DNA vaccination using expression vectors carrying FIV structural genes induces immune response against feline immunodeficiency virus

Vaccine,Vol. 15, No. 10, pp. 1085-1094, 1997 Q 1997 Elsevier Science Ltd. All rights resetved Printed in-Great Britain 0284-410)(/97 $17+0.00

PII: so264-41ox(!qMoo4-2

DNA vaccination using expression vectors carrying FIV structural genes induces immune response against feline immunodeficiency virus A.M. Cuisinier*, V. Mallet, A. Meyer, C. Caldora and A. Aubert Following inactivated virus vaccination trials, the surface glycoprotein gp120 of the feline immunode$ciency virus (FIV) was considered as one of the determinants for protection. However, several vaccination trials using recombinant Env protein OY some peptides failed to induce protection. To understand the role of the gp120 protein in vivo, we vaccinated cats with naked DNA coding for FIV structural proteins gp120 andpl0. We analyzed the ability of these vaccinations to induce immune protection and to influence the onset of infection. Injection in cat muscles of expression vectors coding for the FIV gp120 protein induced a humoral response. Cats immunized twice with the gp120 gene showed d@erent patterns after challenge. Two cats were, like the control cats, infected from the second week after infection onwards. The two others maintained a low proviral load with no modtjication of their antibody pattern. The immune response induced by gp120 DNA injection could control the level of viral replication. This protective-like immune response was not correlated to the humoral response. All the cats immunized with the gp120 gene followed by the ~10 gene were infected, like the control cats, from the second week but they developed a complete humoral response against viral proteins after challenge. Futhermore, they showed a sudden but transient drop of the proviral load at 4 weeks after infection. Under these conditions, one injection of the ~10 gene after one injection of the gp120 gene was not suficient to stimulate protection. On the contrary, after a period, it seems to facilitate virus replication. 0 1997 Elsevier Science Ltd. Keywords:

vaccine;

DNA

immunization;

FIV

Feline immunodeficiency virus (FIV) has been recognized as an important pathogen of the domestic cat. The virus causes severe immunodeficiency with a pathogenesis similar to human AIDS’,*. Genetically, FIV is more closely related to ungulate lentiviruses, such as Visna virus and CAEV, than to the primate lentiviruses SIV and HIV334. In order to study the mechanisms of the immune response and the major protective antigens, vaccination trials were carried out. Vaccination experiments with crude inactivated virus incorporated into ISCOMS showed an enhancement of infection in vaccinated cats in comparison to the control cats. The cats had high anti-Gag protein antibodies but no anti-Env protein antibodies showing that the immune response against Gag protein was insufficient to achieve protection. Thus, the external glycoprotein gp120 was considered as one of the determinants for protection5. Inactivated infected cells containing high amounts of Virbac Laboratories, BP 27, 06511, Carros, Cedex, France. *To whom correspondence should be addressed. (Received 13 May 1996; revised 22 November 1996; accepted 12 December 1996)

Env protein was then used to vaccinate cats. Protection against homologous FIV challenge was demonstrated when using the cell line FL4 to grow the virus637.On the contrary, growing the virus in other cell lines such as CrFk, 4201 or in the primary culture of thymocytes or PBL did not induce protection’. This vaccine formulation induced high virus neutralizing antibodies (VNA) which inhibit or reduce FIV multiplication in vitro or in viva’-’ ‘. The VNA were then recognized as playing an important role in protection. However, their role is controversial since inactivated infected cell vaccine was demonstrated to induce protection, without any VNA production, against TM2 primary FIV isolate propagating in cats’*. To evaluate the role of the gp120 protein, vaccination trials using Env protein were performed. However, they failed to demonstrate protection, and rather several different formulations of the recombinant protein induced enhancement of infection in vaccinated cats compared to the control cats1’3’3-‘6. In one specific case, the observed immuno-enhancement was transferred to naive cats with the plasma of vaccinated catsi7,‘* These experiments suggest that recombinant Env protein essentially induces a strong TH2-like

Vaccine

1997 Volume

15 Number

10 1085

DNA vaccination using expression vectors: A.M. Cuisinier et al. response which favors viral replication. A vaccination trial using recombinant gp120 surface proteins with adjuvant known to stimulate the THl-like immune response showed a transient reduction of the viral load during the first 8 weeks of infection”. On the basis of these experiments, the cellular immune response has to be stimulated to reach protection. Recently, a new vaccination approach was described which stimulates the cellular response by using injection of naked DNA”. The proteins are expressed endogenously leading to the presentation of antigens in natural form. Futhermore, the in viva expression of the proteins avoids the long and costly purification steps which could influence protein integrity. In the mouse model, injection of DNA coding for flu nucleoprotein” and hemagglutinin proteins”.” stimulated not only the humoral response but also a strong CTL response leading to total protection. Protection was also obtained with rabies glycoprotein coding DNA24. In the HIV model, it was shown that injection of expression vector coding for the Env proteins elicited immune responses in miceZ5-“. In macaques, DNA injection coding for SIV errv and gug genes stimulated CTL responses”. This technique is interesting to apply to the FIV model to study the in viva role of the gp120 protein in its physiological status. We vaccinated cats with DNA coding for gp120. We also injected the p10 gene coding for the nucleocapsid. This protein was chosen because it plays an early role in the pre-integration step of the FIV replicative cycle. Moreover, it is a conserved protein which probably induces an early cytolitic response. We studied the influence of these vaccinations on infection development to determine which proteins could be protective and which mechanisms could ensure protection.

