Identification of preferentially targeted tumor-associated antigens in melanoma patients via mRNA stimulation of CD8+ blood lymphocytes

Until now, therapeutic vaccination of cancer patients has mainly relied on rather few T cell epitopes processed from structurally normal shared tumor antigens and presented by frequent HLA alleles. So far the design of these studies has not addressed the individuality of tumor-host interactions, which are not only determined by the antigenic tumor phenotype or the natural HLA polymorphism, but also by the individual T cell repertoire. The procedure described herein was developed to identify the preferential targets of the individual repertoire from a panel of known shared tumor-associated antigens. Lymphocytes were isolated from the peripheral blood of cancer patients or healthy donors and stimulated twice with autologous mRNA-transfected FastDC (Dauer et al., J Immunol. 170:4069, 2003). FastDC were generated from blood monocytes and separately transfected via lipofection with in vitro transcribed mRNAs encoding the panel antigens. Responder lymphocytes were tested on day 12 in a 20-hour IFN-g ELISPOT assay for recognition of 293T cells co-transfected pairwise with plasmids encoding the stimulation antigens and the respective individual’s HLA class I alleles. In a first step, stimulation parameters were optimized for the detection of anti-HCMV pp65 responses. A maximum amplification of pp65-specific CD8+ T cell responses was obtained at a rather low IL-2 concentration (25 IU/ml) and at a minimum APC-to-effector ratio of 1:10. Addition of IL-4, IL-7 or IL-15 did not substantially improve the stimulatory potential. The test was applied to the human melanoma models D05 and MZ2, in both of which multiple T cell-defined antigens had previously been identified by expression screening. Blood lymphocytes were stimulated in parallel with autologous tumor cells and with mRNA-transfected FastDC. In D05, T cell reactivities against three out of eleven epitopes induced by stimulation with tumor cells were also found after stimulation with mRNA-transfected FastDC. Two further T cell target epitopes were identified with mRNA but not with tumor cell stimulation. In MZ2, T cell responses against five distinct epitopes were detected on day 12 after stimulation with mRNA transfectants. The same responses were detectable after stimulation with tumor cells only on day 32. mRNA stimulations against 21 tumor-associated antigens in addition to HCMV pp65 were performed in four healthy individuals. In all cases, CD8+ T cells against HCMV pp65 could be expanded. Among tumor-associated antigens, only reactivity against Melan-A/MART-1 in association with HLA-A*0201 was detectable in one of the donors. The vaccination of patients with targets a priori known to be recognized by their T cell repertoire may help to improve the outcome of therapeutic vaccination.

Mass spectrometric identification of Varicella-Zoster Virus (VZV) proteins recognized by human T cells

Primary varicella-zoster virus (VZV) infection during childhood leads to varicella commonly known as chickenpox. After primary infection has occurred VZV establishes latency in the host. During subsequent lifetime the virus can cause reactivated infection clinically known as herpes zoster or shingles. In immunodeficient patients’ dissemination of the virus can lead to life-threatening disease. Withdrawal of acyclovir drug prophylaxis puts allogeneic hematopoietic stem-cell transplantation (HSCT) patients at increased risk for herpes zoster as long as VZV-specific cellular immunity is impaired. Although an efficient live attenuated VZV vaccine for zoster prophylaxis exists, it is not approved in immunocompromised patients due to safety reasons. Knowledge of immunogenic VZV proteins would allow designing a noninfectious nonhazardous subunit vaccine suitable for patients with immunodeficiencies. The objective of this study was to identify T cell defined virus proteins of a VZV-infected Vero cell extract that we have recently described as a reliable antigen format for interferon-gamma (IFN-γ) enzyme-linked immunosorbent spot (ELISpot) assays (Distler et al. 2008). We first separated the VZV-infected/-uninfected Vero cell extracts by size filtration and reverse-phase high performance liquid chromatography (RP-HPLC). The collected fractions were screened for VZV reactivity with peripheral blood mononuclear cells (PBMCs) of VZV-seropositive healthy individuals in the sensitive IFN-γ ELISpot assay. Using this strategy, we successfully identified bioactive fractions that contained immunogenic VZV material. VZV immune reactivity was mediated by CD4+ memory T lymphocytes (T cells) of VZV-seropositive healthy individuals as demonstrated in experiments with HLA blockade antibodies and T cell subpopulations already published by Distler et al. We next analyzed the bioactive fractions with electrospray ionization mass spectrometry (ESI-MS) techniques and identified the sequences of three VZV-derived proteins: glycoprotein E (gE); glycoprotein B (gB), and immediate early protein 62 (IE62). Complementary DNA of these identified proteins was used to generate in vitro transcribed RNA for effective expression in PBMCs by electroporation. We thereby established a reliable and convenient IFN-γ ELISPOT approach to screen PBMCs of healthy donors and HSCT patients for T cell reactivity to single full-length VZV proteins. Application in 10 VZV seropositive healthy donors demonstrated much stronger recognition of glycoproteins gE and gB compared to IE62. In addition, monitoring experiments with ex vivo PBMCs of 3 allo-HSCT patients detected strongly increased CD4+ T cell responses to gE and gB for several weeks to months after zoster onset, while IE62 reactivity remained moderate. Overall our results show for the first time that VZV glycoproteins gE and gB are major targets of the post-transplant anti-zoster CD4+ T cell response. The screening approach introduced herein may help to select VZV proteins recognized by memory CD4+ T cells for inclusion in a subunit vaccine, which can be safely used for zoster prophylaxis in immunocompromised HSCT patients.

