Comparison of 3D culture systems on mesenchymal stem cell longevity, regenerative potential, and secretome production

ABSTRACT

Background/Aims

Mesenchymal stem/stromal cells (MSCs) are widely investigated for regenerative therapies, yet current expansion methods often compromise their stem-like properties, limiting large-scale clinical translation. We aimed to evaluate whether a novel hydrogel-based Bio-Block® platform preserved intrinsic adipose-derived MSC (ASCs) phenotype and secretome compared with conventional systems.

Materials/Methods

ASCs were cultured for four weeks in 2D, spheroids, Matrigel, or Bio-Blocks. Cultures were assessed for proliferation, senescence, apoptosis, trilineage differentiation, stem-like gene expression, secretome protein and extracellular vesicle (EV) production, and EV potency on endothelial cells (ECs).

Results

Bio-Block ASCs exhibited ~2-fold higher proliferation than spheroid and Matrigel groups, with senescence reduced 30–37% and apoptosis decreased 2–3-fold. Trilineage differentiation and stem-like markers (e.g. LIF, OCT4, IGF1) were significantly higher in Bio-Block ASCs. Secretome protein declined 35%, 47%, and 10% in 2D, spheroid, and Matrigel, respectively, but was preserved in Bio-Blocks. Similarly, EV production increased ~44% in Bio-Blocks, while other systems declined 30–70%. Bio-Block EVs enhanced EC proliferation, migration, and VE-cadherin expression, whereas spheroid EVs induced senescence and apoptosis.

Conclusion

This study highlights the critical influence of culture systems on scaling robust MSC-based therapies and introduces a biomimetic platform that represents a potential scalable strategy for producing high-potency and robust regenerative therapies.

1. Introduction

The intersection of biological and engineering sciences has catalyzed the growth of tissue engineering and regenerative medicine (TERM), a field focused on repairing or replacing damaged tissues in the human body with direct clinical application [1–3]. TERM integrates cells, biomaterials, and biomodulatory factors to facilitate tissue regeneration tailored to specific needs [1–4]. Among the cell types utilized, mesenchymal stem/stromal cells (MSCs) stand out due to their intrinsic multipotency and accessibility from diverse tissue sources [5]. MSCs have the capacity to differentiate into multiple cell lineages, making them pivotal in TERM applications [6,7]. This versatility underpins their use in creating neo-tissues and whole-organ replacements critical for developing the next generation of translational clinical therapies, with ongoing research exploring methods such as scaffold-based implantation into tissue defects, seeding decellularized organs, or scaling functional tissue (or extracellular matrix) production for transplantation [7–14].

In the aforementioned studies, precise control of MSC fate is often paramount. For example, scaffold physical properties can guide osteochondral differentiation [14–16], bioreactor perfusion and dynamic cell culture systems can help promote integration and differentiation of MSCs for cardiac engineering [17–19], and MSC-derived biomodulatory factors can drive pancreatic cell regeneration and functionality for treatment of diabetes [20–22]. Taking a step back, one can see that a fundamental principle for these developing preclinical/clinical therapies is the reliance on the intrinsic properties of MSCs, such as their multipotency, in order to be able to drive regenerative activity for cells/tissues of interest.

Beyond differentiation potential, advancements have highlighted the importance of MSC engraftment and retention in tissue microenvironments [23,24]. Enhanced retention extends MSCs’ interaction with local tissues, where they adaptively modulate regeneration by secreting bioactive factors tailored to environmental cues, such as anti-inflammatory cytokines in inflamed tissues or pro-angiogenic factors in fibrin-rich matrices [25–27]. This dynamic adaptability has inspired therapies that leverage the MSC secretome, which consists of a heterogenous collection of biomodulatory proteins (e.g., growth factors and cytokines), nucleic acids (e.g., microRNA), and small molecules (e.g., antioxidants), to target specific regenerative pathways [28,29].

MSCs typically secrete these factors either in free/soluble forms or encapsulated into vesicular entities, such as extracellular vesicles (EVs) like exosomes [30,31]. EVs offer greater stability and control over target cell populations and have been shown to be key regulators of wound healing and tissue regeneration [32,33]. EVs encapsulate biomodulatory cargo that varies with environmental stimuli, cellular health, and phenotype, further demonstrating the dynamic nature of MSCs and their secretome. More specifically, previous studies have shown that the quality and potency of MSC-derived EVs can depend heavily on culture conditions, surface interactions, and culture system design in the context of biomanufacturing and translation of new therapies [33–35].

Despite some encouraging preliminary studies pertaining to MSC therapies, there are still significant challenges to overcome [36,37]. Inconsistencies and lack of reproducibility in efficacy and outcomes persist [36–38]. Current MSC culture systems (predominantly 2D monolayers) fail to mimic native tissue environments, often compromising MSC multipotency, viability, and secretory function due to their simplistic and non-physiological nature [39–42]. Thus, there remains a need to identify these barriers and develop biomanufacturing systems that are more reliable in generating high output and high-quality MSC populations for translation of novel therapies.

To address these limitations, 3D culture systems have been explored, including spheroids, Matrigel, microparticles in bioreactors, and biomaterial-based scaffolds [43–47]. While these systems offer improvements, they still face significant challenges in scalability, long-term culture, and ability to maintain inherent MSC phenotypic properties. However, hydrogel-based scaffolds, in particular, have garnered attention for their tunable properties, but their focus has largely been on promoting specific MSC functions (forced differentiation, growth factor secretion, etc.) rather than sustaining a broad, native “stem-like” phenotype for an extended period in culture [47–51]. Predictable long-term retainment of “stem-like” cells is crucial for scaling and translating MSC-based for therapies.

Our group has developed Bio-Blocks®, a hydrogel-based culture system designed to replicate in vivo-like environments while addressing the challenges of traditional culture systems [52–54]. The tissue-mimetic Bio-Blocks have a unique micro-/macro-architecture that circumvents diffusional constraints and eliminates the need for subculturing via addition/subtraction of new/old Bio-Blocks as needed (i.e., puzzle piece design). These features in summation help reduce cellular stress, diminish exogenous intervention, and maintain MSC viability and phenotype over longer culture periods. Unlike other systems, the unique micro-architectural design of Bio-Blocks helps facilitate efficient mass transport, promote degrees of freedom for cellular migration and proliferation (i.e., eliminates confinement constraints), and enable easy collection of secreted byproducts, functioning as a continuous “bioreactor” for generating biologics. The authors predict that mass transport is critical to nutrient exchange and the overall longevity of cells needed for in vitro scaling, with loss of adequate nutrient exchange resulting in hypoxia-stressed and non-viable cells populations [55]. The important role of matrix viscoelastic and mechanical properties in conjunction with cellular confinement within a matrix on cells is further elaborated on by Chaudhuri et al. [56].

This study builds on our previous work by comparing Bio-Blocks to 2D culture and commonly used 3D systems, including spheroids and Matrigel, over a four-week period. Using adipose-derived MSCs (ASCs), Bio-Blocks were fabricated to mimic the mechanical properties of adipose tissue. We evaluated MSC viability, phenotype, and secretome composition and quality across systems, focusing on EV potency. To assess functional outcomes of the MSC secretome, such as the angiogenic potential, we dosed endothelial cells (ECs) with MSC-derived EVs and assessed for phenotypic and functional changes in ECs. To our knowledge, this is the first study to comprehensively compare MSC cultures across these different 3D platforms for an extended period of time, providing critical insights into generating scalable and high-quality MSC populations for regenerative therapies.

2. Materials and methods

2.1. Cell and media sources

Human adipose-derived mesenchymal stem cells (ASCs; Cat. #PT-5006, Lot #22TL018258, 42 year-old Female, Caucasian, Lonza, Walkersville, MD, USA), human umbilical vein endothelial cells (ECs; Cat. #FC-0003, age and sex not disclosed, Lifeline Cell Technology, Frederick, MD, USA), were utilized in this study. ASCs were cultured in RoosterNourish MSC-XF (Cat. #K82016, RoosterBio, Frederick, MD, USA) for growth media (MSC-GM) and switched to RoosterCollect EV-Pro (Cat. #K41001, RoosterBio) for serum-free, low particulate media prior to conditioned media collection. Lastly, VascuLife® EnGS Endothelial Medium Complete Kit (#LL-0002, Lifeline Cell Technology) was used for EC culture.

2.2. Initial expansion of ASCs and ECs

The initially cryopreserved cells from suppliers of the ASCs and ECS were termed “Passage 0 (P0).” The P0” cells were expanded one passage by seeding on 2D plastic and cultured until ~80% confluency before subculturing to obtain enough cells for experimental studies. Subculturing of cells was performed by removing culture media, washing 3x with HBSS (Hanks Balanced Salt Solution), and incubating with 0.05% Trypsin/EDTA (Cat. #CC-3232, Lonza) at 37°C for 5 minutes. Trypsin was neutralized with serum-containing media (5% FBS in DMEM) and cells were centrifuged at 500 g for 5 minutes. The resulting cell pellet was resuspended with cell culture media. “Passage 1 ECs (P1 ECs)” were used for subsequent experimental studies. “Passage 1 (P1)” ASCs were assessed for phenotype MSC characteristics as our baseline/control population. Remaining P1 ASCs not used in phenotype characterization were then either reseeded into a new 2D vessel at a seeding density of 5,000 cells/cm² (standard T-150 TCP flasks with filter caps were used) for continued expansion or reseeded into 2D and 3D systems for experimental studies as ‘P1” cells. The ECs were seeded at a density of ~25,000 cells/cm² for experimental assays. For experimental seeding of each culture system with ASCs, a large single batch of ‘P1’ ASC cell suspension was generated initially (in order to promote greater homogeneity of cell seeding across all four culture platforms), followed by diluting with culture media to a working/seeding cell concentration for each system. Initial cell number of the single batch was calculated by hemacytometer calculation. A working cell solution was aliquoted and diluted from there for each system, as described, respectively. Final cell concentrations based on arithmetic of initial cell batch and dilutions.