MATERIAL

AND METHODS

Virus strain The FIV Gasser strain was isolated from a naturally infected asymptomatic cat”. This strain is a primary isolate and was only propagated once in vitro in activated FIV negative peripheral blood lymphocytes (PBL) to establish a challenge virus stock (titer 104.’ TCID?, ml-‘). The PBLs were activated with 10 pug mll of phytohemagglutinin M (Gibco-BRL) and 10 U ml-’ of human recombinant interleukin-2 (IL-2) in RPM1 1640 medium supplemented with 10% fetal calf serum (Gibco-BRL), 2% ultroser HY (GibcoBRL) and 5 PM B mercaptoethanol (Sigma). For neutralizing assay and viremia, the strain was propagated one more time in PBL (titer 104.’ TCIDSO ml- ‘).

Construction of expression vector A 457 base pair (bp) and a 1340 bp DNA fragments of FIV were amplified using, respectively, a primer pair specific for the p10 gene (bp 1713-1735 and 2149-2166) and a primer pair specific for the gp120 gene (bp 6207-6247 and 8270-8295). All the primers contained a Not1 restriction site. One microgram of purified FIV positive DNA was submitted to 30 cycles of PCR amplification in a DNA thermal cycler (PerkinElmer Cetus Norwalk, CN). The mixture reaction was composed of: 200 PM dNTP, 0.5 PM of each primer, (Biolabs). 1.5 mM MgCI,, 2 U of Taq Polymerase

1086

Vaccine

1997 Volume

15 Number

10

DNA samples were denatured at 94°C for 1.30 min, renatured at 55°C for 2 min and elongated at 72°C for 2 min. Amplified DNA fragments were separated by electrophoresis on a 2% agarose gel and purified by electroelution. The purified p10 and gp120 fragments and the pCMV vector (provided by Dr M. Suzan) were digested by NotI. After purification, the p10 and gp120 genes were ligated into the pCMV vector. The recombinant vector was then transferred into JMlOl bacteria by electroporation. The resistant colonies were screened for the gene incorporation by hybridization using the p10 and gp120 genes themselves as probes. Ten colonies, positive by hybridization and restriction analysis, were tested for p10 expression after transfection into COS cells by the DEAE method. Briefly, lo4 COS cells were incubated during 5 h with 20 ,ug of plasmid in serum free medium containing 8 mg of DEAE dextran. The cells were washed and incubated 48 h at 37°C in 5% CO?. After trypsinization, they were pelleted and lyzed in 1 ml PBS SDS 10% at 95°C for 10 min. The cell debris were eliminated by centrifugation 30 min 35000 revs min-’ in a 45Ti rotor (Beckman). The proteins contained in the supernatant were precipitated with one volume of acetone. After drying the pellet, the proteins were immediately dissolved in 1 x loading buffer. One tenth of the protein samples was loaded on a 5-20% acrylamide gel (Biorad) and submitted to electrophoresis. After transfer to a nylon membrane (Biorad), the proteins were analyzed by the Western Blot technique. The membrane was incubated with an FIV positive serum diluted l/50, then washed and incubated with peroxydase conjugated goat anti-cat IgG diluted l/1000 (Jackson). The strips were washed and stained with luminol substrate (NEN). Positive and negative controls were included in each Western Blot.

Vaccination protocol Twelve SPF cats from Harlan US pedigree were divided into three groups. One group (cats 14) was immunized twice with one intramuscular (i.m.) injection into each shin of 300 ,ul (300 pg) of pCMV-120 vector (clone 2). The second group (cats 5-8) was immunized first with one i.m. injection in each shin of 300 ~1 (300 pg) of pCMV-120 vector followed by one i.m. injection in each shin of 300 ~1 (300 pug) of pCMV-10 (clone 4). The third group (cats 9912) was the infection control group and received in the same condition the pCMV vector. Each immunization was performed at 15 day intervals. Two weeks after the last immunization. all the cats were challenged intraperitoneally with 100 TCID,, of Gasser strain (which would be the approximate equivalent of 5 CID,,unpublished results). Every 2 weeks, blood was drawn from cats and the humoral response, as well as the cat infectious status were determined.

