Associated Immune Responses to p-Galactosidase ...

Thl -Associated Immune Responses to p-Galactosidase Expressed by a Replication-Defective Herpes Simplex Virus' Jeffery 0. Brubaker,* Claudette M. Thompson,* Lynda A. Morrison,2t David M. Knipe,+ George R. Siber,* and Robert W. Finberg3* The immunogenic properties of a replication-defective herpes simplex virus HD-2, containing the Escherichia coli lacZ gene under control of the HSV ICP8 early gene promoter were studied in BALB/c mice. Experiments were designed to determine if the HD-2 virus preferentially stimulated either Thl- or Th2-associated immune responses to p-galactosidase @gal). Sera from mice immunized i.p. or S.C. with virus HD-2, pgal on aluminum phosphate adjuvant, or a control ICP8 deletion mutant, d301, were assayed for total and Ag-specific lgGl and lgC2a Abs, pgal-driven lymphocyte proliferation, and in vitro productionof the cytokines IFN-y, 11-4, and 11-2. Viruses HD-2 and d301 preferentially stimulated the productionof total serum lgG2a following two immunizations i.p. ora single immunization s.c., while only HD-2 virus stimulated in vivo production of pgal-specific lgC2a serum Abs. In contrast, pgal adsorbed on AIPO, preferentially stimulated productionof Ag-specific lgCl serum Abs. The HD-2 virus also induced a potent cellular proliferative response to pgal, whichwas still pronounced5 wk after primary immunization. Cultured lymphocytes from HD-2-immunized mice produced IFN-)I after 5 days in culture with soluble pgal in an Ag- and dose-dependent fashion. These resultsdemonstrate that replication-defective mutants of HSV can be usedas vectors for eliciting Thl -associated immune responses to a heterologous Ag expressed from the viral genome. The Journal of Immunology, 1996, 157: 1598-1604. "

T

he successful resolution of infection with microbial pathogens and the induction of cellular protective and/or pathogenic immune responses is determined, in part, by the types of immune cell subsets and associated functions that are stimulated in vivo (1, 2). These include innate immunity from NK cell cytotoxicity, enhanced phagocytosis of Ab-opsonized organisms and immune complexes, Ab-dependent cellular cytotoxicity (ADCC),4 and the activities of APCs such as macrophages, dendritic cells, and B cells that function in the progression to either Thl (delayed type) or Th2 (humoral) acquired immune responses (3-19). Production of immune cytokines by APC, NK cells, and Ag-specific T cells, in turn, influences the nature of immune memory that is called into play during re-infection with the homologous or related heterologous pathogens (20-22). The reciprocal regulation of cytokine production during infection or vaccination has been hypothesized to be the result ofThl and Th2 cell activities. In some animal models, Thl responses are associated with recovery from infection, while Th2 responses are associated with

'Laboratory of Infectious Diseases, Dana-Farber Cancer Institute, and the 'Departmentof Mlcrobiology and Molecular Genetics, Harvard Medical School, Boston, M A 02 1 15 Received for publication March 4, 1996. Accepted for publication June3, 1996. The costs of publication of this article were defrayed in part by the payment of page charges. This artlcle must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

'

This work was supported by National Institutes of Health Grants POAI-24010 and POAI-AG37963.

'

Current address: Department of Mlcrobiology and Immunology, Saint Louis University School of Medicine, 1402 S. Grand Ave., St. Louis, MO 63104.

'

Address correspondence and reprint requests to Dr. Robert W. Finberg, Laboratory of Infectious Diseases, Dana-Farber Cancer Institute, 44 Binney Street, Boston, M A 021 15. Abbreviations used in this paper: ADCC, Ab-dependent cellular cytotoxicity; 1; AIPO,, aluminum pgal, p-galactosidase; HSV-1 herpes simplexvirustype phosphate; RSV, respiratorysyncytial virus; LN, lymph node; RD, replication defective. Copyright 0 1996 by The American Association of Immunologists

