- Email: [email protected]
Staphylococcal enterotoxin B mutants (N23K and F44S): biological effects and vaccine potential in a mouse model
Elsevier PII: S026440X(96)00166-1
ELSEVIER
Vaccine, Vol. 15, No. 2, pp. 133-139, 1997 Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain 0264--410X/97 $17+0.00
Staphylococcal enterotoxin B mutants (N23K and F44S): biological effects and vaccine potential in a mouse model Mary Alice Woody,
Teresa Krakauer
and Bradley
G. Stiles*
Superantigens produced by Staphylococcus aureus can cause food poisoning and toxic shock syndrome. The biological activities and vaccine potential of mutant staphylococcal enterotoxin B (SEB) proteins, N23K and F44S, were studied in a lipopolysaccharidepotentiated mouse model. Although 10 ,ug of SEB per mouse is equivalent to 30 LD,,, the same intraperitoneal dose of either mutant protein was nonlethal and did not elevate serum levels of tumor necrosis factors (TNF). N23K, F44S, and SEB were serologically identical in an enzyme-linked immunosorbent assay with polyclonal anti-SEB. Immunization with alum containing N23K, F44S, or SEB elicited an anti-SEB response that protected 80-87% of the mice against a 10 ,ug SEB challenge. Controls lacking an anti-SEB titer did not survive. Pooled sera from immunized mice eflectively blocked SEB-induced T-cell proliferation in vitro. Naive mice survived a lethal SEB challenge when given pooled antisera 1,2, or 4 h later, whereas the antisera failed to protect animals when administered 4 or 8 h after the toxin. Lethality at the later times was consistent with increased serum levels of TNF observed 6 h after SEB injection. These studies suggest that the N23K and F44S mutant proteins of SEB are less biologically active than the wild-type toxin, yet retain epitopes useful for eliciting a protective antibody response. Published by Elsevier Science Ltd. Keywords:
staphylococcal
enterotoxin
B; mutant
proteins;
superantigen
Staphylococcus aureus is a ubiquitous gram-positive bacterium which produces many protein toxins that are determinants of pathogenicity. Staphylococcal enterotoxins (SE) are heat stable, protease-resistant molecules of 25-30 kDa. The SE are most often associated with food poisoning, although nonmenstrual cases of toxic shock syndrome (TSS) are often linked to SEB and SEC, ‘.‘. Another toxin that is closely related to SE is the toxic shock syndrome toxin-l (TSST-1). Although TSST-1 shares little amino acid homology with SE, it has a crystal structure remarkably similar to SEB3*4. Recently, the SE and TSST-1 have received renewed attention because of their superantigenic properties5. SE and TSST-1 bind to molecules of the major histocompatibility complex class II (MHC II) found on antigenpresenting cells and subsequently cause massive proliferation of T-cells that express specific V,, elements in their antigen receptors. Cytokines6-9 and bacterial lipopolysaccharide (LPS) lo,’ ’ play an important role in toxic shock caused by these toxins. Previous studies with recombinant molecules of SE and TSST-I offer clues to Department of immunology and Molecular Biology, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, MD 21702-5011, USA. *To whom correspondence should be addressed. (Received 12 February 1996; revised 5 July 1996; accepted 5 July 1996)
vaccines
the biologically important regions of these proteins12-16 and recent X-ray crystallography of SEB or TSST-1 complexed with the MHC II molecule’7~1s have greatly resolved the toxin residues intimately involved in binding to this receptor. SEB mutant proteins with single amino acid substitutions at residue 23, asparagine to lysine (N23K), and residue 44, phenylalanine to serine (F44S), have been partially characterized regarding binding to a human MHC II molecule (DRl), T-cell stimulation, and weight loss in micei5. F44S does not effectively bind to the human MHC II molecule and subsequently has diminished T-cell proliferative effects. N23K readily binds to MHC II molecules but this complex is defective in stimulating T-cells. Both mutant proteins react with various SEB monoclonal antibodies, suggesting that N23K and F44S retain epitopes found on wild-type SEB. This previous information encouraged us to further assess the biological activity and vaccine potential of N23K and F44S in a recently characterized LPSpotentiated mouse mode18,9. Neither N23K nor F44S was lethal, unlike wild-type SEB, in LPS-treated mice. These mutant proteins did not increase the levels of tumor necrosis factors (TNF) in sera relative to those found in SEB-injected mice. N23K and F44S were serologically similar to SEB in an enzyme-
Vaccine
1997 Volume
15 Number
2
133
Microbial superantigen
vaccines: M.A. Woody et al.
linked immunosorbent assay (ELISA) with polyclonal anti-SEB. The two mutant proteins were as effective as SEB in eliciting antibodies that recognized the toxin. These antibodies protected mice against a lethal SEB challenge in active and passive immunization experiments.
MATERIALS
AND METHODS
SEB, SEB mutant proteins, and LPS Purified SEB was purchased from Toxin Technology (Sarasota, FL). Purified mutant proteins of SEB (F17S, N23K, F44S, and Y61N) were derived by site-directed mutagenesis’ 5 and kindly provided by Dr Jerry Bill. SEB and mutant proteins were diluted in sterile phosphatebuffered saline, pH 7.4 (PBS), and stored at - 50°C. SEB concentrations were determined by absorbance at 277 nm with an extinction coefficient (E’“‘, ,,) of 14.4. Protein estimates of the mutants were determined in a BCA assay (Pierce Chemical Co., Rockford, IL) with SEB as a standard. Protein homogeneity was assessed by electrophoresis on 10% sodium dodecyl sulfate polyacrylamide gels”. Escherichiu coli 055:B5 LPS was obtained from Difco Laboratories (Detroit, Ml), reconstituted in sterile PBS, and stored at - 50°C. ELISA Mouse anti-SEB serum was produced by immunization procedures described below and routinely used as a positive control. Antisera were diluted in PBS containing 0.1% Tween 20 and 0.1% gelatin. SEB and mutant proteins were diluted in carbonate buffer to a 5 pg ml-’ concentration and adsorbed overnight on to lmmulon 11 microtiter plates (Dynatech Laboratories, Chantilly, VA) at 4°C. Unadsorbed sites were blocked for 30 min at 37°C with 1% gelatin in PBS. Wells were aspirated, and mouse anti-SEB sera were added for 1 h at 37°C. The wells were then emptied, washed with PBS containing 0.1% Tween 20 (PBST), and sheep anti-mouse IgG conjugated to alkaline phosphatase (Sigma Chemical Co., St. Louis, MO) was added for 1 h at 37°C. The plates were finally washed with PBST and p-nitrophenyl phosphate substrate (Kirkegaard and Perry Laboratories, Gaithersburg, MD) added for 30 min at room temperature. Absorbances were read at 405 nm and data represented as the mean&standard deviation (S.D.) of triplicate wells. Mouse lethality assay BALB/c mice (18-20 g) were obtained from the Animal Production Division of Cancer Treatment, National Cancer Institute (Frederick, MD). SEB, mutant proteins, and LPS were diluted in sterile PBS and injected intraperitoneally (i.p.). A 1Opg dose of SEB per mouse, equivalent to 30 median lethal doses (LD,,)“, or mutant protein was followed 4 h later by 85 ,ug of LPS. Controls were injected with either LPS or SEB alone. Deaths were recorded for 72 h. Serum cytokine determinations BALB/c mice were injected i.p. with 5 pug of SEB, N23K, or F44S, plus 85 pg of LPS, as described for the
134
Vaccine 1997 Volume 15 Number 2
mouse lethality assay. Control groups were injected with either 5 pg of a single protein or 85 pg of LPS. Pooled sera were collected from four terminally bled mice per group every 2 h and analyzed for TNF and interleukin-6 (IL-6) in an ELISA (Genzyme, Boston, MA). Sera were pooled to insure sufficient volumes for all cytokine assays. Data are shown as the average of duplicate samples f 5% error.
