Advancing urine‐derived stem cells: Cryopreservation validation and sex‐specific metabolism

Abstract

Background

Urine‐derived stem cells (UDSC) are an emerging, non‐invasive source of human stem cells combining easy collection, broad accessibility and high patient compliance with multilineage differentiation capacity. However, key gaps remain in UDSC research, particularly in understanding sex‐related differences and the lack of a validated cryopreservation protocol, a critical aspect for primary cells, given their variability in colony formation, proliferation rates and experimental timing. To address these limitations, this study aimed to establish, for the first time, a reliable protocol for UDSC cryopreservation and to explore potential sex‐related differences, with a specific focus on glycolysis and mitochondrial respiration.

Methods

UDSC were isolated from urine samples of healthy donors (aged 27–50, 4 males and 4 females), cultured in 1:1 DMEM:KSFM supplemented with 10% fetal bovine serum and cryopreserved at passages 2–4 using the same medium with the sole addition of 5% dimethyl sulfoxide. Cells were evaluated for viability, apoptosis/necrosis, metabolic profile and multilineage differentiation potential. Comparisons were performed based on donor sex, as well as before and after cryopreservation.

Results

Male‐ and female‐derived UDSC displayed no significant differences in viability and cell death or metabolic profile. Moreover, supervised and unsupervised machine learning methods were unable to discriminate between the two groups, allowing for pooled data analysis and improved statistical power. Similarly, fresh and cryopreserved UDSC displayed comparable viability, metabolic activity and multilineage differentiation relative to fresh cells, with no detectable differences in computational analyses.

Conclusions

These findings support UDSC adoption for biobanking, disease modelling and regenerative medicine.

Urine‐derived stem cells (UDSC) are a non‐invasive, easily accessible source of stem cells with robust differentiation potential. We developed a cryopreservation protocol for UDSC isolated from healthy donors, assessing viability, apoptosis/necrosis, metabolic profile and differentiation capacity. No cryopreservation‐induced alterations were observed across the evaluated parameters, confirmed by supervised and unsupervised machine learning. Sex‐based analysis also shows no differences due to donor's sex. These findings support standardized UDSC biobanking and strengthen their application in personalized medicine, disease modelling and regenerative therapies.

1. INTRODUCTION

Stem cells are a valuable resource for biomedical and clinical applications, including personalized medicine, tissue engineering and regenerative therapies. However, despite their potential, obtaining stem cells presents challenges. Traditional sources often involve invasive collection procedures or raise ethical concerns. ¹ For instance, while embryonic stem cells are pluripotent, their derivation from embryos poses ethical dilemmas. ² As an alternative, adult stem cells are available, but they typically exhibit lower potency and are limited in quantity. ³ Their collection often requires invasive methods, such as bone marrow aspiration or adipose tissue harvesting. ¹ Another approach is the generation of induced pluripotent stem cells (iPSCs) from somatic cells. While iPSCs offer pluripotency, their production and maintenance present significant challenges, particularly for modelling highly heterogeneous diseases, as their derivation can be resource‐intensive. ⁴ Furthermore, iPSCs carry risks of genomic instability and oncogenic transformation. ⁴ , ⁵ Given these limitations, there is a growing interest in alternative stem cell sources that are non‐invasive, easily accessible and cost‐effective. One promising option involves urine‐derived stem cells (UDSC). UDSC can be collected non‐invasively from voided urine and isolated through a simple centrifugation‐based protocol. ⁶ , ⁷ This approach eliminates the need for invasive biopsies or surgical procedures. Moreover, urine can be collected repeatedly with minimal restrictions and without patient discomfort, ⁷ making UDSC a renewable and patient‐specific stem cell source with broad applications in research and therapy. In addition to their accessibility, UDSC are relatively easy to culture and expand compared to other stem cell types. ⁸ Notably, UDSC exhibit high differentiation potential, particularly into mesodermal‐derived cell types such as osteogenic, chondrogenic and adipogenic cells. ⁹ They can also differentiate into neuronal ¹⁰ and muscular cells, ⁹ highlighting their versatility and relevance for potential therapeutic applications and disease modelling. Recent studies have demonstrated promising results for UDSC applications in various conditions, including tissue injury repair, ¹¹ renal diseases ¹² and neuronal disorders. ¹³ , ¹⁴ These findings highlight the potential of UDSC as a viable alternative to conventional stem cell sources, paving the way for their broader clinical and research applications. ⁸

Despite their promising potential, UDSC remain less well‐characterized than other, more established stem cell types. This limited understanding poses challenges in ensuring the consistency and reproducibility of UDSC‐based experiments. Notably, UDSC may exhibit heterogeneity in their characteristics, including variations in morphology, colony emergence times and proliferation rates. ⁹ Such variability can arise from multiple factors, including differences in developmental stages, cellular activity and possibly their site of origin within the urinary tract. ⁹ , ¹⁵

