Mucosal immunity and HIV-1 infection: applications for mucosal AIDS vaccine development.

Abstract

Natural transmission of human immunodeficiency virus type 1 (HIV-1) occurs through gastrointestinal and vaginal mucosa. These mucosal tissues are major reservoirs for initial HIV replication and amplification, and the sites of rapid CD4(+) T cell depletion. In both HIV-infected humans and SIV-infected macaques, massive loss of CD4(+) CCR5(+) memory T cells occurs in the gut and vaginal mucosa within the first 10-14 days of infection. Induction of local HIV-specific immune responses by vaccines may facilitate effective control of HIV or SIV replication at these sites. Vaccines that induce mucosal responses, in particular CD8(+) cytotoxic T lymphocytes (CTL), have controlled viral replication at mucosal sites and curtailed systemic dissemination. Thus, there is strong justification for development of next generation vaccines that induce mucosal immune effectors against HIV-1 including CD8(+) CTL, CD4(+) T helper cells and secretory IgA. In addition, further understanding of local innate mechanisms that impact early viral replication will greatly inform future vaccine development. In this review, we examine the current knowledge concerning mucosal AIDS vaccine development. Moreover, we propose immunization strategies that may be able to elicit an effective immune response that can protect against AIDS as well as other mucosal infections.

Full text

Page 1

Mucosal Immunity and HIV-1 Infection:
Applications for Mucosal AIDS Vaccine
Development
Igor M. Belyakov and Jeffrey D. Ahlers
Abstract Natural transmission of human immunodeficiency virus type 1 (HIV-1)
occurs through gastrointestinal and vaginal mucosa. These mucosal tissues are
major reservoirs for initial HIV replication and amplification, and the sites of rapid
CD4+T cell depletion. In both HIV-infected humans and SIV-infected macaques,
massive loss of CD4+CCR5+memory T cells occurs in the gut and vaginal
mucosa within the first 10–14 days of infection. Induction of local HIV-specific
immune responses by vaccines may facilitate effective control of HIV or SIV
replication at these sites. Vaccines that induce mucosal responses, in particular
CD8+cytotoxic T lymphocytes (CTL), have controlled viral replication at mucosal
sites and curtailed systemic dissemination. Thus, there is strong justification for
development of next generation vaccines that induce mucosal immune effectors
against HIV-1 including CD8+CTL, CD4+T helper cells and secretory IgA.
In addition, further understanding of local innate mechanisms that impact early
viral replication will greatly inform future vaccine development. In this review, we
examine the current knowledge concerning mucosal AIDS vaccine development.
Moreover, we propose immunization strategies that may be able to elicit an
effective immune response that can protect against AIDS as well as other mucosal
infections.
I. M. Belyakov (&)
Midwest Research Institute, 110 Thomas Johnson Drive, Suite 170,
Frederick, MD 21702, USA
e-mail: igorbelyakov@yahoo.com
J. D. Ahlers
National Institute of Allergy and Infectious Diseases, National Institutes of Health,
6700 Rockledge Drive, Bethesda, MD 20817, USA
Current Topics in Microbiology and Immunology (2012) 354: 157–179
DOI: 10.1007/82_2010_119
? Springer-Verlag Berlin Heidelberg 2010
Published Online: 4 January 2011
157

Page 2

Contents
1
2
3
4
5
Introduction........................................................................................................................
Role of Mucosal CD8+CTL in Protection Against Local Viral Infection.....................
Cytokine and Adjuvants for Enhancing Mucosal CTL Responses.................................
Mucosal Vaccination for Induction of Protective CTL in the Mucosa ..........................
Functional CD8+CTL for Preventing Immunodeficiency Virus Infection ....................
5.1Prime-Boost Strategies for Generating High-Avidity CTL ....................................
5.2Vaccine-Induced Mucosal High-Avidity CD8+CTL Preventing
Virus Dissemination from Mucosa..........................................................................
5.3Localization of High Quality CD8+CTL at Sites of Vaccine Delivery................
Conclusion.........................................................................................................................
References...............................................................................................................................
158
158
161
166
167
167
168
169
171
172
6
1 Introduction
Gastrointestinal (GI) and vaginal mucosal tissues are major sites of HIV entry and
initial infection (Veazey et al. 1998; Berzofsky et al. 2001; Belyakov and Berzofsky
2004; Neutra and Kozlowski 2006; Wilkinson and Cunningham 2006; Morrow et al.
2007; Belyakov and Ahlers 2008; Ahlers and Belyakov 2009a; Belyakov and Ahlers
2009b). An early sign of immunodeficiency virus infection is depletion of CD4+
CCR5+memory T-cells in the mucosa (Brenchley et al. 2004; Li et al. 2005;
Mehandruetal.2004;Veazeyetal.1998,2003).Dysfunctionofthemucosalimmune
system during the early stages of AIDS leads to major structural abnormalities in the
gutofinfectedindividualsandtothedevelopmentofopportunisticinfections(Clayton
et al. 2001; Heise et al. 1994; Kotler et al. 1984; Sharpstone et al. 1999). Subsequent
systemic immune activation is considered a hallmark of the disease. Thus, AIDS can
be considered primarily as a disease of the mucosal immune system (Belyakov and
Berzofsky 2004). There is substantial evidence indicating that virus replication is
rapidinthemucosa,eliminatingCD4+targetcellsbeforedisseminationintoblood2–7
dayslater(Spiraetal.1996;Zhangetal.1999).Thus,theinductionofCD8+CTLand
CD4+Thelpercellsinconcertwithprotectiveantibodieswillbeimportantcriteriafor
an effective HIV vaccine (Kozlowski et al. 1997; Murphey-Corb et al. 1999; Baba
etal. 2000; Barouch etal. 2000; Amara etal. 2001; Shiver etal. 2002; Kozlowskiand
Neutra 2003; McMichael 2006; Manrique et al. 2009; Sui et al. 2010).
2 Role of Mucosal CD8+CTL in Protection Against Local
Viral Infection
Initially, studies pertaining to the mucosal immunology of HIV infection focused
on induction of envelope specific antibody responses and the role of secretory IgA
(S-IgA) and IgG in viral control (Funkhouser et al. 1993; Mazzoli et al. 1997).
Failure to detect an early neutralizing antibody response in infected individuals,
158 I. M. Belyakov and J. D. Ahlers

