T and B Cell Markers in Dried Blood Spots of Neonates with Congenital Cytomegalovirus Infection: B Cell Numbers at Birth Are Associated with Long-Term Outcomes

Human CMV is the most common cause of congenital infections worldwide, with an overall birth prevalence in industrialized countries between 0.6 and 0.7% (1, 2). Among congenitally infected children, 12.7% are estimated to be symptomatic at birth, with the most common symptoms being petechiae, jaundice, hepatosplenomegaly, thrombocytopenia, chorioretinitis, and microcephaly (1, 2). An estimated 40–58% of these symptomatic children will develop permanent sequelae, such as hearing loss, mental retardation, and developmental delay (1). Importantly, of the 87.3% of neonates that are asymptomatic at birth, ∼13.5% will also develop permanent sequelae (1). Defining markers that may help to predict whether a neonate will develop long-term impairment (LTI) will have profound impact on postnatal policy.

The pathogenesis of fetal damage during congenital infection is still poorly understood due to the complex interplay between viral, maternal, placental, and fetal factors. It has been shown that infections occurring predominantly in the first half of pregnancy are associated with sequelae (3, 4). The developing fetal immune system may play a role in controlling the infection later in pregnancy, thereby preventing the development of long-term sequelae (5). There is evidence of fetal and neonatal immune system activation in the context of congenital CMV infection (cCMV). Previous studies have shown expansions of fetal γδ T cells (6), CD8⁺ T cells (7, 8), and CD4⁺ T cells (7, 9) in cCMV. Upon congenital infection similar types of effector CD4⁺ and CD8⁺ T cells are generated as in adults (10–12), but they appear to be functionally impaired (7) with the CD4⁺ T cell response being more impaired than CD8⁺ T cells (8, 13). Despite the fetal capacity to generate IgM against CMV (14), the fetal B cell response has not been extensively studied in cCMV. Previous studies on cellular and humoral immunity in cCMV have not related their findings to either symptoms at birth or LTI.

In recent years, molecular markers for T and B cells have been used to address several clinical questions. These markers specifically involve the circular excision products produced upon the most common TCR and Ig gene rearrangements during T and B progenitor cell generation in thymus and bone marrow, respectively. δRec-ΨJα signal joints on TCR excision circles (TRECs) are used in newborn screening on dried blood spots (DBS) for primary immunodeficiency such as SCID (15, 16) because they can be readily detected by quantitative PCR (17–19). Furthermore, studies have been carried out to implement screening for agammaglobulinemia (20), using quantitative PCR on DBS for Igκ-deleting recombination excision circles (KRECs). In addition, measurement of TRECs has been used in HIV patients to monitor the cellular recovery after initiation of highly active antiretroviral therapy (21), and in stem cell transplantation patients with primary immunodeficiency, TRECs and KRECs were used to monitor the thymic T cell and bone marrow B cell neogenesis (22, 23). Finally, the replication history of isolated B cell subsets has been used to improve the characterization of immunological disease with aberrant B lymphoid maturation (24–26).

For this study, neonatal DBS from a large cohort of children with cCMV and from non-infected controls were analyzed for molecular markers of T and B cells and the results were related to clinical data from birth until 6 y of age. Using quantitative real-time PCR on DNA isolated from DBS, TCR and Ig gene rearrangements were detected. δRec-ΨJα signal joints on TRECs were detected as a measure of αβ T cell thymogenesis and coding joints of Vδ1-Jδ1 and Vδ2-Jδ1 rearrangements in the genome as a measure of circulating γδ T cells. Furthermore, intron recombination signal sequence k-deleting element coding joints (cj intronRSS-Kde) in the genome and intronRSS-Kde signal joints on KRECs were quantified as a measure of circulating B cells and newly derived bone marrow B cells, respectively. KRECs were additionally used to determine the B cell replication history (27).

In the current study, the quantification of TRECs and KRECs, and molecular markers for γδ T cells was applied to provide new insights into the immune regulation of cCMV and to identify early markers to predict LTI at 6 y of age, such as neurodevelopmental impairment.

