Occlusive dressing-induced secretomes influence the migration and proliferation of mesenchymal stem cells and fibroblasts differently

HS-27, a commercially available human derived fibroblast cell line (ATCC—LGC Standards, Germany), was cultured in DMEM (Biochrom, Germany) supplemented with 10% FCS. HS-27 cells are human foreskin fibroblasts which are growing adherent to cell-culture plastic. We used adipose-derived MSCs also known as adipose-derived stem cells (ASCs or ADSCs) isolated from aesthetic liposuction, as previously described [21]. The cells were characterized in our lab by their expression of CD73, CD90, and CD105 as well as their potency to differentiate into fat cell, chondrocytes, and osteoblasts. For cultivation, MEM-alpha (Biochrom, Germany), containing 10% FCS was used.

MSCs were seeded in a density of 8000 cells/well and HS-27 FBs in a density of 4000 cells/well in 96 well plates using standard cultivation media as described above. Cells were allowed to attach overnight, then the medium was replaced with sample (wound fluid or serum) in a concentration of 10%. Alamarblue assay (Sigma-Aldrich) was performed after 3 days of incubation at 37 °C in 5% CO₂ atmosphere according to the instructions of the manufacturer. In short, Alamarblue reagent was added to the cells and incubated for 30 min, followed by fluorescence measurement at Ex 530 nm/Em 600 nm.

The xCELLigence device enables the continuous analysis of cell proliferation. The proliferation rates of MSCs and HS27 fibroblasts were determined with the xCELLigence-DP Analyser (OLS^® Omni Life Sciences, Germany). We seeded 4000 HS-27 FBs and 8000 MSCs in individual wells and incubated them with standard cultivation medium for 40 h. Wound fluid or serum was prediluted 1:10 in standard medium before it was added to the seeded wells in a ratio of 1:1 to achieve a final sample concentration of 5%. Continuous measurement of impedance (Cell index) started 30 min after seeding and was performed in intervals of 15 min for 100 h. Analysis was done for the first 60 h.

Migration analysis was performed in CIM plates in combination with xCELLigence-DP-Analyzer (OLS^® Omni Life Sciences). The CIM plate consists of an upper chamber with a porous membrane and a lower chamber. 40,000 HS-27 FBs or MSCs were seeded in standard medium in the wells of the upper chamber and left for 30 min to settle. As chemoattractant 5% wound exudate or 5% serum, diluted in standard medium, was filled in the wells of the lower chamber. Impedance measurement at the lower side of the porous membrane was carried out to visualize cell migration from the upper to the lower chambers. Impedance read out was expressed as cell index and executed every 15 min to a maximum of 25 h. Analysis was done for the first 15 h.

We first pipeted 70 μl cell suspension (1.2 × 10⁵ cells/ml) in each of the 24 wells of Ibidi culture inserts (Ibidi, Germany). Next, cells were incubated at 37 °C and 5% CO₂ for 24 h to obtain a confluent cell layer. Culture inserts were removed afterwards and cell layers washed with PBS. Filled plate wells were prepared with culture medium (Dulbecco’s MEM by Merck Millipore, 2% wound fluid and 1% antibiotic/antimycotic ingredient by Capricorn, AAS-B) up to 500 μl. Wells filled with medium without wound fluid were used as negative controls. Microscopy pictures were taken at 0 h, 6 h, 12 h, and 24 h and images were analyzed using the WimScratch software.

Data are shown as mean ± SEM. Statistical analysis between the groups was performed using unpaired t tests. p values of less than 0.05 were considered statistically significant.

The migration of MSCs under the influence of fingertip fluid measured with xCELLigence was significantly increased at 4 h (p = 0.046), 5 h (p = 0.032) and 6 h (p = 0.048) compared to blood serum controls of the same patients. This effect reversed significantly in the MSC population of occlusive treated fingertip injuries after 11 h (p = 0.047) and continued afterwards (see Fig. 2). MSCs under the influence of split skin fluid showed initially a trend towards an increased migration compared to respective serum controls, which was not significant. A significant decrease of MSC migration, influenced by split skin donor site fluid, was found starting at 9 h after incubation (p = 0.049) and continued to do so (see Fig. 2). Combined analysis of MSC migration of both wound fluids compared to both serum controls showed a significant increase from 3 h (p = 0.023) to 6 h (0.017) with the strongest increase at 4 h (p = 0.001). A significant decrease in MSC migration of all fluids compared to all serum values started at 9 h (p = 0.034) onwards (see Fig. 2).

The migration of HS27-FBs of fingertip fluids in xCELLigence was significantly increased at 5 h (p = 0.041), 6 h (p = 0.043) and 7 h (p = 0.049) compared to respective serum controls (see Fig. 3). The migration of HS27 FBs of split skin fluid alone did not show any significant changes, but mimicked the curve of the fingertip fluid values with a less steep increase and decline after a few hours. Analyzing the migration HS27 FBs of all wound fluid samples compared to all serum samples showed a significant increase between 5 and 10 h after sample incubation (p = 0.004) (see Fig. 3).

The in vitro wound-healing assay showed a significantly increased HS27 FB migration at 12 h (p = 0.012) and at 24 h (p = 0.017) compared to the control, supporting our data from the impedance analysis (see Figs. 4, 5).

