RESUMEN
BACKGROUND: The lack of an ideal cell type that can be easily acquired, modified to produce insulin, and re-implanted has been a limitation for ex vivo insulin gene therapy. Canine diabetes is currently treated with human insulin and is a good model for human diabetes. Mesenchymal stromal cells (MSCs) are a promising candidate cell type for gene therapy. In the present study, we optimised insulin production using lentiviral transduced canine MSCs (cMSCs), aiming to evaluate their ability for use as surrogate beta cells. METHODS: Canine MSCs were derived from bone marrow and validated by measuring the expression of MSC lineage specific markers. Lentivirus vectors encoding the proinsulin gene (with or without a Kozak sequence) under the control of spleen focus forming virus, cytomegalovirus, elongation factor 1α and simian virus 40 promotors were generated and used to transduce primary cMSCs and a hepatocyte cell line. The insulin-producing capacity of transduced primary cMSCs was assessed by measuring the concentration of C-peptide produced. RESULTS: Primary cMSC could be readily expanded in culture and efficiently transduced using lentiviral vectors encoding proinsulin. Increasing the multiplicity of infection from 3 to 20 led to an increase in C-peptide secretion (from 1700 to 4000 pmol/l). The spleen focus forming virus promoter conferred the strongest transcriptional ability. CONCLUSIONS: The results of the present study suggest that optimised lentiviral transduction of the insulin gene into primary cMSCs renders these cells capable of secreting insulin over both the short- and long-term, in sufficient quantities in vitro to support their potential use in insulin gene therapy.