The Journal of Immunology
Intranasal HIV-1-gp160-DNA/gp41 Peptide Prime-Boost
Immunization Regimen in Mice Results in Long-Term HIV-1
Neutralizing Humoral Mucosal and Systemic Immunity1
Claudia Devito,* Bartek Zuber,* Ulf Schroder,† Reinhold Benthin,* Kenji Okuda,‡
¨
§
Kristina Broliden, Britta Wahren,* and Jorma Hinkula2*
An intranasal DNA vaccine prime followed by a gp41 peptide booster immunization was compared with gp41 peptide and control
immunizations. Serum HIV-1-specific IgG and IgA as well as IgA in feces and vaginal and lung secretions were detected after
immunizations. Long-term humoral immunity was studied for up to 12 mo after the booster immunization by testing the presence
of HIV-1 gp41- and CCR5-specific Abs and IgG/IgA-secreting B lymphocytes in spleen and regional lymph nodes in immunized
mice. A long-term IgA-specific response in the intestines, vagina, and lungs was obtained in addition to a systemic immune
response. Mice immunized only with gp41 peptides and L3 adjuvant developed a long-term gp41-specific serum IgG response
systemically, although over a shorter period (1–9 mo), and long-term mucosal gp41-specific IgA immunity. HIV-1-neutralizing
serum Abs were induced that were still present 12 mo after booster immunization. HIV-1 SF2-neutralizing fecal and lung IgA was
detectable only in the DNA-primed mouse groups. Intranasal DNA prime followed by one peptide/L3 adjuvant booster immu-
nization, but not a peptide prime followed by a DNA booster, was able to induce B cell memory and HIV-1-neutralizing Abs for
at least half of a mouse’s life span. The Journal of Immunology, 2004, 173: 7078 –7089.
I n establishing preventive strategies that reduce the risk of bicides have been designed to target dendritic cells, reverse tran-
HIV infection, mucosal immune responses induced by vac- scriptase, and HIV fusion in the same way as peptides based on gp
cination are of great importance, considering that HIV, like 41, T-20 and T-1249 (6, 7). HIV produces a chronic infection and
many other pathogens, is generally transmitted through the mu- integrates into the host chromosomes, which indicates that pro-
cosa. Although the incidence of HIV-1 transmission is very low phylactic vaccination is the only way to inhibit the spread of the
(0.0001– 0.0004 in discordant couples), there is the possibility of virus.
becoming infected during each sexual act (1, 2). A high frequency Many current vaccine candidates induce cellular immune re-
of sexual encounters, different viral loads of HIV-1 carriers, and sponses and low levels of neutralizing Abs, meaning that those
sexually transmitted diseases contribute to transmission of the vi- immunogens will be useful only when the infection is already es-
rus. Inflammations and lesions in the mucosal tissues also influ- tablished, but will not have any sterilizing effect (8 –11). Admin-
ence HIV-1 transmission, which occurs with every genetic HIV-1 istration of DNA immunogens has been shown to be a potent al-
subtype or infected cell (3). To prevent infection from different ternative to induce cellular, although poor humoral, responses in
pathogens, the immune system reacts quickly, producing peptides, mice and humans. Prime-booster immunization with DNA vaccine
NK cells, macrophages, and a number of Abs, such as IgM, IgG, encoding the gp of herpes simplex virus (gB)3 expressed and at-
and IgA. However, innate responses, thought to be effective in tenuated recombinant vaccinia virus vector (rvacgB) induced mu-
most cases, are not enough; they are unspecific and do not produce cosal and systemic responses (12). DNA constructs encoding
a secondary response. Therefore, specific immune responses are multi-CTL epitopes from human, macaques, and mice have also
needed to control the spread of the virus. Mucosal microbicides are been administrated i.m., enhancing HIV-specific CTL epitopes
being used as a therapeutic approach (4, 5). Microbicides may acts (13). Abs have been administrated to prevent HIV transmission,
as lubricants, protecting the mucosal surfaces, reducing the trauma, such as mAbs 2F5 and 2G12, which protected macaques against
preventing sexually transmitted diseases, reducing the risk of in- SHIV vaginal challenge (14).
flammation and ulceration, and maintaining a low pH. The micro- To enhance the HIV-1-specific immune responses, immunogens
are used in different combinations that target HIV attachment to
the host cells. Thus, passive infusion of mAbs, which protects
*Swedish Institute for Infectious Disease Control and Microbiology and Tumorbiol- macaques from mucosal chimeric simian-human immunodefi-
ogy Center, Department of Virology, Karolinska Institute, Solna, Sweden; †Eurocine, ciency virus (SHIV) infections, was combined with DNA immu-
Stockholm, Sweden; ‡Department of Bacteriology, Yokohama City University School
of Medicine, Yokohama, Japan; and §Division of Clinical Virology, Huddinge Uni- nization using IL-2 as adjuvant (15).
versity Hospital, Huddinge, Sweden Because i.m. DNA immunization has been shown to induce
Received for publication July 28, 2004. Accepted for publication September 27, 2004. mainly systemic immune responses, different ways of delivering
The costs of publication of this article were defrayed in part by the payment of page the DNA have been evaluated to enhance the properties of the
charges. This article must therefore be hereby marked advertisement in accordance immune responses, mainly in the induction of neutralizing Abs,
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by grants from the Swedish Research Council, the Swed-
ish Medical Society, KI Fonder, and Swedish Physicians against AIDS. 3
Abbreviations used in this paper: gB, gp of herpes simplex virus; HEPS, highly
2
Address correspondence and reprint requests to Dr. Jorma Hinkula, Swedish Insti- exposed, persistently seronegative; rvacgB, gB expressed and attenuated recombinant
tute for Infectious Disease Control, Department of Virology, 17182 Stockholm, Swe- vaccinia virus vector; sIgA, secretory IgA; SHIV, simian-human immunodeficiency
den. E-mail address: jorma.hinkula@smi.ki.se virus; NSI, non-syncytium-inducing.
Copyright © 2004 by The American Association of Immunologists, Inc. 0022-1767/04/$02.00
The Journal of Immunology 7079
where the responses are not strong enough. Thus, nasal immuni- immunization combinations used are described in Table I. All DNA plas-
zation may be the alternative to induce a broader response sys- mids contained the CMV immediate-early promoter for gene expression
temically and in distal mucosa, such as lungs, intestines, and vag- and the kanamycin resistance gene. Eight control mice were included in the
experiments. The intranasal booster immunization was performed with
inal secretions (16, 17). peptides (10 g/peptide) emulsified in mono-oleate/fatty acid (named L3)
We have used a heterologous vaccine strategy in which we first as previously described (25).
immunized with DNA encoding gp160/CCR5, then boosted with This study was performed with ethical permission from the regional
gp41 peptide to induce a long-lasting cellular and humoral re- ethical committee (Stockholm Nord).
sponse against HIV-1. This was based on our previous findings Sample collection
showing that serum and mucosal IgA from highly exposed, per-
Blood obtained by retro-orbital bleedings, vaginal and lung lavages, and
sistently seronegative individuals (HEPS) could neutralize primary feces were collected at 1- to 3-mo intervals from each mouse. Serum was
isolates of HIV-1 (18, 19) and that mucosal IgA could neutralize stored at 20°C until used. Mucosal samples were collected as previously
HIV-1 primary isolates representing many subtypes (20). We have described (16, 17, 26). Briefly, the vaginal and lung lavages were collected
previously reported that secretory IgA (sIgA), the main humoral in PBS with protease inhibitors (Sigma-Aldrich, St. Louis, MO) and stored
effector molecule in the mucosal immune system has a role in at 70°C before use. Feces were weighed and solubilized in PBS (0.1
g/ml) with protease inhibitors (1 mg/ml; Sigma-Aldrich). The debris was
inhibiting HIV-1 epithelial transcytosis (21). It has recently been removed by centrifugation, and the supernatant was stored at 70°C.
shown that IgA of HEPS recognized an epitope located on the When mice were killed, vaginal, lung, and intestinal lavages were collected
gp41 protein that differs from the IgA epitope recognized by HIV- as previously shown (16, 17).
infected individuals restricted to aa 581–584 (LQAR), which cor-
Peptide synthesis
responds to the conserved coiled-coil pocket in the helix region
of gp41 (22). Previous studies have also shown that certain HEPS The peptides corresponding to the gp41-neutralizing epitope synthesized
were aa 661– 675/ELDKWAS (Hybaid T-Peptides), representing clades
individuals may develop Abs against the HIV-1 non-syncytium-
A/(92UG31), B/(MN), C/(92BR25), and D/(92UG21), and a gp41 clade
inducing (NSI) phenotype coreceptor CCR5 (23). The protective B/(MN) peptide (Hybrid T-Peptides) located between aa 578 –595 pep-
capacity of Abs directed against this coreceptor has thus been in- tides, representing the human and simian CCR5 N-terminal region and
vestigated as a vaccine approach with interesting possibilities in second loop aa 168 –182. We also synthesized aa 178 –192 (Hybaid T-
vivo. Considering that experimental approaches inducing long- Peptides) and aa 302–318 from gp120 V3 loop/clade B. The HIV-1 Lai
Rev peptide synthesized, representing aa 65– 84, was used as a negative
term mucosal immune responses may be necessary, we evaluated control. All peptides mentioned were produced using solid phase F-moc
the immunogenicity of a peptide representing the gp41 coiled-coil chemistry (27).