A previously described, nationwide, retrospective cohort was used in this study. It was derived from a total of 31,484 children born in 2008 in the Netherlands who were retrospectively tested for cCMV infection by PCR for CMV DNA in neonatal DBS at 5 y of age (28). After approval by the Medical Ethics Committee of the Leiden University Medical Center, the parents of 125 congenitally CMV–infected children and 263 non-infected children were asked to participate. The parents of 99 congenitally infected children gave informed consent for the use of DBS and of clinical data for this study. In addition, 54 controls without cCMV were randomly sampled from a control group matched for gender, month of birth and region. Children were defined as symptomatic at birth if they had one or more of the following signs or symptoms in the neonatal period: prematurity, small size for gestational age, microcephaly, hepato- or splenomegaly, generalized petechiae or purpura, hypotonia, abnormal laboratory findings (elevated liver transaminases, hyperbilirubinemia, neutropenia or thrombocytopenia), cerebral ultrasound abnormalities, ophthalmologic abnormalities or neonatal hearing impairment. LTI was defined as the presence of impairment in one or more domain (hearing, visual, neurologic, motor, cognitive, and speech-language). Additionally, the severity of the LTI was assessed by accumulating the number of domains affected and indicated as the presence of LTI in two or more domains. The same definitions were used for children with and without cCMV. Finally, in this cohort maternal seroimmunity to CMV before birth was unknown, therefore it was assumed that cCMV infection could have resulted from either maternal primary or secondary infection.

After a first initial CMV PCR screening performed at the National Institute for Public Health and the Environment, a second confirmatory PCR was performed at the Leiden University Medical Center (28). For this purpose, DNA was extracted from DBS by using the QIAamp DNA minikit according to the previously described protocol (29). For each test one full DBS was punched by using an automated DBS puncher (1296-071, Perkin Elmer-Wallac, Zaventem, Belgium). CMV DNA amplification of a 126-bp fragment from the immediate-early Ag region was performed using an internally controlled quantitative real-time PCR as described previously (30, 31) on a CFX96 Real-Time PCR Detection System (BioRad, Veenendaal, The Netherlands). The PCR was performed in triplicate and the CMV viral load expressed in IU/ml.

To study the numbers of T and B cells in neonatal DBS, the most frequently formed TCR and Ig gene rearrangements were quantified by TaqMan-based quantitative PCR. The δRec-ΨJα rearrangement occurs in 70–80% of TCRD alleles in mature αβ T lymphocytes, resulting in TRECs in nearly all newly formed αβ T cells (32). Vδ1-Jδ1 and Vδ2-Jδ1 rearrangements are collectively present in nearly all mature γδ T cells in neonates, but absent in αβ T cells (6, 33). The intronRSS-Kde rearrangement occurs in ∼30% of Igκ⁺ and almost all Igλ⁺ mature B lymphocytes (27).

These rearrangements and the β-globin housekeeping gene were quantified in triplicate from each sample using two multiplex real-time PCR assays (Table I). Each multiplex assay consisted of 5 μl of DNA extract and 20 μl reaction mixture containing 3.5 mM of MgCl₂, 0.04 mg/ml of BSA and 12.5 μl HotStar Master mix (QIAGEN, Hilden, Germany). Both mixtures were optimized by primer limitation, and for probe and MgCl₂ concentration with primers and probes specific for TREC, KREC, cj intronRSS-Kde, Vδ1-Jδ1, Vδ2-Jδ1, and β-globin to ensure equal amplification efficiencies. The phocine herpesvirus (PhHV) amplification was used to check for inhibition of the PCR and β-globin amplification was used to control for the number of nucleated cells. Two PCR mixes were used: the first contained 300 nM primers with 200 nM probe for cj intronRSS-Kde, Vδ1-Jδ1, and 500 nM primers with 200 nM probe for β-globin. The second contained 300 nM KREC, Vδ2-Jδ1 and PhHV primers, 900 nM TREC primers, 200 nM KREC, TREC, and Vδ2-Jδ1 probes, 50 nM PhHV probe.

Quantification was performed using a dilution series of DNA from the following cell lines: DB01+T, a modification of the previously published U698 DB01 cell line (27) that contains one TREC, one KREC, and one cj intronRSS-Kde rearrangement copy per genome; Peer (34), a T lymphoid cell line, containing one Vδ1-Jδ1 rearrangement copy per genome; and T-ALL T032 (34), a T-acute lymphoblastic leukemia cell line, containing one Vδ1-Jδ1 rearrangement construct per genome. For TRECs two positive controls (cord blood with two different levels of TRECs spotted on filter paper) and a negative control (leukocyte-reduced adult blood produced by filtration and spotted on filter paper) were included in each run. These materials were kindly provided by the Centers for Disease Control and Prevention, Atlanta. The real-time PCR was performed on a CFX96 Real-Time PCR Detection System (BioRad) in a 96-well plate using a thermocycling profile as follows: 15 min 95°C followed by 45 cycles of 95°C (30 s), 55°C (30 s), and 72°C (30 s). Data were analyzed with CFX Manager 3.1 software.

