Platelet-Released Growth Factors Induce Differentiation of Primary Keratinocytes

Autologous thrombocyte concentrate lysates, for example, platelet-released growth factors, (PRGFs) or their clinically related formulations (e.g., Vivostat PRF®) came recently into the physicians' focus as they revealed promising effects in regenerative and reparative medicine such as the support of healing of chronic wounds. To elucidate the underlying mechanisms, we analyzed the influence of PRGF and Vivostat PRF on human keratinocyte differentiation in vitro and on epidermal differentiation status of skin wounds in vivo. Therefore, we investigated the expression of early (keratin 1 and keratin 10) and late (transglutaminase-1 and involucrin) differentiation markers. PRGF treatment of primary human keratinocytes decreased keratin 1 and keratin 10 gene expression but induced involucrin and transglutaminase-1 gene expression in an epidermal growth factor receptor- (EGFR-) dependent manner. In concordance with these results, microscopic analyses revealed that PRGF-treated human keratinocytes displayed morphological features typical of keratinocytes undergoing terminal differentiation. In vivo treatment of artificial human wounds with Vivostat PRF revealed a significant induction of involucrin and transglutaminase-1 gene expression. Together, our results indicate that PRGF and Vivostat PRF induce terminal differentiation of primary human keratinocytes. This potential mechanism may contribute to the observed beneficial effects in the treatment of hard-to-heal wounds with autologous thrombocyte concentrate lysates in vivo.


Introduction
The optimal therapy of patients' chronic, hard-to heal wounds is difficult and often not successful. Many patients worldwide experience minor or even major extremity of amputation due to chronic wound complications. In general, treatment options include surgical procedures as well as the application of diverse wound dressings. In the past decade, autologous thrombocyte concentrates lysates, for example, platelet-released growth factors (PRGFs), or their clinically related formulations (e.g., Vivostat PRF) came into the focus of regenerative and reparative medicine because they contain a multitude of chemokines, cytokines, and growth factors and are therefore supposed to support healing of chronic or infected wounds [1][2][3][4][5]. It has been shown that PRGF has the opportunity to stimulate cell proliferation and tissue regeneration, to modify cell and tissue differentiation, and to support angiogenesis [6][7][8][9][10][11][12][13]. Although beneficial clinical effects of a local application of autologous thrombocyte concentrate lysates on the healing of chronic or complicated wounds in vivo have been described [14][15][16], little is known about possible mechanisms involved. Recently, we have shown that platelet-released growth factors induce the antimicrobial peptides human beta-defensin-2 in primary human keratinocytes indicating an enhancement of the epithelial barrier function by PRGF and Vivostat PRF treatment [17]. So far, further investigations on possible mechanisms involved are rare. Therefore, we examined a possible influence of PRGF on human keratinocyte differentiation by analyzing the influence of PRGF on the expression of keratin 1, keratin 10, transglutaminase-1, and involucrin in primary human keratinocytes as indicators of their differentiation status. Keratin 1 and keratin 10 are primarily expressed on keratinocytes of the stratum basale and spinosoum and are therefore used as markers for their early terminal differentiation [18,19]. Involucrin and transglutaminase-1 are mainly expressed on mature human keratinocytes of the stratum granulosum and corneum and are regarded as indicators of the late terminal differentiation of human keratinocytes [20][21][22][23].