persistent or exacerbated disease(20, 23-29). ThI-type responses are characterized by the production of T cell-derived IL-2 and IFN-y, and are influenced by the production of the monokines lymphotoxin and IL-12 (NK cell stimulatory factor) (30-32). Secretion of the T cellderived cytokines IL-4, IL-5, IL-6, and IL-10, in contrast, are correlated with Th2 function (23, 24). At the Ab level, Th2 responses are associated with greater production of IgM, IgGl, IgGZb,and IgE molecules, while Thl responses are associated with production of the IgG2a Ab subclass (18). Therefore, the types of cytokines and Abs produced in vivo and in vitro by lymphocytes during primary and secondary infections can be used to predict the types of cellular immunity thatoccur during microbial pathogenesis (33,34). Information on the profiles of cellular cytokine and Ab production in response to challenge organisms or vaccine Ags helps determine how manipulation of quantity and chemical forms of Ags can influence the outcome of immune responses in model systems of infection (35). The recognition and clearance of many obligate and facultative intracellular parasites, including viruses, appears to require specific cellular immune functions, especially in situations where production of neutralizing Abs alone is not sufficient (9). Viruses and microbial pathogens such as mycobacterial sp. (25), Listeria (36), Brucella (37), Chlamydia (38, 39), Toxoplasma (26). Leishmania (23), and Schistosoma (24, 40) each induce particular types of cellular and humoral immune responses, some of which actually enhance the pathogenic capacity of the invading organisms. The predominance of Th2-associated immune responses during infection with Schistosoma (21, 40), Leishmania (20), and Chlamydia (41, 42), or after vaccination with killed respiratory syncytial virus (RSV), for example, are associated with heightened immunopathology and/or prolonged infection (43,44). In these infection models and several others, examination of the types of Tcells infiltrating into sites of infection in normal or cytokine receptorknockout mice and the depletion of Th2-associated functions using anti-cytokine mAbs in vivo or in vitro has revealed that protective immunity is dependent more on Thl cellular functions (27-29, 0022-1 767/96/$02.00

1599

The Journal of Immunology

45). However, as pointed out by Fazenkas and colleagues (46), the outcome of infection very much relies on the types and activities of immunocompetent cells and APC that reside in local sites of infection in mucosa or draining lymphatics (Le., the route of infection), the metastatic nature of the invading pathogens, and local Ab production. Viruses are notable in theirability to stimulate, to various degrees, brisk type 1 (alp) and type 2 ( y ) IFN production and their respectivenon-MHC-restricted NK cell andMHC-restricted T cell-mediated cytolysis (CTL) (47-54), as well as the heightened production of anti-viral IgG2a serum Abs in vivo (45, 55) The preferential induction of CTL and other Thl-associated functions by viruses has been used recently to study the potential of using viruses as vectors for stimulating protective Th I responses against virulent challenge or the exogenous Ags expressedfrom vector genomes.Thesameexperimental paradigms that allow for differentiation between Thl and Th2 responses duringnatural infectioncan be used to studyhow these geneticmanipulations affect immunity to targetAgsin host animals. Anumber of viruses, including vaccinia, influenza,andretroviruses, aswellasplasmids and wholeDNA, have been used as vectors for delivering exogenous gene products to the host immune system (1, 56-63). These vectors effectively deliver target proteins and their peptides to the host immunesystemand offer the advantages of selectivityand heightened adjuvanticityfor thenativeAgs contained in the vaccine vectors and expressed in host cells or tissues. In addition, the vector-driven expression of target Ags in the milieu of host cells results in the loading of viral peptides onto class I MHC moleculesand the inductionof potentially protective CTL responses. Herpesviruses possess complex replication cycles that progress through immediate-early, early, and late viral gene expression and therefore allow for the expression of exogenous genes from viral promoters associated with any of the three major phases of viral replication. As shown previously, insertional or deletional mutation of selected genes of the herpes simplex virus type 1 (HSV-1) genome results in viral replication up to the point at which the mutation occurs and no further (64-67). These mutants are thus rendered replication defective (RD) in the absence of any complementing genes in the host, and cannot complete their maturation into infectious virions. Interestingly, different RD HSV-I mutants possessing lesions in the ICP4 or ICP27 (both immediate-early genes), or ICP8 (early) genes elicit different levels of total and virus-specific serum IgG2a Ab responses, and correspondingly different levels of immune protection against virulent challenge (6567). The use of RD HSV mutants as vectors for eliciting protective cellular immunity against homologous HSV or foreign Ags is feasible by virtue of their natural host range, their well-defined genomic makeup and replication cycles, and their ability to elicit T cell responses. We have undertaken studies with an ICP8 insertional mutant of HSV-1, referred to as virus HD-2, which contains a functional bacterial lacZ gene for P-galactosidase (Pgal), and have examined its ability to induce pgal-specific, Thl- and Th2-associated Abs and Ag-driven secretion of cytokines following inoculation by different routes into recipient BALB/c mice. The experimental strategies used in these studies were 1) measurement of the amounts and relative proportions of pgal-specific and total IgGl and IgG2a Abs in sera obtained after primary and secondary immunization, and 2) Pgal-driven cellular proliferation and cytokine production in vitro.