Donor‐related variability is another critical factor influencing UDSC characteristics. Attributes such as health status, diet, hydration levels and age can affect stem cell yield, quality and function. One key donor‐related factor remains largely unexplored: the sex of the donor. Although broader stem cell literature rarely reports sex‐based differences in viability or metabolic traits, UDSC represent a unique case because they originate from urine, a biofluid with well‐documented sex‐dependent variability. Male and female urinary tracts differ in hormonal influence, urinary composition, osmolarity, pH and microbiota UDSC, ¹⁶ , ¹⁷ , ¹⁸ all of which could potentially shape UDSC properties. For example, sex hormones such as oestrogen and testosterone have been shown to regulate the proliferation, differentiation potential and behavior of other stem cell types. ¹⁹ Notably, differences in isolation success rates between male and female donors have been consistently reported. Samples from female donors often exhibit lower UDSC isolation efficiency and a higher proportion of epithelial cells, ⁹ , ²⁰ which persist in culture until the first medium change, when they are removed because they do not adhere, in contrast to UDSC. These observations provide a strong rationale for exploring potential sex‐related influences on UDSC viability and metabolism in vitro, as such effects could reflect biological features specific to these cells. Furthermore, understanding whether sex‐specific differences exist in UDSC properties is essential for optimizing experimental design and ensuring study robustness. Overlooking sex‐related variability may introduce confounding effects in experiments using paired patient–control samples or pooled donor‐derived UDSC, potentially influencing results and the choice of data analysis strategy. Exploring the existence of any sex‐derived differences would help improve study controls, enhance reproducibility and strengthen the clinical translation of UDSC‐based therapies.

In addition to the aforementioned sources of variability, the success rate of UDSC colony emergence, while relatively high, is not guaranteed, and proliferation rates can be inconsistent. This unpredictability complicates experiment planning, particularly when working with fresh UDSC samples. These challenges become even more pronounced in batch studies or when incorporating paired patient‐control samples within the same experiment, an approach that is crucial for minimizing assay‐to‐assay variability and ensuring robust, reproducible data. A practical solution to mitigate these limitations is the cryopreservation of UDSC. Cryopreservation in liquid nitrogen enables preservation of cells for months to years or potentially longer, while maintaining biological function when appropriate protocols are implemented. This allows for on‐demand retrieval, facilitating the synchronization of experimental timelines, streamlined batch analyses and improved study consistency. While various cryopreservation protocols have been developed for other mesenchymal stem cells (MSCs), they vary widely and lack standardization, leading to potential inconsistencies. ²¹ More importantly, no cryopreservation protocol has been validated specifically for UDSC, and findings from other MSCs cannot be directly extrapolated. ²² Simply adopting freezing protocols designed for other MSCs may not ensure the preservation of UDSC characteristics. Improper cryopreservation conditions, including suboptimal freezing and thawing protocols, can induce osmotic stress, ice crystal formation and membrane damage, ultimately reducing cell viability, recovery rates and proliferative capacity. ²³ These factors, in turn, may compromise the functionality and therapeutic potential of UDSC. ²⁴ Therefore, the absence of a validated cryopreservation method presents a significant barrier to the widespread adoption of UDSC in research and clinical applications. Addressing this gap is critical to improving experimental reproducibility, scalability and practicality.

Given the potential influence of donor sex on UDSC characteristics and the need for an optimized cryopreservation protocol, this study aims to address these two critical aspects. First, we explored whether UDSC from male and female donors exhibit differences in metabolism and overall cell function, providing insights into the role of sex as a biological variable in UDSC research. Second, we developed and validated a standardized cryopreservation protocol for UDSC storage in liquid nitrogen to enhance the practicality, scalability and reproducibility of UDSC‐based studies. By tackling these factors, this work contributes to refining UDSC applications in both research and potential clinical settings.

2. MATERIALS AND METHODS

The information about all the reagents used in this study is provided in Table S1. A detailed description of the methods can be found in Appendix S1.

2.1. Study cohort

Eight healthy adult volunteers (aged between 27 and 50 years, four males and four females) were recruited for this study (Table 1). All participants signed an informed consent, and the study design was approved by the Ethics Committee of Unidade Local de Saúde Coimbra; (reference 319/CES‐OBS.SF.040.2023). Personal data were treated in compliance with current European personal data protection rules [Regulation (EU) 2016/679 and law No 58/2019].

TABLE 1.

Characterization of the donors in the study.

Donor ID	Age	Sex	Race
A01	35	Male	Caucasian
A02	47	Male	Caucasian
A03	50	Male	Caucasian
A04	27	Female	Caucasian
A05	28	Female	Caucasian
A06	39	Female	Caucasian
A07	27	Female	Caucasian
A08	27	Male	Caucasian

2.2. UDSC isolation and culture

UDSC were isolated and cultured following our protocol described in. ⁷ This protocol ensures the maintenance of mesenchymal stem cell properties, including positivity for canonical MSC markers (CD24, CD44, CD73, CD90, CD105) and negativity for haematopoietic markers (CD14, CD20, CD34, CD45). Briefly, voided urine (up to 100 mL) was collected and isolated. Urine samples were centrifuged at 400 g for 10 min at room temperature (RT), washed with phosphate‐buffered saline (PBS) + 3% (v/v) antibiotic/antimycotic. Another centrifugation with the same settings was performed, and each pellet was resuspended in UDSC medium [1:1 DMEM+KSFM +10% FBS + 1% antibiotic/antimycotic (detailed medium preparation in section 1.2 of Appendix S1)], supplemented with an additional 1% antibiotic/antimycotic (2% total). This suspension was divided into 4 wells of a MW24 per sample and cultured at 37°C, until the appearance of UDSC colonies. Passages were carried out when confluence reached ≥80%.

