• CD56;
  • cytomegalovirus;
  • T cells


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Human T cells expressing CD56 are capable of tumour cell lysis following activation with interleukin-2 but their role in viral immunity has been less well studied. Proportions of CD56+ T cells were found to be highly significantly increased in cytomegalovirus-seropositive (CMV+) compared with seronegative (CMV) healthy subjects (9·1 ± 1·5% versus 3·7 ± 1·0%; P < 0·0001). Proportions of CD56+ T cells expressing CD28, CD62L, CD127, CD161 and CCR7 were significantly lower in CMV+ than CMV subjects but those expressing CD4, CD8, CD45RO, CD57, CD58, CD94 and NKG2C were significantly increased (P < 0·05), some having the phenotype of T effector memory cells. Levels of pro-inflammatory cytokines and CD107a were significantly higher in CD56+ T cells from CMV+ than CMV subjects following stimulation with CMV antigens. This also resulted in higher levels of proliferation in CD56+ T cells from CMV+ than CMV subjects. Using Class I HLA pentamers, it was found that CD56+ T cells from CMV+ subjects contained similar proportions of antigen-specific CD8+ T cells to CD56 T cells in donors of several different HLA types. These differences may reflect the expansion and enhanced functional activity of CMV-specific CD56+ memory T cells. In view of the link between CD56 expression and T-cell cytotoxic function, this strongly implicates CD56+ T cells as being an important component of the cytotoxic T-cell response to CMV in healthy carriers.




culture medium








monoclonal antibody


natural killer


peripheral blood mononuclear cells




peridinin chlorophyll protein


T-cell receptor


tumour necrosis factor-α


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

CD56 is a homologue of neural cell adhesion molecule present on most natural killer (NK) cells and also a small subpopulation of T cells.[1, 2] T cells expressing CD56 have been variously referred to as CD3+ CD56+ cells,[3] NK-like T cells[4] and cytokine-induced killer cells.[5] They have been most studied in relation to tumour cell killing and following activation with cytokines such as interleukin-2 (IL-2), their specific or non-specific cytotoxic function against tumour cells is enhanced.[6-8] However, their role in combating virus infection has not been well studied. Phenotypically, CD56+ T cells are mostly T-cell receptor (TCR)-αβ+ T cells[4, 7] expressing CD8[8] but lacking most NK cell inhibitory and activating receptors.[6, 8] They normally comprise between 1% and 11% of peripheral lymphocytes[8] and hence are considerably more abundant than classical CD1d-restricted invariant NKT cells which, although some express CD56,[9] normally make up < 1% of peripheral lymphocytes.[10] Levels of CD56+ T cells in peripheral blood of healthy subjects have been reported to increase significantly with age.[11, 12]

Relative levels of CD56+ T cells were found to increase during the early course of infection in malaria patients but had normalized after disease resolution.[13] However, there were significant decreases in numbers in patients with the autoimmune diseases psoriasis[14] and systemic lupus erythematosus.[15] In patients with coronary artery disease, levels of interferon-γ (IFN-γ) producing CD56+ CD8+ T cells were significantly increased compared with healthy controls.[16] Although mean levels were increased in lung cancer patients, we previously reported that functional activity in terms of CD25 and IFN-γ expression was impaired.[17] No consistent changes have been seen in patients with a range of viral infections. While CD56+ T-cell levels were increased in chronic hepatitis B virus infection,[18] a decrease was seen in patients suffering from acute hepatitis E virus infection.[19] There were highly significant increases in Chinese children infected with HIV-1 compared with uninfected controls.[20]

Cytomegalovirus (CMV) is a commonly occurring herpesvirus that infects over half of the world's population, mostly during childhood. The virus then remains latent for the lifespan of the individual where it is kept at bay by an ongoing immune response but is never completely eliminated owing to multiple immune evasion strategies.[21] CMV rarely causes significant symptoms except in neonates or immunosuppressed patients.[22] In immunocompetent individuals, CD8+ T cells provide defence against reactivation[23] but NK cells are also important in protection.[24] In some elderly CMV seropositive (CMV+) subjects, CMV-specific CD4+ and CD8+ T cells have been greatly expanded to the detriment of T cells of other specificities, leading to immunosenescence as seen in the ‘immune risk profile’.[25]

There are few studies of CD56+ T cells in relation to CMV infection. Although transient elevations in CD8+ CD56+ T cells were reported in primary CMV infection in renal transplant patients,[26] higher levels of CD56+ cells expressing HLA-DR were found in a similar cohort of patients with CMV reactivation, although these were not specifically identified as T cells.[27] Increased levels of CD56+ T cells were associated with CMV positivity in elderly subjects[28] and in age-related macular degeneration, although in the latter study CMV seropositivity did not have a significant additional effect.[29] Also, an unusual population of CD8+ T cells restricted to CMV peptides in association with HLA-E were found to express CD56 and were designated NK-cytotoxic T lymphocytes.[30] In the present report we have investigated the frequencies, phenotype and function of CD56+ T cells in relation to CMV infection in healthy subjects.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Peripheral blood mononuclear cell samples and cell lines

Peripheral blood was obtained from a panel of 53 healthy volunteers aged between 23 and 60 years (29 men, 24 women), following written informed consent. Subjects known to be suffering from any autoimmune, malignant or immunosuppressive disease were excluded. Between 10 and 20 ml of blood was taken into preservative-free heparin (Wockhardt UK Ltd, Wrexham, UK). Ethical approval for the study was obtained from the Liverpool Research Ethics Committee, Project Ref. 2K/175. Peripheral blood mononculear cells (PBMC) were prepared by density gradient centrifugation using Ficoll-Paque™ Plus (GE Healthcare, Little Chalfont, Bucks, UK). Cells accumulating at the interface were washed twice in PBS and resuspended in culture medium (CM) consisting of RPMI-1640 + 10% heat inactivated fetal calf serum + 2 mm glutamine + antibiotics. K562 erythroleukaemia cells were maintained in suspension culture in CM, subculturing twice weekly.