We observed a non-significant increase of cell proliferation of MSCs exposed to each fluid alone at 5 h (fingertips p = 0.061, split skin donor sites p = 0.511) and 10 h (fingertips p = 0.091, split skin donor sites p = 0.511). By comparing of all fluid samples with the respective serum samples, we observed a significant increase at 5 h (p = 0.034) and 10 h (p = 0,041). However, this effect was reversed at 15 h. A significant decrease in MSC proliferation of both wound fluids combined compared to the controls was seen starting at 45 h of incubation (p = 0.046) and later (see Fig. 6). The proliferation of HS27 FBs showed initially a slow decrease in both fluids, which was significant in the fingertip fluid group compared to the control group at 15 h (p = 0.012) and 20 h (p = 0.030). The pooled analysis of HS27 FB proliferation of all fluids compared to all serum controls was significantly decreased from 15 h (p = 0.022) to 25 h (0.039), continuing to be so at 40 h (p = 0.044) onward (see Fig. 7).

HS27 FBs were significantly reduced in the fingertip fluid group compared to respective controls (p = 0.006), whereas MSCs were significantly decreased only in the split skin fluid group compared to controls (p = 0.0448). We saw a significant decrease in MSC proliferation exposed to both combined wound fluids compared to blood serum controls (p = 0.008). HS27 FB proliferation also decreased in both pooled fluid groups significantly compared to the controls (p = 0.003) after 3 days of sample incubation (see Fig. 8).

We did not observe any statistically significant difference in the behavior of mesenchymal stem cells and fibroblasts between samples from fingertip or split skin donor site wound fluids.

Recent studies showed that hypoxic stimuli seem to mobilize MSCs from the wound bed and that continuous hypoxia increases MSC mobilization by supposedly upregulation of hypoxia-induced factor-1α (HIF-1α) [38]. In our study, we demonstrated a significant increase of MSC mobilization when exposed to fluid from OD wounds, possibly showing one more factor involved in modulating wound repair. Here, we observed a stronger increase of MSC migration in the fingertip fluid group as in the occlusive-dressed split skin fluid group. As MSC motility decreases after reaching a plateau in our experiments, there seem to be further mechanisms in the regulation of MSC activity involved. Occlusive dressings, which maintain a hypoxic wound environment, could be responsible for the observed positive effects on wound healing [14, 15].

It would have been very interesting to study the contents of the secretome in our wound fluids in more detail, but unfortunately, we were restricted by the amount of fluid that could be obtained from our patients. Per patient, we only could obtain around 90 µl of fluid, so we had to decide to either study the proteomic content or do cell-culture analysis. Here, we decided to focus on the cell behavior modulation of fibroblasts and MSCs. Fortunately, there are already data available about the contents of such wound fluids: Vogt et al. analyzed wound fluid from occlusive-dressed fingertip injuries on growth factors and found upregulated bFGF, epidermal growth factor (EGF), interleukin-1α (IL-1α), and TGF-β2 compared to blood serum values, which could be a reason why ODs on fingertip injuries show a strong regenerative potential [39]. It was hypothesized that wounds heal faster due to fibroblast and keratinocyte proliferation, increased angiogenesis, and growth factor expression [22, 23]. As fibroplasia and the formation of granulation tissue mainly bases on an initial burst of fibroblast activity, scar quality should not be increased due to an overabundance of proliferating fibroblasts, which are often seen in hypertrophic scars [35]. Our results show that fibroblast proliferation seems to be inhibited by occlusive-dressed wound fluids, which could be another major factor to initiate a favorable scar modulation process. These data are corroborated by a recent study, showing that inhibiting a known-scar-forming fibroblast lineage during wound healing did not lead to a change of healing rates during the initial days of wound healing. However, a significant difference in wound size developed by day 9 after wounding, with larger and less healed wounds in the treated cohorts. Most importantly, treated wounds showed significantly reduced final scar size after completion of wound healing [40]. Similarly, we could observe a spike of fibroblast mobility in the first 10 h after sample incubation, which afterwards decreased significantly. As fibroblasts are important players in the wound-healing process, their active role in the inflammatory and proliferative phase of wound healing is essential [35]. In dry wound healing, a fast wound closure is important to minimize fluid loss and reduce the risk of infection, thus an increased fibroblast proliferation and activity is necessary to minimize these risks, but leading to a scar [1]. A superficial occlusive barrier, allowing wound fluid to cover the wound, seems to inhibit excessive fibroblast proliferation, as we could show in our study. Interestingly, this effect was more pronounced in the fingertip fluid samples as in the split skin fluid samples, which could be responsible for the good clinical outcomes observed in occlusively dressed fingertip injuries. In addition, split skin fluids showed a significantly decreased MSC proliferation after 3 days, which was not that prominent in the fingertip fluid group, suggesting that also other factors contribute to the positive effects of ODs in wound healing [25].

Our results were generated in a simplified in vitro model of cell migration and proliferation. This is limited by several factors: the wound fluid is only administered once in the model, while it is presumably constantly produced in the in vivo wound situation. Thus, the results in the early hours of the experiment may reflect the wound environment better than the later data. In vivo, the wound is an injured 3D-tissue consisting of extracellular matrix and a multitude of cells that constantly interact with each other through autocrine, paracrine, electrical, and mechanical cues that are not reproduced in the model. On the other hand, this approach allowed us to isolate effects of the wound fluid on two specific cell types that are thought to be important in wound healing, which is not as easily possible in a more complex model.