2.3. UDSC cryopreservation and thawing

Cryopreservation was carried out at the lowest passage possible, predominantly at passage (P) 2, with occasional freezing at P3 and rare cases at P4, once cultures reached at least 90% confluence. UDSC were detached and resuspended in UDSC medium, supplemented with 5% DMSO and stored at −80°C in a cell freezing container with isopropanol. Vials were transferred to a liquid nitrogen cryogenic storage system and kept until use. For thawing, cryovials were placed in a 37°C water bath until a small ice core remained. Then, 10 mL of UDSC medium were added, and a centrifugation was performed at 400 g for 5 min. Cells were then plated into a T25 and cultured until P4–5 to allow for recovery and expansion. The experimental workflow assessing cryopreservation effects on UDSC viability and metabolism is presented in Figure S1.

2.4. Assessment of cellular viability and proliferation

To evaluate cellular viability and proliferation of UDSC, cells were plated in 96‐well plates at a density of 3200 cells/well in UDSC medium. 24 and 48 h afterwards, a resazurin assay was performed in a Cytation 3 spectrophotometer (BioTek Instruments Inc., USA). Afterward, cells were fixed with cold methanol/acetic acid (1%) at −20°C following staining with a 15 μg/mL Hoechst 33342 solution and imaging using the InCell Analyser 2200 to assess the number of nuclei. Finally, to identify cellular protein content, cells were incubated with 0.05% (w/v) sulforhodamine B (SRB). After being washed with 1% acetic acid and resuspended in Tris‐EDTA (pH 10.5), absorbance was read in a Cytation 3 spectrophotometer. Data were analysed using Excel and GraphPad Prism 10.

2.5. Flow cytometric analysis of cell death

Cell death was evaluated using fluorescent labeling with Apopxin Green and 7‐Aminoactinomycin D (7‐AAD), indicative of apoptosis and necrosis, respectively, from the Abcam Apoptosis/Necrosis Assay Kit. The manufacturer's procedures were followed. Data acquisition was performed using a BD FACSAria III cytometer, and the data were then analysed using FlowJo software and GraphPad Prism 10. The gating strategy that was followed is presented in Figure S2.

2.6. Assessment of oxygen consumption and extracellular acidification rates

Agilent Seahorse Mitochondrial and Glycolysis Stress assays were performed, evaluating changes in oxygen consumption rate (OCR) and extracellular acidification rate (ECAR). Assays were performed on a Seahorse XFe96 Extracellular Flux Analyser. Data were analysed using Seahorse Wave Desktop Software and GraphPad Prism 10.

2.7. UDSC differentiation potential analysis

Human Mesenchymal Stem Cell Functional Identification Kit was used following the manufacturer's guidelines. Briefly, cells were plated in glass coverslips (adipocytes, 21,000 cells/cm²; osteoblasts, 4200 cells/cm², coverslips coated with 1 μg/mL of fibronectin) or 15 mL Falcon tubes (chondrocytes, 250,000 cells/tube) and were exposed to specific induction media for 21 days, with periodic media replacement. In the end, cells were fixed with 4% paraformaldehyde (PFA) for further analysis.

2.8. Immunocytochemistry for determination of lineage‐specific markers

Following fixation, UDSC were rinsed and incubated with a blocking solution composed of 0.3% Triton X‐100 and 3% Bovine Serum Albumin (BSA) in PBS. Cells were incubated with different antibodies: anti‐fatty acid binding protein‐4 (FABP4) for adipocytes; anti‐osteocalcin for osteoblasts; anti‐aggrecan for chondrocytes. After incubation, the cells were washed with PBS and then incubated with secondary antibodies: Alexa Fluor anti‐goat 488, Alexa Fluor anti‐mouse 568 and Alexa Fluor anti‐goat 568. At the same time, nuclear staining was performed using Hoechst 33342. Cells were then mounted in fluorescence mounting medium, Dako. Images were captured in a Confocal LSM 710 with the Plan‐Apochromat 40x/1.4 Oil DIC M27 objective. Images were analysed using ImageJ 1.54 f.

2.9. Computational data analysis

36 paired samples (isolation replicates, obtained from 4 male and 3 female donors) were considered for the computational data analysis of the parameters of mitochondrial respiration, with SRB normalization. Samples were evaluated according to the cryopreservation status. Analysis of the effect of donor sex was also performed for fresh samples. Principal Component Analysis (PCA) was first performed, then followed by the application of an unsupervised K‐Means algorithm. Supervised decision trees were also fitted to evaluate the creation of predictors for sex and cryopreservation status.

2.10. Statistical analysis

For data analysis, two approaches were considered. The first approach was performed using individual samples from each donor, termed isolation replicates (results showed as supplementary data). However, since some donors contributed with more samples than others, this method inherently assigned greater weight to donors with more samples. To mitigate this imbalance, a second approach was employed, in which the isolation replicates for each donor were averaged, referred to as donor averages. The primary results presented in this paper are based on donor averages. Data are expressed as means ± standard deviation (SD), with points corresponding to either isolation replicates or donor averages. Shapiro–Wilk normality test was performed and unpaired t‐tests or Mann–Whitney tests were carried out for unpaired groups, while paired t‐tests or Wilcoxon tests were carried out for paired analysis. Statistical significance was considered as p < .05.

