SEARCH

SEARCH BY CITATION

Keywords:

  • Human;
  • Immunodeficiency;
  • Memory;
  • Rodent;
  • T cells

Abstract

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

Deficiency in the guanine nucleotide exchange factor dedicator of cytokinesis 8 (DOCK8) causes a human immunodeficiency syndrome associated with recurrent sinopulmonary and viral infections. We have recently identified a DOCK8-deficient mouse strain, carrying an ethylnitrosourea-induced splice-site mutation that shows a failure to mature a humoral immune response due to the loss of germinal centre B cells. In this study, we turned to T-cell immunity to investigate further the human immunodeficiency syndrome and its association with decreased peripheral CD4+ and CD8+ T cells. Characterisation of the DOCK8-deficient mouse revealed T-cell lymphopenia, with increased T-cell turnover and decreased survival. Egress of mature CD4+ thymocytes was reduced with increased migration of these cells to the chemokine CXCL12. However, despite the two-fold reduction in peripheral naïve T cells, the DOCK8-deficient mice generated a normal primary CD8+ immune response and were able to survive acute influenza virus infection. The limiting effect of DOCK8 was in the normal survival of CD8+ memory T cells after infection. These findings help to explain why DOCK8-deficient patients are susceptible to recurrent infections and provide new insights into how T-cell memory is sustained.


Introduction

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

The recent discovery of the human dedicator of cytokinesis 8 (DOCK8) immunodeficiency syndrome has highlighted DOCK8 and the DOCK super-family of guanine nucleotide exchange factors as important mediators of normal immune function 1–4. As a group, members of the DOCK family interact with Rho GTPases to regulate pathways that are typically involved in the cytoskeletal rearrangements required to maintain cell structure, migration and adhesion; but the details of their individual functions are largely unknown 3, 5. The DHR1 (CZH1) domain in DOCK proteins has been shown to bind phosphatidylinositol (3,4,5)-triphosphate (PIP3) generated by PI3K activity, thus promoting localised membrane binding and activation of DOCK proteins 3, while the DHR2 (CZH2) domain binds to Rho-family GTPase proteins (Rac, Cdc42) and provides catalytic GEF activity to stimulate these proteins and promote integrin reorganisation and adhesion, lamellipodia formation, cell polarisation, phagocytosis or cell fusion 3, 4. DOCK8 had been associated with Cdc42 and found to localise to lamellipodia in fibroblasts; but its role in immune function is largely unknown 6.

Recently, described patients who are deficient in DOCK8 fail to mount an effective immune response and are susceptible to recurrent sinopulmonary and viral infections, hyper-IgE, atopy and allergic disease and some types of cutaneous cancer and lymphoma 1, 2. The majority of patients have low absolute counts of both CD4+ and CD8+ T cells in the blood and impaired proliferation of the cells following mitogen stimulation with a modest decrease in the production of the anti-viral cytokines IFN-γ and TNF-α. These facts suggest a defect in T-cell immunity; yet, the nature of any T-cell dysfunction in the absence of DOCK8 remains to be defined.

DOCK8 was also recently identified as being critical for a functional B-cell response in a screen for N-ethyl-N-nitrosourea (ENU)-induced mutations that cripple the maturation, but not initiation, of antibody production 7. Two independent strains (CPM and pri) were found to have loss-of-function DOCK8 mutations causing the cell intrinsic loss of germinal centre (GC) B cells during affinity maturation 7. Further analysis showed that naïve DOCK8-deficient cells are unable to recruit the intercellular adhesion molecule ICAM-1 into the B-cell immune synapse during recognition of membrane-bound antigen. This deficit may explain the impact on GC B-cell survival, which depends on binding antigen and integrin ligands on follicular DCs 8.

In this study, we use the DOCK8-deficient mice to gain insights into the clinical and laboratory findings of DOCK8 immunodeficiency in humans, extending our study of B-cell immunity to examine the T-cell compartment. Like humans, DOCK8-deficient mice have decreased peripheral T cells. Although development of T cells is normal, mature single-positive CD4+ T cells accumulate in the thymus and show increased chemotaxis to CXCL12; peripheral CD4+ and CD8+ T cells show increased turnover and decreased survival. Despite these defects, DOCK8-deficient animals are able to generate a comparatively normal primary T-cell response to immunisation with Modified Vaccinia Ankara (MVA) or infection with influenza. However, the memory recall response is significantly reduced due to decreased survival of memory CD8+ T cells. These results may explain the relentless pattern of recurrent infection in the human DOCK8 immunodeficiency syndrome, and may provide clues to how T-cell memory is generated and sustained.

