BackgroundThere　is　an　urgent　need　for　targeted　biological　therapies　for　prostate　cancer　with　greater　efficacy　and　less　toxicity,　particularly　for　metastatic　disease,　where　current　therapies　are　not　curative.　Therapeutic　adenoviral　vectors　or　oncolytic　adenoviruses　offer　the　possibility　of　a　competent,　nontoxic　therapeutic　alternative　for　prostate　cancer.　However,　free　viral　particles　must　be　delivered　locally,　an　approach　that　does　not　address　metastatic　disease,　and　they　display　poor　tumor　penetration.　To　fully　exploit　the　potential　of　these　vectors,　we　must　develop　methods　that　improve　intratumoral　dissemination　and　allow　for　systemic　delivery.　This　study　establishes　a　proof-of-principle　rationale　for　a　novel　human　mesenchymal　stem　(stromal)　cell-based　approach　to　improving　vector　delivery　to　tumors.Methods/resultsWe　have　generated　mesenchymal　stem　cell-derived　packaging　cells　for　adenoviruses　(E1-modified　mesenchymal　stem　cells)　by　modifying　human　mesenchymal　stem　cells　with　the　adenovirus　(type　C)　E1A/B　genes　needed　for　viral　replication.　Using　cell-based　assays,　we　have　demonstrated　that　two　adenoviral　vectors,　replication-defective　adenovirus　expressing　p14　and　p53　or　conditionally　replicating　oncolytic　adenovirus,　packaged　by　E1A/B-modified　mesenchymal　stem　cells,　suppress　the　growth　of　prostate　cancer　cells　in　culture.　Using　subcutaneous　xenograft　models　for　human　prostate　cancer　in　mice,　we　have　shown　that　E1A/B-modified　mesenchymal　stem　cells　display　tumor　tropism　in　tumor-bearing　nude　mice,　that　E1A/B-modified　mesenchymal　stem　cells　disseminate　well　within　tumors,　and　that　replication-defective　adenovirus　expressing　p14　and　p53　or　conditionally　replicating　oncolytic　adenovirus-loaded　E1-modified　mesenchymal　stem　cells　suppresses　tumor　growth　in　mice.ConclusionThe　results　show　that　this　approach,　if　optimized,　could　circumvent　the　obstacles　to　efficient　gene　delivery　encountered　with　current　gene　delivery　approaches　and　provide　an　effective,　nontoxic　therapeutic　alternative　for　metastatic　disease.