2.3. Three-dimensional (3D) cell culture systems

Spheroid: To generate uniform 3D ASC spheroids, 96-well Ultra-Low Attachment Spheroid Microplates (Cat. #4520, Corning®, Tewksbury, MA, USA) were utilized rather than using the other commonly utilized technique, “hanging-drop,” in order to increase the standardization of the methodology. In brief, the large single batch of ASC cell suspension was diluted to a working cell solution within MSC-GM. The final concentration (determined by arithmetic) of ASC cell suspension for spheroids was ~16,667 cells/mL in MSC-GM. A total of ~5,000 cells were added per well of a 96-well plate (i.e., each well of a 96-well plate received 5,000 cells that was suspended into 300 µL of MSC-GM derived from the larger ASC cell batch). This yielded a final standardized concentration of 16,667 cell/mL of culture media (calculated based on arithmetic).

Matrigel: Matrigel® Growth Factor Reduced (GFR) Basement Membrane Matrix, Phenol Red-free, LDEV-free (Cat. #356231, Corning) was used for 3D gel suspension. First, the Matrigel matrix was thawed at 4°C overnight as recommended by manufacturer. Using cooled pipette tips and culture plates (pre-chilled in a refrigerator followed by keeping on ice during process), 125 µL of Matrigel was first added to each well of a 24-well glass bottom culture plate to provide a thin coat, the plate was placed in incubator at 37°C for 30 min. During this time, ASCs were prepared from the batch cell suspension. The cell suspension of 1.25 x 10⁶ cells/mL was made with cell culture media and was then added to Matrigel matrix to create a Matrigel/Cell suspension. The cell suspension was added to Matrigel at a ratio of 1 part cell suspension to 9 parts Matrigel. Matrigel-coated plates were removed from the incubator and 400 µL of the Matrigel/Cell suspension was added to all 24 wells providing 50,000 cells/well. Plates were then immediately placed in incubator again for 60 minutes. Plates were removed once again and 3 mL of cell culture media was added. Yielding a final concentration of cells of 16,667 cells/mL (calculated based on arithmetic).

Bio-Block: The tissue-mimetic 3D Bio-Block® system is ~1-cm³ and is a 3D-printed cell culture and expansion system (Cat. #BMX101-101003, Ronawk, Overland Park, KS, USA) that contains a unique macro-/micro-architectural design (please see previous studies for more detail) [52,53], containing continuous interconnecting microchannels with diameters than range from 300 to 1000 µm, with the 300 µm being used in this study. Bio-Blocks can be fabricated to mimic the mechanical properties of tissues of interest, for this study adipose tissue was used for reference for the Storage (E’), Loss (E’’), and Secant (E) Moduli which are ~48, 4, and 8 kPa, respectively (please see references for more explanation [52,53]). The microarchitectural design significantly increases the surface area-to-volume ratio to enhance cellular proliferation and migration within a confined space. The Bio-Blocks are fabricated with a proprietary mixture of biodegradable substrates that contain biologically native cellular binding epitopes (please see associated USTPO patent #WO2023133120A1). Four (4) Bio-Blocks were placed into a glass 6-well culture plate. Next, an ASC cell suspension of 3.33 x 10⁵ cells/mL was made and then 125 µL of cell suspension was added dropwise (Supplemental Figure S1) to each of the four (4) Bio-Blocks in each well for a total of 500 µL per well and 1.67 x 10⁵ cells per well. Bio-Blocks were then submerged in 10 mL of media. Yielding a final concentration of cells of 16,667 cells/mL (calculated based on arithmetic).

2.4. Experimental Expansion of ASCs

Given the variability in volume and surface area between the four different culture systems (2D, Spheroid, Matrigel, and Bio-Blocks), cells were seeded at the same standardized concentration of 16,667 cells/mL of media to allow for analogous comparison between 2D and 3D culture in addition to conditioned media analysis. Experimentally expanded ASCs were initially seeded into each respective culture system as ‘P1’ ASCs. For this study, two time points were utilized, one at 1-week and one at 4-week culture time. For the 2D system, ASCs were seeded at a concentration of 250,000 cells per T-75 flask and utilizing 15 mL of media. Final concentration of cells was 16,667 cells/mL of media. Additionally, for the 2D system there were zero (0) subculturing events between initial seeding (time 0) and the 1-week collection timepoint, giving rise to a P2 ASC population at 1-week collection. 2D ASCs were then sub-cultured every ~5 days until the 4-week timepoint, making the 4-week a P6 population. Only one subculturing event was performed for the 3D systems, which was at 2 weeks for each system, cells were extracted and then reseeded at their initial seeding density (as previously stated). The 2-week subculture timepoint was utilized due to limitations of the spheroid and Matrigel systems to effectively extend beyond this timepoint feasibly or in a controllable manner. It is important to note that Bio-Blocks do not require this subculturing step but were still “subcultured” to allow standardized processing between the groups.

2.5. Extraction of ASCs from each cell culture system

2D Culture: Cell culture media was removed, cells were washed 3x with HBSS, and incubated with 0.05% Trypsin/EDTA (Cat. #CC-3232, Lonza) at 37°C for 5-minutes. Trypsin was neutralized with serum-based media and cells were counted prior to centrifugation at 500 g for 5 minutes. The resulting cell pellet was resuspended.

Spheroid Culture: Cell culture media was removed with micropipette tips gently (not to disturb the cell mass), cells were washed 3x with HBSS, and incubated with 50:50 solution of TrypLE™ Express Enzyme, no phenol red (Cat. #12604013, Gibco) and 0.05% Trypsin/EDTA (Cat. #CC-3232, Lonza) at 37°C for 10 minutes, with gentle pipette agitation at 5 minutes. Trypsin was neutralized with serum-based media and cells were counted prior to centrifugation at 500 g for 5 minutes. The resulting cell pellet was resuspended.

Matrigel Culture: Cell culture media was removed with micropipette tips gently (not to disturb the gel structure), the Cell/Matrigel construct was gently washed 3x with HBSS, and then incubated with Cell Recovery Solution (Cat. #354253, Corning,) 4°C for 30 minutes, with gentle pipette agitation every 10 minutes. The Matrigel digested cell suspension was centrifuged at 500 g for 5 minutes and the residual gelatinous Matrigel components were removed. The resulting cell pellet was collected, washed with HBSS, and recentrifuged. Two (2x) additional washes and centrifugation steps were performed with either media or HBSS depending on whether cells were being replated (media) or analyzed (HBSS).

Bio-Block Culture: Cell culture media was removed with micropipette tips, the Bio-Blocks were gently washed 3x with HBSS submersion. The washed Bio-Blocks were then incubated with X-Tract Block Dissociation Reagent™ (Cat. #R1R01-00101, Ronawk) at 37°C for 120 minutes, with gentle pipette agitation every 30 minutes. The Bio-Block digested cell suspension was centrifuged at 500 g for 5 minutes, supernatant removed and cell pellet collected, washed with HBSS and re-centrifuged. Two (2x) additional washes and centrifugation steps were performed with either media or HBSS depending on whether cells were being replated (media) or analyzed (HBSS).

