Effects of different adjuvants on rotavirus antibody responses and protection in mice following intramuscular immunization with inactivated rotavirus

Vaccine 17 (1999) 1573±1580

E€ects of di€erent adjuvants on rotavirus antibody responses and protection in mice following intramuscular immunization with inactivated rotavirus Monica M. McNeal, Mary N. Rae, Richard L. Ward * Division of Infectious Diseases, Children's Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, OH 45229-3039, USA Received 23 July 1998; received in revised form 9 September 1998; accepted 10 September 1998

Abstract I.m. immunization of mice with inactivated rotavirus particles protects against subsequent infection. To optimize protection, the e€ects of di€erent adjuvants (QS-21, QS-7, QUIL A, PCPP and RAS) with potential for human use were compared. Twenty-eight days after i.m. immunization with 20 mg of puri®ed, UV/psoralen-inactivated murine rotavirus (EDIM), either with or without adjuvant, BALB/c mice were orally challenged with live EDIM and virus shedding was measured. All ®ve adjuvants stimulated large (P < 0.001) increases in rotavirus antibody, but signi®cant di€erences were found between adjuvants. The order of rotavirus IgG responses, i.e. no adjuvant < RAS < QS-7 < Quil A < QS-21 < PCPP, was the same as the order of protection except that QS-21 and PCPP were reversed. These results establish the importance of adjuvants during i.m. immunization with rotavirus and identify those with the greatest potential. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Rotavirus; Intramuscular immunization; Adjuvants

1. Introduction Rotavirus infection is the primary cause of severe infantile gastroenteritis and results in nearly one million deaths worldwide each year. All rotavirus vaccine candidates evaluated to date in clinical trials have been attenuated live viruses that are delivered orally to mimic the route of infection and resulting protection associated with natural rotavirus infections. Because these vaccines have provided only partial protection against subsequent rotavirus disease [1±9], development of more e€ective vaccines is being actively pursued. Until recently, parenteral immunization was not given serious consideration as a method to induce the intestinal immunity believed to be required to protect against rotavirus disease. However, using the adult mouse model developed to study active immunity against rotavirus [10], we reported that parenteral (i.e. intraperitoneal) immunization with live or UV-inacti* Corresponding author. Tel.: +1-513-636-7628; fax: +1-513-6367682; e-mail: [email protected]. 0264-410X/99/$19.00 # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 4 - 4 1 0 X ( 9 8 ) 0 0 3 5 9 - 4

vated rotaviruses stimulated protection against infection with the epizootic diarrhea of infant mice (EDIM) strain of murine rotavirus [11]. Subsequently, other investigators reported that intramuscular (i.m.) immunization of rabbits with live or formalin-inactivated simian rotavirus provided protection against infection with a rabbit rotavirus [12]. More recently, Con et al. [13] also observed that i.m. immunization with live or psoralen-inactivated EDIM provided signi®cant protection against subsequent EDIM infection. Further studies in our laboratory with live and psoralen-inactivated EDIM administered intramuscularly revealed that inclusion of the saponin adjuvant QS-21 greatly enhanced rotavirus antibody responses and stimulated signi®cantly greater protection [14]. In contrast to these highly positive results, Yuan et al. [15] observed that piglets immunized intramuscularly with BEI (binary ethylenimine)-inactivated human rotavirus (strain Wa) in the presence of incomplete Freund's adjuvant were not signi®cantly protected when subsequently challenged with virulent Wa virus. In all these studies, attempts were made to correlate protection with serum and intestinal rotavirus anti-

