Introduction

Cancer, the disease of our age, ranks as a primary cause of death and a significant barrier to increasing global life expectancy. It is characterized by high complexity and various stages driven by copy number aberrations, genetic mutations, and epigenetic modifications that disrupt numerous signaling pathways1,2. Consequently, cancer should not be considered a static disease. The development and progression of cancer are highly dynamic processes3, leading to genetic variations both within individual tumors and between patients4,5. This intrinsic heterogeneity complicates the translation of molecular and genomic data into effective patient treatments6. Despite the identification of numerous drugs that show promise in cancer models, many ultimately fail in clinical trials due to this complexity and variability.

Conventional two-dimensional (2D) in vitro culture models have been utilized for decades in cancer research and drug screening. However, these models face significant limitations in translating findings to clinical phases due to their inability to accurately replicate the physical conditions of the human body. Specifically, 2D models lack the critical cell–cell and cell-extracellular matrix (ECM) interactions and communications necessary to mimic the native tumor microenvironment7. In colorectal cancer (CRC), the ECM is characterized by specific structural proteins such as collagens (types I, III, IV, and VI), elastin, glycoproteins like fibronectin and laminin, and glycosaminoglycans such as hyaluronic acid. These ECM components significantly support and influence tumor progression, invasion, and drug resistance, emphasizing the importance of accurately replicating ECM composition in disease modeling. The limitations arising from the absence of cell–cell and cell-ECM interactions, can be overcome by employing advanced disease modeling systems that accurately recapitulate tumor-specific features and genetic heterogeneity8,9.

In this context, organoid technology has emerged as a promising innovation, producing more realistic three-dimensional (3D) in vitro culture models. This technology closely mimics original tumor architecture, malignancy, and cancer progression, effectively demonstrating the genetic characteristics and indistinguishable physiology of the tumor10. Fundamentally, the process of generating organoids involves four key steps: the collection of human cancer tissue, enzymatic isolation of cells from the tissue, embedding the isolated cells into a basement membrane extract, and culturing these cells11. The isolation of cells from patient specimens, a critical step in this process, can vary significantly as do the other main processes.

During the isolation step, enzymes specifically target and disrupt cell–cell adhesion molecules and cell-extracellular matrix (ECM) interactions, effectively releasing individual cells or small cell clusters from the original tumor architecture. Different enzymes exhibit distinct substrate specificities, with collagenase primarily degrading collagen fibers, hyaluronidase targeting the glycosaminoglycan hyaluronic acid, and trypsin or TrypLE disrupting cell–cell adhesion molecules such as integrins and cadherins, as well as ECM glycoproteins including fibronectin and laminin. This enzymatic dissociation not only influences initial cell viability but also critically determines the preservation of cancer stem cell populations, which is essential for accurately modeling tumor heterogeneity and maintaining the organoid-forming potential of isolated cells12.

Cancer stem cells possess distinct surface markers, including Leucine-rich repeat-containing G-protein-coupled receptor 5 (LGR5) and prominin-1 (CD133), which plays critical roles in the carcinogenesis and progression of CRC13,14. LGR5, a key intestinal stem cell marker, is essential for self-renewal, differentiation, and tumor initiation, while CD133 enhances tumorigenesis by supporting self-renewal and promoting tumor growth. Both markers contribute to maintaining stem-like properties, which are critical for tumor progression and the generation of cancer organoids15.

Numerous studies have presented various isolation protocols to generate organoids from different cancer types, including CRC, which is the focus of this research11,12,15,16,17,18,19,20,21,22,23,24. Determination of the isolation protocol is essential for generating organoids because they influence the quality, viability, and cellular composition of the initial cell population, all crucial for successful organoid formation. High viability ensures that cells can proliferate and organize into three-dimensional structures, while protocols that preserve cancer stem cells retain the heterogeneity and behavior of tumors in vitro. Additionally, minimizing cell clumping and debris promotes uniform seeding, which is vital for the self-organization and differentiation processes within organoids. Effective and optimized isolation methods are also critical for organoid based disease modeling in research applications such as drug testing, ensuring consistency and reproducibility across samples.

Current protocols for enzymatic isolation of cells, which is essential for successful organoid generation, vary widely, and there is a lack of comparative studies to identify the most effective methods. This study aims to bridge this critical knowledge gap by systematically evaluating the comparative effectiveness of four isolation protocols with distinct enzymes namely, Hyaluronidase type IV, Collagenase type II, TrypLE Express, and Trypsin–EDTA, on the establishment and efficiency of patient-derived organoids (PDO) in CRC. Specifically, the impact of each enzyme-based isolation approach on cellular viability, preservation of cancer stem cell populations, and subsequent organoid formation capability from primary CRC tissues will be rigorously assessed and compared.