3. RESULTS

3.1. UDSC viability is not sex‐dependent

The male and female urinary tracts have anatomical and physiological differences ¹⁶ , ¹⁷ , ¹⁸ that can potentially influence the viability and metabolic function of UDSC. ⁹ , ²⁰ Given these distinctions, assessing whether sex impacts UDSC viability and function is crucial to determine whether male‐ and female‐derived UDSC should be analysed as separate groups or can be analysed as a single group in experimental settings.

To explore potential sex‐related differences in UDSC viability, several assays were performed, namely resazurin reduction, which reflects metabolic activity via the reduction of resazurin by intracellular dehydrogenases, the SRB binding assay, which assesses cellular protein content and nuclei quantification, indicative of cell proliferation (Figure 1A). At 24 h post‐plating, analysis of donor replicates' averages revealed no statistically significant differences between male‐ and female‐derived cells in terms of nuclei number (Figure 1Ai), metabolic activity (Figure 1Aii) and protein content (Figure 1Aiii). These data remained consistent when individual isolation replicates were analysed (Figure 3A) and when assays were extended to 48 h post‐plating (Figure S3B,C).

FIGURE 1.

Sex does not influence UDSC viability. (A) Viability and metabolic status. (i) Cell nuclei count (Hoechst 33342); (ii) Global metabolic activity and cell viability (resazurin assay, normalized to nuclei); (iii) Total protein content (SRB assay, normalized to nuclei). (B) Apoptosis/necrosis. (i) Apopxin Green–positive cells (unstressed); (ii) Apopxin Green–positive cells following exposure to 1 mM H₂O₂; (iii) 7‐AAD–positive cells (unstressed); (iv) 7‐AAD–positive cells following exposure to 1 mM H₂O₂. All assays were performed 24 h after plating. Sample information: (A) n = 4 male and 4 female donors; 4–8 isolation replicates per donor (points represent donor averages). (B) (i) and (iii): N = 3 males and 4 females, 2–6 replicates per donor; (ii) and (iv): N = 3 males and 2 females, 2–4 replicates per donor (points represent donor averages). Statistical analysis: Data are presented as mean ± SD; normality was assessed by the Shapiro–Wilk test; between‐group comparisons were performed using either unpaired t‐tests or Mann–Whitney tests, *p < .05. M, Male; F, Female.

FIGURE 3.

Cryopreservation does not affect UDSC viability. (A) Viability and overall metabolism: (i) Nuclei number (Hoechst 33342); (ii) Overall metabolic activity and cell viability (resazurin assay, normalized to nuclei); (iii) Protein content (sulforhodamine B assay, normalized to nuclei). (B) Apoptosis/necrosis: (i) Apopxin Green‐positive cells (unstressed). (ii) Apopxin Green‐positive cells after exposure to 1 mM H₂O₂. (iii) 7‐AAD‐positive cells; (unstressed); (iv) 7‐AAD‐positive cells after exposure to 1 mM H₂O₂. All assays were performed 24 h after plating. Sample information: In panel A n = 8 donors; values represent donor averages, each donor contributing 4–8 isolation replicates; in panel B (i) and (iii) n = 7 donors (1–4 isolation replicates per donor); (ii) and (iv), n = 5 donors (2–4 isolation replicates per donor). Data are shown as mean ± SD. Statistical analysis: Normality was assessed using the Shapiro–Wilk test; paired t‐tests or Wilcoxon signed‐rank tests were applied, as appropriate; *p < .05. BF, Before freezing; AT, After thawing.

To further investigate cell viability, apoptosis and necrosis levels were assessed via flow cytometry using Apopxin Green, which labels apoptotic cells, and 7‐AAD staining, which labels necrotic or secondary/late‐apoptotic cells, under basal conditions and following oxidative stress induction with 1 mM H₂O₂ (Figure 1B). Results evidenced that there were no statistically significant differences between male‐ and female‐derived UDSC in either apoptotic (Figure 1Bi,iii) or necrotic cell populations (Figure 1Bii,iv), both under basal conditions and following H₂O₂ exposure. These findings were consistent in the analysis accounting for individual isolation replicates (Figure S3D).

Some parameters were evaluated to assess the impact of cell passages, which may cause phenotypic changes such as senescence and spontaneous differentiation (Figure S4A,B). Results evidenced no differences at 24 h post‐plating (Figure S4A,B). However, the resorufin fluorescence (generated from the reduction of resazurin by metabolically active cells, indicating cellular metabolic activity) was significantly increased at P5 compared with P4 at 48 h post‐plating (Figure S4Cii,Dii). Thus, data suggest that experiments should be performed 24 h post‐plating whenever possible and at the lowest feasible passage while minimizing passage variability among experimental groups to avoid potential passage‐associated metabolic shifts.

