Enhancing the culture of mouse primary fibroblasts to study age-dependent effects in skin tissue engineering in vitro

Abstract

Introduction:

Aged dermal fibroblasts exhibit reduced migration, proliferation, and extracellular matrix (ECM) production, alongside upregulation of pro-inflammatory signaling that disrupts the proliferative, granulation, and remodeling phases of wound healing. Despite these age-related impairments, tissue engineering biomaterials are rarely evaluated using age-specific fibroblasts. Current in vitro skin aging models commonly rely on acute stressors, such as oxidative stress or DNA damage, which only partially recapitulate chronological aging. Alternatively, fibroblasts are artificially aged through extensive passaging, which may not reflect physiological aging. This work established a system for accurately representing chronological aging of mouse fibroblasts in vitro by isolating and evaluating fibroblasts from age-specific mice. While senescence in human fibroblasts results from telomere shortening, primary mouse fibroblasts senesce by oxidative damage in atmospheric O₂, requiring careful culture conditions.

Methods:

We demonstrate that physiological O₂ levels (3%) abrogate the detrimental effects of 20% O₂ by comparing cell proliferation, senescence markers, and morphology of explant-derived fibroblasts. Next, fibroblasts from young (5wks), middle-aged (47wks), and old (90wks) mice were cultured (3% O₂) and assessed for age-dependent changes in their proliferation, migration, immune signaling, collagen production, and metabolic activity in a 3D collagen matrix.

Results:

Our data demonstrate the impaired metabolic activity and migratory ability of old fibroblasts, as well as an overall amplification of pro-inflammatory proteins (e.g., MCP1 and IL-1α), which shifted tissue repair macrophages (“M2”) towards a pro-inflammatory (“M1”) phenotype. In 3D collagen-glycosaminoglycan scaffolds, old fibroblasts also showed significantly reduced metabolic activity compared to young and middle-aged cells.

Conclusion:

Together, these findings show that primary mouse fibroblasts can retain chronological age-related characteristics when cultured under physiological conditions in vitro, making them relevant in models of skin aging. Studying aging using these robust in vitro methods will be essential for improving the design and translational relevance of biomaterials and therapeutic strategies for wound healing and skin tissue engineering.

Graphical Abstract

Introduction

Aging is a universal biological process marked by a progressive decline in molecular and cellular function [1]. Along with the known systemic changes in individuals as they age, there are also cellular changes: an increase in genomic instability, mitochondrial dysfunction, cellular senescence, and chronic inflammation [2]. These changes impact basic cellular functions, such as migration, mitosis, and extracellular matrix (ECM) production. As wound healing is a complex, highly coordinated biological process that involves initial inflammation, tissue formation, and remodeling, these changes impact central processes in tissue regeneration, leading to delayed and impaired wound healing in older adults [3]. Consequently, these individuals are more prone to the development of chronic wounds (e.g., patients >65 yo account for 72% of pressure ulcers; diabetic foot ulcer related amputation rates rise with age (0.39% < 65 yo vs. 0.66% > 65 yo); and >80% of leg ulcers occur in patients over 65 yo) [4]. This motivates the consideration of age as a factor in designing wound treatments.

The skin is the largest organ in the human body, acting as a protective barrier against external insults. With age, the skin undergoes profound cellular and ECM changes that are visible to the naked eye, including decreased collagen, elastin, and glycosaminoglycans. The main cells of the dermis – fibroblasts – are responsible for the synthesis and remodeling of the ECM and play a central role in tissue regeneration and intercellular signaling. Chronological aging of fibroblasts is characterized by reduced migratory and proliferative capacity, reduced expression of ECM proteins, and upregulation of pro-inflammatory molecules [5]. Despite these critical changes in aged cells – and the increase in skin wounds with age – in tissue engineering and biomaterials research, age is often overlooked, particularly during the in vitro phase of design.

Instead, most in vitro studies on aging are used to study homeostasis and are done on human-derived cells – typically primary human cells or induced pluripotent stem cells (iPSCs). Primary human cells isolated from different age groups make up a reliable aging model since they retain accumulated epigenetic and metabolic memory associated with chronological aging. However, their availability is limited, and the results often show high donor-specific biological variation. On the other hand, while iPSCs offer improved experimental control, they do so at the expense of physiological relevance [6], as the aging phenotype has to be induced in these cells using different exogenous stressors (e.g., long-term culture, exposure to aged ECM, genetic engineering, and telomere shortening) [7]. While these human-derived approaches have the advantage of translational relevance, the overwhelming majority of in vivo aging research relies on mouse models, owing to their accessibility, cost, genetic proximity to humans, and their ability to recapitulate many phenotypes associated with human aging [8]. Laboratory mouse strains, such as C57BL/6J, have been instrumental in establishing models of accelerated aging and enabling genetic mapping studies that have led to the discovery of age-associated transcriptomic and proteomic changes [9,10]. As much of our current understanding of aging mechanisms is derived from murine systems, there is a critical need to bridge in vivo observations with in vitro cellular models to accurately interpret age-related biological processes. This necessitates the successful culture of age-dependent mouse fibroblasts.

One key challenge in successfully culturing primary mouse fibroblasts is that the mechanisms governing cellular senescence differ between mice and humans. While replicative senescence of human cells is triggered by telomere shortening, senescence in mouse fibroblasts is ruled by oxidative damage under atmospheric oxygen levels (20%), the level used in standard cell culture. Notably, mouse strains such as C57BL/6J are extremely susceptible to oxygen [11], which can accelerate senescence and confound interpretations of aging phenotypes in vitro. Hence, culture conditions under physiological oxygen levels (3%) are necessary to isolate and study age-dependent changes in healthy mouse fibroblasts.