pocket region in mice. We also evaluated whether these immune
responses can be enhanced by combining HIV-1 rgp160/Rev ELISA for detection of IgG and IgA Abs
DNA, CCR5 DNA, and CCR5 peptides with other well-conserved Ninety-six-well plates (Maxisorp; Nunc, Copenhagen, Denmark) were
gp41 epitopes of the envelope from HIV-1 clades A, B, C, and D coated with clade A–D gp41 ELDKWAS peptide, gp41 coiled-coil peptide
(24). A spectrum of specific epitopes from subtypes A–D provides (QLQARVL) peptide, gp120 V3/MN (IHIGPGRAFV), and the human
CCR5 second-loop peptides. All peptides were solubilized in 0.1 M
both broad HIV-1 clade immunity combined with autoantibodies NaHCO3 buffer (pH 9.5–9.6) at a coating concentration of 10 g/ml and
against the prominent coreceptor CCR5. We hypothesized that this added at 100 l/well. Plates were stored overnight at room temperature and for
approach would result in a more robust and long-lasting neutral- a minimum of 24 h at 4°C. Mouse sera were diluted in PBS (pH 7.4) with 0.5%
izing mucosal immunity able to resist the development of escape BSA (Roche, Mannheim, Germany) and 0.05% Tween 20 (Sigma-Aldrich),
and 100 l of dilutions of 1/100, 1/200, 1/400, and 1/800 were added to
mutations, which is so commonly seen with the HIV-1 virus. each well and incubated at 37°C for 90 min.
Mucosal samples were tested as previously described (16, 17). HRP-
Materials and Methods conjugated anti-mouse IgG (Bio-Rad, Richmond, CA; dilution, 1/2000) or
Immunization anti-mouse IgA (Southern Biotechnology Associates, Birmingham, AL;
dilution, 1/1000) was added at 100 l/well and incubated for 2 h at 37°C,
Female 10- to 12-wk-old C57BL/6 mice of the H-2b haplotype were im- and 2 mg/ml ophenylenediamine in 0.05 M sodium citric acid, pH 5.5, with
munized intranasally with HIV-1 rgp160BaL DNA, HIV-1 Rev/Lai DNA, 0.003% H2O2 were added as substrate at 100 l/well. After a 30-min in-
human CCR5 DNA, gp41/MN coiled-coil (aa 578 –591, GIKQLQARVLA cubation period, the reaction was stopped by adding 100 l/well 2.5 M
VERY), and gp41 subtype A–D peptides (aa 661– 675, MN, NEQLLEL H2SO4, and optical densities were measured at 490 nm. The positive con-
DKWASLWN, A/92UG31; aa 652– 665, EKDLLALDKWANLWN, trol used was a human HIV-IgG pool collected from HIV-1-infected Ugan-
C/92BR025; aa 651– 665, NEQDLLALDSWNLWN, D/92UG021; aa 643– dan patients, the Kabi 62 serum, mAbs against the gp41 ELDKWAS
657, EQELLKLDQWASLWN), E/93TH976 aa 643– 662, DQQDRNEK epitope mAb 2F5 (28) (donated by Dr. H. Katinger), anti-gp120 V3 mAb
DLLLDKWASLW peptides, and peptides representing the human second F58/H3 (29), and mAb 2D7 directed against the second CCR5 external
CCR5 coreceptor (aa 168 –182, FTRSQKEGLHYTCSSHFPYS) region (Hy- loop region (Coulter Pharmaceuticals, Palo Alto, CA). Sera from vaginal
baid T-Peptides, Ulm, Germany). All animals were immunized twice at a 2-mo secretions and feces from preimmunized animals were used as negative
interval, and the animal experiments were repeated twice. The different controls.
Table I. Immunization schedule of micea
Primary Immunization/Booster Immunization
Mouse Group rgp160/rev DNA CCR5-DNA gp41 ELDKWAS peptide gp41 coil peptide CCR5 second loop peptide
1 / / / / /
2 / / / / /
3 / / / / /
4 / / / / /
5 / / / / /
6 Reversed group 1 / / / / /
a
Six mice were used in each group except in the control group 5 controls receiving saline which contained eight mice. Groups 1, 2, 3, 5, and 6 were repeated twice. The
adjuvant monooleate/oleate L3 (groups 1, 2, 3, and 4) was used for the booster immunizations with peptides. DNA immunizations: 25 g/plasmid; peptide immunizations: 10
g/peptide.
7080 NASAL IMMUNIZATION WITH gp160 DNA/gp41 PROTEIN
IgA purification and quantitation B and T cell memory, IgG/IgA synthesis, and T cell proliferation
The secretions collected from lungs and intestines were used to isolate and
in vitro
analyze the HIV-1-specific IgA content. The Kaptive IgA/IgE reagents For analysis of HIV-specific B cell responses, 1 105 spleen cells were
(Biotech IgG, Copenhagen, Denmark) were purchased and used as recom- cultured in 200 l of RPMI 1640 supplemented with 5% inactivated FCS
mended by the manufacturer. IgA quantities were determined with an in- at 37°C in 96-well, flat-bottom plates (Nunc) with rgp160 (1 g/ml) or
house murine IgA capture ELISA, and a commercial murine IgA (1 mg/ml; peptides (10 g/ml) for 72 h. After washing, Ag-specific Igs were mea-
Sigma-Aldrich) was used for preparing a standard curve. Briefly, the pu- sured by ELISA as described above (14, 26). A positive reactivity was
rified IgA and the standard IgA were diluted in PBS with 5% fat-free dry considered if the ELISA substrate absorbance was higher than the mean
milk and 0.05% Tween 20 at 10-fold serial dilutions. One hundred micro- value OD 2 SD of the negative control. T cell proliferation assay was
liters per dilution was added to a 96-microwell plate precoated with rabbit performed as previously described (4, 11).
anti-murine IgA (Dakopatts, Sollentuna, Sweden) and incubated at 37°C
for 1 h. The plates were washed four times with saline and 0.05% Tween Statistical analyses
20 before 100 l of HRP-conjugated goat anti-murine IgA was added to
each well. After 1-h incubation at 37°C, plates were washed as previously Statistical comparisons between the groups were performed using the non-
described, and the presence of bound conjugate was detected using o- parametric Mann-Whitney U test. A significant difference was considered
phenylenediamine in 0.05 M sodium-citric acid, activated with 0.03% at a value of p 0.05. A one-way ANOVA nonparametric test was per-
H2O2 as substrate. The substrate reaction was terminated with 100 l/well formed using PRISM version 4.0a (GraphPad, San Diego, CA) for MacIn-
of 2.5 M H2SO4, and absorbance was measured at OD490. The amounts of tosh (OS 9; Apple Computers, Redmond, WA) for comparisons of medians
IgA in the mouse samples were determined by comparing the OD values between groups at p 0.05 and p 0.001 levels when comparing fecal
of the test samples with that of the IgA standard with known IgA and vaginal IgA titers.
concentration.
Neutralization assay Results
Systemic humoral immunity
Serum and IgA purified from feces from preimmunized and immunized
mice were tested for the presence of neutralizing activity. Mouse sera were Serum HIV-specific IgG responses against gp41, gp120, gp160,
heat-inactivated (56°C for 30 min) and serially diluted at 3-fold dilutions, and CCR5 were measured 1 year after postboost immunization.
starting at 1/20. Virus aliquots of the dual-tropic T cell line-adapted HIV-1 The geometric mean titers 6 and 12 mo after intranasal prime and
SF2 strain and the primary NSI/CCR5 tropic clade B isolate 6920 (19)
were diluted in quadruplicate 4-fold steps in medium supplemented with
booster immunizations are given in Table II. High Ab titers were
10% inactivated FCS (Invitrogen Life Technologies, Paisley, U.K.), 10 seen against all gp41 peptides. HIV-specific IgG persisted over 12
IU/ml IL-2 (Amersham Biosciences, Little Chalfont, U.K.), 50 g/ml mo, and no significant differences were detected among groups 1,
streptomycin, and 50 IU penicillin (Invitrogen Life Technologies). Seventy- 2, and 3 ( p 0.05) at 6 or 12 mo or in the variations in serum IgG
five microliters of each virus dilution and 75 l of each serum dilution titer at the two time points measured. The lowest IgG titers were
were incubated in duplicate for 1 h at 37°C in round-bottom culture plates
(Nunc). Seventy-five microliters with 1 106/ml PHA-stimulated PBMC detected against gp41 coiled-coil peptide, gp120 V3, gp160Lai,
were added to each well. After 16- to 18-h incubation at 37°C, the cells and CCR5. Mice immunized only with CCR5 peptides (group 4)
were washed with RPMI 1640, and 200 l of fresh medium was added to had equally high serum IgG Ab titers to CCR5 as mice from group
each well. Every 3 days, 100 l of medium was changed. After 6 –7 days, 1, which also received CCR5/DNA, whereas the control mice
100 l of supernatant from each well was collected, and virus production
was measured in a p24 Ag capture ELISA as previously described (30). An
(group 5) had low or undetectable IgG titers ( 100) against all
HIV-1-positive serum pool (HIVIG) and the human mAb 2F5, specific for HIV-1 Ags used. Among the groups of mice in which immuniza-
the gp41 ELDKWAS epitope, were used as a positive control. Neutraliza- tions were repeated, the mice primarily immunized with gp41/
tion was defined as the sample titer resulting in 50 and 90% reduction of CCR5 peptides in adjuvant and 2 mo later immunized with gp160/
p24 Ag in the supernatant compared with p24 Ag content when the virus CCR5 DNA had Ab responses similar to those in the group 3 mice
was incubated in the presence of HIV Ab from negative serum. All assays
were repeated at least twice. receiving peptides only (not shown).
