Immunol. Cell Biol. (1990)68, 113-117
Response of monkeys to vaccination with recombinant vaccinia virus which coexpress HIV gpl60 and human interleukin-2 J. Ruby, C. Brinkman,* S. Jones,''' and I. Ramshaw Viral Engineering Group. Division of Cell Biology. John Curtin School of Medical Research. Australian National University. Canberra. ACT 2601. Australia. *P.sychology Department. Australian National University. Canberra. ACT 2601. Australia: and ^Commonwealth Serum Laboratories. Parkville. Vic. 3052, Australia (Submitted 31 July 1989. Accepted for publication 24 January
1990.)
Summary Immunization of two macaque monkeys with recombinant vaccinia viruses encoding the cnv gene of HlV-1 (VV-gpl60) resulled in demonstrable levels of gpl60, gpl20 and gp41-specific immunoglobulins in both animals. The virus used to immunize one ofthe monkeys additionally expressed Ihc human IL-2 gene, which encoded human [L-2 (VV-gpl60-IL-2). No toxic side-effects of vaecine-delivered IL-2 were observed. Despite marked attenuation of virulence by the coexpresscd lymphokine, the levels of vector-specific antibodies in both animals were similar. Some differences in the HI V-specifie reactivity patterns were detected. Serum reactivityof monkey #A56( VV-gp 160) was directed against gp41, whereas monkey #B58 (VV-gpl60-IL-2) showed a wider range of recognition, with higher antibody titres against the HIV lysate preparation. Furthermore, this study demonstrates the capacity to boost antibody responses to the vector and coexpresscd HIV antigens in primates which are already immune.
INTRODUCTION 11 has been shown previously that expression of niurine IL-2 by vaccinia virus (VV) allowed immunodeficient nude mice to resolve an otherwise lethal infeclioti with this virus (1). This observation has important implications for the engineering of viral delivery systems which can be safely administered to immunosuppressed as well as normal individuals. However, human IL-2 therapy for cancer patients has showti that this iymphokine can have toxic sideeffects, notably, vascular leak syndrome (2). The major objective ofthe present study was to test the toxicity of VV-encoded human IL-2 in a primate model. As an adjunct to extensive studies of these viruses in murine models, we assessed the potential ofthe VV-IL-2 vector to modify the humoral immune response to a coexpressed antigen. Recombinant human IL-2 has been shown to induce increased IL-2 receptor expresCorrcspondencc: Dr J. Ruby, Division of Ceil Biology. J.C.S.M.R., PO Box 334, Canberra. ACT 2601. Australia. .ibbreviations used in this paper b.p.. base pairs; HIV, human immunodeficiency virus; HRP. horseradish peroxidase; i.d., intradermal; IL-2. interleukin2; p.f.u. plaque-forming unit; VV, vaccinia virus.
sion and proliferative responses in T cells of macaque monkeys (3). In order to address these questions we have infected a monkey with recombinant VV which concomitantly encoded human IL-2 and the gpl60 envelope glycoprotein of HIV-1 (VV-gpl60-IL-2). A second monkey was infected with VV which expressed gpl60only (VV-gpl60). MATERIALS AND METHODS Construction of recomhinani vaccinia virus encoding gpl60 and human IL-2 The construction of recombinant VV encoding gpl60 (VV-gpl60) has been described elsewhere (4). VV-gpl6O-lL-2 was constructed by inserting the coding sequence for human lL-2 cDNA, excised from pcD-IL-2 (K. I. Arai. DNAX Research Institute. Palo Alto. CA). into the Hind HI F region of VV-gpl60. A 1kb Xho I fragment was first isolated from pcD-lL-2 and cut with Dra I/Rsa I; the 500 b.p. fragment isolated from an agarose gel was blunt-end ligated into pBC07. The EcoR! fragment of pBCO7/IL-2 was then isolated and ligated into the EcoRI site of pFB-x. Construction of VV-gpl6O-IL-2 was carried out using selection protocols detailed previously (5). Expres.sion of IL-2 hy VV-fipl60-IL-2 lL-2 expression was assessed by infection of 143B cells with VV-gpl60-lL-2. Human ostcosarcoma I43B
14
J, RUBY ET AL.