Materials and Methods Mice Adult male BALB/c mice were obtained from The Jackson Laboratory. Bar Harbor, ME, and were housed and fed in the animal care facility at DanaFarber Cancer Institute. Mice aged 6 to 12 wk were used in all experiments. Mice were anesthetized with Metaphane before injection into the hind footpads of 25 to 50 ELI volumes of virus suspensions in PBS or soluble pgal in AIPO, adjuvant, Sera were collected from the tail veins of mice during on-going experiments or by cardiac puncture of anesthetized mice immediately before the animals were killed by cervical dislocation. Sera were stored at -20°C until assay. Control mice were left untreated or were injected in the footpads with PBS/AIPO,.

Viruses Parental HSV-I strain KOS 1.1 was propagated and titrated on Vero cells as described previously (68), while the ICP8 replication-defective mutant viruses d301 @gal-) and its derivative HD-2 @gal’) were propagated and titrated on S-2 cells. which provide the 1CP8 gene in trans (65). Virus stocks were prepared as clarified lysates of infected Vero cells or by pelleting and resuspending extracellular virions in complete medium with 15% glycerol. Virus stocks were stored at -80°C until their use immediately before injection into recipient mice. The HD-2virus contains the lucZ coding sequences fused to the ICP8 open reading frame so that pgal is expressed as a fusion protein from a non-functional ICP8 gene (65). 0-galactosidase(GradeVIII, from E. coli. Sigma Chemical Co., St. Louis, MO) was adsorbed onto AIPO, (4 mg/ml) by gentle mixing at room temperature for 30 min. Adsorption was determined to be complete by assay of solid phase (AIPO,) and supernatant for pgal activity. All enzyme activity was restricted to the solid phase.

Cellular proliferation assay Single-cell suspensions of lymphocytes were prepared from whole spleens and pooled popliteal, inguinal, and axillary lymph nodes by pressing tissues through sterile 200-gauge stainless steel wire mesh (Small Parts, Inc., Miami, FL) in RPMI medium containing L-glutamine and glucose (CellGro, Washington, DC) supplemented with 109 heat-inactivated fetal bovine serum (BioWhittaker, Walkersville, MD) and penicillin/streptomycin. RBC were lysed with aqueous 0.84% ammonium chloride solution before resuspending lymphocytes in DMEM containing the same ingredients listed above plus IO-’ M 2-ME. Viability of cells was examined by trypan blue dye exclusion and always exceeded 95% viability. Cells were adjusted to a density of 1.0 to 2.5 million/ml and 100 pl volumes of the cell suspensions placed into quadruplicate round-bottom wells of 96-well plates (Costar, Cambridge, MA) along with equal volumes of various concentrations of soluble Pgal in DMEM or medium alone. Cells were labeled for 5 to 6 h with 1 FCi/well of tritiated thymidine at the end of 5 days incubation at 37 C in 10% COz. Data for cellular proliferation were expressed as the arithmetic means 5 SE.

Assays for total Ig and pgal-specific serum Ab subclasses The concentrations of total serum Ig subclasses were determined by a sandwich ELISA method. Immunlon 2 microtiter plates (Dynatech Labs. Chantilly, VA) were coated for 2 hat 37 C with unconjugated goat Abs to mouse IgGI, IgG2a, IgG2b. and IgC3 (Southern Biotechnology, Inc.. Birmingham, AL) (2 pg/ml) and washed three times with pBW0.57~ Tween-20. One hundred microliter volumes of dilutions of mouse serum samples were placed onto the coated wells, incubated for 2 h at room temperature, and washed as before. Wells were then incubated for 2 h at room temperature with appropriately diluted, alkaline phosphatase-conjugated goat antimouse IgG subclass-specific Abs for 2 h at room temperature, washed and then developed with the addition of o-nitrophenyl phosphate substrate (Sigma Chemical Co.). followed by read-out at OD = 405 nm in a microplate reader (Molecular Devices, Inc., Sunnydale, CA).Concentrations of serum anti-pgal Abs were determined by ELlSA on plates coated with 100 pI/ well of soluble Pgal in PBS (3 pglml), followed by the same sequence of steps and reagents listed above for the total Ig assay. Assays were calibrated by a modification of the technique used by Zollinger and Boslego (69) and are expressed as geometric mean titers in @ml.

