• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of clinexpimmunolLink to Publisher's site
Clin Exp Immunol. Nov 2006; 146(2): 234–242.
PMCID: PMC1942048

The T cell response to persistent herpes virus infections in common variable immunodeficiency


We show that at least half of patients with common variable immunodeficiency (CVID) have circulating CD8+ T cells specific for epitopes derived from cytomegalovirus (CMV) and/or the Epstein–Barr virus (EBV). Compared to healthy age-matched subjects, more CD8+ T cells in CVID patients were committed to CMV. Despite previous reports of defects in antigen presentation and cellular immunity in CVID, specific CD4+ and CD8+ T cells produced interferon (IFN)-γ after stimulation with CMV peptides, and peripheral blood mononuclear cells secreted perforin in response to these antigens. In CVID patients we found an association between a high percentage of circulating CD8+ CD57+ T cells containing perforin, CMV infection and a low CD4/CD8 ratio, suggesting that CMV may have a major role in the T cell abnormalities described previously in this disease. We also show preliminary evidence that CMV contributes to the previously unexplained severe enteropathy that occurs in about 5% of patients.

Keywords: common variable immunodeficiency, cytomegalovirus, CD8+ T lymphocytes, perforin


Most patients with primary antibody deficiency (PAD) are prone to bacterial infection, but usually recover uneventfully from common acute virus infections such as measles and mumps [1,2]. In contrast, patients with severe T lymphocyte defects are prone to life-threatening infections with a broad range of viruses [3]. Common variable immunodeficiency (CVID) is the most common PAD, occurring in approximately 1:25 000 of the Caucasian population; the current consensus is that CVID is caused by a variety of inherited genetic defects, the disease in many patients probably being determined by multiple abnormalities in different cell types [4]. In contrast, X-linked agammaglobulinaemia (XLA) is a much rarer condition caused by a single genetic defect in Btk, a cytoplasmic kinase involved in B lymphocyte differentiation; affected patients therefore have a severe antibody deficiency with normal cellular immunity [5].

About 50% of CVID patients have a T cell lymphopenia; this is often associated with a relative expansion of CD4+ CD45RO+ T cells, poor T cell proliferation to mitogens in vitro and low numbers of class-switched B cells in the blood [6]. Many patients who are lymphopenic also have what appears to be a significant defect in cellular immunity, as shown by failure to mount delayed hypersensitivity skin reactions and failure to prime T cells after immunization with protein antigens (reviewed in [4]). This may be explained partly by a recently reported defect in the maturation of monocyte-derived dendritic cells in vitro[79]. These same patients often show signs of persistent T cell stimulation, with raised numbers of circulating T cells expressing human leucocyte antigen D-related (HLA-DR), interleukin (IL)-12Rβ1, cAMP-dependent protein kinase A type 1 and CD57 [1012]. Furthermore, there are other signs of persistent immune stimulation and inflammation in these patients, such as increased levels of plasma soluble CD25, β2-microglobulin, neopterin and low levels of reduced homocysteine, the latter being a marker of increased oxidative stress [4]. Patients with these abnormalities often have splenomegaly and/or unexplained inflammation in liver, lungs or bowel. Patients with XLA do not have these T cell abnormalities, suggesting that they are not secondary to a failure of antibody production, or to the immunoglobulin therapy given to both XLA and CVID patients.

Jaffe et al. [13] investigated a group of CVID patients with raised numbers of CD8+ CD57+ T cells, representing about half their cohort of patients, and showed that these cells produced high levels of IFN-γ after stimulation, had high cytotoxic activity and suppressed B cell IgG production in vitro. The association between raised numbers of ‘activated’ CD8+ T cells and a defect in B cell immunoglobulin class-switching in vitro was confirmed in a larger cohort by Farrant et al. [6]. Serrano et al. [14] investigated similar patients and found high numbers of oligoclonal CD8+ T cells expressing perforin. Similar but less pronounced CD8+ T lymphocyte changes are seen in immunocompetent individuals with cytomegalovirus (CMV) infection [15], raising the possibility that these abnormalities reflect persistent viral infection in a substantial subset of CVID patients.

Against this background, we have investigated the possibility that CMV and the Epstein–Barr virus (EBV), two common herpes viruses that are known to induce a substantial persistent immune response in healthy individuals, may be contributing to the T cell abnormalities observed in CVID patients.

