a / ]e          Pr L    &
              0022-7 7s7164/t 392-o96 ;so2.oo/0
              Tm Jouaau o. luwoMp.ow                                                                                                       Vol 133 . No2 . August 19184
              Copynght 0 1994 by The Amertran Anaoi.°ISUon of Immunolorsts                                                                           Prrnted In U. S. A.


                                               I . lEvidence that a Lymphokine Mediates Hilling'

                                    RICHARD 1B . MARKHAM,• JOSEPH GOELLNER,' AND GERALD B . PIER'
               From the Departn tents of Medicine and M[crobtology and Immunology. Washington University School of,Medicine and The
              Jeurtsh Hospital oJSt. Louis . St . Louts . MO 63l 10: and the'Channing Laboratory . Haroard Medical Schoo(. and the Brigham
                                                           and Women's Hospital . Boston . MA 02115

               Previous studies have demonstrated in vivo that                               cells in protection against these bacteria .
             T cells can provide protective immtutity, in the ab-                              We have identified an inbred motAse strain . BALB/c.
             sence of antibody . against infection with the extra-                           that is incapable of generating an antibody response to
             cellular Gram-negative bacterium Immunotype I                                   low doses of a high m .w . polysaccharide (P5)' antigen
             (IT-I ) Pseudomortas aeru_glinosa . We established an                           isolated from broth cultures of P. aerk.g(nosa (4) . BALB/
             in vitro system in which immune T cells, after re-                              c mice immunized with low doses of PS could not resist
             exposure to bacterial antigens and to macrophages .                             challenge with the live organism arly more effectively
             secrete a product that kiLls the bacteria . Although                            than unimmunized control mice . On theotherhand, mice
             macrophages are required, for in vitro killing, they                            from another inbred strain . C3H/ANF . did produce anti-
             function neither as antigen-presenting nor as phag-                             body after low dose PS immunizatiott ; and when com-
             ocytic cells in this system. T cells from animals                              pared to unimmunized control mice . were protected
             imsnunized against a different P . aeruginosa im                               against live bacterial challenge .
             muenotype will not ki .U lT-,l organismst but the su-
             pernatants produced by IT-1 immune T cells after                                  BALB/c mice that received the cytotoxic agent vinblas-
             exposure to macirophages and IT-I P . aeruginosa                               tine sulfate at the time of low dose PS immunization
                                                                                            failed to produce antibody . but they were protected

         i   organisms are nonspecifically effective in killing
             unrelated bacteria . Because the supernatants from
             immtrne T cells lose their bactericidal properties
             upon minimal tlilution, we conclude that if this
                                                                                            against live bacterial challenge . The pratective immunity
                                                                                            observed in these mice was T cell-mediated (5) .
                                                                                               To understand the mechanisms by which T cells par-
             mechanism is active in vi»o, It must play a role in                            ticipate In the destruction of extracellular ba~cteria, we
             local immunity .                                                               established an in vitro scstem In whicli (mmune T cells
                                                                                            kill P. aeruginosa . We now describe, this system and
                                                                                            define the mechanism b_v which T cells kill these extra-
               Protective immunity to extracellular bacteria has tra-                       cellular bacteria .
             ditionally been attributed to the activity of antibody .
             complement . and phagocytic eells. Other than function-                                            MATERIALS AND METHODS
             ing as helper cells for antibody responses . T cells have                        Bacteria . Fisher-Devlin lmmunot .pes 1 and 4(IT-) and 1T-4) of
             been thought to beof little Importance In such Immunity .                     P . aerugtnosa (originally provided by M Fisher . Parke-Davis C.o .
             T cells do play a critical role in resistance to intracellular                Detroit . 1611) a•ere grown overnight in 50 ml trlypttcase-soy broth
                                                                                           Bacteria from this overnight growth were inocetlated fnto 20 ml of
             bacteria, and analysis of this resistance has provided                        fresh broth to obtain a relative opucal density of 0.05 OD units (35
             new insights into the understanding of basic immune                           Spectrophotometer . Perkin-Elmer Cbrporatton : Coleman Instru-
             mechanisms (reviewed In Reference 1) .                                        ments Drvision .Oak Park . IL) . and were allowed togrow 4oa density
                We have been studying prbtective immunity to the                          of 0 .20 OD units- These bacteria were then harvested, washed once
                                                                                           ln 0 .15 M NaCI . and resuspended to a concentration of 1 O .D . unit .
             extracellular Gram-negative bacterium . Pseudomonas                          which routinely contained 2 to 3 x 10° colony forming units of P .
             aeruginosa (P. aeruginosa) . These bacteria contribute                       aerugtnosa . The bacteria were then diluted In medium and added to
             significantly to the morbiditv and mortality of irnmuno-                     microwells to assess In vrrro killtng . Staphyloccocus aureus (Amer-
t                                                                                         Ican Type Culture Collection )ATCC) 25923) . E . colt (ATCC 25922).
             suppressed patients. burn paYients . and children with
                                                                                          and a strain of P . aeruginosa highly resistant to gentarnicin were
             cystic fibrosis . Most studies of protective immunity                        obtained from stocks maintained In the Bacteriology Laboratory of
             against P. aeruginosa have focused on the role of anti-                      The Jewish Hospital of St . t .outs. St . Louis. MO .
             body directed against the bacteria or their toxic products                       Mice- BALB/c mice were obtained from the Animal Facility of The
                                                                                          Jewish Hospital of St- Louis . and from Cumberland Vifw Farms .
             (2 . 3) . Few studies have examined the importance of T                      Clinton . TN .
                              I                                                             Cell separation . T cells were prepanrd according to Wysocki and
                Rece ived for publication January 3 . 198-0                               Sato (6) . Brlefly. 3 x 10' spleen cells from BALB/c mice were placed
                Accepted for publlcatlon Aprt1 23 . 198-t .                               for 70 mtn on petri dishes that had been precoated with 50w1 affinity-
                The costs of publication of this arucle were defrayed in part by the      purified goat anti-mouse tmmunogiobulin (1g) antiserum (Gateway