Immunization Preimmune sera were collected from BALB/c mice before the first antigen injection. Antigen (3 ,ug per mouse) was combined with alum adjuvant (Pierce Chemical Co.) and groups of 15 mice were injected i.p. with a 200~1 volume per animal. One control group received alum only. Subsequent injections were given at 11 and 27 days after the first injection. Sera were collected 11 and 9 days after the second and third injections, respectively. Anti-SEB titers for each mouse and pooled sera from commonly immunized groups were determined in an ELISA. Serum pools were also used in the T-cell proliferation studies described below. Each immunized mouse was challenged with SEB (10 pug) and LPS (75 pg) 11 days after the third injection, as described in the mouse lethality assay. Naive, agematched mice were injected with SEB plus LPS or LPS only for comparison.
Passive immunization with antisera Pooled antisera from mice immunized with N23K, F44S, or SEB were collected 5 days after the toxin challenge and co-incubated with SEB at room temperature for 30 min before i.p. injection. Each mouse received 10 pg of SEB plus 100 ~1 of serum, followed 4 h later with an injection of 75 pug of LPS. Subsequent experiments included i.p. injection of mice with sera at 1, 2, 4, 6, or 8 h after the SEB dose.
T-cell proliferation assay Pooled sera from naive mice or animals immunized with alum only, SEB plus alum, or either SEB mutant protein plus alum were diluted 128-fold in RPM1 1640 medium containing 10% fetal calf serum. Cells from naive BALB/c mouse spleens (4 x lo6 cells ml- ‘) were plated with or without SEB (1 or 0.2 ,ug ml-‘) in the presence of pooled sera for 48 h at 37°C in 96-well microtiter plates. Cells were pulsed with 1 &i [3H]thymidine per well during the last 7 h of culture. Cells were harvested and the incorporated radioactivity was measured in a liquid scintillation counter. Data are shown as the mean counts per minute (c.p.m.) of triplicate samples f S.D.
RESULTS Toxicity of SEB mutant proteins Four mutant proteins of SEB were initially screened in LPS-potentiated mice to identify nonlethal and consequently suitable antigens for further investigation. N23K and F44S were not lethal at 10 pg per mouse.
Microbial superantigen vaccines: M.A. Woody et al
a
-
30000
8
t
0
-----a----
-2
0
2
4
6
F-.-.-.%._.1 ;
-
;
E s 2
i
8
LPS Injection
b 100000
80000
F44S + LPS N23K + LPS SE6 + LPS
_. _ .- _. _
-4
Time After
=
SEB only LPSonly
*
*_.
IO
(h) SEB only
LpSon,y
8.
-----Cl---- F44S + LPS
‘*. ‘..
----A----.-,-m-.-,
N23K + LPS SEB + LPS
; 60000
-
Time After
LPS Injection
(h)
Figure 1 Comparison of TNF (a) and IL-6 (b) serum levels in BALB/c mice injected with 5 ,ug of SEB, N23K, or F44S plus 85 ,ug of LPS. Controls were given SEB or LPS alone. Sera were pooled from four mice for each time point. The data presented are the mean readings of duplicate samples 25%
The same dose of wild-type SEB was 80-100% lethal. Two other mutant proteins, F17S and Y61N, were biologically active and- killed 60% of the injected mice at 10 ,ug per animal. Controls that each received only 85 pg of LPS did not die. Serum cytokine levels Cytokines, like TNF, play an important role in the toxic effects of SEB and other superantigens. We wanted to correlate the serum cytokine concentrations of TNF and IL-6 elicited by N23K or F44S in LPS-potentiated mice with lack of lethality. After an injection of either N23K or F44S plus LPS, serum TNF levels were comparable to those induced by 85 pg LPS alone (Figure la). Serum TNF levels induced by a lethal dose of SEB plus LPS were 30 times higher than those observed for an equivalent amount of either mutant protein. The maximum IL-6 concentrations after an injection of either mutant protein with LPS were approximately half those observed for SEB plus LPS, but similar to those elicited by LPS alone (Figure Zb).
Serological similarity of N23K, F44S, and SEB It is possible that mutation of a single amino acid could disrupt important epitopes and therefore reduce the effectiveness of a mutant protein as an immunogen. The serological similarities of N23K, F44S, and SEB were therefore compared in an ELISA with mouse anti-SEB serum to confirm that these mutant proteins had not lost major epitopes (Figure 2). Nearly identical results with N23K, F44S, and SEB suggest that the mutations at residues 23 or 44 did not grossly distort the SEB molecule and that the mutant proteins share common epitopes with wild-type SEB. Anti-SEB titers of immunized mice Both N23K and F44S were nontoxic at the doses used in this study and serologically similar to SEB; therefore, we tested each mutant protein for vaccine potential. Immunizing mice with N23K, F44S, or SEB produced very similar anti-SEB titers in an ELISA (Figure 3). There were slightly higher titers among mice immunized with F44S. Controls that were injected with alum alone
Vaccine 1997 Volume 15 Number 2
135
Microbial superantigen
vaccines: M.A. Woody et al Table 1
SEB N23K F44S
l
o A
1.0
0.8 1
t-1
mice after lethal SEB challenge
Immunizationa
Challengeb
!X&wild
SEB+LPS SEB+LPS SEB+LPS SEB+LPS SEB+LPS LPS only
type
F44S Alum only Naive Naive
.#I
1
Survival of immunized
% Survival
:: 80 ?I 100
0.6 =Each group consisted of 15 mice; bthe challenge dose was 10 ,ug of SEB plus 75 pg of LPS per mouse 0.4 Table 2
0.2 t
0.0 ’ 0.00001
i
4 . .
’ . ‘.‘.I 0.0001
Anti-SEB
Passive immunization
.
. . . ..“I
.
.
0.001
Serum
....I
0.01
Dilutions
Figure 2 Serological similarity of SEB, N23K, and F44S in an ELISA with serially diluted anti-SEB. Wells were coated with 5 ,ug ml-’ of SEB, N23K, or F44S in carbonate buffer. Data are shown as the mean A,,&.D. of triplicate samples. The readings for control wells prepared without SEB, yet treated with a 1:300 dilution of anti-SEB. were 0.040*0.008
did not develop any detectable SEB antibodies. Two injections of SEB elicited an anti-SEB response in 73% of the mice, whereas either mutant protein elicited specific antibodies in 93% of the mice. After the third injection of SEB, 93% of the mice seroconverted and all those immunized with N23K or F44S had detectable anti-SEB titers. Protection against lethal challenge Similar anti-SEB titers in mice immunized with N23K, F44S, or SEB suggested that these groups would be likewise protected against a lethal SEB challenge. N23K and F44S were as effective as SEB in eliciting a
against SEB in mice
Treatmenta
Survival (live/total)
Anti-N23K+SEB+LPS Anti-F44S+SEB+LPS Anti-SEB+SEB+LPS NMS’+SEB+LPS LPS only
8/l 5 lO/lO 1.5115 o/15 14115
a Each mouse was injected with 10 ,ug of SEB preincubated with 100 ,uI sera for 30 min at room temperature; bnormal mouse serum
protective immune response (Table I), but naive mice, or those that received alum only, were not protected against SEB. Antisera pooled from survivors in each of the immunization groups were used in passive immunization experiments to assess the kinetics of SEB intoxication in LPS-potentiated mice. Normal mouse serum (NMS) was substituted for sera from the mice injected with alum only. Initial experiments established that 100 ~1 of antiserum neutralized 10 pug of SEB. Mice were completely protected from SEB lethality by antisera produced against either F44S or SEB (Table 2). Antisera to N23K reduced, but did not abolish, SEB lethality, whereas the NMS was not protective. The antisera from mice immunized with SEB or F44S were equally protective and therefore combined for subsequent time-course
1.0
0.8
0.0 1:200
Preimmune
1:200
1:600
Immune
1: 1800
1:5400
1:16200
sera dilutions
Figure 3 Anti-SEB titers of mice immunized with SEB, N23K, or F44S in an ELISA. Wells were adsorbed with 5 ,ug ml-’ of SEB in carbonate buffer. Pooled sera from each group were obtained 9 days after the third immunization. Data are presented as the mean A,,,*S.D. of triplicate readings. The readings for control wells prepared without SEB, yet treated with a 1:200 dilution of anti-SEB, were 0.055*0.005
136
Vaccine 1997 Volume 15 Number 2
Microbial superantigen vaccines: M.A. Woody et al. Table 3
Delayed passive immunization
against SEB in mice
Treatmenta
Serum injection (hours after SEB)
Survival (live/total)
SEB+anti-SEB+LPS SEB+anti-SEB+LPS SEB+(LPS+anti-SEB) SEB+LPS+anti-SEB SEB+LPS+anti-SEB SEB+NMS+LPS NMS+LPSb
1 2 4 8 8 2 -
515 8110 lO/iO 015 O/5 t/t5 14115
