOLOGY June 1999

liver injury in normal mice. Notably, the serum ALT level ofTCRb2/2 mice was significantly less than TCRb1/1 mice onday 7 and remained undetectable on day 12 after salmonellainfection (Fig. 1B). We settled latter comparative experimentson day 7 after Salmonella infection. Histological examinationshowed that infection with S. choleraesuis induced an appre-ciable damage to hepatocytes characterized by cell swelling

and focal apoptotic change showing cell shrinkage, chroma-tin condensation, and hemorrhages in TCRb1/1 mice (Fig. 2Band D), whereas the lesions were much smaller in TCRb2/2

mice on day 7 after infection (Fig. 2A and C).With respect to lymphocyte surface markers, liver lympho-

cytes comprised large proportions of ab T cells (32.5% 63.5%, n 5 5), and NK1.11CD32 cells (13.9% 6 2.2%, n 5 5)but only a small proportion of gd T cells (2.7% 6 0.3%,n 5 5) in TCRb1/1 mice on day 7 after Salmonella infection,whereas large proportions of gd T cells (22.5% 6 2.6%,n 5 5) and NK1.11CD32 cells (16.4% 6 1.3%, n 5 5) but noab T cells were present in liver lymphocytes in TCRb2/2 mice(Fig. 3). These data indicate that ab T cells are important forthe formation of Salmonella induced liver injury, but even ifab T cells do not exist, protection against Salmonella iselicited.

Liver Injury in CD42/2 and CD82/2 Mice After SalmonellaInfection. To further determine which ab T cell subsets areresponsible for liver injury after salmonella infection, serumALT level was examined in CD42/2 and CD82/2 mice afterinfection with S choleraesuis. There were no differences in thenumbers of bacteria between CD42/2 mice and control miceon days 7 (Fig. 4A, Table 1) and 12 after infection (data notshown), whereas serum ALT level was significantly lower in

TABLE 1. Comparison of Serum ALT Levels and Bacterial NumbersAmong Various Gene-Knockout Mice Following Salmonella Infection

MiceSerum ALT

(IU/L)Bacterial Number

(3104 CFU)

1/1 141 6 35 3.926 6 3.899TCRb 2/2 50 6 8* 3.148 6 0.863CD4 2/2 78 6 6* 2.655 6 2.393CD8 2/2 137 6 24 4.656 6 2.362b2m 2/2 106 6 23* 3.085 6 1.198IAb 2/2 140 6 34 3.098 6 1.167Ja281 2/2 57 6 12* 3.239 6 1.058

NOTE. Mice were ip inoculated with 2 3 106 Salmonella choleraesuis31N-1 and serum ALT and bacterial number in liver were measured on day 7after infection. Data are expressed as means 6 SD of 3 or 5 mice of eachgroup.

*P , .05 compared with control (1/1) mice.

FIG. 2. Comparisons of the histological findings of liver specimens between TCRb-knockout and control mice on day 7 after infection with S. choleraesuis31N-1. Formalin fixed, paraffin embedded thin specimens were stained with hematoxilin and eosin, and microscopical analyses were performed. Arrowsrepresent the necroinflammatory foci, and asterisks represent apoptotic hepatocyte that is characterized cell shrinkage and chromatin condensation. (A andC) Liver specimens from TCRb-knockout mice. (B and D) Liver specimens from control mice.

HEPATOLOGY Vol. 29, No. 6, 1999 ISHIGAMI ET AL. 1801

CD42/2 mice than in control mice on day 7 after infection(Fig. 4B, Table 1). In contrast, there was no difference inbacterial growth or serum ALT levels on days 7 (Fig. 4A andB) and 12 (data not shown) after infection between CD81/1

and CD82/2 mice.With respect to lymphocyte surface markers, the propor-

tion of ab T cells increased but gd T cells decreased in liver ofCD42/2 mice compared with CD41/1 mice on day 7 after

infection (Fig. 4C). There was no significant difference inlymphocyte populations except for CD81 cells in liverbetween CD82/2 mice and control mice (Fig. 4C). Theseresults suggest that CD41 ab T cells are important for liverinjury after Salmonella infection.