HIV-1-specific IgA neutralization activity was performed as previously Serum IgA titers against HIV-1 and CCR5 were detected as late
described (28). In brief, purified fecal IgA from control or immunized mice as 1 year postboost immunization (Table III). No significant geo-
was incubated at 37°C for 1 h with 30 –50 tissue-culture infectious doses metric differences in the mean of IgA Ab titers ( p 0.05) were
resulting in 50% infected culture wells of cell-free virus HIV-1SF2 (T cell
line-adapted subtype B isolate) or with the primary HIV subtype B isolate
detected among groups 1, 2, and 3 against the gp41 clade A, B, and
6920 in a 200- l volume in duplicate 96-well cell culture plates (Nunc, C peptides, whereas the IgA titers against clade D gp41 were lower
Aarhus, Denmark). in group 3 ( p 0.05). In addition, group 3 lacked detectable IgA
Table II. Serum Ab levels following intranasal immunizationa
Geometric Mean IgG Titers of Clade Ag
B gp41 MN A gp41 UG31 C gp41 BR25 D gp41 UG21 B gp41 coil B gp120 V3 B na CCR5 loop
Group Time (mo) aa 661–675 aa 652–665 aa 651–665 aa 643–657 aa 578–592 aa 303–322 rgp160Lai aa 168–185
1 6 4,525 5,382 3,805 2,691 336 800 336 1,345
12 3,805 3,200 2,691 1,345 336 336 566 1,903
2 6 3,676 4,850 4,222 4,222 400 696 459 100
12 1,832 3,200 2,111 1,838 200 132 400 100
3 6 2,786 2,786 3,676 1,600 264 528 606 115
12 2,111 1,600 1,600 919 174 100 400 159
4 6 100 100 100 100 100 100 100 770
12 100 100 100 100 100 100 100 696
5 6 100 100 100 100 100 100 100 100
12 100 100 100 100 100 100 100 100
mAb 2F5 Positive controlb 12,400 1,440 680 100 100 100 57,600 100
a
IgG titers against HIV-1 gp41, gp120, rgp160, and CCR5 representing Ags 6 and 12 mo post-boost immunization. rgp160 recombinant gp160, na, not applicable.
b
Positive IgG control 5 g/ml (HRP-anti-IgG conjugate used). Statistical analysis using Mann-Whitney U nonparametric test were used to compare differences between
groups and p 0.05 was considered significant. No significant differences were seen when comparisons were performed between relevant groups.
The Journal of Immunology 7081
Table III. Serum IgA titers against HIV-1 gp41, gp120, and CCR5 representing Ags 6 and 12 mo post-boost immunizationa
Geometric Mean IgA Titers of Clade Ag
B gp41 MN A gp41 UG31 C gp41 BR25 D gp41 UG21 B gp41 coil B gp120 V3 B na CCR5 loop
Group Time (mo) aa 661–675 aa 652–665 aa 651–665 aa 643–657 aa 578–592 aa 303–322 rgp160Lai aa 168–185
1 6 200 598 400 100 400 390 420 370
12 200 800 336 100 425 200 275 350
2 6 303 303 303 132 303 225 160 100
12 303 303 264 152 303 174 152 100
3 6 159 200 200 100 264 100 200 440
12 159 152 174 100 100 100 180 160
4 6 100 100 100 100 100 100 100 790
12 100 100 100 100 100 100 100 300
5 6 100 100 100 100 100 100 100 100
12 100 100 100 100 100 100 100 100
mAb 2F5b Positive control 10,900 1,560 990 120 100 100 65,000 100
a
na not applicable; rgp160 recombinant gp160 Ag.
b
Positive IgG control 5 g/ml (HRP-anti-IgG conjugate used). Statistical analysis using Mann-Whitney U nonparametric test were used to compare differences between
groups and p 0.05 was considered significant. No significant differences were seen when comparisons were performed between relevant groups.
Abs against the V3 loop region. The HIV-specific IgA titers were, primarily immunized with gp41/CCR5 peptides and boosted with
in general, 10-fold lower than the IgG titers. Among the mice in HIV-gp160/CCR5 DNA, the fecal IgA reaction pattern did not
group 4, the highest IgA titers were seen against the CCR5 peptide. differ from the group 3 mice who were given peptides only (not
However, the mean Ab titers of this group were not significantly shown).
different from the titers obtained in group 1 ( p 0.42). In the In group 4 mice, IgA reactivity in feces with the CCR5 second-
negative control (group 5), HIV-specific IgA responses could not loop peptide was seen in all six animals. After 12 mo of follow-up,
be detected against any of the Ags used. all six mice had detectable IgA in the small intestine toward the
CCR5 peptide without reacting with tested HIV-1 Ags. No Ab
HIV-1-specific IgA in feces and intestinal washes after reactivity was seen in group 5 immunized with saline (Fig. 1). The
immunizations total IgA amount in fecal samples ranged between 280 and 418
We analyzed the reactivity of IgA from feces against gp41 g/ml; in the intestinal washes it ranged between 48 and 92 g/ml.
ELDKWAS peptides representing clades A, B, C, D, and E as well No significant differences were found among mice from the five
as CCR5. Fig. 1 shows the median IgA ELISA Ab reactivity at two studied groups.
different time points before and after the second immunization in
groups 1–5. Animals from group 1 developed IgA against clades
A, B, C, and D and CCR5 after the second booster, and the highest HIV-1-specific IgA in vaginal secretions after immunization
reactivity was detected against the CCR5 peptide (Fig. 1). None of Fig. 2 shows the median anti-HIV and CCR5 peptide IgA Ab
the animals developed IgA against the control Rev peptide. When reactivity in vaginal secretions 3–12 mo postbooster in four im-
mice were killed at 12 mo follow-up (Table IV), HIV-specific IgA munized groups and one control group of mice. Significantly
against gp41 ELDKWAS representing subtypes A, B, and C were higher reactivities in group 1 were detected against all HIV gp41
detected in the small intestine of all mice from group 1. This HIV- and CCR5 Ags compared with group 4 or 5. When studying the
specific reactivity was also detectable against the gp41 coiled-coil kinetics of vaginal HIV-specific IgA reactivity, long-term immu-
peptide. Furthermore, at 12 mo, five of six mice had intestinal IgA nity (9 mo) was seen against the gp41 peptides. By 12 mo after the
reactive with gp160, and four of six mice had intestinal IgA reac- peptide booster, HIV-specific IgA in vaginal samples against the
tive against the CCR5 second-loop peptide. gp41 ELDKWAS and the gp41 coiled-coil LQAR peptides were
HIV-1-specific IgA Abs in feces were also found in group 2 seen. In group 2, all mice developed IgA Abs against all Ags
(Fig. 1). After the booster immunization, all mice developed IgA except the CCR5 peptide, although the responses decreased with
Abs against gp41 peptides representing clades A, B, C, and D, time more rapidly than in group 1 (Fig. 2).
whereas in half (three of six) of the animals, the highest reactivity One of five mice in group 3 developed vaginal IgA Abs against
was detected against gp41 ELDKWAS clade B. After 12 mo (Ta- clade C and CCR5 peptide after the first immunization. However,
ble IV), IgA in the small intestine reacted with gp41 ELDKWAS the group median reactivity was low or undetectable against all
peptides of clade A in all six mice, against clade B in five of six Ags at most time points.
mice, against clade C in four of six mice, against the gp41 coil In group 4, CCR5 peptide-specific vaginal IgA response was
peptide in one of six mice, and against gp160 in five of six mice. seen at 1–9 mo postbooster immunization, but did not differ sig-
None of the animals from group 2 developed IgA Abs against nificantly from group 1 vaginal IgA titers. Reactivity at later time
CCR5 or Rev peptides. points and against the other Ags was undetectable (Fig. 2E). The
Low IgA Ab responses were seen in group 3, in which mice amount of sample that can be collected from each mouse is small
were immunized with gp41 ELDKWAS, CCR5, and gp41 coiled- (100 l/sample/occasion), and for this reason the samples were
coil peptides (Fig. 1). In this group, the highest reactivities were pooled for determining IgA amounts within each group (the pools
detected only against clade B and CCR5 peptides. When studying were prepared from three mice). Because total IgA responses were
the intestinal IgA responses 12 mo after the booster, clade A gp41 analyzed from a pool of vaginal wash samples, the amounts of IgA
ELDKWAS was recognized by three of six mice, that against clade measured in the preimmunized animals (3.3– 8.1 g/ml) did not
C was recognized by one of six, and no responses against the gp41 significantly differ from the IgA amounts detected in the postim-
coil peptide, gp160, CCR5, or Rev peptide were found. In mice munization samples (2.7–11.5 g/ml).
