Selective Paracrine Modulation of Stromal Cells: Wharton’s Jelly MSC Secretome Enhances Adipose-Derived MSC Functionality While Maintaining Dermal Fibroblast Quiescence

Aligned with the central aim of this study, we initially explored the impact of the WJ-MSCs on the adhesion and morphological characteristics of adipose-derived mesenchymal stem cells (AD-MSCs) and human dermal fibroblasts (HDFs), all originating from distinct mesenchymal tissues. Collagen was selected as the substrate protein due to its prominent role as a key ECM component, essential for tissue regeneration and the whole stem cell biology field [26].

As depicted in Figure 1A, exposure to the WJ-MSC-derived secretome administered at a concentration equivalent to that present in the native WJ-MSC culture medium resulted in the enhanced cell spreading of both AD-MSCs and HDFs after 24 h of incubation (Figure 1A(b) and Figure 1A(e), respectively) compared to cells cultured on collagen-only substrates (Figure 1A(a,d)). Notable differences in cytoskeletal architecture were also observed. For AD-MSC samples, incubation with secretome led to the development of prominent actin stress fibers, indicating enhanced cytoskeletal organization. Conversely, samples treated with FBS showed less pronounced stress fiber formation, characterized by a primarily cortical arrangement, while cells grown on plain collagen revealed an intermediate cytoskeletal phenotype (Figure 1A(a–c)). Fibroblasts, in the same conditions, exhibited less developed actin networks compared to AD-MSCs, while treatment with FBS induced a more polarized morphology in both cell types (Figure 1A(c,f)), evident by the elongated cell shape, the corresponding increase in the cell aspect ratio (AR), and directionally oriented actin stress fibers. These results suggest the differential modulation of cytoskeletal dynamics depending on the bioactive environment. In this respect, it is important to note that these morphological characteristics were assessed after 24 h of incubation—a time point where the organization of the actin cytoskeleton differs significantly from the early stages of cell adhesion.

The quantitative morphometric analysis shown on Figure 1B and Table 1 revealed statistically significant differences in the cell spreading area (CSA) and AR for both AD-MSCs and HDFs (Figure 1B). In the presence of the WJ-MSC secretome, AD-MSCs exhibited a ~30% increase in the CSA, expanding from 4007 μm² on unmodified collagen to 5081 μm² (Figure 1B(a)), indicative of an enhanced cell–substrate interaction. This expansion was accompanied by an approximate 15% increase in AR, suggestive of a slightly diminished cellular polarization. Conversely, under serum-enriched conditions (the positive control), AD-MSCs demonstrated a ~15% reduction in the CSA, decreasing to 3389 μm² from the baseline of 4007 μm² (Figure 1B(a)), which was correlated with a concomitant decrease in AR values (Figure 1B(b); Table 1), implying increased cellular elongation and potential directional migration.

Fibroblasts, typically over threefold smaller in size compared to AD-MSCs, also demonstrated enhanced spreading in the presence of the WJ-MSC secretome, with the CSA surprisingly increasing by approximately 79% from 971.6 μm² to 1737 μm² (Figure 1B(c)). The increased values are presented in % because of the initial difference in cell size between WJ-MSC and fibroblasts. However, despite this improvement, their CSA remained intermediate and did not reach the values observed in fibroblasts cultured with FBS, where a CSA of 2289 μm² was recorded (corresponding to a 135.6% increase). AR values for fibroblasts remained relatively stable at around 0.350 (Figure 1B(d), Table 1), showing only modest decreases ranging from 5% to 11.9% compared to controls, indicating a slight increase in cellular polarization (Figure 1A(e,f)).

Building upon our preliminary findings that the WJ-MSC-derived secretome enhances the proliferation of AD-MSCs, we now provide a more comprehensive analysis of this effect, including a comparative evaluation of its influence on fibroblasts. This approach is motivated by the involvement of both cell types in tissue regeneration and the clear relevance of this experimental observation to regenerative medicine.

As shown in Figure 2A, the WJ-MSC-derived secretome promotes the proliferation of AD-MSCs cultured on collagen substrates under serum-free conditions, with a statistically significant effect observed by day 3.