1574

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

body titers or with the quantities of rotavirus antibody secreting cells. In all cases, high titers of serum rotavirus IgG were generated but the presence of intestinal rotavirus IgG or IgA or rotavirus antibody-secreting cells were variable. For example, substantial protection was stimulated by either i.p. or i.m. immunization of mice with inactivated EDIM under conditions where intestinal rotavirus IgA or rotavirus IgA secreting cells were not detectable [11, 13]. However, under conditions where i.m. immunization with EDIM did result in measurable intestinal rotavirus IgA, i.e. when QS-21 was included [14] or immunization was performed with live, wild type EDIM [13], protection was substantially increased. The primary goal for these types of studies is to identify conditions that give maximum immune responses and protection in animal models which can potentially be used directly in humans. Intramuscular immunization with inactivated rotavirus in the presence of adjuvants o€ers this possibility. Because QS21 stimulated much greater immune responses and protection than without adjuvant [14] and is being evaluated in human trials with other vaccines [16], it may be the adjuvant of choice. On the other hand, other adjuvants may be more e€ective. The purpose of this report was to compare the immunogenic and protective e€ects of QS-21 as an adjuvant for i.m. immunization with rotavirus in the adult mouse model to that of other selected adjuvants, all but one of which (i.e. Quil A) appear to have potential for human use. 2. Materials and methods 2.1. Mice Female BALB/c mice were purchased from Harlan± Sprague±Dawley (Indianapolis, IN) and used at six weeks of age. Animals contained no rotavirus antibody as determined by ELISA. Microisolation cages were used to house the mice and eight animals were included in each group. All procedures were conducted in accordance with protocols approved by Children's Hospital Research Foundation Institutional Animal Care and Use Committee. 2.2. Virus All experiments were conducted with the EDIM strain of murine rotavirus classi®ed as G3 (P16) which had been adapted to grow in cell culture by passage in MA104 cells [10], a fetal rhesus monkey kidney cell line. A preparation of the 32nd passage was puri®ed by CsCl gradient centrifugation [11], and the fully encapsidated, triple-layered infectious particles were inactivated by UV/psoralen treatment [14]. This treat-

ment reduced the infectivity of the virus preparation from 8.4  108 focus forming units (FFU)/ml to below the limit of detection (i.e. 102 FFU/ml) following 20 min of irradiation, but the virus was irradiated for 40 min in this study to eliminate any residual infectivity. These inactivated particles (20 mg) were used for i.m. immunization of mice. A preparation of the ninth cell culture passage of EDIM containing 2  106 FFU/ml was used to orally challenge the immunized mice. 2.3. Adjuvants Five di€erent adjuvants were used in this study. Two are commercially available and were purchased from the manufacturers. These included the Ribi Adjuvant System (RAS) and Quil A. RAS (RIBI ImmunoChem Research, Hamilton, MT) contains monophosphoryl lipid A (MPL) and synthetic trehalose dicorynomycolate incorporated into squalene and Tween-80. The MPL component of this adjuvant has been used in both animal and human studies and found to induce cellular and humoral immune responses to a variety of antigens [17]. Quil A was originally isolated from the bark of the South American tree Quillaja saponaria Molina. It was prepared by Superfos Specialty Chemicals a/s as a sterile ®ltered solution which was lyophilized and distributed by Accurate Chemical and Scienti®c Co. (Westbury, NY). It has been used in multiple animal studies and is known to enhance antibody titers [16]. Both RAS and Quil A were reconstituted and used according to manufacturer's instructions. QS-7 and QS-21 are puri®ed saponins contained in Quil A [16]. Both have been found to enhance antibody and cell mediated immune responses, but only QS-21 has been incorporated into human trials [18]. QS-7 was provided by Aquila Biopharmaceuticals (Worcester, MA) and reconstituted in phosphate bu€ered saline (pH 7.0) to 1 mg/ml. It was mixed with rotavirus antigen prior to inoculation and each dose contained 20 mg. QS-21, manufactured by Aquila Biopharmaceuticals, was provided by Wyeth-Lederle Vaccines and Pediatrics (Pearl River, NY) and reconstituted in PBS (pH 6.0), mixed with antigen and used at a concentration of 20 mg/dose. Poly (di(carboxylatophenoxy)) phosphazene (PCPP) has been found to augment antibody responses to a variety of antigens [19]. This adjuvant has also been evaluated in human clinical studies. It was provided by Virus Research Institute (Cambridge, MA) as a 0.4% (WIV) solution which was diluted 4-fold with antigen. Each dose contained 100 mg of PCPP. 2.4. Study design Because i.m. inoculation with two doses of inactivated EDIM in the presence of QS-21 had been found

