Characterization of Dermal Stem Cells of Diabetic Patients

Diabetic foot ulcers (DFUs) are lesions that involve loss of epithelium and dermis, sometimes involving deep structures, compartments, and bones. The aim of this work is to investigate the innate regenerative properties of dermal tissue around ulcers by the identification and analysis of resident dermal stem cells (DSCs). Dermal samples were taken at the edge of DFUs, and genes related to the wound healing process were analyzed by the real-time PCR array. The DSCs were isolated and analyzed by immunofluorescence, flow cytometry, and real-time PCR array to define their stemness properties. The gene expression profile of dermal tissue showed a dysregulation in growth factors, metalloproteinases, collagens, and integrins involved in the wound healing process. In the basal condition, diabetic DSCs adhered on the culture plate with spindle-shaped fibroblast-like morphology. They were positive to the mesenchymal stem cells markers CD44, CD73, CD90, and CD105, but negative for the hematopoietic markers CD14, CD34, CD45, and HLA-DR. In diabetic DSCs, the transcription of genes related to self-renewal and cell division were equivalent to that in normal DSCs. However, the expression of CCNA2, CCND2, CDK1, ALDH1A1, and ABCG2 was downregulated compared with that of normal DSCs. These genes are also related to cell cycle progression and stem cell maintenance. Further investigation will improve the understanding of the molecular mechanisms by which these genes together govern cell proliferation, revealing new strategies useful for future treatment of DFUs.


Introduction
Diabetes mellitus is a universal health problem. People suffering from diabetes endure long-term complications such as cardiovascular diseases, nephropathy, retinopathy, neuropathy, and ulcers in the lower limbs. Indeed, diabetic neuropathy and peripheral arterial disease are common factors that cause diabetic foot ulcers (DFUs). In particular, motor neuropathy causes muscle weakness, atrophy, and paresis, sensory neuropathy leads to loss of the protective sensation of pain, pressure, and heat, and autonomic dysfunction causes vasodilation and decreased sweating, resulting in a loss of skin integrity, providing a site vulnerable to microbial infection [1]. It is estimated that between 15% and 25% of patients suffering from diabetes mellitus develop during their lifetime skin ulcers below the ankle, which seriously affect their quality of life, can be limb-and life-threatening, and are responsible for the majority of hospital admissions among diabetics. DFUs constitute a major public health burden in both the developed and developing countries [2]. Current treatment of DFUs encompasses

Patients Recruitment and Samples Collection
The study was conducted at Maria Cecilia Hospital (Cotignola, Ravenna, Italy) following the ethical principles for medical research involving human subjects of the World Medical Association Declaration of Helsinki. The patients participated in the study after signing the consent form. Patients met the following inclusion criteria: older than age 18 years; known diagnosis of diabetes mellitus; assessment of critical limb ischemia and infection was carried out; signature of the consent form. The exclusion criteria were known or suspected cancer diagnosis, chronic renal failure in dialytic treatment, and life expectancy of less than 1 year.
The study involved 20 diabetic patients with critical ischemia of the lower limbs and ulcer. The management of the infected ulcer involved antibiotic therapy targeted to microbiological isolations and surgical debridement to remove the necrotic and infected area. During surgery, the discarded tissue was stored and destined for the laboratory investigations. The dermis at the edge of the ulcer was used for gene expression analysis by real-time PCR array and for the isolation and characterization of resident stem cells.

RNA Extraction and Real-Time PCR Array
Total RNA was extracted from dermis or cultured cells with the RNeasy Mini Kit (Qiagen, Hilden, Germany), including an on-column DNase digestion step by the RNase-Free DNase Set (Qiagen), according to the manufacture procedure. The concentration and purity of the RNA were checked by spectrophotometric measurement on the NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Of each sample, 500 ng of total RNA was reverse-transcribed with RT 2 First Strand Kit (Qiagen) in a SimpliAmp Thermal Cycler (Applied Biosystems, Thermo Fisher Scientific) following the manufacture procedures. The resultant first-strand cDNAs were stored at −20 • C until the next step. Human wound healing RT 2 Profiler PCR Array and human stem cell RT 2 Profiler PCR Array (Qiagen) were performed, in accord with the manufacture protocol. Briefly, the cDNA samples were mixed with RT 2 SYBR Green Mastermix (Qiagen), and then aliquoted into the wells of the RT 2 Profiler PCR Array. A StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) was set up with the following thermal cycling conditions: denaturation at 95 • C for 10 min; followed by 40 cycles of denaturation at 95 • C for 15 s; annealing and elongation at 60 • C for 1 min. A dissociation curve for each well was performed by running the following program: 95 • C for 1 min; 65 • C for 2 min; 65 • C to 95 • C at 2 • C/min. Relative expression was determined using the 2∆∆CT method. Ct values of target genes were normalized to the geometric mean Ct values of five housekeeping genes (ACTB, B2M, GAPDH, HPRT1, and RPLP0). Results were reported as fold regulation of target genes in the test group (pathological) compared with the control group (healthy). Fold regulation values greater than 2 indicate increased gene expression, fold regulation values less than −2 indicate decreased gene expression, and fold regulation values between −2 and 2 indicate indifferently expressed genes. Data analysis was performed by using the GeneGlobe Data Analysis Center (Qiagen). The data analysis software does not perform any statistical analysis beyond the calculation of p values using Student's t-test based on 2∆CT values for each gene of the test group compared with the those of control group. Statistical significance was set at p < 0.05.