In vitro characterization of human AML-reactive CTL clones generated from the naive subset of healthy donors and adoptive transfer into leukemia-engrafted NSG mice

Acute myeloid leukemia (AML) is a very aggressive cancer of the hematopoietic system. Chemotherapy and immunotherapeutical approaches including hematopoietic stem cell transplantation (HSCT) and donor lymphocyte infusion (DLI) are the only curative options available. The beneficial graft-versus-leukemia (GVL) effect of cellular immunotherapy is mostly mediated by donor-derived CD8+ T lymphocytes that recognize minor histocompatibility antigens (mHags) and leukemia-associated antigens (LAAs) presented on the surface of AML blasts (Falkenburg et al. 2008; Kolb 2008). A main complication is graft-versus-host disease (GVHD) that can be induced when cytotoxic T lymphocytes (CTLs) recognize broadly expressed antigens. To reduce the risk of GVHD, specific allogeneic T-cell therapy inducing selective GVL responses could be an option (Barrett & Le Blanc 2010; Parmar et al. 2011; Smits et al. 2011). This requires efficient in vitro strategies to generate AML-reactive T cells with an early differentiation phenotype as well as vigorous effector functions and humanized mouse models to analyze the anti-leukemic potential of adoptively transferred T cells in vivo. In this study, AML-reactive CTL clones and oligoclonal T-cell lines could be reliably generated from the naive subset of healthy HLA-class I-identical donors by stimulation with primary AML blasts in mini-mixed-lymphocyte / leukemia cultures (MLLCs) in eight different patient / donor pairs. These CTLs were promising candidates for cellular immunotherapy because of their relatively early differentiation phenotype and strong proliferative and lytic capabilities. The addition of the common γ-chain cytokine IL-21 to the stimulation protocol enabled more precursors to develop into potent leukemia-reactive CTLs, presumably by its beneficial effects on cell survival and antigen-specific proliferation during the first weeks of cultures. It also strengthened the early-stage phenotype. Three long-term cultured CTLs exemplarily transferred into leukemia-engrafted immunodeficient NSG mice mediated a significant reduction of the leukemic burden after a single transfusion. These results demonstrate that CTL clones with reactivity to patient-derived AML blasts can be isolated from the naive compartment of healthy donors and show potent anti-leukemic effects in vivo. The herein described allo-MLLC approach with in vitro “programmed” naive CTL precursors independent of a HSCT setting is a valuable alternative to the conventional method of isolating in vivo primed donor CTLs out of patients after transplantation (Kloosterboer et al. 2004; Warren et al. 2010). This would make leukemia-reactive CTLs already available at the time point of HSCT, when residual leukemia disease is minimal and the chances for complete leukemia eradication are high. Furthermore, leukemia-reactive CTLs effectively expanded by this in vitro protocol can be used as screening populations to identify novel candidate LAAs and mHags for antigen-specific immunotherapy.