=Each mouse received 10 pg of SEB and 75 ,ug of LPS. The (100 ,uI per mouse) were injected at indicated times after Anti-SEB serum was the combined postchallenge serum pools mice immunized with SEB or F44S; *mice were injected with 2 h before the LPS dose
proteins, N23K and F44S. N23K and F44S were serologically similar to wild-type SEB in an ELISA with polyclonal anti-SEB, which suggested that many epitopes were conserved; these nontoxic proteins could therefore act as potential immunogens. Previous studies show that mutant molecules of SE or TSST-1 may be conformationally, and serologically, different from wildtype toxin’2121. Conformational integrity is an important consideration for vaccine studies, as the vast majority of epitopes are conformationally dependent22. Immunization with N23K, F44S, or SEB equally protected mice against lethal challenge with SEB. Anti-SEB titers, as determined in an ELISA, were a good correlate for survival. The pooled postchallenge sera against N23K, however, were less protective than anti-SEB or antiF44S sera in passive immunization experiments. AntiSEB titers of the postchallenge sera pools were similar, as measured in an ELISA (data not shown), but the level of neutralizing antibodies in the anti-N23K sera was less than those in the anti-SEB or anti-F44S sera. Because antibodies neutralize the biological effects of SE or TSST-12’-25, vaccines against microbial superantigens could be useful for protecting especially susceptible individuals from TSS, a life-threatening disease. There is also a strong relationship between the presence of antibody against TSST-1 and protection against TSS in humans2.’ . Recent vaccine studies with a TSST-1 mutant protein containing a histidine to alanine change at residue 135 offer further experimental evidence for the importance of antibodies in preventing shock due to TSST-19,‘7. This mutant protein of TSST-1 does not cause T-cell proliferation of human peripheral blood mononuclear cells in vitro2* due to defective interactions with T-cell receptors29. Because we identified a TSST-1 mutant protein with vaccine potential in previous studies’, we wanted to apply the same rationale for in vivo analysis of two SEB mutant proteins, N23K and F44S. Like TSST-1, SEB is also linked to TSS1.‘530. Among bacterial toxins, the SE and TSST-1 represent a unique group of proteins that bind to MHC II molecules and subsequently induce the proliferation of T-cells bearing specific V, elements in their receptors. We chose N23K and F44S for vaccine studies because previous data indicate that these mutant proteins are less biologically active than wild-type SEB”. F44S binds to
sera SEB. from NMS
experiments. We wanted to further characterize the effectiveness of the antisera at neutralizing SEB in mice previously injected with SEB. LPS-potentiated mice that were given anti-SEB sera i.p. at 1, 2, or 4 h after a lethal injection of SEB survived, whereas mice that received anti-SEB sera at 6 and 8 h after the SEB injection died (Table 3).
Inhibition of T-cell proliferation in vitro The toxicity of superantigens is linked to T-cell proliferation and increased cytokine levels. To determine whether antisera from mice immunized with N23K, F44S, or SEB could inhibit SEB activity in vitro, murine splenocytes were incubated with antisera and SEB. Neither NMS nor sera from mice injected with alum alone prevented proliferation of murine splenocytes, whereas antisera from mice immunized with SEB, N23K, or F44S were equally effective in reducing proliferation caused by 200 ng ml-’ SEB (Figure 4). Antisera against F44S were slightly more effective at preventing T-cell proliferation at a SEB concentration of This is consistent with the higher anti-SEB 1 ,ug ml titer of anti-F44S sera as measured in an ELISA.
DISCUSSION In this study, we further characterized the biological activity and vaccine potential of two SEB mutant
1
s
12000:
g 10000
-
8000
-
Z ‘E !I
&
SEB + Alum Sera N23K + Alum Sera F44S + Alum Sera Alum Sera Normal Sera
l-
b
E ._ c” 2 E r ‘: Z .E
6000
4000 2000
0
1 OOOng
200ng SEBlml
Figure 4 Anti-proliferative effects of pooled sera from immunized SEB and a 1:128 dilution of pooled sera. The data are represented
mice. Naive BALBk splenocytes were incubated as the mean c.p.m.&.D. of triplicate readings
Vaccine
with medium containing
1997 Volume 15 Number 2
137
Microbial superantigen
vaccines: M.A. Woody et al.
the human MHC II molecule, DRl, approximately lOOO-fold less efficiently than SEB, and subsequently has very low T-cell stimulating activity. Co-crystal data for SEB complexed to the human MHC II molecule reveal that residue 44 of SEB intimately contacts the receptor17. The binding of N23K to a human MHC II molecule (DRl) is not altered, but interaction of the N23K-MHC II complex with murine T-cells bearing VP17and V,$3receptors is evidently defective15. We found that 10 pg of N23K or F44S was not lethal in LPSpotentiated mice, in contrast to SEB. Two other mutant proteins, F17S and Y61N, were lethal in the mouse assay and not further tested as vaccine candidates. F17S binds the DRl MHC II molecule approximately lOOfold less efficiently than SEB and Y61N is defective in T-cell receptor interactions15. F17S and Y61N stimulate T-cells bearing V/&.2 and V/,8.3 elements, but not those with V,]7 or Vf18.1. These data suggest that mutations of SEB that produce molecules with diminished proliferative effects on specific, but perhaps not all, reactive T-cell populations in vitro may not greatly alter in vivo toxicity. Toxicities of the SE and TSST-1 are a consequence of increased production and/or release of various cytokines, like TNF, IL-6 and y-interferon”‘p’3. Increased serum levels of cytokines also correlate well with the toxic effects of SE and TSST-1 in mouse models for toxic shock7-9. Experiments with galactosamine-treated mice reveal that IL-66 and TNF7 play an important role in SEB toxicity. Studies with SEA’ and TSST-l9 in LPS-potentiated mice have also shown that serum levels of TNF are an important marker of toxicity. In our study, mice injected with 5 pg of SEB plus 85 ,ug of LPS developed 30-fold higher serum levels of TNF than those elicited by either SEB mutant protein, N23K or F44S, which accounted for the reduced lethality of these mutant proteins in LPS-treated mice. Wild-type SEB plus LPS also elevated the serum levels of IL-6 at least twofold vs either N23K or F44S. In addition to mice, the biological activity of F44S has been tested in a monkey emetic assay34. F44S has reduced emetic activity compared to that of wild-type SEB. A 300 pg dose of F44S is needed to produce emesis in contrast to 75 ,ug of SEB. An analogous SEA mutant protein, F47S, also elicited emesis in monkeys yet a F47G molecule was inactive, implying that the glycine substitution for phenylalanine at residue 47 of SEA has a profound effect upon emesis. Additional mutational analysis of SEB residue 44 might also result in a molecule lacking emetic activity. SEA equivalents, N25G and N25A, of the SEB mutant protein N23K have also been tested in monkeys”5. The N25G mutant protein of SEA elicited an emetic response, albeit at a dose five times higher than that required for wild-type SEA. Further work needs to be done with additional SEB mutant proteins that have different biological activities in various assays. It is possible that a mutant molecule of any SE possessing greatly diminished biological activity, as determined by cell proliferation assays and cytokine levels in vitro or in vivo, could be tolerated as a vaccine when administered at doses much lower than those used for the monkey emetic assay. Obviously, monkey emesis indicates biological activity among the SE; however, the toxins are placed directly into the gastrointestinal tract which is probably the most biologically sensitive site for any enterotoxin. Further work is ongoing in our laboratory with other
138
Vaccine 1997 Volume 15 Number 2
SEB mutant proteins, devoid of biological activity in various assays, which might elicit a protective antibody response.
ACKNOWLEDGEMENTS The expert laborator ‘skills of Francis Sexton, Marilyn Buckley, and Jay Bri 5sey were invaluable in completing these studies. The SEB mutants were kindly provided by Dr Jerry Bill of NeXstar Pharmaceuticals, Lakewood, Colorado. Mary Alice Woody was supported by a National Research Council postdoctoral fellowship.
