Cryopreservation of human lung tissue for 3D ex vivo analysis

Abstract

Ex vivo culture techniques have assisted researchers in narrowing the translational gap between the lab and the clinic by allowing the study of biology in human tissues. In pulmonary biology, however, the availability of such tissues is a limiting factor in experimental design and constrains the reproducibility and replicability of these models as scientifically rigorous complements to in vitro or in vivo methods. Cryopreservation of human lung tissue is a strategy to address these limitations by generating cryopreserved biobanks of donors in the ex vivo study of pulmonary biology. Modern cryopreservation solutions, incorporating blends of cryoprotective extracellular macromolecules and cell-permeant non-toxic small molecules, have enabled the long-term storage of human lung tissue, allowing repeated experiments in the same donors and the simultaneous study of the same hypothesis across multiple donors, therefore granting the qualities of reproducibility and replicability to ex vivo systems. Specific considerations are required to properly maintain fundamental aspects of tissue structure, properties, and function throughout the cryopreservation process. The examples of existing cryopreservation systems successfully employed to amass cryobanks, and ex vivo culture techniques compatible with cryopreservation, are discussed herein, with the goal of indicating the potential of cryopreservation in ex vivo human lung tissue culture and highlighting opportunities for cryopreservation to expand the utility of ex vivo human lung culture systems in the pursuit of clinically relevant discoveries.

Introduction: ex vivo lung models in pursuit of improved translational capacity

For as long as pulmonary biology has been studied, researchers have attempted to determine the mechanistic causes of lung diseases by replicating characteristics of the human lung in a laboratory environment. This approach has primarily involved the utilization of genetic manipulation or disease models derived from the culture of cells or animals [1,2,3,4]. While these models have advanced contemporary biology by providing a wealth of mechanistic information on pulmonary disease at a molecular level [5,6,7], the translation of these findings into clinical efficacy has demonstrated mixed success [8,9,10]. With around 80% of drugs found to be effective in murine studies ultimately failing clinical trials in humans [11, 12], and a success rate of between 50 and 60% in clinical trials of lung cancer therapies originating from animal studies [13], there is a significant need for additional models to improve the correlation between laboratory observations and clinical outcomes [10, 11, 14]. Ideally, new models will improve upon in vitro techniques by including features such as a retention or recreation of the lung’s characteristic 3D structure [15,16,17], varied cellular composition [18, 19], interactions between the more than 60 constituent lung cell types [20,21,22], relevant mechanical forces and stimuli [20, 23, 24], response to perturbation or infection [25, 26], and elements of any chosen disease states [27, 28]. Reproducing these attributes with as much fidelity as possible to a functional human lung is critical in ensuring in vitro results will properly translate to clinical applications [29, 30]. Recent advances have seen the development of ex vivo human lung culture techniques emerge as a method for satisfying these criteria [25, 31,32,33].

With the appropriate range of human lung cells arranged in a natural 3D architecture [31], ex vivo culture techniques can be utilized to study traits of human pulmonary biology [34, 35], response to drugs [36,37,38] or pathogens specific to humans [18, 39], or investigate the early stages of diseases which are either difficult to faithfully model in animals [40, 41] or challenging to observe in humans [42, 43]. Choice of ex vivo culture method will depend on factors such as availability of tissue, resources, and research question [44]. Common ex vivo lung culture models include human precision-cut lung slices (hPCLS) [45], culture of explanted lung tissue fragments [32], decellularized lung tissue [34], and growth of organoids derived from primary lung cells [29]. Each of these methods has strengths which suit them to different aspects of pulmonary biology, but all have the advantage of utilizing primary human cells in three dimensions for the study of lung disease [44].

In addition to the wealth of potential methods available for ex vivo culture of the human lung, there are numerous reasons ex vivo culture is increasingly employed in the study of human pulmonary biology [4, 25, 44]. Generally, utilization of human specimens is advantageous when the study requires a fidelity to in vivo human lung biology difficult to accurately replicate in animals, such as in viruses specific to human hosts [18, 39, 46, 47]. In this scenario, ex vivo explant culture provides a method for directly observing host-pathogen interactions in the human lung’s component cells [46, 47] within a 3D model containing a natural extracellular matrix (ECM) and resident immune cells [18, 48], a configuration not reproducible in two-dimensional cell culture [49]. With this 3D nature also comes the potential to mimic the breathing motions found within an active lung [24, 50, 51], responsible for critical mechanical stimuli which govern the balance of Type and Type II alveolar epithelial cell (AECI and AECII) identity [52] and extracellular matrix composition [53] in these cellular relationships, and essential to the proper modeling of inhaled drug delivery [54]. In addition, there are various physiological differences between animal models of the lung and human lung models [10, 55], not all of which are known [56,57,58]. In the case of diseases such as cancer, ex vivo culture of organoids or explants can also incorporate autologous immune cells, which is impossible in cell lines [59], recapitulate spatially dependent features of tumorigenesis [60], or function as a personalized testbed for immunotherapies [61]. For these reasons, ex vivo culture methods are often suggested as a way to bridge the translational gap between the lab and the clinic, improving correlation between lab studies and the outcomes of clinical trials [62].