Preclinical in vitro and in vivo characterization of the fully human monoclonal IgM antibody KBPA101 specific for Pseudomonas aeruginosa serotype IATS-O11

Pseudomonas aeruginosa infection in ventilator-associated pneumonia is a serious and often life-threatening complication in intensive care unit patients, and new treatment options are needed. We used B-cell-enriched peripheral blood lymphocytes from a volunteer immunized with a P. aeruginosa O-polysaccharide-toxin A conjugate vaccine to generate human hybridoma cell lines producing monoclonal antibodies specific for individual P. aeruginosa lipopolysaccharide serotypes. The fully human monoclonal antibody secreted by one of these lines, KBPA101, is an IgM/kappa antibody that binds P. aeruginosa of International Antigenic Typing System (IATS) serotype O11 with high avidity (5.81 x 10(7) M(-1) +/- 2.8 x 10(7) M(-1)) without cross-reacting with other serotypes. KBPA101 specifically opsonized the P. aeruginosa of IATS O11 serotype and mediated complement-dependent phagocytosis in vitro by the human monocyte-like cell line HL-60 at a very low concentration (half-maximal phagocytosis at 0.16 ng/ml). In vivo evaluation of KBPA101 demonstrated a dose-response relationship for protection against systemic infections in a murine burn wound sepsis model, where 70 to 100% of animals were protected against lethal challenges with P. aeruginosa at doses as low as 5 microg/animal. Furthermore, a high efficacy of KBPA101 in protection from local respiratory infections in an acute lung infection model in mice was demonstrated. Preclinical toxicology evaluation on human tissue, in rabbits, and in mice did not indicate any toxicity of KBPA101. Based on these preclinical findings, the first human clinical trials have been initiated.

A comparison of acute and long-term effects of industrial multiwalled carbon nanotubes on human lung and immune cells in vitro

The close resemblance of carbon nanotubes to asbestos fibers regarding their high aspect ratio, biopersistence and reactivity increases public concerns on the widespread use of these materials. The purpose of this study was not only to address the acute adverse effects of industrially produced multiwalled carbon nanotubes (MWCNTs) on human lung and immune cells in vitro but also to further understand if their accumulation and biopersistence leads to long-term consequences or induces adaptive changes in these cells. In contrast to asbestos fibers, pristine MWCNTs did not induce overt cell death in A549 lung epithelial cells and Jurkat T lymphocytes after acute exposure to high doses of this material (up to 30 g/ml). Nevertheless, very high levels of reactive oxygen species (ROS) and decreased metabolic activity were observed which might affect long-term viability of these cells. However, the continuous presence of low amounts of MWCNTs (0.5 g/ml) for 6 months did not have major adverse long-term effects although large amounts of nanotubes accumulated at least in A549 cells. Moreover, MWCNTs did not appear to induce adaptive mechanisms against particle stress in long-term treated A549 cells. Our study demonstrates that despite the high potential for ROS formation, pristine MWCNTs can accumulate and persist within cells without having major long-term consequences or inducing adaptive mechanisms.

Disruption of SIRPα signaling in macrophages eliminates human acute myeloid leukemia stem cells in xenografts

Although tumor surveillance by T and B lymphocytes is well studied, the role of innate immune cells, in particular macrophages, is less clear. Moreover, the existence of subclonal genetic and functional diversity in some human cancers such as leukemia underscores the importance of defining tumor surveillance mechanisms that effectively target the disease-sustaining cancer stem cells in addition to bulk cells. In this study, we report that leukemia stem cell function in xenotransplant models of acute myeloid leukemia (AML) depends on SIRPα-mediated inhibition of macrophages through engagement with its ligand CD47. We generated mice expressing SIRPα variants with differential ability to bind human CD47 and demonstrated that macrophage-mediated phagocytosis and clearance of AML stem cells depend on absent SIRPα signaling. We obtained independent confirmation of the genetic restriction observed in our mouse models by using SIRPα-Fc fusion protein to disrupt SIRPα-CD47 engagement. Treatment with SIRPα-Fc enhanced phagocytosis of AML cells by both mouse and human macrophages and impaired leukemic engraftment in mice. Importantly, SIRPα-Fc treatment did not significantly enhance phagocytosis of normal hematopoietic targets. These findings support the development of therapeutics that antagonize SIRPα signaling to enhance macrophage-mediated elimination of AML.

