Human Zika infection induces a reduction of IFN-γ producing CD4 T-cells and a parallel expansion of effector Vδ2 T-cells

The definition of the immunological response to Zika (ZIKV) infection in humans represents a key issue to identify protective profile useful for vaccine development and for pathogenesis studies. No data are available on the cellular immune response in the acute phase of human ZIKV infection, and its role in the protection and/or pathogenesis needs to be clarified. We studied and compared the phenotype and functionality of T-cells in patients with acute ZIKV and Dengue viral (DENV) infections. A significant activation of T-cells was observed during both ZIKV and DENV infections. ZIKV infection was characterized by a CD4 T cell differentiation toward effector cells and by a lower frequency of IFN-γ producing CD4 T cells. Moreover, a substantial expansion of CD3+CD4−CD8− T-cell subset expressing Vδ2 TCR was specifically observed in ZIKV patients. Vδ2 T cells presented a terminally differentiated profile, expressed granzyme B and maintained their ability to produce IFN-γ. These findings provide new knowledge on the immune response profile during self-limited infection that may help in vaccine efficacy definition, and in identifying possible immuno-pathogenetic mechanisms of severe infection.

The aim of this study was to study and compare the phenotype and functionality of T-cells in patients with acute ZIKV and Dengue viral (DENV) infections.

ZIKV infection expanded CD8 and DN T cells and induced T-cell activation. The characterization
of T-cell subsets in healthy donors (HD), ZIKV and DENV-infected patients were performed by multiparametric flow cytometry (Fig. 1). Representative panels from one HD, one ZIKV-and one DENV-infected patient are shown in Fig. 1a. The CD8 T-cells frequency was different in the three groups (Kruskal Wallis, KW < 0.05). In particular, when compared to HD, a significant higher CD8 T-cell frequency was observed both in ZIKV-and in DENV-infected patients (Fig. 1b). In contrast, no difference in CD4 T-cell frequency was observed among groups (Fig. 1c). Finally, the frequency of CD3 + CD4 − CD8 − T-cell population (double negative, DN T-cells) was different in the three groups (KW < 0.05). Specifically, a significant expansion of DN T-cells was observed during ZIKV infection (Fig. 1d).
The impact of ZIKV infection on T-cell activation was evaluated by analyzing the expression of activation markers CD38 and HLA-DR on CD8 (Fig. 2a,b), CD4 (Fig. 2c,d) and on DN (Fig. 2e,f) T-cells. A significant difference among HD, ZIKV and DENV was observed in the activation profile of CD8 (KW < 0.05) and of CD4 T cells (KW < 0.05). Specifically, when compared to HD, a higher frequency of CD38 pos and of CD38 pos /HLA-DR pos CD8 T-cells was observed both in ZIKV and in DENV patients (Fig. 2a). Moreover, a significant higher frequency of CD38 neg /HLA-DR pos CD8 T-cells was observed in ZIKV-patients than in HD (Fig. 2a). CD4 T-cells showed a lower level of activation than CD8 T-cells both in ZIKV and DENV patients but CD38 pos CD4 T-cells were significantly higher in ZIKV and DENV than in HD (Fig. 2c). Although a trend of CD38 increase on DN T cells during DENV infection, no significant difference was observed on the activation of DN T-cells among HD, ZIKV and DENV, probably due to the small sample size (Fig. 2e). The analysis of CD8 T-cells did not show any difference on the differentiation profile between ZIKV, DENV and HD (Fig. 3a). In contrast, a significant difference was observed in the differentiation profile of CD4 (KW < 0.05) and DN (KW < 0.05) T-cells. In particular, ZIKV patients showed a lower frequency of CM-CD4 T-cells in comparison to both HD and DENV patients, and a parallel higher frequency of EM-CD4 T-cells and TEMRA-CD4 T-cells in ZIKV respect to HD (Fig. 3c). Similar results were obtained by analysing DN T-cells: indeed, a lower frequency of CM-DN T-cells was observed in ZIKV in comparison to both HD and DENV patients and a parallel higher frequency of TEMRA-DN T-cells in ZIKV respect to HD (Fig. 3e). Finally, an inverse correlation between days after symptoms onset and EM CD8 T-cell frequency was observed (Pearson R: −0.76, R Squared 0.58, p = 0.04).

