Predominant Th1 response to primary ZIKV infection via the intravenous route in LysMCre⁺Ifnar1^fl/fl mice

We have previously employed the LysMCre⁺Ifnar1^fl/fl C57BL/6 mouse model to study the response of CD8⁺ T cells to ZIKV infection [22]. To validate the model for investigation of CD4⁺ T cells, 5-week-old LysMCre⁺Ifnar1^fl/fl mice were intravenously (retro-orbitally, RO) infected with ZIKV strains MR766 or FSS13025 or were mock-infected with vehicle (10% FBS/PBS). Splenocytes were prepared 7 days later and analyzed for activated (CD44⁺) or antigen-experienced (CD49d⁺CD11a⁺) CD4⁺ T cells by flow cytometry. We found that cells from ZIKV-infected mice contained significantly more (2-fold) activated and antigen-experienced CD4⁺ T cells compared with cells from mock-infected mice (Fig 1A and 1B). We also noted a striking increase in the number of CD4⁺ T cells with cytotoxic potential (granzyme B⁺) in the spleens of ZIKV-infected mice (Fig 1C), which is consistent with the known contribution of cytotoxic effector CD4⁺ T cells in anti-flaviviral control [27, 32]. These results confirmed the suitability of LysMCre⁺Ifnar1^fl/fl mice to study the CD4⁺ T cell response to ZIKV infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Mapping of the CD4⁺ T cell response in the LysMCre⁺Ifnar1^fl/fl mouse model of primary ZIKV infection.

Five-week-old LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were infected retro-orbitally with 10⁴ FFU of ZIKV strain MR766 or FSS13025 in 10% FBS/PBS or were mock-infected (10% FBS/PBS alone). Data are the mean ± SEM of n = 4–6 mice/group. (A–C) Splenocytes were removed on day 7 post-infection and analyzed for the percentage of (A) CD44⁺CD4⁺ T cells, (B) CD49d⁺CD11a⁺ T cells, and (C) granzyme B⁺CD4⁺ T cells. (D–F) Splenocytes were removed on day 7 post-infection and stimulated with the indicated ZIKV-derived peptides and brefeldin A. The percentage CD44⁺CD4⁺ T cells producing (D) IFNγ, (E) IFNγ and TNF, and (F) IL-2 was measured by ICS. Cells stimulated with DMSO or PMA/ionomycin served as negative and positive controls, respectively. (G) Summary of the data shown in (D–F). (H) In vivo killing of ZIKV peptide-pulsed target cells. LysMCre⁺Ifnar1^fl/fl mice were retro-orbitally mock-infected (n = 4) or infected with 10⁴ FFU ZIKV MR766 (n = 5) and FSS13025 (n = 4) for 7 days, and then injected retro-orbitally with naïve C57BL/6 splenocytes (n = 4) pulsed with a pool of ZIKV peptides (E_346-360, E_644-658, NS3_1740-1754, NS4B_2480-2494, NS5_2604-2618, NS5_2738-2752) or treated with DMSO. After 12 h, the splenocytes were harvested from recipient mice, analyzed by flow cytometry, and the percentage ZIKV-specific cytotoxicity was calculated. *P < 0.05, **P < 0.01 by the Mann–Whitney U test. The production of cytokines after stimulation with each peptide was compared to the negative control (DMSO) using one-way ANOVA t-test. **** P < 0.0001. Data are representative of two independent experiments.

https://doi.org/10.1371/journal.ppat.1007474.g001

To identify potential CD4⁺ T cell epitopes in the ZIKV proteome, we screened the Immune Epitope Data Base for predicted class II (H-2 I-A^b)-binding epitopes and selected 49 peptides (the top 1% of candidates) for further testing in vitro (Table 1). Splenocytes were harvested from ZIKV- or mock-infected LysMCre⁺Ifnar1^fl/fl on day 7 post-infection and stimulated with the individual peptides. Flow cytometric intracellular staining (ICS) assays were then performed to determine the frequency of CD44⁺CD4⁺ T cells producing Th1 (IFNγ, TNF, IL-2), Th2 (IL-4, IL-5), or Th17 (IL-17A) cytokines. A strong Th1 response, as indicated by the production of IFNγ (with or without TNF) and IL-2, was induced by both the African and Asian ZIKV strain (Fig 1D–1F). Six immunodominant epitopes derived from the structural E protein (E_346-360, E_644-658) and nonstructural NS3 (NS3_1740-1754), NS4B (NS4B_2480-2494), and NS5 (NS5_2604-2618, NS5_2738-2752) proteins stimulated a particularly vigorous response by both ZIKV MR766- and FSS13025-primed CD4⁺ T cells (Fig 1D–1G). However, none of the immunodominant epitopes stimulated the production of Th2 or Th17 cytokines (S1A–S1C Fig), confirming the primacy of the Th1 response.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of predicted ZIKV-derived CD4⁺ T cell epitopes.

https://doi.org/10.1371/journal.ppat.1007474.t001

We confirmed that the granzyme B⁺CD44⁺CD4⁺ T cells detected in vitro were bona fide cytolytic cells by performing an in vivo cytotoxicity assay. Naïve C57BL/6 splenocytes (CD45.1) were incubated with medium alone or pulsed with a mixture of the six immunodominant ZIKV epitopes and then labeled with a low (control target cells) or high (antigen-specific target cells) concentration of CSFE. The target cells were mixed and injected into mock- or ZIKV-infected mice, and splenocytes were harvested after 12 h and examined by flow cytometry to quantify low- and high-CSFE cells. As shown in Fig 1H, we detected approximately 70% specific killing of ZIKV epitope-pulsed target cells by both MR766- and FSS13025-primed mice. Taken together, these results indicate that Asian and African ZIKV strains both induce a robust expansion of cytokine-secreting and cytotoxic Th1 effector CD4⁺ T cells in LysMCre⁺Ifnar1^fl/fl mice.

