Retroviral Gene Transfer of Suicide Genes in Human T Lymphocytes: Improved Transduction Conditions for a Clinical Protocol.

Aiming to control GVHD by means of the transfer of suicide genes into donor T cells, we seeked to optimise the transduction of human T cells using immobilized anti-CD3i/28i MoAbs for T cell stimulation. In previous studies we showed that reducing the concentration of anti-CD3i, from 1,000ng/ml down to 1ng/ml, helps to preserve the cytotoxicity of the samples to an allogeneic stimulus and maintains the susceptibility of the T cells to the retroviral transduction. Aiming to reduce the cost of the manipulation procedure, we also tried to reduce the amount of anti-CD28 MoAb during the activation (pre-infection) and expansion (post-infection) steps. To investigate this issue, different doses of anti-CD28i were combined either with 1,000 ng/ml or with 10 ng/ml of anti-CD3i. In contrast to the effects observed with anti-CD3i, lowering the dose of anti-CD28i, from 1,000 ng/ml to a range of 500 to 10 ng/ml significantly decreased the transduction efficacy, as well as the total number of transduced T cells, regardless of the concentration of anti-CD3i used in the experiments. With the purpose of minimizing potential effects mediated by retroviral supernatants on the viability and growth of the T cells, we investigated whether standard infections with supernatants could be replaced by alternative or complementary transduction procedures based on the pre-loading of the retroviral vectors in fibronectin-coated bags. In these experiments, fibronectin-coated cell culture bags were pre-loaded with retroviral vectors by means of a 3h-incubation with the infective supernatants. Thereafter, peripheral blood lymphocytes were added to the culture bags, incubated O/N and then subjected to further infection cycles, either with pre-loaded vectors or with infective supernatants. Performing only one cycle of infection with either method, rendered similar low yields of transduced cells compared to results obtained after 2 infection cycles. Under these conditions, we found no differences between both methods, neither with respect to the transduction efficacy nor with the total number of transduced cells. Conducting a third infection cycle, either using the pre-loading or the supernatant infection method, did not increase significantly the transduction efficacy of the samples. Also, to simplify and improve the safety of the manipulation procedure, we studied alternatives to the use of human allogeneic or autologous serum for the supplement of the cell culture medium (RPMI). In these experiments, RPMI supplemented with 10% autologous serum was replaced by the CellGro and X-Vivo 10 serum-free media. Similar transduction efficacies and similar number of transduced cells were generated as a result of the manipulation of the samples with either medium. We conclude that optimised conditions of T cell stimulation and gene expression are obtained with 1ng/ml anti-CD3i and 1,000 ng/ml anti CD28i, and that conventional supernatant infections conducted with serum-supplemented media can be partially or completely replaced by pre-loading infections in serum-free media. We propose these manipulations to simplify and to reduce risks and costs associated to the transduction of human T cells; something that is of particular interest for the development of clinical protocols that require the infusion of large numbers of transduced T lymphocytes.