Vaccination with HIV-1 gp120 DNA induces immune responses that are boosted by a recombinant gp120 protein subunit

0

Ekevier PII: so2f&4lox(%)oo264-2

ELSEVIER

Vaccine, Vol. 15, No. 6, pp. 669-673, 1997 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0264410X/97 $17+0.00

Vaccination with HIV-l gpl20 DNA induces immune responses that are boosted by a recombinant gp120 protein subunit Susan W. BarnetttS, Sabita Rajasekar?, Harold Leggt, Barbara Doe?, Deborah H. Fuller*, Joel R. Haynes*, Christopher M. Walkert and Kathelyn S. Steimert Small animals were immunized with plasmid DNA encoding HIV-l envelope gpI20 either intramuscularly by needle injection (mice and guinea pigs) or epidermally with the Accell@ gene gun (guinea pigs). Subsequently, the animals were boosted with a recombinant gp120 protein subunit vaccine in an oil-in-water based adjuvant, MF59. Antibodies and cytotoxic T-lymphocyte (CTL) immune responses to the HIV envelope glycoprotein were observed in animals immunized with gp120 DNA derived from the laboratory strain or from HIV-I field isolates. Titers of ELBA antibodies and HIV-I,, serum neutralizing antibodies against the HIV-I, laboratory isolate were substantially increased in DNA-immunized animals following a single boost with recombinant gp120 protein subunit. This DNA primelprotein subunit boost immunization approach may be important for vaccination against infectious agents such as HIVfor which it is dtflicult to raise strong antiviral humoral responses with DNA vaccination alone. 0 1997 Elsevier Science Ltd. Keywords: HIV; Env gp120:

DNA vaccine

Preclinical evaluation of HIV vaccination strategies indicates that novel approaches may be required to protect individuals from HIV infection and disease. There is growing evidence that such a vaccine strategy should target both the cellular and humoral arms of the immune system. While some studies suggest that virusspecific cytotoxic T lymphocytes (CTL) play a role in controlling the spread of HIV during early infection’, others indicate that antibody responses to human or primate lentiviruses are important forprotection against infection or disease in animal models . Early work in the field of nucleic acid vaccines has repeatedly demonstrated the effectiveness of this approach in the induction of humoral and cellular immune responses to immunogens from diverse infectious agents including HIV. Here we have confirmed and extended some of these observations for HIV. We show that boosting of animals primed with plasmid DNA encoding B genotype gp120 with recombinant gp120sFz (rgpl20,& subunit substantially elevates antibody responses to gp120 and dramatically increases titers of virus neutralizing antibodies against the HIV- 1sF7 *Geniva Inc., 8520 University Green, Middleton, WI 53562, USA. tChiron Vaccines, 4560 Horton Street, Emeryville, CA 94608-2916, USA.$To whom correspondence should be addressed.

laboratory isolate. Furthermore, this immunization approach results in the induction of strong CTL responses against the Env V3 epitope of these viruses.

MATERIALS

AND METHODS

HIV gp120 plastids and recombinant gp120 protein The pCMV6agp120sFz plasmid used for these DNA immunizations has been described”. The gp120 DNA sequences from the primary isolates, HIV-lCMZX5(herein referred to as CM235; Thai subtype E) and HIV-l,,, (US4: North American subtype B), were cloned into the same pCMV6a expression vector. Plasmid DNA was prepared from bacterial cultures by alkaline lysis and passage over Qiagen plasmid purification columns (Qiagen, Chatsworth, CA). Recombinant HIV-1sr2 gp120 (rgpl20,,,) subunit protein was produced in Chinese hamster ovary (CHO) cells and has been previously described4; rgp 120CM235protein was prepared by similar methods. Animal inoculations For the intramuscular (IM) immunizations, mice were injected bilaterally with 5Opg of DNA per site in the tibialis anterior (TA) muscles as described5; guinea pigs received IOOyg per site. For epidermal (ED)

