Kinetics of Effector Functions and Phenotype of Virus-Specific and γδ T Lymphocytes in Primary Human Cytomegalovirus Infection During Pregnancy
- First Online:
- Cite this article as:
- Fornara, C., Lilleri, D., Revello, M.G. et al. J Clin Immunol (2011) 31: 1054. doi:10.1007/s10875-011-9577-8
- 170 Views
The T-cell response to human cytomegalovirus (HCMV) primary infection was analyzed in 27 pregnant women during the first year after primary HCMV infection. Pregnant women with remote HCMV infection were enrolled as controls. Interferon-γ-producing T cells were readily detected at levels comparable (CD4+) or higher (CD8+) than controls, whereas the CD4+ and CD8+ lymphoproliferative response as well as IL-2 production was significantly reduced with respect to controls for at least 9 months after infection. In addition, CD45RA re-expression as well as cytotoxic T lymphocyte activity and perforin expression were the major components of the adaptive CD4+ and CD8+ T-cell immune response, while Vδ2− γδ T-cell expansion in response to HCMV infection followed kinetics similar to that of CD8+ T cells. Reduced CD45RA re-expression directly correlated with HCMV transmission to the fetus, thus providing an important prognostic parameter.
KeywordsT-cell immune responseγδ T cellsprimary HCMV infectioncytotoxicityperforin
Human cytomegalovirus (HCMV) is a β-herpesvirus that primarily causes mild or silent infection in the immunocompetent host. However, when primary HCMV infection occurs in pregnant women, HCMV is transmitted from mother to fetus in about 32% of cases (with 11.0% to 12.7% of congenitally infected newborns exhibiting symptomatic infection at birth), whereas this risk drops to 1.4% in case of reactivated infection during pregnancy . Conversely, in immunocompromised patients, HCMV infection can lead to severe disease and mortality [2, 3].
Several studies have documented the importance of both virus-specific T-cell-mediated and humoral immune responses for the control of HCMV infection [4–7]. In addition, some aspects of innate immunity during HCMV infection of immunocompromised patients have been evaluated, including natural killer cells and γδ T cells [8, 9]. While the role of antibodies and particularly of neutralizing antibodies as well as that of innate immunity in protection against HCMV infection remains to be more extensively investigated [10, 11], several studies have documented the role of T-cell immune responses and shown that women with preconceptional immunity to HCMV only rarely transmit HCMV to the fetus [12–14]. In this respect, recently, it has been shown that pregnant women with primary HCMV infection transmitting the virus to the fetus show a delayed T-cell lymphoproliferative response to HCMV, compared with nontransmitting women [15, 16]. Furthermore, it has been shown that circulating HCMV-specific effector memory T cells (TEM) may revert to the CD45RA+ phenotype, which appears to be associated with control of HCMV viremia and mother-to-fetus transmission .
Given the relatively poor availability of data relevant to the cellular immune response to HCMV infection in immunocompetent subjects, we decided to evaluate a number of parameters of αβ and γδ T-cell responses over the first year after onset of infection. The collection of multiple data sets should enable a comprehensive evaluation of the T-cell response to HCMV primary infection. Measures determined included lymphoproliferation, T-cell cytokine production, development of long-term (CD45RA+) effector/memory T cells, cytotoxicity and perforin expression, as well as Vδ2+ and Vδ2− γδ T-cell frequency. Results document that, while some parameters reached levels observed in HCMV-seropositive individuals with remote (>5 years) infection, other values were still far from those seen in pregnant control subjects. Thus, 1 year may not be long enough to allow development of a long-term immune response.
Patients and Methods
Twenty-seven pregnant women (median age 33, range 18–39 years) experiencing primary HCMV infection were enrolled in the study. On the whole, a total of 98 heparinized blood samples (2–6 per subject) were collected 11–594 days after the onset of infection and tested, whenever a sufficient number of cells was available, by all different assays. In parallel, 18 HCMV-seropositive pregnant women (median age 33, range 27–42 years) were tested cross-sectionally during the first (n = 7), second (n = 5), and third (n = 6) trimester of pregnancy as pregnant controls with remote (>5 years) HCMV infection. In addition, 19 HCMV-seropositive healthy non-pregnant subjects (median age 34, range 24–46 years) with remote HCMV infection (>5 years) were enrolled as external control subjects. Three of them were tested repeatedly (three to four times) within a time lapse of 1 year. The study was approved by the local ethical committee, and all subjects enrolled in the study gave written informed consent.
