3.2.1. Antibody Response

Antibodies induced after natural infection and vaccination play an important role as correlates of protection against TBEV since they can prevent (lethal) disease (reviewed in [3]). Antibodies can neutralize TBEV in various manners including prevention of viral attachment or fusion and support of pathogen elimination via the complement system (reviewed in [78]). Therefore, the humoral immune response is critical for controlling virus dissemination, virus clearance and long-term protective immunity (reviewed in [79,80]).

Upon TBEV infection, the amount of TBEV-specific antibodies in human serum and cerebrospinal fluid (CSF) increases and antibodies are usually detectable with the onset of neurological symptoms [2]. The highest IgM levels, which persist for around six weeks, are found in the early stages of TBE, whereas IgG levels peak in the late convalescence period [81]. IgG antibodies can persist for several years and protect from reinfection with TBEV (reviewed in [36]). In general, natural infection induces higher virus-neutralizing (VN) antibody titers than vaccination with an inactivated virus [82]. Low serum levels of VN antibodies combined with a high number of cells (segmented granulocytes (60–70%) and lymphocytes (30–40%)) in the CSF at disease onset are thought to be indicative of a more severe clinical course of TBE [83].

Following primary TBEV vaccination, an increase in antibody titers is observed after the second dose which subsequently decline, followed by a strong booster response upon the third immunization [84]. Studies on long-term persistence of vaccine-induced antibodies showed durability of protective or VN antibodies, respectively, up to 5 years in 99% of children and above 98% in young individuals [54,56,57]. In addition, other studies reported persistence for up to 8 or even 10 years [55,85]. However, with increasing age, immunosenescence, i.e., alteration of the innate and adapted immunity, can be observed in the elderly leading to a lower immunogenicity of vaccines (reviewed in [86]). Noteworthy, avidity and functional activity of TBEV vaccine-induced antibodies are apparently neither influenced by age, age at primary immunization nor last vaccination, but are rather affected by the individual [87].

Several human linear B cell epitopes within the structural and non-structural proteins of TBEV (prM, E, NS1, NS2B, NS3, NS4B and NS5) have been identified. Two of these are located in domain I (DI) and II (DII) of the E protein, respectively, and one in the C-terminal RNA-dependent RNA polymerase region of NS5, and showed a positive reaction with almost all positive TBEV sera tested [88]. Intrinsic factors, such as the conformational flexibility (reviewed in [89]) and the maturation state of flaviviruses (reviewed in [90]), as well as extrinsic factors, such as the formaldehyde inactivation of virus particles during the manufacturing process of vaccines (reviewed in [39]; [88,91]), potentially impact sensitivity to antibody-mediated neutralization by changing the accessibility of epitopes. In addition, antibody avidity can be influenced by the addition of adjuvants, e.g., aluminum hydroxide, to the vaccines [92].

Upon TBEV infection and vaccination, epitopes in the E protein are the main target for the induction of VN antibodies (Table 2).

On the viral surface, 180 copies of the E protein structured into 90 dimers can be found in a ”herringbone-like” icosahedral arrangement (reviewed in [109]). The E monomer consists of three distinct domains (DI–DIII) connected by flexible linkers and a membrane anchor domain (DIV) (Figure 1).

Open in a separate window

Figure 1

The structure of the TBEV E protein (PDB-ID 5O6A [5]). (A) Side view of a single TBEV E protein monomer. Depicted are the four domains (DI: red, DII: yellow, DIII: blue, DIV (stem/anchor): gray) and the fusion loop (FL: orange). (B) Top view of a soluble TBEV E protein dimer. Color code same as in (A). Ribbon diagrams were prepared with UCSF Chimera [110].

DII has an elongated finger-like structure, which is formed by two loops connecting DI, the central part of the protein, in the dimeric E protein. The hydrophobic loop at the tip of DII, covered by DI and DIII in mature virions, is responsible for the fusion of the viral and endosomal membranes, therefore, termed fusion loop (FL) [5]. The sequence of the FL is highly conserved among all flaviviruses. DIII has an immunoglobulin-like structure with exposed loops protruding from the viral surface. The structure of the E protein is influenced by the pH leading to increased exposure of previous inaccessible epitopes, such as the FL [5,111]. All three ectodomains are capable of inducing a VN antibody response (reviewed in [109,112]). Epitopes can comprise not only individual domains of the E protein but also residues from adjacent dimers and dimers in the quaternary herringbone-like arrangement of E proteins at the virion surface [109,113]. In addition, the dynamic behavior of flaviviruses, commonly known as “viral breathing”, can have a great impact on antibody binding (reviewed in [89,109]).