Materials and methods

Patient sample collection

To address the identified challenges, freshly resected colorectal cancer (CRC) tissue samples were collected from five patients (aged 50–85 years) who had undergone surgical resection at Acıbadem Izmir Kent Hospital. The CRC tissue samples were preserved in ice-cold transfer medium containing Advanced DMEM/F12 (ADMEM/F12) (Gibco, USA) supplemented with 1X Glutamax (Life Technologies, USA), 10 mM HEPES (Cegrogen, Germany), Penicillin/Streptomycin (Cegrogen, Germany), and 10 µM Y-27632 (Apexbio, Taiwan) at 4 °C until processing.

This study was approved by the Non-Interventional Research Ethics Committee of Dokuz Eylul University, and all patients provided informed consent prior to sample collection. All procedures were conducted in accordance with relevant ethical guidelines and regulations.

Tissue dissociation

Five freshly resected human CRC specimens were transported to the laboratory in the transfer medium, and washed multiple times (> 5 times) with ice-cold DPBS (Ca–/Mg–) (Cegrogen, Germany) containing Penicillin/Streptomycin (Cegrogen, Germany), Amphotericin B (Bristol Myers Squibb, USA), Normocin (Invivogen, USA), Vancomycin (PanReac Applichem, Spain-Germany), and Gentamycin (Biochrom, Germany) until the supernatant was clear. For mechanical and enzymatic digestion, tissue specimens were first chopped into 0.5-1 mm pieces with a sterile scalpel. Tissue fragments were divided into four equal parts by weight and digested with four different enzyme solutions. The following enzymes were used individually on the tissue fragments: 1 mg/ml Hyaluronidase type IV-S (Sigma, USA) and 1 mg/ml Collagenase type II (Sigma, Germany), 1X TrypLE Express (Life Technologies, USA) and 0.005% Trypsin/EDTA (Cegrogen, Germany). Tubes were incubated on a shaking water bath at 200 rpm at 37 °C for 30 min. Digestion was halted by mixing with 10% FBS enriched DMEM to inactivate the enzymes and centrifuged at at 300G for 2 min at 4 °C to pellet cells. Cell pellet mixed with unsupplemented DMEM to wash away residual FBS, and collected dissociated cells were stained with trypan blue dye (Thermo Fisher, USA) and evaluated under microscopy (Zeiss, Germany).

Evaluation of enzymatic isolation efficiency

Several parameters were assessed to determine the quality and quantity of the cells isolated from tissue samples to evaluate enzymatic isolation efficiency. Cell viability was assessed using two complementary techniques: the Trypan Blue exclusion assay (Sartorius, Israel), which relies on the principle that viable cells exclude the dye while non-viable cells take it up due to compromised membrane integrity, and 7-AAD staining for flow cytometry, a fluorescent-based method that operates on a similar principle, allowing for a more detailed and quantitative analysis of membrane integrity and viability within a heterogeneous cell population. In the Trypan Blue assay, isolated cells are mixed with the dye, which penetrates non-viable cells while viable cells exclude it. Non-viable cells stain blue, allowing for their distinction from clear, viable cells. The total number of viable and non-viable cells is then counted using a hemocytometer or automated cell counter, and the percentage of viable cells is calculated to evaluate the enzymatic efficiency of cell isolation. In contrast, the 7-AAD staining method uses flow cytometry to assess cell viability. As a fluorescent DNA-binding dye, 7-AAD selectively penetrates only dead cells with impaired membranes, allowing for the quantification of non-viable cells. Quantification for the isolation process was evaluated by comparing the ratio of stained (non-viable) cells to unstained (viable) cells over 15,000 isolated cells. Unstained control samples are also used to establish the fluorescent threshold for accurate assessment. The cell count per milligram of tissue is calculated by dividing the total number of cells (including both viable and non-viable cells) and the viable cell count (determined using the Trypan Blue exclusion assay) by the weight of the tissue sample used for dissociation. This metric provides a standardized comparison of how well each enzyme can isolate viable cells from a given tissue mass. The quality of cell isolation is visually evaluated using brightfield (BF) images of the isolated cells after the Trypan Blue assay (Zeiss, Germany), focusing on parameters such as the presence of dead cells, cell clumps, and subcellular debris. A scoring system is used (e.g., 1–3 points per parameter), where higher scores reflect better isolation quality, indicating fewer dead cells and cleaner suspensions25.

Detection of cancer stem cells

Flow cytometry is used to assess the presence of specific stem cell markers such as CD133 and LGR5, which are indicative of cancer stem cells. The proportion of these markers in the isolated cell population can further indicate the effectiveness of the enzymatic dissociation in preserving critical cell populations for downstream applications, such as organoid formation. Monoclonal antibodies against CD133 (Beckman Coulter, USA) and LGR5 (BD Bioscience, USA), along with their respective isotype controls, were used following the manufacturers’ protocols. The experiments and analyses were performed using a Navios Ex Flow Cytometer and Kaluza Analysis Software (Beckman Coulter, USA).