Materials and methods

Patients and volunteers

A total of 76 patients with CVID, defined using International Union of the Immunological Societies (IUIS) criteria [16], were enrolled for the study. Eight patients with X-linked agammaglobulinaemia (XLA) were chosen as a ‘disease control’ as they have a severe antibody deficiency and suffer from a similar spectrum of infections to CVID patients. All the CVID and XLA patients were receiving regular immunoglobulin replacement therapy at the time of the study. Twenty-nine healthy volunteers acted as controls.

Ethical permission for the study was granted from the local regional ethical committee (04Q0501119) and blood was taken after informed consent was obtained. The study was conducted according to good clinical practice (GCP) standards.

Flow cytometric analysis

Lymphocyte subsets

Fresh peripheral blood mononuclear cells (PBMCs) isolated on a Ficoll-Hypaque density gradient were used for immunological assays.

Tetramer/streptamer staining

Seventy-two CVID patients were genotyped for HLA alleles: the CD8+ T cells from 29 patients (M = 12, F = 17, mean age: 49 years) who were positive for either HLA-A*0201 or HLA-B*0801 were tested for binding to viral peptide-bound HLA-class I tetramer/streptamers. A control group of 13 healthy volunteers positive for HLA-A2 or B8 alleles was also tested, these being seropositive for EBV and/or CMV (M = 8, F = 5, mean age: 41 years).

EBV-specific T cells were analysed using different tetramer-peptide complexes of HLA-A*0201 or HLA-B*0801 molecules bound to viral peptides. HLA-A*0201 restricted peptides consisted of: GLCTLVAML (derived from the lytic cycle protein BMLF1) and CLGGLLTMV (derived from the latent cycle protein LMP2). HLA-B*0801 restricted epitopes were RAKFKQLL (from the lytic cycle protein BZLF1) and FLRGRAYGL (derived from the latent cycle protein EBNA3A). All tetramers were streptavidin–phycoerythrin (PE)-labelled. EBV tetramer staining was performed by adding 0·5 µg/ml of tetramer to 0·5 × 106 PBMCs and incubating for 30 min at room temperature.

For CMV, a streptamer system was used (IBA Gmbh, Göttingen, Germany) incorporating the HLA-A2-restricted peptide NLVPMVATV, following the manufacturer's recommendations. Briefly, this involved conjugating the peptide loaded HLA-A2 with Strep-Tactin-PE, incubating on ice with PBMCs and then staining with various combinations of monoclonal antibodies (mAbs). To detect tetramer/streptamer binding cells, events from 150 000 lymphocytes were acquired and only those CD8+ T cells in a well-demarcated cluster counted. Negative controls were cells from non-HLA-A*0201 and B*0801 healthy individuals seropositive for EBV or CMV.

Intracellular perforin staining

Briefly, after cell surface staining with anti-CD8-peridinin chlorophyll (PerCP) or CD8-allophycocyanin (APC) (BD Biosciences, CA, USA) and CD27-PE (Dakocytomation, Cambridgeshire, UK) or CD28-PE (BD Biosciences, CA, USA), cells were fixed and permeabilized with Fix and Perm (Caltag Laboratories, CA, USA). Anti-perforin-fluorescein isothiocyanate (FITC) (BD PharMingen) was added to the cells during permeabilization and incubated for 20 min in the dark before analysis on a BD FACSCalibur flow cytometer, using CellQuest software to analyse cell markers immediately after staining.

Perforin release

Perforin release was measured using an enzyme-linked immunospot (ELISpot) assay (Mabtec, Stockholm, Sweden) following the manufacturer's instructions with minor modifications. Briefly, 96-well polyvinylidene difluoride-backed plates (Millipore, Bedford, MA, USA) were coated with anti-perforin mAb overnight at 4°C. After washing in phosphate-buffered saline (PBS) with 0·05% Tween 20 (Sigma, Poole, UK), and blocking with RPMI-10% fetal calf serum (FCS) for 2 h at room temperature, 2 × 105 PBMCs in 100 µl RPMI-1640 with 10% human AB male serum (Sigma) was added to wells. A mixture (100 µl) of 23 HLA class 1 binding CMV, EBV and influenza A peptides (CEF) (Mabtech) [17], or overlapping peptides from the CMV protein pp65 (1 µg/ml) was added to triplicate wells at a final volume of 200 µl/well and incubated overnight at 37°C. Cells stimulated with 10 µg/ml phytohaemagglutinin (PHA) were used as a positive control, the negative control being cells in media alone. After washing, released perforin was detected with biotin-conjugated anti-perforin antibody (1 µg/ml) for 2 h at room temperature. Following a further wash, plates were incubated with streptavidin-conjugated alkaline phosphatase for 1 h at room temperature. The reaction was developed with an alkaline phosphatase chromogenic substrate kit (Bio-Rad, CA, USA) and the spots counted by an automated ELISpot reader (Karl Zeiss − Imaging Associates, Hallbergmoos, Germany).