Vaccine

1997 Volume

15 Number

8

869

Vaccination with HIV-1 gp120 DNA: S. W. Barnett et al.

immunizations, guinea pigs were inoculated using the Accell@ gene fun at 10 sites on the abdomen for a total dose of 1Opg . The HIV rgpl20,,, subunit protein was combined with MF59 adjuvant (5% squalene, 0.5% Tween 80, 0.5% Span 85; ref. 7) and injected bilaterally into the TA muscles (IM experiments) or into the quadriceps (gene gun experiment). Mice received a total dose of 12.5,ug, of protein and guinea pigs received 50 pug of protein per immunization.

1OOOOOO -)_

,..,... *

h

.z

100000 ----o---

52

SF2 SF2

DNA subunit

“84

DNA

SF2

subunit

CM235 CM235

A

DNA subunit

10000 -

5 5

1000 -

B .;

Measurement of antibody responses Enzyme-linked immunosorbent assays (ELISA) to measure gpl20-specific antibody titers were performed on sera collected at two week intervals by a modification of previously described methods4,8. Sera were tested at l/25 or l/l00 dilutions and by serial three- or four-fold dilutions, thereafter. The titers reported are the reciprocal of the dilution of serum that gave a half-maximum optical density (O.D.). Assays for serum neutralizing antibody activity against the HIV-l,,, laboratory strain were performed as previously described’; neutralizing antibody titers reported in Table 2 represent the reciprocal of the dilution at which a 50% inhibition of virus infection was observed. Measurement of cytotoxic T-cell lytic (CTL) activity Spleen cells from immunized BALB/c (H-2d) mice were cultured, restimulated, and assayed for CTL activity against Env V3 peptide-pulsed target cells as described’. The HIV- 1SF2 V3 peptide used was IGPGRAFHT, previously identified to comprise the minimal CTL epitope for this HIV-l strain in BALB/c (H-2d) mice’. The corresponding sequences chosen for the assays performed with cells from mice immunized with gpl20 DNA from the US4 and CM235 strains of HIV-l were IGPGRAFYA and IGPGQVFYR, respectively. Cytotoxic activity was measured in a standard “Cr release assay. The SVBALB (MHC matched, H-2d) and C57BW6 (mismatched, H-2b) fibroblast cell lines used as target cells express class I but not class II MHC molecules. Target (T) cells were cultured with effector (E) cells at various E:T ratios for 4 h and the average cpm from duplicate wells was used to calculate percent specific 5 Cr release as described’. RESULTS Antibody responses in mice immunized with HIV gp120 DNA by the intramuscular route are boosted by rgp120/MF59 protein subunit Groups of twelve BALB/c mice were immunized intramuscularly with 100 pg of plasmid DNA encoding Env gpl20 from three different strains of HIV-l. These viruses were the “laboratory isolate”, HIV-l,,, (SF2), and two field (or “primary”) isolates, HIV-l”,,, (US4) from subtype B, recovered from an individual in the United States, or HIV-1CM235(CM235), from subtype E, recovered from a person in Thailand. Four weeks after a single DNA injection, most animals exhibited low antibody responses to these immunogens, but by eight weeks, all animals showed evidence of an antibody response (Figure IA). At the eight-week time point, a second DNA dose was given. This resulted in 4-, 15- and 60-fold increases in geometric mean antibody titers 870

Vaccine 1997 Volume 15 Number 8

100 -

4 lot”

4

i

DNA

;I&

DNA

i”

