Cryopreservation of Viable Human Tissues: Renewable Resource for Viable Tissue, Cell Lines, and Organoid Development

Abstract

The availability of viable human tissues is critical to support translational research focused on personalized care. Most studies have relied on fresh frozen or formalin-fixed paraffin-embedded tissues for histopathology, genomics, and proteomics. Yet, basic, translational, and clinical research downstream assays such as tumor progression/invasion, patient-derived xenograft, organoids, immunoprofiling, and vaccine development still require viable tissue, which are time-sensitive and rare commodities. We describe the generation of two-dimensional (2D) and three-dimensional (3D) cultures to validate a viable freeze cryopreservation technique as a standard method of highest quality specimen preservation. After surgical resection, specimens were minced, placed in CryoStor^™ media, and frozen using a slow freezing method (−1°C/min in −80°C) for 24 hours and then stored in liquid nitrogen. After 15–18 months, the tissues were thawed, dissociated into single-cell suspensions, and evaluated for cell viability. To generate primary 2D cultures, cells were plated onto Collagen-/Matrigel-coated plates. To develop 3D cultures (organoids), the cells were plated in reduced serum RPMI media on nonadherent plates or in Matrigel matrix. The epithelial nature of the cells was confirmed by using immunohistochemistry for cytokeratins. DNA and RNA isolation was performed using QIAGEN AllPrep kits. We developed primary lines (2D and 3D) of colon, thyroid, lung, renal, and liver cancers that were positive for cytokeratin staining. 3D lines were developed from the same cohort of tumor types in both suspended media and Matrigel matrix. Multiple freeze-thaw cycles did not significantly alter the viability and growth of 2D and 3D lines. DNA/RNA recovery was similar to its fresh frozen cohort. In this study, we validated 2D and 3D tissue cultures as methods to corroborate the feasibility of viable cryopreservation of tumor tissue. This proof-of-principle study, if more widely implemented, should improve accessibility of human viable tumor tissue/cells in a time-independent manner for many basic, preclinical, and translational assays.

Introduction

The availability of viable human tissues has become critical to support translational research focused on personalized (individualized) cancer care.^1,2 Currently many molecular profiling studies and other related analyses are performed using fresh frozen tissues (FFTs) from biobanks. Most studies have relied upon FFT or retrospective analysis of formalin-fixed paraffin embedded tissue, as both can be used for histopathology, genomics, and proteomics. Yet, many basic, translational, and clinical research downstream assays such as tumor progression, invasion, drug response, patient-derived xenograft models, organoid development, immunoprofiling, vaccine development, and adaptive T cell therapy still require viable tissue. Viable frozen tumor tissues not only contain all components of the original tissue, including malignant epithelial cells, endothelial cells, leukocytes, and fibroblasts, but also represent their relevant tumor models. However, fresh viable human tissues are rare and time sensitive commodities and processing them involves complex logistics in real time. Viable tumor tissue cryopreservation provides the opportunity to overcome this complex process, preserving the original tissue for investigators to be able to evaluate any assay in a time-independent manner. Previous studies have reported on the feasibility of viable freezing for head and neck and cervical tumor tissue.^3–7

At the UCSD Moores Cancer Center Biorepository, we have adopted a viable freezing method of cryopreservation using CryoStor^™ media as the routine methodology for tissue banking. CryoStor is a specialized commercially available serum and protein-free cryoprotectant solution (10% dimethyl sulfoxide [DMSO]), which has a composition similar to the intracellular fluid compartment. This product was designed to mitigate temperature-induced cell stress responses by scavenging free radicals, pH buffering, providing oncotic/osmotic support, energy substrates, and ionic concentrations that balance the intracellular state at ultralow temperature environments (−80°C to −196°C).⁸

To characterize and validate this method, we performed a comparative analysis of viability in fresh and viable frozen tissues by evaluating their ability to develop two-dimensional (2D) and three-dimensional (3D; organoid) cultures. Confirmation of the epithelial origin of 2D and 3D cultures was performed using immunohistochemistry (IHC) staining. All 2D and 3D developed cultures were tested for viability through multiple freeze and thaw cycles. 2D and 3D culture of cryopreserved primary tumor tissues enables dynamic studies of cancer development⁹ invasion^10–12 and drug response.¹³ In this study, we report the successful development of tissue culture as a validation process to validate a viable freeze cryopreservation protocol, as a standard method to retain highest quality solid tissue specimen preservation.

Materials and Methods

Human subjects and tissue

The informed consents were obtained by a trained clinical research coordinator under the Moores Cancer Center Biorepository IRB-approved protocol and annotated for pathological diagnosis, anatomic site, and disease and treatment history. Twenty solid tissues were viably frozen tissue (VFT) for 15–18 months.