Assay for IFN-y The concentration of IFN-y in lymphocyte culture supernatants was determined by ELISA using a commercial assay kit (Endogen, Cambridge. MA) that is sensitive to 2 ng/ml or more of mouse IFN-y. Supernatant samples were assayed in three-step, twofold dilution series on Nunc MaxiSorp flatwell plates according to the exact specifications of the kit instructions.

USE OF REPLICATION-DEFECTIVE AS VIRUSES

1600

VACCINES

Table I. Primary and secondary serum lgGI and lgCZa responses to 0-gakctosidase and HSV Ags in BALB/c mice injected i.p. with HD-2 virus or Pgal in AIPO," Antibody Titer"

4 Wk postboost

3 Wk postchallenge HSV-1

0-galactosldase Immunogen

UD 0.09"

Controls UD pgallAIP0, HD-2 UD

IgC 1

lgC2a

IgG 1

6-galactosidase lgG2a

UDUD' UD 0.1 OC' UD UD UD 6) (1.10.360.31 (1.19) 2.241.89

HSV-1

lgCl

lgG2a

IgG 1

lgG2a

9.28

0.23 UD(0.02)'

UD 7.83

9.07 (1.16)

Croups of five BALB/c mice were injected i.p. with either2.5 X 10" pfu of HD-2 virus in 100 pI volumes of sterile PBS, or 1 p g of soluble pgal in 100 pl of AIPO, suspension, and reinjected with the same doses of Ag 3 wk postpriming. "Geometric mean of Ab titer expres5ed in pp/nlI. ' Undetectable ( = less than 0.03-0.06 pdmT specific Ab). "Two of five mice responding only. One of fwe mice responding only, ' Numbers in parentheses indicate the ratio of lgC2a to IgCl. "

Supernatant IFN-y titers were calculated on the basis of standard curves that were run simultaneous with the test samples.

Table II. Anti-&a/ serum lgC7 and lgGZa responses in mice inoculated once in the footpads with pgal/AIPO,, the &a/- ICP8 deletion n7utant d301, or the pgal' HD-2 virus"

Results Serum Ab responses to soluble /!-galactosidase vs pgal expressed from an HSV vector

We hypothesized that expression of Pgal in a herpesvirus vector would stimulate predominantly Thl-associated immune responses 1.14 to pgal. Initially, we examined the serum Ab responses in BALB/c mice that had been immunized with the HD-2 mutant herpes simplex virus that expressed Pgal or with Pgal adsorbed on AIPO,. In mice receiving i.p. doses of virus, detectable serum Ab responses to Pgal were obtained only after booster immunizations (Table I). The secondary Ab response to Pgal in mice immunized with the HD-2 virus was shifted toward IgG2a in the HD-2-infected animals. yielding an IgC2dIgGl ratio of 1.16. In contrast, the secondary Ab response to soluble pgallAlP0, was almost entirely IgCI, with the IgG2dlgG1 ratio being approximately 0.02. Serum Ab responses to HSV-I were detectable after primary immunization, with titers increasing after secondary immunization. The HSV-specific Ab response in the HD-2-infected animals also demonstrated nearly equal titers of IgG2a and IgGI, with IgG2dIgG1 ratios of 1.16-1.19 after primary and secondary immunizations, respectively (Table I). The Ag specificity of the response was determined by challenging mice subcutaneously in the hind footpads with the HD-2 virus strain, containing the 1ucZ gene, or the control d301 virus. Subcutaneous challenge of mice with the HD-2 virus resulted in the development of a primary Ab response to Pgal. while injection of mice with the control d301 virus did not elicit the production of Abs to Pgal (Table IT). This demonstrated the specificity ofthe reagents and a lack of cross-reaction of anti-HSV Abs with Pgal. We then examined the effect of immunization with either virus strain or Pgal onAIPO, on the production of total serum Igs. Consistent with our previous data (67), a preferential increase in total serum IgG2a Ig was seen in mice inoculated once in the footpads with either the HD-2 virus or the parental d301 virus (Fig. I). Injection of soluble Pgal on AIPO,didnot elict preferential production of any single IgG subclass. We conclude from these data that the shift to anti-pgal serum Abs of the IgG2a subclass was specific to the HD-2 virus and was consistent with the ability of the mutant herpesviruses to induce Thl-associated humoral responses, in general. Thus, the Ab response to an exogenous gene product contained within the repli-