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

to provide essentially complete protection against shedding following subsequent EDIM challenge, only one dose was administered in this study. This was anticipated to provide the opportunity to measure both increased and decreased shedding in immunized mice relative to that found when QS-21 was included as the adjuvant. Anesthetized mice were immunized by injection of 20 mg of UV/psoralen inactivated, puri®ed EDIM particles into the thigh muscle of the hind leg. When injections were made in the presence of either of the ®ve adjuvants, the antigen was mixed with the appropriate volume of adjuvant just prior to inoculation. The ®nal volume injected was 100 ml/mouse. Mice were orally challenged by swallowing 4  104 FFU of EDIM four weeks after immunization using 20 ml of the EDIM lysate (ninth passage). Stool specimens were collected daily for seven days after challenge and analyzed for rotavirus antigen. Blood specimens (100±200 ml) were collected by retro-orbital capillary plexus puncture on the day of immunization (day 0), on the day before challenge (day 28), and 21 days after challenge. Stool specimens were also collected at these time points. Both the stool and blood specimens were used to measure rotavirus IgG and IgA titers. Serum neutralizing antibody titers to EDIM were also determined. 2.5. Detection of rotavirus antigen in stool Two stool pellets collected into 0.5 ml of Earle's balanced salt solution (EBSS) on each day following EDIM challenge were stored frozen (ÿ208C). To test for rotavirus antigen, the specimens were thawed, homogenized, and analyzed by ELISA as previously described [20]. The quantity of virus shed was determined by net A490 values (i.e., the average of two rotavirus negative antibody-coated wells subtracted from the average of two rotavirus antibody-coated wells). A490 values of r3.0 were assigned values of 4.0 for quantitation purposes. To determine the amount of rotavirus shed per mouse within a group, the average A490 values above background were determined for every mouse on each of the 7 days after EDIM challenge. 2.6. Determination of rotavirus antibody titers Serum and stool rotavirus IgA and IgG were measured by ELISA as previously described [21]. Brie¯y, EIA plates (Corning Costar Co., Cambridge, MA) were coated overnight at 48C with anti-rotavirus rabbit IgG. After washing with phosphate bu€ered saline plus 0.05% Tween 20, EDIM viral lysate and mock-infected cell lysate were each added to duplicate

1575

wells. Serial two-fold dilutions of pooled sera from EDIM infected mice assigned concentrations of 160,000 or 10,000 units/ml of rotavirus IgG or IgA, respectively, were added to duplicate wells coated with either EDIM-infected or uninfected MA104 cell lysates to generate a standard curve. Serial 10-fold dilutions of mouse sera to be tested were also added to duplicate wells of each lysate. This was followed by sequential addition of biotin-conjugated goat anti-mouse IgG or IgA (Sigma Chemical Co., St. Louis, MO), peroxidase-conjugated avidin-biotin (Vector Laboratories, Burlingame, CA) and o-phenylene-diamine substrate (Sigma Chemical Co.). Color development was stopped after 15 min with 1 M H2SO4 and the A490 was measured. Titers of rotavirus IgG or IgA, expressed as units/ml, were determined from the standard curve generated by subtraction of the average A490 values of the duplicate cell lysate wells from the average of the wells coated with EDIM lysate. This same assay was performed to measure serum rotavirus IgG1 and IgG2a titers except that the biotin-conjugated goat anti-mouse antibody (Caltag Laboratories, Burlingame, CA) was directed against these speci®c subclasses of antibody. In this case, however, the concentration of rotavirus antibody was determined from a standard curve that measured rotavirus antibody in nanograms (ng) rather than units/ml. To generate this curve, the EIA plates were coated overnight with either goat anti-mouse IgG or normal goat serum. After washing the plates, serial two-fold dilutions of puri®ed mouse IgG1 or IgG2a (Caltag Laboratories) were added to duplicate sets of coated wells. The remaining steps were identical to those used to measure rotavirus speci®c IgG1 or IgG2a. Concentrations of the rotavirus antibodies were determined from the standard curve and expressed as ng/ml. No cross reactivity between the two IgG subclasses was found using the speci®ed reagents. For determination of stool rotavirus IgA, two stool pellets were collected into 0.5 ml of EBSS, homogenized, and centrifuged (1500g, 5 min). Stool rotavirus IgA and IgG were then measured by the method described above. Neutralizing antibody to EDIM was determined by using an antigen reduction ELISA assay as described previously [22].