20

24

28

rm120/MF59

_“““““”

b 2 .z looooo_---(t

SF2 SF2

DNA subunit

US4 SF2

DNA subunit

B

-I i DNA

4 t” i”WLfp120,~F59 2o3236 DNA

DNA subunit

Figure 1 Antibody responses to HIV-1 gp120 DNA are boosted by recombinant gp120 protein subunit. Mice (A) and guinea pigs (B) were immunized intramuscularly with plasmid DNA expressing Env gp120 from the indicated HIV-1 strains. Data shown represent geometric mean titers (GMT) and standard errors for each group of animals; 12 mice or 8 guinea pigs were included in each group. Arrows below graphs indicate timing of DNA immunizations; timing of the single rgpl20/MF59 protein subunit immunization is also indicated. The animals immunized with gp120 DNA from SF2 and US4 received the rgp120,,, protein; the mice immunized with gp120 DNA from CM235 received rgp120,,,,, protein. Sera from the SF2 and US4 groups were tested on ELISA plates coated with rgp120,, protein; for the CM235 group, plates were coated with rgp120,,,,,

(GMT) as compared to the levels seen after a single DNA immunization for the groups receiving CM235, US4, and SF2 plasmid DNAs, respectively. These titers remained fairly stable for an additional eight weeks thereafter. For example, SF2 DNA-immunized mice showed only a 50% reduction in GMT during this period. At the 16-week time point, mice were boosted once more, this time with an HIV-l subtype-specific recombinant gpl20 protein subunit adjuvanted with the novel oil-in-water emulsion, MF59. The mice immunized with SF2 and US4 plasmid DNA received HIV-l subtype B rgp120sF2 protein and the CM235 animals received subtype E rgp120,,,,, protein. Following the protein subunit boost, the GMT rose 50-, 32-, and 20-fold for the CM235, US4, and SF2 groups, respectively, relative to the titers observed following two DNA immunizations. All groups exhibited high antibody titers comparable to those observed following multiple subunit immunizations (not shown).

Vaccination with HIV-1 gp720 DNA: S. W. Barnett et al.

Table 1

Cytotoxic T-lymphocyte (CTL) responses nized intramuscularly with HIV-1 gp120 DNA

in mice immu-

Table2 Neutralizing antibody responses in guinea

pigs immunized with HIV-1 gp120 DNA by either the IM or ED (gene gun) routes are boosted by rgp120/MF59 protein subunit

Percent specific lysis of target cells

wgpl60 (positive control) pCMV6agpl 20sF2 DNA rgp12O,,,/MF59 subunit boost #l

#z pCMVGagpl20,s,DNA rgpl20,,JMF59 subunit boost #3

#4 pCMVagp1 20CMZa5DNA rgp1200M,,$MF59 subunit boost #5

E:T”

SVBALB None

SVBALB V3

MC57 v3

Group

Animal

post-DNAa

postsubunitb

5O:l 12.5:1 3:l

14 6 4

4ab 44 15

12 6

pCMV6agp120sF, DNA (IM) rgpl20,,,MF59 subunit boost (IM)

1 2 3 4

275’ 120 35 480

1100 300 2850 8550

5O:l 12.5:1 3:l


28 22 6


pCMV6agpl 20sF2 DNA (ED, gene gun) rgpl20,,,,MF59 subunit boost (IM)

5 6 7 8

1o-50* 1O-50 1O-50 1O-50

250 400 800 1800

5O:l 12.5:1 3:l


25 12 cl


pCMV6agpl 20us4 DNA (IM) ;gp? 20sFZ, MF59 subunit boost ‘JM)

9 10 11 12

25
1550 350 700 1900

5O:l 12.5:1 3:l


46 41 19


rgpl20,J (IM)

13 14 15 16

N.A.e N.A. N.A. N.A.

7150’ 2750 2500 6875

5O:l 12.5:1 3:l


16 6 Cl


5O:l 12.5:1 3:l

22 9 4

28 9 4

19 6

5O:l 12.5:1 3:l

19 11 6

18 6 2

13 6 2

aEffector: target ratio in 4 hour 5’Cr release assay; bSpleen cells from BALB/c mice immunized with gp120 DNA and protein from the designated HIV-1 strain or infected with a vaccinia virus expressing SF2 gp160 (vvgpl60) were restimulated with Env V3 peptide from the homologous HIV-1 strain. Lytic activity was assessed after 7 days of culture against SVBALB (H-2d, MHC matched) or MC57 target ceil lines sensitized with the same (H-2b, mismatched) peptide. Positive CTL responses are in bold.