REFERENCES 1
2 3
4
5 6
7
8
9
10
11
12
13
14
15
16
17
Bohach, G.A., Fast, D.F., Nelson, RD. and Schlievert, P.M. Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses. Crit. Rev. Microbial. 1990,17,251-272 Crass, B.A. and Bergdoll, MS. Toxin involvement in toxic shock svndrome. J. Infect. Dis. 1986. 153. 918-926 Piasad, G.S., Earhart, C.A.‘, M&ray, D.L., Novick, R.P., Schlievert, P.M. and Ohlendorf, D.H. Structure of toxic shock syndrome toxin I. Biochemistry 1993, 32, 13761-l 3766 Swaminathan, S., Furey, W., Pletcher, J. and Sax, M. Crystal structure of staphylococcal enterotoxin B, a superantigen. Nature 1992, 359, 801-806 Marrack, P. and Kappler, J. The staphylococcal enterotoxins and their relatives. Science 1990, 248, 705-711 Matthys, B., Mitera, T., Heremans, H., Van Damme, J. and Billiau, A. Anti-gamma interferon and anti-interleukin-6 antibodies affect staphylococcal enterotoxin B-induced weight loss, hypoglycemia, and cytokine release in D-galactosaminesensitized and unsensitized mice. infect. Immun. 1995, 63, 1158-1164 Miethke, T., Wahl, C., Heeg, K., Echtenacher, B., Krammer, P.H. and Wagner, H. T cell-mediated lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: critical role of tumor necrosis factor. J. Exp. Med. 1992, 175, 91-98 Stiles, B.G., Bavari, S., Krakauer, T. and Ulrich, R.G. Toxicity of staphylococcal enterotoxins potentiated by lipopolysaccharide: major histocompatibility complex class II molecule dependency and cytokine release. Infect. Immun. 1993, 61, 5333-5338 Stiles, B.G., Krakauer, T. and Bonventre, P.F. Biological activity of toxic shock syndrome toxin 1 and site-directed mutant, H135A, in a lipopolysaccharide-potentiated mouse lethality model. Infect. lmmun. 1995, 63, 1229-1234 Pettit, G.W., Elwell, MR. and Jahrling, P.B. Possible endotoxemia in rabbits after intravenous injection of Staphylococcus aureus enterotoxin B. J. Infect. Dis. 1977, 135, 646-648 Stone, R.L. and Schlievert, P.M. Evidence for the involvement of endotoxin in toxic shock syndrome. J. Infect. Dis. 1987, 155, 682-689 Bonventre, P.F., Heeg, H., Cullen, C. and Lian, C-J. Toxicity of recombinant toxic shock syndrome toxin 1 and mutant toxins produced by Staphylococcus aureus in a rabbit infection model of toxic shock syndrome. Infect. Immun. 1993, 61, 793-799 Harris, T.O., Hufnagle, W.O. and Betley, M.J. Staphylococcal enterotoxin type A internal deletion mutants: serological activity and induction of T-cell proliferation. Infect. Immun. 1993, 61, 2059-2068 Hoffmann, M.L., Jablonski, L.M., Crum, K.K. eta/. Predictions of T-cell receptor- and major histocompatibility complex-binding sites on staphylococcal enterotoxin Cl. Infect. Immun. 1994, 62, 3396-3407 Kappler, J.W., Herman, A., Clements, J. and Marrack, P. M&ions defining functional regions of the superantigen staohvlococcal enterotoxin B. J. Exe. Med. 1992. 175. 387-396 M&ray, D.L., Prasad, G.S., Earhai, C.A. eta/. lmmunobiologic and biochemical properties of mutants of toxic shock syndrome toxin-l. J. Immun. 1994, 152, 87-95 Jardetzky, T.S., Brown, J.H., Gorga, J.C. et al. Threedimensional structure of a human class II histocompatibility molecule complexed with superantigen. Nature 1994, 368, 711-718
Microbial superantigen vaccines: MA. Woody et al. 18
19
20
21
22
23
24
25
26
Kim, J., Urban, R.G., Strominger, J.L. and Wiley, D.C. Toxic shock syndrome toxin-l complexed with a class II major histocompatibility molecule HLA-DRl. Science 1994, 266, 187O1878 Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 1970, 227,680-685 LeClaire, R.D., Kell, W., Bavari, S., Smith, T. and Hunt, R.E. Protective effects of niacinamide in staphylococcal enterotoxin B induced toxicity. Toxicology 1996, 107, 69-81 Metzroth, B., Marx, T., Linnig, M. and Fleischer, B. Concomitant loss of conformation and superantigenic activity of staphylococcal enterotoxin B deletion mutant proteins. Infect. lmmun. 1993,61, 2445-2452 Van Regenmortel, M.H.V. Molecular dissection of protein antigens. In: Structure of Antigens, Vol. I (Ed. Van Regenmortel, M.H.V.). CRC Press, Inc., Boca Raton, Florida, 1992, pp. l-27 Bohach, G.A., Hovde, C.J., Handley, J.P. and Schlievert, P.M. Cross-neutralization of staphylococcal and streptococcal pyrogenie toxins by monoclonal and polyclonal antibodies. Infect. Immun. 1988, 56, 400-404 Bonventre, P.F., Thompson, M.R., Adinolfi, L.E., Gillis, Z.A. and Parsonnet, J. Neutralization of toxic shock syndrome toxin-l by monoclonal antibodies in vitro and in vivo. Infect. Immun. 1988, 56, 135-141 Takei, S., Arora, Y.K. and Walker, S.M. Intravenous immunoglobulin contains specific antibodies inhibitory to activation of T cells by staphylococcal toxin superantigens. J. C/in. Invest. 1,993, 91, 602-607 Bonventre, P.F., Linnemann, C., Weckbach, L.S. et a/. Antibody responses to toxic-shock-syndrome (TSS) toxin by patients with TSS and by healthy staphylococcal carriers. J. infect. Dis. 1984, 150,662-666
27
28
29
30
31
32
33
34
35
Bonventre, P.F., Heeg, H., Edwards, C.K. and Cullen, C.M. A mutation at histidine residue 135 of toxic shock syndrome toxin yields an immunogenic protein with minimal toxicity. Infect. Immun. 1995, 63,509-515 Dt-ynda, A., Konig, B., Bonventre, P.F. and Konig, W. Role of a carboxy-terminal site of toxic shock syndrome toxin 1 in eliciting immune responses of human peripheral blood mononuclear cells. Infect. lmmun. 1995, 63, 1095-1101 Cullen, C.M., Blanco, L.R., Bonventre, P.F. and Choi, E. A toxic shock syndrome toxin 1 mutant that defines a functional site critical for T cell activation. Infect. lmmun. 1995, 63,2141-2146 Garbe, P.L., Arko, J.R., Reingold, A.L. et al. Staphy/ococcus aureus isolates from patients with nonmenstrual toxic shock syndrome: evidence for additional toxins. J. Am. Med. Assoc. 1985,253,2538-2542 Carlsson, R. and Sjogren, H.O. Kinetics of IL-2 and interferon-y production, expression of IL-2 receptors, and cell proliferation in human mononuclear cells exposed to staphylococcal enterotoxin A. Cell. lmmun. 1985, 96, 175-183 Jupin, C., Anderson, S., Damais, C., Alouf, J.E. and Parant, M. Toxic shock syndrome toxin 1 as an inducer of human tumor necrosis factors and gamma interferon. J. Exp. Med. 1988, 167, 752-761 Krakauer, T. Differential inhibitory effects of interleukin-10, interleukin-4, and dexamethasone on staphylococcal enterotoxin-induced cytokine production and T cell activation. J. Leuk. Biol. 1995, 57, 450-454 Harris, T.O., Grossman, D., Kappler, J.W., Marrack, P., Rich, R.R. and Betley, M.J. Lack of complete correlation between emetic and T-cell-stimulatory activities of staphylococcal enterotoxins. infect. lmmun. 1993, 61, 3175-3183 Harris, T.O. and Betley, M.J. Biological activities of staphylococcal enterotoxin A mutants with N-terminal substitutions. Infect. Immun. 1995, 63,2133-2140
Vaccine
1997 Volume
15 Number
2
139
ELSEVIER
Vaccine, Vol. 15, No. 2, pp. 133-139, 1997 Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain 0264--410X/97 $17+0.00
Staphylococcal enterotoxin B mutants (N23K and F44S): biological effects and vaccine potential in a mouse model Mary Alice Woody,
Teresa Krakauer
and Bradley
G. Stiles*