This motivated us to identify robust techniques to isolate age-specific cells from mice that can be used to study age-dependent physiology and the impact of chronological aging on skin tissue engineering and biomaterials development. In this study, we first confirmed the detrimental effect(s) that atmospheric oxygen has on isolated primary fibroblasts by comparing cell proliferation, morphology, and senescence markers of cultures maintained at 20% O₂ vs 3% O₂. Next, dermal fibroblasts from young (5 weeks old), middle-aged (47 weeks old), and old mice (90 weeks old) were isolated and assessed for age-dependent changes in their proliferation, migration, immune signaling, collagen deposition, and metabolic activity in a 3D collagen matrix. This work establishes a culture model for accurately representing the aged phenotype of mouse fibroblasts, which can be employed in in vitro tissue engineering skin aging models, biomaterials development, and the study of age-dependent changes in in vitro engineered tissues.

Methods

Cell isolation and culture

Primary skin fibroblasts were obtained from healthy young (5 weeks old, n=6), middle-aged (47 weeks, n=6), and old (90 weeks, n=4) female and male mice. All animal procedures were approved by the University of Massachusetts Amherst Institutional Animal Care and Use Committee. Following euthanasia, 6 dorsal skin samples were collected from each animal using 6 mm biopsy punches. The explants were cultured with Dulbecco’s Modified Eagle Medium/Ham’s F-12 Nutrient Mixture (Gibco, 11-765-054 and 11-885-084), supplemented with 15% FBS (Biowest S1620), 1% penicillin-streptomycin (P/S) (Gibco 15–140–122), and sodium pyruvate (SP) (Gibco 11-360-070). The explants/cells from the three age groups were maintained at either 20% O₂ or 3% O₂ (37°C, 5% CO₂) by adjusting nitrogen levels in the incubator. Individual cells were observed leaving the explants by day 7 (shown in Suppl. Fig. 1). To ensure only fibroblasts were maintained in the culture, on day 14, the culture medium was changed to Eagle’s Minimum Essential Medium (EMEM; ATCC 30–2003) with 15% FBS, 1% P/S, and SP. Pooling of cell samples was used to identify broad differences in aging by smoothing it out across several animals; this additionally matches the way age-dependent cells are often provided from manufacturers. Further, during the isolation of primary mouse fibroblasts from skin explants, high variability was found in the number of cells initially recovered from individual samples, which further motivated the use of a pooled sample approach. Cellular identity was confirmed by immunocytochemistry using characteristic fibroblast markers, vimentin (Cell Signaling 9854 [1:400]), and CD90 (Thy1; ThermoFisher MS21838-RTM1-P1; 1.5μg/mL). For immunodetection of CD90, goat anti-rat IgG Secondary Antibody, DyLight^™ 594 (ThermoFisher SA5–10020; 2μg/mL) was incubated for 1h at RT, and nuclei were counterstained with 2 μg/mL Hoechst (Invitrogen H1938). Passages ≤6 were used for all experiments. Cell numbers and viability were quantified at each passage using a cell counter (Invitrogen, Countess 3) after seeding equal initial densities (200,000 cells for passages 2–4 and 250,000 cells for passages 5–6). Metabolic activity of cells (a combined reading accounting for the metabolic activity of individual cells times the number of cells) was assessed at passages 2, 4, and 6 using Cell Counting Kit-8 (Abcam 228554). In line with manufacturer recommendations, 3,000 cells were seeded per well in a 96-well plate to enable measurable differences in growth as a function of time. After 24h, fresh media was added along with 10μL of CCK8 solution for 2 h before measuring the absorbance (460 nm). Cell identity, number, viability and metabolic activity experiments were conducted using n=3 for each age groups.

Senescence-Associated Beta-Galactosidase Staining (SA-β-gal)

To evaluate cellular senescence, a β-galactosidase assay was performed using a staining kit (Cell Signaling 9860) according to the manufacturer’s instructions. Primary fibroblasts were seeded at a density of 1,800 cells/cm². Cells were allowed to grow for 48 h before commencing the assay, with fresh media changes every 24 h. Cells were then counterstained with 2 μg/mL Hoechst for 10 min at RT, and protected from light. Five images were taken per technical replicate, with each biological repeat (n=3, 15 images per biological repeat) consisting of three technical replicates. The total number of cells was quantified per image using NIS-elements General Analysis 3 (NIS-GA3) and used to normalize the number of SA-β-gal⁺ cells. Experiments were conducted using n=3 for each age group.

Cellular morphology

Primary fibroblasts were seeded as per the Senescence-Associated Beta-Galactosidase Staining experiment. To quantify morphological changes at different O₂ levels and age, cells were washed with 1X PBS and fixed with 4% paraformaldehyde for 10 min at RT. The cytoplasmic and nucleic area, and the nucleus to cytoplasm ratio of cells were calculated using F-actin (Phalloidin) (Abcam 176759) and Hoechst. Three images were taken per technical replicate, with each biological repeat (n=3, 9 images per biological repeat) consisting of three technical replicates using an EVOS M5000 microscope. Image analysis was done using NIS-GA3. Experiments were conducted using n=3 for each age group.