Cytotoxic T-lymphocyte (CTL) responses in mice immunized intramuscularly with gp120 DNA Cytotoxic T-cell (CTL) activity was also measured in splenocytes recovered from the mice immunized intramuscularly with HIV gp120 DNA at 4 (SF2, US4 groups) or 17 weeks (CM235) following the protein subunit boost (Table 1). CTL responses were observed in mice that received either of the North American subtype B (SF2 and US4) plasmid DNA immunizations, but not in the group that received the subtype E CM235 gp120 DNA. Effector cells from the SF2 and US4 DNAimmunized animals exhibited specific lysis of Env V3 peptide-pulsed SVBALB target cells (H-2d, MHC matched) indicative of a CTL response. Target cells that were not peptide pulsed, or were pulsed with an irrelevant peptide (not shown), or were derived from an MHC-unmatched mouse strain (MC57 cells, derived from the H-2b C57B1/6 strain) were not lysed (Table I). In contrast to these results with the mice immunized with DNA plus subunit, animals that received multiple IM injections of rgp120s,,/MF59 protein subunit alone did not exhibit CTL activity (not shown), confirming previous studies’. Therefore, it is likely that the CTL responses observed in the experiment shown in Table 1 were due to the DNA immunizations and not the protein subunit boost. This has been independently confirmed in separate experiments in which groups of mice that received either IM or ED (gene gun) immuni-

MF59 subunit alone

“Sera collected at two weeks after the third DNA immunization with gp120 from indicated HIV-1 strains; bSera collected at two weeks after three gpl20 DNA immunizations plus a single rgpl20,,JMF59 boost; ‘Values represent reciprocal of highest dilution at which a 50% inhibition of HIV-1 sF2 replication in HUT-78 cells was observed. ~10 indicates no detectable neutralizing activity at lowest serum dilution tested; “This group was tested only at 1:lO and 1:50 dilutions at this time point; eN.A., not applicable; ‘Sera collected at two weeks after four IM immunizations with rgpl20,,dMF59 subunit protein vaccine

zations with pCMV6agp120,,? DNA alone in the absence of protein subunit boostmg also showed consistent CTL responses (not shown). Whether protein subunit immunization is able to boost CTL responses in mice previously immunized with DNA remains to be determined. While splenocytes from the mice with CM235 showed some lysis of target cells (especially at the 50: 1 E:T ratio) in these assays, this activity appeared to be non-specific and not representative of CTL activity (Table 2). The lytic activity observed was not MHC-restricted (lysis occurred in MHC-mismatched MC57 cells pulsed with V3 peptide) and it was not dependent upon stimulation of target cells with the CM235 Env V3 peptide. The failure of the CM235 gpl20-immunized BALBlc mice to exhibit CTL activity in this assay is most likely attributable to the sequence divergence of the potential V3 epitope of the Env glycoprotein of this HIV-l strain. There is a change from an arginine (found in the SF2 and US4 epitopes) to a glutamine at position 5 of the putative CM235 V3 epitope which would be an important anchor residue for MHC binding. Neutralizing antibody responses in guinea pigs immunized with HIV gp120 DNA by either the IM or ED (gene gun) routes are boosted with rgp120s&IF59 protein subunit Guinea pig studies were initiated to test the ability of the DNA plus protein immunization strategy to elicit neutralizing antibody responses. These animals provide larger serum volumes and exhibit fewer problems with background activity in our neutralization assays than do mice. Groups of eight guinea pigs were immunized with the SF2 and US4 gp120 plasmids by either 1) intramuscular injection with doses of 2OOpg of DNA at 0, 8

Vaccine

1997 Volume

15 Number

8

871

Vaccination with HIV- 1 gp120 DNA: S. W. Barnett et al.

