Best Practices in Mouse Models of Skin Transplantation

Abstract

The murine skin transplant model has been used for decades as a sensitive in vivo model in which to dissect mechanistic aspects of alloreactivity. Indeed, seminal work defining the nature of alloreactivity was accomplished in the murine skin graft model. In the modern era, the incorporation of genetically altered animals, pharmacologic manipulation, and multi-omics approaches in the murine skin graft model has significantly increased the mechanistic insight that can be gleaned from these studies. Because skin grafts are not primarily vascularized, the immune response against them is dominated by a CD8⁺ T cell response, with a minor but important contribution of alloreactive CD4⁺ T cells. In contrast, there is very little role for donor-reactive antibody in mediating skin transplant rejection. Importantly, the results of pharmacologic inhibition of T cell-directed novel targets for immunosuppression in transplantation in the murine skin graft model have largely reflected the results subsequently obtained in non-human primate renal transplant studies, demonstrating the physiologic relevance of the model. In this minireview, we further discuss the relative benefits and limitations of murine skin transplantation as a model of alloreactivity.

1. Introduction

Skin transplantation in the mouse has been used for decades as a sensitive and convenient in vivo measure of alloimmune responses. While interest in skin transplantation began in response to wartime injuries and was first studied by Medawar in the 1940s (1), during the 1950s it quickly gained traction as an expeditious approach to interrogating the effect of MHC genetics on tissue rejection or acceptance (2–4). In fact, skin transplantation was used in seminal experiments to demonstrate the antigen-specific nature of allograft rejection: Mintz and Silvers grafted skin from type A:B genetic mosaic donors onto type A recipients and showed that only type A regions of the graft survived (5, 6). These key findings illuminated the activity of lymphocytes that kill cells expressing foreign antigens but spare those that do not, even when in close proximity. In the decades that ensued, mouse skin transplantation models allow researchers to investigate the roles of dendritic cells/ antigen processing and presentation, innate immunity, T cell-mediated rejection, and the contribution of regulatory immune cells in graft survival (7–14). Across many studies, the model has also proven instrumental in evaluating novel immunosuppressive therapies, such as costimulatory blockade and tolerance-inducing protocols, which aim to prolong graft survival while minimizing systemic immunosuppression. Genetic manipulation in mice enables precise mechanistic exploration of molecular pathways involved in graft rejection and tolerance, helping to identify potential targets for clinical application. Finally, skin grafting is a cost-effective and technically feasible model, facilitating high-throughput studies on immune responses following transplantation.