γH2AX and SIRT1 Immunostaining

Primary fibroblasts were seeded as per the Senescence-Associated Beta-Galactosidase Staining experiment. Fibroblasts were fixed with 4% PFA for 10 min and washed 3 times with 1X PBS. Non-specific binding was blocked with 5% goat serum in 0.3% Triton X-100/1X PBS for 1 h before immunostaining with γH2AX (Cell Signaling 9718; [1:250]) or SIRT1 (ThermoFisher PA5120542; [1:100]) overnight at 4°C followed by incubation with secondary antibody (Alexa 488; Invitrogen; 2 μg/mL), for 1 h at RT. Following washing and incubation with Hoechst, samples were imaged using an EVOS M5000 microscope. Image collection was done as per the cellular morphology experiment. Percent positive γH2AX cells were counted using ImageJ. The level of SIRT1 expression was determined using mean fluorescence intensity and area in NIS-GA3. Experiments were conducted using n=3 for each age group.

Mitochondria and Reactive Oxygen Species (ROS) Content

Cells were plated at 5000 cells/cm² in a 16-well chamber slide (Thermo Fisher Scientific 178599). At 80% confluence, fibroblasts were incubated for 30 min at 37 °C with 100 nM of MitoTracker^™ Deep Red FM (Cell Signaling 8778) and 5 μM of CellROX^™ Green (Thermo Fisher C10444). Cells were fixed using 4% PFA, and nuclear staining was performed using Hoechst. Images were visualized using an EVOS M5000 microscope. Experiments were conducted using n=3 for each age group.

Cytokine Profiler

The conditioned media from young, middle-aged, and old fibroblasts cultured at a physiological oxygen level (3% O₂) was collected at P3. The level of expression of 40 cytokines (normalized to cell number) was assessed using Mouse Cytokine Proteome Profiler Array (R&D Systems ARY006) according to the manufacturer’s instructions. Experiments were conducted using n=3 for each age group.

Macrophage Polarization Assay

RAW 264.7 mouse macrophages were kindly gifted from the Ashish Kulkarni Lab (University of Massachusetts Amherst, Department of Chemical and Biomolecular Engineering). Macrophages were cultured in expansion media (M0) consisting of Dulbecco’s Modified Eagle Medium (DMEM), 10% FBS, and 1% PS. Upon reaching 80–90% confluency, cells were detached by scraping and seeded into 48-well plate at 3,600 cells/cm². After 24 h, cells were gently washed with 1X PBS and cultured in one of the following conditions: unpolarized macrophages (M0), M1-polarizing medium containing 5 ng/mL interferon-γ (IFN-γ; Gibco 3150520UG) and 100 ng/mL lipopolysaccharide (LPS; Sigma-Aldrich L4391), or M2-polarizing medium containing 20 ng/mL interleukin-4 (IL-4; BioLegend 50170363) and 20 ng/mL interleukin-13 (IL-13; BioLegend 50170418). Cells were incubated under these conditions for 48 h.

Following polarization, cells were washed and treated for an additional 24h with a 1:1 dilution of the same conditioned media from young, middle-aged, and old fibroblasts used in the cytokine prolifer assay and M0-unpolarized media. Control groups consisting of 1:1 dilutions of M0, M1, and M2 media with primary fibroblast media (EMEM without FBS, supplemented with 1% PS and 1% SP) were also included.

After treatment, cells were fixed with 4% paraformaldehyde (PFA) for 10 min and washed three times with 1X PBS. Cells were permeabilized with 0.3% Triton X-100 in 1X PBS for 5 min. Non-specific binding was blocked first with TruStain FcX^™ PLUS (anti-mouse CD16/32; BioLegend 156603; [1:200]) for 10 min, followed by incubation in 5% goat serum in 0.3% Triton X-100/PBS for 50 min. Cells were then immunostained overnight at 4°C with primary antibodies against F4/80 (Cell Signaling 70076; [1:500]), iNOS (ThermoFisher 740025TP647; [1:500]), and CD206 (ThermoFisher MA5–16871; [1:200]). After washing, cells were incubated with secondary antibodies, Alexa Fluor 488 (Invitrogen; 2 μg/mL) and goat anti-rat IgG DyLight^™ 594 (ThermoFisher SA5–10020; 2 μg/mL) for 1h at RT. For imaging, three fields of view were captured per technical replicate using an EVOS M5000 microscope. Each biological replicate (n = 3) consisted of two technical replicates (total of six images per biological replicate). Mean fluorescence intensity was quantified using NIS-Elements GA3 software. Experiments were conducted using n=3 for each age group.

Migration assay

The migratory ability of primary fibroblasts cultured at 3% O₂ was assessed by performing an in vitro scratch assay. Cells were plated at 17,000 cells/cm² to ensure appropriate 80–90% confluency for experiment. The following day, the cell monolayer was scratched using a p200 pipette tip. Debris and dead cells were removed by washing 3 times with 1X PBS, then cells were replenished with fresh media and imaged at 0 h, 4 h, 8 h, 12 h, 20 h, and 24 h. Wound area was quantified using ImageJ, and closure percentage was calculated by normalizing to the wound area at 0 h. Experiments were conducted using n=3 for each age group.

Collagen-Glycosaminoglycan (Collagen-GAG) scaffold fabrication

Collagen slurry was prepared by blending bovine tendon collagen (5 mg/mL; Collagen Solutions, UK) in 0.5 M acetic acid, following established protocols [12,13]. Briefly, the solution was homogenized for 90 min at 15,000 rpm using an Ultra-Turrax T18 (IKA Works) with temperature control (Lauda WKL250). Chondroitin 6-sulfate (Toronto Research Chemicals) was dissolved in 0.05 M acetic acid (0.44 mg/mL) and added dropwise to the slurry during blending, followed by an additional 90 min of blending. The collagen-GAG slurry was degassed under vacuum (0 mmHg, 30 min), freeze-dried (Christ Epsilon 2–4; −10°C, 17 h), and cut into 6-mm cylindrical scaffolds. Scaffolds were thermally crosslinked and sterilized under vacuum (104°C, 24 h, 60 mmHg), hydrated overnight in sterile PBS, and chemically crosslinked for 2 h at room temperature using EDAC/NHS (2.17 mg EDAC and 0.52 mg NHS per scaffold in 2 mL sterile ddH₂O) with orbital shaking. Scaffolds were washed twice in sterile PBS (30 min each) and stored in PBS for up to one week before their use. For fibroblast studies, scaffolds were seeded with 250,000 cells/scaffold. To track cellular viability, alamarBlue^™ (ThermoFisher, A50100) was used at days 3, 6, 9, 12, and 14 after seeding. Experiments were conducted using n=3 for each age group.