Despite the abundance of potential for ex vivo lung culture to produce new models with improved clinical correlation, there are some limitations to these methods. The numerous physical [4, 45, 63], chemical [64], and biological [65, 66] discrepancies which naturally arise during the culture of human lung cells and tissue fragments outside of their native milieu present challenges in the correct modeling of in vivo outcomes and mechanisms [44]. By necessity, all of these models will also only partially represent a region of the lung rather than the entire tissue or organ, limiting conclusions derived from experiments involving them to local effects [66]. Several ex vivo models do not incorporate mechanisms for waste metabolite clearance or circulation, the forces of which provide critical signaling cues in biological processes such as coagulation [67, 68]. This lack of circulation precludes ex vivo models from modeling systemic effects, such as toxicity from a drug treatment stemming from reactivity in other organs, or whole-body immune system dynamics relating to immune cell migration in the context of infection, as the majority of ex vivo models only represent a single organ [45]. Some of these models, such as organoids, are also highly sensitive to the composition of culture media [69, 70]. The density of tissue cultured ex vivo and the necessity of culture media also places many ex vivo lung tissue models in what is at best a partially hypoxic environment, potentially creating challenges for the analysis of certain hypotheses [71]. The formation of specific hypotheses, adapted to the limitations of each model, is thus necessary to ensure relevance and correct translation to clinical outcomes [72].

Ex vivo tissue culture is impossible without access to high-quality donor tissue for experiments, especially those which require multiple replicates [66]. Obtaining human lung tissue requires coordination between clinics and potential donors, and is often time-consuming and laborious [73]. Throughout the entire tissue collection process, care must be taken to document variables which could potentially affect the characteristics of the recovered sample and subsequent experimental results, necessarily including but not limited to patient health history and diagnoses associated with the collected tissue, location and condition of the collected tissue from the donor’s lungs, media formulation and temperature used to transport the tissue, warm ischemia time prior to sample collection, method and instrument of tissue dissection, amount of elapsed time prior to dissection, dimensions and thickness of dissected tissue, length of adaptation period, and tissue viability prior to and following any storage [74]. In addition, cell and tissue composition of the lung is functionally distinct across the alveoli, bronchioles, and bronchi, so the lobe of origin and presence of airways or amount of parenchyma in tissue collected and prepared for ex vivo culture should be considered [22, 74, 75]. As the majority of available lung donors are likely to belong to a disease population, the effects of disease history are also a potential variable in the use of their tissue [66], with historical factors such as smoking or drug treatment history particularly noted for their effects on the tissue received [76, 77]. In certain cases, the availability of diseased tissue may be desirable, such as in the usage of lung recovered from patients with idiopathic pulmonary fibrosis [78, 79] or COPD [80]. Apart from these elements, one of the primary limitations of ex vivo lung models is tissue scarcity, which constrains the ability of these models to offer the scientifically rigorous elements of reproducibility (the ability to obtain the results of multiple experiments within the same donor) and replicability (the ability to obtain the results of the same experiment in multiple donors) [81].

Cryopreservation to enhance the utility of ex vivo lung models

Cryopreservation has the potential to address the scarcity of high-quality human lung donor tissue for ex vivo models and improve their inherent reproducibility and replicability weaknesses [82,83,84]. In addition to storing tissue available at the researcher’s convenience [85], cryobanks allow for the study of multiple donors in parallel [84], the opportunity to observe effects potentially related to donor heterogeneity [18], and afford a platform of study suited to high-throughput experiments [82]. Perhaps most advantageous compared to the use of fresh tissue, cryobanks provide the benefit of allowing repeat experiments to be performed on the same donor, which would otherwise be impossible [18]. These advantages have prompted researchers to amass cryobanks as a basis for clinically relevant studies [76]. However, just as tissue culture in three dimensions presents numerous challenges which do not apply to two-dimensional cell culture [31], there are unique considerations for which researchers must account in the cryopreservation of 3D tissues and organoids. To appropriately address these considerations, a thorough understanding of the principles of cryopreservation [86] is necessary.

For successful cryopreservation, cells must be protected from injury caused by the numerous phase transitions implicit in transfer to and from a sub-freezing environment [87]. The loss of cell viability at or below sub-freezing temperatures during cryostorage occurs due to the mechanical disruption induced by ice crystal formation [88, 89] inside and outside of cells [90, 91], and most cryoprotectants function by preventing the destructive effects of ice crystal formation on critical cellular structures such as membranes [92, 93]. Additional damage to freezing or thawing cells is caused by the cellular sequestration of salt during the formation of extracellular ice crystals, effectively dehydrating cells and inducing severe volume reduction [94, 95] prior to intracellular aqueous salt crystallization and further deleterious effects [96]. The severity of these crystallization events is generally dependent on the rate of cooling [97] throughout the cryopreservation process, as a solution of cooling or heating cells passes through temperature points at which phase changes occur and crystallization or recrystallization is induced [92, 98]. The optimal rate of cooling depends on the type of cryoprotectant used [87]. Slow cooling occurs at a rate of -1 °C/min in specialized containers, whereas vitrification, or fast-cooling, occurs at rates exceeding − 20,000 °C/min [99]. Cooling rate must be considered in the context of osmotic stresses placed upon cells by the permeating components of a cryoprotectant [100], and effective cryoprotectants will ease the mechanical stresses induced by these events over all the phase changes during freezing and thawing in such a way that the plasma membrane remains intact and the overall viability of the cryopreserved cells is not compromised [88, 92, 101].

