Comparative analysis of liquid and gel platelet rich plasma from apheresis and buffy coat in wound healing

Abstract

Platelet-Rich Plasma (PRP) is widely used in wound care and tissue regeneration, yet variability in its preparation and delivery has hindered clinical standardization. This study compares the regenerative potential of PRP formulations derived from apheresis (Aphe PRP) and buffy-coat (BC PRP), in both liquid and gel forms. We characterized the seven-day release profile of platelet-derived effectors and assessed the impact of PRP on key wound-healing cellular processes—proliferation, migration, and differentiation—using keratinocytes, fibroblasts, and microvascular endothelial cell models. Our findings revealed a sustained effector release over seven days, with liquid PRPs showing significantly higher release levels than gels, regardless of PRP source. Aphe and BC PRPs exhibited comparable effects on cell proliferation, migration, and differentiation. However, gel PRPs were less effective than liquid PRPs in promoting proliferation and migration, although both forms significantly enhanced regeneration compared to untreated controls. Notably, differentiation effects were similar across all PRP types. These results suggest Aphe and BC PRPs are interchangeable for wound care, with liquid PRPs offering superior regenerative effects in early treatment stages. This study highlights the need for standardized PRP preparation and application protocols to optimize therapeutic outcomes.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-30402-w.

Introduction

The wound healing process entails the recruitment and activation of various cell types, such as immune cells, fibroblasts, and endothelial cells, all of which contribute to the wound-healing cascade. Initially, immune cells clear pathogens and debris. This is followed by the activation of fibroblasts, the deposition of the extracellular matrix, and re-epithelialization. These processes culminate in tissue remodelling and the restoration of structural integrity. Among the key mediators of tissue regeneration, platelets play a pivotal role in coordinating the complex wound-healing process. They not only initiate hemostasis but also release numerous bioactive molecules, including growth factors (GFs), cytokines, chemokines and extracellular vesicles (EVs), that are crucial for the progression of wound healing through phases of inflammation, cell proliferation, and tissue remodeling¹. GFs such as platelet-derived growth factor (PDGF), transforming growth factor-beta (TGF-β), vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and connective tissue growth factor (CTGF), are essential for regulating cellular processes such as chemotaxis, proliferation, angiogenesis, and matrix synthesis. These proteins act in an autocrine and paracrine manner to recruit immune cells, stimulate fibroblast and keratinocyte activity, and promote re-epithelialization and neovascularization.

In parallel, EVs, which include exosomes and microvesicles, are lipid bilayer-enclosed particles released by activated platelets. They carry a complex cargo of proteins, RNAs (including mRNAs and miRNAs), and lipids, and function as intercellular communicators. EVs can modulate the behavior of target cells by transferring their molecular content, thereby influencing immune modulation, angiogenesis, and matrix remodeling. Their contribution is particularly relevant in the fine-tuning of the wound microenvironment, and they are increasingly recognized as bioactive effectors in regenerative medicine.

Together, this molecular repertoire enables platelets to act as central regulators of tissue repair, bridging innate immune responses and regenerative pathways in both physiological and therapeutic contexts.

Due to their role in modulating the wound-healing microenvironment by releasing bioactive molecules, platelets have been extensively utilized in the form of platelet-derived concentrates for wound treatment. Among these, Platelet-Rich Plasma (PRP) has emerged as a promising therapeutic option in regenerative medicine, with applications across various fields, including dermatological^2,3 conditions, maxillofacial surgery⁴, orthopedic surgery⁵, ocular surface diseases⁶ and some autoimmune diseases⁷. PRP is a blood component characterized by a high concentration of platelets, GFs, EVs and bioactive molecules within a small volume of plasma. The use of PRP in dermatology for alopecia, scarring, acne and skin rejuvenation is attracting increasing interest⁸ while novel applications, such as radio dermatitis and rheumatological skin diseases, have been recently proposed^9–11. Most importantly, PRP has also been largely used in wound care for chronic and poorly healing wounds, such as diabetic ulcers, venous leg ulcers, pressure sores, and post-surgical wounds that fail to progress through the normal phases of healing^2,12. This use is both accepted by the scientific community¹³ and described in clinical guidelines^14–16, and is becoming a valid therapeutic option. The GFs and cytokines released by PRP aid in promoting the re-vascularisation of damaged tissue by inducing the reorganization of endothelial cells into new blood vessels. They also activate fibroblasts to promote fibrogenesis, restore the damaged extracellular matrix and support the proliferation and migration of keratinocytes^17,18. Most of the literature studies agree with the safety and efficacy of PRP treatment of non-healing wounds of various aetiologies, particularly diabetic ulcers, where earlier and more complete wound closure has been observed^2,12,19. In vitro studies have shown that PRP significantly increases the migration ability and proliferation rate of mesenchymal stem cells and fibroblasts²⁰. Despite these evidences, the diffusion of PRP in the treatment of chronic wounds is still hampered by the heterogeneity of PRP preparation and administration protocols (often unspecified or ambiguous), highlighting the need for standardisation for clinical practice^21–23.

The term PRP refers to a range of blood components that vary in the composition of their regenerative effectors, due to differences in key methodological steps²³. Depending on the method of preparation different PRP products have been described (for instance pure PRP, leucocyte-PRP, platelet rich in fibrin, PRF, Freeze–Thaw PRP)^1,24. Even within the same PRP subgroup, numerous variables can influence the quality and efficacy of the product. These variables include the method of platelet collection (e.g., buffy-coat or platelet apheresis) as well as differences in the phlebotomy protocol, centrifugation steps, freeze-thawing processes and activation techniques. The absence of standardized protocols, individual donor variability, and the heterogeneity of preparation methods can result in significant variations in PRP composition, ultimately leading to unpredictable treatment outcomes.

Donor characteristics are also crucial, as factors such as age, sex, underlying health conditions, and medications can affect platelet content, thereby influencing the quality of the PRP²⁵.

Allogeneic PRP products provide greater standardization in terms of donor characteristics and preparation procedures. Although autologous PRP is commonly used in most studies and clinical applications, obtaining autologous blood can impose a greater burden on patients, particularly those who are critically ill or elderly. Further, patient medical conditions can have a significant impact on the quality of PRP^26–28. In this respect, the use of platelet concentrates as prepared by blood banks under Good Manufacturing Practice (GMP) might result in more defined compositions, with minimal erythrocyte and leukocyte contamination²⁹.

Among allogeneic products, buffy-coat-derived PRP (BC PRP) is obtained from pooled platelets collected from 4–5 different donors, which helps reduce inter-individual variability. However, pooling platelets could trigger immune responses, increase the risk of transmitting infectious agents, prions, or triggering an allogeneic immune reaction^11,25, especially if leukocytes or plasma proteins are not adequately removed. In contrast, apheresis-derived PRP (Aphe PRP) is typically processed to reduce immunogenic components, but its collection requires a more complex and invasive procedure, for which only a limited number of donors are eligible or willing to participate³⁰.

These two types of allogeneic PRP products have not been directly compared in studies, nor has it been thoroughly investigated whether and how the activation process of PRP, leading to PRP-gel formation, may impact the functional properties of PRP.

The aim of this study is to characterise and compare the wound healing and regenerative potential of BC and Aphe PRP, both liquid and gel, using in vitro models of human keratinocytes, fibroblasts and microvascular endothelial cells. We evaluated the release kinetics of GFs and EVs from the PRPs over a 7-days period and investigated their effect on cellular processes critical to tissue regeneration on cellular models involved in tissue regeneration including proliferation, migration and differentiation in fibroblasts, keratinocytes and microvascular endothelial cells.

Results

PRPs were obtained from Platelet-apheresis (Aphe PRP) and buffy-coat (BC PRP) as summarized in Fig. 1A and B, respectively; the detailed methodological procedures are described in the Material and Methods section. Liquid and gel PRPs were compared in in vitro studies as summarized in Fig. 1C.

Fig. 1.

Schematic representation of: (A) Procedure to obtain PRPs from a pool of platelet concentrate form Apheresis (Aphe PRP). (B) Procedure to obtain PRPs from a pool of platelet concentrate from Buffy-Coat (BC PRP). (C) Experimental setting for releasate biochemical characterization and cell culture experiments. Image created with Biorender.