eells. a ihymidine kinase negative varianl of eel! line R970-5, originalcd from Dr K, Huebner (Wistar Institute. Philadelphia. PA) and were maintained as deseribed (5), Confluent monolayers of cells grown in 50 mm Pelri dishes were infected with VV-gpl6O-lL-2 at 5 p.f.u,/ccll. The supcrnalanls (2-5 mL) were collected 8 and 24 h later and filtered through 0-22 \ini filters to remove virus. IL-2 was quantilated as proliferation of CTI.t. cells relative to an international standard preparation of reeombinant human lL-2 (National Institute of Biological Standards and Control. Herts. UK). Inoculation of monkeys The monkeys used in this study were 7 year old male pig-tail macaques maintained al the J.C.S.M.K. Both monkeyswere inoculated with iCp.fu. intradermally (i.d.) on Ihe central dorsal region on day 0. Animal #B58 received the VV-gpl60-IL-2 tonstrucl and animal #A56 received the VV-gpl60 construct. The monkeys were reinocuiated with the same recombinanl VV on day 35(2 X IO''p,f,u, intradermally) and day 91 (total of 2 X lO^p.f.u.at four different sites subcutaneously). Blood was collected for measurement of scrum antibodies before the initial infection and on days 8. 20, 35. 49. 63 and 105. Detection of VV-specific antibodies Titres of VV-specific immunoglobulin were determined using an ELISA method. Microtitrc wells were coated wilh 1 O^p.f.u. of wild type VV overnight at 4°C. pH 9.6, Wells were incubated in triplicate with serial I wo-fold dilutions of monkey serum followed by borseradish peroxidase-conjugaled anti-monkey immunoglobulin (HRPO-anti-monkcy Ig. 1:1000. Nordic). Incubations were 90 min at room temperature. Plates were washed between incubations three times wilh phosphate-buffered saline containing 5% bovine serum albumin. The substrate used was o-phenylenediamine. Detection of HIV-specific antibodies .Antibodies agains! the HIV-I env gene products gpl60/gpl20 and gp41 were demonstrated on nitrocellulose strips coated with proteins from Jnaelivated HIV-I using the Western blotting procedure (Diagnostic Biotechnology Ltd. Singapore). Incubation with positive, inactivated human serum (1:100) showed bands corresponding to the major HIV antigens p I 5, p 1 7. p24. p55. p31. p53. p64. gp41. gp 120 and gpl60. Reactivity of monkey sera (1:10) was detected using Ihe procedure accompanying the strips and substituting HRPO-anti-monkey-Ig for the biotinylated goat anti-human IgG and avidin-HRPO steps used for the demonstration of human scrum reactivity. The ELISA procedure was used to measure antibody titres againsl a peptide orgp41 and HIV lysate proteins. The antigens used were prepared as follows: (i)gp41 peptide 12mer peplidc containing an octapeptide sequence derived from gp41 which is believed to represent anepitope which all sera that display antigp41 reactivity recognize. The peptide was conjugated
to a carrier (casein) and coated on to microlitre plates at a protein concentration of 0 2 |jg/well. (ii) HIV lysate prepared from purified HIV-1 virus grown al CSL and shown lo contain all the antigen species necessary for Western blot confirmatory testing. Microtitre plates were coaled al the same protein concentration as ihal used for gp41 peptide-carrier conjugate above, (iii) HIV/gp41 peptide prepared by mixing one part of antigen (i) with one part of antigen (ii) before being coated on to microtitrc plates al the equivalent final protein concentration used for the peptide and viral lysale ELISAs. Triplicate samples of serial two-fold dilulions of monkey scrum were incubaled in the wells, followed by a protein A-HRPO conjugale (SpA-HRPO. 1:200; Kirkegaard & Perry Laboratories Inc.. Gailhersburg. MD). Incubations were maintained over 60 min at room temperature. RESULTS IL-2 expression hy VV-gp160-1L-2 infected cells in vitro High levels of lL-2 activity could be detected in the supernatants of VV-gpl6O-IL-2 infected 143B cells within 8 h of virus infection. The supernatant obtained 24 h after infection contained the equivalent of 1500U of human IL-2. Growth of VV-gpl60-IL-2 The appearance ofthe lesions after initial infection indicated that the growth of VV-gpl6O-IL-2 was markedly attenuated. On day 5 the site of inoculation ofthe control virus VV-gp 160 (monkey #A56) was a swollen lump 4 0 cm in diameter. In contrast, the lesion at the inoculation site of VV-gpl6O-IL-2 (#B58) was 15 cm in diameter and no swelling was observed. Figure 1 illustrates the appearance ofthe lesions on day 8. On subsequent inoculations, the lesions were successively smaller with the lesion formed in response to infection with VV-gp 16O-1L-2 being always smaller. No side-effects of vector-delivered IL-2 were observed in the monkey infected with VVgpI60-lL-2 at any stage during the experiment. We looked for symptoms that could be associated with the vascular leak syndrome, such as obvious development of ascites. dyspnoea and respiratory distress. Other reported side-effects ofclinical lL-2 administration which were under consideration were fever, vomiting, diarrhoea or rash development. Each serum sample was tested for the presence of IL-2 activity in the CTLL assay but no activity was detected {results not presented).
