VIRAL IMMUNOLOGYVolume 13, Number 1, 2000Mary Ann Liebert, Inc.Pp. 57-72

The Immunogenicity of Subunit Vaccines for RespiratorySyncytial Virus after Co-formulation with AluminumHydroxide Adjuvant and Recombinant Interleukin-12

GERALD E. HANCOCK, JASON D. SMITH, and KRISTEN M. HEERS

ABSTRACT

The effects of recombinant interleukin-12 (rIL-12) on immune responses generated by sub-unit vaccines for respiratory syncytial virus (RSV) were evaluated in BALB/c mice. Par-enteral co-administration of rIL-12 with F/AIOH or F/PBS resulted in accelerated clearanceof infectious virus from the lungs 4 days after challenge. The immune responses elicited by0.03 /¿g of F protein plus 10 ng of rIL-12 adsorbed to AlOH were more efficacious thanthose induced by 3 ¿tig of F protein co-formulated with 1,000 ng of rIL-12 in PBS alone. Ad-sorption to AlOH prolonged the presence of rIL-12 in the sera. The resultant systemic hu-moral immune responses after vaccination with F/AIOH or G/AIOH were dependent on thedose of rIL-12 and characterized by heightened serum immunoglobulin G2a (IgG2a) anti-body titers. Co-administration of rIL-12 with F/AIOH was also associated with diminishedprotein-specific IgE titers, elevated neutralizing antibody titers, and interferon-y and (IFN-y) in the sera, and enhanced antigen-dependent killer cell activity in the lungs after chal-lenge. For maximum benefit, the data suggested that rIL-12 must be co-administered withF/AIOH. Collectively, the results indicated that rIL-12 directed immune responses towarda type 1 phenotype. However, examination of pulmonary inflammatory cells after challengesuggested that the type 1 phenotype was not absolute. Co-formulation with rIL-12 did notdiminish pulmonary eosinophilia upon challenge of naive mice primed with F/AIOH,G/AIOH, or FI-RSV, and CD4+ T cells were expanded relative to the CD8+ T-cell com-

partment. These results are important for the future design of subunit vaccines against RSV.

INTRODUCTION

Respiratory syncytial virus (RSV) is a negative-strand RNA virus that is recognized as a major cause ofrespiratory tract disease in human infants, aged adults, and patients with immunological abnormalities

(6,7,19,46). RSV is the major infectious agent for bronchopneumonia, bronchiolitis, and pneumonia in veryyoung infants, and infants and children with congenital heart disease, bronchopulmonary dysplasia, or cys-tic fibrosis. Asthma and atopy are also associated with RSV bronchiolitis early in life (11,44). Despite theacknowledged role of RSV in human disease, development of attenuated or subunit vaccines has not been

Department of Immunology Research, Wyeth-Lederle Vaccines, West Henrietta, NY.

57

HANCOCK ET AL.

successful. For subunit vaccines, efforts have focused on natural or recombinant fusion (F) protein, eitheralone, or in combination with the major surface (G) glycoprotein of RSV (3,20-22). Several studies in hu-mans (8,18,26,36,37) have suggested important roles for both glycoproteins in the generation of protectiveimmune responses against RSV. However, the generation of type 2-dominated immune responses follow-ing vaccination of naive subjects with subunit vaccine antigens may be problematic. The hypothesis is thattype 2 immune responses were responsible for the exacerbated disease that was observed in naive humaninfants immunized with formalin inactivated (FI-RSV) vaccine following subsequent infection (29,31,32,39).Type 2 immune responses to RSV antigens have been associated with pathology and atypical inflamma-tory responses in the pulmonary tissues of rodents (1,2,12,13,17,49) and nonhuman primates (30). Thus,key to the evolution of subunit vaccines against RSV is the nature of T-cell responses induced after im-munization.

Interleukin-12 (IL-12) is a heterodimeric cytokine primarily produced by dendritic cells, macrophages,and neutrophils and a potent inducer of type 1 immune responses and cell-mediated immunity (4,43). Thepotential of recombinant (r) IL-12 to function as an adjuvant for highly purified subunit vaccine antigenswas implied recently in the BALB/c mouse model of RSV infection. Via multiple systemic injections, rlL-12 was shown to alter the type 2 phenotype observed in mice sensitized to G protein expressed by recom-

binant vaccinia virus (25). In a separate report (47), co-administration of rIL-12 simultaneously with FI-RSV vaccine resulted in enhanced anti-RSV immunoglobulin G2a (IgG2a) antibody titers. rIL-12 was alsorecently demonstrated to have adjuvant properties for several other subunit vaccine antigens [e.g., heat-killed Listeria monocytogenes (23), acellular Bordetella pertussis antigens (34), and irradiated Shistosomamansoni cercariae (52)]. Furthermore, it was recently documented that adsorption to aluminum hydroxide(AlOH) increased the adjuvant activity of rIL-12 for the rgpl20 envelope protein of HIV-1 (27). Therefore,we assessed the impact of rIL-12 on both the quantitative and qualitative aspects of the immune responsesgenerated following intramuscular (i.m.) injection of natural F or G proteins, in phosphate-buffered saline(PBS) alone, or after adsorption to AlOH adjuvant. Our results indicate that rIL-12 enhances the ability ofF/AIOH and G/AIOH to induce type 1 immune responses and is therefore a potent adjuvant for AlOH-basedvaccines against RSV. Co-administration of rIL-12 with F/AIOH augmented antigen-dependent killer cellactivity, complement-assisted serum-neutralizing antibody titers, and the secretion of interferon-y (IFN-y)into the sera. The presence of rIL-12 in the vaccine was also associated with a reduction in F protein-spe-cific, serum IgE antibody titers. However, the pulmonary eosinophilia associated with prior injection ofnaive recipients with either natural F/AIOH or G/AIOH was not diminished after challenge. Thus, type 1immune responses, although elevated following co-administration of rIL-12, were not completely dominant.

MATERIALS AND METHODS

Animals. Female BALB/c mice (8-10 weeks of age) and Sprague Dawley rats (retired breeders) were

obtained from Taconic (Germantown, NY). All animals were housed in a facility accredited by the Amer-ican Association for Accreditation of Laboratory Animal Care.

Antigens, cytokines, antibodies. Natural F and G glycoproteins were purified (22) from Vero cells(ATCC No. CCL 81) infected with the A2 strain of RSV. The proteins were greater than 95% pure as es-timated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and antigen-captureenzyme-linked immunosorbant assay (ELISA). The formalin-inactivated vaccines were prepared as de-scribed (22) using the A2 and C243 strains of RSV (FI-RSV) and parainfluenza virus type 3 (FI-PIV), re-

spectively. Recombinant murine IL-12 was obtained from Genetics Institute (Cambridge, MA). Recombi-nant murine IFN-y and monoclonal antibodies (mAb) specific for IFN-y (XMG1.2 and R4-6A2) were

purchased from PharMingen (San Diego, CA).Immunizations and challenges. All immunizations were i.m. (0.1 mL). In experiments that used two

vaccinations, mice were injected at weeks 0 and 4. The F (0.03 or 3.0 pg/dose.) or G (1.0 ¿ig/dose) proteinvaccines were adsorbed to AlOH (100 ¿ig/dose) adjuvant and co-administered with ascending doses of rlL-12 (10-1,000 ng of rIL-12/dose). The formalin-inactivated vaccines, FI-RSV and FI-PIV, were also ad-sorbed to AlOH (1.6 mg/dose). Control mice were vaccinated with F/AIOH, G/AIOH, FI-RSV, or FI-PIV

58

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

in PBS alone, F protein admixed with 20 ¿ug/dose QS-21 (F/QS-21, Aquila BioPharmaceuticals, Inc.,Worcester, MA), F protein in PBS alone (F/PBS), or PBS/AlOH. Additional control mice were immunizedby infection with RSV A2 (~2 X 106 pfu) clarified from HEp-2 cells (ATCC No. CCL 23), or received an

equal amount of mock-infected HEp-2 cell antigen control. In experiments to test the potency of rIL-12 in-jected at a distal site, mice were primed with F/AIOH and 100 ng rIL-12 was co-administered using threedistinct protocols: (i) In the distal protocol, the mice were injected with F/AIOH in PBS alone in one thighand received rIL-12 in the contralateral thigh, (ii) In the proximal protocol, the mice received F/AIOH andrIL-12 in two separate injections. The second injection was administered at approximately the same sitewithin minutes of the first, (iii) In the same protocol, the mice received one injection of F/AIOH co-for-mulated overnight with rIL-12. Studies demonstrated (data not shown) that F protein and rIL-12 were bothadsorbed to AlOH after overnight incubation at 4°C. The A2 strain was used for all intranasal (i.n.) chal-lenges (approximately 2 X 106 pfu). All i.n. challenges (0.05 mL) were 4 or 5 weeks after primary immu-nization and were performed under injectable anesthesia (ketamine, The Butler Co., Dublin OH, 60 mg/kg).