GFs, calcium, and citrate quantification in Aphe and BC PRP

GFs are known to promote tissue regeneration, stimulating proliferation, migration and differentiation. EGF, FGF2, CTGF, TGF-β1, PDGF-AA and PDGF-AB were quantified in the Aphe PRPs and BC PRPs: there were no significant differences in GFs concentrations between Aphe and BC PRPs (Table 1). The calcium concentration in Aphe PRP was 1.66 mM and 1.53 mM in BC PRP. Citrate concentration in Aphe PRP was 33.3 µM and 28.5 µM in BC PRP.

Table 1.

GFs concentrations in Aphe and BC PRPs (n = 2).

	Aphe PRP		BC PRP		p-value
	Mean	StDev	Mean	StDev
EGF (pg/ml)	429.9	42.3	437.3	20.4	ns
FGF-2 (pg/ml)	92.9	4.4	86.2	2.3	ns
CTGF (ng/ml)	177.8	26.1	176.9	22.2	ns
TGF-β1 (ng/ml)	195.6	2.3	187.6	3.3	ns
PDGF-AA (ng/ml)	872.4	25	832.7	32	ns
PDGF-AB (ng/ml)	16.9	0.6	18.0	0.9	ns

Release over time of GFs, EVs and calcium from liquid PRP

Over time release of GFs, EVs and Calcium (Ca²⁺) were measured in liquid BC and Aphe PRPs. The release of GFs was monitored and quantified over time at several time points, every 24 h for a total of 7 days (Fig. 2A). GFs release was normalized to the total amount of each GF in the corresponding PRP source (Supplementary Table 1), and expressed as a percentage of it. When comparing the release of GFs over time between liquid and gel PRP (7 days), we observed different trends for these two products (Fig. 2B and supplementary Table 1).

Fig. 2.

Release of Growth Factors (GFs), Extracellular Vesicles (EVs) and Calcium (Ca²⁺) from Aphe and BC liquid and gel PRP. (A) Schematic representation of the experimental design. (B) EGF, CTGF, TGF-β1, PDGF-AA, PDGF-AB were normalized for PRPs GFs concentrations and expressed as a % of them. (C) Number of EVs released. (D) Ca²⁺ concentration [mM]. Data (mean ± SD) are shown over 7 time points (days 1–7) for four different PRP preparations: Aphe liquid (red dots), BC liquid (blue dots), Aphe gel (pink dots) and BC gel (light blue dots). GFs were measured with ELISA, EVs release over time was evaluated by EVs quantification kit and flow cytometry, and calcium was quantified with a colorimetric calcium assay. Quantification was carried out on PRP releasates collected from the lower chamber of the Transwell insert every 24 h over a period of 7-days.

FGF2 was always under the limit of detection for both types of products. In liquid PRP, the highest percentage of release for EGF and CTGF were observed at the first time point. The level of FGF2 in PRPs was consistent with previous publications³¹. However, the limit of detection of the assay (10 pg/ml) may have prevented the detection of FGF2 in the releasate, as this corresponds to approximately 10% of the daily amount of FGF2 released.

EGF dropped to less than 50% of its initial concentration while CTGF declined to 30% of their starting values. In contrast TGF-β1 levels fluctuated over time, showing an overall increasing trend. PDGF-AA and -AB showed different trends when comparing Aphe and BC products, PDGF-AA decreased to a 30% of its maximum releasing concentration at day 7 although BC PRP has a peak at day 1 and Aphe PRP at day 4. PDGF-AB decreased to 50–60% of its maximum concentration at day 7, although Aphe PRP has a releasing peak at day 3 while BC PRP at day 1.

Intriguingly the total released amount over the 7 days period varied among the different GFs: EGF and CTGF were released at 80–100%, PDGF-AA and -AB at 15–25%, while TGF-β1 was released at 27 and 45% respectively for Aphe and BC PRP (Supplementary Table 1).

In the releasate of PRP gel, GFs were lower than in liquid PRP, often by an order of magnitude or more. However, GFs release from gel PRP increased gradually over time, predominantly during the final 3–4 days, with the exception of EGF, which remained stable throughout. Notably, in BC PRP, the release of PDGF-AA and -AB occurred earlier compared to Aphe PRP. TGF-β1 in gel PRPs started to be released after 6 days and at a higher rate in BC PRP.

In parallel, we quantified the release of EVs, which are fragments derived from the disrupted platelets, and play a critical role in tissue regeneration, transporting GFs and other cellular effectors³² (Fig. 2C). Similar to the release pattern observed for GFs, EVs release was higher in liquid PRP compared to gel PRP. In gel PRP, EVs release remained near the detection limit until day 6, with a notable increase observed only on day 7. By contrast, in liquid PRP, EVs release steadily increased over the various time points.

In addition, we measured Ca²⁺ release (Fig. 2D), given its pivotal function in various cellular processes, including those related to cell–cell junctions and keratinocytes differentiation³³.

The Ca²⁺ release profile exhibited a peak within the initial two days, followed by a gradual decline over the subsequent five days. A comparable trend was observed between Aphe and BC PRP. However, as anticipated, Ca²⁺ release exhibited higher levels for the PRP gel, given that this ion, in conjunction with thrombin, is employed during the production phases (see Fig. 1C and Materials and Methods) to induce platelet coagulation and, consequently, PRP gel formation. In liquid PRP, Ca²⁺ release predominantly occurred within the first four days.

Furthermore, the Ca²⁺ concentration measured on day 5 was near the limit of detection, further supporting the conclusion that most of the release had already taken place.

Effect of PRPs on BJ fibroblast cells behaviour

Wound healing is a dynamic, multistage process involving cell proliferation and tissue remodelling. Fibroblasts migrate into the wound in response to various soluble mediators initially released by platelets. Inflammatory cytokines such as TGF-β and PDGF stimulate fibroblast proliferation, migration to the wound site, and induce phenotypic changes that transform fibroblasts into myofibroblasts³⁴. As demonstrated in in vitro cultures of fibroblast, the process of differentiation of fibroblast to myofibroblast is induced by TGF-β in the wound microenvironment^35–37 with increased levels of ACTA2 (αSMA) as a marker of myofibroblast differentiation^35,38. Accordingly, a significant increase (Fold change, FC, 9.6 and 6.1) of αSMA expression was observed in BJ cells in the FBS 0.1% condition, used as untreated control, at both 48 and 72 h after treatment (Fig. 3A). Therefore, the subsequent PRPs treatments were performed after 72 h of stimulation with TGF-β.

Fig. 3.

Effect of PRPs treatments on BJ cells in the presence (+ TGF-β) or absence (-TGF-β) of TGF-β. TGF-β was added in the culture media at the concentration of 10 ng/ml for 24 h before the addition of PRP treatments to the wells. (A) Gene Expression levels the myofibroblast differentiation markers ACTA2 (αSMA) in the presence (+ TGF-β) or absence (-TGF-β) at different time points in FBS 0.1%, significant differences were indicated for the comparison between –TGF-β and + TGF-β at the same time point. (B) Schematic representation of the experimental design. (C) Proliferation. WST-1 assay was performed after 72 h of PRPs treatment. n ≥ 3 replicates. Data are expressed as % of 10% FBS control. (D) Migration. Number of BJ cells migrated into a scratch test after 48 h of treatment of with PRPs. Data are obtained by constructing a reference area and counting the number of cells migrating within the area over time. n ≥ 3 replicates. (E) Circularity. A cellular shape descriptor, circularity, was measured on n = 14 cells/condition. Circularity is in a range between 1 and 0, where 1 identifies completely circular cells, and 0 sharp cells. (F) Representative phase-contrast images of BJ cellular shape. Images are collected after 72 h of PRPs treatment. White arrows indicate vortex structures of polarized cells. Scale bar 100 μm. (G) Gene Expression levels of specific myofibroblast differentiation markers. ACTA2 (αSMA), COL1A1, VIM and MMP1 expression (2^-ΔCt) in BJ fibroblasts evaluated by qPCR after 72 h of treatment with PRP. n ≥ 3 replicates. Different bar colours indicate the following conditions: black for 10% FBS, grey for 0.1% FBS, red for Aphe liquid , blue for BC liquid , pink for Aphe gel and light blue for BC gel. Data are expressed as mean ± SD. Significance of data differences was established using unpaired two-ways ANOVA test with multiple comparisons. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001. ## indicate a p-value < 0.01 between the FBS 0.1% untreated conditions of the TGFβ- and TGFβ + settings.