7082 NASAL IMMUNIZATION WITH gp160 DNA/gp41 PROTEIN
FIGURE 1. A–E, Median fecal IgA ELISA reactivity in mouse groups 1, 2, 3, 4, and 5 at two different time points, 3 wk after the primary immunization
and 4 wk after the second immunization. Fecal IgA titers are shown against the gp41 peptides representing clades A, B, D, and E and the CCR5 second-loop
peptide. All boxed figures show the median IgA titers and the 25th and 75th quartiles of group 1–5 mice. Significant differences between the control group
5 and the other groups at the p 0.001 or p 0.05 level are indicated above the bars (ns, nonsignificant difference). Values for determining cutoff limits
were determined by testing fecal pellet washes from preimmunization animals by ELISA. Mean background ELISA absorbance values 2 SD ranged from
between 0.095 and 0.150 in the various peptide Ags. A, Against gp41A-EKDLLALDKWANLWN peptide; B, against gp41B NEQLLELDKWASLWN
peptide; C, against gp41D-EQELLKLDQWASLWN peptide; D, against gp41 E-RNEKDLLLDKWASLW peptide; E, and against the second-loop human
CCR5 peptide (amino acids FTRSQKEGLHYTCSSHFPYS).
The Journal of Immunology 7083
Table IV. Long-term HIV- and CCR5 Ag-specific IgA responses in lung and small intestinal washes 12 mos post-boostera
gp41 ELDKWAS IgA Peptide Titers/Clade
CCR5 IgA Concn
Group Sampleb gp41/A gp41/B gp41/C gp41 Coil/B rgp160 Peptide ( g/ml)
1 Lung 5 4 60 16 50 18 16 7 15 8 4 18 6
Intestine 46 46 80 90 50 80 16 12 30 30 20 57 11
2 Lung 2 2 4 3 6 8 2 2 2 2 2 26 9
Intestine 12 12 20 28 16 16 4 4 8 8 4 86 19
3 Lung 2 2 10 2 10 4 2 2 2 2 2 16,6 9
Intestine 4 4 20 28 4 4 4 4 4 8 4 86 27
4 Lung 2 2 2 2 2 2 2 2 2 2 2 19,2 8
Intestine 4 4 4 4 4 4 4 4 4 4 16 78 11
5 Lung 2 2 2 2 2 2 2 2 2 2 16 33 16
Intestine 4 4 4 4 4 4 4 4 4 4 4 48 8
Positive controlsc
mAb2F5 1,490 12,000 710 100 61,000 100
HIVIG 4,900 3,900 6,800 2,900 590,000 140
mAb2D7 nt nt nt nt 100 220
Negative control
Mouse sera 50 50 50 50 50 50
a
Samples collected from the lung were tested at dilution 1/2–1/118 and intestinal washes at dilutions 1/4 –1/256 at 2-fold serial serum dilutions. Sample IgA titer cut-off was
calculated from mean absorbance values of negative control mice plus two SD. nt, not tested.
b
Two IgA pools per mouse group were prepared and tested.
c
Positive controls used were human IgG mAb and HIVIgG (HRP-anti-human IgG conjugate used) end-point titers are shown.
HIV-1-specific IgA in lung lavages after immunization which the two best animals per group are shown in Figs. 3 and 4.
Lung wash IgA was collected 12 mo postbooster immunization When serum was tested for neutralization responses, all three sera
(Table IV). To obtain sample amounts large enough to analyze IgA in group 1 neutralized a primary HIV-1 isolate exemplified by the
reactivity against several HIV-1 and CCR5 peptides, two pools of two best-neutralizing animal sera in Fig. 4. In group 2, three of
lung washes were prepared from three mice per study group. The four sera neutralized a primary HIV-1 isolate with 80% neutraliz-
strongest reactivities against the gp41 ELDKWAS epitopes were ing titers ranging between 20 and 80 (Fig. 4). Sera from mice in
seen with samples from group 1 mice. Group 2 mouse lung IgA groups 3 and 5 and mice primed with gp41/CCR5 peptides fol-
reacted with gp41 ELDKWAS peptides from clades B and C, al- lowed by an HIV-1 gp160/CCR5 DNA booster lacked detectable
though at low titers (IgA titer range, 2– 8). Group 3 had lung IgA HIV-1-neutralizing activity (Figs. 3 and 4). The HIV-1-neutraliz-
against gp41 ELDKWAS peptides from clades B and C only. ing sera were absorbed with HIV-1-infected or uninfected cells
The CCR5 peptide-immunized group 4 had CCR5 peptide-spe- before repeating the HIV-1 neutralization assays. This resulted in
cific IgA only in lungs (IgA titer, 10 –16). The amount of IgA in loss of HIV-1-neutralizing activity when HIV-1-infected cells
lung secretions ranged between 5.5 and 33 g/ml samples in the were used (Fig. 5). When uninfected cells were used to absorb
various groups. serum, a slight decrease in the serum neutralization activity was
seen in the plasmid DNA-primed mice (groups 1 and 2; group 2
IgG and IgA memory B and T cell responses in vitro assays not shown), and a complete loss of neutralizing activity was found
Three groups of mice (1, 2, and 3) had similarly high frequencies among the CCR5 peptide-immunized mice (group 4).
of spleen and inguinal lymph nodes IgG and IgA/B cell responders Fecal IgA purified from all six mice in group 1 were shown to
against HIV-1 Ags, and two groups (1 and 4) had high frequencies neutralize HIV-1 SF2 by 80% at 6 and 9 mo (not shown) after
of IgA responders against the CCR5 loop peptide at 12 mo after booster immunization (Table V). In group 2, the fecal IgA neu-
booster immunization. In the control group 5, one of 10 mice re- tralization capacity reached 50% neutralization titer at 6 mo (four
acted with the CCR5 peptide, whereas there was no response of six mice) and 50% at 9 mo postbooster immunization (not
against the other peptides used (not shown). At 12 mo, HIV Ag- shown). Fecal IgA from all other groups failed to neutralize HIV-1
specific in vitro T cell proliferative responses were detectable only SF2. This neutralization capacity was also seen against primary
in the DNA HIV-1 gp160 animals (groups 1 and 2). Low amounts HIV-1 isolate 6920 at 6 and 9 mo after booster immunization in all
of IFN- were detected in gp120 V3 peptide-stimulated spleen four mice in group 1.
cells of mice from groups 1 and 2 (45–150 pg/ml; not shown). In group 2, the fecal IgA-neutralizing capacity was detectable at
6 mo in two of four mice and was undetectable at 9 mo postbooster
HIV-1 neutralization responses immunization (not shown). HIV-1 SF2-neutralizing IgA in lung
We evaluated the ability of serum to neutralize one laboratory- lavages was detected in two of six mice in groups 1 and 2 at 12 mo
adapted strain and a primary HIV-1 isolate (Figs. 3 and 4). All postbooster immunization (data not shown). All other groups of
three sera tested neutralized the HIV-1/SF2 isolates, with 80% immunized mice lacked detectable neutralizing Abs in feces and
neutralizing titers ranging between 20 and 80 (1, 2, and 3), of lung lavages.
7084 NASAL IMMUNIZATION WITH gp160 DNA/gp41 PROTEIN
FIGURE 2. A–E, Kinetics of vaginal wash IgA ELISA reactivity against gp41 peptides representing HIV-1 clades A, B, and C. The median IgA titer
reactivity is shown at four time points (3, 6, 9 and 12 mo after the second immunization) in groups 1–5. The boxes show the median vaginal IgA Ab titers
with the 25th and 75th quartiles from four HIV Ag-immunized mouse groups (groups 1–3) and the control mice (group 5) Values for determining cutoff
limits were performed by testing vaginal washes from preimmunization animals by ELISA. Mean background ELISA absorbance values 2 SD ranged
from between 0.028 and 0.50 in the various peptide Ags. Significant differences at the p 0.001 or p 0.05 level (ns, nonsignificant difference) between
control group 5 animals and the other groups are indicated at the top of the bars. Statistics were performed using the one-way ANOVA nonparametric
analysis. A, Against the HIV-1 gp41 clade A representing peptide (aa EKDLLALDKWANLWN); B, against the HIV-1 gp41 clade B representing peptide
(aa NEQLLELDKWASLWN; C, against the HIV-1 gp41 clade C representing peptide (aa NEQDLLALDSWNLWN); D, against the HIV-1 gp41 coiled-
coil peptide clade B representing peptide (aa GIKQLQARVLAVERY); E, against the CCR5 second-loop peptide representing the human CCR5 HIV-1
coreceptor (aa FTRSQKEGLHYTCSSHFPYS).