3.2. UDSC metabolic profile is not sex‐dependent

To assess the impact of sex on mitochondrial and glycolytic fluxes, Seahorse Mitochondrial Stress and Glycolytic Stress assays were performed through the measurement of OCR and ECAR, respectively. Data analysis was conducted considering both each donor's samples averages and isolation replicates (Figures 2 and S5, respectively; and Table S2). Results depicted in Figure 2A indicate no statistically significant differences between male‐ and female‐derived UDSC in any of the mitochondrial oxygen respiration‐related parameters, namely basal respiration, proton leak, maximal respiration, bioenergetic health index (BHI), ²⁵ ATP production‐linked OCR, non‐mitochondrial oxygen consumption, spare respiratory capacity, coupling efficiency (Figure 2Aii–ix), baseline OCR and ECAR, and stressed OCR and ECAR (Figure S5A). This suggests that mitochondrial respiration profiles did not differ between sexes. Results were similar when considering donor (Figures 2A and S5A) and isolation replicates (Figure S5B). Regarding glycolytic rates, results depicted in Figure 2B evidenced a lower basal glycolysis (Figure 2Bii) in female UDSC and no statistically significant differences between male and female groups in glycolytic capacity, glycolytic reserve and non‐glycolytic acidification (Figure 2Biii–v). Analysis considering isolation replicates showed no significant differences (Figure S5C) suggesting no biologically relevant sex‐related differences in UDSC glycolytic rates in our cohort. Of note, the effect of passage number on UDSC metabolic characteristics was assessed by comparing cells at P4 and P5 (Figure S4E–H and Table S2). Overall, despite most parameters remaining unchanged, some reached statistical significance. Noteworthily, P5 cells showed a small (~5%) decrease in BHI, which, while statistically significant, likely reflects only a modest decline in cellular bioenergetic capacity without impairing overall function. Therefore, in agreement with Section 3.1, assays should be conducted at the lowest possible passage number, with minimal passage variation among experimental groups to maintain data reliability.

FIGURE 2.

Sex does not produce biologically relevant differences in the metabolic profile of UDSC. (A) Mitochondrial stress test: (i) Time course of the oxygen consumption rate (OCR); (ii–ix) Parameters of mitochondrial respiration derived from the OCR. (B) Glycolytic stress test: (i) Time course of the extracellular acidification rate (ECAR). (ii–v) Parameters of glycolytic function derived from the ECAR. Sample information: For both panels, n = 4 male and 3 female donors; each donor contributed 3–4 isolation replicates; values represent donor averages; data are shown as mean ± SD. Statistical analysis: Normality was assessed using the Shapiro–Wilk test. Group comparisons were performed using unpaired t‐tests or Mann–Whitney tests, as appropriate. *p < .05. M, Male; F, Female.

3.3. Cryopreservation maintains UDSC viability

To assess the impact of cryopreservation on UDSC, we evaluated the effect of a slow freezing cryopreservation protocol on several UDSC functional parameters. First, immediate post‐thaw viability was assessed through trypan‐blue exclusion assay in a subset of samples, showing a decrease in viability from 97.8% ± 0.8367 SD (before freezing) to 82.4% ± 6.768 SD immediately after thawing (before replating), which fits well within expected experimental results for other MSC. ²⁶ Importantly, viability calculated through this assay was fully recovered by P4 (97.0% ± 1.225 SD) when compared to pre‐freezing values at P2 (Figure S6A). Subsequently, to confirm the full recovery of thawed UDSC in culture, functional parameters related to cell viability were evaluated using resazurin and SRB assays, along with nuclei quantification. Results depicted in Figure 3A showed no statistically significant differences between UDSC before freezing and after thawing in any of the assays performed. The average values for positive staining with Apopxin Green (BF: 10.91 ± 4.92; AT: 13.07 ± 7.74) and 7‐ADD (BF: 12.22 ± 6.81; AT: 15.73 ± 9.23), indicate acceptable levels of cell death when compared to other studies with MSC, ²⁷ as this information is lacking for UDSC. Similar results were observed when considering isolation replicates (Figure S6B) and at 48 h post‐plating (Figure S6C,D).

In the same line, flow cytometry was also performed to compare cell death levels before and after cryopreservation, as well as sensitivity to H₂O₂. The results demonstrated no statistically significant differences in Apopxin Green and 7‐ADD labeling between pre‐ and post‐cryopreservation cells under both basal and H₂O₂‐treated conditions (Figures 3B and S6E), suggesting again that our cryopreservation protocol does not affect UDSC viability.

3.4. Cryopreservation keeps metabolic profiles in UDSC

To determine whether UDSC metabolic function is preserved following cryopreservation, Seahorse Mitochondrial and Glycolytic Stress assays were performed on cells before and after the procedure. Data were analysed both as donor averages (Figure 4 and Table S2) to ensure equal contribution from each donor, and as individual isolation replicates (Figure S7 and Table S2).

FIGURE 4.

Cryopreservation does not alter the metabolic profile of UDSC. (A) Mitochondrial stress test: (i) time course of the oxygen consumption rate (OCR); (ii–ix) parameters of mitochondrial respiration derived from the OCR. (B) Glycolytic stress test: (i) time course of the extracellular acidification rate (ECAR); (ii–v) parameters of glycolytic function derived from the ECAR. Sample information: In panel A n = 7 donors, values represent donor averages (1–3 isolation replicates per donor); in panel B n = 7 donors; values represent donor averages (1–4 isolation replicates per donor); data are presented as mean ± SD. Statistical analysis: Normality was assessed using the Shapiro–Wilk test; paired t‐tests or Wilcoxon signed‐rank tests were applied between conditions, as appropriate; *p < .05. BF, Before freezing; AT, After thawing.

