Lymph node or ..... removed and lymph node cells were cultured with virus antigen for 5 days before ... 43 ~ specific 5~Chromium release at 25 : 1 respectively).
J. gen. Virol. (1987), 68, 1103-1114. Printedin Great Britain
1103
Key words: HSV-1/T lymphocyte recognition/glycoproteins B and D
Specificity of the Immune Response of Mice to Herpes Simplex Virus Glycoproteins B and D Constitutively Expressed on L Cell Lines By B. A. B L A C K L A W S , ~* A. A. N A S H 1 AND G. D A R B Y 2 ~Department o f Pathology, Tennis Court Road, Cambridge CB2 I Q P, and 2 Wellcome Research Laboratories, Beckenham, Kent, U.K. (Accepted 11 December 1986)
SUMMARY Mouse L cell lines have been developed which constitutively express glycoproteins B (gB) and D (gD) of herpes simplex virus type 1. When used to study the immune response of mice to the viral glycoproteins, it was found that both gB and gD induce a delayed type hypersensitivity response and both also induce an antibody response, but only the cell line expressing gD could stimulate the production of neutralizing antibody. Virus-specific cytotoxic T lymphocytes (CTLs) recognized gB expressed by the cell line and this line could also induce CTLs in mice. Recognition of gD by major histocompatibility complex class I restricted CTLs was never seen. Vaccination of mice with the cell lines provided protection from viral challenge and inhibited the establishment of a latent infection, although gD proved to be the better protective immunogen. INTRODUCTION The immune response of mice to herpes simplex virus type 1 (HSV-1) is known to involve both antibody and cell-mediated protective mechanisms. In particular, T lymphocytes are considered to play a central role in the recovery from a primary infection (for review, see Nash et al., 1985). However, the particular proteins of the virus recognized by these cells are far from clear. Experiments to date have centred on the use of the virus mutants deficient in the expression of certain glycoproteins to study the nature of the antigens recognized by cytotoxic T lymphocytes (CTLs) (Carter et al., 1981; Lawman et al., 1980; Glorioso et al., 1985), and the affinity purification of glycoproteins from whole virus preparations for use as immunogens to study delayed type hypersensitivity (DTH) responses (Schrier et al., 1983; Chan et al., 1985). The main problem has been how to separate the various antigens of this complex virus in order to dissect the nature of the immune response. Our approach has been to investigate the immune response to individual glycoproteins of HSV-1 by expressing the appropriate genes in mouse L cells. In this system, the glycoproteins are expressed in isolation but are still presented at the cell surface analogous to infected cells. Our reason for choosing glycoproteins is that they are likely to represent major target antigens of the immune system: this is true for the antibody response (for review, see Norrild, 1985) and there is evidence to suggest that glycoproteins are the major targets for both major histocompatibility complex (MHC) class I and class II restricted T lymphocytes (Carter et al., 1981 ; Lawman et al., 1980; Glorioso et al., 1985; Schrier et al., 1983). In this paper we report the establishment of L cells expressing either glycoprotein B (gB) or D (gD) and the use of these lines to study the specificity of the T cell response to HSV-1. The lines were also used to immunize mice against primary virus infections and to prevent the establishment of latent infections. METHODS Mice. Female CBA mice, 8 to 12 weeksof age, were obtained from the Department of PathologyAnimal House, University of Cambridge, U.K. 0000-7473 © 1987 SGM
B. A. BLACKLAWS, A. A. NASH AND G. DARBY
1104
(EcoRl) (EcoRl) ~//Hind II I ~KpnI Clat'~
f /
~ L;~k-b BomHI" ~ ~pm.
SV40 HaelIl C fragment
(b)
SV40 HaelII C fragment
(a)
SV40 HindlII C fragment
~ - BamH1
(EcoRI) (EcoRI) ~...];z;~-~f ~mall~ /
pB ~
KpnI
\
/.~ Kt)
SV40 early promoter fragment
]]1] HSV-1 TK gene fragment
/ ~ ~
JBamHl'~.~
fragment (
BamHl
~
"1
BamHI
Deleted fragment
- - pAT 153 sequences ~5~ SV40 sequences a=~ HSV-I sequences ~
Direction of transcription
Fig. 1. Expression vectors. (a) The plasmid pMI-1 is based on pAT 153 and contains the SV40 HaelII-C and HindlII-C fragments and the HSV-1 [strain CI(101)] BamHI-Q fragment in the pAT153 EcoRI, HindlII and BamHI sites respectively (sites in brackets were lost during cloning). In an anticlockwise direction, the plasmid therefore contains the HSV-I TK promoter, coding sequence and transcription termination signal, the SV40 enhancer, early promoter and early transcription initiation site, the pAT 153 ClaI site and the SV40 early polyadenylation signal. (b) The vector pBB-1 was developed from pMI-1 by inverting the KpnI fragment containing the HSV-1 TK gene relative to the SV40 sequences and then deleting the non-essential 1.3 kb BamHI fragment which was created (inset). This leaves the second HindIII site between the SV40 early promoter and pAT153 ClaI site as an additional cloning site. Transcription from the TK promoter is now in the opposite direction to transcription initiated from the SV40 early promoter.