The Journal of Immunology 7085
FIGURE 3. HIV-1 SF2 SI-phenotype neutralization activity of two rep-
FIGURE 4. HIV-1 6920 NSI phenotype primary isolate neutralization
resentative mouse sera from groups 1–5 collected 9 mo after the second
activity of two representative mouse sera from groups 1–5 collected 9 mo
immunization. Sera collected 6 mo postbooster immunization from two
after the second immunization. Sera collected from two representative mice
representative mice primarily immunized with gp41/CCR5 peptides in L3
primary immunized with gp41/CCR5 peptides in L3 adjuvant and boosted
adjuvant and boosted with HIV-1 gp160/CCR5 DNA intranasally, referred
with HIV-1 gp160/CCR5 DNA intranasally, referred to as Reversed Group
to as Reversed Group 1 mice, are shown. Mouse group and numbers of
1 mice, are shown. Mouse group and numbers of individual mice are pre-
individual mice are presented in Table I. A 50% neutralization was con-
sented in Table I. A 50% neutralization was considered positive.
sidered positive.
Discussion
In this study our aim was to design an immunogen that induces binding breaks down into two coils within the gp41 monomer, and
systemic and mucosal humoral responses capable of eliminating that synthetic peptides against either gp41 helical coil are able to
free virus particles and protect against HIV-1 infection by inducing inhibit viral infectivity (33, 34). We have included the gp41 coiled-
long-lasting B and T cell memory responses. For that reason, we coil peptide to obtain an Ab response that might inhibit fusion and
decided to compare HIV-1 DNA priming/protein or peptide boost possibly be capable of recognizing several HIV-1 subtypes. We
with peptide immunization only, because the first approach has pre- have also included the ELDKWAS peptides, because it was pre-
viously been shown to provide a good memory response and detect- viously shown that mucosal IgA, but not IgG, from HIV-seropos-
able humoral immunity. Peptide or protein immunizations alone have itive individuals neutralized HIV by recognizing the ELDKWAS
shown a wide variability of both efficient and less efficient induc- epitope located on the gp41, thus inhibiting HIV-1 transcytosis
tion of functional Abs or cell-mediated immunity (28, 31). (35), suggesting a substantial role for sIgA.
In this study we have used immunogens in different combina- The broadest systemic and mucosal responses were induced in
tions that contain the HIV-1 envelope gp120 and the transmem- mice primarily immunized with rgp160/Rev and CCR5 DNA and
brane gp41 (gp160), a few gp41 peptides representing HIV-1-neu- boosted with gp41-ELDKWAS and CCR5 peptides in mono-
tralizing regions, and the second external loop of the CCR5 oleate/oleate. High, long-term serum IgG and IgA titers were ob-
receptor. It is known that the gp120 mediates binding and entry tained against the neutralizing gp41 ELDKWAS epitopes from
steps in HIV-1 infection and that the gp120-CD4 coreceptor bind- HIV-1 clades A, B, C, and D. The lower titers of serum IgG still
ing induces conformational changes that activate the gp41 trans- reacted with the gp41 coiled-coil peptide, gp120 V3 clade B, and
membrane regions of the envelope (32). It is also believed that the gp160/Lai 1 year after booster immunization. Mice immunized
fusion structure of the envelope protein after CD4 and coreceptor with HIV-1 peptides only developed high serum IgG titers against
7086 NASAL IMMUNIZATION WITH gp160 DNA/gp41 PROTEIN
Table V. HIV-1 neutralization by purified fecal IgA samples of
immunized micea
HIV Isolate HIV-1 SF2 HIV-1 6920
50% 80% 50% 80%
Group
1 IgA pool 1 12 5 10 6
IgA pool 2 14 5 6 4
IgA pool 3 8 5 nt nt
2 IgA pool 1 6 4 4 4
IgA pool 2 5 4 5 4
IgA pool 3 4 4
3 IgA pool 1 4 4 4 4
IgA pool 2 4 4 4 4
4 IgA pool 1 4 4 4 4
IgA pool 2 4 4 4 4
Controls
mAb2F5 Anti-gp41 900 300 330 160
mAbZA1 Anti-CMV 10 10 10 10
a
HIV-1 neutralizing IgA titers resulting in 50 or 80% reduced HIV-1 p24 Ag
detection are shown with two to three pooled IgA fractions from immunized mouse
groups. nt, not tested; mAb 2F5 human anti-ELDKWAS peptide binding mono-
clonal Ab; mAb ZA1, murine anti-human CMV binding monoclonal Ab. Each pool
of IgA contains IgA from two mice with a final IgA concentration of 22–32 g
IgA/ml.
groups receiving CCR5 Ags ( p 0.01 and p 0.02) 6 mo after
the peptide booster. This difference was lost at 12 mo of follow-up.
Differences in the long-term mucosal IgA responses in the dif-
ferent mouse groups were seen against HIV-1 Ags, and both HIV-
specific gp41 peptides and gp160 were found in lungs and vaginal
secretions. Mice immunized with HIV envelope and CCR5 DNA
developed significantly higher HIV-specific IgA titers to gp41 pep-
tides representing HIV-1 clades A–C in lungs, vaginal mucosa,
and feces, whereas IgA Abs in the small intestine were directed
against clades A and C and against the gp41 coiled-coil peptide. A
significant difference in fecal, but not vaginal or serum, IgA reac-
tivity against clade B gp41 peptide was found between mice given
HIV-1 DNA and gp41 peptide compared with mice receiving HIV-
gp41 peptides only. In mice first immunized intranasally with
HIV-1 gp41/CCR5 peptides and then boosted with HIV-1/CCR5
DNA, a poor functional Ab response was obtained. These data
suggest that priming the immune system with an HIV envelope
immunogen in the form of a DNA plasmid better supports the
development of functional virus-neutralizing Abs than intranasal
priming performed with HIV-gp41 peptides even if the booster is
given as a DNA plasmid. Nasal immunization of mice with HIV-1
FIGURE 5. HIV-1 6920 NSI primary isolate-neutralizing activity with gp160/Rev CCR5 DNA (group 1) induced higher HIV-1-specific
pooled sera collected 9 mo after booster immunization from group 1 (A) IgA Ab titers in vaginal secretions during 6 –12 mo of follow-up
and from group 4 (B) before and after absorption with uninfected and than in the other groups (groups 2–5), even if the numbers of
HIV-1 6920-infected U937 cells. responders were not significantly different in groups 1–3.
Different routes and immunogens are being used to deliver DNA
to induce HIV-1-specific immune responses. It has been shown that to
the gp41 peptides, as also found in the DNA-primed animals. Sim- induce a stronger CTL response, immunizations may be given i.m.
ilarly, serum IgA titers were equal to titers obtained in DNA-primed with DNA encoding CTL epitopes, followed by a gene gun booster of
animals against the gp41-neutralizing ELDKWAS epitopes from modified vaccinia Ankara, and that specific HIV-1 immune responses
clades A, B, and C, but were significantly lower or undetectable 1 year are increased as a result of the boosting strategy (13).
after booster immunization against the gp41 ELDKWAS epitopes Intrarectal immunization has been shown to induce CTL mem-
from clade D, gp41 coiled-coil, and gp120 V3 peptides. ory in Peyer’s patches, lamina propia, and spleen in mice (36), and
Serum IgG and IgA titers against HIV Ags did not differ sig- in macaques, CTL responses have also been found in mesenteric
nificantly among the three groups of mice receiving HIV-1 Ags lymph nodes and PBMC (37); that protection could be enhanced
with or without inclusion of the CCR5 coreceptor Ag. Serum IgG by the use of IL-2 (38). Combinations of IL-1, IL-12, IL-18, and/or
and IgA responses against the CCR5-representing peptide were GM-CSF have been shown to be effective in inducing systemic and
seen in the three mouse groups immunized with CCR5 Ag, with a mucosal CTL-specific responses after nasal immunization (39).
significantly higher serum IgG against CCR5 in the two mouse However, none of these studies reported the presence of mucosa
The Journal of Immunology 7087
HIV-1-IgA-specific Abs or how long these responses could be tected were nonneutralizing, and IgA-neutralizing Abs in the mu-
maintained, as we show in this report where mucosa Abs were cosa have been seen when mice were systemically primed with
enhanced by the peptide protein booster in combination with a DNA and boosted with peptides. The reason for this is not clear,
mono-oleate fatty acid adjuvant (L3; data not shown). Our aim was but it resembles many of the previous studies in which HIV-1
to induce IgA Abs and Th cell memory; to achieve this, we believe peptide immunizations have failed to induce neutralizing Abs. It is
the best way is to use a nasal immunization protocol to target likely that the peptide-induced humoral immunity does not recog-
mucosal B cells responsible for the generation of sIgA organized nize the neutralizing epitopes presented on HIV virions. The sys-
in the MALT. temic and mucosal immunity was still detectable 1 year after the
Indeed, in our study, 12 mo after the last immunization, vaginal booster immunization, which represents at least one-third of the
anti-gp41-specific IgA was still detectable, mainly in mice immu- life span of the mouse.
nized with gp160/CCR5 DNA. There seemed to be a trend toward A number of vaccine studies have focused on gp120 envelope
an enhanced anti-HIV Ag-specific mucosal humoral immunity in protein, because the variable regions (V1–V3) in the rgp120 en-
this same group of mice. The group differs in its immunization velope glycoprotein are the major targets for neutralizing Abs (28,
schedule by the inclusion of the CCR5 DNA and CCR5 peptide 43, 44). However, the use of rgp120 to date has been less success-
pooled with HIV-1 DNA and gp41 peptides. The use of DNA ful in inducing broad clade-recognizing Abs that neutralize HIV-1
plasmid may have resulted in a higher amount of activated sys- primary isolates. Therefore, our immunization strategy included
temic and local lymphocytes, resulting in a longer and better pre- relatively conserved gp41 envelope components and one of the
served immune memory, as the long-term humoral mucosal im- main HIV-1 coreceptors, to which some epitopes probably are ex-
munity indicates. HIV-1-neutralizing serum Abs were highly posed as cryptic and highly conserved determinants (45).