Geometric Mean of Anti-pgal A b Titers [ p d m l ) " Immunogen

Control0.06 &al/AIPO, HD-20.32 (pgal') @gal-) d301 0.06

lgGl

0.06 2.38 0.05 0.28d 0.06

IgC2a

r(lgGZa/lgCl)

0.1 1

" Groups of five male BALB/c mice were inoculated S.C. in each footpad with five million pfu of either virus contained In 25 pl of sterile PBS, or with 0.5 pg of soluble pgal in25 pl of AIPO, suspension, and sera were collected 5 wk later. "All Igs detectable at 20.12 pdml. ' One of five mice responded. Four of five mice responded.

cation-defective herpesvirus is the same type as the responses to viral Ags. Cellular proliferative responses to pgal in HD-2-immunized mice

Lymph node cells from mice inoculated 2 wk (Fig. 2) or 5 wk before assay (Fig. 3) with either Pgal on AIPO, or the HD-2 virus responded in a dose-dependent fashion to soluble Pgal after 5 days in culture. LN cells from mice injected with either the HD-2 or d301 replication-defective viruses had high background proliferation. so the data in Figure 3 are represented as the arithmetic difference between pgal-specific proliferation and background cpm (delta cpm). Figure 3 reveals that at 5 wk postimmunization, the HD-2 virus induced greater cellular proliferation against soluble Pgal thandid Pgal on AIPO,, and no response to Pgal in cells taken from mice injected with the negative control virus d301. These data provide evidence for the vigorous induction of cellular immune responses to the Pgal expressed by replication-defective mutant herpesvirus. Secretion of cytokines by HD-2-immune lymphocytes

To further define the helper T cell responses induced by immunization with a virus vector or soluble Ag, we measured IFN-y and IL-4 secretion by lymphocytes from immunized animals. Figure 4 shows the IFN-y concentrations in 5-day cell culture supernatants from one experiment in which the HD-2-immune LN cells secreted IFN-y in the presence of soluble pgal. No other groups of LN cells produced IFN-y in a pgal dose-dependent fashion, supporting the

1601

The Journal of Immunology 550 1 5004

. -

’

T

lgG’ lgG2a

20000

-

17000

-

14000

-

Control BgallAIP04 HD-2 d301

E

rn

400 350

TI

s 4.a

$ -a

11000-

300

0 250

-a

200

U

8000

-

5000

-

2000

-

W

4.a

al

v)

150

T

4.a

100 50

0

Bgal Control

d 3 0 1H D - 2 Stimulus

18000

Control

14000

BgaVAIP04 HD-2

20

Beta-galactosidase (pglml)

FIGURE 1. The HD-2 andd301replication-defective herpes simplex viruses preferentially stimulate production of total serum lgG2a Ig. Groups of5 male BALBlc mice were inoculated S.C. in each footpad with five million pfu of either virus contained in 25 pI of sterile PBS, or with 0.5 pg of soluble pgal in 25pI of AIPO, suspension, and sera were collected 5 w k later.