Self-assembled matrix collagen content

Cells were seeded at a confluency of 16,000 cells/cm² with cell medium refreshed every 72 hours for 3 weeks, based on previous studies [14]. For the groups receiving ascorbic acid treatment, the medium was supplemented with 100 μg/ml ascorbic acid. For collagen quantification, samples were collected after 3 weeks and resuspended in 0.5M acetic acid with 0.1mg/mL pepsin (Sigma-Aldrich, P6887). Total collagen was analysed with Sircol Soluble Collagen Assay (Biocolor, S1000) according to the manufacturer's instructions. Experiments were conducted using n=3 for each age group.

Statistical analysis

All statistical analyses were carried out using GraphPad Prism 10. Graphs express means ± standard deviation (SD). One-way ANOVA and Tukey’s post hoc test were used to compare the significant differences between three or more groups. One-way ANOVA and Dunnett’s post hoc test were used to compared the significant differences between groups and a control group. Two-way ANOVA and Tukey’s post hoc test was used when more than one factor was involved in the experiment, with select multiple comparisons performed as listed in the relevant figure captions. Results were considered significant at *p < 0.05, **p < 0.01, *** p<0.001.

Results

Aging and Atmospheric Oxygen Negatively Impact Primary Mouse Fibroblast Proliferation

A hallmark of aging is the progressive decline in dermal fibroblast proliferative capacity [2]. To assess age- and oxygen- dependent changes, we evaluated the proliferative capacity, viability, and metabolic activity of aged fibroblasts in vitro. After explant collection, migration of primary cells out of tissues was observed between day 7 and day 14 (shown in Supp. Fig. 1). By day 14, insufficient numbers of middle-aged and old fibroblasts cultured under atmospheric oxygen conditions (20% O₂) were obtained to sustain continued culture; consequently, only young fibroblasts were analyzed in this condition. Fibroblast identity was confirmed by immunostaining for vimentin and CD90, canonical markers of dermal fibroblasts (shown in Fig. 1A). After successful confirmation of cell identity, we followed fibroblasts from young, middle-aged, and old mice cultured at physiological oxygen levels (3% O₂) versus young fibroblasts placed at 20% O₂ over 6 passages. Across all passages, fibroblasts cultured at 3% O₂ consistently reached higher cell numbers than young fibroblasts cultured at 20% O₂ (p < 0.05). When comparing between age groups maintained at physiological oxygen levels, young and middle-aged fibroblasts achieved greater cell numbers at passages 3, 4, and 6 compared to old fibroblasts, indicating an age-associated decline in proliferative capacity (shown in Fig. 1B); however, this was only significant between middle-aged and old cells at P3, 4, and 6 (p < 0.05).

Fig. 1.

Aging and atmospheric oxygen level impact the proliferation, viability, and metabolic activity of primary mouse fibroblasts. (A) Representative images of characteristic fibroblast markers, vimentin (GFP) and CD90 (TX-Red), confirming fibroblast phenotype (nucleus = blue; scale bar = 150μm. (B) Cell number recorded over 6 passages for each cell group. Dashed lines represent the initial cell seeding number at 200,000 cells (P2-P4) and 250,000 cells (P5-P6). (C) Cell viability (%) of primary mouse fibroblasts at 6 passages. (D) Metabolic activity recorded using CCK8 at passages 2, 4, and 6 (mean ± SD; n=3, *p < 0.05, **p < 0.01, *** p<0.001). Data was analyzed using one-way ANOVA (Tukey’s post hoc test).

The reduced cell numbers observed in young fibroblasts cultured at 20% O₂ were accompanied by a significant decrease in cell viability relative to all groups maintained at 3% O₂ (p<0.05) (shown in Fig. 1C). In contrast, no significant differences in viability were detected among age groups cultured at 3% O₂ from passage 3 onward, with viability remaining above 85% across all passages. CCK8 analysis demonstrated that young fibroblasts cultured at 3% O₂ exhibited significantly higher metabolic activity than middle-aged and old fibroblasts at passages 2 and 6 (p < 0.05). Overall, metabolic activity declined in a passage-dependent manner across all groups. Interestingly, young fibroblasts cultured at 20% O₂ displayed a progressive increase in metabolic activity with increasing passages and exhibited significantly higher metabolic activity than all other groups at passages 4 and 6 (p < 0.01) (shown in Fig. 1D).