Detection of antibodies specific for Env and Gag proteins The antibody specificity of the sera was examined by the Western Blot technique using Escherichia coli expressed recombinant proteins: ~26: ~50, the unglycosylated form of gp120; Dx, the extracellular domain of gp41 protein corresponding to TM213 region” and the baculovirus expressed recombinant protein gp80, the

DNA vaccination using expression vectors: A.M. Cuisinier et al. glycosylated form of recombinant gp120. Twelve percent SDS-PAGE was carried out on slab gels according to the Laemmli method. The purified recombinant proteins were separated by electrophoresis and transferred to nitrocellulose membrane paper (Biorad). The blots were,blocked overnight in PBS containing 5% of milk. Serum samples were diluted l/50 and then processed as described in paragraph 2. Positive and negative controls were included in each Western Blot. The titer of the different antibodies was evaluated by ELISA. Briefly, microtiter plates were coated overnight with the corresponding protein dissolved in ELISA coating buffer (0.1 M Na,CO, pH=9.6). Wells were washed and blocked with 10% BSA in TNE buffer. They were incubated for 2 h at 37°C with 200 ,ul of twofold dilutions of the cat sera. Then the plates were washed and incubated for 1 h at 37°C with 200 ~1 of l/1000 diluted peroxydase conjugated goat anti-cat IgG (Jackson). After washes, 100 ~1 of ABTS substrate (Sigma) was added. The reaction was stopped by 50 ~1 of acid solution. The optical density at 650 nm was read in an ELISA photometer (SF400 SFL lab instrument). Supernatants from non-infected and infected cells were used as negative and positive controls, respectively. The sensitivity of this ELISA is similar to commercially available kits.

Neutralization assays One hundred and fifty microliters of twofold dilutions of decomplemented sera from control and infected cats were incubated 1 h at 37°C with 150 yl of l/70 diluted Gasser virus stock (equivalent to 100 TCID,,). Fifty microliters of this mixture were added to 2 x lo4 noninfected PBL and then incubated for 10 days at 37°C in a 5% COZ atmosphere in complete medium. A modified version of the ELISA described by Dandekar et a1.30 was employed to detect FIV p26 antigen in the culture supernatant. Briefly, microtiter plates were coated with the monoclonal antibody FC89 specific for the Gag protein (provided by the CBC Corporation) dissolved in ELISA coating buffer. Wells were washed and incubated first for 2 h with 200 ~1 of culture supernatant diluted 1:2 and then 20 min with 200 ~1 of l/500 of biotylinated anti-FIV p26 monoclonal antibodies 431B9 and 431E2 (provided by N. Pedersen). After washes, the plates were incubated for 20 min at 37°C with 200 ~1 streptavidin peroxidase (l/1000) followed by an incubation with 100 ~1 of TMB substrate (Sigma). The color development was stopped by 50 ~1 of l/200 hydrofluoric acid solution. The optical density at 650 nm was read in an ELISA photometer. The sensitivity of our ELISA is the similar to the IDDEX FIV antigen test kit commercially available. Supernatants from non-infected and infected cells were used as negative and positive controls, respectively. The serum was considered positive for neutralizing antibodies when the O.D. was < 1.5. The VNA titer was calculated by the Spearman and Karber method.

Viremia Two hundred microliters of twofold dilutions of filtered sera were incubated for 21 days at 37°C in a 5% CO, with 2 x lo4 FIV negative PBL in complete medium. Medium was replaced each week. Virus

production was measured by the detection Gag protein in the supernatant as described

of the p26 above.

Virus isolation PBLs from infected and non-infected cats were isolated from 3 ml of blood using Ficoll-Paque gradient (Pharmacia). Estimated PBL recoveries by this cedure were 4 x lo6 to 20 x lo6 per 3 ml. Five 10YOof these PBL were co-cultivated for 21 days at 37°C in a 5% CO, atmosphere with 5 x lo5 PBL of a FIV negative cat. The PBL were fed once a week with 5 ml of complete medium. After 2 weeks of culture, 5 x lo5 fresh FIV negative PBL were added. Each week, the presence of virus in the co-culture supernatant, was monitored for the presence of p26 Gag protein by an ELISA as previously described.

Provirus detection by PCR The DNA was extracted from lo6 freshly isolated PBL. After washing in sterile PBS, lymphocytes were resuspended in 500 ~1 of lysis buffer (1% SDS; 0.1 M NaCl; 0.1 M Tris pH=S; 1 mM EDTA and K) and incubated 30 min at 300 pug ml-’ proteinase 65°C. Estimated DNA recovery from lo6 PBL was 5 pg. DNAs from non-infected or infected cells were used, respectively as negative or positive controls. To quantify the number of proviruses in cells, we developed a quantitative PCR with an internal standard. The competitor was the 70 bp deleted TM213 region cloned in the Bluescript KS vector. It shared the same primer recognition sequences as the sample (bp 81808204 and 8548-8588). The PCR amplification was performed in the conditions described above, using l-0.3 ,ug of purified genomic DNA (depending on the time of infection). Tenfold competitor dilutions were from 1 to 1000 copies. Amplified DNA fragments were separated by electrophoresis on a 2% agarose gel and transferred to a nylon membrane. Southern Blots were hybridized overnight at 65°C with a specific “‘P-labeled probe. The titer was determined by a densitometer reading (Pharmacia). The sensitivity of this technique could be considered to be one FTV provirus copy among 2.5 x lo5 cells.