The percentage of cells that contained the rearrangement in relation to the total amount of nucleated cells in blood was calculated using the ΔΔCt method with the formula as previously described (35): 2^((CT_β-globin − CT_target)_sample − (CT_β-globin − CT_target)_{cell line})*100%. The housekeeping gene β-globin was quantified in both the sample and cell line used as standard.

The average number of B cell divisions, the B cell replication history, was calculated with the formula as previously described (35) (CT_sj − CT_cj)_sample − (CT_sj − CT_cj)_{cell line}.

The copies per μl of whole blood were calculated based on the theoretical recovery of 1 μg DNA from ∼150,000 cells (19, 36) and the assumption that one full spot contains 50 μl of whole blood.

Ten-fold serial dilutions of DNA from the various control cell lines carrying the target rearrangements were used to determine the assay sensitivity. The analytical sensitivity, expressed as the lower limit of detection, was assessed by testing the dilutions in triplicate. Quantification of β-globin levels was possible with 0.05 ng total DNA with fewer than 36 PCR cycles in DB01+T, Peer and T-ALL 032 cell lines. Quantification of TREC and KREC levels was possible with 0.05 ng total DNA with fewer than 37 PCR cycles. Quantification of cj intronRSS-Kde, Vδ1-Jδ1, and Vδ2-Jδ1 levels was possible with 0.05 ng total DNA with <36 PCR cycles.

The similar efficiency between each target and the β-globin is required to use the ΔCt method for correct quantification of normalized target levels (37–39). The amplification efficiency, determined from the slope of the log-linear portion of the calibration curve, was assessed by testing the dilutions in triplicate. The efficiencies of the assays were very similar: 0.99 ± 0.02 for cj intronRSS-Kde, 0.98 ± 0.01 for KREC, 0.96 ± 0.02 for TREC and 0.99 ± 0.01 for β-globin in DB01+T cell line; 1.04 ± 0.01 for Vδ1-Jδ1 and 1.03 ± 0.03 for β-globin in Peer cell line; 1.06 ± 0.03 for Vδ2-Jδ1 and 0.98 ± 0.01 for β-globin in T-ALL 032 cell line; all R² ≥ 0.98.

The differences in the levels of immunological markers between the different categories — CMV status, viral load, symptoms at birth, and LTI — were assessed by using a linear mixed model with random effects to account for the repeated measurements on the same patient. A Pearson’s correlation analysis between viral loads, log (IU/ml), and the different immunological markers was carried out. A p value <0.05 was considered statistically significant. Due to the exploratory nature of this study, the correction for multiple comparison was not applied for multiple statistical testing. Data were analyzed by using the Statistical Package for Social Sciences (version 23, Chicago, IL).

DBS of 99 children with cCMV were tested using two multiplex real-time PCR assays (Table I), as were 54 controls. The study population and presence of symptoms at birth and LTI are shown in Table II. In the control group, seven (12.9%) children showed symptoms at birth and five (9%) had LTI. In the children with cCMV, 16 (16%) were symptomatic at birth and 22 (22%) had LTI.

First, the effect of cCMV was assessed by quantifying TCR and Ig gene rearrangements on DNA from DBS using real time PCR and comparing children with cCMV (cCMV+) and children without cCMV (cCMV−). Supplemental Table I shows the estimated means for the markers in our cohort. The cCMV+ group had a trend toward a significant decrease in the percentage of cells that contained the TREC rearrangement normalized for the presence of β-globin (p = 0.073) (Fig. 1A). In accordance with this finding, the number of TRECs per μl of whole blood was lower than the control group (p = 0.043) (Fig. 1B).