DOCK2 is required for chemokine-promoted human T lymphocyte adhesion under shear stress mediated by the integrin alpha4beta1

The alpha4beta1 integrin is an essential adhesion molecule for recruitment of circulating lymphocytes into lymphoid organs and peripheral sites of inflammation. Chemokines stimulate alpha4beta1 adhesive activity allowing lymphocyte arrest on endothelium and subsequent diapedesis. Activation of the GTPase Rac by the guanine-nucleotide exchange factor Vav1 promoted by CXCL12 controls T lymphocyte adhesion mediated by alpha4beta1. In this study, we investigated the role of DOCK2, a lymphocyte guanine-nucleotide exchange factor also involved in Rac activation, in CXCL12-stimulated human T lymphocyte adhesion mediated by alpha4beta1. Using T cells transfected with DOCK2 mutant forms defective in Rac activation or with DOCK2 small interfering RNA, we demonstrate that DOCK2 is needed for efficient chemokine-stimulated lymphocyte attachment to VCAM-1 under shear stress. Flow chamber, soluble binding, and cell spreading assays identified the strengthening of alpha4beta1-VCAM-1 interaction, involving high affinity alpha4beta1 conformations, as the adhesion step mainly controlled by DOCK2 activity. The comparison of DOCK2 and Vav1 involvement in CXCL12-promoted Rac activation and alpha4beta1-dependent human T cell adhesion indicated a more prominent role of Vav1 than DOCK2. These results suggest that DOCK2-mediated signaling regulates chemokine-stimulated human T lymphocyte alpha4beta1 adhesive activity, and that cooperation with Vav1 might be required to induce sufficient Rac activation for efficient adhesion. In contrast, flow chamber experiments using lymph node and spleen T cells from DOCK2(-/-) mice revealed no significant alterations in CXCL12-promoted adhesion mediated by alpha4beta1, indicating that DOCK2 activity is dispensable for triggering of this adhesion in mouse T cells, and suggesting that Rac activation plays minor roles in this process.

P-glycoprotein expression in lamina propria lymphocytes of duodenal biopsy samples in dogs with chronic idiopathic enteropathies

P-glycoprotein (p-gp) is a transmembrane protein functioning as a drug-efflux pump in the intestinal epithelium. Human patients with inflammatory bowel disease (IBD) who fail to respond to treatment with steroids express high levels of p-gp in lamina propria lymphocytes. The purpose of this study was to investigate p-gp expression in duodenal biopsy samples of dogs with chronic enteropathies and to evaluate the expression of p-gp after treatment with a known inducer of p-gp (prednisolone). Duodenal biopsy samples from 48 dogs were evaluated immunohistochemically with the mouse monoclonal antibody C219 for expression of p-gp in lamina propria lymphocytes. Biopsy samples were available from 15 dogs after treatment with prednisolone and 16 dogs after dietary therapy alone ("elimination diet"). Treatment with prednisolone resulted in an increase in p-gp expression (P=0.005). In contrast, dietary treatment alone produced no significant change in p-gp expression (P=0.59). A low p-gp score before initiation of steroid treatment was significantly associated with a positive response to treatment (P=0.01). These results indicate that lamina propria lymphocyte expression of p-gp is upregulated after prednisolone treatment in dogs with IBD, and that mucosal expression of p-gp may be of value in predicting the response to therapy.

Exovesicles from human activated dendritic cells fuse with resting dendritic cells, allowing them to present alloantigens

Dendritic cells (DCs) can release microvesicles, but the latter's numbers, size, and fate are unclear. Fluorescently labeled DCs were visualized by laser-scanning microscopy. Using a Surpass algorithm, we were able to identify and quantify per cell several hundred microvesicles released from the surface of stimulated DCs. We show that most of these microvesicles are not of endocytic origin but result from budding of the plasma membrane, hence their name, exovesicle. Using a double vital staining, we show that exovesicles isolated from activated DCs can fuse with the membrane of resting DCs, thereby allowing them to present alloantigens to lymphocytes. We concluded that, within a few hours from their release, exovesicles may amplify local or distant adaptive immunological response.