16000

2

0.2

T

FIGURE 3. Cellular proliferative responses in mice inoculated once in the footpads with pgallAIP04, HD-2 virus, or the ICP8 parental deletion control mutant d301. Pooled LN cells from groups of mice injected as described in the legend to Figure 1 were incubated at a density of 250,000 cells per well in the presence of different concentrations of soluble pgal for 5 days. The data for tritiated thymidine uptake are expressed as the 6 cpm ofAg-specific responses to medium background counts in each group. Error bars are the SEM of quadruplicate wells. 50

H

Control BgaVAIP04 HD-2 d301

12000 10000

E

n

0

8000

T

6000 4000

2000

0

0

5

25

0 Beta-galactosidase (pglml)

FIGURE 2. Cellular proliferative response against soluble pgal after a single S.C. injection of HD-2 virus. Pooled LN cells from mice immunized in each hind footpad 13 days before assay with five million pfu of HD-2 virus in 25 pl of PBS or 5 p g of soluble pgal in 25 pI of AIPO, suspension were incubated at a density of 100,000 cells per well for 6 days in thepresence of 0,5, or 25 pg/ml of soluble pgal, and their proliferation measured in a 5 h 13H]TdR incorporation assay. Error bars are the SEM of quadruplicate wells.

Ag-driven nature of this cytokine response and its specificity to immune T cells. A similar result was observed in a second experiment in which only HD-2-immune mouse LN cells secreted IFN-y in the presence of 25 pg/ml of soluble pgal (datanot shown). The IFN-y secretion by lymphocytes of HD-2-infected mice was consistent with a Thl response in these animals. In one experiment, the IL-2 content in culture supernatants of control (nonimmune) and HD-2-immune lymph node cells was

0

0.2

2

20

Beta-galactosidase (pglml)

FIGURE 4. In vitro production of IFN-y by pgal-immune BALB/c LN cells. Pooled cell culture supernatants from lymphocyte proliferation assay plates of a single experiment (Fig. 3) were diluted 1/8, 1/16, and 1/32 in PBS and assayed for IFN-y content by ELlSA (Endogen). Average IFN-y titers ? error bars were calculated on the basis of OD readings obtained from duplicate samples, using a standard curve generated from kit reagents (Endogen) where r = 0.991,

tested at 18 h and 5 days in culture with various concentrations of pgal. The LN cells from HD-2-immune mice produced IL-2 after 5 days in culture, but the levels were not considered significant compared with the high background of IL-2 produced in these cultures. Neither control cells nor HD-2-immune cells produced any measurable IL-2 during the initial 18 h of culture in soluble

1602 pgal, whereas the T cell mitogen Con A did stimulate production of IL-2 within 18 h (results not shown). No IL-4 was detected in any of these culture supernatants using an ELISA assay sensitive to 2 pg/ml or more of IL-4. We were also unable to detect IL-4 production in cultures set up with greater numbers of LN cells, and therefore concluded that the total amounts of IL-4 produced in these cultures were too small to detect by ELISA.

Discussion We have tested the capability of the HD-2 replication-defective mutant of HSV-I (IacZIHSV-I) to stimulate humoral and cellular immune responses in mice against the exogenous gene product, pgal. Several lines of evidence indicate that the HD-2 virus stimulates Thl-associated immune responses against Pgal. These are 1) the potent production of total and /3gal-specific IgG2a serum Abs in vivo, 2) the induction of cellular proliferation during in vitro restimulation of immune LN cells with soluble Pgal in a dosedependent fashion, and 3) the Pgal-dependent in vitro production of IFN-y by immune LN cells. Taken together, these data indicate that the HD-2 virus is an effective vector for presenting peptides of a heterologous gene product to the host immune system. We attempted to normalize the dosage of pgal in the HD-2 virus and &al/AlPO, inocula to allow for a more accurate comparative analysis of their effects on host immunity. Results indicate that it is difficult to predict precisely the amplitude of cellular and Ab responses to Pgal in a given experiment. In comparing the data from the experiments shown in Tables I and 11, for example, the different levels of Pgal-specific serum Abs elicited after primary immunization with Pgal in AIPO, may be explained by the different routes of administration used in these two experiments (i.p. in one and S.C.in the other). More than 1 pg of Pgal on AIPO, may have been required to elicit a greater primary Ab response after immunization by the i.p. route. Given the ditterence in time after primary immunization at which sera were collected in these two experiments (3 wk vs 5 wk), it is also possible that the Pgal Ab titers in mice immunized i.p. would have increased during the period from 3 to 5 wk post-primary immunization. Nevertheless, the proportional switch to Pgal-specific IgG2a serum Abs in the HD-2 immunized mice occurred in every experiment and yielded reproducible IgG2a/IgGI ratios of 0.9 to 1.2 in primary-immunized mice (s.c. route) or greater in mice receiving booster injections. The mechanism(s) of the long-lasting stimulatory effect of the HD-2 virus on in vivo-restimulated immune mouse T cells (Fig. 1 and Table 11) remain to be elucidated. This long-term stimulatory effect might be due to persistence of the virus or viral Ags, or due to induction of early stimulatory cytokines such as IL-2. IFN-y production in vitro was a sensitive and specific measure of pgaldriven cytokine production during cellular restimulation. The tendency of different strains of mice to respond to Ags or pathogens in either a predominantly Thl-like or Th2-like manner (16, 23). would lead us to expect different proportions of cellular and Ab responses to pgal, or any other inserted gene product within the HD-2 virus, in various mouse strains. The fact that BALB/c mice respond so well to the HD2-associated Pgal is testimony to the ability of this mutant virus to switch immune responsiveness toward Thl-associated functions in a mouse strain more characteristically Th2-like in its response to exogenous protein Ags. Previous work has shown that RD HSV mutants can elicit longlasting and protective cellular responses against lethal wild-type herpesvirus infection without vaccine-induced pathologies (65, 66). Thus, RD HSV mutants can be successfully used to circumvent many of the problems inherent to vaccination with live, live-