Jarmer et al. [93] provided insights into individual specificity and the variation of the humoral immune response after TBEV infection or vaccination in humans. Using an immunoassay with recombinant E protein and combinations of the single E domains, they observed strong individual variation in antibody titers as well as immunodominance patterns concerning the individual E domains [93]. The human humoral response was dominated by EDI- and EDII-specific antibodies [88,93], whereas in mice immunized with inactivated, aluminum hydroxide-adjuvanted, purified TBE particles mostly EDIII-specific VN antibodies were detected [92]. Furthermore, depletion of antibodies with the dimeric E protein from sera of both naturally infected and vaccinated individuals led to a strong reduction in VN activity, indicating a minor role of complex quaternary epitopes in the VN antibody response. Additionally, depletion of EDI- and EDII-specific antibodies also significantly reduced the VN activity of post-infection or post-vaccination sera, suggesting that antibodies recognizing antigenic sites independent of the dimeric structure of the E protein display considerable VN activity [93]. The characterization of several flavivirus antibodies identified amino acids of the highly conserved FL in EDII as a dominant antigenic site for cross-reactive, but not cross-neutralizing, antibodies [114,115,116]. After solubilization of TBE virions with a mild detergent, EDII-specific antibodies showed a strong increase in binding avidity, indicating that the EDII-FL epitopes of native viruses may have a limited accessibility due to partial occlusion. The fact that these broadly cross-reactive, non-neutralizing antibodies are present in human sera after infection supports the hypothesis that during natural flavivirus infection, cryptic epitopes become accessible and are presented to the host immune system [116].

In addition to the E protein, the TBEV NS1 protein plays an important role as a soluble antigen. In its pentameric or hexameric form, NS1 is secreted into the extracellular space [11], inducing a NS1-specific antibody response. After natural TBEV infection, NS1-specific IgM and IgG antibodies are detectable at high levels in human sera (Table 2) [94,95,96,97]. Several studies showed that NS1 immunization or passive transfer of anti-NS1 antibodies afforded protection against infections with flaviviruses, such as DENV, YFV and WNV in animal models [117,118,119,120,121,122,123]. These findings have been confirmed for the TBEV NS1 protein [97,101,105,106,107,108,124]. Mouse studies showed that immunization with a synthetic peptide corresponding to the structurally conserved α-helix (aa37–55) of NS1 [108] or with various synthetic fragments of NS1 [101] induced partly protective immunity against lethal challenge infection with TBEV. In addition, significantly prolonged survival of TBEV-infected mice was observed after hyperimmunization with a whole recombinant TBEV NS1 protein, although all mice succumbed to infection [97].

Whether or not an anti-NS1 immune response is induced by TBEV vaccination is matter of debate. In some studies, anti-NS1 antibodies were almost exclusively observed after natural infection but not after vaccination [94,95,96]. However, recent studies on the European vaccines question these findings by demonstrating the presence of NS1 in Encepur by mass spectrometry and detection of NS1-specific antibodies by Western blot in sera of vaccinees who repeatedly received Encepur or FSME-IMMUN. Moreover, vaccination with FSME-IMMUN induced a robust anti-NS1 antibody response in mice after several immunizations [97]. Further investigations are needed to confirm these results and to gain a better understanding of the role of vaccine-induced anti-NS1 antibodies. Due to the robust IgM and IgG response, NS1 might be an interesting target for novel vaccine designs and approaches.

In addition to antibodies against the E and NS1 proteins of TBEV, antibodies against the structural prM protein have been described. Analysis of sera obtained from TBE patients and vaccinees showed the presence of prM-specific antibodies, although titers were higher after infection. However, results from depletion studies indicated that prM-specific antibodies play only a minor role in TBEV neutralization [93]. In flavivirus infection, the protective potential of prM-specific antibodies has been shown e.g., for DENV in mice studies [125,126]. In contrast, more recent studies indicated a role of cross-reacting prM antibodies in antibody-dependent enhancement (ADE) in human DENV infection [127,128]. Thus, further research is required to identify which role anti-prM antibodies, induced after infection or vaccination, have in TBEV infection.

3.2.2. CD4⁺ T Cell Response

Virus-specific CD4⁺ T cell responses are essential for the induction of antiviral immunity. CD4⁺ T cells recognize antigenic peptides of 13–25 residues presented by major histocompatibility complex (MHC) class II molecules, which are expressed on antigen-presenting cells (reviewed in [151]). Upon stimulation, they produce pro-inflammatory and antiviral cytokines which results in the recruitment and priming of other lymphoid cells. They may have cytotoxic activity against virus-infected cells and are essential for promoting virus-specific antibody responses (reviewed in [152]).

Against infections with flaviviruses, CD4⁺ T cells play an important role as was demonstrated e.g., for WNV and ZIKV [153,154]. Studies in mice showed that clearance of primary WNV infection from the CNS is dependent on CD4⁺ T cells [154]. Likewise, CD4⁺ T cells contribute to controlling of ZIKV infection and are pivotal for inducing ZIKV-specific antibodies [153]. For TBEV, adoptive transfer of CD4⁺ T cells to TBEV-infected severe combined immunodeficient (SCID) mice, demonstrated a protective role of these T lymphocytes in limiting the development of TBE and mortality [77].