Preparation of cytomegalovirus lysate extract

A modification of the method of Suni et al. [31] was used to prepare CMV lysate from infected fibroblasts. Briefly, human fetal foreskin fibroblasts were maintained in Dulbecco's modified Eagle's medium supplemented with 2 mm glutamine, antibiotics and 10% fetal calf serum. Cells were infected with AD169 or Towne strains of CMV at a multiplicity of infection of 4 : 1 at 37° for 90 min with occasional rocking. Excess virus was washed off and cells were incubated in flasks for several days, checking daily until a cytopathic effect was noted. From this time-point, supernatants were collected and stored at −80°, replacing the medium. This was repeated until cells showed extreme cytopathic effect when they were removed by a sterile scraper and pooled with the supernatants. Cells and debris were then pelleted and snap frozen in liquid N2 followed by heating to 50° several times. Lysate was exposed to UV light to inactivate any remaining intact virus particles.

Anti-cytomegalovirus antibody assay

An ELISA for measurement of anti-CMV IgG antibody titres was obtained from GenWay Biotech (San Diego, CA) and used according to the manufacturer's instructions. Briefly, serum or plasma samples were collected from healthy subjects and frozen at −20° until analysis. Samples were assayed in duplicate and in all cases produced a clear cut negative or positive result, allowing definitive assignment of CMV serostatus of all healthy subjects.

Flow cytometric analysis

Aliquots of 1 × 106 cells were placed into 5-ml tubes and incubated with a combination of monoclonal antibody (mAb) at a dilution of 1 : 50 in a volume of 100 μl CM for 30 min at 4° before washing. For initial identification of CD56+ T cells the antibodies used were anti-CD3-Peridinin chlorophyll protein-Cy5·5 (PerCP-Cy5·5) and anti-CD56-allophycocyanin (APC) together with appropriate isotype control antibodies (BD Biosciences, Oxford, UK). Lymphocytes were gated based upon forward and side scatter parameters and results were expressed as total % of T cells that were CD56+. Analyses were performed using winmdi 2.8 (

For subsequent phenotypic analyses mAb conjugated to a third colour were used as follows: anti-CD4-phycoerythrin (PE), anti-CD8-PE, anti-CD11a-PE, anti-CD16-PE, anti-CD25-PE, anti-CD45RA-PE, anti-CD45RO-PE, anti-CD57-PE, anti-CD58-PE, anti-CD62L-PE, anti-CD69-PE, anti-CD94-PE, anti-CD127-PE, anti-CD161-PE, anti-CCR7-PE, anti-TCR Vα7.2-PE, anti-TCR-γδ-PE (BD Biosciences), anti-CD158a-PE, anti-CD158b-PE (Beckman Coulter, High Wycombe, UK), anti-CD158e-PE (Miltenyi Biotec, Bisley, UK), anti-CD158f-PE (BioLegend, London, UK) and anti-NKG2C-PE (R & D Systems, Abingdon, Oxford, UK) at a dilution of 1 : 50. Lymphocytes expressing both CD3 and CD56 were gated and cells expressing a third antigen recognized by a PE-conjugated mAb were enumerated as a % of total CD3+ CD56+ lymphocytes.

Stimulation of PBMC and intracellular cytokine staining

PBMC from both CMV+ and CMV healthy subjects were prepared and resuspended in CM at a concentration of 106 cells/ml. They were then cultured in CM alone, CMV lysate at a concentration of 1 : 10 or 200 ng/ml staphylococcal enterotoxin B (Sigma Aldrich, Poole, Dorset, UK) as a positive control, for 24 hr. Brefeldin A at 10 μg/ml was added for the last 4 hr of culture. Cells were then surface labelled with anti-CD3-FITC and anti-CD56-APC before fixing with intracellular fixation buffer and intracellular permeabilization buffer (eBioscience, Hatfield, UK). Next, PerCP-Cy5·5-conjugated mAb against IFN-γ, tumour necrosis factor-α (TNF-α), IL-2 or isotype control antibody were added to separate aliquots. CD3+ CD56+ cells were gated and results expressed as % CD3+ CD56+ cells expressing each individual cytokine. For examination of CD107a exposure, cells were stimulated for 4 hr either with K562 erythroleukaemia cells at an effector : target ratio of 2 : 1 or with CMV antigen extract in the presence of brefeldin A as described above and unfixed cells were stained with anti-CD3-PerCP-Cy5·5, anti-CD56-APC and anti-CD107a-PE before analysis. Lymphocytes were gated on the basis of forward and side scatter to ensure that K562 cells were excluded from the analysis. Cells cultured with CM alone under the same conditions served as controls.

Proliferation assays

PBMC from CMV+ subjects were resuspended in pre-warmed PBS at a concentration of 106 cells/ml. Carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes; Invitrogen, Paisley, UK) was added at a concentration of 5 μm in DMSO and cells were incubated for 5 min at 37°. Then 5 volumes of ice-cold PBS was added and the cells were washed twice and resuspended in CM. Cells were then cultured for up to 7 days either in CM alone, or in the presence of CMV antigen extract or staphylococcal enterotoxin B at a concentration of 1 μg/ml as a positive control. Cells were then labelled with anti-CD3-PerCP-Cy5·5 and anti-CD56-APC and analysed by flow cytometry. Gated CD3+ CD56+, CD3+ CD56 and CD3 CD56+ lymphocytes were analysed for reduction of CFSE staining and results were expressed as % of cells that had undergone one or more divisions.