Figure 4A evidenced that most parameters assessed by the Seahorse Mitochondrial Stress assay showed no statistically significant differences. Although frozen UDSC presented a significant increase in non‐mitochondrial respiration compared with fresh UDSC and considering donor averages (Figure 4Avii), this difference disappeared when analysing isolation replicates (Figure S7Bvii). In the same way, differences between both groups regarding baseline ECAR and BHI when considering isolation replicates (Figure S7) are lost when considering the donor averages (Figure 4).

Similarly, glycolytic activity also showed mostly no changes after cryopreservation (Figure 4B). Nonetheless, the glycolytic capacity was the only parameter with slight changes in frozen UDSC (~12%) (Figure 4Biii), although the same reduction was not statistically significant in the analysis considering isolation replicates (Figure S7Ciii), suggesting again that this change has low to no biological relevance given the overall context of the other evaluated metabolic parameters.

Thus, our findings support the robustness of our cryopreservation protocol, ensuring that UDSC viability and metabolic function are maintained post‐thawing.

3.5. Computational data analysis is unable to find patterns that stratify UDSC

To complement classical statistical analyses, we used machine learning methods to investigate whether donor sex or cryopreservation status could be discriminated across all assessed parameters. This data‐driven approach adds robustness and depth to our interpretation. We applied supervised and unsupervised machine learning methods to seek for patterns that allow for the creation of separate groups regarding the sex of donors (Figure 5A) and/or the cryopreservation status (Figure 5B). We considered several attributes from the Mitochondrial Stress Assay data. Standardization and PCA were applied to the raw data. K‐Means, with k = 2, was applied to the data to determine clusters. After the application of the algorithm, we analysed the composition of the clusters. The results obtained with the 36 paired samples (before and after thawing) reveal that it was impossible to generate well‐structured or homogeneous clusters (Figure 5Bi). There are two clusters, respectively with 24 and 12 samples (different colours), but in both cases, they are perfectly balanced with samples collected before and after thawing (different shapes). The purity score considering this property is .5, revealing that the cryopreservation status is not a relevant pattern to generate different clusters. The same data were used to fit a supervised decision tree. The validation results (Figure 5Bii) presented in the confusion matrix confirm that it was impossible to train a model that accurately predicts the cryopreservation status. The final accuracy is .64, with a precision of .63 and a recall of .67.

FIGURE 5.

Neither donor sex nor cryopreservation status can be discriminated by machine‐learning analysis of UDSC metabolic data. (A) Sex‐based comparison: (i) Standardization followed by principal component analysis (PCA) was applied to mitochondrial respiration parameters obtained from Seahorse assays. K‐means clustering (k = 2) failed to generate homogeneous groups according to donor sex (purity score = .66); (ii) A decision tree classifier trained on the same dataset was unable to accurately predict sex in validation samples (accuracy = .50). n = 13 male and n = 11 female isolation replicates. (B) Cryopreservation comparison: (i) Standardization followed by PCA was applied to the same metabolic parameters. K‐means clustering (k = 2) was unable to cluster samples according to cryopreservation status (purity score = .50). (ii) A decision tree classifier trained on the same data did not reliably classify samples as before freezing or after thawing (accuracy = .64). n = 18 paired isolation replicates (18 BF and 18 AT). M, Male; F, Female; BF, Before freezing; AT, After thawing.

The study was repeated for the 24 samples that were selected for analysing the impact of sex on UDSC. Clustering results (Figure 5Ai) confirm that it was not possible to create homogeneous clusters when considering the sex of the donor (purity score: .66). The same situation happens when training decision trees. The confusion matrix with the validation results (Figure 5Aii) gives credit to the hypothesis that there are no patterns that can help to differentiate between samples collected from donors of different sex. In this test, the validation accuracy is .5, with a precision of .46 and a recall of .55.

These results indicate that neither donor sex nor cryopreservation status produces meaningful differences in the analysed parameters, consistent with the conclusions from the classical statistical analysis. The machine‐learning outputs therefore reinforce these findings and increase confidence in the absence of detectable biological differences, without revealing conflicting or additional patterns.

3.6. UDSC retain mesenchymal stem cell differentiation potential following cryopreservation

Preserving the multilineage differentiation potential of stem cells after cryopreservation is crucial for applications ranging from disease modelling, where cells can be differentiated into specific pathological lineages, to regenerative medicine, to keep the capacity to generate functional cell types for tissue repair and regeneration, thereby ensuring maximal therapeutic efficacy and safety prior to clinical translation. To evaluate whether UDSC maintain differentiation potential following our cryopreservation procedure, frozen cells were thawed and then differentiated into specialized MSC lineages, namely osteoblasts, chondrocytes and adipocytes. Figure 6 evidences that thawed UDSC successfully differentiated into each targeted cell type, undergoing morphological changes and staining positive for FABP4 (adipocytes), osteocalcin (osteoblasts) and aggrecan (chondrocytes). These findings confirm that UDSC retain their MSC differentiation potential following cryopreservation. These results suggest the maintenance of stemness over several months of storage, as the samples used in this assay were preserved in liquid nitrogen for approximately 9 months.

FIGURE 6.

Thawed UDSC have the ability to differentiate into specialized cell types, characteristic of Mesenchymal Stem Cells. Confocal microscope images of immunofluorescence staining of differentiated cells derived from UDSC after 21 days of culture in their respective differentiation media: (I) fatty acid binding protein‐4 (FABP4), marker of adipocytes (green, first row); (ii) osteocalcin, marker of osteoblasts (green, second row); (iii) aggrecan, marker of chondrocytes (green, third row). Nuclei are stained with Hoechst 33342 (in blue). The composite images are presented in the last column. Scale bars: 50 μm.