The earliest discovered cryopreservation methods involved the use of cell-permeant small molecules which prevent cell lysis or damage during freeze-thaw cycles, such as glycerol [102] or DMSO [103]. Despite the widespread, nearly century-long success of these compounds at preserving a wide variety of cells at sub-freezing temperatures [102], these cryopreservatives are poorly suited to some applications [97]. DMSO, for instance, poorly maintains functional macrophages through a freeze-thaw cycle, heavily impacting their ability to generate reactive oxygen species [104], and in some cases inhibits the differentiation potential of embryonic stem cells [105]. Furthermore, there are numerous reports of broad cytotoxic [106], transcriptional [107], and epigenetic [108] side-effects associated with the use of DMSO in various cell types [109]. Glycerol is less toxic than DMSO overall [110], but its high viscosity poses handling challenges upon thawing [111] and alters intracellular protein interactions [112]. In cryopreservation of tissues, which are by nature comprised of a heterogeneous cellular constituency, the differential effect of cryoprotectants on viability and function also varies depending on cell type, with neutrophils and primary fibroblasts notably affected by the cryopreservation process, especially when DMSO is used [113, 114]. Researchers should therefore accordingly account for the uneven alteration of cellular viability and function when assessing the suitability of a cryopreserved ex vivo model to meet research objectives [74]. The detrimental effects observed in cryoprotectants such as DMSO and glycerol have prompted exploration of alternate cryopreservatives for use in research applications where these drawbacks would prevent meaningful observations, potentially with a more consistent preservation of all cell types present in tissues [93, 97].

A major area of focus in the development of new cryopreservatives is that of biologically compatible macromolecules, inspired by their utilization in extremophiles to withstand inhospitably cool temperatures [93, 97, 115, 116]. Nature has served as a guide for the study of several of these macromolecular cryoprotectants [117], with polyols and sugar polymers some of the first observed to help larger organisms such as insects and amphibians survive freezing temperatures [118,119,120]. Types of macromolecular cryoprotectants which have seen laboratory success in preserving viable cells at rates similar to that of DMSO or glycerol include polysaccharides of trehalose [116, 121, 122], sucrose [116], and inulin [123], polymers which contain mixed charges to assist in maintenance of cellular osmotic integrity (polyampholytes) [97, 124] such as proteins [125], or other polymers such as polyethylene glycol [126], polyvinyl alcohol [127], or polyvinylpyrrolidone [128]. In addition to these biologically derived compounds, hydrogels of agarose [129, 130], gelatin, and alginate [131], materials used in the preparation of PCLSs [132], have also been successfully employed as cryopreservation agents [133]. Whereas small-molecule cryoprotectants act by restricting ice crystal formation and regulating osmotic pressure inside and outside of cells [98, 100, 134], macromolecular cryopreservatives are not cell-permeant and function by controlling the flow of solutes and water into and out of cells during the freezing process, effectively preventing excess dehydration associated with extracellular ice formation while still allowing enough dehydration to prevent the formation of intracellular ice crystals and maintenance of cell size [124]. Macromolecules containing mixed charges have also been presumed to act in a manner which preserves the integrity of the cell membrane or other intracellular structures, such as proteins or microtubules, independent of any ability to restrict ice crystal formation during the freezing and thawing process [97, 135, 136]. Within these molecules, charge ratios and the location of charges are critical to their ability to maintain viability of cryopreserved cells [136, 137]. Hydrogel-based cryoprotectants such as those of agarose function by reducing free water and mechanically inhibiting ice crystal formation during freezing while also restricting recrystallization of ice during thawing [129].

Despite the effectiveness of some biologically derived macromolecular cryoprotectants in their host species, these compounds do not achieve biological efficacy in isolation [97], and have been shown to inadequately protect the cell membrane from the formation of sharp extracellular ice crystals when used alone as cryoprotectants [138]. This would indicate that any osmotic control or membrane alteration provided by these compounds is not alone enough to safeguard cellular viability during the freezing and thawing process, and biological mechanisms of tissue cryoprotection rely on the presence of both large and small molecules [139]. Indeed, this phenomenon can be observed in frogs, where a blend of polysaccharides and urea balances the osmotic gradient of solutes within cells and prevents cell shrinkage during freezing due to the departure of water, enabling survival in harsh freezing temperatures [140]. Examples such as these, which combine cryoprotective macromolecules and small molecules, serve as a guide for how cryoprotectants utilized in a laboratory setting could be designed to improve control over detrimental osmotic effects while mitigating intracellular ice formation and cell dehydration [141]. The properties of these biological examples have become a template in the design of synthetic cryoprotectants, such as zwitterionic small molecules [134], which could pair with emerging synthetic macromolecular compounds to achieve the goal of finding non-toxic and biologically inert cryopreservation solutions [87].

Small-molecule and macromolecule cryoprotectants, as well as solutions containing combinations of both, have been shown to effectively cryopreserve ex vivo lung tissue for on-demand culture [18, 77]. As air comprises 80% of lung tissue, most conventional small-molecule cryopreservation solutions such as DMSO easily diffuse through the tissue mass and displace harmful ice crystal formation, enabling a consistent freezing point propagation throughout the explanted tissue that is within 10% of that found in other solid tissues [142]. The choice of appropriate cryopreservation technique and methods with which to validate effectiveness of cryopreservation will vary, however, depending on which ex vivo lung model is employed [77]. While cell-permeant cryopreservatives like DMSO are an appropriate media in which to preserve hPCLSs, for example, these same cryopreservatives may be incapable of penetrating denser explants at a high enough concentration for desired levels of viability after exposure to temperatures found in the vapor phase of liquid nitrogen [18, 143]. Below, ex vivo lung culture techniques suitable for cryopreservation and cryobank accumulation are discussed, as well as the appropriate cryopreservation techniques for each and notable descriptive or hypothesis-driven studies which have been performed using viable lung tissue from cryobanks.