Preparation of PRGF.
In general, PRGF used for one in vitro experiment was obtained from a single donor. PRGF was prepared from supernatants of freshly donated human thrombocyte concentrates (Institute of Transfusion Medicine, University of Schleswig-Holstein, Campus Kiel) derived from leucocyte-depleted haemapheresis according to the officially recommended practice (Richtlinien zur Gewinnung von Blut und Blutbestandteilen und zur Anwendung von Blutprodukten (Hämotherapie), Transfusionsgesetz, Bundesärztekammer, 2010). Thrombocyte concentration exceeded 2-4 × 10 11 per concentrate (200-450 ml). It includes less than 1 × 10 6 leucocytes and 3 × 10 9 erythrocytes. To prepare the PRGF, the freshly donated thrombocyte concentrates were centrifuged for 10 minutes at 2000g. Afterwards, the thrombocyte pellet was washed twice with a sodium citrate buffer (0.11 mM, ph 5.5, 37°C) and centrifuged again for 10 min at 2000g. The thrombocytes were resuspended in half the volume of the initial thrombocyte concentrate volume using Keratinocyte Growth Medium 2 (PromoCell, Heidelberg, Germany) without supplements. The resuspended thrombocytes were stored on ice, lysed by ultrasound, and stored at −80°C for 24 hours. The next day, the suspension was thawn again, the ultrasound procedure was repeated, and the suspension stored again at −80°C for 24 hours. After repeated thawing, the suspension was centrifuged for 1 minute at 18.000g. The PRGF is the supernatant which was stored in aliquots at −20°C.

Preparation and Application of Vivostat PRF (Platelet-Rich Fibrin).
To prepare Vivostat PRF, 120 mL fresh blood from one patient was transferred to the so-called preparation unit and subsequently processed by the processor unit to obtain 5-6 mL of the Vivostat PRF product containing approximately 7 times the base level of the donor's blood thrombocyte concentration (>1 million/μL) and fibrin (average concentration of 18.1 mg/mL) (for details, see http:// www.vivostat.com). The Vivostat PRF was sprayed on the wound surface to completely cover the wound area using the Vivostat spraypen® (see also http://www.vivostat.com).
For EGFR-or IL-6 receptor-blocking experiments, we used the EGFR-blocking antibody cetuximab (Merck, Darmstadt, Germany) or the IL-6 receptor-blocking antibody tocilizumab (Hoffmann-La Roche, Basel, Switzerland) at a concentration of 20 μg/mL and 50 μg/mL, respectively. After stimulation, the cells were washed with 1 mL per well of PBS followed by the isolation of the RNA.
2.4. RNA-Isolation and cDNA Synthesis. We harvested keratinocytes from one well of a 12-well plate and lysed them using 500 μL Crystal RNAmagic reagent. Total RNA was isolated according to the supplier's protocol (BioLab Products, Bebensee, Germany). The concentration of isolated total RNA was photometrically determined using a NanoDrop device (Peqlab, Erlangen, Germany). 1 μg of total RNA was reversely transcribed to cDNA using oligo-dT-primers and 50 Units Maxima Reverse Transcriptase (Thermo Fisher Scientific, Waltham, USA) according to the manufacturer's protocol.

Analyses of the Influence of Vivostat PRF on the Epidermal
Expression of Keratin 1, Keratin 10, Transglutaminase-1, and Involucrin In Vivo. This study was conducted with human in vivo samples that have been already used in a previous study [17]. Briefly, we set bilateral gluteal wounds in five male test persons by punch biopsy (Ø 4 mm) after local anesthesia and treated the freshly generated wounds with either NaCl 0.9% (left) or Vivostat PRF (right) followed by the application of occlusive wound dressings (Biatain®, Coloplast, Germany) on the treated wounds. After 5 days, we removed the dressing and the dried wound exudate by a sterile moistened compress and repeated the treatment as mentioned above. After 10 days, dressings were removed and bilateral wound areas were resected by punch biopsies (Ø 6 mm). RNA was isolated using RNAeasyKit (Qiagen, Hilden, Germany), and reverse transcription of the RNA and real-time PCR was performed as described above. This pilot study was approved by the University Committee for Ethical Affairs Kiel (AZ A 115/13) in accordance with the Helsinki guidelines. All participants included in this investigation provided written informed consent.
2.7. Statistics. GraphPad Prism 6.07 was used for statistical analysis and was carried out by Student's t-test or one-way ANOVA with Tukey's multiple comparison test. A P value of <0.05 was considered statistically significant.

