Leptin and Wound Inflammation in Diabetic ob/ob Mice: Differential Regulation of Neutrophil and Macrophage Influx and a Potential Role for the Scab as a Sink for Inflammatory Cells and Mediators

Female C57BL/6J (wild-type) or C57BL/6J-ob/ob mice were obtained from The Jackson Laboratories (Bar Harbor, ME) and maintained under a 12-h light-dark cycle at 22°C until they were 8 weeks of age. At this time they were caged individually, monitored for body weight, and wounded as described below.

Murine recombinant leptin (Calbiochem, Bad Soden, Germany) was injected intraperitoneally once a day at 8 a.m. (2 μg/g body wt) in 0.5-ml PBS per injection for 13 days. For local treatment, wounds of mice were covered with 1 μg leptin in 20 μl PBS twice a day (8 a.m. and 8 p.m.). Control mice were treated with PBS alone.

Mice were wounded as described previously (17,18). For each experimental time point, tissue from four wounds each from four animals (n = 16 wounds, RNA analysis) and from two wounds each from four animals (n = 8 wounds, protein analysis) were combined and used for RNA and protein preparation. Nonwounded back skin from four animals served as a control. All animal experiments were carried out according to the guidelines and with the permission from the local government of Hessen, Germany.

RNA isolation and RNase protection assays were carried out as described previously (18,19). The murine cDNA probes were cloned using RT-PCR. The probes corresponded to nt 816–1,481 (for lipocalin, accession no. X81627), nt 425 (exon 1) to 170 (exon 2) (for lysozyme M, M21047), nt 481–739 (for interleukin [IL]-1β, NM008361), nt 541–814 (for tumor necrosis factor [TNF]-α, NM013693), nt 163–317 (for GAPDH, NM002046), nt 181–451 (for macrophage inflammatory protein [MIP]-2, NM009140), nt 50–290 (for keratinocyte-derived chemokine [KC], NM008176), nt 63–323 (for macrophage chemoattractant protein [MCP]-1, NM011333), nt 139–585 (for vascular endothelial growth factor [VEGF], S38083), or nt 40–298 (for RANTES, NM013653) of the published sequences.

Wound lysates were prepared as described previously (17,20). Total protein (10–50 μg diluted in lysis buffer [20] to a final volume of 50 μl) from skin lysates was subsequently analyzed for the presence of immunoreactive MIP-2, MCP-1, IL-1β, and TNF-α by enzyme-linked immunosorbent assay (ELISA) using the Quantikine murine ELISA kits (R&D systems, Wiesbaden, Germany). Fifty microliters of supernatants from control and cytokine (2 nmol/l IL-1β, 2 nmol/l TNF-α)-stimulated murine PAM 212 keratinocytes were analyzed for MIP-2 by ELISA (R&D systems, Wiesbaden, Germany).

Wound biopsies were selected from surgery in normal individuals (normal wound), from chronic venous leg ulcers in patients without diabetes (ulcus cruris), and from diabetic foot ulcers in patients with diabetes (diabetic foot ulcers). Histological sections were analyzed from three patients in each group. All wounds were not debrided for at least 2 weeks before biopsy. All nondiabetic individuals had no medication, besides compression in the case of venous leg ulcers, for 2 weeks for biopsy, The diabetic patients were under oral antidiabetic therapy, and the diabetic foot ulcers were not of the typical neuropathic type.

Complete murine wounds were isolated from the back, bisected, and frozen in tissue-freezing medium. Six-micrometer frozen sections were subsequently analyzed using immunohistochemistry as described (17). Antisera against murine F4/80 antigen (Serotec, Eching, Germany), murine Gr-1 (Ly-6G) (Pharmingen, Hamburg, Germany), and murine cyclooxygenase (COX)-2 (Santa Cruz, Heidelberg, Germany) were used for immunodetection. Human paraffin skin sections were analyzed for the leptin receptor ObR (Santa Cruz, Heidelberg, Germany), or macrophage (Dako, Hamburg, Germany). Formal consent was obtained from the subjects after the nature of the procedure was explained.

Thirteen days after injury, wounds were excised and fixed in 4% paraformaldehyde (PFA)/PBS solution. Six-micrometer serial sections were subsequently analyzed for MIP-2 mRNA or MCP-1 mRNA expression using the HybriProbe in situ hybridization assay (Biognostik, Göttingen, Germany). MIP-2- or MCP-1-specific oligonucleotides were derived from the published sequences (see above).

