Published online before print March 30, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
,
,¶
* Division of Immunology and Genetics, The John Curtin School of Medical Research, and
ANU Medical School, The Australian National University, Canberra, ACT; and
Cancer Research Unit and Departments of
Anatomical Pathology and
¶ Neonatology, The Canberra Hospital, ACT, Australia
1Correspondence: Division of Immunology and Genetics, John Curtin School of Medical Research, Australian National University, Mills Rd., Acton, ACT 2601, Australia. E-mail: Hilary.Warren{at}anu.edu.au
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. These data suggest that cord blood CD8 T cells expressing MHC-NKR are a unique subset of cells, distinct from those in adult blood, and may represent a less-differentiated population.
Key Words: NK receptors • KIR • CD27 • granzyme B • TCR-Vß
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/ß T cells, which have a memory phenotype characterized by the absence of CC chemokine receptor 7 (CCR7) and a failure to detect CD27 and CD28 on most of them [2
]. The proportion of MHC-NKR+ CD8 T cells in adult blood increases with age, and the population of MHC-NKR+ CD8 T cells can have a skewed distribution of T cell receptor (TCR)-Vß expression [3
]. KIR+ CD8 T cells reportedly have a high intracytoplasmic perforin and granzyme B content and a reduced proliferative potential [4
, 5
]. KIR cannot be induced on CD8 T cells by antigen exposure during in vitro culture [6
]. NKG2A expression on T cells is up-regulated by cytokines {transforming growth factor-ß, interleukin (IL)-12, or IL-15 [7
8
9
]} and by antigen re-exposure in vitro [9
, 10
]. Expression of inhibitory MHC-NKR on T cells is considered to reflect a prior history of chronic antigen or autoantigen exposure [11
], a notion supported by the oligoclonal nature of MHC-NKR CD8 T cell expansions [10
, 12
]. It is proposed that inhibitory MHC-NKR on CD8 T cells prevents activation-induced cell death and/or limits responses of T cells to chronic antigen stimulation. In this study, we demonstrate that KIR and NKG2A are expressed on some CD8 T cells in cord blood, a finding contrary to earlier reports [13 , 14 ]. We present phenotypic and functional data to suggest that these MHC-NKR+ CD8 T cells in cord blood are less differentiated than the majority of these cells in blood from most adult donors.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Antibodies and flow cytometry
The anti-human receptor mAb with their sources are listed as follows: anti-NKG2A-phycoerythrin (PE), anti-CD158a,h-PE, anti-CD158b1,b2,j-PE, and anti-CD8-PC5 from Beckman-Coulter Immunotech (Fullerton, CA); anti-CD158e1(NKB1)-PE, anti-CD8-peridinin chlorophyll protein (PerCP), anti-CD3-fluorescein isothiocyanate (FITC), anti-granzyme B-FITC, anti-CD45RA-FITC, anti-CD45R0-FITC, anti-interferon- (IFN-)-FITC, and anti-CD3-allophycocyanin (APC) from BD Biosciences (San Jose, CA); anti-CCR7-FITC from R&D Systems (Minneapolis, MN); and anti-CD27-FITC from Cymbus Biotechnology (UK). FITC-conjugated anti-TCR-Vß-specific mAb (Serotec, UK; TCR-Vß 1, 2, 3, 5.1, 5.2, 7, 8, 11, 12, 13.1, 13.6, 14, 16, 17, 20, 21.3, and 22) were a generous gift from Dr. Rajiv Khanna (Queensland Institute of Medical Research, Brisbane, Australia).
