JGV Papers in Press. Published June 9, 2014 as doi:10.1099/vir.0.067280-0

1

Title:

2

A poorly neutralizing IgG2a/c response elicited by a DNA vaccine protects mice

3

against Japanese encephalitis virus

4 5

Running title:

6

Protection by poorly neutralizing IgG2a/c

7 8

Category:

9

Animal viruses-Positive-strand RNA

10 11

Authors:

12

Hsin-Wei Chen1, 3, Hui-Wen Huang2, Hui-Mei Hu1, Han-Hsuan Chung1, Szu-Hsien

13

Wu1, Pele Chong1, Mi-Hua Tao2 and Chien-Hsiung Pan1, 3*

14 15

1. National Institute of Infectious Disease and Vaccinology, National Health Research

16

Institutes, Taiwan 350.

17

2. Institutes of Biomedical Sciences, Academia Sinica, Taipei, Taiwan.

18

3. Graduate Institute of Immunology, China Medical University, Taichung, Taiwan.

19 20

*Corresponding author:

21

35 Keyan Road, Suite R1-7024

22

Zhunan Town, Miaoli County, 350 Taiwan

23

Tel: 886-37-246166 ext. 37727

24

Fax: 886-37-583009

25

Email: [email protected]

26

Word Counts: Summary: 190; Main text (exclude references): 4318.

27

Abstract

28

We previously demonstrated that immunization with a DNA vaccine

29

expressing the JEV envelope (E) protein conferred a high level of protection through

30

a poorly neutralizing antibody response. In this report, we further investigated the role

31

of IgG subclass in this antibody-dependent protection using cytokine co-

32

immunization and cytokine-deficient mice. A significant difference in IgG2a/c but not

33

IgG1 was observed between mice that survived or died following a lethal challenge.

34

Correspondingly, the IgG2a/c response and protection increased in IL-4-deficient

35

mice but decreased in IFN-γ-deficient mice, highlights the importance of IgG2a/c. In

36

addition, the restoration of protection and E-specific IgG2a/c production in IFN-γ-

37

deficient mice by a Th1-biased intramuscular immunization suggests that IgG2a/c but

38

not IFN-γ is the major component for protection. The failure of protection against a

39

direct intracranial challenge indicated that IgG2a/c-mediated protection is restricted to

40

outside the CNS. Consistent with this conclusion, passive transfer of E-specific

41

antisera conferred protection only pre-exposure to JEV. Therefore, our data here

42

provide evidence that IgG subclass plays an important role in the protection against

43

JEV, particular in poorly neutralizing E-specific antibodies; and Th1-biased IgG2a/c

44

confers a better protection than Th2-biased IgG1 to against JEV.

45 46 47 48 49

Keywords: DNA vaccine; JEV vaccine; Antibody therapy; IgG subclass

50 51

Introduction Japanese encephalitis virus (JEV) belongs to the family of Flaviviridae and

52

causes severe encephalitis in southern and eastern Asia [for review, see reference

53

(Mackenzie et al., 2004)]. It is estimated that JEV infection causes 45,000 cases of

54

encephalitis and 10,000 deaths per year (van den Hurk et al., 2009). Even following

55

recovery, approximately 50% of survivors suffer long-term neurological sequelae

56

(Burke & Leake, 1988). Due to the limited therapeutic approaches, vaccination is the

57

most efficient strategy to control JEV, and a significant success has been achieved in

58

many Asian countries (Ding et al., 2003; Hoke et al., 1988; Srivastava et al., 2001).

59

Studies in animals and humans have demonstrated that host immunity correlates with

60

the outcome of JEV infection. It is well known that humoral immunity is a major

61

component of protection against JEV infection. Passive transfer of antisera has been

62

shown to provide strong protection against JEV and other flaviviruses (Engle &

63

Diamond, 2003; Hawkes et al., 1988; Kimura-Kuroda & Yasui, 1988; Kreil & Eibl,

64

1997; Phillpotts et al., 1987). In addition, studies in immunodeficient mice have

65

shown that protection was unchanged in CD8-deficient mice but lost in

66

immunoglobulin-deficient mice (Larena et al., 2011; Pan et al., 2001). It suggests

67

antibodies but not CD8 T cells play an important role in the protective immunity

68

against JEV; however, the protective role of CD8 T cells and the cytokines they

69

produced, particular in interferon (IFN)-γ cannot be completely ruled out, especially

70

in CNS infection (Larena et al., 2013).