Colorectal cancer organoid culture

Each different enzymatically isolated CRC single-cell suspension was processed as described in Fujii et al., with modifications26. After determining viable cell numbers, the cell suspension was resuspended in Matrigel (Corning, Bedford, USA) and 25 µl was dispensed into 48-well culture plates (2 × 103 single cells/well). Plates were incubated inverted at 37 °C in a 5% CO2 humidified incubator (ESCO, Singapore) for 45 min. The minimal culture medium included 1X Glutamax (Life Technologies, USA), 1X B27 (Thermo Fisher Scientific), 1X N2 (Life Technologies, USA), 10 mM HEPES (Cegrogen, Germany), 10 µM Y-27632 (Apexbio, Taiwan), 10 nM Gastrin I (Merck), and 1 mM N-acetylcysteine (Sigma, Germany), 100U/L Penicillin/Streptomycin (Cegrogen, Germany), 50 µg/ml Primocin (Invivogen, USA) in ADMEM/F12. The following niche factors were added to the basal culture medium: 50 ng/ml mouse recombinant EGF (Peprotech, UK), 100 ng/ml mouse recombinant Noggin (Peprotech, UK), 500 nM A8301 (Tocris Bioscience, UK), and 10 µM SB202190 (Sigma, Germany). After Matrigel polymerization, 300 µl of pre-heated complete organoid growth medium was added without disrupting the Matrigel dome. Mediums were refreshed every 3 days, and organoid generation was monitored under the microscope continuously.

Determining organoid yield and size

Organoid yield and size were assessed by evaluating the number and surface area of organoids over a series of time points (days 1, 3, 5, 7, 9, and 11) using confocal microscopy (Zeiss, Germany) in brightfield (BF) mode with Z-stack and extended depth of field (EDF) features. During counting, organoids with a surface area smaller than 500 µm2 were excluded to ensure accurate quantification. Organoid surface area was measured using Zen 2.3 Blue Edition software, with the analysis focusing on the four largest organoids obtained from each enzymatic isolation method to compare the relative efficiency of each protocol in supporting organoid growth. Data were obtained from a total of twenty-seven wells belonging to each experimental group.

Statistical analysis

Statistical analyses were performed using GraphPad Prism version 10.3.1. The Mann–Whitney U test was employed to compare the dissociation quality between different enzyme treatments. Data were presented as means with standard error (SE) unless otherwise specified. A p-value of less than 0.05 was considered statistically significant. The asterisk symbols on graphs *, **, ***, and **** indicate significant differences in the Mann–Whitney U test, corresponding to p-values of < 0.05, < 0.01, < 0.001, and < 0.0001, respectively. The absence of an asterisk indicates no significant difference in the Mann–Whitney U test.

Results

Evaluation of enzymatic isolation efficiency on cell viability

To compare enzymatic performance, colorectal cancer (CRC) tissues were digested using four different enzymatic protocols: Hyaluronidase, TrypLE, Trypsin–EDTA (T/E) and Collagenase. The efficiency of enzymatic tissue isolation was evaluated based on cell viability, cell yields per mg tissue, and the overall quality of the isolation process.

Cell viability percentages were determined using 7-AAD flow cytometry (Fig. 1A,B) and the Trypan Blue exclusion assay (Fig. 1C). Flow cytometry analysis revealed that the highest cell viability was achieved with T/E (91.50 ± 3.43%), followed by TrypLE (83.06 ± 7.22%). In contrast, cells isolated with Collagenase (75.74 ± 4.65%) and Hyaluronidase (70.22 ± 5.18%) exhibited lower viability percentages (Fig. 1B). These findings were corroborated by the Trypan Blue exclusion assay, which showed similar trends: T/E (87.98 ± 2.99%) and TrypLE (86.55 ± 3.93%) yielded higher viability percentages compared to Collagenase (74.45 ± 3.88%) and Hyaluronidase (69.04 ± 4.72%) (Fig. 1C, Supplementary Fig. 1).

Fig. 1
figure 1

Comparative evaluation of enzymatic dissociation methods for tissue isolation efficiency and cell viability. (A) Flow cytometric analysis of viable cells (% viability) using 7-AAD staining for tissues dissociated with Hyaluronidase, TrypLE, Trypsin–EDTA (T/E), and Collagenase. The percentage of viable cells is indicated in each histogram, with 15,000 events analyzed by flow cytometer for each enzymatic dissociation. (B) Bar graph representing the percentage of viable cells determined via 7-AAD staining across the four enzymatic dissociation methods. (C) Bar graph representing the percentage of viable cells determined via Trypan Blue exclusion assay, comparing the dissociation methods. (D) Bar graph representing the total cell count per mg of tissue (by Trypan Blue), reflecting the yield of dissociated cells for each enzymatic method. (E) Isolation quality scoring based on parameters such as the presence of dead cells, subcellular fragments, and cell clumps, with the total score presented for each method. Enzymatic isolation quality is visually assessed based on parameters such as the presence of dead cells (by Trypan Blue), cell clumps, and subcellular debris from the brightfield (BF) images of the isolated cells (Zeiss, Germany). A scoring system was used (e.g., 1–3 points per parameter), where higher scores reflect better isolation quality, indicating fewer dead cells and cleaner suspensions as described by Volovitz et al.25). Data are presented as the mean ± standard error (SE). Statistical significance is denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Full size image