EBV-induced interferon (IFN)-γ production

Specific CD8+ T cell responses to HLA-A*0201 or HLA-B*0801-matched EBV-transformed B lymphoblastoid cell lines were assessed by intracellular IFN-γ in five selected HLA-A*0201 or -B*0801-positive CVID patients. In one case, an autologous B lymphoblastoid cell line was also used; 1 × 106 PBMCs were cultured overnight with 1 × 106 lymphoblastoid cells. Brefeldin A was added at a final concentration of 5 µg/ml after the first hour of incubation. The negative controls were three HLA unmatched lymphoblastoid cell lines. As an internal positive control, PBMCs were stimulated with anti-CD3 and anti-CD28. After 16 h incubation, the cells were stained with anti-CD3 FITC and anti-CD8 PerCP, then fixed and permeabilized. Anti-IFN-γ-APC was added and the cells incubated for 30 min before the final wash.

CMV-induced IFN-γ production

PBMCs were incubated with anti-CD28 (1 µg/ml BD Biosciences) along with overlapping synthetic peptides for either IE1 or pp65 (1 µg/ml) (JPT, Peptide Technologies, Berlin, Germany) or CMV lysate (CMV antigen; Dade Behring, Marburg, Germany) for 2 h at 37°C. GolgiPlug (1 µg/ml BD Biosciences) was then added and cells were incubated for a further 14 h. T cells were surface-stained with CD4-PerCP and CD3-APC for 15 min, fixed and washed then permeabilized in the presence of anti-IFN-γ-FITC.

EBV and CMV genome copy number in blood

EBV or CMV genomic DNA was measured quantitively in whole blood from selected patients. A real-time polymerase chain reaction (PCR) method, with an EBV-infected Raji cell line as a standard, was used to measure EBV copy number in blood with a lower sensitivity of 50 copies/ml (Micropathology Ltd, Coventry, UK). An in-house Tachman real-time PCR was used for CMV copy number with primers for the glycoprotein B gene, with a lower sensitivity of 200 copies/ml of blood [18].

Elution and staining of lymphocytes from intestinal biopsies

Two mucosal biopsies were obtained from each of two patients at endoscopy, rinsed in RPMI-1640 medium and treated for 3 min with 1 mM dithiothreitol in calcium and magnesium-free PBS. After washing in PBS, the samples were incubated in 1 mM ethylenediamine tetra acetic acid (EDTA) for 1 h, washed twice in PBS containing 10% FCS and 60 mM CaCl2 and 55 mM MgCl2. Samples were then treated with 48 u/ml collagenase (Sigma) in RPMI-1640 for 16 h at 37°C. Undigested tissue was removed by passing through a 105-µm nylon mesh and the cells washed in PBS, and then stained for CD3, CD8 and with the CMV streptamer.


The majority of the data was analysed using the non-parametric Mann–Whitney U-test. The χ2 test (Fisher's exact test) was used where indicated.


EBV and CMV tetramer-specific CD8+ T cells in CVID

The percentage of CD8+ T cells recognizing selected EBV and CMV epitopes were measured in HLA-A*0201 or HLA-B*0801-positive patients and seropositive healthy controls. Sixty-nine per cent of the CVID patients had CD8+ T cells specific for the EBV peptide used in the analyses, whereas 55% of the patients recognized the CMV peptide. Overall, there was a 13-fold increase in the frequency of CD8+ T cells recognized by the CMV streptamer in CVID patients compared to healthy seropositive controls (P = 0·005) (Table 1). In contrast, there was no significant difference in the frequency of the CD8+ T cells recognizing the EBV peptides between the positive CVID patients and healthy controls. Representative fluorescence activated cell sorter (FACS) plots illustrating the high frequency of CMV responses are shown in Fig. 1. The majority of the patients were tested at least twice within a 6-month period and similar results were obtained (data not shown).