MONKFY RESPONSE TO VACCINIA ENCODING HIV
AND IL-2
II?
FIR. I. Ihe appearance of vaccinia lesions on day 8. (a) Monkey #B58 was inoculated with VV-gpi60-lL2; (b) monkey #A56 was inoculated with VV-gpl60.
Antibody response lo vaccinia virus The levels of anti-VV immunoglobulin detected in both monkeys increased 8 days after primary inoculation, and continued to increase by day 20 (Table 1). This antibody pattern was maintained through day 35. After the second inoculation, the anti-VV levels of both animals increased by day 49. Antibody levels remained relatively constant despite a third inoculation. Antibody response to HIV The incubation of sera from the final bleeds (day 105) with Western blot strips ofthe major HIV-I proteins demonstrated reactivity against gpl60. gp 120 and gp41. This pattern of react ivity indi-
cates that the env gene product expressed by both VV constructs was correctly processed (Fig. 2). However, reactivity measured to a peptideof gp41 and to HIV lysate using ELISA techniques indicated some differences in the reaction patterns ofthe two monkeys (Table 1). On day 49 both monkeys showed increased titres against gp41 peptideand HIV lysate. The level of reactivity of monkey #A56 to gp41 peptide was higher on day 105. however, this reactivity was not increased by the addition of intact gp4l or other HIV lysate antigens. In contrast, sera from monkey #B58 showed increased reactivity against the HIV lysate preparation, but lower recognition ofthe gp4l peptide.
Table 1. ELISA reactivity of experimentally immune monkey sera against VV. an immunodominanl gp41 octapeptidc. and HlV-l lysate antigens Monkey
Day Inoeulation
A56
0
ELISA reactivity' Bleed
0
8
35
B58
20
91
35 49 63 105
0
0
35
35 49
8
20 63
91 105 *Endpoint titres for each serum (end
VV
Response of monkeys to vaccination with recombinant vaccinia virus which coexpress HIV gpl60 and human interleukin-2 J. Ruby, C. Brinkman,* S. Jones,''' and I. Ramshaw Viral Engineering Group. Division of Cell Biology. John Curtin School of Medical Research. Australian National University. Canberra. ACT 2601. Australia. *P.sychology Department. Australian National University. Canberra. ACT 2601. Australia: and ^Commonwealth Serum Laboratories. Parkville. Vic. 3052, Australia (Submitted 31 July 1989. Accepted for publication 24 January
1990.)
Summary Immunization of two macaque monkeys with recombinant vaccinia viruses encoding the cnv gene of HlV-1 (VV-gpl60) resulled in demonstrable levels of gpl60, gpl20 and gp41-specific immunoglobulins in both animals. The virus used to immunize one ofthe monkeys additionally expressed Ihc human IL-2 gene, which encoded human [L-2 (VV-gpl60-IL-2). No toxic side-effects of vaecine-delivered IL-2 were observed. Despite marked attenuation of virulence by the coexpresscd lymphokine, the levels of vector-specific antibodies in both animals were similar. Some differences in the HI V-specifie reactivity patterns were detected. Serum reactivityof monkey #A56( VV-gp 160) was directed against gp41, whereas monkey #B58 (VV-gpl60-IL-2) showed a wider range of recognition, with higher antibody titres against the HIV lysate preparation. Furthermore, this study demonstrates the capacity to boost antibody responses to the vector and coexpresscd HIV antigens in primates which are already immune.