Cells Isolation and Characterization
Dermis was removed from DFU biopsy and washed in phosphate-buffered saline (PBS, EuroClone, Milano, Italy) added with 1% antibiotic-antimycotic (AA, Thermo Fisher Scientific, Waltham, MA, USA). Dermal tissue was minced and digested with 200 U/mL collagenase type II (Gibco, Thermo Fisher Scientific) in Hanks' balanced salts solution (HBSS, Euroclone) at 37 • C for 16 h. The resulting cells were pelleted, rinsed with PBS, and counted using the trypan blue exclusion assay. They were seeded at a density of 5 × 10 4 cells/cm 2 in Basal Medium (BM) consisting of Dulbecco's modified Eagle's medium (DMEM, Euroclone) supplemented with 10% Fetal Bovine Serum (FBS, EuroClone), and 1% AA. Cell cultures were maintained at 37 • C and 5% CO 2 , and medium was changed twice a week.
For immunofluorescence staining, 2 × 10 4 cells/cm 2 were seeded on glass coverslips put into 24 well plates and cultured in BM. The following day, cells were fixed in 4% paraformaldehyde (Sigma-Aldrich, St. Louis, MO, USA) for 10 min. After three washes, cells were incubated in 3% bovine serum albumin (BSA; Sigma-Aldrich) solution in PBS at room temperature (RT) for 1 h. Then, cells were incubated overnight at 4 • C with the primary antibodies: mouse anti-human CD44 (Thermo Fisher Scientific), rabbit anti-human CD73 (Abcam, Cambridge, UK), rabbit anti-human CD90 (Abcam, Cambridge, UK), and mouse anti-human CD105 (Thermo Fisher Scientific). Then, cells were incubated with the fluorescent secondary antibodies goat anti-rabbit Alexa Fluor 555 (Thermo Fisher Scientific), or goat anti-mouse Alexa Fluor 488 (Thermo Fisher Scientific) at RT for 1 h. Actin staining was performed with Phalloidin Alexa Fluor 555 (Thermo Fisher Scientific) in PBS for 20 min at RT, and nuclear staining with NucBlue Fixed Cell Stain (4 ,6-diamidin-2-fenilindolo, DAPI; Thermo Fisher Scientific) in PBS for 5 min at RT. Immunofluorescence images were acquire on an Upright ECLIPSE Ni Microscope (Nikon, Minato, Tokyo, Japan).
For flow cytometry, as previously described [13], cells were dissociated and resuspended in flow cytometry staining buffer (R&D Systems, Minneapolis, MN, USA) at a final cell concentration of 1 × 10 6 cells/mL. Cells were incubated with the following fluorescent monoclonal mouse anti-human antibodies (eBioscienceTM, Thermo Fisher Scientific): CD14 R-PE; CD34 FITC; CD44 FITC; CD45 APC; CD73 APC; CD90 R-PE; CD105 PE-Cy 7; HLA-DR FITC. Cells were washed twice with 2 mL of flow cytometry staining buffer and resuspended in 500 µL of flow cytometry staining buffer. Flow cytometry analyses were performed on an Attune NxT flow cytometer (Thermo Fisher Scientific) with the Attune NxT software (Thermo Fisher Scientific). Each experiment was performed independently three times. Results were expressed as mean ± standard deviation (SD).