Cells were stained with combinations of optimum concentrations of mAb to CD3, CD8, and KIR (CD158a,h plus CD158b1,b2,j plus CD158e1) or NKG2A by incubating on ice for 30 min and washing three times with phosphate-buffered saline (PBS) containing 5% fetal calf serum (FCS) and 0.01% sodium azide. For some experiments, aliquots of the labeled cells were then stained with mAb to CCR7, CD27, CD45RA, CD45R0, or TCR-Vß mAb or appropriate isotype-matched, control Ig. Prior to staining with anti-granzyme B mAb, anti-IFN- mAb, or isotype-matched, control Ig, the labeled cells were fixed and permeabilized by using the BD Cytofix/CytopermTM kit from BD Biosciences. After staining, washed cells were fixed with 1% paraformaldehyde in PBS and analyzed using a FACScan (for PC5, PE, and FITC) or a FACSCalibur (for PerCP, APC, PE, and FITC, BD Biosciences). In all cases, CD3+ CD8bright T cells were analyzed. Data were processed by using WinMDI Version 2.8 software (http:/facs.scripps.edu/) and then plotted and statistically analyzed by using GraphPad Prism® Version 3.0 software (San Diego, CA).
Cell culture
Isolated blood lymphocytes were labeled with the fluorescent dye 5- and 6-carboxyfluorescein diacetate succidimidyl ester (CFSE; Molecular Probes, Eugene, OR) prior to culture on plastic wells precoated with anti-CD3 mAb using methods detailed previously [16
]. Culture medium was Eagle’s minimal essential medium supplemented with 10% heat-inactivated FCS, 0.1 mM 2-mercaptoethanol, and 20 U/ml recombinant IL-2 (Chemicon, El Segundo, CA). Cells were harvested on Day 4 or 5 of culture and labeled with mAb as described above.
Isolated blood lymphocytes were cultured with phorbol 12-myristate 13-acetate (PMA; 5 ng/ml) and ionomycin (0.5 µg/ml) in the culture medium described above at 37°C for 5 h. At 3 h, GolgiStopTM (BD Biosciences, Cat. 51-2092KZ) was added at the concentration recommended by the manufacturer. Cells were labeled with mAb as described above.
Histopathology
Placentas were collected at delivery, fixed in 10% buffered formalin, processed, and assessed for in utero exposure to infection or inflammation as described previously [17
].
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (18K): [in a new window] |
Figure 1. KIR and NKG2A are expressed on subsets of CD8 T cells in cord blood from term (>37 weeks gestational age) and preterm (25–30 weeks gestational age) infants, independent of placental pathology. The population of NK and T cells isolated from blood was labeled with FITC-conjugated CD3 mAb and PC5-conjugated CD8 mAb and a mix of PE-conjugated KIR mAb (CD158a,h, CD158b1,b2,j, and CD158e1) or PE-conjugated NKG2A mAb. The gated CD3+CD8+ T cells were analyzed for KIR (A) and NKG2A (B) expression. Data for cord blood from preterm and term infants with no placental pathology (H) and infants with placental pathology (P) are grouped separately and compared with adult blood. Bars indicate the mean of the data points. The mean values and number of samples analyzed are indicated along the x-axis. The P values between samples are shown by square brackets and are from a Mann-Whitney analysis of the data.
|
  View larger version (76K): [in a new window] |
Figure 2. Comparative phenotype of KIR+ CD8 T cells in cord blood and adult blood. The population of NK and T cells isolated from blood was labeled with APC-conjugated CD3 mAb and PerCP-conjugated CD8 mAb and a mix of PE-conjugated KIR mAb (CD158a,h, CD158b1,b2,j, and CD158e1) or PE-conjugated NKG2A. The cells were then labeled with FITC-conjugated anti-CCR7 mAb, anti-CD27 mAb, anti-CD45RA mAb, anti-CD45R0 mAb, or after fixation and permeabilization, with FITC-conjugated granzyme B mAb. The gated CD3+CD8+ T cells were analyzed for KIR or NKG2A (x-axis, PE) and expression of CCR7, CD27, CD45RA, CD45R0, and granzyme B (y-axis, FITC). (A) Flow cytometry profiles for a representative term cord blood from an infant with no placental pathology and for two adult blood samples. (B) Summary data for cohorts of cord blood samples from infants with no placental pathology and adult blood samples. Bars indicate the mean of the data points. The mean values and SEM are indicated along the x-axis. (C) Granularity of CD8 T cell subsets measured as side-scatter (SSC). The percentage exceeding the granularity of CCR7+ cells is indicated below the marker (M1).
|
KIR+ CD8 T cells in adult blood were almost entirely CCR7–; however, in contrast to cord blood, KIR+ CD8 T cells in different adult blood samples were more heterogeneous; fewer cells expressed CD27, and more cells expressed granzyme B. Compared with KIR+ CD8 T cells in cord blood, fewer KIR+ CD8 T cells in adult blood expressed CD45RA. NKG2A+ CD8 T cells in adult blood were also CCR7– and with respect to other markers, were also a heterogeneous population. Compared with KIR+ CD8 T cells in adult blood, more NKG2A+ CD8 T cells expressed CD27, and less expressed granzyme B and CD45RA. CD45R0 expression was not expressed reciprocally with CD45RA in many cases, and the reason for this is not known. Collectively, the data for adult blood MHC-NKR+ CD8 T cells suggest that these populations are more differentiated than those in cord blood and that the populations are heterogeneous.
To further assess if MHC-NKR+ CD8 T cells differ in other respects from MHC-NKR– CD8 T cells in cord blood and adult blood, we compared the granularity of the different subsets. Representative data presented in Figure 2C
compare granularity (SSC) for CCR7+ cells, the CCR7– cells which lack KIR and NKG2A, and the KIR+ and NKG2A+ cells which are also CCR7–. The extent of granularity is indicated by the percent positive above the CCR7+ control, set at 
1%. The mean and SEM of a number of cord blood and adult blood CD8 T cells are presented below the histograms in Figure 2C
. For cord blood CD8 T cells, the CCR7– cells were more granular than the CCR7+ cells (3.91% positive compared with 1.26% positive), and the KIR+ and NKG2A+ cells were somewhat more granular (9.41% positive and 11.71% positive, respectively). For adult blood CD8 T cells, the CCR7– cells were considerably more granular than the CCR7+ cells (24.87% positive compared with 1.12% positive), and the KIR+ and NKG2A+ cells were somewhat more granular (35.37% positive and 30.55% positive, respectively). The increased granularity of CCR7– cells in cord blood and adult blood suggests that these cells are more differentiated than CCR7+ cells. It is important that compared with the CCR7+ cells, the CCR7– cells and the KIR+ and NKG2A+ cells in cord blood are less granular than those in adult blood, suggesting that cord blood MHC-NKR+ CD8 T cells are less differentiated than their counterparts in adult blood.
Analysis of KIR expression in different TCR-Vß CD8 T cell subsets
Several TCR-Vß CD8 T cell subsets in cord blood from term infants with no placental pathology were analyzed for expression of KIR (Fig. 3
). In four cord blood samples, a total of 17 TCR-Vß subsets was analyzed, although a complete analysis of all TCR-Vß subsets for any one cord blood sample was not possible, because of limited cell numbers available. Likewise, it was not possible to analyze NKG2A+ CD8 T cells in cord blood for their TCR-Vß expression because of the limited number of cells available. The proportion of KIR+ cells in individual TCR-Vß subsets was similar to the proportion for all CD8 T cells, indicating that there was no skewing of their distribution within the CD8 T cell population. For comparison, two adult blood samples were analyzed for KIR+ cells in 17 TCR-Vß CD8 T cell subsets. KIR+ CD8 T cells in the donors chosen for this analysis were entirely CD27– granzyme B+. For Adult #1 (age 32 years), the distribution of KIR+ cells amongst TCR-Vß subsets ranged between 2% and 11.6%, compared with 8.9% for all CD8 T cells. For Adult #2 (age 58 years), the proportion of CD8 TCR-Vß subsets expressing KIR was low for 16 of the 17 subsets analyzed (range 0.1–1.83%) and was high (74%) for TCR-Vß22 (Fig. 3)
. The expansion of KIR+ cells in the TCR-Vß22 CD8 T cell subset in Adult #2 is consistent with reports of preferential expansion of particular TCR-Vß subsets in adults [12
]. Adult #2 had a prior history of Helicobacter pylori infection and subsequent Hashimoto’s thyroiditis; therefore, the expansion of KIR+ cells in the TCR-Vß22 CD8 T cell population is consistent with a response to chronic antigen or autoantigen stimulation.