Fig. 1
Cytomegalovirus (CMV)-specific CD8+ T cells in common variable immunodeficiency (CVID) patients are CD27 perforin-positive CD57+. Peripheral blood mononuclear cell (PBMC) staining of one CVID patient with a human leucocyte antigen (HLA)-A2-restricted ...
Table 1
Epstein–Barr virus (EBV) and cytomegalovirus (CMV)- specific CD8+ T cell responses in common variable immunodeficiency (CVID) patients and healthy normal seropositive controls.

Phenotypic analysis of viral peptide-specific CD8+ T cells

The distribution of surface markers within the tetramer/streptamer binding T cell populations were analysed to determine whether there were significant differences between CVID patients and controls (Table 2). Generally, surface phenotyping of EBV- and CMV-specific CD8+ T cells from healthy individuals were consistent with those described previously [19,20]. For the EBV (lytic peptide)-specific cells, the majority were CD27+ CD28+/– in both patients and controls. For CMV-specific cells there was an increase in the percentage of late memory cells (CD27 CD28) in CVID patients compared to CMV seropositive controls [Table 2]. A high proportion of the CMV-specific CD8+ T cells from these CVID patients expressed CD57 (median 70%) and perforin (median 90%).

Table 2
Phenotypic analysis of Epstein–Barr virus (EBV)-lytic and cytomegalovirus (CMV)-specific CD8+ T cells.

IFN-γ response to EBV and CMV antigens

To study the functional integrity of viral specific T cells we measured IFN-γ production in stimulated cells.

CD8+ T cell response to EBV lymphoblastoid cell lines

Intracellular IFN-γ was measured in PBMCs from five HLA-A*0201 or B*0801-positive CVID patients following incubation with three HLA-matched heterologous EBV lymphoblastoid cell lines, and in one patient incubated with her own cell line. The numbers of cells producing IFN-γ were similar in frequency to the tetramer-positive cells for the single peptides (data not shown), indicating that the tetramers recognize the majority of the EBV epitopes expressed on the cell lines. The finding that only two of 13 patients with EBV-tetramer binding CD8+ T cells had measurable EBV DNA in the blood (20 and 300 copies/ml) implies good control of this virus in vivo.

Response to CMV peptides

PBMCs from 21 CVID patients (11 HLA-A*0201 with positive CMV streptamer binding) were stimulated with overlapping peptides from pp65 and IE-1 (Fig. 2). The CD3+ CD4 (presumed CD8+) T cell response to the pp65 peptide pool was generally higher in the CVID patients compared to controls (Fig. 2a). In addition we assessed intracellular IFN-γ production by CD4+ T cells in response to these peptides (Fig. 2b). In all cases, when CVID patients demonstrated a CD8+ T cell response there was also a CD4+ T cell response. We tested for CMV DNA in the blood of seven of these patients and all were negative (< 200 copies/ml), suggesting good control of CMV in vivo.

Fig. 2
T cell-derived interferon (IFN)-γ response to cytomegalovirus (CMV) in common variable immunodeficiency (CVID) patients and normal subjects. Peripheral blood mononuclear cells (PBMCs) were stimulated with either pp65 or IE-1 synthetic peptides ...

Perforin release after CMV peptide stimulation

As an additional measure of the functional integrity of viral specific CD8+ T cells, we stimulated PBMCs with a mixed pool of CMV, EBV and influenza A (CEF)-derived peptides and/or pp65. There was a higher frequency of perforin-releasing T cells in response to CEF and pp65 by ELISpot assay in those CVID patients with a high percentage of perforin-positive cells, although patients with both high and low percentages of perforin-positive cells had similar numbers of cells releasing perforin after PHA stimulation (Table 3). We analysed this further by subdividing the patients into those with > 55% CD8+ T cells expressing intracellular perforin (designated ‘high perforin’ group) and with < 55% perforin-positive cells (designated ‘low perforin’ group). The ‘cut-off ’ at 55% was chosen because this represented 2 standard deviations above the mean perforin levels observed in healthy controls. This confirmed an association between the presence of high percentages of CD8+ perforin-positive T cells and a response to CEF and pp65 (Table 3). Furthermore, all six patients in the ‘high perforin’ group had T cells reacting with CMV peptides using other assays (see above).