To investigate how soluble mediators released by PRPs influence fibroblast behaviour, BJ cells³⁹ were used as a model to assess the effects of PRP on fibroblast proliferation, migration, morphology and differentiation (see scheme on Fig. 3B).

WST-1 assay was performed to evaluate BJ cell proliferation (n ≥ 3 replicates, Fig. 3C). After 24 h from seeding, cells were cultured or not with TGF-β and then treated with PRPs.

BJ cells untreated controls viability was significantly reduced when compared to the 10% FBS standard growing control. The absorbance values shown in Fig. 3C indicate that PRP treatments significantly enhance BJ cell proliferation compared to untreated controls, with the exception of Aphe PRP gel. Growth stimulation was significantly higher in BJ cells treated with liquid PRP, compared to gel formulations, regardless of whether the PRP was derived from apheresis or buffy coats sources (WST-1 Absorbance values: FBS 10% 0.46 ± 0.10, FBS 0.1% 0.17 ± 0.06, Aphe liquid 0.50 ± 0.08, Aphe gel 0.28 ± 0.07, BC liquid 0.54 ± 0.09, BC gel 0.30 ± 0.05). The pro-proliferative effect of PRP was further validated in the presence of TGF-β (+ TGF-β), where all the tested PRP treatments increased cellular proliferation. Liquid PRP formulations showed a significantly greater effect compared to gel formulations (WST-1 Absorbance values: FBS 0.1% 0.19 ± 0.03, Aphe liquid 0.52 ± 0.009, Aphe gel 0.30 ± 0.04, BC liquid 0.53 ± 0.07, BC gel 0.30 ± 0.04) (Fig. 3C).

To evaluate the potential of different PRP preparations to facilitate the fibroblast-mediated wound healing, a wound scratch assay was performed comparing the same PRP treatments performed for the proliferation assay. Migration in FBS 0.1% was not lower than in FBS 10%, suggesting that the effect of FBS supplementation acts mainly on proliferation. After 48 h, none of the PRP treatments significantly influenced cell migration compared to the controls (Fig. 3D). Representative phase-contrast images of the scratch are reported in Supplementary Figure 1. The wound closure rates for both liquid and gel PRPs were similar to those of the untreated control (Number of migrating cells: FBS 10% 144.0 ± 5.2, FBS 0.1% 128.2 ± 21.1, Aphe liquid 167.2 ± 32.0, Aphe gel 164.5 ± 22.5, BC liquid 186.2 ± 81.3, BC gel 154.0 ± 22.9, ≥ 3 replicates). Similar results were observed for the + TGF-β condition (Fig. 3D). Although not significant, BC liquid seemed to induce greater effect compared to other PRP treatments (Number of migrating cells: FBS 0.1% 103.0 ± 30.6, Aphe liquid 106.0 ± 35.4, Aphe gel 139.0 ± 11.8, BC liquid 221.7 ± 102.4, BC gel 114.0 ± 26.2, n ≥ 3 replicates).

Since the effect of both TGF- β and PRP was negligible on BJ cell migration, we assessed how PRP treatment affects cell morphology in the presence or not of TGF-β. Cellular shape changes were quantified using circularity measurement^40,41 (n ≥ 14 cells for each condition), where low circularity levels indicate more elongated/spindle shape cells. Fibroblast circularity was not significantly different when comparing the + TGF β and –TGF-β (untreated) conditions, although both showed a decreased circularity when compared with the FBS10% culture. However, BJ cells became significantly more elongated when PRPs were added in TGF-β treated cells compared to -TGF-β cells, demonstrating a cumulative effect of PRPs and TGF-β (Fig. 3E). Circularity values for the − TGF-β were FBS10% 0.32 ± 0.05, FBS 0.1% 0.23 ± 0.06, Aphe liquid 0.27 ± 0.06, Aphe gel 0.26 ± 0.06, BC liquid 0.19 ± 0.05, BC gel 0.22 ± 0.09, for the + TGF-β were FBS 0.1% 0.20 ± 0.06, Aphe liquid 0.14 ± 0.04, Aphe gel 0.13 ± 0.06, BC liquid 0.13 ± 0.05, BC gel 0.10 ± 0.04.

Representative phase-contrast images (Fig. 3F) showed how liquid PRPs in co-treatment with TGF-β (lower line, (+ TGF-β) induced BJ to produce vortex structures (white row), suggesting an influence on cell polarization and a shift towards apico-basal polarity⁴².

Among the key genes identified as marker of the differentiation of fibroblast into myofibroblast^35,38, we analysed the expression of αSMA, COL1A1, VIM and MMP1 after 3 days of treatment with PRPs (Fig. 3G). Typically, αSMA, COL1A1 and VIM are expected to increase during with myofibroblast differentiation, while MMP1 is expected to decrease^35,38.

Our analysis revealed that in the untreated control (FBS 0.1%), the expression of α-SMA, COL1A1 and VIM remained comparable, regardless of pre-treatment with TGF-β1. However, MMP1 was significantly downregulated in the + TGF-β untreated condition (FBS 0.1%, Fold change, FC, 1.3, p-value < 0.01). Conversely, COL1A1 decreased and MMP1 increased in 0.1% FBS when compared with 10% FBS, in both − and + TGF-β conditions, suggesting that this effect is caused by the FBS concentration and linked to cell viability. VIM increased instead only in the + TGF-β, which might thus be the cause of this gene expression difference (COL1A1: FBS 10% vs FBS 0.1% FC 1.5, p-value < 0.001, FBS 10% vs FBS 0.1% + TGF-β FC 1.7, p-value < 0.001. VIM: FBS 10% vs FBS 0.1% + TGF-β FC 0.5, p-value < 0.05. MMP1: FBS 10% vs FBS 0.1% FC 0.8, p-value < 0.001, FBS 10% vs FBS 0.1% + TGF-β FC 0.1, p-value < 0.001. ).

Interestingly, in the absence of TGF-β, some PRP conditions were able to modify gene expressions related with myofibroblast differentiation. For example, COL1A1 increased following treatment with PRP gels (both Aphe and BC), with fold changes of 2.6 and 2.0, respectively (p-values < 0.001).

Similarly, MMP1 increased in both BC liquid PRP and Aphe gel PRPs (FC = 1.7 for both, p-values < 0.01), and αSMA decreased in Aphe liquid PRP treatment (FC = 1.7, p-value < 0.05), which is contrary to the typical pattern seen in the myofibroblast differentiation.

Nevertheless, in the presence of TGF-β, PRPs treatment modified gene expression with a pattern consistent with myofibroblast differentiation. In particular, PRP gels (both Aphe and BC) significantly increased the expression of αSMA compared to the untreated control (FBS 0.1%), with a fold change of 3 and 10, respectively (p-values < 0.001). Similarly, COL1A1 expression increased across all PRP conditions, with a more pronounced effect observed in the PRP gels (FC for Aphe liq, BC liq, Aphe gel and BC gel were 6.5, 4.8, 9.6 and 17.1 respectively, p-values < 0.001). VIM expression was also more elevated in the PRP-treated cells, although it was significantly different from the untreated control only in the case of Aphe PRPs (FC for Aphe liq and gel were 1.78 and 1.74, respectively, p-values < 0.01).

MMP1 decreased in all the conditions, with the most substantial reductions seen in Aphe liquid and BC gel PRPs (FC for Aphe liq, BC liq, Aphe gel and BC gel of 0.26, 0.78, 0.86 and 0.36, p-values respectively of < 0.001, < 0.001, < 0.05 and < 0.001).