2.6. Assessment of ASC phenotype

Assessment of adipogenic, chondrogenic, and osteogenic trilineage differentiation potential of ASCs was performed via culture with differentiating media, according to the manufacturer’s instructions and as previously described for ASCs at P1 (baseline) and P6 (4-week equivalent in 3D systems). Adipogenic differentiation was performed using hMSC Adipogenic Differentiation BulletKit™ (Cat. #PT3004, Lonza). Chondrogenic differentiation was performed using hMSC Chondrogenic Differentiation Medium BulletKit™ (Cat. #PT-3003, Lonza), and was supplemented with TGF-β3 (Cat. PT-4124, Lonza) at a concentration of 10 ng/mL. Osteogenic differentiation was performed using hMSC Osteogenic Differentiation Medium BulletKit™ (Cat. #PT-3002, Lonza). After conclusion of adipogenic differentiation, cells were fixed in 4% paraformaldehyde and stained with Oil Red O (Cat. #0843, ScienCell, Carlsbad, CA, USA), per the manufacturer’s protocol. After conclusion of chondrogenic differentiation, cells were fixed in 4% paraformaldehyde and stained with Alcian Blue (ScienCell; Cat. #8378), per the manufacturer’s protocol. After conclusion of osteogenic differentiation, cells were fixed in 4% paraformaldehyde and stained with Alizarin Red S (Cat. #0223, ScienCell), per the manufacturer’s protocol. A total of four (4) replicates were performed for each differentiation challenge, with 10 total images taken per replicate for a total of up to 40 images per cell culture system. All primary antibodies were obtained from Abcam (Cambridge, UK) unless otherwise stated. Evaluation of ASC “stem-like” phenotype was performed with the initial population at P1 via positive immunofluorescent labeling for CD73/90/105/271 and negative staining for CD34/45, as previously described. In brief, ASCs at P1 were seeded in 2D, fixed with 4% paraformaldehyde, washed 3x with HBSS, blocked with 1% donkey-serum, and immunolabeled for CD34 (ab81289), CD45 (ab40763), CD90 (ab181469), CD105 (ab231774), CD271 (ab52987), and CD73 (Cat. #41–0200, Invitrogen, Waltham, WA, USA) and allowed to incubate overnight at 4°C. The next day cells were washed 3x with blocking buffer followed by application of the appropriate secondary antibodies for 1-hour (Cat. #A-21206 and Cat. #A-21202, Invitrogen), then 3x washes with HBSS again. Cells were then counterstained with a nuclear stain, Hoechst 33342 (Cat. #H3570, Invitrogen). Additionally, ASCs from P1 (baseline) and P6 (4-week equivalent in 3D systems) were assessed via an RT² Profiler™ PCR Array for Human Mesenchymal Stem Cells (Cat. #330231; PAHS-082ZC-24, Qiagen, Germantown, MD, USA) to evaluate expression of 84 MSC and MSC-associated genes. RNA was isolated and purified via an RNeasy Mini Kit (Cat. # 74,104, Qiagen). Only RNA samples with a 260/280 ratio of > 1.8 were used for this study. Cycle threshold (Ct) values were recorded and analyzed via the Delta-Delta-Ct method (Supplemental Table S1). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Beta-actin (ACTB), and Beta-2-Microglobulin (B2M) were the endogenous control genes utilized by the array (n = 3).

2.7. Viability characterization of ASCs

Cellular health and viability is a product of a multi-faceted process and can encompass a variety of predetermined cellular endpoints, including decreased cellular proliferation, induction of senescence, or apoptosis. Thus, to aptly describe the cellular state, ASC viability was evaluated via multiple pathways, including protein-based with microplate spectroscopy (SpectraMax iD3, Molecular Devices, San Jose, CA, USA) and RNA-based with qRT-PCR analysis (Analytik Jena, qTOWER3 G Touch, Tewksbury, MA, USA). Microplate assays were carried out per manufacturer’s instructions and as previously described. For proliferation, extracted ASCs were replated at a standardized concentration of 5,000 cells/well (~25% confluency) and incubated for 24 hours in culture media (to prevent artificial impact of changes in cellular confluency on proliferation rate), followed by analysis with PicoGreen (Cat. # P7589, Invitrogen) fluorescence obtained at 435/535 nm (n = 4) to quantify DNA (cell number) at 24 hours, as a surrogate measurement of proliferation via changes in cell number. For senescence, replated ASCs were stained with CellEvent™ Senescence Green Detection Kit (Cat. #C10850, Invitrogen), per manufacturer’s instructions. Senescence levels were quantified via fluorescence obtained at 435/535 nm (n = 4), which was paired with a Hoechst 33342 counterstain as a nuclear marker for cell number, in order to standardize average senescence intensity per cell. The last microplate assay for ASCs was an apoptotic caspase assay for Caspase-3, −8, and −9 (Abcam; Cat. # ab219915). Assay carried out per manufacturer’s instructions. In brief, replated ASCs were replated at a standardized concentration of 5,000 cells/well and incubated overnight (~12 hours) to allow for adherence to plates. Reagents were added and a caspase inhibitor was used for a control group. Fluorescence was quantification of relative senescence for Caspase-3 (Ex/Em: 535/620 nm), Caspase-8 (Ex/Em: 490/525 nm), and Caspase-9 (Ex/Em: 370/450 nm).

Microplate assays were further validated with gene expression analysis via qRT-PCR with Qiagen RT² qPCR Primer assays (Cat. #330001) for CCND1 (GeneGlobe ID: PPH00128F), CDK2 (GeneGlobe ID: PPH00117F), CDKN2A (GeneGlobe ID: PPH00207C), TP53 (GeneGlobe ID: PPH00213F), and BAX (GeneGlobe ID: PPH00078B). As previously described, RNA was isolated and purified via an RNeasy Mini Kit. Only RNA samples with a 260/280 ratio of > 1.8 were used for this study. Cycle threshold (Ct) values were recorded and analyzed via the Delta-Delta-Ct method with a housekeeping Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene (GeneGlobe ID: PPH00150F).

2.8. Isolation of ASC conditioned media

When conditioned medium (CM) from ASC culture was desired (ASC-CM), MSC-GM growth media was removed, cells were washed 3x with HBSS and serum-/xeno-free RoosterCollect EV-Pro media was added for an additional 24-hr wash. The 24-hr wash was removed and RoosterCollect EV-Pro media was added followed by collection after 24-hr intervals for two consecutive days (48 hours total). Combination of both collection samples were from the same wells. Otherwise, ASC-CM was kept separate from each well replicate for Bio-Blocks and 2D groups, but consolidation of spheroid and Matrigel down to 6 replicates (e.g., for Matrigel combine 4 wells together per replicate for 6 replicates in one 24well plate), in order to achieve adequate volumes for all experiments. Collected ASC-CM was centrifuged at 1,500 g for 10 min to eliminate cell debris, Steriflip filtered with a 0.22-μm filter, and stored at −80°C for long-term storage until use.

2.9. ASC-CM total protein and EV quantification

ASC-CM samples were quantified via total protein analysis with QuickDrop absorbance at 280 nm (SpectraMax QuickDrop Micro-Volume Spectrophotometer, Molecular Devices, San Jose, CA, USA), as previously described. EVs were then isolated per manufacturer’s instruction. In brief, an ASC-CM aliquot was centrifuged at 4,000 g for 30 minutes through a Vivaspin 20 MWCO 100,000 kDa centrifuge cutoff filter (Cat. #28932363, Cytiva, Marlborough, MA, USA) followed by washing with PBS and re-centrifugation at 4,000 g for 5 minutes through the 100-kDa filter, for a total of 3 washes. The EVs were precipitated from the remaining > 100-kDa concentrate overnight using a ExoQuick-TC kit (Cat. #EXOTC10A-1, SBI, Palo Alto, CA, USA), per the manufacturers’ protocol. Unused (not exposed to cells) RoosterCollect EV-Pro media underwent the same processing and was used to establish/determine baseline EV/particulate levels, which were negligible. The EV samples (n = 4) were resuspended in PBS and aliquots were removed and used to quantify relative protein content as a surrogate measure of EV content with QuickDrop spectrophotometer absorbance at 280 nm in order to allow for a more direct comparison to total ASC-CM protein content. Remaining purified EV samples were utilized for ASC-EV treatment of ECs.

2.10. EC phenotypic characterization after ASC-EV treatment

After isolation of EVs from ASC-CM from each culture system group, EVs were diluted with PBS to a standardized concentration. An aliquot from each sample was taken and utilized as a “supplement” for the VascuLife® EnGS Endothelial Medium (Cat. No. LL-0087, LifeLine Cell Technology) for the ECs at a volume of 50-µL of EVs per mL of EC media (1:20). The final concentration of EVs was ~250 µg EV protein/mL EC media for each ASC-EV culture group in order to prevent discrepancies based on concentration. For all assays, a non-enriched baseline EC media and a PBS-sham enriched media were utilized as controls. In brief, ECs were plated in 2D in 24-well plates and allowed to acclimate and achieved desired confluences for one of two applications, either “wound scratch” or RNA analysis via qRT-PCR (wound study ~100%, RNA study ~80%). For the “wound scratch” assay ECs, the EC media was then removed and “wound scratch” with a pipette tip was utilized to recapitulate a wounding response within a confluent monolayer. Subsequently, the ECs were washed with HBSS 3x and ASC-EV enriched EC media was applied for 24 hours. Migration images were taken using an ImageXpress Micro XLS Imaging System (Molecular Devices) and the difference in wound area was calculated from time 0-hr to 24-hr and denoted as percent % wound recovery based on area of original “wound.” Image analysis performed with ImageJ (National Institutes of Health, Bethesda, MD, USA). Similarly, for ECs dedicated for qRT-PCR were replated and acclimated for 48 hours (until approximately 60–75% confluency). EC media was then removed, the cells were washed 3x with HBSS, and ASC-EV enriched EC media was applied for 24 hours. After 24 hours, media were removed and cells were washed and collected for RNA analysis, as previously described. Markers for proliferation, viability and phenotype were utilized and included CCND1 (GeneGlobe ID: PPH00128F), CDK6 (GeneGlobe ID: PPH00119C), BAX (GeneGlobe ID: PPH00078B), CDKN2A (GeneGlobe ID: PPH00207C), CDH5 (GeneGlobe ID: PPH00668F), and HIF1A (GeneGlobe ID: PPH01361B). GAPDH was used as an endogenous control for these samples as well (n = 3).

