Application of an Indoleamine 2,3-Dioxygenase|[ndash]|Expressing Skin Substitute Improves Scar Formation in a Fibrotic Animal Model

TO THE EDITOR Any delay in wound closure increases the probability of developing dermal fibrotic conditions such as hypertrophic scars (Deitch et al., 1983). For patients with extensive burn injuries, one of the most promising approaches is the application of an engineered skin substitute containing both epidermal and dermal cells (Coulomb et al., 1998). Where allogeneic skin substitutes can provide a rapid, patient-ready wound coverage, they are susceptible to immune rejection. Our approach has been to engineer an allogeneic skin substitute that expresses an immunomodulating enzyme to protect the graft. Indoleamine 2,3-dioxygenase (IDO) is the rate-limiting enzyme of tryptophan catabolism that converts tryptophan to kynurenine (Taylor and Feng, 1991), and is well known to have a key role in producing immunoprivileged tissue environments. For example, it has been shown that IDO has a role in the prevention of the immune rejection of the semi-allogeneic fetus (Munn et al., 1998) and in the immune resistance of tumors (Uyttenhove et al., 2003). In our earlier work, we demonstrated that the expression of IDO can not only downregulate the expression of MHC class I in IDO-transduced keratinocytes (Li et al., 2004), but also that it can prevent immune rejection of grafted, simple xenogeneic fibroblast-populated scaffolds (Li et al., 2006). Conversely, depletion of tryptophan by IDO jeopardizes the survival of CD4+ lymphocytes and THP-1 monocytes (Ghahary et al., 2004). In this study, primary dermal fibroblasts were transduced with IDO gene using a lentiviral vector (Rezakhanloo et al., 2010). Kynurenine catabolite concentration was found to be 36.5 times higher in conditioned media from IDO-transduced cells compared with controls (14.6±0.103 vs. 0.4±0.007, n=4; P<0.001). Bilayered skin substitutes were engineered to have both an epidermal (primary keratinocytes) and dermal layer containing either IDO-transduced fibroblasts (ESSIDO) or non-transduced fibroblasts (ESS). Using the rabbit ear model (Morris et al., 1997), we created four 8-mm punch wounds on the ventral surface of each ear and either covered it with skin substitutes or left it untreated (nontreated (NT)). Wounds were evaluated over a period of 35 days. Untreated wounds were considered as the control for wound healing progression. Clinical and histological examination of the wounds showed an increase in scar elevation for NT and ESS wounds compared with those of ESSIDO wounds, which showed a flatter scar (Figure 1a). Wounds were analyzed for epidermal thickness and cellularity using hematoxylin and eosin staining. Normal rabbit skin was used as control (C) (Figure 1b top panel). Epidermal thickness was found to be greater in NT and ESS samples compared with that of control and ESSIDO (Figure 1c). In addition, both NT and ESS displayed increased cellularity compared with ESSIDO, which was more comparable to normal skin (Figure 1b lower panel). A decreased scar elevation index was evident in ESSIDO wounds compared with those of NT and ESS samples (1.4±0.04 vs. 2.3±0.25 and 1.72±0.26, respectively; n=6; P<0.05). Moreover, ESSIDO wounds showed a thinner epidermal layer when compared with those of ESS and NT wounds (1.99±0.16 vs. 9.01±1.4 and 6.63±0.58, respectively; n=6, P<0.001). A significant decrease in cellularity was also found in the ESSIDO wounds compared with either NT or ESS wounds (108.25±5.34 vs. 238.88±24.01 and 164.30±4.68, respectively; n=6, P<0.001; Figure 1c). To further elucidate the role that ESSIDO might have in mitigating fibrotic conditions, biopsies were examined for MMP-1 and pro-α1 collagen messenger RNA. At the gene level, we found that MMP-1 expression in the ESSIDO-treated wounds was significantly higher compared with control, NT, and ESS-treated wounds (1.21±0.06 vs. 0.38±0.16, 0.13±0.15, and 0.73±0.03, respectively; n=3, P<0.001; Figure 2a). In contrast, there was no significant difference in pro-α1 collagen gene expression. Therefore, these data suggest that local IDO expression in vivo is correlated with an increase in MMP-1, without changing the production of collagen (pro-α1). To ensure that grafted cells were present in the rabbit tissue for 35 days, biopsy sections were stained with a mAb for HLA-DR, revealing the presence of human keratinocytes from the ESSIDO in rabbit tissue but not in those from NT or ESS tissue samples (Figure 2b). In comparison, wound sections were also stained with a CD3 lymphocyte mAb demonstrating a random distribution of T cells throughout the dermis of both NT and ESS. Interestingly, ESSIDO-treated tissue sections were infiltrated with fewer T cells, present only in the deep dermis and as if to form a linear distribution (Figure 2c). These results suggest that the expression of IDO by the ESSIDO creates a barrier-impeding migration of T cells into the transplanted graft. Using a well-documented rabbit ear fibrotic model, we showed that the application of a skin substitute containing IDO-transduced cells improves scar elevation, epidermal thickness, and cellularity, which are the known characteristics of dermal fibroproliferative disorders. In addition, we demonstrated that IDO-expressing cells significantly stimulated MMP-1 expression in bystander fibroblasts (Supplementary Figure S1 online). This finding suggests that IDO likely induces MMP-1 expression in the host fibroblasts. Although it remains to be further elucidated, the mechanism for this response could be partly due to depletion of tryptophan or an increase in IDO metabolites. Again, we are able to demonstrate that IDO is able to protect a multicellular xenograft from rejection, similar to our previous study (Jalili et al., 2010). CD3 staining in the ESSIDO samples revealed the presence of these cells only in the deep dermis, corresponding with our previous in vitro findings (Forouzandeh et al., 2008) and suggesting that the IDO may create a niche that restricts T-cell infiltration. In conclusion, this study describes a possible alternative role for IDO in scar improvement through MMP-1 stimulation in host cells. Moreover, it demonstrates the survival of an ESSIDO graft for up to 35 days, (Supplementary Figures S2, S3 online) suggesting the possibility of serving as a permanent wound coverage. Further studies are suggested to address the mechanism through which these phenomena occur. STB is the named inventor on patents and patent applications assigned to the University of Cincinnati and Shriners Hospitals for Children according to their intellectual property policies. Patents, patent applications, and other intellectual property pertaining to engineered skin substitutes are licensed to Cutanogen Corporation, which was founded by STB, and in which he has past and present financial interests. STB resigned as an officer of Cutanogen's current activities. STB also serves currently as a paid consultant to Aderans Research. This study was supported by the Canadian Institute of Health Research (CIHR). CC-M holds a CIHR-doctoral graduate award and Michael Smith Foundation for Health Research (MSFHR) junior award. The authors are grateful to Dr A Ghaffari, Dr A Hurtado-Coll, and Dr L Fazili for their unconditional support and patience. This research was funded by the Canadian Institute for Health Research (CIHR) CIHR-MOP-74483 and SRTC-CIHR. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper