Phenotypical characterization of regulatory T cells in acute Zika infection

Abstract

Zika virus (ZIKV), alongside Dengue virus (DENV), Chikungunya virus (CHIKV), and Yellow Fever Virus (YFV) are prevalent arboviruses in the Americas. Each of these infections is associated with the development of associated disease immunopathology. Immunopathological processes are an outcome of counter-balancing impacts between effector and regulatory immune mechanisms. In this context, regulatory T cells (Tregs) are key in modulating the immune response and, therefore, in tissue damage control. However, to date, Treg phenotypes and mechanisms during acute infection of the ZIKV in humans have not been fully investigated. The main aim of this work was to characterize Tregs and their immunological profile related to cytokine production and molecules that are capable of controlling the exacerbated inflammatory profile in acute Zika infected patients. Using whole blood analyses of infected patients, an ex vivo phenotypical characterization of Tregs, circulating during acute Zika virus infection, was conducted by flow cytometry. We found that though there are no differences in absolute Treg frequency between infected and healthy control groups. However, pro-inflammatory cytokine up-regulation such as IFN-γ and LAP was observed in the acute disease. Furthermore, acute ZIKV patients expressed increased levels of CD39/CD73, perforin/granzyme B, PD-1, and CTLA-4, all markers involved in mechanisms used by Tregs to attempt to control strong inflammatory responses. Thus, the data indicates a potential contribution of Tregs during the inflammatory ZIKV infection response.

Graphical abstract

1. INTRODUCTION

Arboviruses cause a range of serious infections in humans and constitute a major public health concern in tropical and subtropical areas [1]. The Zika virus (ZIKV) belongs to the Flaviviridae family and Flavivirus genus. This virus was first isolated in 1947 from a monkey (Macaca mulatta), captured in the Zika forest in Uganda, and then, from humans in 1968, in Nigeria. Mosquitoes of the Aedes genus are responsible for ZIKV transmission. During the 2016 pandemic, ZIKV was one of the most prevalent arboviruses in the Americas, along with Dengue (DENV), Chikungunya (CHIKV), and Yellow Fever (YFV) [2,3]. Zika is predominantly an asymptomatic disease, only 20% develop a mild illness characterized by an array of symptoms, such as fever, skin rash, conjunctivitis, muscle and joint pain, and other sporadic symptoms [1]. Although Zika often only causes mild symptoms, our group recently reported that arbovirus cases, including ZIKV cases, may be associated with induction of neuropathology [4]. It has been demonstrated that type I and II interferon responses play an essential role in the initial control of Flavivirus infections [3]. In the absence or delay of these first responders, acute viral infection may often progress to immunopathology [5]. ZIKV-infected patients show increased levels of pro-inflammatory cytokines such as IL-1β, IL-6, IL-9, IFN-γ, and IL-17, as well as chemokines CXCL10, CXCL12, CCL2, and CCL3 [6]. In ZIKV infection, as with the other human arboviruses, there is evidence for immunopathological damage by antiviral T cells [7].

Mechanisms behind these pathologies are commonly associated with excessive, poorly controlled immune effector responses, highlighting the importance of regulatory mechanisms, especially by regulatory T cells [8]. The role of T cell subsets (including Tregs) during acute ZIKV infection is poorly understood. Thus, the objective of this study was to phenotypically characterize and assess the potential role of the Treg subset (CD4⁺CD25^highFOXP3⁺) during acute ZIKV infection.

2. MATERIAL AND METHODS

2.1. Patient recruitment and blood collection

Whole blood (WB) was harvested from 42 patients suspected of arboviruses infection, of which 7 were confirmed for ZIKA infection by quantitative real-time PCR (qPCR). These patients with Zika-like fever symptoms were collected at the Hospital da Polícia Militar General Edson Ramalho, João Pessoa and Hospital Materno Infantil João Marsicanano, Bayeux, both in Paraíba, Brazil with informed consent and institutional ethical approval. Samples were tested by qPCR for ZIKV (Hospital Clinical Laboratory) and blood was further analyzed using flow cytometry. All patients answered questionnaires detailing their clinical symptoms. Patients that were diagnosed with ZIKV (4 females, mean age=43.75 years, and 3 males, mean age=32.66 years) declared symptoms such as fever, myalgia, edema, conjunctivitis, headache, pruritus, rash, retro-orbicular pain, and diarrhea. A healthy control (HC) group was included which consisted of 13 individuals (7 females and 6 males, mean age=25.53 years). Individuals in the HC group had a negative ZIKV qPCR and presented no signs or previous clinical history of arbovirus infection (Table 1).