 View larger version (18K): [in a new window] |
Figure 3. CD8 T cells expressing KIR from term cord blood and adult blood are represented in a range of TCR-Vß subsets. The population of NK and T cells isolated from blood was labeled with APC-conjugated CD3 mAb and PerCP-conjugated CD8 mAb and a mix of PE-conjugated KIR mAb (CD158a,h, CD158b1,b2,j, and CD158e1). The cells were then labeled with FITC-conjugated mAb to TCR-Vß subsets. Data show KIR expression on TCR-Vß CD8 T cell subsets in cord blood from infants with no placental pathology (C #1–4) and adult blood (A #1 and A #2). The individual TCR-Vß subsets tested were TCR-Vß 1, 3, 7, 17, 22 (C #1); TCR-Vß 5.1, 8, 12, 20, 21.3 (C #2); TCR-Vß 1, 2, 3, 5.1 (C #3); TCR-Vß 2, 7, 13.1, 14, 17, 21.3, 22 (C #4); and TCR-Vß 1, 2, 3, 5.1, 5.2, 7, 8, 11, 12, 13.1, 13.6, 14, 16, 17, 20, 21.3, 22 (A #1 and A #2). TCR-Vß 22 expansion is seen for A #2. The percentage of all CD8 T cells that express KIR is shown on the x-axis.
|
when stimulated by PMA and ionomycin (Fig. 5
).
 View larger version (38K): [in a new window] |
Figure 4. Proliferation of cord blood and adult blood CD8 T cells in relation to KIR expression. The population of NK and T cells isolated from blood was labeled with CFSE prior to culture in medium supplemented with 20 U/ml IL-2 on plastic wells precoated with anti-CD3 mAb. Cultures were harvested on Day 4 (A #1, A #2, C #5) or Day 5 (A #3, C #6) and labeled to identify CD8 T cells and KIR as described in the legend to Figure 3
. The percentages at the point of each quadrant are the percentage of the analyzed cells that are divided (left quadrants) or are undivided (right quadrants) in relation to KIR expression (y-axis). The values within the upper left- and lower left-hand quadrants are the percentage of cells that have entered division. These values are calculated from the percentage of cells at each division peak divided by the expected progeny of those divisions and compared with the percentage of cells that are undivided (right-hand quadrants).
|
 View larger version (35K): [in a new window] |
Figure 5. Comparison of the ability of cord blood and adult blood CD8 T cells to produce IFN- in relation to KIR expression. The population of NK and T cells isolated from blood was incubated for 5 h in medium supplemented with 20 U/ml IL-2 and containing PMA (5 ng/ml) and ionomycin (0.5 µg/ml). GolgiStopTM was added at 3 h. After harvest, cells were labeled to identify CD8 T cells and KIR as described in the legend to Figure 3
and then fixed and permeabilized prior to labeling with FITC-conjugated anti-IFN- mAb. The values within the right-hand quadrants are the percentages of KIR+ (upper right quadrant) or KIR– (lower right quadrant) cells that are IFN-+.