Statistical analysis revealed that T/E yielded significantly higher cell viability than Collagenase (*p < 0.05) and Hyaluronidase (**p < 0.01), suggesting it may be the most effective enzymes for cell isolation in this context. Similarly, TrypLE demonstrated significantly better performance than both Collagenase (*p < 0.05) and Hyaluronidase (*p < 0.05), though it did not differ significantly from T/E. While Collagenase showed lower viability compared to T/E and TrypLE, its performance was slightly better than Hyaluronidase, albeit not statistically significant (p > 0.05). Hyaluronidase shows the lowest percentage of viable cells, indicating this enzyme may be less efficient at preserving cell viability during the isolation process compared to the others.

Focusing solely on cell viability percentages regardless of the total cell count, TrypLE and T/E (Trypsin–EDTA) were found to be the most effective enzymes in preserving cell viability during tissue digestion, while Hyaluronidase exhibited the lowest cell viability percentage. Collagenase demonstrated an intermediate performance, falling between these two extremes in maintaining cell viability (Fig. 1B,C). These observations are complemented by analyzing the morphology and uniformity of the isolated cells, ensuring that the enzymatic digestion method effectively dissociates the tissue while maintaining cell integrity. Together, these metrics provide a comprehensive evaluation of the enzymatic isolation efficiency.

Evaluation of cell yield per milligram of CRC tumor tissue

The total cell counts per milligram of CRC tumor tissue, as determined by the Trypan Blue exclusion assay, are shown in Fig. 1D. Among the tested enzymes, Collagenase yielded the highest total cell count per mg tissue (3664 ± 1520 cells/mg tissue), demonstrating its superior efficiency in tissue dissociation. Hyaluronidase also exhibited a high total cell count (3249 ± 1420 cells/mg tissue), indicating its capability for effective tissue digestion. The lack of a statistically significant difference in total cell counts obtained using Collagenase and Hyaluronidase suggests that both enzymes exhibit comparable efficacy in tissue dissociation (p > 0.05).

In comparison, TrypLE and T/E produced significantly lower total cell counts than both Collagenase and Hyaluronidase (*p < 0.05). T/E achieved a total cell count of 883 ± 333 cells/mg tissue, while TrypLE yielded 663 ± 231 cells/mg tissue. Notably, the difference in total cell counts between TrypLE and T/E was not statistically significant (p > 0.05). This also indicates that TrypLE and T/E exhibit similar tissue dissociation capacities (p > 0.05).

In summary, Collagenase and Hyaluronidase emerged as the most effective enzymes for tissue dissociation, generating higher total cell yields per mg tumor tissue. However, Hyaluronidase appeared to be slightly less balanced in preserving cell viability compared to Collagenase, which achieved a more favorable balance between high cell yield and cell integrity (Fig. 1B,C). Conversely, TrypLE and T/E produced lower total cell counts, but they demonstrated superior preservation of cell viability. This observation suggests that the lower enzymatic activity of TrypLE and T/E, while less effective in fully dissociating tissue, may contribute to better preservation of cellular integrity by minimizing the enzymatic stress during digestion.

Evaluation of enzymatic isolation efficiency on isolation quality

The quality of isolation was assessed by using a scoring system based on the presence of dead cells, subcellular fragments, and cell clumps, as evaluated from brightfield (BF) images obtained during the Trypan Blue exclusion assay (Fig. 1E). For each parameter (dead cells, subcellular fragments, and cell clumps), a score ranging from 1 (lowest) to 3 (highest) was assigned, as described by Volovitz et al. The total isolation quality score was calculated by summing these individual scores, resulting in a range from 3 (minimum) to 9 (maximum). Higher total scores indicate better isolation quality.

Based on the ‘Total Score’ results, Collagenase demonstrated the highest performance with a score of 6.4 ± 0.41, followed by T/E with a score of 6.0 ± 0.36. Hyaluronidase and TrypLE showed lower performance, with scores of 5.6 ± 0.32 and 5.2 ± 0.30, respectively. Despite these differences, statistical analysis revealed no significant differences between the groups.