Table 1.

Demographic and symptomology of healthy controls and ZIKV infected patients

Characteristic		Total patients
Sex % (n)	Healthy (n=13)	Infected patients(n=7)
Female	53.8% (n=7)	57.1% (n=4)
Male	46.2% (n=6)	42.9% (n=3)
	Mean age (range)
	Healthy	Infected patients
Total (range)	25.53 (20–45)	39 (29–53)
Female (range)	26 (20–45)	43.75 (37–53)
Male (range)	25 (20–30)	32.66 (29–40)
Signs and Symptoms % (n)	Famale (n=4)	Male (n=3)	Total (n=7)
Fever	100% (n=4)	100% (n=3)	100% (n=7)
Conjunctivitis	50% (n=2)	33.33% (n=1)	42.85% (n=3)
Myalgia	25% (n=1)	33.33% (n=1)	28.57% (n=2)
Headache	50% (n=2)	66.66% (n=2)	57.14% (n=4)
Pruritus	25% (n=1)	33.33% (n=1)	28.57% (n=2)
Edema	50% (n=2)	0	28.57% (n=2)
Arthralgia	50% (n=2)	66.66% (n=2)	57.14% (n=4)
Rash	50% (n=2)	33.33% (n=1)	42.85% (n=3)
Retro-orbicular pain	25% (n=1)	33.33% (n=1)	28.57% (n=2)
Diarrhea	0	33.33% (n=1)	14.28% (n=1)

2.2. PCR assay

Viral RNA extractions were obtained from WB or serum samples using the QIAmp Viral Mini Kit (QIAGEN, Inc., Valencia, USA). ZIKV qPCR was performed with previously reported protocols and primers [9,10]. Experiments contained positive (nucleic acid extracted from virus stocks) and negative controls (blank reagent and water). All samples were stored at −80ºC until use.

2.3. Monoclonal antibodies (mAbs)

The antibodies used for staining were immunoglobulin FITC, PE, and PERCP controls (BD, San Diego, CA, USA), anti-CD4 (APC-Cy7- Clone RPA-T4), anti-CD25 (PE-Cy7-Clone M-A251), anti-CD39 (PE-Clone TU66), anti-CD73 (APC- Clone- AD2), and anti-PD1 (FITC- Clone MIH4) obtained from BD Biosciences (USA), along with anti-FOXP3 (Alexa fluor 488- Clone 236A/E7, Alexa Fluor 647- Clone 236A/E7), anti-Perforin (PE- Clone Delta G9), anti-Granzyme B (FITC- Clone GB11), anti-CTLA-4 (CD152) (PE- Clone BNI3), anti-TGF-β (PE- Clone TW4–2F8), anti-IL10 (APC- Clone JES3–19F1) and anti-IFN-γ (PE- Clone B27) antibodies (BD Biosciences, Thermo Fisher Scientific, USA).