Effect of PRPs on HaCaT keratinocytes cells behaviour

Keratinocytes proliferation and migration are crucial in wound healing, and various kinds of GFs released by PRP, such as EGF or TGF-β, are involved in the regulation of human keratinocyte function⁴³.

The immortalized human keratinocyte cell line HaCaT has been largely used as a model to study the response of keratinocytes to treatments, including PRP^44–46. HaCaT cells can switch between a differentiated and basal state based on Ca²⁺ concentration in the medium⁴⁷. We investigated the effects of PRPs on HaCaT cells cultured in both standard (DMEM) and low Ca²⁺ DMEM with limited proliferative stimuli (FBS 0.1%) to evaluate cellular proliferation, migration, morphology and differentiation (Fig. 4A). To evaluate proliferation, untreated controls (FBS 0.1%) were compared to PRP treatments (n ≥ 3), FBS 0.1% controls (both in low and high Ca²⁺) had a decreased viability when compared to 10% FBS. In standard DMEM culture conditions, PRPs stimulated HaCaT proliferation, although this increase was not significant for BC PRP gel. Liquid PRPs had a stronger effect than gel PRPs. Under low Ca²⁺ conditions, proliferation for the untreated cells was significantly lower compared to standard DMEM cultured cells (#, p-value < 0.05, Fig. 4B) although PRP treatments significantly increased HaCaT proliferation, with a greater effect measured for gel PRPs (Fig. 4B). WST-1 absorbance values were as follows: FBS 10% 0.94 ± 0.06, Normal Ca²⁺ (FBS 0.1% 0.49 ± 0.18, Aphe liquid 1.48 ± 0.12, Aphe gel 1.02 ± 0.17, BC liquid 1.45 ± 0.21, BC gel 0.67 ± 0.15); Low Ca²⁺ (FBS 0.1% 0.21 ± 0.02, Aphe liquid 0.44 ± 0.10, Aphe gel 0.60 ± 0.02, BC liquid 0.43 ± 0.07, BC gel 0.64 ± 0.03). While PRPs stimulated cellular growth in low Ca²⁺ conditions, levels did not reach those observed in the standard medium (DMEM), while in high Ca²⁺ cell viability was comparable to viability in FBS 10% condition.

Fig. 4.

Effect of PRPs treatments on HaCaT cells in normal DMEM and Low Ca²⁺ medium (A) Schematic representation of the experimental design with HaCaT. (B) Proliferation. WST-1 assay evaluated after 72 h of PRPs treatment. n ≥ 3 replicates. Data are expressed as % of 10% FBS control. (C) Migration. Number of HaCaT cells migrated into a scratch test after 72 h of treatment with PRPs. Data are obtained by constructing a reference area and counting the number of cells migrating within the area over time. n ≥ 3 replicates. (D) Circularity. A cellular shape descriptor, circularity, was measured on n ≥ 14 cells/condition. Circularity is in a range between 1 and 0, where 1 identifies completely circular cells, and 0 sharp cells. (E) Representative phase-contrast images of HaCaT cellular shape. Images are collected after 72 h of PRPs treatment. Scale bar 100 μm. (F) Gene expression levels of TGM1 from HaCaT grown in Low and High Ca²⁺ condition. Cells were treated for 72 h with PRPs and compared with the untreated control (n = 3). Different bar colours indicate the following conditions: black for 10% FBS, grey for 0.1% FBS, red for Aphe liquid, blue for BC liquid PRP, pink for Aphe gel and light blue for BC gel. Data are expressed as mean ± SD. Significance of data differences was established using unpaired Two-Way ANOVA test with multiple comparisons. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001.

A wound scratch assay was conducted to assess PRPs effect on wound healing (n ≥ 3) after 72 h of treatment (Fig. 4C). In low Ca²⁺, untreated HaCaT cells exhibited reduced wound closure capacity compared to standard DMEM (##, p-value < 0.01). FBS 0.1% controls showed also a reduced wound closure when compared to 10% FBS (although this difference was significant only for the low Ca²⁺ condition).

PRPs promoted wound closure in all low Ca²⁺ and standard media (DMEM) conditions, although in the latter this was not significant for PRP gels. The results demonstrated that liquid PRPs are able to induce a superior wound closure compared to PRP gels (Fig. 4C). However, this difference was diminished under conditions of low Ca²⁺ concentration and was no longer statistically significant for Aphe gel. Wound closure values were as follows: DMEM (FBS 10% 75.6 ± 27.4, FBS 0.1% 39.2 ± 9.8, Aphe liquid 100 ± 0, Aphe gel 60.9 ± 6.2, BC liquid 100 ± 0, BC gel 57.6 ± 20.5); Low Ca²⁺ (FBS 0.1% 11.9 ± 0.4, Aphe liquid 100 ± 0, Aphe gel 87.9 ± 12.3, BC liquid 100 ± 0, BC gel 78.4 ± 3.8). Representative phase-contrast images are reported in Supplementary Figure 1.

In order to assess HaCaT morphology, cellular circularity was measured (n ≥ 14 cells for each condition, Fig. 4D). HaCaT cells in standard medium displayed a circular morphology, with no significant differences among conditions. As expected, in low Ca²⁺ conditions, untreated control cells became more elongated⁴⁸ compared to high Ca²⁺ FBS 0.1% control and to 10% FBS, while there was not any difference between these two latter conditions; however, PRPs reversed this effect, partially with liquid PRPs and fully with gel PRPs (Fig. 4D). Circularity values were: FBS 10% 0.84 ± 0.11, High Ca²⁺: FBS 0.1% 0.84 ± 0.09, Aphe liquid 0.80 ± 0.10, Aphe gel 0.84 ± 0.07, BC liquid 0.84 ± 0.05, BC gel 0.79 ± 0.13; Low Ca²⁺: FBS 0.1% 0.41 ± 0.16, Aphe liquid 0.67 ± 0.11, Aphe gel 0.85 ± 0.05, BC liquid 0.66 ± 0.14, BC gel 0.81 ± 0.10.

Phase-contrast images (Fig. 4E) after 72 h of treatment, revealed more defined cellular borders in low Ca²⁺ and in HaCaT cells treated with liquid PRPs, probably as a result of altered cell–cell junctions. The observed morphological changes in low calcium conditions induced by PRP gels may be attributed to their elevated calcium content, which is a consequence of their production methodology (see Fig. 1C and Materials and Methods).

The gene expression of TGM1 (Transglutaminase 1), a marker of keratinocytes and HaCaT differentiation^49–51, was assessed to determine the effect of the low/high Ca²⁺ in combination with the treatment with PRPs. Its expression was significantly lower in low Ca²⁺ compared to both high Ca²⁺ and FBS 10%.

All the PRP-treated cells promoted an increase in TGM1 expression compared to the FBS 0.1% control, in both low and high Ca²⁺. This effect was slightly higher in liquid PRPs than in gel PRPs and in some cases this increase was significant (Fig. 4F) (FBS 10% 0.0036 ± 0.0001, High Ca²⁺: FBS 0.1% 0.0037 ± 0.0010, Aphe liquid 0.0085 ± 0.0051, Aphe gel 0.0068 ± 0.0001, BC liquid 0.0090 ± 0.0003, BC gel 0.0061 ± 0.0005; Low Ca²⁺: FBS 0.1% 0.0015 ± 0.0006, Aphe liquid 0.0051 ± 0.0006, Aphe gel 0.0048 ± 0.0008, BC liquid 0.0057 ± 0.0009, BC gel 0.0031 ± 0.0005).

Effect of PRPs on HDMEC dermal microvascular endothelial cells behaviour

In addition to keratinocytes and fibroblasts, skin function in both normal and pathological conditions is regulated by microcirculation, which play a key role in mediating inflammatory responses, cell signalling, and migration⁵². Notably, endothelial cells themselves are often involved in tissue regeneration through tissue micro-vascularization and angiogenesis⁵³. Herein we exploited a cellular model of skin microcirculation, specifically human dermal microvascular endothelial cells, HDMEC^54,55, to investigate the effects of PRPs on their cellular behaviour (treatment scheme is represented in Fig. 5A).

Fig. 5.

