Cord Blood T Cells Retain Early Differentiation Phenotype Suitable for Immunotherapy After TCR Gene Transfer to Confer EBV Specificity

Abstract

Adoptive T cell therapy can be effective for Epstein–Barr virus (EBV)-associated posttransplant lymphoproliferative disease and melanoma. Transducing high-affinity TCR genes into T lymphocytes is an emerging method to improve potency and specificity of tumor-specific T cells. However, both methods necessitate in vitro lymphocyte proliferation, generating highly differentiated effector cells that display reduced survival and antitumor efficacy postinfusion. TCR-transduction of naive lymphocytes isolated from peripheral blood is reported to provide superior in vivo survival and function. We utilized cord blood (CB) lymphocytes, which comprise mainly naive cells, for transducing EBV-specific TCR. Comparable TCR expression was achieved in adult and CB cells, but the latter expressed an earlier differentiation profile. Further antigen-driven stimulation skewed adult lymphocytes to a late differentiation phenotype associated with immune exhaustion. In contrast, CB T cells retained a less differentiated phenotype after antigen stimulation, remaining CD57-negative but were still capable of antigen-specific polyfunctional cytokine expression and cytotoxicity in response to EBV antigen. CB T cells also retained longer telomeres and in general possessed higher telomerase activity indicative of greater proliferative potential. CB lymphocytes therefore have qualities indicating prolonged survival and effector function favorable to immunotherapy, especially in settings where donor lymphocytes are unavailable such as in solid organ and CB transplantation.

Abbreviations

7AAD

7-aminoactinomycin

ACT

adoptive cellular immunotherapy

CB

cord blood

CBMC

cord blood mononuclear cells

CFSE

carboxyfluorescein succinimidyl ester

CMV

cytomegalovirus

DCs

dendritic cells

DMSO

dimethyl sulfoxide

EBV

Epstein–Barr virus

FCS

fetal calf serum

GMCSF

granulocyte-macrophage colony-stimulating factor

GVHD

graft versus host disease

IFNγ

interferon gamma

PB

peripheral blood

PBMC

peripheral blood mononuclear cells

PBS

phosphate buffered solution

PI

propidium iodide

PTLD

posttransplant lymphoproliferative disease

SOT

solid organ transplantation

T_N

naive T lymphocytes

T_CM

central memory T lymphocytes

T_EM

effector memory T lymphocytes

T_EMRA

effector memory CD45RA+ T lymphocytes

TNFα

tumor necrosis factor alpha

Introduction

Epstein–Barr virus (EBV)-associated posttransplant lymphoproliferative disease (PTLD) complicates up to 13% of solid organ transplantation (SOT) (2010). Anti-CD20 immunotherapy is an effective first line treatment but around 40% of cases remain refractory. Cord blood (CB) transplantation is also associated with significant EBV and cytomegalovirus (CMV) reactivation and PTLD. Adoptive cellular immunotherapy (ACT) using third-party partially HLA-matched EBV-specific T cells is an effective treatment for PTLD in both settings (2007, 2010, 2010). However, for PTLD following SOT only 50% respond to ACT (2007) possibly because the polyspecific T cells generated in vitro using an EBV-transformed B-lymphoblastoid cell line (LCL) contain too few effectors specific for the limited set of EBV antigens expressed by the tumor.

An alternative approach to rapidly generate large numbers of potent and specific effectors is to engineer T cells to express an appropriate antigen-specific TCR or a chimeric antigen receptor (2010, 2006). This has been used successfully to treat cancers such as melanoma (2011), where naturally occurring tumor-specific T cells are rare and of low avidity (1999, 2005, 2008). Using retroviral vectors, human T cells can be reliably transduced with TCR genes enabling them to recognize viral or tumor antigens. Adoptive transfer of engineered T cells is currently undergoing clinical trials with encouraging results (2011, 2006, 2011).