RESULTS Expression of env and p10 proteins after transfection with env and p10 expression vectors The gp120 and p10 genes from the FIV Gasser strain were cloned into the pCMV plasmid directly under the control of the CMV promotor. The gp120 gene was inserted with its own leader sequence. The p10 gene was cloned without any leader sequence. The expression of the corresponding proteins was assayed by Western Blot analysis of crude protein extract from transiently transfected COS cells (Figure 1). A clear band was visualized respectively at 12-15 kDa for p10 and at loo-120 kDa for gp120 (lane 3 Figure la, lanes 2 and 3 Figure lb). which was not present in the negative control lysates (lanes 1, 2, 4 Figure la and lane 1 and 4 Figure Zb). The low level expression observed for the 2 viral proteins could be due to a low number of transfected cells and/or to a non-optimized construct.

Vaccine

1997 Volume

15 Number

10 1087

DNA vaccination using expression vectors: A.M. Cuisinier et al.

PM

33 28

15 kD

PM

120

-

205

-

140

-

80

-

45

-

32

-

18

kD

2

3

4

5

Figure 1 gp120 and ~10 protein expression in COS cells. COS cells were transiently transfected by the DEAE protocol. After cell lysis, the protein extract was analyzed by Western Blot. Proteins were detected with a 58 week infected cat serum and revealed by a chimioluminescence process. Cell lysate from COS cells transfected by the pCMV vector was used as a negative control. fnactivated virus or recombinant p10 protein were used as a positive control of the Western Blot analysis. (a) pl0 expression: lane 1, non-transfected cells; lane 2, pCMV transfected cells; lane 3, pCMV-10 clone 4 transfected cells; lane 4, pCMV-10 clone 2 (anti-sens) transfected cells; lane 5, recombinant p10 protein. (b) gp120 expression; lane 1, pCMV transfected cells; lane 2, pCMV-I 20 clone 2 transfected cells; lane 3, pCMV-120 clone 7 transfected cells; lane 4, pCMV-120 clone 5 (anti-sens) transfected cells; lane 5, inactivated virus

Vaccination Humoral response induced by DNA immunization. The vaccination intake was checked by determination of humoral response by ELISA and Western Blot assays (Figure 2 and Table I). After one injection of pCMV-120 plasmid, two out of eight vaccinated cats (cats 2 and 3 in gp120 group) showed an anti-gp120 response visualized only by Western Blot (Table I). A second injection was necessary to induce a clear response visualized by Western Blot and ELISA in all four cats of the gp120

1088

Vaccine 1997 Volume 15 Number 10

group (Table 1 and Figure 2A, respectively). On the contrary, in the gpl2O+plO group (Figure 2B), anly one cat (No. 7) showed a low anti-gp120 antibody on the challenge date (titer=50). The second injection of pCMV-120 is thus necessary to boost the anti-gp120 humoral response and consequently induce a significant humoral response (Table I). No neutralizing antibodies against the FIV Gasser strain were detected the day of challenge (data not shown). The p10 gene in the pCMV-10 construct was cloned without any leader sequence. Consequently, its

d

A

38. m

I

T2

. T4

T6__

T6 TlO

TO

T2

T4

Tb

T8

TIO

weeks post challenge

TO

T2 T4 T6 T8 weeks post challenge

control group

TlO

Figure 2 Detection of antibodies specific for FIV gp120 by ELBA. Sera obtained from all animals before and after infection were tested for anti-gpl20 antibodies by ELBA using E. colirecombinant gp120 protein. The titer is the last dilution giving an O.D. threefold superior to the negative sera. Sera from non-infected and infected cats were used as negative and positive controls, respectively

weeks post challenge

0 !-

500

1500

2000

2500

titer 3000

500

3

500

TO

gp120 f p10 group

1000

1500

2000

2500 1

titer 3000

B)

1

01

m2 83 EI4

1000

4-

gp120 group

1000

1500

2000 t

2500

titer 3000

9)

DNA vaccination using expression vectors: A.M. Cuisinier et al. Table 1

Kinetics of appereance

of antibodies

specific for Env and Gag proteins

Weeks before and after infection Group

gpl20

gp12O+plO

Control

Antigen

T-4

T-2

TO

T2

T4

T6

T8

TlO

gpl20 ~26 PlO TM213

0 0 0 0

233 0 0 0

1, 2, 3, 4 0 0 0

1, 2, 3, 4 0 0 0

1, 2, 3, 4 0 0 0

1, 2, 3, 4 0 0 0

1, 2, 3, 4 0 0 3, 4

1, 2, 3, 4 0 0 3, 4

gpl20 ~26 PlO TM213

0 0 0 0

0 0 0 0

7 0 0 0

7 0 0 0

7 0 0 0

5, 6, 7 0 0 5, 6, 7

5, 6, 7, 8 0 0 5, 6, 7, 8

5, 5, 5, 5,

gpl20 ~26 PlO TM213

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

9. 11 11 0 9, 11

9, 11, 12 11 0 9, 11, 12

9, 10, 11, 12 11 0 9, 10, 11, 12

9, 10, 11, 12 11 0 9,10, 11, 12

6, 6, 6, 6,

7, 8 7 7 7, 8

lmmunoblots using E. co/i expressed recombinant proteins: ~26, the capsid protein, p50 (the unglycosylated form of gp120) and TM2/3, the extracellular domain of gp41 protein (E. Young 1992) and baculovirus recombinant protein gp80 (the glycosylated form of gp120) were reacted with sera from cats. The blots were then incubated with 111000 diluted peroxydase conjugated goat anti-cat IgG and stained with luminol substrate (NEN). A positive and negative control were included in each Western Blot. The results are presented as the number of positive cats