Expansion of follicular helper CD4⁺ T cells and reduction of regulatory CD4⁺ T cells during primary ZIKV infection

To further dissect the CD4⁺ T cell response to ZIKV, we investigated two subsets that regulate Ab and cytolytic responses. T_FH cells are the major cell subset that provides help to B cells and promotes antiviral Ab production [33], while Treg cells play crucial roles in limiting the immune response during infection to avoid extensive tissue damage [34]. We analyzed the CD44⁺CD4⁺ T cells for expression of the T_FH surface markers CXCR5 and PD-1 or the Treg markers CD25 and FoxP3 (gating strategy shown in S2A and S2B Fig). We observed a significant expansion of T_FH cells that peaked on day 7 post-infection (~5-fold increase) and remained elevated thereafter for the duration of the experiment (Fig 2A and 2B and S2A & S2C Fig). In contrast, we detected a significant (~2-fold) reduction in the frequency of CD44⁺CD4⁺ CD25⁺FoxP3⁺ Treg cells in day 7 splenocytes from ZIKV-infected compared with mock-infected mice (Fig 2C and S2B & S2D Fig). Interestingly, this reduction was preceded by a marked and transient increase in the frequency of Treg cells early in the response (day 3; Fig 2D). This decrease in splenic Treg cells is consistent with the findings of a recent study of ZIKV infection in wildtype mice [30] and with our earlier study in Ifnar1-deficient mice infected with DENV [28].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Kinetics of the follicular helper and regulatory CD4⁺ T cell responses in the LysMCre⁺Ifnar1^fl/fl mouse model of primary ZIKV infection.

Five-week-old LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were infected retro-orbitally with 10⁴ FFU of ZIKV strain FSS13025 or were mock-infected. (A and C) Splenocytes were removed on day 7 post-infection and the percentage (A) CXCR5⁺PD1⁺CD44⁺CD4⁺ T_FH cells or (C) CD25⁺FoxP3⁺CD44⁺CD4⁺ Treg cells was assessed by flow cytometry. (B and D) Splenocytes were analyzed on the indicated days post-infection for the percentage (B) CXCR5⁺PD1⁺CD44⁺CD4⁺ T_FH cells or (D) CD25⁺FoxP3⁺CD44⁺CD4⁺ Treg cells. (E) Splenocytes were removed on day 7 post-infection and analyzed for the frequency of IL-10-producing CD44⁺CD4⁺ cells. Data are the mean ± SEM of n = 7 (A), n = 4 (B), n = 5 (C), n = 4 (D), n = 5 (E) mock-infected and n = 9 (A), n = 4–6 (B), n = 8 (C), n = 4–6 (D), n = 6 (E) ZIKV-infected mice. **P < 0.01, ***P < 0.001, by the Mann–Whitney U test. For the kinetic analysis, each time point was compared to day 0 using the Mann–Whitney U test, *P < 0.05, ***P < 0.001. Data are representative of two independent experiments.

https://doi.org/10.1371/journal.ppat.1007474.g002

IL-10 is a key CD4⁺ T cell-secreted immunoregulatory cytokine that controls viral immunity by inhibiting proinflammatory responses and preventing tissue damage (Reviewed in [35, 36]; therefore, we also examined their frequency in ZIKV-infected mice. Unexpectedly, we saw a significant expansion of IL-10-producing CD44⁺CD4⁺ splenocytes from day 7 ZIKV-infected compared with mock-infected mice (Fig 2E), which contrasts with the pattern of Treg cell retraction on day 7. The expression of IFNγ by IL-10⁺CD44⁺CD4⁺ T cells indicated that a substantial proportion of the IL-10-producing cells were Th1 effector cells with a regulatory phenotype, rather than Treg cells (S2E Fig). Collectively, these results demonstrate that, at the peak of the T cell response, ZIKV infection induces expansion of Ab-promoting T_FH cells and IL-10-producing CD4⁺ T cells, but suppresses the population of Treg cells.

Requirement for CD4⁺ T cells for induction of virus-specific IgG, but not for the CD8⁺ T cell response or viral control, after intravenous infection with ZIKV

T_FH cells play important roles in generating mature B cells and supporting long-lived Ab-producing plasma cells [33]. Since ZIKV infection leads to a rapid expansion in the frequency of T_FH cells, we next investigated the requirement for CD4⁺ T cells in generating ZIKV-specific Abs during primary infection. LysMCre⁺Ifnar1^fl/fl mice were treated with isotype control or anti-CD4 Ab, and serum samples were taken on day 7 post-infection for analysis of IgM and IgG reactivity to ZIKV E protein by ELISA. As an estimate of anti-ZIKV E Ab concentrations, we determined the endpoint titer (defined as the reciprocal of the highest serum dilution to give a reading twice the cutoff absorbance). Although the ZIKV E-specific IgM titer was the same in sera from control Ab- and anti-CD4-treated mice, depletion of CD4⁺ T cells significantly reduced the ZIKV E-specific IgG endpoint titer compared with the control sera (Fig 3A and 3B), suggesting that CD4⁺ T cell help is required for the development of the IgG, but not IgM, response to ZIKV. Surprisingly, we also found that CD4⁺ T cells were not required for generation of ZIKV-specific neutralizing Abs (Fig 3C). Chemical inactivation of IgM reduced the neutralizing capacity of serum samples from anti-CD4-treated mice, indicating that this isotype was largely responsible for the neutralizing activity at day 7 after infection (Fig 3C). To verify these findings, we also analyzed the Ab response at 10 days post-infection (S3A and S3B Fig). Although we again observed that CD4⁺ T cells were not required to generate ZIKV-specific IgM Abs, there was a marked reduction in neutralizing activity in the sera from anti-CD4 Ab-treated compared with control mice, consistent with production of IgG neutralizing Abs at this later time point during the primary infection (S3C Fig). These data therefore show that CD4⁺ T cells are not required for generation of the neutralizing Ab response at the peak of the T cell response at day 7 post-infection; however, they do not exclude the possible involvement of CD4⁺ T cells at later times. Indeed, analysis of splenocytes on day 7 post-infection revealed a significant reduction in the frequencies of plasma and GC B cells in CD4⁺ T cell-depleted mice compared with control mice (Fig 3D and 3E), supporting the potential role of CD4⁺ T cells in virus-specific IgG production at later time points after infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Contribution of CD4⁺ T cells to Ab and CD8⁺ T cell responses and to viral control during primary ZIKV infection in LysMCre⁺Ifnar1^fl/fl mice. LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were treated with a depleting anti-CD4 Ab (n = 8) or isotype control Ab (n = 8) on days −3 and −1 prior to and every 2 days after retro-orbital infection with 10⁵ FFU of ZIKV FSS13025.