In contrast, HDFs cultured under identical conditions did not exhibit a notable increase in cell number (Figure 2B), except in the presence of FBS, where cell counts were significantly elevated on both day 2 and day 3, reaching approximately double the control levels by day 3 for both cell types.

The influence of the secretome on cell proliferation kinetics is more clearly illustrated by analyzing the cell doubling time (Figure 2C,D). For AD-MSCs, secretome treatment resulted in a substantial reduction in the doubling time, approximately twofold compared to the controls (Figure 2C), again indicating an accelerated proliferation. However, no significant change in the cell doubling time was observed for HDFs under the same conditions (Figure 2D). In both cell types, FBS supplementation consistently yielded the shortest doubling times, confirming its potent mitogenic effect.

To better clarify the effect of WJ-MSCs' secretome, we further performed an image-based fluorocytometric analysis of the nuclear DNA content, as viewed by Hoechst staining of the same AD-MSCs and HDF cells cultured for 24, 48, and 72 h (Figure 3). WJ-MSCs' secretome treatment leads to faster proliferation in stem cells compared to the control with plain media, observable at 48 h and more so at 72 h after the start of the incubation, with a doubling time 1.6 times faster than that of the non-treated samples but still about 1.9 times slower than the cells incubated with FBS 10%. This tendency was not present when fibroblasts were treated with the secretome, where no difference was found in the proliferation rate.

As shown in Figure 3, the WJ-MSC-derived secretome exerts distinct modulatory effects on cell cycle progression in MSCs while having minimal impact on HDFs. MSCs exposed to the secretome under serum-free conditions display a transient accumulation in the S phase at 24 and 48 h, followed by a pronounced shift toward G0/G1 by 72 h (Figure 3B, upper row), indicative of entry into a quiescent state. In contrast, fibroblasts treated with the secretome remain consistently arrested in the G0/G1 phase throughout the entire observation period, with no detectable transition into S or G2/M phases (Figure 3B, lower row). These cell cycle profiles are consistent with the previously noted differences in cell proliferation and doubling time.

A key hallmark of mesenchymal cell function is the production and organization of the ECM [34], with collagen playing a central role. In this context, we investigate the influence of the WJ-MSC secretome on the intracellular processing and substratum deposition of collagen by AD-MSCs and HDFs, again in a comparative plan. Specifically, we examine the fate of the endogenous type I collagen produced by these cells using an anti-human collagen antibody while they are cultured on an RTC, which is not detected by the antibody.

Figure 4 illustrates the intracellular organization and extracellular deposition of endogenous human type I collagen following 24 h incubation of AD-MSCs (Panel A) and HDFs (Panel B) on RTC-coated substrata. The impact of the WJ-MSC secretome (images in Figure 4A(b,f),B(b,f)) is assessed relative to plain collagen-coated controls (Figure 4A(a,d),B(a,d)) and FBS supplementation (Figure 4A(c,g),B(c,g)), which serves as a positive control.

In both Figure 4A,B, the bottom row displays the same cells as on the top row but with outlined contours to delineate intracellular and extracellular compartments—an essential distinction for the subsequent microfluorimetric quantifications shown in diagrams (d and h) on both panels (Figure 4A,B), respectively.

An analysis of Figure 4A revealed that the WJ-MSC secretome significantly enhanced the extracellular deposition of collagen type I by AD-MSCs. This was evidenced by the clustered fluorescence signals (indicated by an arrow in image b on Figure 4A) and quantitatively supported by elevated fluorescence intensity (images d and h on Figure 4A). The deposited collagen appeared morphologically disorganized, forming dense clusters without clear spatial alignment (see the 5× augmented inset on image b). These differences were statistically significant, as shown in graph (Figure 4A(d)).

In contrast to AD-MSCs, HDFs cultured under identical conditions (Figure 4B) demonstrated significantly lower levels of extracellular collagen type I deposition in response to the WJ-MSC secretome (Figure 4B(b,f)). This reduction was evident when compared to both the plain collagen control (Figure 4B(a,e)) and the FBS-supplemented samples (Figure 4B(c,g)), with statistical significance confirmed in Figure 4B(d).