HLA typing

The class I HLA type of healthy donors was determined serologically using monoclonal antibodies against several of the common HLA types for which HLA pentamers were available. Antibodies against HLA-A2 and HLA-B7 conjugated to FITC were obtained from Serotec (Kidlington, Oxford, UK). Anti-HLA-B8 conjugated to biotin was obtained from AbCam (Cambridge, UK) and was used in conjunction with streptavidin-FITC (Serotec). Aliquots of 1 × 106 PBMC were incubated with conjugated antibody for 30 min at 4° and washed. For assessment of staining with anti-HLA-B8, a further incubation with streptavidin-FITC was carried out before washing and analysis by flow cytometry. All three antibodies gave clear-cut positive or negative staining of ~ 100% of gated lymphocytes in comparison with isotype control antibodies (data not shown), permitting definitive identification of HLA-A2+, -B7+ or -B8+ subjects.

HLA pentamer staining

To determine what proportion of CD8+ T cells were specific for common immunogenic epitopes of CMV antigens, HLA pentamers were used. PBMC were labelled with one of three APC-conjugated HLA pentamers, depending upon donor HLA type. Pentamers used were: HLA-A2+ CMV pp65495–503 (NLV), HLA-B7+ CMV pp65417–426 (TPR) and HLA-B8+ CMV IE188–96 (QIK), all conjugated to APC (Proimmune, Oxford, UK). Cells were incubated with pentamer at a dilution of 1 : 20 for 10 min in the dark at 20° followed by washing and subsequent incubation with anti-CD56-PE and anti-CD8-FITC (Beckman Coulter) for 30 min at 4°. Lymphocytes were gated on CD8bright cells and these were further gated on CD56+ and CD56 populations. Results were expressed as % of CD8bright CD56+ or CD8bright CD56 cells positive for each pentamer.

Statistical analysis

For comparisons between two groups of data, a paired Student's t-test was used (StatsDirect) with P < 0·05 being regarded as statistically significant.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Increased % of CD56+ T cells in CMV+ subjects

The PBMC were prepared from a panel of healthy subjects previously screened for anti-CMV antibodies. Cells were dual labelled with anti-CD3 and anti-CD56 and analysed by flow cytometry. Lymphocytes were gated on the basis of forward and side scatter and the % of double-positive lymphocytes was expressed as a % of total lymphocytes and also as % of total T cells. There were significantly greater proportions of CD56+ T cells in CMV-seropositive healthy subjects (n = 32) than CMV-seronegative subjects (n = 22; Fig. 1; P < 0·0001), whether the results were expressed as ‘% total lymphocytes (Fig. 1b) or ‘% T cells’ (Fig. 1c). There was no difference in mean age between CMV+ and CMV subjects and no significant differences in % CD56+ NK cells or CD56 T cells were noted between CMV+ and CMV subjects (data not shown).


Figure 1. Proportions of CD56+ T cells in relation to cytomegalovirus (CMV) status in healthy subjects. (a) Two-colour flow cytometry profiles of gated lymphocytes stained with anti-CD3 and anti-CD56 from representative CMV and CMV+ healthy subjects. (b) Mean % CD56+ T cells ± SE from CMV (n = 22) and CMV+ (n = 32) healthy subjects expressed as % total lymphocytes. (c) As in (b) but expressed as % T cells. Open symbols = CMV and filled symbols = CMV+ subjects.

Download figure to PowerPoint

Phenotypic differences in CD56+ T cells between CMV+ and CMV subjects

To investigate phenotypic differences between CD56+ T cells from CMV+ and CMV subjects, three-colour staining of PBMC was carried out using a combination of anti-CD3-PerCP-Cy5·5, anti-CD56-APC and a panel of third antibodies conjugated to PE. Several cell surface markers were expressed by a significantly lower proportion of CD56+ T cells in CMV+ compared with CMV subjects (P < 0·05). These included: CD161, CD62L, CD28, CD127 and CCR7 (Fig. 2a). Other cell surface molecules, including: CD4, CD8, CD45RO, CD57, CD58, CD94 and NKG2C were present on a significantly increased proportion of CD56+ T cells in CMV+ compared with CMV subjects (Fig. 2b; P < 0·05). The majority of CD56+ T cells (> 60%) from CMV subjects were CD4 CD8 whereas in CMV+ subjects almost all were single CD4+ or CD8+ (Fig. 2b). CD94 was present on a higher % of cells than NKG2C, which was almost absent from CD56+ T cells from CMV subjects. This would indicate that a similar proportion of cells expressed NKG2A paired with CD94 in both CMV+ and CMV subjects. Other cell surface molecules tested but showing no significant differences between CMV+ and CMV subjects included CD45RA and CD11a (Fig. 2b), although mean fluorescence intensity of the latter was significantly higher on CD56+ T cells from CMV+ than CMV subjects (data not shown). Several other cell surface markers, including CD25, CD69 and TCR-γδ showed < 3% positive cells among CD56+ T cells with no significant differences between CMV and CMV+ subjects. Proportions of cells expressing CD16, TCRVα7.2 and several killer immunoglobulin-like receptors (CD158a, b, e and f) were somewhat higher at between 15 and 25% but also showed no significant difference in relation to CMV status (data not shown).


Figure 2. Phenotypic differences between CD56+ T cells from cytomegalovirus seropositive (CMV+) and CMV healthy subjects. (a) Cell surface markers which were significantly decreased in CMV+ subjects (filled bars) compared with CMV subjects (open bars). (b) Cell surface markers which were mostly increased in CMV+ subjects. Results are expressed as mean % of CD56+ T cells ± SE with P values above indicating level of significance between CMV and CMV+ subjects (n ≥ 10 in all cases). Open symbols = CMV and filled symbols = CMV+ subjects.

Download figure to PowerPoint

Functional responses in CD56+ T cells from CMV+ and CMV subjects

Expression of pro-inflammatory cytokines in response to stimulation with CMV antigens was tested in PBMC populations from CMV+ and CMV healthy subjects. The proportions of CD56+ T cells individually expressing each of the cytokines tested, IFN-γ, TNF-α and IL-2 were higher in cells from CMV+ compared with those from CMV subjects although in the case of IL-2 this was not statistically significant (Fig. 3a). Under the same experimental conditions, production of IFN-γ and TNF-α was also significantly higher in the major CD56 T-cell subpopulation as well as NK cells in CMV+ than CMV subjects (Fig. 3b,c). There were no significant differences in cytokine production by any cell population between CMV+ and CMV subjects in response to the polyclonal stimulator staphylococcal enterotoxin B (data not shown).