Lymphoproliferative Response (LPR)
The LPR was investigated, as previously described , by analysis of CFSE (Molecular Probes, Eugene, OR, USA) dilution after a 7-day culture in the presence of HCMV antigen (AD169-infected cell lysate) or control antigen (uninfected cell lysate). Analysis was performed using a Cytomics FC500 flow cytometer and CXP software (Beckman Coulter). A cell division index (CDI) was calculated as described elsewhere .
HCMV-Specific T-Cell Cytokine and Perforin Production and Phenotype Determination
HCMV-specific T-cell cytokine and perforin production as well as surface phenotype were determined ex vivo  following stimulation with autologous, monocyte-derived , HCMV-infected (VR1814) immature dendritic cells (DC). Peripheral blood mononuclear cells (PBMCs) were tested for frequencies of HCMV-specific interferon-γ (IFN-γ)-, interleukin (IL)-2-, and tumor necrosis factor-α (TNF-α)-producing CD4+ and CD8+ T cells by cytokine flow cytometry. In addition, the differentiation phenotype cell surface marker CD45RA and perforin expression were determined.
The following fluorochrome-conjugated mAbs were used: anti-CD3 PC5, anti-CD8 PC5, anti-IFN-γ FITC, and anti-IL-2 PE (Beckman Coulter Immunotech, Marseilles, France); anti-CD4 ECD and anti-CD8 PC7 (Beckman Coulter, Fullerton, CA, USA); anti-CD45RA FITC, anti-IFN-γ PE-Cy7, and anti-TNF-α PE (BD Pharmingen, San Diego, CA, USA); and anti-perforin PE (Diaclone, clone B-D48, Besançon, France).
Following incubation with HCMV-infected DC, PBMCs were washed, and surface staining was performed for 30 min on ice with anti-CD45RA, anti-CD4, and anti-CD8 in phosphate buffered saline (PBS) supplemented with 5% FBS containing 5% human immunoglobulin and 0.01% sodium azide. Then, cells were washed, fixed, and permeabilized with Cytofix/Cytoperm (BD Pharmingen) for intracellular staining with mAbs: anti-CD3, anti-IFN-γ, anti-TNF-α, anti-IL-2, and anti-perforin. PBMCs were then resuspended in 1% paraformaldehyde and analyzed as above. As a routine, 1–2 × 105 viable lymphocytes were collected.
Absolute CD3+CD4+ and CD3+CD8+ T-cell counts were determined on whole blood samples by a direct immunofluorescence flow cytometry method (Beckman Coulter Inc, Fullertone, CA, USA). The total number of HCMV-specific CD4+ and CD8+ T cells was calculated by multiplying the percentages of HCMV-specific T cells positive for IFN-γ by the relevant absolute CD4+ and CD8+ T-cell counts. The percentages of HCMV-specific T cells producing TNF-α, IL-2, perforin, and re-expressing CD45RA were calculated by determining the percentages of T cells positive for the different parameters within the IFN-γ+ T-cell population (assumed to represent the total number of functional HCMV-specific T cells).
Activated Caspase-3 Cytotoxicity Assay
Following 24-h infection with the HCMV VR1814 strain, autologous DC were used partly as PBMC stimulators and partly as target cells. For the latter purpose, HCMV-infected and mock-infected DC were frozen in liquid nitrogen until use. After 7-day stimulation with HCMV-infected DC in round bottom wells at a ratio of 20 (PBMC):1 (DC) in RPMI supplemented with 5% autologous serum, PBMCs (effector cells) were harvested, washed, and resuspended in RPMI+10% fetal calf serum (FCS). In parallel, target DC were thawed, resuspended in 200 μl of PBS in polypropylene tubes, labeled with 200 μl (2 μM concentration) of PKH26 (Sigma), and incubated for 3 min at room temperature with shaking. Then, an equal volume (400 μl) of PBS+2% human serum albumin was added, and DC were incubated 1 min at room temperature with shaking. PKH26-labeled DC were washed three times with RPMI+10% FCS (changing the tube each time), counted, and resuspended in RPMI+10% FCS. Effector (PBMC) and target (DC) cells (either HCMV-infected or mock-infected) at a ratio of 3:1, respectively, were then cultured in round bottom wells. After a 3-h incubation at 37°C in a 5% CO2 atmosphere, effector and target cells were harvested, washed in PBS, fixed, and permeabilized using Cytofix/Cytoperm. Anti-active caspase-3 FITC mAb (BD Pharmingen) was added, and cells were incubated 30 min at room temperature, washed, and resuspended in 1% paraformaldehyde . Cells were analyzed using a FACS-Calibur instrument (BD) and the Cell Quest software (version 3.2, BD). In preliminary experiments, it was found that target DC, frozen and thawed, gave results comparable to fresh DC as target cells, although they produced a somewhat higher background.