BAL/flow cytometry. BAL was performed as previously described (22) 5 or 7 days after challenge. Therelative percentage of eosinophils was determined after staining cytopreps with Diff-Qik (Dade Interna-tional, Miami, FL) and the examination of at least 400 leukocytes. Flow cytometry (FACSort, Becton Dick-inson, Mountain View, CA) was performed on pooled BAL cells from 5 mice per group, 5 days after chal-lenge. Standard flow cytometric techniques employed phycoerythrin (PE)-conjugated anti-mouse CD3(17A2) and CD4 (H129.19), and fluorescein isothiocyanate (FITC) conjugated anti-mouse CD8a (53-6.7)antibodies to detect T lymphocyte surface markers. The BAL cells were preincubated with Fc block reagent(anti-CD 16/CD32, 2.4G2) prior to staining, and propidium iodide was added prior to collection. The rela-tive percentage of CD3+, CD4+, or CD8+ cells was determined after gating on live lymphocytes. The mAband reagents were purchased from PharMingen. The data are presented as the mean percent (±1 SD) ofthree independent studies.

Determination of percent cytotoxicity. Antigen-dependent killer cell activity was determined ex vivoas previously described (20) 5 days after challenge in a standard 4-hour 51Cr (Amersham Corp., ArlingtonHeights, IL) release assay using an actively infected (22) syngeneic cell line (15) as target cells. The tar-

gets were infected (1 hour at 37°C, multiplicity of infection of 10) with the A2 strain of RSV and incu-bated overnight to allow optimal protein expression. Percent specific release was calculated by 100 X [(meancpm experimental)

—

(mean cpm spontaneous release)]/[(mean cpm total release)—

(mean cpm sponta-neous release)].

Determination of RSV infectivity. The detection of infectious virus in the pulmonary tissues after chal-lenge was assessed in a plaque assay as described employing HEp-2 cell monolayers (20). Briefly, the lungand trachéal tissues were removed en bloc 4 days after challenge, homogenized, clarified, snap-frozen, andstored at

—

70°C until assayed.Serum antibody determinations. Four and 2 weeks after primary and secondary vaccination, respec-

tively, geometric mean serum end point IgG and neutralizing antibody titers were determined as previouslydescribed by ELISA and the plaque reduction neutralization test, respectively (20-22). The neutralizing an-

tibody titers were determined against the A2 strain of RSV in the presence or absence of 5% (vol/vol) guineapig serum (GenTrak, Inc., Plymouth Meeting, PA) as a source of complement (C). The neutralization titerswere calculated as the reciprocal of the serum dilution that showed 60% reduction (relative to the virus con-

trol) in the number of foci per well. The passive cutaneous anaphylaxis (PCA) model was used to ascer-

tain F protein-specific IgE antibody titers (10). Briefly, pooled sera from 5 mice per group were serially di-luted three-fold in PBS and injected intradermally (0.1 mL) in the shaved backs of rats sedated with ketamine(40-80 mg/kg) and xylazine (2.5-5 mg/kg) (The Butler Co.). Approximately 48 hours thereafter, the ratswere challenged intravenously (i.v.) with a 1.0 mL of solution containing 2 mg of natural F protein and 1 %Evans Blue dye in PBS. The presence of F protein-specific IgE antibodies was indicated by the blue dyeat the injection site. The titer is the reciprocal of the greatest dilution showing no evidence of Evan's BlueDye in the skin.

Serum IFN-y and IL-12. For the determination of IFN-y and IL-12, the sera were collected pre-injec-tion and 1, 2, 3, 4, 5, and 7 days after injection of F/AIOH plus rIL-12. IFN-y was determined by ELISAas previously described (22). The optical density was measured on an ELISA reader (Dynatech) at 490 nm

59

HANCOCK ET AL.

with a reference of 630 nm. The amount of cytokine was determined by interpolating on the linear portionof a standard curve (log-log plot) that was generated using rIFN-y. The limit of detection of theseassays is defined as the amount of recombinant cytokine that gives an optical density (OD) value equal to3 SD above the average OD490 of eight wells containing no cytokine. IL-12 was detected using the Quan-tikane ™ M mouse IL-12 (p70) Immunoassay kit following manufacturer's (R&D Systems, Minneapolis,MN, Cat. No. M1270) instructions.

Statistical analyses. Significant differences (<0.05) were determined after log transformation by Tukey-Kramer HSD multiple comparison or Student's r-test using JMP® statistical discovery software (SAS In-stitute Inc., Cary, NC). Unless otherwise noted, the data are expressed ± 1 SD.

RESULTS

The effect of rIL-12 on protective immune responses. The results indicated that rIL-12 enabled 3 pgof natural F protein, prepared in PBS alone, to elicit immune responses in BALB/c mice that were more

efficacious after primary immunization (Table 1). Significant reductions in infectious virus were observedin the lungs after immunization with F/PBS plus either 100 or 1,000 ng of rIL-12, 4 days after challenge(Table 1). However, clearance was not complete and the infectious virus titer diminished only 10-fold. Incomparison, primary immunization with 3 pg of F protein adsorbed to AlOH was able to accelerate theclearance of virus from the pulmonary tissues to background levels. Thus, to determine if rIL-12 could aug-ment protective immune responses, additional studies with a suboptimal dose (100-fold less) of F proteinadsorbed to AlOH were performed. Primary immunization with 0.03 pg of natural F protein adsorbed toAlOH (in PBS alone) conferred only partial protection (Table 1, Exp. 2). Although infectious virus titers

Table 1. rIL-12 Augments the Capacity of NaturalF Protein Prepared in PBS Alone, or Adsorbed to

AlOH, to Elicit Immune Responses That InhibitVirus Replication in the Lungs of BALB/c Mice

Experimentnumber Vaccine3 rIL-12 (ng) GMT RSV (logw)h1 F/PBS None 5.4 ± 0.3

F/PBS 10 5.0 ± 0.4F/PBS 100 4.4 ± 0.6CF/PBS 1,000 4.1 ± 0.7°

F/AIOH None <1.6PBS/AIOH None 5.5 ± 0.3

RSV None <1.6

2 F/AIOH None 3.1 ± 0.3F/AIOH 10 1.8 ± 0.2dF/AIOH 100 2.2 ± 0.2d

PBS/AIOH None 4.5 ± 0.2

aBALB/c mice were primed with 3 pg of F protein (experiment number1) or 0.03 pg of F protein (experiment number 2). The vaccines were ad-mixed with the indicated 10-fold ascending doses of rIL-12. Control micewere injected with either F/AIOH, PBS/AIOH, or experimentally infectedwith the A2 strain of RSV.

bThe mice were challenged 4 weeks after primary vaccination. GMT isthe geometric mean titer ofRSV per gram ofpulmonary tissue. The data inexperiment number 2 are the results of three independent experiments with5-10 mice per group.

cp < 0.05 vs. F/PBS, PBS/AIOH, and F/AIOH.dp < 0.05 vs. PBS/AIOH, and F/AIOH.

