Autologous administration of adipose stromal cells improves skin flap survival through neovascularization: An experimental study

Abstract

One of the most severe complications in aesthetic and reconstructive surgeries is the partial or total necrosis of a skin flap. In our experimental study, we demonstrated the use of adipose‐derived stem cells in the increase of skin flap survival rates. Stem cells were isolated from the fat of Wistar rats and genetically modified to permanently produce a green fluorescent protein (GFP). Two random‐pattern skin flaps (2 cm × 8 cm) were elevated on the dorsal area of the spine, and after being separated from the surgical wounds with a thin silicone sheet, they were placed back onto their original location. Then, the autologous GFP‐producing cells were injected intradermally into the dorsal area of the rats. At the seventh day, after the implantation of the stem cells, a clinical and immunohistochemical control was performed. The fluorescence microscopy revealed green vascular formations, suggesting that autologous GFP stromal cells were converted into endothelial cells through neovascularization. In the control skin flaps, where no stromal cells were used, no fluorescence was observed. The statistical analysis showed significantly lower necrosis rates in the right‐sided flaps (i.e., the flaps where adipose‐derived stromal cells were injected) compared with the left‐sided ones. Findings from our study demonstrate that adipose‐derived stem cells play an important role in the improvement of skin flap survival. Neovascularization is an effective way of achieving it.

Abbreviations

(GFP)

green fluorescent protein

(ADSCs)

adipose‐derived stromal cells

(PBS)

phosphate‐buffer saline

(SB)

sleeping beauty

(VEGF)

vascular endothelial growth factor

(PE)

anti‐rabbit phycoerythrin

1. INTRODUCTION

Skin flaps are used in reconstructive surgeries to cover areas with inadequate blood supply or to resurface exposed bones or areas where no transplant can be used. Such examples are the reconstruction of eyelids, lips, ears, noses, and cheeks as well as the repair of body‐wide defects.1, 2 Skin flaps are usually preferred over skin grafts in plastic surgeries, because they have been proven superior in terms of aesthetic and functional results.

Despite the indisputable usefulness of skin flaps in aesthetic and reconstructive surgeries, as with any other surgical operations, complications may occur, and partial or total flap loss is the most serious. The main cause of the necrosis is inadequate blood supply to the flap because of various factors, such as insufficient surgical technique and comorbidities (e.g., smoking, diabetes, and peripheral angiopathy).3 Another issue related to randomly perfused local flaps is that their design is restricted by the length‐to‐width ratio to avoid partial necrosis.4

To eliminate the problem of flap necrosis, various studies have been conducted, including on the use of adipose‐derived stromal cells (ADSCs). Ischemic tissues in which stromal cells are administered show a significant survival improvement through the increase of endothelial cells. The exact mechanism of this effect remains controversial. Several studies have indicated the direct conversion of stromal cells into endothelial cells, whereas others have reported indirect increase of the endothelial cells. However, the contribution of neoangiogenesis through in vivo stem cells differentiation in skin flap survival has not be adequately studied.5 In our study, we investigated the role of neoangiogenesis on ischemic random skin flap survival, believing that the ability of adipose stem cells to differentiate in vivo to endothelial cells and new blood vessels can contribute positively to the survival of skin flaps.

2. MATERIALS AND METHOD

2.1. Subjects/Ethics

Twenty Wistar rats, each weighing 200 to 250 g and with a mean age of 40 weeks (range 30‐50 weeks), were obtained from the Animal Division of the Institut Pasteur of Athens. All rats were kept in metal cages in the experimental laboratories of the Hospital Papageorgiou in Thessaloniki. The rats were maintained in weather‐controlled chambers under controlled lighting (natural light with day/night changes) and stable temperature (20°C). Their nutrition was controlled and consisted of standard portions of solid pellet containing a full nutrient supplementation. The rats had free‐range water intake. The selection of these specific criteria was made based on the existing literature on related contemporary experimental models. The study was approved by the Committee for the Approval of Protocols Using Animals of the Division of Veterinary Medicine of the Aristotle University of Thessaloniki (No. 13/6843/22.06.2010). All animals were anaesthetised by intraperitoneal administration of ketamine/xylazine solution, as a rapid injection, at a dose of 0.05 to 0.1 mL/100 g body weight.