In addition, the isolation quality was analyzed separately for each parameter, including dead cells, subcellular fragments, and cell clumps. When analyzing the results for dead cells, the scores were as follows: T/E (2.8 ± 0.20), TrypLE (2.2 ± 0.20), Collagenase (1.2 ± 0.20), and Hyaluronidase (1.2 ± 0.20). A higher score for the presence of dead cells indicates fewer dead cells and a greater proportion of viable cells. Accordingly, the findings demonstrate that isolation performed using TrypLE and T/E resulted in a statistically significant reduction in the presence of dead cells compared to isolations performed with Collagenase and Hyaluronidase (TrypLE vs Hyaluronidase- *p = 0.011; T/E vs Hyaluronidase- **p = 0.004; TrypLE vs Collagenase- *p = 0.011; T/E vs Collagenase- **p = 0.004). Notably, these results align with the findings of viable cell percentage analyses (Fig. 1B,C), further supporting the observations.

According to the score analysis of subcellular fragments, the ranking is as follows: Collagenase (2.6 ± 0.20), Hyaluronidase (2.0 ± 0.25), T/E (1.6 ± 0.20), and TrypLE (1.4 ± 0.20). Statistical comparisons showed that isolation with Collagenase resulted in significantly fewer subcellular fragments compared to T/E (*p = 0.022) and TrypLE (*p = 0.013). No significant differences were observed between the other groups.

The ranking according to the results regarding the presence of cell clusters is as follows: Collagenase (2.6 ± 0.20), Hyaluronidase (2.4 ± 0.20), TrypLE (1.6 ± 0.20), and T/E (1.6 ± 0.20). Statistical analysis revealed that isolation with Collagenase resulted in significantly fewer cell clumps compared to TrypLE (*p = 0.022) and T/E (*p = 0.021). Similarly, Hyaluronidase-based isolation also demonstrated a significantly lower presence of cell clumps compared to TrypLE (*p = 0.043) and T/E (*p = 0.043). However, no statistically significant difference was observed between Collagenase and Hyaluronidase, indicating that both enzymes yielded similar results regarding the presence of cell clumps. Since the cell clump outcomes are associated with enzymatic tissue dissociation, the results for them demonstrate a statistical similarity to the total cell count per mg of tissue data presented in Fig. 1D. These findings further emphasize the superior efficacy of Collagenase and Hyaluronidase in facilitating tumor tissue dissociation.

Impact of enzymatic isolation on cancer stem cell markers

To assess the impact of different isolation methods on LGR5positive and CD133positive cells which are critical for organoid formation capacity, flow cytometry was performed to analyze stem cell populations expressing LGR5 and CD133, as illustrated by scatter plots in Fig. 2B,D.

Fig. 2
figure 2

Impact of different isolation methods on LGR5 positive and CD133positive cells. (A, C) Bar graphs representing the percentages of viable LGR5positive and CD133positive cells isolated using different enzymatic protocols-Hyaluronidase, TrypLE, Trypsin–EDTA (T/E), and Collagenase-as determined by flow cytometry. (B, D) Corresponding flow cytometry scatter plots illustrating the distribution of LGR5positive and CD133positive cell populations following enzymatic dissociation. Data are presented as the mean ± standard error (SE). Statistical significance is denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Full size image

In Fig. 2A, the percentage of LGR5positive cells varied across different enzymatic isolation methods, with Collagenase yielding the highest percentage (5.97% ± 0.83%), followed by Hyaluronidase (5.47% ± 0.76%), TrypLE (3.1% ± 0.08%), and Trypsin–EDTA (1.23% ± 0.02%). Notably, Collagenase resulted in a significantly higher LGR5positive cell yield compared to TrypLE (*p = 0.029) and T/E (*p = 0.028). However, the difference in the percentage of LGR5positive cells between Collagenase and Hyaluronidase was not statistically significant. Hyaluronidase also yielded a significantly higher percentage of LGR5positive cells compared to TrypLE (*p = 0.028) and T/E (*p = 0.029), demonstrating its effectiveness in preserving LGR5positive cell populations. Hyaluronidase also resulted in a significantly higher percentage of LGR5positive cells compared to TrypLE (*p = 0.028) and T/E (*p = 0.029), highlighting its effectiveness in preserving LGR5positive cell populations. TrypLE yielded a moderate percentage of LGR5positive cells, which was lower than that observed with Collagenase and Hyaluronidase but significantly higher than T/E (*p = 0.029). Among the tested enzymes, T/E exhibited the lowest percentage of LGR5positive cells, indicating its limited effectiveness in preserving stem cell populations. These findings emphasize the superior ability of Collagenase and Hyaluronidase to maintain LGR5positive stem cells during enzymatic isolation, while TrypLE demonstrated intermediate effectiveness, and T/E was the least effective.