Results

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

T-cell lymphopenia in DOCK8-deficient humans and mice

To establish the effects of DOCK8 deficiency on T cells in humans, we counted naïve CD45RA and activated effector/memory CD45RO CD4+ and CD8+ T cells in the blood of patients of different ages. This showed a reduction in both naïve and effector/memory numbers, although the latter were less affected and even normal in some patients (Fig. 1A). However, it is difficult to interpret the significance of these results or establish the causes of immunodeficiency when T-cell phenotype is continually altered by persistent infection and antigenic stimulation. Therefore, we turned to a recently described ENU mutant, CPM, carrying homozygous null alleles of DOCK8 7. The CPM splice-site mutation truncates the native protein, removing the DHR2 GEF domain. To investigate the effects of DOCK8 deficiency on T cells, we enumerated CD4+ and CD8+ T cells in the secondary lymph organs of DOCK8-deficient CPM mice (Fig. 1B–E).

thumbnail image

Figure 1. T-cell lymphopenia in DOCK8-deficient patients and mice. (A) Absolute numbers of total naïve (CD45RA+) and effector/memory (CD45RO+) CD4+ and CD8+ T cells in the blood of DOCK8-deficient patients. Unique symbols represent individual patients over time; grey bars show the normal 10–90 percentiles for age. (B–D) Absolute numbers of total CD4+ and CD8+ T cells and activated CD44+ subsets in the peripheral lymphoid organs of WT and CPM mice. Dots represent individual animals aged 9–14 wks. Bars show arithmetic means and SE. (E) The percentage of CD4+ and CD8+ T cells in the blood of WT and CPM mice was calculated as a fraction of total lymphocytes (top). Histograms show typical surface CD44 expression on CD4+ lymphocytes in the blood of WT (filled histogram) and CPM mice (solid line histogram) (bottom). (B–E) *p<0.05, **p<0.01, ***p<0.001, t-test.

Download figure to PowerPoint

Examination of spleen, mesenteric lymph nodes (MLNs) and peripheral lymph nodes (PLNs) revealed a 2–3-fold reduction in CD4+ and CD8+ T cells compared with WT C57BL/6 controls (Fig. 1B–D). Total CD4+ and CD8+ T cells were reduced in the blood as a fraction of total lymphocytes (Fig. 1E, upper panel) and in absolute terms, where CD4+ counts were 5.28×105/mL (± SE 0.541) in CPM versus 14.53×105/mL (± SE 2.132) in WT (p=0.014), and CD8+ 3.32×105/mL (± SE 0.240) in CPM versus 5.62×105/mL (± SE 0.404) in WT (p=0.008). In contrast, the absolute number of activated CD44hi/CD62Llo CD4+ and CD8+ T cells was more similar in the spleen and PLNs and 1.5–2-fold higher in the MLNs of CPM mice compared with that of WT (Fig. 1B–D). The increased fraction of activated CD44hi CD8+ and CD4+ T cells was also seen in the blood of CPM mice (Fig. 1E, lower panel and data not shown). We conclude that DOCK8 deficiency in mice leads to peripheral T-cell lymphopenia predominantly affecting the number of naïve rather than activated cells, reminiscent of the phenotype in the human immunodeficiency syndrome.

Normal activation and differentiation of peripheral DOCK8-deficient T cells

Studies in DOCK8-deficient patients have demonstrated decreased T-cell activation and proliferation, but are inconclusive regarding function due to the difficulty in isolating naïve T cells from the peripheral blood 1, 2. Therefore, we analysed the response of DOCK8-deficient T cells from CPM to antigenic stimulation to see whether loss was associated with a defect in activation or differentiation. Overnight activation of naïve DOCK8-deficient mouse T cells with anti-CD3 showed no difference in the upregulation of CD69 and CD25 on CPM and WT CD4+ and CD8+ T cells (Fig. 2A); moreover, when stimulated with anti-CD3 and anti-CD28 in vitro, CPM and WT CD4+ and CD8+ T cells underwent the same number of divisions, after five days, in culture (Fig. 2B). These findings show that DOCK8-deficient T cells from uninfected mice are able to respond normally to antigenic stimulation and proliferate freely in vitro.

thumbnail image

Figure 2. Normal T-cell activation and proliferation in DOCK8-deficient mice. (A) Histograms showing CD25 and CD69 expression on CPM and WT CD4+ (top) and CD8+ (bottom) splenocytes, stimulated overnight with plate-bound anti-CD3 (10 ug/mL), and unstimulated controls (open histograms). Analysis was gated on cells with an intact cell membrane using the LIVEDEAD® Cell Stain, and representative of three experiments, with at least three animals. (B) Proliferation of WT (top) or CPM (bottom) CD4+ or CD8+ T cells (filled histograms), stimulated with 10 μg/mL plate-bound anti-CD3 and 10 μg/mL soluble anti-CD28 for 5 days, compared with unstimulated controls (broken line). Analysis gated as in (A). (A, B) Data are representative of at least five replicates using separate mice.