expression was cytoplasmic. Futhermore, the p10 protein being a poor immunogen, no anti-p10 antibody csponse was expected after one injection of pCMV- 10 plasmid. The ELISA and Western Blot were negative in the gp12O+plO group (Table I). All the cats Hurnorcd response after FIV challenge. were challenged with 100 TCID,,, of the Gasser strain 15 days after the last immunization. After challenge, no boost of the anti-gp120 response was observed for the gp120 vaccinated cats. This response was low and stable throughout the complete experiment (Figure 2A). On the contrary, the four cats receiving gp12O+plO vaccine showed no (cats 5, 6 and 8) or a low response (cat 7) during the first 6 weeks after infection. Then, all the cats presented a dramatic increase in their anti-gp130 response (Figure 2B). In parallel, the control cats had a significant anti-gpl20 response starting at week 6 which increased slowly except in the case of cat 11 (Figure 2C). We also studied the humoral response against the FIV Gag proteins p10 and ~26. The results from Western Blot and ELISA analyses are summarized in Table 1. None of the cats from the gp120 vaccinated group showed a humoral response directed against the Gag proteins throughout the experiment. Only one cat in the control group (cat 11) had an anti-p26 response. No cats from this group presented an anti-p10 response. On the contrary, three out of the four cats (Nos 5, 6 and 8) in the gp12O+plO group had an anti-p10 and anti-p26 humoral response at 10 weeks of infection. The humoral response against the TM2/3 region of the gp41 protein, an immunodominant domain3’, which correlates with infection was also analyzed (Tuble I). All the control cats and gp12O+plO vaccinated cats were positive for anti-TM2/3 antibodies at week 8 after infection. In the gp120 group, only two out of four cats (Nos 3 and 4) presented anti-TM2/3 antibodies at week 8. The two other cats in this group (Nos 1 and 2) had no anti-TM213 antibodies throughout the experiment. Based on this data, all the cats except cats 1 and 2 were infected. The neutralizing antibodies appeared only at week 8 after challenge with low titers and similar between cats (data not shown). They were most probably due to the

1090

Vaccine 1997 Volume 15 Number IO

Table 2

Virus isolation after infection Weeks post challenge

Group

TO

T2

T4

T6

T8

gpl20 gp12o+p10 Control

0 0 0

0 0 0

0 7 11, 12

3 5, 6, 7 9, 11, 12

3, 4 5, 6, 7, 8 9, 10, 11, 12

5~10~ PBLs of infected cats were mixed with 5x10s non-infected cells and cultivated in vitro during 21 days. The production of virus in PBL culture supernatant was analyzed by p26 ELISA detection. The results are presented as the number of positive cats

infection because all the control cats presented neutralizing antibodies whereas only one cat in each vaccinated group (respectively cat 4 in the gp120 group and cat 7 in the gp12O+plO) had neutralizing antibodies. Infectiosis status of the cats. Virus isolation confirmed the anti-TM2/3 response data. All the cats except 2 from the gp120 group (cats 1 and 2) were positive for virus isolation at week 8 after infection (Table 2). No significant difference in the time of infection between the control and gp12O+plO groups was observed. There was a correlation between the time of appearance of the anti-TM2/3 antibodies and the time of positive virus isolation. The virus isolation technique seems, however, to be more sensitive than the antibody detection since the virus is isolated before the anti-TM2/3 antibodies appear. Virus load was determined by detection of free virions able to infect cells after incubation of sera with noninfected cells. By this means, only free infectious viruses can be detected. We could not detect viruses coated on cells, mutant viruses or viruses linked to antibodies. The virus load detection was transient and so low that we could not calculate a titer. As a result, the data were presented as the number of positive cats (Table 3). Some cats were viremic only once during the infection analysis. No difference in the time of positivity or in the number of viremic cats could be observed among the three groups. No correlation with the virus isolation or the TM2/3 antibodies could be determined.

DNA vaccination using expression vectors: A.M. Cuisinier et al. Table 3

Viral load after infection Weeks post challenge

Group

TO

T2

T4

T6

T8

gpl26 gp12O+plO Control

0 0 0

0 0 0

3 0 0

3 5, 7 10.12

4 5, 8 10, 11, 12

Twofold dilution of sera was incubated with non-infected PBL for 21 days. The FIV p26 antigen produced in the supernatant is an indicator for the presence of FIV infectious virus in sera. Supernatant from non-infected and infected cells were used as negative and positive controls respectively. Viral load was analyzed every 2 weeks throughout the experiment. The results are presented as the number of positive cats

The proviral load was also determined by a competitive and quantitative PCR on DNA using env primers. Despite some differences, results demonstrated that all the vaccinated cats had a lower proviral load than the control cats (Figure 3). The control cats (Figure 3C) rapidly rose to a plateau 2 weeks after challenge with an average of 500 copies per lo6 cells. The gp120 vaccinated cats showed a heterogeneous pattern (Figure 3A). Two out of four cats (nos 3 and 4) showed a slow increase to reach a proviral load similar to that of the control group. One cat (no. 1) which was also negative for anti-TM2/3 response, virus isolation and viral load, had a low proviral load throughout the experiment (five copies per lo6 cells). The other one (no. 2) had a transient high proviral load which decreased to five copies per lo6 cells. In the gp12O+plO group, as in the control group, the viral load increased rapidly (Figure 3B). However, at week 4 after challenge a significant decrease was observed. Then the proviral load increased again to reach a plateau (average of 250 copies per lo6 cells). This sudden decrease could not be correlated with the humoral response which appeared at only week 6 when the provirus load increased again.