Figure 3. Expression of intracellular cytokines following stimulation with cytomegalovirus (CMV) antigens. (a) CD56+ T cells; (b) CD56 T cells; (c) natural killer (NK) cells from CMV+ (filled bars) and CMV subjects (open bars). Results are expressed as mean % positive cells ± SE with P values above indicating level of significance between CMV and CMV+ subjects (n ≥ 10 in all cases). Open symbols = CMV and filled symbols = CMV+ subjects.

Download figure to PowerPoint

Following stimulation of PBMC with the NK and cytokine-induced killer cell target K562, or with CMV antigen extract, CD107a was detectable on a significantly higher proportion of CD56+ T cells from CMV+ than from CMV subjects (Fig. 4). For comparison, results for NK cells and CD56 T cells are also shown. NK cells showed similar differences between CMV+ and CMV subjects whereas CD56 T cells showed very low levels of CD107a expression following stimulation with K562 (Fig. 4a). However, following stimulation with CMV antigens, the majority CD56 T-cell population showed an increase in CD107a expression, although this was not as pronounced as in CD56+ T cells and did not reach statistical significance (Fig. 4b).


Figure 4. Cell surface exposure of CD107a in response to stimuli. Cells were incubated with K562 cells (a) or cytomegalovirus (CMV) antigen extract (b) and gated lymphocyte populations from CMV+ and CMV subjects analysed. Results are expressed as mean % of natural killer (NK) cells, CD56+ T cells and CD56 T cells expressing CD107a ± SE with the P value above indicating level of significance between CMV (open bars) and CMV+ subjects (filled bars; n = 10). Open symbols = CMV and filled symbols = CMV+ subjects.

Download figure to PowerPoint

Proliferative responses of CD56+ T cells to CMV antigens

The proliferative responses of CD56+ T cells from CMV+ subjects following stimulation with CMV antigens were investigated using CFSE labelling and flow cytometry. After 7 days of antigen stimulation, significant proliferation of both CD56 and CD56+ T cells as well as CD3 NK cells was found, although not to such a great extent as with the polyclonal mitogen staphylococcal enterotoxin B (Fig. 5a). The % of CD56+ T cells that had proliferated was similar to that for the majority CD56 T-cell population (Fig. 5b).


Figure 5. Proliferation of CFSE-labelled natural killer (NK) cells and CD56+ and CD56 T cells from a cytomegalovirus seropositive (CMV+) subject. (a) Cells were incubated for 7 days with culture medium alone, CMV antigen extract, or staphylococcal enterotoxin B (SEB) as a positive control; representative data from one of five experiments. (b) Mean % proliferation of unstimulated (open columns) and CMV-stimulated (filled columns) NK cells, CD56+ and CD56 T cells from CMV+ subjects (n = 5). Results are expressed as % cells with reduced CFSE staining ± SEM. *P < 0·05. Open symbols = CMV and filled symbols = CMV+ subjects.

Download figure to PowerPoint

Antigen specificity of CD8+ CD56+ T cells in CMV+ subjects

HLA pentamers comprising three common HLA types together with immunodominant 9mer or 10mer CMV peptides were used to investigate antigen specificity of CD56+ T cells. Healthy CMV+ subjects were HLA typed using mAb against HLA-A2, HLA-B7 and HLA-B8 and PBMC labelled with anti-CD56, anti-CD8 and HLA pentamer for flow cytometric analysis. Lymphocytes were gated for CD8bright+ cells and both CD56+ and CD56 populations were analysed for HLA pentamer staining. In all subjects, HLA pentamer stained cells were detected in both CD8+ CD56+ and CD8+ CD56 populations (Fig. 6). There was no clear preponderance of either CD56+ or CD56 pentamer+ cells with some subjects having a greater proportion of CD56+ cells and others more CD56 cells. However, for all four HLA-B8+ subjects tested, CD56+ cells made up the majority of pentamer+ cells (Fig. 6b). Several subjects were positive for HLA-A2 and either HLA-B7 or HLA-B8 and in each case there were approximately 0·5–2·5% of pentamer+ cells of both HLA types. However, one HLA-A2+ and HLA-B8+ subject had between 4% and 15% of CD8+ cells positive for each pentamer so that almost 20% of CD8+ T cells were specific for the two immunodominant peptides used. The results overall suggest that CD56+ T cells comprise a similar proportion of the antigen-specific CD8+ T-cell population to CD56 T cells.


Figure 6. HLA pentamer staining of cytomegalovirus (CMV) -specific CD8+ CD56+ and CD8+ CD56 T cells from CMV+ healthy subjects. (a) Representative staining of gated CD8+ CD56+ T cells with HLA-A*0201-NLV, HLA-B*0702-TPR and HLA-B*0801-QIK pentamers; percentages indicate the % of total CD8+ CD56+ T cells. FSc = forward scatter. (b) % HLA pentamer staining of CD56 CD8+ and CD56+ CD8+ T cells from several healthy CMV+ subjects; lines join data for the same subject.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Relative numbers of CD56+ T cells were found to be significantly higher in CMV+ than in CMV healthy subjects. Phenotypic analysis showed that there were increased proportions of CD56+ T cells negative for CD161, CD62L, CD28, CD127 and CCR7 in CMV+ subjects. The CD62L CD28 CCR7 phenotype is consistent with some of the cells being effector memory T cells rather than central memory T cells,[8] and also cytokine-induced killer cells generated following treatment with IFN-γ, IL-2 and anti-CD3 in vitro[7] but is also similar to the phenotype of senescent cells[32] or effector memory CD45RA+ T cells,[8] which have been found to be increased in elderly CMV+ subjects.[33] The decreased proportion of CD161+ cells in CD56+ T cells from CMV+ subjects would argue against any virus-associated expansion of T helper type 17 cells, invariant NKT cells or mucosa-associated invariant T cells, all of which are CD161+.[34, 9, 35] Interestingly, although CD161 has been reported to be increased on NK cells in CMV+ children,[36] it is absent from CMV-specific cytotoxic T lymphocytes.[37] Loss of CD127 expression also characterizes an effector population of CD8+ T cells in HIV-infected patients.[38]