Viruses. HSV-I strains HFEM, KOS, SC16 and CI(101)TK- were all passaged at a low multiplicity of infection in BHK-21 cells and stocks of virus were stored at - 7 0 °C. All viruses were titrated on BHK-21 cells. Cells. LMTK- and BHK-21 cells were grown in Glasgow modified Eagle's medium (GMEM). Selection of L cells containing thymidine kinase (TK) was carried out by supplementing the GMEM with thymidine, adenosine, guanosine, glycine and methotrexate and all transfected cell lines were maintained in this medium (Minson et al., 1978). Injection of mice. The various cell lines were harvested by mechanical displacement from growing surfaces with glass beads. The cells were washed three times in phosphate-buffered saline (PBS) and then resuspended to the required concentration in PBS. Mice were injected intraperitoneally with 0.1 ml of cells at 2 x 107 to 5 × 107 cells/ml or were injected subcutaneously in the hind footpad with 20 ~tl of cells at 2.5 x l0 s cells/ml. The immunized mice were challenged with SC16 injected into the ear pinna (Hill et al., 1975) or by scarifying the virus onto the neck of depilated animals (Simmons & Nash, 1984). Molecular cloning ofgB-1 and gD-1. Fig. 1 shows the expression vectors used to clone the glycoprotein genes. Based on pAT153, each contains the simian virus 40 (SV40) enhancer, early promoter, early transcription initiation site (SV40 HindllI-C fragment) and early polyadenylation signal (SV40 HaelII-C fragment) (sequence: Fiers et al., 1978). Between the transcription initiation site and polyadenylation signal, the ClaI site of pAT153 provides a suitable cloning site for the insertion of the structural genes; the unique HindIII site of pBB-1 is also between the two SV40 sequences. In addition both vectors contain the HSV-I TK gene of strain CI(101) including the promoter, coding sequence and transcription termination signal (Wagner et al., 1981) for use as a selectable marker in LMTK- cells. In pMI-I (Fig. I a), TK transcription is in the same direction as that from the SV40 promoter whilst in pBB-1 (Fig. 1b), transcription from the two promoters is in opposite directions. An EcoRI-F clone of HSV-1 (strain KOS) in pBR325 (a gift from M. Levine, Ann Arbor, Mich., U.S.A.) was double digested with XhoI and SphI (Boehringer Mannheim) and the 3-0 kb XhoI-SphI fragment containing the gB-1 structural gene isolated. The fragment was subcloned into the ClaI site ofpMI-1 (fragment and vector bluntended) and a plasmid, pMI-B3, containing the gB gene in the correct orientation to the SV40 promoter and polyadenylation signal was isolated. A BamHI-J clone of HSV-1 (strain Patton) in pBR322 (a gift from R. Watson, Minneapolis, Mn., U.S.A.) was double digested with HindIII and NruI (Boehringer Mannheim) and the 1.4 kb fragment containing the gD structural gene was isolated. This fragment was subcloned into the HindIII site ofpBB-I and a plasmid, pBB-DI, containing the gD gene in the correct orientation to the SV40 sequences was isolated.
The T cell response to H S V glycoproteins
1105
Transfection of LMTK- cells. Calcium phosphate precipitation of DNA onto LMTK- cells was performed as described by Minson et al. (1978) with a minor modification. In these experiments the DNA dose per dish was made up to 10 ~tg using LMTK- DNA as carrier and contained 0,1 ~tg of linearized plasmid [pMI-B3 or pBB-DI linearized with PvuI or XmnI (Boehringer Mannheim) respectively]. Transfected cells were selected in supplemented GMEM. Only one colony per dish was harvested (so each cell line was uniquely derived) and then cloned by limiting dilution to establish a cell line. The cell line, LTK +, was established by transfection of the parental LMTK- cells with the expression vector pMI-1. Polyaerylamide gel electrophoresis of immunoprecipitated antigens. Radiolabelled infected cell lysates were prepared by infecting confluent BHK or LMTK- cell monolayers (grown in 3.5 cm dishes) at a multiplicity of infection of 10 or 20 p.f.u./cell respectively. Starting at 4 h post-infection, infected or uninfected cells were incubated with methionine-free minimum essential medium supplemented with 1 ~ GMEM containing 50 p,Ci of [35S]methionine (1315 Ci/mmol; Amersham) for 20 h at 37 °C, harvested and then lysed in 1 ml RIPA buffer (50 mM-Tris pH 7.2, 150 mM-NaC1, 1% sodium deoxycholate, 0.1~ SDS, 1 ~ Triton X-100) plus 2 mMphenylmethylsulphonyl fluoride and 10 units of micrococcal DNase (Sigma) per mL Lysates were clarified by centrifugation at 100000g for 1 h at 4 °C. Immunoprecipitations were performed as described by Richman et al. (1986) on 250 ~tl samples of lysate with 5 ~tl of ascites fluid containing either anti-gB (gB2) or anti-gD (LP2) monoclonal antibodies (gifts from M. Levine, Ann Arbor, Mich., U.S.A. and A. Minson, Cambridge, U.K., respectively). All samples were boiled for 3 min before use. Samples of 20 ~tl were electrophoresed for 4 h at 35mA on 10~ acrylamide-0.1~ N,N'-methylenebisacrylamide resolving gels with 3 ~ acrylamide-0.13~ N,N'methylenebisacrylamide stacking gels using the SDS buffer system of Laemmli (1970). The gels were then fixed, fluorographed (Bonnet & Laskey, 1974) and dried before exposure to Kodak X-Omat S film at - 7 0 °C. Determination of antibody titres in serum Two methods to determine antibody titre were used. Radioimmunoassay. This assay was based on the method reported by Colombatti & Hilgers (1979). Briefly, antigen plates contained HSV-1 strain HFEM-infected or uninfected BHK-21 cells whose non-specific binding sites were blocked with heat-inactivated foetal calf serum. Serum and ~251-1abelled Protein A ( > 30 mCi/mg; Amersham) dilutions were made in PBS plus 10~ heat-inactivated foetal calf serum. Endpoint titres of antibody were taken as the dilution at which the counts from infected wells were double that of uninfected wells. Plaque reduction assay. Neutralizing antibody titres were determined by incubating 5 x 103 p.f.u. HFEM with serum and exogenously added rabbit serum as a source of complement before titrating the remaining viable virus on BHK-21 cells. Antibody endpoint titres were taken as the dilution that gave 50~o reduction in virus titre. Detection of virus-specific CTLs. CTLs were prepared as described by Pfizenmaier et al. (1977) with modifications. Briefly, the cervicular and auricular draining lymph nodes or spleens from experimental mice were removed and made into single cell suspensions as previously described by Nash et al. (1980a). Lymph node or spleen cells were cultured for 5 days in RPMI 1640 medium with SC16-infected, X-irradiated spleen cells (2:1 cells :feeder ratio) and then harvested for use in a cytotoxicity assay. Assay. 