Antioxidant Preconditioning Improves the Paracrine Responsiveness of Mouse Bone Marrow Mesenchymal Stem Cells to Diabetic Wound Fluid

Abstract

Mesenchymal stem cells (MSCs) are a promising therapeutic tool for the treatment of nonhealing diabetic wounds. The pathological nature of the niche microenvironment limits the use of autologous cell therapy in diabetic patients. Prolonged exposure of endogenous MSCs to a pathological microenvironment in vivo reduces their ability to respond to environmental cues. This study investigated the effectiveness of ex vivo antioxidant treatment [N-acetylcysteine (7.5 mM NAC) and Ascorbic acid 2-phosphate (0.6 mM AAP)] to restore the paracrine function of diabetic MSCs. Healthy control [bone marrow stem cells derived from wild-type mice (SC^WT)] (source: wild-type C57BL/6J mice) (n = 12) and impaired/dysfunctional [bone marrow stem cells derived from ob/ob mice (SC^ob)] (source: obese diabetic, B6.Cg-Lep^ob/J mice) (n = 12) MSCs were isolated. Ex vivo treatment groups (SC^WT vs. SC^ob) were as follows: (1) no treatment (baseline phenotype), (2) stimulated with diabetic wound fluid (DWF) (baseline response), (3) antioxidant preconditioning (preconditioned phenotype), and (4) antioxidant preconditioned with subsequent stimulation with DWF (preconditioned response). The paracrine responsiveness on both the molecular (mRNA expression of 80 cytokines and receptors, quantitative polymerase chain reaction microarray) and protein (23-plex bead-array Luminex assay) level was assessed. At baseline, 31 genes were overexpressed (> × 2-fold) and 39 genes were underexpressed (> × 2-fold) in SC^ob versus SC^WT. In conditioned media, significant differences (P < 0.05) were detected at baseline for two proinflammatory cytokines [tumor necrosis factor alpha (TNFα) and interferon gamma (IFNγ)], four chemokines [keratinocyte chemoattractant (KC), granulocyte colony-stimulating factor (GCSF), Eotaxin, and macrophage chemoattractant protein (MCP1)], and one anti-inflammatory cytokine [interleukin 10 (IL10)]. Following stimulation with DWF, significant differences (P < 0.05) were detected in the secretion of two chemokines [granulocyte macrophage colony-stimulating factor (GMCSF) and Eotaxin], three proinflammatory cytokines (TNFα, IFNγ, and IL9), and four anti-inflammatory cytokines (IL10, IL4, IL13, and IL3). Antioxidant preconditioning significantly dampened the excessive TNFα response observed in SC^ob and improved the secretion of IL10. Taken together these data suggest that the combined ex vivo treatment of autologous stem cells with NAC and AAP could potentially be an effective strategy to restore the paracrine function of impaired diabetic MSCs before transplantation.

Introduction

Obesity and type 2 diabetes mellitus (DM) are regarded as chronic low-level inflammatory states that, in combination with hyperglycemic conditions, cause severe co-morbidities such as nonhealing wounds. Lower extremity wounds [such as diabetic foot ulcers (DFU)], which affect the musculoskeletal system and skin, occur in 4%–10% of diabetic patients and have a 20%–80% recurrence rate [1 –4]; DFUs are currently the leading cause for lower limb amputations globally.

Mesenchymal stem cell (MSC) therapies aimed at improving wound healing are widely pursued in preclinical trials. This is due to the immune-modulatory, growth-promoting, and anti-inflammatory properties that MSCs possess [3,5 –7]. However, in this respect, the use of autologous stem cells from patients with DM are not as effective in promoting healing as those from nondiabetic controls [8]. Based on the findings from various in vitro studies, it is now known that DM alters the characteristics of endogenous MSCs to such an extent that the changes remain evident, despite removal of these cells to a more optimal in vitro environment [8,9]. Bone marrow stem cells derived from a DM microenvironment have impaired viability, have refractory mobility, and are unable to close an in vitro wound effectively [9 –12]. The mechanism underlying these impairments is, however, not yet clear. Furthermore, it is unknown if these impairments can be reversed using in vitro treatments, and if so, whether preconditioned MSCs will retain a repaired status when reintroduced to the pathological microenvironment in vivo.

Wound healing is a complex process that consists of four main phases: hemostasis, inflammation (destructive/phagocytic), proliferation (extracellular matrix formation, angiogenesis, and reepithelialization), and remodeling (matrix reorganization) [13,14]. The inflammatory phase is essential for the phagocytosis of bacteria, dead cells, and tissue debris. This phase is characterized by immune cell (neutrophil/M1 macrophage) infiltration and the presence of proinflammatory signaling molecules [15]. If the transition from M1 (proinflammatory) to M2 (proregenerative) macrophages does not occur timeously, healing is impaired, and the wound may become chronic. This is the case with diabetic ulcers, where a combination of factors, including an excessive and delayed inflammatory response, dysfunctional supporting cells (fibroblasts, keratinocytes, and epithelial cells), ischemia, impaired angiogenesis, and neuropathy, prevents the wound from transitioning into the proliferative and remodeling phases [5,15,16]. The paracrine properties of exogenously applied allogeneic (healthy) bone marrow stem cells have been shown to restore the wound microenvironment by inducing the proliferative and remodeling phases of healing through anti-inflammatory and proregenerative signaling [7]. Bone marrow stem cells isolated from pathological environments (such as those existing during DM) lose their ability to promote repair. We hypothesize that prolonged exposure of autologous MSCs to a pathological microenvironment in vivo reduces the ability of these cells to respond to environmental cues and thereby effect repair. Assessing whether the paracrine (cytokine) responsiveness of MSCs is skewed toward a prodestructive or proregenerative response could potentially be used as diagnostic parameters to predict the degree of MSC dysfunction in a diabetic patient.

The pathogenesis of DM, which contributes to the functional decline of stem cells, is driven by hyperglycemia (advanced glycation end products), chronic inflammation, and oxidative stress [8]. Ascorbic acid 2-phosphate (AAP) and N-acetylcysteine (NAC) are antioxidants that have been shown to preserve the “stemness” of healthy MSCs during prolonged culture, by preventing apoptosis and promoting proliferation, telomerase activity, and increasing intracellular levels of antioxidant enzymes such as superoxide dismutase [17 –20]. It is hypothesized that these protective agents could potentially be used to restore the function of impaired autologous stem cells.

This study therefore investigated whether in vitro antioxidant preconditioning could restore the paracrine function of impaired diabetic bone marrow stem cells. To achieve this, an in vivo murine model of impaired wound healing was established, and diabetic wound fluid collected. The paracrine (secretome; extracellular) and molecular (intracellular) responsiveness of healthy and impaired/dysfunctional stem cells were then assessed at baseline and following ex vivo exposure to diabetic wound fluid. The successful restoration of the paracrine function of impaired/diabetic stem cells ex vivo before use for autologous cell therapy could improve the predictability of cell therapy in diabetic ulcer patients in the clinical setting.