4. DISCUSSION

Urine is a promising source of stem cells due to its ease of collection, which contrasts with other conventional stem cell sources. ⁸ However, as a relatively recent model, some of the UDSC characteristics require further clarification. We previously demonstrated that donor's age influences UDSC properties ⁷ ; thus, in this work, we questioned whether donor sex might also affect UDSC. Subsequently, given the primary cell nature of UDSC and its influence on colony emergence timing and proliferation rates—factors that affect sample coordination for downstream applications—we tested and evaluated the impact of cryopreservation on these cells. Establishing a biobank is essential not only for the long‐term preservation of UDSC but also for their on‐demand retrieval, facilitating synchronization of experimental timelines and study consistency. Regarding sex‐induced differences, very limited research exists on how the donor's sex affects the functional parameters of UDSC. In particular, one study showed no significant sex‐based differences in MSC markers in UDSC but lacked further analysis on cell viability and metabolic status. ²⁸ When considering overall non‐UDSC MSC, studies evidence sex‐related changes in cellular characteristics such as proliferation, immunomodulation and metabolism. ²⁹ , ³⁰ , ³¹ , ³² , ³³ For example, a systematic meta‐analysis using Transcriptome Mapper software showed that male and female MSCs express different genes related to metabolic processes. ³⁰ In adipose tissue and bone marrow‐derived stem cells, sex has been identified as a factor for differences in those MSC, with higher metabolic activity and proliferation for cells from male donors. ³¹ Another study that evaluated umbilical stem cells from twins also suggested a higher cell proliferation rate for males, as well as higher adipogenic ability and different expression of stemness markers. ³² On the other hand, human skeletal stem cells showed faster proliferation rates for female samples while having a different phenotype ³³ suggesting that sex‐related cellular characteristics can be heterogeneous across different MSC cell types. Importantly, in our study, under our conditions, we observed that sex has no effect on cell viability, proliferation, mitochondrial and glycolytic function.

Biological differences between males and females, including differences in urine composition itself, such as in patterns of baseline urinary metabolites between sexes (e.g. α‐ketoglutarate, and 2‐hydroxyglutaric acid levels ³⁴ ) are well‐known. Urine‐specific sex‐related changes were not the direct subject of this study, but one cannot overlook the possibility that cells derived from urine may be affected by such environment. Thus, to allow sample pooling of the study population, and enhance reproducibility and statistical power, UDSC sex‐derived differences must be studied to be excluded. Our findings evidence that no biologically relevant sex‐based metabolic differences exist in UDSC within our study's cohort. This is further supported by the results obtained with the supervised and unsupervised machine learning methods that were applied to several mitochondrial respiratory attributes. Clustering and prediction results clearly suggest that there are no discernible patterns linking individual values to donor sex. Having all the observed outcomes in mind, one limitation of our study, concerning sex‐based analysis, is the relatively small number of male and female donors who participated in the study (n = 4 per group), composed of healthy subjects in the age range of 27–50 years old. However, while the cohort size and demographics may influence the generalizability of these findings, the consistent results within this well‐defined cohort offer valuable insights and a strong foundation for future studies involving larger and more diverse populations. Importantly, at the cost of having a lower number of participants, the main advantage of our study design is that it accounted for intra‐donor variability by collecting multiple samples from each individual across different days. This approach mitigates the high heterogeneity reported in UDSC characteristics, even within the same donor, ¹⁵ allowing for a more accurate representation of individual UDSC properties for this group of participants. Our findings are crucial for experimental planning, as they support the pooling of data from both sexes, which enhances statistical robustness.

Then, we addressed the lack of a validated cryopreservation protocol for UDSC, which represents a major barrier to their broader translational use. A robust and reproducible freezing method is essential for enabling future applications, including disease modelling for precision drug‐response testing and mechanistic studies ¹⁴ , ¹⁵ ; high‐throughput drug screening to capture phenotypic heterogeneity ¹⁵ ; pharmacogenomic analyses to identify patient subgroups with differential drug sensitivity ¹⁵ ; and regenerative medicine approaches that leverage the immunomodulatory properties of UDSC. ⁸ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ Although temporary cooling of UDSC at 4°C is adequate for short delays between sample collection and processing, ³⁵ no study to date has validated a full cryopreservation protocol. Establishing such a protocol is critical for preserving stem‐cell functionality over extended storage periods and for supporting reproducible research and potential clinical translation. The use of protocols validated for cell lines in general does not guarantee proper freezing. Additionally, even if one protocol has been validated for a given type of MSC, it does not guarantee that it would be equally successful in maintaining UDSC characteristics, given that functional differences exist between MSCs, including, for example, differences between UDSC and adipose tissue‐derived stem cells. ³⁶  We evaluated a slow‐freezing cryopreservation protocol using UDSC medium identical in composition to the culture medium, supplemented with 5% DMSO. Viability assays conducted before and after thawing revealed no statistically significant differences in overall metabolic activity, protein content or proliferation at both 24 and 48 h post‐plating. Analysis of apoptosis and necrosis also showed no significant differences in basal conditions or after oxidative stress induction with H₂O₂. Metabolic profiling through mitochondrial and glycolytic stress assays demonstrated that most parameters remained unchanged post‐thawing. Further supporting the robustness of our cryopreservation protocol, bioinformatics analysis failed again to cluster samples based on freezing status, reinforcing the lack of major cryopreservation‐induced alterations. Additionally, thawed UDSC retained multipotency, successfully differentiating into chondrocytes, adipocytes and osteoblasts, as evidenced by characteristic morphological changes and lineage‐specific marker expression at the end of the differentiation process.