ACT studies indicate that antitumor response is linked to long term in vivo persistence of infused cells (2004, 2005, 2005). The factors influencing in vivo persistence of lymphocytes are not fully understood, but evidence suggests that the differentiation status of the T cell is critical. Less differentiated naive (T_N) and central memory (T_CM) T cell subsets display superior proliferation, persistence and antitumor responses following infusion when compared to the more differentiated effector memory (T_EM) subset (2005, 2005, 2008). This raises an important issue for ACT using genetically engineered T cells because in vitro activation of adult lymphocytes, required for retroviral transduction, drives the majority of peripheral blood-derived T cells into highly differentiated effector. Thus current approaches using transduced T cells may be suboptimal because the majority of cells infused will be differentiated and may therefore be of limited efficacy in vivo (2009, 2010).

The challenge for ACT with genetically engineered T cells, or with any protocol involving cell expansion, is therefore to generate cells with a minimally differentiated phenotype. Recent studies (2009, 2010) suggest that CD8 T_N lymphocytes selected from adult peripheral blood (PB) are optimal for this purpose because, in contrast to T_CM and T_EM cells, they display minimal differentiation following TCR transduction. Human umbilical CB T cells, unlike adult-derived PB lymphocytes, are mostly T_N. It is therefore reasonable to speculate whether CB might be an alternative source of T-lymphocytes for genetic engineering. As both solid organ and CB transplant recipients cannot access lymphocytes from the original donors, third-party allogeneic CB is a convenient alternate source of lymphocytes for ACT against EBV-PTLD. Such cells can also be used in lymphopenic cancer patients where autologous lymphopheresis is not possible.

This study utilizes cryopreserved CB units from an unrelated cord blood bank and assesses the feasibility of using cord T cells to transduce EBV-specific TCR, and to analyze their functional capacity for in vivo use in immunotherapy.

Materials and Methods

Cell isolation and culture

Frozen umbilical CB units, unsuitable for transplantation, were provided by NHS Cord Blood Bank, UK. Units were thawed in cold RPMI 1640 (Sigma-Aldrich, St. Louis, USA) plus 10% fetal calf serum (FCS) (PAA, Pasching, Austria). After washing, mononuclear cells were isolated using Ficoll. Adult PB mononuclear cells (PBMC) were isolated from aphereses cones (2006) collected at the blood donor centre, where the mean donor age was 42 years. The study was approved by the West Midlands Research Ethics Committee (05/Q2706/91). Dendritic cells (DCs) were generated from adherent mononuclear cells after incubating in plates for 2 h. Adherent cells were cultured in the medium supplemented on days 0, 3 and 6 with 50 ng/mL GMCSF and 500 U/mL IL-4. On day 6, DCs were matured by adding 2 ng/mL IL-1β, 1000 U/mL IL-6 and 10 ng/mL TNFα (R&D Systems, Minneapolis, USA) (1997). DCs were recovered after a further 24/48 h. To confirm maturation, they were stained for CD14, CD83, CD86 and HLA class II antigens. DCs were pulsed with the SSCSSCPLSK (SSC) peptide epitope (1997) at a concentration of 10 μg/mL for 2 h, washed and used together with 100 U/mL IL-2 to stimulate transduced T cells at a responder: stimulator ratio of 20: 1. Cell counts were performed using ABX-Pentra 60 (Horiba, Kyoto, Japan).

Lymphocyte activation and TCR gene transduction

CB mononuclear cells (CBMC) or PBMC were resuspended in RPMI containing 10% FCS, 1% pooled human AB serum (TCS Biosciences, Buckingham, UK), 2 mM L-glutamine, 100 mg/mL streptomycin and 100 IU/mL penicillin and activated with 30 ng/mL anti-CD3 antibody (OKT3) plus 600 U/mL IL-2 (Chiron, Emeryville, USA). Cells were transduced with retrovirus 48 h later. The retrovirus used was the pMP71-PRE vector (2003) (provided by C. Baum, Hannover, Germany) into which we had inserted genes encoding TCR α and β chains isolated from an EBV-specific CD8+ T cell clone that targets the HLA A*1101-restricted epitope SSC derived from the viral protein LMP2 (manuscript in preparation, Zheng, Lee et al.). To generate the retrovirus, Phoenix amphotropic packaging cells (1996) were transfected with the pMP71-PRE vector using FuGENE HD (Roche, Basel, Switzerland). After 48 h the retroviral supernatant was recovered. Preactivated cells were seeded at 4–6 × 10⁶ cells/well in 1 mL RPMI onto 6-well plates coated with retronectin (Takara, Shiga, Japan). Retroviral supernatant (1.5 mL/well) or medium alone (mock-transduced) was added to each well and centrifuged for 1 h × 800 g at 30°C. Medium supplemented with IL-2 (100 U/mL) was added three times weekly. Cells from six CB and six adult PB samples were assayed. Seven and 15 days after transduction, T cells were stimulated with SSC peptide-pulsed DCs. Sixteen days after transduction, CD8+ lymphocytes were isolated using immunomagnetic beads (Miltenyi Biotech, Bergisch Gladbach, Germany) according to the manufacturer's instructions.