Materials and Methods

This study was approved by the animal research ethics committee (SU-ACUD17-000016) at Stellenbosch University and complied with the South African Animal Protection Act (Act no. 71, 1962). All experimental procedures were conducted according to the ethical guidelines and principles of the declaration of Helsinki.

Induction of full-thickness wounds and collection/analysis of wound fluid

Full-thickness excisional wounds were induced dorsally on obese prediabetic mice (B6.Cg-Lep^ob/J) (ob/ob) with subcutaneous injections of recombinant neutral endopeptidase (NEP) to (1) prevent wound contraction and (2) more closely mimic diabetic ulcers in patients [21,22].

Seven (n = 7) ob/ob mice (8-week old; weight 43 ± 2 g) were anesthetized using isoflurane (Safeline Pharmaceuticals Pty, Ltd., SA) gas. Dorsal hair was shaved, washed with povidone-iodine and a local anesthetic, and 2% lignocaine (7 mg/kg; Bodene Pty, Ltd., SA) applied before making two identical contralateral full-thickness skin excisions (including the underlying panniculus carnosus) (6 mm²) (Fig. 1). Immediately postwounding and on day 7, NEP (0.3 mg/mL; enzyme activity 50 pmol/h/μg) (SRP6450, Neprilysin/CD10; Sigma-Aldrich, St. Louis, MO) was injected subcutaneously at the wound edges to prevent skin contraction and further delay healing. The enzyme activity of the injected NEP is equivalent to that observed in the foot ulcers of diabetic patients [21,22]. Postsurgery pain (3 days) was managed by adding acetaminophen (300 mg/kg) (GSK, Cape Town, SA) to the drinking water of individually housed animals. Both excisional wounds were covered by a vapor-permeable polyurethane film (Hydrofilm; Paul Hartmann AG, Heidenheim, Germany) and the accumulated exudate (diabetic wound fluid, DWF) harvested by needle puncture [23,24] at 7, 14, and 21 days postwounding. At these time points, DWF was collected by injecting 100 μL sterile saline onto the wounds and then collecting the wound exudate together with the saline. The collected DWF was pooled, sterile filtered, and centrifuged to eliminate all debris and stored at −80°C. Note: on day 14 postwounding, the start of reepithelialization was evident, and wounds were therefore reopened (by removing the epithelial layer) at this time point, following the collection of the DWF.

FIG. 1.

Full-thickness excisional diabetic wound model. (A) The weight of the B6.Cg-Lep^ob/J (ob/ob) mice over time after initial wounding (day 0) and rewounding (day 14). (B) Representative images of contralateral dorsal full-thickness excisional wounds that were induced on ob/ob mice. Images were taken immediately (day 0) and on days 7, 14, and 21 postwounding. (C) Percentage wound closure (%). Wound surface area at each time point was determined using ImageJ software (version 1.46, nih.gov) and the percentage wound closure calculated using the following formula: (wound area d0 − wound area dn)/(wound area d0) × 100. Statistical analysis: data are presented as (mean ± SE). Repeated-measures ANOVA with Bonferroni post hoc test. **P < 0.01. Color images available online at www.liebertpub.com/scd

The total protein concentration of the sterile filtered DWF (9.6 μg/mL) was determined to be threefold less than the total amount of protein in fetal bovine serum (FBS) (27.9 μg/mL) using a standard Bradford assay with BSA (Hyclone BSA; Thermo, Logan) as control (Supplementary Fig. S1; Supplementary Data are available online at www.liebertpub.com/scd). For the in vitro experiments, bone marrow-derived stem cells (SCs) were therefore stimulated with media containing 30% DWF [600 μL DWF + 1,400 μL Dulbecco's modified Eagle's medium (DMEM) with ultra-glutamine containing 1% penicillin/streptomycin], while control growth media contained the standard 10% FBS. The total protein content in the media used to stimulate SCs with DWF was thus equivalent to that of standard growth media. The cytokine concentrations within DWF were determined using a Bio-Plex Pro Mouse Cytokine 23-plex assay (No. M60009RDPD; Bio-Rad, Hercules, CA) [Eotaxin, granulocyte colony-stimulating factor (GCSF), granulocyte macrophage colony-stimulating factor (GMCSF), interferon gamma (IFNγ), interleukin (IL)1α, IL1β, IL2, IL3, IL4, IL5, IL6, IL9, IL10, IL12 (p40), IL12 (p70), IL13, and IL17A, keratinocyte chemoattractant (KC), macrophage chemoattractant protein (MCP1) (MCAF), macrophage inflammatory protein (MIP)1α, MIP1β, RANTES, and tumor necrosis factor alpha (TNFα)] as per manufacturer's instructions.

As a parallel control, to assess whether wound closure was truly delayed in our diabetic wound model, acute wounds (without NEP injection) were induced on six (n = 6) healthy control mice (C57BL6/J) (8-week old; weight 25 ± 0.2 g) using the same procedure as described above. Complete macroscopic wound closure and significant healing were evident in the control animals on day 7 (Supplementary Fig. S2).

Determination of the safest, nontoxic doses of NAC and AAP for in vitro antioxidant preconditioning

The dose-response experiment was performed using a mouse MSC line (C3H/10T1/2, clone 8, ATCC ccl-226) (passage 5–7) as part of the refinement criteria of this study to limit the required number of animals. MSCs were cultured in standard growth media (SGM) consisting of DMEM with ultra-glutamine (4.5 g/L high-glucose DMEM; BioWittaker, Lonza, Basel, Switzerland) containing 10% FBS (Biochrom, Berlin, Germany) and 1% penicillin/streptomycin (BioWittaker, Lonza). Cells were maintained at 37°C, in 90% humidified air with 5% CO₂. The effect of different doses of NAC (7.5, 15, and 30 mM) (Sigma-Aldrich, St. Louis, MO) or AAP (0.15, 0.3, 0.6 mM) (Sigma-Aldrich, St. Louis, MO) on MSC proliferation (BrdU ELISA; Roche, Basel, Switzerland) was assessed over a period of 24 and 48 h, respectively. This was done to determine the optimum dose of each individual antioxidant. To assess the synergistic effect of the optimum doses, the survival rate of MSCs following exposure to the combined antioxidant treatment was determined over a period of 6 days using Crystal Violet staining (0.01% in dH₂O) (Sigma-Aldrich, Steinheim, Germany) [25].