Download figure to PowerPoint

To explore T-cell differentiation, we analysed the cytokines produced by activated/memory CD45RO CD4+ T cells from the blood of DOCK8-deficient patients, parents and controls stimulated ex vivo with PMA and ionomycin. A higher proportion of activated cells secreting TH2 cytokines IL-4 or IL-13 were observed in 3/5 and 4/5 DOCK8-deficient patients, respectively, a finding that may be linked to the phenotype of hyper-IgE in DOCK8 immunodeficiency, while the number of cells secreting IFN-γ was variable (Supporting Information Fig. 1A). However, the degree to which these phenotypes reflect ongoing infection, over-stimulation/exhaustion or other secondary effects caused by the immunodeficient state was again uncertain. Therefore, we asked whether CPM CD4+ T cells could differentiate normally under TH1 or TH2 polarising conditions in vitro. Under these conditions, Tbet and GATA3 were induced appropriately (Supporting Information Fig. 1B). IFN-γ production by TH1 cells was equivalent, and only the proportion of IL-4 secreting cells was higher in CPM mice under TH2 conditions (Supporting Information Fig. 1C and D). It is possible that DOCK8 and integrin-dependent effects on the signalling threshold of the TCR or other receptors, or small differences in the basal and induced levels of T-bet and GATA3, might contribute to such a bias in humans and mice. Overall, however, the differentiation of effector cell subsets is not intrinsically unimpaired and therefore unlikely to contribute to the state of immunodeficiency.

T-cell development in DOCK8 deficiency

We next asked whether abnormal thymic development might explain the reduction in peripheral T cells. We found no difference in the numbers of developing T cells at the double negative (DN) or double positive (DP) stage (Fig. 3A) or in the number of mature CD8+ T cells (Fig. 3A and 3B); and no consistent differences were seen in CD3, CD5, TCRβ, CD44 or CD25, staining (data not shown). The sole abnormality observed was an accumulation of mature single-positive CD4+ T cells, particularly the most mature CD4+CD62LhiCD69lo population (Fig. 3A, right panel, and 3B). On its own this may contribute to the peripheral lymphopenia, but cannot explain the deficiency of CD8+ T cells. Labelling CPM and WT mice with BrdU showed no difference in the absolute number of BrdU+ single-positive CD4+ T cells generated over a 40-h period (1.00×106±SE 0.14 in CPM versus 1.33×106±SE 0.18 in WT, p=0.1674; unpaired t-test), implying that the accumulation of mature DOCK8-deficient CD4+ T cells is due to delayed egress.

thumbnail image

Figure 3. T-cell development and accumulation of CD4+ thymocytes in DOCK8-deficient mice. (A) Representative FACS plots showing T-cell development in the thymus of WT and CPM mice: left, total thymocytes; middle, CD4-CD8- double-negative cells; and right, single-positive CD4+ T cells. Numbers show the percentage of gated cells of parent populations. (B) Absolute number of total CD4+ and CD8+ single-positive T cells, mature CD4+ and CD8+ CD62Lhi subsets in the thymus of WT and CPM mice. Dots represent individual animals aged 9–14 wks. Bars show arithmetic means and SE. **p<0.01, ***p<0.001, t-test. Data are representative of at least five experiments.

Download figure to PowerPoint

T-cell lymphopenia in DOCK8 deficiency is an intrinsic defect

To examine whether T-cell lymphopenia in DOCK8 deficiency was cell intrinsic, we reconstituted lethally irradiated CD45.1 allotype-marked recipients with 50:50 or 20:80 mixtures of WT (CD45.1) and CPM or WT (CD45.2) BM. While the 50:50 WT:CPM BM generated the expected 1:1 ratio of follicular B cells, it failed to reconstitute normal numbers of DOCK8-deficient T cells due to a block in T-cell development prior to the DN stage, which was not previously seen in the absence of competition (data not shown). The cause of this competitive disadvantage in establishing early thymic development is unknown; but it is not limiting in non-chimeras, where the number of cells at the DN stages of thymic development are similar in CPM and WT mice (Fig. 3A). Analysis of 20:80 WT:CPM mixed chimeras showed the same competitive failure to reconstitute thymic development relative to the B-cell lineage but no effect of DOCK8 in the transition from DN to DP and mature single-positive cells (Fig. 4A and data not shown). However, naïve peripheral T-cell numbers were reduced in CPM compared with WT mice (Fig. 4A), which suggests a cell intrinsic defect in peripheral CD4+ and CD8+ survival underlying the immunodeficiency. The absence of CD4+ cell accumulation in the thymus of mixed chimeras also supports the idea that this is not the principal cause of peripheral lymphopenia (Fig. 4A). Moreover, the absence of greater numbers of peripheral CPM CD44hi T cells in the mixed chimeras (Fig. 4B) implies this is due to homeostatic proliferation in unmanipulated DOCK8-deficient animals, where T-cell numbers are low.

thumbnail image

Figure 4. The T-cell lymphopenia in DOCK8 deficiency is cell intrinsic. (A) The percentage of CD45.1 and CD45.2 T cells in thymus, spleen and MLNs of mice reconstituted with an 20:80 mixture of CD45.1 WT and CD45.2 WT T cells (top) or 20:80 mixture of CD45.1 WT and CD45.2 CPM T cells (bottom). Bars represent individual chimeric mice. (B) Histograms showing typical surface CD44 expression on WT (filled) and CPM (solid black line) CD8+ (top) and CD4+ (bottom) T cells from mixed chimeras. Data are representative of at least three experiments.