2.4. Flow cytometry assay

Peripheral blood cell samples (4 mL/volunteer) were collected in heparinized vacutainer tubes for the ex vivo analysis. Leukocytes were obtained through erythrocyte lysis using a lysis solution at a volume suggested by the manufacturer (BD Biosciences, USA), during 10 min at room temperature (25°C). After centrifugation (5 min, 577 × g, 25°C), the supernatant was discarded, and leukocytes were washed with Saline Phosphate Solution (PBS) three times. At each wash cycle, the sample was centrifugated at 577 × g at 25°C for 5 min. Cells were plated in 96-well U-bottom plates with 175 μL of supplemented RPMI-1640 medium (supplemented with 10% of BFS and 1% of Penicillin/Streptomycin). A Human FOXP3 buffer set (BD Biosciences, USA) was used for intracellular staining, cells were incubated at 5% CO₂ at 37°C for 4 h. (Sigma Aldrich, USA). Next, to follow the ex vivo protocol, the plaque was centrifugated (8 min, 244 × g, 4°C), and the supernatant was removed. Extracellular conjugated antibodies (BD Biosciences, USA) were added at the volume suggested by the manufacturer. Cells were incubated for 15 min at 4°C and later washed with 150 μL of PBS/well. After centrifugation (8 min, 577 × g, 4°C) and supernatant removal, extracellular staining was fixed using 100 μL of formaldehyde 4% diluted in 100 μL of PBS and incubated at room temperature (25°C) for 20 min. Following centrifugation (8 min, 577 × g, 4°C) the supernatant was discarded and samples were washed with 150 μL/well of PBS. Again, the plaque was centrifugated (8 min, 577 × g, 4°C) and the supernatant was discarded. To perform intracellular staining, cells were permeabilized with 150 μL of permeabilization buffer (PBS+BSA (0.5%) + Saponin (0.5%)) for 10 min at room temperature (25°C). After centrifugation (8 min, 577 × g, 4°C) and removal of the supernatant, intracellular antibodies were added at a volume suggested by the manufacturer (BD Bioscience, CA, USA). Then, the plate was incubated for 30 min at room temperature (25°C) and after this 150 μL/well of permeabilization buffer was added. Following centrifugation (8 min, 577 × g, 4°C), the supernatant was removed. Finally, 200 μL/well of Wash B (PBS/BSA) was added and the samples were transferred to FACS tubes and stored at 4°C. At least 70,000 gated events were acquired using FACS CANTO II (BD Biosciences, USA), and analyzed using FlowJo software version 10.4 (TreeStar, USA).

2.5. Flow cytometry data analysis

Tregs were analyzed for their production of cytokines, enzymes, and receptors using FlowJo software (v10.0.7). Limits for the quadrant markers were set based on negative populations (cells) and isotype controls. Six different fluorochromes were associated with each analysis. The fluorochromes, anti-CD3 PE-Cy7, anti-CD4 APC-Cy7, anti-CD25 PE-Cy7, and anti-FOXP3 Alexa Fluor 488, were used to identify the cell population of interest. In this manner, we first selected the total lymphocyte gate through size-FSC-A and granularity-SSC-A profiles followed by singlets separation through FSC-Ax FSC-H parameters. Thus, we selected the CD3⁺ cells and, within this subpopulation, we analyzed the total CD4+ T cells. Next, CD4⁺CD25^high T cells were assessed and within this subpopulation, we analyzed the FOXP3+ cells. Another two fluorochromes were used to assesses intracellular or surface markers. The markers used were CD39, CD73, CTLA-4, Granzyme B, Perforin, IL-10, LAP, and IFN-γ. This analysis strategy is shown in Figure 1. All analyses were performed ex vivo and at least 70,000 gated events were acquired for later analysis using a FACS CANTO II (BD Biosciences, USA).

Figure 1. Treg cells profile in acute Zika.

Tregs were analyzed for their production of cytokines, enzymes, and receptors using FlowJo software (v10.0.7). Limits for the quadrant markers were set based on negative populations (cells) and isotype controls. Six different fluorochromes were associated with each analysis. The fluorochromes, anti-CD3 PE-Cy7, anti-CD4 APC-Cy7, anti-CD25 PE-Cy7, and anti-FOXP3 Alexa Fluor 488, were used to identify the cell population of interest. Representative plots of gating strategy for flow cytometry showing different characterization of human Tregs. Through analysis in the lymphocytes gate, the dot-plot was selected for FSC-A versus side-scatter area (SSC-A) (A) followed by the forward-scatter area (FSC-A) versus forward-scatter height (FSC-H) plot for doublet exclusions (B). Then, a CD3 versus CD4 plot was used to gate CD4 T cells (C), followed by CD4+ T cells expressing CD25^high (D), and finally, FOXP3+ cells were gated within CD4⁺CD25^high cells (E). This strategy was used in both groups (healthy controls [HC] and Zika virus patients). The bars represent differences between the means of the ZIKV group compared to the control group (F).