2. Description of the model: kinetics, impact of genetic disparity, and role of immunosuppression

2.1. Skin grafting procedure and elicitation of alloimmunity

Skin is historically classified as the most immunogenic tissue (15) and the hardest in which to induce immune regulation, eliciting a potent cellular adaptive immune response within 3-7 days following transplantation (16). Factors contributing to the high immunogenicity of skin include its extensive network of tissue-resident T cells (17) and a dense, dynamic population of antigen-presenting cells (APCs) (18, 19), which possess an enhanced capacity to induce T cell responses (20). Standard practice is to transplant full-thickness donor ear, tail, or trunk skin onto a shaved, sterilized and well-vascularized graft bed by making an incision on the dorsal thorax of the recipient (21, 22). Most commonly, two skin grafts are placed on each recipient. Tail skin is commonly used due to its ease of preparation and resistance to ischemic necrosis (23, 24). It is also easy to monitor and score tail skin grafts due to the characteristic hair texture of the skin. Tail skin also differs from ear and trunk skin in that it contains a reduced density of skin-resident myeloid cells (19). Due to this greater resistance to minor ischemic and inflammatory insults and lower density of innate cells, tail skin may have slightly delayed rejection kinetics and may be less sensitive at detecting minor histocompatibility differences relative to ear or trunk skin (23, 24). Common pitfalls of the skin graft procedure include excessive bleeding, which may cause hematoma formation under the graft and prevent proper revascularization, or damage to the underlying muscle fascia, which can impede healing. Grafts are then typically bandaged and allowed to heal for 6 days before removal of the bandage (25). Proper bandaging technique is critical to the success of the procedure in that it prevents the mouse or cage mates from disturbing the graft and provides optimal pressure for revascularization; this procedure can be visualized in Cheng et al. (25). Alternatives to bandaging include the use of sutures (25) (which lengthen the time of the procedure, can cause additional tissue trauma at the stitch sites, and requires re-anesthetizing the mouse for the removal of non-adsorbable sutures) and surgical adhesives (which can be difficult to manage with thin graft tissue and have the potential for seepage into the wound, interfering with healing) (26). The resultant cellular adaptive immune response to skin transplantation is highly localized (27). Donor-reactive CD4⁺ and CD8⁺ T cells are primed in the graft draining lymph nodes (axillary, brachial and inguinal for this model) via direct, indirect, or semi-direct allorecognition, can be detected there by day 3 post-transplant, and continue to divide there until about day 7-10, where they become evident in the spleen and circulation (16, 28). Donor-reactive antibodies become detectable in the blood at approximately day 10-14 post-transplant and continue to increase until they plateau around day 25-30 post-transplant during unmodified rejection. By day 7-10 post-transplant, donor-reactive T cells infiltrate into the skin graft parenchyma, where they re-encounter donor antigen and mediate cytokine and cytolytic function. While both CD4⁺ and CD8⁺ T cells contribute to skin graft rejection, CD8⁺ T cells are the dominant population of graft-infiltrating T cells, comprising about 70-90% of graft-infiltrating T cells (29). Cytokine secretion results in local inflammation that functions to further recruit cells of the innate immune system, while CD8⁺ T cell cytolytic function can kill parenchymal cells directly. Ultimately, occlusion of small blood vessels results in ischemia and fibrosis and the graft scabs over.

2.2. Effects of MHC disparity on the strength of the alloimmune response to skin transplantation

The magnitude of the alloimmune response and kinetics of allograft rejection are highly dependent on both recipient genetics and the degree of donor:recipient disparity. Certainly, the degree of MHC disparity is the largest contributing factor to the strength of the immune response to a skin allograft. For example, C57BL/6 mice (H-2^b) reject BALB/c (H-2^d) skin grafts with rapid kinetics (MST 12d) in the absence of immunosuppression (27). In contrast, if the BALB/c skin instead expresses H-2^b (BALB.B strain), the MST is extended to C57BL/6 costimulation blockade-treated recipients of BALB/c grafts experienced prolonged graft survival with an MST of 26 days, but C57BL/6 recipients of BALB.B grafts enjoyed long-term graft survival of > 70 days (Table I)(27). This is a result of the low vs high pre-transplant precursor frequency of donor-reactive CD4⁺ and CD8⁺ cells in the two strain combinations, as demonstrated by CFSE dilution in an in vitro MLR (27). Partially MHC-matched skin graft combinations, such as bm12→B6, also exhibit relatively rapid rejection (MST 10- 13d) (13), while surrogate minor antigens, such as the OVA disparity, have longer rejection kinetics (MST 19d) in the absence of immunosuppression (29), again owing to the low pre-transplant precursor frequency of donor-reactive T cells. Of note, pre-transplant donor-reactive T cell precursor frequency has been demonstrated to be a critical factor in determining the susceptibility of T cell responses to immunosuppression, with low pre-transplant donor-reactive T cell precursor frequencies being more susceptible to inhibition by immunosuppression than high precursory frequency responses (16). An example of this is the low precursor frequency present in recipients of B6.OVA skin grafts are highly susceptible to CTLA-4Ig + anti-CD154 costimulation blockade (Table I)(16). In this way, degree of MHC mismatch determines not only the strength of alloreactivity but also contributes to resistance to immunosuppression.

Table I.