Download figure to PowerPoint

Migration and survival of early naïve T cells in the absence of DOCK8

B and T cells from DOCK2-deficient mice fail to migrate to a variety of chemoattractants, resulting in the loss of T-cell thymic egress in response to S1P, and peripheral lymphopenia 9, 10, therefore we considered whether similar effects could contribute to immunodeficiency in the absence of DOCK8. To assess chemotaxis, we mixed equivalent numbers of mature allotype-marked thymic T cells from CPM and WT and assessed their ability to migrate in response to constitutively expressed cytokines CXCL12, CCL19/21 and S1P. In contrast to DOCK2-deficient mice, the migration of thymic and peripheral T cells to these chemokines was normal, with the exception of a consistent increase in the migration of mature DOCK8-deficient thymic CD4+CD62Lhi T cells towards CXCL12, (Fig. 5A; p=0.0224 by ANOVA, combining data from three experiments). As CXCL12 signalling via CXCR4 is important in lymph node entry 11, we transferred 50:50 mixtures of mature thymic CPM & WT T cells into WT recipients and plt/plt recipients, which lack expression of the other main T cell attractants CCL21 and CCL19 11. In these experiments, the proportion of thymic CPM CD4+ CD62Lhi T cells in peripheral lymphoid organs after 90 min was increased in comparison to WT cells (Fig. 5B). These findings are consistent with an increased chemotactic response to CXCL12 but also confirm that entry and localisation of cells to peripheral lymphoid organs is normal in unimmunised mice.

thumbnail image

Figure 5. The migration and survival of early naïve T cells in the absence of DOCK8. (A) Transwell migration of mixed CD4+CD62Lhi thymocytes from CPM (CD45.2) and WT (CD45.1) in response to various chemoattractants at indicated concentrations. Data are presented as mean of n=2 and are representative of at least three experiments. (B) Co-transfer for 90 min of 1×106 fluorescently labelled thymocytes from CPM (CD45.2) and WT (CD45.1) into WT or plt/plt recipient mice. Data show proportions of CD4+CD62Lhi cells at input (inp), and after transfer in spleen (spl), MLNs, PLNs and blood: each bar represents an individual mouse. (C) Co-transfer for 24 h of fluorescently labelled thymocytes from WT or CPM (CD45.2) and WT (CD45.1) mice into WT recipients, adjusted so approximately equal numbers of mature (CD62Lhi) single-positive T cells were transferred. Data show proportions of CD4+CD62Lhi cells at input (inp), and after transfer in spleen (spl), MLNs, PLNs and blood (Bld). Data are presented as mean+SEM of n=4-5 and are representative of at least three experiments. (D) Intracellular Bcl-2 expression on total CD4+ and CD8+ T cells in the thymus, spleen and PLNs of WT and CPM mice shown as representative histograms (left) and median fluorescence (right), where n=4–5, columns are means, and bars are SE. *p<0.05, **p<0.001, t-test.

Download figure to PowerPoint

Twenty-four hours after transfer, however, the proportions of mature thymic CPM CD4+ and CD8+ T cells were reduced relative to WT (Fig. 5C), suggesting that DOCK8 deficiency causes these emigrants to survive less well after localisation to secondary lymphoid organs. We found that expression of the survival factor Bcl-2 is reduced in CPM single-positive CD4+ and CD8+ thymocytes, suggesting that DOCK8-dependent activation of anti-apoptotic pathways may contribute to survival at this stage in T-cell development (Fig. 5D).

Decreased survival of peripheral DOCK8-deficient T cells

Together, the above results suggested that defects in cell survival were the principal reason for the loss of mature T-cell populations in DOCK8 immunodeficiency. To evaluate this further, CPM and WT mice were fed BrdU for 15 days and the survival of labelled cells was followed for a further 30 days. In the first phase, BrdU labelling of CD4+ and CD8+ T cells was equivalent or more rapid in the spleen and lymph nodes of CPM compared with WT (Fig. 6A–D). This is consistent with a higher level of homeostatic proliferation and the majority of newly labelled cells were CD44hi (data not shown). More importantly, during the washout phase, DOCK8-deficient CD4+ and CD8+ T cells disappeared much faster in all tissues, confirming a survival defect in mature T cells (Fig. 6A–D).

thumbnail image

Figure 6. Decreased survival of DOCK8-deficient peripheral CD4+ and CD8+ T cells. (A–D) The percentage of BrdU-labelled CD4+ T cells (A, B) and CD8+ T cells (C, D) in the spleens and pooled mesenteric and peripheral lymph nodes of CPM and WT mice fed BrdU for up to 15 days. Cohorts of 3-5 mice were analysed in each group at day 3, 9, 15, 20, 30 and 45. Data show mean and SE. *p<0.05, **p<0.01, t-test. Data are representative of one experiment with n of 2–3 per timepoint.