In Fig. 2C, the results indicate that the highest percentages of CD133positive cells were observed in samples isolated using Collagenase (2.40% ± 0.69%) and Hyaluronidase (1.97% ± 0.33%), whereas lower percentages were detected in cells isolated with T/E (1.30% ± 0.15%) and TrypLE (1.13% ± 0.24%). Statistical analysis confirmed that Collagenase and Hyaluronidase yielded significantly higher percentages of CD133positive cells compared to TrypLE (*p = 0.016 and *p = 0.015, respectively).

In summary, Collagenase was the most effective enzyme for preserving CD133positive cancer stem cells during isolation, followed by Hyaluronidase. In contrast, TrypLE and T/E were significantly less effective, suggesting their limited suitability for experiments focusing on CD133positive cancer stem cells.

Organoid growth and expansion dynamics under different enzymatic isolation

Organoid generations were observed and captured their images at the 1st, 3rd, 5th, 7th, 9th, and 11th days of cultures under confocal microscopy (Fig. 3A). In order to compare the expansion rates of CRC organoids, administered with different isolation steps, the organoid numbers (Fig. 3B) and their surface areas of organoids (Fig. 3C) were measured.

Fig. 3
figure 3

Evaluation of organoid generation from tissues dissociated using different enzymatic methods. (A) Representative bright-field microscopy images of organoids derived from colorectal cancer tissues dissociated with Hyaluronidase, TrypLE, T/E (Trypsin–EDTA), and Collagenase were captured on days 1, 3, 5, 7, 9, and 11 of culture using confocal microscopy in bright-field mode. Images were taken at 20× magnification, with a scale bar of 20 µm. (B) Comparison of the of the number of organoids (per well) generated by each enzymatic dissociation method. (C) Comparison of organoid surface area (µm2) at day 11 of culture for each enzymatic dissociation method. Data were obtained from a total of twenty-seven wells belonging to each experimental group. Data are presented as the mean ± standard error (SE). Statistical significance is denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Full size image

Figure 3B illustrates the organoid count per well obtained from CRC tissues processed with four different enzymatic isolation protocols highlighting the efficiency of each enzyme in generating organoids as a key metric for assessing isolation performance and tissue regeneration capacity. Collagenase showed the highest organoid count per well, approximately 54.70 ± 1.57 organoids, significantly outperforming all other enzymes. Statistical analysis confirms that Collagenase produces a significantly higher organoid count compared to: Hyaluronidase (****p: 0.0008), TrypLE (****p < 0.0001), T/E (****p < 0.0001). Hyaluronidase demonstrated the second-highest organoid count, approximately 44.44 ± 2.13 organoids per well. The organoid count for Hyaluronidase was significantly higher than that for TrypLE (****p < 0.0001) and T/E (***p < 0.001) but lower than Collagenase (****p: 0.0008). Both TrypLE and T/E yielded significantly lower organoid counts, approximately 16.0 ± 0.77 organoids per well for TrypLE and slightly above 20.59 ± 0.73 organoids per well for T/E. The organoid count for T/E was significantly higher than that for TrypLE (****p < 0.0001), indicating T/E has better performance in generating organoids.

As conclusion, Collagenase is the most effective enzyme for generating organoids from CRC tissues, likely due to its superior tissue dissociation and preservation of stem cell populations which is crucial for organoid formation. Hyaluronidase also performs well but is less efficient than Collagenase. TrypLE and T/E are less effective in generating organoids, suggesting their limited ability to dissociate tissue and preservation of stem cell populations to support organoid formation. These findings highlight the suitability of Collagenase and Hyaluronidase for experiments requiring high organoid yields, while TrypLE and T/E may not be optimal for such purposes.

Figure 3C represents the total average area (µm2) of the four largest CRC organoids on day 11 of culture, generated using four enzymatic isolation protocols. The data highlights the effectiveness of each enzyme in promoting organoid expansion, as measured by the organoid area. Hyaluronidase-treated organoids have the largest total area, approximately 5852 ± 327 µm2, making it the most effective enzyme for promoting organoid expansion. Statistical analysis shows that the organoid area for Hyaluronidase is significantly larger than those for TrypLE (****p < 0.0001) and T/E (****p < 0.0001) but not shows difference with Collagenase. This suggests that hyaluronidase and collagenase exhibit comparable efficacy in promoting organoid expansion, as reflected by the similarity in organoid area measurements. Organoids isolated with Collagenase show a high total area of approximately 5535 ± 297 µm2, indicating strong support for organoid expansion. The organoid area for Collagenase is significantly higher compared to TrypLE (****p < 0.0001) and T/E (****p < 0.0001). T/E-treated organoids show a total area of approximately 3156 ± 118 µm2, slightly higher than TrypLE which is around 2891 ± 111 µm2, but not shows a significantly difference.

In conclusion, when evaluated in terms of organoid area, the organoids treated with Hyaluronidase and Collagenase exhibited the largest areas, indicating that these enzymes are the most efficient enzymatic dissociation methods for promoting organoid expansion.