γδ T-Cell Determination
Anti-CD45 PC7, anti-TCR pan γδ PC5, and anti-TCR Vδ2 FITC (Beckman Coulter Immunotech) were used to determine Vδ2− and Vδ2+ γδ T cells in whole blood. The total number of γδ T cells was determined by multiplying the percentage of γδ T cells by the absolute lymphocyte count calculated in whole blood.
Diagnosis and Timing of Primary HCMV Infection
Diagnosis of primary HCMV infection was based on one or more of the following criteria: HCMV-specific IgG seroconversion, HCMV-specific IgM antibody detection and low IgG avidity, and the presence of HCMV or HCMV products in blood . Timing of primary HCMV infection was based on the following criteria: decreasing levels of HCMV-specific IgM antibody, increasing levels of IgG avidity, and presence of non-HCMV-specific clinical symptoms along with laboratory findings . Early HCMV clearance was defined as HCMV DNA disappearance from blood within 3 months after onset of infection, whereas disappearance of virus from blood beyond 3 months was considered late virus clearance.
Diagnosis of Congenital HCMV Infection
Congenital HCMV infection was diagnosed before birth by HCMV isolation or viral DNA detection in amniotic fluid or fetal blood , or within 2 weeks after birth by virus recovery from urine or viral DNA detection in neonatal blood . Overall, of the 21/27 pregnant women analyzed with primary HCMV infection, 8 transmitted virus to the fetus, and 13 did not. Of the remaining six women, HCMV transmission to the fetus was not known.
Variability of immunological assays
LPR and Cytokine Production by HCMV-Specific T Cells in Pregnant Women with Primary HCMV Infection and with Remote Infection (Pregnant Controls)
IFN-γ-producing CD4+ and CD8+ T cells were readily detected during the first month after onset of infection, and their levels did not change with time. However, while the number of IFN-γ-producing CD4+ T cells was not significantly different at any time from that of pregnant women with remote CMV infection (Fig. 2a), HCMV-specific CD8+ T cells were significantly higher in number in pregnant women with primary infection at all time points tested (Fig. 2b). On the other hand, within the aliquot of IFN-γ-producing T cells, HCMV-specific CD4+ and CD8+ T cells producing TNF-α increased progressively, although remaining significantly (p < 0.01) lower than CD4+ of pregnant controls until day 60 and CD8+ until day 270 (Fig. 2c, d). Conversely, HCMV-specific IL-2-producing T cells remained significantly lower (p < 0.05) than controls until 270 days after infection for CD4+ (Fig. 2e), and the same trend (p < 0.05) was also observed until day 360 for IL-2-producing CD8+ (Fig. 2f) T cells. No T-cell producing TNF-α or IL-2 alone (in the absence of IFN-γ) was observed. No difference was observed for any cytokine tested between transmitter and nontransmitter women, as well as between women showing early or late virus clearance (data not shown).
Cytotoxicity and Perforin Expression
Initially, three HCMV-seropositive and three seronegative subjects were tested in parallel by comparing the Cr51 release assay and the activated caspase-3 assay. The latter assay appeared to be more sensitive than the Cr51 assay. Subsequently, the activated caspase-3 assay was used routinely by employing VR1814-infected autologous DC as target cells with a 3:1 effector/target ratio. Cytotoxic T lymphocyte (CTL) activity was already present 30 days after the onset of infection, and its level was lower than that of pregnant controls until day 90. After that time point, the CTL percentage was comparable with that of a small number of controls (Fig. 3a). However, CTL activity within 90 days after infection onset was significantly (p = 0.002) higher in a group of pregnant women clearing the virus early compared with a group of women clearing (>90 days after onset of infection) the virus late (Fig. 3b).
