Introduction
Organoids are three-dimensional (3D) in vitro cultures derived from tissue stem cells. They form through stem cell-driven self-organization and can simulate the structure and function of natural organs [1]. Organoid culture originated from the isolation and culture of Lgr5+ intestinal stem cells by Clevers et al. in 2009, who constructed intestinal crypt villous organoids that could produce a continuously expanding and self-organizing epithelial structure similar to the human intestine. This novel in vitro model paved the way for regenerative medicine and gene therapy [2]. However, the study of thyroid morphogenesis and development lacks a stem cell-derived thyroid organoid model system that can recapitulate the differentiation and assembly of thyroid follicular cells into functional thyroid follicles. In 2012, Antonica et al. induced mouse embryonic stem cells (ESCs) to produce thyroid organoids displaying the morphological and functional properties of thyroid follicles for the first time, and demonstrated their ability to metabolize iodide, thereby establishing the applicability of stem cell regenerative medicine in the treatment of hypothyroidism [3]. In addition to its putative application in thyroid regeneration, organoid technology also has potential for use in research on the occurrence and development of thyroid tumours. The thyroid is an organ with a high incidence of malignant tumours, among which papillary thyroid carcinoma (PTC) is the most common. PTC patients generally have a good prognosis, with a 5-year survival rate of 95–97%; however, approximately 20% exhibit tumour recurrence, metastasis, and radioactive iodine-refractory disease (RAIRD) within 10 years [4]. In addition, some patients with total bilateral thyroid resection experience severe side effects during hormone replacement therapy [5]. Anaplastic thyroid cancer (ATC) is a very rare and aggressive form of thyroid carcinoma, with a median survival of only 6 months due to its highly invasive nature [6]. Organoids can be utilized to clarify the mechanisms underlying tumour occurrence and development as well as for the identification of effective intervention targets that can be applied to individualized diagnosis and treatment of patients with refractory and highly invasive thyroid tumours. In addition, thyroid regeneration using organoids may help patients with congenital hypothyroidism or postoperative patients with thyroid cancer, who experience serious side effects after hormone replacement therapy. In this review, we focus on the difficulties associated with thyroid organoid technology, as well as on the current applications of thyroid organoids in regenerative medicine and tumour research. We also discuss the limitations and prospects of this technology.
Organoid technology
Since Clevers et al. [2] first generated intestinal organoids from Lgr5+ intestinal adult stem cells (ASCs), organoid technology has been applied to a variety of human organs. Organoids have unique advantages over traditional two-dimensional (2D) culture and animal models (Fig. 1). First, compared with 2D culture, organoids are complex, 3D, multicellular, self-organizing structures with specific tissue organization and function, which allows for the close connection and frequent interaction of multiple cell types. For instance, interactions among a variety of cell types result in intestinal crypt villous organoids forming a central lumen lined with villous epithelium surrounded by crypt-like domains [2]. Secondly, organoids can be widely amplified during culture while maintaining genomic stability [7]. For instance, the number of nephron progenitor cells (NPCs) is limited, and these cells are difficult to obtain with poor in vitro expansion capacity, such as that seen in 2D culture, which hampers research into kidney development and diseases. Under organoid culture conditions, long-term in vitro expansion of NPCs can be achieved, while maintaining genome stability, molecular homogeneity, and renogenic potential [8]. Thirdly, compared with animal models, organoids better replicate human physiology and have a higher degree of operability and experimental flexibility. For example, the microinjection of Helicobacter pylori into human gastric organoids can successfully recapitulate the typical symptoms of infection by this bacterium, whereas H. pylori infection in mice does not progress to ulceration and cancer as it does in humans [9]. Traditional 2D culture does not have accurate cell–cell and cell–extracellular matrix (ECM) interactions [10], uniform distribution of cells in the culture dish, or a nutritional gradient, which weakens the physiological correlation between cells [11]. Furthermore, differences in cell structure and physiology mean that animal models cannot accurately reflect the regulatory mechanisms and cell interactions occurring in human tissues [12, 13]. Combined, these observations indicate that organoids are more suitable for the construction of biobanks and high-throughput screening, and have gradually become a representative tool for studying living organs in vitro.