2.7. Statistical analyses Di€erences in the mean quantity of rotavirus antigen shed per mouse between groups, di€erences in geometric mean titers of rotavirus antibody between groups of mice and di€erences in serum rotavirus IgG1/IgG2a ratios between groups were all evaluated by Student's t-test.

1576

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

Table 1 Rotavirus antibody titers following intramuscular immunization with inactivated EDIM and di€erent adjuvantsa Rotavirus antibody Geometric mean titers2standard error. Adjuvant used during immunization

Serum IgG Serum IgA Serum neutralizing Stool IgG Stool IgA

unimmunized none

RAS

QS-7

Quil A

QS-21

PCPP

<100b <100c 24g23 <2c <5i

100,000d219,300 254e251 35h28 5e21 6i21

272,400e239,000 161f 2 24 108e240 6e21 7i21

272,700e252,200 400 275 59e219 8e22 22j212

339,100e250,300 4722 108 66e231 13e26 77232

1,281,1002124,100 679 250 381 2169 46 26 10j25

12,600c22,000 <100c 62e220 <2c <5i

a Mice (8/group) were administered a single i.m. inoculation with 20 mg of inactivated EDIM either with or without adjuvant and blood and stool specimens collected 27 days after immunization, the day before oral challenge with live EDIM, were analyzed for rotavirus antibodies. The limit of detection was 100, 100, 2 and 5 units/ml for serum rotavirus IgG, serum IgA, stool IgG and stool IgA, respectively, and these values were used to calculate GMTs when rotavirus antibody titers were below the limit of detection. bSigni®cantly (P < 0.001) less than all immunized groups. cSigni®cantly (P R0.004) less than the adjuvant groups. dSigni®cantly (P < 0.001) less than other adjuvant groups. eSigni®cantly (PR 0.05) less than PCPP group only. fSigni®cantly (P R0.001) less than Quil A, QS-21 and PCPP groups only. gSigni®cantly (PR 0.03) less than all but the RAS adjuvant group. hSigni®cantly (P R0.01) less than QS-7 and PCPP groups only. iSigni®cantly (P R0.01) less than Quil A and QS-21 groups only. jSigni®cantly (P R0.05) less than QS-21 group only.

3. Results 3.1. Increased rotavirus antibody responses stimulated by the di€erent adjuvants Intramuscular immunization with one dose of psoralen-inactivated, triple-layered EDIM particles stimulated moderate serum rotavirus IgG responses and signi®cant (P = 0.008) serum neutralizing antibody responses but no detectable serum rotavirus IgA or stool rotavirus IgG or IgA (Table 1). Inclusion of any of the ®ve adjuvants to be compared in this study greatly (P < 0.001) increased serum rotavirus IgG responses, but only PCPP stimulated a signi®cant (P = 0.02) increase in serum neutralizing antibody to EDIM relative to that found in the absence of adjuvant. All adjuvants stimulated small but measurable quantities of serum rotavirus IgA and stool rotavirus IgG and IgA. In all cases except the stool rotavirus IgA titers stimulated by the RAS, QS-7 and PCPP adjuvants, these responses were signi®cant (P R 0.01). The IgA responses were, however, much lower than those stimulated by 21 days following oral challenge of the unimmunized mice with live EDIM during the course of this same study. These geometric mean titers (GMTs) were 6976 and 2002 units/ml for serum and stool rotavirus IgA, respectively. When the increased antibody responses stimulated by the adjuvants were compared, signi®cant di€erences were found but none were strikingly superior or inferior to the others. The RAS and QS-7 adjuvants almost always produced the smallest increases and PCPP and QS-21 stimulated the largest responses. Exceptions were that QS-7 stimulated the second largest titer of neutralizing antibody and Quil A stimu-