Flow cytometry

Transduced lymphocytes were identified using HLA A*1101: SSC peptide-specific pentamers (Proimmune, Oxford, UK). Cells were also costained with fluorochrome-conjugated antibodies: anti-CCR7 (FITC; R&D), anti-CD4 (PerCP-Cy5.5 or PE-Cy7; eBioscience, San Diego, CA, USA), anti-CD3 (PE or APC-H7), anti-CD8 (APC or APC-H7), anti-CD27 (APC), anti-CD45RA (PE-Cy7), anti-CD62L (PE) (all from BD) and anti-CD57 (APC; BioLegend, San Diego, CA, USA). 7-aminoactinomycin D (7AAD from BD) was used as a viability marker. For intracellular staining, cells were fixed and permeabilized using the FIX&PERM kit (ADG, Kaumberg, Austria), followed by anti-Perforin FITC (eBioscience).

For intracellular cytokine staining, cells were first stimulated with either anti-CD3 plus IL-2 or autologous peptide-pulsed DCs, and 1 h later Monensin (Golgi Stop; 1 μL/mL; BD) was added. After overnight incubation cells were then fixed and permeabilized, and stained with anti-IL-2 FITC, anti-IFNγ APC and anti-TNFα PE-Cy7 (BD).

Proliferation was evaluated by staining cells for 2 min with 1 μM carboxyfluorescein succinimidyl ester (CFSE) prior to activation. The enumeration of cells in the different phases of cell cycle or in apoptosis was performed using propidium iodide (PI) (1998). Briefly, the cell pellet was incubated for 30 sec in 200 μL 0.1% Triton ×100 in PBS. Afterward, the same volume of PBS plus PI 50 μg/mL and RNAse 500 μg/mL was added, and samples were analyzed after 10 min at RT. Data were acquired using a FACSCanto II flow cytometer (BD).

Cytotoxicity assay

Cytotoxicity of transduced T cells was assessed in a standard 5-h chromium release assay at known effector: target ratios using 2500 target cells/well. HLA A*1101-transduced T2 cells (1992) were used as targets and were pulsed with the SSC peptide or another A*1101-restricted peptide epitope (IVTDFSVIK) (1993) as a control. The percentage-specific lysis and lytic units (LU) per 10⁶ effector cells were calculated. One lytic unit was defined as the number of effectors required to achieve lysis of 50% of targets (1971).

Telomere length measurements by flow FISH

Telomere length measurements using automated multicolor flow-fluorescence in situ hybridization (flow FISH) was performed by Repeat Diagnostics Inc. (North Vancouver, Canada) as described by Baerlocher et al. (2006).

Telomerase repeat amplification protocol (TRAP) assay

Telomerase activity in extracts from cultured lymphocytes was measured using the telomerase detection assay kit (TRAPeze® telomerase Detection, Millipore) according to the manufacturer's instructions.

Statistics

The statistical analyses were performed using the paired, two-tailed t-test.

Results

Activated CB-derived T-lymphocytes have greater proliferative capacity compared to activated adult T cells

Initially we determined the relative proportions of naive and memory T cells within CB and compared this to blood from adult donors. T-lymphocytes were categorized into: T_N: CCR7+CD27+CD45RA+CD62L+; T_CM: CCR7+CD27+CD45RA-CD62L+; T_EM: CCR7-CD27+CD45RA-CD62L-; or T_EMRA: CCR7-CD27-CD45RA+CD62L- (2006). CCR7 and CD45RA provided sufficient discrimination among the subsets and therefore results based on expression of these two markers are shown hereafter. Phenotypic analysis confirmed that a mean of 81% of T cells in CB samples (n = 6) displayed a T_N phenotype in contrast to healthy adult blood donors (n = 6) where T_N constituted a mean of 29% (Figure 1A). The memory cell subset within CB comprised predominantly T_CM with very few T_EM.