(A–C) Sera were collected on day 7 post-infection to measure anti-ZIKV IgM (A) and IgG (B) titers by ZIKV E-specific ELISA and (C) ZIKV neutralizing activity using a U937 DC-SIGN cell-based assay with sera from both groups and sera from the anti-CD4 depleted mice group after inactivation of IgM. Data are the mean ± SEM. (D and E) Splenocytes were collected on day 7 post-infection and analyzed by flow cytometry for the percentage CD138⁺IgD⁻ plasma cells (D) or GL7⁺Fas⁺ germinal center B cells (E). Data are the mean ± SEM of n = 6 (D) or n = 3 (E) isotype control mice and n = 7 (D) or n = 3 (E) anti-CD4-treated mice. (F) Splenocytes were collected on day 7 post-infection, stimulated with the class I-binding ZIKV peptides PrM_169-177, E_297-305, or NS5_2783-2792 and analyzed by flow cytometry for the percentage IFNγ-producing (left) or IFNγ + TNF-producing (right) CD8⁺ T cells. Data are the mean ± SEM of n = 8 for both isotype control and anti-CD4-treated mice. (G) Serum, brain, and testes were harvested on day 7 post-infection from mice treated with isotype control or depleting anti-CD4 Ab and inoculated with 10⁵ FFU of ZIKV FSS13025, and infectious ZIKV titers were determined using a focus-forming assay. Data are the mean ± SEM of n = 8 (serum and brain), n = 4 (testes) for both isotype control and anti-CD4-treated mice. (H) Serum, brain, and testes were harvested on day 7 post-infection from mice treated with isotype control or depleting anti-CD4 Ab and inoculated with 10³ FFU, and infectious ZIKV levels were measured using a focus-forming assay; n = 7 (serum, brain, testes) for both isotype control and anti-CD4-treated mice. *P < 0.05, **P < 0.01, ***P < 0.001 by the Mann–Whitney U test. Data were pooled from two and three independent experiments for high and low viral challenge dose challenge, respectively.

https://doi.org/10.1371/journal.ppat.1007474.g003

We next asked whether CD4⁺ T cells are required for the primary CD8⁺ T cell response to ZIKV infection, as we have previously demonstrated a critical role for CD8⁺ T cells in protecting against primary ZIKV infection [22]. Splenocytes from control and anti-CD4 Ab-treated mice were isolated on day 7 post-ZIKV infection, stimulated in vitro with three immunodominant class I-restricted CD8 epitopes (PrM_169-177, E_297-305, or NS5_2783-2792), and analyzed for the frequencies of CD8⁺ T cells producing IFNγ alone or and IFNγ and TNF. We observed no difference in the frequency of either CD8⁺ T cell subset between splenocytes from control and CD4⁺ T cell-depleted mice (Fig 3F) or in the absolute number of total CD8⁺ T cells (S3D Fig), effector memory (CD44^highCD62L^low) CD8⁺ T cells (S3E Fig), or IFNγ- or IFNγ⁺TNF⁺-producing CD8⁺ T cells (S3F & S3G Fig). Thus, the primary CD8⁺ T cell response to ZIKV does not require CD4⁺ T cell help.

To investigate the impact of CD4⁺ T cell depletion on viral clearance, LysMCre⁺Ifnar1^fl/fl mice were treated with control or anti-CD4 Ab prior to intravenous infection with 10⁵ or 10³ FFU of ZIKV FSS13025. Seven days later, infectious ZIKV particles in the serum, brain, and testes were quantified using a cell-based focus-forming assay. Although the serum and brain were devoid of infectious particles in the majority of animals at this time point, ZIKV particles were detectable in the testes at levels not significantly different between the CD4⁺ T cell-sufficient and -depleted animals (Fig 3G and 3H).

To better mimic viral transmission via a mosquito bite in this animal model, we also evaluated the role of CD4⁺ T cells in the anti-ZIKV response of mice infected via the intrafootpad route. Here, too, we found as shown for the intravenous route that depletion of CD4⁺ T cells reduced the magnitude of the ZIKV-specific IgG, plasma cell, and GC B cell responses (S4A–S4E Fig) and there were no difference between control and CD4⁺ T cell-depleted mice in the CD8⁺ T cell response (S4F and S4G Fig) or in viral titers in serum, brain, and testes (S4H Fig). Overall, these results suggest that, although CD4⁺ T cells are required for the differentiation of plasma and GC B cells and for the production of virus-specific IgG, they are not involved in viral clearance during primary ZIKV infection via either the intravenous (RO) or intrafootpad routes, which contrasts with the role of CD8⁺ T cells [22].

Contribution of memory CD4⁺ T cells to viral clearance in mice after intravenous ZIKV infection

Since our data indicate that CD4⁺ T cells do not contribute to ZIKV clearance early (day 7) after primary infection via the RO route, we asked whether memory CD4⁺ T cells, which become detectable at days 21–28 after primary infection, might be more effective. To this end, we infected mice RO with ZIKV FSS13025 and isolated splenocytes after 34 days. CD4⁺ T cells were purified and injected (10⁷ or 1.5 × 10⁷) RO into naïve LysMCre⁺Ifnar1^fl/fl mice, which were then inoculated RO with high or low doses of ZIKV FSS13025 (10⁵ or 10³ FFU, respectively). Although the transferred CD4⁺ memory T cells had no effect on infectious viral burden in the serum, brain, or testes on day 3 post-infection in the high dose-infected mice (Fig 4A), we observed a significant reduction in viral burden in some tissues (e.g., spleen, sciatic nerve, and female reproductive tract [FRT]) of mice injected with memory CD4⁺ T cells and inoculated with a low dose of ZIKV (Fig 4B). Brain and serum did not show a significant reduction in viral particles and are not represented. These data suggest that memory CD4⁺ T cells can contribute to ZIKV clearance from some, but not all, tissues during infection with low doses of ZIKV.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Contribution of memory CD4⁺ T cells to ZIKV clearance in LysMCre⁺Ifnar1^fl/fl mice.

LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were retro-orbitally infected with 10⁴ FFU of ZIKV strain FSS13025. After 30 days, CD4⁺ T cells were purified from the spleens and 10⁷ or 1.5 × 10⁷ cells were transferred into naïve LysMCre⁺Ifnar1^fl/fl C57BL/6 mice. One day later, mice were infected retro-orbitally with (A) 10⁵ FFU (n = 9 for serum and brain and n = 4 for testes) or (B) 10³ FFU of ZIKV strain FSS13025 (n = 4 for mice receiving CD4⁺ T cells and n = 5 for control mice receiving no T cells). The indicated organs were isolated 3 days later and infectious ZIKV titers were determined using a focus-forming assay. (C) LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were immunized subcutaneously with a mixture of six immunodominant peptides (n = 5) or DMSO (n = 5) on days −28 and −14, and infected with 10⁵ FFU of ZIKV FSS13025 on day 0. Three days later, the brain and testes were removed and infectious ZIKV titers were determined using a focus-forming assay. (D) LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were infected with 10⁴ FFU of ZIKV FSS13025 on day 0 and treated with CD4-depleting antibody or isotype control Ab on day-3 and day-1 prior to challenge with 10³ FFU of ZIKV FSS13025 on day 28 post-priming. Three days later, infectious ZIKV titers in serum, brain, and testes were determined using a focus-forming assay. Data are the mean ± SEM. *P < 0.05, by the Mann–Whitney U test.

https://doi.org/10.1371/journal.ppat.1007474.g004

To explore the protective role of the anti-ZIKV CD4⁺ T cell response, we next employed a variety of additional approaches. First, LysMCre⁺Ifnar1^fl/fl mice were immunized with the six immunodominant Th1 ZIKV epitopes on days −28 and −14 (or mock-immunized with adjuvant alone), followed by inoculation with 10⁵ FFU of ZIKV FSS13025 on day 0. Three days later, viral titers in the brain and testes were measured. Notably, immunization with the CD4⁺ T cell epitopes resulted in a significant reduction in viral burden in both organs of ZIKV epitope-immunized mice compared with the mock-immunized mice (Fig 4C), confirming that memory CD4⁺ T cells could contribute to viral clearance.

Second, we asked whether memory CD4⁺ T cells contribute to viral clearance during secondary ZIKV infection. LysMCre⁺Ifnar1^fl/fl mice were mock- or ZIKV-inoculated RO on day 0, and then treated with a control or CD4-depleting Ab on day-3 and day-1 prior to challenge with a non-lethal dose of ZIKV on day 28 post-priming. Measurement of viral particles in the serum, brain, and testes revealed almost complete eradication of virus in ZIKV-primed mice compared with mock-primed mice on day 3 after secondary ZIKV infection, with no apparent difference between anti-CD4-treated and control Ab-treated mice (Fig 4D). The anti-CD4-treated mice showed reduced frequencies of plasma cells and GC B cells compared with control mice (S5A and S5B Fig), consistent with the role of CD4⁺ T cells in production of memory anti-ZIKV Ab responses. In contrast, the number of memory CD8⁺ T cells producing IFNγ⁺TNF⁺ was higher in CD4-depleted compared with control mice (S5C and S5D Fig). Analysis of CD4⁺ T cell subsets revealed an increase in T_FH cells and IFNγ-, IFNγ⁺ TNF⁺-, and IL-2-producing CD4⁺ T cells and a decrease in Treg cells in Mock-immune compared with ZIKV-immune mice (S5E–S5G Fig), which is consistent with the observation that Treg cell number peaks at day 3 after primary infection (Fig 2D). These data thus reveal the presence of a memory CD4⁺ T cell response in ZIKV-immune mice, but indicate that this memory CD4⁺ T cell response is not required to control viral clearance early after a secondary ZIKV challenge.

Third, we examined whether CD4⁺ T cells could contribute to protection against ZIKV infection in fully deficient Ifnar1^−/− mice, which are more susceptible than LysMCre⁺Ifnar1^fl/fl mice to ZIKV infection. Ifnar1^−/− mice were immunized with a mixture of the ZIKV peptides, boosted two weeks post-immunization, and then challenged 14 days later with a lethal dose of ZIKV. Body weight, clinical signs of disease, and mortality were monitored daily for 15 days post-challenge. Although fewer peptide-immunized mice succumbed to ZIKV infection compared with mock-immunized mice (~50% and 75%, respectively), both groups of mice showed similar clinical disease scores and no significance difference was observed between the two groups (S6A–S6D Fig). To assess a possible role for CD4⁺ T cells in protection, Ifnar1^−/− mice were injected with a CD4⁺ T cell-depleting Ab or isotype control Ab on days −3 and −1 prior to lethal dose of ZIKV. We daily monitored the infected mice and detected no significant difference in weight loss, clinical disease score, or mortality between mice treated with anti-CD4 Ab or isotype control Ab (S7A–S7D Fig). Both groups succumbed to ZIKV infection after nine days of infection (S7A–S7D Fig). Similar results were obtained even upon secondary challenge of ZIKV-primed Ifnar1^−/− mice (S7E and S7F Fig). Specifically, Ifnar1^−/− mice were injected with a CD4⁺ T cell-depleting Ab or isotype control Ab on days −3 and −1 prior to priming with a sub-lethal dose of ZIKV, and then every week thereafter for 4 weeks. On day 30, the mice were challenged with a lethal dose of ZIKV. We detected no significant difference in weight loss, clinical disease score, or mortality between mice primed with ZIKV and treated with anti-CD4 Ab or isotype control Ab for 30 days prior to secondary ZIKV infection (S7E and S7F Fig). Thus, CD4⁺ T cells do not appear to play an essential role in protecting against lethal primary or secondary infection with ZIKV.

Taken together, the results of the adoptive cell transfer and peptide immunization experiments using LysMCre⁺Ifnar1^fl/fl mice indicate that CD4⁺ T cells do play a role in controlling viral burden in ZIKV-infected mice under certain conditions, such as low-dose secondary infection and peptide vaccination. Additionally, experiments using LysMCre⁺Ifnar1^fl/fl and Ifnar1^−/− mice show that CD4⁺ T cells also play a role in Ab production during secondary infection but are not required to protect against systemic lethal challenge.

