“In design sometimes one plus one equals three” J. Albers
“Come together” J. Lennon and P. McCartney
Adjuvants are usually defined as compounds that can
increase and/or modulate the intrinsic immunogenicity
of an antigen, and the term adjuvant is itself derived
from the latin adjuvare, meaning to help. Why do vac-
cines need help, and why do new vaccines need more
help than vaccines that are well established? To reduce
reactogenicity, new vaccines have a more defined com-
position that is often linked to lower immunogenicity
compared with previous whole-cell or virus-based
vaccines. Adjuvants are therefore required to assist
new vaccines to induce potent and persistent immune
responses, with the additional benefits that less antigen
and fewer injections are needed. Moreover, new vaccine
targets often require the induction of strong cellular
responses, including T helper (TH) cells and sometimes
cytotoxic T lymphocytes (CTLs), in addition to anti-
bodies. As aluminium salts, the adjuvants that have
been most widely used in humans so far, predominantly
induce antibody responses, discovering new adjuvants
is crucial for the development of vaccines that require
a cell-mediated response (for reviews on vaccines and
adjuvants see REFS 1–3).
Although much of the adjuvant research that was
carried out in the past can be seen as empirical, this
research did sometimes give rise to potent and useful
products. Nevertheless, there is a need to develop a new
generation of adjuvants that are rationally designed on
the basis of the recent progress that has been made in our
understanding of the immune response, particularly the
innate immune response. The first part of this Review
will present an overview of this progress, which has had
a dramatic impact on adjuvant research. The second part
will present the practical applications of this knowledge
and the new perspectives that have been opened in vac-
cinology as a result. As live vaccines and vectors usually do
not need adjuvants this Review will focus on inactivated
and inert vaccines.
It is important to stress that adjuvants are a heteroge-
neous group of compounds. The term adjuvant is often
used as a synonym for immunostimulant, but whereas
immunostimulants are generally single compounds with
intrinsic immunostimulant and/or immunomodula-
tory properties, adjuvants can be composed of differ-
ent constituents with different functions and activities,
including carrier/depot or targeting functions and
immunostimulant and/or immunomodulatory activity.
The current challenge facing adjuvant research is to find
the ‘perfect mix’; an optimal, safe formulation, the dif-
ferent components of which will be not only additive but
synergistic, and which will eventually drive the desired
immune response. In this Review I will attempt to show
that in adjuvant research, as in design, one plus one often
equals more than two, and that combining different vac-
cine constituents in appropriate formulations can further
enhance and target such synergy.
Research Department, sanofi
pasteur, Campus Merieux,
69280 Marcy l’Etoile, France.
Published online 11 June 2007
T helper cell
(TH cell). A T cell that has cell-
surface antigen receptors that
bind fragments of antigens
displayed by MHC class II
molecules, which are
expressed at the surface of
Activated TH cells express
cytokines and membrane-
molecules that help other
immune cells carry out their
specific functions. TH cells can
be divided into subsets
according to their cytokine-
The perfect mix: recent
progress in adjuvant research
Abstract | Developing efficient and safe adjuvants for use in human vaccines remains both
a challenge and a necessity. Past approaches have been largely empirical and generally
used a single type of adjuvant, such as aluminium salts or emulsions. However, new vaccine
targets often require the induction of well-defined cell-mediated responses in addition
to antibodies, and thus new immunostimulants are required. Recent advances in basic
immunology have elucidated how early innate immune signals can shape subsequent
adaptive responses and this, coupled with improvements in biochemical techniques, has
led to the design and development of more specific and focused adjuvants. In this Review,
I discuss the research that has made it possible for vaccinologists to now be able to choose
between a large panel of adjuvants, which potentially can act synergistically, and combine
them in formulations that are specifically adapted to each target and to the relevant
correlate(s) of protection.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 505
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
(DC). ‘Professional’ antigen-
presenting cells that are found
in the T-cell areas of lymphoid
tissues and as minor cellular
components in most tissues.
They have a branched or
dendritic morphology and are
the most potent stimulators of
naive T-cell responses.
CD4+ T cell
A subpopulation of T cells that
express the CD4 receptor.
These cells aid in immune
responses and are therefore
referred to as T helper cells.
CD8+ T cell
A subpopulation of T cells that
express the CD8 receptor.
CD8+ T cells recognize antigens
that are presented on the
surface of host cells by MHC
class I molecules, leading to
their destruction, and are
therefore also known as
cytotoxic T lymphocytes.
Regulatory T (TReg) cell
A population of CD4+ T cells
that naturally express high
levels of CD25 (the interleukin-
2 receptor α-chain) and the
transcription factor forkhead
box P3 (Foxp3), and that have
suppressive regulatory activity
towards effector T cells and
other immune cells.
(T helper 1 cell). A type of
activated TH cell that promotes
responses associated with the
production of a particular set of
cytokines, including interleukin
(IL)-2, interferon (IFN)-γ and
tumour-necrosis factor (TNF),
the main function of which is to
mediated defences against
(T helper 2 cell). A type of
activated TH cell that
participates in phagocytosis-
that are induced by TH1 cells.
TH2 cells secrete interleukin
(IL)-4, IL-5 and IL-6.
A fresh look at innate immunity
Until recently, innate immunity was largely seen as less
‘noble’ than adaptive immunity, except for the few com-
ponents, such as complement and phagocytes, that had
the privilege of cooperating with antibodies to eliminate
pathogens. Innate immunity was thought of as merely the
first line of defence, providing non-specific microbicidal
activity and early inflammatory signals, and thus giving
more time for adaptive immunity to develop. However,
it is now clear that the adaptive immune response mainly
depends on the level and specificity of the initial ‘dan-
ger’ signals4 perceived by innate immune cells following
infection (and vaccination). Some innate and adaptive
cells have long been known to be mandatory for initiat-
ing immune responses by presenting antigen to T cells.
Among these antigen-presenting cells (APCs), dendritic
cells (DCs) are the key sentinels for priming naive
In parallel with the renaissance in innate immunity,
researchers have continued to characterize the adap-
tive response and clarify the respective roles of B cells,
and CD4+ T cells and CD8+ T cells. In particular, vari-
ous CD4+ TH cell subsets have been defined through
their cytokine profiles and ability to modulate B cell
and CD8+ CTL responses6–8 (BOX 1). In the continuous
football match involving the human immune system
and pathogens as the opposing teams, TH cells are the
key midfielders who conduct and organize the game
between the B-cell defenders and CTL strikers. It is
important to have referees keeping each game under
control and eventually signalling full time, and this
role is filled by regulatory T (TReg) cells, which are capable
of modulating both TH1 and TH2 responses9. Finally,
T- and B-cell memory responses have also been
dissected, allowing the identification of various means
of inducing different subsets of memory cells with
The overall consequence of these developments was
to strengthen the link between the initial innate and
subsequent adaptive response. The crucial steps
and signals required to induce T- and B-cell responses
have been defined, demonstrating that innate immune
signals modulate not only the magnitude of the adap-
tive response but also the quality of this response. For
instance, secretion of pro-inflammatory cytokines such
as interleukin (IL)-12 will drive TH1-cell polarization
and subsequent interferon (IFN)-γ secretion by T cells.
These findings are summarized in BOX 1 and FIG. 1,
which present the different signals required to initiate a
potent immune response as signal 0 (antigen recognition
and APC activation), signal 1 (antigen presentation) and
signal 2 (co-stimulation)12. Adjuvants can act on each
of these three signals. Although targeting co-stimula-
tory molecules (signal 2) directly through antibodies,
cytokines or chemokines is an interesting possibility13,
caution is advised as this might result in non-focused
over-stimulation (discussed in more detail in the later
section on safety). This approach might be more usefully
applied in therapeutic settings rather than preventative
settings, and will not be discussed any further here.
TLRs and innate immunity
Innate immunity has regained a primary role, not only
chronologically but also functionally through its impor-
tance in shaping the adaptive response. Applied research
has directly benefited from these basic findings, thanks
in particular to the discovery of the Toll-like receptors
(TLRs) and other pattern-recognition receptors (PRRs).
Box 1 | Innate and adaptive immunity: from signals 0, 1 and 2 to TH0, 1 and 2 and beyond
The initiation of T helper (TH)-cell responses requires different signals from antigen-presenting cells (APCs). Signal 1 is
triggered by specific peptide presentation by class II major histocompatibility (MHC) molecules to the T-cell receptor (TCR).
However, unless another signal is given, anergy and abortive responses are induced. To avoid anergy, the co-stimulatory
signal, signal 2, is needed, through receptor–ligand interaction between APCs and T-cell antigens, such as CD40–CD40L
(CD154), OX40 (CD134)–OX40L (CD134L), CD80 (B7.1) or CD86 (B7.2)–CD28. Alternatively, interactions between CD80 or
CD86 and CD152 (CTLA4) or between PD1L (B7H1) or PDL2 (B7DC) and PD1 downregulate T-cell activation.
It was subsequently discovered that an additional, and earlier, signal called signal 0, was mandatory to activate APCs and
orientate the subsequent TH response. Signal 0 is mostly induced through the recognition of pathogen-associated
molecular patterns (PAMPs) by pathogen-recognition receptors (PRRs), including Toll-like receptors (TLRs)4,5. These signals
might also apply to class I MHC presentation by APCs for CD8+ T cell stimulation.
The TH1/TH2 cell dichotomy was first described 20 years ago6,7. The TH1-type response is accompanied by secretion of
interleukin-2 (IL-2) and interferon-γ (IFN-γ), favouring cell-mediated immune responses, activation of CD8+ cytotoxic T cells
(CTLs), and the generation in mice of the complement-fixing antibody IgG2a (IgG2c in strains lacking this IgG subclass),
and IgG3. The TH2-type response is characterized by secretion of IL-4, IL-5 and IL-6, providing B-cell help, and the
preferential induction of IgG1, IgE and IgA in mice, and IgE and IgG4 in humans. It is important to note that although TH1-
cell adjuvants promote the activation of CD4+ IFN-γ-secreting cells, this is not sufficient to trigger the activation of CTLs,
which also require class I MHC antigen presentation. More recently, IL-17-expressing TH17 cells have been linked to early
inflammation, defence against extracellular bacteria and parasites, and autoimmunity7.
In addition, regulatory T (TReg) cells, initially called suppressor cells, are undergoing a re-discovery and several types have
been identified, including CD4+ CD25+ Foxp3+ TReg cells, and Tr1 and TH3 TReg cells9. These cells regulate the balance
between tolerance and immune responses involving both T and B cells, in particular through the action of transforming
growth factor β and/or IL-10 and the membrane-bound molecules cytotoxic T-lymphocyte-associated antigen 4 (CTLA4)
and glucocorticoid-induced tumour necrosis factor receptor (GITR).
