In a recent report in Nature, Nemazee and colleagues1 put forward a proposal that verges on heresy to some immunologists involved in vaccine development. Working with mice, they show that it is possible to stimulate strong antibody responses using an experimental antigen and common vaccine adjuvants without any contribution from Toll-like receptor (TLR) pathways. Although TLR ligands have been an important focus of recent vaccine research, the authors propose that their exclusion from vaccines may avoid unwanted side effects.

Vaccination has been the single greatest success of biomedical science, enabling eradication or control of smallpox, polio and many other diseases. Nevertheless, there is still much room for further development as we have no effective vaccines against malaria, AIDS and many other infectious diseases. Recent research on vaccine development has focused on improving adjuvants (from the Latin adjuvare, to help), immune boosters that trigger an effective response against the vaccine antigen.

Vaccine efficacy is generally linked to B cell–dependent production of protective antibodies that block pathogen infection. B cells normally require help from antigen-specific T cells to differentiate into antibody-secreting cells (Fig. 1). For some types of vaccines under development, such as therapeutic HIV or cancer vaccines, different antigen-specific T cells, killer T cells, are desired for their ability to directly and specifically kill infected or tumor cells. A natural infection normally triggers a protective immune response—a person can get infected with the measles virus only once, because the immune system develops 'memory' against the virus (antibodies against measles antigens) that prevents reinfection with the same virus. In addition, memory is specific—a person immune to measles could still be infected by a different virus, such as chickenpox. Immune memory and specificity characterize 'adaptive' immune responses, in contrast to 'innate' immune responses, which act to control most infections until adaptive immunity is generated.

Figure 1: Development of an effective vaccine.
figure 1

Kim Caesar

At a minimum, a vaccine must have an antigen and an adjuvant (the adjuvant is inherent in the case of live, attenuated vaccines2). To produce antibodies to the vaccine, B cells are thought to require two signals: (i) binding to the antigen through the surface receptor and (ii) a co-stimulatory signal provided by T cells activated against the same antigen by dendritic cells. Adjuvants are essential for the provision of the second signal, but some adjuvants, such as TLR9 agonists, also can provide a third stimulatory signal through TLR9 expressed in the B cell. Some NOD-like receptors (NLRs) and RIG-I-like receptors (RLRs) are thought to be expressed within all immune cell types and can trigger cell activation11, making them interesting candidate targets for the development of new vaccine adjuvants.

Full size image

Early vaccines were attenuated, or weakened, strains of the pathogen that stimulate a protective response without causing sickness. However, because attenuated vaccines can be lethal for people with weakened immune systems, the trend in vaccination has been to move to the use of 'subunit vaccines', which are composed of highly purified antigens that can be targeted specifically by the immune system. The role of the adjuvant is to stimulate the immune system to trigger an effective response to the purified subunit, which by itself is ignored by the immune system. The mechanisms of action of adjuvants remain unclear.

For many years, the gold-standard adjuvant was an emulsion of killed mycobacteria in oil, known as complete Freund's adjuvant (CFA). Although effective, CFA is too toxic for human use, especially if injected repeatedly, and even the oil component alone (incomplete Freund's adjuvant, IFA) causes injection-site reactions severe enough to preclude its use except in cancer vaccines. Adjuvant discovery was a matter of trial and error for many years, which may explain why alum, an aluminum salt developed in the early 1920s, is still the most widely used adjuvant. An especially promising class of adjuvants in development today is TLR ligands, which are thought to work by binding to TLRs, a family of ten immune proteins in humans.

A pivotal advance in understanding how vaccines work was the realization that the optimal development of an adaptive immune response requires that the vaccine activate a concomitant innate immune response. One of the most effective vaccines known, the live attenuated yellow fever vaccine 17D, stimulates multiple TLRs, the activation of which is essential for the generation of the adaptive immune response2. Another study even suggested that the production of antibody responses requires TLR activation in B cells3.