Liver Injury in MHC Class IIAb2/2 and b2m2/2 Mice AfterSalmonella Infection. Differentiation of CD41 T cells wasseverely impaired in MHC class II2/2 mice,38 whereas b2m2/2

FIG. 3. Comparisons of lympho-cyte surface markers analyzed byflow cytometry between TCR-bknockout and control mice. Datawere obtained from mice inoculatedip with 2 3 106 S. choleraesuis31N-1 7 days previously and repre-sentative of 5 mice of each group.TCRb2/2, TCRb-knockout mice;TCRb1/1, control mice.


mice have no CD81 T cells39 or NK1.11 T cells specific forCD1 because of lack in MHC class I and MHC class I-likemolecules associated with b2m.14,15 We next examined serumALT levels in MHC class II Ab2/2 and b2m2/2 mice afterSalmonella infection. The numbers of bacteria in liver ofMHC class IIAb2/2 and b2m2/2 mice were much the same asthose seen in control mice on days 7 and 12 after infection(Fig. 5A, Table 1). Notably, serum ALT level was significantlylower in b2m2/2 mice but there was no difference in MHCclass II Ab2/2 mice from those in control mice on day 7 afterSalmonella infection (Fig. 5B, Table 1). Histological examina-tion confirmed that cell death and lymphocyte infiltrationwere obviously less in b2m2/2 mice (Fig. 6C and F), but were

much the same in IAb2/2 mice as control mice (data notshown).

With respect to lymphocyte surface markers, the relativenumbers of CD41NK1.11 T cells were severely reduced(0.8% 6 1.2%, n 5 5), whereas the numbers of NK1.11CD32

cells (22.0% 6 5.8%, n 5 5) and gd T cells (7.0% 6 1.6%,n 5 5) were increased in b2m2/2 mice compared with thosein control mice on day 7 after Salmonella infection (Fig. 7).These results suggest that CD41NK1.11 T cells specific forMHC class I like molecules associated with b2m may beresponsible for the formation of liver injury.

Liver Injury in Ja2812/2 Mice After Salmonella Infection. Alarge number of NK1.11 T cells express an invariant TCR

FIG. 4. Comparisons of bacterialgrowth, the serum ALT levels, andlymphocyte surface markers amongCD4-knockout, CD8-knockout, andcontrol mice. All data were obtainedfrom mice inoculated with 2 3 106

S. choleraesuis 31N-1 7 days previ-ously. Data in (A) and (B) are ex-pressed as mean 6 SD for 5 mice ofeach group. A representative profileanalyzed by flow cytometry from 5mice of each group was shown in(C). CD42/2, CD4-knockout mice;CD82/2, CD8-knockout mice;1/1, control mice. N.S., no signifi-cant difference vs. control mice. *Sig-nificantly different from the valuesfor control mice; P , .05.


encoded by the Va14 and Ja281 gene segments and areselected by the nonpolymorphic MHC class I-like surfaceprotein, CD1d.12 Our results suggest that CD41NK1.11 Tcells specific for MHC class I like molecules associated withb2m may be responsible for the formation of liver injuryfollowing Salmonella infection. To elucidate this issue, weexamined serum ALT in Ja2812/2 mice, which have almostno NK1.11 T cells in liver.36 The numbers of bacteria did notdiffer from those in control mice on days 7 and 12 afterSalmonella infection (Fig. 5A, Table 1). As shown in Fig. 5B,ALT level was severely reduced in Ja2812/2 mice on day 7after infection as compared with control mice. There was nodifference in serum ALT levels between Ja2812/2 mice andcontrol mice on day 12 after Salmonella infection (Fig. 5B).Histological examination confirmed that cell death andlymphocyte infiltration were obviously less in Ja2812/2 miceon day 7 after infection (Fig. 6B and E).

With respect to lymphocyte surface markers, the propor-tions of CD41NK1.11 T cells were severely reduced

(0.9% 6 0.2%, n 5 3), whereas the relative numbers ofNK1.11 CD32 cells (16.5% 6 5.8%, n 5 3) and gd T cells(3.5% 6 2.8%, n 5 3) were not changed but CD81 T cells(19.5% 6 3.5%, n 5 3) increased in Ja2812/2 mice comparedwith those in control mice on day 7 after Salmonella infection(Fig. 7). Collectively these results suggest that NK1.11 T cellsbearing Va14/Ja281 contribute to liver injury induced bySalmonella infection.