Superantigens produced by Staphylococcus aureus can cause food poisoning and toxic shock syndrome. The biological activities and vaccine potential of mutant staphylococcal enterotoxin B (SEB) proteins, N23K and F44S, were studied in a lipopolysaccharidepotentiated mouse model. Although 10 ,ug of SEB per mouse is equivalent to 30 LD,,, the same intraperitoneal dose of either mutant protein was nonlethal and did not elevate serum levels of tumor necrosis factors (TNF). N23K, F44S, and SEB were serologically identical in an enzyme-linked immunosorbent assay with polyclonal anti-SEB. Immunization with alum containing N23K, F44S, or SEB elicited an anti-SEB response that protected 80-87% of the mice against a 10 ,ug SEB challenge. Controls lacking an anti-SEB titer did not survive. Pooled sera from immunized mice eflectively blocked SEB-induced T-cell proliferation in vitro. Naive mice survived a lethal SEB challenge when given pooled antisera 1,2, or 4 h later, whereas the antisera failed to protect animals when administered 4 or 8 h after the toxin. Lethality at the later times was consistent with increased serum levels of TNF observed 6 h after SEB injection. These studies suggest that the N23K and F44S mutant proteins of SEB are less biologically active than the wild-type toxin, yet retain epitopes useful for eliciting a protective antibody response. Published by Elsevier Science Ltd. Keywords:
staphylococcal
enterotoxin
B; mutant
proteins;
superantigen
Staphylococcus aureus is a ubiquitous gram-positive bacterium which produces many protein toxins that are determinants of pathogenicity. Staphylococcal enterotoxins (SE) are heat stable, protease-resistant molecules of 25-30 kDa. The SE are most often associated with food poisoning, although nonmenstrual cases of toxic shock syndrome (TSS) are often linked to SEB and SEC, ‘.‘. Another toxin that is closely related to SE is the toxic shock syndrome toxin-l (TSST-1). Although TSST-1 shares little amino acid homology with SE, it has a crystal structure remarkably similar to SEB3*4. Recently, the SE and TSST-1 have received renewed attention because of their superantigenic properties5. SE and TSST-1 bind to molecules of the major histocompatibility complex class II (MHC II) found on antigenpresenting cells and subsequently cause massive proliferation of T-cells that express specific V,, elements in their antigen receptors. Cytokines6-9 and bacterial lipopolysaccharide (LPS) lo,’ ’ play an important role in toxic shock caused by these toxins. Previous studies with recombinant molecules of SE and TSST-I offer clues to Department of immunology and Molecular Biology, United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, MD 21702-5011, USA. *To whom correspondence should be addressed. (Received 12 February 1996; revised 5 July 1996; accepted 5 July 1996)
vaccines
the biologically important regions of these proteins12-16 and recent X-ray crystallography of SEB or TSST-1 complexed with the MHC II molecule’7~1s have greatly resolved the toxin residues intimately involved in binding to this receptor. SEB mutant proteins with single amino acid substitutions at residue 23, asparagine to lysine (N23K), and residue 44, phenylalanine to serine (F44S), have been partially characterized regarding binding to a human MHC II molecule (DRl), T-cell stimulation, and weight loss in micei5. F44S does not effectively bind to the human MHC II molecule and subsequently has diminished T-cell proliferative effects. N23K readily binds to MHC II molecules but this complex is defective in stimulating T-cells. Both mutant proteins react with various SEB monoclonal antibodies, suggesting that N23K and F44S retain epitopes found on wild-type SEB. This previous information encouraged us to further assess the biological activity and vaccine potential of N23K and F44S in a recently characterized LPSpotentiated mouse mode18,9. Neither N23K nor F44S was lethal, unlike wild-type SEB, in LPS-treated mice. These mutant proteins did not increase the levels of tumor necrosis factors (TNF) in sera relative to those found in SEB-injected mice. N23K and F44S were serologically similar to SEB in an enzyme-
Vaccine
1997 Volume
15 Number
2
133
Microbial superantigen
vaccines: M.A. Woody et al.
linked immunosorbent assay (ELISA) with polyclonal anti-SEB. The two mutant proteins were as effective as SEB in eliciting antibodies that recognized the toxin. These antibodies protected mice against a lethal SEB challenge in active and passive immunization experiments.
MATERIALS
AND METHODS
SEB, SEB mutant proteins, and LPS Purified SEB was purchased from Toxin Technology (Sarasota, FL). Purified mutant proteins of SEB (F17S, N23K, F44S, and Y61N) were derived by site-directed mutagenesis’ 5 and kindly provided by Dr Jerry Bill. SEB and mutant proteins were diluted in sterile phosphatebuffered saline, pH 7.4 (PBS), and stored at - 50°C. SEB concentrations were determined by absorbance at 277 nm with an extinction coefficient (E’“‘, ,,) of 14.4. Protein estimates of the mutants were determined in a BCA assay (Pierce Chemical Co., Rockford, IL) with SEB as a standard. Protein homogeneity was assessed by electrophoresis on 10% sodium dodecyl sulfate polyacrylamide gels”. Escherichiu coli 055:B5 LPS was obtained from Difco Laboratories (Detroit, Ml), reconstituted in sterile PBS, and stored at - 50°C. ELISA Mouse anti-SEB serum was produced by immunization procedures described below and routinely used as a positive control. Antisera were diluted in PBS containing 0.1% Tween 20 and 0.1% gelatin. SEB and mutant proteins were diluted in carbonate buffer to a 5 pg ml-’ concentration and adsorbed overnight on to lmmulon 11 microtiter plates (Dynatech Laboratories, Chantilly, VA) at 4°C. Unadsorbed sites were blocked for 30 min at 37°C with 1% gelatin in PBS. Wells were aspirated, and mouse anti-SEB sera were added for 1 h at 37°C. The wells were then emptied, washed with PBS containing 0.1% Tween 20 (PBST), and sheep anti-mouse IgG conjugated to alkaline phosphatase (Sigma Chemical Co., St. Louis, MO) was added for 1 h at 37°C. The plates were finally washed with PBST and p-nitrophenyl phosphate substrate (Kirkegaard and Perry Laboratories, Gaithersburg, MD) added for 30 min at room temperature. Absorbances were read at 405 nm and data represented as the mean&standard deviation (S.D.) of triplicate wells. Mouse lethality assay BALB/c mice (18-20 g) were obtained from the Animal Production Division of Cancer Treatment, National Cancer Institute (Frederick, MD). SEB, mutant proteins, and LPS were diluted in sterile PBS and injected intraperitoneally (i.p.). A 1Opg dose of SEB per mouse, equivalent to 30 median lethal doses (LD,,)“, or mutant protein was followed 4 h later by 85 ,ug of LPS. Controls were injected with either LPS or SEB alone. Deaths were recorded for 72 h. Serum cytokine determinations BALB/c mice were injected i.p. with 5 pug of SEB, N23K, or F44S, plus 85 pg of LPS, as described for the
134
Vaccine 1997 Volume 15 Number 2
mouse lethality assay. Control groups were injected with either 5 pg of a single protein or 85 pg of LPS. Pooled sera were collected from four terminally bled mice per group every 2 h and analyzed for TNF and interleukin-6 (IL-6) in an ELISA (Genzyme, Boston, MA). Sera were pooled to insure sufficient volumes for all cytokine assays. Data are shown as the average of duplicate samples f 5% error.
Immunization Preimmune sera were collected from BALB/c mice before the first antigen injection. Antigen (3 ,ug per mouse) was combined with alum adjuvant (Pierce Chemical Co.) and groups of 15 mice were injected i.p. with a 200~1 volume per animal. One control group received alum only. Subsequent injections were given at 11 and 27 days after the first injection. Sera were collected 11 and 9 days after the second and third injections, respectively. Anti-SEB titers for each mouse and pooled sera from commonly immunized groups were determined in an ELISA. Serum pools were also used in the T-cell proliferation studies described below. Each immunized mouse was challenged with SEB (10 pug) and LPS (75 pg) 11 days after the third injection, as described in the mouse lethality assay. Naive, agematched mice were injected with SEB plus LPS or LPS only for comparison.