In parallel, HCMV-specific CD4+ and CD8+ IFN-γ-producing T cells expressing perforin were measured. The percentage of CD4+ T cells expressing perforin increased significantly from day 30 to day 180, then declined slowly (Fig. 3c). On the other hand, the percentage of perforin-expressing CD8+ T cells was significantly higher (p < 0.01) than pregnant controls from day 60 through day 360 (Fig. 3d). Since pregnant controls exhibited lower levels of IFN-γ+ perforin expressing CD8+ T cells, one can argue that a physiological reduction in this T-cell phenotype occurs more than 1 year after infection onset. No difference in CTL activity or perforin expression was observed between transmitter and nontransmitter women (data not shown).
Effector Memory T Cells (TEM) Reverting to the RA CD45 Isoform (TEMRA Cells)
The percentage of IFN-γ-producing CD4+ and CD8+ T cells re-expressing CD45RA during the convalescent phase of primary HCMV infection increased until day 90 and day 360 after onset of infection, respectively (Fig. 4a, b). Cytokine and perforin production by CD45RA− and CD45RA+ HCMV-specific T cells was analyzed in 11 pregnant women at different time points after onset of infection. Grouping of results from all samples tested evidenced that both CD4+ and CD8+ TEMRA cells displayed a significantly (p < 0.001) higher production of TNF-α (Fig. 4c) and perforin (Fig. 4d) compared to TEM cells, whereas IL-2 production was not analyzed due to low production.
In addition, the percentages of both CD4+ and CD8+ TEMRA cells at 60 (41–110) days after onset of infection were investigated in 26 pregnant women with primary infection who did not transmit infection to the fetus and in 20 women who did transmit the virus. The group of transmitters (n = 20) included 12 new cases and 8 old cases , whereas the group of nontransmitters included 16 new cases and 10 old  cases (n = 26). As shown in Fig. 4e, the percentage of HCMV-specific CD4+ CD45RA+ T lymphocytes was significantly higher (p = 0.0071) in the group of pregnant women not transmitting the infection to the fetus (median value 17%, range 5–46%) compared to the group of pregnant women transmitting the virus (median value 8%, range 0–36%). Thus, these data from a larger series of pregnancies confirm previous results. In addition, a significant difference between the two groups was (for the first time) observed for CD8+ CD45RA+ T lymphocytes (p = 0.0373). Median percent values were 40% (range 12–72%) for nontransmitters vs 25% (range 11–66) for transmitters (Fig. 4f).
γδ T Lymphocytes
The study of the kinetics of Vδ2− T lymphocytes (median value 50, range 7–151 cells/μl blood) as well as the Vδ2−/Vδ2+ ratio (median value 4.06, range 1.69–10.69) showed that values were significantly greater (p < 0.04) in 23 pregnant women 30 days after onset of primary HCMV infection than in 17 HCMV-seropositive pregnant controls (median value 18.00, range 3–55, Vδ2−/μl; ratio median value 0.60, range 0.1–2.70). Both values were consistently higher than controls until 270 days after onset of infection (Fig. 5a, c), showing a kinetics similar to that of HCMV-specific CD8+ αβ T cells reported in Fig. 2b. On the contrary, no significant difference with respect to pregnant control subjects was observed for Vδ2+ γδ T cells during the first year after onset of infection (Fig. 5b). No difference in number/percentage of Vδ2− was observed between transmitter and nontransmitter women (data not reported). In addition, no difference was observed in the number of Vδ2− T cells between women clearing virus early (<90 days) or late (>90 days).
Comparison of Different Immunological Parameters Analyzed in a Group of Pregnant Women and a Group of Non-pregnant Subjects, Both with Remote HCMV Infection
In order to verify whether pregnancy was a physiological condition inducing in some aspect a certain degree of immunodeficiency, a group of pregnant women with remote (>5 years) HCMV infection (n = 18, with no significant difference observed in a cross-sectional analysis according to the gestational trimester) and a group of age-matched non-pregnant subjects with remote infection (n = 19, showing no significant variations in results from the sequential immunological assays performed in three of them) were examined in parallel for the different immunological parameters analyzed in this study. A single parameter was significantly lower (p < 0.01) in magnitude in the group of pregnant women with no current HCMV infection compared with the group of non-pregnant subjects: IL-2 producing HCMV-specific CD4+ T cells (data not reported). All the other parameters were comparable between the two groups (data not reported).