In contrast, several cell surface markers were found on an increased proportion of CD56+ T cells in CMV+ subjects, including CD4, CD8, CD45RO, CD57, CD58, CD94 and NKG2C. Previous studies have found that around 70% of CD56+ T cells are CD8+;[1, 8] this is similar to findings for CMV+ subjects in this study whereas in CMV subjects a significantly lower proportion of CD56+ T cells co-expressed CD8. This discrepancy may be because previous work has used predominantly CMV+ healthy subjects. Some cells had a phenotype resembling that of a recently described population of CD8+ regulatory T cells, which are CD8+ CD56+ CD161 CD45RA+ CD45ROCD28 CCR7 but, unlike the population increased in CMV+ subjects, also express CD62L.[39] Previous work has shown increased numbers of both CD4+ leucocyte function antigen-1hi T cells[40] and CD4+ CD28 and CD8+ CD28 T cells in CMV+ subjects[41] and CD57 is well known to be increased on CD8+ T cells following infection with CMV and other viruses.[42] Some strains of CMV have long been known to up-regulate CD58 on infected target cells.[43]

As well as being increased on NK cells from CMV+ subjects,[44, 45] NKG2C is also expressed by a higher percentage of CD56+ T cells in CMV+ compared with CMV adults[45, 46] and children.[36] The ligand for NKG2C is HLA-E, which is bound by the leader sequence from the CMV-encoded UL40 protein,[47] although the inhibitory ligand NKG2A has a much higher affinity for HLA-E. Culture with CMV-infected cells led to preferential expansion not only of NKG2C+ over NKG2A+ NK cells but also of CD56+ T cells[48] so that the relative balance would be shifted towards activation of cytotoxic function following CMV infection. However, in another report the relative balance between activating and inhibitory CD94-associated receptors on CD8+ T cells was unchanged following CMV infection.[49] In the present experiments, the induction of NKG2C expression by CD56+ T cells following CMV infection may enhance overall cytotoxic capacity, depending upon continued expression of HLA-E by CMV-infected target cells.

Both the pro-inflammatory cytokines IFN-γ and TNF-α were produced by a higher proportion of CD56+ T cells following stimulation with CMV antigens in CMV+ compared with CMV subjects, possibly reflecting a CMV-specific memory response. Non-specifically stimulated liver CD56+ T cells were found to produce the T helper type 1 cytokines IFN-γ, TNF-α and IL-2,[50] as did those in hepatitis B-related liver disease.[51] However, CMV antigens were able to induce IFN-γ expression by CD4+ CD8dim CD56dim T cells[52] and CMV-specific HLA-E restricted CD8+ CD28 CD45RA+ CD56+ T cells also produced IFN-γ.[30] In the present experiments, CMV antigens induced substantial proliferation of CD56+ T cells from CMV+ subjects to almost the same extent as CD56 T cells.

CD107a expression was induced on a significantly higher proportion of CD56+ T cells from CMV+ than CMV subjects following stimulation with the NK and cytokine-induced killer cell target K562 and also with CMV antigens. This is supportive of the concept that CD56+ cells comprise a major cytotoxic T-cell population in CMV+ subjects. CD107a has previously been reported to be induced in pro-inflammatory cytokine-producing CD8+ T cells from stem cell transplant patients recovering from CMV infection in response to CMV antigen.[53] Interestingly, CD3 CD56+ NK cells from CMV+ subjects also showed significantly greater induction of CD107a expression following stimulation with both K562 cells and CMV antigens, the latter being suggestive of a degree of immunological memory to CMV.

HLA pentamers were used to show that a substantial proportion of the CD8+ CD56+ T cells were specific for immunodominant CMV peptides in healthy subjects of three different HLA types. In the case of HLA-B8+ subjects, a higher proportion of QIK-specific T cells were CD56+ than CD56 in all four subjects tested. CD56+ T cells therefore comprised a major component of the antigen-specific CD8+ T-cell population, although this was not statistically significant. Indeed, CD56 expression has been correlated with anti-CD3-induced cytotoxic function in CD8+ T cells from healthy subjects[54] and in human papillomavirus-specific CD8+ T cells.[55] CD56+ T cells have been implicated as effector cells in protection against HIV,[56] hepatitis C virus[57] and Japanese encephalitis virus,[58] as well as the autoimmune inflammatory condition Behçet's disease.[59] Also, CD4+ CD28 CD56+ cells were expanded in patients with chronic hepatitis B virus infection[60] and mast cell stimulation by reovirus was found to lead to recruitment of CD56+ T cells in vitro.[61]

With regard to CMV-associated immunosenescence,[25] proportions of CD56+ T cells were found to increase progressively with age in vivo and also following IL-2-dependent T-cell proliferation in vitro,[62] probably derived from CD56 precursors.[8] A recent report has shown that CD8+ CD45RA+ T cells with lower affinity for CMV peptide are expanded in older subjects and also in vitro in response to CMV antigens or IL-15.[63] Purification of CMV class I HLA tetramer+ CD8+ T cells and expansion in IL-2 generated high numbers of CD56+ cytokine-induced killer cells that retained anti-CMV-specific cytotoxic function.[6] This is strongly suggestive of CD56+ T cells contributing to the cytotoxic cell response to CMV.