2 x 106 target cells were labelled with 100 laCi 51Chromium (sodium chromate in aqueous solution, 250 to 500 ~tCi/mg Cr; Amersham) as in Nash et al. (1980b). Target cells were then incubated with effectors for 8 to 10 h at 37 °C in 5 ~ CO2 and the percentage specific 5tChromium release was determined. All values were obtained from quadruplicate samples. DTH response. Mice were injected subcutaneously in the left hind footpad with 20 ~tl PBS containing 5 x 106 cells or 5 × 104 p.f.u. CI(101)TK-. Seven days later 5 × 105 p.f.u. CI(101)TK- was injected into the left ear and the ear thickness was measured at 24, 48, 72, 96 and 120 h post-infection as previously described (Nash et al., 1980a). All results are expressed as the difference in thickness between infected and uninfected ears. Titration of virus in mouse tissues. Ear pinnae were removed 5 days post-infection, homogenized in 1 ml 10~ tryptose phosphate broth (ETC) and stored at - 7 0 ° C before dilution for independent assay of virus on BHK-21 cells (Nash et al., 1980a). To determine latent virus in ganglia, mice received 105 p.f.u. SC16 scarified onto the neck. The mice were sacrificed during the latent phase of infection (i.e. > 30 days post-infection) and the left cervical ganglia (CII, III and IV pooled) were removed. Ganglia were placed into 0.5 ml GMEM with 10% tryptose phosphate broth and cultured for 5 days at 37 °C, in 5 ~ CO2. The ganglia were then homogenized in 1 ml total of ETC and stored at - 7 0 °C before dilution for independent assay on BHK-21 cells. For the purpose of these experiments, if any infectious virus was recovered from the ganglia this was scored as positive for a latent infection. RESULTS
Cloning the structural genes for H S V - 1 g B and g D T h e c l o n i n g v e c t o r s used to e x p r e s s t h e HSV-1 g l y c o p r o t e i n g e n e s are s h o w n in Fig. 1. E a c h v e c t o r c o n t a i n s t h e S V 4 0 early p r o m o t e r f r o m w h i c h c l o n e d g e n e s are e x p r e s s e d a n d t h e H S V - 1 T K g e n e as a s e l e c t a b l e m a r k e r .
1106
B. A. BLACKLAWS, A. A. NASH AND G. DARBY (a) gB-1 (strain KOS) 0.35 re.u_ SphI
XhoI,a "~7m.u.
signal
i
3.1 kb mRNA
411
mRNA initiation site
(b) gD-1 (strain Patton) NruI ~10.93 m.u.
0.90 m iu!HindIII
~
Coding sequences of glycoprotein genes
f
!ATG mRNA initiation site
3-0 kb mRNA
1 kb
j
//'---I~
Fig. 2. Structural genes for gB-1 and gD-1. (a) HSV-1 strain KOS (Bzik et al., 1984) was sequenced between 0.345 and 0.370map units (m.u.) in the UL region of the HSV-1 genome. The sequence spanned the gB gene which is transcribed from right to left in the prototypic form of the viral genome. The XhoISphI sites used to excise the structural gene from an EcoRI-F clone are indicated and the fragment generated contains the coding region and polyadenylation signal, (b) HSV-1 strain Patton (Watson et al., 1982)was sequenced between 0.90 and 0.93 m.u. of the Us region, spanning the 5' non-coding and coding regions of the gD gene. The HindlII and NruI sites used to excise only the coding region of the gene are shown. In the prototypic arrangement of the HSV-1 genome, the gD gene is transcribed from the left to right.
Using the published sequences for gB-1 (strain KOS: Bzik et al., 1984), suitable restriction sites for subcloning the structural gene ofgB-1 (strain KOS) from an EcoRI-F clone were chosen. The XhoI site 5' of the coding region was used. This is 252 bases from the initiation codon and is 3' of the putative T A T A box and m R N A initiation site of gB. The chosen 3' terminal of the cloned fragment was the SphI site which is 48 bases 3' of the predicted polyadenylation signal. Therefore the cloned XhoI-SphI fragment contains the complete gB coding sequence, some 5' leader sequence (but not the promoter) and some 3" flanking sequence (including the putative gB polyadenylation signal) (Fig. 2a). This 3 kb fragment was cloned into the ClaI site of pMI-1 and a plasmid containing the gB gene in the correct orientation to the SV40 promoter was identified by restriction enzyme analysis. The published sequence of gD-1 (strain Patton: Watson et al., 1982) was used to choose suitable restriction sites to excise the structural gene of gD-1 (strain Patton) from a BamHI-J clone. The 5' and 3' restriction sites chosen were the HindIII and NruI sites respectively (Fig. 2b). The HindIII site is 75 bases 5' of the initiation codon and the NruI site is 188 bases 3' of the termination codon. Therefore the HindIII-NruI fragment chosen contains only the gD-1 coding sequence and some 5' and 3' flanking sequence (Fig. 2b). This 1-4 kb fragment was cloned into the HindIII site of pBB-1 and all recombinant plasmids isolated contained the gD gene in the correct orientation to the SV40 promoter. The HindIII-Nru] fragment has also been cloned into the ClaI site of pMI-1 to give a plasmid, pMI-D2. Establishment o f L cell lines expressing HSV-1 glycoproteins L cell lines recovered after transfection with the vectors all contained plasmid D N A integrated into the genome of the cells. Two cell lines have been used in this study: LTKgB (clone 22) and L T K g D (clone 2). L T K g D contained the complete gD gene at about three copies/genome .whilst LTKgB contained the complete gB gene at about five copies/genome (determined by Southern blot analysis, data not shown). Plasmid D N A was maintained stably in the cell genome throughout passage (LTKgB passed 20 times and L T K g D passed 23 times which is the equivalent of approximately 5 to 6 months in culture). To demonstrate the expression of the glycoprotein genes in the cell lines, radiolabelled cell
The T cell response to H S V glycoproteins
e,l
tL
~
e-i
1107
t~
(~
Fig. 3. Immunoprecipitates from LTKgB and LTKgD. LTKgB (clone 22) and LTKgD (clone 2) are cell lines established from LMTK- cells transfected with pMI-B3 (gB gene) and pBB-D1 (gD gene) respectively. [3sS]Methionine-labelled cell lysates were immunoprecipitated with (a) gB2 and (b) LP2 monoclonal antibodies and precipitates were electrophoresed on 10% SDS-polyacrylamide gels crosslinked with N,N'-methylenebisacrylamide. Fluorograms of gels were exposed to Kodak X-Omat S film for (a) 4 and (b) 3 days. MI, Mock-infected LMTK- cells; HFEM, HFEM-infected LMTK- cells; 2, LTKgD; 22, LTKgB. Positions of molecular weight marker proteins are shown.