2.2. Isolation of stem cells

The first step was the surgical removal of the adipose tissue from the inguinal region. The fur was removed by shaving after antisepsis with alcoholic solution 70%. The operation field was covered with sterile surgical drapes. Sections were made in the regions 38 and 39 according to Krinke G. J.6 with a #11 scalpel blade. The subcutaneous fat was then transferred to a sterilised test tube (Eppendorf type) containing sterile normal saline. The inguinal fat obtained was finely minced and digested with 0.5 mg/mL type‐1 collagenase (Sigma‐Aldrich, St. Louis, Missouri) for 1 hour at 37°C. Then, it was washed with water under constant shaking. After the addition of an equal volume of phosphate‐buffer saline (PBS), the mixture was isolated for 15 minutes at room temperature, until three distinct layers were formed. The middle layer, containing the stromal cells, was aspirated with a syringe and then centrifuged for 10 minutes at 600 × g. The sample was, thereby, divided into two layers, the lower one including the stromal cells.7

2.3. Identification of the stromal cells

Flow cytometry was performed for the detection of typical surface markers expression. As a first step, the cells were detached using Trypsin–EDTA ×1 in PBS and mild centrifugation. Second, staining with monoclonal antibodies CD34, CD45, CD44, and CD73 (BD Pharmingen) was performed for 15 minutes in the absence of light. The final results were obtained with a FACSCalibur device (Becton Dickinson, BD, Franklin Lakes, New Jersey) and analysed with the CellQuest Pro6 software. ADSCs were characterised by the coexpression of the CD44/CD73 markers and the absence of CD45/CD34 markers (Figure 1).

Figure 1.

Immunophenotypic characterisation of the administered MSCs. ADSCs were characterised by the coexpression of the CD44/CD73 markers and the absence of CD45/CD34 markers

The induction of differentiation towards osteocytes and adipocytes was accomplished by introducing osteogenesis or adipogenic induction medium in the culture for 28 and 30 days, respectively. Cells plated in plastic surfaces were used as a control group, in the same number as the rest of the other groups. The successful induction of differentiation towards osteocytes was verified with Alizarin Red staining while the respective differentiation towards adipose cells was estimated upon oil red staining (Figure 2).

Figure 2.

Staining of induced differentiated MSCs towards osteocytes and adipocytes

2.4. Culture of stromal cells and green fluorescent protein plasmid addition in their genome

The stromal cells were cultured with the addition of Dulbecco's modified Eagle medium supplemented with 5% foetal calf serum, penicillin (100 IU/mL), and streptomycin (100 μg/mL). The medium was changed every 4 days, and continuous measurements of the viable cells under an optical microscope were performed with the Trypan‐Blue technique, which stains only living cells with intact cell membranes.

After confluency had reached 60%, green fluorescent protein (GFP) plasmid was added into the genome of the stromal cells, by means of the Xfectreagent and the Sleeping Beauty (SB 100x/pT2‐CAGGSßGFP‐1/10) transposase/transposon system as previously described.8 The culture was continued until the number of the GFP stromal cells reached 1 million (Figure 3).

Figure 3.

Image of stem cells under fluorescence microscopy. The stem cells are genetically engineered to produce the green fluorescent GFP protein using the “Sleeping Beauty” transfusion/transposon system (scale bar = 150 μM). The red arrows show GFP stem cells, which emit green fluorescence

2.5. Random‐pattern skin flap model

Two identical random‐pattern skin flaps, 8 cm long and 2 cm wide, were raised on the dorsal spine of the rats, at areas 40 and 42, according to Krinke's definition.6 The flaps were divided by a narrow skin bridge to prevent cell transport between the flaps. After their elevation, a 0.13‐mm‐thick silicone sheet was placed at the bottom of every wound. The flaps were then sutured back into their initial position. The harvested autologous GFP stromal cells (1.000.000) were intradermally injected into the middle dermis along the long axis of the right‐sided skin flap (Figure 4). Each flap received one injection of 1 mL of PBS solution including the ASCs with a 26‐gauge needle. In each left skin flap, 1 mL of cell‐free PBS was administered.