71

Neutralization by blocking of the binding to the viral receptor on host cells is

72

the major mechanism through which antibodies limit virus infection and becomes the

73

primary criterion to evaluate vaccine efficacy (Hombach et al., 2005). Most

74

neutralizing antibodies recognize envelope (E) or pre-membrane protein on the

75

surface of the virion. However, the protection conferred through passive transfer of

76

anti-nonstructural protein (NS) 1 antibodies or a poorly neutralizing E-specific

77

monoclonal antibody suggests that poorly or non-neutralizing antibodies also play a

78

role in the protective mechanism against flaviviruses (Chung et al., 2006; Li et al.,

79

2012; Schlesinger et al., 1985; Vogt et al., 2011). Interestingly, a dominant IgG2a

80

subclass was observed among the protective NS1-specific monoclonal antibodies

81

(Chung et al., 2007). IgG subclass switching is regulated by cytokines, and plays

82

different role in infectious and autoimmune diseases. For example, IgG1 is regulated

83

by the Th2 cytokine-interleukin (IL)-4 and associated with allergy (Miyajima et al.,

84

1997). In contrast, IgG2a and IgG2c in some mouse strains are regulated by Th1

85

cytokines- IFN-γ and IL-12 (Barr et al., 2009), and becomes the most efficient

86

subclass at fixing complement and binding to the Fc gamma receptor (FcγR) on

87

macrophages and NK cells (Heusser et al., 1977; Kipps et al., 1985; Neuberger &

88

Rajewsky, 1981). FcγR engagement with antibody-antigen immune complexes

89

activates effector cells that destroy virus-infected cells by complement, antibody-

90

dependent cytotoxicity (ADCC) or phagocytosis (Clynes et al., 1998; Fossati-Jimack

91

et al., 2000). There are four murine FcγRs, including three activating FcγRs- FcγR I,

92

FcγR IIIA and FcγR IV, and one inhibitory FcγR- FcγR IIB (Ravetch, 2003). The

93

binding affinity to FcγRs also varies among IgG subclasses. The low affinity FcγR

94

IIIA and the inhibitory FcγR IIB are capable of binding a broad range of IgG

95

subclasses, including IgG1, IgG2a, IgG2b and IgG2c. By contrast, high affinity FcγR

96

I preferentially binds mouse IgG2a, and FcγR IV can bind IgG2a, IgG2b and IgG2c

97

with an intermediate to high affinity (Hamaguchi et al., 2006; Nimmerjahn &

98

Ravetch, 2005). Although FcγR and complement have been reported to involve in the

99

protective mechanism of poorly or non-neutralizing antibodies (Vogt et al., 2011), the

100

importance of IgG subclass in the protection mediated by poorly or non-neutralizing

101

antibodies is still unclear.

102

We have previously developed a JEV DNA vaccine based on a plasmid (pE)

103

expressing E protein lacking the signal peptide required for native conformation.

104

Immunization of pE elicited a poorly neutralizing antibody response but still provided

105

approximately 90% protection in a mouse model (Chen et al., 1999). We also

106

demonstrated that the protective mechanism was antibody dependent (Pan et al.,

107

2001). In the present study, we examined whether IgG subclass play a role in pE-

108

induced protection by using either cytokine co-delivery or cytokine-deficient mice.

109

The results can expand our understanding in the protective mechanism of the JEV

110

vaccine and provide useful information for the development of flavivirus vaccines.