efficient in neutralizing HIV-1 SF2 and the primary HIV-1 isolate The two chosen gp41 epitopes included in our present immu-
in both plasmid DNA-primed mouse groups where the HIV-1- nization strategy thus show a high or relatively high degree of
neutralizing capacity could be removed by adsorbing the serum conservation, as shown in a study by Dong et al. (46). They com-
with HIV-1-infected U937 cells, but only partially when unin- pared 862 HIV-1 strains and showed that 90 –97% of the isolates
fected cells were used. These results may suggest that the main contain the LQAR sequence in the coiled-coil region, whereas the
functional Abs were directed against HIV-1-neutralizing epitopes. LELDKWAS region showed a more variable mix of 50 –97% con-
However, it is important to remember that the way the neutraliza- served amino acids (with amino acids L-LD-W being 97% con-
tion assay was designed was not optimal for studying the anti- served). This combination of epitopes included in a vaccine may
CCR5-blocking capacity of the serum. Despite this, in the group 4 provide the conserved epitopes that have been so difficult to find in
the serum neutralizing capacity reached 60 –70% against the pri- other regions of the HIV-1 envelope even though highly conserved
mary HIV-1 isolate used. This could indicate that the CCR5 co- cross-neutralization Ab-inducing epitopes have been described in
receptor may be present at the virion envelope, as previously sug- the gp120 V3 region aa GPGR (47– 49) and within the CD4-bind-
gested, or that anti-idiotypic Abs for the CCR5 peptide may have ing region of gp120 (50). Additional evidence suggesting the im-
induced V3-neutralizing Abs. HIV-1-neutralizing IgA purified portance of inducing Abs against these gp41 epitopes is the low
from feces and in the lungs was detected only in mice intranasally frequency of Ab reactivities against these regions in HIV-infected
immunized with HIV DNA with and without CCR5 plasmid. individuals (51). In the mice used in our study the gp41 peptides
It could be argued that the human CCR5 Ag in this mouse study are poor MHC class I binders; thus, CTL and IFN- responses
would act as any foreign Ag, because the human and murine CCR5 were low or undetectable, whereas in the DNA-immunized mice,
proteins differ by 20% (40). It has been previously shown that gp120 V3-specific IFN- was low, but detectable. The gp41 pep-
anti-human or macaque CCR5-specific serum and mucosal IgG tides were able to stimulate T cell proliferative and IL-4 responses,
and IgA inhibit HIV-1 entry in vitro (41, 42). Serum HIV-specific suggesting that a Th2-type Th cell memory may have been evoked
IgA against the CCR5 coreceptor has been shown to develop in (not shown). With the inclusion of the CCR5 DNA in our vaccine
humans exposed to HIV as well as in individuals with the 32-bp formulation, this combination has every crucial constituent in-
CCR5 deletion (23). In preliminary studies we have seen how long volved in the fusion HIV-1 process.
serum IgG and IgA could last in macaques immunized with the The results reached by our intranasal administration of the im-
human CCR5 as plasmid DNA and CCR5 peptide booster. Al- munogens are consistent with previous studies that showed that i.p.
though only two monkeys were studied, both animals developed immunization followed by intranasal or intragastric boosts with
IgG and IgA Abs toward CCR5 N and second variable loop re- gp41 induced mucosal and systemic Abs recognizing HIV-1 pri-
gions of the HIV-1 outer envelope protein gp120, whereas long- mary isolates (31). Previous studies have shown that oral or intra-
term serum Ab responses were seen in one of the animals (data not nasal administration of PLG-encapsulated or water-dissolved plas-
shown). In macaques, Abs to CCR5 remain highly stable and mid DNA encoding gp160/Rev also induced Env-specific serum
clearly detectable toward both the N-terminal and the second Abs, and that an increased level of IgA directed to gp160 was
CCR5 loop region. detected in the feces of immunized mice (52).
Previous studies have shown that plasmid HIV-1 DNA prime The finding that some of the HEPS individuals had serocon-
protein booster immunizations have resulted in long-term HIV-1- verted shortly after giving up sex work, have brought to light the
specific immunities, showing that it would be possible to induce importance of B and T cell memory to achieve protection (53). The
long-term immunity to both HIV Ags and the CCR5 coreceptor seroconversion in this small group of previously resistant sex
without causing side effects (42). In addition, prime immunization workers was associated with a reduction in CTL, indicating that
with herpes simplex gB DNA and systemic boosting with rvacgB the presentation of Ag even in small doses is important to induce
have been shown to be effective in inducing mucosal IgA re- a long-term immunological response. Because memory B cells can
sponses in mucosally immunized mice, and these responses were persist after immunization with low Ab production (55, 56), we
increased when animals received the rvacgB as prime, followed by measured the presence of IgA and IgG B cell synthesis in vivo in
a booster of gB DNA (12). It is important to note that when we spleen and lymph nodes 1 year after intranasal immunization. We
primed with synthetic gp41-CCR5 HIV-1 peptides and boosted could see that the memory B cells against HIV-1 peptide and pro-
with DNA/gp160-CCR5, the HIV-specific IgG or IgA Abs de- tein Ags persisted up to 1 year after booster immunization in both
7088 NASAL IMMUNIZATION WITH gp160 DNA/gp41 PROTEIN
DNA-primed and HIV gp41 peptide-boosted mice, whereas the 5. Turpin, J. A. 2002. Considerations and development of topical microbicides to
inhibit the sexual transmission of HIV. Exp. Opin. Invest. Drugs 11:1077.
CCR5 peptide-immunized mice had a lower frequency of long- 6. Kilby, J. M., S. Hopkins, T. M. Venetta, B. DiMassimo, G. A. Cloud, J. Y. Lee,
term B cell memory toward CCR5. Memory B cells play an im- L. Alldredge, E. Hunter, D. Lambert, D. Bolognesi, et al. 1998. Potent suppres-
portant role in controlling and preventing infection. They prolif- sion of HIV-1 replication in humans by T-20 a peptide inhibitor of gp 41-medi-
ated virus entry. Nat. Med. 4:1302.
erate immediately in the presence of Ags and differentiate into 7. Baldwin, C. E., R. W. Sanders, and B. Berkhout. 2003. Inhibiting HIV-1 entry
plasma B cells that will produce specific Abs. B cells can also with fusion inhibitors. Curr. Med. Chem. 10:1773.
present Ag, stimulating T cell and cytokine responses (57–59). To 8. Amara, R. R., F. Villinger, J. D. Altman, S. L. Lydy, S. P. O’Neil, S. I. Staprans,
D. C. Montefiori, Y. Xu, J. G. Herndon, L. S. Wyatt, et al. 2001. Control of a
evaluate the specific immune responses induced by our immuno- mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA. Sci-
gens, we performed a number of immunizations with rgp160/Rev ence 292:69.
DNA/ELDKWAS peptide, with and without the CCR5 peptide. 9. Barouch, D. H., S. Santra, J. E. Schmitz, M. J. Kuroda, T. M. Fu, W. Wagner,
M. Bilska, A. Craiu, X. X. Zheng, G. R. Krivulka, et al. 2000. Control of viremia
No responses detected against the CCR5 coreceptor were seen in and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA
animals without the CCR5 Ag, whereas specific B cell responses vaccination. Science 290:486.
10. Hel, Z., W. P. Tsai, A. Thornton, J. Nacsa, L. Giuliani, E. Tryniszewska,
were detected against rgp160; gp41 clades A, B, C, and D; gp41 M. Poudyal, D. Venzon, X. Wang, J. Altman, et al. 2001. Potentiation of simian
coiled-coil peptide; and whole recombinant gp41 when CR5 was immunodeficiency virus (SIV)-specific CD4 and CD8 T cell responses by a
included. DNA-SIV and NYVAC-SIV prime/boost regimen. J. Immunol. 167:7180.
11. Kim, J. J., J. S. Yang, L. K. Nottingham, D. J. Lee, M. Lee, K. H. Manson,
Previous investigations have shown that Abs play a crucial role M. S. Wyand, J. D Boyer, R. E. Ugen, and D. B. Weiner. 2001. Protection from
in prevention against infection. Antiviral and protective activities immunodeficiency virus challenges in rhesus macaques by multicomponent DNA
were detected when neutralizing Abs were i.v. administered pas- immunization. Virology 285:204.
12. Eo, S. K., M. Gierynska, A. A. Kamar, and B. T. Rouse. 2001. Prime-boost
sively (14, 49, 59, 60, 61). These data clearly suggest that systemic immunization with DNA vaccine: mucosal route of administration changes the
Abs can provide protection against mucosal virus exposure, but rules. J. Immunol. 166:5473.
because IgG or monomeric IgA in serum and plasma will not be 13. Hanke, T., V. C. Neumann, T. J. Blanchard, P. Sweeney, A. V. S. Hill,
G. L. Smith, and A. McMichael. 1999. Effective induction of HIV-specific CTL
actively transported across mucosa, whereas secretory IgA or IgM by multiple-epitope using gene gun in a combined vaccination regimen. Vaccine
will, we believe that mucosal IgA Abs could play an important role 17:589.
