gms | German Medical Science

Introduction: Therapeutic angiogenesis aims at generating new blood vessels by delivering growth factors such as Vascular Endothelial Growth Factor (VEGF), and thus is a promising treatment for patients with peripheral arterial disease. It has previously been shown that a discrete threshold in VEGF dose exists, below which normal stable capillaries are induced and above which angioma growth occurs [1]. A discrete threshold in VEGF dose exists, below which normal stable capillaries are induced and above which angioma growth occurs. However, VEGF dose must be homogeneously distributed in the microenvironment, as VEGF remains localized around each producing cell. By using retrovirally-transduced myoblasts, normal angiogenesis can be induced by implanting clonal populations, in which every cell expresses the same VEGF level. While providing proof-of-concept, this approach is not directly applicable to clinical practice. Therefore, we developed a Fluorescence Activated Cell Sorting (FACS)-based method to allow the rapid isolation of engineered myoblasts homogeneously expressing predictable levels of VEGF from a heterogeneous primary population in vitro, in order to achieve controlled microenvironmental VEGF doses in vivo.

Materials and methods: Primary mouse myoblasts were transduced with a retroviral construct in which the VEGF164 gene is linked through an Internal Ribosomal Entry Site (IRES) to a truncated version of CD8a, acting as a reporter gene. Therefore, changes in the level of VEGF expression are reflected by the cell-surface expression of CD8a, which can be detected and quantified on live cells by FACS.

Results: To determine the relationship between expression of VEGF (by ELISA) with that of CD8 (by FACS) at the single cell level, a standard curve was produced by isolating clones from all regions of the expression spectrum of the heterogeneous polyclonal population. VEGF and CD8 expression was found to be stable during expansion over 49 cell doublings in 5 sample clones producing levels of VEGF in a range of 2-142 ng/106 cells/day. A set of clones was implanted in the ear muscle of SCID mice to correlate VEGF expression with the 3-dimensional morphology of induced vessels after four weeks in whole-mount preparations. We selected the clone expressing the highest VEGF level (40 ng/106 cells/day) which was still safe (below the threshold). This clone was used as a reference to sort the cells expressing similar levels of VEGF out of the heterogeneous polyclonal population, based on their CD8a expression. We tested up to 3 successive rounds of sorting, with 2 different gates (large and narrow), to optimize the purity of the selection. VEGF and CD8a expression of the sorted cells was stable during in vitro expansion over 23 doublings. When implanted in vivo, the heterogeneous population always caused angioma growth by four weeks. In contrast, the sorted populations caused exclusively normal angiogenesis in 85% of the animals four weeks and three months after implantation, respectively. In 8 (out of 43) implantation sites, single instances of small aberrant vessels were found, but only with large-gate selected populations. Therefore, selective sorting with a narrow gate appears to reliably avoid expression of excessive VEGF levels.

Discussion: The ability to isolate homogeneous populations expressing predictable levels of the transgene might allow the optimization of both efficacy and safety in cell-based strategies for VEGF gene delivery.