ZIKV infection induced the expansion of DN T-cell expressing Vδ2 TCR.
To further characterize the phenotype of expanded DN T-cells observed during ZIKV infection, the expression of the γδ TCR (Vδ1 and Vδ2) and CD56 NK marker was analysed by flow cytometry. Results showed that the large majority of DN T-cells expressed a Vδ2 TCR (Fig. 4a), suggesting that the DN T-cell population expanded during ZIKV infection belongs to the Vδ2 T-cell subset. Accordingly, the frequency of Vδ2 T-cells within CD3 T-cells was significantly different among HD, ZIKV and DENV (KW < 0.05). In particular, Vδ2 T-cells was significantly higher in ZIKV patients than in HD and in DENV patients (Fig. 4b). Notably, the Vδ2 T-cell expansion was more pronounced in patients sampled during the first 2-3 days from symptoms onset [Pt1 and Pt2: Range (23.8-26.6%)] than in the other patients sampled after 6 days [Pt3-6: Range (1.5-13.1%)], and an inverse correlation between days after symptoms onset and Vδ2 T-cell frequency was observed (Pearson R: −0.84, R Squared 0.70, p = 0.01, Fig. 4c).
Although not reaching the statistical significance, the analysis of granzyme expression showed a trend of increase of Granzyme B positive Vδ2 T-cells in ZIKV patients compared to HD (KW = 0.07, Fig. 4d).

ZIKV infection induced a reduction in IFN-γ production by T-cells.
In order to define the functional properties of αβ and γδ T-cells during ZIKV infection, we tested their ability to produce cytokines by EliSpot assay (Fig. 5a) and flow cytometry (Fig. 5b,e) after mytogenic stimulation. Representative cytometric panels are shown (Fig. 5e). As shown in Fig. 5a, a significant reduction in the frequency of IFN-γ producing T-cells was observed in ZIKV patients than in HD [ZIKV: median 593 SFC/10 6 PBMC (IQR: 517-620) vs. HD: median 1165 SFC/10 6 PBMC (IQR: 950-1180), p < 0.05]. We then analysed three different cytokines by flow cytometry, focusing on those that were found elevated in acute ZIKV patients 10 . The analysis of IFN-γ-producing CD4 T-cells revealed a significant difference among HD, ZIKV and DENV (KW < 0.05). In particular, ZIKV infection was associated to a reduction of IFN-γ production by CD4 T-cells (Fig. 5c). Of note, Vδ2 T-cells maintained their ability to produce IFN-γ as well as HD and DENV (Fig. 5d). No significant differences were observed in IL-17A and IL-2 producing CD4, CD8 and Vδ2 T-cells.