14. Mascola, J. R., G. Stiegler, T. C. VanCott, H. Katinger, C. B. Carpenter,
and also inhibit epithelial HIV transcytosis (21). Probably the most C. E. Hanson, H. Beary, D. Hayes, S. S. Frankel, D. L. Birx, et al. 2000. Pro-
robust immunity against a virus that causes chronic infection such tection of macaques against vaginal transmission of a pathogenic HIV-1/SIV
as HIV would be to have a systemic and mucosal B cell repertoire chimeric virus by passive infusion of neutralizing antibodies. Nat. Med. 6:207.
15. Mascola, J. R., M. G. Lewis, T. C. VanCott, G. Stiegler, H. Katinger, M. Seaman,
capable of neutralizing the virus, and a systemic as well as local T K. Beaudry, D. H. Barouch, B. Korioth-Schmitz, G. Krivulka, et al. 2003. Cel-
cell response, with both Th and cytotoxic properties, should be lular immunity elicited by human immunodeficiency virus type 1/simian immu-
able to prevent entry of the virus. In most cases where the inhibi- nodeficiency virus DNA vaccination does not augment the sterile protection af-
forded by passive infusion of neutralizing antibodies. J. Virol. 77:10348.
tion of virus entry via mucosa fails, a systemic HIV-neutralizing B 16. VanCott, T. C., R. W. Kaminski, J. R. Mascola, V. S. Kalyanaraman,
and T cell response may still be preventive, but we believe that this N. M. Wassef, C. R. Alving, T. J. Ulrich, G. H. Lowell, and D. L. Birx. 1998.
level of immunity, when lacking the immunity in the mucosal HIV-1 neutralizing antibodies in the genital and respiratory tracts of mice intra-
nasally immunized with oligomeric gp160. J. Immunol. 160:2000.
compartments, may have a smaller chance of being efficient over 17. Lundholm, P., Y. Asakura, J. Hinkula, E. Lucht, and B. Wahren. 1999. Induction
the long term. Our study suggests the intranasal route for admin- of mucosal IgA by a novel jet delivery technique for HIV-1 DNA. Vaccine
17:2036.
istering HIV-1 DNA plasmid vaccines combined with relevant 18. Mazzoli, S., L. Lopalco, A. Salvi, D. Trabattoni, S. Lo Caputo, F. Semplici,
synthetic HIV-1 peptides as a non-live, safe, and promising vac- M. Biasin, C. Bl, A. Cosma, C. Pastori, et al. 1999. Human immunodeficiency
cine combination with the capacity to provide potent functional virus HIV-specific IgA and HIV neutralizing activity in the serum of exposed
seronegative partners of HIV-seropositive persons. J. Infect. Dis. 180:871.
antiviral humoral immunity. 19. Devito, C., J. Hinkula, R. Kaul, L. Lopalco, J. J. Bwayo, F. Plummer, M. Clerici,
In conclusion, HIV-1 gp41-specific IgA were found in feces and and K. Broliden. 2000. Mucosal and plasma IgA from HIV-exposed seronegative
lung and vaginal secretions up to 12 mo after immunization. HIV- individuals neutralize a primary HIV-1 isolate. AIDS 14:1917.
20. Devito, C., J. Hinkula, R. Kaul, J. Kimani, P. Kiama, L. Lopalco, C. Barrass,
specific IgG and IgA Abs and memory B cells directed against S. Picconi, D. Trabattoni, J. J. Bwayo, et al. 2002. Cross-clade HIV-1 specific
putative neutralizing epitopes were detectable in spleen and drain- neutralizing IgA in mucosal and systemic compartments of HIV-1 exposed, per-
ing lymph nodes against gp41/MN; gp41 clades A, B, and C; sistently seronegative subjects. J. AIDS 30:413.
21. Devito, C., K. Broliden, R. Kaul, L. Svensson, K. Johansen, P. Kiama, J. Kimani,
rgp160; and CCR5 loop peptide. This response was induced by a L. Lopalco, P. Stefania, J. J. Bwayo, et al. 2000. Mucosal and plasma IgA from
vaccine that includes critical fusion-dependent determinants. Be- HIV-exposed uninfected individuals inhibit HIV-1 transcytosis across human ep-
ithelial cells. J. Immunol. 165:5170.
cause memory B cells accumulate with every immunization even 22. Clerici, M., C. Barassi, C. Devito, C. Pastori, S. Piconi, R. Longhi, J. Hinkula,
when Abs are not detectable in serum, we measured IgG and IgA K. Broliden, and L. Lopalco. 2002. Serum IgA of HIV-exposed uninfected in-
B cell stimulation in vitro in spleen and lymph nodes that reflected dividuals inhibit HIV through recognition of a region within the -helix of gp41.
AIDS 16:1731.
the presence of long-term immunity. Although innumerable vac- 23. Lopalco, L., C. Barassi, C. Pastori, R. Longhi, S. E. Burastero, G. Tambussi,
cine studies promoted the idea of boosting cellular immunity, the F. Mazotta, A. Lazzarin, M. Clerici, and A. G. Siccardi. 2000. CCR5-reactive
induction of long-lasting memory B cells is still an unresolved task antibodies in seronegative partners of HIV-seronegative individuals down-mod-
ulate surface CCR5 in vivo and neutralize infectivity of R5 strains of HIV in
in the development of a preventive HIV vaccine; our mucosal ap- vitro. J. Immunol. 164:3426.
proach could provide some clue of how this may be accomplished. ¨ ¨
24. De Wolf, F., E. Hogervorst, J. Goudsmit, E. M. Fenyo, H. Rubsamen-Waigmann,
H. Holmes, B. Galvao-Castro, E. Karita, C. Wasi, S. D. Sempala, et al. 1994.
Syncytium-inducing and non-syncytium-inducing capacity of human immunode-
ficiency virus type 1 subtypes other than B: phenotypic and genotypic charac-
References teristic. AIDS Res. Hum. Retroviruses 10:1387.
1. Vittinghoff, E., J. Douglas, F. Judson, D. McKirnan, K. MacQueen, and ¨
25. Schroder, U., S. B. Svenson. 1999. Nasal and parenteral immunizations with
S. P Buchbinder. 1999. Per-contact risk of human immunodeficiency virus trans- diphteria toxoid using monoglyceride/fatty acid lipid suspensions as adjuvants.
mission between male and sexual partners. Am. J. Epidemiol. 150:306. Vaccine 17:2096.
2. Gray, R. H., M. J. Wawer, R. Brookmeyer, N. K. Sewankambo, D. Serwadda, ¨
26. Asakura, Y., P. Lundholm, A. Kjerrstrom, R. Benthin, E. Lucht, J. Fukushima,
F. Wabwire-Mangen, T. Lutalo, X. Li, T. VanCott, T. C. Quinn, et al. 2001. K. Okuda, S. Schwartz, B. Wahren, and J. Hinkula. 1999. DNA-plasmids of
Probability of HIV transmission per coital act in monogamous, heterosexual, HIV-1 induce systemic and mucosal immune responses. J. Biol. Chem. 380:375.
HIV-1 discordant couples in Rakai, Uganda. Lancet 357:1149. ¨
27. Sallberg, M., U. Ruden, L. O. Magnius, E. Norrby, and B. Wahren. 1991. Rapid
3. Norvell, M. K., G. I. Benrubi, and R. J. Thompson. 1984. Investigation of mi- “tea-bag” peptide synthesis using 9-fluorenylmetoxycarbonil (Fmoc) protected
crotrauma after sexual intercourse. J. Reprod. Med. 29:269. amino acids. Immunol. Lett. 30:59.
4. Stone, A. 2002. Microbicides: a new approach to preventing HIV and other sex- 28. Bukawa, H., K. Sekigawa, K. Hamajima, J. Fukushima, Y. Yamada, H. Kiyono,
ually transmitted infections. Nat. Rev. Drug Discov. 1:977. and K. Okuda. 1995. Neutralization of HIV-1 by secretory IgA induced by oral
The Journal of Immunology 7089
immunization with a new macromolecular multicomponent peptide vaccine can- 44. Bou-Habib, D. C., G. Roderiquez, T. Oravecz, P. W. Berman, P. Lusso, and
didate. Nat. Med. 1:681. M. A. Norcross. 1994. Cryptic nature of envelope V3 region epitopes protects
29. Åkerblom, L., J. Hinkula, P.-A. Broliden, B. Makitalo, T. Fridberger, J. Rosen, primary monocytotropic human immunodeficiency virus type 1 from antibody
M. Villacres-Eriksson, B. Morein, and B. Wahren. 1990. Neutralizing cross- neutralization. J. Virol. 68:6006.
reactive and non-neutralizing monoclonal antibodies to HIV-1 gp120. AIDS 45. Burton, D. R., and D. C. Montefiori. 1997. The antibody response in HIV-1
4:953. infection. AIDS 11:S87.