2.11. Imaging analysis

Imaging was acquired with Molecular Devices (ImageXpress Micro XLS) or Revolve (Discover Echo, San Diego, CA, USA) microscopes. Processing and analysis of images was performed with ImageJ (National Institutes of Health). All sets of images and image analyses were standardized and treated the same way across all similar image sets.

2.12. Statistical analysis

All data were reported as means with standard error of mean (s.e.m.). Characterization analyses of ASC populations for phenotype, proliferation, viability, and senescence were evaluated with One-way ANOVA, except for gene expression data with MSC phenotyping PCR array which a Two-Way ANOVA was utilized. ASC secretome and conditioned media data (including EVs) were analyzed with a Two-way ANOVA. All ASC-EV treated EC data were also evaluated with a One-way ANOVA (n = 3). A minimum of four replicates (n = 4) was used unless otherwise specifically stated. Data were tested for normality via Shapiro-Wilk and Kolmogorov-Smirnov tests and plotted with a QQ plot. GraphPad Prism 9.4.2 software (La Jolla, CA, USA) was used for the analyses and a p < 0.05 was considered significant. ImageJ were utilized for image processing.

3. Results

3.1. Evaluating ASC viability and longevity in different systems

Adipose-derived MSCs (ASCs) were initially screened for MSC characteristics based on the International Society for Cellular Therapy criteria of adherence, cell surface markers (positive CD73, 90, 105, and 271; negative for CD34 and 45), and trilineage potential (osteogenic, chondrogenic, and adipogenic) (Figure 1A) [57,58]. Subsequently, the ASCs were plated into one of four different culture systems − 2D flask, 3D spheroid, 3D Matrigel, or 3D Bio-Blocks (Figure 1B; Supplemental Figure S2). After 7 days within each culture system, the ASCs were imaged to evaluate their morphological structure and spatial distribution. ASCs within the 2D flask became flat 2D sheets of bipolar spindle-like cells (Figure 1C, far-left). Spheroid cell morphology was difficult to assess with light microscopy but cells arranged into well-formed sphere-like masses (Figure 1C, middle-left). Within the Matrigel system, ASCs formed multipolar stellate-like cells forming clusters of tubular like satellite colonies (Figure 1C, middle-right). Lastly, ASCs cultured in Bio-Blocks maintained a spindle-like morphology with cells evenly dispersed throughout the Bio-Blocks with migration down into the pre-formed microchannel networks (Figure 1C, far-right).

Figure 1.

Schematic diagram of adipose-derived mesenchymal stem cells (ASCs) within respective culture systems: ASCs at passage 1 (P1) were plated in 2D and stain for positive cell surface markers (A, green), including CD73 (left), CD90 (left-middle), CD105 (right-middle), and CD271 (right) and was paired with a nuclear counterstain using Hoechst (blue). A representative schematic of each culture system and the cellular arrangement within that system are denoted (B). Photo micrographs of ASCs after 7 days within each culture system (C) with 2D (left), spheroid (left-middle), Matrigel (right-middle), and Bio-Blocks (right) depicting cellular organization and morphology; scale bar = 100 µm top row (A) and scale bar = 200 µm bottom row (C).

The viability and overall health of the ASCs were then assessed after 4 weeks to demonstrate long-term longevity of each system. This was achieved via ASC extraction from each respective culture system and processing for RNA analysis or re-plating into a standardized assay for further evaluation. ASCs within the 2D and Bio-Block systems retained the greatest proliferative capabilities after 4 weeks; Bio-Blocks exhibited 1.98-fold and 2.04-fold relative to spheroid and Matrigel (Figure 2A). When looking at expression of proliferative gene markers, 2D-ASCs had the highest expression of CCND1 while Bio-Block ASCs had a significantly lower expression of CDK2 relative to the other culture groups (Figure 2B).

Figure 2.

Assessment of ASC viability via quantification of proliferation, senescence, and apoptosis: ASCs at passage 1 (P1) were then seeded into each respective cell culture system for 4 weeks. After 4 weeks (P6 for 2D), ASCs were extracted and either re-plated in 2D for microplate analysis (A/C/E) or further processed for qRT-PCR analysis (B/D/F). Microplate analysis (n = 4) included quantification of cell number via PicoGreen (A), relative senescence (C), and apoptosis via caspase-3 (E, left), caspase-8 (E, middle), and caspase-9 (E, right) multiplex array. ASCs were processed for qRT-PCR gene expression analysis for markers of proliferation, senescence, and apoptosis. Proliferation was analyzed via CCND1 (B, left) and CDK2 (B, right). Senescence was analyzed via CDKN2A (D, left) and TP53 (D, right). Apoptosis was analyzed via BAX (F). GAPDH was used as an internal control. Values are represented as relative Fold change to baseline control ASCs at P1. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (n = 4).

ASC cellular activity was further investigated via assessment of senescence after 4 weeks of culture. The relative burden of ASC senescence within spheroid and Matrigel culture was significantly higher compared to that of 2D and Bio-Block culture. ASCs from 2D exhibited senescence at 55% and 66% relative to spheroid and Matrigel, whereas Bio-Block ASCs were at 31% and 37%, respectively (Figure 2C). This data set was further validated with gene expression of CDKN2A and TP53, with Bio-Block ASCs expressing the lowest levels of these senescent markers with no significant change in CDKN2A expression after 4 weeks from the baseline expression. Whereas all three other groups had a significant increase in expression of this senescent marker, with spheroid ASCs having the highest expression at ~5-fold increase in expression relative to baseline levels. Similarly, Bio-Block ASCs had a significantly lower expression of TP53 relative to all other groups, with Spheroid ASCs once again having the highest (Figure 2D).

Next, apoptotic activity of ASCs was investigated via caspase 3, 8, and 9 activity. ASCs from Bio-Blocks had the lowest caspase activity for all three caspases relative to the other culture groups, with ASCs in spheroid and Matrigel culture having 2–3x increase in caspase activity relative to Bio-Block ASCs. Notably, 2D ASCs exhibited a similar level of caspase activity as Bio-Blocks, with significantly less caspase activity relative to spheroid and Matrigel (Figure 2E). Verification of apoptotic activity in ASCs was then assessed via expression of the apoptotic marker BAX. BAX expression in ASCs from Bio-Blocks was significantly decreased relative to the time-zero baseline group and relative to all other culture systems, where 2D ASCs had the most significant increase in BAX expression across all groups, including baseline expression levels (Figure 2F).

3.2. Effect of different culture systems on ASC “stem-like” properties

After assessing the overall health and viability of the ASCs after 4 weeks in each culture system, ASCs were evaluated for multipotent capacity via expression of key trilineage markers RUNX2 (osteogenic), PPARG (adipogenic), and SOX9 (chondrogenic), in addition to trilineage media differentiation. ASCs from Bio-Block culture exhibited a significant increase and/or retainment in expression of all three lineage markers relative to the other culture systems, whereas spheroid culture exhibited the lowest expression for all transcripts (Figure 3A). After extraction, ASCs were replated for differentiation media challenge down each of the three lineages and subsequently stained with Alizarin Red (osteogenic), Oil Red (adipogenic), and Alcian Blue (chondrogenic). ASCs from Bio-Blocks demonstrated the highest propensity for differentiation down each of the three lineages relative to ASCs from the other culture systems. ASCs from Matrigel exhibited the lowest capacity to differentiate down all three lineages (Figure 3B).

Figure 3.

Characterization of ASC trilineage differentiation potential: ASCs at passage 1 (P1) were seeded into each respective cell culture system for 4 weeks. After 4 weeks (P6 for 2D), ASCs were extracted and either processed for qRT-PCR gene analysis (A) or re-plated into 2D for trilineage media differentiation challenge (B). ASCs gene expression analysis for markers of RUNX2 (osteogenic), PPARG (adipogenic), or SOX9 (chondrogenic) was performed (n = 4). GAPDH was used as an internal control. Values are represented as relative Fold change to baseline control ASCs at P1 (dashed line). Re-plated ASCs were further assessed for trilineage potential via culture with osteogenic, adipogenic, and chondrogenic differentiating media, according to the manufacturer’s instructions. After conclusion of differentiation media culture periods, cells were fixed in 4% paraformaldehyde and stained with alizarin red S (B, top row), Oil red O (B, middle row), or Alcian Blue (B, bottom row), per manufacturer’s instructions. A total of four (4) replicates were performed for each differentiation challenge, with 10 total images taken per replicate for a total of up to 40 images per cell culture system. Representative images are utilized. Objective = 10x, scale bar = 50 µm. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

ASC “stem-like” phenotype was then further evaluated with an MSC profiler array after 4 weeks in culture. Notably, ASCs extracted from the Bio-Block system demonstrated a higher propensity for maintaining the baseline expression of key MSC “stem-like” markers including TBX5, MMP2, ICAM1, IGF1, and POU5F1 (OCT4); whereas ASCs from all 3D systems increased expression of other “stem-like” markers, CSF2 (GM-CSF), VCAM1, and BMP4. Notably, only ASCs from Bio-Block culture significantly increased the expression of the key “stemness” marker LIF, where the other systems exhibited a loss of expression (Figure 4A).

Figure 4.