This investigation was relevant to HCMV-specific CD4+ and CD8+ αβ T cells as well as γδ T cells. The study population included pregnant women only. It is a commonly shared opinion that pregnancy is associated with a period of transitory immune deficiency. In a recent study, we compared the LPR and TNF-α production in primary HCMV infection in a group of pregnant vs a group of non-pregnant women and men . It was concluded that there was no difference between pregnant and non-pregnant individuals, both at the qualitative and the quantitative level. This conclusion was further confirmed and extended in a subsequent study, where both CD4+ and CD8+ T-cell LPR and IFN-γ production were evaluated . Similar results were observed in the present study, except for IL-2 production by CD4+ T cells, which appeared lower in pregnant women. Thus, whether pregnancy itself may impair the immune response to HCMV (and to what extent) should be further investigated.
In the present study, as in the two above-mentioned studies, where the LPR was significantly delayed in mothers transmitting the virus as compared to nontransmitting mothers, it took about 1 year to reach both the CD4+ and CD8+ LPR observed in pregnant controls. In parallel, IL-2 producing CD4+ and CD8+ T cells were detected at a significantly lower level than those observed in healthy HCMV-seropositive controls until up to 1 year (or more) after onset of infection. It is well known that IL-2 production by T cells is important for the induction of CD4+ and CD8+ T-cell proliferation . On the contrary, it was found that IFN-γ- and TNF-α-producing CD4+ and CD8+ T cells were already present in blood at a level comparable to that of pregnant controls within the first month (IFN-γ) or within 3–6 months (TNF-α) after onset of infection.
The delay in the appearance of both the CD4+ and CD8+ LPR suggests that first-line defense mechanisms against HCMV infection do not require major LPR activity by peripheral blood lymphocytes. Instead, the majority of HCMV-specific T cells circulating in blood in the first months after infection resemble either early effector or “exhausted” T cells (low proliferative capacity, production of IFN-γ alone) , while HCMV-specific memory T cells with high proliferative activity and polyfunctional capacity  appear in peripheral blood only months after onset of infection. By contrast, during the early phase of the immune response, these cells might be confined to lymphoid organs, where CD4+ T cells do exert their helper function on both protective arms (B cell and CD8+ T cell) of the immune response already actively operating in the early post-infection period. In fact, neutralizing antibodies, until recently believed to appear late (2–3 months) after onset of infection, have been shown to appear very early if the in vitro cellular detection system (endothelial, epithelial cells) reproduces the in vivo cellular substrate .
In addition, we demonstrated that CTL activity is present the first month after infection onset, although at a level lower than that of pregnant controls. The method employed in this study, using antibody to activated caspase-3, was found to be more sensitive, more informative and a safer alternative to the standard 51Cr release assay, and enables the study of antigen-specific cellular immune responses in real time at the single-cell level [25, 26]. Furthermore, we observed an increase with time after infection of perforin upregulation in both IFN-γ-producing CD4+ and CD8+ T cells. The presence of distinct polyfunctional effector profiles (perforin+IFN-γ upregulation and IL-2+IFN-γ upregulation) was recently reported for CD8+ T cells as an indicator of the protective immune response to HCMV . It appears mandatory at this time to reiterate that a polyfunctional profile including simultaneous production of TNF-α, IFN-γ (and MIP-1β in the absence of IL-2) and direct cytolytic activity associated with surface mobilization of CD107a (a degranulation marker), and intracellular expression of perforin and granzymes, has been reported for mature HCMV-specific CD4+ T cells exhibiting functional properties reminiscent of those of HCMV-specific CD8+ T lymphocytes . This functional profile is not only a characteristic of the CD4+ T-cell response in the chronic phase of HCMV infection , but appears in this study to be even more predominant in the first months after onset of primary infection. In fact, the percentage of CD4+ and CD8+ T cells upregulating perforin is lower in pregnant controls than in subjects analyzed 6 months after infection.
