BackgroundThe efficacy and biosafety of lentiviral gene transfer is influenced by the design of the vector. To this end, properties of lentiviral vectors can be modified by using cis-acting elements such as the modification of the U3 region of the LTR, the incorporation of the central flap (cPPT-CTS) element, or post-transcriptional regulatory elements such as the woodchuck post-transcriptional regulatory element (WPRE). Recently, several studies evaluated the influence of the incorporation of insulators into the integrating lentiviral vector genome on transgene expression level and position effects.MethodsIn the present study, the influence of the matrix attachment region (MAR) of the mouse immunoglobulin-κ (Ig-κ) or the chicken lysozyme (ChL) gene was studied on three types of HIV-1-derived lentiviral vectors: self-inactivating (SIN) lentiviral vectors (LV), double-copy lentiviral vectors (DC) and non-integrating lentiviral vectors (NILVs) in different cell types: HeLa, HEK293T, NIH-3T3, Raji, and T Jurkat cell lines and primary neural progenitors.Results and DiscussionOur results demonstrate that the Ig-κ MAR in the context of LV slightly increases transduction efficiency only in Hela, NIH-3T3 and Jurkat cells. In the context of double-copy lentiviral vectors, the Ig-κ MAR has no effect or even negatively influences transduction efficiency. In the same way, in the context of non-integrating lentiviral vectors, the Ig-κ MAR has no effect or even negatively influences transduction efficiency, except in differentiated primary neural progenitor cells.The ChL MAR in the context of integrating and non-integrating lentiviral vectors shows no effect or a decrease of transgene expression in all tested conditions.ConclusionsThis study demonstrates that MAR sequences not necessarily increase transgene expression and that the effect of these sequences is probably context dependent and/or vector dependent. Thus, this study highlights the importance to consider a MAR sequence in a given context. Moreover, other recent reports pointed out the potential effects of random integration of insulators on the expression level of endogenous genes. Taken together, these results show that the use of an insulator in a vector for gene therapy must be well assessed in the particular therapeutic context that it will be used for, and must be balanced with its potential genotoxic effects.