|
IFN- production was assessed after 5 h stimulation with PMA and ionomycin by intracellular staining (Fig. 5)
. The proportion of cells within the KIR+ and KIR– CD8 T cell subsets producing IFN- is indicated in the upper right and lower right quadrants, respectively. For individual adult blood samples, there was variation in the relative proportions of KIR+ and KIR– CD8 T cells, which produced IFN-. For the three adult donors, the percentages of KIR+ and KIR– CD8 T cells producing IFN- were 29.6% compared with 17.1%, 26.9% compared with 40.7%, and 66.7% compared with 33.7%. These data suggest that factors other than expression of KIR also regulate IFN- production. For the two cord blood samples analyzed, the proportion of KIR+ CD8 T cells producing IFN- was tenfold higher than the proportion of KIR– CD8 T cells producing IFN-: 45.9% compared with 4.5% and 30.6% compared with 2.8%. Thus, KIR+ CD8 T cells in cord blood are more responsive to PMA and ionomycin stimulation for production of IFN- than KIR– CD8 T cells and in this respect, are similar to KIR+ CD8 T cells in adult blood.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As expected, compared with cord blood, there was a higher proportion of MHC-NKR+ CD8 T cells in adult blood, but it is interesting that there was considerable overlap in the range of values for the cord and adult blood samples. MHC-NKR+ CD8 T cells in cord blood were a relatively homogeneous population compared with those in adult blood. Cord blood KIR+ CD8 T cells were almost entirely CD27+ and granzyme B–, consistent with a naïve phenotype [19 ]. By contrast, KIR+ CD8 T cells in adult blood were heterogeneous for CD27 and granzyme B expression. For some adult donors, a majority of KIR+ CD8 T cells was CD27+ and granzyme B–, and others lacked CD27, consistent with cells that have undergone extensive rounds of cell division [20 ] and expressed granzyme B as reported by others [5 ]. A wide range of CD27 expression was reported previously for KIR+ T cells in adult blood [2 ]. A relatively homogeneous population of NKG2A+ CD8 T cells in cord blood also contrasted with a more heterogeneous population of these cells in adult blood with respect to CD27 and granzyme B expression. These data indicate that cord blood MHC-NKR+ CD8 T cells are less differentiated than their counterparts in adult blood.
Based on the expression of various cell surface markers, CD8 T cells in adult blood can be subdivided into subsets designated as naïve (CD45RA+, CCR7+), effector (CD45RA+, CCR7–), central memory (CD45RA–, CCR7+), and effector memory (CD45RA–, CCR7–). The presence or absence of CD27 subdivides the CCR7– subsets [21 ]. Effector cells, which express CD27, have been termed pre-effector cells, as these cells have characteristics that place them intermediate in differentiation between naïve cells and effector cells [21 ]. Pre-effector cells have undergone a limited number of cell divisions, resulting in relatively long telomeres, and have reduced levels of TCR excision circles compared with naïve cells. Pre-effector cells differ from naïve cells in that they are more granular, subsets express KIR and CD94 receptors, and a proportion express granzyme B and are cytolytic [21 ]. Cord blood MHC-NKR+ CD8 T cells appear to share some characteristics with these pre-effector cells described in adult blood, but there are some differences. Similar to pre-effector cells, cord blood MHC-NKR+ CD8 T cells are CCR7– and CD27+. Although MHC-NKR+ CD8 T cells in cord blood are more granular than CCR7+ cells, they are relatively less granular than MHC-NKR+ CD8 T cells in adult blood when compared with CCR7+ cells. However, MHC-NKR+ CD8 T cells in cord blood do not express granzyme B and in this respect, differ from pre-effector cells in adult blood [21 ]. Thus, MHC-NKR+ CD8 T cells in cord blood appear more naïve than the pre-effector cells described in adult blood.
A CCR7– subset of a CD8 T cell cord blood subset was described recently by Zippelius et al. [22
], which was defined as naïve based on high levels of TCR excision circles, long telomeres, and highly polyclonal TCRs. These CCR7– cells were not cytolytic and were unable to produce IFN-. Thus, these cells differ from the MHC-NKR+ CD8 T cells described in this study, which although lacking the cytolytic effector molecule granzyme B, were able to produce IFN-. Thus, MHC-NKR+ CD8 T cells in cord blood are a unique subset with limited differentiation, giving partial effector functions.
KIR expression on CD8 T cells in adult blood is considered to be associated with a loss of proliferative potential, a diminished ability to produce IFN-, and an enhanced cytolytic ability as indicated by perforin expression [4
]. In our study of adult donors, we observed a strongly reduced, proliferative response of KIR+ CD8 T cells for one donor, a moderately reduced, proliferative response for a second donor, and an enhanced, proliferative response in a third donor when compared with KIR– CD8 T cells. This donor-to-donor variability could reflect variability in differentiation states within the KIR+ CD8 T cells in adult blood. Donor-to-donor variability was also evident in the ability of KIR+ CD8 T cells in adult blood to produce IFN- compared with KIR– CD8 T cells. In contrast to KIR+ CD8 T cells in adult blood, KIR+ CD8 T cells in the two cord blood samples analyzed were not impaired in their ability to proliferate in response to CD3 ligation compared with KIR– CD8 T cells and were as capable as adult KIR+ CD8 T cells in their ability to produced IFN-. However, KIR+ CD8 T cells in adult blood and cord blood were different in granzyme B expression. Thus, KIR expression on CD8 T cells does not necessarily define functional ability.