lysates were immunoprecipitated with type-common monoclonal antibodies to gB and gD (gB2 and LP2 respectively). The two cell lines used, LTKgB (Fig. 3a, lane 22) and L T K g D (Fig. 3b, lane 2), expressed gB and gD respectively as clearly precipitable bands which co-migrated with infected cell gB (Fig. 3a, lane H F E M ) and gD (Fig. 3b, lane HFEM). Uninfected cell peptides which cause background bands upon immunoprecipitation are shown in mock-infected (MI) lanes of Fig. 3. In Fig. 3(a), lanes 22 and H F E M show bands that do not correspond to background bands. These are approx. 49K, 33K and 27K in lane 22 and approx. 110K, 87K and
1108
B. A. BLACKLAWS, A. A. NASH AND G. DARBY Table 1. Serum antibody titres from mice immunized with gB or gD Antibody titre* A
(
Group LMTKLTKgB LTKgD LMTK- + HFEM
Radioimmunoassay
Neutralization
10 104 to 105 103 to 105 l0 s to 106
< 10 < 10 10 to 103 102 to 103
* Endpoints were expressed as the reciprocal of the highest dilution of antiserum which gave an infected :uninfected B H K binding ratio of 2 for radioimmunoassay, or 5 0 ~ reduction in plaque n u m b e r for neutralization assay.
48K in lane HFEM and may be due to degradation products ofgB. In Fig. 3(b), the infected cell lane (HFEM) shows a band of approx. 120Kd which may correspond to a dimer of gD (Eisenberg et al., 1982). Each cell line expressed the precursor and mature forms of the glycoprotein throughout cell passage. Pulse labelling of the glycoproteins before immunoprecipitation showed that the precursor forms of the glycoproteins in the cell lines were processed more swiftly to the mature forms than in infected cells (data not shown) as was the case for gD-2 expressed in transfected L cells by Johnson & Smiley (1985). However, the relative intracellular locations of the precursor and mature species and the post-translational modification events taking place were not studied further. Both gD and gB were present in the plasma membrane of LTKgD and LTKgB as detected by immunofluorescence (data not shown). From a series of cell lines studied, pMI-1 was found to be the more efficient expression vector; for example 100 ~ of the cell lines screened expressed gD in pMI-1 compared to 33 ~o in pBB-1. Previous evidence suggests that this effect may be due to the orientation of the TK gene compared to the glycoprotein gene (Darby et al., 1981 ; Minson et al., 1978).
Antibody response induced by celt lines The immune response of mice to individual HSV-1 glycoproteins, gB and gD, was studied using the cell lines established in the previous section. Firstly, the antibody response of mice to the cells was investigated. Mice received two injections of 5 × 106 live cells (LMTK-, LTKgB, LTKgD, or L M T K - infected with HFEM, m.o.i. 2) intraperitoneally, 14 days apart. Seven days after the second injection, serum samples were taken and the antibody levels determined by radioimmunoassay or by a plaque reduction assay (Table 1). Sera were analysed individually and the results for each immunization group were expressed as the range obtained from individual mice in the group. Each group contained five mice. As can be seen in Table 1, immunization of mice with live cells expressing gB or gD induced comparable antibody titres, when measured by radioimmunoassay. Non-specific binding of the sera to uninfected and infected BHK-21 cells caused a high background at high serum concentrations and this caused the apparent 'HSV-1specific antibody' titre seen in mice immunized with only the parental L M T K - cells. Subsequent immunoprecipitation of radiolabelled HSV-1-infected BHK cell lysate with the sera has shown this binding to be non-specific (data not shown). Immunopreeipitation of infected BHK lysates has also shown that the anti-HSV-1 antibody in sera from mice immunized with LTKgD or LTKgB was specific for gD and gB respectively (data not shown). No detectable neutralizing antibody was found in sera from mice immunized with LTKgB. In contrast, mice immunized with LTKgD produced clearly detectable neutralizing antibody (Table 1). This neutralizing antibody was complement-independent. Specificity of HSV-l-induced cytotoxic T lymphocytes CTLs induced in vivo by infection of mice with strain CI(101)TK- were re-stimulated with virus antigen in vitro and tested against infected or transfected cells in a 51Chromium release assay. The effector cell population induced by this method has previously been shown to be Thyl ÷ L3T4- Lyt2 + (Nash et al., 1980b; A. A. Nash, unpublished observation). As shown in Fig. 4, lysis of SC16-infected cells was 39~ at an effector :target ratio of 50 : 1. Although gB is
The T cell response to H S V glycoproteins 50
I
I
I
1109
I
80
I
I
I
70
40
6O 30
50
G ?,
~, 40 20 3o
20
l0
10 0
50:1
25:1 12:1 Effector :target ratio
6:1
0
I
I
I
25:1
12:1 Effector:target ratio
6:1
Fig. 4
Fig. 5
Fig. 4. Specificityof HSV-l-induced cytotoxic T cells. Mice were injected twice with 5 x 104 p.f.u. CI(101)TK- in the ear at a 7 day interval. Four days after the last injection, draining lymph nodes were removed and lymph node cells were cultured with virus antigen for 5 days before use in the cytotoxicity assay. Target cells were (O) LMTK-, (11) LTKgB, (0) LTKgD and (A) SC16-infectedLMTK- cells. The assay mixture was incubated for 10 h at 37 °C and the s~Cr release determined. Minimum S~Cr release into the supernatant was 15~ or less of the maximum values obtained. Error bars show one standard deviation of experimental values. Fig. 5. Specificityof cytotoxic T cells induced in vivo with LTKgB. Mice were injected in the ear with 5 x 104 p.f.u. CI(101)TK- 7 days before injection intraperitoneally with 5 x 106 LTKgB cells (two doses at a 7 day interval). Five days after the last immunization spleens were removed and spleen cells cultured for 5 days with HSV-1 antigen. Cells were harvested, separated on lymphocyte separation medium (Flow Laboratories), and the lymphocyte fraction was used as effectors in the cytotoxicity assay. Target cells were (O) LTK+, (11) LTKgB, (A) SC16-infected LTK + and (0) KOS-infected LTK + cells. The assay mixture was incubated for 9 h at 37 °C and the 5~Cr release determined. Minimum 5~Crrelease valuesobtained were 20~ or less than the maximum values. Error bars show one standard deviation of experimental values.