Quantity of senescent cells in the culture of primary mouse fibroblasts is oxygen-dependent

With aging, senescent cells progressively accumulate in the skin, contributing to tissue dysfunction through permanent growth arrest and paracrine signaling to neighboring cells [15]. Here, we wanted to clarify whether mouse fibroblasts isolated from young, middle-aged, and old mice cultured under 3% oxygen exhibited different rates of senescence in vitro. Under 3% oxygen, SA-β-gal staining and analysis of fibroblasts from all age groups revealed low levels of senescence, with no significant differences observed between young (2.5 ± 0.5%), middle-aged (2.2 ± 0.1%), and old (2.6 ± 0.6%) cultures (shown in Fig. 2A–B). By contrast, young fibroblasts cultured under 20% oxygen displayed a marked increase in senescence compared to all age groups maintained at 3% oxygen (20.2 ± 3.7%, p < 0.001). Apart from the expression of SA-β-gal, senescent cells are characterized by a flat and enlarged morphology [16]. Young fibroblasts under 20% oxygen had significantly larger nucleic area (p<0.05), nucleic to cytoplasmic ratio (p<0.01), as well as a 3-fold increase in cytoplasmic area (p<0.001) compared to those maintained at 3% oxygen (shown in Fig. 2C–D). The presence of senescence was further assessed by the increased percentage of γH2AX-positive cells (p<0.05) compared to young fibroblasts maintained at 3% oxygen (shown in Fig. 2E) and reduced expression of SIRT1 (p<0.01) compared with middle-aged fibroblasts (shown in Fig. 2F).

Fig. 2.

The quantity of senescent cells in the culture of primary mouse fibroblasts is oxygen-dependent. (A) Representative images of senescent associated-β Galactosidase (SA-b-gal) assay on age groups placed at 3% oxygen and 20% oxygen (scale bar = 300μm). (B) Quantification of SA-b-gal positive percentage (%) of cells. (C) Representative images of oxygen concentration and age effects on cellular morphology, stained for F-actin (CY5) (nucleus = blue; scale bar = 300μm). (D) Nucleic area, cytoplasmic area, and nucleic to cytoplasm ratio (N:C ratio) of aged fibroblasts under different oxygen levels. (E) Quantification of γH2AX-positive nuclei. (F) Quantification of the level of expression of SIRT1 in aged fibroblasts cultured under different oxygen levels (mean ± SD; n=3, *p < 0.05, **p < 0.01, *** p<0.001). Data was analyzed using one-way ANOVA (Tukey’s post hoc test).

Aging and atmospheric oxygen disrupt mitochondrial metabolism of primary mouse fibroblasts

Mitochondrial dysfunction has been identified as a primary hallmark and driver of aging, as well as cellular senescence [17]. To evaluate the age- and oxygen-dependent changes in mitochondrial metabolism, we labeled these organelles with a mitochondrial potential-dependent dye. Therefore, differences in fluorescence intensity reflect changes in mitochondrial polarization rather than mitochondrial mass. Overall, we observed a decrease in fluorescence intensity in older cells, associated with a decrease in mitochondrial membrane potential (shown in Fig. 3A). Primary mouse fibroblasts cultured at 3% oxygen exhibited an extensive, elongated, and interconnected mitochondrial network distributed throughout the cytoplasm. On the contrary, mitochondria from young fibroblasts cultured at 20% oxygen had pronounced perinuclear accumulation, visualized as a high-intensity MitoTracker Deep Red signal surrounding the nucleus only. This change in mitochondria distribution was accompanied by an increase in the quantity of ROS in this area, as observed by fluorescent analysis of 20% oxygen fibroblasts compared to 3% oxygen fibroblasts. We did not observe an increase in the quantity of ROS between age groups placed at 3% oxygen (shown in Fig. 3B).

Fig. 3.

Aging and atmospheric oxygen disrupt the mitochondrial metabolism of primary mouse fibroblasts. (A) Representative images of mitochondrial potential and morphology in young, middle-aged and old fibroblasts cultured at 3% and 20% O₂ using Mitotracker (n=3; scale bar = 75μm). (B) Representative images of the presence of ROS in aged mouse fibroblasts cultured under different oxygen levels (n=3; scale bar = 75μm).

Aging negatively impacts the migration of primary mouse fibroblasts

We evaluated age-dependent changes in the migratory ability of young, middle-aged, and old primary mouse fibroblasts by performing a scratch assay and analyzing wound closure percentages intermittently between 0 and 24 h. We observed an age-dependent decline in the migratory ability of primary mouse fibroblasts. Old fibroblasts exhibited reduced migration at all time points compared to young and middle-aged fibroblasts, with significant differences observed between middle-aged and old fibroblasts at 4, 8h and 24 hours (4h; 9.32% vs 6.78%, 8h; 16.16% vs 11.76%, 24h; 74.6% vs 59.1%, p < 0.05) and between young and old fibroblasts at 20 hours (57.79% vs 51.57%, p < 0.05) (shown in Fig. 4A–B).

Fig. 4.

Aging negatively impacts the migration of primary mouse fibroblasts. (A) Representative images of cell migration (scratch assay) in young, middle-aged, and old mouse dermal fibroblasts. (scale bar = 750μm). (B) Quantification of wound closure percentage (%) of different groups from 0 – 24 hrs post wounding (mean ± SD; n=3, *p < 0.05, **p < 0.01, *** p<0.001). Data was analyzed using Two-way ANOVA test (Tukey’s post hoc test) with multiple comparisons used to only compare between ages at each individual timepoint.

Aging is linked with an overall increase in immune signaling of primary mouse fibroblasts

The secretomic and proteomic profiles of dermal fibroblasts shift with age [18]. Here, we did a screen of the secretome of 2D cultured young, middle-aged, and old fibroblasts cultured under physiological 3% oxygen level. Old fibroblasts had an all-around upregulation of pro-inflammatory chemokines and cytokines compared to young and middle-aged fibroblasts. Notably, Monocyte Chemoattractant Protein 1 (MCP-1/JE; also known as CCL2), secretion was significantly elevated in old fibroblasts relative to both younger groups (p<0.01 vs middle aged cells and p<0.05 vs young cells). Additionally, IP-10 (Interferon Gamma-induced Protein 10), C5/C5a (p<0.01), IL-1α, I-TAC/CXCL11, IL-3 and GM-CSF (Granulocyte-macrophage colony-stimulating factor) (p<0.01) were significantly increased compared to middle-aged fibroblasts (p<0.05). B-lymphocyte chemoattractant (BLC) expression was upregulated in young vs middle aged fibroblasts (p<0.05), while still reaching higher levels in old fibroblasts compared to both (p<0.01 vs young and p<0.001 vs middle aged cells; shown in Fig. 5A).