Figure 1. Lymphocytes from cord blood display increased proliferative activity in comparison to lymphocytes from adult blood. (A) The distribution, evaluated by flow cytometry, of CD4 and CD8 lymphocytes into T_N, T_CM, T_EM and T_EMRA subsets is shown for CBMC (white bars) and PBMC (gray bars). Asterisks indicate p < 0.05. (B) CFSE-stained cells were activated with anti-CD3 plus IL-2, and proliferation was assessed at the indicated time points. One representative experiment out of six is shown. (C) Lymphocytes, either from CBMC (white bars) or PBMC (gray bars) were activated, and the distribution of cells in the different phases of the cell cycle was measured 48 h later by propidium iodide uptake. The percentage of cells in apoptosis or in different phases of the cell cycle is indicated. Asterisks indicate p < 0.05.

The proliferation prior to retroviral transduction was monitored by the dilution of CFSE over the next 3 days. This showed an increased proliferation rate within CB cells compared to adult T cells (Figure 1B) and was confirmed with PI staining 48 h after activation which showed more than twice as many CB cells in the S to M phases of cell cycle compared to cells from adult donors. Importantly, this difference was not due to cell death as the proportion of apoptotic cells was the same in both cultures (Figure 1C)

Cord and adult lymphocytes show comparable levels of TCR expression following retrovirus-mediated gene transfer

For the transduction of EBV-specific TCR genes, CBMC and adult PBMC were activated for 2 days and then infected with the retroviral vector encoding the TCR. Expression of the transduced TCR was determined using HLA-A*1101: SSC pentamers (Figure 2A). The TCR transduction efficiency was similar for both cord and adult lymphocytes (Figure 2B). Pentamer staining of CB CD8 and CD4 T cells showed mean transduction efficiencies of 9.8% and 5.6% respectively whereas for adult CD8 and CD4 cells they were 10.7% and 6.1%, respectively. There was no difference in the mean fluorescence intensity of pentamer staining of T cells from cord or adult blood, indicating comparable levels of surface TCR expression (Figure 2C).

Figure 2. Cord blood lymphocytes can be transduced as effectively as lymphocytes from adult blood. (A) The results from a representative transduction of cord blood and adult blood lymphocytes with the EBV-specific TCR are shown at day 6 postactivation. The transduction efficiency was assessed by staining with a specific pentamer using mock-transduced T cells as controls. The proportions of pentamer-stained cells among CD4 and CD8 T cell subsets are indicated. One representative experiment out of six is shown. (B) The mean percentage (+ SD) of transduced T-lymphocytes are shown for CBMC (n = 6; white bars) and PBMC (n = 6; gray bars). (C) The mean fluorescence intensity (MFI) of transduced T-lymphocytes derived from CBMC (n = 6; white bars) and PBMC (n = 6; gray bars) that stained positive with the pentamer.

TCR-transduced CB T cells acquire a predominantly central memory phenotype

We then explored how retroviral transduction affected the differentiation status of CB and adult T cells. Among transduced CB cells, 60% were T_CM and most of the remaining cells were T_N (Figures 3A and B). In contrast, the majority of transduced adult T cells had a T_EM phenotype. This pattern was similar for both CD4 and CD8 T cells and for those T cells that failed to express the TCR (Figure 3A and B).

Figure 3. After activation and TCR transduction, cord blood lymphocytes display a less differentiated phenotype compared with adult T-lymphocytes. (A) The expression of CCR7 and CD45RA before cells were activated with anti-CD3 and IL-2 and then transduced compared with the same cells 6 days later. Black dots indicate transduced lymphocytes, gray dots nontransduced lymphocytes. One representative experiment out of six is shown. (B) The distribution of transduced CD4 and CD8 among T_N, T_CM, T_EM and T_EMRA subsets on day 6 is shown as mean (+ SD) percentages for CBMC (n = 6; white bars) and PBMC (n = 6; gray bars). Asterisks indicate p < 0.05.