30. Devito, C., M. Levi, K. Broliden, and J. Hinkula. 2000. Mapping of B-cell 46. Dong, X.-N., Y. Xiao, M. P. Dierich, and Y.-H. Chen. 2001. N- and C-domains
epitopes in rabbits immunized with gag antigens for the production of HIV-1 gag of HIV-1 gp41: mutations, structure and functions. Immunol. Lett. 75:215.
capture ELISA reagents J. Immunol. Methods 238:69. 47. Conley, A. J., P. Conrad, S. Bondy, C. A. Dolan, J. Hannah, W. J. Leanza,
31. Mantis, N. J., P. A. Kozlowski, D. W. Mielcarz, W. Weissenhorn, and S. Marburg, M. Rivetna, V. K. Rusiecki, E. E. Sugg, et al. 1994. Immunogenecity
M. R. Neutra. 2001. Immunization of mice with recombinant gp41 in a systemic of synthetic HIV-1 gp120 V3-loop peptide-conjugate immunogens. Vaccine
prime/mucosal boost protocol induces HIV-1 specific serum IgG and secretory 12:445.
IgA antibodies. Vaccine 19:3990.
48. Alfsen, A., P. Iniguez, E. Bouguyon, and M. Bomsel. 2001. Secretory IgA spe-
32. LaCasse, R. A., K. E. Follis, M. Trahey, J. D. Scarborough, D. R. Littman, and
cific for a conserved epitope on gp41 envelope glycoprotein inhibits epithelial
J. H. Nunberg. 1999. Fusion-component vaccines: broad neutralization of pri-
transcytosis of HIV-1. J. Immunol. 166:6257.
mary isolates of HIV. Science 283:357.
33. Berger, E. A., P. M. Murphy, and J. M. Farber. 1999. Chemokine receptors as 49. Emini, E. A., W. A. Schleif, J. H. Nunberg, A. J. Conley, Y. Eda, S. Tokiyoshi,
HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu. Rev. Immu- S. D. Putney, S. Matsushita, K. E. Cobb, C. M. Jett, et al. 1992. Prevention of
nol. 17:657. HIV-1 infection in chimpanzees by gp120 V3 domain-specific monoclonal anti-
34. Wild, C., T. Oas, C. McDanal, D. Bolognesi, T. A. Mathews. 1992. Synthetic body. Nature 355:728.
peptide inhibitor of human immunodeficiency virus replication: correlation be- 50. Trkola, A., A. B. Pomales, H. Yuan, B. Korber, P. J. Maddon, G. P. Allaway,
tween solution structure and viral inhibition. Proc. Natl. Acad. Sci. USA H. Katinger, C. Barbas III, D. R. Burton, and D. D. Ho. 1995. Cross-clade neu-
89:10537. tralization of primary isolates of human immunodeficiency virus type 1 by mono-
35. Bomsel, M., M. Heyman, H. Hocini, S. Lagaye, L. Belec, C. Dupont, and clonal antibodies and tetrameric CD4-IgG. J. Virol. 69:6609.
C. Desgranges. 1998. Intracellular neutralization of HIV transcytosis across tight 51. Calarota, S., M. Jansson, M. Levi, K. Broliden, O. Libonatti, H. Wigzell, and
epithelial barriers by anti-HIV envelope protein IgA or IgM. Immunity 9:277. B. Wahren. 1996. Immunodominant glycoprotein 41 epitope identified by sero-
36. Belyakov, I. M., M. A. Derby, J. D. Ahlers, B. L. Kelsall, P. Earl, B. Moss, reactivity in HIV type 1-infected individuals. AIDS Res. Hum. Retroviruses
W. Strober, and J. A. Berzofsky. 1998. Mucosal immunization with HIV-1 pep- 12:705.
tide vaccine induces mucosal and systemic cytotoxic T lymphocytes and protec- 52. Kaneko, H., I. Bednarek, A. Wierzbicki, I. Kiszka, M. Dmochowski, T. J. Wasik,
tive immunity in mice against intrarectal recombinant HIV-vaccinia challenge. Y. Kaneko, and D. Kozbor. 2000. Oral DNA vaccination promotes mucosal and
Proc. Natl. Acad. Sci. USA 95:1709. systemic immune responses to HIV envelope glycoprotein. Virology 267:8.
37. Murphey-Corb, M., L. A. Wilson, A. M. Trichel, D. E. Roberts, K. Xu, 53. Kaul, R., S. L. Rowland-Jones, J. Kimani, T. Dong, H. B. Yang, P. Kiama,
S. Ohkawa, B. Woodson, R. Bohm, and J. Blanchard. 1999. Selective induction T. Rostron, E. Njagi, J. J. Bwayo, K. S. MacDonald, et al. 2001. Late serocon-
of protective MHC class I-restricted CTL in the intestinal lamina propria of version in HIV resistant Nairobi prostitutes despite pre-existing HIV-specific
rhesus monkeys by transient SIV infection of the colonic mucosa. J. Immunol. CD8 responses. J. Clin. Invest. 107:341.
162:540. 54. Margolis, H. S. 1993. Prevention of acute and chronic liver disease through
38. Belyakov, I. M., J. D. Ahlers, B. Y. Brandwein, P. Earl, B. L. Kelsall, immunizations: hepatitis B and beyond. J. Infect. Dis. 168:9.
B. L. Kelsall, B. Moss, W. Strober, J. A. Berzofsky. 1998. The importance of 55. Mainwright, R. B., B. J. McMahon, and L. S. Bulkow. 1989. Duration of im-
local mucosal HIV-specific CD8 cytotoxic T lymphocytes for resistance to mu- munogenecity and efficacy of hepatitis B in a Yupik Eskimo population. JAMA
cosal viral transmission in mice and enhancement of resistance by local admin- 261:2362.
istration of IL-12. J. Clin. Invest. 12:2072. 56. West, D. J., and G. B. Calandra. 1996. Vaccine induced immunologic memory for
39. Staats, H. F., C. P. Bradney, W. M. Gwinn, S. S. Jackson, G. D., Sempowski, hepatitis B surface antigen: implications for policy on booster vaccination. Vac-
H.-X. Liao, N. L. Letvin, and B. F. Haynes. 2001. Cytokine requirements for cine 1:1019.
induction of systemic and mucosal CTL after nasal immunization. J. Immunol.
57. Mittrucker, H. W., B. Raupach, A. Kohler, and S. H. Kaufmann. 2000. Cutting
167:5386.
edge: role of B lymphocytes in protective immunity against Salmonella
40. Lee, B., M. Sharron, C. Blanpain, B. J. Doranz, J. Vakili, P. Setoh, E. Berg,
typhymurium infection. J. Immunol. 164:1648.
G. Liu, H. R. Guy, S. R. Durrel, et al. 1999. Epitope mapping of CCR5 reveals
multiple conformational states and distinct but overlapping structures involved in 58. Yang, X., and R. C. Brunham. 1998. Gene knockout B-cell-deficient mice dem-
chemokine and coreceptor function. J. Biol. Chem. 274:9617. onstrate that B-cells play an important role in the initiation of T-cell responses to
41. Zuber, B., J. Hinkula, D. Vodros, P. Lundholm, C. Nilsson, A. Morner, M. Levi,
¨ ¨ ¨ Chlamydia trachomatis (mouse pneumonitis) lung infection. J. Immunol.
R. Benthin, and B. Wahren. 2001. Induction of immune responses and break of 161:1439.
tolerance by DNA against the HIV-1 coreceptor CCR5 but no protection from 59. Moore, J. P., and A. Trkola. 1997. HIV type 1 coreceptors, neutralization sero-
SIVsm challenge. Virology 278:400. types, and vaccine development. AIDS Res. Hum. Retroviruses 13:733.
42. Lehner, T., C. Doyle, Y. Wang, K. Babaahmady, T. Whittall, L. Tao, 60. Ruprecht, R. M., R. Hofmann-Lehmann, B. A. Smith-Franklin, R. A. Rasmussen,
L. Bergmeier, and C. Kelly. 2001. Immunogeniticy of the extracellular domains V. Liska, J. Vlasak, W. Xu, T. W. Baba, A. L. Chenine, L. A. Cavacini, et al.
of C-C chemokine receptor 5 and the in vitro effects on simian immunodeficiency 2001. Protection of neonatal macaques against experimental SHIV infection by
virus or HIV infectivity. J. Immunol. 166:7446. human neutralizing monoclonal antibodies. Transfus. Clin. Biol. 8:350.
43. Beddows, S., Louisirirotchanakul, R. Cheingsong-Popov, P. J. Easterbrook, 61. Baba, T. W., V. Liska, R. Hofmann-Lehmann, J. Vlasak, W. Xu, S. Ayehuni,
P. Simmonds, and J. Weber. 1998. Neutralization of primary and T-cell line L. A. Cavacini, M. R. Posner, H. Katinger, G. Stiegler, et al. 2000. Human
adapted isolates of human immunodeficiency virus type 1: role of V3-specific neutralizing monoclonal antibodies of the IgG1 subtype protect against mucosal
antibodies. J. Gen. Virol. 79:77. simian-human immunodeficiency virus infection. Nat. Med. 6:200.