The findings regarding organoid count and area demonstrate statistical consistency with total cell count per mg of tumor tissue, cell clump rate (reflecting dissociation efficiency), and the presence of LGR5- and CD133-positive cells, which are indicative of stem cell populations. These results underscore the critical role of these parameters in driving organoid growth and expansion. Notably, collagenase and hyaluronidase stand out as the most effective enzymatic agents for optimizing the tissue dissociation process in organoid culture systems.

Discussion

Patient-derived colorectal cancer organoids have emerged as powerful preclinical models due to their ability to retain the phenotypic, genetic, and intratumoral heterogeneity of the original tumors27. These models offer a physiologically relevant platform for studying disease mechanisms, evaluating drug responses, and guiding personalized therapeutic strategies. Although numerous protocols are available for generating CRC organoids, the choice of dissociation method is often driven by practical factors rather than comparative, evidence-based assessments. This lack of standardized methodological comparisons presents a challenge in optimizing organoid generation workflows and understanding the implications of each approach28.

Among the various steps in organoid generation, tissue dissociation represents a critical determinant of organoid formation efficiency, as it directly affects the quality, viability, and composition of the resulting cell population. Effective dissociation enables the isolation of viable single cells or small clusters, which are essential for initiating organoid development and preserving the regenerative capacity of tissue-derived stem cells11,25,29.

In summary, selecting and optimizing tissue dissociation methods tailored to the specific biological characteristics of colorectal tumors is fundamental for improving PDO formation. Such refinements enhance not only the efficiency and reproducibility of organoid generation but also the translational value of PDOs as models for cancer biology, drug testing, and personalized medicine.

This study aimed to assess the impact of four enzymatic isolation methods—TrypLE, T/E, Collagenase, and Hyaluronidase—on the capacity to generate PDOs in CRC. The evaluation focused on critical parameters such as cell viability, total cell yield per milligram of tumor tissue, and isolation quality, which included the presence of dead cells, subcellular fragments, and cell clumps. Methods such as 7-AAD staining for flow cytometry and the Trypan Blue exclusion assay were employed to analyze these parameters.

Among the tested enzymes, TrypLE and T/E were most effective in preserving cell viability, as evidenced by their higher scores in the dead cell analysis. This suggests that their mild enzymatic activity minimizes cellular damage during tissue digestion, making them ideal for applications requiring viable single cells, such as organoid culture. However, their limited dissociation efficiency was evident in lower total cell counts, and suboptimal performance in reducing subcellular fragments and cell clumps varying sizes, which are not consistently reproducible. These results indicate that while TrypLE and T/E preserve cellular integrity, they are less effective in fully dissociating dense or fibrotic tissues.

Conversely, Collagenase and Hyaluronidase demonstrated superior tissue dissociation capabilities, yielding higher total cell counts and effectively minimizing subcellular fragments and cell clumps. The results obtained in this study are closely related to the specific extracellular matrix (ECM) components targeted by the enzymes used. The finding that collagenase resulted in the lowest amount of subcellular fragments and cell clumps can be attributed to its effectiveness in breaking down collagen fibers, which are abundant in CRC tissues. Similarly, the high efficacy of Hyaluronidase may be explained by its specific targeting of hyaluronic acid, another significant ECM component in CRC tissues. Both enzymes likely facilitate effective tissue dissociation by degrading these major structural components of the dense ECM environment. On the other hand, T/E and TrypLE primarily target cell–cell adhesion molecules (such as integrins and cadherins) and glycoproteins within the ECM (e.g., fibronectin and laminin)8,9. Given the high density of core structural ECM components like collagen and hyaluronic acid in CRC tissues, the comparatively lower effectiveness of T/E and TrypLE observed in reducing cell clusters and subcellular debris is consistent with expectations. These results underscore that substrate specificity of enzymes significantly influences dissociation efficiency, particularly in relation to the ECM composition. The findings emphasize that the type and abundance of ECM components in colorectal cancer tissues are critical determinants of enzymatic dissociation efficiency. This efficiency in tissue dissociation allowed these enzymes to access deeper tissue regions where cancer stem cells, such as those expressing LGR5 and CD133, are predominantly located. Cancer stem cells are crucial for organoid propagation due to their tumorigenic and self-renewal properties30,31. The higher presence of these markers in cells isolated with Collagenase and Hyaluronidase underscores their ability to preserve critical stem cell populations, ultimately contributing to better organoid generation outcomes. For example, Piwocka et al. demonstrated the effectiveness of enzymatic and mechanical isolation methods in primary breast cancer cultures, showing that Collagenase and Hyaluronidase achieved high cell viability while preserving essential cellular characteristics32. Similarly, this study found that Collagenase-treated samples yielded the highest organoid counts, while Hyaluronidase-treated samples exhibited the largest organoid expansion. These results align with the concept that enzymatic methods capable of preserving stem cells while efficiently dissociating tissues are critical for achieving high organoid yields and robust expansion.