2.6. Statistical analysis

The data were normalized for all parameters evaluated by flow cytometry. For comparison of multiple groups, statistical analysis was performed using GraphPad Prism, Inc. Software (Version 7.0) by one-way ANOVA and Tukey post-hoc test. Statistical significance was determined by p<0.05.

2.7. Ethics

All experiments were performed in compliance with relevant regulations, institutional guidelines, and under the Declaration of Helsinki ethical standards. This study was approved by the National Commission of Ethics in Research (Certificate CAAE: 59833416.6.0000.5183).

3. RESULTS

Initial reports regarding Tregs and viral infection showed that T cells’ protective function was compromised by the presence of Tregs during the acute phase [11]. To investigate whether Tregs play a role during ZIKV infection, we evaluated their presence in peripheral blood. Our data demonstrated that there were no changes in Treg cell frequency during acute ZIKV infection compared to the HC group (Figure 1 A–F). In parallel, we used intracellular cytokine staining to assess cytokine frequency profiles within Treg subsets. Our data showed that Tregs from ZIKV acute patients had a lower frequency of IL-10 in Tregs (Figure 2 A–B) compared to the HC group, and a higher frequency of LAP (Figure 2 C–D), and IFN-γ (Figure 2 E–F). Note that the cytokine programs of Tregs are appreciated to be relatively plastic, encompassing IFN-γ secretion [12]. Other markers of regulatory function in Tregs were also considered to evaluate other suppression mechanisms of Tregs, such as cell-cell contact, metabolic disruption, and induction of cell death [8]. Analyses of %CD39⁺CD73⁺ co-expression on Tregs demonstrated that Tregs of ZIKV-infected patients expressed higher levels of these ectonucleotidases compared to HC Tregs (Figure 3 A–B). Moreover, %granzyme B⁺perforin⁺ in Tregs from ZIKV-infected patients showed significantly higher levels of these co-expression than the controls (Figure 4 A–B). The same pattern of up-regulation was observed in %PD-1 expression on Tregs during the acute phase infection (Figure 5 A–B). Finally, we considered the expression of CTLA-4 (CD152) in Tregs from these cohorts, identifying increased frequency in Tregs from ZIKV-infected patients compared to controls (Figure 5 C–D).

Figure 2. Treg cytokine production profile in acute Zika patients.

Representative contour plots of gates of Treg molecules associated with intracellular cytokine secretion of IL-10 (A), TGF-β (C), and INF-γ (E). Frequency of comparative expression of IL-10 (B), TGF-β (D), and INF-γ (F) by Tregs in HC and ZIKV groups. Representative bars demonstrate significant differences (p<0.05) between the means of the ZIKV group and the control group.

Figure 3. Treg metabolic disruption profile in acute Zika patients.

Representative contour plots of gates of Treg molecules associated with CD39⁺CD73⁺ co-expression (A). Frequency of comparative co-expression of CD39 and CD73 in healthy control [HC] and ZIKV groups. Representative contour plots of gates of Treg molecule associated markers.

Figure 4. Treg mechanism of cell death induction in Zika disease.

Representative contour plots of gates of Treg molecules associated granzyme and perforin production (A). Frequency of comparative expression of granzyme/perforin coproduction (B) in healthy control [HC] and ZIKV groups. Representative bars demonstrate significant differences (p<0.05) between the means of the ZIKV group and the control group.

Figure 5. Treg cell-cell mechanism profile in acute Zika patients.

Representative contour plots of gates of Treg molecules associated PD-1 (A) and CTLA-4 (C) expression. Frequency of comparative expression of PD-1 in Tregs in healthy control and ZIKV groups (B). Frequency of comparative expression of CTLA-4 in Tregs in HC and ZIKV groups (D). Representative bars demonstrate significant differences (p<0.05) between the means of the ZIKV group and control group.