It has been shown that prematurity, defined as birth before 37 wk gestational age, is related to a lower amount of TRECs (15, 40) and that intrauterine growth retardation, or dysmaturity, is associated with a small thymus (41). To exclude their influence on the differences in TREC levels, an additional sensitivity analysis was performed by first excluding the group of premature newborns, 10 (10%) with cCMV and 4 (7.4%) in the control group, and then by excluding the group of dysmature newborns, two (2%) with cCMV and two (3.7%) in the control group. No differences were found in the estimates and significance compared with the whole cohort (data not shown). Therefore, our data suggest that the reduction in the amount of TRECs is not confounded by prematurity or by being small for gestational age.

The percentage of γδ T cells that contained the Vδ1-Jδ1 rearrangement was significantly higher in the cCMV+ group (p = 0.019). To exclude the possibility that this higher percentage of Vδ1-Jδ1 rearrangements was the result of fewer αβ T cells, we determined the absolute number of Vδ1-Jδ1 per microliter of whole blood. Indeed, significantly more Vδ1-Jδ1 copies per μl of whole blood were present in the infected group (p = 0.038). No statistically significant difference between the groups was observed for the percentage of cells with the Vδ2-Jδ1 rearrangement or for the Vδ2-Jδ1 copies per μl of whole blood.

In the cCMV+ group, the percentage of cells that contained the cj intronRSS-Kde rearrangement was higher than in the cCMV− group (p = 0.055). As the cj intronRSS-Kde copies per microliter of whole blood were slightly higher in the cCMV+ group, but not statistically significant, the small increase in percentage might be due to a decrease in αβ T cells. The KREC copies and the percentage of cells that carried a KREC were no different in the cCMV+ and cCMV− groups (p = 0.297, p = 0.139, respectively) even though the same trend of higher numbers in the cCMV+ group was observed. Consequently, the B cell replication history was not significantly different between cCMV+ and controls.

To study the relation between T and B cell numbers and CMV viral load, the cCMV+ group was divided into three by taking two cut-points at the first and third quartile of the viral load in DBS, resulting in a low (n = 24), medium (n = 50), and high (n = 25) viral load group. The mean viral loads for each group were 254, 2907 and 27,121 IU/ml, and the estimated means for the molecular T and B cell markers according to these groups are shown in Table III.

The high viral load group did not show statistically significant differences compared with the low and medium viral load groups for the TRECs copies per microliter of whole blood or the percentage of cells that contained the TREC rearrangement.

DBS from neonates with high CMV viral loads showed a significantly higher percentage of cells containing Vδ1-Jδ1 rearrangements than the low viral load group (p = 0.022), and slightly higher than the medium viral load group (p = 0.124) (Fig. 2A). In addition, the Vδ1-Jδ1 copies per microliter of whole blood were significantly higher in this group than in the low and medium viral load groups (p = 0.033, p = 0.041, respectively) (Fig. 2B). Furthermore, a Pearson’s r data analysis suggested a positive correlation between Vδ1-Jδ1 percentage as well as Vδ1-Jδ1 copies per microliter, and viral loads (p < 0.001, r = 0.25, and r = 0.23, respectively). No statistically significant differences between the groups were observed for the percentage of cells that carried Vδ2-Jδ1 rearrangements or in the copies Vδ2-Jδ1 copies per microliter of whole blood.

The percentage of cells that contained KRECs (Fig. 2C) and the KREC copies per microliter (Fig. 2D) were significantly higher in the high viral load group than in the low and medium viral load groups (p = 0.002, p = 0.001, p < 0.001, respectively). A Pearson’s r data analysis suggested a positive correlation between percentage of KREC, as well as KREC copies per microliter, and viral load (p < 0.001, r = 0.33, and r = 0.41, respectively). Additionally, in the high viral load group, the percentage of cells that contained the cj intronRSS-Kde rearrangement was significantly higher than the low viral load group (p = 0.024), and trend significantly higher than the medium viral load group (p = 0.068) (Fig. 2E). These differences were even more significant for the cj intronRSS-Kde copies per μl of whole blood (p = 0.001, p = 0.004, respectively) (Fig. 2F). A Pearson’s r data analysis suggested a positive correlation between the percentage of cj intronRSS-Kde, as well as cj intronRSS-Kde copies per microliter, and viral loads (p < 0.01, r = 0.26, and r = 0.37, respectively). No differences in the B cell replication history were observed between the viral load groups.