Blood glucose levels were determined using the Accutrend sensor (Roche Biochemicals, Mannheim, Germany), serum leptin and insulin by ELISA (Crystal Chemicals, Chicago), and serum cortisol and corticosterone by radioimmunoassay (RIA) (ICN, Enschede, Germany) as described by the manufacturer.

Quiescent murine PAM 212 keratinocytes were stimulated with the cytokines IL-1β (1 ng/ml), TNF-α (5 ng/ml), or interferon (IFN)-γ (100 units/ml) as indicated. After 24 h of stimulation, conditioned cell culture supernatants were analyzed by ELISA. Cytokines were from Roche (Mannheim, Germany). Scab lysates were prepared in lysis buffer (20) and given to quiescent keratinocytes in a final concentration of 750 μg/ml.

Wound and cell culture lysates were prepared as described previously (17,20). Fifty micrograms of total protein from skin or cellular lysates was separated using SDS-gel electrophoresis. COX-2 and P-Tyr₇₀₅-STAT3 protein were detected using polyclonal antibodies (anti-COX-2 sc-1746; Santa Cruz, Heidelberg, Germany; and anti-P-Tyr₇₀₅-STAT3; New England Biolabs, Bad Schwalbach, Germany).

Data are shown as means ± SD. Data analysis was carried out using the unpaired Student’s t test with raw data and the Sigma Plot statistics computer program (Jandel Scientific, Erkrath, Germany).

First, we determined the potency of leptin to improve the metabolic syndrome as well as the impaired wound healing conditions observed in ob/ob mice. ob/ob mice were injected intraperitoneally with recombinant leptin (2 μg/g body wt, once a day) for 13 days. Three hours after injection, we determined serum leptin levels (320 ± 164 ng/ml). Additionally, to circumvent systemic effects of leptin, wounds from a second group of mice were directly covered with leptin (1 μg/20 μl PBS, twice a day). Following systemic leptin treatment, blood glucose levels readjusted to normal levels (100 mg/dl) (Fig. 1A). Moreover, a 13-day systemic leptin regimen resulted in a marked loss of body weight in wounded ob/ob mice, which could not be observed after topical treatment (Fig. 1B). In line with recently published data from our laboratory, we found a complete reepithelialization of the injured areas in both the systemically and the topically treated mice. This observation could be well documented by the loss of the scab from reepithelialized wounds (Fig. 1C and D). Additionally, we assessed wound diameters in PBS- and leptin-treated ob/ob mice. We determined an average wound size of 5.5 ± 0.8 mm (for PBS-treated mice) compared with 2.8 ± 0.7 mm (for leptin-treated mice) (n = 24 wounds, P < 0.01). A detailed wound size kinetics (day 1–8 postwounding) for leptin-treated ob/ob mice versus control animals has been published recently (16).

In this study, we focused on the role of inflammation under conditions of leptin-driven skin repair. First, we assessed PMN and macrophage populations at the wound site by determining the constitutively expressed molecular markers lipocalin (for PMNs) and lysozyme M (for macrophages). Impaired healing conditions in PBS-treated and untreated ob/ob mice revealed an increase in PMNs (Fig. 2A and B, left panels) and macrophages (Fig. 2A and B, right panels) during late repair (days 7 and 13 postwounding). This observation was confirmed by immunohistochemistry staining specific protein markers for PMN (GR-1) or macrophage (F4/80) (Fig. 2C, as indicated). Systemic administration of leptin, surprisingly, led to a strong decline only in PMN (Fig. 2A and B, left panels), but not in macrophage (Fig. 2A and B, right panels), cellular numbers at late time points of repair. Granulation tissues of wounds from leptin-treated mice were depleted from PMNs (Fig. 2C, upper right panel), but had tremendous numbers of macrophages (Fig. 2C, lower right panel). However, these cells were most likely to reflect an inactivated state. As shown in Fig. 2D, macrophages represented the COX-2-expressing cell type in the developing stroma of wounds (left panel). As we observed equal numbers of macrophages in the wound stroma for PBS- and leptin-injected mice, the massive expression of COX-2 mRNA in PBS-treated ob/ob mice probably reflected an activated state of the macrophage, which was markedly attenuated by leptin-treatment of the animals. Moreover, subsequent analysis of both the tissue and scab wound compartments revealed the predominant presence of PMNs within the wound scab, whereas macrophages tended to accumulate in the underlying wound tissue (Fig. 2E).