TCR-transduced CB CD8+ T cells proliferate after exposure to antigen but are less differentiated compared with adult T cells

The ultimate aim of these studies would be to use TCR-transduced CB T cells in ACT. We therefore studied the proliferation, differentiation phenotype and cytokine production of such cells after prolonged in vitro culture and exposure to antigen. TCR-transduced CB and adult T cells were expanded using two rounds of stimulation with SSC peptide-pulsed autologous DCs. After 23 days of culture, both CB and adult T cells had expanded 50-fold with cell numbers continuing to rise (Figure 4A). Preferential expansion of transduced cells occurred so that by day 23 they comprised 50% of the CD8+ T cell population (Figure 4B).

Figure 4. After activation and transduction cord blood T-lymphocytes proliferate as effectively as T-lymphocytes from adults. (A) Graph showing the mean (+ SD) fold increase in cell number of lymphocytes after activation with anti-CD3, retroviral transduction and two restimulations with peptide-pulsed DCs for CD8 (continuous lines, filled symbols) and CD4 (dotted lines, open symbols) T-lymphocytes from CBMC (n = 6; squares) and PBMC (n = 6; triangles). (B) The expansion of transduced cells is shown as the mean percentage (+ SD) of pentamer-stained cells within CD8 (continuous lines, filled symbols) and CD4 (dotted lines, open symbols) T cell subsets of lines generated from CBMC (n = 6; squares) and PBMC (n = 6; triangles).

Comparing the differentiation status of transduced CB and adult T cells after stimulation with peptide-pulsed DCs we observed consistent differences especially within the CD8+ T cell subset (Figure 5A and B). Transduced CD4+ CB T cells were initially dominated by T_N and T_CM, the latter twice as common, while adult T cells were dominated by equal proportions of T_CM and T_EM with minimal T_N. However, following stimulation with antigen the T_EM expanded markedly in both populations while T_CM proportions fell. After 16 days the final product was comparable in both cultures with means of 55–65% T_EM cells.

Figure 5. After antigen stimulation TCR-transduced CBMC retain a less differentiated phenotype than TCR-transduced PBMC. (A) The changes in phenotype of transduced and nontransduced cells from the same culture are shown at two time points (day 6 and day 23 postactivation). Representative data from a single CBMC- and PBMC-derived culture are shown. (B) The relative proportion of TCR-transduced T_N, T_CM, T_EM and T_EMRA cells within the CD4 and CD8 subsets from CBMC (n = 6, open symbols, upper panel) and PBMC (n = 6; closed symbols, lower panel). Data shown are the mean percentage (+ SD). (C) The mean percentage of transduced CD8 lymphocytes expressing CD27 at the indicated time points is shown for CBMC (n = 6; white bars) and for PBMC (n = 6; gray bars). Asterisks indicate p < 0.05. (D) The expression of CD57 on the same cell lines at day 13 postactivation. Asterisks indicate p < 0.05.

However, the differentiation status of CD8+ T cells during expansion was markedly different between cord and adult blood. CD8+ T cells in CB showed a memory cell transition that was similar to CD4+ CB T cells, with the proportion of T_CM decreasing and T_EM increasing until they dominated the population reaching a mean of 54.8% at the end of the culture period. Very few T_EMRA were observed. In contrast, adult T cells showed a sharp decline in the proportion of both T_CM and T_EM subsets during expansion, and a marked increase in T_EMRA which comprised a mean of 66.2% of the final product.

At the end of the culture period, while 32.1% of CB CD8+ T cells retained the least differentiated T_CM or T_N phenotype, these two subsets accounted for only 5.4% of adult CD8+ cells. Note that only the transduced lymphocytes moved towards a more differentiated phenotype indicating that differentiation was driven by antigen-specific recognition (Figure 5A). Transduced CB CD8+ T cells therefore display a less differentiated phenotype compared to adult T cells throughout the in vitro culture period although this difference is not seen within the CD4 population.