Page 3

and the possibility that infection was primarily amplified by cell-to-cell spread,
suggested that an effective antibody response was subverted by HIV infection.
Furthermore, S-IgA antibodies were infrequently detected in mucosal secretions of
HIV-infected individuals (Mestecky and Jackson 1994). In contrast, a major role
for CD8+CTL in initial virus control was suggested in seminal studies showing
that the depletion of peripheral CD8+cells in SIV-infected macaques significantly
increased virus loads (Castro et al. 1992; Jin et al. 1999; Schmitz et al. 1999).
However, evidence of a role for mucosal CD8+CTL in control of immunodefi-
ciency virus infection was limited (Gallichan and Rosenthal 1996; Porgador et al.
1997; Belyakov et al. 1998a, b; Berzofsky et al. 1999; Berzofsky et al. 2004;
Belyakov and Ahlers 2008; Ahlers and Belyakov 2009a).
Early studies in our laboratory attempted to understand the role of mucosal
CD8+T cells in reducing mucosal viral loads and delaying the appearance of virus
in the blood. We asked whether immunization with CD4Th-CD8 epitope peptide
constructs through different mucosal routes (e.g., intrarectal, intragastric, intra-
nasal) could elicit HIV-specific CD8+CTL in small intestinal Peyer’s patches (PP)
or lamina propria (Belyakov et al. 1998b). Mice were immunized with 4 doses of
the synthetic HIV-1 Th-CTL envelope peptide construct, PCLUS3-18IIIB, on day
0, 7, 14 and 21 in combination with cholera toxin (CT) mucosal adjuvant. In a
comparison of the three mucosal routes of immunization, only intrarectal (I.R.)
immunization induced long-lasting, antigen-specific CD8+CTL memory in both
the inductive PP and lamina propria effector sites as well as in the spleen. The CTL
responses in spleen after I.R. immunization were similar to those induced by
subcutaneous (S.C.) immunization (Belyakov et al. 1998b). In contrast, S.C.
immunization with PCLUS3-18IIIB induced systemic CD8+CTL responses with
little evidence of mucosal CD8+T cell responses (Belyakov et al. 1998b).
Strong long-lasting mucosal CD8+CTL responses can also be generated by
mucosal immunization with recombinant vaccinia virus vectors (Belyakov et al.
1998d, 1999; Wyatt et al. 2008). In a follow up study, we demonstrated the
mucosal immunogenicity of replication-defective modified vaccinia Ankara
(MVA) virus expressing the HIV89.6 gp160 envelope protein in mice (Belyakov
et al. 1998d). A single I.R. immunization with MVA89.6 generated antigen-
specific CD8+CTL in both PP and intestinal lamina propria, at least as efficiently
as a replication-competent recombinant vaccinia virus expressing 89.6 gp160
(Belyakov et al. 1998d). Furthermore, CD8+CTL responses were detected in PP
up to 6 months after I.R. immunization with MVA89.6 and were slightly higher
than those after immunization with WR89.6 virus (Belyakov et al. 1998d). In
contrast, intraperitoneal (I.P.) immunization with MVA89.6 induced CTL in the
PP but not in the intestinal lamina propria. The magnitude of the response in PP of
I.P. immunized animals was modest compared to the spleen, and this result was
reproducible in three independent experiments (Belyakov et al. 1998d).
Transcutaneous immunization has also been shown to induce immune
responses in the GI tract (Glenn et al. 1998; Scharton-Kersten et al. 2000). The
application of antigen and adjuvant directly onto the skin has induced robust IgG
and S-IgA responses, as well as CD8+CTL in PP and lamina propria (Glenn et al.
Mucosal Immunity and HIV-1 Infection159