Vigorous CD4⁺ T cell response after primary intravaginal ZIKV infection in LysMCre⁺Ifnar1^fl/fl mice

Given the increasing evidence that ZIKV can be transmitted via sexual contact [1], we were also interested in determining whether a protective CD4⁺ T cell response is activated via this route of infection. We previously described a model of sexual transmission of ZIKV using LysMCre⁺Ifnar1^fl/fl mice [5]. In this model, hormonal changes were found to influence ZIKV transmission, since mice were more susceptible to intravaginal (IVag) ZIKV infection and the virus persisted in the female reproductive tract (FRT) for longer periods during the diestrus-like phase relative to estrus-like phase [5]. Therefore, for the experiments performed here, we induced a diestrus-like phase by treating LysMCre⁺Ifnar1^fl/fl mice with progesterone at day −3 prior to intravaginal infection with a high dose (10⁵ FFU) of ZIKV FSS13025. Mice were confirmed to be in a diestrus-like phase by vaginal cytology prior to inoculation [37]. Splenocytes were prepared on days 3, 7, 10, 15, and 21 after infection, stimulated in vitro with the immunodominant epitope E_644-658, and analyzed by flow cytometry. As shown in Fig 5, an increase in the frequency of IFNγ- and IFNγ + TNF-producing CD44⁺CD4⁺ T cells was detectable on day 6, peaked at day 10, and waned thereafter. In contrast, IL-2-producing CD44⁺CD4⁺ T cells were already detectable on day 3 but were most abundant on day 15 post-infection (Fig 5A–5C). To investigate the localized immune response to IVag ZIKV infection, we analyzed CD4⁺ T cell activation in the local (iliac) draining lymph nodes (LNs) on day 10 after infection. Despite the much lower absolute number of cells compared with the spleen, we detected a similar percentage of IFNγ-, IFNγ + TNF-, and IL-2-expressing CD44⁺CD4⁺ T cells in the iliac LNs and the spleen (Fig 5D). Notably, we also observed the same pattern of expansion of CD4⁺ T cell subsets in the spleen on day 10 after IVag ZIKV infection as we had on day 7 after RO infection. Thus, the frequencies of activated (CD44⁺), antigen-experienced (CD49d⁺CD11a⁺), cytolytic (granzyme B⁺), and T_FH CD4⁺ T cells in the spleen were all increased (Fig 5E–5H), whereas the Treg cell frequency was reduced (Fig 5I). Moreover, we also observed an increase in IL-10-producing CD4⁺ T cells (Fig 5J). CD4⁺ T cells in the iliac LNs of intravaginally infected mice showed an identical pattern of expansion of T_FH and IL-10-producing CD4⁺ T cells and reduction in Treg cells on day 10 post-infection (S8A–S8C Fig). Thus, IVag ZIKV infection induced a vigorous Th1 CD4⁺ T cell response in the spleen and draining LNs comparable to that observed in the spleen after RO infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Characterization of the CD4⁺ T cell response after primary intravaginal ZIKV infection in LysMCre⁺Ifnar1^fl/fl mice.

LysMCre⁺Ifnar1^fl/fl C57BL/6 female mice (8- to 9-week-old) were treated with progesterone and infected via IVag route with 10⁵ FFU of ZIKV FSS13025 3 days later. (A–C) On the indicated days, splenocytes (n = 4–8 mice) were stimulated in vitro with E_644-658 FSS13025 peptide and analyzed by flow cytometry for the frequency of CD44⁺CD4⁺ T cells producing (A) IFNγ, (B) IFNγ + TNF, and (C) IL-2. (D) As described for (A–C) except cells were harvested from the spleen and iliac lymph nodes (n = 8) on day 10 post-infection, stimulated in vitro, and analyzed for the frequency of CD44⁺CD4⁺ T cells producing IFNγ (white bars), IFNγ + TNF (black bars), or IL-2 (gray bars). (E–J) On day 10 post-infection, splenocytes were analyzed for the frequency of (E) CD44⁺CD4⁺ T cells, (F) CD49d⁺CD11a⁺ cells, and (G) granzyme B⁺ cells, (H) T_FH (CXCR5⁺PD1⁺CD44⁺CD4⁺) cells, (I) Treg cells (FoxP3⁺CD25⁺CD4⁺CD44⁺) cells, and (J) IL-10-producing CD44⁺CD4⁺ T cells. All experiments were performed twice. Data are the mean ± SEM of n = 3 (E), n = 4 (F), n = 4 (G), n = 4 (H), n = 4 (I), n = 4 (J) mock-infected and n = 4 (E), n = 7 (F), n = 4 (G), n = 7 (H), n = 5 (I), n = 6 (J) ZIKV-infected mice. *P < 0.05, **P < 0.01 by the Mann–Whitney U test.

https://doi.org/10.1371/journal.ppat.1007474.g005

Since the T cell response peaked later after IVag ZIKV infection than after RO infection, we examined the frequency of CD8⁺ effector memory (CD44^highCD62L⁻) and CD8⁺ central memory (CD44^highCD62L⁺) T cell subsets on day 10 after IVag infection with ZIKV FSS13025. We found a significant expansion (~2.5-fold) of effector memory CD8⁺ T cells and antigen-experienced (CD11a^high) CD8⁺ T cells (~6-fold), and a concomitant reduction (~2.5-fold) in central memory CD8⁺ T cells in the ZIKV-infected compared with mock-infected mice on day 10 post-infection (S9A & S9B Fig). In addition, splenic CD8⁺ T cells producing IFNγ and IFNγ + TNF were present after in vitro stimulation with the class I-restricted ZIKV E_297-305 epitope (S9C Fig). Taken together, these results indicate that IVag infection with ZIKV promotes vigorous systemic CD4⁺ and CD8⁺ T cell responses.

CD4⁺ T cell-mediated regulation of the antiviral Ab response, control of local viral burden, and protection from lethality following intravaginal infection with ZIKV

Our results thus far reveal a mixed role for CD4⁺ T cells in promoting Ab and viral clearance to ZIKV infection via the RO route. To determine whether this was also the case for the primary response to IVag infection, we performed a similar analysis using progesterone-pretreated LysMCre⁺Ifnar1^fl/fl mice that were infused with a control or depleting anti-CD4 Ab before IVag ZIKV infection. Analysis of serum on day 10 post-infection revealed that, in contrast to the findings with mice infected RO, depletion of CD4⁺ T cells significantly reduced the titers of ZIKV-specific IgM and IgG as well as the anti-ZIKV neutralizing Ab activity compared with control mice (Fig 6A–6C). We observed a reduction in the frequency of splenic plasma cells and GC B cells (Fig 6D and 6E) in CD4-depleted mice but saw no effect on the CD8⁺ T cell response, as reflected by the frequency of IFNγ- and IFNγ+ TNF-producing cells after in vitro stimulation of splenocytes with E_297-305 ZIKV epitope (Fig 6F), on day 7 post-infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Contribution of CD4⁺ T cells to Ab production, CD8⁺ T cell response, and local viral control during primary intravaginal ZIKV infection of LysMCre⁺Ifnar1^fl/fl mice.