lated more stool rotavirus IgA than PCPP. Even though di€erences in antibody titers stimulated by these ®ve adjuvants were not dramatic, they were highly signi®cant in many instances (Table 1). One possible reason for di€erences in antibody responses could be due to modi®cation of the relative quantities of T-helper cell subsets stimulated by these immunizations. TH1 responses are typically associated with the release of speci®c cytokines such as interferon-g and the stimulation of speci®c classes of antibody such as IgG2a [23]. Conversely, TH2 responses typically result in IL-4 and IL-5 release and stimulation of IgG1. Therefore, the ratio of IgG1/IgG2a should re¯ect the predominance of each helper cell subset. This ratio was found to be slightly (P = 0.02± 0.07) higher following immunization with inactivated EDIM without adjuvant but no signi®cant di€erences in ratios were found between groups of mice immunized with the adjuvants (Table 2). Thus, di€erences in antibody responses due to the use of di€erent adjuTable 2 IgG1/IgG2a ratios of rotavirus antibodies stimulated by immunization with di€erent adjuvantsa Adjuvant used

Average ratio of IgG1/IgG2a

None RAS QS-7 Quil A QS-21 PCPP

0.28b 0.12 0.14 0.10 0.11 0.10

a

Methods of immunization and analysis of specimens are described in the legend to Fig. 1 and Section 2. bSigni®cance of di€erence compared to groups with adjuvant varied between P = 0.02 to P = 0.07.

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

1577

Fig. 1. Shedding of rotavirus following oral EDIM challenge of either unimmunized or EDIM-immunized mice without adjuvant or with PCPP. Shedding is expressed as the average A490 for each group of eight mice in each of the 7 days after challenge as determined by ELISA. Bars represent standard error.

vants could not be correlated with the relative quantities of T-cell subsets induced in response to immunization as measured by IgG subclass ratios. 3.2. Increased protection against rotavirus stimulated by the di€erent adjuvants Twenty-eight days after i.m. immunization with inactivated EDIM, mice were orally challenged with live EDIM. Rotavirus shedding during the seven days after challenge, using optical density measurements as determined by an ELISA, were compared for all

groups of mice. The average shedding/day for three groups (i.e., unimmunized, EDIM immunized and EDIM immunized with PCPP) clearly demonstrate that inactivated EDIM alone provides some protection, but that inclusion of adjuvant greatly enhances this protection (Fig. 1). When the average shedding/ mouse/day was determined for each group, highly signi®cant di€erences were found (Table 3). Immunization with inactivated EDIM alone provided signi®cant (P < 0.001) protection, but inclusion of all adjuvants except RAS caused further signi®cant (P R 0.001) reduction in rotavirus shedding. Although

Table 3 Shedding of rotavirus antigen following EDIM challenge of immunized micea Immunization group

Shedding (average OD/mouse/day)

Statistical analyses

Unimmunized Immunized, no adjuvant Immunized + RAS Immunized + QS-7 Immunized + Quil A Immunized + QS-21 Immunized + PCPP

2.231 0.99 0.89 0.51 0.32 0.08 0.16

signi®cantly signi®cantly signi®cantly signi®cantly

(P < 0.001) greater than immunized groups (P R0.001) greater than adjuvant groups except RAS (P R0.005) greater than other adjuvant groups (P R0.01) greater than QS-21 and PCPP groups

a Groups of mice (8/group) were immunized and, 4 weeks later, challenged with EDIM. Shedding was determined by ELISA in stools collected during the following 7 days. Shedding values represent the average net optical density measurement on each stool specimen for every mouse within a group during each of the 7 collection days. Di€erences in shedding between groups were determined by a Student t-test using the total net OD values for each mouse during the 7-day collection period.