recognized by the HSV-l-specific CTL ( 1 4 ~ specific 51Chromium release at 5 0 : 1 ; LTKgB, Fig. 4), the low level of killing indicates that gB-specific CTLs constitute only a minor population. There was, however, no recognition of the cell line expressing gD (LTKgD, Fig. 4).
Specificity of CTLs stimulated in vivo with gB As only a minor population of HSV-l-specific CTLs recognized LTKgB, it was decided to amplify this population by stimulating with antigen in vivo. Mice were injected with live virus [strain CI(101)TK-] and then LTKgB (twice) at 7 day intervals before removal of the spleens and re-stimulation of the lymphocytes in vitro with SC16-infected spleen cells. Fig. 5 shows that CTLs generated by this method now recognized LTKgB as their major target (25:1, 71 specific 51Chromium release) and SC 16- and KOS-infected cells were also recognized (45 ~o and 43 ~ specific 5~Chromium release at 25 : 1 respectively). The effector T lymphocytes in these experiments were characterized as Thyl + L3T4- Lyt2 ÷ (data not shown). (KOS-infected cells were included as targets as the gB gene cloned into pMI-1 was originally from strain KOS and therefore any strain-specific recognition of gB would be shown as a difference in recognition between the SC16- and KOS-infected targets.)
1110
B. A. B L A C K L A W S ,
20
I
I
I 24
I 48
A. A. N A S H A N D G . D A R B Y I
I
I
10 e-
0
I i 72 96 Time post-infection (h)
I 120
Fig. 6. DTH response to gB and gD. Mice were injected subcutaneously in the hind footpad with 5 × 106 cells (O, LTK+; II, LTKgB; 0 , LTKgD) or 5 x 104 p.f.u. CI(101)TK-(A). Seven days after the injection, the mice were challenged with 5 x 105 p.f.u. CI(101)TK- in the ear. Ear thickness was measured at 24, 48, 72, 96 and 120 h post-challenge and the difference between infected and uninfected ears was calculated. Values were expressed as the mean differences in ear thickness + 1 standard deviation of five mice/group.
Induction of CTLs specific for gD In view of the results of the above experiment we investigated whether L T K g D could stimulate a CTL response. However stimulation of CTLs in vivo with this line by the method used to induce gB-specific CTL (Fig. 5) produced lymphocytes which gave a 7 ~ , 5 ~ and 3 specific 51Chromium release from L T K ÷, SC 16 infected L T K + and L T K g D cells respectively at a 50 : 1 effector : target ratio. In this and other experiments CTLs specific for gD were never observed. A cell line expressing HSV-1 T K was included as a target in many experiments (see above) and the presence of the transfected T K gene had no a p p a r e n t effect on the recognition of infected cell proteins by HSV-l-specific CTLs.
DTH response of mice primed with gB or gD To study the D T H response to HSV-1 glycoproteins, mice were injected subcutaneously with live cells (LTK ÷, L T K g B or L T K g D ) or virus and challenged 7 days later in the ear pinna with live virus. Fig. 6 clearly shows that priming with whole virus, L T K g D or L T K g B all produce a maximal response at 48 h post-infection. The swelling induced by immunization with gB or gD was as great as that induced by whole virus antigen. Background swelling caused by viral replication is not seen (Fig. 6, L T K ÷) as the virus strain used to challenge the mice was T K negative and replicates poorly in vivo.
Protection of mice against viral challenge by immunization with gB or gD Two methods were used to look at the protection afforded to mice by immunization with the glycoproteins: a virus clearance from the site o f infection and protection from establishment of a latent infection.
Virus clearance from the site of infection & mice immunized with gB or gD Mice were injected once with 2 × 106 live cells (LTK ÷, L T K g B or L T K g D ) or 8 × 104 p.f.u. SC16 subcutaneously in the hind footpad 7 days before challenge in the ear with 5 × 104 p.f.u. SC16. Virus remaining in the ear at 5 days post-infection was determined and is expressed in Table 2 as the virus titre from individual mice. Previous immunization of animals with live virus
The T cell response to H S V glycoproteins
1111
Table 2. Virus clearance from the ear pinna of mice immunized with gB or gD Group LTK+ LTKgB LTKgD SC16
Virus titre* (log10 p.f.u./ear) 5.10, 5.38, 5.39, 5.58, 5.70 3.26, 3,37, 4.10, 4.30, 4.75 < 1.0, 1.00, 1.30, 1.30, 2-38 < 1.0, < 1.0, < 1.0, < 1.0, 1-00
* Virus titres of less than 1 log~0 were not determined. Table 3. Effect of immunization with gB or gD on establishment of latent infections ( Group LTK + LTKgB LTKgD CI(101)TK-
No. of immunizations ), 1
2
100" 100 100 78
80 40 10 11
* Values given are the percentage of mice latently infected.
gave a > 4 lOglo reduction in virus titre in the ear: in some cases, this was a total clearance of detectable virus. Immunization with gD caused a 3 to 4 log~o reduction in virus titre compared to background (LTK + values) and in one mouse this was a total clearance of detectable virus. Immunization with gB only gave a 0.5 to 1.5 log1 o reduction in virus titre at 5 days post-infection. These levels of virus clearance were obtained from three separate experiments and immunization twice (a 14 day interval between injections was used) with the antigens 7 days before viral challenge also gave similar responses (data not shown).