Female LysMCre⁺Ifnar1^fl/fl C57BL/6 mice were treated with a depleting anti-CD4 Ab (n = 7) or isotype control Ab (n = 6) on days −3 and −1 prior to intravaginal infection with 10⁵ FFU of ZIKV FSS13025. Mice were also treated with progesterone on day −3. (A–C) Sera were collected on day 10 post-infection to measure anti-ZIKV IgM (A) and IgG (B) titers by ZIKV E-specific ELISA and neutralizing activity (C) using a U937 DC-SIGN cell-based flow cytometric assay. (D and E) Splenocytes were collected on day 10 post-infection and analyzed by flow cytometry for the percentage plasma cells (CD138⁺IgD⁻) (D) or germinal center B cells (GL7⁺Fas⁺) (E). Data are the mean± SEM of n = 7 (D) or 4 (E) for isotype control mice and n = 6 (D) or 3 (E) for anti-CD4-treated mice. (F) Splenocytes were collected on day 10 post-infection, stimulated with the immunodominant CD8⁺ T cell ZIKV epitope E_297-305, and analyzed by flow cytometry for the percentage IFNγ- or IFNγ + TNF-producing CD8⁺ T cells. Data are the mean ± SEM of n = 8 mice/group. (G) Infectious ZIKV particles were measured in serum (n = 7, n = 9), vaginal washes (n = 9, n = 9), and the female reproductive tract (FRT) (n = 4, n = 6) on day 10 post-infection for the isotype control or anti-CD4 treated groups. (H) As described for (G) except mice were treated with an anti-CD8 Ab (n = 6) or isotype control Ab (n = 8) before infection, only serum and vaginal washes are represented. *P < 0.05, **P < 0.01 by the Mann–Whitney U test.

https://doi.org/10.1371/journal.ppat.1007474.g006

To assess the requirement for CD4⁺ T cells for viral clearance after IVag infection, we quantified infectious ZIKV particles in serum, vaginal washes, and the FRT on day 10 post-IVag infection of control or anti-CD4 Ab-treated LysMCre⁺Ifnar1^fl/fl mice. Although the viral burden in serum and FRT was unaffected by CD4⁺ T cell depletion, ZIKV RNA levels were significantly higher in the vaginal washes of anti-CD4-treated compared with isotype control Ab-treated mice (Fig 6G). To determine whether viral clearance from vaginal washes was mediated by CD4⁺ T cells alone, CD8⁺ T cells were depleted by injection of anti-CD8 Ab on days −3 and −1 before IVag infection with ZIKV, and serum and vaginal washes were analyzed for ZIKV RNA levels at 10 days post-infection. We detected no difference in viral RNA levels between anti-CD8 Ab and isotype control Ab-treated mice in either sample (Fig 6H), indicating that CD8⁺ T cells are likely not involved in control of viral burden following IVag infection. However, animals lacking CD8⁺ T cells showed an enhanced CD4⁺ T cell response, with increased IFNγ- and IFNγ+ TNF-producing cell frequencies and numbers in the spleen (S10A & S10B Fig) and increased frequencies in the iliac lymph nodes (S10C Fig), suggesting that the CD4⁺ T cell response to ZIKV compensates for the absence of CD8⁺ T cells by increasing the abundance of Th1 cells. To confirm and extend this result showing CD4⁺ T cell contribution to protection against IVag ZIKV infection in LysMCre⁺Ifnar1^fl/fl mice, we depleted CD4⁺ T cells in Ifnar1^−/− mice on days −3 and −1 before IVag infection with a lethal dose of ZIKV. Only 22% of CD4-depleted mice survived to day 15 post-infection compared with 78% of control Ab-treated mice (S11A Fig), and the CD4⁺ T cell-depleted animals showed greater body weight loss and more severe clinical disease scores than the control animals (S11B–S11D Fig). Collectively, the results of the experiments with LysMCre⁺Ifnar1^fl/fl and Ifnar1^−/− mice infected via the IVag route indicate that CD4⁺ T cells are required to mount an efficient Ab response, to control the local viral burden, and to reduce clinical signs and mortality following IVag infection.

Ethics statement

All experiments were performed in strict accordance with recommendations set forth in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee at the La Jolla Institute for Allergy and Immunology (protocol number APO28-SS1-0615 and AP00001029). Both ZIKV strains (MR766 and FSS13025) were obtained from the World Reference Center for Emerging Viruses and Arboviruses with Institutional Review Board approval. All samples were anonymized.

Mouse experiments

LysMCre⁺Ifnar1^fl/fl and Ifnar1^−/− mice were bred under pathogen-free conditions at La Jolla Institute for Allergy & Immunology. Sample sizes were based on similar studies [5, 22]. None of the experiments were randomized or blinded.

Virus culture and titration

Asian lineage strain FSS13025 was isolated from a 3-year-old boy in Cambodia in 2010 [52], and African lineage strain MR766 was isolated from a sentinel rhesus monkey in Uganda in 1947 [53]. Both viruses were cultured in C6/36 Aedes albopictus mosquito cells (American Type Culture Collection [ATCC]; Manassas, Virginia). Viral supernatants were collected 7–10 days after infection, clarified by centrifugation, and concentrated by ultracentrifugation. Viral titers were measured using a baby hamster kidney (BHK)-21 (ATCC) cell-based focus-forming assay (FFA). In brief, BHK-21 cells were plated at 2 × 10⁵ cells per well in a 24-well plate and incubated overnight in complete MEM-α medium (containing 10% fetal bovine serum [FBS], 1% penicillin/ streptomycin, and 1% HEPES) at 37°C in a 5% CO₂ atmosphere. Cells were infected with serial dilutions of virus for 1.5 h with gentle shaking every 15 min. The medium was then aspirated and replaced with fresh complete MEM-α medium supplemented with 1% carboxymethyl cellulose (Sigma), and the cells were cultured for 2 days. Cells were then fixed with 4% formalin (Fisher), permeabilized with 1% Triton X-100 (Sigma) and blocked by addition of 10% FBS in phosphate-buffered saline (PBS). ZIKV was detected by incubation of cells with 4G2, a pan-flavivirus E protein-specific monoclonal Ab (BioXcell) for 1.5 h. Cells were washed and incubated for 1.5 h with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (Sigma). Finally, foci were detected by incubation with True Blue substrate (KPL) and counted manually. Viral titers were expressed as log FFU/g tissue or log FFU/ml serum. Next-generation sequencing of viral stocks confirmed the absence of competing pathogens.