Download figure to PowerPoint

Loss of memory T cells in the absence of DOCK8

During a normal immune response to pathogens in vivo, the activation, proliferation and differentiation of T cells leads to short-term effector cells and long-lived memory cells, which survive after contraction of the effector population and act to curtail recurrent infection. The in vitro experiments, we have described, suggest that the primary effect of DOCK8 might not be on the generation of short-term effector cells but rather the survival of long-lived memory cells. To investigate these possibilities, we compared the response of CPM and WT mice to vaccination with replication deficient MVA virus. After 9 days there was no significant difference between CPM and WT in IFN-γ, IL-2 or TNF-α production by antigen-specific CD8+ T cells directed against MVA (Fig. 7A and data not shown) or IFN-γ ELISpot (data not shown), either as percentage of CD8+ T cells or in absolute terms. In three separate experiments, this analysis showed no difference between CPM and WT confirming that the initial phase of the T-cell response is not functionally limited by the lymphopenia in pre-immune animals. Anti-MVA specific CD8+ T-cell numbers also remained similar after the contraction phase (Fig. 7A); therefore, there was no premature collapse of the effector T-cell population in the absence of DOCK8. However, 42 days after the primary immunisation, during the phase of long-lived memory cell survival, the number of antigen-specific cells was significantly reduced (Fig. 7A). The loss of surviving T cells was not due to a block in memory cell differentiation as numbers of memory precursor cells (MPECS), defined as KLRG1low CD127hi (IL-7Rhi), were unaltered (Fig. 7B). Furthermore, expression levels of memory cell markers CD127, CD62L, KLRG1 and Bcl-2 were similar on MVA-specific WT and CPM T cells at day 9, 14, 42 and after rechallenge supporting this conclusion (Fig. 7C and data not shown). After 55 days, the response to rechallenge with MVA in vivo was reduced in DOCK8-deficient mice compared with WT (Fig. 7A). However, the fold-increase in the secondary response was proportionate to the number of surviving memory cells (Fig. 7A and data not shown), once again suggesting that it is the survival of memory cells rather than the generation of the secondary effector response that is limiting for CD8+ T-cell immunity in the absence of DOCK8.

thumbnail image

Figure 7. DOCK8-deficient mice mount normal anti-viral responses but fail to sustain memory. (A) Numbers of LIVEDEAD® negative splenic CD8+ T cells secreting IFN-γ in response to the MVA specific peptide B8R taken at various time points before and after primary and secondary immunisation of mice with intradermal MVA. WT, circles and CPM, triangles. Data show mean+SEM of data combined from three experiments where n>4. *p<0.05, t-test; however, there is no statistical difference in fold change between the memory response at day 42 and the boost at day 70 comparing WT (mean 15.83 ± SE 1.81) and CPM (mean 26.22 ± SE 4.98). (B) Numbers of memory precursor cells (defined as KLRG1low CD127hi) at day 9 post immunisation from (A). Data are presented as mean and SE and are representative of three experiments with n=5 samples per time point. (C) CD127, CD62L, KLRG1 and Bcl-2 expression on B8R-responsive IFN-γ-secreting CD8+ splenic T cells from WT (solid black line) and CPM (grey-filled histogram) mice at day 9 and day 42 post immunisation and after rechallenge. Dotted line shows staining on naïve non-responding cells. (D) The percentage of CD45.1 and CD45.2 B8R specific CD8+ cells in the spleen of chimeric mice reconstituted with a 20:80 mixture of CD45.1 WT and CD45.2 WT T cells (top) or CD45.1 WT and CD45.2 CPM T cells (bottom). Left panel shows B220+, CD4+ and CD8+ populations in blood of reconstituted animals two wks preimmunisation; right panels show B8R antigen-specific CD8+ T cells from animals culled 9 days and 42 days post immunisation with the percentage of ‘naïve’ non-responding CD8+ T cells for comparison. Effector (EFF) and memory precursor (MPEC) are defined as KLRG1hi CD127low and KLRG1low CD127hi, respectively. B8R-specific CD8+ T cells at day 42 are defined as ‘Memory’. Data are presented as means+SE of n=5 samples with statistical values determined by t-test. (E) Histograms showing expression of CD127, KLRG1 and Bcl-2 on ‘Memory’ cells on CD45.2 WT (top, filled histogram) and CD45.2 CPM (bottom, filled) compared with CD45.2 ‘Naïve’ cells (solid black line). (C–E) Data from one experiment with n=5 samples.