Page 4

2000; Gockel et al. 2000; Belyakov et al. 2004b). In addition, studies have
demonstrated protection against mucosal challenge with toxin or live virus fol-
lowing transcutaneous immunization (Glenn et al. 1998; Gockel et al. 2000;
Scharton-Kersten et al. 2000; Belyakov et al. 2004b). Transcutaneous vaccination
targets antigen to bone marrow-derived Langerhan’s dendritic cells (DC) resident
in the outer epidermal layers of skin. In a recent study, we demonstrated that
activated DC carrying skin-derived antigen migrate from the skin to PP and
present antigen directly to resident lymphocytes (Belyakov et al. 2004). By using
an in vivo pulsed antigen-presenting cell (APC)/T cell co-culture model for
tracking migrating APC by flow cytometry, we demonstrated that CD11c+DC
carrying skin-derived antigens can be isolated from PP inductive sites in intestinal
mucosa within 24 h following transcutaneous immunization with HIV peptide
vaccine and CT mucosal adjuvant (Belyakov et al. 2004). The ex vivo treatment of
bone marrow-derived CD11c+DC with vitamin D3, CT, or forskolin has been
shown to increase the ability of DC to migrate to inductive mucosal sites and to
induce mucosal immune responses (Enioutina et al. 2000).
Direct evidence supporting the ability of local CD8+CTL to mediate protection
against mucosal viral transmission has been difficult to obtain (Belyakov et al.
1998a). Furthermore, previous studies demonstrating protection against mucosal
viral challenge had not elucidated immune mechanisms involved in protection
(Marx et al. 1993). A number of studies have shown a role for CD8+CTL in
protection against mucosal infections, such as influenza (Gao et al. 1991; Ulmer
et al. 1993). However, understanding the role of CD8+CTL at mucosal sites of
infection in control and resolution of infection where antibody plays a prominent
role in protection is complex (Eichelberger et al. 1991; Lukacher et al. 1984;
Taylor and Askonas 1986). In an early study, we were able to demonstrate CTL-
mediated protection against mucosal viral challenge and showed that CD8+CTLs
present at the mucosal site of challenge were required for protection (Belyakov
et al. 1998a). Using a novel model mucosal viral challenge system with recom-
binant vaccinia virus that expresses HIV-1 gp160 in infected cells but not in the
virus particle (in order to eliminate the contribution of antibody-mediated
responses), we found that I.R. immunization with the synthetic HIV envelope
peptide vaccine, PCLUS3-18IIIB, induced a mucosal CD8+CTL response that
protected mice against vaccinia-gp160 challenge up to 6 months after mucosal
immunization (Belyakov et al. 1998a). Protection was attributed to a specific T cell
response against gp160 since mice were not protected against challenge with
vaccinia virus expressing an unrelated protein (Belyakov et al. 1998a). Impor-
tantly, protection against I.R. challenge with vaccinia-gp160 was dependent on
CD8+CTL as it was abrogated by treatment of I.R. immunized mice with anti-
CD8 antibody (Belyakov et al. 1998a). Because S.C. HIV peptide immunization,
which elicited a similar level of CD8+CTL in the spleen but not in the mucosa, did
not protect, we concluded that protection against mucosal challenge requires local
CD8+CTL. Local mucosal (but not systemic) delivery of IL-12 with CT in the
vaccine formulation significantly increased mucosal and systemic HIV-specific
CTL activity as well as the level of protection (Belyakov et al. 1998a).
160 I. M. Belyakov and J. D. Ahlers

Page 5

Furthermore, we showed that the effect of IL-12 was dependent on induction of
IFN-c, as no effect of IL-12 in enhanced protection was seen in IFN-c knock-out
mice. This was the first study to demonstrate that mucosal CD8+CTL can mediate
protection following local virus challenge in the mucosa (Belyakov et al. 1998a).
It is important to note that mucosal T cell responses and partial protection can
be achieved with systemic immunization routes (Kaufman et al. 2008; Lin et al.
2007; Pal et al. 2006; Tatsis et al. 2007). However, it is the authors’ opinion that
optimal mucosal immune responses and protective immunity are achieved through
oral, nasal, rectal, or vaginal mucosal immunization routes (Ahlers and Belyakov
2009a; Belyakov and Ahlers 2008, 2009). Also, it is important to state that
mucosal DC are the main target for mucosal vaccination and systemic immuni-
zation may have a limited effect on these DC. DC precursors that are recruited to
mesenteric lymph nodes (MLN) during inflammation are fully capable of secreting
IL-12 and are potent inducers of Th1 IFN-c responses. A recent study identified a
population of CD11chiCD11bhilamina propria DC that express toll-like receptor
(TLR)-5 and produce proinflammatory cytokines such as IL-6 and IL-12, but not
IL-23 or IL-10, in response to flagellin (Uematsu et al. 2008). Understanding the
unique properties of mucosal DC and the mucosal milieu in regulating local T and
B cell responses and ‘‘mucosal memory’’ will be important for the delivery of
vaccines that can provide protection against mucosal infections. Although the
magnitude, quality of response, and tissue residency of cells that migrate to
mucosal sites following systemic immunization needs further investigation, the
targeting of mucosal DC for induction of local immune responses by mucosal
vaccination has proven more effective for containing mucosal infections. Thus,
next generation HIV-1 vaccines and vaccines against other mucosal pathogens will
require formulations and delivery strategies which can effectively induce frontline
mucosal immune responses and memory (Belyakov and Ahlers 2009a).
3 Cytokine and Adjuvants for Enhancing Mucosal CTL
Responses
Cytokines and mucosal adjuvants are two major factors that can significantly
augment mucosal CD8+CTL responses and protective efficacy of mucosal vac-
cines (Table 1) (Beagley and Elson 1992; Belyakov et al. 1999a; Belyakov et al.
2000; Staats et al. 2001; O’Neill et al. 2002; Ahlers et al. 2003; Belyakov et al.
2004a; Zhu et al. 2008; Ahlers and Belyakov 2009b; Zhu et al. 2010). Identifi-
cation of cytokines, chemokines, and immunomodulatory molecules that augment
mucosal CTL responses and resistance to mucosal viral challenge has largely been
empirical (Belyakov et al. 1998c; Belyakov et al. 2004c; Belyakov et al. 2006a, b).
We asked whether the combination of GM-CSF, which recruits DC to inductive
sites, and IL-12, which drives CD4+Th1 function and CD8+CTL responses, could
enhance mucosal immunogenicity and protective efficacy of an HIV-1 peptide
vaccine given with mucosal adjuvants CT or LT(R192G). CT, one of the most
Mucosal Immunity and HIV-1 Infection161