DISCUSSION

The mechanisms whereby Salmonella infection inducesliver injury remain unknown. We here show that b2m2/2 andJa2812/2 mice lacking NK1.11 T cells are resistant to liverdamages induced by infection with S. choleraesuis. Further-more, liver injury was suppressed in CD42/2 mice but not inMHC class II Ib2/2 mice. These results suggest a significantfraction of ab T cells such as T cells bearing NK1.1 and CD4is mainly responsible for the formation of liver injury inducedby Salmonella infection.

Salmonella spp. are Gram-negative bacteria of which cell-wall components contain LPS.39 Endotoxin/LPS derived fromGram-negative bacteria play an important role in the patho-genesis of liver injury associated with sepsis. LPS-inducedacute liver injury in mice pretreated with Propionibacteriumacne has been widely used as a model of hepatic failureassociated with endotoxin.41,42 Athymic nude mice are re-ported to be susceptible to this liver injury,41 raising apossibility that NK cells and extrathymic liver T cells such asTCR-intermediate T cells may be involved in liver injury.TCR-intermediate T cells contain a large population ofNK1.11 T cells in liver.43,44 A large fraction of NK1.11 T cellsexpress invariant TCR encoded by Va14 and Ja281 genesegment10,11 and are selected by the nonpolymorphic MHC-class I-like surface protein, CD1d.12 NK1.11 T cells are almostcompletely absent in the thymus and liver of mice lackingb2m,14,15 CD1d,45,46 and Ja281.36 We found that b2m2/2 orJa2812/2 mice were resistant to liver injury induced bysalmonella infection. These results strongly suggest thatNK1.11 T cell expressing a canonical TCR Va14/Ja281 maybe a critical role in the formation of liver injury after injuryafter salmonella infection.

Cytotoxic pathways including Fas-mediated and perforin-mediated are considered for the formation of hepatocyteinjury. In recent report of hepatitis C virus (HCV)-specifichuman CTL, HCV-specific CTLs would kill HCV-infectedcells by Fas- and perforin-mediated cytolysis.47 In ConA-induced hepatitis, both perforin/granzyme pathway and Fas-FasL interaction are important in the development of hepati-tis.48,49 NK1.11 T cells have abilities for cytotoxity viaperforin and Fas-dependent pathways.50 Therefore, the cyto-toxic pathways including Fas/FasL and perforin-dependentcytotoxities were assumpted as important pathways in liverinjury of murine salmonellosis.

We have also obtained evidence that CD42/2 mice wereresistant to liver injury against Salmonella infection. Theseresults suggest that CD41 T cells specific for Salmonellaantigens in context of MHC class II molecules may contributeto liver injury as well as NK1.11 T cells. However, MHC classIIAb2/2 mice, which have no CD41 T cells specific for MHC

FIG. 5. Kinetics of bacterial growth and serum ALT levels in b2-microglobulin-knockout, MHC class IIAb-knockout, Ja281-knockout, andcontrol mice. Serum and livers were collected on days 0, 7, and 12 afterinfection, and bacterial counts (A) and serum ALT levels (B) were measured.Day 0 means the data from mice without infection. Each point and verticalbar indicates the mean 6 SD for 3 or 5 mice of each group. (d) b2m2/2,b2-microglobulin knockout mice; (n) IAb2/2, IAb knockout mice; (s)Ja2812/2, Ja281 knockout mice; (h) 1/1, control mice. *P , .05 vs.control mice.


class II-binding antigens, are susceptible to the liver injuryinduced by Salmonella infection, excluding this possibility.NK1.11 T cells are consisted of CD41 and CD4282 popula-tions.15,51-54 Therefore, it is possible that NK1.11 T cellsbearing CD4 may be responsible for liver injury after Salmo-nella infection. CD41 population of NK1.11 T cells in liverpreferentially produce IL-4 on TCR engagement,55,56 whereasCD42 T cells produce IFN-g as a result of such stimulation.57