         •   pay-ment of page charges. This article must therefore be hereby marked
             aduerttsement in accordance with 18 U .S .C . Sectton 1734 solely to rndi-
             cate this fact .
                                                                                          (mmunosera. St . Louis. MO) . The nonadherent cells were collected
                                                                                          and placed for 70 min on a second anti-mouse Ig antiserum coated
                                                                                          petri dish . The nonadherent cells collected from this second cycle of
                ' Thls work was supported by a grant from the Council for Tobacco         adherence routinely contained fewer than 5% cells that reacted with
             Research . by funds provided by the Auwliary of the Jewish Hospital of
             SL l.ouis . by Publtc Health Service Grants At-15353 and At-15835 from          ' Abbrevrations used in this paper : ATCC. American Type Culture
             the Nattonal Institutes of Health. and DAMD-17-79C-9050 from the U .S        Collection : IT-l . immunotype 1 : IT-4. immunotype 4 ; PS. high m .w.
             Army Research and Development Command .                                      polysacchandetsolated from P . aerugtnosa broth cultures-

                                                                                PUBLICATIONS 024480
                                                      IN VITRO KILLING OF P . AERUGINOSA                                                            963
      fluorescent goat anU-mour lg antiserum . The response of the non-          the ne .xt day . Resident peritoneal macrophages were har-
      adherent cells to the T cell mitogen concanavaltn A(Con A : Phar-          vested from non-imniune mice . allowed to adhere to mi-
      maeta Fine Chemicals . Ptsrataavay. NJ) was equivalent to the re-
      sponse of unseparbted cells . whereas their response to the B cell         crowells overntght . and washed free of nonadherent cells .
      mitogen lipopolysaccharidt (LPS) . obtained from Salmonella enter-         To microwe)!s containing various cell combinations wk
      tditts (Dlfco. Iktro11- MI) was routinely reduced by at least 75% when     added 2 x 102 bacteria and compared their survival aftei
      compared to the rpsponse of unseparated spleen cells . After the
      second cycle of adiurence, nonadherent cells were incubated over-          4 hr incubation to bacteria surviving In wells containing
      night at 37°C in 20 mi tissue culture flasks in a 5% COs atmosphere .     only tissue culture medium . The combinaUons were Im-
      and the nonadherent cells from this overnight Incubation were              mune T cells alone, resident macrophages alone . or both
      harvested and washed . Fewer than I of 300 of these cells were
      nonspecific esterase stain-posltive .                                     resident macrophages and immune T cells . We also stud-
        Macrophages were obtatned by lavagtng the peritoneum of non-            ied bacterial survival In the presence of resident maP-ro .
      immune BA1 .B/c mice with tlanks' balanced salt solution (HBSS) .         phages. immune T cells . and ] 0a heat-killed bacteria :
      These resident ce[Is, were allowed to adhere overnight in mtcrowells .    Both macrophages and immune T cells had to be present
      and the nonadherer't cells were then washed free the next day with
      warm medium . The remaining adherent cells were routinely 93 to           to effect a significant (p < 0 .05) reduction in viable bac-
      95% nonspecific esterase-poaluve .                                        teria. compared to wells containing only medium (Fig . 1) .
        T cells were depleted from immune spleen cell populations by            The number of bactesia remaining In wells containing
      Incubation for l hr at 4°C In a monoclonal ant/-Thy-1 .2 antibody
      (dlluted 1/250 In cytotox/clty medium) followed by a 45-min incu-         only immune T cells or only macrophages did not differ
      batton at 37°C In low tox, rabbit complement diluted 1/10 . The           significantly from wells containing medium only (p >
      reagents used in these Tvel"epleting steps were obtained from             0.30) .
      Accurate Scientific and Chemical . Westbury . NY . The efficiency of         In similar experimehts . survival of bacteria In wells
      the depleuon ts Indicated In the legend to Table 1 .
         Reagents. The bacterial PS used in these studies were prepared         containing only resident macrophages was quite variable .
       by described methods 15 . 7). VtnblasUne sulfate is a product of Ell     but was always significantly greater than In the presence
      Lilly Co. . IndianapolAs .1N .                                            of resident macrophages and immune T cells . Further-
          TYssue culture medium. R.P5t1 1640 medium (GIBCO. Grand Is-           more, the number of viable bacteria in wells containing
      land. NY) was supplemented witlt 2.0 mM t-glutamine . 25 mM
      HEPES buffer (GIBCO). 50 pg/ml gentamicln sulfate (Schering Corp . .      only immune T cells was always significantly greater
      Kentlworth . N.l)- and lO% heat-inactivated fetal calf serum (KC          than In wells containing both macrophages and immune
      Biological. Inc . . Lenexa. K5,). which had been absorbed six Umes        T cells .
      a-1th 1 O° P. aeruglnosa/ml to retnoveany anUbody to these bacteria
      that might be present in the serum. Before absorption, no antibody          Theaddition of 10° heat -killed bacteria to macrophages
      could be detected In this serum by a radtolmmunoassay that detects        and immune T cells did not further decrease bacterial
      antibody to the PS antigens {8) .                                         survival. suggesting that 2 x 10' live bacteria are suffi-
         Assay of bactertal killing . T celltnrlched lymphocyte popula-         cient to stimulate the immune response . On the other
      tions collected from overnight incubaUon of doubly 'panned' spleen