Download figure to PowerPoint

To investigate further the nature of DOCK8 deficiency in the maintenance CD8+ memory T cells, we repeated our MVA immunisation protocol with chimeric mice reconstituted as before with 20:80 mixtures of WT (CD45.1) and CPM or WT (CD45.2) BM. In this competitive setting, DOCK8 deficiency resulted in reduced primary CD8+ responses to MVA relative to WT cells (Fig. 7D). The loss of DOCK8 CD8+ response in this competitive situation is consistent with an underlying defect in survival, which is normally compensated for by increased proliferation in non-chimeras. The greatest effect of competition is on the KLRG1hi short-lived effectors, which are turning over most rapidly (Fig. 7D middle panel). During the maintenance phase of memory, day 42, DOCK8 responders were further affected when compared with day 9 MPEC. Together these data support an intrinsic role for DOCK8 in the maintenance of CD8+ T-cell memory. The WT and DOCK8 CD8+ memory T cells expressed similar levels of markers such as CD127+ (IL-7R) and KLRG1 as well as the survival factor Bcl-2 (Fig. 7E).

We also examined T-cell responsiveness during pathogenic infection by challenging CPM and WT mice with mouse adapted influenza strain X31(A/HK-X31 (H3N2)). Using daily weight loss as a clinical parameter, we found no significant difference in disease severity or the ability to survive the acute infection through the nadir of infection at day 6 to recovery by day 10 (Fig. 8A). 11 of 12 WT mice survived the challenge, whereas 9 of 12 CPM mice survived. Seven days post immunisation, there was no significant difference between WT and CPM in IFN-γ production and numbers of LAMP1hi CD8+ T cells to T-cell restricted epitopes with similar expression profiles of CD127, KLRG1 and CD62L (Fig. 8B and data not shown). The number of MPECS were also produced to normal levels (data not shown). These findings again indicate that the immunodeficiency in the absence of DOCK8 is not due to an inability to mount an effective response to acute infection. To assess the memory response, we bled mice 56 days after primary infection and counted LAMP-1hi CD8+ T cells responsive to conserved internal flu T-cell restricted epitopes (Fig. 8B). While WT mice produced a robust response to X31, memory T cells were not sustained above background in CPM mice; these data were substantiated by blood ELISpot (data not shown). Together these findings show that the functions of DOCK8 are not limiting during primary T-cell immune response against viral infection but are critical for maintenance of memory to vaccination and viral infection.

thumbnail image

Figure 8. DOCK8-deficient mice survive live influenza challenge with loss of memory CD8+ responses. (A) Weight loss and recovery in CPM and WT mice infected with influenza strain X31 (n=12 in each group). (B) The ex vivo influenza X31 antigen-specific response in CPM and WT mice from (A), showing numbers of LAMP1+ CD8+ T cells 7 days (left) and LAMP1+ at 8 wks (right). Numbers are corrected for background levels observed in unstimulated controls, (ND, not detectable). Data are presented as means and SE and are representative of two experiments.

Download figure to PowerPoint

Discussion

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

The results presented here contribute to our understanding of how deficiency in DOCK8 can increase susceptibility to disease, particularly viral infection 1, 2. Our findings show that DOCK8 operates at multiple steps in cellular immunity, but its main effects are on T-cell survival, and it is most limiting in the survival of memory T cells. The inability to generate high-affinity antibodies to pathogens due to a failure to sustain GC B cells, shown previously, and the absence of T-cell memory shown here may explain the increased susceptibility of human DOCK8 immunodeficiency syndrome patients to recurrent infections 1, 2, 7.

The relative paucity of peripheral T cells in DOCK8-deficient mice is not due to defective thymic development. Although DOCK8-deficient T cells have a selective disadvantage in populating early thymic development relative to WT cells in mixed chimeras this is not limiting in natural mutants or single BM chimeras. B-cell development is normal in the mixed chimeras and so competitive effects on the early thymus might be due to the defective survival, proliferation and migration of T-cell precursors. However, at terminal thymic development, DOCK8-deficient single-positive CD4+ T cells accumulate in the thymus and these CD4+ cells show increased migration to SDF-1/CXCL12. Emigration from the thymus is mediated primarily by trafficking to S1P, and while CXCL12-dependent signals have been suggested to play a role in thymic egress it remains to be formally demonstrated in physiological circumstances 12, 13.

The discovery that mature naïve thymic CD4+ T cells show increased migration to CXCL12 does raise the interesting possibility of a connection between DOCK8 and the human WHIM syndrome, which is a combined immunodeficiency defined by warts, hypogammaglobulinemia, immunodeficiency and myelokathexis causing neutropenia 14. WHIM- and DOCK8-deficient patients share a similar spectrum of recurrent respiratory and cutaneous viral infections 14 although not all clinical features are comparable. T cells from patients with the WHIM syndrome also show augmented T-cell responses to CXCL12, and a substantial number of these patients have mutations in the CXCR4 chemokine receptor gene 14, 15. DOCK8 may, therefore, be a regulator of CXCR4 signalling and speculatively, WHIM patients with no obvious abnormality in CXCR4-dependent signalling may have mutations in DOCK8-dependent signalling pathways 15, 16.