In addition, after 23 days of in vitro expansion, CD27, a marker of less differentiated T cells, was expressed on 85.5% of transduced CB CD8+ cells but only 39.1% of transduced adult CD8+ cells (Figure 5C). Similarly, after 13 days of in vitro expansion, CD57, a marker of replicative senescence and antigen-induced apoptotic death of T cells (2003), was expressed on only 17.8% of transduced CD8+ CB T cells but 56.9% of adult CD8+ cells (Figure 5D).

Telomere length dynamics and telomerase activity in TCR-transduced CB and adult T cells

TCR-transduced adult and CB T cells were taken at day 0 (prestimulation) and day 21 (4 days after last stimulation) and assayed for telomere length (Figure 6A). Both adult and CB samples (with the exception of CB1) showed comparable rates of telomere length decline over 21 days of culture, consistent with expected shortening following cell proliferation, but telomeres were consistently longer in CB cells. Telomerase activity was assessed at day 3 after stimulation and again at day 21 (Figure 7B and C). Two out of three CB cells maintained relatively high levels of telomerase activity, indicative of greater proliferative potential, whereas all adult T cells showed a decline in telomerase activity to relatively low levels at day 21. Interestingly CB 1 lost telomerase activity by day 21, and was the only sample to show an increase in telomere length during the culture period.

Figure 6. TCR-transduced CBMC have longer telomere length than TCR-transduced PBMC and generally retain elevated telomerase activity after antigen stimulation. (A) Changes in telomere length measured by Flow-FISH (Kb) of transduced cells from three CBMC and three PBMC cultures are shown at two time points (start of culture day 0 and day 21 postactivation). (B) Corresponding telomerase activity from cell extracts of the same three CBMC and three PBMC cultures recovered at days 3 and 21 after initial antigen stimulation. Data shown are the total TRAP assay telomerase products generated (TPG) from 2 × 10⁴ cell equivalents measured by densitometry and expressed as arbitrary units. (C) Representative telomerase activity gels including assay controls as well as enzyme heat inactivated controls (H).

Figure 7. TCR-transduced cord blood T-lymphocytes can secrete cytokines and mediate cytolytic function. (A) Intracellular staining for IFNγ, IL-2 and TNFα was performed 24 h after activation with anti-CD3 (day 1) and 24 h after the first restimulation with autologous peptide-pulsed DCs (day 10). Mean percentage plus standard deviation of CD8 T lymphocytes staining for IFNγ and for the different combinations of cytokines are displayed for CBMC (n = 3, white symbols) and for PBMC (n = 3, gray symbols). The percentage of cells staining for all three cytokines was evaluated by first gating on double positive IFNγ/TNFα staining cells and then analyzed for IL-2 expression. (B) Intracellular perforin was assessed by flow cytometry on days 6 and 23 postactivation. The percentages of perforin-positive cells are indicated. One representative experiment of three is shown. (C) CD8 T-lymphocytes from CBMC (left panel) and from PBMC (right panel) were assayed for specific cell lysis activity against HLA A*1101-transduced T2 cells at the effector to target (E:T) ratios indicated. The ⁵¹Cr release assay was performed in the presence of an irrelevant peptide, DMSO or different concentrations of the cognate SSC peptide. Results show the mean% specific lysis plus standard deviation from three separate experiments. (D) The same data are expressed as mean lytic units with gray bars for PBMC and white bars for CBMC.

TCR-transduced CB T cells mediate multicytokine production and antigen-specific cytolysis

To investigate the effector function of the transduced cells we determined their cytokine production profile. T cells which secrete multiple cytokines are considered polyfunctional and this correlates with optimally functioning effector cells (2007, 2007). As such, the cultured cells were analyzed for production of IFNγ, IL-2 and TNFα on day 1 (24 h after stimulation with anti-CD3) and on day 10 (24 h after stimulation with peptide-pulsed DCs) (Figure 7A). On day 1 IFNγ secretion was negligible in CBMC but detectable in PBMC. However, after restimulation, the proportion of IFNγ−secreting CD8 cells in both CBMC and PBMC increased to approximately half of both populations. Polyfunctional CBMC secreting more than one cytokine also increased several fold upon restimulation, exceeding multicytokine secretion by PBMC. Compared to PBMC, the proportion of CD8+ CBMC secreting all three cytokines was significantly greater (p = 0.025). This pattern likely reflects the earlier differentiation phenotype of CB T cells and the need for antigen rechallenge before full effector function is activated.