Peptide prediction and synthesis

ZIKV MR766 and FSS13025 sequences were obtained from the NCBI protein database. MHC class II peptide binding affinity predictions were performed using the Immune Epitope Database (www.iedb.org) website tools using the “IEDB-recommended” method selection, as previously described [54]. Predicted binding affinities were obtained for all non-redundant 15-mer peptides that bound H-2 I-A^b, and the peptide list was sorted by increasing consensus percentile rank and restricted to the top 1%. Peptides were synthesized by Synthetic Biomolecules as crude material (1 mg scale) and validated by mass spectrometry. Peptides for in vitro stimulation followed by flow cytometric analyses were synthesized and purified by reverse-phase high performance liquid chromatography to ≥95% purity. Peptides were dissolved in DMSO for use.

Mouse infection

LysMCre⁺Ifnar1^fl/fl and Ifnar1^−/− mice (5- to 6-week-old males and females for RO and intrafootpad infection; 8- to 9-week-old females for IVag infection) were infected with ZIKV FSS13025 or MR766 at 10¹, 10², 10³, 10⁴, or 10⁵ FFU RO, 10⁵ or 10⁶ FFU IVag, and 10⁵ FFU via intrafootpad route in 10% FBS/PBS. Three days prior to IVag infection, mice were injected subcutaneously with 2 mg of progesterone (Millipore Sigma) in 100 μl of 5% ethanol, 5% Kolliphor, and 90% H₂O to induce a diestrus-like phase, which was confirmed as previously described [5].

CD8⁺ and CD4⁺ T cell depletion

T cell depletion Abs (CD8⁺ T cell-depleting clone 2.43 or CD4⁺ T cell-depleting clone GK1.5) were purchased from BioXCell and administered on days −3 and −1 pre-infection and every 2 days post-infection for the duration of the experiment. For long-term (30 day) depletion, mice were treated on days −3 and −1 and weekly thereafter until the end of the experiment. A rat IgG2 isotype control Ab (clone LTF-2) served as the control and was administered on days −3 and −1 pre-infection only.

Adoptive T cell transfer

CD4⁺ T cells were isolated from donor mice 34 days after infection with 10⁴ FFU of ZIKV FSS13025. Spleens were harvested, and CD4⁺ T cells were positively selected using a CD4 T Cell Isolation Kit (Miltenyi Biotec). A total of 1 × 10⁷ or 1.5 × 10⁷ purified CD4⁺ T cells were then injected RO into the recipient mice 1 day prior to ZIKV infection. CD4⁺ T cells purified from naïve LysMCre⁺Ifnar1^fl/fl mouse spleens were harvested and injected into recipient mice as controls.

Tissue collection

For preparation of serum samples, mice were sacrificed by CO₂ inhalation and blood was collected by cardiac puncture. For vaginal washes, the vaginal canal was rinsed 3–5 times with 40 μl PBS and the washes were combined. Mice were then perfused with PBS and the desired organs were collected. For tissue FFA, organs were transferred to pre-weighed tubes containing complete MEM-α medium and a metal bead, and the tubes were then stored at −80°C until analyzed. For qRT-PCR, organs were stored in RNAlater (Invitrogen) at 4°C until analyzed.

Quantification of virus in tissues

For the FFA, pre-weighed tubes containing frozen organs were thawed, homogenized (Tissuelyser II; Qiagen), and centrifuged. The supernatant was serially diluted, added to BHK-21 cells, and the titers were determined as described above. Titers were expressed as the log FFU/g of tissue. For qRT-PCR, RNA was isolated from tissues using an RNeasy Mini Kit (Qiagen) or from serum or vaginal washes using a Viral RNA Isolation Kit (Qiagen). qRT-PCR was performed using a qScript One-Step qRT-PCR Kit (Quanta, Bioscience) with a CFX96 Touch™ real-time PCR detection system (Bio-Rad CFX Manager 3.1). ZIKV-specific primers have been previously described [55]. Cycling conditions were: 45°C for 15 min, 95°C for 15 min, followed by 50 cycles of 95°C for 15 s and 60°C for 15 s, and a final extension of 72°C for 30 min. Viral RNA concentration was calculated using a standard curve composed of five 100-fold serial dilutions of in vitro-transcribed RNA from ZIKV strain FSS13025.

ZIKV-binding IgG/IgM ELISA

ELISA plates (96-well, Costar) were coated with ZIKV E protein (1 μg/ml, ZIKVSU-ENV, Native Antigen) in coating buffer (0.1 M NaHCO₃) overnight at 4°C and then blocked for 1 h at room temperature (RT) with 5% Blocker Casein in PBS (Thermo Fisher Scientific). Mouse serum samples were diluted three-fold (from 1:30 to 1:65,610) in 1% bovine serum albumin (BSA)/PBS, added to the coated wells, and incubated for 1.5 h at RT. Wells were then washed with wash buffer (0.05% Tween 20 [Promega] in PBS), and HRP-conjugated goat anti-mouse IgG Fc or goat anti-mouse IgM (1:5000 in 1% BSA/PBS) was added to each well for 1.5 h at RT. TMB chromogen solution (eBioscience) was added to the wells, the reaction was stopped by addition of sulfuric acid, and the absorbance at 450 nm was read on a Spectramax M2E microplate reader (Molecular Devices). The ZIKV-specific Ab endpoint titers were calculated as the reciprocal of the highest serum dilution that gave a reading twice the cutoff absorbance based on the negative control (BSA/PBS).