Page 6

Table 1 Strategies for optimizing HIV vaccines
Strategy
Mechanism of activity
References
Mucosal route of immunization for
induction of CTL in the mucosa
Protection against mucosal viral transmission was
accomplished by establishing CD8+CTL in the
mucosal tissue prior to exposure. Generation of
functional CD8+CTL and compartmentalized
immunity by mucosal vaccination was associated
with the preservation of CD4+T cells in the
colonic lamina propria after mucosal challenge
with pathogenic virus
Barnett et al. (2008), Belyakov et al. (1998a, b, d,
1999, 2000, 2001a, 2006b, 2007a), Bruhl et al.
(1998), Caputo et al. (2008), Egan et al. (2004),
Kaneko et al. (2000), Li et al. (2008), Manrique
et al. (2009), Mercier et al. (2007), Pinczewski
et al. (2005), Ranasinghe et al. (2006), Sharpe et al. (2003), Shata et al. (2001), Vajdy et al.
(2001)
Heterologous mucosal prime/boost
High quality CD8+CTL responses were generated in
the intestinal mucosa after mucosal priming with
HIV gp160 envelope DNA vaccine and mucosal
boosting with recombinant viral vector expressing
the same envelope gene. A single systemicimmunization with rMVA was sufficient forinduction of high-avidity CD8+CTL in systemic
lymphoid organs, whereas a single mucosalimmunization with rMVA was not able to elicit
high-avidity CD8+CTL in the mucosa. A
heterologous mucosal DNA prime-viral vector
mucosal boost strategy was needed to induce
functional HIV-1-specific CD8+CTL in intestinal
mucosa
Allen et al. (2000), Amara et al. (2001), Belyakov
et al. (2008), Eo et al. (2001), Evans et al. (2003),
Gherardi et al. (2003), Gherardi et al. (2004),Hanke et al. (1998), Masopust et al. (2006),
Neeson et al. (2006), Peacock et al. (2004),
Ranasinghe et al. (2006), Ranasinghe et al. (2007),
Sharpe et al. (2003), Zhou et al. (2007), Huang
et al. (2007)
(continued)
162 I. M. Belyakov and J. D. Ahlers

Page 7

Table 1 (continued)
Strategy
Mechanism of activity
References
Generation of high avidity CTL
High-avidity CTL are readily activated by low
concentrations of peptide/MHC presented on targetcells, while low avidity CTL require higherconcentrations of peptide to become fully activated and exert effector function. Strong costimulation
skewCTLtoward higheraviditycells.High-avidity
CTL exert selective pressure on HIV during the
acute phase of infection, resulting in the emergence
of escape variants. Vaccines that induce high-
avidity mucosal CTL reduce dissemination of virus
from the mucosa to the blood
Alexander-Miller et al. (1996), Belyakov et al.
(2006b, 2007a, b, 2008), Bennett et al. (2007),
Dzutsev et al. (2007), Estcourt et al. (2002),
O’Connor et al. (2002), Oh et al. (2003),
Ranasinghe et al. (2007), Sedlik et al. (2000),
Snyder et al. (2003), Yoshizawa et al. (2003)
Inclusion of cytokines, chemokines,
costimulatory molecules, and TLR
ligands that enhance vaccine
efficacy
Cytokine and chemokine combinations and TLR-
triggering can help recruit monocytes,
macrophages and neutrophils to local lymph
nodes. TLR fusion proteins may help target
antigen to the appropriate APC, stimulate
maturation of DC, steer cellular immune responses
toward Th1-type, and enhance mucosal S-IgA and
IgG antibodies and isotype balance. Synergistic
combinations of cytokines and immunomodulating
molecules may be required for protection against mucosal challenge with virus. Mucosal adjuvant
LT(R192G) alone was as effective as CT plus
IL-12. GM-CSF synergized with LT(R192G).
A triple cytokine combination of GM-CSF, IL-12,
and TNF-a was synergistic for induction of CD8+
CTL and for antiviral protection. Choice of
adjuvants affects the interplay of cytokines and chemokines in regulation of mucosal CTL
Ahlers et al. (2001a, b, 2003), Belyakov et al. (1998c,
2000), Biragyn et al. (2002), Lena et al. (2002),
O’Neill et al. (2002), Staats et al. (2001), Staats
and Ennis 1999), Trumpfheller et al. (2008)
(continued)
Mucosal Immunity and HIV-1 Infection 163

Page 8

Table 1 (continued)
Strategy
Mechanism of activity
References
Counteracting Treg mechanisms that
dampen immune responses
Depletion of Treg with anti-CD25 antibody
significantly enhanced CD8+T cell
immunodominant responses in both the acute and
memory phases of the immune response. The
depletion of CD4+T cells enhanced long-lasting
CD8-mediated protective immunity upon protein
vaccination. In vivo inactivation experiments
attributed enhancement primarily to MHC class
II-restricted CD4+Treg cells which suppress the
differentiation process towards effector memory
CD8+T cells. Controlling suppressor effects at the
time of vaccination may produce more effective long-term immunity
Denning et al. (2007), Heit et al. (2008), Shevach
2002), and Suvas et al. (2003)
Push–pull approach to maximize
vaccine efficacy
A synergistic enhancement of vaccine mediated CD8+
CTL generation and antiviral protection by GM-CSF and the costimulatory molecule CD40L,
combined with the relief of suppression mediated
by CD4+Treg cells, including CD4+NKT cells
may provide optimum induction of CD8+CTL.
Cytokines, costimulatory molecule agonists, and
TLR synergies in combination with antibodies that
block IL-10 or IL-13 produced by Type II NKT
cells may provide optimum long term memory and
protection
Ahlers et al. (2002) and Sutmuller et al. (2001)
164 I. M. Belyakov and J. D. Ahlers