The statistical correlation observed between organoid count, area, total cell yield, dissociation efficiency, and stem cell marker expression further highlights the critical role of enzymatic isolation methods. Although Collagenase and Hyaluronidase were associated with slightly higher dead cell counts, this trade-off is acceptable given their ability to access deeper tissue regions and isolate stem cells necessary for organoid propagation. In contrast, the milder enzymatic activity of TrypLE and T/E maintained higher cell viability but limited their dissociation efficiency, making them more suitable for applications prioritizing single-cell viability.

Optimizing enzymatic tissue dissociation methods holds substantial clinical and translational significance, particularly in the context of generating PDOs for precision oncology. Efficient enzymatic isolation not only improves cell yield and viability but also supports the preservation of key cancer stem cell populations, which are critical for maintaining tumor heterogeneity and mimicking in vivo tumor behavior. These qualities are essential for developing PDO models that faithfully recapitulate patient-specific tumor biology, thereby enhancing their utility in drug screening and personalized therapy design33,34,35. Although this study did not directly assess drug responses, the enhanced organoid formation achieved through optimized enzymatic protocols establishes a crucial platform for future translational applications. High-quality organoid models derived through refined isolation methods are not only essential for faithfully recapitulating patient-specific tumor biology but also hold significant promise for advancing drug screening, predicting treatment responses, and ultimately guiding personalized therapeutic strategies in clinical oncology36.

Building on these translational implications, it is important to acknowledge certain limitations of the present study. First, the relatively small sample size and lack of integrated clinical data limit the generalizability of the results across the diverse molecular subtypes and stages of CRC. Considering the pronounced inter- and intra-tumoral heterogeneity characteristic of CRC, patient-specific differences in extracellular matrix (ECM) composition and tumor microenvironment may significantly affect enzymatic isolation efficiency and subsequent organoid formation. Therefore, future studies should incorporate larger and more diverse patient cohorts, accompanied by comprehensive clinical annotations—including tumor subtype, stage, and ECM features—to better understand how patient variability influences enzyme performance and organoid generation capacity. Moreover, the current study did not directly assess the effects of enzymatic digestion on specific ECM components at the molecular level. Given the critical role of ECM integrity in modulating cell release and organoid-forming potential, future investigations should include targeted analyses of key ECM constituents—such as collagen, hyaluronic acid, fibronectin, and laminin—following enzymatic treatment. Such analyses would provide valuable mechanistic insights into how distinct enzymes interact with tumor-specific ECM structures. Additionally, the development and testing of enzyme combinations tailored to the ECM characteristics of CRC tissues may further enhance dissociation efficiency, improve the preservation of cancer stem cell populations, and ultimately support the generation of high-quality, clinically relevant patient-derived organoid models.

This study was designed to evaluate the effects of different enzymatic isolation methods on the efficiency of colorectal cancer organoid formation. While the primary focus was on the initial establishment phase, it is equally important to consider the potential impact of enzymatic dissociation on the long-term maintenance, genetic integrity, and functional applicability of PDOs. Building on our findings, future studies should aim to assess the sustained viability, growth dynamics, and structural consistency of organoids across multiple passages to determine their long-term stability. In parallel, molecular analyses—such as mutation profiling and gene expression studies—can provide insights into the preservation of tumor-specific features over time. Furthermore, evaluating drug response profiles in PDOs derived through different enzymatic protocols would offer valuable information regarding their predictive power in personalized therapy and preclinical drug development. Together, these efforts will contribute to a more comprehensive understanding of how enzymatic isolation influences the translational potential of PDOs in cancer research.

In conclusion, this study highlights the pivotal role of enzymatic dissociation methods in the successful establishment of PDO in CRC, which are increasingly recognized as powerful tools in cancer research and precision oncology. Among the tested enzymes, Collagenase and Hyaluronidase demonstrated superior performance in yielding viable cells, preserving critical cancer stem cell populations, and supporting efficient and consistent organoid formation. These findings not only provide a practical framework for optimizing organoid culture protocols but also address a key technical bottleneck that directly influences the fidelity and translational relevance of PDO-based models. Importantly, this work brings attention to critical avenues for future research, such as dissecting enzyme-specific effects on individual ECM components and accounting for patient-specific differences in ECM composition and tumor heterogeneity. Advancing our understanding in these areas will be essential for refining tissue dissociation strategies and establishing more standardized, reproducible, and clinically representative PDO platforms. By enabling the generation of high-quality organoids that more accurately mirror patient tumors, these improvements will significantly enhance the application of PDOs in drug screening, biomarker discovery, and individualized therapeutic decision-making—ultimately accelerating the translation of benchside findings into bedside benefits.