60

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

Table 2. The Systemic Humoral Immune Responses of BALB/c Mice Immunized with F/AIOH Plus rIL-12

Geometric mean serum antibody titers (log¡o)*

Vaccine

Anti F Protein NeutralizingrIL-12 (ng) IgG IgG, IgG2 IgEb (+C) (-C)

F/AIOHF/AIOHF/AIOHF/AIOHPBS/AIOH

1,000100

1000

6.76.45.85.4

0.10.10.1C0.1C

<1.7

5.45.45.45.1

0.10.10.10.1

NT

6.0 ± 0.25.5 ± 0.24.4 ± 0.2C3.8 ± 0.2f

NT

<54515

135<5

2.1 ± 0.1h1.4 ± 0.11.4 ± 0.1

<1.3<1.3

<1.3<1.3<1.3<1.3<1.3

F/AIOHF/AIOHF/AIOHF/AIOHPBS/AIOH

1,000100

1000

7.1 ± 0.1d6.7 ± 0.16.5 ± 0.16.2 ± 0.1e

<2.0

5.95.96.15.8

0.10.10.10.1

6.25.75.34.7

0.20.20.2«

: 0.2fNT NT

<5>135>135>135<5

2.7 ± 0.2'2.3 ± 0.2"2.0 ± 0.21.7 ± 0.2

<1.3

1.7 ± 0.11.5 ± 0.1

1.31.5

<1.3

aBALB/c mice were vaccinated on weeks 0 and 4 with natural F protein (3 /xg/dose) adsorbed to AlOH and the indi-cated doses of rIL-12. Control mice were injected with PBS/AIOH. The upper and lower panels depict geometric mean

antibody titers 4 and 2 weeks after primary and secondary vaccination, respectively. The numbers are the end point IgGand neutralizing antibody titers ± 1 SE. There were 5 mice per group. NT denotes not tested. The data were confirmed inseveral studies with similar results.

bThe F protein-specific IgE antibody titers were determined by PCA reaction on pooled serum samples. The resultswere confirmed in two other studies with similar results.

cp < 0.05 vs. the IgG or IgG2a titers elicited after primary vaccination with F/AIOH plus 100 or 1,000 ng of rIL-12.dp < 0.05 vs. the IgG titers elicited after secondary vaccination with F/AIOH plus 0, 10, or 100 ng of rIL-12.ep < 0.05 vs. the IgG titers elicited after secondary vaccination with F/AIOH plus 100 ng IL-12.{p < 0.05 vs. the IgG2a titers elicited after vaccination with F/AIOH plus 10, 100, or 1,000 ng of rIL-12.8p < 0.05 vs. the IgG2a titers elicited after vaccination with F/AIOH plus 1,000 ng of rIL-12.hp < 0.05 vs. C assisted neutralizing antibody titers elicited after vaccination with F/AlOH plus 0, 10 or lOOngof rIL-12.lp < 0.05 vs. the C assisted neutralizing antibody titers elicited after vaccination with F/AIOH plus 0 or 10 ng of rIL-12.>p < 0.05 vs. the C assisted neutralizing antibody titers elicited after vaccination with F/AIOH plus 0 ng of rIL-12.

in lungs were significantly less when compared to mice injected with PBS/AIOH, the titers were not re-

duced to background levels as observed after immunization with 3 pg of F protein adsorbed to AlOH (Table1, Exp. 1). The data from three independent studies demonstrated that co-administration of 10 or 100 ng ofrIL-12 significantly enhanced the protective immune responses elicited by 0.03 pg of natural F protein ad-sorbed to AlOH (Table 1, Exp. 2). For example, only 1.8 ± 0.2 logio pfu virus per gram of pulmonary tis-sue were recovered from mice primed with 0.03 pg of natural F protein adsorbed to AlOH plus 10 ng ofrIL-12. In contrast, the lungs of control mice injected with the same dose of F/AIOH in PBS alone, or

PBS/AIOH, had 3.1 ± 0.3 and 4.5 ± 0.2 logio pfu virus per gram of tissue, respectively, 4 days after chal-lenge.

Humoral and cell-mediated immune responses after co-administration of rIL-12 with F/AIOH or

G/AIOH. The results from the analysis of systemic humoral immune responses (Table 2) 4 and 2 weeksafter primary and secondary vaccination, respectively, agreed with the studies on efficacy (Table 1). Thedata suggested that rIL-12 transformed the quality of protective immune responses elicited by 3 pg of Fprotein adsorbed to AlOH adjuvant. For example, 4 weeks after primary immunization with F/AIOH co-

formulated with 100 or 1,000 ng of rIL-12, the total IgG end point antibody titers were significantly in-creased 10 and 20 times, respectively, when compared to the serum end point titers of mice vaccinated withF/AIOH alone (Table 2, upper panel). In similar fashion, the total IgG antibody titers were elevated 3 and8 times, respectively, 2 weeks after secondary vaccination (Table 2, lower panel). The data further impliedthat the increase in IgG antibody titers was dependent on the dose of rIL-12 in the vaccine. This was bestexemplified after the determination of F protein-specific IgG2a titers 4 weeks after primary immunization.

61

HANCOCK ET AL.

When contrasted with the serum anti-F protein IgG2a antibody titers elicited with F/AIOH alone, thevaccines formulated with F/AIOH plus either 10, 100, or 1,000 ng of rLL-12 were significantly elevated A-,50-, and 158-fold, respectively (Table 2, upper panel). Elevations in IgG2a antibody titers were also ob-served 2 weeks after secondary vaccination with F/AIOH plus rIL-12 (Table 2, lower panel). Most impor-tantly, the data suggested that the addition of rIL-12 could increase the capacity of F/AIOH to generate

B.40-1

40-1

30 H

20 H

îo H

OH

-10 ~~I I I I I I10 20 30 40 50 60

E:T Ratio

-#- F/ALOH

F/ALOH/IL12(0.01^g)

-H- F/ALOH/IL12(0.1pg)

-_I- F/ALOH/IL12 (1pg)

F/ALOH

F/ALOH/IL12(0.01tig)

F/ALOH/IL12(0.lMg)

-£- F/ALOH/IL12 (1(jg)

FIG. 1. The effect of rIL-12 on the ability of F/AIOH to promote cell-mediated immune responses in the lungs ofBALB/c mice after challenge. BALB/c mice were primed with 3 /ixg of F protein adsorbed to AlOH plus the indicatedamounts of rIL-12. The mice were challenged with the A2 strain of RSV 4 weeks later. BAL cells were obtained 5days thereafter, pooled, and tested ex vivo for cytolytic activity against RSV-infected (filled symbols) and control (opensymbols) syngeneic target cells. A and B show the results from two independent experiments. There were 10 mice pergroup.

62

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

Table 3. The Systemic Humoral Immune Responses of BALB/c Mice Immunized with G/AIOH Plus rIL-12

Serum anti-G protein antibody titers (log¡o)bVaccine3 rIL-12 (ng) IgG IgG] IgG2a

G/AIOH None 5.3 ± 0.2 5.0 ± 0.2 <3.0G/AIOH 10 5.6 ± 0.4 5.0 ± 0.4 3.9 ± 0.6dG/AIOH 100 5.7

_

0.4 4.3 ± 0.4C 4.7 ± 0.5d

"Naive BALB/c mice were primed with natural G protein ( 1 /Ltg/dose) adsorbed to AlOH plus the indicated doses ofrIL-12.

bThe geometric mean antibody titers were determined on sera collected 4 weeks after primary vaccination. There were

5 mice per group. The results are representative of three separate studies.cp < 0.05 vs. the IgGi antibody titers from mice primed with G/AIOH alone.dp < 0.05 vs. IgG2a antibody titers from mice primed with G/AIOH alone.

functional serum antibody titers. The complement-assisted neutralizing antibody titers in the sera of miceinjected with F/AIOH plus 1,000 ng of rIL-12 were significantly heightened at least six-fold, 4 weeks af-ter primary immunization (Table 2, upper panel). Following secondary vaccination with F/AIOH formu-lated with either 100 or 1,000 ng of rIL-12, the neutralizing antibody titers were significantly increased 4and 10 times, respectively (Table 2, lower panel).