Finally, important progress has been made in elucidating how T- and B-cell memory is induced, identifying T effector and
central memory subsets, and still involving innate immune signals, in particular TLRs10,11,34 (see also BOX 2).
506 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
(PRR). A host receptor (such as
Toll-like receptors (TLRs) or
NOD-like receptors (NLRs))
that can sense pathogen-
associated molecular patterns
and initiate signalling cascades
that lead to an innate immune
response. These can be
membrane bound (such as
TLRs) or soluble cytoplasmic
receptors (such as NLRs).
(PAMP). A molecular pattern
that is found in microorganisms
but not mammalian cells.
Examples include bacterial
hypomethylated DNA, flagellin
and double-stranded RNA.
Following host invasion, microorganisms express-
ing pathogen-associated molecular patterns (PAMPs) are
recognized by innate immune cells through specific
and non-polymorphic PRRs. Among the PRRs, TLRs
have been shown to have a crucial role in the early steps
of the immune response to infection. These receptors
were discovered in human cells14 and named after the
Drosophila melanogaster Toll gene, the protein product
of which is involved in innate immunity and develop-
ment15. The importance of Toll signalling in mammals
was confirmed by the observation that TLR4 is involved
in lipopolysaccharide (LPS) recognition16. This seminal
finding opened the way for an ever-increasing number of
investigations focusing on the different roles of TLRs in
APC activation and T- and B-cell responses (for reviews
see REFS 17,18).
BOX 2 and FIG. 2 present some key features of TLR dis-
tribution and localization, TLR agonists and downstream
signalling pathways and their role in the modulation of
the adaptive response. TLRs are expressed on the surface
or in the endosomes of different APC subtypes19,20 and
they respond to, among other stimuli, specific bacterial,
viral, fungal or protozoan signals. This activates APCs
and modulates and shapes the adaptive response, for
instance, by shifting the balance towards TH1 or TH2 cells
(REFS 21,22) and activating or inhibiting TReg cells.
Importantly, TLRs not only trigger APC activation
(signals 0 and 2), they also have a role in antigen pre-
sentation (signal 1). The presence of both antigen and
TLR is required for optimal antigen presentation and
activation: TLRs control the generation of T-cell receptor
(TCR) ligands from the phagosome, which ensures that
the contents derived from microbial pathogens are prefer-
entially presented to T cells by the activated APC. This
was first demonstrated in vitro23, and the importance of
such selection has also been demonstrated in vivo with
the generation of an immunodominant response against
Toxoplasma profilin, a TLR11 agonist24. This highlights
the need for antigen and TLR-agonists to be co-delivered,
in order to be present in the same phagosome cargo and
induce optimal antigen presentation and stimulation of
the subsequent T-cell response. Moreover, direct TLR
stimulation is also needed to fully activate B cells25,26.
Figure 1 | Where do adjuvants act? The initiation of T helper (TH)-cell responses
requires three signals, referred to as signal 0, signal 1 and signal 2 (see BOX 1).
In theory adjuvants can act — alone or in combination — on each of these three signals,
and an adjuvant acting on each signal is shown. Most of the recently developed
specific adjuvants, such as Toll-like receptor (TLR) agonists, can be considered as
type A adjuvants: they act on signal 0, and indirectly on signal 2, by activating antigen-
presenting cells (APCs) and triggering the secretion of cytokines such as interleukin
(IL)-12. In addition, TLR agonists can act on signal 1 by favouring efficient presentation of
the co-administered antigen (Ag). Some TLR agonists also directly trigger signal 0 on
regulatory T cells and B cells expressing some of the corresponding receptors. Adjuvants
and formulations targeting APCs or favouring Ag capture can be viewed as type B
adjuvants acting on signal 1, as their effect is eventually mediated by enhanced Ag
presentation to T cells. Liposomes, microspheres and some emulsions are in this category.
As stressed in this Review, targeting signal 1 is not sufficient and an immunostimulatory
signal should be co-delivered for an optimal response. In this regard, some specific
ligands of co-stimulatory molecules can directly enhance signal 2, acting as type C
adjuvants; such compounds must however be used with caution.
Box 2 | Toll-like receptors: expression, ligands and signalling
Toll-like receptors (TLRs) are pathogen-recognition receptors (PRRs). They are membrane-anchored proteins exposing
leucine-rich-repeat (LRR) domains in extracellular or luminal (endosomal) compartments, and can therefore interact
with pathogen-associated molecular patterns (PAMPs). The cytoplasmic tail of TLRs contains Toll–interleukin 1
receptor (TIR) domains, that can interact with downstream signalling pathways. TLRs mostly recognize foreign
antigens, although it has been shown that endogenous proteins such as heat-shock proteins can stimulate some TLRs.
Evidence of direct binding of agonists to the corresponding TLRs is still lacking in most cases; for instance, TLR4 does
not bind lipopolysaccharide (LPS) directly123.
As shown in FIG. 2, TLRs 1, 2, 4, 5, 6 and 11 are surface-exposed, whereas TLRs 3, 7, 8 and 9 are located within
endosomes. In addition to differences in cellular sub-compartmentalization, there are also differences in their cellular
distribution: in humans, myeloid dendritic cells (DCs) express all TLRs except TLR7 and 9, which are selectively
expressed by plasmacytoid DCs20. Epidermal Langerhans cells express mRNA encoding TLRs 1, 2, 3, 5, 6 and 10, but do
not directly sense TLR4 or TLR-7/8 agonists19.
TLR sensing of their natural or synthetic agonists can orientate the immune response towards a TH1- or TH2-cell
response21,22 (FIG. 2) and promote regulatory T (TReg) cell activity29–31. In particular, LPS agonists, imidazoquinolines and
unmethylated CpG oligonucleotides induce TH1 responses after sensing by TLR4, 7/8 and 9, respectively.
In this regard, all TLRs were initially considered to transduce similar transactivation and phosphorylation pathways
through their common adaptor protein MyD88, leading to the activation of the transcription factor nuclear factor (NF)-κB
and the JNK and p38 kinases. This pathway is crucial for inducing antigen-specific TH1 responses, as observed in MyD88-
deficient mice. However, although most TLRs trigger MyD88-dependent pathways, TLR3 stimulates MyD88-independent
pathways and TLR4 can signal through both pathways; for both TLR3 and TLR4 this depends on their coupling with the TRIF
(TIR-domain-containing adaptor protein inducing IFN-β). In addition, TLR negative regulators have progressively been
identified (interleukin-1 receptor-associated kinase M (IRAK-M) and Tollip), which help to keep inflammation under control.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 507
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
Gram positive and Gram negative
polyanionic ligands, LDL, apoptotic cells
5′-triphosphate RNA, ssRNA
influenza virus, JEV)
Poly (I:C), picornavirus
Peptidoglycan, muropeptides, DAP
Peptidoglycan, muropeptides, MDP
Bacterial RNA, uric acid crystals, toxins, MDP
dsRNA, Poly (I:C)
(Imiquimod, resiquimod (R848))
Unmethylated CpG DNA,
Endosome TLRs and agonists
Triacyl lipopeptides, Pam3Cys
Lipoteichoic acids, Zymosan
LPS, MPL, LPS analogues, Taxol
Surface TLRs and agonists
However, some researchers have questioned the need
for TLRs in inducing antibody responses27,28, and this
will be discussed in more detail later.
TLR stimulation also regulates TReg activity, directly
through TLR recognition on TReg cells, or indirectly
through APC–TReg interactions29. The consequence of
direct stimulation is to activate TReg cells, as has been shown,
for example, for TLR2 and TLR4 (REFS 30,31), or to reverse
their inhibitory effect, as has been shown for TLR8 (REF. 32).
Indirect stimulation through TLR-linked APC stimulation
inhibits TReg cell differentiation33. This shows that TLR
recognition not only induces inflammatory responses,
but can also regulate these responses by triggering
the expansion of TReg cells, which will limit potentially
harmful responses through a feedback mechanism once
the infection is resolved. Finally, it has been shown that
TLR-induced signals are important for the generation of
memory CD4+ T cells, but not for their activation34.
It was then logical to translate these basic discover-
ies into practical applications in adjuvant research. TLR
agonists can act as a coach for the immune system team:
they select and stimulate the most appropriate players
with respect to the optimal strategy to oppose each par-
ticular team of pathogens. Vaccines could benefit from
having such coaches, and this will be developed in the
Figure 2 | Potential targets for adjuvants and formulations. Different types of receptors on antigen-presenting cells
(APCs) can be targeted by adjuvants. Each particular type of APC can express a restricted array of these receptors, which
can regulate each other’s signalling through some cross-talk within the cell. First, some synthetic agonists of Toll-like
receptors (TLRs) (the names of synthetic agonists are highlighted in red) are among the most promising and efficient
adjuvant candidates, able to sense surface- and endosome-located TLRs, the activation of which will orientate T helper
(TH) 1 and/or TH2 responses. The cytosolic NOD-like receptors (NLRs) and RIG-like helicases (RLHs) can sense signals
from intracellular bacteria and viruses, respectively. Although, so far, less development has been carried out in humans
with synthetic agonists of these receptors, compounds such as muramyl dipeptide (MDP) have been used with some
success in the veterinary field. Other receptors such as C-type lectin receptors (CLRs) and scavenger receptors can
constitute targets for formulations aiming at enhancing antigen capture and presentation should a stimulatory signal be
co-delivered to avoid tolerance induction. Finally, TREM receptors (triggering receptors expressed on myeloid cells) are
also involved in early inflammation, although their ligands are still unknown. DAP, diaminopimelic acid; JEV, Japanese
encephalitis virus; LDL, low density lipoprotein; LPS, lipopolysaccharide; MPL, monophosphoryl lipid A; NOD,
nucleotide-binding oligomerization domain; Pam3Cys, tripalmitoyl-S-glyceryl cysteine; VSV, vesicular stomatitis virus.
508 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
(central or peripheral) results
from different mechanisms
preventing the immune system
mounting responses against
(self) antigens. Peripheral
tolerance occurs when mature
antigens and undergo anergy,
deletion or suppression. In
particular, T-cell anergy is
defined by defective
proliferation by previously
primed T cells following
restimulation, reflecting a
selective defect in the
activation of some TCR-
induced signalling pathways.
Natural killer T cells
(NKT cells). A subpopulation of
T cells that expresses both
NK-cell and T-cell markers.