Figure 4.

Intraoperative image depicting the mode of injection of stem cells, isolated from the fat of each rat, on the right‐sided skin flaps. The PBS solution of 1 mL containing (1.000.000) autologous GFP stromal cells was injected intradermally along the long axis using a 26‐gauge needle. Equal volumes of cell‐free PBS were injected in each left‐sided flap (the control flap) in the same way

2.6. Clinical control

Seven days after stromal cell injection, flap necrosis was assessed. Digital photos were obtained with the same digital camera for all rats with standardised settings at a constant distance, with a ruler next to the wound for scale. Flap necrosis, in terms of necrotic surface, was quantified. The measurement process was independently performed by two of the authors (KV and FP). The surface areas of necrotic tissue were calculated using ImageJ software which is an image processing program for Optical and Computational Instrumentation. In case of discrepancies, the worst score was accepted and further analysed.

2.7. Immunohistochemical control

Immunohistochemistry was conducted by means of cryostat sections from the same histological skin flap samples. First, in the blocking phase, a regulatory PBS solution containing 5% w/v bovine serum albumin and 10% v/v normal goat serum was administered. All samples remained this way for 1 hour at room temperature to allow blocking to be completed. Next, the samples were incubated with Anti‐Von Willebrand antibody (SC‐14014 Santa Cruz).9 The slides were washed in PBS and incubated with secondary anti‐rabbit phycoerythrin (PE) conjugated antibody (sc‐3745, Santa Cruz).10 At the next stage, DAPI ProLong Gold anti‐fade reagent (Invitrogen) was administered, and the samples were observed with an electronic optical microscope (Axiovert CFL40 Zeiss microscope) equipped with an HBO 50 mercury lamp reflector, with fluorescence filter set for GFP (excitation 488 nm; emission 517 nm) and PE (excitation 565 nm; emission 575 nm). Images were acquired with the Fluorescence Lite software module of AxioVision LE (Carl Zeiss). PE fluoresces red, DAPI, which colours the cell nuclei, fluoresces blue, and the GFP plasmid fluoresces green. The immunohistochemical study was performed to determine if the injected GFP stem cells had been converted in vivo into endothelial cells. In this case, vascular formations would be observed as red fluorescence turned into green fluorescence with special filters.

2.8. Statistical analysis

Data normality was determined with histograms, Q‐Q plots, and the Shapiro test. The paired t‐test was used to compare the mean values. A P‐value less than 5% (P < .01) was considered to be statistically significant. All analyses were performed with the statistical package SPSS 17.0.

3. RESULTS

Because the data were approximately normally distributed, the mean necroses of the flaps were compared. The right‐sided flaps (i.e., the flaps where ADSCs were injected) showed lower rates of necrosis compared with the left‐sided flaps (n = 20) (Figure 5). This difference was statistically significant (P < .01, Table 1).

Figure 5.

Image of flaps 1 week after GFP‐ADSCs injection into the right‐sided skin flaps. Skin flaps injected with GFP‐ADSCs showed a significant improvement of tissue survival

Table 1.

Necrosis in both groups (left‐sided and right‐sided flaps)

Outcome	Left‐sided flaps	Right‐sided flaps	P‐value
Necrosis surface (cm2)	6,9 (4,2)	3.1 (2,8)	<.01
Necrosis surface (%)	43 (26)	19 (18)	<.01
Necrosis surface (range, cm2)	0‐14	0‐8	‐

Note: All data are approximately normally distributed and therefore expressed in mean (deviation) form. Statistical significance: P‐value < .01.

