What translatable knowledge from dengue vaccine design can we pass onto future anti-parasitic vaccine development?

KEYWORDS:

• Dengue
• dengue vaccines
• antibody
• antibody dependent enhancement
• cell-mediated immunity
• parasite vaccines

1. Introduction

The global burden of infectious diseases remains high as measured in the loss of human lives and productivity. Vaccination has been key to the reduction of infectious disease burden. Successful vaccination programs against childhood infections have significantly increased the survival rates of children globally. These achievements notwithstanding, effective vaccines against important infectious agents such as dengue viruses and most human parasites remain elusive. The development and licensing of the first dengue vaccine in 2015 (Dengvaxia, CYD-TDV) was a major milestone in the effort to combat dengue. However, it has become evident that this vaccine may not protect but rather increase the risk of infection in some individuals. The lessons learned from the development of dengue vaccines highlight the complex interactions between pathogens and the immune system, and these may be applicable to the development of vaccines against parasitic infections.

2. Dengue and dengue vaccines

Dengue viruses (DENV) are a group of four serologically distinct flaviviruses that are transmitted to human through mosquito vectors [1]. In endemic areas where all four serotypes co-circulate, repeated infections are common. The first or primary infection occurs early in childhood and is usually asymptomatic or manifests as a mild febrile illness. The immunity resulting from a first, or primary infection provides only short-lived cross-protection to other serotypes, rendering the host susceptible to a secondary infection. A secondary infection is usually associated with more severe illness ranging from dengue fever (DF), a febrile illness with a constellation of symptoms to a more severe form of the disease characterized by plasma leakage, severe thrombocytopenia, and bleeding [2]. Studies have demonstrated that the risk of severe dengue is greatly increased during a secondary infection [3,4]. Immune responses elicited during primary infections have been postulated to enhance the infection and severity through multiple mechanisms including infection-enhancing antibodies and serotype cross-reactive memory CD4 and CD8 expressing T cells that produce potentially harmful mediators such as TNF-α [5–7]. On the other hand, T cell response has also been associated with protection against severe illness [8,9]. Th1 cytotoxic CD4 + T cell response directed against DENV capsid antigen which may afford protection has been reported [10]. The protective or injurious effects of T cell response appear to depend on several factors including host genetic background such as HLA haplotypes [10,11].

The challenge in dengue vaccine development is the induction and maintenance of protective immunity against multiple serotypes of DENV. Efforts to develop dengue vaccines have been on-going for decades culminating in the first licensed vaccine in 2015. The first dengue vaccine was licensed on the basis of the results of phase III clinical trials in Southeast Asia and Latin Americas. CYD-TDV is a chimeric tetravalent recombinant dengue vaccine in which prM and envelope genes of each serotype of DENV are inserted into the backbone of yellow fever 17D vaccine, a live attenuated viral vaccine with a proven safety record. The result of the clinical trials indicates that the overall efficacy of the vaccine for clinically apparent infection was approximately 60% [12,13]. The vaccine was 80% protective against severe disease. However, the efficacy for each serotype differed significantly with low efficacy against DENV type 1 and 2 in spite of the presence of serotype-specific neutralizing antibodies in vaccine recipients [12]. Due to the observed increased risk of hospitalized dengue in young vaccine recipients (age 2–5 years), the initial recommendation for vaccination was for children over 9 years of age and adults. A follow-up study demonstrated an increased risk of severe infection in vaccine recipients who were determined retrospectively to be dengue immune naïve prior to vaccination regardless of age.

Other vaccines that are in phase III clinical trial include 1) TV003/TV005, a tetravalent live attenuated vaccine with a deletion of the 3ʹ untranslated region of the genomes. Earlier studies have shown that this vaccine induced neutralizing antibodies to all four serotypes after one single dose in 90% of flavivirus naïve adult subjects and protected against subsequent challenge infection with DENV-2 [14], 2) TAK-003 (DenVax) a live attenuated chimeric vaccine based on attenuated DENV-2 PDK-53 strain [15,16]. These vaccines are currently undergoing phase III clinical trials. Other vaccines in the pipeline include inactivated whole virus vaccine, tetravalent envelope subunit vaccine, DNA vaccines, and virus-like particles (VLP). These vaccines are undergoing early phase clinical trials (Table 1).

Table 1. Dengue vaccines currently in clinical trials.