Protection from establishment of a latent infection Scarification of virus onto the skin of mice has been shown to be a very efficient method of introducing virus into the nervous system and inducing latent infections (Simmons & Nash, 1984). This model, rather than the ear model of infection, was therefore used in the study. Two groups of mice were injected intraperitoneally with 2 × 106 LTK +, LTKgB or L T K g D cells, or 5 × I04 p.f.u. CI(101)TK-. One group of mice was given a second injection with the above dose of cells or virus 14 days after the first. Fifteen days after the last injection, 10s p.f.u. SC16 were scarified onto the neck. Cervical ganglia were removed 30 days after challenge, explanted for 5 days and the virus was assayed. Ganglia were scored as positive for a latent infection if infectious virus was recovered from the explant. In direct contrast to protection as measured by virus clearance, one immunization of antigen (LTKgB or LTKgD) did not protect the mice from establishment of a latent infection (Table 3). However, two doses (at a 14 day interval) of gB or gD (Table 3) did provide protection against establishment of latency. Again gD induced the greatest protection with only one of 10 mice showing a latent infection compared with gB in which four of 10 mice showed latency (although it must be noted that only 8 0 ~ of the unimmunized mice showed a latent infection). DISCUSSION
In view of the complexity of HSV-1 it is necessary to separate its antigenic determinants in order to unravel the nature of the immune response to this virus. In this report, we describe the use of L cell lines consistitutively expressing gD or gB of HSV-1 as probes to study the type of immune response induced to these glycoproteins and their effect as immunogens in the prevention of infection and the establishment of latent infections. An intriguing observation with LTKgD (clone 2; cell line expressing gD) was that CTLs obtained from virus-infected animals would not recognize this cell line, neither could this cell line induce a CTL response when used to immunize mice. Nevertheless, mice injected with
1112
B.A.
B L A C K L A W S , A. A. NASH AND G. DARBY
LTKgD were able to eliminate virus rapidly from the skin and when mice were immunized twice with LTKgD they were protected from the establishment of a latent infection following challenge with a high dose of wild-type virus. Vaccination with gD (whether synthetic peptides, purified protein or vaccinia virus recombinants expressing gD) has also previously shown that this protein is a potent protective immunogen (Eisenberg et al., 1985; Berman et al., 1985; Long et al., 1984; Cremer et al., 1985; Paoletti et al., 1984). Although human (T4 +) CTL clones have been produced which will proliferate in response to cloned, truncated forms of gD-1 (Zarling et al., 1986a) and are cytotoxic for cells infected with vaccinia virus recombinants expressing gD (Zarling et al., 1986 b), these T4 + cells are MHC class II restricted (Zarling et al., 1986 a, b). Classical (MHC class I restricted) T cell killing specific for gD-1 has not been observed in man or the mouse. This could reflect a failure of gD to associate with MHC class I molecules or could represent a 'hole' in the T cell repertoire for this antigenMHC combination. Whatever the explanation for the failure to induce CTLs to the glycoprotein, it is obvious that the mice are not compromised by this. So in the prevention of infection, CTLs may be considered a redundant force. Support for this suggestion is found in mice deficient in Lyt2 + cells where these animals are well able to control a primary HSV-1 infection (Nash et al., 1987). The protective mechanism stimulated by gD-1 is uncertain; LTKgD induces a good neutralizing antibody response and a DTH response. There is strong evidence favouring either or both in the control of a primary cutaneous infection (Simmons & Nash, 1985; Nash et al., 1980a). However, mice deficient in B cell function and hence the ability to make antiviral antibodies, were able to resist re-infection with HSV-1 which was scarified into the skin (Simmons & Nash, 1987). This would suggest that the antiviral activity here could be mediated predominantly via MHC class II restricted 'helper' T cells. In contrast to LTKgD, lymph node cells from virus-infected mice were able to recognize and lyse LTKgB (clone 22; cell expressing gB) albeit weakly. That gB is a target for Lyt2+ CTLs was shown clearly when this clone was used to immunize mice. Both the clone and infected syngeneic cells were recognized. Why the infected cells are lysed to a lesser extent than LTKgB is unclear. One possible explanation is that on the membrane of infected cells, gB interacts with other glycoproteins and is unavailable for interacting with MHC class I gene products. Alternatively, in the transfected cell, antigen 'processing' may occur in a different way than in the infected cell. These arguments may also be applied to the recognition of LTKgD by virus-specific CTLs but this would not explain why LTKgD will not induce CTLs which recognize itself. Using HSV-1 mutants temperature-sensitive for gB, it has previously been inferred that gB is recognized by CTLs (Lawman et al., 1980). However, Glorioso et al. (1985) have shown that gC is the major target antigen recognized by MHC class I restricted T cells. This is particularly intriguing in view of the evidence in other viruses where there is a dominant recognition of nonmembrane proteins, such as nucleoprotein of influenza (Townsend et al., 1984), immediate early gene product of murine cytomegalovirus (Reddehase & Koszinowski, 1984), and the nuclear protein of vesicular stomatitis virus (Yewdell et al., 1986). It remains to be determined conclusively which is the dominant antigen recognized by HSV-l-specific CTLs. LTKgB also induces a protective response which is less efficient than that to LTKgD. This difference could be related to the amount of antigen expressed by the two cell lines, although preliminary evidence suggests that there is little difference between the two. Unlike