1578

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

QS-21 stimulated the greatest reduction in shedding, no signi®cant di€erences were found between this adjuvant and either PCPP or Quil A. The titers of no speci®c rotavirus antibody could be correlated with a speci®c adjuvant following immunization, as previously noted (see Table 1). Even so, there were signi®cant di€erences in the overall quantities of rotavirus antibodies measured within the di€erent immunized groups of mice. More importantly, the groups that developed the highest antibody titers (i.e. the QS-21 and PCPP adjuvant groups) also shed the lowest quantity of rotavirus antigen. Likewise, unimmunized groups that produced lower antibody responses (i.e. those administered RAS, QS-7, or Quil A) shed larger quantities of rotavirus. Finally, the group immunized without adjuvant that developed the lowest antibody titers shed the most virus. 4. Discussion Because rotavirus vaccines have all been live, orally deliverable rotavirus strains, immune responses and subsequent protection elicited by these viruses depend largely on their abilities to replicate in intestinal mucosal cells. Although these vaccines have provided signi®cant short-term protection, there is concern about their long-term e€ectiveness, particularly in light of the limited duration of high titers of intestinal rotavirus antibody found after natural infection [24]. Furthermore, subjects infected with vaccine strains of rotavirus are at least partially protected against subsequent infection with the same strains, thus limiting their potential to boost immune responses. This limitation should not, however, a€ect immune responsiveness following parenteral immunization where virus replication is not required. Therefore, a potential advantage of parenteral immunization is that it may be more e€ective than oral immunization for boosting immunity to rotavirus. Recent studies in mice and rabbits have demonstrated that parenteral immunization can elicit excellent protection against rotavirus infection [11±13]. Furthermore, the extent of protection was found to be augmented by inclusion of adjuvant [14]. Limited protection against rotavirus shedding was found in mice challenged after two i.m. immunizations with murine rotavirus but shedding was undetectable when the saponin adjuvant QS-21 was included during these immunizations. Because QS-21 has low toxicity and has boosted immune responses with other antigens, it is being evaluated in clinical trials [16, 18]. To determine the relative e€ectiveness of QS-21 as an adjuvant for i.m. immunization with rotavirus, it was compared to other adjuvants with potential for human use utilizing the adult mouse model [10].

The adjuvants selected for this comparative study have all been found to stimulate increased antibody production to a variety of antigens when administered parenterally [16±19, 25]. However, the mechanisms responsible have not been clearly de®ned for any of the ®ve adjuvants investigated. Quil A has been used as a component of immunostimulating complexes which are believed to sequester antigen for site-speci®c retention and slow release following injection [26]. Although this mechanism may play some role with other adjuvants, they are all believed to function as immunomodulators which induce local cytokine secretion. The subtypes of antibody elicited by immunization is dependent on the expansion of helper T cell subsets (i.e. TH1 and TH2) either directly or indirectly stimulated by inclusion of the adjuvant [23]. Interestingly, the relative expansion of T helper cell subsets in this study appeared to be identical for all ®ve adjuvants as determined by the nearly constant ratios of serum rotavirus IgG1/IgG2a found after immunization. In all cases the TH1 response dominated because there was ten times more IgG2a than IgG1 produced. Although inclusion of the adjuvants greatly increased the concentrations of both IgG subtypes, the ratio of rotavirus IgG1/IgG2a decreased relative to that found following vaccination in the absence of adjuvant, thus indicating greater stimulation of TH1 relative to TH2 cells with all adjuvants. As found with other antigens, all ®ve adjuvants increased the antigen-speci®c serum IgG titers found after i.m. immunization with inactivated rotavirus. Likewise, all adjuvants signi®cantly stimulated both serum rotavirus IgA and stool rotavirus IgG titers and both Quil A and QS-21 caused signi®cant increases in stool rotavirus IgA. However, the IgA responses were much lower than found after oral inoculation with live murine rotavirus which provided life-long protection of mice [27]. For all rotavirus antibodies measured except stool IgA, PCPP stimulated signi®cantly higher titers than the other adjuvants. This agrees with a previous report that a single dose of PCPP with a variety of antigens elicits persistent high titers of antibody, a response that appears to require more doses with other adjuvants [25]. We next attempted to correlate post-immunization rotavirus antibody titers with protection against shedding following subsequent EDIM challenge. As noted, there was an overall trend between higher rotavirus antibody responses and protection. However, this correlation was imperfect since protection in the PCPP group, which had the highest rotavirus antibody titers, was no better than in the QS-21 group and not signi®cantly better than in the Quil A group. Correlations between antibody titers and protection against rotavirus have been found in several studies including those associated with i.m. immunization [12, 13].