DISCUSSION Injection in cat muscles of mammalian expression vectors coding for the FIV gp120 protein induced a humoral response in cats. This response was, however, low in comparison to a classical protein vaccination protocol. We need two immunizations to stimulate a clear humoral response. This low humoral response could be explained by a faint protein expression encoded by the inoculated DNA. Furthermore, we did not use any muscle preparation or mechanic delivery to increase DNA uptake by cells. However, DNA expression seems to depend on the introduced gene. Certain recombinant proteins are difficult to express in an in vitro system regardless of the technical improvement used as in the case of the gp120 protein of different lentiviruses32 (D. Bishop communication). Increasing our vaccinal efficiency could be achieved by improving constructions such as using a strong leader sequence or a specific muscle promotor to increase the protein expression or an intronic sequence to stabilize the RNA transcripts. We tested the DNA vaccination technique in the FIV model to analyze the in vivo ability of two different proteins gp120 and ~10, presented in their physiological status, to induce protection. Cats immunized twice with the gpl20 gene showed different patterns after challenge.

Two cats were infected from the second week. They had an anti-TM2/3 response. We were able to isolate the virus from their PBL. The proviral load deduced from PCR analysis was the same as in control cats. In the two other cats, we could not detect the virus by viremia or co-culture assays throughout the experiment. However, these cats were considered infected because we could detect a low level of provirus in PBL. The immune response induced by DNA injection could control the level of viral replication or select a mutant virus with a poor replicative ability. This protective-like immune response was not correlated to the humoral response since these two cats presented neither anti-Gag antibodies nor a neutralizing response. The low neutralizing response could not be explained by anti-p26 antibodies interfering with our revelation system33534 because all the sera were clearly and significantly negative for anti-p26 antibodies during the first 8 weeks of infection. We cannot ignore the fact that these two cats were more resistant to infection than the other ones. Their condition was very similar to that of seronegative cats housing with FIV seropositive cats30. One injection of the p10 gene after one injection of the gp120 gene was not sufficient to induce protection. The cats belonging to the gp12O+plO group showed a pattern similar to that of the control group concerning viremia, virus isolation and the anti-TM2/3 antibodies. However, three out of four cats developed a complete humoral response against Gag proteins at the end of the trial. On the contrary, only one control cat developed anti-p26 antibodies. The humoral response directed against p10 appeared on an average of 20 weeks after the early experimental infection35. In our experiment, detection of p10 antibodies (week 10 after challenge) in the gp12O+plO group suggested that priming of the immune response occurred after the p10 gene inoculation. More probably, this complete response against Gag proteins added to the strong anti-gp120 humoral response in this group could be the result of an immunoenhancement. Indeed, the proviral load of these cats was increasing while this complete humoral response developed. The exact mechanism of antibody participation in viral dissemination is not yet understood. In the HIV model, some works described the participation of complement activating antibodies which could increase the efficiency of virus entry and could expand the virus tropism to cell bearing low or no CD4 receptors36. A transient drop of the proviral load was observed in the gp12O+plO group at 4 weeks after infection. It could be due to ADCC or CD8 cells. In the HIV model, ADCC was described as an important tool to fight viral disease37p39. However, in our model the drop of the proviral load was not correlated to the ADDC response because no humoral response was developed at this moment. Whether the drop of the proviral load could be due to a cellular response remains to be determined. DNA vaccination with the FIV gp120 gene did not induce a neutralizing response. In fact, the neutralizing response was only due to viral replication and appeared at the eighth week after challenge. However, the in vitro technique for detecting neutralizing antibodies did not represent a good tool for determining the in vivo neutralizing antibodies. In fact, neutralizing antibody detection depends on the sensitivity of the technique used33734. Furthermore, only the passive transfer of sera inducing protection against challenge could really

Vaccine 1997 Volume 15 Number 10 1091

T8 TO

T2

T4

T6

T8

weeks post challenge

weeks post challenge

1

T6

1

T4

10

10

T2

100

100

100

1

10

1000

10000 1000

titer

1000

titer

gp120 + p10 group

10000

TO

B)

2 q4

03

n2

01

10000

titer

gp120 group

TO

T4

T6

weeks post challenge

T2

control group

T8

c

EI12

n l1

HlO

q9

Figure 3 Proviral load. DNA of infected cats was extracted. PCR amplification was carried out from 1 pug of purified DNA in the presence of l-1000 copies of competitor (94°C 1.30 min, 54°C 2 min and 72°C 2 min). Amplified DNA fragments were separated by electrophoresis and transferred to a nylon membrane. Southern Blots were hybridized overnight at 65°C with a specific 3zP-labeled probe. The titer (copy number for 1O6 cells) correspond to the signal of the competitor reduced by 50%. *Represent non-determined data