Cell growth has been investigated by the cumulative population doubling (CPD) assay. Briefly, 1.2 × 10 5 cells at passage 2 (p2) were seeded into 6 well plates. Every two days, cells were detached, counted, and seeded again at the same density in a new 6 well plate. This was repeated until the cells reached p6. The population doubling (PD) of the cells was calculated according to the formula: where PD represents the number of cell divisions that occur in each passage; N t corresponds to cell number on the second day, and N 0 is the initial seeding number of cells. To determine the CPD, the PD level for each passage was calculated and added to the levels of the previous passages. The experiment was performed independently three times. Results were expressed as mean ± standard deviation (SD). Student's t-test was performed to determine the statistical significance. Statistical significance was set at p < 0.05. Cell migration ability was investigated by in vitro wound healing assay. Briefly, 2 × 10 4 cells were seeded in 24 well plates and cultured until they reached confluence. Cell monolayers were scratched by a 100 µL sterile pipette tip. After a wash with PBS, fresh BM was added. Images of the whole wound were taken immediately after the scratch and after 24 h with a Nikon Inverted Microscope Eclipse Ti-E equipped with a Digital Sight camera DS-03. Scratch areas were measured at 0 and 24 h, and migration rate was calculated according to the formula: where A 0 corresponds to whole wound taken immediately after the scratch, and A t corresponds to whole wound taken after 24 h. At least three independent experiments were performed. Results were expressed as mean ± standard deviation (SD). Student's t-test was performed to determine the statistical significance. Statistical significance was set at p < 0.05.

Gene Expression Profile of Diabetic Dermis
In order to define the molecular alterations underlying the development of chronic ulcers in diabetic patients, the expression of 84 key wound healing associated genes was probed by real-time PCR array. The array comprises genes expressed during the four phases of the inflammatory response: coagulation, inflammatory, proliferative, and maturation phase. The expression profile of wound healing genes of diabetic dermis taken from the edge of 20 ulcers was compared with that of normal dermis. Plotting of the 84 detected transcripts on a volcano plot ( Figure 1) indicated that 31 genes were differentially expressed in diabetic dermis and healthy dermis by 2-fold or greater. A total amount of 11 genes was upregulated, whereas 20 genes were downregulated. The complete list of genes investigated and the fold regulation values are reported in Table 1.    Among the genes involved in the coagulation phase, coagulation factor III (F3), fibrinogen alpha chain (FGA), and plasminogen (PLG) were downregulated, whereas plasminogen activator urokinase receptor (PLAUR) was upregulated. Transcripts of inflammatory chemokines and cytokines, such as chemokine (C-C motif) ligand 2 (CCL2), chemokine (C-X-C motif) ligand 1 (CXCL1), chemokine (C-X-C motif) ligand 2 (CXCL2), and interleukin 1 beta (IL1B) were upregulated, and colony stimulating factor 2 (CSF2) and interleukin 2 (IL2) were downregulated. The proteolytic enzymes cathepsin G (CTSG), matrix metallopeptidase 1 (MMP1), and matrix metallopeptidase 9 (MMP9) were upregulated, whereas matrix metallopeptidase 2 (MMP2) downregulated. Overall, the proteins constituting the extracellular matrix showed a reduction in the gene expression profile. In particular, collagen type IV alpha 3 (COL4A3), collagen type V alpha 3 (COL5A3), and vitronectin (VNT) were downregulated. Also cell adhesion molecules integrin alpha 1 (ITGA1), integrin alpha 2 (ITGA2), integrin beta 3 (ITGB3) showed a downregulation. Moreover, the expression of growth factors registered a broad downregulation: fibroblast growth factor 2 (FGF2), platelet-derived growth factor (PDGF), connective tissue growth factor (CTGF), and colony stimulating factor 2 (CSF2) were particularly downregulated. downregulated. Also cell adhesion molecules integrin alpha 1 (ITGA1), integrin alpha 2 (ITGA2), integrin beta 3 (ITGB3) showed a downregulation. Moreover, the expression of growth factors registered a broad downregulation: fibroblast growth factor 2 (FGF2), platelet-derived growth factor (PDGF), connective tissue growth factor (CTGF), and colony stimulating factor 2 (CSF2) were particularly downregulated.

Stem Cell Isolation and Morphological Characterization
Cells were isolated from diabetic and normal dermis by enzymatic digestion and plated under basal conditions in cell culture flasks. Like the normal cells, the diabetic cells have adhered to flask's plastic forming a monolayer. The immunostaining of actin filaments with phalloidin revealed a spindle-shaped fibroblast-like morphology both in diabetic and in normal cells ( Figure 2).  Table 2.  Table 2. downregulated. Also cell adhesion molecules integrin alpha 1 (ITGA1), integrin alpha 2 (ITGA2), integrin beta 3 (ITGB3) showed a downregulation. Moreover, the expression of growth factors registered a broad downregulation: fibroblast growth factor 2 (FGF2), platelet-derived growth factor (PDGF), connective tissue growth factor (CTGF), and colony stimulating factor 2 (CSF2) were particularly downregulated.