Median Survival Times (MST) of Murine Skin Transplant Models

Donor	Recipient	Immunosuppression	Median Survival Time	Reference
BALB/c	C57BL/6	None	14d	(101)
BALB/c	C57BL/6	CTLA-4Ig+anti-CD40L (clone MR-1)	26d	(101)
BALB/.B	C57BL/6	None	16d	(27)
BALB/.B	C57BL/6	CTLA-4Ig+anti-CD40L (clone MR-1)	70d	(27)
Bm.12	C57BL/6	None	10-13d	(13)
B6.OVA	C57BL/6	CTLA-4Ig (500ug) +anti-CD40L (clone MR-1)	>100d (100% grafts viable at 100d)	(16)
B6.OVA	C57BL/6	CTLA-4Ig (250ug)	21d	(77)
C57BL/6 Male	C57BL/6 Female	None	~30-40d	(102)
BALB/c	C3H/HeJ	CTLA-4Ig+anti-CD40L (clone MR-1)	>100d (100% grafts viable at 100d)	(103)
BALB/c	C57BL/6	Tacrolimus	~70d	(104, 105)

2.3. Effect of non-MHC genetic disparity on the strength of the alloimmune response following skin transplantation

Genetic factors outside of the MHC also play a role in determining the strength of alloimmune response against a skin graft. For example, C57BL/6 mice execute a more robust alloimmune response as compared to C3H or DBA mice (30). Combined blockade of the CD154 and CD28 costimulatory pathways effectively promotes skin allograft survival in C3H/HeJ mice, extending median survival times (MSTs) even greater than 100 days (30). However, this treatment was shown to be significantly less effective in C57BL/6 mice, with reported MSTs ranging between 20 and 30 days. The authors of these studies went on to demonstrate that the underlying genetic cause of this disparity was independent of both the MHC locus and the pre-transplant donor-reactive T cell precursor frequency (30). Instead, they found that genetic differences in cells of hemopoietic origin dictated the rejection kinetics and that the costimulation blockade-resistant phenotype was dominant. Further genetic analysis revealed that 17 loci on four chromosomes contributed to the resistance phenotype (30). Additional work using modern tools to elucidate the genes underlying the observed resistance phenotype is warranted.

3. Defining skin transplant rejection

Murine skin graft models commonly employ a 4-point scoring rubric which defines rejection by gross changes upon physical examination of the graft. Many studies have defined rejection as the day on which <10% of the original allograft tissue is viable (13, 29, 31). Although this scoring system is relatively subjective, it is likely to yield reasonably consistent results when utilized by the same group of researchers scoring all mice within a given experiment. A key advantage of the model is that grafts can be monitored longitudinally, and mice do not need to be euthanized for assessment of rejection, allowing for secondary rechallenge experiments. While this approach has been a mainstay of many laboratories due to its ease and minimal invasiveness, recent work has pioneered the use of more objective approaches to quantify skin graft rejection. Skin allografts can be processed for histology and scored for acute allograft rejection according to the Banff 2007 working classification of skin-containing composite tissue allograft pathology (32). A potential downside of this approach is the requirement for euthanizing the experimental animals in order to obtain histological examination of allografts (33). In addition, recent studies have implemented the use of digital image analysis to better quantify, magnify, and record changes in graft color and/or texture over time (34) or the use of thermal imaging to measure blood flow as an indicator of viability (35). Overall, the use of these more quantitative and potentially more objective approaches to score skin transplant rejection in the mouse may increase standardization across practices in laboratories in the field.

Mechanistic aspects of skin transplant rejection are also frequently assessed by reporting the number of graft-infiltrating cells. As mentioned above, CD8⁺ T cells predominate as the primary graft-infiltrating T cell subset. Thus, metrics of rejection often include the measurement of absolute numbers of CD8⁺ T cells, in addition to other graft-infiltrating T cell subsets such as CD4⁺ T cells, CD11b⁺ monocytes, CD11c⁺ dendritic cells, and F4/80⁺ macrophages. Careful phenotypic and transcriptomic analysis of these cell subsets can yield additional information about their activation and differentiation status to further characterize the rejection.