Page 9

commonly used mucosal adjuvants in experimental animals, is unsuitable for
humans because of potent toxicity associated with a massive luminal secretory
response(Xu-Amanoetal.1993;Marinaroetal.1995;Elson1996;Braunetal.1999).
A number of studies have introduced mutations into CT in an attempt to eliminate
the toxicity associated with the ADP-ribosyltransferase activity of the A subunit
and induction of cAMP in cells. These efforts have only been partially successful
since induction of cAMP has been shown to be necessary for adjuvant activity.
Dickinson and Clements constructed the LT(R192G) attenuated derivative of
Escherichia coli heat-labile enterotoxin (LT) using site-directed mutagenesis
(Dickinson and Clements 1995). They found that a single amino acid substitution
of Glycine for Arginine at position 192 within the disulfide-subtended region of
the LT A subunit separating A1 from A2 was effective for reducing toxicity while
retaining significant adjuvant properties (Dickinson and Clements 1995; Dickinson
and Clements 1996; Morris et al. 2000).
Using either CT or LT(R192G) and the HIV PCLUS3-18IIIB envelope peptide
construct, we were able to demonstrate a synergistic effect between these mucosal
adjuvants and the combination of GM-CSF and IL-12 for generating mucosal
CD8+CTL in I.R. immunized mice (Belyakov et al. 2000). Furthermore, I.R.
immunization with the HIV peptide plus LT(R192G) proved to be as effective for
induction of HIV-specific CD8+CTL in PP and intestinal lamina propria as native
LT or CT (Belyakov et al. 2000). After just two mucosal immunizations, the
combination of the two cytokines synergistically enhanced the CD8+CTL
response to the HIV-1 peptide vaccine (Belyakov et al. 2000). Co-administration
of GM-CSF and IL-12 with peptide also markedly enhanced protection against
mucosal challenge with vaccinia-gp160 when compared to animals immunized
with each cytokine alone (Belyakov et al. 2000). However, supplementation of
peptide vaccine with LT(R192G) and both cytokines afforded the greatest pro-
tection. The LT(R192G) adjuvant may produce a more favorable cytokine profile
since, in contrast to CT, it did not inhibit IL-12 production (Belyakov et al. 2000).
Also, much less IL-4 was induced by LT(R192G) than CT. Thus, the selection of
mucosal adjuvant may be critical for influencing the cytokine environment and the
induction of mucosal T cell responses that prevent viral transmission. It is
important to note that different strategies from those used to elicit optimal cellular
responses may be required for induction of humoral responses at mucosal sites.
New HIV vaccine design incorporating CD4+and CD8+T cell epitopes with
appropriate envelope immunogen structures and TLR agonists and the utilization
of mucosal delivery strategies that target mucosal DC may induce both high-
avidity CD8+CTL and local IgA and IgG neutralizing antibodies.
In a subsequent study, we demonstrated that GM-CSF, IL-12 and TNF-a also act
synergistically in the induction of CD8+ CTL following systemic immunization
(Ahlers et al. 2001a). The combination of IL-12 and TNF-a was essential for the
optimal development of Th1 responses to provide help for CD8+CTL induction in
vivo, while GM-CSF increased the number and activityof antigen-presenting DC in
draining lymph nodes where the immune response was initiated (Ahlers et al.
2001a). Most importantly, significant improvement in protection against viral
Mucosal Immunity and HIV-1 Infection 165