Stem Cell Isolation and Morphological Characterization
Cells were isolated from diabetic and normal dermis by enzymatic digestion and plated under basal conditions in cell culture flasks. Like the normal cells, the diabetic cells have adhered to flask's plastic forming a monolayer. The immunostaining of actin filaments with phalloidin revealed a spindle-shaped fibroblast-like morphology both in diabetic and in normal cells (Figure 2).  Table 2.  The rate of growth and migration of diabetic cells was compared with that of normal cells ( Figure  4). A CPD assay was performed to establish growth potential of diabetic and normal dermal cells during five consecutive passaging (from p2 to p6). The CPD corresponds to the total number of  The rate of growth and migration of diabetic cells was compared with that of normal cells (Figure 4). A CPD assay was performed to establish growth potential of diabetic and normal dermal cells during five consecutive passaging (from p2 to p6). The CPD corresponds to the total number of estimated divisions during the considered interval. The CPD tended to be lower for diabetic cells with respect to normal cells at all passages examined (Figure 4a). The migration rate was investigated by in vitro wound healing assay (Figure 4b-f). Even the migration rate of diabetic cells was lower than that of normal cells (Figure 4f). estimated divisions during the considered interval. The CPD tended to be lower for diabetic cells with respect to normal cells at all passages examined (Figure 4a). The migration rate was investigated by in vitro wound healing assay (Figure 4b-4f). Even the migration rate of diabetic cells was lower than that of normal cells (Figure 4f).

Gene Expression Profile of Diabetic Dermal Cells
The expression profile of genes characterizing stem cells was investigated in diabetic dermal cells by real-time PCR array. The array comprises genes that control self-renewal, cell division, cell cycle, cell communication, cell adhesion, metabolism, and differentiation. The transcripts of diabetic cells were compared with those of healthy DSCs. The volcano plot in Figure 5 shows that 29 genes were differentially expressed in diabetic DSCs and healthy DSCs by 2-fold or greater. In particular, 19 genes were upregulated and 10 were downregulated. The fold regulations of the 84 investigated genes are reported in Table 3.

Gene Expression Profile of Diabetic Dermal Cells
The expression profile of genes characterizing stem cells was investigated in diabetic dermal cells by real-time PCR array. The array comprises genes that control self-renewal, cell division, cell cycle, cell communication, cell adhesion, metabolism, and differentiation. The transcripts of diabetic cells were compared with those of healthy DSCs. The volcano plot in Figure 5 shows that 29 genes were differentially expressed in diabetic DSCs and healthy DSCs by 2-fold or greater. In particular, 19 genes were upregulated and 10 were downregulated. The fold regulations of the 84 investigated genes are reported in Table 3.     In diabetic DSCs, the genes related to self-renewal, symmetric, and asymmetric cell division showed a comparable expression profile to those of healthy DSCs. The transcription of NOTCH1, NOTCH2, WNT1, SOX2, and desert hedgehog (DHH) were equivalent to that in normal DSCs. In addition, the genes involved in the WNT signaling including APC, AXIN1, cyclin D1 (CCND1), frizzled family receptor 1 (FZD1), FRAT1, and MYC were equally expressed. The genes connected to the Notch signal were investigated: the transcription of delta-like 1 (DDL1) gene was equivalent to that in normal DSCs, whereas jagged1 (JAG1) and deltex 2 (DXT2) transcription was upregulated. The transcription of genes regulating cell cycle progression, such as cyclin A2 (CCNA2), cyclin D2 (CCND2), and cyclin-dependent kinase 1 (CDK1) were downregulated. The gene expression of growth factors, including fibroblast growth factor 1 (FGF1), fibroblast growth factor 2 (FGF2), and fibroblast growth factor 3 (FGF3), was equal in diabetic DSCs compared to normal DSCs. Instead the expression of bone morphogenetic protein 1 (BMP1), bone morphogenetic protein 3 (BMP3), and growth differentiation factor 2 (GDF2, alias BMP9) was upregulated in diabetic DSCs. Moreover, the gene expression of the metabolism genes such as ATP-binding cassette sub-family G member 2 (ABCG2) and aldehyde dehydrogenase 1 family member A1 (ALDH1A1) were downregulated.