Isolation and characterization of bone marrow-derived SCs

Healthy control (source: wild type, C57BL/6J mice, weight 23.6 ± 0.9 g) (n = 12) and impaired/dysfunctional bone marrow-derived SCs (source: obese diabetic, B6.Cg-Lep^ob/J mice, weight 37.2 ± 1.7 g) (n = 12) were isolated as previously described [9]. Briefly, following sacrifice of the animals, the femurs were dissected out and the proximal and distal ends of each femur cut open in a sterile environment. Bone marrow aspirates were flushed into tissue culture plates (six-well plates; Nest Biotechnology, NJ) using a 25-gauge needle and syringe containing growth media with 20% FBS (Biochrom). Growth media consisted of DMEM with ultra-glutamine (4.5 g/L high glucose; BioWittaker, Lonza), containing 1% penicillin/streptomycin (BioWittaker, Lonza). Bone marrow aspirates were maintained at 37°C, in 90% humidified air with 5% CO₂. After 96 h, nonadherent cells were washed off with phosphate-buffered saline (PBS) and the media replaced with SGM containing 10% FBS (Biochrom). Media were then changed every 3–4 days until the cells reached confluence in passage 0. All subsequent experiments were performed on SCs derived from at least three animals per group (n = 3 biological repeats).

SCs (90% confluent; passage 0) were harvested through trypsinization and resuspended in flow cytometry staining solution (PBS containing 20% FBS). Cell suspensions at a concentration of 1 × 10⁶ cells per 100 μL were co-labeled with rat anti-mouse monoclonal antibodies against the stem cell marker, Ly-6A/E (Sca-1) (PE conjugated; BD Pharmingen, San Diego, CA) and the hematopoietic lineage marker, CD45 (FITC conjugated; BD Pharmingen). Flow cytometry was performed on a BD FACSCalibur instrument using CellQuest software. A total of 15,000 events were recorded before data analysis. Fluorescent compensation settings were established through a compensation experiment using comp beads (BD™CompBead Plus Anti-Mouse Ig, k; BD Biosciences, San Jose, CA). An unstained control sample was used as a negative control for gating purposes. Data analysis was performed using Flow Jo Vx (Treestar, OR) software.

For all subsequent experiments, the isolated bone marrow SCs were expanded in culture under either standard growth media (control) conditions or in the presence of antioxidant preconditioning before stimulation with DWF.

In vitro treatment conditions and study design

Antioxidant preconditioning consisted of combined treatment with NAC (7.5 mM) (Sigma-Aldrich, St. Louis, MO) and AAP (0.6 mM) (Sigma-Aldrich, St. Louis, MO) added to the growth media (SGM), with media being changed every 3–4 days until the cells reached 80% confluence. The inflammatory phenotype and paracrine responsiveness of healthy versus impaired SCs (with and without preconditioning) to environmental cues were assessed following ex vivo exposure to DWF. For the treatment of SCs with DWF, the FBS (10%) in the growth media was replaced by DWF (30%) with an equivalent total protein concentration. The treatment groups were as follows: (1) no treatment (baseline phenotype): SCs expanded in standard growth media (±8 days) and only exposed to growth media, (2) DWF: SCs expanded in standard growth media (±8 days) followed by acute exposure to DWF for a period of 4 h, (3) antioxidant preconditioning: SCs expanded in the presence of NAC/AAP (±8 days), And (4) antioxidant preconditioning+DWF: SCs expanded in the presence of NAC/AAP (±8 days) followed by acute exposure to DWF for a period of 4 h. Refer to Supplementary Fig. S3 for an overview of the study design and treatment groups.

Collection and analysis of conditioned media

After MSCs reached 80% confluence, the growth media (with or without NAC/AAP) were replaced with serum-free media for a period of 24 h. The complete serum-free growth media consists of Stem Pro MSC SFM basal media (No. A10334-01; Gibco, Life Technologies), Stem Pro MSC SFM growth factor Supplement CTS (A10333-01; Gibco, Life Technologies), 2 mM l-Glutamine (No. 25030-081; Gibco, Life Technologies), and 5 μg/mL penicillin/streptomycin (No. DE17-6026; BioWhittaker, Lonza). After 24 h, the conditioned media were collected, centrifuged to remove cellular debris, and stored at −80°C. The cytokine concentrations within conditioned media were assessed using the Bio-Plex Pro Mouse Cytokine 23-plex Assay (No. M60009RDPD; Bio-Rad) [Eotaxin, GCSF, GMCSF, IFNγ, IL1α, IL1β, IL2, IL3, IL4, IL5, IL6, IL9, IL10, IL12 (p40), IL12 (p70), IL13, and IL17A, KC, MCP1 (MCAF), MIP1α, MIP1β, RANTES, and TNFα] according to the manufacturer's instruction, using a Multiplex reader (MAGPIX; Bio-Rad) and Bio-Plex Manager™ MP software.

RNA isolation, cDNA synthesis, and RT2 quantitative polymerase chain reaction microarray

Total RNA was isolated using the RNeasy RNA isolation kit (Qiagen, Berlin, Germany) with on-column DNAse digestion (Qiagen). RNA quantification and quality control were performed using the Eukaryote Total RNA Nano Series II (2100 Bioanalyzer; Agilent Technologies, Inc., Santa Clara, CA). Total RNA (0.5 μg) with an integrity of above 9 (RIN >9) was used as a template for cDNA synthesis, using the RT² First Strand Kit (Qiagen) that includes a genomic DNA elimination step. cDNA samples for each treatment condition were analyzed with an RT² Profiler polymerase chain reaction (PCR) inflammatory cytokine and receptor array (96-well plate format; Qiagen) and RT² SYBR Green Master-mix (Rox PCR Master mix; Qiagen) using an ABI 7900HT Fast real-time PCR system (Applied Biosystems, Life technologies), and SDS software (version 2.3; Life technologies). The web-based PCR Profiler Array data analysis package was used for ΔΔCt-based fold change calculations (Qiagen, www.qiagen.com). Relative gene expression was calculated by comparison with the following housekeeping gene: heat shock protein 90α class B member 1 (Hsp90ab1) (NM_008302). The PCR array analysis furthermore included a mouse genomic DNA contamination (SA_00106), reverse transcriptase (SA_00104), and positive PCR (SA_00103) controls. Genes of interest were identified using Student's t-test of the replicate ΔΔCt values for each gene in the control and treatment groups. The level of significance was accepted at P < 0.05.

Statistical analysis

Values are presented as mean ± standard error (mean ± SE). Statistical analysis was performed using Statistica software (Statsoft, version 13). Repeated-measures ANOVA with Bonferroni post hoc test was used to assess animal weight and wound closure over time. Factorial ANOVA with Tukey post hoc test was used to assess group, treatment, and group × treatment effects. Genes of interest were identified using a Student's t-test of the replicate 2^(ΔCt) values of each gene in the control and treatment groups. The level of significance was accepted at P < 0.05.