Vaccine	Vaccine type	Antigens	Phase of development
CYD live recombinant viruses (Dengvaxia-Sanofi Pasteur)	Chimeric vaccine of DENV1-4 with yellow fever viral vaccine backbone.	prM/E proteins of DENV1, 2, 3, and 4	Licensed, post phase III surveillance.
TV003/TV005 Tetravalent live, attenuated/recombinant vaccine (US-NIH)	Whole virus DENV1–3 and recombinant DENV2 in DENV4 backbone	Complete antigens of DENV-1, 3, 4, prM/E proteins of DENV-2.	Phase III
DENVax (Takeda)	Tetravalent live, attenuated/recombinant (attenuated DENV2 and recombinant DENV1/3/4 in DENV2 backbone)	Complete antigens of DENV-2, prM/E proteins of DENV-1, 3, 4.	Phase III
DPIV Tetravalent purified inactivated vaccine (GSK-WRIAR)	Inactivated whole virions	Capsid, prM/E proteins	Phase I
DEN-80E (Hawaii Biotech-Merck)	Tetravalent E protein subunit vaccine,	Truncated envelope proteins	Phase I
D1ME-VR-P (Naval Medical Research Center)	DNA vaccine	prM/E genes	Phase I

The majority of the vaccines that have entered phase III clinical trials are live viral vaccines. Vaccination with live viruses has been thought to induce balance immunity mimicking a natural infection and long-lasting immunity. However, the findings from CYD-TDV indicate that this is not always the case. They also bring into focus several deficits in our understanding of dengue immunity: 1) the mechanisms that confer protection and disease enhancement, 2) immune correlates for protection, 3) factors that contribute to the sustainability of the immune response. Although antibodies are thought to play a key role in the protection against dengue, the characteristics that distinguish between protective and infection-enhancing antibodies are currently being elucidated [17]. The lack of meaningful correlates of protection is underscored by the poor correlation between the neutralizing antibody titers in the CYD-TDV recipients and protection. Further, the characteristics of protective cell-mediated immune responses have not been well delineated and the effects of cell mediated immune responses may be HLA and disease context dependent [10,18,19].

Antigens included in a vaccine are critical in vaccine-mediated protection or disease enhancement. This is notable in the design of CYD-TDV which includes only two structural proteins of DENV and none of the non-structural proteins. Recent studies have implicated nonstructural protein-1 (NS-1) of dengue as a pathogenic factor and that antibody against NS-1 protected against disease in animal models [20]. However, some antigens may induce an undesirable immune response. Studies have suggested that antibodies to prM are serotype cross-reactive and may contribute to infection enhancement during a secondary infection [21]. The limited breadth of CYD-TDV induced immunity may be particularly relevant in vaccine recipients who were DENV naïve prior to vaccination. In an immunogenicity study of CTD-TDV recipients, DENV-specific CD4+ and CD8 + T cells producing IFN-γ and IL-13 were detected four years after vaccination. However, cytokine response to nonstructural protein-3 derived peptides was very low reflecting the lack of priming to DENV nonstructural proteins [22]. The relatively weak T cell help in DENV naïve individuals may result in rapidly waning protection. In support of this notion, TAK-003 (DenVax), a live attenuated tetravalent vaccine based on PDK-53 DENV-2 virus strain that expresses the full panel of DENV proteins, has been reported to be effective in DENV naïve vaccine recipients [15]. In addition to the breadth of antigens, how antigens are processed and presented in a vaccine is critical in shaping the immune response. A recent study in rhesus macaques vaccinated with inactivated DENV vaccine demonstrating enhancement of infection and cytokine production in vaccinated animals when subsequently challenged with live DENV [23].

3. Lessons from dengue vaccine development and their applicability to the design of parasite vaccines

Dengue differs from most parasitic infections in many aspects including the chronicity of infection, the basic biology of the organisms, the comparatively limited number of target antigens, and the presence of co-circulating multiple serotypes. However, immunity against intracellular parasites such as malaria and leishmania shares common features with immunity against DENV in that antibody and cell-mediated immunity contribute to infection prevention and resolution [24,25].

Antibody plays an important role in protection against DENV and malaria infection. In dengue, antibodies directed against conformation-sensitive epitopes appear to contribute to protection. CYD-TDV vaccine recipients developed both serotype-specific and cross-reactive antibodies depending on the viral serotypes and host pre-vaccination dengue immune status [26]. The targeted epitopes of CYD-TDV induced antibodies have not been well characterized. The specificities of protective antibodies in malaria are best characterized to P. falciparum circumsporozoite antigens (PfCSP) and some blood stage antigens such as P. falciparum reticulocyte-binding protein homologue 5 (PfRH5) [27]. RTS/S/As01, the only available human malaria vaccine, contains part of the circumsporozoite antigen which induces antibody response [28,29]. The fine specificities and the phagocytosis enhancing activity of these antibodies have been shown to correlate with protection [30]. Most vaccines developed against leishmania induce antibody responses but the role of antibodies in protection against leishmanial infection or disease resolution is not clear. However, Th1 type CD4 + T cell response characterized by the production of IFN-γ and TNF-α, and the production of reactive oxygen species have been shown to be critical in disease resolution in animals and in humans [25,31]. Current leishmania vaccines used in animals and those in human clinical trials contain either fractioned proteins of leishmania (CaniLeish, Leishmune, Leish-Tec) or recombinant proteins (LEISH-F1) formulated with adjuvants such as monophosphoryl lipid A (MPL) and QA-21 designed to enhance Th1 type response [32].

