Islet transplantation represents a viable treatment for type 1 diabetes. However, due to loss of substantial mass of islets early after transplantation, islets from two or more donors are required to achieve insulin independence. Islet-extracellular matrix disengagement, which occurs during islet isolation process, leads to subsequent islet cell apoptosis and is an important contributing factor to early islet loss. In this study, we developed a fibroblast populated collagen matrix (FPCM) as a novel scaffold to improve islet cell viability and function post-transplantation. FPCM was developed by embedding fibroblasts within type-I collagen and used as scaffold for islet grafts. Viability and insulin secretory function of islets embedded within FPCM was evaluated in vitro and in a syngeneic murine islet transplantation model. Islets embedded within acellular matrix or naked islets were used as control. Islet cell survival and function was markedly improved particularly after embedding within FPCM. The composite scaffold significantly promoted islet isograft survival and reduced the critical islet mass required for diabetes reversal by half (from 200 to 100 islets per recipient). Fibroblast embedded within FPCM produced fibronectin and growth factors and induced islet cell proliferation. No evidence of fibroblast over-growth within composite grafts was noticed. These results confirm that FPCM significantly promotes islet viability and functionality, enhances engraftment of islet grafts and decreases the critical islet mass needed to reverse hyperglycemia. This promising finding offers a new approach to reducing the number of islet donors per recipient and improving islet transplant outcome.

OBJECTIVEThe requirement of systemic immunosuppression after islet transplantation is of significant concern and a major drawback to clinical islet transplantation. Here, we introduce a novel composite three-dimensional islet graft equipped with a local immunosuppressive system that prevents islet allograft rejection without systemic antirejection agents. In this composite graft, expression of indoleamine 2,3 dioxygenase (IDO), a tryptophan-degrading enzyme, in syngeneic fibroblasts provides a low-tryptophan microenvironment within which T-cells cannot proliferate and infiltrate islets.RESEARCH DESIGN AND METHODSComposite three-dimensional islet grafts were engineered by embedding allogeneic mouse islets and adenoviral-transduced IDO–expressing syngeneic fibroblasts within collagen gel matrix. These grafts were then transplanted into renal subcapsular space of streptozotocin diabetic immunocompetent mice. The viability, function, and criteria for graft take were then determined in the graft recipient mice.RESULTSIDO-expressing grafts survived significantly longer than controls (41.2 ± 1.64 vs. 12.9 ± 0.73 days; P < 0.001) without administration of systemic immunesuppressive agents. Local expression of IDO suppressed effector T-cells at the graft site, induced a Th2 immune response shift, generated an anti-inflammatory cytokine profile, delayed alloantibody production, and increased number of regulatory T-cells in draining lymph nodes, which resulted in antigen-specific impairment of T-cell priming.CONCLUSIONSLocal IDO expression prevents cellular and humoral alloimmune responses against islets and significantly prolongs islet allograft survival without systemic antirejection treatments. This promising finding proves the potent local immunosuppressive activity of IDO in islet allografts and sets the stage for development of a long-lasting nonrejectable islet allograft using stable IDO induction in bystander fibroblasts.

Matrix metalloproteinases (MMPs) are key elements in extracellular matrix (ECM) degradation and scar remodeling during the wound-healing process. Our previous data revealed that keratinocyte-releasable factors significantly increased the expression of fibroblast MMPs in monolayer-cultured fibroblasts. In this study, we analyzed the differences in the MMP expressions of fibroblasts in a three-dimensional fibroblast-populated collagen gel (3D FPCG) from that in a two-dimensional monolayer-cultured fibroblasts when both co-cultured with keratinocytes. Differential mRNA and protein expression of fibroblasts were examined by microarray, RT-PCR, and western blot. Our results showed that fibroblasts co-cultured with keratinocytes in a 3D FPCG expressed significantly higher MMP1 and MMP3 at the gene and protein levels. Due to the physiological advantages of a 3D FPCG model to a 2D system, we concluded that the 3D FPCG model may provide a better means of understanding the fibroblast-keratinocyte cross-talk during the wound-healing process.

Interaction between keratinocytes and fibroblasts plays a critical role in maintaining skin integrity under both normal and pathological conditions. We have previously demonstrated that keratinocyte-releasable factors influence the expression of key extracellular matrix components, such as collagen and matrix metalloproteinases in dermal fibroblasts. In this study, we utilized DNA microarray analysis to examine the effects of keratinocyte-releasable factors on the expression of several cytokines in human dermal fibroblasts. The results revealed significantly higher granulocyte colony-stimulating factor (G-CSF) expression in fibroblasts co-cultured with keratinocytes relative to mono-cultured cells, which was verified by RT-PCR and western blot. G-CSF is an important hematopoietic factor also thought to play a beneficial role in wound healing through stimulating keratinocyte proliferation. To partially characterize the keratinocyte-releasable factors responsible for stimulating G-CSF production, keratinocyte-conditioned medium (KCM) was subjected to thermal treatment and ammonium sulfate precipitation before treating fibroblasts. The results showed that keratinocyte-releasable G-CSF-stimulating factors remain stable at 56°C and upon 50% ammonium sulfate precipitation. Knowing that keratinocytes release IL-1, which stimulates G-CSF expression in various immune cells, several experiments were conducted to ask whether this might also be the case for fibroblasts. The results showed that the addition of recombinant IL-1 markedly increased G-CSF expression in fibroblasts; however, IL-1 receptor antagonist only partially abrogated KCM-stimulated G-CSF expression, indicating the role of additional keratinocyte-releasable factors. These findings underline the importance of cross-talk between keratinocytes and fibroblasts, suggesting that communication between these cells in vivo modulates the production of cytokines required for cutaneous wound healing and maintenance. J. Cell. Biochem. 118: 308-317, 2017. © 2016 Wiley Periodicals, Inc.

Indoleamine 2,3-dioxygenase (IDO) can locally suppress T cell-mediated immune responses. It has been shown that defective self-tolerance in early prediabetic female non-obese diabetic (NOD) mice can be attributed to the impaired interferon-gamma (IFN-γ)- induced IDO expression in dendritic cells of these animals. As IFN-γ can induce IDO in both dendritic cells and fibroblasts, we asked the question of whether there exists a similar defect in IFN-γ-induced IDO expression in NOD mice dermal fibroblasts. To this end, we examined the effect of IFN-γ on expression of IDO and its enzymatic activity in NOD dermal fibroblasts. The results showed that fibroblasts from either prediabetic (8 wks of age) female or male, and diabetic female or male (12 and 24 wks of age respectively) NOD mice failed to express IDO in response to IFN-γ treatment. To find underlying mechanisms, we scrutinized the IFN- γ signaling pathway and investigated expression of other IFN-γ-modulated factors including major histocompatibility complex class I (MHC-I) and type I collagen (COL-I). The findings revealed a defect of signal transducer and activator of transcription 1 (STAT1) phosphorylation in NOD cells relative to that of controls. Furthermore, we found an increase in MHC-I and suppression of COL-I expression in fibroblasts from both NOD and control mice following IFN-γ treatment; indicating that the impaired response to IFN-γ in NOD fibroblasts is specific to IDO gene. Finally, we showed that an IFN-γ-independent IDO expression pathway i.e. lipopolysaccharide (LPS)-mediated-c-Jun kinase is operative in NOD mice fibroblast. In conclusion, the findings of this study for the first time indicate that IFN-γ fails to induce IDO expression in NOD dermal fibroblasts; this may partially be due to defective STAT1 phosphorylation in IFN-γ-induced-IDO signaling pathway.