Additionally, intracellular collagen fluorescence in both cell types (Figure 4A,B) exhibited a consistent perinuclear localization. The fluorescence intensity remained intermediate between the plain collagen and FBS-supplemented conditions, with no statistically significant differences observed (Figure 4A(h),B(h)). This distribution pattern is morphologically indicative of Golgi-associated intracellular processing, where newly synthesized collagen undergoes post-translational modifications such as hydroxylation, glycosylation, etc., that ensure proper trimer assembly and trafficking [35,36].

Another essential hallmark of the regenerative potential of mesenchymal stem cells and fibroblasts is their capacity for directed migration toward sites of tissue injury—an ability that can be effectively simulated using the in vitro wound healing assay, commonly referred to as the scratch assay [37]. Figure 2 illustrates the impact of the WJ-MSC-derived secretome on the migratory behavior of AD-MSCs and HDFs, as evaluated through the time-lapse microscopy of live cells (see the Materials and Methods section). The figure presents only representative images captured at the initiation of the assay and following 24 h of incubation.

As can be seen in Figure 5, both cell types demonstrated markedly increased migration in response to the WJ-MSC secretome relative to the untreated control, although the extent of migration was lower than that observed under FBS-supplemented conditions. These findings suggest that the WJ-MSC secretome exerts a stimulatory effect on cell motility in both MSCs and HDFs, consistent with a paracrine-mediated pro-migratory mechanism.

While fibroblasts also responded to the WJ-MSC secretome, their activation was modest compared to AD-MSCs. This differential responsiveness likely reflects intrinsic differences in cellular plasticity and receptor expression. AD-MSCs, as multipotent progenitors, possess dynamic transcriptional and epigenetic landscapes that allow rapid adaptation to paracrine cues. In contrast, fibroblasts, being more terminally differentiated, exhibit limited transcriptional flexibility and are primarily geared toward ECM maintenance and wound closure. Mechanistic studies have shown that fibroblast activation is tightly regulated by mechanotransduction pathways such as RhoA/ROCK and YAP/TAZ, which respond to substrate stiffness and ECM tension [39]. Moreover, the fibroblast-to-myofibroblast transition (FMT), a hallmark of fibrotic remodeling, is often triggered by persistent TGF-β signaling and mechanical stress. Our data suggests that the WJ-MSC secretome does not induce this transition nor excessive collagen secretion, thus preserving fibroblast quiescence while selectively activating progenitor traits in AD-MSCs.

Paracrine stimulation influenced both collagen secretion and its spatial organization within the ECM, underscoring the therapeutic relevance of MSC-derived secretomes in ECM-focused regenerative strategies. The use of species-specific collagen substrates, such as RTC, allowed immunological distinction between endogenous human collagen and exogenous scaffold components, providing a refined platform for studying matrix remodeling. However, under our experimental conditions, RTC appeared suboptimal for supporting the organized fibrillogenesis of newly secreted collagen. Its lack of native crosslinking and nucleation sites, combined with a pre-adsorbed and non-physiological surface topology, likely disrupted the alignment of secreted collagen, resulting in clustered aggregates rather than structured fibrils. Collagen self-assembly is highly sensitive to substrate architecture, as well as to environmental factors like pH and ionic strength [40]. To enhance ECM remodeling and functional integration, future studies should explore advanced 3D scaffolds such as decellularized human matrices or biomimetic hydrogels, which have demonstrated superior support for collagen alignment and structural organization [41].