The results demonstrated that rIL-12 could also enhance the capacity of F/AIOH to generate cell-medi-ated immune responses in BALB/c mice. In the experiment depicted in Fig. 1A, ex vivo antigen-dependentkiller cell activity (40% cytolysis) in the lungs of mice primed with F/AIOH plus 10 ng of rIL-12 was de-tected at the 10:1 effector to target ratio 5 days after challenge. The antigen-dependent killer cell activityelicited after primary vaccination with F/AIOH plus 100 ng of rIL-12 was nearly four-fold less (Fig. 1A).

Table 4. The Requirement for F/AIOH and rIL-2 at the Same Injection Site to GenerateAugmented Systemic Humoral Immune Responses in a Single Injection Protocol"

Serum antibody titers (logjo)Anti F Protein Neutralizing

Vaccine rIL-12 (ng) Site IgG¡ IgG2a (+C) (-C)

F/AIOH 100 Distal 5.3 ± 0.2b 4.7 ± 0.3b 1.8 ± 0.4be <1.3F/AIOH 100 Proximal 5.7 ± 0.4b 5.6 ± 0.4d 1.7 ± 0.2b'e <1.3F/AIOH 100 Same 6.0 ± 0.1b 6.2 ± 0.2C 2.5 ± 0.5f <1.3

F/AIOH 0 Same 5.5 ± 0.2 5.2 ± 0.3 1.6 ± 0.3 <1.3F/PBS 0 Same 3.3 ± 0.4 <3.0 <1.3 <1.3PBS 0 Same NT NT <1.3 <1.3

"Naive BALB/c mice were primed with natural F protein (3 /ag/dose) adsorbed to AlOH. F/AIOH was co-administeredwith 100 ng of rIL-12 using three distinct protocols as described in the Materials and Methods. Additional control micewere vaccinated with F/AIOH alone, F/PBS, or PBS/AIOH. The geometric mean antibody titers were determined on seracollected 4 weeks after primary immunization. NT denotes not tested. There were 5 mice per group. The results wereconfirmed in a separate study with similar results.

bp > 0.05 vs. the serum antibody titers of mice vaccinated with F/AIOH alone.cp < 0.05 vs. the serum antibody titers of mice vaccinated with F/AIOH alone or F/AIOH plus 100 ng of rIL-12 injected

at a distal site.dp < 0.05 vs. the serum IgG2a antibody titers of mice vaccinated with F/AIOH plus 100 ng of rIL-12 injected at a distal

site."p < 0.05 vs. the serum-neutralizing antibody titers of mice vaccinated with F/AIOH plus 100 ng of rIL-12 injected at

the same site.fp < 0.05 vs. the serum antibody titers of mice vaccinated with F/AIOH alone.

63

HANCOCK ET AL.

The effect of rIL-12 on cell-mediated immune responses was confirmed in a separate study (Fig. IB). How-ever, in the second experiment, the data implied that 100 ng of rIL-12 was the appropriate dose to gener-ate antigen-dependent killer cells. The ex vivo killer cell activity was approximately 30% cytolysis at an ef-fector to target ratio of 50:1. Killer cell activity appeared less in the pulmonary tissues of mice vaccinatedwith F/AIOH plus either 10 or 1,000 ng of rIL-12 (Fig. IB). In neither experiment was killer cell activitydetected in the lungs of mice primed with F/AIOH in PBS alone 5 days after challenge. Thus, in a primaryimmunization protocol with F/AIOH, the results implied that the optimum dose for heightened cell-medi-ated immune responses was between 10 and 100 ng of rIL-12. In both experiments, the killer cell activitywas antigen-dependent because syngeneic targets not infected with RSV were not lysed (Fig. 1A, B).

The systemic humoral immune responses of BALB/c mice primed with G/AIOH plus rIL-12 were alsoexamined. In three independent experiments, primary immunization with G/AIOH plus either 10 or 100 ngof rIL-12 induced statistically augmented anti-G protein IgG2a antibody titers when compared to G/AIOHalone (Table 3). In addition, serum anti-G protein IgGi antibody titers were significantly less in mice primedwith G/AIOH plus 100 ng of IL-12. Co-administration of rIL-12 with G/AIOH did not result in statisticallyaugmented complement-assisted serum-neutralizing antibody titers (data not shown).

The effect of rIL-12 administration at a distal site. Important to the formulation of future vaccineswas the question of whether antigen and rIL-12 were required in close proximity, at the site of injection,for the generation of augmented anti-F protein antibody titers. To that end, studies were performed whereinthe immune responses elicited after injection of rIL-12 at a site distal to that of antigen administration were

compared to those generated after co-administration of vaccine antigen plus rIL-12 at the "same" site. Sim-ilar to previous data (Table 2), the results demonstrated that adsorption of rIL-12 to AlOH adjuvant en-hanced the capacity of F/AIOH to generate systemic humoral immune responses in BALB/c mice (Table4). The most meaningful impact of rIL-12 was on the generation of serum IgG2a antibody titers. A singleinjection of F/AIOH plus 100 ng of rIL-12 resulted in end point anti-F protein IgG2a antibody titers thatwere statistically elevated 4 weeks after primary vaccination (Table 4). When compared to immunizationwith F/AIOH in PBS alone, protein-specific IgG2a antibody titers were 10 times higher. More importantly,the addition of 100 ng of rIL-12 to F/AIOH elicited complement-assisted neutralizing antibody titers thatwere also statistically augmented nearly 10-fold (Table 4).

A.

E'c3

EInCO

Q.

CM

E

CO

10000 -i

7500 H

5000 H

2500 H

2 4 6

Days Post Vaccination2 4

Days Post Vaccination

FIG. 2. IL-12 and IFN-y in the sera of BALB/c mice primed with F/AIOH plus rIL-12. BALB/c mice were primedwith 3 pg of natural F protein. The protein was adsorbed to AlOH or prepared in PBS alone. rIL-12 was added to thevaccine formulation at 1,000 ng per dose. At the indicated time points IFN-y (A) and IL-12 (B) were determined inthe sera. Each data point represents the mean of 5 mice. When compared to all groups, adsorption to AlOH resulted insignificantly (p < 0.05) increased amounts of IL-12 in the sera 2, 3, and 4 days after injection.

64

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

The data further suggested that in a one-injection protocol rIL-12 must be proximal to F/AIOH to gen-erate augmented systemic humoral immune responses. For example, the anti-F protein IgG2a antibody titersof mice primed with F/AIOH and administered 100 ng of rIL-12 at a distal site were significantly less thanthose of mice primed with F/AIOH and given a separate injection of 100 ng of IL-12 at approximately thesame site (Table 4). The antibody titers of the former group were not significantly different from those ofmice vaccinated with F/AIOH in PBS alone. Indeed, to achieve maximal benefit, the data implied that rlL-12 must be administered together with antigen to have its greatest impact on the generation of systemic hu-moral immune responses. Although injection of 100 ng of rIL-12 proximal to F/AIOH resulted in elevatedanti-F protein IgG2a antibody titers (compared to distal administration), they were not significantly greaterthan those elicited after vaccination with F/AIOH in PBS alone. Moreover, injection of 100 ng of rIL-12proximal to F/AIOH did not augment functional antibody titers (Table 4). The complement-assisted neu-

tralizing antibody titers were not significantly different from those of mice vaccinated with F/AIOH in PBSalone 4 weeks after primary immunization.

The effects of rIL-12 on the generation of type 1 immune responses. The enhanced cell-mediated im-mune responses (Fig. 1) and statistically significant increases in IgG2a and functional antibody titers (Ta-bles 2^4) suggested that rIL-12 increased the generation of "type 1" immune responses in BALB/c mice.For further clarification, the effects of rIL-12 on the generation of F protein-specific IgE antibodies were

examined after vaccination with F/AIOH. The data from the PCA test suggested that rIL-12 diminished thepotential of F/AIOH to elicit IgE antibodies in a dose-dependent manner (Table 2). The serum anti-F pro-tein IgE antibody titer 4 weeks after primary vaccination with F/AIOH in PBS alone was 135. In contrast,co-formulation of F/AIOH with 10, 100, or 1,000 ng of rIL-12 resulted in protein-specific IgE antibodytiters of 15, 45, and <5, respectively (Table 2, upper panel).