INTRODUCTION 11 has been shown previously that expression of niurine IL-2 by vaccinia virus (VV) allowed immunodeficient nude mice to resolve an otherwise lethal infeclioti with this virus (1). This observation has important implications for the engineering of viral delivery systems which can be safely administered to immunosuppressed as well as normal individuals. However, human IL-2 therapy for cancer patients has showti that this iymphokine can have toxic sideeffects, notably, vascular leak syndrome (2). The major objective ofthe present study was to test the toxicity of VV-encoded human IL-2 in a primate model. As an adjunct to extensive studies of these viruses in murine models, we assessed the potential ofthe VV-IL-2 vector to modify the humoral immune response to a coexpressed antigen. Recombinant human IL-2 has been shown to induce increased IL-2 receptor expresCorrcspondencc: Dr J. Ruby, Division of Ceil Biology. J.C.S.M.R., PO Box 334, Canberra. ACT 2601. Australia. .ibbreviations used in this paper b.p.. base pairs; HIV, human immunodeficiency virus; HRP. horseradish peroxidase; i.d., intradermal; IL-2. interleukin2; p.f.u. plaque-forming unit; VV, vaccinia virus.
sion and proliferative responses in T cells of macaque monkeys (3). In order to address these questions we have infected a monkey with recombinant VV which concomitantly encoded human IL-2 and the gpl60 envelope glycoprotein of HIV-1 (VV-gpl60-IL-2). A second monkey was infected with VV which expressed gpl60only (VV-gpl60). MATERIALS AND METHODS Construction of recomhinani vaccinia virus encoding gpl60 and human IL-2 The construction of recombinant VV encoding gpl60 (VV-gpl60) has been described elsewhere (4). VV-gpl6O-lL-2 was constructed by inserting the coding sequence for human lL-2 cDNA, excised from pcD-IL-2 (K. I. Arai. DNAX Research Institute. Palo Alto. CA). into the Hind HI F region of VV-gpl60. A 1kb Xho I fragment was first isolated from pcD-lL-2 and cut with Dra I/Rsa I; the 500 b.p. fragment isolated from an agarose gel was blunt-end ligated into pBC07. The EcoR! fragment of pBCO7/IL-2 was then isolated and ligated into the EcoRI site of pFB-x. Construction of VV-gpl6O-IL-2 was carried out using selection protocols detailed previously (5). Expres.sion of IL-2 hy VV-fipl60-IL-2 lL-2 expression was assessed by infection of 143B cells with VV-gpl60-lL-2. Human ostcosarcoma I43B
14
J, RUBY ET AL.
eells. a ihymidine kinase negative varianl of eel! line R970-5, originalcd from Dr K, Huebner (Wistar Institute. Philadelphia. PA) and were maintained as deseribed (5), Confluent monolayers of cells grown in 50 mm Pelri dishes were infected with VV-gpl6O-lL-2 at 5 p.f.u,/ccll. The supcrnalanls (2-5 mL) were collected 8 and 24 h later and filtered through 0-22 \ini filters to remove virus. IL-2 was quantilated as proliferation of CTI.t. cells relative to an international standard preparation of reeombinant human lL-2 (National Institute of Biological Standards and Control. Herts. UK). Inoculation of monkeys The monkeys used in this study were 7 year old male pig-tail macaques maintained al the J.C.S.M.K. Both monkeyswere inoculated with iCp.fu. intradermally (i.d.) on Ihe central dorsal region on day 0. Animal #B58 received the VV-gpl60-IL-2 tonstrucl and animal #A56 received the VV-gpl60 construct. The monkeys were reinocuiated with the same recombinanl VV on day 35(2 X IO''p,f,u, intradermally) and day 91 (total of 2 X lO^p.f.u.at four different sites subcutaneously). Blood was collected for measurement of scrum antibodies before the initial infection and on days 8. 20, 35. 49. 63 and 105. Detection of VV-specific antibodies Titres of VV-specific immunoglobulin were determined using an ELISA method. Microtitrc wells were coated wilh 1 O^p.f.u. of wild type VV overnight at 4°C. pH 9.6, Wells were incubated in triplicate with serial I wo-fold dilutions of monkey serum followed by borseradish peroxidase-conjugaled anti-monkey immunoglobulin (HRPO-anti-monkcy Ig. 1:1000. Nordic). Incubations were 90 min at room temperature. Plates were washed between incubations three times wilh phosphate-buffered saline containing 5% bovine serum albumin. The substrate used was o-phenylenediamine. Detection of HIV-specific antibodies .Antibodies agains! the HIV-I env gene products gpl60/gpl20 and gp41 were demonstrated on nitrocellulose strips coated with proteins from Jnaelivated HIV-I using the Western blotting procedure (Diagnostic Biotechnology Ltd. Singapore). Incubation with positive, inactivated human serum (1:100) showed bands corresponding to the major HIV antigens p I 5, p 1 7. p24. p55. p31. p53. p64. gp41. gp 120 and gpl60. Reactivity of monkey sera (1:10) was detected using Ihe procedure accompanying the strips and substituting HRPO-anti-monkey-Ig for the biotinylated goat anti-human IgG and avidin-HRPO steps used for the demonstration of human scrum reactivity. The ELISA procedure was used to measure antibody titres againsl a peptide orgp41 and HIV lysate proteins. The antigens used were prepared as follows: (i)gp41 peptide 12mer peplidc containing an octapeptide sequence derived from gp41 which is believed to represent anepitope which all sera that display antigp41 reactivity recognize. The peptide was conjugated