4b    cells were washed and placed In anOblotlrfree tissue culture me-
      diurn at a concentration of 4 x lOs/ml . Then 0.1 ml of these cells
     was placed in flat-bottomed microculture wells (Falcon 3072 . Bec-
                                                                                hand . the observation that 10e heat-killed bacteria did
                                                                                not significantly increase bacterial survival provides ev-
                                                                                idence that the reduction in viable bacteria is not medi-
      ton-Dickinson Labware. Oxnard. CA) that either did or did not con-
     tatn adherent cells from the peritoneal lavages of non-immune mice .       ated by antibody or any other antigen-specific product .
      Live P. aerugtnosa (I to 2 x 10') In a volume of 0 . 1 ml of tissue       Such a huge excess of dead bacteria would be expected
     culture medium weree then added to the wells containing various            to adsorb and deplete any antigen-specific product in the
     combinations of cells, and/or medium . Unless otherwise indicated .        medium . unless that product were present In great eio-
     the bacteria used in all experiments were Fisher-Devlin IT-1 . After
     a 4-hr tncubation at,?7°C in a moist 5% COr atmosphere . the wells         cess.
     were pipetted vigorously and then a 50-µl sample was removed and             In the experiment show-n in Figure 1 . 2 x 102 bacteria
     plated on trypUcase-soy agar . A second 50-NI sample was removed .         had been added to each well . After 4 hr incubation . the
     diluted 1/10/n HBSS . and plated on agar The next day . colonies on
     each plate were counted and the number of surviving bacteria in
     each well was calculated. For those studies in which culture super-
     natants were used to assay kllling . either I 0' heat-k/lled or 10' live
     bacteria were added to the wells in which the supernatants were
     prepared . After 4 hr .,the supernatants were collected and passed
     through a 0.45-pm sterile ftlter (Gelman . Ann Arbor. MI) to remove
     bacteria and cells . Then 0 .1 ml of each supernatant was added to
     0.1 ml of the bacteria (10'/ml in ttssue culture medium) In flat-
     bottomed mlcrocultur» wells and bacterial survival was assessed
     after 4 hr tncubatton at 37°C In a 5`a COo atmosphere .
        In all experiments. the percent of bacteria killed was determined
     by subtrsctingthe mean number of bacteria surviving in experirnen-
     ta1 wells from the mean number surviving in wells containing only
     medium, dividing this difference by the mean number of bacteria
     surviving In the media control wells . and multiplying by }00.
        Statistics. A one-way analysis of variance using an F test (9) was
     employed to analyze the difference between experimental groups . P
     values < 0 .05 were considered significant .


       In vitro T ceTl-mediated killing of PA aerug(nosa . We
     first Investigated the cell combinations and conditions
0    required for in vitro T cell-mediated killing of P . aerugi-                                           CELLS IN CULTURE
     nosa IT-1 . BALB/c mice were injected with 125 kg vin-                        Figure 1 Killing of 2 x 10° P . aerugtnosr: (P.a.) in the presence of
     blastlne (i .v .) and 10 µg 1T-1 PS (i .p.): 6 to 7 days later,            macrophages (M0) plus tmmune T cells: immune T cells alone: macro-
                                                                                phages alone: or macrophages. immune T crlls . and 10' heat-killed P .
     their spleen cells were harvested, enriched for T cells .                  aeruginosa . Control wells contained only tissue culture medium . Each
     cultured overnight . and then added to microwell cultures                  point represents the mean s SEM of four mlcrowells .