The Epidermal Growth Factor Receptor (EGFR) Influences the Expression of Differentiation Markers in
Primary Human Keratinocytes Treated with PRGF. To examine the underlying signal transduction pathways of the PRGF-mediated keratin 1, keratin 10, transglutaminase-1, and involucrin gene expression in primary human keratinocytes, we analyzed the influence of the epidermal growth factor receptor (EGFR) using a specific monoclonal EGFRblocking antibody (cetuximab). Treatment of the keratinocytes with cetuximab significantly induced the gene expression of keratin 1 (Figure 4(a)) and keratin 10 ( Figure 4(b)), but this induction was inhibited by PRGF. In contrast, the inhibition of the EGFR using cetuximab did not induce transglutaminase-1 and involucrin expression but resulted in a significant decrease of PRGF-induced transglutaminase-1 (Figure 4(c)) and involucrin (Figure 4(d)) gene expression.

The Influence of PRGF on the Expression of Differentiation Markers Does Not Involve the Interleukin-6
Receptor (IL-6R). We have recently shown that PRGF treatment induced IL-6 in primary human keratinocytes already after 4 hours. In addition, the PRGF-mediated induction of the antimicrobial peptide hBD-2 partially depended on IL-6 signaling [17]. Therefore, we sought to determine whether IL-6 signaling plays a role in the expression of differentiation markers in keratinocytes treated with PRGF. To this end, we blocked the IL-6 receptor with the IL-6 receptor-neutralizing antibody tocilizumab and stimulated the keratinocytes with PRGF. These experiments revealed a significant decrease of keratin 1 (Figure 5(a)) and keratin 10 ( Figure 5

Treatment of Human Cutaneous Wounds with Vivostat PRF Induced the Gene Expression of Involucrin and
Transglutaminase-1 In Vivo. To investigate if our in vitro data could be transferred into the in vivo setting, we performed the above mentioned human in vivo study. This study revealed a significant transglutaminase-1 (Figure 7(c)) and involucrin (Figure 7(d)) gene induction in the wounds treated with Vivostat PRF. Keratin 1 (Figure 7(a)) and keratin 10 (Figure 7(b)) gene expression was slightly but not significantly reduced by Vivostat PRF treatment.