The concentration of the secretome applied in vitro does not necessarily reflect its clinical efficacy. Achieving optimal dosing requires a careful balance between biological activity and safety, avoiding both overstimulation and unintended off-target effects. As noted by Gwam et al. (2021) [42], the regenerative potency of MSC-derived secretomes is dose-dependent, with maximal effects typically observed at intermediate concentrations ranging from 50 to 200 µg/mL of the total protein. Higher doses may provoke cellular stress or immune activation, while insufficient concentrations may fail to elicit a meaningful therapeutic response. In our study, we utilized the native concentration of secretomes present in the conditioned medium collected from WJ-MSCs after 48 h of culture (approx. 300 µg/mL total protein) prior to any concentration step, thus reflecting a physiologically relevant exposure level. To ensure reproducibility and translational relevance, the standardization of MSC secretome production is essential. This involves tightly controlling variables such as cell density, culture duration, and the composition of the conditioning medium, all of which significantly influence the secretome's bioactivity and therapeutic potential [43,44]. Without harmonized protocols, batch-to-batch variability can compromise both experimental outcomes and clinical efficacy, underscoring the need for good manufacturing practice (GMP)-compliant manufacturing strategies [43]. Furthermore, quantitative proteomic profiling and functional bioassays should be employed to define therapeutic windows tailored to specific clinical applications.

Our findings support the development of combination MSC therapies that leverage the complementary strengths of different MSC sources. WJ-MSCs offer robust immunomodulatory and proliferative signals, while AD-MSCs contribute lineage-specific differentiation and ECM remodeling. Clinical trials have demonstrated that WJ-MSC-derived exosomes accelerate wound healing and reduce inflammation in diabetic foot ulcers [45], whereas AD-MSCs show promise in cartilage regeneration and osteoarthritis management [46]. A combinatorial approach, either through co-culture, sequential secretome delivery, or engineered hybrid formulations, could enhance therapeutic outcomes by targeting multiple regenerative pathways simultaneously. Furthermore, the ability of the WJ-MSC secretome to activate both the progenitor and stromal cells suggests a dual mechanism of action: recruiting stem cells to sites of injury while mobilizing resident fibroblasts for matrix reorganization. This aligns with emerging paradigms in regenerative medicine that favor cell-free therapies capable of orchestrating endogenous repair processes without the logistical and immunological challenges of cell transplantation, as highlighted by González-González and colleagues (2020) [47].

Although secretome-based therapies are generally considered less immunogenic than whole-cell approaches, the use of allogeneic secretomes still warrants careful immunological evaluation. Donor-derived components—including extracellular vesicles, proteins, and nucleic acids—may retain immunogenic epitopes capable of eliciting host immune responses. This is particularly relevant in allogeneic applications, where inter-donor variability and the absence of standardized secretome profiles may influence immunocompatibility. Emerging strategies to mitigate these risks include preclinical immunogenicity screening, selective depletion of immunostimulatory molecules, and the development of engineered or synthetic secretomes with defined compositions [48]. Furthermore, the route of administration, dosage, and target tissue microenvironment can modulate the host response and should be optimized during translational development. Addressing these immunological aspects is critical to ensure the safety and efficacy of secretome-based interventions in clinical settings.

Despite the promising in vitro findings, several limitations must be acknowledged to contextualize the translational relevance of this study.

Human adipose tissue derived mesenchymal stem cells (AD-MSCs) at passage 2 were obtained from the Tissue Bank BulGen (Sofia, Bulgaria) with informed consent from donors prior to liposuction. Cells were cultured in DMEM/F12 medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS) (both from Sigma-Aldrich, USA) and 1% antibiotic–antimycotic solution (Sigma-Aldrich, USA) in humidified thermostat at 37 °C, 5% CO₂. Medium was replaced every 2 to 3 days until cells reached ~90% density and then passaged using 0.05% trypsin/0.6 mM EDTA (Sigma-Aldrich, USA). Cells used for experiments were between passages 5 and 7.

Primary human dermal fibroblasts (HDFs) were isolated from fresh skin biopsy also with informed donor consent following the procedure of Ningsih [51]. Briefly, the donor skin was sectioned into approximately 2 × 2 mm fragments, left to attach in a 12-well TC polystyrene plate (Corning, Corning, NY, USA) on an area marked by scratches, which were made to enhance the cell attachment. The explants were cultured in DMEM/F12 medium (Sigma-Aldrich, USA) supplemented with 10% FBS (Sigma-Aldrich, USA) and 1% antibiotic–antimycotic solution (Sigma-Aldrich, USA) in humidified thermostat at standard conditions (37 °C, 5% CO₂). Once fibroblasts began migrating out of the explant, the culture medium was replaced every 3–5 days. Primary cultures were maintained for approximately 10 days until reaching ~70–80% density. Cells were then harvested with trypsin–EDTA solution (Sigma-Aldrich, USA), and its activity was neutralized by adding an equal volume of FBS, then the cell suspension was centrifuged at 300× g for 5 min. The pellet was resuspended in fresh medium and seeded in a T25 TC polystyrene flask (Corning, USA) and left for subsequent cultivation in standard conditions.