USE OF REPLICATION-DEFECTIVE VIRUSES AS VACCINES attenuated, and killed viruses, such as live virus-induced morbidity, latent infection and reactivation of attenuated viruses in host animals, and the poor immunogenicity of killed virus vaccines. The mechanism by which live viruses induce the production of IgG2a Abs is not entirely clear. Coutelier et al. (69) have provided evidence for a role for IL-12 produced by virus-infected macrophages in the switch to Th 1-associated immune responses in vivo. We have shown that lymph node cells induced by live virus are IFN-y-producing and that infection with replication-defective HSV leads to the expansion of Thl cells. Further experiments are necessary to determine how even replication-defective viruses lead to IL-12 production. Theoretically, the genetic construction of RD HSV mutants containing exogenous genes isolated from clinically important microorganisms may allow for the successful delivery of these gene products to the portion of the host immune system that elicits protective cellular responses and serum/secretory Abs functional in microbial neutralization and/or ADCC. To maximize the safety of these RD mutants, viruses with multiple lethal mutations are being designed to reduce the possibility of recombination with naturally occurring HSV strains to generate a more virulent virus. Since Pgal is a potent Ag that is not immunologically important in humans, additional experiments need to be done to test the efficacy of replication-defective herpesviruses in eliciting Th 1-associated immune responses to poorly immunogenic molecules from clinically important human pathogens. The RDHSV mutants appear to constitute effective vaccine vectors for stable in vivo transcription and expression of selected microbial genes, with subsequent loading of immunogenic peptides into class 11, and likely class I, MHC molecules. Genetic construction of these viral mutants precludes the need for biochemical purification of the native or recombinant forms of exogenous gene products. Additionally, the Ab isotype switching seen in mice immunized in the footpads was also observed in mice inoculated via the nasal mucosa and gut (not shown). The ability of RD HSV viruses to stimulate Thl-associated mucosal immunity is an important benefit to these vectors, because adjuvants that enhance gut immunity tend to favor the stimulation of Th2-associated responses (70, 71). The potent gut-stimulatory activity of RD HSV mutants may be useful in models of vaccination against such important human respiratory/mucosal pathogens as RSV, HIV, and others, for which potent and safe vaccines do not yet exist (1, 72).

Acknowledgments The authors are indebted to Dr. Abul K. A b b a s for his assistance in establishing in vitroanalyses of mouse cytokinea and hisassistance with interpretation of experimental results.