The precise role played by the CD56 molecule in specific or non-specific cytotoxicity by CD56+ T cells is unclear. CD56 can exert homotypic adhesion and cold target inhibition with CD56+ cells inhibited cytotoxic function of CD56+ T cells.[64] However, this interaction would only be relevant to lysis of CD56+ cells, which comprise cells of neuronal origin as well as NK cells and CD56+ T cells themselves. There is some evidence for a role for fibroblast growth factor receptor-1 as a ligand for CD56.[65, 66] This receptor is widely expressed on cells of mesodermal, endodermal and ectodermal origins,[67] which extensively overlaps the range of cellular targets of cytomegalovirus.[68] Future work will investigate in more detail the cytotoxic capacity of CD56+ T cells in CMV infection with reference to the role of the CD56 molecule. It would also be of interest to follow changes in CD56+ T-cell numbers and phenotype during the course of a primary CMV infection as well as in elderly patients with the immune risk profile.[25]


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

The work was funded by a grant from the Cultural Bureau of Saudi Arabia. MA conducted the experiments and analysed the results, BFF and NK advised on experimental strategy and suggested improvements to the manuscript, SA advised on experimental techniques and suggested improvements to the manuscript and SEC designed the research and wrote the paper.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  • 1
    Lanier LL, Le AM, Civin CI, Loken MR, Phillips JH. The relationship of CD16 (Leu-11) and Leu-19 (NKH-1) antigen expression on human peripheral blood NK cells and cytotoxic T lymphocytes. J Immunol 1986; 136:44806.
  • 2
    Lanier LL, Testi R, Bindl J, Phillips JH. Identity of Leu-19 (CD56) leukocyte differentiation antigen and neural cell adhesion molecule. J Exp Med 1989; 169:22338.
  • 3
    Lu PH, Negrin RS. A novel population of expanded human CD3+CD56+ cells derived from T cells with potent in vivo antitumor activity in mice with severe combined immunodeficiency. J Immunol 1994; 153:168796.
  • 4
    Pilla L, Squarcina P, Coppa J et al. Natural killer and NK-Like T-cell activation in colorectal carcinoma patients treated with autologous tumor-derived heat shock protein 96. Cancer Res 2005; 65:39429.
  • 5
    Schmidt-Wolf IG, Lefterova P, Mehta BA, Fernandez LP, Huhn D, Blume KG, Weissman IL, Negrin RS. Phenotypic characterization and identification of effector cells involved in tumor cell recognition of cytokine-induced killer cells. Exp Hematol 1993; 21:16739.
  • 6
    Pievani A, Borleri G, Pende D, Moretta L, Rambaldi A, Golay J, Introna M. Dual-functional capability of CD3+CD56+ CIK cells, a T-cell subset that acquires NK function and retains TCR-mediated specific cytotoxicity. Blood 2011; 118:330110.
  • 7
    Satoh M, Seki S, Hashimoto W, Ogasawara K, Kobayashi T, Kumagai K, Matsuno S, Takeda K. Cytotoxic γδ or αβ T cells with a natural killer cell marker, CD56, induced from human peripheral blood lymphocytes by a combination of IL-12 and IL-2. J Immunol 1996; 157:388692.
  • 8
    Franceschetti M, Pievani A, Borleri G, Vago L, Fleischhauer K, Golay J, Introna M. Cytokine-induced killer cells are terminally differentiated activated CD8 cytotoxic T-EMRA lymphocytes. Exp Hematol 2009; 37:61628.
  • 9
    O'Reilly V, Zeng SG, Bricard G et al. Distinct and overlapping effector functions of expanded human CD4+, CD8α+ and CD4-CD8α invariant natural killer T cells. PLoS One 2011; 6:e28648.
  • 10
    Karadimitris A, Patterson S, Spanoudakis E. Natural killer T cells and haemopoiesis. Br J Haematol 2006; 134:26372.
  • 11
    Takayama E, Koike Y, Ohkawa T et al. Functional and Vbeta repertoire characterization of human CD8+ T-cell subsets with natural killer cell markers, CD56+ CD57 T cells, CD56+ CD57+ T cells and CD56 CD57+ T cells. Immunology 2003; 108:2119.
  • 12
    Lemster BH, Michel JJ, Montag DT, Paat JJ, Studenski SA, Newman AB, Vallejo AN. Induction of CD56 and TCR-independent activation of T cells with aging. J Immunol 2008; 180:197990.
  • 13
    Watanabe H, Weerasinghe A, Miyaji C et al. Expansion of unconventional T cells with natural killer markers in malaria patients. Parasitol Int 2003; 52:6170.
  • 14
    Koreck A, Surányi A, Szöny BJ, Farkas A, Bata-Csörgö Z, Kemény M, Dobozy A. CD3+CD56+ NK T cells are significantly decreased in the peripheral blood of patients with psoriasis. Clin Exp Immunol 2002; 127:17682.
  • 15
    Green MR, Kennell AS, Larche MJ, Seifert MH, Isenberg DA, Salaman MR. Natural killer T cells in families of patients with systemic lupus erythematosus: their possible role in regulation of IgG production. Arthritis Rheum 2007; 56:30310.
  • 16
    Bergström I, Backteman K, Lundberg A, Ernerudh J, Jonasson L. Persistent accumulation of interferon-γ-producing CD8+CD56+ T cells in blood from patients with coronary artery disease. Atherosclerosis 2012; 224:51520.
  • 17
    Al Omar SY, Marshall E, Middleton D, Christmas SE. Increased numbers but functional defects of CD56+CD3+ cells in lung cancer. Int Immunol 2012; 24:40915.
  • 18