Effect of PRPs treatments on HDMEC investigating: (A) Schematic representation of the experimental design with HDMEC. (B) Proliferation. WST-1 assay evaluated after 72 h of PRPs treatment. n ≥ 3 replicates. Data are expressed as % of 10% FBS control. (C) Migration. Number of HDMEC cells migrated into a scratch test after 64 h of treatment with PRPs. Data are obtained by constructing a reference area and counting the number of cells migrating within the area over time. n ≥ 3 replicates. (D) Circularity. A cellular shape descriptor, circularity, was measured on n ≥ 15 cells/condition. Circularity is in a range between 1 and 0, where 1 identifies completely circular cells, and 0 sharp cells. (E) Representative phase-contrast images of HDMEC cellular shape. Images are collected after 64 h of PRPs treatment. Scale bar 100 μm. (F) Gene Expression levels of specific EMT and normal endothelial markers. ACTA2 (αSMA), VWF, SELE and COL1A1 expression (2^-ΔCt) in HDMEC fibroblasts evaluated by qPCR after 72 h of treatment with PRP. Different bar colours indicate the following conditions: black for 5% FCS, grey for 0.1% FBS, red for Aphe liquid, blue for BC liquid, pink for Aphe gel and light blue for BC gel. Data are expressed as mean ± SD. Significance of data differences was established using unpaired Two-Way ANOVA test with multiple comparisons and simple t-test for the comparison between FCS 5% and FCS 0.1%. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001.

Proliferation of HDMEC was evaluated by WST-1 assay performed culturing cells for three days in the presence of the four PRP formulations under test (n ≥ 3 replicates, Fig. 5B). As previously described for BJ and HACAT cell lines, HDMEC were maintained in a medium with low concentrations of nutrients (0.1% FCS). HDMEC were treated with PRPs products 72 h and compared with the untreated control (0.1% FCS only). HDMEC cultured with 5% FCS were used as standard comparative control. Untreated control showed reduced proliferation compared to cells cultured under standard conditions with 5% FCS (p-value < 0.001), whereas all PRPs significantly promoted HDMEC proliferation, with liquid PRPs showing a greater effect than gel PRPs. The WST-1 absorbance values were as follow: FCS 5% 0.41 ± 0.05, 0.1% FCS 0.11 ± 0.04, Aphe liquid 0.45 ± 0.04, Aphe gel 0.31 ± 0.11, BC liquid 0.48 ± 0.08, BC gel 0.26 ± 0.09.

PRP treatment significantly stimulated the migration ability of HDMEC, as demonstrated by the scratch assay (n ≥ 3 replicates, Fig. 5C), in all the conditions except for the BC gel formulation. As observed with cell proliferation, liquid PRPs induced a stronger migratory response than PRP gel. The experiment was conducted over a 64-h period, by which time the wound had already closed in several conditions (Wound closure: FCS 5% 100 ± 0.0, FCS 0.1% 39.3 ± 6.0, Aphe liquid 100 ± 0.0, Aphe gel 75.7 ± 22.7, BC liquid 96.5 ± 5.6, BC gel 52.6 ± 20.4).

In order to evaluate how PRPs treatment influence HDMEC cellular morphology, we assessed circularity after 64 h of treatment. HDMEC culture in FCS 0.1% showed reduced circularity compared to the FCS 5% condition (Fig. 5D). Interestingly, treatment with all the PRPs reverted this morphological elongation, causing HDMEC to regain their circularity values, measured on n ≥ 15 cells per condition; values were as follows: FCS 5% 0.68 ± 0.12, FCS 0.1% 0.34 ± 0.09, Aphe liquid 0.69 ± 0.13, Aphe gel 0.71 ± 0.09, BC liquid 0.60 ± 0.10, BC gel 0.63 ± 0.14 (Fig. 5D).

The expression of some key genes involved in vascular endothelial cell regulation (αSMA, COL1A2, SELE, VWF, n ≥ 3 replicates) indicated that PRPs significantly promoted the expression of Selectin E (SELE) (FCS 0.1% vs BC liq FC 8, p-value < 0.001, FCS 0.1% vs Aphe gel FC 4.7, p-value < 0.05). No significant effects were observed on other genes, with the exception of an increase of αSMA expression in the Aphe liquid condition (FC 1.5, p-value < 0.05) (Fig. 5E). SELE is an endothelial-leukocyte adhesion molecule involved in inflammatory responses that can have a role also in endothelial cell proliferation⁵⁶. In accordance with proliferation data (Fig. 5A), we found an increase of SELE expression after PRP treatments (Fig. 5D). PRP treatment restored HDMEC proliferation, migration and their endothelial phenotype compared to the elongated shape of HDMEC in 0.1% FCS condition.

Discussion

In this study, we compared the effects of different formulations of PRP on wound healing. Our study aims to fill knowledge gaps regarding various basic PRP products for wound healing to improve clinical decision-making and pave the way for innovative strategies in PRP administration. The heterogeneity of PRP products^2,12,57,58, both in terms of preparation and administration, complicates the establishment of consensus clinical guidelines.

The most diffused formulation for treating wound repair is autologous PRP, which is generally considered safe across many randomised clinical trials (RCTs), with no significant increases in infections or other adverse events. Allogeneic PRP, which can be produced from both apheresis and BC (or, in some contexts, derived from umbilical cord blood), has also shown a favourable safety profile in clinical use⁵⁹. For instance, Liao et al.⁶⁰ and Mohseni et al.⁶¹ reported no major safety concerns. Akbarzadeh et al.⁶² conducted a comprehensive review of allogeneic PRP and concluded that, when properly processed, it is a safe and potentially effective wound-healing agent. Nonetheless, large-scale RCTs specifically designed to assess the safety of allogeneic PRP are lacking. While the initial evidence is encouraging, broader and longer-term safety data are still needed.

PRP can be administered as either a liquid or a gel, depending on the tissue being treated and clinical judgment². Liquid PRPs may also be applied to lesion sites using different delivery strategies (e.g. polymers and hydrogels⁶³) soaked with PRP to enhance contact time and facilitate controlled release of PRP effectors (i.e., GFs). Herein we specifically focused on PRPs produced with a freeze and thaw cycle derived from platelet apheresis and buffy coat, produced in both liquid and gel forms. To evaluate the differences among these formulations, we quantified the release of GFs, EVs and Ca²⁺ ions over time. We then assessed the effects of PRPs on in vitro models of fibroblasts (BJ), keratinocytes (HaCaT), and microvascular endothelial cells (HDMEC). To mitigate donor variability, PRP products were prepared from pooled samples of four donors.

We first quantified the release of GFs from the PRPs at the baseline and over a period of seven days. At baseline, no significant differences were observed between apheresis-derived and buffy coat-derived PRP in terms of GFs content. The use of transwells to test GFs release through a permeable barrier mimics GFs release from PRPs towards the wounded tissue. With this experimental setup we demonstrated that PRP products provided sustained GFs release over seven days.

Nevertheless, liquid PRP exhibited an early and sustained release of GFs, while the gel formulation demonstrated a late and sensibly lower release. Moreover, we observed considerable heterogeneity in the total percentage of GFs released over the entire seven-day period: approximately 100% of EGF and CTGF were released, while TGF-β1 and PDGF releases were below 50%. Although all the GFs analyzed are stored in platelets α-granules, these granules are heterogeneous and consist of distinct subpopulations that undergo differential release depending on the stimulus⁶⁴. The different kinetics observed may be partly explained by the preferential localization of EGF and CTGF in pro-angiogenic α-granules, which are more readily released than the pro-fibrotic granules containing PDGF and TGF-β. Comparisons between Aphe and BC PRPs revealed no differences in EGF and CTGF release, while higher amounts of TGF-β1 and PDGF-AA were released in liquid BC compared to liquid Aphe. Conversely, previous studies found that PRPs (both with and without freeze and thaw) released the majority of GFs in the supernatant in the first 24–72 h^31,65,66. The difference with a sustained release over the whole seven days period observed herein may be caused by a controlled diffusion of the GFs from the PRP matrix over the permeable membrane of the transwell. The fibrin mesh of gel PRPs may limit GFs diffusion; nevertheless, although previous studies found a sustained release over time of GFs form PRP gels^67,68 a clear mechanistic explanation has never been investigated. As discussed above, other studies have reported divergent release kinetics depending on the GFs. However, it is not possible to exclude the possibility that this is due also to a different retention of the transwell membrane. Further investigation is required in this respect, as it constitutes a limitation of the study.