The impact of rIL-12 on the induction of type 1 immune responses was further assessed in separate stud-ies that determined the presence of IL-12 and IFN-y in the sera, and IL-5-dependent eosinophilia in the

Table 5. Eosinophils in the Lungs after Challenge of BALB/cMice Immunized with Subunit Vaccines Prepared with rIL-12

Experiment number Vaccine3 rIL-12 (ng)3 Percent EOSb

1 FI-RSV 100 20 ± 10cFI-RSV 0 40±lldG/AIOH 100 33 ± 5deG/AIOH 0 35 ± 2dFI-PIV3 0 9 ± 9F/QS-21 0 7 ± 5

RSV 0 <12 F/AIOH 100 23 ± 14f

F/AIOH 10 36 ± 15fF/AIOH 0 22 ± 12

RSV 0 <1

"BALB/c mice were challenged with RSV 4 (experiment number 1) or 2 (experimentnumber 2) weeks after primary or secondary immunization, respectively, with the indi-cated vaccines.

bPulmonary eosinophilia (EOS) was assessed after BAL following the enumerationof at least 400 leukocytes and expressed as the mean percent (±1 SD). There were 5mice per group. The data were confirmed in separate studies with similar results. In ex-

periment number 2, the numbers are the results of three separate experiments withF/AIOH.

cp < 0.05 vs. the eosinophils detected in mice vaccinated with FI-RSV.dp < 0.05 vs. the eosinophils detected in mice vaccinated with FI-PIV3, F/QS-21, or

infectious RSV.ep > 0.05 vs. the eosinophils detected in mice vaccinated with G/AIOH.f/> > 0.05 vs. the eosinophils detected in mice vaccinated with F/AIOH.

65

HANCOCK ET AL.

lungs after challenge. Striking amounts of IL-12 and IFN-y were observed in the sera after administrationof F/AIOH plus 1,000 ng of rIL-12 (Fig. 2). Maximum amounts of IFN-y were observed in the sera 2 daysafter injection of rIL-12 (Fig. 2A). The results indicated that the quantity and duration of IFN-y in the serawere not dependent on the presence of F protein or AlOH in the vaccine. The amount of IFN-y in the seraof mice administered rIL-12 in PBS alone was not significantly different from that of mice immunized withF/PBS or F/AIOH plus rIL-12. IFN-y was not detected in the sera of mice injected with F/PBS or F/AIOHalone. The results suggested, however, that adsorption to AlOH prolonged the bioavailability of IL-12. Peakamounts of IL-12 were observed 1 day after injection of F/AIOH plus rIL-12, and remained elevated in thesera 4 days thereafter (Fig. 2B). In contrast, IL-12 was not detected in the sera of mice administered F/PBS,F/AIOH alone, or rIL-12 in PBS alone 2 days after injection.

The presence of IFN-y and IL-12 in the sera, however, did not diminish the type 2 immune responsesassociated with IL-5-dependent pulmonary eosinophilia (Table 5). In several studies, significant reductionsin pulmonary eosinophilia were not observed upon challenge of mice co-administered 100 ng of rIL-12with FI-RSV, F/AIOH, or G/AIOH. Even 1,000 ng of rIL-12 failed to diminish eosinophilia in the lungs ofmice vaccinated with F/AIOH or G/AIOH (data not shown). As previously reported (22), pulmonaryeosinophilia upon challenge of mice vaccinated with F/QS-21, FI-PIV, or infectious RSV was not observed(Table 4).

In separate studies, the pulmonary infiltrates of mice primed with F/AIOH plus 1,000 ng of rIL-12 were

analyzed for T-cell phenotype via flow cytometry 5 days after challenge. The results suggested that co-for-mulation with rIL-12 resulted in an expansion of CD4+ T cells relative to the CD8+ T-cell compartment(Table 6). When compared to F/AIOH alone (11 ± 4%), the lungs of mice primed with F/AIOH plus rlL-12 contained a significantly increased percentage of CD4+ T lymphocytes (20 ± 1%) after challenge. Thepercentage of CD8+ T cells (12 ± 8%) was not significantly elevated. In comparison, the percentage ofCD4+ or CD8+ T cells detected in the lungs after challenge of mice primed with F/QS-21 was 16 ± 1%and 17 ± 6%, respectively (Table 6). When compared to mice primed with F/PBS, the percentage of CD4+or CD8+ T cells in the lungs of mice primed with F/QS-21 was significantly elevated. It is noteworthy thatimmunization with F/QS-21 is not associated with pulmonary eosinophilia (Table 5, ref. 22) after challenge.

DISCUSSION

RSV has been recognized as a major etiologic agent of human respiratory tract disease for more than 40years (6,7,46). Unfortunately, development of safe and efficacious vaccines has not been successful. Theprimary factor limiting the evolution of subunit vaccines is the history of enhanced disease in seronegativerecipients immunized with a nonreplicating FI-RSV vaccine adsorbed to AlOH adjuvant (38). On the ba-

Table 6. The Phenotypic Analysis of Pulmonary T Cells afterChallenge of Mice Primed with F/AIOH Plus rIL-12a

Vaccine rIL-12 (ng) Percent lymphocytes Percent CD3 Percent CD4 Percent CD8

F/AIOH 1,000 43 ± 8 31 ± 9 20 ± 1 12 ± 8F/AIOH None 29 ± 5b-c 17 ± 2b-c 11 ±4" 6 ± 2CF/PBS 1,000 30 ± Ie 15 ± lbc 9 ± 3b-c 7 ± 2CF/PBS None 29 ± 2b-c 14 ± 3b-c 8 ± 3b'c 6 ± IeF/QS-21 None 50 ± 4 31 ± 5 16 ± 1 17 ± 6

aBALB/c mice were primed with 3 pg of F protein adsorbed to AlOH, admixed with QS-21, or prepared in PBSalone. The indicated vaccines were co-formulated with rIL-12 after overnight adsorption to AlOH. The data are themean percent lymphocytes, CD3+, CD4+, or CD8+ T cells 5 days after challenge. Statistical differences (p <0.05)were determined by Student's r-test after log transformation. There were 5 mice per group.

bp < 0.05 vs. F/AIOH plus rIL-12.cp< 0.05 vs. F/QS-21.

66

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

sis of immunological analyses of sera (39) and peripheral blood mononuclear cells (32) from the recipients,the prevailing hypothesis is that immunization with the FI-RSV vaccine primed for immune responsesskewed toward a type 2 phenotype when compared to natural infection. Thus, following seasonal exposureto RSV, type 2 immune responses were dominant and resulted in atypical disease. Studies in rodents alsosupport this hypothesis (1,2,12,13,17,49). Immunization with a facsimile FI-RSV vaccine elicits elevatedhumoral immune responses composed primarily of nonneutralizing antibodies, diminished cytolytic T-cellactivities, and a bias for atypical pulmonary inflammatory responses after challenge.

To develop subunit vaccines composed of highly purified natural or recombinant RSV antigens, novelformulations are required. Currently, the only adjuvants approved for humans are the aluminum-based salts(51). Natural F protein adsorbed to AlOH adjuvant was shown to be safe and immunogenic in seropositivehuman volunteers after i.m. vaccination (50). In addition, F/AIOH induced protective immunity in naiveBALB/c mice (20). Overt disease, as evinced by loss of body weight, was neither observed nor associatedwith the inhibition of virus replication. However, when compared to experimental infection the protectiveimmune responses elicited after vaccination with F/AIOH were characterized by low to moderate levels ofantigen-dependent killer cell and serum-neutralizing antibody activities, elevated serum anti-protein IgGiantibody titers, and atypical pulmonary eosinophilia after challenge (20,22). Recently we described our ef-forts to alter the quality of immune responses to natural F protein by co-formulation with the saponin, QS-21 (21). In contrast to F/AIOH, i.m. vaccination with F/QS-21 generated significantly elevated serum anti-F protein IgG2a and complement-assisted neutralizing antibody titers. More importantly, the morphology ofcellular constituents of the pulmonary inflammatory infiltrates was modified. Antigen-dependent killer cellactivity mediated by CD8+ T cells was enhanced and eosinophilia was significantly diminished in the pul-monary tissues after challenge. Even more striking were the data from a subsequent report (22), whichdemonstrated that serum-neutralizing antibody titers could be significantly enhanced if highly purified nat-ural G protein was added to F/QS-21. However, the combination vaccine biased naive recipients on sub-sequent challenge for pulmonary eosinophilia. Thus, although QS-21 clearly augmented the ability of nat-ural F protein to elicit type 1 immune responses and associated local antigen-dependent CD8+ killer cellactivities, type 2 immune responses to natural G protein were not diminished.