LTKgD, LTKgB does not induce a neutralizing antibody response. As these mice are able to clear the virus from the inoculation site, this supports the arguments above, that neutralizing antibodies are not an absolute requirement for protection. Nevertheless, there is evidence in murine HSV infection for a role of non-neutralizing antibodies in protection against ocular (Rector et al., 1982), subcutaneous (Balachandran et al., 1982) and intracerebral infection (Kumel et al., 1985). By using cell lines expressing individual gtycoproteins, a start has been made on dissecting the nature of the immune response to this complex virus. There are at least seven glycoproteins of HSV-1 so far determined (for review, see Spear, 1985; Richman et al., 1986; Frame et al., 1986; Buckmaster et al., 1984), and gD and gB have already shown that the type of immune response to different glycoproteins may be as varied as the type of immune response to different viruses. We
The T cell response to H S V glycoproteins
1113
are now undertaking a systematic dissection of the immune response to other HSV-1 glycoproteins for comparative purposes. Protective responses induced by these glycoproteins will also be studied as this information will be important for the development of subunit vaccines to HSV-1. However, it should be borne in mind that these are just seven genes out of more than 50 encoded by the virus genome, many of whose proteins may prove to be important target structures for the immune system. The expression vector, pMI-1, was developed and kindly given to the authors by Dr M. Inglis, Department of Pathology, University of Cambridge, U.K. The authors would like to thank the Medical Research Council of Great Britain for funding the work reported in this paper. B.B. was a recipient of a MRC research studentship. REFERENCES BALACHANDRAN,N., BACCHETTI, S. & RAWLS, W. E. (1982). Protection against lethal challenge of Balb/c mice by passive transfer of monoclonal antibodies to five glycoproteins of herpes simplex virus type 2. Infection and Immunity 37, 1132 1137. BERMAN, P. W., GREGORY, T., CRASE,O. & LASKY,L. A. ([985). Protection from genital herpes simplex virus type 2 infection by vaccination with cloned type 1 glycoprotein D. Science 227, 1490-1492. BONNER, W. M. & LASKEY,R. A. (1974). A film detection method for tritium-labeled proteins and nucleic acids in polyacrylamide gels. European Journal of Biochemistry 46, 83-88. BUCKMASTER, E. A., GOMPELS, U. & MINSON, A. (1984). Characterisation and physical mapping of an HSV-1 glycoprotein of approximately 115 x 103 molecular weight. Virology 139, 408~,13. BZIK, D. J., FOX, B. A., DELUCA,N. A. & PERSON,S. (1984). Nucleotide sequence specifying the glycoprotein gene, gB, of herpes simplex virus type 1. Virology 133, 301-314. CARTER, V. C., SCHAFFNER, P. A. & TEVETHIA, S. S. (1981). The involvement of herpes simplex virus type 1 glycoproteins in cell-mediated immunity. Journal oflmmunology 126, 1655-1660. CHAN, W. L., LUKtG, M. L. & LmW, F. Y. (1985). Helper T cells induced by an immunopurified herpes simplex virus type 1 (HSV-1) 115 kilodalton glycoprotein (gB) protect mice against HSV-1 infection. Journal of Experimental Medicine I62, 1304- t 318. COLOMBATTI,A. & HILGERS,J. (1979). A radioimmunoassay for virus antibody using binding of t 25i_labelled Protein A. Journal of General Virology 43, 395401 . . . . CREMER, K. l., MACKETT, M., WOHLENBERG, C., NOTKINS, A. L. & MOSS, B. (1985). Vaccinia virus recombinant expressing herpes simplex virus type 1 glycoprotein D prevents latent herpes in mice. Science 228, 737-740. DARBY, G., BASTOW,I(. F. & MINSON,A. C. (1981). Transfections with HSV D N A fragments and D N A from HSV transformed cells. In Developments in Molecular Virology, vol. 1, Herpesvirus DNA, pp. 223 253. Edited by Y. Becker. The Hague, Boston & London: Martinus Nijhoff Publishers. EISENBERG,R. J., PONCEDE LEON,M., PEREIRA,L., LONG, D. & COHEN, G. H. (1982). Purification of glycoprotein gD of herpes simplex virus types 1 and 2 by use of monoclonal antibody. Journal of Virology, 41, 1099-1104. EISENBERG,R. J., CERINI, C. P., HEILMAN,C. J., JOSEPH, A. D., DIETZSCHOLD, B., GOLUB, E., LONG, D., PONCEDE LEON, M. & COHEN, G. n. (1985), Synthetic glycoprotein D-related peptides protect mice against herpes simplex virus challenge. Journal of Virology 56, 1014-1017. FIERS, W., CONTRERAS,J., HAEGEMAN,G., ROGIERS,R., VANDE VOORDE,A., VANHEUVERSWYN,H., VANHERREWEGHE, I., VOLCKAERT,G. & YSEBAERT,i . (1978). Complete nucleotide sequence of SV40 DNA. Nature, London 273, 113-120. FRAME,i . C., MARSDEN,H. S. & McGEOCH,D. I. (1986). Novel herpes simplex virus type 1 glycoproteins identified by antiserum against a synthetic oligopeptide from the predicted product of gene US4. Journal of General Virology 67, 745-751. GLORIOSO, J'., KEES,U., KUMEL,G., KIRCHNER,H. & KRAMMER,P. H. (1985). Identification of herpes simplex virus type 1 (HSV-1) glycoprotein gC as the immunodominant antigen for HSV-1 specific memory cytotoxic T lymphocytes. Journal of Immunology 135, 575-582. HILL, T. J., FIELD,H. J. & BLYTH,W. A. (1975). Acute and recurrent infection with herpes simplex virus in the mouse: a model for studying latency and recurrent disease. Journal of General Virology 28, 341-353. JOHNSON, D. C. & SMILEY,J. R. (1985). Intracellular transport of herpes simplex virus gD occurs more rapidly in uninfected cells than in infected cells. Journal of Virology 54, 682-689. KUMEL, G., KAERNER,H. C., LEVINE, M., SCHRODER, C. H. & GLORIOSO, J. C. (1985). Passive immune protection by herpes simplex virus-specific monoclonal antibodies and monoclonal antibody-resistant mutants altered in pathogenicity Journal of Virology 56, 93(V937. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, 680-685. LAWMAN,M. J. P., COURTNEY,R. J., EBERLE, R., SCHAFFER,P. A., O'HARA,M. K. & ROUSE,B. T. (1980). Cell mediated immunity to herpes simplex virus: specificity of cytotoxic T cells. Infection and Immunity 30, 451~461. LONG, D., MADARA, T. J., PONCE DE LEON, M., COHEN, G. H., MONTGOMERY, P. C. & EISENBERG, R. J. (1984).