4. Uses and Benefits of the Skin Allotransplantation Model

4.1. Technical and logistical considerations of the murine skin transplant model

The skin allotransplantation model has several benefits that render it useful for the investigation of the cellular and molecular basis of allograft rejection. Since it does not require vascular anastomoses, the technique is relatively easy to learn and therefore accessible to incoming laboratory personnel (Table II). The relatively short time needed to complete each procedure also allows a single researcher to perform many transplants in one day, establishing it as a reliable and widely used model for cellular immunology studies. Being able to complete 20-40 skin grafts in a single day allows investigators to generate multiple groups of mice to be sacrificed at multiple timepoints post-transplant, allowing for well-controlled longitudinal analyses of multiple treatment groups in a single experiment. This property of the model certainly facilitates mechanistic investigation of mechanisms of cellular alloimmunity.

Table II.

Key Features of Murine Skin Transplantation Model

	Skin Transplant	Vascularized Transplant
Source of Endothelium	Recipient-derived	Donor-derived
Requirement for Secondary Lymphoid Organs	Required	Required
Primary Rejection Mechanism	Cell-mediated	Cell-mediated and antibody-mediated
Cost	Lower	Higher
Technical Difficulty	Lower	Higher
Suitability for Tolerance Studies	More stringent	Less stringent

4.2. Investigation into mechanisms of allorecognition, acute cellular rejection, and T cell trafficking

Skin transplantation is a useful model for studying the mechanisms of wound healing responses to isogenic tissue (36), as well as acute cellular responses to allogeneic tissue (24). First, skin transplantation has afforded insights into the innate immune mechanisms involved in the initiation of an adaptive immune response, including the recognition of danger-associated molecular patterns (DAMPs), and revealed that following implantation, neutrophils are recruited to the graft site and contribute to the developing inflammatory response by releasing mediators, such as reactive oxygen species (ROS), neutrophil elastase (NE) and profibrotic factors, which serve to amplify the response by attracting other immune cells to infiltrate the graft (37). It has also demonstrated that phagocytic macrophages engulf and digest donor-derived cells, cellular debris and extracellular vesicles, functioning as antigen-presenting cells to activate the adaptive immune response (38). Skin transplantation has been used as a model to interrogate how alloreactive T cells are activated by different pathways of allorecognition. For example, researchers have recently used skin transplant models to describe that the recipients’ APCs cross-decorated with intact donor MHC: peptide molecules are a major mechanism in activation of alloreactive T cells (39, 40). Recipient-derived Batf3-dependent DCs are essential for the rejection of minor histocompatibility antigen-mismatched skin grafts, particularly through the semi-direct pathway (41, 42). Borges et al. extended these insights by showing that donor-derived CD103⁺ Batf3⁺ DCs are major contributors in transporting donor MHC class II to host skin-draining lymph nodes (dLNs), thereby initiating robust direct allorecognition (43). Indeed, seminal studies showed that rejection of both non-primarily vascularized skin transplants as well as primarily vascularized transplants requires the presence of secondary lymphoid organs (Table II) (44, 45). In addition to these myeloid-derived innate immune cells, skin-resident dendritic epidermal T cells (DETCs) are a unique population of gamma-delta T cells that reside in the epidermis, acting as sentinel cells to monitor and rapidly respond to skin injury and infection (46). These cells possess a dendritic shape that facilitates communication with neighboring epidermal cells and secrete cytokines such as IL-17A and insulin-like growth factor-1 (IGF-1), which are critical for maintaining skin homeostasis, promoting wound healing by stimulating keratinocyte proliferation, and orchestrating immune responses.