Discussion
Skin wound healing is a dynamic process tightly regulated. The regular order in which tissue healing takes place is divided into four stages, corresponding roughly to the four waves of predominant cell type appearing in the wound bed. These are the hemostasis, inflammatory, proliferative, and maturation phases [14]. In the hemostasis phase, platelets and circulating coagulant factors accumulate at the site of tissue injury. The blood clotting is achieved by a cascade of enzymatic reactions, which involves prothrombin and a series of factors, which are converted to active proteases by hydrolysis. F3 is the primary initiator of the extrinsic blood coagulation. Upon injury of the vessel and surrounding tissue, F3 is exposed to blood coagulation factors. The complex with coagulation factor VII catalyzes the proteolytic activation of the coagulation factors X and IX, leading to thrombin generation [15]. Thrombin converts the soluble protein FGA to an insoluble fibrin gel. Thrombin also activates Factor XIII (F13A1), which crosslinks the fibrin polymers and forms a fibrin mesh that traps circulating platelets, leucocytes, and red blood cells [16]. The actions of thrombin and several other activated coagulation factors are inhibited by circulating antithrombin, e.g., serpin peptidase inhibitor clade E member 1 (SERPINE1). Once sufficient thrombin is produced to overcome the effect of circulating antithrombin, the coagulation is able to initiate. The clot dissolution is carried out by plasmin, which catalyzes the fibrin hydrolysis. Plasmin is generated from PLG by extracellular protease plasminogen activator urokinase (PLAU) and tissue-type PLAU (PLAT) that are directly activated and released in extracellular matrix (ECM) by a number of growth factors, e.g. hepatocyte growth factor (HGF) [17]. PLAU also plays a pivotal role in cell adhesion and migration by binding PLAUR. Even a modest increase in PLAUR expression is correlated with the ability of monocytes to penetrate stromal tissue in a PLAU-dependent process. In addition, PLAUR mediates cellular adhesion to VTN, promotes integrin-dependent migration, and initiates intracellular signaling events [18]. In diabetic dermis, the enzymes involved in the creation (F3 and FGA) and the dissolution (PLG) of fibrin clot were downregulated. On the contrary, PLAUR was upregulated to favor the migration of monocyte in the site of injury. The inflammatory phase is characterized by immune cells infiltration. Neutrophils are initially captured from the fast-flowing blood stream in the postcapillary microvasculature to the endothelial vessel wall, followed by rolling along the endothelium, integrin-dependent arrest and firm adhesion, intravascular crawling, and trans-endothelial migration into the inflamed tissue. Inflammatory stimuli, such as tumor necrosis factor alpha (TNFα) and lipopolysaccharide (LPS), induce the upregulation of adhesion molecules on the surface of platelets and vascular endothelial cells that enables these cells to stick to each other and to neutrophils [19]. The migration, adhesion, and activation of leucocytes also depends on the expression of specific integrins. Each integrin consists of one α (ITGA) subunit and one β (ITGB) subunit. Leucocytes by integrin α4β1 (alias CD49d/CD29) bind vascular cell adhesion protein 1 (VCAM-1) and FGA, whereas integrin αVβ3 (alias CD51/CD61) binds intercellular adhesion molecule 1 (ICAM-1), platelet endothelial cell adhesion molecule (PECAM-1), VCAM-1, VTN, and FGA [20]. Compared with the control condition, in the diabetic dermis the expression of ITGB3 was downregulated while ITGA4, ITGAV, and ITGB1 genes were indifferently expressed. This could be related to a lesser lymphocyte activation and a lesser adhesion of them to platelets and endothelial cells. In addition to physical contacts, platelets secrete proinflammatory chemokines such as, CXCL1 with a chemotactic activity for neutrophils, chemokine (C-X-C motif) ligand 5 (CXCL5) a neutrophils activator, CCL2 a potent chemoattractant for monocytes and basophils, and chemokine (C-C motif) ligand 7 (CCL7) a chemotactic factor for monocytes and eosinophils. Platelets also release soluble CD40 ligand (CD40L), which induces in endothelial cells the expression of inflammatory adhesion receptors (e.g., E-selectin, VCAM-1, ICAM-1), the production of chemokines and interleukins (ILs) (e.g., CCL2, IL6, and IL8), and the production of MMP9 [21]. Moreover, immune cells, including monocytes, macrophages, and lymphocytes secrete many proinflammatory cytokines and growth factor to amplify the inflammatory response. Proinflammatory ILs (e.g., IL1B, IL6, and IL8) and chemokines (e.g., CXCL2, CXCL11), TNFα, CSF2, colony-stimulating factor 3 (CSF3) are released to stimulate cytokine production, cell proliferation, macrophage activation, and increase neutrophil and monocyte function [22,23]. Compared with in the control group, the dermis of diabetic patients showed an upregulation of the pro-inflammatory chemokines and cytokine (e.g. CCL2, CXCL1, CXCL2, and IL1B), but a downregulation of CSF2 and IL2. CSF2 can enhance macrophage proliferation as well as modulate their differentiation and function, while IL2 modulates T-cell proliferation and activation. Despite an increase in chemokine expression, the downregulation of CSF2 and IL2 