The resected specimens were transported to the pathology grossing room on ice and parsed for research or grossing purposes (Fig. 1). The diagnoses for the sample were obtained from electronic medical records system.

FIG. 1.

Cryopreservation—methods and workflow. The patients are consented for tissue collection under an UCSD IRB-approved protocol (IRB No. 181755) that allows access to clinical data. Tissues are collected at surgery time under minimal ischemia time by a research PA assisted by a trained technician. The tissues are then slowly frozen at −1°C/min for 24 hours using an isopropanol freezing container and then transferred to a vapor phase liquid nitrogen freezer for permanent storage. PA, Pathologist Assistant.

Viable freezing of tissue

The tissues were diced into 3–5 mm pieces within an hour of resection, placed in 1.5 mL cryovials, and aliquoted in 1.0 mL of CryoStor (C2874; Sigma) media per tube, placed in freezing containers (5100-0001; Thermo Scientific), and cooled at a rate of −1°C/min until reaching −80°C, and then in liquid nitrogen (LN₂) for long-term vapor phase storage (Fig. 1).

Viability comparison of fresh and VFT

The samples were diced into 3-mm pieces within an hour of resection. Cell viability measured with a TC20 cell counter (Bio-Rad) and the samples were then aliquoted into three groups, FFT, fresh tissue (FT), and VFT. The FT samples were immediately processed following dissociation steps; cell viability was evaluated and the cells plated for 2D and 3D culture. Tissues from the FFT group were placed in 1.5 mL cryovials with no media, and snap frozen in LN_2. After 7 days of storage in LN₂, FFT tissues were processed for DNA/RNA extraction. Tissues from the VFT group were placed in 1.5 mL tubes, and aliquoted in 1 mL of CryoStor (C2874; Sigma) media, transferred to freezing containers (5100-0001; Thermo Scientific) and cooled at a rate of −1°C/min until reaching −80°C. At that point, the samples were placed in LN₂ for long-term vapor phase storage for later processing. Small samples of tissue from each case were fixed in formalin and embedded in paraffin for histological evaluation.

Culture media

Primary culture media (R10F) consisted of 10% fetal bovine serum (FBS; 35-011-CV; Corning), antibiotic/antimycotic (15240-062; Fisher Scientific), sodium pyruvate (S8638; Sigma-Aldrich), MEM-NEAA (M7145; Sigma-Aldrich), Insulin (I2643; Sigma), human Epidermal Growth Factor (EGF; E9644; Sigma-Aldrich), and Hydrocortisone (H0888-1G; Sigma), all of which were diluted in RPMI-1640 supplemented with l-Glutamine. Organoid media (R5F) was similar in composition to R10F with the change of 5% FBS and doubled concentration of EGF.

DNA/RNA evaluation

DNA/RNA was extracted from FFT and VFT using a QIAGEN AllPrep Kit (80204; QIAGEN). Nucleic acid concentration and quality were measured using a Nanodrop, DNA, and RNA integrity number (DIN and RIN) values were measured with an Agilent TapeStation. DNA and RNA concentrations were normalized by dividing concentration in ng/μL against starting amount in mg of tissue fragment to create a value termed DNA and RNA yield in ng/(μL·mg).

Primary cultures

VFT were thawed in 37°C water bath and FTs within 30 minutes of procurement dissociated using Miltenyi Biotec Tumor Dissociation Kit human protocol 2.2.3. In brief, tumor pieces were mixed with Miltenyi proprietary enzyme mix (130-095-929; Miltenyi) and mechanically agitated in a MACS dissociator (130-093-235; Miltenyi) as per the manufacture's instruction, then incubated for 30 minutes at 37°C under constant rotation. The procedure was then repeated, and the tissues were then mechanically agitated again in the dissociator. Contents were filtered through a 70 μm strainer (130-110-916) and the filtrate suspended in R10F. Cell viability was measured using a TC20 cell counter. The cells were plated onto collagen- or Matrigel-coated six-well plates and incubated at 37°C 95% humidity, and the medium was changed every 3–5 days.

Upon reaching 80%–90% confluence, the cells were split using Trypsin (15090-046; Life Technologies Corporation) diluted to 0.05% in Versene (15040066; Gibco). R10F was aspirated from the wells and the cells were rinsed four times with phosphate-buffered saline (PBS; 21-040-CM; Fisher Scientific). Trypsin-Versene (TV) mix was added to the wells, and the cells were then incubated at 37°C until cells had detached from coating matrix. TV mixes were inactivated using R10F media; suspended mixes were pipetted into conical tubes. Cell mixes were centrifuged at 250 × g for 5 minutes to collect pellets. Media was aspirated and the cell pellets (1 × 10⁶ cells) were resuspended in R10F media. Cells were split in a 1:3 or 1:5 dilution, based on needs of experiment.