and 12 weeks, or, 2) gold-particle mediated epidermal deliver of 10 microgram doses of DNA using the Accell”tl gene gun at 0, 7 and 14 weeks. All animals were subsequently boosted with 25 pg of rgpl20,,, protein in the MF59 adjuvant at 8 or 6 weeks after the third DNA immunization, respectively. Anti-gp 120 antibody responses were first detectable by ELISA after the second DNA immunization by either injection method, and did not appear to be boosted significantly by the third DNA dose (IM results, Figure ZB; gene gun results, not shown). As in the mouse studies, modest antibody titers were observed following the gp120 DNA immunizations and were boosted substantially (20- to 30-fold) following a single rgpl20,,, p rotein in MF59 immunization (Figure IB). Antibody titers in the guinea pigs immunized with three DNA immunizations followed by a single subunit boost were comparable to those seen in animals that received three or four immunizations with rgp120s,,/MF59 alone (not shown). Moreover, the elevation of antibody titers observed in response to a single subunit boost in these DNA-primed animals was sustained at increased levels above the baseline seen after the three DNA immunizations for up to 10 weeks post-subunit boost. Both the SF2 and US4 DNA immunized groups showed only a 50% reduction in GMT at 32 weeks (12 weeks post-subunit boost) compared to the 22 week time point (2 weeks post-subunit; Figure 1B). This contrasted with the relatively rapid decline in antibody titers following the subunit boost in the mice (Figure IA). This difference in the longevity of the elevated antibody response following subunit boosting may be attributable to the different animal species or dosing regimens used in these two experiments. Neutralizing antibody activity against the HIV- 1SF2 laboratory isolate was measured in a HUT-78 cell line-based neutralization assay*, using sera collected from guinea pigs before and after rgp120s,,/MF59 protein subunit boosting (Table 2). At two weeks following three DNA immunizations, most guinea pigs inoculated by either the IM or ED (gene gun) routes showed modest neutralizing activity against this HIV-1 isolate. At two weeks after the rgpl20,,, subunit boost, neutralizing antibody titers in all animals increased (2 to 200-fold), approaching those measured in control animals injected with four doses of subunit protein. Furthermore, immunization with plasmid DNA from the subtype B primary isolate US4 primed for crossstrain neutralizing activity against the heterologous SF2 isolate. After a single rgp120s,,/MF59 boost, the HIV1SF2 neutralizing antibody titers in the US4 group were similar to those seen in the guinea pigs primed with DNA derived from the homologous SF2 strain.

DISCUSSION These studies clearly demonstrate the ability of DNA immunization combined with protein subunit boosting to elicit vigorous cellular and humoral immune responses against HIV gp120 in small animals. This combination prime/boost strategy has the potential to overcome possible deficiencies that may arise using a single vaccination modality against HIV (for example, employing a DNA, protein subunit or live recombinant viral-vectored vaccine alone). Previous studies have shown that the CHO cell-derived glycosylated

872

Vaccine 1997 Volume 15 Number 8

recombinant gp120,,, protein subunit employed here was able to induce strong humoral immune responses in various species (2, 4, 8), but was not effective in the induction of Env V3-specific CTL responses in mice’. In contrast, the present study has shown that HIV gp120 DNA immunization consistently induced these CTL responses for B genotype viruses (Table 1; Barnett et al., unpublished results), but elicited only modest antibody responses (Figure 1) that correlated with weak neutralizing antibody responses in guinea pigs (Table 2). However, ELISA and neutralizing antibody responses in these animals were dramatically increased following a single protein subunit boost after which they showed neutralizing titers comparable to those observed in infected HIV-l individuals (not shown) and in control animals immunized with multiple subunit alone (Table 2). Furthermore, in related studies performed in Rhesus macaques, substantial increases in antibody responses to Env and Gag proteins were also observed in animals immunized with HIV-l gugpol and env DNA using the gene gun and subsequently boosted with both Env gp120/MF59 and Gag p24/MF59 protein subunit vaccines (Fuller et al., these Proceedings). For HIV, the evaluation of prime/boost strategies that employ nucleic acid vaccination is particularly relevant at this time. Efforts are currently in place to move forward into expanded clinical trials with a prime/boost strategy that employs a live Canarypox HIV vaccine (Pasteur-Merieux/Connaught) in combination with the rgpl20,,, protein subunit vaccine described here”. The ability of HIV DNA vaccines to effectively and reproducibly induce CTL responses may indicate their potential usefulness as a safe and less costly alternative to live viral vectors in similar prime/boost strategies.