U

DNA vaccination using expression vectors: A.M. Cuisinier et al.

measure the in viva neutralizing response. Whatever, our data did not support a correlation between neutralizing antibodies and protection as already proposed’* and in contrast to the results of Yamamoto’s group’,“. Injection of DNA coding for some FIV proteins is efficient to induce an immune response even if some aspects of DNA vaccination are still misunderstood. For example, the precise mechanism underlying DNA uptake and expression as well as the duration of the protein expression upon DNA injection remain to be established. However, DNA injection is a good tool to screen which proteins play a role in inducing the protection. It also permits an understanding of the mechanisms liable to play a role in protection.

ACKNOWLEDGEMENTS We wish to thank Dominique Grousson, RCgis Petitjean and Thierry Crosio for their technical assistance with the cats. We also thank Marie Suzan for her fruitful discussion and Jane de Vaugelas for proofreading the manuscript.

REFERENCES 1

2

3

4

5

6

7

8 9

10

11

12

13

14

Bendinelli, M., Pistello, M. and Lombardi, S. et al. Feline immunodeficiency virus: an interesting model for AIDS studies and an important cat pathogen. C/in. Microbial. Rev. 1995, 8, 87-l 12 Pedersen, N.C. Feline immunodeficiency virus infection. In: Animal Mode/ in AlDS (Eds Schellekers, H. and Horzinek, M.C.). Elsevier, Amsterdam, 1991, pp. 165-l 83 Elder, J.H. and Phillips, T.R. Molecular properties of feline immunodeficiency virus. infectious agents Dis. 1994, 2, 361374 Olmsted, R.A., Hirsch, V.M., Purcell, R.H. and Johnson, P.R. Nucleotide sequence analysis of feline immunodeficiency virus: genome organization and relationship to other lentiviruses. PNAS 1989,86,8088-8092 Hosie, M.J., Osborne, R., Reid, G., Neil, J.C. and Jarrett, 0. Enhancement after feline immunodeficiency virus vaccination. Vet. Immunol. Immunopathol. 1992, 35, 19t-197 Yamamoto. J.K.. Okuda. T. and Acklev. C.D. et a/. Experimental vaccine protection against feline immunodeficiency virus. A/DS 1991,7,911-922 Yamamoto, J.K., Hohdatsu, T. and Olmsted, R.A. et a/. Experimental vaccine protection against homologous and heterologous strains of feline immunodeficiency virus. J. Viral. 1993, 67, 601-605 Hosie, M.J. The development of a vaccine against feline immunodeficiency virus. Br. Vet. J. 1994, 150, 25-39 Hohdatsu, T., Pu, R., Torres, B., Trujillo, S., Gadner, M.B. and Yamamoto, J.K. Passive antibody protection of cats against feline immunodeficiency virus infection. J. Viral. 1993, 67, 2344-2348 Pu, R., Okada, S., Little, E.R., Xu, B., Stoffs, W. and Yamamoto, J.K. Protection of neonatal kittens against feline immunodeficiency virus infection with passive maternal antiviral antibodies. AIDS 1994, 9, 235-242 Tozzini, F., Matteucci, D. and Bandecchi, P. ef a/. Neutralizing antibodies in cats infected with feline immunodeficiency virus. J. C/in. Microbial. 1993, 31, 1626-1629 Matteucci, D., Pistello, M. and Mazzetti, P. et a/. Vaccination protects against in r&o-grown feline immunodeficiency virus even in the absence of detectable neutralizing antibodies. J. Virol. 1996, 70, 617-622 Hoffmann-Lehmann, R., Holznagel, E., Aubert, A., Bauer-Pham, K. and Lutz, H. FIV vaccine. II. Clinical findings, hematological changes and kinetics of blood lymphocyte subsets. Vet. Immmuol. 1995, 46, 115-125 Lombardi, S., Garzelli, C. and La Rosa, C. eta/. Identification of a linear neutralization site within the third variable region of the feline immunodeficiency virus envelope. J. Viral. 1993, 67. 4742-%749