Ex vivo lung models which can be cryopreserved

Human precision-cut lung slices (hPCLSs)

Fig. 1

Diagram illustrating the cryopreservation of human precision-cut lung slices for cryobank generation and experimental use. The thin nature of precision-cut lung slices enables the mass storage of several slices from the same donor in a multitude of cryoprotectants. Figure illustrations were generated in BioRender

Perhaps the most-studied model in cryopreservation of ex vivo human lung tissue is that of precision-cut lung slices, sections of lung tissue measuring around 1 mm or less in thickness [144]. PCLSs were first employed in animal studies of lung toxicology during the 1980s [145, 146] before being utilized in studies of human toxicology in 1994 [147]. hPCLSs are generally best obtained from a fresh lobe of lung which has been inflated with low melting point agarose or gelatin to prevent the collapse of the alveoli, sectioned, and hole-punched, before the hole punches are sliced by a vibratome or other precision slicing instrument to produce slices of the desired thickness [45]. Several slices can be created from each hole punch, and several hole punches can be taken from each section, generating a large number of replicates from a given region of the lung in the same patient [45, 63]. In addition to the parameters which should be considered in the collection of any tissue for ex vivo culture, characteristics of collected hPCLSs, including composition of filling gel, gel temperature, tissue sampling method, dimensions of tissue selected for slicing, elapsed time prior to slicing, resultant PCLS thickness, and slicing instrument, blade, and angle should be documented to account for potential variability [74]. Each hPCLS represents a near-3D spatial snapshot of the donor lung, complete with any airways, blood vessels, or parenchyma present in the hole punch [148], a full complement of cellular heterogeneity including structurally critical cells and resident immune cells [149, 150], an air-liquid interface [151], the potential inclusion of mechanical stimuli associated with breathing [24, 50, 51], and the ability to function as a model of physiological phenomena such as airway constriction [152,153,154,155,156]. In addition to their usefulness in toxicology and drug discovery studies [74, 157], these features have made hPCLSs an excellent model for the study of a wide variety of lung diseases, including bacterial [25] and viral [149] infections, chronic inflammatory disorders [4], allergic reactions [158], and cancer [159]. hPCLSs can also serve as a platform for comparing murine and human response to the same disease [160], and can be generated from healthy or diseased donors, meaning that disease phenotypes can either be studied directly as they existed in situ or induced experimentally for analysis [28, 161].

Despite the utility of the hPCLS model, there are limitations which must be considered in the course of experimental design [45]. Though the thinness of hPCLSs allows for the formation of an air-liquid interface, the 3D structural detail of this model is limited compared to that of tissue fragments, potentially making that model a better choice for studies which require greater spatial detail [18, 144]. This lack of dimensionality in hPCLSs also allows any factors with which they are treated in culture media, including viruses or drugs, to pass above epithelial barriers and reach all cells of the slice in a way that would not occur physiologically [45]. While it was initially believed that culture of PCLS was limited to less than 7 days without specific culture conditions, such as within hydrogels coated in integrin-binding peptide sequences to replicate extracellular cues [162], more recent experiments have demonstrated a viable lifespan of around 4 weeks in these tissues, with media formulation presumed to be a determining factor in hPCLS survival in culture [85].

There is enormous scope for enhanced reproducibility and replicability in the hPCLS model due to the large volume of slices generated from each donor [45, 63]. Prior to cryopreservation, cold storage of PCLS was explored as a possible method to maintain these tissues for later study [163, 164]. Initial efforts to cryopreserve PCLSs occurred in animals, with a 2014 study evaluating cryostorage of murine PCLS in 10% DMSO and Dulbecco’s modified Eagle medium (DMEM-F12) reporting a nearly 50% reduction in cellular metabolic activity post-thaw despite no significant loss in viability and preservation of airway contractility in response to chemical stimulation [132]. These findings were corroborated in a 2016 study also conducted in animal PCLSs, which further reinforced the ability of cryopreserved lung to appropriately respond to stimuli by analyzing PCLS response to zinc toxicity [157]. Another 2016 study using a similar preparation technique and the same cryopreservation medium in human PCLSs compared fresh and frozen slices from three donors, and found frozen tissue to have comparable phagocytic function in stimulated resident immune cells, comparable proliferative capacity in stimulated T cells, and comparable modulation of airway contractility in smooth muscle cells as a response to TAS2R agonists [82]. All of these studies utilized a DMSO-based (small-molecule) cryopreservative. With an appropriate protocol for cryopreservation, the high quantity of slices which can be generated from a single lobe of human lung thus gives this method immense potential for reproducibility in a single donor, and an equivalently high potential for replicability in a cryobank of multiple donors, overcoming the main shortcoming of ex vivo tissue models (Fig. 1) [163].

A 2023 study utilizing a proprietary cryopreservation medium provided the most detailed analysis to date on cryopreservation in hPCLSs. The authors determined there was no difference in viability, protein concentration, response to lipopolysaccharide (LPS), tissue structure, or surfactant production between fresh or frozen hPCLSs which were successfully cultured in DMEM-F12 supplemented with 1% insulin-transferrin-selenium for a period of 4 weeks [85]. The results of this study indicate that choice of cryopreservative and culture media are critical to effective utilization of cryobanks in the hPCLS model, suggesting the possibility that the thin nature of hPCLSs may allow for successful cryopreservation using DMSO, but side-effects associated with altered cell metabolism will be present unless a less toxic cryopreservative is used. Proper cryopreservation of hPCLSs thus greatly expands their utility as a tool for modeling human biology ex vivo by addressing the issue of tissue availability, and subsequently the time constraints limiting the reproducibility and replicability of this model. The usefulness of appropriate cryopreservation techniques to this end was confirmed in a 2024 study, where the ability of cryopreserved hPCLS to serve as a model of infection for viral pathogens such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was explored as part of an analysis of the hPCLS platform in general [165]. Cryopreservation with non-toxic formulations thus holds great promise as a method for removing the greatest obstacles to the use of hPCLSs in studies of pulmonary biology.