Fig. 5.

Aging is linked with an overall increase in the immune signaling of primary mouse fibroblasts, which influences mouse macrophage polarization. (A) Level of expression of select pro-inflammatory, (B) anti-inflammatory, and (C) dual role proteins graphed as pixel density/cell number. Data was analyzed using multiple one-way ANOVA (Tukey’s post hoc test). (D) Representative images of M0, M1 or M2-polarized RAW 264.7 mouse macrophages stained with M1 marker, iNOS (Cy5) and M2 marker, CD206 (TX-Red) after treatment with 1:1 dilution of young, middle-aged, or old fibroblast conditioned media (scale bar = 150μm). (E) Quantification of the level of expression of iNOS and CD206. Data was analyzed using one-way ANOVA (Tukey’s post hoc test) (mean ± SD; n=3, *p < 0.05, **p < 0.01, *** p<0.001).

Simultaneously, the proteomic secretome from old fibroblasts also exhibited an overall increased expression of anti-inflammatory proteins compared to young and middle-aged fibroblasts (e.g., IL-4 (p<0.01) and Granulocyte- Colony Stimulating Factor (G-CSF) (p<0.001 vs. middle-aged fibroblasts) (shown in Fig. 5B). Analysis of proteins with context-dependent immunomodulatory roles revealed that old fibroblasts secreted significantly higher levels of Tissue Inhibitor Metalloproteinase-1 (TIMP-1) compared to middle-aged fibroblasts (p< 0.05). TARC/CCL17 was also significantly increased in old fibroblasts relative to middle-aged fibroblasts (p < 0.05) (shown in Fig. 5C). The complete cytokine profiler results can be found shown in Supp. Fig. 2.

Age-dependent Fibroblast Signaling influences macrophage polarization

The secretome of dermal fibroblasts can influence surrounding macrophages towards a more pro-inflammatory or tissue repair phenotype that drives wound healing. To study this relationship, we treated M0, M1 or M2-polarized macrophages with conditioned media derived from young, middle-aged, or old fibroblasts, and analyzed the expression of characteristics phenotype markers (iNOS- M1 and CD206-M2; fig 5E). An overall expansion in macrophage population was observed across all treatment groups, including controls treated with 1:1 dilutions of unconditioned fibroblast media. M0 macrophages exposed to conditioned media from old fibroblasts exhibited reduced expression of CD206 compared to those treated with conditioned media from young fibroblasts (p=0.07). There were no significant differences in the expression of M1 or M2 markers in M1-polarized macrophages. However, M2-polarized macrophages treated with conditioned media from old fibroblasts had an increased expression of iNOS (p<0.05, vs. conditioned media from young fibroblasts) accompanied by a reduction in the expression of CD206 marker (p<0.05, vs. conditioned media from middle-aged fibroblasts) (shown in Fig. 5D–E). As we were changing the baseline media macrophages were cultured in (i.e., 50% macrophage media:50% conditioned fibroblast media, 1:1), we ran controls with 1:1 non-conditioned fibroblast media (Supp. Fig. 3). For all macrophage phenotypes, control 1:1 media was not significantly different from full macrophage media. When 1:1 control media was compared with conditioned media, M1-polarized macrophages had a reduction in iNOS expression (all groups; p < 0.05), and CD206 expression was significantly decreased in M1 and M2 polarized macrophages treated with old fibroblast conditioned media (shown in Supp. Fig. 3).

Ascorbic acid increases collagen deposition of primary mouse fibroblasts in culture, while aging negatively impacts metabolic activity of cells in a 3D collagen-GAG scaffold.

Fibroblasts are the main cells responsible for collagen production for ECM deposition and remodeling. In vivo during aging, there is a decrease in collagen production accompanied by an increase in collagen fragmentation by matrix metalloproteinases (MMPs) [19]. Treatment of fibroblasts with ascorbic acid (AA) has been shown to increase collagen production and deposition in self-assembled matrices in in vitro cell cultures [20]. Collagen mass in all age groups was enhanced by the external addition of AA when compared to the no-treatment group (p<0.05). When supplemented with AA, middle-aged fibroblasts had a significant increase in total soluble collagen mass compared to young fibroblasts (p<0.001). Collagen mass in old fibroblasts with ascorbic acid exhibited greater variability in results than in young and middle fibroblasts in the same condition (shown in Fig. 6A).

Fig. 6.

Ascorbic acid treatment increases collagen deposition in primary mouse fibroblasts, while aging negatively impacts metabolic activity of cells in a 3D collagen-GAG scaffold. (A) Quantification of total soluble collagen (μg) produced by young, middle-aged, and old fibroblasts with or without ascorbic acid treatment (AA). Data was analyzed using Two-way ANOVA with preselected pairs to compare with and without AA at the same age, or between age groups either with or without AA. (B) Metabolic activity of aged fibroblasts seeded on collagen-GAG scaffolds during days 3, 6, 9, 12 and 14 (mean ± SD; n=3, *p < 0.05, **p < 0.01, *** p<0.001). Data was analyzed using Two-way ANOVA test (Tukey’s post hoc test) with multiple comparisons used to only compare between ages at each individual day.