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

However, the importance of antibody as an immune e€ector has only been de®nitely established following oral immunization with live rotavirus and even there the mechanism of protection is uncertain [21, 28]. Quite possibly, antibody may only be a marker of immunity and not the e€ector following i.m. immunization. Although the ®ve adjuvants examined in this study stimulated signi®cantly di€erent antibody responses which generally correlated with increased protection, no adjuvant was clearly outstanding relative to the others. Most notably, under the conditions tested, Quil A, QS-21 and PCPP all provided comparable protection. Furthermore, even though QS-7 was evaluated at the concentration recommended by its manufacturer, a very recent result suggests that better responses may have been attained if the concentration of this adjuvant had been increased from 20 to 60 mg/dose [18]. Therefore, even though this study provides evidence for better protection against rotavirus with certain adjuvants following i.m. immunization, these results could be modi®ed if conditions such as dose number, concentration of adjuvant or antigen, or the route of immunization are altered.

Acknowledgements This work was funded in part by NIH-NIAID Contract N01-AI 45242 to Children's Hospital Medical Center.

References [1] De-Mol P, Zissis G, Butzler JP, Mutwewingabo A, Andre FE. Failure of live, attenuated oral rotavirus vaccine. Lancet 1986;ii:108. [2] Christy C, Madore HP, Pichichero ME, Gala C, Pincus P, Vosefski D, Hoshino Y, Kapikian A, Dolin R. Field trial of rhesus rotavirus vaccine in infants. Pediatr Infect Dis J 1988;7:645±50. [3] Bernstein DI, Smith VE, Sander DS, Pax KA, Schi€ GM, Ward RL. Evaluation of WC3 rotavirus vaccine and correlates of protection in healthy infants. J Infect Dis 1990;162:1055±62. [4] Bernstein DI, Glass RI, Rodgers G, Sack DA. Evaluation of rhesus rotavirus monovalent and tetravalent reassortant vaccines in US children. J Am Med Assoc 1995;273:1191±6. [5] Santosham M, Letson GW, Wolfe M, Reid R, Gahagan S, Adams R, Callahan C, Sack RB, Kapikian AZ. A ®eld study of the safety and ecacy of two candidate rotavirus vaccines in a Native American population. J Infect Dis 1991;163:483±7. [6] Santosham M, Moulton LH, Reid R, Croll J, Weatherbolt R, Ward RL, Forro J, Zito E, Mack M, Brenneman G, Davidson BL. Ecacy and safety of high-dose rhesus-human reassortant rotavirus vaccine in Native American populations. J Pediatr 1997;131:632±8. [7] QHBRV Study Group, Clark HF, White CJ, Ot PA, Stinson D, Eiden J, Weaver S, Cho I, Shaw A, Krah D, Ellis R.

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16] [17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