Passive immunization with antisera Pooled antisera from mice immunized with N23K, F44S, or SEB were collected 5 days after the toxin challenge and co-incubated with SEB at room temperature for 30 min before i.p. injection. Each mouse received 10 pg of SEB plus 100 ~1 of serum, followed 4 h later with an injection of 75 pug of LPS. Subsequent experiments included i.p. injection of mice with sera at 1, 2, 4, 6, or 8 h after the SEB dose.
T-cell proliferation assay Pooled sera from naive mice or animals immunized with alum only, SEB plus alum, or either SEB mutant protein plus alum were diluted 128-fold in RPM1 1640 medium containing 10% fetal calf serum. Cells from naive BALB/c mouse spleens (4 x lo6 cells ml- ‘) were plated with or without SEB (1 or 0.2 ,ug ml-‘) in the presence of pooled sera for 48 h at 37°C in 96-well microtiter plates. Cells were pulsed with 1 &i [3H]thymidine per well during the last 7 h of culture. Cells were harvested and the incorporated radioactivity was measured in a liquid scintillation counter. Data are shown as the mean counts per minute (c.p.m.) of triplicate samples f S.D.
RESULTS Toxicity of SEB mutant proteins Four mutant proteins of SEB were initially screened in LPS-potentiated mice to identify nonlethal and consequently suitable antigens for further investigation. N23K and F44S were not lethal at 10 pg per mouse.
Microbial superantigen vaccines: M.A. Woody et al
a
-
30000
8
t
0
-----a----
-2
0
2
4
6
F-.-.-.%._.1 ;
-
;
E s 2
i
8
LPS Injection
b 100000
80000
F44S + LPS N23K + LPS SE6 + LPS
_. _ .- _. _
-4
Time After
=
SEB only LPSonly
*
*_.
IO
(h) SEB only
LpSon,y
8.
-----Cl---- F44S + LPS
‘*. ‘..
----A----.-,-m-.-,
N23K + LPS SEB + LPS
; 60000
-
Time After
LPS Injection
(h)
Figure 1 Comparison of TNF (a) and IL-6 (b) serum levels in BALB/c mice injected with 5 ,ug of SEB, N23K, or F44S plus 85 ,ug of LPS. Controls were given SEB or LPS alone. Sera were pooled from four mice for each time point. The data presented are the mean readings of duplicate samples 25%
The same dose of wild-type SEB was 80-100% lethal. Two other mutant proteins, F17S and Y61N, were biologically active and- killed 60% of the injected mice at 10 ,ug per animal. Controls that each received only 85 pg of LPS did not die. Serum cytokine levels Cytokines, like TNF, play an important role in the toxic effects of SEB and other superantigens. We wanted to correlate the serum cytokine concentrations of TNF and IL-6 elicited by N23K or F44S in LPS-potentiated mice with lack of lethality. After an injection of either N23K or F44S plus LPS, serum TNF levels were comparable to those induced by 85 pg LPS alone (Figure la). Serum TNF levels induced by a lethal dose of SEB plus LPS were 30 times higher than those observed for an equivalent amount of either mutant protein. The maximum IL-6 concentrations after an injection of either mutant protein with LPS were approximately half those observed for SEB plus LPS, but similar to those elicited by LPS alone (Figure Zb).
Serological similarity of N23K, F44S, and SEB It is possible that mutation of a single amino acid could disrupt important epitopes and therefore reduce the effectiveness of a mutant protein as an immunogen. The serological similarities of N23K, F44S, and SEB were therefore compared in an ELISA with mouse anti-SEB serum to confirm that these mutant proteins had not lost major epitopes (Figure 2). Nearly identical results with N23K, F44S, and SEB suggest that the mutations at residues 23 or 44 did not grossly distort the SEB molecule and that the mutant proteins share common epitopes with wild-type SEB. Anti-SEB titers of immunized mice Both N23K and F44S were nontoxic at the doses used in this study and serologically similar to SEB; therefore, we tested each mutant protein for vaccine potential. Immunizing mice with N23K, F44S, or SEB produced very similar anti-SEB titers in an ELISA (Figure 3). There were slightly higher titers among mice immunized with F44S. Controls that were injected with alum alone
Vaccine 1997 Volume 15 Number 2
135
Microbial superantigen
vaccines: M.A. Woody et al Table 1
SEB N23K F44S
l
o A
1.0
0.8 1
t-1
mice after lethal SEB challenge
Immunizationa
Challengeb
!X&wild
SEB+LPS SEB+LPS SEB+LPS SEB+LPS SEB+LPS LPS only
type
F44S Alum only Naive Naive
.#I
1
Survival of immunized
% Survival
:: 80 ?I 100
0.6 =Each group consisted of 15 mice; bthe challenge dose was 10 ,ug of SEB plus 75 pg of LPS per mouse 0.4 Table 2
0.2 t
0.0 ’ 0.00001
i
4 . .
’ . ‘.‘.I 0.0001
Anti-SEB
Passive immunization
.
. . . ..“I
.
.
0.001
Serum
....I
0.01
Dilutions
Figure 2 Serological similarity of SEB, N23K, and F44S in an ELISA with serially diluted anti-SEB. Wells were coated with 5 ,ug ml-’ of SEB, N23K, or F44S in carbonate buffer. Data are shown as the mean A,,&.D. of triplicate samples. The readings for control wells prepared without SEB, yet treated with a 1:300 dilution of anti-SEB. were 0.040*0.008
did not develop any detectable SEB antibodies. Two injections of SEB elicited an anti-SEB response in 73% of the mice, whereas either mutant protein elicited specific antibodies in 93% of the mice. After the third injection of SEB, 93% of the mice seroconverted and all those immunized with N23K or F44S had detectable anti-SEB titers. Protection against lethal challenge Similar anti-SEB titers in mice immunized with N23K, F44S, or SEB suggested that these groups would be likewise protected against a lethal SEB challenge. N23K and F44S were as effective as SEB in eliciting a
against SEB in mice
Treatmenta
Survival (live/total)
Anti-N23K+SEB+LPS Anti-F44S+SEB+LPS Anti-SEB+SEB+LPS NMS’+SEB+LPS LPS only
8/l 5 lO/lO 1.5115 o/15 14115
a Each mouse was injected with 10 ,ug of SEB preincubated with 100 ,uI sera for 30 min at room temperature; bnormal mouse serum
protective immune response (Table I), but naive mice, or those that received alum only, were not protected against SEB. Antisera pooled from survivors in each of the immunization groups were used in passive immunization experiments to assess the kinetics of SEB intoxication in LPS-potentiated mice. Normal mouse serum (NMS) was substituted for sera from the mice injected with alum only. Initial experiments established that 100 ~1 of antiserum neutralized 10 pug of SEB. Mice were completely protected from SEB lethality by antisera produced against either F44S or SEB (Table 2). Antisera to N23K reduced, but did not abolish, SEB lethality, whereas the NMS was not protective. The antisera from mice immunized with SEB or F44S were equally protective and therefore combined for subsequent time-course
1.0
0.8
0.0 1:200
Preimmune
1:200
1:600
Immune
1: 1800
1:5400
1:16200
sera dilutions
Figure 3 Anti-SEB titers of mice immunized with SEB, N23K, or F44S in an ELISA. Wells were adsorbed with 5 ,ug ml-’ of SEB in carbonate buffer. Pooled sera from each group were obtained 9 days after the third immunization. Data are presented as the mean A,,,*S.D. of triplicate readings. The readings for control wells prepared without SEB, yet treated with a 1:200 dilution of anti-SEB, were 0.055*0.005
136
Vaccine 1997 Volume 15 Number 2
Microbial superantigen vaccines: M.A. Woody et al. Table 3
Delayed passive immunization
against SEB in mice
Treatmenta
Serum injection (hours after SEB)
Survival (live/total)
SEB+anti-SEB+LPS SEB+anti-SEB+LPS SEB+(LPS+anti-SEB) SEB+LPS+anti-SEB SEB+LPS+anti-SEB SEB+NMS+LPS NMS+LPSb
1 2 4 8 8 2 -
515 8110 lO/iO 015 O/5 t/t5 14115
=Each mouse received 10 pg of SEB and 75 ,ug of LPS. The (100 ,uI per mouse) were injected at indicated times after Anti-SEB serum was the combined postchallenge serum pools mice immunized with SEB or F44S; *mice were injected with 2 h before the LPS dose