Explant tissue fragments

Fig. 2

Diagram illustrating the sample preparation process for ex vivo lung microexplants. With the appropriate cryoprotectants, tissue microexplants can be up to approximately 0.5 cm³ in volume. Figure illustrations were generated in BioRender

Direct ex vivo culture of explanted lung tissue can be accomplished by culturing a whole explanted lung in its entirety or by culturing smaller fragments of dissected tissue [18, 47, 79]. While ex vivo lung perfusion systems that maintain a complete human lung for study are perhaps the method most faithful to human lung biology, this is also the most expensive, least reproducible, and least replicable way to culture human lung ex vivo, and removes available viable lungs from the transplant pool [166, 167]. Culturing dissected fragments of lung tissue from the same donor is therefore a preferable method to improve reproducibility, similar to the approach used in the generation of hPCLSs [168]. To create a 3D ex vivo tissue model of the lung parenchyma, the pleura and airways are removed from a fresh lobe of lung and the alveolar space is dissected into fragments of various dimensions which can then receive different experimental treatments [18, 47, 168, 169]. Ex vivo tissue fragments can vary in diameter from 1 mm or less (microexplants) [18] to nearly 10 cm [169], and carry the fully functional cellular diversity of the alveolar space in an accurate 3D structure [170] more relevant to whole lung tissue than that found in hPCLS or organoids [18]. Because the 3D architecture of the lung is the best preserved of all ex vivo models in this method, it is the most useful in studies where the spatial arrangement of lung cells is critical, such as those involving viral tropism [47, 170, 171], bacterial infections [172, 173], fibrogenesis [42], or cell migration [174]. Cryopreserved ex vivo tissue fragments can also function as the basis for the generation of decellularized tissue scaffolds [80] or organoids [175], both applications which will be discussed in later sections.

There are numerous experimental design considerations necessary when directly culturing ex vivo lung tissue fragments. Firstly, even in microexplants, the 3D nature of the tissue fragments makes the development of an air-liquid interface difficult due to the density of the tissue [18]. Reliable long-term (> 7 day) culture methods for ex vivo lung tissue fragments also remain elusive, though perfusion culture has been suggested as potentially extending the lifespan of these tissue fragments in other organs to a length comparable to that of hPCLSs [176]. While lung tissue fragments prepared in this manner retain resident innate and adaptive immune cells [18], the study of any phenomena involving the recruitment of immune cells from other sources to these tissues is challenging, though the incorporation of a hydrogel matrix as a scaffold to support these tissues does make the study of cell migration possible [172, 174, 177, 178]. Additionally, the higher concentration of developed ECM in these models compared to organoids or hPCLSs can obfuscate certain analyses contingent on the detection of molecules such as chemokines, some of which like CXCL8 might ultimately remain bound to the ECM throughout the culture period [18].

Similar to hPCLSs, cryopreservation of ex vivo tissue fragments in a cryobank can address the reproducibility and replicability shortcomings common to ex vivo lung culture models (Fig. 2) [31, 179]. Unlike in hPCLSs, however, the enhanced density of 3D ex vivo lung tissue fragments poses an additional challenge to their cryopreservation, as temperature changes throughout these fragments are not uniform [180] and small molecules are incapable of fully penetrating to the core of larger tissue fragments [18]. While smaller fragments (~ 5 mm³) may be adequately cryopreserved using small-molecule cryoprotectants such as DMSO [48, 77, 84], viability of larger fragments with this method is less favorable and the potential for undesired side-effects exists as noted in hPCLSs [79]. The successful cryopreservation of larger (up to ~ 0.5 cm³) ex vivo lung tissue fragments has been reported with the use of macromolecular cryoprotectants such as trehalose supplemented with surfactants, perfluorocarbons, and protease inhibitors [79], or the commercially available CryoSOFree [18, 174]. A 2014 study compared lung tissue fragments cryopreserved with a homebrew solution containing small molecule and macromolecular cryoprotectants and fresh lung tissue fragments below 0.5 cm³ from the same donors, and found similar protein profiles, cell viability, and tissue structure in healthy donors and those with idiopathic pulmonary fibrosis [79]. Although the exact mechanism for this improved viability remains unclear, based on the principles by which macromolecular cryoprotectants operate, we speculate it is possible that the extracellular control of osmotic pressure exerted by these molecules independent of their ability to reach the innermost cells of the tissue fragments is enough to prevent the formation of intracellular ice crystals deep within the core of the fragments and retain favorable viability.

Ex vivo culture of cryopreserved lung tissue fragments has been used to study viral infection of human lung tissue by SARS-CoV-2 and the migration of lung cancer cells along the alveolar surface [18, 174]. A 2019 study detailed the use of cryopreserved lung tissue fragments as microexplants to analyze the response to SARS-CoV-2 infection and characterized the cellular makeup of the microexplants, the transcription and translation of inflammatory mediators in response to drug treatment in the microexplants, and the viral titer of infected microexplants, determining that the microexplants maintained Type I and Type II alveolar epithelial cells, endothelial cells, T cells, and alveolar macrophages and monocytes throughout the cryopreservation process and demonstrated unimpaired cytokine production in response to infection. This information indicated that the microexplants were an appropriate model for the study of antiviral drugs, and the study further concluded that dexamethasone reduced SARS-CoV-2 viral titer in infected microexplants without affecting the production of inflammatory mediators [18]. A 2024 study analyzing cancer cell migration demonstrated that cells adherent to previously cryopreserved microexplants could also be quantified using intracellular dyes and flow cytometry [174].