Emerging evidence highlights the role of EVs as carriers of bioactive molecules that stimulate skin wound healing⁶⁹. Our findings show that EVs release is significantly higher in liquid PRP compared to gel formulations, suggesting greater regenerative potential for the liquid formulation related to it. The delayed release of EVs aligns with their nature as byproducts of platelet degradation⁷⁰. The increase in EVs release over time may be due to the progressive degradation of platelets. Although the progressive release of EVs from platelets over time has been previously described^71,72, to our knowledge, the kinetics of EVs release from freeze–thaw PRP have never been described. Ca²⁺ is involved in the regulation of keratinocytes function^33,47 and it was predominantly released during the first 72 h; increased Ca²⁺ levels in gel PRPs is consistent with its addition during gel preparation.

We further analyzed how the different PRP formulations affected cell culture by investigating their impacts on key factors influencing wound healing.

Human fibroblasts (BJ) treated with various PRP formulations showed differences in cellular proliferation; liquid PRPs demonstrated superior capacity to restore BJ cell proliferation than gel PRPs. This effect was observed both with and without TGF-β pre-treatment. Gel PRPs also induced cell proliferation, indicating that even lower amounts of GFs can stimulate this response. However, no effects on BJ migration were noted. Regarding myofibroblast differentiation, TGF-β induced significant changes in expression markers^35,38 but did not alter cellular morphology. In contrast, cumulative treatment with TGF-β and PRPs resulted in more elongated BJ cells and to a sustained dysregulation of several myofibroblast differentiation markers. This effect is more pronounced in gel PRPs suggesting that a different balance of GFs stimulation can finely tune cell fate towards proliferation and/or differentiation. Analysing MMP1, a marker of extracellular matrix (ECM) remodelling⁷³, it was found that its expression was upregulated with PRP and downregulated with a combination of PRPs and TGF-β. Accordingly, a previous research found that MMP1 was upregulated in dermal fibroblast treated with PRP⁷⁴, herein we confirmed this result, suggesting also that pre-treatment of fibroblast with TGF-β may hamper the capability of GFs present in PRPs to induce fibroblast to become more prone to modulate ECM turnover and remodelling.

In contrast to BJ cells, HaCaT keratinocytes exhibited both stimulated proliferation and migration upon treatment with PRPs; again, liquid formulations outperformed gels without significant differences between apheresis and BC sources. Interestingly, liquid PRPs increased HaCaT cell viability under low calcium conditions but had less impact compared to normal DMEM (high Ca²⁺). Notably, both high and low Ca^2⁺ conditions resulted in similar wound closure capabilities with liquid PRP treatment.

The differential effects on proliferation and migration may be related to the rate at which the PRP releasates restore these respective functions. Ca²⁺concentrations measured during initial days of treatment were comparable or exceeded those found in normal DMEM (~ 1–2 mM)⁷⁵, suggesting that higher Ca²⁺concentrations in gel formulations could explain their enhanced effects on cell proliferation and morphology. Conversely, TGM1 expression, a marker for cellular differentiation^49–51, was slightly higher in liquid than gel formulations, indicating a combined effect from Ca²⁺and GFs released at higher concentrations from liquid PRPs influencing HaCaT differentiation from basal states.

The impact of different PRP formulations on microcirculation neo-angiogenesis was assessed using HDMEC as an in vitro model. All tested PRPs promoted viability and wound closure capabilities along with the typical endothelial morphology, suggesting sustained HDMEC function due to PRP treatment. Additionally, all formulations promoted SELE expression, an adhesion marker correlated with cell growth⁵⁶, while no changes were observed in epithelial-mesenchimal transition (EMT) markers (αSMA and COL1A2) except for a slight increase in αSMA following BC liquid treatment.

Future work will explore the vesicle cargo composition and bioactivity in greater detail, gene expression analysis also revealed contradictory results that should be better elucidated. In particular, we investigated BJ cells transformation upon TGF-β treatment. The effect was confirmed assessing αSMA expression at 48 and 72 h from the beginning of the treatment (Fig. 3A). Nevertheless, EMT genes did not show a sustained expression and the FBS 0.1% control has a low level of expression of αSMA, MMP1 and COL1. PRPs sustain EMT for a longer time in the +TGF-β condition. In particular PRP gels promote αSMA expression and COL1 expression, suggesting that a low TGF-β dose may contribute to amplify the TGF-β pre-stimulation^76,77. A different effect of TGF-β depending on the concentration/dose and on the interaction with other GFs has been previously documented^77,78. Conversely, the effect on VIM was not influenced by the TGF-β dose, with VIM significantly higher than the control in the Aphe PRPs. In this case the differential effect of Aphe and BC may not be directly link to the GFs dosage, since the major differences we found were between liquid and gel products. Nevertheless, we may hypothesize a possible contribution of co-regulatory mechanisms due to the co-stimulation of two or more GFs, for instance both liquid and gel Aphe PRP, compared to the BC counterparts, had lower concentrations of released TGF-β over time and a delayed release of PDGF-AA.

MMP1 showed a different regulation pattern, in the -TGF-β condition the gene is in general up-regulated by PRPs, while it is down-regulated by the pre-treatment with TGF-β. An exception is represented by the Aphe condition not treated with TGF-β(-TGF-β), where MMP1 has a level of expression comparable with the control. In this case as well, it is challenging to formulate a definitive hypothesis based on the analysis conducted. As with other observations, it is challenging to formulate a definitive hypothesis based on the current data. However, one possible explanation could involve PDGF-AB secretion, which appears to be more sustained in the Aphe liquid PRP, along with the influence of endogenous TGF-β. A strong stimulation with TGF-β, such as that induced by pre-treatment, may significantly suppress MMP1 expression, as observed in our experiments. Additionally, the higher PDGF-AB levels and delayed PDGF-AA release may have enhanced TGF-β signaling, consistent with previous studies showing TGF-β overexpression in PDGF-treated fibroblasts⁷⁶. A similar mechanism may underlie SELE regulation. In HDMECs, SELE was upregulated in response to BC liquid PRP. Again, considering the known synergistic interactions among growth factors, particularly between TGF-β and PDGF^76,78,79, we hypothesize that the high levels of TGF-β and PDGF-AA found in BC liquid PRP may act cooperatively to induce SELE expression. Nonetheless, to the best of our knowledge, there are no studies that specifically investigate the fine regulation of endothelial SELE in response to stimulation by different growth factors. Altogether, these findings suggest a complex regulatory landscape in which growth factors exert synergistic or antagonistic effects on gene expression, depending not only on their individual concentrations but also on the timing and context of their administration.

Overall, our results demonstrate that all tested PRPs enhance the regenerative capacity of cells involved in wound healing without significant differences between BC-derived and Aphe-derived products; however, noteworthy distinctions exist between gel and liquid formulations. This substantial difference in GF release kinetics must also consider application methods for optimal therapeutic outcomes.

For chronic wounds requiring topical application, gel formulations are often preferred; however, liquid PRPs may be utilized when absorbed into biocompatible materials such as hydrogel dressings. In cases such as lichens sclerosus, where injections are administered at lesion sites⁸⁰, it is crucial to consider the duration of GFs exposure at injection sites since this may significantly reduce cellular exposure times.

This study has several limitations that should be acknowledged. The number of PRP units and experimental replicates was constrained by the availability of clinical-grade platelet concentrates and the logistical complexity of multi-day release experiments and in vitro assays, some of which were performed in duplicate. Furthermore, all functional assays were conducted using in vitro wound-healing models which, although standardized and reproducible, do not fully capture the physiological intricacies of tissue repair in vivo. As such, while our findings provide meaningful insights into the differential behavior of PRP formulations, they should be interpreted within the context of these experimental conditions. Future studies involving larger sample sizes and in vivo models will be necessary to confirm and expand upon these observations.