proteins, N23K and F44S. N23K and F44S were serologically similar to wild-type SEB in an ELISA with polyclonal anti-SEB, which suggested that many epitopes were conserved; these nontoxic proteins could therefore act as potential immunogens. Previous studies show that mutant molecules of SE or TSST-1 may be conformationally, and serologically, different from wildtype toxin’2121. Conformational integrity is an important consideration for vaccine studies, as the vast majority of epitopes are conformationally dependent22. Immunization with N23K, F44S, or SEB equally protected mice against lethal challenge with SEB. Anti-SEB titers, as determined in an ELISA, were a good correlate for survival. The pooled postchallenge sera against N23K, however, were less protective than anti-SEB or antiF44S sera in passive immunization experiments. AntiSEB titers of the postchallenge sera pools were similar, as measured in an ELISA (data not shown), but the level of neutralizing antibodies in the anti-N23K sera was less than those in the anti-SEB or anti-F44S sera. Because antibodies neutralize the biological effects of SE or TSST-12’-25, vaccines against microbial superantigens could be useful for protecting especially susceptible individuals from TSS, a life-threatening disease. There is also a strong relationship between the presence of antibody against TSST-1 and protection against TSS in humans2.’ . Recent vaccine studies with a TSST-1 mutant protein containing a histidine to alanine change at residue 135 offer further experimental evidence for the importance of antibodies in preventing shock due to TSST-19,‘7. This mutant protein of TSST-1 does not cause T-cell proliferation of human peripheral blood mononuclear cells in vitro2* due to defective interactions with T-cell receptors29. Because we identified a TSST-1 mutant protein with vaccine potential in previous studies’, we wanted to apply the same rationale for in vivo analysis of two SEB mutant proteins, N23K and F44S. Like TSST-1, SEB is also linked to TSS1.‘530. Among bacterial toxins, the SE and TSST-1 represent a unique group of proteins that bind to MHC II molecules and subsequently induce the proliferation of T-cells bearing specific V, elements in their receptors. We chose N23K and F44S for vaccine studies because previous data indicate that these mutant proteins are less biologically active than wild-type SEB”. F44S binds to
sera SEB. from NMS
experiments. We wanted to further characterize the effectiveness of the antisera at neutralizing SEB in mice previously injected with SEB. LPS-potentiated mice that were given anti-SEB sera i.p. at 1, 2, or 4 h after a lethal injection of SEB survived, whereas mice that received anti-SEB sera at 6 and 8 h after the SEB injection died (Table 3).
Inhibition of T-cell proliferation in vitro The toxicity of superantigens is linked to T-cell proliferation and increased cytokine levels. To determine whether antisera from mice immunized with N23K, F44S, or SEB could inhibit SEB activity in vitro, murine splenocytes were incubated with antisera and SEB. Neither NMS nor sera from mice injected with alum alone prevented proliferation of murine splenocytes, whereas antisera from mice immunized with SEB, N23K, or F44S were equally effective in reducing proliferation caused by 200 ng ml-’ SEB (Figure 4). Antisera against F44S were slightly more effective at preventing T-cell proliferation at a SEB concentration of This is consistent with the higher anti-SEB 1 ,ug ml titer of anti-F44S sera as measured in an ELISA.
DISCUSSION In this study, we further characterized the biological activity and vaccine potential of two SEB mutant
1
s
12000:
g 10000
-
8000
-
Z ‘E !I
&
SEB + Alum Sera N23K + Alum Sera F44S + Alum Sera Alum Sera Normal Sera
l-
b
E ._ c” 2 E r ‘: Z .E
6000
4000 2000
0
1 OOOng
200ng SEBlml
Figure 4 Anti-proliferative effects of pooled sera from immunized SEB and a 1:128 dilution of pooled sera. The data are represented
mice. Naive BALBk splenocytes were incubated as the mean c.p.m.&.D. of triplicate readings
Vaccine
with medium containing
1997 Volume 15 Number 2
137
Microbial superantigen
vaccines: M.A. Woody et al.
the human MHC II molecule, DRl, approximately lOOO-fold less efficiently than SEB, and subsequently has very low T-cell stimulating activity. Co-crystal data for SEB complexed to the human MHC II molecule reveal that residue 44 of SEB intimately contacts the receptor17. The binding of N23K to a human MHC II molecule (DRl) is not altered, but interaction of the N23K-MHC II complex with murine T-cells bearing VP17and V,$3receptors is evidently defective15. We found that 10 pg of N23K or F44S was not lethal in LPSpotentiated mice, in contrast to SEB. Two other mutant proteins, F17S and Y61N, were lethal in the mouse assay and not further tested as vaccine candidates. F17S binds the DRl MHC II molecule approximately lOOfold less efficiently than SEB and Y61N is defective in T-cell receptor interactions15. F17S and Y61N stimulate T-cells bearing V/&.2 and V/,8.3 elements, but not those with V,]7 or Vf18.1. These data suggest that mutations of SEB that produce molecules with diminished proliferative effects on specific, but perhaps not all, reactive T-cell populations in vitro may not greatly alter in vivo toxicity. Toxicities of the SE and TSST-1 are a consequence of increased production and/or release of various cytokines, like TNF, IL-6 and y-interferon”‘p’3. Increased serum levels of cytokines also correlate well with the toxic effects of SE and TSST-1 in mouse models for toxic shock7-9. Experiments with galactosamine-treated mice reveal that IL-66 and TNF7 play an important role in SEB toxicity. Studies with SEA’ and TSST-l9 in LPS-potentiated mice have also shown that serum levels of TNF are an important marker of toxicity. In our study, mice injected with 5 pg of SEB plus 85 ,ug of LPS developed 30-fold higher serum levels of TNF than those elicited by either SEB mutant protein, N23K or F44S, which accounted for the reduced lethality of these mutant proteins in LPS-treated mice. Wild-type SEB plus LPS also elevated the serum levels of IL-6 at least twofold vs either N23K or F44S. In addition to mice, the biological activity of F44S has been tested in a monkey emetic assay34. F44S has reduced emetic activity compared to that of wild-type SEB. A 300 pg dose of F44S is needed to produce emesis in contrast to 75 ,ug of SEB. An analogous SEA mutant protein, F47S, also elicited emesis in monkeys yet a F47G molecule was inactive, implying that the glycine substitution for phenylalanine at residue 47 of SEA has a profound effect upon emesis. Additional mutational analysis of SEB residue 44 might also result in a molecule lacking emetic activity. SEA equivalents, N25G and N25A, of the SEB mutant protein N23K have also been tested in monkeys”5. The N25G mutant protein of SEA elicited an emetic response, albeit at a dose five times higher than that required for wild-type SEA. Further work needs to be done with additional SEB mutant proteins that have different biological activities in various assays. It is possible that a mutant molecule of any SE possessing greatly diminished biological activity, as determined by cell proliferation assays and cytokine levels in vitro or in vivo, could be tolerated as a vaccine when administered at doses much lower than those used for the monkey emetic assay. Obviously, monkey emesis indicates biological activity among the SE; however, the toxins are placed directly into the gastrointestinal tract which is probably the most biologically sensitive site for any enterotoxin. Further work is ongoing in our laboratory with other
138
Vaccine 1997 Volume 15 Number 2
SEB mutant proteins, devoid of biological activity in various assays, which might elicit a protective antibody response.
ACKNOWLEDGEMENTS The expert laborator ‘skills of Francis Sexton, Marilyn Buckley, and Jay Bri 5sey were invaluable in completing these studies. The SEB mutants were kindly provided by Dr Jerry Bill of NeXstar Pharmaceuticals, Lakewood, Colorado. Mary Alice Woody was supported by a National Research Council postdoctoral fellowship.