                                                                       Pi1BLICRTICNS 024481
         964                                                  IN VITRO KILLING OF P. AERUGINOSA

         wells with medium alone contained a mean of 532 bac-                             mice . We added 2 x 102 live IT-4 or IT-1 P. aeruginosa
         teria, in contrast to a mean of 216 bacteria in wells with                       to these cult ures and assayed bacterial survival4 hr la'ter .
         macrophages and Immune T cells . Subsequent experi-                              The results (Table 111) show that macrophages and dT-4
         ments Indicated that bacterial killing. rather than growth                       immune T cells effectively kill IT-4 organisms (98r
         inhibit.ion, occurs In wells containing macrophages and                          killed). although these macrophages and IT-4 immttne T
        immune T cells .                                                                  cells are ineffective in killing 1T-1 organisms (21 % killed) .
i          Cofn partsdn of ktll tng by im rrdu ne T cells and immune                      We conclude that activation of T lymphocytes for in vitro
        B cells . We next compared bacterial killing by T cell-                           killing requires re-exposure to the homologous immuniz-
        depleted Immune spleen cell populations with bacterial                            ing antigen .
        killing by B cell-depleted immunc spleen cell populations .                         Killing by supernatants from cultures of [mmune T
        The data (Table I) tndicate that, when compared with                             cells and nfracrophages. To examine whether the ob-
        media controls . 60% of the bacteria In wells with mac-                          served killing resulted from the secretion by cells of a
        rophages and B cell-depleted populations were killed .                           lethal product or from direct Interaction of bacteria and
        Only 31 % of the bacteria were killed by macrophages and                         cells, we assayed the ability of cell-free culture superna-
        T cell-depletcd populations (p = 0 .01) . Thus, macro-                           tants to kill P. aeruginosa . Macrophages and T cells were
        phages and T cell-enriched populations of immune spleen                          prepared as in previous experiments and five different
        cells are more effective at killing bacteria than are mac-                       groups of microwell cultures were set up (Fig . 2) : 1) mac-
        rophages and T cell-depleted populations .                                       rophages . Immtule T cells, and 103 heat-killed P . aerugi-
    t      Comparison of killing by immune and non-immune T                              nosa : 2) macrophages and tmmune T cells without bac-
        cells . We next cmmpared the abi)ity of Immune and non-                          terla : 3) Immune T cells and 101 heat-killed P. aerugi-
        Immune T cells to promote bacterial killing (Table II) .                         nosa ; 4) macrophages and 109 heat-killed P . aeruginosa :
        Wells containing non-immune T cells and peritoneal                               and 5) 101 heat-killed P. aerugtnosa in medium without
        macrophages from non-immune mice contained 28%                                   cells .
        fewer viable bacteria than control wells . Wells with T
                                                                                                                       TABLE ill
        cells from imprlunized mice and macrophages from non-                             Spec!ftctty of protection achieved by tmmunization with IT-A PS and
        Immunized mice contained 63% fewer bacteria (p = 0 .02 .                                                      utnblast : ne
        non-immune vs immune T cells) . These observations are                                                                  1T• I Bacteria           IT-4 Bacteria
                                                                                                Oells In Culture
        consistent with our previous results showing that non-                                                              Survlving° (g6 killed)   Survlvtng° f4b luuedl
        Immune BALB/c mice are highly resistant to P . aerugi-                                       Mm                  3 .400 f 138 (15)°    3.824 f 166 (17)°
        nosa Infection and that this resistance can be enhanced                             MO + 4 x 10° T cells        3.160 t 106 (21)°          92 ± 14 (98)-
                                                                                            MO + 2 x 10" T cells        3 .436 t 183 (14)      1 .224 * 196 (74)
        by immunizaUon with PS and vinblastine (4 . 5) .                                    Mm + l x 10° T cells        3.954 3 190 (1)        3 .464 Y 114 (25)
           Specificity ~nf killing by T cellsfrom mice immunized                            Media alone                 4.000 3 186 (-)        4 .624 ± 260 (-)
        with PS and ufnblastine. BALB/c mice were immunized                                ° Mean :t SEM of bacteria surviving in four mlcrowells .
        with a PS isolated from Fisher-Devlin IT-4 P . aeruginosa,                         ° p> 0-3 for significance of difference in killing of IT-1 P . aentgtnosa
                                                                                        between wells containing macrophages (Mm) and wells containing Mo and
        and 6 days later, their T cells were harvested and placed                       4 x 10° IT-4 immune P. aerugtnosa-
        In tissue culture with macrophages from non-immune                                 ' p value <0 .O00I for significance of difference in killing of 1T-4 P .
                                                                                        aeruqlnosa between wells containing Mo and wells containing Mts and 4
                                        TABLE 1                                         x 100 IT-4 /mmur9e T cells .
         Comparison of ktiUn_0 of P . aerugtnosa by tmmun.e T cells vs immune
                                         B cells
          G~                  Cetls in Final
                                                           Survlvl           p Value°
                P                Cltlture
                                                       In Well i°~ kl~ing)
                    None (media control)                189 :t 1 O° (-)      <0 .0001
            I       Macrophages                         173 m 5 (81          <0 .0001
            3       Antl-Thy-1 .2-treatedimmune         131 f 14 (31)         0 .01
                      spleen cells + enscrophages'
            4       AnU-Ig-treated lmmune spleen         76 t 13 (6)
                      cells + macrophages'
          ° This value expresses the significance of the difference between the
        number of bacteria surviving In wells contalning antl-Ig-treated immune
        spleen cells and macrophages (group 4) and the number surviving in the
        given experimental group.
           °Mean :t SEM of number of bacteria surviving in four mtcrowells
           ° Prollferative responses (± SEM) to T and B cell mitogens of the
        separated cells were : group 3: Con A 55 .977 t 5904 . LPS 35 .867 f 1034 :
        group 4 : Con A 137.014 :t 9987. LPS 2 .670 ± 372 : unseparated spleen
        cells : Con A 218.508 ± 3503. LYS 60.470 s 1509.

                                    TABLE n
        Kfntng oJP. aerugtn.osa by TcellsJrotn immunized vs non•tmmunlzed
                                       mice                                                                    T Cens .   T Cdls   TCNn               MeO~o
                                                                                                                             .       .                   .
                                                     Number of Bacteria                                        MO .Pa
                      Cknn in Culture                                                                                       tl~     PO                  P.a
                                                     surviving f% kiIlingl   P Value°
                                                                                                                      SUPERNOTE SOURCE
         None (media control)                         836 t 48' (-)          <0.0001
         immune T cells and macrophages               308 t 8 163)                         Figure 2. Killing of I x 102 P . aeruglnosa (P .a .) by supernatants from
                                                                                        mlcrowelts eontalnln various combinations of Immune T cells. macro•
         Non-tmmune T cells and macrophages           604 * 44 (28)           0 .0005   phages (Mm) . and l0~ heat-killed tracterla- After 3 hr co-incubation of
          ° This value expresses the significance of the difference between bac-        cells and bacterla . the supernatants were collected . transferred to ml-
        teria surviving ln wells containing immune T cells and macrophages              crowells. inoculated with I x 10° 1[ve bacteria, and reincubated for 3 hr .
        compared with the other experimental groups .                                   Control wells contained fresh tissue culture medium. Each point repre-
          b Geometric mean ± SEM of bacteria surviving in four mlcrowells               sents the mean 3 SEM of four microwells .