Serum neutralization assay

Sera from naïve and ZIKV-immune mice were inactivated by incubation for 30 min at 56°C and then serially diluted and added to 96-well round-bottom plates. Some sera were incubated for 30 min with dithiothreitol (0.01 M, Sigma) prior to infection to inactivate IgM. A sufficient amount of ZIKV FSS13025 causing 7–15% of infection in U937 DC-SIGN, was added to the sera and incubated for 1 h at 37°C (titration of virus is determined for each batch of cells). U937 cells expressing dendritic cell-specific intracellular adhesion molecule-3-grabbing non-integrin (U937 DC-SIGN, ATCC) were then added at 1 x 10⁵ cells/well and the plates were incubated for 2 h at 37°C with rocking every 15 min. Cells and ZIKV FSS13025 incubated in the absence of serum served as the positive control. Plates were then centrifuged, the supernatants were aspirated, fresh RPMI medium supplemented with 10% FBS was added, and the cells were incubated for 16 h at 37°C. Finally, cells were harvested, stained with PE-conjugated anti-CD209 (DC-SIGN, clone DNC246), incubated with Cytofix/Cytoperm solution (BD Biosciences), and stained intracellularly with Alexa Fluor 647-conjugated 4G2 (to ZIKV E protein). The cells were analyzed on an LSRII flow cytometer (BD Biosciences) and the percentage of infected cells was determined using FlowJo 10.4.2 software (Tree Star, Ashland, OR). The percentage serum inhibition was calculated as 100 − (% infected cells in the presence of serum) / (% infected cells in the absence of serum) x 100.

Intracellular cytokine staining (ICS) assay

Splenocytes were plated into 96-well round-bottom plates at 1 × 10⁶ cells/well in complete RPMI 1640 medium (containing 10% FBS, 1% penicillin/streptomycin, and 1% HEPES (all from Gibco) and stimulated with 1 μg of the indicated ZIKV peptides. After 1 h at 37°C, brefeldin A (1000X, BioLegend) was added at a 1:1000 dilution and the cells were incubated for an additional 5 h. Cells incubated with RPMI medium or with RPMI medium + PMA/ionomycin (500X) served as negative and positive controls, respectively. After incubation, the splenocytes were stained with efluor 455 (UV) viability dye, (Invitrogen) and fluorophore-conjugated Abs against mouse CD3e (clone 145-2C11, Tonbo), CD8α (clone 53–6.7, BioLegend), CD4 (clone GK1.5, Invitrogen), CD44 (clone IM7, BioLegend), CD62L (MFL-14, BioLegend) CD25 (clone PC61, BioLegend), Biotin-CD185 (clone SPRCL5, Invitrogen), CD279 (clone 29F.1A12, BioLegend), CD19 (clone ebio1D3, eBioscience), CD11a (clone M17/4, eBioscience), CD45.1 (clone A20, eBioscience), CD49d(clone R1-2, eBioscience), IgD (clone 11-26C.2a, BD Pharmingen), CD138 (clone 281.2, BioLegend) and streptavidin-conjugated BV421 (BD Pharmingen), all at 1:200 dilution. The cells were then fixed and permeabilized with Cytofix/Cytoperm (BD Bioscience) (or Fixation/Permeabilization Concentrate [eBioscience] for analysis of Treg cells) and stained with fluorophore-conjugated Abs against mouse IFNγ (clone XMG1.2, Tonbo), TNF (clone MP6-XT22, eBioscience), IL-10 (clone JES5-16E3, BioLegend), IL-2 (clone JES6-5H4, BioLegend), IL-17A (clone eBio17B7, eBioscience), IL-4 (clone 11B11, BioLegend), IL-5 (clone TRFK5, Invitrogen), granzyme B (clone NGZB, eBioscience), Bcl-6 (clone K112-91, BD Pharmingen), and/or FOXP3 (clone FJK-16S, eBioscience). Data were collected on an LSR II (BD Bioscience) and analyzed using FlowJo software.

Peptide immunization

Five-week-old LysMCre⁺Ifnar1^fl/fl and Ifnar1^-/- mice were injected subcutaneously with 100 μg each of six immunodominant peptides (E_346-360, E_644-658, NS3_1740-1754, NS4B_2480-2494, NS5_2604-2618, NS5_2738-2752) in complete Freund’s adjuvant. Two weeks later, the mice were boosted by injection of the same peptides in incomplete Freund’s adjuvant. Fourteen days later, the mice were infected retro-orbitally with 10⁵ FFU ZIKV. Three days post-infection, organs were collected and infectious ZIKV particles were quantified using FFA.

In vivo cytotoxicity assay

To prepare target cells, splenocytes were harvested from naïve donor mice (C57BL6, CD45.1) and incubated for 3 h at 37°C with a pool of H2-IA^b-restricted peptides (E_346-360, E_644-658, NS3_1740-1754, NS4B_2480-2494, NS5_2604-2618, NS5_2738-2752) or DMSO. Cells were then washed and labeled with CSFE (Invitrogen) in PBS/0.1% BSA for 10 min at 37°C. To distinguish between target cells, cells incubated with DMSO and ZIKV peptides were labeled with 100 nM CSFE (low) and 1 μM CSFE (high), respectively. Cells were then washed and 5 × 10⁶ each of CSFE-low and CSFE-high target cells were mixed and injected RO into recipient LysMCre⁺Ifnar1^fl/fl mice that had been infected retro-orbitally with 10⁴ FFU ZIKV FSS13025, 10⁴ FFU ZIKV MR766, or vehicle (10% FBS/PBS; mock-infected). Twelve hours later, splenocytes were harvested from the recipient mice and the number of CSFE-labeled cells was quantified by flow cytometry. The specific percentage killing was calculated as follows: 100 − ([percentage specific (CSFE-high) target cells present in infected mice]/[percentage non-specific (CSFE-low) target cells present in infected mice]/[percentage specific (CSFE-high) target cells present in mock-infected mice]/[percentage non-specific (CSFE-low) target cells present in mock-infected mice] x 100).

Clinical scoring of disease

Mice were weighed on the day of infection and then weighed and scored for disease daily post-infection. Clinical features were based on a 7-point scale: 1, healthy; 2, slightly ruffled coat around head and neck; 3, ruffled coat over the entire body; 4, severely ruffled coat and slightly closed eyes; 5, sick with closed eyes and slow movement (mice were euthanized); 6, no movement and slow breathing; 7 dead. Weight loss was calculated by comparison with the weight on the day of infection.

Statistical analyses

All data were analyzed with Prism software, version 7.0 (GraphPad Software). Results are expressed as the mean ± standard errors. A non-parametric Mann–Whitney test was used to compare differences between two groups, and the Wilcoxon test was used to compare two parameters from the same group. One-way or two-way ANOVA or a Kruskal–Wallis test was used to compare more than two groups. P < 0.05 was considered significant.