Decellularized lung

Fig. 3

Diagram illustrating possible experimental protocols used for experiments involving decellularized lung. Tissues cryopreserved prior to decellularization can still be used in experiments as whole tissues or subjected to alternate decellularization protocols, while cryopreserved decellularized scaffolds are less adaptable for other protocols but a more rigorous model as all thawed scaffolds from a single donor would have undergone the same decellularization process. Figure illustrations were generated in BioRender.

Decellularization of lung lobes or whole lung from cadavers, followed by recellularization through seeding with stem cells from the target donor, was originally conceived as a method for generating viable lung tissue to be used in transplantation with minimal risk of rejection by the patient [181]. Currently, this application of this technique is limited by several factors, including the inability to completely remove all cells or cellular debris generated from the decellularization process without critically damaging the ECM [181, 182], damage potentially caused to collagens and proteoglycans in the ECM during the decellularization process [182, 183], the inability of seeded stem cells to properly form long-term functional vasculature [184], and the lack of innervation being key examples. Despite these shortcomings, the leftover ECM in the generation of decellularized lung scaffolds from animal or human tissue alike provides an excellent substrate onto which either cell lines or cryopreserved suspensions of primary cells can be seeded [185], a strategy not dissimilar to that used in the formation of organoids within the ECM-rich framework of Matrigel [186]. Decellularized human lungs are a particularly desirable basis for tissue engineering studies due to tissue-specific cues in the proteins, glycoproteins, and proteoglycans of the human ECM, collectively known as the matrisome [185, 187, 188]. Gentle decellularization of human lung tissue can thus be used as a source of 3D scaffolds [80, 185, 189] comprising matrisomes representative of disease states in afflictions such as asthma, chronic obstructive pulmonary disease, and idiopathic pulmonary fibrosis [78, 190, 191].

Decellularization of lungs is typically accomplished by perfusion of a detergent designed to separate cells from the lung ECM through either the airways, vasculature, or both [189, 192]. Depending on the decellularization method used, primary cells removed from lung tissue can be cryopreserved for later use [175] and seeded onto an acellular ECM scaffold [80]. For cryopreservation of these primary cells from any organ, small-molecule based cryopreservatives suitable for cell lines are appropriate and represent a well-covered ground in cryopreservation techniques [175, 193]. A central question in the cryopreservation of lung tissue expressly for decellularization is how the characteristics of the ECM are maintained [80]. Cryopreservation of small tissues as a source for small decellularized lung scaffolds has been successfully performed for laboratory experiments [185], and while cryopreservation of decellularized lung scaffolds themselves has not been well explored, cryopreserved decellularized pulmonary heart valves have been deployed clinically in allografts to avoid a deleterious immune response to lingering donor cells [194], suggesting promise for this approach. Decellularization could thus be performed prior to cryopreservation of tissues with the goal of cryopreserving scaffolds, or performed on tissues which have already been cryopreserved, as applicable to research objectives (Fig. 3). Construction of a cryobank with either approach would be useful in the utilization of tissue engineering techniques to study lung diseases in which critical factors may be related to the unique features of the matrisome in various donors [80, 185].

A 2023 study employed the snap-freezing and cryopreservation of lung tissue from healthy donors as a basis for a decellularized lung scaffold, onto which primary AECIIs from chronic obstructive pulmonary disease (COPD) donors were seeded in the study of ECM production and AECII behavior associated with the disease [80]. This study utilized a procedure in which peripheral lung tissue fragments of approximately 8 mm in diameter were flash-frozen and stored in a homebrew cryoprotectant containing 30% v/v glycerol, 30% v/v ethylene glycol, and 0.1 M sodium phosphate buffer, chosen for its ability to preserve the architecture of the lung tissue [185]. Post-thaw, the authors decellularized the tissue fragments before generating precision-cut lung slices of the remaining ECM and seeding them with primary AECIIs retrieved from a digestion of lung tissue of COPD patients. The authors used this model to successfully culture adherent AECIIs from healthy and COPD-afflicted donors for one week or more, and determined that there were few differences between seeded AECIIs from healthy donors and seeded AECIIs of COPD donors, implying the importance of the ECM composition in the behavior of these cells in COPD. The authors noted that the use of an acellular scaffold from COPD donors seeded with AECIIs from either healthy or COPD donors would provide a clearer picture as to the role of the ECM in this disease [80]. It is conceivable that macromolecular cryoprotectants appropriate for the preservation of lung tissue fragments would also allow cryostorage of lung tissue for the purpose of decellularization, therefore, further analysis of the effect of these cryopreservatives on the key signaling features of the ECM would prove useful in the construction of a cryobank for use in acellular scaffold generation, allowing the contributions of the ECM to lung disease to be more easily studied.