Characterizing gene expression of the ASC phenotype: ASCs after 4 weeks (P6 for 2D) were extracted from their respective culture systems and processed for analysis with RT² profiler PCR array for human mesenchymal stem cells (84 genes). ASCs were evaluated for changes in gene expression of key markers (27 selected) and depicted with a heatmap as fold change relative to baseline P1 ASCs (A). Downregulation of gene expression denoted with “red” color and upregulation denoted with “blue” color. A Fold change of ≥ 10 is denoted as the highest increase in Fold change (dark blue). Specific gene were then further displayed in graphical form for quantitative comparison in one of three (3) key MSC related pathways, including “stem-like phenotype” (B), “Secretome proteins” (C), and “Angiogenesis” (displayed in Supplemental Figure S3). GAPDH, ACTB, and B2M were the endogenous control genes utilized by the array. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (n = 3).

Interestingly, all four culture systems demonstrated a significant increase in expression of HGF and CSF3 (G-CSF) with equal fold changes in the 2D (78-fold), Matrigel (77-fold), and Bio-Block (74-fold) in HGF expression relative to baseline, but significantly less fold increase in spheroid (50-fold). ASCs from Bio-Blocks had over ~7-fold increase in expression of CSF3 relative to all other culture groups with a 586-fold increase overall, whereas the next highest was 2D ASCs at 85-fold increase and the lowest was Matrigel at 56-fold (Figure 4A; Supplemental Table S1).

MSC profiler array data were subdivided into specific pathways to characterize expressional patterns, resulting in a “Stem-like Phenotype” group (Figure 4B), a “Secretome Proteins” group (Figure 4C), and an “Angiogenesis” group (Supplemental Fig. S3). Within the “Stem-like Phenotype” group, Bio-Block ASCs expressed significantly higher levels of LIF, IGF1, TBX5, ICAM1, and MMP2, relative to all other culture groups; whereas both Matrigel and Bio-Blocks resulted in an increased expression of POU5F1 (OCT4) relative to 2D and spheroid ASCs (Figure 4B; Supplemental Table S1). The intrinsic regenerative characteristics of MSC paracrine activity was also evaluated via expression of pro-regenerative growth factors and cytokines. ASCs from Bio-Blocks exhibited the highest expression for a large proportion of genes, including IL6, IL10, TNF, CSF2 (GM-CSF) and CSF3 (G-CSF); whereas 2D and spheroid cultured ASCs had the lowest expression of critical biomodulatory factors, such as the cytokines IL6, IL10, and CSF2 (GM-CSF). Matrigel cultured ASCs had more variability with low expression of factors like IL1B, IL6, TGFB1, and BDNF; whereas they had significantly elevated levels of BMP4 (Figure 4C; Supplemental Table S1).

3.3. Culture systems affect ASC secretome

Given the findings of altered expression of paracrine factors from ASCs (Figure 4), specific factors were quantified within the ASC secretome including antioxidants/oxidants (Figure 5), soluble proteins (Figure 6A), and EVs (Figure 6B-C). Previous studies have highlighted the importance of proper antioxidant/oxidant balance for proper cell function and regenerative activity [59,60]. Three total cellular and extracellular fractions, which included the cell lysate, secreted fraction, and the EV lysate, were obtained to determine the location and balance of antioxidant/oxidant activity. Bio-Block ASCs significantly increased production of antioxidant compounds in all three compartments, with over a 2-fold increase in production within the ASC cytoplasm and EV compartments relative to the baseline levels (Figure 5, left). Interestingly, there was an increase in ROS/RNS oxidative activity within the ASCs cultured in Matrigel and Bio-Blocks relative to 2D and spheroid culture systems. However, only Bio-Block cultured ASCs correlated with retainment of ROS/RNS oxidative species within the EV fraction when compared to the baseline population (Figure 5, right), whereas the other groups all saw a significant decline (relative fold change < 1.0). Notably, ASCs from Bio-Blocks did not exhibit an increase in expression of HIF1A in response to elevated oxidative activity within the cytoplasm (Supplementary Fig. S4).

Figure 5.

Evaluating effect of culture system on ASC antioxidant/oxidant balance: ASC conditioned media (ASC-CM) was collected over the 4-week culture period at pre-determined timepoints, including ASC-CM from baseline P1 ASCs and at 4 weeks (P6 for 2D) for each culture system. ASC-CM at 4 weeks was evaluated for total antioxidant (left column) and oxidant (right column) content via microplate assays assessing for total activity (see methods section). The antioxidant/oxidant activity was broken down into three compartments, the secreted soluble fraction (top row), the ASC lysate (middle row), and the EV lysate (bottom row). Schematic of cell displayed to represent the three different compartments of antioxidant/oxidant activity. The total antioxidant/oxidant activity is based on relative activity to baseline P1 ASCs and displayed as relative Fold change. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (n = 4).

Figure 6.

Quantifying total protein and extracellular vesicles secreted from ASCs: ASC conditioned media (ASC-CM) was collected over the 4-week culture period at pre-determined timepoints, including ASC-CM from baseline P1 ASCs, then after 1 week (P3 for 2D) and 4 weeks (P6 for 2D) in each culture system. The total protein content (µg/mL) was calculated for the ASC-CM collected from P1 ASCs (A, dashed line), 1-week (A, silhouette bars/leftmost bars), and 4-week (A, hatched bars/rightmost bars) via QuickDrop protein quantification with absorbance at 280 nm. Subsequently, the EV fraction was isolated and purified from the total ASC-CM fraction (as described in the methods section) and the protein content was quantified to provide a concentration of EV based on protein (B). Additionally, the relative quantity of EV in each sample relative to the total amount of secreted protein within each ASC-CM sample was then calculated to determine relative compositional changes for each group (C). ASC-EVs were isolated from ASCs at baseline (P1), 1-week, and 4-week and the relative concentration of EVs per sample was displayed as protein content (µg/mL) via QuickDrop absorbance at 280 nm. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (n = 4).

The total amount of soluble protein secreted by the ASCs into the conditioned media (ASC-CM) was evaluated via spectroscopy absorbance at 280 nm. The baseline level of protein secreted by ASCs was 911 µg/mL. All three 3D culture systems enhanced the production of secretive protein activity from ASCs at the 1-week timepoint in culture, whereas 2D culture had no significant change (Figure 6A). However, when evaluating for sustained ability to produce paracrine proteins overtime at 4 weeks, only Bio-Block ASCs retained the propensity for secreting proteinaceous factors, whereas 2D, spheroid, and Matrigel cultured ASCs all exhibited significant declines in their secretory activity relative to their week 1 levels (Figure 6A).

ASC production of secreted EVs was then evaluated at 1-week and 4-week culture time points. Similarly, at 1-week timepoint all three 3D culture systems increased the production of secreted EVs relative to the baseline production of 175 µg/mL. However, ASCs from Bio-Blocks had a significantly higher increase in EV production relative to the other culture systems, with a 71% increase; whereas 2D had a 29%, spheroid at a 42% and Matrigel had a 47% increase relative to the initial baseline ASC-EV levels (Figure 6B). Notably, 2D, spheroid, and Matrigel ASCs exhibited a significant decline in production of EVs over time, with 2D ASCs demonstrating a 68% loss, spheroid a 70% loss, and Matrigel a 30% loss in EV production after 4 weeks in culture. Yet, Bio-Block cultured ASCs demonstrated a significant increase in EV production by 44% (300 µg/mL to 432 µg/mL) (Figure 6B). When accounting for relative EV production per amount of total secreted protein, only Bio-Block ASCs exhibited a sustained increase in EV-to-Protein (secreted) production over time, whereas both decreased protein and EV production was observed by all other culture groups over time (Figure 6C).

3.4. Potency of ASC-EVs from different culture systems on endothelial activity

To assess for overall quality and potency of ASC-EVs from each system after 4 weeks of culture, a standardized dose of ASC-EVs from each system were used to treat endothelial cells (ECs), which were then assessed for functional and phenotypic changes after 24 hours of treatment (Figure 7A). Measurement of ASC-EV functional activity on ECs was assessed via EC migratory activity after “wounding” and dosing with ASC-EVs for 24 hours. ECs treated with spheroid-based ASC-EVs were found to have a significant decline in ability to migrate to close off “scratch wound,” with Bio-Block ASC-EVs exhibiting a significant increase in stimulating migratory activity in ECs, relative to spheroid ASC-EVs (Figure 7B).

Figure 7.

ASC-EVs from tissue-mimetic system provide a more potent regenerative response in ECs: ASC conditioned media (ASC-CM) was collected at the 4-week timepoint (P6 for 2D) for each culture system. Purified EVs from each system were used to dose ECs at a final concentration of 250 µg/mL for 24–48 hours (A). ECs were then evaluated for ability to migrate and recover a “wound area” via a scratch assay over 24 hours (B). Baseline EC media without ASC-EVs (dashed line) was utilized as a control. Additionally, ECs were evaluated for changes in expression of key markers important for EC regenerative functionality, including proliferation (C), senescence (D), and phenotype (E) after 48 hours of ASC-EV treatment. GAPDH was used as an internal control for qRT-PCR. Values are represented as relative Fold change to baseline control ECs without ASC-EV treatment. Significance denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (n = 3).

Next, expression of key markers of EC activity were assessed after treatment of the ASC-EVs, including transcripts for proliferation, viability, and phenotype. Bio-Block ASC-EVs were found to increase the expression of both markers of proliferation, CCND1 and CDK6, relative to 2D ASC-EVs, where 2D ASC-EVs exhibited the most significant decline in expression of these markers (Figure 7C). Next, EC viability and robustness was measured via expression of apoptotic and senescent markers, BAX and CDKN2A. Spheroid-based ASC-EVs resulted in a significant increase in expression of both markers relative to baseline expression and all other culture systems while the other three groups retained a similar expression as the untreated baseline ECs (Figure 7D). Phenotypic changes of ECs were also evaluated, where ECs treated with Bio-Block ASC-EVs exhibited an increased expression of CDH5 (VE-Cadherin), a marker of EC differentiation and angiogenesis. Conversely, spheroid and Matrigel-based ASC-EVs were found to promote expression of HIF1A relative to treatment with 2D and Bio-Block ASC-EVs (Figure 7E).

4. Discussion

Mesenchymal stem cells (MSCs) represent a heterogeneous group of progenitor cells whose multipotent and secretory characteristics have made them a focal point in TERM, offering a wide range of potential clinical applications [6,7]. To utilize MSCs effectively, they must be first extracted from donor tissues and cultured in vitro, outside their native environment. This process is crucial for generating the large cell numbers required for therapeutic use, typically ranging from 10⁷ to 10¹² cells [61,62]. However, the current methods for expanding MSCs are time-consuming and labor-intensive, leading to a loss of regenerative capabilities and “stem-like” properties. Thus, diluting the desired final populations with cells that are no longer “stem-like” and potentially interfering with regenerative tissue processes. Therefore, suboptimal culture conditions continue to represent a significant bottleneck in the development and large-scale translation of MSC-based therapies [37,39].

Research is increasingly focused on understanding how two-dimensional (2D) and three-dimensional (3D) culture conditions ultimately influence MSC phenotype and functionality in order to improve translatability. Studies have shifted focus to the potential advantages of 3D approaches over traditional 2D culture. The main 3D-like systems utilized at GMP-level research for investigating potential new clinical therapies are scaffold-free/spheroid-based, gel-encapsulation-based (commonly Matrigel), hollow-fiber-based and microcarriers with/without large stirred-tank bioreactors (notably, both spheroid and hydrogel encapsulation can both be used with stirred-tank bioreactors as well) [63–66].

Within this study, we utilize two of the major methodologies, spheroids and Matrigel (i.e., gel-encapsulation), for MSC culture [66]. However, the long-term efficacy and scalability of these two systems remains uncertain. Major limitations of these systems include the continuous need for harsh subculturing, lack of tailorable and reproducible products, and restrictive scalability due to limitations of mass transport and diffusion that results in cellular heterogeneity and decreasing viability [66–70]. These systems have provided significant benefits over traditional 2D systems but still have their limitations. Thus, even though these other 3D-like systems allow for potentially large cell and/or biologic production (e.g., large stirred-tank bioreactors), the quality, robustness, and reproducibility of said products represents a significant barrier with the preexisting modalities, as suggested in this study and others. Tissue-engineered hydrogel systems offer a promising avenue for circumventing these issues creating customizable tissue-like environments that can mimic the chemical, physical, and mechanical micro-environments of native tissue to better retain the viability, multipotency, and regenerative features of MSCs and improve translatability of MSC-based research [56,71,72].

Our research group has recently explored the Bio-Block hydrogel system as a means to achieve a more efficacious cell cultivation system given its elimination of subculturing (demonstrated in previous studies), tailorability, efficiency of mass transport and diffusion (relative to other 3D systems), and modular nature [52,53]. This study aimed to assess and compare how the aforementioned systems affect three key principles of the MSC-like populations, 1) overall cell health and viability, 2) inherent multipotency and “stem-like” phenotype, and 3) regenerative paracrine functionality, in particular the “quality/potency” of EVs. Predictability of these three properties are critical for future large-scale production of any potential translational therapy. Of note, “benchmark” values and studies designs are often variable between previous published literature, in part because of the vast arrays of assays to characterize cellular populations. Additionally, cross-comparison studies such as this one are lacking when it comes to newer 3D systems and MSC culture over longer culture time periods. This further highlights the importance for elucidating relative impacts of individual systems from each culture system when controlling for inter-experimental variability that naturally occurs over time, such as that highlighted in this study.

As mentioned, the first aspect investigated was the ability to maintain healthy, viable, and proliferative cellular phenotypes. It is generally accepted that spheroids (and many other 3D systems) can hinder nutrient exchange and cause hypoxia near the center of the 3D cell mass [66]. However, previous studies suggest that spheroids improve viability and proliferation of MSCs, relative to 2D, due to the upregulation of “hypoxia-adaptive molecules” [73]. However, these studies typically studied short-term culture (3–5 days) and neglect to evaluate longer culture periods necessary to produce cellular yields adequate for clinical therapies. Even in the absence of visible necrotic cores, spheroids are inherently subject to gradients in nutrient and oxygen diffusion, as well as waste accumulation, which introduce physiological stress. This study reveals the potential negative impact long-term hypoxic stress and inefficient nutrient exchange can have on MSC populations, where spheroid and Matrigel cultures both exhibited a significant decline in ASC viability and proliferative capacity at the 4-week timepoint. This is likely due to the increased induction of senescent and apoptotic populations in response to the sustained hypoxic stress and inadequate nutrient exchange the cells were exposed to. Moreover, it is well established that senescent cells can influence neighboring cells through the senescence-associated secretory phenotype (SASP), potentially amplifying the decline in multipotent characteristics across the entire spheroid and Matrigel systems. Whereas Bio-Blocks have previously been shown to promote equilibration of oxygen distribution throughout which may offer insight into the improved cell properties of these cells [52].

Interestingly, although Bio-Block ASCs maintained a higher proliferative capacity (based on changes in cell number), when compared to spheroid and Matrigel cells, their expression of cell cycle promoters, CCND1 (cyclin D1) and CDK2, was similar or lower. It is possible that proliferation for Bio-Block ASCs is being driven by an increase in other cell cycle promoters. Alternatively, Bio-Block ASCs may promote a decrease in cell cycle inhibitors, as suggested by the significant decrease in CDKN2A (p16^ink4a) and TP53 (p53) in Bio-Block ASCs. Further studies expanding the suite of gene expression targets will be needed to assess potential dynamic mechanisms at play, such as induction of telomerase, the role of oxidative stress, changes in metabolic demand, and changes in DNA repair mechanisms to further establish the protective role of Bio-Blocks from senescence and apoptosis seen in this study and prior studies.

Next, the impact of each culture system on multipotency was assessed after 4 weeks. ASCs from Bio-Blocks demonstrated a phenotype suggestive of an increased retainment of multipotent capacity relative to all other culture systems. Conversely, spheroid and Matrigel culture were inferior to both the Bio-Blocks and 2D culture systems. The loss of multipotency with Matrigel-cultured ASCs is not surprising when placed in the context of Matrigel’s previously established propensity to promote angiogenic activity and differentiation of MSCs. Interestingly, spheroids were inferior to 2D culture when it came to multipotent capacity. This is converse to previously suggested studies that state that spheroid-cultured MSCs demonstrate improved “stem-like” properties, including multipotency [74–77]. However, these previous comparative studies did not culture out to 4 weeks and only looked at up to 7–14 days maximum in culture, except for one more recent study that showed MSCs becoming senescent within spheroids during a 20-day culture period [78]. The decline in ability of ASCs to differentiate down osteogenic, adipogenic, or chondrogenic lineages after just 4 weeks for the spheroid and Matrigel systems suggests that these cell populations have significantly impaired multipotent potential and are likely lacking “stem-like” properties and suboptimal for use in tissue engineering applications that require in vitro expansion of cells.

To further assess for “stem-like” characteristics and phenotypic profiles, an MSC microarray was used consisting of 84 key genes. The heatmap analysis of 27 key genes demonstrates a general trend of ASCs from the 3D systems that were associated with increased expression of specific MSC-like phenotypic markers to a greater extent than 2D culture. Interestingly, VIM (Vimentin), a marker associated with mesenchymal-like cellular phenotype, exhibited minimal differences between the culture groups and no significant change relative to baseline ASC expression (i.e., no loss of expression). However, Bio-Blocks were associated with elevated expression of multiple critical genes associated with “stem-like” cell populations, whereas spheroid and Matrigel culture resulted in a more pronounced downregulation in a number of key genes, including LIF and IGF1 which were significantly downregulated in all systems except for Bio-Blocks. LIF and IGF1 have previously been shown to be strongly associated with “stem-like” cell populations [79,80]. These data suggest that MSC-like cells potentially retain their inherent “stem-like” expression profiles within Bio-Blocks. Whereas spheroids and Matrigel are potentially driving ASCs down pathways that are no longer MSC-like and/or depriving these cells of adequate nutrient exchange that may be promoting senescence and apoptosis.