ACKNOWLEDGEMENTS The authors would like to thank Louisa Leung for her assistance with the ELISA, Elizabeth Miller, Christine Safieddine, Yvonne Young and Charles Vitt for their assistance with the animal care and handling, Keith Higgins for help with assay development and animal protocols, Heather Davis for teaching us how to inject DNA into the TA muscle, and Julie Klinger for her critical reading of this manuscript.

REFERENCES Koup, R.A., Safrit, J.T., Cao, Y., Andrews, C.A., McLeod, G., Borkowski, W., Farthing, C. and Ho, D.D. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J. Viral. 1994, 66, 4650-4655 El-Amad, Z., Murthy, K.K., Higgins, K., Cobb, E.K., Haigwood, N.L., Levy, J.A. and Steimer, K.S. Resistance of chimpanzees immunized with recombinant gp120,,, to challenge by HIV-I.,,. ADS 1995, 9, 1313-l 322 Chapman, B.S., Thayer, R.M., Vincent, K.A. and Haigwood, N.L. Effect of intron A from human cytomegalovirus (Towne) immediate-early gene on heterologous expression in mammalian cells. Nucleic Acids Res. 1991, 15, 3979-3966 Haiawood. N.L.. Nara. P.L.. Brooks, E., Van Nest. G.A.. Ott. G.. Higgins, K.W., Dunlop, N.,‘Scandeila, C.J., Eichberg, J.W.‘and Steimer, K.S. Native but not denatured recombinant HIV-1 gp120 generates broad spectrum neutralizing antibodies in baboons. J. Virol. 1992, 66, 172-182 Davis, H.L., Whalen, R.G. and Demeneix, B.A. Direct gene transfer into skeletal muscle in vivo: Factors affecting efficiency of transfer and stability of expression. Human Gene Therapy 1993,4,151-159

Vaccination with HIV- 1 gp120 DNA: S. W. Barnett et al. 6

7

Fuller, D.H. and Haynes, J.R. A qualitative progression in HIV type 1 glycoprotein 120specific cytotoxic cellular and humoral immune responses in mice receiving a DNA-based glycoprotein 120 vaccine. AIDS Res. Hum. Retroviruses 1994, 11, 14331441 Van Nest, G.A., Steimer, K.S., Haigwood, N.L., Burke, R.L. and Ott, G. Advanced adjuvant formulations for use with recombinant subunit vaccines. In Vaccines 92, Modem Approaches to New Vaccines. Ed. by Brown, F., Chanock, FL, Ginsberg, H.S., Lerner, R. Cold Spring Harbor Laboratory Press, 1992, pp. 57-62.

8

9

10

Steimer, K.S., Scandella, C.J., Skiles, P.V. and Haigwood, N.L. Neutralization of divergent HIV-l isolates by conformationdependent human antibodies to gp120. Science 1991, 254, 105-l 08 Doe, B., Steimer, K.S. and Walker, C.M. Induction of HIV-1 envelope (gpl20)specific cytoxic T lymphocyte responses in mice by recombinant CHO cell-derived gp120 is enhanced by enzymatic removal of N-linked glycans. fur. J. Immunol. 1994, 24, 2369-2376 Cohen, J. Vaccine drought spurs NIAID Plan to improve industry ties. News and Comments. Science 1996, 271, 1227-l 228

Vaccine 1997 Volume 15 Number 8

873