Next, TCR and Ig gene rearrangements were studied in relation to symptoms at birth to evaluate whether they reflected differences in symptoms (Supplemental Table I). The comparison of T and B cell markers between asymptomatic cCMV+ and symptomatic cCMV+ individuals did not show statistical differences for any of the markers. Additionally, when comparing the asymptomatic cCMV+ with asymptomatic cCMV− groups, a similar trend was observed in the same markers as shown in the overall comparison between cCMV+ and cCMV− children (data not shown). Finally, no significant differences in viral loads were found between symptomatic and asymptomatic subjects.

Next, the TCR and Ig gene rearrangements were correlated with LTI (Supplemental Table I). First, we evaluated the immunological markers in relation to the development of any disorder in one or more of the following domains of impairment: hearing, visual, neurologic, motor, cognitive, and speech-language. When comparing the group of children with cCMV infection that develop any LTI to those who do not, a significantly lower percentage of cells that contain the KREC rearrangement was observed (p = 0.008) (Fig. 3A). Also, the KRECs copies per μl of whole blood were significantly lower in the patients with LTI (p = 0.005) (Fig. 3B). A similar trend, although not statistically significant, was observed for the percentage of cells that contained the cj intronRSS-Kde rearrangement as well as cj intronRSS-Kde copies per μl of whole blood (p = 0.137, p = 0.073). Second, TCR and Ig gene rearrangements were assessed in relation to the development of more extensive impairments, defined as LTI in two or more domains. Similar trends were observed when children with more extensive impairments were compared with children without any LTI. Lower percentages and numbers of KRECs (p = 0.04, p = 0.02, respectively) and slightly lower percentages and numbers of cells containing cj intronRSS-Kde rearrangements were detected in cCMV+ children with more extensive LTI than in cCMV+ children without any LTI. No differences in B cell replication history and TCR rearrangements were observed in relation to LTI.

Next, we evaluated if the same markers that were significant in the overall cCMV+ and cCMV− comparison were still present in the absence of any LTI. Indeed, a similar trend was observed in the same markers, TREC and Vδ1-Jδ1 (data not shown). In addition, the cCMV+ group had a significant increase in the percentage of cells that contain the KREC rearrangement (p = 0.021) and KREC copies (p = 0.044) as well as the percentage of cells that contain the cj intronRSS-Kde rearrangement (p = 0.017) and slightly higher cj intronRSS-Kde copies per microliter of whole blood (p = 0.075).

Finally, no significant differences in viral loads were found between congenitally infected children with or without LTI.

The analysis of molecular markers for T and B cells from DBS in this large cohort of children shows that cCMV resulted in reduced thymic production of αβ T cells, increased numbers of γδ T cells, and a trend toward increased numbers of B cells. Children with cCMV and LTI did show lower numbers of B cells. The observed trend of increasing B cell numbers in the infected group was further emphasized when excluding the patients with LTI, who had a lower number of B cells and might have diluted the effect.

The reduced number of TRECs in the DBS from children with cCMV suggests that intrauterine infection leads to reduced thymic production of T cells. Indeed, the sensitivity analysis indicated that the reduction in the amount of TRECs in our cohort is not confounded by prematurity or by being small for gestational age. CMV infection is known to induce a shift from naive toward more differentiated αβ T cells with a reduction in the pool of naive T cells, as has been shown in immunosuppressed individuals and the elderly (42, 43), as well as in pregnant CMV IgG-seropositive women (44). However, these all concern adults in whom the reduction of the pool of naive T cells might result from a process that takes place over a longer time than the gestational period and where the longevity of naive T cells might also play a role in masking differences in thymic output, with the reduced output being visible only after several years (45). Therefore, this process is unlikely to be the cause of the reduced amount of TRECs in our cohort. Interestingly, in vitro CMV has been shown to be capable of infecting thymic epithelial cells that play a central role in T cell development and maturation, both during gestation and early stage of life (46, 47). Moreover, in newborn infants with cCMV, hypoplastic thymuses have been described and in both guinea pig and mouse models pathologic changes of the thymus have been shown (48–50). Although the TREC numbers were only moderately lowered, the effect of cCMV on thymopoiesis certainly deserves further study.