                                                                                   PuBLICRrXONs                                                          024482
                                                    IN VITRO KILLING OF P . ;4ERUGINOSA                                                965
             After 3 hr incubation . 0 .1 nll of the supernatants from TABLE IV
         the five different cultures was placed in mieroculture.Ability of T cells with rnacrophages to ktll P . aierugtnosca after previous
                                                                 The Incubation and and without heat•kutea txicFerw
         wfth 102 live P. aeruginosa in 0 .1 rrrl of medium                .p e`l>8 in F„'al c~'h°rc sa,.t""~ sti","`"n~ P v8'°`~
   ~ control group eont.~ined fresh tissue e~ulture medium in- cr°`
                                                                          1 lImmune T cells previously cultured !8 t 4(99) <0 .0001 (2)
         oculated with 1021 ive P . aeruginosa . After 3 hr incuba- wuh macrophages and heat-lulled
          tion . the number of viable bacteria was determined for bacteria
                                                               con- 2 Immune T cells not prevlously cul- 456 3 76 (29) -
          each supernatant group. When compared with the . mac- ~y`~ ""n""ac"°pt~ag`s or bacte
          trol . the supernatant from wells containing T celis
          rophages . and bacteria reduced the number of viable 3 Immune T cells previously cultured 29 . t 8(96) <o.oool (2)
                                                                         only with macrophages
          bacteria by 52% ; this killing was significantly greater 4 Macrophages previously cultured 236' s 32 (63) 0 .07 (5)
          than that observed in any of the other groups (p < 0 .005) . with Immune T cells and heat-
          Supernatants from wells containing T cells and macro- killed bacteria
                                                                     5 Macrophages not previously cultured 344 s 50 (46)
          phages killed 21 % of P . aerugfnosa, those froth wells with T cells or bacteria
         containing macrophages and bacteria killed 23% : those 6 Media alone (control) 640 :t 44 i-) -
         from wells containing T cells and heat-killed bacteria . ° Mean ± SEM of bacteria surviving In four mtcrowelts .
         29% ; and those from wells containing heat-killed bacteria °Thisvalueexpressessignificanceofdtfferencetxtweenthepanicular
                                                                      experimental group and the group indicated within the parentheses .
         but no cells . 6% of the P. aeruginosa .
           There are two points of note in these results : First, been exposed to macrophages and bacteria (group 1)
         bacteria] killing in this in vitro system is mediated by a killed 99% of the bacteria (p < 0 .000) . group 1 vs group
         soluble product and does not require phagocytosis by 2) . These data indicate that the final bactericidal product
         macrophages . Second . maximal killing occurs with su- resulting from the interaction of immune T cells and
         pernatants obtained from cultures cor .ltaining T cells . macrophages is a T cell lymphokine.
         macrophages . and bacteria, whereas supernatants from A surprising finding was that immur9e T cells exposed
         cultures containing T cells and macrophages but rio bac- to macrophages In the absence of bacteria could, upon
        teria produced significantly less killing [p = 0.0003) . subsequent exposure to live bacteria, reduce the number
           Determination of which cell secretes the soluble bac- of viable organisms by 96% compared with media con-
        terictdat product. To determine whether the soluble bac- trols (p < 0 .0001 . group 3 us group 6) . This observation
        tericidal product was a lymphokine or a monokine, we suggests that the critical interaction between macro-
        conducted two separate experiments . To analyze which phages and immune T cells that must occur to achieve
        cell was responsible for the bactericidal effect observed optimal T cell bacterial killing does not involve presen-
        in the presence of immune T cells, we assayed three tation of antigen to T cells by macrophages . In fact, T
 ~ groups of immune T cells (4 x l05/well) and two groups cells do not have to be simultaneously exposed to mac-
        of macrophages . rophages and bacteria to be activated for bacterial killing .
           The Immune T cell groups were : 1) T cells that had although exposure of T cells to macrophages is clearly
         been incubated previously for 3 hr with 4 x 103 macro- required for the T cells to kill the bacteria to which they
         phages and l09 heat-ltifled bacteria : 2) T cells that had are subsi'quently exposed (p < 0 .0001, group 1 vs group
         not been incubated with either macrophages or bacteria ; 2) .
         and 3) T cel)s that had been incubated fot 3 hr with 4 x The reduction in viable bacteria in this experiment was
         103 macrophages without bacteria . The T cells were har- so profound that it enables us to confirm that the reduc-
         vested from their respective cultures, washed . and placed tion in bacterial survival observed in these in vitro ex-
         in fresh culture medium with 10' live P. aeruginosa ; periments actually represents bacterial killing . We added
         bacterial survival was determined after 3 hr . 2 x 102 live bacteria to the cells in group I and only eight
            The two groups of macrophages were : 1) 4 x] 03 mac- bacteria remained in these wells after 3 hr incubation .
          rophages that had been incubated with 4 x) 05 immune Although this experiment suggests that the T cell se-
         T cells and 10' heat-killed bacteria . and 2) 4 X 103 cretes a bactericidal product, we sought to confirm this
         macrophages that had t,ot been exposed to T cells or conclusion directly . We therefore again co-cultured resi-
         bacteria . The macrophages were washed with warm me- dent peritoneal macrophages and immune T cells over-
         dium to remove T cells and heat-killed bacteria . and then night, and the following day we separated the T cells
          102 live P. aeruginosa ^were added to the wells . Bacterial from the adherent macrophages . After washing . each cell
         survival was assayed after 3 hr incubation at 37°C . A type was placed in separate microcultures with fresh
         media control group was Inoculated with 102 live bacte- media and 300 live P . aeruginosa . After 4 hr, these
         ria . and bacterial survival was assayed at the end of 3 cultures were harvested and filtered to remove bacteria
         hr .            and                      cells . We then examined the ability of the different
          The results (Table IV) show that resident macrophages culture supernatants to kill P . aeruginosa (Table V) . We
        that had not been exposed to T cells and heat-killed also investigated whether the combination of T cell su-
        bacteria (group 5) killed 46% of the bacteria when com- pernatant and macrophage supernatant produced syn-
        pared with control wells . Macrophages that had been ergistic killing .
       exposed to immune T cells and heat-killed bacteria (group Supernatants obtained from T cells that had been in-
~ 4) killed 63% of the bacteria compared with the control, cubated overnight with macrophages . separated from
       but this number did not differ significantly from the 46% those macrophages . and then exposed to live bacteria,
       killing seen In group 5 (p = 0 .07) . On the other hand, killed 50°-0 of the bacteria compared to a control well
       immune T cells that had not been exposed to macro- containing only tissue culture media . Macrophage super-
       phages or bacteria (group 2) killed 29% of the bacteria natants . obtained from macrophages that had been ex-
       compared to the media controls, whereas T cells that had posed to T cells overnight and subsequently to live bac-