1114
B.A.
B L A C K L A W S , A. A. NASH AND G. DARBY
Glycoprotein D protects mice against lethal challenge with herpes simplex virus types 1 and 2. Infection and
Immunity 43, 761 764. MINSON,A. C., WILDY, P., BUCHAN,A. & DARBY,G. (1978). Introduction of the herpes simplex virus thymidine kinase gene into mouse cells using virus D N A or transformed cell DNA. Cell 13, 581 587. NASH, A. A., FIELD, H. J. & QUARTEY-PAPAFIO,R. (1980a). Cell-mediated immunity in herpes simplex virus-infected mice: induction, characterization and antiviral effects of delayed type hypersensitivity. Journal of General Virology 48, 351-357. NASH, A. A., QUARTEY-PAPAFIO,R. & WILDY, P. (1980b). Cell-mediated immunity in herpes simplex virus-infected mice: functional analysis of lymph node cells during periods of acute and latent infection, with reference to cytotoxic and memory cells. Journal of General Virology 49, 309-317. NASH,A. A., LEUNG,K-N. & WILDY, P. (1985). The T-cell-mediated immune response of mice to herpes simplex virus. In The Herpesvh'uses, vol. 4, pp. 87-102. Edited by B. Roizman & C. Lopez. New York: Plenum Press. NASH, A. A., JAYASURIYA,A., PHELAN, J., COBBOLD,S. P., WALDMANN,H. & PROSPERO, T. (1987). Different roles for LDT4 ÷ and Lyt 2 + T cell subsets in the control of an acute herpes simplex virus infection of the skin and nervous system. Journal of General Virology 68, 825-833. NORRILD, B. (1985). Humoral response to herpes simplex virus infections. In The Herpesviruses, vol. 4, pp. 69-86. Edited by B. Roizman & C. Lopez. New York: Plenum Press. PAOLETTI, E., LIPINSKAS,B. R., SAMSONOFF,C., MERCER,S. & PANICALI,D. (1984). Construction of live vaccines using genetically engineered poxviruses: biological activity of vaccinia virus recombinants expressing the hepatitis B virus surface antigen and the herpes simplex virus glycoprotein D. Proceedings of the National Academy of Sciences, U.S.A. 81, 193 197. PF1ZENMAIER,K., JUNG, H., STARZINSKI-POWITZ,A., ROLLINGHOFF,M. & WAGNER,H. (1977). The role of T cells in anti-herpes simplex virus immunity. I. Induction of antigen-specific cytotoxic T lymphocytes. Journal of Immunology 119, 939 944. RECTOR, J. T., LAUSCH,R. N. & OAKES,J. E. (1982). Use of monoclonal antibodies for analysis of antibody-dependent immunity to ocular herpes simplex virus type 1 infection. Infection and Immunity 38, 168-174. REDDEHASE, M. J. & KOSZINOWSKI,U. H. (1984). Significance of herpesvirus immediate early gene expression in cellular immunity to cytomegalovirus infection. Nature, London 312, 369-371. RICHMAN,D. D., BUCKMASTER,A., BELL,S., HODGMAN,C. & MINSON,A. C. (1986). Identification of a new glycoprotein of herpes simplex virus type 1 and genetic mapping of the gene that codes for it. Journal of Virology 57, 647 655. SCHRIER, R. D., PIZER, L. I. & MOORHEAD, J. W. (1983). Type-specific delayed hypersensitivity and protective immunity induced by isolated herpes simplex virus glycoprotein. Journal oflmmunology 130, 1413-1418. SIMMONS,A. & NASH,A. A. (1984). Zosteriform spread of herpes simplex virus as a model of recrudescence and its use to investigate the role of immune cells in prevention of recurrent disease. Journal of Virology 52, 816-821. SIMMONS,A. & NASH,A. A. (1985). Role of antibody in primary and recurrent herpes simplex virus infection. Journal of Virology 53, 94~948. SIMMONS,A. & NASH,A. A. (1987). The effect of B cell suppression on primary infection and re-infection of mice with herpes simplex virus. Journal of Infectious Diseases (in press). SPEAR, P. G. (1985). Glycoproteins specified by herpes simplex virus. In The Herpesviruses, vol. 3, pp. 315-356. Edited by B. Roizman. New York: Plenum Press. TOWNSEND, A. R. M., McMICHAEL,A. J., CARTER,N. P., HUDDLESTON,J. A. & BROWNLEE,G. G. (1984). Cytotoxic T cell recognition of the influenza nucleoprotein and hemagglutinin expressed in transfected mouse L cells. Cell 39, 13-25. WAGNER, M. J., SHARP,J. A. & SUMMERS,W. C. (1981). N ucleotide sequence of the thymidine kinase gene of herpes simplex virus type I. Proceedings of the National Academy of Sciences, U.S.A. 78, 1441-1445. WATSON,R. J., WEIS,J. H., SALSTROM,J. S. & ENQUIST,L. W. (1982). HSV-1 glycoprotein D gene: nucleotide sequence and expression in E. coll. Science 218, 381 384. YEWDELL, J. W., BENNINK, J. R., MACKETT,M., LEFRANCOIS,L.,LYLES,D. S. & MOSS,B. (1986). Recognition of cloned vesicular stomatitis virus internal and external gene products by cytotoxic T lymphocytes. Journal of Experimental Medicine 163, 1529-1538. ZARLING,J. M., MORAN,P. A., BURKE,R. L., PACHL,C., BERMAN,P. W. & LASKY,L. A. (1986a). Human cytotoxic T cell clones directed against herpes simplex virus-infected cells. IV. Recognition and activation by cloned glycoproteins gB and gD. Journal of Immunology 136, 4669-4673. ZARLING, J. M., MORAN,P. A., LASKY,L. A. & MOSS,B. (1986b) Herpes simplex virus (HSV)-specific human T-cell clones recognise HSV glycoprotein D expressed by a recombinant vaccinia virus. Journal of Virology 59, 506509.
(Received 25 September 1986)