Organoid culture

1.5 × 10⁵ cells were suspended in 50 μL of growth factor reduced Matrigel (354230, Lot 8015322; Corning). Fifty microliters Matrigel domes were plated on 24-well plates and incubated at 37°C for 10 minutes to solidify Matrigel domes. Five hundred microliters of R5F was added to each well, and organoids were incubated at 37°C. Media was changed every 3 days. After 30–45 days of growth, media from organoid wells was discarded. The wells were rinsed twice with PBS, and Cell Recovery solution (354352; Corning) was added to each well. The Matrigel domes were dislodged and disrupted with gentle pipetting action and incubated at 4°C for 30 minutes, with gentle pipetting action every 10 minutes. The solution was recovered from each well after 30 minutes and spun down at 250 × g. The supernatant was discarded, and pellet (up to 1 × 10^4–5 cells) was suspended in PBS. Vigorous pipetting action was used to shear organoids into a single-cell suspension, and solution was spun at 250 × g. Supernatant was discarded and the pellet suspended in enough Matrigel for a 1:2 passage of cells.

Immunohistochemistry

Cells were plated in 8-well chamber slides (154534; Thermo Fisher) pan-cytokeratin filament stain (Rabbit, Cat. No. Z0622; Agilent).

The wells were washed three times with distilled water, followed by three washes with TBST. The cells were the allowed to cool down to room temperature and then washed three times with PBS. Antigen retrieval was performed in a Diva Decloaker (DV2004 MX; Biocare) at pH = 6.2 with deionized water at 95°C for 30 minutes.

Peroxidase was inhibited with Bloxall (SP-6000; Vector) for 10 minutes, followed by three washes in TBST. IHC was performed on an IntelliPATH Automated IHC stainer (Biocare). Background punisher (BP974; Biocare) was added to cells for 10 minutes for protein block followed by incubation for 1 hour with the specific primary antibody (anti-cytokeratin; 1:8000). The wells were subsequently washed twice in TBST, incubated with anti-rabbit HRP Polymer Secondary Antibody (RHRP520 L; Biocare) for 30 minutes, and washed twice again. The reaction was developed using DAB (IPK5010 G80; Biocare) for 5 minutes, and washed twice in distilled water. The cells were subsequently counterstained with Modified Mayer's Hematoxylin (Cat. No. 72804; Thermo) for 5 minutes, and finally washed twice with distilled water. Standard hemotoxylin and eosin staining was done on formalin-fixed, paraffin-embedded tissue and analyzed for histology confirmation.

Image/data analysis

Cell line images were taken using an Olympus CKX53 microscope with an Olympus XM10 in-line camera. Special and immune stained slides were scanned with an Aperio AT2 Digital Scanner (Leica Biosystems). Differences were evaluated with the unpaired Student's t-test using GraphPad Prism version 7.00 for Windows (GraphPad Software, La Jolla, CA).

Results

Human subjects and tissue

Informed consent was obtained by a trained clinical research coordinator under the Biorepository IRB-approved protocol and annotated for pathological diagnosis, anatomic site, disease history, and treatment history. Fifteen samples of solid tumors were VFT for 15–18 months (Fig. 2).

FIG. 2.

Viability and growth validation. Six freshly resected tissues and eight VFTs were dissociated using the Miltenyi Biotec human tumor dissociation kit. No significant difference in viability between VFTs and fresh tissues was recorded, or in between tissues with treatment before surgery and tissues with no treatment before surgery. VFT, viably frozen tissue.

Cell viability evaluation

Nine dissociated VFTs had a cell viability average of 41% ± 19%, while cells from six dissociated FTs returned a cell viability average of 41% ± 23% (Fig. 2). There were no differences in viability between VFT and FT. Neither were there significant viability differences in cells from tissues that had been treated before surgery and those treatment naive.

DNA/RNA comparison of snap frozen and VFT

No significant differences were recorded in DNA yield between FFT and VFT. (Fig. 3). There was significant improvement (p = 0.015) in average RNA yield from FFT vs RNA yield from VFT. Average RNA yield from VFT was significantly higher than that of RNA FFT (p = 0.015). DIN were 7.10 ± 0.44 for FFT and 8.47 ± 1.08 for VFT. RIN were 7.10 ± 0.36 for FFT and 9.62 ± 0.58 for VFT.

FIG. 3.

Nucleic acid quality and quantity assessment. DNA and RNA were harvested from FFT and VFT. Significant differences in DNA yield were found between FFT and VFT in one of the tissues tested. RNA yield from VFT was significantly (p = 0.015) higher than from FFT group. No significant differences in nucleic acid integrity were found between VFT and FFT. FFT, fresh frozen tissue.