                                                                 PUBLICRTIONS 024A~3
 966           IN        VITRO                   KILLING                        OF     P . AERUGINOSA
                                TABLE V                                                IT-1 . in a separate experiment . we compared survival of
      Ability of cuiture supernatarttsJrom isolate¢t populations oJ
                                                                                       our original strain of P. aert.lginosa IT-1 vvith a strain of
          me.crophages or intrnune T cells to kill P : aerugtnosa ,
                                                                                      P. aeruginosa that is highly resistant togentamicin . The
          Supernatant Souree°                                           p Valu e       latter experiment was performed to detei-mtne whether
                                              Surviving 146 knledl'
     T cells                                    513                                   the killing we were observiing was due to the overnight
                                                      :1- 79 (50)          -
     Macrophages                                804   t 71 (21)         0.02          incubation of the T cells and macrophages in me¢iium
     T cells and tnacrophages°                  653   f 18 (36)         0 .2          containing gentamicin . The results (Fig . 3) demonstrate
     Medta control                             1017   ~ 22 (-)          0 .0~08
                                                                                      that the killing is not specific . The supernstant was able
    ° Supernatants were prepared by oo-lncubaung vnacrophages and Im-
 mune T cells overnight . separating and washing the two ce11 populations .           to reduce the number of viable S . auretis and E. colt .
 exposing them to live bacteria for 4 hr . and then ha.rvesting and filtering         compared with their respective media controls . even inore
 the supernatant . All cultures contained 200 ul of swpernatant .                     effectively than it reduced the number of viAble P. aerllgi -
    ° Mean Y SEM of bacteria surviving in three microwells .
    `This value compares the significance of dtff0rences between the                  nosa. The gentamicin-resistant strain was also killed b}•
 bacterial survival In wells containing T cell supernatants and the bacte-            the supernatant, confirming that the killing was not
 rial survival in the other experimental groups .                                     related to the previous exposure of immurAe cells to that
    d This group of wells contained 100 kl of macroptiage supernatant and
 100 Iil of T cell supernatant .                                                      antibiot ic .

                               TABLE VI                                                                                 DISCUSSION
    Abtltry oJdituted superrtatantsJrom cultures o,~ tmmune Tcells.
    ntacrophages . and heat-killed bacteria to kill live P. aeruginosa In this report we described the in vitro conditions undct
    Supernatant          Bacteria Surviving        p Value vs       which T cells can kill the extracellular bacterium F
                                                                      p Value vs
      Dilut/on               f°b lullcd)"         Media Control     aeruginosa . The killing is mediated by immune T cells .
                                                                      1A Dilution
      Media            1 .256 3 146 (-)         -          0 .0002  which must interact with nAacrophages In vitro before
      ' .2                664 3 12 (47)      0.0002           -     optimal bacterial killtng can occur . Macrophages are, not
      1/4                 824 t 26 (34)      0 .003 .      0 .2
      1ie                 7362 42(41)        0 .0006       0 .5     required, however, during the exposure of T cells to bac-
      V12              1 .0081 128120)       0 .06         0.01     teria, ilhdicating that in this system . macrophages do not
   ° Mean :t SEM of bacteria surviving in four microwells. function to present antigen to T cells . T cells from tfliee
                                                                    immunized with an unrelated P . aerugino5a PS are not
teria, killed only 21 % of the bacteria . sigtlificantly fewer      stimulated to kill IT-I P. aeruginosa in vitro. Our im-
bacteria than were killed by the T cell supernatants (p =           munization protocol does not induce detect .able antibody
0 .02) . Adding macrophage supernatant Io T cell super-             production in vivo . and in viRro B cells are significantly
natant did not significantly alter the killing by the 'T cell       less effective than T cells in promoting bacterial killing_
supernatant (p = 0 .2 . T cell supernatant vs combined T            Soluble factors in supernatants from cultura's of immiune
cell and macrophage supernatants) . and if anything,                T cells that had been exposed in vitro to both macro-
served only to dilute and reduce the bactericidal effect of         phages and bacteria can kill the bacteria in a cell-free
the T cell supernatant . This experiment therefore pro-             environment . These supernatants are nonspecific In
vides direct evidence that T cells secrete a bactericidal           their activity .
factor, the effect of which is not enhanced by macro-                 Our system represents a heretofore undescribed mech-
phage supernatants.                                                 anism by which immune cells can kill bacteria . Described
   Ability oJdfferent concentrations of supernatant to              n,echanisms of killing of extracellular bacteria include
kill bacteria . To determine the potential physiologic role direct lysis by complement (10) . phagocytosis by macro-
of the bactericidal product secreted by T cells, we asked
whether this product is produced in concentrations that
would be effective systemically in vtvo . We compared
killing at various dilutions of supernatant with the killing
observed In media control wells and with the killing
observed in wells with the highest supernatant concen-
tration . Because live bacteria had to be added to the
undiluted supernatants to evaluate killing, the most con-                                                      a
centrated supernatant evaluated was 1/2 . The results
(Table VI) demonstrate that significant killing . In com-
parison to media controls, is achieved until the superna-
tant is diluted 1/12 . At that point the killing is also
significantly less than that achieved with the highest                                                              ,
supernatant concentration (1/2 dilution) . This observa-
tion suggests that the lethal component in the culture
supernatants Is not potent enough to be effective system-
ically, and if it does function in vtvo, it functions, like
other lymphokines . only over short distances .
  SpeciJicity of killing by culture supernatants . To in-                                                                                 pp
vestigate whether the bactericidal T cell product is spe-
cific in its activity . we collected supernatants from 3-hr
                                                                                                                   BACTERIAL STRAIN

                                                                                       Ftgure3. Ability of supernatantsfrotn microwells containmgmacxo-
cultures of immune T cells, macrophages, and heat-killed                            phagesg Immune T cells, and heat-k111ed P . aerctg[nosa (P.a .) IT• I to kill
P. aerugtnosa, and placed 100 jul In microwells contain-                            other bacteria in each experlment . the survival of bacteria In wells
                                                                                    containing the cuiture supernatant and the given bacterial strain was
ing approximately 102 colony-forming units of either S .                            compared to wells containing fresh tissue culture medium and that strain
aureus . E . colt, or the original strain of P. aeruginosa                          of bacteria . Each point represents the mean :t SEM of four microwells .