Human umbilical cord tissue from healthy donors was obtained with informed consent, approved by the Ethics Committee of Ob/Gyn Hospital Dr. Shterev, Sofia, Bulgaria. Mesenchymal stem cells (MSCs) were isolated and characterized by Tissue Bank BulGen based on the expression of MSC-specific markers and absence of hematopoietic ones. Briefly, umbilical cord segments (2.5 cm) were dissected to remove blood vessels, minced, and enzymatically digested with 0.25% Collagenase I, 4 mg/mL hyaluronidase, and 1% Pen/Strep/A at 37 °C for 3 h in a humidified 5% CO₂ incubator. The digested tissue was diluted with saline, filtered through a 70 μm strainer, and centrifuged at 876× g for 10 min. The cell pellet was re-suspended in DMEM/F12, supplemented with 10% FBS and 1% antibiotic–antimycotic solution (all from Sigma-Aldrich), and then seeded at density 1 × 10⁴ cells per 25 cm² flask. Medium was refreshed every 48 h until cells reached around 80–90% density. At that point, the medium was replaced with fresh serum-free medium and cultured for 48 h before being collected. The collected medium containing secretome was centrifuged at 876× g for 10 min (to remove debris) and concentrated 10 times using Amicon Ultra centrifugal filter units (Amicon, Schorndorf, Germany). The protein concentration was measured using Bradford assay (Sigma-Aldrich, USA) and determined to be approx. 0.3 mg/mL. The secretome was then lyophilized and stored at −80 °C until use.

Collagen type I was produced from rat tail tendon by acetic acid extraction and salting out with NaCl, as described elsewhere [52]. The pellets were collected by centrifugation at 4000 rpm at 4 °C for 30 min and re-dissolved in 0.05 M acetic acid, then dialyzed to remove the excess NaCl. All procedures were performed at 4 °C. The collagen concentration in the solutions was measured by the modified Lowry assay [53] and from the optical absorbance at 220–230 nm.

Collagen-coated 24-well culture plates (Greiner Bio-One, Frickenhausen, Germany) or 22 × 22 mm glass coverslips (ISOLAB GmbH, Frankfurt am Main, Germany, placed in 6-well plates), were coated with 100 µg/mL rat tail collagen (isolated as above) dissolved in 0.05 M acetic acid via incubation at 37 °C for 1 h, before samples were rinsed three times with phosphate-buffered saline (PBS).

AD-MSCs and HDFs were seeded at 1.2 × 10⁴ cells/well in 600 µL of serum-free DMEM/F12 (Sigma-Aldrich, USA) medium with 1% antibiotic–antimycotic solution (Sigma-Aldrich, USA) on collagen-coated 24-well plates or alternatively on coverslips in 6-well plates. After 2 h of incubation (37 °C, 5% CO₂), WJ-MSCs' secretome was added in final protein concentration of 0.03 mg/mL without adding FBS. Control groups received only serum-free DMEM/F12 medium as a negative control or DMEM/F12 medium supplemented with 10% FBS as a positive control.

Cell proliferation was assessed at 24, 48, and 72 h of culture. Cells were fixed with 4% paraformaldehyde and stained with Hoechst 33258 (1:2000, Sigma-Aldrich) to visualize cells' nuclei, then viewed with fluorescence microscopy (Thunder Imager Live Cell, Leica, Wetzlar, Germany) with a 10× objective and counted using the CellProfiler software version 4.2.8 [54]. The doubling time was calculated by using the following formula:Doubling Time=T×ln2/[ln(Ne/Nb)] where Nb is the initial cell density, Ne is the final cell density, and T is the time passed between the two measurements.