Lastly, the intrinsic regenerative functionality of ASCs to secrete biomodulatory factors to promote tissue growth and healing (i.e., a positive regenerative profile) was evaluated. The ASC expressional profiles of secretory proteins were stratified between the culture systems. Growth factors like HGF and BMP4 were elevated in all culture systems, yet other commonly investigated proteins involved in the EGF and FGF family of proteins were heavily downregulated in all systems (supplemental table). HGF commonly plays a role in angiogenesis [81–83] and immunoregulation [83–85] via promoting the production of factors like VEGF and IL10, respectively. Whereas BMP4 is involved in cell growth and differentiation via the BMP/TGF/SMAD regulatory pathway [86,87].

Other factors were significantly elevated in all three 3D systems relative to 2D culture, including CSF2 (GM-CSF), IL10, and IGF1. When looking specifically at immunoregulatory factors (both anti- and pro-inflammatory), Bio-Blocks commonly promoted their highest expression, including CSF2 (GM-CSF), CSF3 (G-CSF), IL1B, IL6, IL10, and TNF, suggesting a likely strong immunoregulatory role for ASCs cultured within Bio-Blocks. However, it is also important to note that several of the immunoregulatory cytokines, such as IL1B, IL6, and IL10, also play a role in promoting angiogenesis in varying ways, including promoting endothelial cell migration, proliferation, and differentiation [88–91].

Therefore, the angiogenic-like signaling profile of ASCs was further investigated. Notably, Bio-Blocks were found to be associated with a greater expression of key angiogenic markers, even when compared to Matrigel. This was an interesting finding given the known angiogenic nature of Matrigel. When looking at angiogenic-related structural proteins, there was not much difference in expression levels (Supplemental Fig. S3). However, when examining secretory/signaling proteins, like VEGFA, IL6, and MMP2, their expression was downregulated in Matrigel below that of the baseline P1 ASCs and similar to that of 2D and spheroid expression levels. Whereas expression of these factors is upregulated significantly for Bio-Block ASCs. The authors believe this could be a result of two possible principles. First, an over-saturation of the signaling pathways over time may lead to suppression/downregulation in the protein expression within Matrigel. Second, cells within the less conducive spheroid and Matrigel systems lose their adaptive and secretory capabilities secondary to decreased cellular activity and protein production (as seen in Figure 6), while Bio-Block ASCs maintain a highly adaptive and secretory MSC-like phenotype capable of orchestrating cellular signaling pathways suggestive of angiogenesis (i.e., Matrigel ASCs are no longer MSC-like and thus not adaptively responding to any potential angiogenic stimulus of Matrigel). Further studies are needed to identify a temporal relationship for the differences seen in angiogenic-like expressional profiles and to investigate how expression of other markers not included in this microarray compare to get a more holistic perspective.

The balance between antioxidant and oxidative activity is another important aspect in regulating MSC-like properties, which was broken down into three compartments for this study (i.e., EV lysate, cell lysate, and secreted soluble fraction). Bio-Blocks were found to promote production of antioxidant compounds relative to the other culture systems to varying degrees in each of the three cellular compartments. Conversely, all culture groups resulted in an increase in oxidative compounds within the ASC lysate fraction after 4 weeks, relative to baseline P1 ASCs levels. However, only Bio-Block ASCs maintained the ability to excrete oxidative compounds within EVs, relative to baseline ASCs. Previous studies have suggested that packaging harmful compounds, such as oxidative species, into EVs can be protective for cells from induction of senescence and/or apoptosis [92,93]. Thus, loss of this ability may play a role in the increased senescence and/or apoptosis seen in spheroid and Matrigel culture, but further studies are needed to investigate this hypothesis.

Given the nonspecific nature of the antioxidant/oxidant assays used in this study, further investigations into what specific compounds are shifting would help determine if specific factors are driving regenerative functions while others are driving anti-regenerative functions. Previous studies have established how oxidative activity can impair tissue regeneration [94]. Conversely, studies have also shown how oxidative compounds help drive angiogenic and other regenerative functions [95]. Thus, the exact stoichiometric balance of antioxidant/oxidant compounds in addition to the specific factors being varied are likely both playing a role in the ultimate functional outcomes seen.

Next, a comparison of the temporal effect of long-term culture on the production of secreted factors, such as soluble protein and EVs, was evaluated. ASCs within Bio-Blocks were associated with retained secretory functionality over time and exhibited an increase in overall EV production. This suggests that cells within 2D, spheroid, and Matrigel culture may lose their secretory functionality as they likely become senescent, apoptotic, and/or lose their “stem-like” features due to culture conditions, which may negatively impact their long-term production of secreted biologics for translational therapies. Whereas the Bio-Block system demonstrated the potential for long-term exploration and production of acellular biologics. It is important to note that assays were designed to standardize to each “culture system” as a whole and thus the conditioned media assays were standardized to media volume per cell at seeding. Given inherent variability in proliferation within each system noted, there could be variation in absolute cells numbers to some extent. However, the decline in secretory activity from 2D, spheroid, and Matrigel would likely be more drastic given the likely increase in cell number (relative to time of seeding) but decreased concentration of protein and EVs at the later timepoints.

Given the preferential shift in EV production and changes in ASC angiogenic-like profile, the authors evaluated the quality/potency of ASC-EVs produced on their ability to modulate EC activity associated with a pro-regenerative phenotype. ASC-EVs from spheroid culture were found to significantly inhibit migratory activity of ECs, whereas Bio-Block EVs maintained a higher migratory stimulus. ASC-EVs from spheroids were also found to promote senescence and apoptosis in the ECs, in addition to inhibiting the induction of CDH5 (VE-Cadherin) expression, a key phenotypic marker of EC activity, relative to Bio-Block EVs. Previous studies have shown that senescent cell populations can impair the regenerative functionality of surrounding cell populations via SASP [96,97]. As depicted in Figure 2, spheroid culture promoted the highest level of ASC senescence, which could explain a mechanism for EC induction of senescence and apoptosis and inhibition of migration. Notably, ASCs within Bio-Blocks have previously been shown to secrete factors protective against senescence [54]. These data highlight the potential transitive impact that cell culture systems have on cultured cell populations and their associated secretome. The use of ECs as a secondary cell population was intended to be a surrogate measurement to identify that there are in fact quantifiable differences in the quality/potency of ASC-EVs from each system, and was not intended to provide an in-depth analysis of EC functionality or angiogenic changes. Future studies will be needed to assess for EC (and other cell type) responsiveness to ASC-EVs and to investigate the exact contents within the EVs that are changing from each culture system.

Thus, this body of work aimed to demonstrate the criticality of appropriate cell culture system design in recapitulating the stem cell niche, including the advantages of incorporating specific tissue-mimetic properties paired with the impact of improving mass transport on the long-term regenerative characteristics of MSC-like populations. Although the mechanotransductive contributions of 3D systems are widely recognized as central regulators of MSC activity, long-term culture with these systems results in either 1) over-confluency of cells (and nutrient limitations) or 2) continuous need for stressful subculturing events that impair cell functionality. These inherent properties of current 3D modalities should be taken into account when considering longer culture timeframes (e.g., timeframes needed for developing translational therapies) given the potential for limitations.

The summative impact of dimensionality, tissue-mimetic features, and effective nutrient exchange was highlighted in this study. The loss of viability, proliferation, and MSC-like characteristics in the 3D spheroid and 3D Matrigel systems over time relative to 2D demonstrates the long-term culture success of MSCs for translational therapies is multifactorial but effective nutrient exchange and mass transport may play a key role. Moreover, these studies were also designed to evaluate the transitive properties of the culture conditions for MSC secretory activity and quality. As previously mentioned, Bio-Blocks offer a number of unique physical features, which includes the macro-architectural puzzle-piece design that allows continuity of culture and eliminates subculturing (a significant drawback of other 3D/hydrogel systems), the micro-architecture design does not significantly hinder mass transport and allows cellular dissemination, and the modular nature of the system as a whole which permits addition/subtraction of Bio-Blocks ad libitum for longer experiments (outside the scope of this study). These features, in combination with the inherent benefits of 3D hydrogels (e.g., substrate tailorability, mechanical and viscoelastic control, etc), provide a more sustainable system for exponential scaling of cells and/or biological cellular byproducts for translational purposes. The body of work of this study provided new insight into secretory dynamics of MSC-like populations over time, in particular the augmentation of EV production. The goal of this research is to demonstrate how systems, like the Bio-Blocks, could help standardize cell culture and potentially provide a continuous “bioreactor” system that can be modified to tailor the culture of “stem-like” populations and their subsequent secretion of regenerative acellular biologics.

Therefore, the future utility of research systems, like that of the Bio-Blocks, can be highlighted across all three major branches of biomedical research, Basic, Translational, and Clinical. Bio-Blocks aid in early-stage basic science research by providing a more physiological environment that recapitulates/models the native tissue and cellular signaling response with higher fidelity than traditional 2D systems and permits longer culture periods relative to other 3D systems due to elimination of the need for subculturing. The Bio-Blocks are tailorable and therefore offer an opportunity, like other 3D systems controlling cells through system manipulation, whether to model native physiology or generate specific cellular bioactivity. From a Translational perspective, this research provides insight into important parameters to consider when designing MSC-based studies requiring large scale production of cells or biologics. Whether utilizing Bio-Blocks or not, it is important to understand that not all biologics are created equal, MSCs are dynamic, so understanding their “functional output” is a good way to screen for efficacy for specific application. However, as discussed in this study, Bio-Blocks do offer advantages over other systems when wanting to scale production of cells and acellular biologics and thus offer a more robust and efficient biomanufacturing platform for future regenerative applications. In part, because Bio-Blocks provide 1) a system that more efficiently scales the cells of interest (i.e., MSCs) while 2) maintaining their desired phenotype (i.e., stem-like properties). When a researcher can trust that they have a desired cell population, that allows for higher predictive value and external validity of data within studies.