Analysis of the CD4⁺ T cell response to TBEV structural proteins demonstrated that the E and C proteins, but not the prM/M protein, were a major target in infected and vaccinated study subjects (Table 2) [98,99]. Correlation of protein structure and immunodominance patterns revealed that peptides corresponding to helices α2 and α4 of the C protein dominated the CD4⁺ T cell response in both vaccinated and infected individuals, whereas the E protein-specific CD4⁺ T cell response was less focused on selected protein domains. Most E protein-specific CD4⁺ T cells recognized epitopes in β-sheets of EDIII as well as exposed loops protruding from the surface of the viral particle. Additionally, vaccination induced significant EDI- and EDII-specific CD4⁺ T cell responses, whereas CD4⁺ T cell responses against peptides from the stem region of E were predominantly observed in TBE patients [99]. Besides the identified CD4⁺ T cell epitopes in the structural proteins, several predicted T helper cell epitopes were located in the NS1 protein of TBEV [101]. In general, individual variation in CD4⁺ T cell peptide specificity as well as cytokine profile was seen in TBEV naturally infected and vaccinated persons [98,99]. Individuals infected with other flaviviruses, like JEV, revealed a comparable CD4⁺ T cell response, also against the JEV structural proteins and NS1 [155]. Comparison of immunodominance patterns from TBEV-, ZIKV-, DENV- and YFV-specific CD4⁺ T cell epitopes revealed similarities in the distribution of some epitopes. In summary, all four flaviviruses shared epitopes in the two helices α2 and α4 of the C protein and in EDI and EDII, respectively, although some amino acid sequence variation was observed in the respective epitopes [156].

Seven days post hospitalization, an elevated number of activated CD4⁺ T cells with effector functions (measured by higher human leukocyte antigen- (HLA-) DR, PD-1 and perforin expression and decreased Bcl-2 expression) was observed in TBE patients compared to control samples [104]. CD4⁺ T cells responding to DENV, ZIKV, WNV, JEV and YFV are characterized by the production of interferon-gamma (IFN-γ), interleukin-2 (IL-2) and tumor necrosis factor-alpha (TNF-α), indicating a T helper (Th) 1 cell response [153,155,157,158,159,160,161,162]. This Th1 response was also observed after TBEV infection and vaccination but differed in its specificity and functionality [98,99,100,102].

Overall, the frequency of TBEV-specific CD4⁺ T cells was higher after vaccination than after infection [99]. Antagonistic immunomodulatory properties of viral proteins, such as NS5, during viral infection [163] and/or an enhanced T cell response after booster vaccinations may explain these differences. On the other hand, a more polyfunctional phenotype of CD4⁺ memory T cells was found in study subjects who had recovered from acute TBEV infection, compared to vaccinees. In recovered TBE patients, a triple-positive phenotype (IL-2⁺TNF-α⁺IFN-γ⁺) was observed predominantly, while in vaccinated subjects a mono- or bifunctional phenotype predominated (TNF-α⁺ or IL-2⁺TNF-α⁺) [100]. Vaccination induced high levels of IL-2, TNF-α and CD40L expression in CD4⁺ T cells but the IFN-γ response was considerably lower compared to that observed in TBE patients [98,100]. The CD4⁺ T cell response patterns were comparable after all three vaccine doses, with the highest response after the second immunization [100]. The VN antibody response after vaccination correlated with CD4⁺ T cell functions, especially with frequencies of TBEV-specific IL-2⁺ and TNF-α⁺ CD4⁺ T cells [98,99,102]. Whether this is also the case after infection is unclear [98,99].

Studies on cross-reactivity of CD4⁺ T cells specific for mosquito-borne flaviviruses (reviewed in [164]) indicate that e.g., JEV-specific CD4⁺ T cells can recognize other flaviviruses which may result in stronger CD4⁺ T cell responses upon secondary infection with a heterologous virus [165,166]. Data on cross-reactivity of TBEV-specific CD4⁺ T cells with other flaviviruses are lacking.

Overall, knowledge about CD4⁺ T cell response after infection or vaccination is limited. CD4⁺ T cell subpopulations are important for protection against TBEV, however, their correlation with antibody responses and their cross-reactive potential require further evaluation.

3.2.3. CD8⁺ T Cell Response

In addition to virus-specific antibodies and CD4⁺ T cells, CD8⁺ T cells are an important immune correlate of protection against viral infections. CD8⁺ T cells recognize viral peptides of about 8–10 amino acid residues that are presented by MHC class I molecules (reviewed in [151]). So far, seven TBEV CD8⁺ T cell epitopes have been identified, all located in the non-structural proteins (NS2A, NS3, NS4B and NS5; Table 2) [103,104]. Almost all of these epitopes are highly conserved among European, Siberian and Far-Eastern TBEV subtypes [103].