The purpose of the studies presented herein was to test the potential of rIL-12 to serve as an adjuvantfor subunit vaccines against RSV. Hence, our goal was to determine if formulation with rIL-12 could in-crease type 1 immune responses elicited by natural RSV antigens after i.m. administration. To that end, IgGantibody subclasses (16) and cytokines (35,41) associated with type 1 immune responses were determinedin sera. In addition, infectious virus, the phenotype of T cells, relative percentages of IL-5-dependenteosinophils (9), and the presence of antigen-dependent MHC class I-restricted killer cells were monitoredin the pulmonary tissues of vaccinated mice after challenge. When viewed in sum, the data strongly sug-gest that rIL-12 increased the capacity of F/AIOH to induce type 1 immune responses in naive BALB/cmice. The results further demonstrate that rIL-12 can be used locally as adjuvant in a single or two injec-tion protocol. Salient for future vaccine development, the results demonstrate that maximum effect was ob-served when rIL-2 was administered together with antigen in a single injection protocol. No increases inanti-F protein IgG2a and neutralizing antibody titers were observed when rIL-12 was administered at a sitedistal from F/AIOH.

The results presented herein are similar to previously published reports where naive BALB/c mice were

co-administered rIL-12 with either G protein expressed in recombinant vaccinia virus (25) or FI-RSV vac-

cine (47,48). However, the present results differ significantly from the earlier reports in several aspects.First, instead of formalin-treated whole virus, or recombinantly expressed protein, highly purified naturalF or G proteins are used as subunit vaccine antigens. Second, in two separate studies noteworthy increasesin antigen-dependent killer cell activity were observed ex vivo, in the lungs after challenge. Third, dramaticincreases in complement-assisted neutralizing antibody titers were detected in the sera, prior to challenge,following primary and secondary vaccination. The presence of antigen-dependent killer cells ex vivo in thelungs after challenge, and serum-neutralizing antibody titers before challenge, were not reported followingco-administration of rIL-12 with FI-RSV vaccine (47,48). Fourth, co-formulation of rIL-12 with F/AIOHor G/AIOH did not diminish pulmonary eosinophilia after challenge. This is in contrast to the reported two-

67

HANCOCK ET AL.

to three-fold reduction in pulmonary eosinophilia associated with G protein expressed in recombinant vac-

cinia virus after multiple (five) intraperitoneal injections of rIL-12 (25). Finally, co-administration of rlL-12 with F/AIOH (or G/AIOH) did not consistently result in a reduction in serum IgGi antibody titers.

In agreement with others (27), the results also suggest that the greatest potential of rIL-12 for subunitvaccines against RSV is after adsorption to alum adjuvant. Adsorption of 10 ng of rIL-12 to alum enableda 100-fold reduction (3.0 to 0.03 pg) in the amount of F protein required to achieve complete protectionof the lungs 4 days after challenge. When prepared in PBS alone, significant reductions in infectious viruswere observed only after 100 or 1,000 ng of rIL-12 was added to 3 pg of F protein. Moreover, clearanceof infectious virus was not complete and the reduction was only 10-fold. In comparison, primary immu-nization with 3 ¿ng of F protein-adsorbed AlOH, without rIL-12, was able to accelerate the clearance ofvirus from the pulmonary tissues to background levels 4 days after challenge. In contrast to a previous re-

port (27), there was no evidence that alum extended the persistence of IFN-y in the sera. Indeed, similaramounts of IFN-y were observed in sera following i.m. injection of rIL-12 in PBS alone, with or withoutantigen. However, adsorption to alum did appear to prolong the biological half-life of rIL-12. Why increasedpersistence of IL-12 did not lead to an extended presence of IFN-y in the sera is unknown, but may be ex-

plained by the fact that F/AIOH and rIL-12 were co-administered i.m. In the earlier report (27), the sub-cutaneous route was used.

Important for the future development of subunit vaccines composed of highly purified antigens of RSVare the data that suggest that co-administration with rIL-12 does not result in the complete dominance oftype 1 immune responses. Despite the transformation of systemic humoral immune responses, the presencerIL-12 did not alleviate the pulmonary eosinophilia associated with prior immunization of naive mice withF/AIOH (or G/AIOH). These results were unpredicted in view of published data (27), which demonstratedthat adsorption of rIL-12 to AlOH augmented type 1 cytokine responses and diminished the in vitro secre-

tion of IL-5 after stimulation with HIV gpl20. Moreover, the data presented herein indicated that co-ad-ministration of rIL-12 with F/AIOH resulted in an increased amount of ffN-y and prolonged presence ofIL-12 in the sera, respectively. In addition, augmented antigen-dependent killer cell activity was observed.Several scenarios may explain the unexpected results observed in the lungs after challenge of mice vacci-nated with F/AIOH plus rIL-12. Phenotypic analyses of lymphoid populations infiltrating the lungs afterchallenge suggest that the primary effect of rIL-12 on immune responses elicited by F/AIOH is the expan-sion of IFN-y-secreting CD4+ T cells, relative to the CD8+ T-cell compartment. That is, CD8+ T cells are

not significantly increased when compared to vaccination with F/AIOH alone. Thus, one explanation is thatco-administration of rIL-12 does not generate sufficient numbers of IFN-y-secreting CD8+ T cells to reg-ulate the eosinophilic response. In support of this hypothesis, the lungs of mice immunized with F/QS-21after challenge contain significantly increased CD8+ T cells when compared to mice vaccinated with F/AIOHalone. The pulmonary CD4:CD8 T-cell ratios upon challenge of mice vaccinated with F/AIOH plus rIL-12or F/QS-21 were 1.7 and 0.9, respectively. Pulmonary eosinophilia is diminished after formulation with QS-21. Thus, the data are in agreement with previous reports (24,45,48), which suggest that IFN-y-secretingCD8+ T cells are extremely important for regulating the pulmonary eosinophilia associated with RSV in-fection.

The failure to diminish pulmonary eosinophilia may also be explained by observations that suggest thatrIL-12 has little impact on the capacity of CD4+CD3 CD1 lc~ dendritic cells (DC2) to induce type 2 helperT cells in an allogeneic mixed lymphocyte reaction (40). In the presence of rIL-12, the capacity of CD2 toinduce IL-4-secreting T cells was not diminished. Instead, the presence of rIL-12 was associated with an

increase in T cells secreting both IL-4 and IFN-y. Thus, although rIL-12 increases the secretion of IFN-ysufficient to govern systemic humoral immune responses, this may occur without a diminution in type 2responses that regulate pulmonary eosinophilia. Therefore, the dramatic increase in IgG2a antibody titersmay be an over-estimation of the capacity of rIL-12 to generate IFN-y-secreting T cells. It is possible thatcells other than T cells of the type 1 phenotype may secrete IFN-y. Natural killer cells (5,43) and B cells(14,28,33,53) may be significant sources of IFN-y after activation with IL-12. The data are also intriguingin view of a recent report, which suggests that IL-5-secreting CD8+ T cells play an important role in thepulmonary eosinophilia associated with acute RSV infection (42). Thus, it is plausible that co-administra-tion of rIL-12 is associated with the generation of IL-5-secreting QD8+ x cells. However, this hypothesis

68

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

must be reconciled with the pulmonary eosinophilia observed upon challenge of mice primed with natural(22) or a synthetic peptide (49) of G protein. In the latter scenario, the data suggest that IL-5-secretingCD4+ T cells play an essential role.