Data in the literature emphasize the oligoclonal nature of CD8 T cells expressing MHC-NKR and the skewed distribution of these cells in particular TCR-Vß subsets. Consistent with this was our finding for one adult donor with a prior history of chronic antigen or autoantigen stimulation, that the KIR+ CD8 T cells had a severely skewed distribution of TCR-Vß subsets. However, for a second adult donor, there was no skewed distribution of TCR-Vß subsets in the KIR+ CD8 T cells, indicating that KIR+ CD8 T cells could be polyclonal. Analysis of cord blood samples did not indicate a skewed distribution of TCR-Vß subsets in the KIR+ CD8 T cells, suggesting that these cells are polyclonal. These data are consistent with a more naïve phenotype for KIR+ CD8 T cells in cord blood compared with adult blood.
There is evidence that naïve CD8 T cells can proliferate into cells with memory characteristics without passing through an effector stage. This had been shown for CD8 T cells undergoing an IL-7-dependent, homeostatic proliferation in neonatal mice [23
] and for IL-15-stimulated human naïve CD8 T cells [24
], and mouse CD8 T cells in vitro [25
]. The expanded cells survive through to adulthood in mice [23
] and have the capacity to produce IFN- [23
, 25
] and in the case of human CD8 T cells, are also cytolytic [24
]. The proliferating naïve cord blood CD8 T cells retained CD27 expression and for the most part were CCR7+, and CCR7 was lost after several rounds of division [24
]. Therefore, a possibility is that a subset of naïve CD8 T cells in cord blood, which have responded to cytokines as a consequence of homeostatic proliferation, has acquired MHC-NKR. This suggestion would be in line with observations that KIR diversification in CD8 T cells occurs after clonal expansion [3
, 6
].
In conclusion, MHC-NKR+ CD8 T cells are present in cord blood, and these appear to be relatively naïve and distinct from MHC-NKR+ CD8 T cells in adult blood. These cord blood cells probably form part of the pool of MHC-NKR+ CD8 T cells in adult blood, which acquire MHC-NKR in response to chronic antigen stimulation.
Received September 29, 2005; revised December 22, 2005; accepted January 31, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ß TCR rearrangement express highly diverse killer cell Ig-like receptor patterns J. Immunol. 166,3923-3932This article has been cited by other articles:
 |
|
 |
|
  M. P. Correia, E. M. Cardoso, C. F. Pereira, R. Neves, M. Uhrberg, and F. A. Arosa Hepatocytes and IL-15: A Favorable Microenvironment for T Cell Survival and CD8+ T Cell Differentiation J. Immunol., May 15, 2009; 182(10): 6149 - 6159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. T. van der Veken, M. Diez Campelo, M. A. W. G. van der Hoorn, R. S. Hagedoorn, H. M. E. van Egmond, J. van Bergen, R. Willemze, J. H. F. Falkenburg, and M. H. M. Heemskerk Functional Analysis of Killer Ig-Like Receptor-Expressing Cytomegalovirus-Specific CD8+ T Cells J. Immunol., January 1, 2009; 182(1): 92 - 101. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Tang, B. Grzywacz, H. Wang, N. Kataria, Q. Cao, J. E. Wagner, B. R. Blazar, J. S. Miller, and M. R. Verneris Umbilical Cord Blood T Cells Express Multiple Natural Cytotoxicity Receptors after IL-15 Stimulation, but Only NKp30 Is Functional J. Immunol., October 1, 2008; 181(7): 4507 - 4515. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