Yoneyama et al.58 have recently reported that IL-4 producingCC chemokine receptor1CD41 T cells may be a key cellcausing massive liver injury after systemic LPS administra-tion. Therefore, it is speculated that CD41NK1.11 T cells mayinduce liver injury during Salmonella infection via IL-4production. LPS-induced liver injury is also known to bemediated by excessive amounts of TNF-a and IFN-g, whichinduce apoptosis of hepatocytes.22,40 NK1.11 T cells candrastically change in their ability for cytokine productionfrom IL-4 to IFN-g in the presence of IL-12.59 Therefore, Th1type CD41NK1.11 T cells producing IFN-g may contribute toliver injury after Salmonella infection, which may induce alarge amount of IL-12 production by macrophages. In ourpresent results, however, serum TNF-a and IFN-g levels werenot correlated with liver injury in b2m2/2 or Ja2812/2 mice

and control mice after Salmonella infection. Furthermore, wecould not detect any IL-4 activity in serum of normal mice atany stage after infection. These cytokines may not be respon-sible for causing liver injury induced by Salmonella infection.Alternatively, IFN-g or IL-4 may be secreted in a very lowamount or TNF-a may be associated to the cell membrane,thus, acting in a narrow intracellular range in the liver. Butthis is only a speculation until now, and further investigationwill be needed to confirm this speculation.

TCRb2/2, Ja2812/2, and b2m2/2 mice equally clearedbacteria after Salmonella infection as compared with controlmice, suggesting that protection against Salmonella infectionmay be compensated by NK1.11 T cells in these knockoutmice. NK cells and gd T cells are known to contribute toprotection against bacterial infection, especially at the earlyphase of infection.60,61 The significant increases in NK and gdT cells were evident in liver of TCRb2/2 and b2m2/2 micemice after Salmonella infection, and CD81T cells wereincreased in Ja2812/2 mice infected with Salmonella. It isspeculated that liver injury may be mainly mediated NK1.11

T cells expressing Va14/Ja281, whereas protection may becompensated by various cells including NK,gd T cells andCD8 T cells in mice lacking NK1.11 T cells.

FIG. 6. Comparisons of the histological findings of liver specimens among b2-microglobulin-knockout, Ja281-knockout, and normal control mice onday 7 after infection with S. choleraesuis 31N-1. Formalin fixed, paraffin embedded thin specimens were stained with hematoxilin and eosin, andmicroscopical analyses were performed. Arrows represent the necroinflammatory foci. (A and D) Liver specimens from control mice. (B and E) Liverspecimens from Ja281-knockout mice. (C and F) Liver specimens from b2m-knockout mice. Original magnification (A, B, and C), 3100; (D, E, and F),3200; (D), 3400.


REFERENCES

1. Lanata CF, Ristori C, Jimenez L, Garcia J, Levine MM, Black RE,Sotomayor V, et al. Vi serology in detection of chronic Salmonella typhicarriers in an endemic area. Lancet 1983;1:441-443.

2. Saphla L, Winter JW. Clinical manifestations of salmonellosis in man. Anevaluation of 7779 human infections identified at the New YorkSalmonella Center. N Engl J Med 1957;256:1128-1134.

3. Nauciel C, Espinasse-Maes F. Role of gamma interferon and tumornecrosis factor alfa in resistance to Salmonella typhimurium infection.Infect Immun 1992;60:450-454.

4. Mastroeni P, Skepper JN, Hormaeche CE. Effect of anti-tumor necrosisfactor alpha antibodies on histopathology of primary Salmonella infec-tions. Infect Immun 1995;63:3674-3682.

5. Mastroeni P, Villarreal-Ramos B, Hormaeche CE. Role of T cells, TNF

FIG. 7. Comparisons of lympho-cyte populations among b2-micro-globulin-knockout, Ja281-knock-out and control mice analyzed byflow cytometry. Data are obtainedfrom mice inoculated ip with 2 3106 S choleraesuis 31N-1 7 dayspreviously and are representative of3 or 5 mice of each group. b2m2/2,b2-microglobulin knockout mice,Ja2812/2; 1/1, control mice.


alpha and IFN gamma in recall of immunity to oral challenge withvirulent samonellae in mice vaccinated with live attenuated aro2

Salmonella vaccines. Microb Pathog 1992;13:477-491.6. Tite JP, Dougan G, Chatfield SN. The involvement of tumor necrosis

factor in immunity to Salmonella infections. J Immunol 1991;147:3161-3164.

7. Matsumoto Y, Emoto M, Usami J, Maeda K, Yoshikai Y. A protective roleof extrathymic ab TcR cells in liver in primary murine salmonellosis.Immunology 1994;81:8-14.

