Mesenchymal stem cells (MSC) are good candidates for cell therapies due to their immunomodulatory properties, ability to home to/engraft damaged tissues, and potential to differentiate into different cell types. However, when transplanted (Tx) in an allogeneic setting, MSC can elicit an immune response, activating the recipient's cytotoxic T lymphocytes (CTL) and Natural Killer (NK) cells, resulting in rejection of the Tx cells and reduced therapeutic efficacy. Human cytomegalovirus (HCMV, has developed several strategies to evade CTL and NK cell recognition. HCMV avoids CTL attack by producing proteins that downregulate MHC-I surface expression. These proteins are coded for by the unique short regions (US) 2, 3, 6 and 11 of HCMV's genome. We have previously shown that when MSC are transduced with retroviral vectors encoding each one of these US proteins, US6 and US11 were the most effective in reducing MSC's HLA-I surface expression and allogeneic CTL recognition and proliferation. However, HLA-I downregulation may render MSC transduced with US6 (MSC-US6) and US11 (MSC-US11) more susceptible to NK killing, undermining MSC's inherent ability to inhibit function of allogeneic NK cells. Here, we first investigated the role of US6 or US11 on MSC allorecognition by NK cells, and on MSC in vivo engraftment capability. NK killing assays demonstrated that US11 generated the most protective effect at the highest NK concentration (E:T ratio 20:1) (% specific lysis for MSC-US6: 60.4 ± 5.7 %; MSC-US11: 45.5 ± 2.4 % vs. MSC: 88.5 ± 3.4 % respectively). However, at an E:T ratio of 10:1 and 5:1 US11 produced the same degree of protection as US6 (E:T ratio of 10:1; % specific lysis for MSC-US6: 30.1 ± 5.6 %; MSC-US11: 26.3 ± 1.9 % vs. MSC: 54.7 ± 1.9 %); (E:T ratio 5:1; % specific lysis for MSC-US6: 11.9 ± 4.2; MSC-US11: 13.4 ± 2.3; vs. MSC: 25.5 ± 4 respectively). Only at an E:T ratio of 1:1 were US6 and US11 similar to untransduced MSCs (% specific lysis for MSC-US6: 4.7 ± 1.6; MSC-US11: 2.1 ± 0.5; vs. MSC: 4.9 ± 1.8; respectively) in terms of inhibition of NK killing. We also studied the role of US6 and 11 on the expression of beta-2-microglobulin (b2m) and other HLA-I molecules, and we found that US6 reduced b2m by 87± 2 % and HLA-G1 by 44±4.7 %, while US11 reduced b2m by 70± 0.6 % but increased HLA-G1 expression by 176.6±1.9 %. Therefore, the increase in HLA-G1 expression induced by US11 may explain the decrease in NK killing observed in the MSC-US11 cells. Furthermore, we investigated whether US6 or US11 could play a role in mediating complement resistance. While US6 increased the expression of CD59 in transduced cells (Mean fluorescence intensity (MFI) increased by 123.3±1), US11 increased the number of cells expressing CD59 by 121.4 ± 0.8 %, but did not modify their MFI. We next compared the in vivo engraftment potential of MSC, MSC-US6 and MSC-US11 by Tx 5.6×10^4 of each cell population into fetal sheep at 60 days of gestation (n=6). Since we have previously reported the ability of MSC to generate liver cells, we first investigated whether the expression of US6 and 11 would allow higher levels of liver engraftment and hepatocyte formation when compared to MSC (MSC-E) transduced with a retroviral vector encoding only NPT-II. Two months after Tx, liver tissues were collected and stained with NPT-II antibody. This revealed that US6 and US11 increased engraftment efficiency by 241% for MSC-US6 and 277% for MSC-US11 (MSC-E: 5.3 ± 0.4 %, MSC-US6: 12.8 ± 0.9 % and MSC-US:11 14.7 ± 0.8 %). Despite the higher level of liver engraftment seen with MSC-US6 and MSC-US11, co-expression of NPT-II and albumin (MSC-US6: 57% MSC-US1: 50% MSC-E: 75%) or NPT-II and Ov-6 was found at significantly lower levels in MSC-US11 and MSC-US6 Tx animals than in those Tx with MSC-E. Nevertheless, similar numbers of NPT-II/CD34 double-positive cells were found in the liver of MSC-US6 and MSC-US11 Tx animals when compared to MSC-E alone. In conclusion, engineering MSC to over-express US6 or US11 is an effective way to reduce CTL proliferation, NK killing and destruction of engrafted cells by the complement membrane attack complex. In agreement with the in vitro studies, transplantation of these cells into a large animal sheep model resulted in significantly higher levels of overall cell engraftment, but not differentiation towards a hepatocytic phenotype. Studies are underway to determine the mechanism by which HCMV proteins are interfering with MSC differentiation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.