Table 3
Perforin release in response to mitogen or viral peptides.

Non-specific changes in blood lymphocytes

We looked for evidence that those CVID patients with lymphocyte responses to CMV peptides belonged to the subgroup with non-specific changes in circulating lymphocytes, suggesting persistent immune activation.

Lymphocyte subset markers

There were no significant differences in the absolute numbers of CD4+ and CD8+ T cells in 18 CVID patients with evidence of a CMV-specific response (measured by positive binding to streptamer and/or IFN-γ expression after peptide and CMV lysate stimulation) and 14 CVID patients without such evidence. However, those patients whose lymphocytes reacted with CMV peptides had a significantly decreased CD4/CD8 ratio (median = 1·1) compared to those not reacting (median = 1·87; P = 0·0038) (data not shown).

Perforin-positive cells

Intracellular perforin was measured ex-vivo in the unstimulated PBMCs of 38 randomly selected CVID patients. Figure 3a shows a significant increase in the percentage of perforin-positive cells in CVID patients compared to healthy controls. Surface phenotyping showed that most of these cells were CD27 CD28. In CVID patients with greater than 55% perforin-positive CD8+ T cells, a median of 12% of CD4+ CD3+ cells expressed perforin. Because suitable reagents were not available to stain cells simultaneously for both CD57 and perforin, we determined the correlation between CD8+ T cells stained separately for these markers. As shown in Fig. 3b, we found a close relationship between the percentage of perforin-positive CD8+ T cells and CD57+ CD8+ T cells.

Fig. 3
Relationship between circulating perforin-positive CD8+ T cells and CD8+ CD57+ T cells. Fresh peripheral blood mononuclear cells (PBMCs) were stained with anti-CD8 and anti-perforin antibodies. (a) Common variable immunodeficiency (CVID) patients (n = ...

Correlation between percentage of perforin-positive CD8+ T cells and CMV-specific CD8+ T cell response

The proportion of CMV and EBV-specific CD8+ T cells (tested with tetramers/streptamers) in CVID patients with high (> 55%) and low (< 55%) percentages of perforin-positive CD8+ T cells was investigated (Table 4). Those in the ‘high perforin’ group were much more likely to have a T cell response to CMV, confirmed with CMV peptide-induced IFN-γ expression. A T cell response against CMV was seen in only two patients in the ‘low perforin’ group, and both these had levels of perforin-positive cells close to the 55% ‘cut-off ’ (Fig. 4). In contrast, most of the patients with an EBV-specific response were in the ‘low perforin’ group (Table 4) One of the eight XLA patients had very low numbers of responding cells; the others were below the limit of detection (Fig. 4).

Fig. 4
Higher interferon (IFN)-γ production in response to cytomegalovirus (CMV) in common variable immunodeficiency (CVID) patients with a high percentage of perforin-positive CD8+ T cells. Peripheral blood mononuclear cells (PBMCs) from 10 CVID patients ...
Table 4
Associations between cytomegalovirus (CMV)-specific, Epstein–Barr virus (EBV)-specific CD8+ T cells and perforin-positive CD8+ T cells in CVID.

Clinical evidence for CMV-induced inflammatory response in bowel

A 58-year-old patient (HLA-A*0201-positive) with high numbers of CMV specific tetramer-positive CD8+ T cells in the blood (Fig. 1) had suffered from severe watery diarrhoea (bowel frequency 8–15/day) for 12 years. Histology of biopsies taken from the colon showed an active colitis with a moderate increase in CD8+ T cells in both the lamina propria and intraepithelial area. Most of the lymphocytes stained positive for perforin (data not shown). A biopsy was positive for CMV DNA by PCR for the glycoprotein B(gB) gene, although the blood was persistently negative (< 200 copies/ml), and there were no typical CMV inclusions within macrophages in the lamina propria. The patient was given a 7-day course of intravenous ganciclovir followed by oral valganciclovir for 35 days. There was a marked improvement in diarrhoea symptoms within 7 days, with reduction in bowel frequency to twice daily by 14 days. After discontinuing the valganciclovir there was a gradual relapse of these symptoms over the subsequent month. A further colon biopsy showed an active colitis. Analysis of the lymphocytes eluted from two colon biopsies at this time showed that 33% were CD8+, and ~5% of these bound the CMV (NLV) streptamer. The patient remains in remission (2 months) after further treatment with a combination of steroids and valganciclovir for 4 weeks. Lymphocytes eluted from a colon biopsy in another patient in this study with diarrhoea and evidence of circulating T cell reactivity to CMV showed similar results; this patient has responded partially to anti-tumour necrosis factor (TNF)-α (Infliximab) therapy and has not yet been treated with ganciclovir. A third patient with a similar T cell response to CMV and colitis subsequently developed overt fatal CMV enteritis with macrophage inclusion bodies after being given cytotoxic therapy for a B cell lymphoma.