As a next step, we hypothesized that differences in PMN-specific CXC- and macrophage-specific CC-chemokine expression might contribute to the differential infiltration behavior of PMNs or macrophages into wounded sites of leptin-treated ob/ob mice. Evidently, systemic administration of leptin strongly attenuated mRNA expression for the PMN attractants MIP-2 (Fig. 3A, left panels, and B) and keratinocyte-derived chemokine (KC) (Fig. 4A, right panels). In situ hyridization (ISH) in 13-day wounds revealed the developing neoepithelium as the MIP-2-expressing wound compartment (Fig. 3C, lower right panels). Whereas leptin completely shut off the MIP-2-specific mRNA signals in keratinocytes, the epithelia from PBS-injected control mice were still markedly immunopositive for MIP-2 mRNA (Fig. 3C). Note the well-developed granulation tissue and multilayered neoepithelium that was associated with an improved skin repair in leptin-administered mice (Fig. 3C, lower left panel). The overall decline in MIP-2 mRNA (Fig. 3B) was paralleled by a reduction of MIP-2 protein in total wound tissue of leptin-treated animals (Fig. 3D, left panel). To further analyze a compartmentalization of MIP-2 expression, we subsequently separated the scab from the underlying wound tissue before ELISA (Fig. 3D, middle and right panels). Interestingly, the scab isolated from delayed-healing wounds was characterized by high concentrations of this chemokine (Fig. 4D, right panel). Moreover, we also assessed MIP-2 protein levels from the few scabs that remained on wounds obtained from the leptin-treated ob/ob mice (see Fig. 1C). As these few scab-covered wounds of leptin-administered mice resembled the status of wounds from untreated control animals, we found, not unexpectedly, comparable amounts of MIP-2 for both healing conditions (Fig. 3D, right panel). These data suggest that the scab might function as a sink or reservoir for epithelial-derived MIP-2. Moreover, topical treatment of wounds with leptin clearly failed to attenuate PMN influx directly (Fig. 3E, upper panel), although we observed an activation of the signal transducer and activator of transcription (STAT)-3 transcription factor at the wound site (Fig. 3E, lower panel). Cytokine-stimulated cultured murine PAM 212 keratinocytes did not respond to leptin exposure with an attenuated MIP-2 expression, although PAM keratinocytes nicely responded to a leptin stimulus with activation of the janus kinase (JAK)/STAT signaling cascade (Fig. 3F).

Unlike the situation reported here for CXC chemokines, we observed no changes in mRNA expression for MCP-1 (Fig. 4A, left panels, and B) and Rantes (Fig. 5A, right panels) in PBS- and leptin-treated as well as in untreated ob/ob mice. Keratinocytes of the neoepithelium did not respond to a leptin-stimulus with reduction of MCP-1 mRNA expression (Fig. 4C, middle panel). In contrast to MIP-2 regulation, we found a persisting MCP-1 protein expression in wound tissue (Fig. 4D, middle panel). High levels of MCP-1 in isolated scabs again indicated that the wound crust might function as a source or sink for protein mediators (Fig. 4D, right panel). Obviously, this phenomenon may present an explanation of reduced MIP-2 (Fig. 4) and MCP-1 (Fig. 5) protein levels in total wound tissue. Leptin-treatment accelerated wound repair and treated animals lost their wound scabs earlier in time than PBS-injected control mice (Fig. 1C).

As a next step, we determined macrophage-derived proinflammatory mediators in the PBS- and leptin-treated experimental setups. We found a massive decrease in TNF-α and IL-1β mRNA and protein levels for both healing conditions (Fig. 5A–D, left panels). The overall reduction of both cytokines was probably due to the loss of scabs from improved healing conditions (Fig. 1C). Nevertheless, TNF-α and IL-1β proteins were much more enriched in isolated scabs than in the underlying wound tissue (Fig. 5B and D).

To investigate the possibility that the scab might function as a source for bioactive molecules during repair, we isolated fresh scab tissue from control mice on day 5 after wounding. As we had determined concentrations of ∼200 pg for IL-1β and TNF-α in 50 μg of isolated scab protein (Fig. 5B and D, right panels), we now added a total of 750 μg of scab protein to stimulate PAM 212 keratinocytes with a similar amount of scab-derived cytokines as compared with recombinant cytokine control conditions (f.c. 1–3 ng/ml) (Fig. 6). Interestingly, we were able to observe an induction of VEGF mRNA expression in cultured keratinocytes (18) in the presence of scab proteins, clearly indicating that the scab indeed included bioactive protein mediators that were capable to drive gene expression in keratinocytes (Fig. 7).