Results

Full-thickness excisional diabetic wound model: collected wound fluid is representative of the in vivo microenvironment of an inflammatory diabetic wound

No weight loss was apparent in the wounded animals (start weight: 43 ± 2 g; end weight: 41 ± 7 g) for the duration of the study (Fig. 1A). Impaired wound closure (macroscopic) was evident in the ob/ob animals over the initial recovery period of 14 days (day 0 to 7: 33% ± 7% wound closure; day 0 to 14: 77% ± 4% wound closure) as well as in the subsequent period after rewounding (day 14 to 21: 48% ± 7% wound closure) (Fig. 1B, C). In contrast, significant wound closure was evident after only 7 days in the WT control animals (Supplementary Fig. S1). Wound fluid was collected and analyzed as indication of the clinical condition of the wound. The DWF collected from the ob/ob animals had a cytokine profile that mimics the clinical characteristics of a nonhealing diabetic wound. Refer to Table 1 for the quantification of cytokine concentrations detected in the DWF. Proinflammatory cytokines (IL1α, IL1β, IL6, IL12p40, IL12p70, IL17A, RANTES, and TNFα) and chemokines (Eotaxin, GCSF, GMCSF, MCP1, MIP1α, and MIP1β) were present in high concentrations, whereas anti-inflammatory cytokines (IL2, IL3, IL4, IL5, IL9, and IL10) had much lower levels within the DWF.

Table 1.

Cytokine Secretion Profile in Response to Diabetic Wound Fluid

	Secreted protein: response to DWF
		Control		Preconditioned
	DWF	Sc^WT	Sc^ob	Sc^ob
Proinflammatory
IL1α	6,992	2,737 (31)	2,640 (560)	2,974 (165)
IL1β	3,360	1,144 (22)	1,184 (125)	1,131 (133)
IL6	1,004	11,305 (2,119)	7,290 (3,240)	10,574 (3,827)
IL9	127	204 (9)^a	228 (4)^b	242 (5)^a,b
IL12p40	4,962	2,944 (71)	3,051 (174)	2,934 (151)
IL12p70	598	1,030 (71)	1,068 (32)	1,064 (118)
IL17A	361	207 (6)	238 (19)	241 (20)
IFNγ	43	57 (3)^a	68 (2)	77 (6)^a
RANTES	128	3,971 (217)	5,376 (687)	12,833 (6,862)
TNFα	932	5,081 (671)^c	20,441 (1,461)^b,c	9,190 (487)^b
Chemokines
Eotaxin	572	179 (6)^a,c	217 (6)^c	229 (9)^a
GCSF	>47,590^d	14,160 (605)	14,942 (781)	15,484 (1,564)
GMCSF	672	399 (10)^a	486 (26)	510 (30)^a
Kc	>21,315^d	11,231 (3,397)	11,165 (1,767)	18,305 (5,378)
MCP1	18,568	35,772 (678)	34,706 (10,271)	35,432 (7,406)
MIP1α	>7,288^d	2,689 (257)	3,095 (493)	6,357 (2,603)
MIP1β	1,187	18,344 (1,776)	26,690 (7,447)	31,905 (12,064)
Anti-inflammatory
IL2	23	49 (1)	60 (1)	120 (19)
IL3	27	33 (1)^a,c *	38 (1)^{c *}	39 (2)^a
IL4	91	53 (1)^c	65 (3)^c	60 (3)
IL5	27	53 (3)	66 (4)	95 (23)
IL10	263	6,823 (575)^a	5,415 (317)^b	10,207 (812)^a,b
IL13	329	617 (14)^a,c	745 (7)^c	747 (29)^a

Data are presented as mean (SE) in pg/mL. Statistical analysis: factorial ANOVA with Tukey post hoc test. Level of significance accepted at P < 0.05.

Significant difference between SC^WT and SC^ob with preconditioning.

Significant difference in SC^ob with and without preconditioning.

Significant difference between SC^WT and SC^ob. ^c ^*Indicates P = 0.06.

The concentration of these chemokines was above the detection limit as indicated.

GCSF, granulocyte colony-stimulating factor; GMCSF, granulocyte macrophage colony-stimulating factor; IFNγ, interferon gamma; IL, interleukin; KC, keratinocyte chemoattractant; MCP, macrophage chemoattractant protein; MIP, macrophage inflammatory protein; SC^ob, bone marrow stem cells derived from ob/ob mice; SC^WT, bone marrow stem cells derived from wild-type mice; TNFα, tumor necrosis factor alpha.

The safest nontoxic doses of NAC and AAP for in vitro antioxidant preconditioning

Compared to standard growth conditions, of all the concentrations tested (NAC 7.5, 15, and 30 mM; AAP 0.15, 0.3, and 0.6 mM), 7.5 mM NAC had the most significant effect on the proliferation rate of MSCs over a period of 24 h (∼2-fold increase) (P < 0.05) and 48 h (∼2.5-fold increase) (P < 0.05) (Fig. 2A, B). Similarly, 0.6 mM AAP was shown to promote MSC proliferation over a period of 24 h (∼1.3-fold increase) compared to standard growth conditions (Fig. 2A, B). The combined treatments of 7.5 mM NAC and 0.6 mM AAP maintained cell viability over a period of 6 days (Fig. 2C) and were thus identified as the optimum concentrations for preconditioning.

FIG. 2.

Safest, nontoxic dosages of NAC and AAP. (A, B) The independent effect of NAC (7.5, 15, and 30 mM) and AAP (0.15, 0.3, and 0.6 mM) on MSC proliferation (BrdU incorporation) over a period of 24 h (A) and 48 h (B). (C) The synergistic effect of combined NAC and AAP treatment on cell viability (crystal violet staining) over a period of 6 days. Statistical analysis: data (mean ± SE) are presented as percentage with SGM (control) set at 100. Factorial ANOVA with Tukey post hoc test. *P < 0.05. AAP, ascorbic acid 2-phosphate, CV, crystal violet; NAC, N-acetylcysteine; SGM, standard growth media.

Characterization of healthy control (source: wild type, C57BL/6J mice) and impaired/dysfunctional (source: obese diabetic, B6.Cg-Lep^ob/J mice) bone marrow-derived SCs

Flow cytometry surface marker characterization demonstrated that 88% ± 4% of isolated bone marrow-derived cells expressed the stem cell marker, Sca1 (Fig. 3A–C). A percentage of cells did, however, also express the hematopoietic lineage marker, CD45. Immediately, following isolation (passage 0), the bone marrow-derived SCs therefore consisted of a heterogenous population of 60% ± 3% MSCs (Sca1⁺CD45⁻) and 39% ± 3% hematopoietic stem cells (Sca1⁺CD45⁺) (Fig. 3A–D). There was no difference in surface marker expression in the bone marrow SCs derived from either WT control (SC^WT) or ob/ob (SC^ob) animals (Fig. 3D). SC^ob had impaired viability and growth with fewer cells visible 96 h postisolation compared to their SC^WT counterparts at the same time point (Fig. 3E, F).

FIG. 3.

Characterization of healthy control (source: wild type, C57BL/6J mice) and impaired/dysfunctional (source: obese diabetic, B6.Cg-Lep^ob/J mice) bone marrow-derived SCs. (A) Representative flow cytometry dot plot demonstrating the forward (size) versus side (granularity) scatter of the isolated cells. (B, C) The expression of Sca1 (PE) versus CD45 (FITC) in unstained negative control sample (B) and bone marrow SCs (C). Q1—MSCs expressing Sca1⁺CD45⁻; Q2—HSCs expressing Sca1⁺CD45⁺; Q3—Mature hematopoietic cells Sca1⁻CD45⁺; Q4—Cells expressing neither Sca1⁻ nor CD45⁻. (D) Quantification of the percentage MSCs and HSCs within the isolated population. (E, F) Representative images of the healthy control (E) and impaired/dysfunctional (F) bone marrow SCs with and without antioxidant preconditioning at 96 h and day 16 postisolation. Statistical analysis: data are presented as (mean ± SE). Factorial ANOVA with Tukey post hoc test. HSC, hematopoietic stem cells; MSCs, mesenchymal stem cells; Precond, antioxidant preconditioning. Color images available online at www.liebertpub.com/scd

Secretome differences at baseline: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)

The secretion of 23 cytokines from SC^WT and SC^ob was assessed at baseline with and without antioxidant preconditioning (Table 2). Of the 23 cytokines assessed within conditioned media, significant differences (P < 0.05) were detected at baseline (without stimulation) for two proinflammatory cytokines (TNFα and IFNγ), four chemokines (KC, GCSF, Eotaxin, and MCP1), and one anti-inflammatory cytokine (IL10).

Table 2.

Cytokine Secretome Differences at Baseline

	Secreted protein: conditioned media
	Control		Preconditioned
	SC^WT	SC^ob	SC^WT	SC^ob
Proinflammatory
IL1α	—	4.0 (2.1)	—	—
IL1β	24 (8)	34 (7)	21 (11)	25 (5)
IL6	1,007 (254)	1,254 (514)	942 (316)	1,176 (263)
IL9	—	13 (12)	—	—
IL12p40	75 (22)	109 (11)	99 (10)	154 (31)
IL12p70	130 (70)	325 (65)	130 (70)	186 (61)
IL17a	—	—	—	—
IFNγ	—^a	15 (4)^a,b	—	—^b
RANTES	127 (13)	226 (48)	122 (17)	229 (25)
TNFα	40 (14)^a	113 (21)^a,b	34 (19)	51 (14)^b
Chemokines
Eotaxin	—^a	20 (7)^a	—^c	9.1 (4.5)^c
GCSF	10 (8)^a	67 (12)^a	10 (7.5)^c	30 (15)^c
GMCSF	85 (32)	137 (22)	77 (40)	84 (17)
KC	353 (50)^a	778 (151)^a	356 (48)^c	578 (71)^c
MCP1	1,721 (113)	5,171 (835)^b	2,571 (901)^c	12,912 (1 772)^b,c
MIP1α	11 (2)	8.9 (0.4)	15 (4)	13 (0.4)
MIP1β	50 (2)	65 (4)	88 (36)	129 (11)
Anti-inflammatory
IL2	2.7 (2.6)	7.9 (2.7)	2.7 (2.6)	6.3 (3.5)
IL3	2.9 (1.7)	9.5 (1.6)	2.9 (1.7)	4.7 (1.8)
IL4	—	1.3 (1.2)	—	—
IL5	13 (7)	16 (2)	15 (6)	38 (8)
IL10	36 (7)^{a *}	87 (8)^{a *}	32 (11)	74 (19)
IL13	138 (25)	207 (37)	107 (55)	135 (23)

Data are presented as mean (SE) in pg/mL. Statistical analysis: factorial ANOVA with Tukey post hoc test. Level of significance accepted at P < 0.05.

Significant difference at baseline (control) between SC^WT and SC^ob. ^a ^*Indicates P = 0.06.

Significant difference in SC^ob with and without preconditioning.

Significant difference between SC^WT and SC^ob with preconditioning.

—, not detected.

Proinflammatory TNFα and IFNγ concentrations were significantly higher (P < 0.05) in the conditioned media derived from SC^ob (TNFα 2.9-fold higher; IFNγ 15 ± 4 pg/mL) compared to SC^WT (IFNγ not detected). Preconditioning SC^ob with antioxidants did, however, downregulate the secretion of TNFα and IFNγ to levels comparable to that of SC^WT. The chemokines KC, GCSF, and Eotaxin remained significantly higher (P < 0.05) in the conditioned media derived from SC^ob (KC 2.2-fold higher; GCSF 6.7-fold higher; Eotaxin 20 ± 7 pg/mL) compared to SC^WT (Eotaxin not detected), regardless of preconditioning. Preconditioning did, however, promote MCP1 secretion in SC^ob (2.5-fold increase) (P < 0.05), whereas only a small increase (1.5-fold, not significant) was evident in SC^WT. The anti-inflammatory cytokine, IL10, tended to be higher (2.4-fold) (P = 0.06) in the conditioned media derived from SC^ob compared to SC^WT and was unaffected by preconditioning.

Gene expression (mRNA) differences at baseline and in response to stimulation with DWF: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)

At baseline, under standard growth conditions, 31 genes were overexpressed (>2-fold) and 39 genes underexpressed (>2-fold) in the SC^ob versus SC^WT (Fig. 4A). The overexpression of colony-stimulating factor (Csf1) (also known as MCSF) (29-fold) was, however, the only significant difference (P < 0.05) between the groups at baseline (Fig. 4A, C). Overall, preconditioning of SC^ob downregulated the expression of 32 genes and upregulated the expression of 48 genes (Fig. 4B). The expression of Csf1 in SC^ob was downregulated with preconditioning to levels comparable to that observed in SC^WT (Fig. 4C). Preconditioning induced the expression of Ccl8 (also known as MCP2) (P < 0.05), which was previously undetectable in SC^ob (Fig. 4B, D).

FIG. 4.

Gene expression differences at baseline and in response to stimulation with DWF: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob). (A, B) A comparison of mRNA expression at baseline in SC^ob versus SC^WT (A) and in SC^ob with and without antioxidant preconditioning (B). (C, D) Quantification of Csf1 (C) and ccl8 (D) mRNA ( × fold difference compared to control) at baseline within the respective groups. (E) The gene expression responsiveness of SC^WT to stimulation with DWF. (F) The gene expression responsiveness of SC^ob to stimulation with DWF. (G) The gene expression response of preconditioned SC^ob following stimulation with DWF. (H) Summary of the responsiveness of SC^WT, SC^ob, and preconditioned SC^ob to DWF (only genes that had a significant change in mRNA expression is indicated). Statistical analysis: genes of interest were identified using a Student's t-test of the replicate (n = 3) 2^(ΔCt) values of each gene in the control and experimental groups. *P < 0.05. Csf, colony-stimulating factor; SC^ob, bone marrow stem cells derived from ob/ob mice; SC^WT, bone marrow stem cells derived from wild-type mice. Color images available online at www.liebertpub.com/scd

When exposed to DWF, both the SC^WT and SC^ob responded by significantly (P < 0.05) increasing the gene expression of macrophage chemoattractant proteins, Ccl2 (also known as MCP1) and Ccl3 (also known as MIP1α) (Fig. 4E–H). In SC^WT, DWF furthermore induced the expression (P < 0.05) of nine additional genes (Ccl24, Ccl7, Ccr1, Cxcl10, Cxcl5, Cxcl9, IL15, IL1α, and TNF) associated with proinflammatory chemotactic responses. In contrast, SC^ob responded by increasing (P < 0.05) the expression of only four additional genes (Cxcl1, IL11rn, IL27, and Cxcl11) (Fig. 4F, H). Preconditioning did, however, restore the TNF response (P < 0.05) and induced the expression of Ccl17, Ccl19, Csf3, macrophage inhibitory factor (Mif), and B2m (Fig. 4G, H) in SC^ob. A summary of the relevant molecular responses (P < 0.05) within these groups is indicated in Fig. 4H.

Differences in the paracrine responsiveness following DWF stimulation: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)

Following stimulation with DWF, significant differences (P < 0.05) were detected in the conditioned media concentrations of two chemokines [GMCSF and Eotaxin (also known as ccl11)], three proinflammatory cytokines (TNFα, IFNγ, and IL9), and four anti-inflammatory cytokines (IL10, IL4, IL13, and IL3) between groups (SC^WT vs. SC^ob vs. preconditioned SC^ob) (Fig. 5 and Table 1).

FIG. 5.

Differences in the paracrine responsiveness following DWF stimulation: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob). The cytokine concentrations in conditioned media poststimulation with DWF were assessed in SC^WT, SC^ob, and preconditioned SC^ob. (A) GMCSF (pg/mL), (B) Eotaxin (pg/mL), (C) TNFα (pg/mL), (D) IFNγ (pg/mL), (E) IL9 (pg/mL), (F) IL10 (pg/mL), (G) IL4 (pg/mL) (H) IL13 (pg/mL), and (I) IL3 (pg/mL). Statistical analysis: factorial ANOVA with Tukey post hoc test. *P < 0.05, **P < 0.01. DWF, wound fluid; GMCSF, granulocyte macrophage colony-stimulating factor; IFNγ, interferon gamma; IL, interleukin; TNFα, tumor necrosis factor alpha.

Similar to baseline, the chemokines GMCSF and Eotaxin in conditioned media were detected at higher concentrations in SC^ob (GMCSF 486 ± 26 pg/mL and Eotaxin 217 ± 6 pg/mL) and preconditioned groups (GMCSF 510 ± 30 pg/mL and Eotaxin 229 ± 9 pg/mL), following stimulation compared to SC^WT (GMCSF 399 ± 10 pg/mL and Eotaxin 179 ± 6 pg/mL) (Fig. 5A, B). Despite these differences, the concentration of these chemokines poststimulation was lower than that detected in DWF (Table 1).

Stimulation with DWF induced a much greater proinflammatory TNFα response in SC^ob (20,441 ± 1,461 pg/mL) than SC^WT (5,081 ± 671 pg/mL); this excessive response in SC^ob was dampened with preconditioning (9,190 ± 487 pg/mL) (Fig. 5C). Preconditioning furthermore increased the release of proinflammatory IFNγ and IL9 following stimulation with DWF in the SC^ob (IFNγ 77 ± 6 pg/mL and IL9 242 ± 5 pg/mL) compared to SC^WT (IFNγ 57 ± 3 pg/mL and IL9 204 ± 9 pg/mL) (Fig. 5D, E).

A slight difference in the anti-inflammatory IL3, IL4, and IL13 response to DWF was evident between SC^ob (IL3 38 ± 1 pg/mL; IL4 65 ± 3 pg/mL; and IL13 745 ± 7 pg/mL) and SC^WT (IL3 33 ± 1 pg/mL; IL4 53 ± 1 pg/mL; and IL13 617 ± 14 pg/mL) (Fig. 5F–I). IL4 in conditioned media poststimulation was, however, below the concentration detected in DWF (Table 1). Preconditioning significantly improved the secretion of anti-inflammatory IL10 poststimulation in the SC^ob group (SC^ob 5,415 ± 317 pg/mL; Precond 10,207 ± 812 pg/mL; SC^WT 6,823 ± 575 pg/mL) (Fig. 5F).

Discussion

Tissue repair and regeneration using endogenous stem cells represents the ultimate goal in regenerative medicine; however, the pathological nature of the in vivo microenvironment has hampered the use of autologous cell therapy in diabetic patients. This study demonstrated that bone marrow stem cells derived from an obese diabetic microenvironment have a skewed inflammatory phenotype; this may contribute to their inability to perform the imperative function of restoring the localized (wounded area) conditions needed for normal healing. Wound fluid is generated by enhanced capillary leakage and local activity of a variety of resident and migratory cell types that react to a breach in the skin barrier. The composition of wound fluid is therefore broadly assumed to reflect the clinical condition of a wound [26,27]. In the context of this ex vivo study, wound fluid therefore represented the extracellular fluid space within the microenvironment of a diabetic wound. Exposing healthy and impaired/dysfunctional bone marrow stem cells to DWF in vitro enabled an investigation into the paracrine responsiveness of these cells on both molecular and protein level. This study was furthermore the first to demonstrate the efficacy of antioxidant preconditioning in restoring the paracrine function of impaired/dysfunctional stem cells toward that seen under control nonpathological conditions.

Although the ob/ob animal model has been used in numerous studies investigating diabetes-related impaired wound healing [28 –31], the endogenous stem cells derived from these animals have not yet been characterized. At baseline, the secretion of proinflammatory cytokines (TNFα, IFNγ) and chemokines (KC, GCSF, and Eotaxin) was more pronounced in the impaired/dysfunctional bone marrow stem cells. These factors all function to amplify the first stage of a wound (inflammatory response), by recruiting neutrophils, lymphocytes, eosinophils, and keratinocytes to the wounded area [5,13,32]. At the mRNA level, a similar pattern was observed; however, only the overexpression of Csf1 was significant between healthy versus dysfunctional stem cells and corresponded with higher chemokine (GMCSF) levels detected in the conditioned media. These and other chemokines were already present within the wound microenvironment (DWF) at very high concentrations and the additional release thereof is thus unwanted. Antioxidant preconditioning was able to reduce the secretion of proinflammatory TNFα and IFNγ at baseline, but did not reduce chemokine secretion.

In the context of cell therapy, the paracrine responsiveness of stem cells to the wound microenvironment will determine their effectiveness in promoting healing [3,6,7]. The bone marrow stem cells (healthy and dysfunctional) responded by upregulating the gene expression and protein secretion of macrophage chemoattractants MCP1 (Ccl2) and MIP1α (Ccl3). In chronic diabetic wounds, a significant delay in macrophage infiltration related to a deficiency in ccl2 has been demonstrated [33]. Wood et al. [33] furthermore indicated that acute ccl2 treatment could restore the macrophage response and promote healing. This is supported by Yin et al. [34], indicating that the use of scaffolds that actively release MCP1 promoted macrophage infiltration and improved wound healing in a mouse model of streptozotocin-induced diabetes. MCP1 is also known to influence the effector state of macrophages [35], to mediate the angiogenic effect of transforming growth factor beta (TGFβ) by recruiting vascular smooth muscle cells to the endothelium [36], and to recruit/improve the migration of bone marrow MSCs [37]. In this study, preconditioning further enhanced the MCP1 response of all the bone marrow stem cells.

Dysfunctional bone marrow stem cells did, however, have an excessive TNFα response to DWF. TNFα is a potent proinflammatory cytokine that amplifies inflammation through NFκB signaling [38]. Persistent TNFα signaling is known to occur in diabetic wounds [39] and prevents the phenotype switch of macrophages form phagocytic (M1) to proregenerative (M2). Ashcroft et al. [40] demonstrated that targeting TNFα in a mouse model (secretory leukocyte protease inhibitor null mice) of severely impaired wound healing-blunted leukocyte recruitment reduced NFκB signaling, altered the M1/M2 macrophage balance, and improved healing. In this study, antioxidant preconditioning of dysfunctional bone marrow stem cells blunted the TNFα response and improved the secretion of anti-inflammatory IL10.

Taken together, these data suggest that the combined ex vivo treatment of autologous stem cells with NAC and AAP could be an effective strategy to restore the paracrine function of impaired bone marrow stem cells before transplantation. This study was, however, done with acute exposure to the pathological microenvironment and future studies should assess whether these cells will retain their restored ability over longer periods of exposure.

Footnotes

Acknowledgments

This research was supported by grants from the National Research Foundation (NRF) (grant no. 105921) and the Harry Crossley foundation. YMA received bursaries from the NRF and the Faculty of Medicine and Health Sciences, Stellenbosch University. We would like to thank Prof. K.H. Myburgh for her insight and role as mentor.

Author Disclosure Statement

No competing financial interests exist.

References

van Acker

, Léger

, Hartemann

, Chawla

and Siddiqui

. (2014). Burden of diabetic foot disorders, guidelines for management and disparities in implementation in Europe: a systematic literature review. Diabetes Metab Res Rev, 30:635–645.

Ahmad

. (2016). The diabetic foot. Diabetes Metab Syndr, 10:48–60.

Mulder

, Lee

and Jeppesen

. (2012). Comprehensive review of the clinical application of autologous mesenchymal stem cells in the treatment of chronic wounds and diabetic bone healing. Int Wound J, 9:595–600.

Wild

, Roglic

, Green

, Sicree

and King

. (2004). Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care, 27:1047–1053.

Baltzis

, Eleftheriadou

and Veves

. (2014). Pathogenesis and treatment of impaired wound healing in diabetes mellitus: new insights. Adv Ther, 31:817–836.

Heublein

, Bader

and Giri

. (2015). Preclinical and clinical evidence for stem cell therapies as treatment for diabetic wounds. Drug Discov Today, 20:703–717.

Marfia

, Navone

, Di Vito

, Ughi

, Tabano

, Miozzo

, Tremolada

, Bolla

, Crotti

, et al. (2015). Mesenchymal stem cells: potential for therapy and treatment of chronic non-healing skin wounds. Organogenesis, 11:183–206.

van de Vyver

. (2017). Intrinsic mesenchymal stem cell dysfunction in diabetes mellitus: implications for autologous cell therapy. Stem Cells Dev, 26:1042–1053.

van de Vyver

, Niesler

, Myburgh

and Ferris

. (2016). Delayed wound healing and dysregulation of IL6/STAT3 signalling in MSCs derived from pre-diabetic obese mice. Mol Cell Endocrinol, 426:1–10.

10.

Shin

and Peterson

. (2012). Impaired therapeutic capacity of autologous stem cells in a model of type 2 diabetes. Stem Cells Transl Med, 1:125–135.

11.

Silva

, Sampaio

, Fernandes

and Gomes

. (2015). The osteogenic priming of mesenchymal stem cells is impaired in experimental diabetes. J Cell Biochem, 116:1658–1667.

12.

Zhang

, Liu

, Shi

, Wu

, Gao

, He

, Gu

and Liu

. (2016). High glucose microenvironments inhibit the proliferation and migration of bone mesenchymal stem cells by activating GSK3β. J Bone Miner Metab, 34:140–150.

13.

Dreifke

, Jayasuriya

and Jayasuriya

. (2015). Current wound healing procedures and potential care. Mater Sci Eng C Mater Biol Appl, 48:651–662.

14.

Dulmovits

and Herman

. (2012). Microvascular remodeling and wound healing: a role for pericytes. Int J Biochem Cell Biol, 44:1800–1812.

15.

Martin

and Nunan

. (2015). Cellular and molecular mechanisms of repair in acute and chronic wound healing. Br J Dermatol, 173:370–378.

16.

Pence

and Woods

. (2014). Exercise, obesity, and cutaneous wound healing: evidence from rodent and human studies. Adv Wound Care (New Rochelle), 3:71–79.

17.

Ali

, Khan

and Riazuddin

. (2016). N-Acetyl cysteine protects diabetic mouse derived mesenchymal stem cells from hydrogen-peroxide-induced injury: a novel hypothesis for autologous stem cell transplantation. J Chin Med Assoc, 79:122–129.

18.

Bae

, Ryu

, Rhee

K-J

, Oh

J-E

, Baik

, Shim

, Kong

, Hyun

, Pack

, et al. (2015). L-ascorbic acid 2-phosphate and fibroblast growth factor-2 treatment maintains differentiation potential in bone marrow-derived mesenchymal stem cells through expression of hepatocyte growth factor. Growth Factors, 33:71–78.

19.

Jeong

S-G

and Cho

G-W

. (2015). Endogenous ROS levels are increased in replicative senescence in human bone marrow mesenchymal stromal cells. Biochem Biophys Res Commun, 460:971–976.

20.

Zheng

, Sui

, Hu

and Jin

. (2015). [Vitamin C promotes in vitro proliferation of bone marrow mesenchymal stem cells derived from aging mice]. Nan Fang Yi Ke Da Xue Xue Bao, 35:1689–1693.

21.

Antezana

, Sullivan

, Usui

, Gibran

, Spenny

, Larsen

, Ansel

, Bunnett

and Olerud

. (2002). Neutral endopeptidase activity is increased in the skin of subjects with diabetic ulcers. J Invest Dermatol, 119:1400–1404.

22.

Spenny

, Muangman

, Sullivan

, Bunnett

, Ansel

, Olerud

and Gibran

. (2002). Neutral endopeptidase inhibition in diabetic wound repair. Wound Repair Regen, 10:295–301.

23.

Löffler

, Schuster

, Bühler

and Beckert

. (2013). Wound fluid in diabetic foot ulceration: more than just an undefined soup?. Int J Low Extrem Wounds, 12:113–129.

24.

Schmohl

, Beckert

, Joos

, Königsrainer

, Schneiderhan-Marra

and Löffler

. (2012). Superficial wound swabbing: a novel method of sampling and processing wound fluid for subsequent immunoassay analysis in diabetic foot ulcerations. Diabetes Care, 35:2113–2120.

25.

Feoktistova

, Geserick

and Leverkus

. (2016). Crystal violet assay for determining viability of cultured cells. Cold Spring Harb Protoc, 2016: pdb.prot087379.

26.

Staiano-Coico

, Higgins

, Schwartz

, Zimm

and Goncalves

. (2000). Wound fluids: a reflection of the state of healing. Ostomy Wound Manage, 46:85S–93S; quiz:94S–95S.

27.

Yager

, Kulina

and Gilman

. (2007). Wound fluids: a window into the wound environment?. Int J Low Extrem Wounds, 6:262–272.

28.

Goren

, Müller

, Pfeilschifter

and Frank

. (2006). Severely impaired insulin signaling in chronic wounds of diabetic ob/ob mice: a potential role of tumor necrosis factor-alpha. Am J Pathol, 168:765–777.

29.

Ring

, Scully

, Davis

, Baker

, Cullen

, Pelleymounter

and Danilenko

. (2000). Systemically and topically administered leptin both accelerate wound healing in diabetic ob/ob mice. Endocrinology, 141:446–449.

30.

Schürmann

, Goren

, Linke

, Pfeilschifter

and Frank

. (2014). Deregulated unfolded protein response in chronic wounds of diabetic ob/ob mice: a potential connection to inflammatory and angiogenic disorders in diabetes-impaired wound healing. Biochem Biophys Res Commun, 446:195–200.

31.

Stallmeyer

, Pfeilschifter

and Frank

. (2001). Systemically and topically supplemented leptin fails to reconstitute a normal angiogenic response during skin repair in diabetic ob/ob mice. Diabetologia, 44:471–479.

32.

Behm

, Babilas

, Landthaler

and Schreml

. (2012). Cytokines, chemokines and growth factors in wound healing. J Eur Acad Dermatol Venereol, 26:812–820.

33.

Wood

, Jayaraman

, Huelsmann

, Bonish

, Burgad

, Sivaramakrishnan

, Qin

, DiPietro

, Zloza

, Zhang

and Shafikhani

. (2014). Pro-inflammatory chemokine CCL2 (MCP-1) promotes healing in diabetic wounds by restoring the macrophage response. PLoS One, 9:e91574.

34.

Yin

, Ding

, Shi

, Guo

, Ni

, Fu

and Fu

. (2016). A bioengineered drug-Eluting scaffold accelerated cutaneous wound healing In diabetic mice. Colloids Surf B Biointerfaces, 145:226–231.

35.

Low

, Drugea

, Duffner

, Quinn

, Cook

, Rollins

, Kovacs

and DiPietro

. (2001). Wound healing in MIP-1alpha(−/−) and MCP-1(−/−) mice. Am J Pathol, 159:457–463.

36.

, Wang

, Fei

, Han

J-DJ

and Chen

Y-G

. (2007). MCP-1 mediates TGF-beta-induced angiogenesis by stimulating vascular smooth muscle cell migration. Blood, 109:987–994.

37.

Ishikawa

, Ito

, Kitaori

, Murata

, Shibuya

, Furu

, Yoshitomi

, Fujii

, Yamamoto

and Matsuda

. (2014). MCP/CCR2 signaling is essential for recruitment of mesenchymal progenitor cells during the early phase of fracture healing. PLoS One, 9:e104954.

38.

Liu

. (2005). Molecular mechanism of TNF signaling and beyond. Cell Res, 15:24–27.

39.

Landis

, Evans

, Chaturvedi

and Haskard

. (2010). Persistence of TNFalpha in diabetic wounds. Diabetologia, 53:1537–1538.

40.

Ashcroft

, Jeong

M-J

, Ashworth

, Hardman

, Jin

, Moutsopoulos

, Wild

, McCartney-Francis

, Sim

, et al. (2012). Tumor necrosis factor-alpha (TNF-α) is a therapeutic target for impaired cutaneous wound healing. Wound Repair Regen, 20:38–49.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

0.07 MB

Antioxidant Preconditioning Improves the Paracrine Responsiveness of Mouse Bone Marrow Mesenchymal Stem Cells to Diabetic Wound Fluid

Abstract

Introduction

Materials and Methods

Induction of full-thickness wounds and collection/analysis of wound fluid

Determination of the safest, nontoxic doses of NAC and AAP for in vitro antioxidant preconditioning

Isolation and characterization of bone marrow-derived SCs

In vitro treatment conditions and study design

Collection and analysis of conditioned media

RNA isolation, cDNA synthesis, and RT2 quantitative polymerase chain reaction microarray

Statistical analysis

Results

Full-thickness excisional diabetic wound model: collected wound fluid is representative of the in vivo microenvironment of an inflammatory diabetic wound

The safest nontoxic doses of NAC and AAP for in vitro antioxidant preconditioning

Characterization of healthy control (source: wild type, C57BL/6J mice) and impaired/dysfunctional (source: obese diabetic, B6.Cg-Lepob/J mice) bone marrow-derived SCs

Secretome differences at baseline: healthy control (SCWT) versus impaired/dysfunctional (SCob)

Gene expression (mRNA) differences at baseline and in response to stimulation with DWF: healthy control (SCWT) versus impaired/dysfunctional (SCob)

Differences in the paracrine responsiveness following DWF stimulation: healthy control (SCWT) versus impaired/dysfunctional (SCob)

Discussion

Footnotes

Acknowledgments

Author Disclosure Statement

References

Supplementary Material

Characterization of healthy control (source: wild type, C57BL/6J mice) and impaired/dysfunctional (source: obese diabetic, B6.Cg-Lep^ob/J mice) bone marrow-derived SCs

Secretome differences at baseline: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)

Gene expression (mRNA) differences at baseline and in response to stimulation with DWF: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)

Differences in the paracrine responsiveness following DWF stimulation: healthy control (SC^WT) versus impaired/dysfunctional (SC^ob)