Finally, for use in in vitro tissue engineering models and biomaterials research, we wanted to study the effect of age when grown in a clinically-proven collagen-glycosaminoglycan scaffold [21]. When seeding the age-specific cells in 3D Collagen-GAG scaffolds, we observed a continuous significant reduction in metabolic activity of old fibroblasts throughout the duration of the study (p<0.01 on days 3 –14 vs. middle-aged fibroblasts and p<0.05 on days 3–12 vs young fibroblasts) (shown in Fig. 6B), confirming that age dependent effects can be recapitulated in 3D scaffolds in vitro.

Discussion

While the interplay of aging and skin health has been extensively studied, there is less consideration for age as a factor when developing biomaterial-based tissue engineering solutions, particularly during in vitro studies. Since most initial preclinical studies are performed in mouse models, we wanted to develop methods to successfully isolate and culture age-specific fibroblasts and to analyze the behavior of these cells. To do this, we established a model that accurately conserves the chronological aging features of primary mouse fibroblast in vitro. Existing 2D in vitro skin aging models rely on extrinsic manipulation of primary human and mouse cells, using stressors such DNA damage and replicative senescence, to induce an aging-like features [22]. However, these methods only offer a partial view into the complex mechanisms of skin aging by these short-term inductions, which limits the ability to fully recapitulate a chronologically aged cell. Outcomes depend on the type, intensity and duration of the applied stressor, often lacking information about physiologically relevant doses, with some models exposing cells to supra-physiological oxygen levels [23]. Herein, we first demonstrated the advantage of culturing primary fibroblasts at physiological levels of O₂. We then compared young, middle-aged, and old fibroblasts across several assays, which broadly showed a decline in function for the old cells versus the young and middle-aged cells in terms of proliferation, metabolic activity, migratory capacity, and the presence of a more pro-inflammatory phenotype. Excitingly, this matches in vivo data on age-dependent fibroblast changes, and importantly, these age-dependent effects were maintained in a clinically relevant 3D tissue engineering scaffold, as observed by a decreased metabolic activity of old cells. Given the high need for biomaterials and tissue engineering treatments, particularly in aged individuals, in vitro screening of biomaterials should consider these age-related changes. Furthermore, given the increasing focus on the use of in vitro engineered tissues to replace in vivo preclinical models, enhancing the physiological relevance of these models to account for age is critical [24]. While ultimately this will be best done with human cells, most initial in vivo studies are completed in mice and given the abundant availability of these cells compared to human cells, there is a clear benefit to developing these age-dependent in vitro models.

A major limitation for the in vitro study of mouse dermal fibroblasts is their high sensitivity to atmospheric oxygen. This susceptibility complicates the reproduction of in vivo aging phenotypes observed in murine models for in vitro study and validation. It has previously been shown that mouse fibroblasts cultured under atmospheric oxygen (20% O₂) undergo premature proliferative arrest after 12–14 population doublings, due to persistent DNA damage and activation of senescent pathways [25]. Consistent with this, our data demonstrated that mouse fibroblasts exposed to 20% oxygen exhibited pronounced senescence-like changes as early as passage 3 (approximately 3 weeks in culture). Indeed, in our middle-aged and old cultures, we were unable to isolate sufficient fibroblasts from these tissues when cultured at 20% O₂. While we were unable to study this in detail, given the low cell number, it can be interpreted that the combination of age-related changes and oxidative stress greatly limited the migratory capacity or viability of aged cells. For the young cells, exposure to 20% oxygen resulted in diminished cell proliferation and viability in comparison to mouse fibroblasts cultured under physiological 3% oxygen. This decline was accompanied by robust expression of markers of oxidative stress-induced senescence, including increased SA-β-gal activity, elevated γH2AX expression, and altered cellular morphology. Unexpectedly, this senescent transition caused increased metabolic activity as detected by CCK-8. We hypothesized that this was partially caused by mitochondrial dysfunction and redox imbalance, as CCK-8 quantifies the reduction of water-soluble WST-8 salt by intracellular NAD(P)H dehydrogenases. Mitochondrial dysfunction in senescence is characterized by inefficient oxidation of NADH and a reduced NAD⁺/NADH ratio. In parallel, accumulation of DNA damage activates NAD⁺ consuming enzymes, depleting intracellular NAD⁺ while accumulating NADH [17]. Thus, we believe that the increased CCK-8 signal observed here comes from these contributions and is driven by senescence rather than representing improved cell viability or mitochondrial function. Indeed, the cell number and viability were decreased for 20% cultured cells versus 3% cultured cells. Together, these results demonstrate that senescence induction in primary mouse fibroblast cultures is strongly influenced by oxygen level and can be minimized by maintaining cells under physiological oxygen conditions. Importantly for the culture of age-dependent cells, this observation is independent of donor age, which allows us to more accurately conserve the chronological aging phenotype of the cells by culturing at 3% O₂.

In vivo, aging is characterized by a diminished proliferation of dermal fibroblasts. In this context, we expected primary mouse fibroblasts isolated from old mice to show reduced proliferation when cultured in vitro for an extended period. In this study, we observed a significant reduction in cell numbers of old fibroblasts compared to young and middle-aged counterparts when plated at the same initial cell density. Thus, primary fibroblasts from old mice seemed to exhaust their proliferative capacity much faster than those from young or middle-aged mice in vitro, in agreement with previous reports [26,27]. We did not observe any difference in the quantity of senescent cells between age groups or in the accumulation of DNA damage. However, important differences were observed in the level of expression of sirtuin 1 (SIRT1). First, our results show a decrease in SIRT1 expression in old fibroblasts (3% oxygen; significant versus middle-aged fibroblasts). SIRT1 is heavily involved in attenuating inflammation and senescence in human dermal fibroblasts [28]. SIRT1-deficiency in mouse fibroblasts promotes senescence and fibrosis pathways, while its expression is associated with protection from age-related diseases [29,30]. SIRT1 was similarly low in young 20% fibroblasts, consistent with increased senescence. Moreover, the expression of SIRT1 is NAD⁺ dependent [31], which is decreased in senescent cells and correlates with our discussion above relating to increased NADH.