Antigenic stimulation of transduced CD8 T cells from both cord and adult blood led to increased expression of intracellular perforin (Figure 7B) a marker of cytotoxic potential. Cytotoxic function was assessed using a chromium release assay. Lysis of HLA-A*1101-transfected T2 cells pulsed with SSC peptide was observed at relatively low E:T ratios and decreased with titration of the peptide (Figure 7C). The results were expressed in equivalent lytic units (Figure 7D) and compared for cord and adult TCR-transduced CD8 lymphocytes. Although there was a trend toward increased cytotoxicity with adult effector cells, reflecting the increased percentage of T_EMRA and perforin-positive cells within this population (Figures 5B and 7B), this difference was not statistically significant (Figure 7D)

Discussion

The clinical efficacy of adoptively transferred T-lymphocytes correlates with their ability to persist in vivo (2004). Several studies indicate that in vivo persistence correlates with a less differentiated T cell phenotype (2011, 2006, 2005, 2005, 2005, 2008, 2003, 2009) and more recent work suggests T_N may be optimal in this setting, especially where retroviral transduction of T cells is required to engineer the appropriate antigenic specificity (2009, 2010). For this reason we studied CB as a potential source of T cells for TCR engineered effectors since the vast majority of these cells are T_N.

Using a protocol adopted for clinical trials (2011, 2006, 2009) we found that retroviral transduction of CB and adult T cells led to comparable EBV-TCR expression (Figure 2) in both CD8 and CD4 CB T cells. This is important, as TCR-transduced CD4 T cells can provide helper functions in vivo to maintain an effective CD8 T cell response, as well as mediate direct antitumor effects (2005, 2010).

Immediately posttransduction, the differentiation phenotype of CB T cells differed from that of adult T cells, shifting from T_N to a predominantly T_CM rather than T_EM phenotype. Moreover, phenotypic differences between the two cell sources were maintained throughout the in vitro culture. Within CD8 T cells CB cells differentiated predominantly to T_EM by day 23 whereas adult CD8 T cells shifted to a T_EMRA phenotype typical of late differentiated cells. It is not clear to what extent the increased proportion of more differentiated T cells was due to increased proliferation of these cells or maturation of T cells from subsets with a less differentiated phenotype. The relative ‘youth’ of expanded CB lymphocytes when compared to adult T cells was supported by reduced expression on cord CD8+ cells of CD57, a marker of replicative senescence and antigen-induced apoptotic death of T cells (2003). Furthermore, 85.5% of transduced CD8 CB T cells retained expression of CD27, a marker recently identified as predictive of clinical response following infusion of T cells to treat melanoma and CMV (2005, 2009).

Telomere shortening is associated with lymphocyte differentiation eventually leading to senescence and apoptosis (2012) and has been observed with cell culture in vitro and with age in vivo (1999, 2000, 2012). Telomerase maintains telomere length and supports proliferative potential but cannot fully prevent telomere shortening (2005). Ectopic telomerase expression supports extended lymphocyte proliferation in vitro (2008). In our study (Figure 6), CB T cells had longer telomeres than adult T cells and this difference was maintained after 3 weeks of antigen-driven in vitro expansion. Similarly, in 2/3 CB samples, high telomerase activity was maintained over the same culture period, whereas this activity decreased in 3/3 adult T cell samples. In one CB culture there was telomere elongation but suppressed telomerase activity at day 21. We speculate that this may be related to a negative feedback mechanism that prevents uncontrolled telomere elongation by telomerase and warrants further investigation (2009). Taken together, our results indicate that in contrast to TCR-transduced adult T-lymphocytes, CB T-lymphocytes have longer telomeres and generally maintain higher levels of telomerase activity during culture, supportive of greater proliferative and survival potential in vivo (2008).