In addition, studies evaluating the clearance kinetics of PRP components at cutaneous and subcutaneous sites are warranted to better understand their in vivo dynamics. This work highlights the relevance of characterizing GFs and EVs release profiles from different PRP formulations. Liquid PRP, characterized by rapid GFs release and a higher EVs content, may be particularly suited for acute or early-phase wound-healing applications. In contrast, gel-based PRP offers a more sustained release profile, which may be advantageous for prolonged regenerative processes.

Future research should focus on optimizing dosages, standardizing preparation methods, and integrating additional bioactive agents to maximize the therapeutic potential of PRP treatments based on specific clinical contexts.

Methods

Study design, platelets collection and PRP preparation.

The study was conducted by the Transfusion Medicine Unit of the Azienda USL-IRCCS di Reggio Emilia in collaboration with the Unit of Clinical Immunology, Allergy and Advanced Biotechnologies, and approved by the Institutional Review Board on 12 May 2022 (Reggio Emilia Ethics Committee, protocol number 2022/0062204). All recruited donors signed an informed consent according to the Declaration of Helsinki. Only male donors were included, as all platelet concentrates used in this study were sourced from volunteer donors recruited through our hospital transfusion service, where current selection policies and standard operating procedures result in a predominance of male donors.

Apheresis-derived platelet concentrates were collected with an automated blood collection system (Mobile Collection System MCS + , Haemonetics Corp., Boston, MA, USA)²¹ in anticoagulant ACD-A (40 ml of Acid Citrate Dextrose Solution A, ACD-A, Haemonetics, Italy) and plasma (160 ml). Prior to PRP preparation, platelets were further diluted in plasma (200 ml of platelet concentrate + 150 ml of Plasma. V_{PlateletConcentrate}:V_Plasma = 4:3) (Fig. 1A). Aphe PRP was obtained pooling together leucodepleted platelet concentrates from 4 different donors. Apheresis units were collected from 0 + male donors with a mean age of 52 ± 7. Aliquots obtained from the pool were stored at − 80 °C and used after a single freeze–thaw cycle.

Buffy Coats (BC) were obtained from separation of whole blood donations with Compomat G4 separator (Fresenius Kabi Medicare, Italy), as previously described⁸¹. Each BC was diluted in a solution having the same composition used for platelets from apheresis (ACD-A and plasma). 4 different diluted BC were combined in a Transfer bag (Transfer bag BB*T200BM, Terumo, Italy) and centrifuged using a specific kit (TACSI PL kit, Terumo, Italy) in order to obtain a platelet concentrate. In the final step the platelet concentrate was diluted in plasma with the same ratio used for platelet concentrate from apheresis (V_{PlateletConcentrate}:V_Plasma = 4:3) (Fig. 1B). BC units were collected from 0 + male donors with a mean age of 49 ± 5. The pooled platelet concentrates prepared were then aliquoted and stored at− 80 °C in order to be used as BC PRPs after a single freeze–thaw cycle within one year.

Aphe and BC PRPs were prepared in order to have the same final platelet count. The latter was tested after preparation on both Apheresis and BC platelet concentrates before storage and resulted to be 580,000 cells/µl, which falls within the commonly accepted PRP range of 500,000–1,500,000 cells/µL⁸².

Gel PRPs were prepared using thrombin and Calcium Gluconate. Briefly, PRPs were mixed 3:1 with a mixture of calcium gluconate (Galenica Senese, Industria Farmaceutica, Italy), thrombin (Homologous thrombin is A by-product of blood routinely obtained at Transfusion Medicine Unit of the AUSL-IRCCS of Reggio Emilia⁹) in order to achieve complete platelets activation and coagulum formation. Final concentration of calcium gluconate was 7.7 mg/ml while thrombin was diluted 6.5 times before usage. The final concentration of thrombin in the PRP gel preparations was approximately 25 IU/m. The volume of thrombin and calcium gluconate was optimized to achieve full clotting, and no residual fluid was observed after gel formation.

PRP preparation for in vitro experiments

In our experimental conditions we investigated the release of effectors from PRPs and their effects on different cellular models. To reach this aim liquid and gel PRPs were prepared in 6-well for releasates experiments and 24-well Transwell for cell culture experiments, respectively (Fig. 1C) (6-well Cell Culture Inserts 0.4 µm PET clear, 24-well Cell Culture Inserts 0.4 µm PET clear, cellQART, Germany). Aphe liquid, BC liquid, Aphe gel and BC gel PRPs were prepared as following in order to compare products with an equal final volume.

Liquid PRPs were obtained by mixing PRPs with PBS (EuroClone, Italy) 3:1.

PRPs in the transwells had a volume of 2 ml for release experiments conducted in 6-well plates, while PRPs had a volume of 0.4 ml for cell culture experiments, conducted in 24-well plates. Experimental conditions are summarized in Fig. 1C.

Characterization of GFs and EVs released by PRPs

To evaluate the concentration of GFs and EVs released by PRPs (PRP releasates), liquid and gel PRPs were prepared in the upper chamber of 6-well transwells as described in the previous section while in the lower chamber were added 3 mL of filtered PBS. The PRP releasates from the lower chamber were collected every 24 h for 7 consecutive days. At each time point, the transwell was moved to a new well with the same amount of PBS to allow the collection of the GFs and EVs released during the previous 24 h (Fig. 1C). Approximately 6 mL of releasate was obtained per condition at each time point. Immediately after collection, samples were supplemented with protease inhibitor cocktail (1:200, Abcam, USA) and stored at − 20 °C until analysis. The concentrations of selected GFs—including EGF, CTGF, TGF-β1, PDGF-AA, PDGF-AB, and FGF2—were quantified in both Aphe PRP and BC PRP stocks, as well as in the releasates collected at days 1 through 7. For TGF-β1, PDGF-AB, and CTGF, samples were concentrated prior to quantification using Amicon Ultra-15 centrifugal filters (Sigma-Aldrich, USA) with dilution factors ranging between 2.8 and 3.3, depending on the condition. ELISA assays were employed for quantification, following manufacturer protocols. Specifically, the assays used for the quantification of EGF, PDGF-AB, PDGF-AA were purchased from Sigma-Aldrich (USA). The assays used for the quantification of TGF-β1 and FGF2 (FGF basic/FGF2/bFGF) were purchased from R&D Systems (USA). Data were collected using the GloMax® Discover Microplate Reader (Promega, USA). The assay used for the quantification of CTGF was purchased from Abcam (UK, Human CTGF SimpleStep ELISA kit). Data were collected with the FLUOstar OMEGA instrument (BMG LABTECH, Germany). All GFs concentrations were normalized according to concentration factors (measured values were multiplied for the dilution factor).

EVs released from PRPs were assessed in the releasates at each time point using a FACSLyrics flow cytometer (BD Bioscience, USA). EVs were quantified through the customized EVs kit for flow cytometry (Custom Kit #626267 BD Biosciences, USA) as previously described^72,83. Briefly, EVs were captured on antibody-coated beads targeting platelet markers (e.g., CD41), stained with fluorescent secondary antibodies, and analyzed by flow cytometry. Calibration was performed using size-standardized beads to define the EVs gate (< 1 μm). EVs concentrations were normalized to sample volume and expressed as events per microliter.

Ca²⁺concentration was assessed using a Calcium Microplate Assay Kit (Biorbyt, USA), according to manufacturers’ instructions. Citrate levels were quantified only on Aphe PRP and BC PRP at T0 (i.e. on PRP stock units) using a colorimetric assay (MAK057 Citrate Assay Kit, Sigma-Aldrich, USA), in accordance with the manufacturer’s guidelines.

Cell cultures

BJ cell line (human fibroblasts) was cultured at 37 °C in high glucose Dulbecco’s modified Eagle’s medium (DMEM High Glc, EuroClone, Italy) supplemented by 10% fetal bovine serum (FBS, EuroClone, Italy) and 1% Penicillin–Streptomycin (Pen-Strep, Sigma Aldrich, USA). Cells were detached 1 or 2 times per week keeping the confluency approximately under 90% in T75 flasks (CytoOne, Italy), growth medium was renewed 1–2 times a week.