Organoids

While originally referring to organ-like structures originating from tumors [195], the term “organoid” has since come to refer to organ-like structures which emerge from the 3D culture of stem cells within a gel matrix containing basement membrane components [69, 186, 196, 197]. The first animal lung organoids were cultured using murine fetal pulmonary cells in 2006 [198], and the first human lung organoids derived from AECIIs were produced in 2013 [199], followed by examples originating from pluripotent stem cells in 2015 [200]. Organoids are utilized in tissue engineering approaches to lung disease modeling [201, 202], and in the study of tissue development [203,204,205] and cancer [206, 207]. Lung organoids initiate from spheroid clusters of stem cells [200, 208], and with the correct signals from culture media or the surrounding matrix, can ultimately undergo a budding process through which structures resembling fetal tracheal [209], bronchial [210], and alveolar features can arise [205], with a thorough complement of intercellular interactions and gene expression critical to the emergence of these features available for analysis throughout [69, 70, 211]. Human lung organoids in particular can be derived from either adult lung stem cells [199], fetal lung stem cells [212], or induced pluripotent stem cells [200], allowing for the study of tissue-like structures in living patients with a minimal requirement for donor tissue [61]. Organoids are the ex vivo lung culture model for which there exists the most variety in cryopreservation strategies, as cryopreservation can be used to store the adult stem cells from which organoids originate [213], the tissues which are used as the source of these stem cells [18, 80], and the organoids themselves [214]. The cryopreservation of lung-sourced stem cells derived from donor tissue, the cryopreservation of lung organoids themselves, and the cryopreservation of small tissue expressly for the purpose of deriving organoids will be the focus of this section. A general discussion of the differentiation techniques and signaling cascades utilized to steer stem cells towards lung growth, as well as cryopreservation techniques particular to stem cells in general, are beyond the scope of this review.

By virtue of being stem cell-derived, organoids are highly customizable models of 3D ex vivo culture, with a great degree of heterogeneity in size and shape amongst cultures originating from the same cluster of cells [29]. Unlike the culture methods discussed above, gene editing with tools such as CRISPR/Cas9 is feasible in lung organoids [70] and has been used in studies of cancer [26, 61, 215], development [216], and idiopathic pulmonary fibrosis [217]. The matrix in which lung organoids reside can also be configured to accept an air-liquid interface [83]. Growth of human organoids requires a minimally invasive amount of tissue from donors, obtainable by biopsy [218], and can be used as a rapidly available, easily expandable, patient-matched testbed for therapeutic interventions in individualized disease conditions such as those found in cancer [61]. Lung cancer is a particularly useful utilization for organoid models of the lung, as the development process observed in cancer organoids can closely mirror patterns related to oncogenesis [219]. The broadly manipulable nature of organoids, however, can also be a limitation, as faithful reproduction of organ structures and disease characteristics requires a fine-tuned media and matrix formulation [70], which can potentially involve specific concentrations of numerous sensitive signal factors to develop organoids with the desired features [69]. Regardless of their origin, organoids are also essentially fetal in nature [205], meaning that they are anatomically simplified structures which do not possess the full complement of cells found in a developed organ, and the cells within them do not fully resemble those of a developed adult organ [220]. Organoids are thus currently incomplete models of adult tissue morphology, pending the discovery of other factors which regulate cellular differentiation [221]. Other models are therefore perhaps a better choice in the study of diseases which require the accurate simulation of relationships within complicated 3D structures such as alveoli.

While it is possible to maintain live biobanks of serially passaged organoids [70], cryopreservation would provide the additional benefits of storage and transport to the use of such systems [214, 222] mitigate the effects of phenotypic drift [223], and can be incorporated into organoid preparation in numerous ways, all of which can be employed to form a cryobank [77, 212, 221]. Adult or fetal lung cells can be freed from fresh ex vivo tissue obtained from various donors and cryopreserved for later use in organoid generation [221]. In this instance, reproducibility is afforded by all thawed stem cells originating from the same donor and undergoing organoid growth in the presence of identical media factors and conditions for each experiment, whereas replicability is enabled through the sourcing of stem cells from multiple donors. Cryopreserved lung tissues themselves can also serve as the source of lung stem cells in this approach, which presents the advantage of allowing factors relating to organoid development to be tested as variables during organoid growth [48, 77, 83]. Alternatively, human lung organoids can be generated from fresh adult or fetal lung tissue and then themselves collected in a cryobank [212, 214, 222]. The benefit of this approach is that all banked organoids for each donor in this instance emerge from the same experimental conditions, providing superior reproducibility for studies not related to development, as this approach prevents the modification of factors associated with the growth stage of organoids (Fig. 4).

Several studies have characterized the effect of cryopreservation on tissue used as the basis for organoid models [48, 77, 212]. A 2017 paper utilizing lung organoids to study development employed a DMSO-based cryoprotectant and slow cooling to freeze human alveolar and bronchial organoid lines derived from either human fetal stem cells or adult lung tissue, in the identification of functional differences in the murine and human transcriptome of the lung distal tip epithelium. The authors cultured the organoids in Matrigel prior to dissolution of the matrix and eventual cryostorage, and reported no critical differences to the expression of SOX genes within the organoids or effects on the ability to incorporate plasmids into cryopreserved organoids post-thaw [212]. A 2021 study utilized DMEM supplemented with 10% DMSO and 20% fetal bovine serum (FBS) to slow-freeze parenchyma fragments approximately 4 mm in diameter prior to thawing and dissociation for organoid generation. The authors noted from single-cell RNA sequencing data that there were no significant changes in cellular identity, function, transcriptional, or epigenetic signatures as a result, but did not derive organoids from their thawed tissues [77]. Another 2021 study employing a similar method, but instead cryopreserving tissue fragments in CryoStor CS10, corroborated these findings before successfully forming organoids from cryopreserved tissue [48]. A 2023 protocol built upon these findings to establish a protocol by which cryopreserved tissues could be used as the basis for infection of lung organoids with SARS-CoV-2 in air-liquid interface culture, highlighting the promise of this method in the study of viral lung disease [83]. As they can be cryopreserved in multiple phases of their development with no apparent deleterious effects on critical genetic factors, organoids thus represent the most flexible ex vivo lung model in terms of cryopreservation protocols. Future analysis on the role of less-toxic cryoprotectants in the cryobanking of organoids, however, would confirm whether or not DMSO has significant effects on the developmental state represented in this model [105].