A preliminary analysis of clinical complications suggests an association of CMV infection, determined by the presence of streptamer binding CD8+ T cells or an IFN-γ response by CD4+ T cells to CMV lysate, with other inflammatory complications (six of 16 patients with granulomatous/inflammatory disease affecting lungs, liver or eyes had a T cell response to CMV, compared to two of 14 with no evidence of prior infection).


Given the immune defects associated with CVID and the treatment of patients with regular immunglobulin infusions, it has been difficult to estimate the exposure of patients to herpesvirus infections such as EBV and CMV using serological methods. Using an approach based on enumerating antigen-specific T cells we have shown for the first time that over 50% of CVID patients have circulating CD8+ T cells primed to EBV and/or CMV peptides. The phenotype of the EBV and CMV peptide-specific CD8+ T cells was similar to that reported in healthy individuals, with the response to EBV being predominantly in the early/intermediate memory T cell subset (CD27+), with an activated and terminally differentiated phenotype for CMV (CD28 perforin-positive) [19,20]. The magnitude of the peripheral blood EBV-specific CD8+ T cell response was similar between CVID patients and seropositive healthy individuals. In contrast, there was a 13-fold increase in the frequency of CD8+ T cells reacting to CMV peptides in comparison to healthy age-matched seropositive controls. Our results suggest that the viral-specific CD8+ T cells in CVID are functionally intact, in that they express IFN-γ and release perforin after viral peptide stimulation, although cytotoxic assays are needed to confirm normal CTL function.

We found an association between a CMV T cell response, high relative numbers of CD8+ CD57+ perforin-positive T cells and a low CD4+/CD8+ T cell ratio in the blood. These T cell phenotypic changes delineate a subset of CVID patients who often have other evidence of persistent lymphocyte activation [4,10,13]. Similar but less marked changes have been described in CMV-infected healthy individuals who have circulating oligoclonal CD8+ CD57+ T cells that proliferate to CMV antigens in vitro[15,21,22]. Oligoclonal expansion within the total CD8+ T cell population has been demonstrated previously in many CVID patients [14], and we found similar oligoclonality within the CD8+ perforin-positive T cell subset in three patients tested in our cohort (data not shown). Although we found a clear association between CMV T cell reactivity in CVID and a relative expansion of CD8+ perforin-positive T cells, only a minority of these cells reacted with CMV peptides. Because these cells may be involved in the inflammatory complications of CVID, there is a need to identify their antigenic specificity using a larger repertoire of CMV peptides, and possibly peptides from viruses such as HHV-8 that has been linked recently to inflammatory complications in CVID [23].

Interpretation of the mainly negative specific T cell responses to CMV peptides in the XLA patients is confounded by their younger age and the fact that most started their immunoglobulin therapy in childhood, in contrast to those with CVID who started treatment as adults, in some cases probably after exposure to CMV. However, if our data on a small number of XLA patients is confirmed in a larger cohort, it will show for the first time that passive antibody given regularly from an early age protects against CMV infection, and that vaccines stimulating a humoral response against CMV are likely to be protective.

During our study we investigated three patients with severe enteropathy, a complication in CVID that is difficult to manage [24]. These patients had circulating T cells reacting to CMV peptides, with 12% of total CD4+ T cells in one case reacting to CMV lysate. Although this patient's symptoms improved dramatically after ganciclovir therapy, none of the three patients had CMV inclusion bodies in bowel biopsies, a finding which most pathologists accept as evidence of significant CMV disease. Nevertheless, CMV viral DNA was found by PCR in colon biopsies from two patients, and a third subsequently developed fulminating overt CMV enteropathy after cytotoxic therapy for a lymphoma. Further investigation of the role that CMV plays in this clinical condition is warranted, including the possibility that an aggressive CD8 T cell response in the gut may be contributing to the pathology and that other immune cells such as natural killer (NK) cells, which are also important for complete control of CMV infection, may be deranged in these patients [25,26].