γδ T cells
A minor population of T cells
that express the γδ T-cell
receptor (TCR), and that are
more abundant in epithelial-
rich tissues such as the skin,
gut and reproductive tract.
Like NKT cells, γδ T cells can
be cytolytic and produce high
levels of cytokines and
Non-TLR innate receptors
TLRs sense extracellular signals, and it therefore fol-
lows that intracellular cytosolic counterparts should
sense intracellular signals. Such cytosolic sensors
include NODs (nucleotide-binding oligomerization
domain proteins), NOD-like receptors (NLRs) and
RIG-like helicases (RLHs), which sense signals from
intracellular bacteria and viruses, respectively35 (FIG. 2).
As for TLRs, agonists of such receptors could in theory
be used as adjuvants, although their effects would be
less focused as NLRs and RLHs are more ubiquitous
NODs and NLRs. NOD proteins and NLRs have in com-
mon a C-terminal leucine-rich-repeat (LRR) domain,
a central nucleotide-binding domain and N-terminal
protein–protein interaction CARD (caspase activation
and recruitment domain) and pyrin or BIR (baculovirus
inhibitor-of-apoptosis repeat) domains36. NOD1 and
NOD2 detect distinct substructures from bacterial pep-
tidoglycan: NOD2 detects muramyl dipeptide (MDP)
from Gram-negative and Gram-positive bacteria and
NOD1 senses meso-diaminopimelic acid (meso-DAP),
which is found in Gram-negative bacteria and in Gram-
positive rods37. NLRs sense danger-associated host
components such as uric acid crystals35. MDP has been
successfully used as an adjuvant in veterinary vaccines,
such as in leishmaniasis, inducing TH1 responses and
protection38. A synthetic lipophilic glycopeptide, mur-
was also shown to have adjuvant and immunostimulatory
properties in humans.
RIG/MDA. Retinoic acid-inducible gene I (RIG-I) and
melanoma-differentiation-associated gene 5 (MDA5) are
RLH proteins that share two N-terminal CARD domains
followed by an RNA helicase domain. These sensors are
activated by viral infection, and trigger cooperative
activation of nuclear factor (NF)-κB and IFN regula-
tory factors 3 and 7 (IRF3/7) to induce antiviral type I
IFNs (IFNα and IFNβ). RIG-I responds to RNA viruses,
including paramyxoviruses, influenza virus and Japanese
encephalitis virus, whereas MDA5 detects picorna-
viruses39. The ligand for RIG-I was recently shown to be
triphosphate RNA (3pRNA)40,41.
Other receptors involved in antigen capture and rec-
ognition. APCs express other receptors involved in
antigen capture and recognition, such as scavenger
receptors, C-type lectin receptors (CLRs) and trig-
gering receptors expressed on myeloid cells (TREMs)
(FIG. 2). Scavenger receptors bind polyanionic ligands
and can internalize not only pathogens but also host
components such as apoptotic cells and modified
low-density lipoproteins42. CLRs, including DC-SIGN
and the mannose receptor, can bind a wide range
of viruses, bacteria and fungi through recognition of
sugar moieties, such as N-acetyl-glucosamine, man-
nose, N-acetyl-mannosamine, fucose and glucose43.
However, it appears that the first role of scavengers and
CLRs is not in pathogen recognition, and pathogens
seem to be able to use these host–host recognition
systems for their own benefit. In contrast to TLRs,
the primary function of which is to discriminate
between self and non-self, targeting scavengers
and CLRs require a co-stimulatory signal to induce
immunity rather than tolerance. This has been shown,
for instance, for the DEC-205 receptor, a decalec-
tin involved in uptake and presentation by DCs.
Targeting this receptor in the absence of a maturation
stimulus leads to CD8+ T-cell tolerance44. A ‘yin yang’
regulation has been proposed to describe the crosstalk
between TLRs and CLRs45. These receptors neverthe-
less have a practical interest for vaccine researchers,
as their targeting could focus antigen and adjuvant
to the target cells, increasing both immunogenicity
and safety. Finally, TREMs have been characterized
as amplifiers of immune responses but so far their
specific ligands have not been identified46.
In addition to these receptors, two additional players
acting at the border between innate and adaptive immu-
nity are worth mentioning, natural killer T (NKT) cells and
γδ T cells. NKT cells can be activated in particular by a
synthetic glycolipid, α-galactosylceramide (α-GalCer).
This compound and its derivatives have demonstrated
immunomodulatory properties in various settings,
including experimental models of cancer47. The γδ
T cells are specifically activated by small antigens, such
as organic phosphoesters, which are also potential adju-
vants. Moreover, some γδ T cells can themselves act as
APCs (for a review see REF. 48).
Practical applications to the adjuvant field
Synthetic TLR agonists. The identification of natural
TLR agonists has led to the design of synthetic ligands
that can target TLRs more precisely and safely than
pathogen-derived ligands, and which were selected by
their ability to bind receptors and activate downstream
signalling pathways49. Interestingly, some adjuvants and
immunostimulants that had been identified previously
were later shown to be TLR agonists: unmethylated viral
or bacterial CpG DNA and oligonucleotides are agonists
for TLR9; poly (I:C) for TLR3; lipopeptides and tripal-
mitoyl-S-glyceryl cysteine (Pam3Cys) for TLR2; LPS
and its derivatives for TLR4; and imidazoquinolines for
TLR7/TLR8 (REF. 50).
Additional TLR agonists have since been synthe-
sized and tested in different experimental models, such
as second-generation oligonucleotide TLR9 agonists,
with sequences that have been optimized with respect
to the species (for instance mouse or human), type of
immune response51,52 and stability. The identification of
TLRs also allowed the generation of new tools such as
transformed cell lines expressing a given array of TLRs,
in which activation of downstream signalling pathways
can be followed using, for example, an NF-κB–luciferase
reporter gene. Such cell lines have been used to investi-
gate the activity of new TLR4 agonists with a modified
backbone53. Other second-generation TLR4 agonists
have been obtained, such as the RC529 adjuvant54. New
synthetic TLR7/8 agonists such as 3M-019 have also
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 509
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
Low HLB surfactants
High HLB surfactants
Combinations of PRR agonists. Different PRR agonists
can synergize and/or balance each other’s immunomod-
ulatory activity (for a comprehensive review see REF. 56).
Synergy was observed with combinations such as TLR2–
TLR4 (REF. 57), and TLR3–TLR4–TLR7/8 (REF. 58), and
with associations between TRIF (Toll–interleukin 1
receptor (TIR)-domain-containing adaptor protein
inducing IFN-β)-coupled TLRs and non-TRIF-coupled
TLRs (see BOX 2), such as TLR4–TLR7/8, TLR4–TLR9 or
TLR3–TLR7/8 (REF. 59). These latter associations strik-
ingly increased the levels of the pro-TH1 cytokines IL-12
and IL-23 in human and mouse DCs, with the mRNA
levels of the rate-limiting components IL-12p35 and
IL-23p19 increased up to 50-fold over that induced by
single agonists59. In agreement, a recent report showed
that synergy was induced by combining agonists that
act on MyD88-dependent and MyD88-independent
pathways (BOX 2), whereas combining agonists that act
on a single pathway induced tolerance60. A vaccine
that stimulates complementary TLR pathways can thus
broaden the TH-cell response that is induced, as shown
with the yellow fever YF17D vaccine61, although viral
replication itself has some role in this case. This latter
study illustrates the fact that live vaccines and inactivated
whole-cell-based vaccines, expressing or containing
multiple TLR agonists, can induce broader and stron-
ger immune responses, and thus do not usually require
Cross-talk exists not only between TLRs but
also between TLRs and NODs62. Synergy has been
observed between NOD1/2 and sub-active doses of
TLR4 agonists63, as well as between NOD2 and TLR9
agonists. This synergy is lost in Crohn’s disease patients
who are homozygous for NOD2 mutations64, and one
hypothesis suggests that the impaired development of
TReg cells might explain the paradoxical consequence
of this lack of synergy in an inflammatory disease65. A
full immune response might thus require engagement
of member(s) of more than one PRR family. However,
combining TLR agonists with NLR or RLH agonists in
the same formulation, although an attractive proposi-
tion, requires that all the consequences of such associa-
tions are documented. Recent developments in clinical
adjuvant research have instead used combinations of
single TLR agonists within carrier/depot formulations,
and this will be discussed in a later section.
Adjuvants and formulations
As stated in the introduction, the carrier activities of
adjuvants can be different from their immunostimulant
and/or immunomodulatory activities, and each formula-
tion is different in this respect. Those developed before
the advent of specific PRR ligands include mineral salts
(aluminium), emulsions (BOX 3) and immunostimulating
complexes (ISCOMs), which have both immunostimu-
lant and carrier activities. By contrast, liposomes and
microparticles are inert carriers, unless they have a spe-
cific composition or carry immunostimulants. FIGURE 3
shows some of the properties of these adjuvants, includ-
ing those of the more specific PRR ligands. The following
paragraphs will address some of the main properties of the
TLR-independent ‘classical’ adjuvants28, which could have
synergistic effects if combined with TLR agonists.
Aluminium. Paradoxically, although aluminium salts are
by far the most widely used adjuvants in humans, their
mechanism of action is less well characterized than that
of TLR agonists. Aluminium salts induce antibodies and
TH2 responses, which was initially thought to require
that the antigen be adsorbed to the adjuvant surface
(the depot effect, that is, the adjuvant forms a depot at
the vaccination site, from which antigen can be released),
transforming a soluble antigen into a particulate one to
favour APC uptake. Most antigens are rapidly desorbed
from aluminum salts following exposure to interstitial
fluid, however66, and adsorption is not required before
vaccine administration for adjuvanticity67. However,
adsorption or entrapment in aggregates might favour a
Box 3 | Emulsions
An emulsion is a mixture of two immiscible substances. One substance (the dispersed
phase) is dispersed in the other (the continuous phase) and stabilized with one or several
surfactant(s) or emulsifier(s), present at the interface between the two phases.
Emulsions used in vaccines can be water-in-oil emulsions (W/O) or oil-in-water
emulsions (O/W) (see figure), depending on the volume fraction of both phases and on
the type of surfactant. Such compounds contain both a hydrophilic polar group and a
hydrophobic (lipophilic) non-polar group, and can be defined by their hydrophilic/
lipophilic balance (HLB): low HLB surfactants with high affinity for oily phases will
drive the formation of W/O emulsions and high HLB surfactants with high affinity for
the aqueous phase will drive the formation of O/W emulsions, although this can also
depend on the nature of the emulsion oil. In the former case, aqueous droplets (shown
in blue in the figure) will be dispersed in the continuous oily phase (shown in yellow),
and the opposite will be true in the latter case. Antigen will usually be contained in
the water phase although some amphiphilic immunostimulants can be at the W/O
interface, such as surfactants.