Fig. 4

Diagram illustrating the cryopreservation of lung organoids in various stages of growth or preparation. Lung tissue fragments can be used to directly derive stem cells or instead cryopreserved and used later as a source of stem cells. These stem cells can either be cryopreserved or used to develop organoids, which can then either be used in experiments or cryopreserved themselves for convenient use at a later date. Figure illustrations were generated in BioRender

Conclusions and future directions

Cryopreservation presents an answer for many of the challenges researchers face in the use of human lung tissue for ex vivo culture [48, 79, 179, 222]. Fresh human lung tissue is scarce, its procurement is difficult, and its viable window for experimental use in parallel with multiple donors is severely limited [66, 73]. Cryopreserved lung tissue is conveniently available, able to be used in various quantities as necessary, and able to be pooled with multiple donors [82,83,84]. These advantages are evident in the above studies, which have demonstrated the use of cryopreserved lung tissue in explant, lung slice, acellular, or organoid form to answer crucial questions about the biology of pulmonary diseases directly in humans. These new approaches to the study of human lung biology represent a movement closer to the goal of attaining greater fidelity to clinical settings in research studies, addressing the commonly cited problems of reproducibility and replicability in the use of fresh human lung tissue. Ex vivo models will continue to become more powerful as cryopreservation protocols evolve, methods to prolong the viability and translational relevance of cultured tissue improve, and strategies for synergistic utilization of each model become apparent.

Methods discovered to effectively cryopreserve and model lung tissue ex vivo could provide a template for the modeling of other organs ex vivo, providing differences in tissue density, structure, and cellular function are appropriately considered as variables affecting the cryopreservation process. Tissue characteristics, including porosity, cell density, and ECM structure, will affect the choice of cryoprotectant and method for adequately infusing frozen tissue with cryopreservative [87]. For smaller, less dense tissues in which diffusion readily occurs, formulations consisting primarily of small-molecule cryoprotectants will perform appropriately [82], while larger or dense sections of tissue will likely require a combination of small molecules and macromolecules to better control osmosis and ice crystal formation [18, 79]. In the case of large tissues specifically, where limited diffusion is an obstacle to fully protecting frozen tissue, new methods, techniques, or equipment designed to quickly perfuse fresh tissue with cryoprotectants and just as quickly flush it upon thawing may also improve viability. For any tissue, but particularly those in which local signaling cues are critical, the development of novel, biologically inert molecules capable of exerting the stabilizing effects necessary for survival of the freezing process without inducing unknown off-target effects on cellular biology will also alleviate existing concerns related to toxicity or undesirable alterations to transcription and translation routines after cryopreservation and thawing, such as occurs in some cell types with DMSO [104, 105]. These improvements in biological fidelity would not only apply to laboratory experiments involving tissue stored in cryobanks, but also to tissue stored for clinical purposes, highlighting effective cryopreservation strategies as a critical area for clinically focused research.

Despite the immense promise for human lung cryobanks, there is much that is yet to be explored about the effects of cryopreservation on human lung tissue. While the effects of specific cryoprotectants on viability, metabolism, transcription, and translation in lung tissue of various sizes has been evaluated [18, 48, 77, 79, 82, 85, 132, 212], there is much left to be determined about the effects of specific cryoprotectant formulations on the general biology of ex vivo human lung tissue. Indeed, there is no consensus method for stockpiling of lung tissue in cryobanks, and if not DMSO, many laboratories currently rely on homemade cryoprotectants designed to be specifically compatible with their research objectives. The specific effects of various cryopreservative formulations on certain disease states of the lung, such as idiopathic pulmonary fibrosis, are therefore elusive and present an obstacle to fully realizing the translational potential of ex vivo models. Studies which directly compare fresh tissue to frozen tissue remain the most effective benchmark in the observation of effects stemming from these individualized circumstances, and as long as validation against fresh tissue is necessary to confirm that cryopreserved tissue is safe to use in the testing of each unique hypothesis in the near term, the widespread benefit from the aforementioned advantages inherent in these ex vivo research methods will be underutilized due to lingering dependence on availability of rare fresh tissue. Large-scale -omics studies on tissues cryopreserved using the current most-studied cryopreservation methods might help to eliminate concerns about the effects of cryoprotectants on ex vivo lung biology [48], and further insight into the effect of cryopreservative formulation on the biology of ex vivo lung tissue is necessary to address this potential remaining hurdle in bridging the gap between the lab and the clinic.

Though cryopreservation is potentially a universally applicable solution to the reproducibility and replicability issues facing ex vivo lung models, it is still critical to consider the limitations inherent in each model when forming hypotheses for study. The aphorism of statistician George Box, “All models are wrong, some are useful,” [224] is especially relevant to the study of pulmonary disease in ex vivo models. However, as confined as these models may be in some areas, the direct use of human tissue holds the promise that these models will prove more useful than existing murine models which precede them. Human ex vivo models will never fully replace cell culture or animal models, but by complementing them, the translational gap can be narrowed and higher success rates for therapeutic development can eventually be achieved. Improvements in cryopreservation techniques and greater understanding of how they affect stored tissues will amplify the rate at which researchers are able to utilize ex vivo models in the study of disease, and subsequently accelerate our understanding of their shortcomings and fuel the discovery of methods which allow us to accommodate them. As the formation of cryobanks becomes commonplace, institutions for which the acquisition of human tissue is all but impossible will also be able to join the study of human ex vivo models, further enlarging the pool of data from which we can refine our knowledge of these models. Cryopreservation is therefore a fundamental tool in the utilization of ex vivo models and will ultimately play a key role in their broad adoption throughout the world of pulmonary biology.