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35 36

Lutz, H., Hoffmann-Lehmann, R. and Bauer-Pham, K. eta/. FIV vaccine. I. Immune response to recombinant FIV env gene products and outcome after challenge infection. Vet Immunol. 1995, 46, 115-125 Verschoor, E.J., Willemse, M.J. and Stam, J.G. eta/. Evaluation of subunit vaccines against feline immunodeficiency virus infection. Vaccine 1995, 14, 285-289 Osterhaus, A.D.M.E., lijiiaar, E. and Huisman, R.C. et al. Accelerated viremia in cats vaccinated with recombinant vaccinia virus expressing envelope glycoprotein of feline immunodeficiency virus. AIDS Res. Human Retroviruses 1996, 12, 437-441 Siebelink, K.H.J., Tihaar, E. and Huisman, R.C. ef a/. Enhancement of feline immunodeficiency virus infection after immunization. J. Viral 1995, 69, 3704-3711 Lutz, H., Hoffmann-Lehmann, R. and Leutnegger, C. et a/. Vaccination of cats with recombinant envelope glycoprotein of feline immunodeficiency virus: decreased viral load after challenge infection. A/LX Res. Human Retroviruses 1996, 12, 431-433 Wolff, J.A., Malone, R.W. and Williams, P. et a/. Direct gene transfer into mouse muscle in vivo. Science 1990, 247, 14651468 Ulmer, J.B., Donnelly, J.J., Parker, S.E. et a/. Heterologous protection against influenza by injection of DNA encoding a viral protein. Science 1993, 259, 1745-1748 Fynan, E.F., Webster, R.G., Fuller, D.H., Haynes, J.R., Santoro, J.C. and Robinson, H.L. DNA vaccines: protective immunizations by parenteral, mucosal and gene-gun inoculations. PNAS 1993, 90, 11478-11482 Robinson, H.L., Hunt, L.A. and Webster, R.G. Protection against a lethal influenza virus challenge by immunization with a haemagglutinin expressing plamid DNA. Vaccine 1993, 11, 957-960 Xiang, Z.Q., Spitalik, S., Tran, M., Wunner, W.H., Cheng, J. and Ertl, H.C.J. Vaccination with a plasmid vector carrying the rabies virus glycoprotein gene induces protective immunity against rabies virus. Viro/ogy 1994, 199, 132-140 Lu, S., Santoro, J.C., Fuller, D.H., Haynes, J.R. and Robinson, H.L. Use of DNAs expressing HIV-1 env and noninfectious HIV-l particles to raise antibody responses in mice. virology 1995, 209, 147-l 54 Okuda, K., Bukawa, H. and Hamajima, K. et a/. Induction of potent humoral and cell mediated immune responses following direct injection of DNA encoding HIV type 1 env and rev gene products. AIDS Res. Human Retroviruses 1995, 11, 933-943 Wang, B., Ugen, K.E. and Srikantan, V. et a/. Gene inoculation generates immune response against human immunodeficiency virus type 1. PNAS 1993, 90,4156-4160 Yasutomi, Y., Robinson, H.L. and Lu, S. et a/. Simian immunodeficiency virus specific cytotoxic T lymphocyte induction through DNA vaccination of rhesus monkeys. J. Viral. 1996, 70, 678481 Morikawa, S., Lutz, H., Aubert, A. and Bishop, D.H.L. Identification of a conserved and variable regions in the envelope glycoprotein sequences of two feline immunodeficiency viruses isolated in Zurich. Virus Res. 1991, 21, 53-63 Dandekar, S., Beebe, A. and Barlough, M.J. et a/. Detection of feline immunodeficiency virus (FIV) nucleic acids in FIVseronegative cats. J. Viral. 1992, 66, 4040-4049 Pancino, G., Camoin, L. and Sonigo, P. Structural analysis of the principal dominant domain of the feline immunodeficiency virus transmembrane glycoprotein. J. Viral. 1995, 69, 21102118 Gonin, P., Fournier, A., Oualikene, W., Moraillon, A. and Eloit, M. Immunization trial of cats with a replication defective adenovirus type 5 expressing the env gene feline immunodeficiency virus. Vet. Microbial. 1995, 45, 395-401 Burns, D.P.W. and Desrosiers, R.C. A caution on the use of SIV/HIV gag antigen detection systems in neutralization assays. AIDS 1992,8, 1189-1192 Mascola, J.R. and Burke, D.S. Antigen detection in neutralization assays: high levels of interfering anti-p24 antibodies in some plasma. AlDS 1993, 9, 1173-1174 Hosie, M.J. and Jarrett, 0. Serological responses of cats to feline immunodeficiency virus. A/DS 1990, 4, 215-220 Montefieri, D.C., Reimann, K.A., Letvin, N.L., Zhou, J. and Hu, S.L. Studies of complement-activating antibodies in the SIV/ macaque model of acute primary infection and vaccine protection. AIDS Res. Human Retroviruses 1995, 11, 963-970

Vaccine 1997 Volume 15 Number 10 1093

DNA vaccination using expression vectors: A.M. Cuisinier et al. 37

38

Evans, L.A., Thomso Honneghier, G. and Steimer, K. et al. Antibody-dependant-cellular-cytotoxicity is directed against both the gp120 and gp41 envelope proteins of HIV. ADS 1989, 3,273-276 Liunggren, K., Brouden, P.A., Morfeldt Manson, L., Jondal, M. and Wahren, B. IgG subclass response to HIV response to HIV in relation to cellular cytotoxicity at different clinical stages. C/in. Exp. Immunol. 1988, 73,343-347

1094

Vaccine 1997 Volume 15 Number 10

39

Tyler, D.S., Nastala, C.L. and Stanley, S.D. et a/. gp120 specific cellular cytotoxicity in HIV-l seropositive individuals: evidence for circulating CD16+ effector cells armed in vivo with cytophilic antibodies. J. Immunol. 1989, 142, 1177-1182