Page 10

challenge was achieved when the triple combination of cytokines (GM-CSF, IL-12
and TNF-a) was co-administered with peptide vaccine (Ahlers et al. 2001a).
The increased magnitude in CTL responses and protection against viral infection
affordedbysynergisticcombinationsofcytokinescouldbefurtherimprovedusinga
‘‘push–pull’’ approach to counteract natural negative regulatory mechanisms which
dampen Th1-type immune responses (Table 1) (Sutmuller et al. 2001; Ahlers et al.
2002).Inarecentstudy,weshowedthatbothTregulatory(Treg)andNaturalKiller-
T (NKT) cells suppress vaccine-induced immune responses (Ahlers et al. 2002).
We found that relief of suppression through in vivo depletion of regulatory CD4+
cells, including CD4+NKT cells, or blockade of IL-13 with an IL-13 receptor
competitive inhibitor significantly improved vaccine-mediated CD8+T cell
responses and protection against surrogate viral challenge. These results were
confirmed in CD1-deficient animals that lack NKT cells (Ahlers et al. 2002). We
reasoned that in mice in which CD4+T cells were depleted by antibody, the com-
binationofGM-CSFandCD40L mightsubstituteforCD4+Tcellhelp.Wededuced
that GM-CSF would recruit more professional APC to the draining lymph nodes
wheresolubleCD40Lwouldprovidematurationandactivationsignals(Ahlersetal.
2002). Indeed, GM-CSF and CD40L given with HIV peptide vaccine did act syn-
ergistically to enhance CTL responses in CD4-depleted mice. The improved CTL
responses achieved by this push–pull strategy translated to significant protection
against vaccinia-gp160 challenge (Ahlers et al. 2002). Thus, mucosal vaccination
strategiesthatutilizebothsynergisticcombinationsofTh1-promotingcytokinesand
approaches that inhibit negative regulatory mechanisms couldsignificantly enhance
protective immunity against mucosal HIV transmission.
4 Mucosal Vaccination for Induction of Protective CTL
in the Mucosa
Immune correlates of protection against HIV-1 are still not very well understood
(Acierno et al. 2006; Ahlers, 2009 #8052; McMichael 2006; Neutra and Kozlowski
2006; Belyakov et al. 2008). However, accumulating experimental evidence sug-
gests that mucosal immune responses (including S-IgA, IgG and CD8+CTL) in
combination with circulating HIV neutralizing antibodies, CD8+CTL and CD4+T
helper cells, and innate immune responses can exert varying degrees of control of
HIV or SIV replication (Baba et al. 2000; Mascola et al. 2000; Bafica et al. 2004;
Acierno et al. 2006; McMichael 2006). Disappointing results of the Merck vaccine
trial underscore the importance of eliciting frontline mucosal immune responses
that can significantly reduce virus load in the intestinal mucosa and subsequent viral
dissemination to blood and peripheral lymphoid tissues.
We next asked whether mucosal CD8+CTL induced by mucosal immunization
of rhesus macaques could impact the course of mucosal pathogenic retroviral
infection similar to our studies in the surrogate viral challenge model in mice
(Belyakov et al. 2001a). We found that SIV-specific CTL could be induced in the
166 I. M. Belyakov and J. D. Ahlers

Page 11

colon and mesenteric lymph nodes by I.R. immunization of monkeys with a syn-
thetic HIV envelope/SIV gag, pol peptide vaccine and LT(R192G) adjuvant
(Belyakov et al. 2001a). The SIV-specific intestinal CD8+CTL were able to traffic
to systemic lymphoid tissues. Interestingly, S.C. immunized monkeys also devel-
oped significant CD8+CTL in the mesenteric lymph nodes. However, after rectal
infection with chimeric simian–human immunodeficiency virus (SHIV)-ku2, the
monkeys immunized by the I.R. route demonstrated a more rapid decline in blood
and intestinal virus loads and a significantly lower viral load set point in blood when
compared to S.C. immunized animals or adjuvant only controls. We speculated that
more CD8+CTL induced by I.R. immunization were in the right place at the time of
I.R. challenge with SHIVku2 (Belyakov et al. 2001a). The numbers of CD4+and
CD8+T cells in I.R. immunized animals were also better preserved after SHIV
challenge when compared to S.C. immunized animals. Thus, CD8+CTL induced at
sites of mucosal challenge can significantly reduce immunodeficiency virus
infection in primates, and mucosal immunization may be optimal to parenteral
immunization for generating these cells. Rhesus macaques immunized by the
intranasal route with a T cell-inducing SIV DNA/MVA vaccine have similarly
demonstrated better control of rectal SIVmac251 infection when compared to
macaques given the same vaccine by the I.M. route (Manrique et al. 2009). These
studies provide strong rationale for the development of mucosal vaccines that
generate HIV-specific CTL and T helper cells at sites of HIV exposure in humans.
5 Functional CD8+CTL for Preventing Immunodeficiency
Virus Infection
5.1 Prime-Boost Strategies for Generating High-Avidity CTL
CD8+T cells that can be activated after recognition of peptide/MHC class I at low
peptide concentration are defined as high-avidity CD8+CTL, whereas those that
require high peptide concentrations are termed low-avidity CD8+CTL (Alexander-
Miller et al. 1996; Snyder et al. 2003; Belyakov et al. 2006b; Belyakov et al. 2007b;
Ahlers and Belyakov 2010c). It is well known today that high-avidity CD8+CTL
are more effective for preventing viral infections (Alexander-Miller et al. 1996;
Belyakov et al. 2006b, 2007a; Estcourt et al. 2002; Gallimore et al. 1998) and
eliminating tumors (Yee et al. 1999; Zeh et al. 1999). Also, a functional impairment
of HIV-specific CD8+CTL has been associated with clinical AIDS progression
(Acierno et al. 2006; Ahlers and Belyakov 2010a; Appay et al. 2000; Hel et al.
2001; McKay et al. 2002). Thus, immunization strategies that generate high-avidity
CD8+T cells in mucosal and systemic lymphoid tissues could significantly impact
initial virus infection and progression of disease. Different combinations of heter-
ologous prime-boost immunization protocols are currently being investigated in
multiple experimental and clinical trials for HIV-1, other infectious diseases and
Mucosal Immunity and HIV-1 Infection 167