It is possible that the expansion of T cell responses to CMV is an attempt to compensate for the failure to produce specific antibodies, which in the murine lymphocytic choreomeningitis viral (LCMV) model of persistent infection are necessary for optimal control of LCMV [27]. However, all the patients described here receive regular replacement IgG immunoglobulin therapy that contains anti-gB CMV antibodies, and probably a spectrum of antibodies to other viral epitopes because batches are manufactured from pools of plasma obtained from approximately 10 000 donors. Five patients in our cohort with evidence of CMV-reactive T cells, including the three patients with colitis, were positive for anti-gB antibody immediately prior to their 3-weekly infusion of immunoglobulin, showing that this antibody is not being exhausted during the intervening period (data not shown).

In conclusion, we provide evidence of normal T cell responses to EBV and CMV in most patients with CVID, the response to CMV being exaggerated in some patients. These data contrast with previous reports of failure to generate a T cell response to proteins and contact sensitizers, raising the possibility that there may be a ‘crowding out’ of such responses by CMV-stimulated cells [21,28]. Our data suggest that CMV infection is an important factor in the circulating T cell abnormalities described in CVID by many workers, and that this needs to be taken into account in future attempts to delineate subsets of patients based on T cell markers and function. Preliminary clinical observations suggest that an exaggerated T cell response to CMV may cause or exacerbate enteropathy in CVID, and may also contribute to other inflammatory complications.


We thank Eira Rawlings for technical assistance and the staff of the routine immunpathology laboratory for measuring the CD4/8 T cell ratios. We also thank Dr Marcus Harbord for obtaining mucosal specimens and Dr James Thaventhiran for help in identifying lymphocyte subsets in these samples. The work was part-funded by a grant from the Primary Immunodeficiency Association, UK.