In theory, the decreased number of naïve T cells in the peripheral T-cell compartment could be due to defects in thymic egress, differentiation and proliferation or the increased death of cells. However, our data suggest that decreased survival is the principal reason for the loss of T cells in DOCK8-deficient animals. The increased turnover of DOCK8-deficient T cells in vivo and the increase in activated or memory-like (CD44hi/CD62Llo) T cells is likely to be caused by lymphopenia-induced proliferation, which occurs when T cells in lymphopenic hosts divide and irreversibly acquire a ‘memory-like’ phenotype 17. The increase in CD44hi/CD62Llo T cells is not present in mixed chimeras where cell numbers are normal, and so, in this case, the loss of naïve peripheral DOCK8 T cells must be due to an intrinsic defect in survival rather than forced differentiation and division. More importantly, lymphopenia affecting the primary repertoire does not seem to be limiting in the primary response to vaccination or infectious challenge, when proliferation can presumably compensate for the reduction in T cells, whereas the most striking effect of DOCK8 deficiency is on the survival of memory CD8+ T cells.

The CD8+ T-cell immune response can be divided into three phases 18: the first phase is of growth and differentiation of naïve CD8+ T cells into effector cells with lytic activity and cytokine production; the second is a phase of 10–20-fold contraction leading to a smaller population of long-lived memory cells, which arise by the differentiation of effectors 19, 20; and the third, a phase in which the memory cells are maintained in the host 20, 21. Long-lived CD44+CD62LhiCD127+ central memory T cells (TCM) arise after 5 days following a typical infection and survive within secondary lymphoid organs where they can respond swiftly in a recall response 20–23. The generation of the memory cells and the expansion of the immune response depend on TCR signalling as well as costimulatory molecules including CD28, CD40, 4-1BB, CD27, ICOS and/or OX40, IL-2 and inflammatory cytokines including IL-12 and IFN-α 24–29. The normal response following MVA vaccination or influenza infection and the initial generation of memory cells suggests that DOCK8 is not limiting downstream of these pathways. The long-term survival of memory CD8+ T cells does not require the presence of antigen but is dependent on cytokines, particularly IL-15 and IL-7 30–33. However, the absence of defects in lymphopenia-induced homeostatic expansion and NK cell development (Greg Crawford, unpublished observations), which depend on IL-7 and IL-15 17 argues against a deficiency in cytokine signalling as the cause of CD8+ T-cell memory failure in the absence of DOCK8.

Since DOCK8 mutant B cells are unable to cluster ICAM-1 following antigen-receptor signalling it is tempting to speculate that similar defects in integrin-dependent signalling might affect the memory T-cell response. Although ICAM-1-dependent interactions between T cells and DCs play a role in the formation of CD8+ memory T cells 34, 35, the absence of ICAM-1/LFA-1 interactions has already been described as having its greatest effect on the survival of memory cells 36–38. It is interesting to speculate that a dependence on external signals from within the cellular niche might also account for the low Bcl-2 expression in DOCK8-deficient single-positive thymocytes and their decreased survival on transfer into the periphery.

How DOCK8-dependent reorganisation of the cytoskeleton could affect lymphocyte survival is also not yet clear. However, a decreased memory response to viral infections is also reminiscent of Wiskott–Aldrich Syndrome (WAS), which is an X-linked, immunodeficiency syndrome principally caused by mutations in the WAS protein (WASP) gene 39. WASP regulates the actin cytoskeleton in a variety of immunological functions including cell migration and the formation and the preservation of T-cell memory 40. WASP and WASP-family verprolin-homologous (WAVE) proteins are fundamental actin-cytoskeleton reorganisers found in eukaryotes. The conserved function across species is to receive upstream signals from Rho-family GTPases and then activate the actin-related protein (ARP) 2/3 complex, leading to rapid actin polymerisation 41, 42. There are a number of striking similarities between WAS and DOCK8-deficiency including peripheral T-cell lymphopenia, a lack of marginal zone B cells and poor immunological synapse formation in WASP-deficient B cells as a consequence of impaired BCR and integrin signalling 43, 44. Cdc42, the probable target of DOCK8, is an important regulator of WASP function 45. Therefore, it is an intriguing possibility that DOCK8 augments the formation of active Cdc42, at the membrane interface thereby signalling through WASP. Further investigation of such pathways and their relationship to the human immunodeficiency syndromes will provide important insights into how T-cell memory is sustained.

Materials and methods

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

Human studies, mouse strains and procedures

Human studies were approved by the institutional review board of the National Institute of Allergy and Infectious Diseases in accordance with Declaration of Helsinki principles. Mouse experiments were performed under UK Home Office Licence or in accordance with UCSF Institutional Animal Care and User Committee approved protocols. CPM mice were generated by ENU-mutagenesis 7; C57BL/6 plt/plt mice were used for transient transfer experiments 46. BM chimeras, C57BL/6 (CD45.1+) mice were irradiated with two fractions of 4.5 Gy spaced by 3 h and reconstituted with 5–10×106 BM cells, comprising 50:50 or 20:80 mixtures of C57BL/6 (CD45.1+) BM and either CPM or C57BL/6 (CD45.2+) cells and allowed to reconstitute for 8–10 wks.