As noted earlier, the graft-infiltrating T cell populations are comprised mostly of CD8⁺ T cells. This is likely because skin is not a primarily vascularized tissue, meaning there is no donor endothelium that is transplanted into the recipient. This lack of donor endothelium is relevant in that there is far less exposure of donor MHC class II molecules to the recipient immune system early following transplantation. Thus, the involvement of CD4⁺ T cells is less in skin transplantation than in a primarily vascularized organ, such as a heart or kidney. Nonetheless, CD4⁺ T cells do play a role in skin transplant rejection (47). Seminal studies by the Jenkins laboratory using adoptively transferred TCR transgenic donor-reactive CD4⁺ (OT-II) and CD8⁺ T cell (OT-I) populations have shown that either population alone is capable of mediating rejection of skin allografts (29, 48). The kinetics of allograft rejection are faster in recipients bearing solely donor-reactive CD8⁺ T cells compared to those bearing solely donor-reactive CD4⁺ T cells (29). In contrast, rejection kinetics were significantly accelerated when recipients contained both donor-reactive CD4⁺ and CD8⁺ T cells, signifying an important role for CD4⁺ T cell help for the generation of alloreactive CD8⁺ T cell populations (29). Indeed, previous work has shown that variations in alloreactive CD4⁺ T cell precursor frequency significantly impact the provision of help for the development of donor-reactive B cell and CD8⁺ T cell responses (49). CD4⁺ T cells can also influence allograft rejection through effector mechanisms. Both Th1 and Th17 cells play distinct but significant roles in skin transplant rejection by inducing inflammation and tissue damage (50, 51). Th1 cells primarily promote acute rejection through IFN-γ secretion, enhancing leukocyte infiltration and macrophage activation (51). In contrast, Th17 cells induce a different form of rejection characterized by epidermal hyperplasia and neutrophil infiltration, with the mechanism possibly involving IL-17A and other effector molecules (50), although the specific mechanisms of Th17-mediated damage are still being explored.

While CD4⁺ T cells clearly influence skin transplant rejection, the role of antibodies in mediating graft damage is far less significant in models of skin transplantation than in primarily vascularized transplant models (1, 52–54). Because a skin allograft will derive its vascular supply via the ingrowth of recipient blood vessels as well as the spontaneous anastomosis of graft and recipient capillaries, most of the endothelium of these tissues will be recipient-derived and thus not bound by anti-donor MHC-specific antibodies (55). Furthermore, because antibodies are largely confined to vascular spaces, alloreactive antibodies have minimal direct impact on parenchymal cells. As such, anti-donor antibodies are thought not to play a measurable role in skin graft rejection (55) (Table II).

Skin transplant models have also been used to define the mechanisms and therapeutic potential of Foxp3⁺ CD4⁺ regulatory T cells (Treg) in transplantation. Through this work, Treg have been shown to play an important role in the induction and maintenance of immunological tolerance to alloantigens (56–59). Moreover, murine and humanized mouse transplant models have demonstrated that Treg, both polyclonal and to a greater extent donor-specific, can control acute and delayed allograft rejection (57, 60). These data laid the experimental foundation for the translation of Treg as a cellular therapy for the treatment or prevention of clinical transplant rejection.

The skin transplant model has been used to interrogate the specificity and mechanisms underlying alloreactivity in vivo. For example, recent studies used the skin allograft model to validate and characterize polyclonal graft-specific CD4⁺ and CD8⁺ T cell populations using peptide: MHC tetramer technology. Burrack et al. used NOD hosts grafted with C57Bl/6 H-2^b skin to define five distinct peptide:I-A^b direct pathway CD4⁺ T cells epitopes (61). Zhanzak et al. used a bioinformatics approach to identify and H-2K^d-derived peptides in the fully allogeneic BALB/c to C57BL/6 skin graft model (62). Using peptide: MHC tetramers for one dominant I-A^b restricted peptide, K^d287, the authors showed that CD4⁺ T cells specific for this epitope were sufficient to provide CD4⁺ T cell help to elicit anti-donor antibody responses. Moreover, in a model of liver-induced tolerance, Son et al. showed that the peptide repertoire expressed from H-2K^d MHC shapes the direct pathway-directed alloimmune response (63). The authors found that a pool of five MHC class I H-2K^b tetramers was sufficient to identify approximately 40% of the CD8⁺ T cells that were activated following skin transplantation, and that the deletion of these tetramer-reactive cells resulted in donor-specific tolerance (63). Finally, using MHC class I tetramers specific for an L^d-restricted peptide “QL9”, Cohen et al. showed that endogenous L^d: QL9⁺ CD8⁺ T cells were activated and expressed Ki67 and activation markers following BALB/c skin grafting (64). Taken together, these data demonstrate how skin allotransplantation has been used to rigorously dissect the molecular basis of alloreactivity.