1. Hermaszewski RA, Webster D. Primary hypogammaglobulinaemia: a survey of clinical manifestations and complications. Q J Med. 1993;86:31–42. [PubMed]
2. Cunningham-Rundles C, Bodian C. Common variable immunodeficiency: clinical and immunological features of 248 patients. Clin Immunol. 1999;92:34–48. [PubMed]
3. Stiehm ER, Chin TW, Haas A, Peerless AG. Infectious complications of the primary immunodeficiencies. Clin Immunol Immunopathol. 1986;40:69–86. [PubMed]
4. Webster ADB. Common variable immunodeficiency. In: Roifmann C, editor. Humoral immunodeficiencesImmunol Allergy Clin North Am. Vol. 21. 2001. pp. 1–21.
5. Sideras P, Smith CI. Molecular and cellular aspects of X-linked agammaglobulinemia. Adv Immunol. 1995;59:135–223. [PubMed]
6. Farrant J, Spickett G, Matamoros N, et al. Study of B and T cell phenotypes in blood from patients with common variable immunodeficiency (CVID) Immunodeficiency. 1994;5:159–69. [PubMed]
7. Bayry J, Lacroix-Desmazes S, Kazatchkine MD, et al. Common variable immunodeficiency is associated with defective functions of dendritic cells. Blood. 2004;104:2441–3. [PubMed]
8. Scott-Taylor TH, Green MRJ, Eren E, Webster AD. Monocyte derived dendritic cell responses in common variable immunodeficiency. Clin Exp Immunol. 2004;138:484–90. [PMC free article] [PubMed]
9. Cunningham-Rundles C, Radigan L. Deficient IL-12 and dendritic cell function in common variable immunodeficiency. Clin Immunol. 2005;115:147–53. [PubMed]
10. Wright JJ, Wagner DK, Blaese RM, Hagengruber C, Waldmann TA, Fleisher TA. Characterization of common variable immunodeficiency: identification of a subset of patients with distinctive immunophenotypic and clinical features. Blood. 1990;7:2046–51. [PubMed]
11. McQuaid A, Tormey VJ, Trafford B, Webster AD, Bofill M. Evidence for increased expression of regulatory cytokine receptors interleukin-12R and interleukin-18R in common variable immunodeficiency. Clin Exp Immunol. 2003;134:321–7. [PMC free article] [PubMed]
12. Aukrust P, Aandahl EM, Skalhegg BS, et al. Increased activation of protein kinase A type 1 contributes to the T cell deficiency in common variable immunodeficiency. J Immunol. 1999;162:1178–85. [PubMed]
13. Jaffe JS, Strober W, Sneller MC. Functional characterisation of CD8+ T cells define a unique subset of patients with common variable immunodeficiency. Blood. 1993;82:192–201. [PubMed]
14. Serrano D, Becker K, Cunningham-Rundles C, Mayer L. Characterization of the T cell receptor repertoire in patients with common variable immunodeficiency: oligoclonal expansion of CD8+ T cells. Clin Immunol. 2000;97:248–58. [PubMed]
15. Wang ECY, Moss PAH, Frodsham P, et al. CD8high CD57+ T lymphocytes in normal, healthy individuals are oligoclonal and respond to human cytomegalovirus. J Immunol. 1995;155:5046–56. [PubMed]
16. Chapel H, Geha R, Rosen F. IUIS PID (Primary Immunodeficiency) Classification Committee. Primary immunodeficiency diseases: an update. Clin Exp Immunol. 2003;132:9–15. [PMC free article] [PubMed]
17. Currier JR, Kuta EG, Turk E, et al. A panel of MHC class I restricted viral peptides for use as a quality control for vaccine trial; ELISPOT assays. J Immunol Meth. 2002;260:157–72. [PubMed]
18. Kidd IM, Fox JC, Pillay D, Charman H, Griffiths PD, Emery VC. Provision of prognostic information in immunocompromised patients by routine application of the polymerase chain reaction for cytomegalovirus. Transplantation. 1993;56:867–71. [PubMed]
19. Hislop AD, Gudgeon NH, Callan MFC, et al. EBV-specific CD8+ T cell memory: relationships between epitope specificity, cell phenotype, and immediate effector function. J Immunol. 2001;167:2019–29. [PubMed]
20. Appay V, Dunbar PR, Callan M, et al. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat Med. 2002;8:379–85. [PubMed]
21. Sylwester AW, Mitchell BL, Edgar JB, et al. Broadly targeted human cytomegalovirus-specific CD4+ and CD8+ T cells dominate the memory compartments of exposed subjects. J Exp Med. 2005;202:673–85. [PMC free article] [PubMed]
22. Khan NN, Shariff M, Cobbold R, et al. Cytomegalovirus seropositivity drives the CD8 T cell repertoire toward greater clonality in healthy elderly individuals. J Immunol. 2002;169:1984–92. [PubMed]
23. Wheat WH, Cool CD, Morimoto Y, et al. Possible role of human herpesvirus 8 in the lymphoproliferative disorders in common variable immunodeficiency. J Exp Med. 2005;202:479–84. [PMC free article] [PubMed]
24. Teahon K, Webster AD, Price AB, Weston J, Bjarnason I. Studies on the enteropathy associated with primary hypogammaglobulinaemia. Gut. 1994;35:1244–9. [PMC free article] [PubMed]
25. Desrosiers M-P, Kielczewska A, Loredo-Osti J-C, et al. Epistasis between mouse Klra and major histocompatibility complex class 1 loci is associated with a new mechanism of natural killer cell-mediated innate resistance to cytomegalovirus infection. Nat Genet. 2005;37:593–9. [PMC free article] [PubMed]
26. Iversen A-C, Norris PS, Ware CF, Benedict CA. Human NK cells inhibit cytomegalovirus replication through a noncytolytic mechanism involving lymphotoxin-dependent induction of IFN-β J Immunol. 2005;175:7568–74. [PubMed]
27. Bachmann MF, Hunziker L, Zinkernagel RM, Storni T, Kopf M. Maintenance of memory CTL responses by T helper cells and CD40–CD40 ligand: antibodies provide the key. Eur J Immunol. 2004;34:317–26. [PubMed]
28. Khan N, Hislop A, Gudgeon N, et al. Herpes virus specific CD8 T cell immunity in old age: cytomegalovirus impairs the response to coresident EBV infection. J Immunol. 2004;173:7481–9. [PubMed]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...