                                                                             PUDOLICRTI0N5                                                     024484
                                                         IN VITRO KILLING OF P.               AERUGINOSA                                  967

          phages or granulocytes of opsonizL-d bacteria (11) . or                 challenge with live bacteria (5) . We can only speculate on
          destruction of antibody-coated ba .ctt:;ria by various types            whether the mechanism of In vitro T cell killing we
~ of killer cells (12) . None of tl~d/se mechanisms plays a role                  describe here is the same as that involved in in vivo
       in the current system of kilf,ing .                                        protection . It appears that if It functions in viao, this
          Neither antlbody noi- complement is required to achieve                 mechanism of T cell killing must be active only in local
       killing in our assay . The BALB/c nilce from which the                     im¢nunity, at sites at which T cells. macrophages, and
       imrtlune T cells were derived do flot produce detectable                   bacteria would eongregate, and at which the lymphokine
       serum antibody In response to tmmunization (5) . an ob-                    would be present in concentrations great enough to kill
       servation that was reconfirriied bn Rhe present studies                    bacteria . The rapid loss of bactericidal activity with di-
       (data not shown). Fewer than 5% of the cells In the culture                lution of the lymphokine suggests that, like many other
       are B . and the addition of large numbers of heat-                  lymphokines, this bactericidal factor cannot function at
       killed bacteria to the cultures, which would be expected                   long distances or within the bloodstream, vtyhere its activ-
       to adsorb any antibody secreted by contaminating B cells .                 ity might be diluted . T cell immunity to extracellular
       does not reduce the magnitude of the observed killing .                    bacteria could then function In a complementary manner
       Furthermore, the fetal calf serum used in the culture                      with antibody, which does promote destruction of bacte-
       medium was heat-inactivated and adsorbed multiple                          ria that enter the circulation .
       times to ensure that no exogenous source of antibody or                      Studies In rats (17) and mice (18) have demonstrated
       complement was present . The direct addition of B cells                    that T cell immunity is critical In resistance to infection
       to this assay system yielded significaritly less killing than              with another extracellular bacterial pathogen . Bacte-
       was observed with B cell-depleted . T ce .ll-enriched popu-                roidesJragilis . The pathologic features of Infection with
       lations .                                                                 B. fragilis, characterized by the development of ab-
         '1'wo points about this killing mechanism are of partic-                scesses, are distinct from that of P. aerugirtosa . and
       ular interest . First, the final product that kills the bac-              preliminary reports indicate that the T cell mechanism
       teria is a lymphokine . Most studies of bacterial killing                 invcalved in killing of B.Jragilts is complement-dependent
       have focused on macrophage or granLilocyte products .                     and mediated by a non-secreted, antigen-specific T cell
       either lysosomal or extracellularly secreted (13 . 14) .                  product (19) . It appears that more than one T cell mech-
       Whereas T cell products such as interferon may be potent                  anism may be involved In resist .ance to infection with
       inhibitors of viral replication (15) . the notion that T cells            extracellular bacteria .
       may secrete bactericidal products is a novel one .                           The Importance of T cell immunity In resistance to
         Second . our observations define a new role for the                     infection remains to be determined . There are well-de-
       macrophage in bacterial killing . In this system, some type               scribed clinical situations In which localized infection
ft of lnteraction between macrophages and T cells Is re-                         develops and progresses despite circulating antibody (20-
       quired before T cells can respond to bacteria by secretion                22) . Most relevant to P . aeruginosa are children with
      of a lethal product . In studies of bacterial kllling, macro-              cystic fibrosis, who develop chronic, localized P . aerugi-
      phages have traditionally functioned to present antigen                    nosa pulmonary infections despite their extraordinarily
      to T or B cells and to phagocyt ize bacterio . Because in the              high levels of antibody to the infecting organisms . Al-
      current system the necessary Interaction between T cells                   though these high antibody levels prevent bacteremia .
      and macrophages can occur in the absence of antigen .                      they do not prevent the recurrent exacerbation of lung
      macrophages are not functioning to present antigen to T                    disease associated with P . aeruginosa infectton In cystic
      cells . Furthermore, because bacterial killing occurs in                   fibrosis patients . Whether a failure of local T cell immu-
      the absence of macrophages . those cells do not have                       ntty accounts for this persistent infection 1s of consider-
      phagocytic function . Characterization of the essential                    able interest and can now be readily evaluated by adap-
      role macrophages play In this killing mechanism requires                   tation of this in vitro system to address the clinical
      further study .                                                            problem .
         How vinblastine functions in the elicitation of this
      Immune response is, as yet . undefined . From preliminary           Acknowledgments . We thank Drs . Judith Kapp and J .
      evidence (unpublished observations) . we suspect that by        Russell Little for critically reviewing this manuscript .
      removal of a suppressor cell, this agent enhances prolif-
      eration of the clone of immune T cells that responds to                                         REFERENCES
      the immunizing PS . This hypothesis is based on the               l . Hahn . H. . and S . H. E. Kaufmann . 1981 . The role of cell-medlated
      observation that T cells from vinbiastine-treated mice                tmmunlty In bacterial infections. Rev . InJect . Dis . 3 : 1221 .
      spontaneously proliferate In 'uitro, and that adoptive            2 . Pavlovskis. O. R. . M . Pollack. L . T. Cs+tlahpn, and B . H . Iglewski .
                                                                             1977 . Passtve protection by antitoxin in experimental Pseudornonas
      transfer of normal spleen cells to vinblastine-treated                aeruglnosa burn Infection . In/ect . lmmun. 18,596-
      mice prevents the subsequent in vitro proliferation . It         3. Alexandv. J .. and M. W. Pisher. 1974 . Immunlzauon against Pseu-
                                                                            domonas Infection after thermal 1njury . J. InJect . Dts . 130:5152 .
      would appear that by expandiing the clone of P . aerugi-         4 Markham . R. B . . and G . B . Pler . 1983. Characterization of the
      nosa-immune T cells, vinblas'tine acts to magnify and                 antibody response in lnbred mice to a htgh-molecular-welght pDly-
      make more observable the imhttune response that may                   saccharide from Pseudomonas aerugtnosa Immunotype 1 . Inject .
                                                                            Im mu n. 41 .232 .
     already be occurring after immuniiation with the PS               5- Pier . 0 . B., and R . B. Markham . 1982 . Induction (n mlce of cell-
     alone . This effect would be similar to that described for             mediated immunity to Pseudomonas aeruginosa by high molecular
                                                                            weight polysaccharyde and vtnblasttne . J. lmntunol . 128:2121 .
     cyclophosphamide with sheep red blood cells used as an            f; . Wysocki. L. J. . and V. L. Sato . 1978. -Panning- for lymphocytes : a
     antigen (16) .                                                         method for cell selection . Proc . Natt. Acad . Sct- USA 75 :2844-
        In a previous study, we showed that T cells from mice          7 . Pler, G. H . 1982 Safety and Immunogenidty of high molecular
                                                                            weight polysaccharide vaccbpe from lmmunotype I Pseudemonas
     given PS and vinblastine could,n in the~ absence of specific           aen.gtnosa J Citn . Invest . 69 .303 .
     antibody, adoptively transfer i protective immunity to           8. Pier . (i . B.. R. B . Markham. and D. D. Eardley . 1981 . Correlation of