Flow cytometry

Human PBL and mouse lymphocytes were isolated and stained as previously described 1, 7. Antibodies were from B.D. Pharmingen or eBioscience® unless otherwise stated; FITC-conjugated antibodies were against: CD24, BrdU, CD8α, CD62L and CD44 (Caltag). Phycoerythrin-(PE) conjugated antibodies were against: CD5, CD3, CD69, FoxP3, CD62L, CD45.1, CD25 and CD4 (Caltag). PE-Cy7 conjugated antibodies were against B220, CD69, CD8α, CD4 and CD45.1. PE-Cy5.5-conjugated antibodies were against CD8α. Allophycocyanin (APC)-conjugated antibodies were against: CD44 and CD4 (Caltag). APC-Cy7-conjugated antibodies were against CD4. APC-Alexa-Fluor® 750-conjugated antibodies were against: CD45.1. Tricolour (TC)-conjugated CD4 and CD8α (Caltag). BD TruCOUNT tubes were used to enumerate the leucocyte count as per manufacturer's instructions. Data were collected on FACSCantoTM or LSRII machines (B.D.) and analysed with the FlowJo (Tree Star) or FACSDiva (B.D.).

T-cell chemotaxis, activation, proliferation

Chemotaxis assays were performed as described 47. To assess T-cell activation, splenocytes were plated in 96-well plates coated with anti-CD3 antibody at 10 μg/mL (145-2C11) and 10 μg/mL anti-CD28 ascites (37.51) in PBS. Cells were incubated at 37°C overnight in RPMI-containing 10% FCS. To assess proliferation, WT or CPM lymphocytes (2×106 cells/mL) and labelled for 8 min at RT with an equivalent volume of CFSE (final 2 μM). The reaction was quenched by the addition of FCS and washed three times with RPMI-1640 containing 10% FCS. 2×105 CFSE labelled cells were cultured on anti-CD3- and CD28-coated plates for 3 days.

Transient transfers

In transient transfer experiments, ∼1–2×107 cells per mL were labelled with either 3.3 or 10 μM carboxyfluorescein-diacetate-succinimidyl-ester (CFSE; Invitrogen/Molecular Probes) or 10 μM 5-(and-6)-(((4-chloromethyl)benzoyl)amino)-tetramethylrhodamine (CMTMR, Invitrogen/Molecular Probes) in RPMI-1640-containing 2% FCS for 10–20 min at 37°C then washed by centrifugation through FCS. For cell transfers, recipient mice received ∼1–2×107cells by i.v., normalised to account for T-cell number. Recipients were sacrificed at designated times.

BrdU Incorporation

BrdU (Sigma/B.D.) was given at 0.25 mg/mL in drinking water and supplemented with 1% glucose (Sigma), for up to 15 days. Cells were fixed in 0.5% paraformaldehyde, permeabilised in 3 M HCl/0.5% Tween20 (for 20-min at RT), neutralised with 0.1 M Borax (Sigma), washed, and stained with FITC-conjugated anti-BrdU 3D4 (B.D. Pharmingen).

Immunisation and influenza challenge experiments

For MVA immunisation or influenza challenge, mice were anaesthetised and immunised intradermally with MVA expressing green fluorescent protein at a dose of 1×106 PFU/mouse (2×107 PFU/mL) 48 or infected intranasally with 25 μL of Influenza X31 (A/Aichi/68; H3N2) at a dose of 512 HAU/mL. Mice were weighed daily as an indicator of health. Animals reaching 80% of original body weight were humanely culled. Antigen-specific cells were identified by intracellular staining for IFN-γ (early response) or LAMP-1 (memory) after stimulation with anti-CD3 or epitopes from MVA B8R epitope (sequence TSYFKESV) or a 1-μg/well of pooled 20mer peptides spanning conserved internal nucleoprotein and matrix 1 protein epitopes.

Statistical data analysis

Statistical comparisons were performed with the Prism 4.0 Software (GraphPad), using ANOVA or two-tailed paired t-tests for comparisons.

Acknowledgements

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

The authors thank the staff of the Oxford Biomedical Services Unit for animal husbandry and Dr. Helen Chapel, Prof. Sarah Gilbert, Prof. Adrian Hill and Dr. Ronald Schwartz, for advice and reagents. This work was supported by the Medical Research Council, NIHR Biomedical Research Centre Programme, and the Intramural Research Program of the NIH, NIAID.

Conflicts of interest: The authors declare no financial or commercial conflict of interest.

References

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

Supporting Information

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

Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors.

FilenameFormatSizeDescription
eji_201141759_sm_SupplInfo.pdf3977KSupplInfo

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.