Primary monolayer and organoids development and characterization

Primary monolayers developed beyond the 20th passage from a metastatic lung squamous cell carcinoma to the liver (Fig. 4A), a clear cell renal carcinoma (Fig. 4B), a metastatic mucinous colon adenocarcinoma (Fig. 4C), a papillary thyroid carcinoma (Fig. 4D), and a lung adenocarcinoma (Fig. 4E). By IHC, all cultures were positive for pan-cytokeratin (Fig. 4). Organoids could be reliably produced from these cultures and passaged for at least three times (Fig. 4). Organoids from clear cell renal carcinoma, colon, breast, pancreas, and liver cancer maintained cytokeratin positivity throughout all (three) passages (data not shown). Viability and growth of 2D and 3D cultures were validated for early passage cryopreserved freeze and thaw cycle for each line at least three times and in progress as quality control measure.

FIG. 4.

Primary tumor and cell lines of five VFTs. Each panel corresponds to one line. (A) Top: 2D monolayer primary culture of lung cancer metastases to liver (phase contrast); left inset, positive CK staining is seen in the cytoplasms; right inset; organoid-type growth; CK staining of the organoid also shown positive staining (not shown). The lower panel shows an H&E staining of the original tumor, with solid sheets and cords of malignant infiltrating cells, with an inset at higher magnification to depict the cytological features of the cancer cells. (B) Clear cell renal cell carcinoma 2D monolayer primary culture in the background. CK IHC and 3D organoid culture are seen as described for (A). The H&E lower panel and inset show the characteristic clear cell aspect of the neoplasia and atypical nuclei. (C) 2D monolayer of a mucinous colon carcinoma. CK IHC and 3D organoid culture are seen as described for (A). (Bottom picture) The inset shows islet with signet ring cells (arrow). (D) 2D monolayer of a papillary thyroid carcinoma. CK IHC and 3D organoid culture are seen as described for (A). The lower panel shows an H&E staining of the carcinoma and a higher magnification inset depicting the malignant papillary pattern (arrows). (E) 2D monolayer primary culture of lung adenocarcinoma. CK IHC and 3D organoid culture described as in (A). The H&E stained slide (bottom picture) depicts the malignant proliferation infiltrating the surrounding tissues and showing glandular differentiation (arrows). 2D, two-dimensional; 3D, three-dimensional; CK, cytokeratin; H&E, hemotoxylin and eosin; IHC, immunohistochemistry.

Discussion

This improved viable freeze cryopreservation method for human tissues led to success. CryoStor media was specifically selected for this study as a result of a recently published study,¹⁴ suggesting that CryoStor is a superior media compared to the traditional laboratory-based RPMI+FBS +10% DMSO for reanimation of patient-derived xenograft tissues.

We were able to develop and characterize primary 2D and 3D cell cultures from colon, thyroid, lung, liver, renal, (as represented in Fig. 4) pancreas, and breast cancers. Recovery of viable dissociated cells from VFT was similar when compared to dissociated cells from FFT. These results validate the viable freeze cryopreservation method as a quality cryopreservation method for developing a high-quality viable tissue biobank. Current methods of solid tissue biobanking, standard fresh freezing, or formalin-fixation limit the ability to develop primary cultures, hence limiting downstream assays using viable cells. This standardized approach provides a time-independent renewable resource for viable cells. Furthermore, development of spheroids/organoids (3D cultures) was successful from the same cohort of tumor in both suspended media and Matrigel matrix. Ex vivo 3D models of tumors is a proven culture model, and tumor organoids are being increasingly used in many fields, as they can simulate the physiological structure and function of tissue in situ and maintain the characteristics of tumor cells in vivo,¹⁵ overcoming the disadvantages of traditional experimental tumor models. Organoids can mimic pathological features of tumors and maintain genetic stability, making them suitable for both molecular mechanism studies and pharmacological experiments of clinical transformation. Thus, the viable cryopreservation method is a valid answer to the need for a time-independent source of organoids.

Flash frozen tumor tissue (FFT) is considered to be the best cryopreservation method, and it is a widely accepted method for DNA and RNA preservation.¹⁶ The approach described herein yields DNA of equivalent quality, and significantly increases the RNA yield, while maintaining equal or superior quality due to a significantly better preservation of cell viability. The viable cryopreservation method is therefore a superior method regarding RNA yield over flash-freezing cryopreservation method. This proof-of-principle study demonstrated that viable freeze cryopreservation method could create a time-independent renewable resource for 2D primary cell culture and 3D organoid culture. If more widely implemented, researcher accessibility to human tumor tissue/cells will significantly improve for many basic, preclinical, translational, biomarker discovery and drug discovery assays.

Author Disclosure Statement

No conflicting financial interests exist.

Funding Information

This work was supported by CCSG Grant P30CA23100.