                                                                         PuBl,IcHrrONS Q2AAAS _
968                                                 IN VITRO KILLING OF P. AERUGINOSA
      the blologycal responses of C3H/HeJ mice , to endotmttn with the          15. Frledmaa. R . M. 1977 . Antlv[ral activity of tnterfesvns . Bactertol-
      chemicay and structural properties of the Lpopolysacchar[des from             Reu. 41 :543.
                                                                                16- Shand. F. C., and F. Y. Lfew. 1980. lnfferentlal susceptlq111ty to

      Pseudontonas nerecginnsa and Escherlchta colt . J. lmrnuno4
      127:184 .                                                                    tt~ po ~hamlde of helper T cells for humoral and sup r T
 9. (1lAta . S. A. 1981 . Prtmer of Btostattsttcs . McGraw-Hlll Bpok Com-
     pany . New York . Pp. 30-45 .                                               17 . Ond frsdonl~ B ..~it. ~Markbam. DZatesnik, R. L. Cisaeros. and
]O. Mairau, S. C . . and R. C . Skarnes. 1975. Complement-medtated                    D. L ; Kasper. 1982. Eviidence for T-cSll dependent tmmuntty In an
     bactericidal system : evidence for a new pathway of complement                   Intra-abdominal abscerss model . J. Cltn. lnvest. ta.9:9.
     action . Sctence 190:278.                                                   18 . 9Dapiro.lN.'. B . . A. B. Onderdonk. D. L. Kasper. and R. W. Flpbag.
                                                                                      1982. Cellular Immunity to BacterofdesJragtlts capsular polysac-
11 . Horwit2. M. A. 1982 . Phagocytos/s of mtcroorganlsms. Reu. Infect.               charide . J. Exp. Med . 155:1188.
     Dis. 4 :1Q4.                                                                19. Zaltzalck. p. F. . R. W., Fiaberg. S . P. Ls11y. A . B. Onderdonk. and
12. Lowell . 0. H.. L. F. Smfth. >b . S . ArYeasteln: d. S . Nash. and R P .          D. L . Kasper. 1983. Cooperation of T celle and complement In
     MacDermott. Jr. 1979. Ant]body-dependent cell-mediated antlbar                   immunity to [nva-abdominal abscess formatton . Ctta Res . 31 :379A
     tertal activity of human mononuclear cells . 1. K lymphocytes and               IAbstr.l.
     monocytes are effective against mentngococcl !n cooperation with           20. Gorbach. S'. L., and J. O . Bartlett. 1974 . Anaerobic Infections . lY.
     human Immune sera. J. Exp. Med. 150:127 . ,                                     EngL J. Med. 90:1177. ,
13 . MwTay. lAi. W.. C. W. J,•aagi+hnni ..h . C. F. Nathan . and Z. A . Cohn.   21 . Polk, B. F .. and D . L. Kasper. 1977 . BacterotdesJrugllts subspecies
     1979 . Macrophage oxygen-dependent antlmfcroblal activity . II. The             tn cl[ntaal tsolates . Ann. intern . Med. ti5:569.
     role of oxygen Intermedtatea. J. Exp. Med. 150:950.                        22 . Diez. F.. L . L. Mosovich . and B. 1970 . Serogroups of Pseu-
14 . Root, R B:._ and J. A. Metcalf. 1977 . H2Og release from human                  donionas aerugtnosa and the Immurne response of patients with
     granulocytes during phagocytosls . J. C1tn. Invest . 60: 1266.                  cystic ftbros9s- J. beJect. Dts . 121:269.



                                                                        PuBLICflrroNS 024486

To top