in diabetic dermis could underline an impairment in cell-mediated immune response. However, in diabetic dermis, the serine protease CTSG was upregulated. CTSG is released by activated neutrophils to clear pathogens and regulate inflammation by modifying chemokines, cytokines, and cell surface receptors. CTSG changes CXCL5 and chemokine (C-C motif) ligand 15 (CCL15) into more potent chemotactic factors by a proteolytic processing. In addition, CTSG causes cell shape modification and intercellular gap formation in endothelial cell that increase endothelium permeability, supporting the migration of neutrophils, monocytes and antigen-presenting cells in the injured site [24]. CTSG affects also tissue remodelling directly by degrading components of the matrix and indirectly by cleaving and activating matrix metalloproteinases (MMPs), such as MMP1 and MMP2 [24]. Activated neutrophils also release many angiogenic molecules, such as IL8, CXCL1, and MMP9. In particular, MMP9 and IL8, by degrading the basement membrane and promoting the endothelial cells migration, respectively, are responsible for the proangiogenic effects of neutrophils [25]. MMPs and their tissue inhibitors (TIMPs) are involved in various stages of the wound healing process, and their expression is regulated by several growth factors, including TNFα, transforming growth factor beta 1 (TGFB1), insulin-like growth factor 1 (IGF1), PDGF, epidermal growth factor (EGF), FGF2, vascular endothelial growth factor A (VEGFA), IL6, and IL10 [26]. During the inflammatory phase, MMPs activate or inhibit several cytokines and improve leukocyte invasion, creating a chemotactic gradient. Several chemokines, including CCL7 and chemokine (C-X-C motif) ligand 12 (CXCL12) are substrates for MMP2. MMP9 cleaves and activates chemokine (C-X-C motif) ligand 6 (CXCL6) and IL8, whereas it inactivates CXCL1 and chemokine (C-X-C motif) ligand 4 (CXCL4) [27]. The gelatinase MMP9 and MMP2 are also responsible for the breaking down of type IV collagen in the basement membrane, thus are involved in keratinocyte migration and re-epithelialization [26]. In diabetic dermis, the expression of MMP9 gene was upregulated, while their inhibitor TIMP1 was indifferently expressed. Previous studies on DFUs have reported that a high concentration of MMP9 in wound liquid and a high ratio of MMP9 to TIMP1 are related to poor wound healing [28,29]. A 5-fold increase in the expression of MMP1 was also detected in diabetic dermis. MMP1 is responsible for proteolytic degradation of type-1 and type-3 collagens as well as elastic fibres. It is crucial for wound re-epithelialization, and their dysregulation is associated to several pathological conditions, including cutaneous ulcer [30]. In diabetic dermis, the upregulation of CTSG, MMP9, and MMP1 may depend on bacterial infection at the site of DFU and on a dysregulation of growth factors.
Compared with the healthy control, diabetic dermis showed a downregulation of various growth factors, particularly PDGF, FGF2, and CTGF. PDGF works as a chemoattractant for inflammatory cells and fibroblasts. It induces fibroblast migration and proliferation, matrix production, production of granulation tissue proteins and provisional ECM, and angiogenesis during the healing process. Various clinical studies have shown that a downregulation of PDGF and its receptor are associated to an enhanced healing time [26]. FGF2 stimulates the growth and differentiation of fibroblasts, vascular smooth muscle cells, and endothelial cells. It increases the rate and degree of granulation tissue formation, stimulating the healing process [31]. The formation of new blood vessels in new-formed tissue is initiated by PDGF, FGF2, as well as VEGFA [32]. Instead, the production of ECM and the inhibition of proteases are influenced predominantly by the fibrogenic growth factors, including PDGF, IGF1, and TGFB1 [33]. CTGF modulates the interaction of cells with the ECM, promoting collagen deposition, mesenchymal cell activation and differentiation, and tissue remodelling [34]. The balance between matrix degradation and matrix formation is crucial in the wound healing process. In the diabetic dermis, different types of collagen were downregulated, including COL4A3, COL5A3, and VTN. Type IV collagen is the major structural component of basement membranes. Type V collagen regulates the assembly of heterotypic fibers composed of both type I and type V collagen. Therefore, the collagen fibers involved in the organization of the ECM, in angiogenesis and re-epithelialization appear to be compromised in diabetic dermis. VTN interacts with glycosaminoglycans and proteoglycans and is recognized by certain integrins and serves as a cell-to-substrate adhesion molecule. In diabetic dermis, even the ITGA1, ITGA2, and ITGB1 integrin subunits able to bind collagens are downregulated.
Smooth muscle cells, fibroblasts, and microvascular endothelium express integrin α1β1 to bind type IV collagen, whereas integrin α2β1 has predominantly epithelial distribution, showing a preference for type I collagen [35]. Among investigated integrins, only ITGB6 gene showed upregulation in diabetic dermis. However, integrin αvβ6 is an epithelium-restricted molecule expressed at low levels in healthy skin that is rapidly upregulated in response to inflammation and injury [36].