The conclusions from the cumulative data identify the benefits and potential limitations of rIL-12 as a

subunit vaccine adjuvant for RSV glycoproteins. The conclusion is that rIL-12 is able to transform the type2 immune responses that regulate the generation of systemic humoral immune responses, rIL-12 also ap-pears to augment the generation of MHC class I-restricted killer cells. These studies do not address, how-ever, the effect of rIL-12 on the generation of major histocompatibility complex (MHC) class II-restrictedkiller cells. For naive populations, the data imply that rIL-12 as tested, may not be adequate enough todown-modulate the type 2 immune responses necessary for the pulmonary eosinophilia elicited after prim-ing with F/AIOH. Thus, seropositive populations may be the preferred recipients of subunit vaccines forRSV containing rIL-12 as an immune response modifier. Finally, the data illustrate the complexities of im-munoregulatory circuits at systemic versus local levels, and, importantly, emphasize the key role that CD8+T cells play in controlling the nature of immune responses to RSV infection.

ACKNOWLEDGMENTS

The authors acknowledge the technical assistance of Dr. D.D. Duncan, K. McGuire, G. Maine, and K.Pryharski. In addition, the authors thank Dr. P.W. Tebbey for a critical review of the manuscript.

REFERENCES

1. Alwan, W.H., W.J. Kozlowska, and P.J.M. Openshaw. 1994. Distinct types of lung disease caused by functionalsubsets of antiviral T cells. J. Exp. Med. 179:81-89.

2. Alwan, W.H., F.J. Record, and P.J.M. Openshaw. 1993. Phenotypic and functional characterization of T cell linesspecific for individual respiratory syncytial virus proteins. J. Immunol. 150:5211-5218.

3. Brideau, R.J., R.R. Walters, M.A. Stier, and M.W. Wathen. 1989. Protection of cotton rats against human respira-tory syncytial virus by vaccination with a novel chimeric FG glycoprotein. J. Gen. Virol. 70:2637-2644.

4. Brunda, M.J. 1994. Interleukin-12. J. Leukocyte Biol. 55:280-288.

5. Chan S.H., B. Perussia, J.W. Gupta, M. Kobayashi, M. Pospisil, H.A. Young, S.F. Wolf, D. Young, S.C. Clark,and G. Trinchieri. 1991. Induction of interferon gamma production by natural killer cell stimulatory factor: char-acterization of the responder cells and synergy with other inducers. J. Exp. Med. 173:869-879.

6. Chanock, R.M., and R.H. Parrott. 1965. Acute respiratory disease in infancy and childhood: present understandingand prospects for prevention. Pediatrics 36:21-39.

7. Chanock, R.M., R.H. Parrott, M. Connors, P.L. Collins, and B.R. Murphy. 1992. Serious respiratory tract diseasecaused by respiratory syncytial virus: prospects for improved therapy and effective immunization. Pediatrics90:137-143.

8. Cherrie, A.H., K. Anderson, G.W. Wertz, and P.J.M Openshaw. 1992. Human cytotoxic T cells stimulated by anti-gen on dendritic cells recognize the N, SH, F, M, 22K, and lb proteins of respiratory syncytial virus. J. Virol.66:2102-2110.

9. Coffman, R.L., B.W.P. Seymour, S. Hudak, J. Jackson, and D. Rennick. 1989. Antibody to interleukin-5 inhibitshelminth-induced eosinophilia in mice. Science 245:308-310.

10. Colby, W.D., and G.H. Strejan. 1980. Immunological tolerance of the mouse IgE system: dissociation between Tcell tolerance and suppressor cell activity. Eur. J. Immunol. 10:602-608.

11. Colocho Zelaya, E.A., C. Orvell, and O. Strannegard. 1994. Eosinophil cationic protein in nasopharyngeal secre-

tions and serum of infants infected with respiratory syncytial virus. Pediatr. Allergy Immunol. 5:100-106.

12. Connors, M., N.A. Giese, A.B. Kulkarni, C.-Y. Firestone, H.C. Morse, and B.R. Murphy. 1994. Enhanced pul-monary histopathology induced by respiratory syncytial virus (RSV) challenge of formalin-inactivated RSV-im-munized BALB/c mice is abrogated by depletion of interleukin-4 (IL-4) and IL-10. J. Virol. 68:5321-5325.

69

HANCOCK ET AL.

13. Connors, M., A.B. Kulkarni, C-Y. Firestone, K.L. Holmes, H.C. Morse III, A.V. Sotnikov, and B.R. Murphy. 1992.Pulmonary histopathology induced by respiratory syncytial virus (RSV) challenge of formalin-inactivated RSV-immune BALB/c mice is abrogated by depletion of CD4+ T cells. J. Virol. 66:7444-7451.

14. Dubois, B., C. Massacrier, B. Vanbervliet, J. Fayette, F. Briere, J. Banchereau, and C. Caux. 1998. Critical role ofIL-12 in dendritic cell-induced differentiation of naive B lymphocytes. J. Immunol. 161:2223-2231.

15. Fernie, B.F., E.C. Ford, and J.L. Gerin. 1981. The development of BALB/c cells persistently infected with respi-ratory syncytial virus: presence of ribonucleoprotein on the cell surface (41129). Proc. Soc. Exp. Biol. Med.167:83-86.

16. Finkleman, R.D., J. Holmes, I.M. Katona, J.F.J. Urban, M.P. Beckmann, L.S. Park, K.A. Schooley, R.L. Coffman,T.R. Mosmann, and W.E. Paul. 1990. Lymphokine control of in vivo immunoglobulin isotype selection. Annu.Rev. Immunol. 8:303.

17. Graham, B.S., G.S. Henderson, Y-W Tang, X. Lu, K.M. Neuzil, and D.G. Colley. 1993. Priming immunizationdetermines T helper cytokine mRNA expression patterns in lungs of mice challenged with respiratory syncytialvirus. J. Immunol. 151:2032-2040.

18. Hall, C.B., E.E. Walsh, CE. Long, and K.C Schnabel. 1991. Immunity to and frequency of reinfection with res-

piratory syncytial virus. J. Infect. Dis. 163:693-698.

19. Han, L.L., J.P. Alexander, and L.J. Anderson. 1999. Respiratory syncytial virus pneumonia among the elderly: an

assessment of disease burden. J. Infect. Dis. 179:25-30.

20. Hancock, G.E., DJ. Hahn, D.J. Speelman, S.W. Hildreth, S. Pillai, and K. McQueen. 1994. The pulmonary im-mune response of BALB/c mice vaccinated with the fusion protein of respiratory syncytial virus. Vaccine12:267-274.

Hancock, G.E., D.J. Speelman, P.J. Frenchick, M.M. Mineo-Kuhn, R.B. Baggs, and D.J. Hahn. 1995. Formulationof the purified fusion protein of respiratory syncytial virus with the saponin QS-21 induces protective immune re-

sponses in BALB/c mice that are similar to those generated by experimental infection. Vaccine 13:391^100.

Hancock, G.E., D.J. Speelman, K. Heers, E. Bortell, J. Smith, and C. Cosco. 1996. Generation of atypical pul-monary inflammatory responses in BALB/c mice after immunization with the native attachment (G) glycoproteinor respiratory syncytial virus. J. Virol. 70:7783.

Hsieh, S., S.E. Macatonia, C.S. Tripp, S.F. Wolf, A. O'Garra, and K.M. Murphy. 1993. Development of Thl (CD4+T cells through IL-12 produced by Lwfiria-induced macrophages. Science 260:547.

Hussell, T., C.J. Baldwin, A. O'Garra, and P.J.M. Openshaw. 1997. CD8+ T cells control Th2-driven pathologyduring pulmonary respiratory syncytial virus infection. Eur. J. Immunol. 27:3341-3349.

Hussell T, U. Khan, and P. Openshaw. 1997. IL-12 treatment attenuates T helper cell type 2 and B cell responsesbut does not improve vaccine-enhanced lung illness. J. Immunol. 159:328.

Issacs, D., N.E. MacDonald, C.R.M. Bangham, A.J. McMichael, P.G Higgins, and D.A.J Tyrrell. 1991. The spe-cific cytotoxic T-cell response of adult volunteers to infection with respiratory syncytial virus. Immunol. Infect.Dis. 1:5-12.