REFERENCES 1
2 3
4
5 6
7
8
9
10
11
12
13
14
15
16
17
Bohach, G.A., Fast, D.F., Nelson, RD. and Schlievert, P.M. Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses. Crit. Rev. Microbial. 1990,17,251-272 Crass, B.A. and Bergdoll, MS. Toxin involvement in toxic shock svndrome. J. Infect. Dis. 1986. 153. 918-926 Piasad, G.S., Earhart, C.A.‘, M&ray, D.L., Novick, R.P., Schlievert, P.M. and Ohlendorf, D.H. Structure of toxic shock syndrome toxin I. Biochemistry 1993, 32, 13761-l 3766 Swaminathan, S., Furey, W., Pletcher, J. and Sax, M. Crystal structure of staphylococcal enterotoxin B, a superantigen. Nature 1992, 359, 801-806 Marrack, P. and Kappler, J. The staphylococcal enterotoxins and their relatives. Science 1990, 248, 705-711 Matthys, B., Mitera, T., Heremans, H., Van Damme, J. and Billiau, A. Anti-gamma interferon and anti-interleukin-6 antibodies affect staphylococcal enterotoxin B-induced weight loss, hypoglycemia, and cytokine release in D-galactosaminesensitized and unsensitized mice. infect. Immun. 1995, 63, 1158-1164 Miethke, T., Wahl, C., Heeg, K., Echtenacher, B., Krammer, P.H. and Wagner, H. T cell-mediated lethal shock triggered in mice by the superantigen staphylococcal enterotoxin B: critical role of tumor necrosis factor. J. Exp. Med. 1992, 175, 91-98 Stiles, B.G., Bavari, S., Krakauer, T. and Ulrich, R.G. Toxicity of staphylococcal enterotoxins potentiated by lipopolysaccharide: major histocompatibility complex class II molecule dependency and cytokine release. Infect. Immun. 1993, 61, 5333-5338 Stiles, B.G., Krakauer, T. and Bonventre, P.F. Biological activity of toxic shock syndrome toxin 1 and site-directed mutant, H135A, in a lipopolysaccharide-potentiated mouse lethality model. Infect. lmmun. 1995, 63, 1229-1234 Pettit, G.W., Elwell, MR. and Jahrling, P.B. Possible endotoxemia in rabbits after intravenous injection of Staphylococcus aureus enterotoxin B. J. Infect. Dis. 1977, 135, 646-648 Stone, R.L. and Schlievert, P.M. Evidence for the involvement of endotoxin in toxic shock syndrome. J. Infect. Dis. 1987, 155, 682-689 Bonventre, P.F., Heeg, H., Cullen, C. and Lian, C-J. Toxicity of recombinant toxic shock syndrome toxin 1 and mutant toxins produced by Staphylococcus aureus in a rabbit infection model of toxic shock syndrome. Infect. Immun. 1993, 61, 793-799 Harris, T.O., Hufnagle, W.O. and Betley, M.J. Staphylococcal enterotoxin type A internal deletion mutants: serological activity and induction of T-cell proliferation. Infect. Immun. 1993, 61, 2059-2068 Hoffmann, M.L., Jablonski, L.M., Crum, K.K. eta/. Predictions of T-cell receptor- and major histocompatibility complex-binding sites on staphylococcal enterotoxin Cl. Infect. Immun. 1994, 62, 3396-3407 Kappler, J.W., Herman, A., Clements, J. and Marrack, P. M&ions defining functional regions of the superantigen staohvlococcal enterotoxin B. J. Exe. Med. 1992. 175. 387-396 M&ray, D.L., Prasad, G.S., Earhai, C.A. eta/. lmmunobiologic and biochemical properties of mutants of toxic shock syndrome toxin-l. J. Immun. 1994, 152, 87-95 Jardetzky, T.S., Brown, J.H., Gorga, J.C. et al. Threedimensional structure of a human class II histocompatibility molecule complexed with superantigen. Nature 1994, 368, 711-718
Microbial superantigen vaccines: MA. Woody et al. 18
19
20
21
22
23
24
25
26
Kim, J., Urban, R.G., Strominger, J.L. and Wiley, D.C. Toxic shock syndrome toxin-l complexed with a class II major histocompatibility molecule HLA-DRl. Science 1994, 266, 187O1878 Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 1970, 227,680-685 LeClaire, R.D., Kell, W., Bavari, S., Smith, T. and Hunt, R.E. Protective effects of niacinamide in staphylococcal enterotoxin B induced toxicity. Toxicology 1996, 107, 69-81 Metzroth, B., Marx, T., Linnig, M. and Fleischer, B. Concomitant loss of conformation and superantigenic activity of staphylococcal enterotoxin B deletion mutant proteins. Infect. lmmun. 1993,61, 2445-2452 Van Regenmortel, M.H.V. Molecular dissection of protein antigens. In: Structure of Antigens, Vol. I (Ed. Van Regenmortel, M.H.V.). CRC Press, Inc., Boca Raton, Florida, 1992, pp. l-27 Bohach, G.A., Hovde, C.J., Handley, J.P. and Schlievert, P.M. Cross-neutralization of staphylococcal and streptococcal pyrogenie toxins by monoclonal and polyclonal antibodies. Infect. Immun. 1988, 56, 400-404 Bonventre, P.F., Thompson, M.R., Adinolfi, L.E., Gillis, Z.A. and Parsonnet, J. Neutralization of toxic shock syndrome toxin-l by monoclonal antibodies in vitro and in vivo. Infect. Immun. 1988, 56, 135-141 Takei, S., Arora, Y.K. and Walker, S.M. Intravenous immunoglobulin contains specific antibodies inhibitory to activation of T cells by staphylococcal toxin superantigens. J. C/in. Invest. 1,993, 91, 602-607 Bonventre, P.F., Linnemann, C., Weckbach, L.S. et a/. Antibody responses to toxic-shock-syndrome (TSS) toxin by patients with TSS and by healthy staphylococcal carriers. J. infect. Dis. 1984, 150,662-666
27
28
29
30
31
32
33
34
35
Bonventre, P.F., Heeg, H., Edwards, C.K. and Cullen, C.M. A mutation at histidine residue 135 of toxic shock syndrome toxin yields an immunogenic protein with minimal toxicity. Infect. Immun. 1995, 63,509-515 Dt-ynda, A., Konig, B., Bonventre, P.F. and Konig, W. Role of a carboxy-terminal site of toxic shock syndrome toxin 1 in eliciting immune responses of human peripheral blood mononuclear cells. Infect. lmmun. 1995, 63, 1095-1101 Cullen, C.M., Blanco, L.R., Bonventre, P.F. and Choi, E. A toxic shock syndrome toxin 1 mutant that defines a functional site critical for T cell activation. Infect. lmmun. 1995, 63,2141-2146 Garbe, P.L., Arko, J.R., Reingold, A.L. et al. Staphy/ococcus aureus isolates from patients with nonmenstrual toxic shock syndrome: evidence for additional toxins. J. Am. Med. Assoc. 1985,253,2538-2542 Carlsson, R. and Sjogren, H.O. Kinetics of IL-2 and interferon-y production, expression of IL-2 receptors, and cell proliferation in human mononuclear cells exposed to staphylococcal enterotoxin A. Cell. lmmun. 1985, 96, 175-183 Jupin, C., Anderson, S., Damais, C., Alouf, J.E. and Parant, M. Toxic shock syndrome toxin 1 as an inducer of human tumor necrosis factors and gamma interferon. J. Exp. Med. 1988, 167, 752-761 Krakauer, T. Differential inhibitory effects of interleukin-10, interleukin-4, and dexamethasone on staphylococcal enterotoxin-induced cytokine production and T cell activation. J. Leuk. Biol. 1995, 57, 450-454 Harris, T.O., Grossman, D., Kappler, J.W., Marrack, P., Rich, R.R. and Betley, M.J. Lack of complete correlation between emetic and T-cell-stimulatory activities of staphylococcal enterotoxins. infect. lmmun. 1993, 61, 3175-3183 Harris, T.O. and Betley, M.J. Biological activities of staphylococcal enterotoxin A mutants with N-terminal substitutions. Infect. Immun. 1995, 63,2133-2140
Vaccine
1997 Volume
15 Number
2
139