One example of a W/O emulsion is Freund’s adjuvant, which is a solution emulsified in
mineral oil with a manide monooleate surfactant. Complete Freund’s adjuvant (CFA)
contains inactivated and dried mycobacteria, usually Mycobacterium tuberculosis, and
incomplete Freund’s adjuvant (IFA) does not contain mycobacteria. CFA cannot be used
in humans because of its high reactogenicity, and should be replaced in animals by other
adjuvants when possible. Another type of W/O emulsion, Montanide ISA51 (50:50 O:W
ratio), is a mixture of a mineral oil with a mannide monooleate family surfactant.
Montanide ISA720, which is a mix of vegetable oil with the same type of surfactant, also
gives W/O emulsions (70:30 O:W ratio70).
One prototype O/W emulsion is MF59, which is a microfluidized oil-in-water emulsion,
consisting of squalene oil and Tween 80 and Span 85 as surfactants71. It gives small (<250
nm), uniform, stable droplets. MF59 is a component of an adjuvanted influenza vaccine
that is marketed in Europe (Fluad).
510 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
Depot /carrier Immunostimulation
ISCOMs Saponins PRR /TLR agonists
TH1 or TH1 /TH2 TH2
Immunostimulation and carrier
high local antigen concentration and improved uptake
by APCs. The role of such cells in aluminium adjuvan-
ticity has been highlighted: macrophages are activated
by aluminium to present antigen68 and a previously
unknown population of IL-4-producing cells was shown
to be required for alum-induced in vivo priming and
expansion of antigen-specific B cells69.
Emulsions. After aluminium, emulsions are among the
most frequently used adjuvants in humans and animals
(BOX 3). One can mention two prototype water-in-oil
(W/O) and oil-in-water (O/W) emulsions, Montanide
720 and MF59, respectively70,71. As for aluminium, the
adjuvanticity of emulsions was proposed to be linked in
part to the depot effect, but this does not seem to apply
to O/W emulsions71.
Saponins, QS21 and ISCOMS. Saponins, which are
derived from the bark of the South American tree
Quillaja saponaria Molina, have been tested and used
in both veterinary and human medicine for decades.
Triterpenoid saponins (Q saponins), including QS21,
efficiently drive TH1 cell and antibody responses but
their widespread application has sometimes been ham-
pered by their reactogenicity. Progress has been made in
this respect72, and unfractionated GPI-0100-containing
semi-synthetic saponin derivatives (DS saponins) have
also been developed, with two fractions that stimulate
TH1 and TH2 responses73. Saponins also constitute
active components of particulate formulations such
as ISCOMs, which are cage-like structures containing
antigen, cholesterol, phospholipid and saponin. ISCOM-
based vaccines promote both antibody and cell-mediated
responses, including CTLs74,75, which might depend in
part on tight antigen binding to ISCOMs.
Liposomes and microparticles. Particulate compounds
can either encapsulate antigen or carry it on their surface
through adsorption or covalent linkage76,77. In both cases,
if an antibody response against conformational epitopes is
required, the conformation and accessibility of the antigen
should be maintained. This also holds true for specific
immunostimulants or ligands that need to interact with
their target properly. In addition to antigen, liposomes and
microparticles must also deliver activatory signals. This
can be linked to their composition (for instance through
cationic or anionic charges78,79), or by the addition of an
immunostimulant. The particulate nature of these com-
pounds might allow some targeting to APCs80, although
specific targeting requires specific ligands.
Antigen/formulation targeting. Targeting APCs, for
example through CLRs, can enhance subsequent
immune responses, should this occur in the presence
of a co-stimulatory signal, to avoid tolerance induction,
as seen after DEC-205 targeting of DCs in the steady
state44. By contrast, delivering proteins by antibodies to
this receptor in the presence of a maturation stimulus
such as poly (I:C) and anti-CD40 antibodies combines
targeting, stimulation of signals 0 and 2, overcomes sup-
pression and induces strong immunity81. Using tightly
defined sizes of nano-beads (40 nM) might also favour
targeting of DCs, as it is proposed that these cells prefer-
entially capture particulate antigens of such well-defined
Some methods can specifically target the class I major
histocompatibility complex (MHC) antigen presenta-
tion pathway to stimulate CD8+ CTLs (signal 1). A TH1
response is necessary but not sufficient to induce CD8+
CTLs. Achieving this stimulation with ‘inert’ antigen
remains a challenge, and different options have been
proposed, including coupling to Bordetella pertussis
adenylate cyclase (CyaA83) or to negative charges84,
incorporation into ISCOMs, co-administration with
TLR9 ligands52 and with a cationic peptide such as in the
synthetic adjuvant IC31 (REF. 85), or the use of liposomes
of particular composition79,86.
The optimal formulation? Clinical developments
Adjuvant activity and safety can be enhanced by com-
bining formulations and immunostimulants87. As stated
previously, the absolute requirement for TLR signalling
in the generation of potent antibody responses has been
questioned by using TLR-signalling-deficient mice
(MyD88–/– TrifLps2/Lps2) and the chemically modified
antigens trinitrophenol-haemocyanin (TNP-Hy) and
Figure 3 | Properties of adjuvants. The main type of adjuvants with respect to their depot/carrier and immunostimulatory
properties are shown. Some compounds can possess both characteristics whereas others possess only one. In addition,
some of the adjuvants shown (red background) can have immunomodulatory properties beyond their ability to trigger
global immune stimulation, by directing responses specifically towards a T helper (TH) 1 or TH2 response. A third dimension
(not represented here) is the specific targeting ability of adjuvants, although carrier/depot activity and ligand specificity
can contribute to targeting. ISCOMs, immunostimulating complexes; O/W, oil-in-water emulsion; PRR, pattern-recognition
receptor; TLR, Toll-like receptor; W/O, water-in-oil emulsion.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 511
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
(Microfold cells). Specialized
epithelial cells that deliver
antigens by transepithelial
vesicular transport from the
gut lumen directly to
and to subepithelial lymphoid
trinitrophenol–keyhole limpet haemocyanin (TNP–
KLH)28. It was shown that under these conditions TLR
recognition does not account for all of the adjuvantic-
ity of some compounds, including the Ribi emulsion
containing the TLR4 agonist monophosphoryl lipid A
(MPL) and the immunostimulatory glycolipid trehalose
dicorynomycolate. The adjuvant Pam3Cys-lipidated
OspA protein88, which presumably acts through TLR2,
has also been shown to act independently of TLRs89.
Although TLR-independent mechanisms might have
been anticipated for ‘classical’ adjuvants such as alum
and emulsions, these studies showed that even for some
more complex compounds containing TLR agonists
there are alternative, and possibly complementary, path-
ways to stimulate antibody responses. Nevertheless, the
work of Gavin and collaborators28 confirmed that TLRs
control TH-cell orientation and the antibody isotypes that
are produced, as seen by an overall bias towards a TH2
response in the TLR-signalling-deficient mice: compared
with wild-type C57BL/6 controls, MyD88–/– TrifLps2/Lps2
mice showed a marked increase in TH2-cell-linked IgG1
and IgE responses accompanied with a marked decrease in
TH1-cell-linked IgG3 levels and a less pronounced decrease
in IgG2c and IgG2b. Specific anti-TNP IgG2c and IgG2b
responses were also significantly lower in the deficient
mice after immunization with MPL. Neither the nature
and duration of T- and B-cell memory nor the antibody
functionality was investigated in this study. These find-
ings support the proposal that associating TLR-dependent
and TLR-independent adjuvants could be beneficial, by
triggering different and potentially synergistic pathways:
TLR-independent ‘classical’ adjuvants would increase the
global level of the immune response, and TLR agonists
would modulate its quality (TH1/TH2-cell bias).
However, finding the optimal association between
adjuvants is difficult as antagonism or anergy can occur
rather than synergy60. Co-formulating the antigen and
immunostimulant might also be required, as shown by the
need for TLRs and antigens to be in the same compartment
for optimal antigen presentation, as stated previously23,24.
Moreover, stimulating TLRs without antigen activates
APCs and decreases their cross-presentation ability90.
Combining antigens in the same formulation also
allows otherwise non-compatible antigens to be asso-
ciated in multivalent vaccines91 and can ensure that
all antigens will be presented simultaneously. This can
reduce the risks of immunodominance, should one anti-
gen be presented before the others and take the lead92.
As stated before, both antigen(s) and immunostimulant
should retain their native conformation and ability to
interact with their ligands (TLRs or membrane-bound
antibodies on B cells for instance). In this respect, the
antigen–adjuvant couple can be crucial, as the antigen
itself might increase or counteract adjuvant efficacy
or affect the induced TH-cell bias. For instance, some
antigens trigger TH1 responses even with alum93, and
different antigens can induce TH1 cell or mixed TH1/TH2-
cell responses even though they are present in identical
formulations94. Finally, formulating immunostimulants
in carriers such as liposomes or microparticles targeted
to specific cells and organs might also increase the safety
profile of these molecules, by limiting their distribution
in vivo, thereby minimizing systemic toxicity.
Combined formulations have thus been developed,
some from before TLRs were identified, and have reached
clinical trials and market approval. Relevant examples
of clinical and preclinical developments are presented in
TABLE 1. Many are aimed at introducing a cellular TH1
response in addition to an antibody response to increase
vaccine efficacy against targets such as certain viruses, for
instance, through TLR4 or TLR9 agonists. For example,
a subunit herpes vaccine adjuvanted with AS04 com-
bining alum and the TLR4 agonist MPL showed partial
efficacy in HSV1- and HSV2-seronegative women but a
similar vaccine adjuvanted with an MF59 emulsion was
Can we re-direct immune responses?
Adjuvants that can selectively trigger TH1 or TH2
responses offer the possibility of re-directing the TH-cell
response once it has been established, although this
is more difficult in primed individuals than in naive
non-biased individuals. This approach has been used
successfully in human allergy to revert TH2 responses,
using TH1-cell agonists such as MPL96 or by coupling an
allergen to CpG oligonucleotides97. This gives hope that
such immunotherapy could contribute to the treatment
of already established immune disorders. In the case of
chronic infectious diseases, re-orientating or reactivating
TH responses might also favour more efficient elimination
mechanisms. For instance, TLR agonists can stimulate
T cells directly or indirectly98, particularly targeting TReg
cells to enhance or counteract their action99. Targeting
the inhibitory PD1–PDL1 pathway can also reactivate
exhausted effector cells100, as has been shown for HIV-
specific CD8+ T cells101,102. The induction of stronger
T-cell responses is also the goal of therapeutic vaccination
against hepatitis B103 and the addition of new adjuvants
to the existing vaccines could be helpful in this respect.