The development of the memory T-cell response was also investigated by studying the differentiation of TEMRA cells during the first year after primary infection. Among HCMV-specific TEM cells, the proportion of those re-expressing CD45RA  progressively increased until reaching median levels similar to those observed in pregnant controls . It has been shown that the reversion of TEM cells to the CD45RA+ phenotype correlates with virus clearance from blood and that CD45RA is re-expressed on memory T cells with homeostatic proliferation in the absence of antigen . Recently, data have been collected both in transplanted as well as HIV-infected patients indicating that the development of TEMRA cells and control of HCMV infection are closely related [21, 22]. Previous preliminary data showing that a higher percentage of HCMV-specific CD4+ TEMRA cells was associated with a lower chance of transmitting HCMV infection to the fetus have been extended and confirmed in this study with larger study populations of pregnant women transmitting and not transmitting the virus to the fetus. Results of this study document that not only CD4+ CD45RA+ IFN-γ+ (p = 0.0071) but also CD8+ CD45RA+ IFN-γ+ (p = 0.0373) percentages are significantly higher in nontransmitting women. Thus, besides the delay in LPR development in transmitters, also a delayed re-expression of CD45RA in both CD4+ and CD8+ T cells appears to be a crucial factor for vertical HCMV transmission. Although this trend cannot be applied to the prediction of virus transmission in individual pregnant women, given the large overlap of results in the two groups (positive and negative predictive values around 75%), it may shed some light on future research efforts, which would better define the biological significance of this particular T-cell subset within the context of immune protection against vertical HCMV transmission. In this respect, apart from the negligible IL-2 production already reported for CD8+  and confirmed here also for CD4+ TEMRA cells, we observed a higher cytokine and perforin production in CD45RA+ cells as compared to CD45RA− cells.
As for γδ T cells, for more than a decade, these cells have been thought to represent a first-line defense mechanism against HCMV infection in kidney allograft recipients whose αβ T-cell response was weakened by immunosuppression . This defense consisted of γδ T-cell expansion which was concomitant with the resolution of HCMV infection and disease, regardless of HCMV serologic status of the donor and recipient prior to transplantation . Only Vδ2− γδ T cells underwent long-lasting expansion and were able to kill HCMV-infected target cells in vitro . Until recently, γδ T cells have been considered to be innate-like effectors unable to establish antigen-specific memory. However, long-term expansion of effector/memory Vδ2− γδ T cells has been shown to be a specific signature of an adaptive immune response to HCMV infection in both immunocompetent and immunocompromised patients . In addition, HCMV-specific CD8+ αβ T cells and Vδ2− γδ T cells have been shown to share common expansion kinetics and a common effector phenotype, suggesting that these cell types act similarly in response to HCMV infection . This is also the first study to demonstrate a Vδ2− γδ T cell expansion in primary HCMV infection of immunocompetent pregnant women, who had Vδ2− levels significantly greater than those of controls at least until 9 months after onset of infection, whereas Vδ2+ γδ T cells of HCMV-seropositive subjects were comparable in number over the entire follow-up period. This finding resembles the expansion previously reported for Vδ2− γδ T cells in transplanted patients. In addition, recently, a strong γδ T-cell response to congenital HCMV infection has been shown during development in utero suggesting that this T-cell subset could participate in antiviral defense in early life . A contraction of the expanded Vδ2− pool occurs after 1 year post-infection. However, the higher number of Vδ2− γδ T cells observed in HCMV-seropositive controls with remote infection as compared to HCMV-seronegative subjects (data not reported) may indicate the persistence of a “memory” γδ T-cell response or a chronic activation due to intermittent HCMV reactivation.
In conclusion, the complete polyfunctional profile of the HCMV-specific memory response in peripheral blood T cells is achieved only after months (or years) post-infection. In this respect, the different signatures of the HCMV-specific T-cell memory response should be taken into account when measuring the in vitro and in vivo immune response to a candidate HCMV vaccine.
The entire technical staff of the Virology Unit is gratefully acknowledged. In addition, we are indebted to Daniela Sartori for careful preparation of the manuscript and to Laurene Kelly for competent revision of the English. This work was partially supported by the Ministero della Salute, Fondazione IRCCS Policlinico San Matteo, Ricerca Corrente (grants 80425 and 80513) and Ricerca Finalizzata (grant 89301), and by Fondazione Carlo Denegri (FCD), Torino, Italy.