Mitochondria regulate cellular metabolism and oxygen consumption. Aging has been associated with reduced mitochondrial membrane potential (Δψm) in both human and mouse fibroblasts [32–34], which consequently causes an increase in the generation of reactive oxygen species (ROS) in aged skin [35]. In agreement with these reports, we observed an age-dependent decrease in mitochondrial membrane potential. However, this reduction was not accompanied by a measurable increase in cellular ROS in old fibroblasts under the conditions examined. Nevertheless, mitochondrial dysfunction led to an observable increase in the quantity of ROS in young fibroblasts cultured under 20% oxygen.

Fibroblast aging is associated with impaired interactions with the surrounding ECM, leading to diminished migratory capacity during wound repair [36]. Consistent with this, old fibroblasts exhibited significantly reduced migration in our assay, hinting at a slower migratory or proliferative response after injury. Aging is also characterized by a decrease in the production of collagen, a major component of the ECM. The addition of supplemental ascorbic acid in our study revealed differences in total soluble collagen mass between age groups. We were surprised to observe that middle-aged fibroblasts exhibited increased collagen deposition with the treatment of ascorbic acid compared to young fibroblasts. This can be due to a variety of reasons, including differences between matrix deposition and assembly, collagen turnover, increased presence of myofibroblasts, or differences in the expression of matrix metalloproteinases (MMPs) and Tissue Inhibitor of Metalloproteinases (TIMPs), as TIMP-1 was significantly upregulated in the secretory profile of old fibroblasts compared to middle-aged fibroblasts. Additionally, while pepsin was used to solubilize the tissue prior to the measurement, it may be more accurate to examine the insoluble collagen directly in future studies.

Dermal fibroblasts may further contribute to cutaneous aging through the production of cytokines, chemokines, and other soluble factors that exert both autocrine and paracrine effects on surrounding cells, thereby promoting a pro-inflammatory environment and extracellular matrix (ECM) dysregulation [18]. Chronic low-grade inflammation, known as inflamm-aging is now a recognized hallmark of aging [37]. The analysis of the secretory profile of old fibroblasts revealed an overall increase in the expression of pro- and anti-inflammatory proteins for these cells, compared to both young and middle-aged fibroblasts. Previously, the analysis of conditioned media from ear-derived mouse fibroblasts showed increased levels of both pro, and anti- inflammatory cytokines, including IL-1α and IL-4 [38]. Among the elevated pro-inflammatory cytokines in our data, MCP-1 – which had an increased expression in old fibroblasts – has been used as a biomarker of chronological aging in both humans and mice. It is also part of the senescence-associated secretory phenotype (SASP) and involved in multiple age-related diseases [39,40]. These immune signaling changes are important to consider; fibroblasts are increasingly being recognized and studied as sentinel cells, and as initiators and amplifiers of the immune response. In turn, the secretion of these proteins could activate and recruit surrounding immune cells. Resolution of inflammation in wound healing is impaired in aged skin and is strongly related to a systemic dysregulation of the immune system [19]. Our data shows that when treating M2-polarized macrophages with conditioned media from old fibroblasts, these cells had increased expression of the inflammatory marker iNOS and reduced expression of the tissue repair marker CD206, hinting at macrophage phenotype switching. The effect of fibroblast secretome on macrophage polarization has previously being studied in the context of quiescent vs activated fibroblasts [41]. To our knowledge there are no previous reports on the effect of aging fibroblasts on macrophage polarization in vitro.

In terms of the use of these cells in 3D models, we successfully grew cells from each age group in a 3D collagen-scaffold and found a reduction in metabolic activity in older cells, consistent with 2D findings. This motivates the use of age-dependent cells in the in vitro study of tissue engineering products that are planned for use in the treatment of aged individuals. It also suggests that we can recapitulate the effects of chronological aging in in vitro tissue mimics, thereby enhancing physiological relevance. While the work here demonstrates the potential for this, further model systems and analyses will be necessary to fully elucidate the age-related effects of these cells in a broader spectrum of in vitro models. For example, the use of collagen-coated culture surfaces for fibroblast culture may better mimic the native extracellular matrix environment in which dermal fibroblasts reside following explant isolation. This model can also be applied to investigate potential sex-dependent differences in the aging process. Lastly, age-specific cells can be incorporated into in vitro skin equivalents or other skin-on-chip models to further model the multicellular and structural complexity of the skin. These directions could contribute to the physiological relevance and applicability of the model for studying mechanisms of skin aging.

Collectively, our data show that chronological aging features in mouse dermal fibroblasts are mostly retained in an in vitro model of cell culture by adjusting the oxygen level to physiological conditions, making this approach a reliable model for the age-dependent assessment of cell behaviour and function. This method can be used in in vitro skin tissue engineering models to elucidate further the impact of aging on wound healing and enable the design of therapeutic strategies to combat it, as well as the potential screening of anti-aging compounds. Additionally, the exposure of mouse fibroblasts to standard atmospheric oxygen levels reliably induces oxidative stress-mediated senescence, offering a simple and reproducible in vitro platform to study senescent-associated mechanisms under physiologically relevant yet treatment-free conditions.

Data Availability Statement

All data generated or analyzed during this study are included in this article and its supplementary material files. Further enquiries can be directed to the corresponding author.