Adjuvants whose reactogenicity would be problematic
in prophylaxis might also be more easily applied in
Mucosal immunization. Although most vaccines are
systemic, much attention has been given to the devel-
opment of mucosal vaccines for practical and target-
driven reasons (for a review see REF. 104). Although
mucosal routes might require specific formulations,
some ‘systemic’ immunostimulants and carriers could
also be used. The properties required for a mucosal
formulation are in fact similar to those required for
systemic immunization: the adjuvant must carry anti-
gen and bind and stimulate APCs, bearing in mind that
a TH2/TH3-cell bias is usually associated with mucosal
inductive sites. Some cells at mucosal sites, called
M cells or M-like cells, are specialized in the uptake
of particulate antigens105. Although liposomes and
microparticles can encapsulate and protect antigens
from harsh conditions encountered in some mucosal
environments, they do not really target mucosal
inductive sites. This might be achieved by coating
512 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
formulations with specific M-cell ligands, and dif-
ferent candidates have been identified, including
lectins and some viral proteins (such as the reovirus
σ1 protein106,107). The MPL adjuvant can be included
to immunopotentiate targeted liposomes106. Regarding
cell activation, cholera toxin (CT, from Vibrio cholerae)
is the prototype ‘gold standard’ mucosal adjuvant, along
with the related heat-labile toxin from Escherichia coli
(LT; for a review see REF. 108). The CT toxin almost
perfectly fulfils the criteria mentioned earlier: the CT B
subunit first binds to the widely expressed GM1 ganglio-
side, and the A subunit subsequently activates mucosal
epithelial cells by catalysing ADP ribosylation of the
adenylate cyclase regulatory protein Gsα, increasing
in turn the synthesis of the 3′-, 5′-cyclic AMP (cAMP)
second messenger. The main problem linked to CT and
LT is their high toxicity in humans, although this is
not the case in mice. Different approaches have been
used to solve this problem, including the construction
of non-toxic mutants retaining adjuvanticity108.
Table 1 | Examples of preclinical and clinical evaluations of combined immunostimulants and formulations*
TLR4 agonists: MPL, RC529
and other synthetic agonists
TLR 7/8 agonists:
and other ligands)
TLR9 agonists: synthetic
unmethylated CpG oligonucleotides
Alum plus MPL (AS04) is a
component of Fendrix to prevent
HepB in high-risk patients125
and Cervarix to prevent cervical
cancer with HPV16/18 L1
virus-like particles126. RC529 in
alum-adjuvanted HepB vaccine
increased seroprotection rates
(up to 99%) and anti-HepB titre in
CpG 7909 mixed with HepB vaccine
induce faster and higher specific
antibody responses; trend towards
higher CTL responses128. In mice, HCV
polyprotein adsorbed to alum plus CpG
induced a 20-fold increase in antibodies
compared with alum alone129.
51 and 720
HIV Gag protein in ISA
51 with 3M synthetic
agonists had higher
frequencies of TH1
with emulsion alone130.
TH1-linked protection in mice with MSP1
malaria antigen administered with CpG
oligos in ISA51 (REF. 131). Monkeys
receiving HIV Gag protein in ISA51
with CpG had higher frequencies of
TH1 responses compared with emulsion
In mice HCV polyprotein with MF59
plus CpG induced a 20-fold increase in
antibodies compared with protein in
MF59 without CpG129.
Efficacy of MPL plus squalene
emulsion plus QS21 (AS02)
demonstrated in humans with
malaria vaccine candidates132,133
and other Ags such as HBsAg, HIV,
HPV and melanoma Ags.
Efficacy of MPL plus
plus QS21 (AS02)
in humans with
and other Ags such
as HBsAg, HIV, HPV
and melanoma Ags.
(DTH) of MPL
plus QS21 (AS01)
in monkeys with
Immunogenicity (DTH) of MPL
plus QS21 (AS01) in liposomes
demonstrated in monkeys with
malaria vaccine candidates134.
in liposomes increased
humoral and cellular
responses in mice
compared with 3M-019
or liposomes alone55.
CpG in liposomes with influenza or
HBsAg up to 30 times more effective
in mice than formulations containing
unencapsulated oligos. TH1-dominant
response with influenza and mixed TH1/
TH2 with HBsAg94. Cationic liposomes
with TLR3 (poly (I:C)) or TLR9 (CpG)
agonists potentiate responses in mice,
including CD8+ T-cell responses in a
therapeutic tumour vaccine model
(B16 melanoma tumours) and in an
Mycobacterium tuberculosis aerosol
challenge model (ESAT-6 antigen135).
In mice, CpG adsorbed on cationic
microparticles induces stronger and
faster immune responses against anthrax
than CpG alone137.
TLR4 agonists (MPL/RC529)
encapsulated in anionic PLG
microparticles enhance Ab
responses and bactericidal
activities (MenB) in mice136.
*The far-left column mainly lists formulations with carrier/depot and/or TH2-biased immunostimulatory effects. The top row corresponds to compounds having
mostly a TH1-biased immunostimulatory effect. The intersections correspond to combinations of these two types of compounds, evaluated in preclinical models
and in clinics. Further details and synthesis of clinical data obtained with other individual adjuvants/formulations can be found elsewhere124. Ab, antibody; Ag,
antigen; CTL, cytotoxic T lymphocyte; DTH, delayed-type hypersensitivity; HBsAg, hepatitis B soluble antigen; HCV, hepatitis C virus; HepB, hepatitis B; HPV, human
papilloma virus; MenB, Neisseria meningitidis serotype B; MPL, monophosphoryl lipid A; MSP1, merozoite surface protein 1; Oligos, oligonucleotides; O/W, oil in
water; PLG, polylactide-co-glycolide; TH, T helper cell; TLR, Toll-like receptor; W/O, water in oil.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 513
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
In vitro induction
of T-cell responses
In vitro evaluation
in cell culture
In vitro/in vivo
in animal models
of formulated antigen
cells that contain characteristic
known as Birbeck granules,
and which express the CD1a
antigen. Principally found in
the stratified squamous
Epidermal and intradermal routes. Epidermal and intra-
dermal vaccination routes are being explored as they could
have an adjuvant effect, thanks to the high density of DCs
present in the epidermis (Langerhans cells) and dermis.
These aspects deserve a review on their own, but one can
mention the topical application of antigen onto the skin
through a non-abrasive patch, as this usually requires the
use of adjuvants such as CT or LT (for a review see REF.
109). Langerhans cells express a selective array of TLRs19,
which should guide in part the choice of adjuvants. CT and
LT derivatives can also be used as adjuvants by the classi-
cal subcutaneous route, and have induced protection in an
animal model of Helicobacter pylori infection110.
Practical aspects of adjuvant development
Besides efficacy and immunogenicity, there are other
critical factors that affect the ability of any adjuvant or for-
mulation to go beyond preclinical studies. Constituents
should be produced according to Good Manufacturing
Practices, and should be chemically and bio logically
defined and stable, thus assuring safe and consistent
responses in vaccinees. The adjuvant combined with
antigen(s) should induce stronger protective immunity
than the same antigen(s) alone. Adjuvant safety must be
documented, including immediate and long-term side
effects. The adjuvant should be intrinsically non-immu-
nogenic, biodegradable and biocompatible. Although
there are some exceptions, adjuvants are generally
not developed as drugs on their own, and toxicology
studies must be designed on a case-by-case basis, as
the adjuvant–antigen couple can also affect safety and
reacto genicity. Some specific guidelines already exist
and others are under evaluation111.
Safety versus efficacy — risk:benefit ratio.The ben-
efits of using an adjuvant must outweigh the risks.
The development of most adjuvants is stopped or
hampered because of acute reactogenicity or toxicity
issues. However, the induction of long-term immuno-
pathology such as autoimmunity is more difficult to
document, and this has been widely discussed for TLR
agonists (for review see REF. 112). TLR agonists might by
themselves limit the risk of uncontrolled reactions
by triggering both activatory and inhibitory signals. To
return to the analogy used previously, TLR ligands can
define the game’s strategy and at the same time attract
the referee’s attention: they modulate inflammatory
responses by triggering the expansion of TReg cells and,
through a feedback mechanism, can limit potential
autoimmunity resulting from over-activated cells once
the infection has been controlled. In addition, one has to
consider the theoretical risks linked to a few immuniza-
tions and short-term stimulation with TLR agonists, in
comparison to life-long stimulation through the same
receptors by environmental pathogens or commensals.
The good safety record of adjuvants such as the TLR4
agonist MPL — up to 120,000 doses of which have been
administered to 30,000 volunteers in the past 15 years
— suggests that the use of such agonists in vaccination
is of low risk.
The situation might be different when specific
T-cell co-stimulatory pathways are triggered directly.
The FDA recently approved CTLA4-Ig, an antagonist
of CD28 co-stimulation, for the treatment of rheu-
matoid arthritis113, which opens the way for similar
compounds. However, it is certainly safer to inhibit
co-stimulatory pathways than to directly activate them,
and the recent outcome of a trial using a super CD28-
agonist (the TGN1412 antibody114) suggests the need
for prudence when using such strong signal 2 agonists.
This antibody, which triggered T-cell activation with-
out the need for antigen–TCR stimulation (signal 1),
satisfied preclinical evaluations in different species.
Its goal was to stimulate TReg cells, which was achieved
in reconstituted immunodeficient mice with a related
antibody (5.11A1)115. Moreover, TG1412 had a posi-
tive safety record in Cynomolgus monkeys, in which
the sequences of the extracellular external and intracel-
lular domains of CD28 are identical to human CD28
Which preclinical model? The predictability of animal
models is therefore questionable in some circum-
stances. Mice are the most widely used preclinical
model, but although they often generate helpful infor-
mation, there are some significant differences between
the immune systems of mice and humans (for a review
see REF. 116). For example, TLR9 is expressed in dif-
ferent DC subsets in mice and humans and optimal
CpG sequences are also species-specific. TLR4 is
expressed at significant levels in mouse B cells but not
in human B cells under normal conditions, and human
and mouse TLR4 respond differently to some LPS117.
TLR8 responsiveness is also different in mice, which
seem to respond poorly to TLR8 agonists, although a
Figure 4 | In vivo and in vitro preclinical evaluation of adjuvants. In addition to
preclinical in vivo evaluation in animal models, it is of interest to investigate adjuvants
and formulations in parallel in vitro in human cells, in particular in antigen-presenting
cells such as dendritic cells. PK/PD, pharmacokinetic/pharmacodynamic.