Granzyme B, a novel mediator of allergic inflammation: its induction and release in blood basophils and human asthma

Histamine, leukotriene C4, IL-4, and IL-13 are major mediators of allergy and asthma. They are all formed by basophils and are released in particularly large quantities after stimulation with IL-3. Here we show that supernatants of activated mast cells or IL-3 qualitatively change the makeup of granules of human basophils by inducing de novo synthesis of granzyme B (GzmB), without induction of other granule proteins expressed by cytotoxic lymphocytes (granzyme A, perforin). This bioactivity of IL-3 is not shared by other cytokines known to regulate the function of basophils or lymphocytes. The IL-3 effect is restricted to basophil granulocytes as no constitutive or inducible expression of GzmB is detected in eosinophils or neutrophils. GzmB is induced within 6 to 24 hours, sorted into the granule compartment, and released by exocytosis upon IgE-dependent and -independent activation. In vitro, there is a close parallelism between GzmB, IL-13, and leukotriene C4 production. In vivo, granzyme B, but not the lymphoid granule marker granzyme A, is released 18 hours after allergen challenge of asthmatic patients in strong correlation with interleukin-13. Our study demonstrates an unexpected plasticity of the granule composition of mature basophils and suggests a role of granzyme B as a novel mediator of allergic diseases.

Telomere length in human natural killer cell subsets

Natural killer (NK) cells are cytotoxic cells that play a critical role in the innate immune response against infections and tumors. In the elderly, the cytotoxic function of NK cells is often compromised. Telomeres progressively shorten with each cell division and with age in most somatic cells eventually leading to chromosomal instability and cellular senescence. We studied the telomere length in NK cell subsets isolated from peripheral blood using "flow FISH," a method in which the hybridization of telomere probe in cells of interest is measured relative to internal controls in the same tube. We found that the average telomere length in human NK cells decreased with age as was previously found for human T lymphocytes. Separation of adult NK cells based on CD56 and CD16 expression revealed that the telomere length was significantly shorter in CD56(dim)CD16(+) (mature) NK cells compared to CD56(bright)CD16(-) (immature) NK cells from the same donor. Furthermore, sorting of NK cells based on expression of activation markers, such as NKG2D and LFA-1, revealed that NK cells expressing these markers have significantly shorter telomeres. Telomere fluorescence was very heterogeneous in NK cells expressing CD94, killer inhibitory receptor (KIR), NKG2A, or CD161. Our observations indicate that telomeric DNA in NK cells is lost with cell division and with age similar to what has been observed for most other hematopoietic cells. Telomere attrition in NK cells is a plausible cause for diminished NK cell function in the elderly.

The cytochrome P450 aromatase lacking exon 5 is associated with a phenotype of nonclassic aromatase deficiency and is also present in normal human steroidogenic tissues

OBJECTIVE: The previously described c655G>A mutation of the human cytochrome P450 aromatase gene (P450aro, CYP19) results in aberrant splicing due to disruption of a donor splice site. To explain the phenotype of partial aromatase deficiency observed in a female patient described with this mutation, molecular consequences of the c655G>A mutation were investigated. DESIGN: To investigate whether the c655G>A mutation causes an aberrant spliced mRNA lacking exon 5 (-Ex5), P450aro RNA was analysed from the patient's lymphocytes by reverse transcription polymerase chain reaction (RT-PCR) and by splicing assays performed in Y1 cells transfected with a P450aro -Ex5 expression vector. Aromatase activity of the c655G>A mutant was predicted by three dimensional (3D) protein modelling studies and analysed in transiently transfected Y1 cells. Exon 5 might be predicted as a poorly defined exon suggesting a susceptibility to both splicing mutations and physiological alternative splicing events. Therefore, expression of the -Ex5 mRNA was also assessed as a possibly naturally occurring alternative splicing transcript in normal human steroidogenic tissues. PATIENTS: An aromatase deficient girl was born with ambiguous genitalia. Elevated serum LH, FSH and androgens, as well as cystic ovaries, were found during prepuberty. At the age of 8.4 years, spontaneous breast development and a 194.6 pmol/l serum oestradiol level was observed. RESULTS: The -Ex5 mRNA was found in lymphocytes of the P450aro deficient girl and her father, who was a carrier of the mutation. Mutant minigene expression resulted in complete exon 5 skipping. As expected from 3D protein modelling, -Ex5 cDNA expression in Y1 cells resulted in loss of P450aro activity. In addition, the -Ex5 mRNA was present in placenta, prepubertal testis and adrenal tissues. CONCLUSIONS: Alternative splicing of exon 5 of the CYP19 gene occurs in the wild type (WT) as well as in the c655G>A mutant. We speculate that for the WT it might function as a regulatory mechanism for aromatization, whereas for the mutant a relative prevalence of the shorter over the full-length protein might explain the phenotype of partial aromatase deficiency.