Skin transplant models have also been used as tools to interrogate T cell trafficking to the site of allograft using In Vivo Imaging System (IVIS) technology, a non-invasive imaging system that monitors disease progression, cell trafficking, and gene expression in vivo. The use of T cells from mice that constitutively express luciferase allows these cells to be visualized in the skin at various intervals longitudinally post-transplant (65). Alternatively, immunotherapeutic reagents can be labeled with bioluminescent or bear-infrared dyes and imaged within the skin allograft at various timepoints post-transplant during rejection or tolerance, thus facilitating measurement of the drug within the allograft itself (Yao and Ford, manuscript in preparation). Thus, owing to their anatomic location at the external surface, IVIS imaging of skin transplantation is a unique opportunity to visualize some of the cellular events occurring during rejection and/or tolerance.

4.3. Skin transplantation to test novel targets for T cell-directed immunosuppression

The considerations listed above make skin transplantation well-suited for the testing of novel immunosuppression regimens in the mouse model. Seminal studies on the efficacy of both CD28 and CD154 costimulation blockade were made in the murine skin allograft model (66). Testing of additional T cell costimulatory targets, including OX40 (67–69), 4-1BB (69), and ICOS (69, 70), was also performed in murine skin transplant models, as well as testing of integrin targets such as LFA-1 (71, 72) and VLA-4 (73). Importantly, these targets were later tested in the non-human primate (NHP) model of renal transplantation. More recently, investigations into the impact of CD122 blockade were done in the murine skin graft model and also verified in the NHP model of renal transplantation. Overall, the murine skin graft studies correctly predicted the general level of efficacy of therapeutic blockade of these pathways as reflected in subsequent non-human primate studies (74, 75), and in some cases, human clinical trials (76). Murine models may also be useful for dose-finding studies prior to costly NHP trials; for example it is clear that lower doses of CTLA-4Ig (250ug) (16)are suboptimal compared to higher doses (500ug) (77) in the minor antigen mismatch murine skin graft model (77). Taken together, these findings demonstrate that the murine skin transplant model has sufficient physiologic relevance for testing immunosuppressive regimens directed against cellular alloreactivity.

4.4. Skin transplantation as a model for vascularized composite tissue allotransplantation

Skin is the primary target of immune rejection in vascularized composite allotransplantation (VCA), including face and limb transplants. Skin transplantation in mice serves as a widely used and highly informative model for VCA, providing critical insights into immune tolerance, rejection mechanisms, and therapeutic strategies (33). Unlike solid organ transplantation, VCA involves the transfer of multiple tissue types, including skin, muscle, blood vessels, and nerves, making skin grafts particularly relevant due to their high immunogenicity and complex immune response. For example, Ashraf et al. recently investigated the impact of donor-derived conventional DCs (cDCs) and APCs on the immunogenicity of skin and skin-containing VCA grafts, using mouse models of skin and hind limb transplantation (78). Their data reveal that the skin component exhibited heightened immunogenicity when compared to the entire VCA, evidenced by increased frequencies of DC subsets in the lymphoid tissues and the blood of skin transplant recipients (78). The involvement of human dendritic cells in VCA rejection was recently suggested in recipients of limb and face transplants (14). These findings highlight the value of the less resource-intensive skin transplant model as a surrogate platform for investigating novel immunosuppressive strategies aimed at preventing rejection in VCA. While the skin model offers practical advantages, it also presents limitations, such as differences in vascularization and the absence of functional integration of complex tissues like muscle and nerves. Nonetheless, mouse skin transplantation remains a powerful tool in VCA research, enabling detailed analysis of immune responses to the skin component of grafts and supporting the development of more effective therapeutic approaches for clinical application.