For PRPs experiments BJ cells were seeded and after 24 h cultured for, 24, 48 and 72 h in presence or absence of TGF-β (TGF-β1, GF345, Millipore, Germany) at 10 ng/mL, prior the addition of PRPs. The 72 h TGF-β treatment was selected and used for most of the experiments as described in the Results section.

HaCaT cell line (human keratinocytes) was cultured at 37 °C in DMEM High Glucose (EuroClone, Italy) supplemented by 10% fetal bovine serum (FBS, EuroClone, Italy) and 1% Penicillin–Streptomycin (Pen-Strep, Sigma Aldrich, USA). Cells were detached twice a week and kept at approximately 80% confluence in T75 flasks, with growth medium renewed 2–3 times a week.

HaCaT cells were cultured and seeded in standard DMEM, at 24 h from seeding DMEM was replaced by fresh standard DMEM or DMEM Low Ca²⁺ (Gibco, USA), in order to obtain HaCaT with differentiated and basal phenotype respectively and treated with PRPs.

Human Dermal Microvascular Endothelial Cells cell line (HDMEC, Cat.No.C-12212PromoCell, Germany), were cultured in the Endothelial Cell Basal Medium MV (Cat.No.C-22220, PromoCell, Germany) supplemented with the Endothelial Cell Growth Medium MV Supplement Pack (Cat.No.C-39220, PromoCell, Germany) at 37 °C, 5% CO2. Cells were detached with Trypsin 0.025%—EDTA 0.25 mM (Thermo Fisher Scientific, USA). Trypsin was neutralized using PBS + 10% FCS and cells centrifuged at 220 g for 5 min. HDMEC cells were seeded and after 24 h cultured for 72 h with the addition of PRPs. Heparin 2U/mL (Millipore, Germany) was added to the cell culture medium together with PRPs for all the experimental conditions and for cellular types in order to avoid fibrin activation. Treatment of PRP was evaluated in the presence of 0.1% FCS as a minimal basal condition to reduce external growth factor influence and highlight the effects of PRP treatments. The usage of FCS instead of FBS was reported for HDMEC culture⁸⁴.

Cell proliferation

Cell viability was assessed using the WST-1 Cell Proliferation Assay (Cell Proliferation Reagent WST-1, Merck, Germany), which measures mitochondrial metabolic activity as an indirect indicator of viable cell number. BJ fibroblasts, HaCaT keratinocytes, and HDMECs were seeded in 24-well plates (Euroclone, Italy) at densities of 3000, 30,000, and 10,000 cells/cm², respectively, and allowed to adhere overnight in complete growth medium. The following day, cells were treated with PRP formulations diluted in culture medium supplemented with 0.1% FBS (for BJ and HaCaT) or 0.1% FCS (for HDMECs). After 72 h of treatment, 300 µL of fresh medium containing WST-1 reagent (1:20 dilution) was added to each well, and cells were incubated at 37 °C for 4 h. Absorbance was measured at 450 nm using a GloMax® Discover Microplate Reader (Promega, USA). Results were expressed as relative viability compared to untreated controls (0.1% serum alone), and data were normalized accordingly.

Scratch assay and cell circularity

To assess the effect of PRP formulations on cell migration, scratch wound assays were performed in 24-well plates. Cells were seeded to reach approximately 90% confluence: HaCaT keratinocytes at 80,000 cells/cm², BJ fibroblasts at 17,000 cells/cm², and HDMECs at 60,000 cells/cm². A linear scratch was created using a sterile 10 µL pipette tip (ClearLine, Dominique Dutscher, France) for HaCaT and BJ, and a 1000 µL pipette tip (TipOne Filter Tips, Starlab, Italy) for HDMECs, to account for their higher migratory capacity.

Following scratch creation, wells were gently washed twice with PBS (EuroClone, Italy) to remove detached cells and debris. Cells were then cultured in medium supplemented with 0.1% fetal bovine serum (FBS) for BJ and HaCaT cells, or 0.1% fetal calf serum (FCS) for HDMECs, to minimize proliferation and isolate the migratory effect of PRP. PRP treatments were added immediately after washing.

Scratch closure was monitored at 48 h (BJ), 72 h (HaCaT), and 64 h (HDMEC), and digital images were acquired at each time point using an EVOS FL microscope (Thermo Fisher Scientific). Image analysis was performed using ImageJ software⁸⁵. Wound closure was calculated differently depending on cell type; for HaCaT and HDMECs, which migrate as a compact monolayer, the percentage of wound closure was calculated using the formula:

For BJ fibroblasts, which migrate individually, the number of cells that migrated into the wound area was manually counted from calibrated images and expressed as cell count per wound area.

To evaluate cell morphology and cytoskeletal remodelling, cell circularity was assessed at the same time points. Bright-field images were acquired at 20X magnification for BJ and HaCaT cells, and 4X magnification for HDMECs. Circularity analysis was performed on approximately 15 randomly selected cells per condition, across replicates. Cell outlines were manually traced, and area and perimeter were measured using ImageJ. Circularity was calculated using the standard formula:^40,41:

A circularity value of 1.0 indicates a perfect circle, while values approaching 0 indicate an elongated or irregular shape, consistent with migratory or activated morphology.

Quantitative polymerase chain reactions (qPCR)

BJ, HaCaT and HDMEC were analysed for gene expression: all cell types were treated as described in the Results section prior to gene expression analysis. Cell culture conditions were the same described in the Cell Proliferation paragraph of the Material and Methods.

RNA samples were isolated from BJ and HaCaT cells pellet with the Monarch® Total RNA Miniprep Kit (New England Biolabs, USA) following manufacturer’s instruction. cDNA was obtained by reverse-transcribing the same amount of total RNA (200 ng) using the RevertAid RT Kit (Thermo Fisher Scientific, USA).

Regarding HDMEC, RNA was extracted from the cells with the Quick DNA/RNA MicroPrep Plus kit (ZymoResearch, USA). The NanoDrop instrument (Thermo Scientific, USA) was used for RNA quantification. For gene expression, 45 ng of RNA were reverse transcribed with the PrimeScrip RT Reagent Kit with gDNA Eraser (Takara, Japan).

For the real-time PCR, the following QuantiTect Primer Assay from QIAGEN (USA) were used: selectin E (SELE, QT00015358—Hs_SELE_1_SG), von Willebrand factor (vWF, QT00051975—Hs_VWF_1_SG), collagen type I alpha-2 (COL1A2, QT00072058—Hs_COL1A2_1_SG).

qPCR was performed on cDNA samples by using the GoTaq® qPCR Master Mix (Promega, USA), the reactions were carried out on the CFX Duet Real-Time PCR System (BioRad, USA). The other primers for qPCR were purchased from Eurofins (Italy) and their sequences are listed in the Supplementary Table 2. GAPDH was used for normalization.

Statistical analysis

Statistical analysis was performed using GraphPad (Prism 8.4.2, GraphPad Software, USA) and Microsoft Excel 2010 (USA). One-way ANOVA with multiple comparisons test or two-tailed Student’s t test were performed to assess differences among samples. Data were presented as mean ± StD. A p-value of < 0.05 was considered statistically significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

D.S., conceptualized the study, C.M. and G.G. performed experiments, analyzed data, and drafted the manuscript. C.C., I.F., and S.C. conducted analyses on microvascular endothelial cells; S.C. and D.S. supervised *in vitro* analyses. A.R. and M.G. supported experimental work and data collection. D.S., and L.M. contributed to study design and manuscript writing. B.I. and E.D.B contributed to data interpretation and manuscript revisions. R.B. acquired funding. All authors reviewed and approved the final manuscript.

Funding

This research did not receive any specific graft from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All data from this study are included in this article. Original data are kept on file and are not publicly available, due to privacy data legislation. Requests can be directed to the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The study was approved by the Institutional Board Review on 12 May 2022 (Reggio Emilia Ethics Committee, protocol number 2022/0062204), all recruited donors signed an informed consent according to the Declaration of Helsinki.