Determinants of phosphatidylinositol-4-phosphate 5-kinase type Iγ90 uropod location in T-lymphocytes and its role in uropod formation

We have previously identified phosphatidylinositol-4-phosphate 5-kinase type I (PIPKI)γ90 as a T cell uropod component. However, the molecular determinants and functional consequences of its localization remain unknown. In this report, we seek to better understand the mechanisms involved in PIPKIγ90 uropod targeting and the role that PIPKIγ90 plays in T cell uropod formation. During T cell activation, PIPKIγ90 cocaps with the membrane microdomain-associated proteins flotillin-1 and -2 and accumulates in the uropod. We report that the C-terminal 26 amino acid extension of PIPKIγ90 is required for its localization to the uropod. We further use T cells from PIPKIγ90(-/-) mice and human T cells expressing a kinase-dead PIPKIγ90 mutant to examine the role of PIPKIγ90 in a T cell uropod formation. We find that PIPKIγ90 deficient T cells have elongated uropods on ICAM-1. Moreover, in human T cells overexpression of PIPKIγ87, a naturally occurring isoform lacking the last 26 amino acids, suppresses uropod formation and impairs capping of uropod proteins such as flotillins. Transfection of human T cells with a dominant-negative mutant of flotillin-2 in turn attenuates capping of PIPKIγ90. Our data contribute to the understanding of the molecular mechanisms that regulate T cell uropod formation.

Depressed type 1 cytokine synthesis by superantigen-activated CD4+ T cells of women with human papillomavirus-related high-grade squamous intraepithelial lesions.

Carcinoma of the cervix is causally related to infection with the human papillomavirus (HPV), and T cells play a pivotal role in the immune response of the host to rid itself of HPV infection. Therefore, we assessed the T-cell function of women with HPV-related cervical neoplasia against a superantigen, Staphylococcus enterotoxin B (SEB). Each woman provided a cervical brush specimen for HPV DNA testing and Papanicolaou (Pap) smears for the staging of cervical lesions. They also provided a blood specimen for determination of the ability of CD4(+) T and CD8(+) T cells to synthesize Th1 (interleukin-2 [IL-2], gamma interferon [IFN-gamma], and tumor necrosis factor alpha [TNF-alpha]) and Th2 (IL-10) cytokines in response to activation with SEB. Compared with control subjects with self-attested negative Pap smears, women with high-grade squamous intraepithelial lesions (HSIL) had significantly lower percentages of activated CD4(+) T cells that produced IL-2 (P = 0.045), IFN-gamma (P = 0.040), and TNF-alpha (P = 0.015) and a significantly lower percentage of activated CD8(+) T cells that produced IL-2 (P < 0.01). These data indicate that women with HPV-related cervical HSIL show a decrease in Th1 cytokine production by activated CD4(+) T cells and suggested that compromised T-helper functions may negatively impact the function of cytotoxic CD8(+) T cells.

Treatment of solid organ transplant recipients with autologous Epstein Barr virus-specific cytotoxic T lymphocytes (CTLs).

We have investigated the in vivo safety, efficacy, and persistence of autologous Epstein Barr virus (EBV)-specific cytotoxic T lymphocytes (CTLs) for the treatment of solid organ transplant (SOT) recipients at high risk for EBV-associated posttransplantation lymphoproliferative disease (PTLD). EBV-CTLs generated from 35 patients expanded with normal kinetics contained both CD8 and CD4 lymphocytes and produced significant specific killing of autologous EBV-transformed B lymphoblastoid cell lines (LCLs). Twelve SOT recipients at high risk for PTLD, or with active disease, received autologous CTL infusions without toxicity. Real-time polymerase chain reaction (PCR) monitoring of EBV-DNA showed a transient increase in plasma EBV-DNA suggestive of lysis of EBV-infected cells, although there was no consistent decrease in virus load in peripheral-blood mononuclear cells. Interferon-gamma enzyme-linked immunospot (ELISPOT) assay and tetramer analysis showed an increase in the frequency of EBV-responsive T cells, which returned to preinfusion levels after 2 to 6 months. None of the treated patients developed PTLD. One patient with liver PTLD showed a complete response, and one with ocular disease has had a partial response stable for over one year. These data are consistent with an expansion and persistence of adoptively transferred EBV-CTLs that is limited in the presence of continued immunosuppression but that nonetheless produces clinically useful antiviral activity.