Organoids can be derived from 2 types of stem cells: pluripotent stem cells (PSCs), which include ESCs and induced PSCs (iPSCs); and ASCs, which are resident stem cells specific to differentiated tissues [14]. PSCs can form all tissues of the body and will spontaneously differentiate in vivo into a disorganized mass of differentiated tissue, called a teratoma, through the manipulation of factors that control embryonic organogenesis [15]. Methods have been developed to guide the stepwise differentiation of PSCs into embryonic germ layer-restricted organoids, organ-specific organoids, and even specific cell types such as hepatocytes, neurons, and cardiomyocytes [15]. The induction of PSCs for organoid culture requires their differentiation into the 3 germ layers (endoderm, mesoderm, and ectoderm) under specific differentiation signals, followed by multiple induction steps toward a fully differentiated structure, which requires a specific growth factor mixture at each step [16]. Gastric organogenesis requires the addition of activin to generate endothelial progenitor cells from PSCs, followed by the addition of Wnt3a and fibroblast growth factor 4 (FGF4) activators and bone morphogenetic protein (BMP) inhibitors to differentiate these cells into foregut cells, and then retinoic acid to yield posterior foregut cells [17]. Finally, high concentrations of epidermal growth factor (EGF) are applied to generate gastric organoids [17]. The whole process of organogenesis in the human embryonic period is simulated in PSC-derived organoids, rendering them representative of living organs for in vitro studies. ASCs are found in differentiated tissues and have the capacity for division and self-renewal as well as for tissue regeneration [18]. Generating organoids from ASCs does not require transcription factor transduction; organoids can be directly induced, making them physiologically compatible with the normal tissues of the host [19, 20]. Stem cells must grow in a specialized microenvironment; namely, a stem cell niche, composed of fibroblasts, immune cells, endothelial cells, perivascular cells or their precursors, ECM, cytokines, and growth factors [21]. Additionally, the growth of stem cells must be regulated by providing cell-to-cell contact and secretory factors [22]. Thyroid organoid culture involves an artificially created microenvironment that allows for the formation of stem cell niches by providing the corresponding cytokines and ECM. The ECM can support cell proliferation, enable cell adherence, and allow nutrient and growth factor diffusion [23]. Stem cells must be in strict contact with ECM components, such as collagen, laminin, and fibronectin, which are important regulators of stem cell behaviour, migration, and differentiation [24]. The cytokines required for thyroid organoid generation include Wnt-3a, R-spondin-1, Noggin (a BMP antagonist), EGF, FGF, thyroid-stimulating hormone (TSH), and Y-27632, a specific ROCK protein inhibitor [25]. Various classes of cytokines play different roles during organogenesis. Wnt and R-spondin-1 play a key role in the self-renewal of multiple types of adult stem or progenitor cells [26]. When Wnt proteins bind to members of the Frizzled family of receptors, the canonical Wnt pathway is activated [27], and this signalling pathway can be further enhanced through the activity of the R-spondin protein [28]; jointly, these factors drive the differentiation of stem cells in culture. Noggin interferes with the binding of BMP to its receptor, thereby antagonizing the function of cytokines that limit stem cell proliferation [29]. EGF is a growth factor for epithelial tissues, and its binding to EGF receptors can induce hyperplasic changes [30]. Saito et al. exposed thyroid cells to different concentrations of TSH for thyroid organoid culture and found that TSH could significantly promote organoid formation in a concentration-dependent manner [25]. Y-27632 has been reported to inhibit premature apoptosis in organoids, enhance the ability of cells to form spheroids, and improve the ability of stem cells to survive and proliferate in vitro [31]. To date, organoid technology has been applied to the in vitro culture of a variety of human organs, including the establishment of an organoid culture system for thyroid-related research.
Applications of thyroid organoids
Regenerative medicine
Since its establishment, organoid culture technology has experienced rapid development and shown strong application value in the field of regenerative medicine, including by providing a new graft source. Many types of organ models have been applied to regenerative medicine to date, including liver, intestine, pancreas, kidney, skin, and, importantly, thyroid models [32–37]. Longmire et al. reported the directed differentiation of thyroid cells from mouse ESCs in 2012. To achieve this, the authors inhibited bone morphogenetic protein (BMP) and transforming growth factor-beta (TGF-b) in ESC-derived endodermal cells, yielding NK2 homeobox 1 (Nkx2.1)+ endothelial progenitor cells. The subsequent combinatorial induction of BMP and FGF signalling in these Nkx2.1+ cells promoted thyroid lineage specification [37]. This report laid the foundation for ensuing studies on the construction of thyroid organoids. Later, Antonica et al. reported that mouse ESCs could be directed to differentiate into thyroid cells through the transient overexpression of the transcription factors Nkx2.1 and paired-box gene 8 (PAX8) under doxycycline induction, and that these thyroid cells formed thyroid organoids with 3D follicular structures under thyrotropin treatment [38]. This study confirmed for the first time that mouse ESC-derived thyroid follicular cells could form thyroid organoids in vitro [38]. Moreover, based on the results obtained by Longmire et al. with differentiated thyroid cells [37], Kurmann et al. differentiated mouse PSC-derived endoderm cells into thyroid follicular organoids [39]. In the 2 studies [38, 39], the constructed thyroid organoids were transplanted into mice with hypothyroidism induced by radioiodine ablation of their thyroid tissue, leading to the generation of follicular-like tissue that secreted thyroid hormone and the restoration of the level of thyroid hormone to normal; moreover, the regenerated thyroid tissue could be regulated by TSH. These studies suggest that PSC-derived thyroid organoids have a strong regenerative capacity in vivo and can compensate for the hypothyroidism caused by the lack of thyroid tissue in situ.