5. Humanized mouse models of skin transplantation

Humanized mouse models of skin transplantation have the potential to be a powerful tool with which to study the immunological mechanisms underlying transplant rejection in a system that recapitulates aspects of human physiology and the human immune response. These models are typically generated by engrafting immunodeficient mice - such as NSG (NOD-scid IL2Rγ^null) mice, which have no T, B, or NK cells - with human hematopoietic stem cells or peripheral blood mononuclear cells, leading to the development of a functional human immune system within the murine host (79). Human skin can then be transplanted onto these mice, and the engrafted human immune cells mount an allogeneic immune response against the human skin graft (12). This experimental approach allows researchers to dissect the contributions of various human immune cell subsets, such as T cells (80) and antigen-presenting cells (14), to graft rejection. Importantly, it also allows the investigation of the human HLA molecules and costimulatory receptors that drive allorecognition. Moreover, humanized mouse models of skin transplantation are a valuable platform for testing novel human-specific immunosuppressive therapies, such as adoptive transfer of Tregs (57, 81), as well as identifying novel human therapeutic targets, like Notch1 blockade (82) and rejection biomarkers, including the chemokine CCL18 (83). These models also enable the study of tolerance-inducing strategies that would be impossible to evaluate in traditional murine systems. Furthermore, humanized skin transplant models can be tailored by using PBMC from different patient populations, such as those with prior immune sensitization to alloantigens. Of course, these models do have important limitations, including the potential for incomplete immune system maturation (in the case of use of hematopoietic stem cells), lack of functional B cells, and variability in engraftment. However, we submit that humanized models may serve as a useful bridge between pre-clinical murine studies and clinical trials in humans, offering insights into the cellular and molecular dynamics of human transplant rejection in a controlled, yet physiologically relevant, in vivo setting.

6. Limitations of the Model and Methods for Increasing Physiologic Relevance

The major limitation of skin transplant models is the fact that skin is not a primarily vascularized tissue; thus, there is far less exposure of donor class II MHC molecules to the recipient immune system early following transplantation (55). As mentioned earlier, this fact alters the involvement of CD4⁺ T cells in skin transplantation as compared to vascularized transplants such as a heart or kidney. Moreover, it significantly curtails the direct involvement of anti-donor antibodies in mediating skin graft pathology (55). Because of this, skin transplantation in the mouse is not a model of antibody-mediated rejection. This limitation should be kept at the forefront when interpreting the results of pre-clinical studies of immunosuppression in the murine skin allograft model. However, skin transplantation does elicit the development of donor-specific B cell responses and DSA and, therefore, may be useful in studying mechanisms of alloantigen sensitization in vivo. For instance, it has been used in NHP studies to generate sensitized recipients for use in subsequent renal transplantation (84).

In contrast, as mentioned above, one of the main processes that can be studied using murine transplant models is the role of cellular immunity in mediating allograft rejection, in particular, the role of CD8⁺ effector T cells trafficking into allografts and mediating inflammation and parenchymal cell destruction. However, it is well known that memory CD8⁺ T cells acquire surface receptors that allow more efficient trafficking into peripheral tissues. Thus, one limitation of using laboratory mice is that they contain far fewer numbers of memory T cells as compared to immunologically mature human adults (85), including tissue resident memory T cells (86) that have recently been shown to contribute to allograft rejection (87, 88). Seminal studies have demonstrated that rendering recipients “immunologically experienced” by exposure to environmental or microbial organisms prior to transplantation can result in the differentiation of the T cell compartment to better reflect the maturation status of adult human transplant recipients (89). Thus, depending on the question being addressed, one may consider employing the use of immunologically experienced murine transplant recipients in skin transplant models.