8. Weintraub BC, Eckmann L, Okamoto S, Hence M, Hedrick SM, Fierer J.Role of ab and gd T cells in the host response to Salmonella infection asdemonstrated in T-cell-receptor-deficient mice of defined Ity genotypes.Infect Immun 1997;65:2306-2312.

9. Nauciel C. Role of CD41 T cells and T-independent mechanisms inacquired resistance to Salmonella typhimurium infection. J Immunol1990;145:1265-1269.

10. Koseki H, Asano H, Inaba T, Miyashima N, Moriwaki K, Lindahi KF,Taniguchi M, et al. Dominant expression of a distinctive V141 T-cellantigen receptor a chain in mice. Proc Natl Acad Sci U S A 1991;88:7518-7522.

11. Makino Y, Kanno R, Ito T, Hisakado K, Taniguchi M. Predominantexpression of invariant V alpha 141 TCR alpha chain in NK1.1 T cellpopulations. Int Immunol 1995;7:1157-1161.

12. Bendelac A, Rivera MN, Se-Ho P, Roark JH. Mouse CD-1 specific NK1 Tcells: development, specificity, and function. Annu Rev Immunol 1997;15:535-562.

13. Budd RC, Miescher GC, Lees RC, Bron C, MacDonald HR. Developmen-tally regulates expression of T cell receptor beta chain in immunethymocytes. J Exp Med 1987;166:577-582.

14. Ohteki T, MacDonald HR. Major histocompatibility complex classII-related molecules control the development of CD4182 and CD4282

subsets of natural killer 1.11 T cell receptor ab1 cells in the liver ofmice. J Exp Med 1994;180:699-704.

15. Bendelac C, Killeen N, Littmann D, Schwartz RH. A subset of CD41

thymocytes selected by MHC class II molecules. Science 1994;263:1774-1777.

16. Ogasawara K, Takeda K, Hashimoto W, Satoh M, Okuyama R, Yanai N,Seki S, et al. Involvement of NK11 T cells and their IFN-gammaproduction in the generalized Shwarzman reaction. J Immunol 1998;160:3522-3527.

17. Mackness GB. Cellular resistance to infection. J Exp Med 1962;116:381.18. Conlan JW, North RJ. Early pathogenesis of infection in the liver with

the facultative intracellular bacteria Listeria monocytogenes, Francisellaturanesis and Salmonella typhimurium involves lysis of infected hepato-cytes by leucocytes. Infect Immun 1992;63:3674-3682.

19. El-Newihi HM, Alamy ME, Reynolds TB. Salmonella hepatitis: analysisof 27 cases and comparison with acute viral hepatitis. HEPATOLOGY

1996;24:516-519.20. Mizuhara H, Uno M, Seki N, Yamashita M, Yamaoka M, Ogawa T,

Fujiwara H, et al. Critical involvement of Interferon gamma in thepathogenesis of T-cell activation-associated hepatitis and regulatorymechanisms of Interleukin-6 for the manifestations of hepatitis. HEPATOL-OGY 1996;23:1608-1615.

21. Toyonaga T, Hino O, Sugai S, Wakasugi S, Abe K, Shichiri M, YamamuraK. Chronic active hepatitis in transgenic mice expressing interferon-g inthe liver. Proc Natl Acad Sci U S A 1994;91:614-618.

22. Leist M, Gantner F, Tiegs G, Germann PG, Wendel A. Tumor necrosisfactor-induced hepatocyte apoptosis precedes liver failure in experimen-tal murine shock models. Am J Pathol 1995;146:1220-1234.

23. Jiang X, Gregory SH, Wing EJ. Immune CD81 T lymphocytes lyseListeria monocytogenes-infected hepatocytes by a classical MHC classII-restricted mechanism. J Immunol 1997;158:287-293.

24. Arai T, Hiromatsu K, Kobayashi N, Takano M, Ishida H, Nimura Y,Yoshikai Y. IL-10 is involved in the protective effect of dibutyryl cyclicadenosine monophosphate on endotoxin-induced inflammatory liverinjury. J Immunol 1995;155:5743-5749.