Discussion
The definition of the immunological response to ZIKV infection in humans represents a key issue to identify a protective profile useful for vaccine development and for pathogenesis studies. Our study is the first to report the dynamics of T-cell profile and function in patients with acute ZIKV infection. There are 4 important findings from our study: 1) activation of both CD4 and CD8 T-cells; 2) CD4 T cell differentiation toward effector cells 3) substantial expansion of effector Vδ2 T-cells in the first days after symptoms onset 4) cytokine modulation in CD4 T-cells with a reduction of IFN-γ production.
Activation of CD8 and CD4 T-cells has been extensively described in several viral infections and likely represents the efforts of the immune system to counteract viral replication. Nevertheless, an excessive T-cell activation observed during severe viral diseases such as severe DENV 15 or Ebola viral (EBOV) infection [16][17][18] may be harmful to the host, by increasing inflammation and promoting anergy of protective T-cells. We found that ZIKV infection induced a significant activation of CD8 and CD4 T-cells, confirming recent data obtained in mice 14 . The frequency of CD8 T-cells co-expressing both CD38 and HLA-DR was similar in ZIKV and in DENV patients 15 , but was lower than during EBOV infection 16,18 , suggesting that a moderate activation may be associated to a protective activity and to a mild profile of the disease. In acute infection such as of DENV 19 or Ebola virus [16][17][18] , the activation of T-cells is associated to an increased expression of apoptotic markers. The role of apoptosis during ZIKV infection remains to be defined.
The ability of T-cells to differentiate into effectors is a key feature of a protective response. During ZIKV infection, a CD4 T-cells differentiation towards effectors was described, suggesting the induction of a well coordinate immune response. In particular, CD4 T-cells differentiated in effector memory and terminally differentiated cells, suggesting the acquisition of a cytokine-producing and cytotoxic profiles. Other study are necessary to deeply explore the functional properties of effector CD4 T-cells. An effector phenotype of T-cells was also described in the mouse model of ZIKV infection 14 . Nevertheless, although the expression of an effector phenotype, a significant reduction of IFN-γ producing CD4 T-cells was observed respect to both HD and DENV patients. Was can speculate that CD4 T-cells during Zika infection were polarized to produce cytokine and chemokines other than IFN-γ. The ability of ZIKV to inhibit Type-I IFNs production and response was reported 9, 20 but the reduction of IFN-γ production by T-cells represent a new finding whose role in the context of protection/pathogenesis needs further investigations. No differences in the frequency of IL-2 and IL-17-producing CD4 and CD8 T-cells was observed, suggesting that ZIKV did not modify the ability of T-cells to produce these cytokines that indeed were found higher in the sera of ZIKV patients 10 .
The current study reports for the first time a substantial expansion of Vδ2 T-cells during ZIKV infection, that was more pronounced during the first days from symptoms onset. Expanded Vδ2 T-cells presented an effector phenotype and expressed granzyme B. A pillar role of Vδ2 T-cells in a well orchestrated immune response to viral infection is well documented 21 . Indeed, human γδ T-cells may affect the progression and outcome of infectious diseases 22,23 . Activated Vδ2 T-cells are able to exert direct antiviral activities and to perform several stimulatory activities on both innate and adaptive immune cells [24][25][26][27] . Several reports suggest a protective role of γδ T-cells during other acute viral infections 22,23,[28][29][30] . γδ T-cells expand quickly in response to WNV infection, produce significant amount of IFN-γ, limiting the viral load and protecting the host from lethal encephalitis 29,30 . We could not directly correlate the Vδ2 T-cell expansion with a protective effect during ZIKV infection, since all the patients included in this study showed a mild clinical course of the disease and no sequelae were observed. A direct comparison between severe and mild diseases may be helpful in supporting a protective role of γδ T-cells during ZIKV infection. The analysis of Vδ2 T-cells may also be important to better understand the pathogenesis associated with severe clinical complications of ZIKV infection such as microcephaly in foetuses 2, 3 and/or Guillain Barré syndrome in adults 31 . Interestingly, an expansion of Vδ2 T-cells has been associated with recurrent abortions 32,33 . Furthermore, the Th1 profile of expanded Vδ2 T-cells might contribute to placenta damage and/or inflammation that has been correlated with foetal brain damage during ZIKV infection 34 . Finally, γδ T-cells may play a role during autoimmune diseases 35 and in particular during Guillain Barré syndrome [36][37][38] , contributing to autoimmune damage.   Table 1. The serological and virological data are reported in Table 2. Elispot assay. T-cell functionality during Zika acute infection was assessed by detecting interferon-gamma (IFN-γ) production using an enzyme-linked immunosorbent spot-forming cell assay (ELISpot) after PHA stimulation. Peripheral blood mononuclear cells (PBMCs) were thawed in culture medium (RPMI 1640, 10% FCS, 2 mmol/liter L-glutamine) and assessed for vitality by Trypan Blue exclusion, counted, and plated at 3 × 10 5 cells/ well in ELISpot plates (AID GmbH, Strabberg, Germany). PBMCs were then stimulated with PHA, included in the Elispot kit, for 24 hours with 5% of CO 2 . At the end of incubation, the ELISpot assay was developed according to manufacturer's instructions. Spontaneous cytokine production (background) was assessed by incubating PBMC with 1 μg/ml αCD28 and αCD49d (IgG1, clones CD28.2 and 9f10, respectively; Becton Dickinson, Mountain View, CA). Results are expressed as spot forming cells (SFC)/10 6 PBMCs in stimulating cultures after subtracting spontaneous background.