In our study, we did not find an increase in Vδ2 T cells in the cCMV infected group, however, we looked at Vδ2-Jδ1 so a role of γδ T cells with other Vδ-Jδ rearrangements, as reported previously, cannot be excluded (6). Our observation of increased percentages of Vδ1 T cells in children with cCMV is in accordance with a previous study that showed a γδ T cell response upon cCMV after primary maternal CMV infection. These fetal γδ T cells, detected as early as 21 wk of gestation, were shown to be activated, undergo cell division and become differentiated with highly restricted repertoires (6). Moreover, γδ clones derived from cCMV-infected newborns showed antiviral activity when incubated with CMV-infected cells (6) suggesting a role in controlling viral replication (51). These unconventional cells react rapidly upon activation (52) and develop earlier than αβ T cells during immune ontogeny. Therefore, they might have an important role in early life (51) and, possibly, in a context where the αβ T-cell response is impaired, they might be more efficient in controlling the early phases of cCMV. In the mouse model, when adaptive mechanisms are impaired or absent, γδ T cells can provide effective control over CMV infection (53, 54). To further support this, γδ T cell expansion in solid organ transplanted patients in response to CMV was associated with the resolution of infection and less symptomatic CMV disease, whereas late γδ T cell expansion correlated with a more intense and durable CMV infection (55). In our cohort, this was not associated with fewer symptoms at birth or LTI. However, the symptoms that define CMV disease in solid organ–transplanted patients are different to the clinical signs in the cCMV setting and the ability of γδ T cells to control long-term CMV disease has not yet been elucidated (52).

CMV viral load in DBS was not correlated to symptoms at birth or to LTI in our cohort. Some previous studies have demonstrated a relation between CMV viral load with clinical outcome (56, 57), whereas others have not (58–60). The predictive role of CMV viral load in blood for congenital CMV disease may differ depending on the timing of infection and whether there was a primary maternal infection or recurrent infection and therefore it still needs to be clarified. In our cohort it is impossible to establish the trimester of infection or if a maternal primary infection occurred. Interestingly, none of our tested molecular B and T cell markers were associated with symptoms at birth in cCMV infected children. cCMV infected children with LTI did show significantly lower absolute and relative numbers of KRECs as well as slightly decreased, though not statistically significant, absolute and relative numbers of cj intronRSS-Kde, compared with cCMV infected children without LTI. However, there were no differences in B cell replication. The same trend was observed when considering more severe LTI. These findings suggest that cCMV infection does not induce a notable intrauterine B cell proliferation and possibly no considerable Ab production, but rather hint to an increase in B cell production. Unfortunately, little information is available on fetal B cell immunity in relation to cCMV. IgM positive B cells have been shown to emerge in the peripheral circulation as early as 12 wk of gestation (61) and CMV infected fetuses can produce IgM (62–64), but the antiviral activity and the role in CMV disease control have not yet been evaluated (5). Whether the different numbers of B cells are associated with postnatal differences in the capacity to generate long-lived plasma cells, memory B cells or support effector functions of immune cells remains to be elucidated. The potential protective role of Ab can be illustrated by the fact that primary CMV infection in pregnancy is associated with a vertical transmission rate of 30–40% and that this risk is at least 10-fold lower in seropositive pregnant women (2). Although it is uncertain whether Abs are capable of influencing an ongoing CMV infection, a possible better initial B cell response might be beneficial in controlling the progressive tissue damage responsible for LTI development, possibly due to a sustained viral replication and spread. A positive correlation between B cell numbers and viral load was observed. In a group of pregnant women with primary infection, an expansion of a large pool of activated memory B cells enriched for CMV specificity and higher in viremic women was shown, further supporting a causal relationship between high viral loads and cell activation (65). On the other hand, the difference in B cell numbers that we observed at birth between infected children with and without LTI may also be related to a different timing of infection, with earlier infections leading to a more extensive cCMV and inflammation influencing the early B cell lymphopoiesis in the fetal liver.

To our knowledge, this is the first study on molecular markers for T and B cells in neonatal DBS of cCMV infected children in relation to long-term outcomes. A reliable marker for long-term outcomes could provide the means to introduce the long-debated (66) newborn screening program for CMV in DBS by defining subgroups that would benefit from clinical, audiological follow-up and possibly antiviral treatment. Whether KRECs, which are related to LTI, have enough discriminative power needs to be assessed in other CMV cohorts. Finally, this study on molecular markers generates new hypotheses on the effects of CMV infection on fetal, and possibly child, immunity and on the potential protective role of B cells in cCMV infection.

We thank Professor Arnaud Marchant for critically reading the manuscript.

The authors have no financial conflicts of interest.