It has been shown that flavivirus-specific CD8⁺ T cells directed to conserved non-structural proteins but also to structural proteins, display cross-reactivity across various flaviviruses. It is unclear if cross-reactive CD8⁺ T cells can afford protection against a heterologous flavivirus or contribute to a more severe disease outcome (reviewed in [164]). Especially for DENV infections this is a matter of debate (reviewed in [167]). However, the potential role of cross-reactive TBEV-specific CD8⁺ T cells is largely unknown.

The peak of CD8⁺ T cell responses in the peripheral blood of acutely infected TBE patients was about seven days after hospitalization during the second, neuroinvasive stage. TBEV-induced effector cells were characterized by increased HLA-DR, PD-1, perforin and granzyme B levels in combination with decreased expression of CD127, Bcl-2 and CD27. Furthermore, most of the CD8⁺ T cells (>50%) showed monofunctional properties in virus control (degranulation, cytokine or chemokine expression) 7, 21 and 90 days after hospitalization. During the course of infection, these cells maintained their monofunctionality but altered their function. The change in the functional composition was accompanied by the differentiation of TBEV-specific CD8⁺ T cells with a dominant Eomes⁺Ki67⁺T-bet⁺ effector phenotype (peak at day seven) into Eomes⁻Ki67⁻T-bet⁺ memory cells as the infection resolved (day 21 and 90) [104].

The MHC haplotype dictates the epitope and T cell repertoire as well as magnitude of the virus-specific CD8⁺ T cell response (reviewed in [151]). This was also demonstrated for TBEV by studies investigating CD8⁺ T cell responses against non-structural TBEV proteins during acute TBEV infection [103,104]. Depending on HLA alleles, the differentiation of effector memory CD8⁺ T cells in two distinct phenotypes was observed. One week after hospitalization, HLA-A2- and HLA-B7-restricted TBEV epitope-specific effector cells showed a common effector memory phenotype (CD45RA^–CCR7^–CD27⁺CD57^–). Two weeks later, mainly effector memory (CD45RA^–CCR7^–CD27⁺CD57⁺; approximately 25% of the cells) and terminally differentiated effector memory RA (CD45RA⁺CCR7^–CD27⁺CD57⁺; approximately 40% of the cells) phenotypes for HLA-A2- and HLA-B7-restricted cells, respectively, were found [103].

In the brain of TBE patients with a fatal disease outcome, TBEV and activated CD8⁺ T cells have been demonstrated [168,169]. Expression of α4β1 integrin (=VLA-4) on CD8⁺ T cells during the neuroinvasive phase of TBEV infection may indicate the migratory capacity of these cells into the CNS [103]. Since TBEV infects neurons of the CNS [169], the recruitment of CD8⁺ T cells may exert, in addition to beneficial effects, detrimental effects and may contribute to the pathogenesis through the killing of infected neurons and the local secretion of cytokines such as TNF-α. This hypothesis is supported by the observation that there were no differences in the extent of virus replication in the brains of mice surviving and in those that succumbed to experimental infection [170]. Since activated TBEV-specific CD8⁺ T cells show upregulation of perforin and granzyme B [104] and CD8⁺ granzyme B⁺ T cells have been found in the brain tissue of fatal TBE cases associated with neuronal damage [168], indeed they may contribute to the pathogenesis, although uncontrolled virus replication may result also in a worse outcome of disease. A correlation between infiltrated CD8⁺ T cells and brain damage was also described for other flaviviruses [171,172,173]. Other evidence for the detrimental effects of virus-specific CD8⁺ T cells stems from work with CD8^–/^– knockout and SCID mice. Upon infection with a neurovirulent TBEV strain, a significantly prolonged mean survival time compared to immunocompetent mice was observed. However, all mice succumbed to infection [77]. Comparable findings have been made for WNV. Infection of CD8⁺-deficient (β2-m⁻/⁻) mice inoculated with a high virus dose resulted in a prolonged mean survival time compared to immunocompetent mice. In contrast, CD8⁺-deficient mice inoculated with a low dose of WNV showed, despite an extended mean survival time, an enhanced mortality rate compared to wild type mice [173]. These findings indicate that during natural flavivirus infections, which are most likely of low virus dose, CD8⁺ T cells may exert beneficial protective effects. Of note, it has been shown that the currently used inactivated TBEV vaccines induce virus-specific CD8⁺ T cell responses inefficiently, if at all [100,174,175].

In summary, CD8⁺ T cells are thought to contribute to protective immunity and recovery from TBEV infection. However, further research should rule out potential downsides of CD8⁺ T cell-mediated immunity.