Collectively, these results confirm that the applied cryopreservation protocol effectively preserves UDSC viability, metabolic function and differentiation potential. Applied to the broader field of MSC cryopreservation, our findings are consistent with and extend previous research. Yong et al. ³⁷ demonstrated that adipose‐derived stem cells cryopreserved in 5% DMSO diluted in DMEM/Ham F‐12, without the addition of FBS, retained both viability and functionality after 3 months, supporting the notion that lower DMSO concentrations can mitigate toxicity while preserving cellular integrity. Similarly, Fujisawa and colleagues reported that synovial MSCs cryopreserved in 5% DMSO with 10% FBS in α‐MEM maintained colony‐forming capacity, but less effectively than when preserved with 95% FBS and 5% DMSO. ³⁸ In contrast, our protocol employs exclusive addition of 5% DMSO as the cryoprotectant and achieves comparable preservation outcomes without the need for additional supplements, underscoring its applicability to this MSC subtype. Furthermore, although many cryopreservation protocols for other mesenchymal stem cell types rely on higher DMSO concentrations (7.5%–10%), ³⁹ our results show that UDSC can be cryopreserved efficiently with only 5% DMSO, minimizing cytotoxicity without compromising cell quality. The validated protocol developed here, using healthy adult donors (27–50 years) as a physiological baseline, constitutes an important step toward UDSC clinical translation by enabling large‐scale biobanking while maintaining viability and multipotency after several months of storage.

In conclusion, this study comprises two major advances. First, it explored the effect of donor sex on UDSC viability and metabolic properties. Across healthy adult donors, we observed no sex‐dependent differences in UDSC viability or metabolic properties. This finding suggests that data pooling across male and female donors is biologically justified and can improve statistical power in future studies. Validation in larger cohorts spanning diverse demographic backgrounds would further establish the generalizability of sex‐independent UDSC biology, but this initial evidence provides a compelling rationale for sex‐integrated experimental designs. Second, we validated a cryopreservation protocol that preserves UDSC viability, metabolism and differentiation potential. Compared to classical stem cell sources, UDSC offer non‐invasive collection ⁸ enabling rapid biobank expansion with diverse cohorts, especially critical for pathologies with high patient heterogeneity. This advancement promotes reproducibility in research, facilitates standardization of UDSC applications and provides a foundational platform for translational applications.

AUTHOR CONTRIBUTIONS

Conceptualization: J. Valero, S. I. Mota, E. Ferreiro. Methodology: G. J. M. Afonso, C. Cavaleiro, S. M. Nunes, F. Pereira, J. Valero, S. I. Mota, E. Ferreiro. Software: G. J. M. Afonso, F. Pereira, J. Valero. Validation: G. J. M. Afonso, C. Cavaleiro, S. M. Nunes, S. I. Mota, E. Ferreiro. Formal Analysis: G. J. M. Afonso, F. Pereira, S. I. Mota, E. Ferreiro. Investigation: G. J. M. Afonso, C. Cavaleiro, S. M. Nunes. Resources: P. J. Oliveira, J. Valero, S. I. Mota, E. Ferreiro. Data Curation: G. J. M. Afonso, C. Cavaleiro, Writing—original draft: G. J. M. Afonso, F. Pereira, S. I. Mota, E. Ferreiro. Writing—review and editing: G. J. M. Afonso, P. J. Oliveira, J. Valero, S. I. Mota, E. Ferreiro. Supervision: F. Pereira, P. J. Oliveira, J. Valero, S. I. Mota, E. Ferreiro, Project Administration: S. I. Mota, E. Ferreiro. Funding Acquisition: J. Valero, E. Ferreiro.

FUNDING INFORMATION

This work was co‐funded by the European Regional Development Fund (ERDF) through the project 0145_CROSS_3DTOOL_4ALS_3_P [Programa Interreg España‐Portugal (POCTEP)]; through the Centro 2030 and Centro 2020 Regional Operational Programme under project CENTRO2030‐FEDER‐00670400 and by the EU Recovery and Resilience Facility and Portuguese national funds via FCT—Fundação para a Ciência e a Tecnologia, under projects, LA/P/0058/2020 [DOI: https://doi.org/10.54499/LA/P/0058/2020], UID/PRR/4539/2025 [DOI: https://doi.org/10.54499/UID/PRR/04539/2025], UID/04539/2025, 2022.13281.BD, DL57/2016/CP1448/CT0027 and 2022.00011.CEECIND.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

Appendix S1.

Appendix S2.

Afonso GJM, Cavaleiro C, Nunes SM, et al. Advancing urine‐derived stem cells: Cryopreservation validation and sex‐specific metabolism. Eur J Clin Invest. 2026;56:e70172. doi: 10.1111/eci.70172

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are openly available in figshare at https://doi.org/10.6084/m9.figshare.30272311.