Here, we determined to what extent changes in immune cell influx, cytokines, and skin repair might also be influenced by leptin-mediated reduction of food intake. To this end, we performed a wounding study in PBS- or leptin-treated and in a control group of pair-fed (caloric-restricted) ob/ob mice, which received equal amounts of nutriment as leptin-treated animals. Pair-feeding indeed resulted in a reduction of body weight comparable to the leptin-treated group of animals. However, overall wound repair was not improved (Fig. 7A, upper panels). Food restriction significantly reduced serum insulin, but not glucose, cortisol, or corticosterone levels (Fig. 7A). Analysis of cytokines (TNF-α and IL-1β), chemokines (MIP-2 and MCP-1), and immune cells (PMNs, macrophages) revealed that caloric restriction did not contribute to beneficial effects of leptin treatment (Fig. 7B). Moreover, it became clear that systemic levels of immune-modulatory glucocorticoids were not significantly changed by those amounts of leptin that were able to regulate immune function differentially and improve skin repair (Fig. 7A, lower panels).

Finally, we intended to make a case for leptin-mediated movements in chronic healing conditions in humans as well. Immunohistochemistry revealed that the leptin receptor was located in keratinocytes of the wound margins and within the dermis in undisturbed human wounds (Fig. 8A and B). Interestingly, and in clear contrast to chronic wound conditions in ob/ob mice (15), we could not detect ObR protein in wound margin epithelia of venous as well as diabetic ulcers. More importantly, whereas immune cells within the granulation tissues of venous ulcers expressed the ObR (Fig. 8C and D), we could not find any staining within diabetic foot ulcer tissue (Fig. 8E and F), although both conditions were characterized by large numbers of infiltrating macrophages and also PMNs (data not shown). Additionally, ObR-responsive cells within the scabs of diabetic ulcers did not represent macrophages, which failed to accumulate in large numbers within the scab (Fig. 8G and H).

Chronic wound-healing disorders represent a serious problem of growing clinical importance. There are a number of known pathological conditions that severely interfere with a coordinated and successful wound closure. Three prototypic chronic wounds are of clinical importance: decubitus or pressure, venous, and the diabetic ulcers (21). Moreover, disturbances in the inflammatory and proliferation phases also delay acute-wound repair from progressing to later stages of healing (22,23). A naturally occurring mouse mutant, the db/db mouse, which suffers from a diabetes-obesity syndrome (10), has long been used as an animal model system to investigate wound healing deficiencies during acute repair. The most important function of leptin is its inhibitory ability to suppress appetite. Thus, leptin reverses the diabetes-obesity syndrome in leptin-deficient ob/ob mice and also attenuates food intake and increases activity in healthy mice (24–26). Additionally, the ob/ob and db/db mice develop syndromes such as inhibition of reproductive functions (14), hormonal disturbances related to the thyroid, hypothalamic-pituitary-adrenal, and growth hormone axes (13), as well as alterations in the hematopoietic and immune systems (14). It is reasonable to suggest that the complexity of the diabetes-obesity syndrome in these animals is most likely to interfere with normal skin repair at various levels of regulation and must not be functionally connected to the hyperglycemia of the animals.

Leptin administration led to serum leptin levels that were about fourfold higher than those described for the db/db mice model (27). However, data from this study demonstrate that changes in serum glucocorticoid and insulin levels did not contribute to leptin-mediated improvement of repair. However, our leptin/pair-fed experiments strongly suggested that attenuation of blood glucose might play a major role in acceleration of healing. Accordingly, some of the proposed mechanisms that drive hyperglycemia-induced pathological changes, such as glycation end products, hyperosmolarity, abnormal myoinositol metabolism, oxidant formation, or protein kinase C activation (28–32), might be improved by leptin treatment. However, leptin-driven skin-tissue responses in ob/ob mice could be transferred, at least partially, to normal wound-healing conditions, as reepithelialization of wounded sites also could be further accelerated by leptin in healthy mice (15). We found the leptin receptor ObR to be expressed in wound-margin keratinocytes of undisturbed human wounds. Nevertheless, a potential role for leptin in acute skin repair in humans has to be determined. However, circulating leptin levels have been described to be present in humans and even to rise in direct correlation to the body’s white fat tissue (33,34). Thus, diabetes-impaired healing disorders in humans develop in the presence of leptin, suggesting a leptin resistance or insensitivity in peripheral tissues. In line, we observed the absence of leptin receptor expression in diabetic foot ulcers. It appears that deficiencies in leptin responses in ob/ob mice as a model system and humans might be determined by different mechanisms, thus highlighting a potential shortcoming of the diabetes-impaired ob/ob mouse model.