The research highlighted in this study, and previously with Bio-Blocks, offers insight into the next steps toward advancing the next generation of biomedical research and clinical therapies, including the importance of culture design. However, there remain important shortcomings and barriers with utilization of the 3D Bio-Block system to be cognizant of. Many assays were built and refined based on 2D culture (e.g., microplate spectroscopy) and provide challenges and alterations in the assay protocol to be modified for 3D systems. Thus, depending on the in situ characterization assays of interest, utilization of Bio-Blocks can lack a robust set of in situ modalities, relative to traditional 2D systems. Another important consideration being the limitations of current imaging modalities and the depth they can penetrate into a 3D system. Therefore, sectioning is a necessary added step that is not needed in 2D systems, but this is common for tissue biopsies as well, so not insurmountable. Ultimately, within the context of biomanufacturing and scaling of robust translational and clinical research products (i.e., cells, biomass, biologics, etc) in situ is not always necessary and the quality of the cells and/or biologic byproducts can be evaluated independently.

There are a few components of this study that require further investigation that are limiting in their current form but do not ultimately detract from the study findings. For instance, the use of only in vitro analyses can be limiting, but the utilization of a wide-range of paired functional, protein-based, and RNA assays provided a holistic perspective that can be used as a framework for future targeted in vivo studies. The studies performed lacked subjective and qualitative datasets (such as imaging analysis for senescence or phenotype) and focused more on objectively quantifiable assays. Future studies will aim to combine both qualitative and quantitative datasets. Additionally, although ASCs from a single donor were used, the comparative study design ensured that any donor-specific effects would influence all culture conditions equally, allowing for meaningful relative comparisons between systems. Similarly, the EV potency was only tested on ECs. Follow up studies with additional secondary cell populations would help provide a better understanding as to the broader impact of ASC-EVs. This study utilized only one modality to quantify EV concentration (Absorbance at 280 nm), although a valid approach and previously established, a secondary validation modality would have provided additional support of the data. Any differences in sensitivity would effect each culture group EV concentration equally. This study also only evaluated the secreted protein and EVs as a whole and did not investigate how the exact contents were dynamically changing. Future studies with mass spectroscopy, transcriptomic, and proteomic quantification will be used to cross-compare and provide a more detailed understanding of the specific factors changing within the ASCs and ASC secretome. Lastly, studies extending out past 4 weeks are of interest to the authors. The 4-week timepoint was utilized due to limitations of the spheroid and Matrigel systems and their limited efficacy extending beyond this timepoint feasibly and to screen whether studying longer timepoints would be valuable for Bio-Blocks. Future studies will aim to extend out to 6 months to investigate how Bio-Block culture impacts MSC populations and their use in biomanufacturing and translational research in in vivo studies.

5. Conclusion

In conclusion, this study highlights the transitive impact of the physical culture system on MSC-like cell behavior, phenotype, and capacity to secrete pro-regenerative compounds with potential for therapeutic use. It is important to note that “MSC-like” populations are dynamic and highly adaptive to their environments; thus, the secretome from one population may vary depending on the system utilized and the timepoints assessed. Therefore, prior to evaluating the translational impact of any MSC-derived therapy, the influence of the culture system design should be considered before attempting to generalize any specific MSC-derived product for future therapies. A direct comparison of the summative impact of specific 3D culture systems demonstrated that spheroid and Matrigel system resulted in a precipitous decline in MSC-like phenotype and viability over time, where an initial positive impact was seen on secretory activity but rapidly resulted in a significant loss in secretory activity by 4 weeks, with ASC-EV quantity and quality having an inverse relationship to culture time. This is potentially a significant limitation to consider when utilizing these systems for large scale therapy production. Whereas ASCs from the Bio-Block system were associated with a phenotype that correlated more closely with their initial MSC-like phenotype, exhibited enhanced viability, and higher secretory activity. The improved robustness of the ASC population in Bio-Blocks correlated with both higher quantity and quality EVs over time, capable of positively impacting a secondary cell population (e.g., ECs in this study). The authors believe these findings are in part due to the summative effects of the dimensionality and tissue-mimetic features to provide the stem-like niche but also the importance of effective nutrient exchange for long-term culture success as cell numbers increase. This study provides critical insight into considerations for future experiments investigating the utility of MSC populations for potential therapies and the importance and transitive properties of culture systems on both the quantity and the quality/potency of the secretome byproducts generated.

Funding Statement

This research was partially funded by both Ronawk Inc. and an NIH MIRA award from NIGMS [R35GM143081] grant through Dr. Jenny Robinson. Members of Ronawk lab team (Aidyn Medina-Lopez, Heather Decker, Christopher Neal, and Adam J. Mellott) were involved with the manuscript revisions and the decision to publish. The NIGMS had no role in study.

Article highlights

• Conclusion: Bio-Blocks outperformed traditional 2D cultures in addition to 3D spheroids (commonly used with stirred tank bioreactors) and 3D Matrigel (representing gel encapsulated systems) in maintaining MSC viability, multipotency, phenotype, and secretome quality over time, representing a more robust and translatable biomanufacturing platform for scaling next-generation regenerative therapies.

• Mesenchymal stem/stromal cells (MSCs) are central to TERM applications because of their multipotency and secretory activity, but current culture systems (2D and traditional 3D) compromise long-term viability, phenotype, and secretome production over time resulting in decreased regenerative potential.

• Existing 3D systems like spheroids and Matrigel provide short-term improvements over 2D because of their dimensionality and tissue-mimetic properties, but suffer long-term culture from diffusion limitations, hypoxia, senescence, and scalability issues that impair reproducibility and translational use.

• Bio-Block hydrogel scaffolds were designed to address these barriers, incorporating modular “puzzle-piece” architecture, enhanced nutrient/mass transport, and tissue-mimetic mechanical properties while eliminating the need for subculturing.

• Cell viability and proliferation: Bio-Blocks sustained ASC viability and proliferative potential at 4 weeks, whereas spheroid and Matrigel cultures declined significantly due to hypoxic stress and senescence.

• Multipotency: Bio-Blocks preserved osteogenic, chondrogenic, and adipogenic differentiation potential; spheroid and Matrigel ASCs lost lineage potential over time, even performing worse than 2D controls in some assays.

• Phenotypic retention: Bio-Blocks maintained expression of key “stem-like” genes (e.g., LIF, IGF1), while spheroid and Matrigel systems showed pronounced downregulation and phenotypic drift.

• Secretory profile: Spheroid and Matrigel 3D cultures were similar in secretome productivity within one week of culture but by 4 weeks exhibited significant diminishment of secretome production. Bio-Blocks promoted sustained and enhanced production of extracellular vesicles (EVs) and biomodulatory proteins with angiogenic-like and immunoregulatory properties throughout 4 weeks.

• EV quality and potency: EVs from Bio-Block ASCs enhanced endothelial cell migration and viability, whereas EVs from highly senescent spheroid ASCs promoted endothelial senescence and apoptosis.

• Oxidative balance: Bio-Blocks increased production of anti-oxidant compounds and maintained the ability to package oxidative species into EVs (a protective mechanism), unlike spheroid and Matrigel systems which saw a decline in packaging oxidative species into EVs.

• Key mechanistic insight: Long-term culture success is strongly dependent on culture system design, such as tissue mimetic properties and recapitulating native niche, however, long-term culture must account for effective nutrient exchange and permit cellular freedom to migrate and proliferate, features that were enhanced in the Bio-Block platform.

• Translational potential: Bio-Blocks function as a modular, scalable “bioreactor” system capable of sustaining stem-like MSC populations and continuously generating high-quantity and high-quality biologics (e.g., EVs) for regenerative applications because of retention of more robust cell populations.

Author contributions

The original study conceptualization and design was performed by Jacob Hodge, Jennifer Robinson, and Adam J. Mellott. The experimental layout, methodology, and investigational procedures were performed primarily by Jacob Hodge with the assistance of Jennifer Robinson and Adam J. Mellott. Data was analyzed and reviewed by Jacob Hodge, Heather Decker, Aidyn Medina-Lopez, Jennifer Robinson, and Adam J. Mellott. Statistical analyses was performed by Jacob Hodge. Jennifer Robinson and Adam J. Mellott served supervisory roles throughout the project and were responsible for funding acquisition. The original manuscript draft was drafted by Jacob Hodge with the assistance of Heather Decker, Aidyn Medina-Lopez, Christopher Neal, Jennifer Robinson, and Adam J. Mellott. Jacob Hodge and Christopher Neal were responsible for figure generation for the manuscript.

All authors reviewed and edited the manuscript drafts multiple times and provided final approval for the submitted version.

Disclosure statement

Ms. Heather Decker and Dr. Adam J. Mellott declare that they do have financial interests in Ronawk Inc. as co-founders of this biotechnology company. All other authors declare that they do not have any conflict of interest in Ronawk, Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17460751.2025.2572177