Overall, the gene expression profile of proximal dermis to DFUs showed a dysregulation in growth factors, MMPs, collagens, and integrins involved in the proliferation phase, angiogenesis, re-epithelialization, and remodeling. In this scenario, the resident stem cells of dermal tissue around DFUs were isolated and analyzed. The cells of diabetic dermis comply with the parameters of dermis stem cells (DSCs) defined by Vapniarsky and colleagues [37], although they showed a lower proliferation and migration rates compared with the cells of nondiabetic DSCs. Diabetic DSCs have adhered to culture flasks with a spindle-shaped fibroblast-like morphology under the basal condition. Moreover, in basal medium diabetic DSCs were positive to the mesenchymal stem cells markers CD44, CD73, CD90, and CD105, but negative for the hematopoietic markers CD14, CD34, CD45, and HLA-DR.
Despite the morphology and phenotype of diabetic DSCs reflecting that of mesenchymal stem cells, the expression profile of stemness genes in diabetic DSCs showed some differences compared with that in DSCs of nondiabetic donors. Stem cells are defined as cells able to continuously divide to produce more undifferentiated stem cells, the so-called self-renewing property, and to differentiate into specialized cells, namely multilineage differentiation potency [38]. These properties are closely related to the ability of stem cells to proliferate by symmetric and asymmetric division. Asymmetric cell division gives rise to two daughter cells with distinct cell fates: one daughter maintains stem cell properties and functions, while the other daughter loses these characteristics. Symmetric cell division enables stem cells to generate two daughter cells having less potency than the parental stem cell. This mode of division is critical for expanding stem cell reservoirs, leading to rapid production of cells but potential depletion of the stem cell pool. Both symmetric and asymmetric cell divisions are employed in vivo to maintain the fine balance between self-renewal and differentiation of stem cells [39]. Notch, WNT, FGF, BMP, and DHH signaling play a key role in maintaining tissue homeostasis by regulating self-renewal of stem cells as well as proliferation or differentiation of progenitor cells [40]. In diabetic DSCs, the transcription of several genes related to cell division and self-renewal, including NOTCH1, NOTCH2, WNT1, SOX2, and DHH was equivalent to that in normal DSCs. The gene expression of growth factors, including FGF1, FGF2, and FGF3, involved in cell proliferation and differentiation was equal in diabetic DSCs as compared with normal DSCs. Instead, the expression of BMP1, BMP3, BMP9 (alias GDF2) was upregulated in diabetic DSCs.
Notch1 and Notch2 coordinately maintain the stem cell pool in the quiescent state by preventing activation [41], whereas WNT proteins promote the proliferation of stem cells and regulate stem cell fate [42]. The WNT proteins, by promoting the ß-catenin translocation in nucleus, activate the transcription of target genes involving cell proliferation control, including FGF20, DKK1, WISP1, MYC, and CCND1 [43]. The expression of MYC and CCND1 genes in diabetic DSCs was like to that in normal DSCs. Moreover, in diabetic DSCs the transcription of FZD1 and FRAT1 genes (positive regulators of the canonical pathway of Wnt signaling), and of APC and AXIN1 genes (negative regulators) was equivalent to that in normal DSCs.
Asymmetric cell divisions can be controlled by Notch signaling through the activating ligands Jagged1 (JAG1) and Delta-like 1 (DDL1), and the negative regulator Deltex (DXT2) [40,44]. Compared to normal DSCs, the transcription of DDL1 gene was equal in diabetic DSCs, whereas JAG1 and DXT2 transcription was upregulated.
Cell division and proliferation are also dependent by cyclins (CNNs) and cyclin-dependent protein kinases (CDKs), which together control cell cycle progression. In diabetic DSCs, the transcription of CCNA2, CCND2, and CDK1 was downregulated compared with in normal DSCs. The cell cycle is a tightly regulated process that orchestrates genome duplication and accurate distribution of DNA and other factors into daughter cells after mitosis. CCNA2 binds to and activates its catalytic partners, CDK2 and CDK1. These complexes phosphorylate proteins like pocket proteins (RB, p107, p130) and proteins involved in DNA synthesis, thereby driving S-phase progression. In line with its role in regulating S phase, CCNA2 expression is induced upon entry into S phase, persists through the S and G2 phases, and is degraded upon entry into mitosis. During early mitosis, CCNA2 associates with CDK1 and drives chromosome condensation and nuclear envelope breakdown [45]. Instead, CCND2 regulating CDK4 and CDK6 drives the G1-to-S phase transition of the cell cycle [46]. Several works report that cyclins and CDKs are upregulated in different tumors such as breast, liver, and lung cancers, and the downregulation of them represents a good strategy to counteract the proliferation of the cancer stem cells [47,48]. Instead, the expression of cyclins and CDKs in stem cells had not yet been investigated.