Page 12

cancer (Belshe et al. 1998; Allenet al. 2000; Barouch et al. 2000; Amara etal. 2000;
Shiver et al. 2002; Gherardi et al. 2003; Dale et al. 2006) (Table 1).
In our studies, we employed a prime-boost immunization strategy consisting of
a DNA prime and a rMVA or recombinant adenovirus (rAd) virus boost, all
encoding an HIV-1 envelope protein, to evaluate immunization routes for ability to
induce mucosal as well as systemic HIV-specific CD8+CTL in mice (Belyakov
et al. 2008). A systemic I.M. prime-boost approach induced a strong CTL response
in the spleen specific for the immunodominant CTL envelope epitope P18-I10
(measured by51Cr-release assay, 7 days after in vitro stimulation with P18-I10-
peptide) and high-avidity CTL (determined by IFN-c ELISPOT assay using
titrated concentrations of P18-I10 peptide). However, the HIV-specific CTL
responses in Peyer’s patches were very low after I.M. DNA-MVA prime-boost
immunizations (Belyakov et al. 2008). When the prime and boost routes were
distinct, the delivery site of the boost had a greater impact than the site of DNA
priming. For example, I.M. DNA prime and I.R. MVA boost was more effective
than I.R. DNA prime and I.M. MVA boost for eliciting high-avidity CD8+CTL in
the intestine. The optimal CTL response in the gut was observed after I.R. priming
with HIV DNA vaccine and I.R. boosting with MVA (Belyakov et al. 2008). The
I.R. prime-boost strategy also induced a very strong systemic P18-I10-specific
CTL response. A single I.M. immunization with MVA was sufficient to elicit high-
avidity CD8+CTL in systemic lymphoid organs (Belyakov et al. 2008). However,
a single I.R. immunization with MVA was not able to elicit high-avidity CD8+
CTL in the mucosa. These results indicate that a mucosal prime-mucosal boost
strategy might be crucial to induce optimal cellular immunity in the mucosa. The
requirement of a booster vaccination for induction of functionally active CTL in
mucosal tissues using mucosal immunization routes may also be more stringent
than that for generating high-avidity CTL in systemic tissues using systemic
immunization routes (Belyakov et al. 2008).
5.2 Vaccine-Induced Mucosal High-Avidity CD8+CTL Preventing
Virus Dissemination from Mucosa
It is a strongly debated subject whether mucosal or systemic CD8+CTL are
necessary to prevent or reduce virus dissemination from the initial mucosal
infection site to systemic tissues (Belyakov et al. 2004a; Belyakov and Berzofsky
2004; Belyakov et al. 2006b; Neutra and Kozlowski 2006; Belyakov and Ahlers
2008; Kaufman et al. 2008; Ahlers and Belyakov 2010b; Ahlers and Belyakov
2010c). Also, the role of CTL avidity in control of mucosal AIDS virus trans-
mission is unknown. To address these questions, we used rhesus macaques to
compare a peptide-based vaccine, a viral vector-based vaccine, and a combination
peptide-prime/viral vector boost regimen (Belyakov et al. 2006b). We chose a
prime-boost strategy with replication-incompetent recombinant NYVAC poxvirus
expressing HIV envelope and SIV gag, pol proteins because similar systemic
168I. M. Belyakov and J. D. Ahlers

Page 13

prime-viral vector boost strategies have been shown to elicit strong systemic CD8+
CTL responses (Amara et al. 2001; Hanke et al. 1998; Hel et al. 2002; Shiver et al.
2002). The peptide vaccine contained a mixture of HIV and SIV CTL epitopes
presented by Mamu-A*01, the class I antigen expressed by the macaques selected
for this study (Belyakov et al. 2006b). All vaccines were delivered I.R. with a
combination of GM-CSF, IL-12 and CpG oligodeoxynucleotides (ODN) as adju-
vants. Four weeks after the last immunization, all macaques were challenged I.R.
with SHIVku2 and monitored for plasma viral loads and CD8+CTL responses.
Two weeks after the last immunization, we analyzed avidity of mucosal CD8+
CTL in the MLN. To examine avidity, the T cell responses were evaluated by
plotting specific lysis versus epitope concentration on peptide-coated target cells
(Belyakov et al. 2001b; Belyakov et al. 2006b). We found that both vaccination
regimens which included the peptide vaccine, GM-CSF, IL-12, and CpG ODN led
to similar avidity, whereas I.R. NYVAC immunization alone produced CTL
responses of lower magnitude and avidity. Thus, the peptide immunization and a
combination of cytokines and CpG ODN improved CTL avidity and functionality,
whereas boosting with NYVAC improved the magnitude of the CD8+CTL
responses but not the quality. The prime-boost regimen was necessary to obtain
responses with both the highest magnitude and avidity.
Next, we performed I.R. challenge and measured acute-phase peak viremia in
blood as an indicator of systemic dissemination (Belyakov and Ahlers 2008;
Belyakov et al. 2006b). We found that macaques given the peptide prime-poxvirus
boost exhibited a significant (2.5 week) delay in peak viremia compared to
macaques immunized with peptide or poxvirus alone. We interpreted this delay in
the peak viremia to most likely reflect a temporary local mucosal control of initial
virus replication by high-avidity CD8+CTL preventing rapid dissemination of
virus from the intestinal mucosa into the bloodstream. At day 17 after challenge,
when viral loads were near their peak, we found a strong inverse correlation
between viremia and the numbers of antigen-specific CD8+T cells in the colon but
not those in the blood. In addition, the animals that had CD8+CTL with the
highest avidity in MLN were those that demonstrated the best viral control
(Belyakov et al. 2006b). Thus, we demonstrated for the first time that a peptide-
prime and poxvirus-boost vaccine that induced high levels of high-avidity mucosal
CD8+CTL can delay dissemination of I.R. administered pathogenic SHIVku2 in
macaques, and that such protection correlates better with mucosal rather than
systemic CD8+CTL (Belyakov et al. 2006b).
5.3 Localization of High Quality CD8+CTL at Sites
of Vaccine Delivery
A number of recent studies have demonstrated that antigen-specific CD8+CTL and
partial protection against mucosal challenge with pathogenic SHIV can be achieved
after systemic vaccination (Horner et al. 2001; Pal et al. 2006; Stevceva et al. 2002;
Mucosal Immunity and HIV-1 Infection169