27. Jankovic, D., P. Casper, M. Zweig, M. Garcia-Moll, S.D. Showalter, F.R. Vogel, and A. Sher. 1997. Adsorptionto aluminum hydroxide promotes the activity of IL-12 as an adjuvant for antibody as well as type 1 cytokine re-

sponses to HIV-1 gpl20. J. Immunol. 159:2409.

28. Jelinek, D.F., and J.K. Braaten. 1995. Role of IL-12 in human B lymphocyte proliferation and differentiation. J.Immunol. 154:1606-1613.

29. Kapikian, A.Z., R.H. Mitchell, R.M. Chanock, R.A. Shevdoff, and CE. Stewart. 1968. An epidemiological studyof altered clinical reactivity to respiratory syncytial (RS) virus infection in children vaccinated with an inactivatedRS virus vaccine. Am. J. Epidemiol. 89:405-413.

30. Kakuk, T.J., K. Soike, R.J. Brideau, R.M. Zaya, S.L. Cole, J-Y. Zhang, E.D. Roberts, P.A. Wells, and M.W. Wa-then. 1993. A human respiratory syncytial virus (RSV) primate model of enhanced pulmonary pathology inducedwith a formalin-inactivated RSV vaccine but not a recombinant FG subunit vaccine. J. Infect. Dis. 167:553-561.

31. Kim, H.W., J.G Canchóla, CD. Brandt, G. Pyles, R.M. Chanock, K. Jensen, and R.H. Parrott. 1969. Respiratory

70

rIL-12 AND SUBUNIT VACCINES FOR RESPIRATORY SYNCYTIAL VIRUS

syncytial virus disease in infants despite prior administration of antigenic inactivated vaccine. Am. J. Epidemiol.89:422-^134.

32. Kim, H.W., S.L. Leikin, J. Arrobio, CD. Brandt, R.M. Chanock, and R.H. Parrott. 1976. Cell-mediated immunityto respiratory syncytial virus induced by inactivated vaccine or infection. Pediatr. Res. 10:75-78.

33. Li L., D. Young, S.F. Wolf, and Y.S. Choi. 1996. Interleukin-12 stimulates B cell growth by inducing IFN-gamma.Cell. Immunol. 168:133-140.

34. Mahon, B.P., M.S. Ryan, F. Griffin, and K.H.G. Mills. 1996. Interleukin-12 is produced by macrophages in re-

sponse to live or killed Bordetella pertussis and enhances the efficacy of an acellular pertussis vaccine by pro-moting induction of Thl cells. Infect. Immun. 64:5295-5301.

35. Mosmann, T.R., J.H. Schumacher, N.F. Street, R. Budd, A. O'Garra, T.A.T. Fong, M.W. Bond, K.M.W. Moore,A. Sher, and D.F. Fiorentino. 1991. Diversity of cytokine synthesis and function of mouse CD4+ T cells. Immunol.Rev. 123:209-229.

36. Muelenaer, P.M., F.W. Henderson, V.G. Hemming, E.E. Walsh, L.J. Anderson, G.A. Prince, and B.R. Murphy.1991. Group-specific serum antibody responses in children with primary and recurrent respiratory syncytial virusinfections. J. Infect. Dis. 164:15-21.

37. Murphy, B.R., D.W. Ailing, M.H. Snyder, E.E. Walsh, G.A. Prince, R.M. Chanock, V.G. Hemming, W.J. Rod-riguez, H.W. Kim, B.S. Graham, and P.F. Wright. 1986. Effect of age and preexisting antibody on serum antibodyresponse of infants and children to the F and G glycoproteins during respiratory syncytial virus infection. J. Clin.Microbiol. 24:894-898.

38. Murphy, B.R., S.L. Hall, A.B. Kulkarni, J.E. Crowe, P.L. Collins, M. Connors, R.A. Karron, and R.M Chanock.1994. An update on approaches to the development of respiratory syncytial virus (RSV) and parainfluenza virustype 3 (PIV3) vaccines. Virus Res. 32:13-36.

39. Murphy, B.R., G.A. Prince, E.E. Walsh, H.W. Kim, R.H. Parrott, V.G. Hemming, W.J. Rodriguez, and R.M.Chanock. 1986. Dissociation between serum neutralizing and glycoprotein antibody responses of infants and chil-dren who received inactivated respiratory syncytial virus vaccine. J. Clin. Microbiol. 24:197-202.

40. Rissoan, M.-C, V. Soumelis, N. Kadowaki, G. Grouard, F. Briere, R. de Waal Malefyt, and Y.-J. Liu. 1999. Rec-iprocal control of T helper cell and dendritic cell differentiation. Science 283:1183-1186.

41. Sad, S., R. Marcotte, and T.R. Mosmann. 1995. Cytokine-induced differentiation of precursor mouse CD8+ T cellsinto cytotoxic CD8+ T cells secreting Thl or Th2 cytokines. Immunity 2:271-279.

42. Schwarze, J., G. Cieslewicz, A. Joetham, T. Ikemura, E. Hamelmann, and E.W. Gelfand. 1999. CD8 T cells are

essential in the development of respiratory syncytial virus-induced lung eosinophilia and airway hyperresponsive-ness. J. Immunol. 162:4207^4211.

43. Scott P., and G. Trinchieri. 1997. IL-12 as an adjuvant for cell-mediated immunity. Sem. Immunol. 9:285.

44. Sigurs, N., R. Bjarnason, F. Sigurbergsson, B. Kjellman, and B. Bjorksten. 1995. Asthma and immunoglobulin Eantibodies after respiratory syncytial virus bronchiolitis: a prospective cohort study with matched controls. Pedi-atrics 95:500.

45. Srikiatkhachorn, A., and T.J. Braciale. 1997. Virus-specific CD8+ T lymphocytes downregulate T helper cell type2 cytokine secretion and pulmonary eosinophilia during experimental murine respiratory syncytial virus infection.J. Exp. Med. 186:421-132.

46. Stott, E.J., and G Taylor. 1985. Respiratory syncytial virus, brief review. Arch. Virol. 84:1-52.

47. Tang, Y.W., and B.S. Graham. 1995. Interleukin-12 treatment during immunization elicits a T helper cell type 1-like immune response in mice challenged with respiratory syncytial virus and improves vaccine immunogenicity.J. Infect. Dis. 172:734.

48. Tang, Y.W., and B.S. Graham. 1997. T cell source of type 1 cytokines determines illness patterns in respiratorysyncytial virus-infected mice. J. Clin. Invest. 99:2183-2191.

49. Tebbey, P.W., M. Hagen, and G.E. Hancock. 1998. Atypical pulmonary eosinophilia is mediated by a specificamino acid sequence of the attachment (G) protein of respiratory syncytial virus. J. Exp. Med. 188:1967-1972.

50. Tristram, D.A., R.C. Welliver, C.K. Mohar, D.A. Hogerman, S.W. Hildreth, and P. Paradiso. 1993. Immunogenicity

71

HANCOCK ET AL.

and safety of a respiratory syncytial virus subunit vaccine in seropositive children 18-36 months old. J. Infect. Dis.167:191-195.

51. Warren, H.S., and Chedid, L.A. 1988. Future prospects for vaccine adjuvants. Crit. Rev. Immunol. 141, 1720-1727.

52. Wynn T.A., D. Jankovic, S. Hieny, A.W. Cheever, and A. Sher. 1995. IL-12 enhances vaccine-induced immunityto Schistosoma mansoni in mice and decreases T helper 2 cytokine expression, IgE production, and tissueeosinophilia. J. Immunol. 154:4701^709.

53. Yoshimoto, T., H. Okamura, Y.I. Tagawa, Y. Iwakura, and K. Nakanishi. 1997. Interleukin 18 together with in-terleukin 12 inhibits IgE production by induction of interferon-gamma production from activated B cells. Proc.Nati. Acad. Sei. USA 94:3948-3953.

Address reprint requests to:Dr. Gerald E. Hancock

Department of Immunology ResearchWyeth-Lederle Vaccines

211 Bailey RoadWest Henrietta, NY 14586-9728

E-mail: hancocg@war.wyeth.comReceived for publication July 30, 1999; accepted September 2, 1999.

72