25. Collins FM. Vaccines and cell-mediated immunity. Bacterial Rev 1974;38:371-402.

26. Maskell DJ, Hormasche CE, Harrington KA, Joysey HS, Liew FY. Theinitial suppression of bacterial growth in a Salmonella infection ismediated by a localized rather than a systemic response. Microb Pathog1987;2:295-305.

27. Maier T, Oels HC. Roles of the macrophage in natural resistance tosalmonellosis in mice. Infect Immun 1972;6:438-443.

28. Hochadel JF, Keller KF. Protective effects of passively transfected

immune T- or B-lymphocytes in mice infected with Salmonella typhimu-rium. J Infect Dis 1977;135:813-823.

29. Kawahara K, Haraguchi Y, Tsuchino M, Terakado N, Danbara H.Evidence of correlation between 50-Kilobase plasmid of Salmonellacholeraesuis and its virulence. Microb Pathog 1988;4:155-163.

30. Yeung RS, Penninger J, Mak TW. T-cell development and function ingene-knockout mice. Curr Opin Immunol 1994;180:298-307.

31. Ryffel B. Gene knockout mice as investigate tools in pathophysiology. IntJ Exp Path 1996;77:124-141.

32. Rahetmulla A, Fung-Leung W-P, Schilham W. Normal development andfunction of CD81 cells but marked decreased helper cell activity in micelacking CD4. Nature 1991;353:180-184.

33. Fung-Leung W-P, Schilham MW, Rahetmura A. CD8 is needed fordevelopment of cytotoxic T cells but not helper T cells. Cell 1991;65:443-449.

34. Kreigler NR, Fathman CG. The use of CD4 and CD8 knockout mice tostudy the role of T-cell subsets in allotransplant rejection. J Heart LungTransplant 1997;16:263-267.

35. Momberts P, Clarke AR, Rudnicki MA, Lacomini J, Itohara S, Lafaille JJ,Tonegawa S, et al. Mutations in T cell antigen receptor genes a and bblock thymocyte development at different stages. Nature 1992;360:225-231.

36. Cui J, Shin T, Kawano T, Sato H, Kondo E, Toura I, Taniguchi M, et al.Requirement for Va14 NKT cells in IL-12-mediated rejection of tumors.Science 1997;278:1623-1625.

37. Emoto M, Danbara H, Yoshikai Y. Induction of gamma/delta T cells inmurine salmonellosis by an avirulent but not by a virulent strain ofSalmonella choleraesuis. J Exp Med 1992;176:363-372.

38. Grusby JM, Glimcher LH. Immune responses in MHC calls II-deficientmice. Annu Rev Immunol 1995;13:417-435.

39. Raulet DH. MHC calls I-deficient mice. Adv Immunol 1994;55:381-421.40. Liang-Takasaki C, Saxen H, Makela H, Lieve L. Complement activation

by polysccharide and lipopolysaccharide: an important virulence deter-minant of Salmonellae. Infect Immun 1983;41:563-569.

41. Tsutsui H, Matsui K, Kawada N, Hyodo Y, Hayashi N, Okamura H,Nakanishi K, et al. IL-18 accounts for both TNF-a- and Fas ligand-mediated hepatotoxic pathways in endotoxin-induced liver injury inmice. J Immunol 1997;159:3961-3967.

42. Mizoguchi Y, Tsutsui H, Miyajima K, Sakagami Y, Seki S, Kobayashi K,Morisawa S, et al. The protective effects of prostaglandin E1 in anexperimental massive hepatic cell necrosis model. HEPATOLOGY 1987;7:1184-1188.

43. Hashimoto W, Takeda K, Anzai R, Ogasawara K, Sakihara H, Seki K,Kumagai K. Cytotoxic NK1.1 Ag1 alpha beta T cells with intermediateTCR induced in the liver of mice by IL-12. J Immunol 1995;154:4333-4340.

44. Tsukahara A, Seki S, Iiai T, Moroda T, Watanabe H, Suzuki S, Abo T.Mouse liver T cells: their change with aging and in comparison withperipheral T cells. HEPATOLOGY 1997;26:301-309.

45. Chen Y-H, Chiu NM, Mandai M, Wang N, Wang CR. Impaired NK11 Tcell development and early T cell development and early IL-4 productionin CD1-deficient mice. Immunity 1997;6:459-467.

