Adoptively transferred transplant donor or third party donor derived CMV-specific T-cells (CMV-CTLs) can effectively prevent and treat CMV disease in HSCT recipients. T-cells (TC) respond to specific viral epitopes when presented by HLA class-I and class-II alleles on infected cells. We have infused third party CTLs from HLA partially matched donors that were chosen based on in-vitro CTL activity against epitopes presented by HLA alleles shared by the recipient, and this approach has afforded a 60% response rate among patients treated for CMV viremia and/or disease. A recurring feature of TC generated in-vitro or directly selected in-vivo is the striking preponderance of TC specific for 1-2 immunodominant epitopes presented by specific HLA alleles. However, the functional activity of such immunodominant CMV-CTLs as compared to CTLs directed against subdominant CMV epitopes has not been evaluated. To directly compare the activity of CMV-CTLs directed against immunodominant and subdominant epitopes, we developed an in-vivo model using human colon carcinoma cells transduced with CMVpp65 as a surrogate system. Human colon carcinoma cells co-expressing HLA A0201 and A2402 (cocapp65), and a melanoma cell line lacking expression of these HLA alleles (melpp65) were each transduced to also express CMVpp65 and a GFP-firefly luciferase transgene. Subcutaneous (s.c.) inoculation of 1 x 105 tumor cellsinto NOD/Scid-IL2Rgc-KO/J mice (NSG) provided consistent engraftment. CMV-CTLs responsive to either the immunodominant NLV epitope presented by HLA A0201 (A2-NLV) or the subdominant QYD epitope presented by HLA A2402 (A24-QYD) were generated from 2 donors, each co-inheriting HLA A0201 and A2402, using NIH 3T3 artificial antigen presenting cells, each expressing a single HLA class-I allele (A0201 or A2402), B7.1, LFA-3 and ICAM1.

Groups of 5-6 NSG mice, each bearing 2 established tumors: cocapp65 and melpp65 (control), were intravenously injected with 2-4 x 106 of tetramer+ A2-NLV or A24-QYD specific CMV-CTLs per mouse. Control animals either did not receive any TC, or received HLA B0801- LTM specific CMV-CTLs, and IL-2 was given 2 x/week to all groups. Tumor growth was monitored for 6-8 weeks by bioluminescent imaging.

Both A2-NLV and A24-QYD CMV-CTLs significantly suppressed growth of cocapp65 tumors in all treated animals with infusion of equivalent doses of epitope specific tet [+] T-cells (p = ns). No tumor suppression was observed in control animals, and there was no evidence of GvHD in CTL treated animals. A2-NLV CTLs eradicated cocapp65 tumor in 4/10 treated animals, while A24-QYD CTLs did not completely eradicate the tumor in any treated animal. In all A2-NLV treated animals, the cocapp65 remained suppressed until end of study (6 wks), while the melpp65 continued to grow. Treatment with A24-QYD CTLs induced tumor suppression after a time lag, as evidenced by initial cocapp65 growth for 10-14 days followed by suppression for 4-5 weeks and then stabilization of tumor size. In subsequent studies with animals bearing single cocapp65 xenografts, and infused with either A2-NLV or A24-QYD CTLs, the tumors remained suppressed for upto 8 weeks after A2-NLV CTL infusion, while recurrent tumor growth was observed 5 weeks after A24-QYD CTL infusion. Studies detailing the relative accumulation of CTLs within tumor tissue, as well as engraftment of CTLs in these animals are in progress. Phenotypic analysis of the infused epitope specific TC demonstrated 91-95% TEM with 5-9% TCM, with a highly restricted oligoclonal TCRVβ repertoire.

This model provides a platform for direct comparative evaluation of the in-vivo cytotoxic activity of epitope specific TC. These studies demonstrate that CMV-CTLs responsive to both immunodominant as well as subdominant epitopes, that are generated using AAPC from the same donor co-inheriting the presenting HLA alleles, can suppress the growth of clonogenic human carcinoma cells co-expressing an immunogenic viral antigen in-vivo at equivalent doses of antigen specific TC. However, less robust antigen specific cytotoxic activity was demonstrated by the subdominant A24-QYD CMV CTLs, which maybe reflective of reduced in-vivo proliferation, effector function or persistence of such TC. Experiments of define these variables contributing to disparities in the in-vivo CTL activity are in progress.


Hasan:Atara Biotherapeutics: Research Funding. O'Reilly:Atara Biotherapeutics: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution