Acute and chronic wound fluids inversely influence adipose‐derived stem cell function: molecular insights into impaired wound healing

Abstract

Wound healing is a complex biological process that requires a well‐orchestrated interaction of mediators as well as resident and infiltrating cells. In this context, mesenchymal stem cells play a crucial role as they are attracted to the wound site and influence tissue regeneration by various mechanisms. In chronic wounds, these processes are disturbed. In a comparative approach, adipose‐derived stem cells (ASC) were treated with acute and chronic wound fluids (AWF and CWF, respectively). Proliferation and migration were investigated using 3‐(4,5‐Dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) test and transwell migration assay. Gene expression changes were analysed using quantitative real time–polymerase chain reaction. AWF had a significantly stronger chemotactic impact on ASC than CWF (77·5% versus 59·8% migrated cells). While proliferation was stimulated by AWF up to 136·3%, CWF had a negative effect on proliferation over time (80·3%). Expression of b‐FGF, vascular endothelial growth factor (VEGF) and matrix metalloproteinase‐9 was strongly induced by CWF compared with a mild induction by AWF. These results give an insight into impaired ASC function in chronic wounds. The detected effect of CWF on proliferation and migration of ASC might be one reason for an insufficient healing process in chronic wounds.

Introduction

Nowadays, chronic wounds represent a major problem in medical field as their incidence is continuously increasing because of an aging population and a rise in the incidence of underlying diseases ¹. As estimated, about 2% of the population in industrialised countries will be affected by chronic wounds in their lifetime ². Physiological wound healing is a complex biological process proceeding from the stage of inflammation through proliferation to maturation ³. It requires a well‐orchestrated interaction of mediators, resident and infiltrating cells. Basal keratinocytes and fibroblasts play an important role in this process as they proliferate and migrate and finally lead to a complete wound closure. Mesenchymal stem cells contribute to the regeneration process as they have the ability to replace damaged tissue ⁴. In chronic wounds, these physiological mechanisms are disturbed by mediators of the wound environment. It has been shown that in chronic wound fluid (CWF), proinflammatory cytokines such as tumour necrosis factor (TNF)‐α, IL‐1 as well as matrix metalloproteinases (MMPs) and neutrophil elastase are elevated, whereas in acute wound fluids (AWFs), the levels of tissue inhibitors of metalloproteinases are reduced ^{5, 6, 7, 8}. MMP2 and MMP9 are the pivotal effectors in wound healing, both during remodelling and reepithelialisation ⁹. They belong to the family of gelatinases and are involved in the breakdown of extracellular matrix during the physiological regeneration process ¹⁰. The growth factor content of wound fluids is still subject to controversy in literature. However, in most studies, the average level of growth factors in CWF was shown to be significantly lower than that in AWF ^{7, 11}. Among these, VEGF and b‐FGF are the two key players of wound healing ^{12, 13, 14}. VEGF has been shown to influence angiogenesis via the HIF‐1α pathway induced by hypoxia ¹⁵, whereas b‐FGF has positive effects on proliferation, migration and angiogenic processes ^{16, 17, 18}. Earlier studies have already shown that adipose‐derived stem cells (ASC) can positively influence wound healing, as they are attracted to the wound site and influence wound healing processes via paracrine mechanisms as well as fusion and differentiation, for example, into keratinocytes or fibroblasts ^{4, 19, 20}. They were shown to promote reepithelialisation and angiogenesis, and thereby accelerate wound healing ¹⁹. In chronic wounds, these physiological mechanisms are disturbed. To elucidate differences between ASC behaviour in acute and chronic wounds, we chose a comparative approach, treating human ASC with AWFs and CWFs. We analysed the proliferation and migration capacity, and gene expression pattern of ASC.

Methods

Adipose‐derived stem cells

Ethical approval for abdominal liposuction was received from the local ethics commission (39/2007). Informed consent was obtained from all patients for sample collection. ASC were isolated from lipoaspirate after abdominal liposuction. Liposuction was performed using the tumescence technique. The fat fraction of the sample was homogenised using collagenase I (Sigma‐Aldrich, Munich, Germany) and ASC were isolated by centrifugation. The cells were cultured in modified eagle medium (α‐MEM) (PAN Biotech, Aidenbach, Germany) containing 20% fetal calf serum FCS (Sigma‐Aldrich), 1 ng/ml b‐FGF, 1 ng/ml endothelial growth factor (EGF) and 1% penicillin/streptomycin. For further experiments, they were cultured in dulbecco modified eagle medium (DMEM) containing 2% (FCS) (Sigma, Hamburg, Germany) and 2% kanamycin. Passages of 2 and 3 of ASC were used for experiments.

Identification of ASC

ASC were identified as human mesenchymal stem cells (MSC) according to the minimal criteria of the International Society for Cellular Therapy: plastic‐adherence, expression of CD73, CD90 and CD105, lack of CD45 expression, and the ability to differentiate into adipocytes ²¹.

Immunophenotyping

ASC were cultured as described above. After 2 days, the cells were fixed with 4% paraformaldehyde (PFA). Immunofluorescence staining was performed using anti‐human CD45, CD90 and CD105 primary antibodies (Stemgent, San Diego, CA) and an Alexa Fluor 568‐coupled secondary antibody (Stemgent). A phycoerythrin (PE)‐conjugated anti‐human CD73 antibody (BioLegend, Fell, Germany) was used for CD73 staining and a PE‐coupled IgG antibody (BioLegend) served as isotype control. Nuclei were stained with bisbenziamide (Sigma). Fluorescence microscopy was carried out using the fluorescence microscope Leica CTR400, Wetzlar, Germany.

Adipogenesis

ASC were cultured in human NH AdipoDiff Medium (Miltenyi Biotec, Bergisch‐Gladbach, Germany). After 3 weeks, the cells were fixed with 4% PFA and stained with Oil‐Red‐O (Sigma). A microscopic analysis was carried out for the evidence of fat differentiation.

Preparation of wound fluids

Informed consent was obtained from all patients for wound fluid collection. Patient‐specific data are shown in Table 1.

Table 1.

Patient data

Sample ID	Age (years)	Sex	BMI (kg/m2)	Diabetes	Arterial hypertension	Vessel disease	Smoking	GC therapy	Microbiology of wound swab
AWF I	42	F	32·5	−	−	−	−	−	Negative
AWF II	34	M	19·2	−	−	−	−	−	Negative
AWF III	46	F	39·6	−	−	−	−	−	Negative
AWF IV	34	F	37·1	−	+	−	−	−	Negative
AWF V	65	F	26	+	+	−	−	−	Negative
Mean	44·2		30·88
CWF I	61	F	35·7	+	+	−	+	−	Enterococci, pseudomonads
CWF II	87	M	16·3	−	−	−	+	−	Enterobacteria, enterococci, staphylococci
CWF III	63	F	35·6	+	−	−	−	−	Pseudomonads, staphylococci, streptococci
CWF IV	60	F	22·6	−	+	−	−	+	Enterobacteria, enterococci, pseudomonads, staphylococci
CWF V	61	F	26·9	−	+	−	−	−	Enterobacteria, enterococci
Mean	66·4		27·42

AWF, acute wound fluid; BMI, body mass index; CWF, chronic wound fluid; F, female; GC, glucocorticoid; M, male.

Acute wound fluid

AWF was collected from five patients who underwent abdominoplasty. Wound fluid that was subcutaneously drained during the first 8 hours after operation was discarded to exclude blood contamination. Wound fluid drained within the following 8 hours was collected and centrifuged. The supernatant was diluted with DMEM and filter sterilised. Protein concentrations of all samples were analysed by Bradford test according to the manufacturers' instructions and pooled afterwards.

Chronic wound fluid

CWF was collected from patients, who were admitted for surgical treatment of chronic sacral decubitus or showed a sacral decubitus as secondary diagnosis. All included patients met the following criterion: presence of a chronic sacral decubitus for minimum 6 weeks without prior vacuum therapy. Wound fluid was collected after applying an occlusive dressing for 24 hours. After wound fluid centrifugation, the supernatant was diluted with DMEM and passed through a sterile filter. Wound fluid from five patients was subjected to Bradford assay for protein quantification and pooled for further experiments.

To identify the best‐qualified wound fluid concentration for our experiments, we evaluated different concentrations of AWF and CWF with respect to their impact on ASC proliferation using MTT assay. Based on these results, we chose a concentration of 2% AWF and CWF, respectively.

Phase contrast microscopy for the evaluation of cell morphology

Cell culture was performed as described above. A total of 20 000 ASC were placed on a six‐well microplate. They were then incubated with 2% AWF, 2% CWF or left untreated. Cell morphology was observed for 6 days using the microscope Leica CTR400 and Leica Application Suite V3.6 (software, Wetzlar, Germany).

MTT assay

MTT test was performed to assess cell proliferation. Therefore, ASC were seeded on a 96‐well microplate. They were allowed to attach overnight and were then incubated with 2% AWF or 2% CWF. MTT assay was performed after 1, 2 and 3 days using MTT reagent (5 µg/ml; Sigma‐Aldrich, Hamburg, Germany) according to the manufacturers' instructions. Extinction was measured at 570 nm using the ELISA reader μQuant (Biotek, Bad Friedrichshall, Germany).

Transwell migration assay

ASC were cultured as described above. The cell culture inserts (FalconTM FluoroBlokTM, pore size 8 µm; BD Bioscience, Heidelberg, Germany) were placed in a 24‐well microplate. A total of 10 000 ASC were seeded into the upper chamber. After 4 hours, 2% AWF or 2% CWF were added to the bottom chamber; a standard culture medium served as a control. Following 24 hours of incubation, the membranes were fixed with 4% PFA and stained with 4',6‐diamidino‐2‐phenylindole (Sigma‐Aldrich, Hamburg, Germany). Migrated and non migrated cells were counted in eight randomly selected high‐power fields at 400× magnification using the fluorescence microscope Leica CTR400 and Leica Application Suite V3.6 software.

Quantitative real time–polymerase chain reaction (qRT–PCR)

A total of 250 000 ASC were placed on a six‐well microplate and cultured according to standard protocols. After 24 hours, they were exposed to 2% AWF or 2% CWF for 6 hours or left untreated. Cells were dissolved in guanidine thiocyanate lysis (RLT) buffer and RNA was isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers' instructions. cDNA synthesis of 1 µg total RNA was performed using the RevertAid First Strand cDNA synthesis Kit (Fermentas, St. Leon‐Rot, Germany). qRT–PCR was performed using the Brilliant II SYBR Green QRT–PCR Master Mix (Agilent Technologies, Böblingen, Germany). Data were acquired using Stratagene Mx3005P QPCR System (Agilent Technologies). Primers were obtained from Biomers (Ulm, Germany) (Table S1, Supporting Information). Expression was normalised to the housekeeping gene ribosomal protein L (RPL). The comparative threshold cycle (C _t) method was applied to determine relative expression differences ²².

Statistical analysis

For statistical reasons, all experiments were performed in triplets. Data are shown as mean ± standard error of the mean (SEM) unless otherwise indicated. Continuous variables were compared using Student's t‐test. A two‐tailed P‐value <0·05 was considered significant.

Results

Identification of ASC

By immunophenotyping, ASC were shown to be positive for the cell surface markers CD73, CD90 and CD105, whereas they lack expression of the pan‐leucocyte marker CD45 (Figure 1A). Furthermore, they were plastic‐adherent in standard cell culture (not shown) and adipogenesis could be induced (Figure 1B). Therefore, ASC were characterised as human MSC according to the minimal criteria of the International Society for Cellular Therapy ²¹.

Figure 1.

(A) Expression of cell surface marker by adipose‐derived stem cells (ASC). ASC show expression of CD73, CD90 and CD105, whereas lacking CD45 expression. Scale bars represent 200 µm. (B) Adipogenesis of ASC. ASC were cultured in AdipoDiff medium for 3 weeks, fixed with 4% paraformaldehyde and stained with Oil‐Red‐O. The staining shows intracellular lipid droplets under phase contrast microscopy (left side) as well as after Oil‐Red‐O staining (right side), showing fat differentiation. Scale bars represent 200 µm.

Protein concentration of wound fluids

Protein concentrations of AWF obtained from five different patients ranged from 35·82 to 41·72 g/l (38·77 ± 4·17 g/l), whereas they were slightly lower for CWF samples, varying from 28·64 to 38·25 g/l (33·45 ± 6·80 g/l) as summarised in Table 2.

Table 2.

Protein concentration of different wound fluid samples

Sample	Protein concentration (g/l)	Sample	Protein concentration (g/l)
AWF I	38·91	CWF I	31·34
AWF II	41·72	CWF II	35·81
AWF III	37·16	CWF III	34·30
AWF IV	36·17	CWF IV	32·26
AWF V	35·82	CWF V	38·25
Mean ± SD	38·77 ± 4·17	Mean ± SD	33·45 ± 6·80

AWF, acute wound fluid; CWF, chronic wound fluid.

Cell morphology

After 6 days of incubation with 2% AWF, ASC were clearly proliferated. They appeared more granulated, suggesting previous phagocytosis (Figure 2). Under the influence of CWF, ASC showed no proliferation activity over 6 days. They appeared smaller but much broader. Cell extension organelles increased in size and number and more granules were apparent (Figure 2).

Figure 2.

Cell morphology of adipose‐derived stem cells (ASC) under the influence of acute wound fluid (AWF) and chronic wound fluid (CWF). The morphology of ASC after 6 days of incubation with 2% AWF and 2% CWF at 200× magnification is shown. Culture media served as a control (ctrl). Images show the results of one experiment representative of three independent ones.

ASC recruitment is impaired by CWF compared with AWF

ASC were attracted by AWF and CWF. However, AWF had significantly stronger chemotactic impact on ASC than CWF (Figure 3). While 77·5% of ASC migrated towards AWF after 24 hours, only 59·8% migrated to the CWF‐containing chamber.

Figure 3.

Adipose‐derived stem cells (ASC) migration after 24 hours. The percentage of migrated (dark grey) and non‐migrated (light grey) ASC after 24 hours is shown. 2% acute wound fluid (AWF) and 2% chronic wound fluid (CWF) were used as chemotactic stimuli; culture media served as control (ctrl). Results of three independent experiments are plotted as mean ± SEM. Asterisk indicates statistically significant P‐value ≤ 0·05.

Proliferation of ASC is inversely influenced by AWF and CWF over time

The proliferation capacity of ASC was inversely influenced by AWF and CWF. Whereas, proliferation was stimulated by AWF, CWF had an inhibiting effect on ASC proliferation over time. At day 1, proliferation of ASC was already stimulated by AWF (121·6%), whereas CWF incubation had no effect (101·8%). After 3 days, the proliferation capacity of ASC increased to 136·3% under the influence of AWF compared with untreated cells and decreased to 80·3% when incubated with CWF (Figure 4).

Figure 4.

Proliferation capacity of adipose‐derived stem cells (ASC) during incubation with acute wound fluid (AWF) and chronic wound fluid (CWF). The proliferation of ASC during incubation with 2% AWF (grey bars) and 2% CWF (black bars) is shown. Proliferation is presented as percentage related to the untreated control. Results of five independent experiments are plotted as mean ± SEM. Asterisks indicate significant differences between ASC proliferation under the influence of AWF and CWF (P ≤ 0·05).

Gene expression is differentially influenced by AWF and CWF

After 6 hours of treatment, b‐FGF and VEGF expressions were induced by AWF and CWF, respectively. However, CWF had a stronger inducing effect on these growth factors than AWF. MMP9 expression was also up‐regulated by both AWF and CWF, with CWF having a stronger inducing effect. The expression of MMP2 was only marginally affected by AWF and CWF (Figure 5).

Figure 5.

Gene expression changes by acute wound fluid (AWF) and chronic wound fluid (CWF). Adipose‐derived stem cells (ASC) were treated with 2% AWF (grey bars) and 2% CWF (black bars) for 6 hours or left untreated. Fold‐changes (FC) of gene expression were analysed by qRT–PCR and plotted as mean ± SEM. Asterisks indicate significant differences in gene expression between AWF and CWF with P ≤ 0·05 (n = 3).

Discussion

CWF basically differs from AWF. In CWF, pro‐inflammatory cytokines and proteases are elevated while proteinase inhibitors are reduced compared with AWF ^{5, 6, 7, 8}. In addition, several growth factors are decreased in CWF in comparison with that of AWF ^{6, 23}. Wound healing is a complex physiological process that includes inflammation, tissue formation and remodelling ³. Besides a whole subset of different cell types involved in this process, ASC have been shown to positively influence wound healing, once they are attracted to the wound site ^{4, 19}. This study elucidated the differences in ASC function under the influence of either AWF or CWF. Evaluating proliferation capacity showed that proliferation of ASC was significantly impaired by CWF in contrast to a stimulating effect of AWF. The positive effect of AWF in terms of ASC proliferation has already been shown by Scherzed et al. ²⁴ who, however, did not investigate the influence of CWF. The anti‐proliferative effect of CWF has previously been shown for fibroblasts, endothelial cells and keratinocytes, and has been assigned to a toxic effect of CWF rather than a lack of growth factors ^{25, 26}. Furthermore, the recruitment of ASC was reduced by CWF compared with AWF. An overload or lack of certain factors or even the combination of different components in CWF might diminish ASC recruitment, which is clearly a prerequisite for their positive impact on wound healing. For AWF, a strong chemotactic activity has been demonstrated previously ²⁴. Chemokines such as CXCL12 and their receptors play an important role in stem cell migration to the site of injury ^{20, 27} and ASC have been shown to express a subset of chemokine receptors ²⁰. Therefore, a lack of chemoattractants in CWF might be responsible for an impaired ASC migration towards chronic wounds. This might be due to an increased MMP level, leading to cleavage of chemokines and therefore reduced activity ²⁸.

ASC that are still able to migrate to the wound site, are then affected by the cytotoxic environment of the chronic wound. As ASC are likely to influence wound healing mainly via paracrine mechanisms, we investigated their gene expression pattern under the influence of AWF and CWF. Performing qRT–PCR, we focused on growth factors that are essential in wound healing, such as b‐FGF and VEGF that have previously been shown to be secreted by ASC ⁴. Although VEGF is induced by both wound fluids, CWF has a stronger effect than AWF. This might be due to a stronger activation of the HIF‐1α pathway induced by tissue hypoxia in chronic wounds. Diminished blood supply is a critical factor that contributes to and maintains non healing wounds. b‐FGF and VEGF are potent angiogenic factors that are mandatory for the formation of new blood vessels. However, these growth factors alone do not lead to a functional and stable vascular network ^{29, 30, 31}. Maturation and remodelling processes are also required ³². Moreover, b‐FGF activity is primarily regulated by the level of its receptors ³¹ and VEGF has already been shown to be overexpressed in chronic wounds but is proteolytically degraded in the chronic wound environment ³³.

Additionally, we investigated the expression of the proteases MMP2 and MMP9 that are involved in extracellular matrix (ECM) degradation. They have been shown to be overrepresented in CWF compared with AWF ^{5, 6, 7}. MMP2 expression was only minimally affected by AWF and CWF, whereas MMP9 expression was significantly up‐regulated by both wound fluids. However, MMP9 induction by CWF was significantly stronger which might be due to a stimulating effect of IL‐1 and TNF‐α, which were shown to be overrepresented in CWF ⁶. This is consistent with the idea that an adequate MMP9 expression is essential for epithelialisation and tissue remodelling during the physiological wound healing process ¹⁰. Nevertheless, highly increased levels of MMP9 lead to a permanent degradation of ECM, growth factors and their receptor, and thereby prevent wounds from healing ²³.

In conclusion, these results give more insight into impaired ASC function in chronic wounds. The anti‐proliferative and migration impeding effect of CWF in contrast to a stimulating effect of AWF on ASC might be one reason for an insufficient healing process in chronic wounds. Further studies are required to understand the interplay of mediators and cells in acute and chronic wound healing.

Supporting information

Table S1. Primer sequences used for qRT–PCR