111 112

Results

113

Modulation of IgG class switching and protection by co-delivery of cytokine genes

114

To address the role of IgG subclass in the protective mechanism of pE, co-

115

delivery of plasmid encoding cytokine gene, such as IL-4 (pIL-4) or IL-12 (pIL-12)

116

was used to shift IgG class switching to IgG1 and IgG2a or IgG2c. Since IgG2a and

117

IgG2c expression pattern varies in different strains of mice, for example C57BL/6

118

mice express IgG2c but not IgG2a (Zhang et al., 2012). Given that IgG2a and IgG2c

119

share 85% of amino acid homology (Schreier et al., 1981) and can be detected by the

120

same monoclonal antibody used in ELISA, the term of IgG2a/c is used to represent

121

the Th1-biased IgG subclass in this study. C3H/HeN mice were administered pE with

122

pIL-4, pIL-12 or vector pcDNA3 by either gene gun or intramuscular (im) injection.

123

For control, mice were immunized with pcDNA3 alone at the same dose. A specific

124

antibody response was detected in all pE-immunized mice but not the pcDNA3

125

control (table 1). There was no difference in the E-specific antibody response between

126

im or gene gun-immunized mice with co-delivery of pE and pcDNA3. Co-

127

administration with pIL-4 significantly enhanced the antibody response in gene gun-

128

immunized mice but not im immunized mice. The neutralizing activity in all pE-

129

immunized mice was below the detection limit (50% of plaque reduction at a 1:10

130

dilution), as we observed previously (Chen et al., 1999). Eight weeks after

131

immunization, mice received an intraperitoneal (ip) challenge, and all pcDNA3-

132

immunized mice died within 5-9 days after challenge. Although a comparable IgG

133

response was observed between pE+pcDNA3 gene gun and im immunization, the

134

latter induced an IgG2a/c-dominant response, with the highest IgG2a/c titer among all

135

immunized mice, and provided 100% protection against JEV challenge. Co-delivery

136

with pIL-4 did not alter the IgG2a/c-dominant response, although a slight decrease in

137

the IgG2a/c titer and protection (80%) compared to pE+pcDNA3 immunized mice

138

were observed. In contrast to im immunization, mice immunized by gene gun

139

developed high IgG1 and low IgG2a/c titers and showed an 80% of protection. Co-

140

delivery with pIL-4 by gene gun resulted in a significant increase in both IgG1 and

141

IgG2a/c; however, the protection was similar to that induced by pE and pcDNA3

142

(80%). Compared to pIL-4, co-delivery of pIL-12 suppressed E-specific antibody

143

responses and resulted in a decrease in both IgG1 and IgG2a/c, and the induced

144

protection was also reduced (40%).

145 146 147

Correlation of E-specific IgG2a/c in mice surviving JEV challenge To understand whether the IgG2a/c titer correlated with protection, the

148

individual IgG1 and IgG2a/c titers and the outcome following a JEV challenge of

149

C3H/HeN mice immunized with pE combined with pcDNA3, pIL-4 or pIL-12 in the

150

previous experiments were analyzed. There was no significant difference in the E-

151

specific IgG1 titer between mice that survived or died following the JEV challenge

152

(fig. 1a). However, a significant difference (p=0.0066 by Mann-Whitney t-test) was

153

observed in E-specific IgG2a/c between mice that survived and mice that died (fig.

154

1b).

155 156 157

Loss of IgG2a/c and protection in pE-immunized IFN-γ-deficient mice We further used IL-4- or IFN-γ-deficient mice to examine the roles of the IgG

158

subclass in pE-induced protection. After gene gun immunization, a comparable

159

antibody response at week 8 was observed in wild-type C57BL/6 (B6) mice, IL-4 KO

160

and IFN-γ KO mice, with IgG titers of 368±319, 361±239 and 491±334 U/ml,

161

respectively (fig. 2a). Following challenge, the survival rate of B6 mice was 80%

162

(n=5; fig. 2b), similar to previous results in C3H/HeN mice (table 1). The protection

163

induced by pE was a little but not significant higher in IL-4 KO mice (100%; n=7)

164

than B6. Surprisingly, the protection of pE in IFN-γ KO mice was diminished (0%;

165

n=5) and significant lower than B6 mice (p