46. Mendiratta SK, Martin WD, Hong S, Boesteanu A, Joyce S, Kaer LV.CD1d mutant mice are deficient in natural T cells that promptly produceIL-4. Immunity 1997;6:469-477.

47. Ando K, Hiroishi K, Kaneko T, Moriyama T, Muto Y, Kayagaki Y,Imawari M. Perforin, Fas/Fas ligand, and TNF-a pathways as specificand bystander killing mechanisms of Hepatitis C Virus specific humanCTL. J Immunol 1997;158:5283-5291.

48. Watanabe Y, Morita M, Akaike T. Concanavalin A induces perforin-mediated but not Fas-mediated hepatic injury. HEPATOLOGY 1996;24:702-710.

49. Seino K, Kayagaki N, Takeda K, Fukao K, Okumura K, Yagita H.Contribution of Fas ligand to T cell-mediated hepatic injury in mice.Gastroenterology 1997;113:1315-1322.

50. Arase H, Arase N, Kobayashi Y, Nishimura Y, Yonehara S, Onoe K.Cytotoxicity of fresh NK1.11 T cell receptor alpha/beta1 thymocytesagainst a CD4181 thymocyte population associated with intact Fasantigen expression on the target. J Exp Med 1994;180:423-432.

51. Takahama Y, Sharrow SO, Singer A. Expression of an usual T cellreceptor Vb repertoire by Ly-6C subpopulations of CD41 and/or CD81

thymocytes. Evidence for a developmental relationship between CD4/CD8 positive Ly-6C1 thymocytes and CD42CD82 TCRab1 thymocytes.J Immunol 1991;147:2883-2891.


52. Bendelac A, Matzinger P, Seder RA, Paul WE, Schwarz RH. Activationevents during thymic selection. J Exp Med 1992;175:731-742.

53. Hayakawa K, Lin BT, Hardy RR. Murine thymic CD41 T cell subsets: asubset (Thy0) that secrets diverse cytokines and overexpress the Vb8 Tcell receptor gene family. J Exp Med 1992;176:269-274.

54. Arase H, Arase N, Ogasawara K, Good RA, Onoe K. An NK1.11

CD41CD82 single-positive thymocyte subpopulation that express ahigh skewed T cell antigen receptor family. Proc Natl Acad Sci U S A1992;89:6506-6511.

55. Flesch IE, Wandersee A, and Kaufmann SH. IL-4 secretion by CD41

NK11 T cells induces monocyte chemoattractant protein-1 in earlylisteriosis. J Immunol 1997;159:7-10.

56. Emoto M, Emoto Y, Kaufmann SH. Bacille Calmette Guerin andinterleukin-12 down-regulate interleukin-4-producing CD41 NK11 Tlymphocytes. Eur J Immunol 1997;27:183-188.

57. Sugie T, Kubota H, Sato M, Nakamura E, Imamura M, Minato N. NK11

CD42 CD82 alphabeta T cells in the peritoneal cavity: specific T cellreceptor-mediated cytotoxity and selective IFN-gamma productionagainst B cell leukemia and myeloma cells. J Immunol 1996;157:3925-3935.

58. Yoneyama H, Harada A, Imai T, Baba M, Yoshie O, Zhang Y, Higashi H, etal. Pivotal role of TARC, a CC chemokine, in bacteria-induced fulminanthepatic failure in mice. J Clin Invest 1998;102:1933-1941.

59. Kawamura T, Takeda K, Mendriatta SK, Kawamura H, Van Kaer L, YagitaH, Okumura K, et al. Critical role of NK11 T cells in IL-12-inducedimmune responses in vivo. J Immunol 1998;160:16-19.

60. Skeen MJ, Ziegler HK. Induction of murine peritoneal g/d T cells andtheir role in resistance to bacterial infection. J Exp Med 1993;178:971-984.

61. Hiromatsu K, Yoshikai Y, Matsuzaki G, Ohga S, Muramori K, MatsumotoK, Nomoto K. A protective role of gd T cells in primary infection withListeria monocytogenes in mice. J Exp Med 1992;175:49-56.


The roles of intrahepatic V?14+ NK1.1+ T cells for liver injury induced bySalmonella infection in...

Documents

Transcript of The roles of intrahepatic V?14+ NK1.1+ T cells for liver injury induced bySalmonella infection in...