Likewise, underlying disease can also impact memory T cell differentiation. For instance, it is now well-established in human patients that chronic diseases such as coronary disease, hypertension, and chronic kidney disease are associated with perturbations in the T cell repertoire (90–93). These alterations in the degree of T cell activation, differentiation, and senescence are likely to impact the magnitude and functionality of alloreactive T cell responses following transplantation (94). Of note, we recently demonstrated altered T cell differentiation programs in the presence of chronic kidney disease in a murine 5/6^th nephrectomy model in the mouse (95). Thus, induction of relevant underlying co-morbidities could be considered when designing skin transplantation experiments in the mouse model to further increase physiologic relevance and translatability of findings.

Significant differences between mouse and human skin exist regarding their structure, composition, and functions, particularly in hair follicles, sweat glands, and the mechanism by which wound healing occurs. Moreover, murine skin contains dendritic epidermal T cells (DETCs), a unique cell type which are essential for its healing process, as well as higher numbers of dermal γδ T cells compared to human skin (96). These differences may limit the relevance of skin transplantation as a model of wound healing in humans.”

Finally, another limitation of the murine skin transplant model is a limitation common to most rodent models based upon standard practice for conducting mechanistic studies. That is, most rodent experiments are set up as “single-center” studies, and initial results are not validated across multiple centers. Moreover, it is likely advisable that laboratory staff responsible for scoring skin transplant rejection are blinded to the experimental groups. It has been proposed that the generation of consortia to cross-validate key experimental results across multiple centers could improve the reproducibility and rigor of murine experimental transplant data.

7. Looking ahead: Use of the murine transplant model in future investigations

Despite these limitations, the murine skin transplant model remains a powerful and cost-effective tool for interrogating mechanisms of immunity during transplantation. One area of emerging investigation is the role of the skin microbiome in immunity. Seminal studies have shown that the skin microbiome plays a significant role in skin transplant rejection by modulating the host’s immune response (97). A diverse microbiome was associated with accelerated rejection due to enhanced T-cell activation, while reducing microbial diversity via germ-free environments or antibiotic treatment improved transplant survival. Skin transplantation will continue to be a powerful tool to interrogate the mechanisms by which commensal microbes influence immune cells to either promote or dampen the immune response to the transplanted tissue, and potentially alter susceptibility to immunosuppressive therapies.

Moving forward, the incorporation of multi-omics approaches in the murine skin graft model will continue to significantly increase the mechanistic insight that can be gleaned from these studies. For example, single-cell RNA sequencing (scRNA-seq) to map the heterogeneity of graft-infiltrating immune cells, TCR/BCR sequencing to understand the clonal diversity of the alloreactive repertoire, and proteomics/metabolomics to identify new rejection biomarkers are ongoing areas of investigation for which the skin transplant model could be employed (98, 99). Looking further ahead, artificial intelligence (AI) may be used to develop a digital twin for the skin transplant rejection model: a virtual, real-time replica of an animal’s skin and immune system (100). By continuously integrating data, this model would predict the likelihood and progression of graft rejection following perturbation of one or more immunologic pathways. This technology is still emerging, but is supported by advancements in AI, genomics, and computational biology, which may facilitate further refinement of the skin transplant model in studies of alloimmunity.

Abbreviations

CFSE

Carboxyfluorescein succinimidyl ester

DAMP

danger associated molecular patterns

DC

Dendritic cells

DETCs

dendritic epidermal T cells

dLN

draining lymph nodes

MST

Median survival time

NHP

non-human primate

NOD

non-obese diabetic

OT-I

OVA-specific TCR restricted by class I

OT-II

OVA-specific TCR restricted by class II

OVA

Ovalbumin