Discussion
Chronic lower extremity skin ulcers affect about 3% of the population in Western countries and cause an immense personal, financial (approximately 2.5% of total healthcare budgets in Europe and America), and social burden [25]. Despite optimal causal and topical therapy, wound healing     Figure 5: The IL-6 receptor is not involved in the PRGF-mediated modulation of differentiation markers in keratinocytes. PRGF stimulation of primary human keratinocytes caused a reduced keratin 1 and keratin 10 gene expression and an induced transglutaminase-1 and involucrin gene expression. Costimulation with the IL-6 receptor-blocking antibody tocilizumab revealed that these effects were not mediated via the IL-6 receptor (real-time PCR analysis, * P < 0 05, * * P < 0 01, * * * P < 0 001, ns = not significant, one-way ANOVA with Tukey's multiple comparison test).
remains often difficult and unsuccessful. One new therapeutic option for the therapy of chronic wounds is the local application of an autologous thrombocyte concentrate lysate (e.g., Vivostat PRF). In general, thrombocytes play a major role during the initial uncomplicated wound healing process through their release of a multitude of cytokines, chemokines, and growth factors that initiate and coordinate the complex wound healing process [1][2][3][4][5]. For this reason, thrombocyte concentrate lysates are supposed to have the opportunity to optimize the unfavorable local wound milieu and support wound healing [26][27][28]. Due to their regenerative, reparative, and angiogenetic potential, thrombocyte concentrate lysates are recently used in many medical disciplines [9,26,[29][30][31][32][33]. In the context of chronic lower extremity ulcers, Steenvoorde et al. described the use of autologous platelet-rich fibrin on a range of hard-to-heal wounds which caused complete wound healing or a significant reduction in wound diameter in the majority of treated patients [14]. Despite these positive clinical experiences with Vivostat PRF for the therapy of chronic or complicated wounds, little is known about possible mechanisms involved [12,[26][27][28].
Recently, we have demonstrated that thrombocyte concentrate lysates induce the antimicrobial peptide human beta-defensin-2 in keratinocytes [17] indicating an improved cutaneous innate defense function as one explanation for the observed positive clinical effects of autologous thrombocyte concentrate lysates on wound healing. We now aimed to investigate if thrombocyte concentrate lysates may also directly affect the formation of the epidermal barrier. In particular, we were interested to evaluate a potential influence on the epidermal differentiation process. Therefore, we analyzed gene expression patterns of early differentiation markers (keratin 1 and keratin 10) and late differentiation markers (transglutaminase-1 and involucrin) in keratinocytes [18][19][20][21][22][23]34]. We observed that in vitro treatment of primary keratinocytes with PRGF pronounced late differentiation processes in primary human keratinocytes as documented by a PRGF-mediated decrease of keratin 1 and keratin 10 gene expression paralleled by an increase of transglutaminase-1 and involucrin gene expression. These effects were concentration dependent and time dependent with maximal effects seen after 12-24 hours of stimulation which is in concordance with data from Liew and Yamanishi [35] reporting that transglutaminase-1 gene expression peaked after 16 hours of 12-O-tetradecanoylphorbol-13-acetate-(TPA-) induced differentiation in keratinocytes. Thus, one may speculate that gene expression of transglutaminase-1 is induced during the differentiation process and decreased when the terminal differentiation phase is reached.
Accordingly, we observed in our in vivo study a significant gene induction of involucrin and transglutaminase-1 in Vivostat PRF-treated skin, thus translating the in vitro findings in the in vivo setting. In concordance with these cell culture and in vivo data, microscopic analyses revealed that PRGF-treated human keratinocytes-unlike untreated keratinocytes-displayed a morphology typical for late terminal differentiated keratinocytes. Taken together, these experiments identified PRGF and Vivostat PRF as a potent inducer of human keratinocyte differentiation.
To investigate possible mechanisms involved, we analyzed the underlying signal transduction pathways. As keratinocyte differentiation was demonstrated to be EGFR dependent [36][37][38], we analyzed the influence of the EGFR on the PRGF-mediated alteration of keratin 1, keratin 10, transglutaminase-1, and involucrin gene expression in human keratinocytes by blocking the EGFR via a monoclonal antibody (cetuximab). These studies demonstrated that the PRGF-mediated decrease of keratin 1 and keratin 10 gene expression as well as the increase of transglutaminase-1 and involucrin gene expression in keratinocytes-and thus the PRGF-induced keratinocytes differentiation-were predominantly EGFR mediated. In line with these results, it has been reported that the EGFR ligand HB-EGF, which is upregulated in the margin of skin wounds, induced involucrin gene expression in keratinocytes, whereas keratin 10 gene expression was downregulated [39]. These data suggest that EGFR ligands present in PRGF and Vivostat PRF activate the EGFR in keratinocytes leading to the observed involucrin and decreased keratin 10 expression. It has been reported that EGFR activation and signaling improve the integrity of the skin barrier by enhancing terminal keratinocyte differentiation and the cross-linking activity of transglutaminases [37]. Thus, PRGF may facilitate the reconstitution of an intact skin barrier by promoting terminal differentiation of the keratinocytes.
In former experiments, we detected a strong expression of IL-6 in primary human keratinocytes already after 4 hours of PRGF treatment [17]. In addition, IL-6 has been reported to negatively affect terminal keratinocytes differentiation [40]. Therefore, we asked if IL-6 may have an influence on the observed PRGF-mediated induction of keratinocyte differentiation. To address this question, we used tocilizumab, a monoclonal antibody directed against the IL-6-receptor (IL-6R), to inhibit the IL-6 signaling pathway. However, tocilizumab had no significant influence on the expression of differentiation markers in PRGF-treated keratinocytes indicating that IL-6 signaling plays no role in this process.

Conclusion
We demonstrated that PRGF stimulation of primary keratinocytes and Vivostat PRF application on artificially generated human skin wounds cause an accelerated differentiation process of primary human keratinocytes that may-in addition to the described induction of antimicrobial peptides [17]-contribute to the observed beneficial effects in the treatment of hard-to-heal wounds with autologous thrombocyte concentrate lysates in vivo.