Page 14

Vogel et al. 2003). However, as we and others have shown, mucosal vaccination
can be even more effective for the clearance of virus from the major site of repli-
cation in the mucosa (Belyakov et al. 2006b, 2007a). The mechanism of protection
generated by mucosal vaccination is not well understood. For example, the local-
ization of CD8+effector and memory T cells after mucosal vaccination has not been
well characterized. However, for many mucosal intracellular pathogens (including
HIV and Herpes), an effective vaccine strategy will require induction and long-term
maintenance of antigen-specific B cells, CD4+Th cells and CTL at the site of viral
transmission. In one study, we performed mucosal versus systemic immunization
and compared CD8+CTL avidity in lymphoid tissues proximal and distal to the site
of immunization (Belyakov et al. 2007a). We observed a novel compartmentali-
zation of functional HIV-specific CD8+CTL in tissue most proximal to the site of
immunization (Belyakov et al. 2007a). In this study, mice were immunized with
MVA by the S.C. or I.R. immunization routes. To determine the extent of com-
partmentalization, we measured vaccinia B8R peptide-specific CD8+T cells by
tetramer staining (Tscharke et al. 2005) and by IFN-c production using ELISPOT
with cells from the spleen, small intestinal epithelium, and lamina propria
(Belyakov et al. 2007a). We found that both systemic and mucosal routes of
immunization generated vaccinia-specific CD8+T cells in both systemic and
mucosal compartments, when measured as total numbers of B8R tetramer-positive
CD8+T cells. However, when we characterized the functional activity of the cells
by IFN-c production, the cell distribution was asymmetric (Belyakov et al. 2007a).
The S.C. vaccination with MVA induced a significant number of IFN-c-producing
cells in the spleen, but not in the gut, while I.R. immunization generated greater
numbers of IFN-c secreting CD8+T cells in the intestinal epithelium and lamina
propria. Thus, mucosal immunization produced a much higher ratio of IFN-c-
secreting cells to the total B8R-tetramer positive cells in the gut when compared to
systemic immunization (Belyakov et al. 2007a). We also found that I.R. immuni-
zation induced more IL-12-producing DC in the colon, while S.C. vaccination
induced more IL-12-producing DC in axillary lymph nodes (ALN). Thus, differ-
ences in local DC activation could account for the differences in functional T cells
in proximal versus distal tissues (Belyakov et al. 2007a) (Fig. 1).
We also characterized the function and avidity of CTL in mucosal MLN and
systemic ALN of Mamu-A*01+ rhesus macaques after I.R. and S.C. vaccination
with an HIV–SIV Th-CTL peptide vaccine by using a functional51Cr release assay
with different concentrations of peptide (Kuroda et al. 1998)-coated target cells and
ELISPOT assay for IFN-c secretion (Belyakov et al. 2007a). We observed that after
I.R. immunization, specific lysis by MLN cells was very high against both low and
high concentrations of CTL peptide on target cells, indicating a large proportion of
high-avidity CD8+CTL. In contrast, S.C. vaccination induced greater levels of
high-avidity CD8+CTL in ALN and was less effective for induction of functional
CD8+CTL in MLN. Moreover, we found a strong inverse correlation between the
number of high-avidity CD8+CTL in the mucosal compartment and viral load in
the colon 200 days after I.R. challenge with SHIVku-2 (Belyakov et al. 2007a).
There was also a strong positive correlation between the percentage of CD4+T cells
170I. M. Belyakov and J. D. Ahlers

Page 15

in colonic lamina propria and the number of high-avidity antigen-specific CD8+
CTL in the same location (Belyakov et al. 2007a). Thus, a mucosal AIDS vaccine
reduced viral load and the depletion of CD4+T cells in the intestinal mucosa.
Control of mucosally transmitted immunodeficiency virus infection can be gener-
ated by local mucosal immunization, and the mechanism for this control can be
attributed to the focusing of high quality cellular responses at sites of viral exposure
(Belyakov et al. 2007a). The additional induction of mucosal and systemic anti-
bodies should improve protection even more significantly by working in concert
with CTL to prevent viral entry and replication in mucosal tissues.
6 Conclusion
Local mucosal CD8+CTL and antibody may completely prevent HIV transmission
at mucosal surfaces or potentially control virus replication within mucosal tissues
prior to systemic dissemination. We believe that a number of approaches can be
Fig. 1 Avidity of CTL in relation to vaccination route and protection. High-avidity CD8+CTL
are optimal to low-avidity CTL for containing immunodeficiency virus infection. Suboptimal
mucosal immunization (e.g., with vaccine lacking adjuvants) can induce high-avidity CTL at the
vaccination site but it also produces low-avidity CD8+CTL or T cell anergy, which results in
limited mucosal protection. Systemic immunization produces high-avidity CTL in systemic
tissues but not mucosal tissues, allowing rapid CD4+T cell depletion and viral spread to the
periphery. Optimal mucosal immunization establishes high-avidity CD8+CTL in both mucosal
and peripheral lymphoid tissues. Multiple variables influence the quality of the CTL response
achieved by mucosal vaccination including: dose, frequency, the use of synergistic combinations
of mucosal adjuvants, cytokines, chemokines, and TLR ligands to enhance vaccine efficacy.
Furthermore, heterologous prime-boost strategies and push–pull approaches will be needed to
maximize vaccine efficacy
Mucosal Immunity and HIV-1 Infection171