    Weng PJ, Ying H, Hong LZ, Zhou WH, Hu YR, Xu CH. An analysis of CD3+CD56+ lymphocytes and their subsets in the peripheral blood of patients with chronic hepatitis B. Zhonghua Gan Zang Bing Za Zhi 2008; 16:6546.
  • 19
    Srivastava R, Aggarwal R, Bhagat MR, Chowdhury A, Naik S. Alterations in natural killer cells and natural killer T cells during acute viral hepatitis E. J Viral Hepat 2008; 15:9106.
  • 20
    Fu GF, Chen X, Hu HY et al. Emergence of peripheral CD3+CD56+ cytokine-induced killer cell in HIV-1-infected Chinese children. Int Immunol 2012; 24:197206.
  • 21
    Cannon MJ, Schmid DS, Hyde TB. Review of cytomegalovirus seroprevalence and demographic characteristics associated with infection. Rev Med Virol 2010; 20:20213.
  • 22
    Boeckh M, Geballe AP. Cytomegalovirus: pathogen, paradigm, and puzzle. J Clin Invest 2011; 121:167380.
  • 23
    Moss P, Khan N. CD8+ T-cell immunity to cytomegalovirus. Hum Immunol 2004; 65:45664.
  • 24
    Biron CA, Byron KS, Sullivan JL. Severe herpesvirus infections in an adolescent without natural killer cells. N Engl J Med 1989; 320:17315.
  • 25
    Pawelec G, Derhovanessian E. Role of CMV in immune senescence. Virus Res 2011; 157:1759.
  • 26
    Labalette M, Salez F, Pruvot FR, Noel C, Dessaint JP. CD8 lymphocytosis in primary cytomegalovirus (CMV) infection of allograft recipients: expansion of an uncommon CD8+ CD57 subset and its progressive replacement by CD8+ CD57+ T cells. Clin Exp Immunol 1994; 95:46571.
  • 27
    van den Berg AP, van Son WJ, Janssen RA et al. Recovery from cytomegalovirus infection is associated with activation of peripheral blood lymphocytes. J Infect Dis 1992; 166:122835.
  • 28
    Looney RJ, Falsey A, Campbell D et al. Role of cytomegalovirus in the T cell changes seen in elderly individuals. Clin Immunol 1999; 90:2139.
  • 29
    Faber C, Singh A, Krüger Falk M, Juel HB, Sørensen TL, Nissen MH. Age-related macular degeneration is associated with increased proportion of CD56+ T cells in peripheral blood. Ophthalmology 2013; 120:23106.
  • 30
    Mazzarino P, Pietra G, Vacca P, Falco M, Colau D, Coulie P, Moretta L, Mingari MC. Identification of effector-memory CMV-specific T lymphocytes that kill CMV-infected target cells in an HLA-E-restricted fashion. Eur J Immunol 2005; 35:32407.
  • 31
    Suni MA, Picker LJ, Maino VC. Detection of antigen-specific T cell cytokine expression in whole blood by flow cytometry. J Immunol Methods 1998; 212:8998.
  • 32
    Tarazona R, DelaRosa O, Alonso C, Ostos B, Espejo J, Peña J, Solana R. Increased expression of NK cell markers on T lymphocytes in aging and chronic activation of the immune system reflects the accumulation of effector/senescent T cells. Mech Ageing Dev 2000; 121:7788.
  • 33
    Derhovanessian E, Maier AB, Hähnel K, Beck R, de Craen AJ, Slagboom EP, Westendorp RG, Pawelec G. Infection with cytomegalovirus but not herpes simplex virus induces the accumulation of late-differentiated CD4+ and CD8+ T-cells in humans. J Gen Virol 2011; 92:274656.
  • 34
    Cosmi L, De Palma R, Santarlasci V et al. Human interleukin 17-producing cells originate from a CD161+CD4+ T cell precursor. J Exp Med 2008; 205:190316.
  • 35
    Dusseaux M, Martin E, Serriari N et al. Human MAIT cells are xenobiotic-resistant, tissue-targeted, CD161hi IL-17secreting T cells. Blood 2011; 117:12509.
  • 36
    Monsiváis-Urenda A, Noyola-Cherpitel D, Hernández-Salinas A, García-Sepúlveda C, Romo N, Baranda L, López-Botet M, González-Amaro R. Influence of human cytomegalovirus infection on the NK cell receptor repertoire in children. Eur J Immunol 2010; 40:141827.
  • 37
    Northfield JW, Kasprowicz V, Lucas M et al. CD161 expression on hepatitis C virus-specific CD8+ T cells suggests a distinct pathway of T cell differentiation. Hepatology 2008; 47:396406.
  • 38
    Paiardini M, Cervasi B, Albrecht H et al. Loss of CD127 expression defines an expansion of effector CD8+ T cells in HIV-infected individuals. J Immunol 2005; 174:29009.
  • 39
    Hu D, Weiner HL, Ritz J. Identification of Cytolytic CD161CD56+ Regulatory CD8 T Cells in Human Peripheral Blood. PLoS One 2013; 8:e59545.
  • 40
    Chidrawar S, Khan N, Wei W, McLarnon A, Smith N, Nayak L, Moss P. Cytomegalovirus-seropositivity has a profound influence on the magnitude of major lymphoid subsets within healthy individuals. Clin Exp Immunol 2009; 155:42332.
  • 41
    Kuparinen T, Marttila S, Jylhävä J, Tserel L, Peterson P, Jylhä M, Hervonen A, Hurme M. Cytomegalovirus (CMV)-dependent and -independent changes in the aging of the human immune system: a transcriptomic analysis. Exp Gerontol 2003; 48:30512.
  • 42
    Ibegbu CC, Xu YX, Harris W, Maggio D, Miller JD, Kourtis AP. Expression of killer cell lectin-like receptor G1 on antigen-specific human CD8+ T lymphocytes during active, latent, and resolved infection and its relation with CD57. J Immunol 2005; 174:608894.
  • 43
    Grundy JE, Downes KL. Up-regulation of LFA-3 and ICAM-1 on the surface of fibroblasts infected with cytomegalovirus. Immunology 1993; 78:40512.
  • 44
    Lopez-Vergès S, Milush JM, Schwartz BS et al. Expansion of a unique CD57⁺NKG2Chi natural killer cell subset during acute human cytomegalovirus infection. Proc Natl Acad Sci U S A 2011; 108:1472532.