CD8 + T cells are elicited in response to viral and parasitic infections. DENV nonstructural proteins are processed efficiently though MHC class I pathway and elicit strong CD8 + T cell responses to DENV. The absence of DENV non-structural antigens in CYD-TDV likely leads to relatively weak CD8 + T cell response in DENV naïve vaccine recipients. Although parasites reside mainly in phagosomal vacuoles, parasite-specific CD8 + T cells can be primed though antigen cross presentation. Malaria-specific CD8 + T cells can be effectively induced by attenuated sporozoite vaccination and have been shown to recognize sporozoite and liver stage antigens [24]. The current malaria vaccine, RTS/S/As01, has been shown to induce IFN-γ producing CD8 + T cells [33,34]. The roles of CD8 + T cells in disease prevention and progression in leishmanial infection are complex and appear to depend on the stages of the infection and parasite burden. CD8 + T cells produce IFN-γ early in infection, which can promote protective Th1 type cytokine responses. However, cytotoxic CD8 + T cells have been shown to mediate tissue damage when the parasite load is high such as in cutaneous and mucosal leishmaniasis. A recent leishmanial vaccine employing simian adenovirus (ChAd63) expressing L. donovani antigens has been shown to induce antigen-specific CD8 + T cells [35].

4. Expert opinion

Findings from the first licensed DENV vaccine underscore the gaps in our understanding of immune-mediated protection to this relatively simple virus. Identifying the key immune functions involved in protection of infection is critical to the development of an effective vaccine. The life cycles of parasites and the repertoire of antigens expressed by any given parasite are an order of magnitude more complex than those of DENV. Due to the complex life cycle and the distinct antigens expressed during the different phases of the life cycle of a given parasite, selection of antigens that will induce immune responses targeting these antigens is an important consideration. For organisms which are exposed to extracellular environment, or express antigens on the surface of infected cells, vaccines should induce strong neutralizing antibodies or antibodies that can mediate elimination of infected cells by antibody-dependent cytotoxicity. T cells are critical in providing help to B cells for antibody production and play an important role in pathogen elimination by direct cytolysis or through cytokine effects. The direct cytolytic effects of T cells are limited to cells expressing MHC molecules and therefore may be less important in a setting where non-nucleated cells are the major targets of infection such as the erythrocytic stage of malaria. Given the complexity and the logistics challenges of live attenuated whole-organism vaccines, the likely approach is subunit vaccines such as purified proteins, virus-like particles, or vectors expressing selected antigens. To achieve these goals, selecting the methods of antigen delivery is critical in the induction of desirable immune response. Novel forms of antigen delivery such as immunization with DNA or RNA or recombinant viral vectors which can induce both CD8+ and CD4 + T cell immunity as well as a strong antibody response offer potential effective alternatives for vaccination.

Another important deficit in our understanding involves factors contributing to the maintenance of immune effector functions. For example, the duration of antigen persistence and the requirement for natural boosting for the continuing stimulation of B cells to produce long-lasting plasma cells, and to maintain effector functions of T cells are not known. The maintenance for T cell effector functions appears to be critical for the protection and disease resolution in parasitic infection such as leishmania and malaria. For organisms that are transmitted through insect vectors, the immune response to vector derived antigens and the microbiome of the vectors and of the host skin likely influence the immune response to organisms both in term of the quality and the magnitude of the response. None of the existing vaccines truly replicate natural immune induction. A better understanding of naturally acquired immune responses in humans at the site of inoculation will be informative in the development of vaccines against vector-borne pathogens.

In the next five years, at least two additional dengue vaccines which are in advance stage of clinical trials will likely be licensed. Currently, implementation of the RTS/S/As01 malaria vaccine in a larger population of children in Africa is on-going [36]. The experience with the first licensed DENV vaccine and the first and only malaria vaccine for human highlights the need for a long-termed post vaccination follow up in order to evaluate the vaccine efficacy and risks. The experience with the first dengue vaccine has demonstrated that the efficacy of a vaccine is determined by the immunological experience of the host. The pre-vaccination adaptive immunity likely plays a key role in shaping the breadth and the durability of the immune response induced by vaccination. Therefore, selecting the right hosts for vaccination is a critical issue. Vaccines that induce immunity that requires natural boosting to maintain protection may be less effective in immune naïve individuals and put them at risk for increased disease severity. This at-risk population will likely be expanding if a partially effective vaccine is widely administered leading to a gradually decline in pathogen exposure and natural boosting. The effects of age on immune response (independent of the pathogen-specific immune status) to a vaccine will also need to be addressed. This is particularly relevant to pathogens with a high degree of transmission early in life such as malaria and dengue. The information from the long term follow up of vaccine recipients will be instrumental in the improvement in the designs of future vaccines.

Declaration of interest

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgement

The opinions expressed are those of the authors and do not represent the official position of the National Institutes of Health.

Additional information

Funding

A Srikiatkhachorn is supported in part by the National Institutes of Health (grant P01 AI034533).