This study provides evidence that endocrine mediator molecules may strongly influence tissue movements in injured areas of the body. A leptin-driven improvement of inflammation during repair was not altogether unexpected, as we had hypothesized a normalization of inflammatory conditions that might contribute to leptin-accelerated repair. Again, we found a severe dysregulation of the macrophage subset of immune cells in not only delayed-healing wounds, but, surprisingly, in also well-developed wounds of leptin-treated animals. Leptin here mediates improved tissue movements, as shown convincingly for epithelial sites (15), in the presence of tremendous numbers of persisting macrophages (this study). Evidently, the number of circulating macrophages was increased about fourfold in ob/ob mice (35). This over-representation might contribute to the large numbers of macrophages observed in wounds of these animals. It is noteworthy that the MIP-2 and MCP-1 chemokines and the inflammatory mediators TNF-α and IL-1β were predominantly enriched in the wound scab. Early studies with wound dressings suggested that the formation of a scab in air-exposed wounds was superior in supporting granulation tissue formation when compared with wound dressings (36,37). Accordingly, the release of bioactive components from the scab might provide one explanation of the beneficial effects of the scab for repair.

Evidently, a systemic effect of leptin is supported by the observation that topical treatment of skin wounds did not result in an attenuation of PMN infiltration behavior. The observation that endocrine signals interfere with the immune cell population of skin wounds might be of importance for chronic wound conditions in humans. Histological analyses of chronic leg ulcers in diabetic humans indicated increased numbers of granulocytes and macrophages which were located in the center and at the edges of the wound (6,7). However, the observed macrophages in chronic diabetic wounds were probably in an inactive state (8). This observation is consistent with the absence of leptin receptor (ObR) expression in diabetic foot ulcers. In fact, debridement of diabetic ulcers impacts inflammatory cells, but does not always solve the problem of impaired healing conditions. It is reasonable to suggest that inactivation of macrophages might be due to a downregulation of several receptors for extracellular signals that control macrophage activation during repair. Data from this study clearly indicate that cells (keratinocytes, immune cells) must have lost their potency to respond to a leptin stimulus. Moreover, loss of ObR expression might be confined to diabetic conditions, as venous ulcers were characterized by ObR-expressing cells within the granulation tissue.

However, macrophages isolated from ob/ob mice express phenotypic abnormalities such as increased superoxide and hydrogen peroxide production as well as elevated IL-6 and COX-2 levels in vitro (38). Our in vivo data support this in vitro situation. Leptin-treatment triggered a downregulation of COX-2 expression at the wound site, as COX-2 expression completely disappeared in the presence of unaltered numbers of resident macrophages. Thus, the observed macrophage population of nontreated and treated ob/ob mice (which did not differ in total numbers) might be distinguished markedly by their inflammatory potencies. The role of leptin as a pivotal mediator of inflammatory processes is strengthened by a recent study of intestinal inflammation in ob/ob mice (39). However, the outcome of leptin-controlled intestinal inflammation in ob/ob mice completely contrasted the situation in skin repair in this animal model, as leptin replacement in ob/ob mice increased the severity of intestinal inflammation by augmentation of PMN influx and cytokine production. It is surprising that the same individual leptin-replaced animal would, however, attenuate MIP-2 and PMN influx at sites of inflammation in the context of skin repair (this study). It is reasonable to suggest leptin as an important endocrine signaling molecule that is tightly connected to the immune system, but the final outcome of leptin’s regulatory potency on different axes of immune responses might be fundamentally different for different organ systems.

This work was supported by the Deutsche Forschungsgemeinschaft (SFB 553, grant FR 1540/1–1).

We thank M. Kock for help with the animal experiments, L. Raspe for critically reading the manuscript, and N. Kämpfer-Kolb and K. Weinelt for excellent technical assistance.