514 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
recent report showed that the murine receptor might
be functional118. TLR expression can even vary between
different strains of mice119.
The pre-immune status of the individuals to be vac-
cinated is another important point to consider. Most
adjuvant evaluation is conducted in naive animals, in
which a dramatic benefit is often observed. However,
the differences between adjuvanted and unadjuvanted
vaccines can be decreased or even absent in primed
animals120. This could be consistent with the fact that
TLR stimulation is required to generate memory
CD4+ T cells, but is not required for their activation34.
One thus needs to consider whether the target human
or animal population is pre-immune for the vaccine
antigen(s), and primed preclinical models could be
used to assess this.
In vitro models. Investigating adjuvant activity in human
cells in vitro can add valuable information to animal
models. As stated previously, transformed cells expressing
specific arrays of TLRs can be used to evaluate new ago-
nists. Many publications have used primary human cells to
evaluate adjuvants and vaccines, in particular monocyte-
derived DCs121. Flow-cytometry allows cells that have been
activated after capturing a formulation to be distinguished
from the surrounding cells. Analysing how focused the
stimulation is can be an indirect means of evaluating safety,
in addition to immunogenicity. This has been used with live
vaccines122, and can also be applied to inert vaccines if the
formulation constituents can be directly labelled or tracked
by fluorescent antibodies. Confocal microscopy also allows
the subcellular localization of antigen and/or formula-
tions to be tracked123, and can identify which pathways
are followed on antigen/formulation capture, for instance,
the class I or class II MHC antigen presentation path-
ways86. DNA microarrays can also identify a ‘stimulation’
signature that can be distinguished from a ‘toxicity’ sig-
nature. FIGURE 4 summarizes which in vitro assays can be
used for adjuvant evaluation.
Conclusions and perspectives
Observation in Nature often leads to practical appli-
cations. In the early 1900s in Switzerland, Georges de
Mestral observed the burdock burrs that stuck to his
clothes and his dog’s coat. He noticed that they did so
because of hundreds of tiny hooks, and then created
Velcro products by making synthetic hooks and loops.
Observation of the different ways pathogens ‘grab’ our
innate immune system has led adjuvant researchers to
create synthetic ligands that can selectively bind some
PRRs and trigger their associated downstream signalling
pathways. Moreover, one can now design and associate
such ‘burr hooks’ of different size, shape and specific-
ity, allowing the optimal formulation to be defined on a
During the past two decades adjuvant development
has shifted from empiricism to more focused research.
Among the many immunostimulants and formula-
tions available, researchers can now rationally identify,
characterize and combine those that will give vaccines
the necessary help. The choice of adjuvant is guided by
the nature of the antigen(s) and the known or expected
correlates of protection, keeping in mind that the
induction of an immune response is not a black or
white process. The development of these sometimes
complex products follows a long and often uncertain
road, but we now have more basic and applied tools
to analyse the potential issues and propose new solu-
tions. There is however still room and need for more
research to identify the combination of ligands that
will give the safest and most efficient response with
respect to the considered pathogen.
Ulmer, J. B., Valley, U. & Rappuoli, R. Vaccine
manufacturing: challenges and solutions. Nature
Biotechnol. 24, 1377–1383 (2006).
Hunter, R. L. Overview of vaccine adjuvants: present
and future. Vaccine 20, S7–S12 (2002).
Schijns, V. & O’Hagan, D. (Eds) Immunopotentiators in
Modern Vaccines. (Academic Press, Elsevier, 2006).
Bianchi, M. E. DAMPs, PAMPs and alarmins: all we
need to know about danger. J. Leukoc. Biol. 81, 1–5
Steinman, R. M. & Hemmi, H. Dendritic cells:
translating innate to adaptive immunity. Curr. Top.
Microbiol. Immunol. 311, 17–58 (2006).
Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin,
M. A. & Coffman, R. L. Two types of murine helper
T cell clone. I. Definition according to profiles of
lymphokine activities and secreted proteins.
J. Immunol. 136, 2348–2357 (1986).
Mosmann, T. R. & Sad, S. The expanding universe of
T-cell subsets: Th1, Th2 and more. Immunol. Today
17, 138–146 (1996).
Harrington, L. E., Mangan, P. R. & Weaver, C. T.
Expanding the effector CD4 T-cell repertoire: the TH17
lineage. Curr. Opin. Immunol. 18, 349–356 (2006).
Belkaid, Y. & Rouse, B. T. Natural regulatory T cells in
infectious disease. Nature Immunol. 6, 353–360
10. Lanzavecchia, A. & Sallusto, F. Understanding the
generation and function of memory T cell subsets.
Curr. Opin. Immunol. 17, 326–332 (2005).
11. Pulendran, B. & Ahmed, R. Translating innate
immunity into immunological memory: implications for
vaccine development. Cell 124, 849–863 (2006).
12. Schijns, V. E. Induction and direction of immune
responses by vaccine adjuvants. Crit. Rev. Immunol.
21, 75–85 (2001).
13. Barr, T. A., Carlring, J. & Heath, A. W. Co-stimulatory
agonists as immunological adjuvants. Vaccine 24,
14. Medzhitov, R., Preston-Hurlburt, P. & Janeway, C. A. Jr.
A human homologue of the Drosophila Toll protein
signals activation of adaptive immunity. Nature 388,
15. Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J. M.
& Hoffmann, J. A. The dorsoventral regulatory gene
cassette spatzle/Toll/cactus controls the potent
antifungal response in Drosophila adults. Cell 86,
16. Poltorak, A. et al. Defective LPS signaling in C3H/HeJ
and C57BL/10ScCr mice: mutations in Tlr4 gene.
Science 282, 2085–2088 (1998).
References 14, 15 and 16 identified and
demonstrated the critical role of Toll and TLRs in
immune defence, which opened new perspectives in
basic and applied immunology.
17. Iwasaki, A. & Medzhitov, R. Toll-like receptor control of
the adaptive immune responses. Nature Immunol. 5,
18. Akira, S. TLR signaling. Curr. Top. Microbiol. Immunol.
311, 1–16 (2006).
19. Flacher, V. et al. Human langerhans cells express a
specific TLR profile and differentially respond to
viruses and gram-positive bacteria. J. Immunol. 177,
20. Jarrossay, D., Napolitani, G., Colonna, M., Sallusto, F.
& Lanzavecchia, A. Specialization and complementarity
in microbial molecule recognition by human myeloid
and plasmacytoid dendritic cells. Eur. J. Immunol. 31,
21. Agrawal, S. et al. Cutting edge: different Toll-like
receptor agonists instruct dendritic cells to induce
distinct TH responses via differential modulation of
extracellular signal-regulated kinase-mitogen-
activated protein kinase and c-Fos. J. Immunol. 171,
22. Dillon, S. et al. A Toll-like receptor 2 ligand stimulates
TH2 responses in vivo, via induction of extracellular
signal-regulated kinase mitogen-activated protein
kinase and c-Fos in dendritic cells. J. Immunol. 172,
23. Blander, J. M. & Medzhitov, R. Toll-dependent
selection of microbial antigens for presentation by
dendritic cells. Nature 440, 808–812 (2006).
24. Yarovinsky, F., Kanzler, H., Hieny, S., Coffman, R. L. &
Sher, A. Toll-like receptor recognition regulates
immunodominance in an antimicrobial CD4+ T cell
response. Immunity 25, 655–664 (2006).
References 23 and 24 show that in addition to
their role in cell activation TLRs also have a direct
role in foreign antigen presentation, in vitro
(Ref. 23) and in vivo (Ref. 24). Activated TLRs and
antigen should preferentially be in the same cargo
for optimal presentation to T cells.
25. Ruprecht, C. R. & Lanzavecchia, A. Toll-like receptor
stimulation as a third signal required for activation of
human naive B cells. Eur. J. Immunol. 36, 810–816
26. Pasare, C. & Medzhitov, R. Control of B-cell responses
by Toll-like receptors. Nature 438, 364–368 (2005).
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 515
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
27. Nemazee, D., Gavin, A., Hoebe, K. & Beutler, B. Toll-
like receptors and antibody responses. Nature 441,
28. Gavin, A. L. et al. Adjuvant-enhanced antibody
responses in the absence of Toll-like receptor
signaling. Science 314, 1936–1938 (2006).
29. Sakaguchi, S. Control of immune responses by
naturally arising CD4+ regulatory T cells that express
toll-like receptors. J. Exp. Med. 197, 397–401
30. Sutmuller, R. P. et al. Toll-like receptor 2 controls
expansion and function of regulatory T cells. J. Clin.
Invest. 116, 485–494 (2006).
31. Caramalho, I. et al. Regulatory T cells selectively
express toll-like receptors and are activated by
lipopolysaccharide. J. Exp. Med. 197, 403–411
32. Peng, G. et al. Toll-like receptor 8-mediated reversal of
CD4+ regulatory T cell function. Science 309,
33. Pasare, C. & Medzhitov, R. Toll pathway-dependent
blockade of CD4+CD25+ T cell-mediated suppression
by dendritic cells. Science 299, 1033–1036 (2003).
34. Pasare, C. & Medzhitov, R. Toll-dependent control
mechanisms of CD4 T cell activation. Immunity 21,
35. Meylan, E., Tschopp, J. & Karin, M. Intracellular
pattern recognition receptors in the host response.
Nature 442, 39–44 (2006).
36. Fritz, J. H., Ferrero, R. L., Philpott, D. J. & Girardin,
S. E. Nod-like proteins in immunity, inflammation and
disease. Nature Immunol. 7, 1250–1257 (2006).
37. Traub, S., von Aulock, S., Hartung, T. & Hermann, C.
MDP and other muropeptides — direct and synergistic
effects on the immune system. J. Endotoxin Res. 12,
38. Lemesre, J. L. et al. Protection against experimental
visceral leishmaniasis infection in dogs immunized
with purified excreted secreted antigens of Leishmania
infantum promastigotes. Vaccine 23, 2825–2840
39. Kato. H. et al. Differential roles of MDA5 and RIG-I
helicases in the recognition of RNA viruses. Nature
441, 101–105 (2006).
40. Hornung, V. et al. 5′-Triphosphate RNA is the ligand
for RIG-I. Science 314, 994–997 (2006).
41. Pichlmair, A. et al. RIG-I-mediated antiviral responses
to single-stranded RNA bearing 5′-phosphates.
Science 314, 997–1001 (2006).
42. Peiser, L., Mukhopadhyay, S. & Gordon, S. Scavenger
receptors in innate immunity. Curr. Opin. Immunol.