to a carrier (casein) and coated on to microlitre plates at a protein concentration of 0 2 |jg/well. (ii) HIV lysate prepared from purified HIV-1 virus grown al CSL and shown lo contain all the antigen species necessary for Western blot confirmatory testing. Microtitre plates were coaled al the same protein concentration as ihal used for gp41 peptide-carrier conjugate above, (iii) HIV/gp41 peptide prepared by mixing one part of antigen (i) with one part of antigen (ii) before being coated on to microtitrc plates al the equivalent final protein concentration used for the peptide and viral lysale ELISAs. Triplicate samples of serial two-fold dilulions of monkey scrum were incubaled in the wells, followed by a protein A-HRPO conjugale (SpA-HRPO. 1:200; Kirkegaard & Perry Laboratories Inc.. Gailhersburg. MD). Incubations were maintained over 60 min at room temperature. RESULTS IL-2 expression hy VV-gp160-1L-2 infected cells in vitro High levels of lL-2 activity could be detected in the supernatants of VV-gpl6O-IL-2 infected 143B cells within 8 h of virus infection. The supernatant obtained 24 h after infection contained the equivalent of 1500U of human IL-2. Growth of VV-gpl60-IL-2 The appearance ofthe lesions after initial infection indicated that the growth of VV-gpl6O-IL-2 was markedly attenuated. On day 5 the site of inoculation ofthe control virus VV-gp 160 (monkey #A56) was a swollen lump 4 0 cm in diameter. In contrast, the lesion at the inoculation site of VV-gpl6O-IL-2 (#B58) was 15 cm in diameter and no swelling was observed. Figure 1 illustrates the appearance ofthe lesions on day 8. On subsequent inoculations, the lesions were successively smaller with the lesion formed in response to infection with VV-gp 16O-1L-2 being always smaller. No side-effects of vector-delivered IL-2 were observed in the monkey infected with VVgpI60-lL-2 at any stage during the experiment. We looked for symptoms that could be associated with the vascular leak syndrome, such as obvious development of ascites. dyspnoea and respiratory distress. Other reported side-effects ofclinical lL-2 administration which were under consideration were fever, vomiting, diarrhoea or rash development. Each serum sample was tested for the presence of IL-2 activity in the CTLL assay but no activity was detected {results not presented).
MONKFY RESPONSE TO VACCINIA ENCODING HIV
AND IL-2
II?
FIR. I. Ihe appearance of vaccinia lesions on day 8. (a) Monkey #B58 was inoculated with VV-gpi60-lL2; (b) monkey #A56 was inoculated with VV-gpl60.
Antibody response lo vaccinia virus The levels of anti-VV immunoglobulin detected in both monkeys increased 8 days after primary inoculation, and continued to increase by day 20 (Table 1). This antibody pattern was maintained through day 35. After the second inoculation, the anti-VV levels of both animals increased by day 49. Antibody levels remained relatively constant despite a third inoculation. Antibody response to HIV The incubation of sera from the final bleeds (day 105) with Western blot strips ofthe major HIV-I proteins demonstrated reactivity against gpl60. gp 120 and gp41. This pattern of react ivity indi-
cates that the env gene product expressed by both VV constructs was correctly processed (Fig. 2). However, reactivity measured to a peptideof gp41 and to HIV lysate using ELISA techniques indicated some differences in the reaction patterns ofthe two monkeys (Table 1). On day 49 both monkeys showed increased titres against gp41 peptideand HIV lysate. The level of reactivity of monkey #A56 to gp41 peptide was higher on day 105. however, this reactivity was not increased by the addition of intact gp4l or other HIV lysate antigens. In contrast, sera from monkey #B58 showed increased reactivity against the HIV lysate preparation, but lower recognition ofthe gp4l peptide.
Table 1. ELISA reactivity of experimentally immune monkey sera against VV. an immunodominanl gp41 octapeptidc. and HlV-l lysate antigens Monkey
Day Inoeulation
A56
0
ELISA reactivity' Bleed
0
8
35
B58
20
91
35 49 63 105
0
0
35
35 49
8
20 63
91 105 *Endpoint titres for each serum (end
VV