4. DISCUSSION

Thus far there have been only limited studies regarding immune cell phenotype from patients with clinical Zika virus infection [13–15]. In the context of severe viral infection, modulation of Treg function may be observed as a pathogen-mediated adaptation to promote immune evasion or as a host adaptation to minimize collateral damage from the immune response [8]. Treg evaluation during acute ZIKV infection in the present study demonstrated several striking alterations to this subset associated with the acute phase of the disease. In our study context, Tregs might be considered phenotypically active when regarding their main mechanisms, in addition to an up-regulation of its still controversial pro-inflammatory modulation.

Crucial cytokine expression up-regulation of IFN-γ, LAP, and IL-10 marks the regulatory activity by Tregs observed in these individuals. Previous studies have demonstrated that IFN-γ production by FOXP3⁺ Tregs seems to happen in Th1 type inflammatory environments [16]. Zhao et al. proposed that IFN-γ production identified virus-specific Tregs that would be especially effective in inhibiting immune response during the peak of infection [17,18]. Although this dichotomous role of IFN-γ has been suggested [19], it is still unclear whether Treg cell-derived IFN-γ could perform functional immunoregulatory activity. Perhaps, IFN-γ, produced by Tregs during ZIKV infection, supports regulatory function, optimizing migratory properties for the suppression of Th1 responses [20]. On the other hand, it is possible that Tregs may adopt more of a Th1 effector phenotype [21]; further functional studies will need to be performed.

Additionally, it has been previously demonstrated that a significant presence of IL-10 and TGF-β marks a full suppression of Th1 responses by Tregs during the acute phase of ZIKV infection [22]. Studies have demonstrated that anti-inflammatory cytokines act through T lymphocyte proliferation inhibition, as shown by the non-responsiveness of macrophages to IFN-γ [23,24]. Additionally, IFN-γ was also associated with the polarization of effector T cells in peripheral Tregs, as the main stressors in the acute phase of the disease [25]. Thus, our results suggest that an increase of IFN-γ and LAP and a decrease of IL-10 in Tregs may be related to the attempt to control the initial acute phase of the immune response in ZIKV infection [17,18].

It has previously been observed that the ability of Tregs to regulate lung inflammation during virus infection is critically dependent on their ability to produce perforin/granzyme B, cytotoxic mediators normally associated with antiviral CD8⁺ T, and natural killer (NK) cells [26]. Therefore, acute viral infections of the lungs can often lead to high production of perforin/granzyme B by Tregs [26]. A similar phenomenon seems to occur in study patients infected with ZIKV during the acute phase of the disease.

Purinergic molecules producing regulation mediators are a major factor for immune suppression in highly inflammatory environments [27,28]. Accumulation of extracellular ATP is a common feature in microbe-infected tissues promoting tissue damage signs and cellular stress [29]. Thus, Treg expression of CD39 and CD73 promotes hydrolysis of extracellular ATP leading to adenosine production, an important anti-inflammatory immune regulator [30,31]. In this study, our results demonstrated that during acute ZIKV infection, the metabolic disruption mechanism might also have an important role, as Tregs exhibited an upregulation of CD39/CD73 co-expression compared to controls.

Another potentially supporting homeostatic mechanism was observed in our ZIKV dataset, whereby there is upregulation of PD-1 and CTLA-4 by Tregs. In a similar manner to what was observed during the hepatitis C virus infection, PD-1 upregulation might be acting as a contra-regulatory mechanism indicative of immune exhaustion [32]. However, CTLA-4 expression by CD4⁺ FOXP3+ T cells is indicative of potent Treg regulation activity [33]. Therefore, our results regarding CTLA-4 expression may indicate a potentially functional regulatory state of these cells during ZIKV infection.

5. CONCLUSION

The data indicate that the Zika virus infection develops an enhanced Treg profile, perhaps explaining the more controlled signs and symptoms this disease entails. Our findings may offer a better understanding of clinical differential immunopathogenesis. Therefore, offering new directions to better comprehend the divergent outcomes following ZIKV exposure, where illness ranges from asymptomatic, or mildly symptomatic characteristics, and severe disease.

List of Abbreviations:

WB

Whole blood

HC

Healthy controls

ICS

Intracellular Cytokine Staining