It has become widely accepted that innate immunity can easily be triggered by stimulating TLRs and will lead to strong adaptive immunity. Indeed, synthetic ligands for almost all of the TLRs have been reported to be effective vaccine adjuvants in animal studies, and ligands for TLR4 and TLR9 have enhanced vaccine responses in human clinical trials.

A TLR4 agonist, monophosphoryl lipid A (MPL; often called 'Ribi adjuvant'when mixed in trehalose dicorynomycolate, a formulation that provides additional adjuvant activity), boosts the percentage of people who develop protective antibody levels to a hepatitis B vaccine within three months of initial vaccination from ∼60% of subjects receiving the TLR-free commercial vaccine to 100% of the subjects receiving the MPL-containing vaccine4. The hepatitis B vaccine containing MPL was recently licensed in Europe (Fendrix).

Two TLR9-agonist adjuvants, oligodeoxynucleotides containing unmethylated CpG motifs (CpG-ODN), have been tested in clinical trials as adjuvants to the hepatitis B vaccine5,6 and induce protective antibodies in most normal subjects within just two weeks of the first vaccine dose, compared with none in the subjects receiving a conventional vaccine. For one of the CpG-ODN, CPG 7909, 100% of the vaccinated subjects achieved protective antibody levels within just six weeks5. Remarkably, hepatitis B vaccination with CPG 7909 rapidly induced long-lasting (>three years) protective antibody titers even in HIV-infected subjects who had previously failed to respond7. Co-administration of CPG 7909 with a one-tenth normal dose of an influenza vaccine in normal volunteers restored the full level of antigen-specific IFN-γ secretion8. In a tumor vaccine, CPG 7909 with IFA increased by ∼tenfold the frequencies of antigen-specific CD8+ T cells in melanoma patients, making this the strongest adjuvant yet reported for inducing this type of immune response9.

However, the use of MPL or CpG ODN vaccine adjuvants can increase the frequency of the usual vaccine-associated toxicities: injection-site reactions and transient flu-like symptoms, such as muscle and joint aches, fatigue, nausea, malaise and headache, which are usually graded as minimal to mild4,5,6,7,8,9.

Nemazee and colleagues used mice that were genetically deficient in all the known TLR pathways to determine whether these pathways were needed for the mice to make an antibody response against a chemically modified antigen (not an actual vaccine) in combination with any of four of the standard adjuvants mentioned above: alum, CFA, IFA and the Ribi adjuvant. Both CFA and Ribi adjuvant contain TLR ligands (along with other components that could have TLR-independent adjuvant activity), and therefore might have been expected to be inactive in TLR-deficient mice. Surprisingly, all these adjuvants drove strong antigen-specific antibody responses in the absence of any functional TLR pathways.

However, no assays were performed to examine the T-cell response to the antigen, and the duration of the immunity was not tested. In addition, there was no challenge experiment to determine whether the antibodies would be equally effective against a pathogen—previous studies in monkeys vaccinated with the FDA-approved anthrax vaccine, AVA, with or without CPG 7909 have shown that the TLR9-induced antibodies provide improved protection against lethal anthrax challenge upon passive transfer of the immune serum into mice10.

What do these results mean for investigators developing new vaccines? The authors did not address the question of the safety and efficacy of TLR agonists as vaccine adjuvants. In fact, based on the limited human clinical trials performed to date, it appears likely that adding a specific TLR4 or TLR9 agonist can dramatically enhance the efficacy of a vaccine without adding major toxicity. However, TLRs are no longer the only game in town—the new findings point to the existence of other pathways for the activation of specific antibody production that might also be harnessed for the development of new generations of adjuvants and vaccines (Fig. 1). At present, it is unclear whether these TLR-independent pathways are independent of the B cell, whether they are mediated by soluble factors produced by some other cell type or whether they involve cooperative interactions of several immune cells. Candidates for these other pathways are the recently described families of NOD-like receptors (NLRs) and RIG-I-like receptors (RLRs)11, but further studies are now needed to clarify the possible contributions of these and other pathways in vaccination.