Cell cycle distribution was determined in accordance with the protocol of Vassilis Roukos [55]. Briefly, the fluorescent images were taken on a fluorescent microscope with 10× objective. Then a pipeline provided by the team of Roukos was applied to the image set, which allowed the quantification of the DNA in each nucleus. In accordance with the protocol, after acquiring the fluorescent images, they were fed into the first pipeline which determined the illumination correction function to be used for removing the variance in the background between images. Next, all the images, together with the illumination correction function file, were analyzed using a second pipeline. It consists of the following important steps—(1) applying the illumination function, (2) determining the borders of the cells' nuclei and selecting them as objects, (3) measuring the total internal intensity of each object, and (4) exporting the information in '.csv' files. These files were remastered manually to fit the standard flow cytometry file format. For each experimental condition, at least 600 nuclei were analyzed. Cell cycle analysis was performed using Floreada.io (https://floreada.io/↗, accessed on 24 April 2025). This was performed using the flow cytometry analysis tool. The nuclear intensity data in the flow cytometry file format was uploaded to the website, and the standard cell cycle analysis Michael H. Fox algorithm [56], provided by the website, was applied. Nuclei were further classified into 3 groups according to the cell cycle phase (G1/G0, S, and G2/M) and plotted in 3 colored graphs (blue, green, and red, accordingly).

After 24 h of incubation in defined conditions, HDFs and MSCs were fixed in 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 for 5 min before being washed and treated with blocking solution (PBS with 10% FBS (Sigma-Aldrich, USA), 15 min), then incubated (37 °C) with rabbit anti-human collagen 1 (AB745, Sigma-Aldrich, USA) followed by Alexa Fluor^® 555-conjugated donkey anti-rabbit IgG (406412, BioLegend, San Diego, CA, USA) as the secondary antibody. All antibodies were applied in 1:100 dilution in PBS containing 10% FBS (Sigma-Aldrich, USA). The actin cytoskeleton was visualized using Alexa Fluor™ 488 phalloidin (Invitrogen, Carlsbad, CA, USA), and the nuclei were counterstained with Hoechst 33258 (1:2000). Fluorescent images were acquired using a fluorescence microscope (Thunder Imager Live Cell, Leica) with 20× objective. In order to determine the overall cell shape, the CellProfiler software [54] was used.

MSCs and HDFs were seeded in 12-well plates and incubated until rich around confluent cell density (90%) when a linear scratch was made across the cell monolayer using a sterile 200 µL pipette tip. The samples were studied either plain (medium only) or medium-supplemented with 10% WJ-MSCs secretome (from 10× concentrate) or with 10% FBS as positive control. The cells were incubated using live cell chamber of the inverted microscope (Thunder Imager Live Cell, Leica) in time-laps mode (each 15 min) during 24 h. Images of the wound area were captured at selected time points (0 and 24 h) with 10× objective under phase contrast.

The wound area was quantified using ImageJ software version 1.54p, Wayne Rasband, National Institute of Mental Health, NIH, Bethesda, MD, USA, with a plugin for the high throughput image analysis of in vitro scratch assays [57] and the percentage of wound closure was calculated as follows:% Closure=[(A0−At)/A0]×100 where A0 is the wound area at 0 h and At is the area at time t.

Statistical analysis was performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). After the Shapiro–Wilk test for normality was performed, the homogeneity of variances of the data was assessed using Bartlett's test, which confirmed that the assumption of equal variances was not violated (p > 0.05). Then the results were analyzed using one-way analysis of variance (ANOVA) followed by a correction for multiple comparisons using Tukey's Honest Significant Difference test. Unless otherwise stated, the experimental conditions have been performed in triplicate. Results were displayed as column bar graphs, as scatter plots with each point representing an independent replicate, as line graphs, or as percentage stacked bar charts. Mean ± SEM (or SD) is indicated as described in the figure legends. The confidence limit reflecting the significant difference between experimental groups was considered 95% when p < 0.05.