Fluorescence microscopy shows green vascular formations, suggesting that autologous GFP stromal cells were converted into endothelial cells through neovascularization (Figure 6A‐D). In the control skin flaps, where no stromal cells were used, no fluorescence was observed.

Figure 6.

A‐D, Immunohistochemical control under fluorescence microscopy (scale bar = 150 μM). A, Image depicting a red fluorescence formation, expressing the endothelial marker von Willebrand, which indicates a blood vessel. B, Image depicting a green fluorescence formation, indicating formation derived from GFP‐autologous stem cells. C, Image depicting a blue fluorescence formation, indicating cellular structures stained with DAPI anti‐fade reagent. D, Image composition, in which the green formation coincides completely with the red formation, indicating the existence of a new blood vessel

4. DISCUSSION

In this experimental study, we examined the effect of adipose‐derived stem cells on the survival of random ischemic skin flaps. Researchers have tried many times to find methods to deal with flap necrosis, including utilisation of angiogenetic growth factors (e.g., the vascular endothelial growth factor). Despite their advantages, angiogenetic growth factors have major restrictions, mainly because of their short half‐life time.10

The application of stem cells has also been proposed to help flap necrosis with promising results.11 Adipose stromal stem cells have found wide use in many medical disciplines, because they have the advantage of easy isolation and multipotency compared with stem cells of different origins, such as bone marrow stem cells.12

In our study, we used a modified version of the previously described McFarlane ischemic model.8 Two identical random‐pattern skin flaps (2 cm × 8 cm) were elevated on the dorsal area of the spine of 20 Wistar rats. We preferred this ratio, because it has been proven to result in significant tissue necrosis in all random‐pattern skin flaps. The elevation of the flaps was performed bilaterally along the spinal cord at regions 40 and 42, as defined by Krinke.6The selection of symmetrical regions minimised anatomical differences that could lead to biased results. Another advantage of this model was that all rats served as controls for themselves.

In this study, we further modified the aforementioned model by applying a thin silicon film to the wounds before placing the skin flaps back in their original position to minimise the possibility of reperfusion from the underlying blood vessels. Moreover, reperfusion through the side of the wound was not possible, because the choke vessels, after the elevation of the flaps, lost their architecture and became unable to feed them with new vascular connections.

In our model, stem cells were strictly intradermally injected in the middle dermis along the long axis of the right‐sided skin flap, although subcutaneous injection has been used in other studies.13 Equal volumes of cell‐free PBS were injected in each left‐sided flap (the control flap) in order to have the same vascular supply interruption to the injected area by compression, injury, or obstruction of the vessels in comparison with the right‐sided skin flaps.14

ADSCs were identified by the flow cytometric analysis. We also proceeded to a successful in vitro differentiation of those cells in other cell lineages, therefore demonstrating their multipotency.

The positive effect of GFP ADSCs on neoangiogenesis is well known. These cells are found to be part of the endothelial layer of new‐build vessels.8 In our study, we proved that GFP ADSCs have resulted in the creation of whole new blood vessels through neovascularization.

However, a limited number of stable research protocols and other experimental studies demonstrating a possible association of adipose stem cells with carcinogenesis,15 require further investigation until the clinical use of adipose stem cells can be achieved.

5. CONCLUSIONS

The present experimental study showed a statistically significant improvement of skin flap survival after injection with adipose stromal cells. Moreover, as immunochemistry showed, one of the mechanisms of action of ADSCs is the direct in vivo conversion into endothelial cells, contributing to the improvement of ischemic skin flap survival through neovascularization. These results in combination with existing scientific studies emphasise the beneficial impact of ADSCs on ischemic skin flap survival and will encourage future clinical application of ADSCs in reconstructive surgery.

Foroglou P, Demiri E, Koliakos G, Karathanasis V. Autologous administration of adipose stromal cells improves skin flap survival through neovascularization: An experimental study. Int Wound J. 2019;16:1471–1476. 10.1111/iwj.13216