Transduced CB cells expanded well in vitro (Figure 4) and were capable of multiple cytokine production and cytotoxic activity following antigen-specific stimulation (Figure 7). Nevertheless, they expressed lower levels of perforin and, though not statistically significant, there was a trend toward reduced cytotoxic function in CB T cells in vitro when compared to adult T cells (Figure 7). This mirrors that seen in mouse studies where less differentiated T cells display reduced cytotoxic function in vitro compared with more differentiated effectors. However, in the same study, the less differentiated T cells possessed more potent antitumor activity in vivo probably reflecting the reduced proliferative and survival potential of more differentiated cells (2005). By analogy, our results suggest that the less differentiated CB T cells, which may have reduced cytotoxic function in vitro, may prove more effective in vivo than adult T cells, although further studies are required to confirm this. In this work HLA A11-restricted EBV-specific TCR was used to explore the function of TCR-transduced CB T cells, but further studies are required to confirm that these properties are generally applicable to any TCR.

When we explored activation of CB T cells as a necessary step for retroviral transduction, we found CB T cells initially proliferated more rapidly than adult T cells (Figure 1B and C). This contradicts some reports that CB cells have increased propensity to apoptosis following activation (1999, 1999). It is unclear whether increased proliferation of CB cells reflects the differing proportions of naive and memory cells within cord and adult blood, or whether cord T_N have a distinct response to mitogenic stimulation. Recent work suggests that the development of the immune system occurs in distinct waves derived from different stem cell populations, and that fetal lymphopoiesis differs from adult lymphopoiesis with enhanced proliferation after exposure to allo-stimulation (2010). CB is at the transition between fetal and adult hematopoiesis, and the greater proliferation we observed in CB T cells may be a reflection of this.

The clinical implication of this study is the possibility of CB providing potent third-party T cells with good replicative and functional reserve for TCR engineering. Third-party ACT is effective in transplant settings where matched donor lymphocytes are unavailable. Third-party, partially HLA-matched EBV-specific T cell lines have demonstrated safety and efficacy in eradicating PTLD following SOT with minimal GVHD risk (2007, 2010). In vitro expanded third-party CMV-specific T cells have also been given successfully to a CB transplant patient with CMV encephalitis with no adverse effects (2008). Immunotherapy with CB T-lymphocytes may also be appropriate for cancer patients whose prior treatment with radio/chemotherapy and age-related thymic involution have rendered them lymphopenic with reduced numbers of T_N and T_CM subsets and CD27 expression (2010, 1997). Such cells may not have the capacity for in vivo persistence and clinical efficacy. CB T cells may also benefit patients with primary T cell dysfunction and where autologous PBMC are difficult to handle ex vivo (e.g. HIV-infected blood). Notwithstanding the potential benefits, TCR gene transfer with third-party T cells carries significant theoretical risks. Although GVHD risk is reduced in CB transplantation (2000, 2009), introducing TCRs could induce heterologous immunity including the risk of graft-versus-graft effects in SOT recipients as the introduced TCR may display unanticipated alloreactivity to normal cells (2004). The HLA A11-restricted EBV-specific TCR described here is not known to cross-react with other antigens, but some EBV-specific TCRs recognize particular alloantigens (1994). Transduced cells should therefore be checked for reactivity to patient or transplant donor cells before infusion. Since TCR-transduced CB T cells may have greater proliferative capacity, fewer cells may need to be infused compared with adult T cells thus reducing the risk of GVHD. Selective enrichment of TCR-transduced T cells with HLA: peptide multimers could reduce the required dose still further.

In summary, we have demonstrated that human CB lymphocytes have qualities suited for adoptive therapy using retrovirally transduced T cells since they can be engineered to express high-avidity functional TCR while maintaining an early differentiation phenotype that could lead to long-term in vivo persistence after infusion.

Acknowledgments

The work was supported by grants from the Howard Ostins Fund (FEC) and National Institute for Health Research (NIHR) (FEC) and by the Cancer Research UK senior research fellowship (SPL). The flow cytometry was funded by the NIHR Birmingham Biomedical Research Unit. The authors would like to thank Christine James, University of Birmingham, for providing technical support and Dr. Jennie Coppock, NHSBT, for arranging the collection of donor blood. The work received laboratory support from NHSBT Stem Cells Immunotherapy and from the NHSBT Histocompatibility and Immunogenetics Laboratory, Birmingham.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation except for G.A., who holds part time employment at Repeat Diagnostics, and P.M.L., who is a founding shareholder in Repeat Diagnostics, a company specializing in leukocyte telomere length measurements using flow-FISH.