  • 45
    Gumá M, Angulo A, Vilches C, Gómez-Lozano N, Malats N, López-Botet M. Imprint of human cytomegalovirus infection on the NK cell receptor repertoire. Blood 2004; 104:366471.
  • 46
    Romo N, Fitó M, Gumá M et al. Association of atherosclerosis with expression of the LILRB1 receptor by human NK and T-cells supports the infectious burden hypothesis. Arterioscler Thromb Vasc Biol 2011; 31:231421.
  • 47
    Tomasec P, Braud VM, Rickards C et al. Surface expression of HLA-E, an inhibitor of natural killer cells, enhanced by human cytomegalovirus gpUL40. Science 2000; 287:10313.
  • 48
    Gumá M, Budt M, Sáez A, Brckalo T, Hengel H, Angulo A, López-Botet M. Expansion of CD94/NKG2C+ NK cells in response to human cytomegalovirus-infected fibroblasts. Blood 2006; 107:362431.
  • 49
    van Stijn A, Rowshani AT, Yong SL, Baas F, Roosnek E, ten Berge IJ, van Lier RA. Human cytomegalovirus infection induces a rapid and sustained change in the expression of NK cell receptors on CD8+ T cells. J Immunol 2008; 180:455060.
  • 50
    Doherty DG, Norris S, Madrigal-Estebas L, McEntee G, Traynor O, Hegarty JE, O'Farrelly C. The human liver contains multiple populations of NK cells, T cells, and CD3+CD56+ natural T cells with distinct cytotoxic activities and Th1, Th2, and Th0 cytokine secretion patterns. J Immunol 1999; 163:231421.
  • 51
    Barnaba V, Franco A, Paroli M et al. Selective expansion of cytotoxic T lymphocytes with a CD4+CD56+ surface phenotype and a T helper type 1 profile of cytokine secretion in the liver of patients chronically infected with Hepatitis B virus. J Immunol 1994; 152:307487.
  • 52
    Suni MA, Ghanekar SA, Houck DW, Maecker HT, Wormsley SB, Picker LJ, Moss RB, Maino VC. CD4+CD8dim T lymphocytes exhibit enhanced cytokine expression, proliferation and cytotoxic activity in response to HCMV and HIV-1 antigens. Eur J Immunol 2001; 31:251220.
  • 53
    Muñoz-Cobo B, Solano C, Benet I et al. Functional profile of cytomegalovirus (CMV)-specific CD8+ T cells and kinetics of NKG2C+ NK cells associated with the resolution of CMV DNAemia in allogeneic stem cell transplant recipients. J Med Virol 2012; 84:25967.
  • 54
    Pittet MJ, Speiser DE, Valmori D, Cerottini JC, Romero P. Cutting edge: cytolytic effector function in human circulating CD8 T cells closely correlates with CD56 surface expression. J Immunol 2000; 164:114852.
  • 55
    Santin AD, Hermonat PL, Ravaggi A et al. Expression of CD56 by human papillomavirus E7-specific CD8+ cytotoxic T lymphocytes correlates with increased intracellular perforin expression and enhanced cytotoxicity against HLA-A2-matched cervical tumor cells. Clin Cancer Res 2001; 7:804s10s.
  • 56
    Hou W, Ye L, Ho WZ. CD56+ T cells inhibit HIV-1 infection of macrophages. J Leukoc Biol 2012; 92:34351.
  • 57
    Doskali M, Tanaka Y, Ohira M, Ishiyama K, Tashiro H, Chayama K, Ohdan H. Possibility of adoptive immunotherapy with peripheral blood-derived CD3⁻CD56+ and CD3+CD56+ cells for inducing antihepatocellular carcinoma and antihepatitis C virus activity. J Immunother 2011; 34:12938.
  • 58
    Sooryanarain H, Ayachit V, Gore M. Activated CD56+ lymphocytes (NK+NKT) mediate immunomodulatory and anti-viral effects during Japanese encephalitis virus infection of dendritic cells in vitro. Virology 2012; 432:25060.
  • 59
    Ahn JK, Chung H, Lee DS, Yu YS, Yu HG. CD8brightCD56+ T cells are cytotoxic effectors in patients with active Behçet's uveitis. J Immunol 2005; 175:613342.
  • 60
    Wang Y, Bai J, Li F et al. Characteristics of expanded CD4+CD28null T cells in patients with chronic hepatitis B. Immunol Invest 2009; 38:43446.
  • 61
    McAlpine SM, Issekutz TB, Marshall JS. Virus stimulation of human mast cells results in the recruitment of CD56⁺ T cells by a mechanism dependent on CCR5 ligands. FASEB J 2012; 26:12809.
  • 62
    Brzezinska A. Does in vitro replicative senescence of human CD8+ cells reflect the phenotypic changes observed during in vivo ageing? Acta Biochim Pol 2005; 52:9315.
  • 63
    Griffiths SJ, Riddell NE, Masters J et al. Age-associated increase of low-avidity cytomegalovirus-specific CD8+ T cells that re-express CD45RA. J Immunol 2013; 190:536372.
  • 64
    Vergelli M, Le H, van Noort JM, Dhib-Jalbut S, McFarland H, Martin R. A novel population of CD4+CD56+ myelin-reactive T cells lyses target cells expressing CD56/neural cell adhesion molecule. J Immunol 1996; 157:67988.
  • 65
    Kos FJ, Chin CS. Costimulation of T cell receptor-triggered IL-2 production by Jurkat T cells via fibroblast growth factor receptor 1 upon its engagement by CD56. Immunol Cell Biol 2002; 80:3649.
  • 66
    Chan A, Hong DL, Atzberger A et al. CD56bright human NK cells differentiate into CD56dim cells: role of contact with peripheral fibroblasts. J Immunol 2007; 179:8994.
  • 67
    Coutts JC, Gallagher JT. Receptors for fibroblast growth factors. Immunol Cell Biol 1995; 73:5849.
  • 68
    Adler B, Sinzger C. Endothelial cells in human cytomegalovirus infection: one host cell out of many or a crucial target for virus spread? Thromb Haemost 2009; 102:105763.