14, 123–128 (2002).
43. Robinson, M. et al. Myeloid C-type lectins in innate
immunity. Nature Immunol. 7, 1258–1265 (2006).
44. Bonifaz, L. et al. Efficient targeting of protein antigen
to the dendritic cell receptor DEC-205 in the steady
state leads to antigen presentation on major
histocompatibility complex class I products and
peripheral CD8+ T cell tolerance. J. Exp. Med. 196,
45. ‘t Hart, B. A. & van Kooyk, Y. Yin-Yang regulation of
autoimmunity by DCs. Trends Immunol. 25, 353–359
46. Klesney-Tait, J., Turnbull, I. R. & Colonna, M. The
TREM receptor family and signal integration. Nature
Immunol. 7, 1266–1273 (2006).
47. Fujii, S. et al. Glycolipid α-C-galactosylceramide is a
distinct inducer of dendritic cell function during innate
and adaptive immune responses of mice. Proc. Natl
Acad. Sci. USA 103, 11252–11257 (2006).
48. Moser, B. & Brandes, M. γδ T cells: an alternative type
of professional APC. Trends Immunol. 27, 112–118
49. Dougan, G. & Hormaeche, C. How bacteria and their
products provide clues to vaccine and adjuvant
development. Vaccine 24, S13–S19 (2006).
50. Kaisho, T. & Akira, S. Toll-like receptors as adjuvant
receptors. Biochim. Biophys. Acta 1589, 1–13 (2002).
51. Kandimalla, E. R., Yu, D. & Agrawal, S. Towards
optimal design of second-generation
immunomodulatory oligonucleotides. Curr. Opin. Mol.
Ther. 4, 122–129 (2002).
52. McCluskie, M. J& Krieg, A. M. Enhancement of
infectious disease vaccines through TLR9-dependent
recognition of CpG DNA. Curr. Top. Microbiol.
Immunol. 311, 155–178 (2006).
53. Hawkins, L. D. et al. A novel class of endotoxin
receptor agonists with simplified structure, toll-like
receptor 4-dependent immunostimulatory action, and
adjuvant activity. J. Pharmacol. Exp. Ther. 300,
This reference shows how new synthetic TLR4
agonists can be designed, characterized and
evaluated in vitro and in vivo using tools such as
transformed cell lines expressing TLR4 and MD2,
and NF-κ κB reporter genes.
54. Persing, D. H. et al. Taking toll: lipid A mimetics as
adjuvants and immunomodulators. Trends Microbiol.
10, S32–S37 (2002).
55. Johnston, D., Zaidi, B. & Bystryn, J. C. TLR7
imidazoquinoline ligand 3M-019 is a potent adjuvant
for pure protein prototype vaccines. Cancer Immunol.
Immunother. 01 Dec 2006 (doi:10.1007/500262-
56. Trinchieri, G. & Sher, A. Cooperation of Toll-like
receptor signals in innate immune defence. Nature
Rev. Immunol. 7, 179–190 (2007).
This reference presents a comprehensive review on
the positive and negative interactions existing
between TLRs and other PRRs.
57. Sato, S. et al. Synergy and cross-tolerance between
Toll-like receptor (TLR) 2- and TLR4-mediated
signaling pathways. J. Immunol. 165, 7096–7101
58. Roelofs, M. F. et al. The expression of Toll-like
receptors 3 and 7 in rheumatoid arthritis synovium is
increased and costimulation of Toll-like receptors 3, 4,
and 7/8 results in synergistic cytokine production by
dendritic cells. Arthritis Rheum. 52, 2313–2322
59. Napolitani, G., Rinaldi, A., Bertoni, F., Sallusto, F. &
Lanzavecchia, A. Selected Toll-like receptor agonist
combinations synergistically trigger a T helper
type 1-polarizing program in dendritic cells. Nature
Immunol. 6, 769–776 (2005).
60. Bagchi, A. et al. MyD88-dependent and MyD88-
independent pathways in synergy, priming, and
tolerance between TLR agonists. J. Immunol. 178,
References 59 and 60 show that stimulation of
TLRs coupled to complementary pathways is
synergistic, and that some combinations can be
preferred to others, depending on which type of
immune response is to be induced.
61. Querec, T. et al. Yellow fever vaccine YF-17D activates
multiple dendritic cell subsets via TLR2, 7, 8, and 9 to
stimulate polyvalent immunity. J. Exp. Med. 203,
62. Takada, H. & Uehara, A. Enhancement of TLR-mediated
innate immune responses by peptidoglycans through
NOD signaling. Curr. Pharm. Des. 12, 4163–4172
63. Fritz, J. H. et al. Synergistic stimulation of human
monocytes and dendritic cells by Toll-like receptor 4
and NOD1- and NOD2-activating agonists. Eur. J.
Immunol. 35, 2459–2470 (2005).
64. van Heel, D. A. et al. Synergy between TLR9 and
NOD2 innate immune responses is lost in genetic
Crohn’s disease. Gut 54, 1553–1557 (2005).
65. Watanabe, T., Kitani, A. & Strober, W. NOD2
regulation of Toll-like receptor responses and the
pathogenesis of Crohn’s disease. Gut 54, 1515–1518
66. HogenEsch, H. Mechanisms of stimulation of the
immune response by aluminum adjuvants. Vaccine 20
(Suppl. 3), 34–39 (2002).
67. Romero Méndez, I. Z., Shi, Y., HogenEsch, H. & Hem,
S. L. Potentiation of the immune response to non-
adsorbed antigens by aluminum-containing adjuvants.
Vaccine 25, 825–833 (2007).
68. Rimaniol, A. C. et al. Aluminium hydroxide adjuvant
induces macrophage differentiation towards a
specialized antigen-presenting type. Vaccine 22,
69. Jordan, M. B., Mills, D. M., Kappler, J., Marrack, P. &
Cambier, J. C. Promotion of B cell immune responses
via an alum-induced myeloid cell population. Science
304, 1808–1810 (2004).
70. Aucouturier, J., Dupuis, L., Deville, S., Ascarateil, S. &
Ganne, V. Montanide ISA 720 and 51: a new
generation of water in oil emulsions as adjuvants for
human vaccines. Expert Rev. Vaccines. 1, 111–118
71. Podda, A. & Del Giudice, G. MF59-adjuvanted
vaccines: increased immunogenicity with an optimal
safety profile. Expert Rev. Vaccines 2, 197–203
References 70 and 71 review the different
steps followed to develop two prototype W/O
and O/W emulsions from the bench to human
use, taking safety and immunogenicity into
72. Kim, Y. J. et al. Synthetic studies of complex
immunostimulants from Quillaja saponaria: synthesis
of the potent clinical immunoadjuvant QS-21Aapi.
J. Am. Chem. Soc. 128, 11906–11915 (2006).
73. Marciani, D. J. et al. Fractionation, structural studies,
and immunological characterization of the semi-
synthetic Quillaja saponins derivative GPI-0100.
Vaccine 21, 3961–3971 (2003).
This reference illustrates how ‘classical’ adjuvants
such as saponins can be further optimized and
characterized with respect to safety and
immunogenicity, using semi-synthesis and
74. Sanders, M. T., Brown, L. E., Deliyannis, G. & Pearse,
M. J. ISCOM-based vaccines: the second decade.
Immunol. Cell Biol. 83, 119–128 (2005).
75. Ennis, F. A. et al. Augmentation of human influenza A
virus-specific cytotoxic T lymphocyte memory by
influenza vaccine and adjuvanted carriers (ISCOMS).
Virology 259, 256–261 (1999).
76. Leserman, L. Liposomes as protein carriers in
immunology. J. Liposome Res. 14, 175–189 (2004).
77. O’Hagan, D. T. & Singh, M. Microparticles as vaccine
adjuvants and delivery systems. Expert Rev. Vaccines
2, 269–283 (2003).
78. Singh, M. et al. Polylactide-co-glycolide
microparticles with surface adsorbed antigens as
vaccine delivery systems. Curr. Drug Deliv. 3,
79. Guy, B. et al. Design, characterization and preclinical
efficacy of a cationic lipid adjuvant for influenza split
vaccine. Vaccine 19, 1794–1805 (2001).
80. Reddy, S. T., Swartz, M. A. & Hubbell, J. A. Targeting
dendritic cells with biomaterials: developing the next
generation of vaccines. Trends Immunol. 27, 573–579
81. Boscardin, S. B. et al. Antigen targeting to dendritic
cells elicits long-lived T cell help for antibody
responses. J. Exp. Med. 203, 599–606 (2006).
This reference shows that one needs to co-deliver a
maturation stimulus when targeting receptors such
as CLRs on APCs in contrast to previous studies
showing that targeting these receptors without
stimulus can induce tolerance.
82. Fifis, T. et al. Size-dependent immunogenicity:
therapeutic and protective properties of nano-vaccines
against tumors. J. Immunol. 173, 3148–3154
83. Moron, G., Dadaglio, G. & Leclerc, C. New tools for
antigen delivery to the MHC class I pathway. Trends
Immunol. 25, 92–97 (2004).
84. Yamasaki, Y., Ikenaga, T., Otsuki, T., Nishikawa, M. &
Takakura, Y. Induction of antigen-specific cytotoxic T
lymphocytes by immunization with negatively charged
soluble antigen through scavenger receptor-mediated
delivery. Vaccine 25, 85–91 (2007).
85. Schellack, C. et al. IC31, a novel adjuvant signaling via
TLR9, induces potent cellular and humoral immune
responses. Vaccine 24, 5461–5472 (2006).
86. Taneichi, M. et al. Antigen chemically coupled to the
surface of liposomes are cross-presented to CD8+
T cells and induce potent antitumor immunity.
J. Immunol. 177, 2324–2330 (2006).
87. Moingeon, P., Haensler, J. & Lindberg, A. Towards the
rational design of Th1 adjuvants. Vaccine 19,
88. Erdile, L. F. & Guy, B. OspA lipoprotein of Borrelia
burgdorferi is a mucosal immunogen and adjuvant.
Vaccine 15, 988–996 (1997).
89. Yoder, A. et al. Tripalmitoyl-S-glyceryl-cysteine-
dependent OspA vaccination of toll-like receptor
2-deficient mice results in effective protection from
Borrelia burgdorferi challenge. Infect. Immun. 71,
90. Wilson, N. S. et al. Systemic activation of dendritic
cells by Toll-like receptor ligands or malaria infection
impairs cross-presentation and antiviral immunity.
Nature Immunol. 7, 165–172 (2006).
