Engineering the Pericellular Matrix: Lentiviral Transduction of Human Mesenchymal Stem Cells

8. Data was normalized to corresponding non-infected controls. Engineering the Pericellular Matrix: Lentiviral Transduction of Human Mesenchymal Stem Cells Thakore, P I; Twomey, J D; Rastogi, A +Hsieh, A H +University of Maryland, College Park, MD; University of Maryland, Baltimore, MD hsieh@umd.edu INTRODUCTION: Human mesenchymal stem cells (hMSCs) have demonstrated great potential in the development of regenerative therapies for load-bearing tissues, such as articular cartilage (AC) and the intervertebral disc (IVD). Chemical and mechanical factors have been successful in inducing differentiation to the general chondrocyte phenotype. More finely-tuned control over the cellular micromechanical environment, however, would be useful for generating heterogeneous regions of mechanosensitivity. This may be useful for replicating the distinct features within each tissue type, including collagen fibril organization and proteoglycan content, which contribute to mechanical and biological function. We have previously shown that hMSCs undergoing chondrogenic differentiation develop a pericellular matrix (PCM), a biochemical bridge that mediates interactions between cells and their surrounding environment and contributes to cell stiffness. The PCM is composed primarily of type VI collagen (ColVI) and various proteoglycans, including decorin (Dcn). Dcn has been found to regulate collagen fibrillogenesis as well as play a role in linking ColVI in the PCM with proteins in the extracellular matrix. To elucidate the function of PCM components in differentiating hMSCs, ColVI and Dcn were silenced through RNA interference. The effect of lentiviral transduction on hMSC chondrogenesis and extent of gene knockdown were determined through gene expression. Our study provides a first step in the successful manipulation of specific mechanotransduction pathways in differentiating hMSCs which can advance the development of engineered load-bearing tissue replacements. METHODS: hMSC Culture: Passage 2 hMSCs (Lonza, MD) were expanded in monolayer according to manufacturer’s protocols, using Mesenchymal Stem Cell Growth Medium (MSCGM) (Lonza, MD). Lentiviral Transduction of hMSCs: A lentiviral expression vector containing the gene for GFP was prepared using the Vivid Colors pLenti6.2-GW/EmGFP kit (Invitrogen, CA). Passage 3 hMSCs were infected at an MOI of 1. On day 4, the population was evaluated for transduction efficiency through fluorescence-activated cell sorting (FACS), using propidium iodide (PI) to identify non-viable cells. A BD FACSAriaII cell sorter was used to separate pure, viable populations of transduced [GFP(+)] and non-transduced [GFP(-)] cells. Gating parameters were set with 3 control hMSC groups: non-infected; PIstained, apoptotic; and GFP-infected cells. Chondrogenesis Assay for Transduced Cells: Separate populations of GFP(+) and GFP(-) hMSCs were resuspended in 2% alginate and expelled dropwise into 100 mM CaCl2 for bead formation. The beads were cultured in chondrogenic medium containing 2 μg/mL TGF-β3 for 14 days. Cells were subsequently released with 100 mM sodium citrate and, along with an undifferentiated hMSC control, analyzed for gene expression levels of GAPDH and 3 markers for chondrogenic differentiation: type II collagen, sox9, and aggrecan. RNA was isolated using the RNeasy Micro kit (Qiagen, CA) and reverse transcribed to obtain gene expression levels using a MyiQ Real-Time PCR System (BioRad, CA). Results were analyzed using the ∆∆Ct method and expressed as the fold difference using the exponential relation 2 . shRNA Construct Preparation: Four 21 nt long complementary oligos for col6a1 and dcn were cloned into lentiviral vectors using the BLOCK-iT U6 Entry Vector Kit and Lentiviral RNAi Expression System (Invitrogen, CA). Knockdown Assay: Each viral construct (designated colVI and dcn AD) was used to infect 30% confluent, P4 hMSCs. Cells were lysed at 1, 5, and 8 days. Each time point also included a corresponding noninfected control and a nonsense control containing a virus with nonspecific shRNA. Gene expression analysis for GAPDH and col6a1 or dcn was performed as described above. RESULTS: Lentiviral infection exhibited 20% transduction efficiency in hMSCs at an MOI of 1 (Fig 1). The GFP(+) and GFP(-) populations were not distinguishable through their forward and side scatter profiles. The overall population was 97% viable. Although there were slight differences GFP(+) and GFP(-) cells both demonstrated increased expression for chondrogenesis markers relative to hMSC controls (Fig. 2). Varying levels of knockdown of the target gene were achieved for col6a1 (Fig 3) and for dcn (Fig 4). ColVI D and Dcn B resulted in the greatest knockdown of gene expression. Progressive downregulation was observed in ColVI C and D and Dcn A, and relatively stable downregulation was maintained in Dcn D. The nonsense displayed upregulation initially for ColVI and downregulation for Dcn.