References I . Alwan, W. H., W. 1. Kozlowska. and P. J . M. Openshaw. 1994. Distinct types of lung disease caused by functlonal subsets of antiviral T cells. J . Exp. Med. 179: 81. 7. Graham, B. S., L. A. Bunton. P. F. Wright. and D. T. Karzon. 1991. Role of T lymphocyte subsets in the pathogenesiq of primary infection and rechallenge wlth respiratory syncytial viruc in mice. J. Clin. hvesr. RN:1026 3. Bukowski. J. F,, B. A. Woda. and R. M . Welsh. 1984. Pathogenesis of murine cytomegalovirus infection in natural killer cell-depleted mice. J. Vim/. 52:IlY. 4. Bukowski. J . F., J. F. Warner, G. Dennert, and R. M. Welsh. 1985. Adoptive transfer studies demonstrating the antiviral effect of natural killer cells in vivo. J. Exp. Mrd. 161:40. 5. De Bruijn, M. L. H., J . D. Nieland, T N. M. Schumacher, H.L. Ploegh, W. M. Kast, and C. I. M.Melief. 1992. Mechanisms of induction of primary virucspecitic cytotoxic T Lymphocyte responses. Eur. J. lmmunol. 22:.3013. 6. DeVelasco, E. A., H . A. Th. Dekker, P. Anta], K. P. Jalink, J. A . G. van Strijp. A. F. M. Verheul, J. Verhoef, and H. Snippe. 1994. Adjuvant Qui1 A irnprovea protection in mice and enhances opsonic capaclty of antisera induced by pneumococcal polysaccharide conjugate vaccines. Vnrcine /2:/419.

The journal of Immunology 7 Sousa. C. R., and R. N. Germain. 1995. Major histocompatibility complex class I persentation of peptides derived from soluble exogenous antigen by a subset of cells engaged in phagocytosis. J. Exp. Med. 182:WI. 8. Harding. C. V., and R. Song. 1994. Phagocyt~cprocessing of exogenous particulate antigens by macrophages for presentation by class IMHC molecules. J. Imnrunol. 153:4Y25. 9. Kohl, S.. N. C. J . Strynadka. R. S. Hodges, and L. Pereira. 1990. Analysls of the role ot antibody-dependent cellular cytotoxic antibody activity in murine neonatal herpes simplex vwus infection with antibodies to synthetic peptides of glycoprotein D and monoclonal antibodies to glycoproteln B J. Clin. I n I'CSI. X6:273. IO. Kurt-Jones, E. A,,D. Liano. K. A. HayGIass, B. Benacerraf, M:S. Sy, and A. Ahhas. 1988. The role of antigen-presenting B cells in T cell pr~mingin VIVO: studies of B cell-deficient mice. J. hnmunol. 14013773. I 1 Macatonia, S. E.. P. M. Taylor, S . C. Knight, and B. A. Askonas. 1989. Primary stimulation by dendrltic cells induces antiviral proliferative and cytotoxic T cell responses in vitro. J. Exp. M d 169:1255. 12. Manischew~tz.J. E., and G . V. Quinnan. Jr. 1980. Antivirus antibody-dependent cell-mediated cytotoxicity during murine cytomegalovirus infectton. 1nfec.t. I m INUII. ZY: IOSO. 13. Miklos, K.. M. Tolnay, H. Bazin, and G. A. Medgyesi. 1993. Rat IgG subclasses mediating binding and phagocytosis of target cells by homologous macrophages. Mol. / l l ~ l i l l ~ ~30: ~ J /127.3. . 14. Mitchell, R. N.. K. A. Barnes, S . A. Grupp, M. Sanchez. Z. Misnlov~n. M.C. Nuswnzweig. and A. K. Abbas. 1995. Intracellular targeting of antigens inlernalizcd by nlemhrane imn~unoglohulinin Blymphocytes. J. EA^. Mead. IX1:17OS. 15. Qu~nnan. G. V. Jr., J. E. Manlschewitz, and N . Kirnlani. 1982. Involvement of natural M e r cclls in the pathogenesis of murine cytomegalovirus interititial pneumonltih and the immune response to infection. 1. Gen. V i m l . 58: 173. 16. Rossi-Bergmann. B.. 1. Muller, and E. B. Godinho. 1993. THI and TH2 T-cell subsets are differentially activated by macrophages and B cells i n murine leishmaniasis. Infect. lnmrun. 6/:2266. 17. Scheicher. C., M Mehlig. H.-P. D~enes.and K. Reske. 1995. Uptake of microparticle-adsorbed protein antigen by hone marrow-derived dendritic cells results in up-regulatlon of interleukin-la and interleukin-I2 p4O/p35 and tr~ggersprolonged. efficient antigen presentation. Eur. J. Immunol. 25:1566. I 8. Snapper. C. M., and W. E. Paul. 1987. Interferon-y and B cell stimulatory factor-1 reciprocally regulate Ig so type production. Nufure 2.