1579

Preliminary evaluation of safety and ecacy of quadrivalent human-bovine reassortant rotavirus vaccine (QHBRV). Pediatr Res 1995;37:1016. Rennels MB, Glass RI, Dennehy PH, Bernstein DI, Pichichero ME, Zito ET, Mack ME, Davidson BL, Kapikian AZ. Safety and ecacy of high-dose rhesus-human reassortant rotavirus vaccines ± Report of the National Multicenter Trial. Pediatrics 1996;97:7±13. Lanata CF, Midthun K, Black RE, Butron B, Huapaya A, Penny ME, Ventura G, Gil A, Jett-Goheen M, Davidson BL. Safety, immunogenicity, and protective ecacy of one and three doses of the tetravalent rhesus rotavirus vaccine in infants in Lima. Peru J Infect Dis 1996;174:268±75. Ward RL, McNeal MM, Sheridan JF. Development of an adult mouse model for studies on protection against rotavirus. J Virol 1990;64:5070±5. McNeal MM, Sheridan JF, Ward RL. Active protection against rotavirus infection of mice following intraperitoneal immunization. Virology 1992;191:150±7. Conner ME, Crawford SE, Barone C, Estes MK. Rotavirus vaccine administered parenterally induces protective immunity. J Virol 1993;67:6633±41. Con SE, Moser CA, Cohen S, Clark HF, Ot PA. Immunologic correlates of protection against rotavirus challenge after intramuscular immunization of mice. J Virol 1997;71:7851±6. McNeal MM, Rae MN, Conner ME, Ward RL. Stimulation of local immunity and protection in mice by intramuscular immunization with triple- or double-layered rotavirus particles and QS-21. Virology 1998;243:158±66. Yuan L, Kang S-Y, Ward LA, To TL, Saif LJ. Antibody-secreting cell responses and protective immunity assessed in gnotobiotic pigs inoculated orally or intramuscularly with inactivated human rotavirus. J Virol 1998;72:330±8. Kensil CR. Saponins as vaccine adjuvants. Crit Rev Therap Drug Carrier Systems 1996;13:1±55. Ulrich JT, Myers KR. Monophosphoryl lipid A as an adjuvant. In: Powell MF, Newman MJ, editors. Vaccine design: the subunit and adjuvant approach. New York, NY: Plenum Press, 1995. p. 495±524. Kensil CR, Wu J-Y, Anderson CA, Wheeler DA, Amsden J. QS21 and QS-7: puri®ed saponin adjuvants. In: Brown F, Haaheim LR, editors. Modulation of the immune response to vaccine antigens. Dev. Biol. Stand. Basel: Karger, 1998. p. 41±47. Payne LG, Jenkins SA, Woods AL, Grund EM, Geribo WE, Loebelenz JR, Andrianov AK et al. Poly [di(carboxylatophenoxy)phosphazene] (PCPP) is a potent immunoadjuvant for an in¯uenza vaccine. Vaccine 1998;16:92±8. McNeal MM, Broome RL, Ward RL. Active immunity against rotavirus infection in mice is correlated with viral replication and titers of serum rotavirus IgA following vaccination. Virology 1994;204:642±50. McNeal MM, Barone KS, Rae MN, Ward RL. E€ector functions of antibody and CD8+ cells in resolution of rotavirus infection and protection against reinfection in mice. Virology 1995;214:387±97. Knowlton DR, Spector DM, Ward RL. Development of an improved method for measuring neutralizing antibody to rotavirus. J Virol Meth 1991;33:127±34. Mosmann TR, Co€man RL. Th1 and Th2 cells: di€erent patterns of lymphokine secretion lead to di€erent functional properties. Annu Rev Immunol 1989;7:145±74. Ot PA, Rotaviruses: immunological determinants of protection against infection and disease. In: Maramorosch K, Murphy FA, Shatkin AJ, editors. Advances in virus research. San Diego, CA: Academic Press, Inc., 1994. p. 161±202.

1580

M.M. McNeal et al. / Vaccine 17 (1999) 1573±1580

[25] Payne LG, Van Nest GA, Barchfeld GL, Siber GR, Gupta RK, Jenkins SA. PCPP as a parenteral adjuvant for diverse antigens. In: Brown F, Haaheim LR, editors. Modulation of the immune response to vaccine antigens. Dev. Biol. Stand., Basel: Karger, 1998. p. 79±87. [26] Barr IG, Mitchell GF. ISCOMs (immunostimulating complexes): The ®rst decade. Immunol Cell Biol 1996;74:8±25.

[27] McNeal MM, Ward RL. Long-term production of rotavirus antibody and protection against reinfection following a single infection of neonatal mice with murine rotavirus. Virology 1995;211:474±80. [28] Franco MA, Greenberg HB. Role of B cells and cytotoxic T lymphocytes in clearance of an immunity to rotavirus infection in mice. J Virol 1995;69:7800±6.