91. Sanchez, V. et al. Formulation of single or multiple
H. pylori antigens with DC Chol adjuvant induce
protection by systemic route in mice. Optimal
prophylactic combinations are different from
therapeutic ones. FEMS Immunol. Med. Microbiol.
30, 157–165 (2001).
92. Yewdell, J. W. Confronting complexity: real-world
immunodominance in antiviral CD8+ T cell responses.
Immunity 25, 533–543 (2006).
93. Pattnaik, P. et al. Immunogenicity of a recombinant
malaria vaccine based on receptor binding domain of
Plasmodium falciparum EBA-175. Vaccine 25,
516 | JULY 2007 | VOLUME 5
© 2007 Nature Publishing Group
94. Joseph, A. et al. Liposomal immunostimulatory DNA
sequence (ISS-ODN): an efficient parenteral and
mucosal adjuvant for influenza and hepatitis B
vaccines. Vaccine 20, 3342–3354 (2002).
95. Aurelian, L. Herpes simplex virus type 2 vaccines: new
ground for optimism? Clin. Diagn. Lab. Immunol. 11,
96. Drachenberg, K. J., Wheeler, A. W., Stuebner, P. &
Horak, F. A well-tolerated grass pollen-specific allergy
vaccine containing a novel adjuvant, monophosphoryl
lipid A, reduces allergic symptoms after only four
preseasonal injections. Allergy 56, 498–505 (2001).
97. Creticos, P. S. et al. Immunotherapy with a ragweed-
Toll-like receptor 9 agonist vaccine for allergic rhinitis.
N. Engl. J. Med. 355, 1445–1455 (2006).
98. Krieg, A. M. Therapeutic potential of Toll-like receptor
9 activation. Nature Rev. Drug Discov. 5, 471–484
99. Kretschmer, K. et al. Inducing and expanding
regulatory T cell populations by foreign antigen.
Nature Immunol. 6, 1219–1227 (2005).
100. Barber, D. L. et al. Restoring function in exhausted
CD8 T cells during chronic viral infection. Nature 439,
101. Trautmann, L. et al. Upregulation of PD-1 expression
on HIV-specific CD8+ T cells leads to reversible immune
dysfunction. Nature Med. 12, 1198–1202 (2006).
102. Freeman, G. J., Wherry, E. J., Ahmed, R. & Sharpe,
A. H. Reinvigorating exhausted HIV-specific T cells via
PD-1-PD-1 ligand blockade. J. Exp. Med. 203,
103. Pol, S. & Michel, M. L. Therapeutic vaccination in
chronic hepatitis B virus carriers. Expert Rev. Vaccines
5, 707–716 (2006).
104. Freytag, L. C. & Clements, J. D. Mucosal adjuvants.
Vaccine 23, 1804–1813 (2005).
105. Jones, B., Pascopella, L. & Falkow, S. Entry of microbes
into the host: using M cells to break the mucosal
barrier. Curr. Opin. Immunol. 7, 474–483 (1995).
106. Clark, M. A., Blair, H. Liang, L. et al. Targeting
polymerised liposome vaccine carriers to intestinal M
cells. Vaccine 20, 208–217 (2001).
107. Lavelle, E. C. Targeted delivery of drugs to the
gastrointestinal tract. Crit. Rev. Ther. Drug Carrier
Syst. 18, 341–386 (2001).
108. Pizza, M. et al. Mucosal vaccines: non toxic derivatives
of LT and CT as mucosal adjuvants. Vaccine 19,
109. Glenn, G. M. et al. Transcutaneous immunization and
immunostimulant strategies: capitalizing on the
immunocompetence of the skin. Expert Rev. Vaccines
2, 253–267 (2003).
110. Weltzin, R., Guy, B., Thomas, W. D. Jr, Giannasca, P. J. &
Monath, T. P. Parenteral adjuvant activities of
Escherichia coli heat-labile toxin and its B subunit for
immunization of mice against gastric Helicobacter pylori
infection. Infect. Immun. 68, 2775–2782 (2000).
111. Sesardic, D. Regulatory considerations on new
adjuvants and delivery systems. Vaccine 24, S86–S87
112. Marshak-Rothstein, A. Toll-like receptors in systemic
autoimmune disease. Nature Rev. Immunol. 6,
113. Bluestone, J. A., St Clair, E. W. & Turka, L. A. CTLA4Ig:
bridging the basic immunology with clinical
application. Immunity 24, 233–238 (2006).
114. Suntharalingam, G. et al. Cytokine storm in a phase 1
trial of the anti-CD28 monoclonal antibody TGN1412.
N. Engl. J. Med. 355, 1018–1028 (2006).
115. Hanke, T. Lessons from TGN1412. Lancet 368,
116. Mestas, J. & Hughes, C. C. Of mice and not men:
differences between mouse and human immunology.
J. Immunol. 172, 2731–2738 (2004).
This reference presents a useful synthesis of the
similarities and differences between mouse and
human immunology, which should be considered
when evaluating vaccines and adjuvants in this
117. Nahori, M. A. et al. Differential TLR recognition of
leptospiral lipid A and lipopolysaccharide in murine
and human cells. J. Immunol. 175, 6022–6031
118. Gorden, K. K., Qiu, X. X., Binsfeld, C. C.,
Vasilakos, J. P. & Alkan, S. S. Cutting edge:
activation of murine TLR8 by a combination of
imidazoquinoline immune response modifiers and
polyT oligodeoxynucleotides. J. Immunol. 177,
119. Liu, T., Matsuguchi, T., Tsuboi, N., Yajima, T. &
Yoshikai, Y. Differences in expression of toll-like
receptors and their reactivities in dendritic cells in
BALB/c and C57BL/6 mice. Infect. Immun. 70,
120. Potter, C. W. & Jennings, R. Effect of priming on
subsequent response to inactivated influenza vaccine.
Vaccine 21, 940–945 (2003).
121. Spisek, R. et al. Maturation of dendritic cells by
bacterial immunomodulators. Vaccine 22,
122. Sanchez, V., Hessler, C., DeMonfort, A., Lang, J. &
Guy, B. Comparison by flow cytometry of immune
changes induced in human monocyte-derived
dendritic cells upon infection with dengue 2 live-
attenuated vaccine or 16681 parental strain.
FEMS Immunol. Med. Microbiol. 46, 113–123
123. Dunzendorfer, S., Lee, H. K., Soldau, K. & Tobias, P. S.
TLR4 is the signalling but not the lipopolysaccharide
uptake receptor. J. Immunol. 173, 1166–1170
124. Burdin, N., Guy, B. & Moingeon, P. Immunological
foundations to the quest for new vaccine adjuvants.
BioDrugs 18, 79–93 (2004).
125. Jacques, P. et al. The immunogenicity and
reactogenicity profile of a candidate hepatitis B
vaccine in an adult vaccine non-responder population.
Vaccine 20, 3644–3649 (2002).
126. Harper, D. M. et al. Sustained efficacy up to 4.5 years
of a bivalent L1 virus-like particle vaccine against
human papillomavirus types 16 and 18: follow-up
from a randomised control trial. Lancet 367,
127. Dupont, J. et al. A controlled clinical trial comparing
the safety and immunogenicity of a new adjuvanted
hepatitis B vaccine with a standard hepatitis B vaccine.
Vaccine 24, 7167–7174 (2006).
128. Cooper, C. L. et al. CPG 7909, an immunostimulatory
TLR9 agonist oligodeoxynucleotide, as adjuvant to
Engerix-B HBV vaccine in healthy adults: a double-
blind phase I/II study. J. Clin. Immunol. 24, 693–701
129. Vajdy, M. et al. Hepatitis C virus polyprotein vaccine
formulations capable of inducing broad antibody and
cellular immune responses. J. Gen. Virol. 87,
130. Wille-Reece, U. et al. Toll-like receptor agonists
influence the magnitude and quality of memory T cell
responses after prime-boost immunization in
nonhuman primates. J. Exp. Med. 203, 1249–1258
131. Kumar, S. et al. CpG oligodeoxynucleotide and
Montanide ISA 51 adjuvant combination enhanced the
protective efficacy of a subunit malaria vaccine. Infect.
Immun. 72, 949–957 (2004).
132. Garcon, N., Heppner, D. G. & Cohen, J. Development
of RTS, S/AS02: a purified subunit-based malaria
vaccine candidate formulated with a novel adjuvant.
Expert Rev. Vaccines 2, 231–238 (2003).
This reference presents an example of the
co-development of a vaccine antigen with a
complex adjuvant formulation, composed
of an oil-in-water emulsion containing MPL
and QS21. It presents the different steps
followed to select the best formulation, from
preclinical evaluation to application in clinical
133. Alonso, P. L. et al. Duration of protection with RTS,
S/AS02A malaria vaccine in prevention of Plasmodium
falciparum disease in Mozambican children: single-
blind extended follow-up of a randomised controlled
trial. Lancet 366, 2012–2018 (2005).
134. Stewart, V. A. et al. Cutaneous delayed-type
hypersensitivity (DTH) in a multi-formulation
comparator trial of the anti-falciparum malaria vaccine
candidate RTS, S in rhesus macaques. Vaccine 24,
135. Zaks, K. et al. Efficient immunization and cross-
priming by vaccine adjuvants containing TLR3 or TLR9
agonists complexed to cationic liposomes. J. Immunol.
176, 7335–7345 (2006).
136. Kazzaz, J. et al. Encapsulation of the immune
potentiators MPL and RC529 in PLG microparticles
enhances their potency. J. Control Release 110,
137. Xie, H. et al. CpG oligodeoxynucleotides adsorbed
onto polylactide-co-glycolide microparticles improve
the immunogenicity and protective activity of the
licensed anthrax vaccine. Infect. Immun. 73, 828–833
I wish to acknowledge J. Almond, N. Burdin, P. Chaux, F.
Dalençon, J. Haensler and E. Trannoy for their support, discus-
sions and critical reading of the manuscript. They synergized
efficiently to help me formulate my work and target it in the
right direction. Innate immunity and adjuvants are large fields
and I apologize for not having mentioned numerous important
works because of space constraints.
Competing interests statement
The author declares no competing financial interests.
The following terms in this article are linked online to:
Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.
NOD1 | NOD2 | TLR2 | TLR3 | TLR4 | TLR7 | TLR8 | TLR9 | TLR11
Entrez Genome Project: http://www.ncbi.nlm.nih.gov/
Escherichia coli | Vibrio cholerae
MDA5 | RIG-I
Access to this links box is available online.
NATURE REVIEWS | MICROBIOLOGY
VOLUME 5 | JULY 2007 | 517
FOCUS ON VACCINES — PROGRESS & PITFALLS
© 2007 Nature Publishing Group
Page 14 Download full-text