The proliferation and differentiation of stem cells are also influenced by the activity of enzymes such us ALDH1A1 and ABCG2, which are downregulated in diabetic DSCs. Aldehyde dehydrogenase (ALDH) is considered a biomarker for stem cells, and its expression is also thought to closely correlate with the stemness of cancer stem cells [49]. ALDHs are family members of NAD-dependent enzymes that catalyze the oxidation of aldehydes to acids. They are localized in the cytoplasm, mitochondria, or nucleus and have been implicated in a wide variety of biological processes, including the detoxification of exogenously and endogenously generated aldehydes and the metabolism of vitamin A, alcohol, and reactive oxygen species. Of these, ALDH1A1 mainly catalyzes the conversion of retinaldehyde to retinoic acid (RA) in vitamin A metabolism. RA enters the nucleus, binds to and activates the RA receptors (RARs) or the retinoid X receptors (RXRs), which are nuclear transcription factors that promote target gene expression. The genes downstream of RA are involved in many important biological processes, including cell differentiation, proliferation, and lipid metabolism [50]. RXR functions as an obligate heterodimeric partner for multiple nuclear receptors including the peroxisome proliferator-activated receptor-gamma (PPARG). The nuclear receptor PPARG is a master regulator of adipogenesis that controls the expression of multiple genes within complex transcriptional networks [51]. In the absence of ALDH1A1, both adipogenesis in vitro and diet-induced fat formation in vivo are markedly impaired. ALDH1A1 deficiency increases retinaldehyde levels that inhibits RXR and PPARG activation, inhibiting PPARG-induced adipogenesis [52]. In diabetic DSCs, adipogenic differentiation is also prevented by the downregulation of the PPARG gene. Moreover, ALDH1A1 controls the cell proliferation by regulating notch signaling. Notch genes have been shown to directly activate cyclins and CDK2, which control the cell cycle progression. Li et al. demonstrated that the deletion of ALDH1A1 inhibits cell cycle progression and cell proliferation by the suppression of the Notch/CDK2/Cyclin pathway [53]. Moreover, ALDH1A1 isozyme has been shown to play an important functional role in maintaining cancer stem cells. In cancer stem cells, the overexpression of ALDH1A1 is associated with enhanced invasion, colony formation, and chemoresistance. It was demonstrated that the stable downregulation of ALDH1A1 isozyme alone dramatically decreased their ability to form colonies and to proliferate [54]. Therefore, the downregulation of ALDH1A1 observed in diabetic DSCs could be the cause of a reduced proliferative capacity due to a downregulation of cyclin and CDKs. In addition, the ABCG2 gene was downregulated in diabetic DSCs. The ABCG2 transporters belong to the ATP Binding Cassette (ABC) superfamily of transporters and use ATP hydrolysis to catalyze the transport of a wide range of substrates from the intracellular to the extracellular milieu. The overexpression of ABCG2 has been shown to promote stemness, while the loss of ABCG2 expression has been shown to promote lineage commitment in several adult stem cells. While the importance of ABCG2 in stem cell maintenance has been shown, the precise mechanism by which ABCG2 prevents stem cells from undergoing differentiation is unknown [55]. It seems that the high ABCG2 activity increases the levels of transcripts involved in stemness such as MEF, SOX2, OCT4, ID1, and HES1, and that the activity of ABCG2 is required for maintaining these stem markers and self-renewal property via Notch-independent manner [56]. Both enzymes have been broadly studied in cancer stem cells, however, data concerning their expression profile in diabetic stem cells are not previously reported in literature.
Overall, the resident stem cells of dermal tissue around DFUs showed morphology and phenotype equal to normal DSCs. However, diabetic DSCs showed reduced proliferative and migration capability. This could be related to the downregulation of the ALDH1A1 enzyme that inhibits cell cycle progression and cell proliferation. Only further investigations could explain the mechanism by which the inflammatory and ischemic condition, typical of DFU, may cause a downregulation of this enzyme.

Conclusions
In conclusion, the present study has demonstrated that in diabetic dermal tissue around DFUs there is an impairment in the gene expression profile of growth factors, metalloproteinases, collagens, and integrins involved in the wound healing process. Moreover, we have isolated and characterized the resident DSCs. Although diabetic DSCs have the same morphology and phenotype as normal DSCs, the gene expression has revealed a downregulation of cyclins, CDK1, ALD1H1A, and ABCG2. These proteins are related to cell cycle progression, cell proliferation, and stem cell maintenance. Further investigation will improve the understanding of the molecular mechanisms by which these proteins together govern cell proliferation and will reveal new strategies aimed at enhancing the expression or activity of these proteins that will be useful for future treatment of DFUs.