In addition to PSCs, organoids can also be generated from ASCs derived from differentiated tissues (Fig. 2A). Lan et al. isolated a side population of cells that can express adenosine triphosphate binding box transporter G2 (ABCG2) from differentiated thyroid cells [40]. These cells had a high nucleus/cytoplasm ratio; they had strong expression of the stem cell markers ABCG2 and octamer binding transcription factor 4 (Oct4); and they could proliferate, form spheres, and differentiate into thyroid cells under the action of TSH [40]. Oct4 is the major transcription factor determining the fate of ESCs as well as that of some ASCs [41]. Although no specific marker of thyroid ASCs has been identified, the above-described research suggests that ASCs may exist in the thyroid. Thyroid organs can be directly cultured from thyroid tissue without undergoing cell reprogramming. Saito et al. cultured surgically excised mouse thyroid tissue fragments after mechanical and enzymatic digestion and established mouse thyroid organoids with normal thyroid function under the action of TSH [25]. These functions included thyroglobulin synthesis, iodine uptake, and the production and release of thyroid hormone [25]. The authors transplanted cultured thyroid organoids into hypothyroic mice, resulting in the generation of thyroid follicle-like tissue capable of iodide uptake [25]. In addition to their normal thyroic function, the thyroid organoids stably expressed thyroid immune markers during both in vitro and in vivo culture [25]. Ogundipe et al. used mouse and human thyroid tissue for organoid culture and found that the thyroid markers NKX2.1, PAX8, sodium iodide symporter (NIS), thyroglobulin (TG), and thyroxine (T4) were expressed in constructed organoids and remained stable in passage [42]. The transplantation of organoid-derived cells under the renal capsule of hypothyroidic mice led to the formation of thyroid follicular structures that expressed NKX2.1, TG, and T4 [42]. The above results show that thyroid organoids derived from mature thyroid tissue also have a strong regenerative ability. Furthermore, regenerated thyroid tissue can function normally in vivo and restore thyroid hormone levels in hypothyroic mice.
Congenital hypothyroidism is the most common congenital endocrine disease in humans, caused by a dysfunctional thyroid gland (15%) or dysplasia (85%) [43]. The incidence rate of neonatal hypothyroidism is approximately 1/2000 [43]. Patients who have undergone bilateral thyroidectomy for thyroid cancer treatment require lifelong hormone replacement therapy owing to the absence of thyroid hormone postoperatively. However, some patients experience serious side effects from hormone replacement therapy after total bilateral thyroidectomy [5]. Organoids consisting of PSC-derived thyroid cells or primary thyroid gland-derived cells can be transplanted into the body and form follicular-like tissue with normal thyroid function, and they may represent an effective means of treating hypothyroidism. Yang et al. [44] developed a method for microencapsulating porcine thyroid cells in alginate-poly-L-ornithine-alginate microcapsules as a thyroid hormone replacement approach. The semipermeable microcapsule membrane allowed the diffusion of thyroid hormones, TSH, nutrients, oxygen, electrolytes, and wastes while blocking that of immunoglobulins, antibodies, and host cells that mediate the immune response [44]. Under TSH stimulation, the encapsulated porcine thyroid cells formed 3D follicular spheres in the inner core of the liquefied alginate microcapsules that released thyroid hormone [44]. In the future, microencapsulation technology may be combined with organoids for in vivo transplantation, thereby allowing the transplant recipient to retain normal thyroid function without immunosuppression. This provides a basis for the application of thyroid organoids in the field of regenerative medicine.
Organoids and thyroid tumours
Tumour cells have unique mutational and epigenetic features, exhibiting heterogeneity in gene expression, metabolism, proliferation, and metastatic potential [45]. Organoid cultures can recapitulate this heterogeneity in vitro while maintaining genomic stability during passaging. Patient-derived organoids (PDOs) can be used to enlarge small tumour samples and enable the analysis of cancer at any stage in culture, greatly expanding the types of tumour samples that can be propagated and studied in the laboratory [46]. Studies on thyroid cancer organoids are scarce and have mainly focused on surgically removed thyroid cancer tissues. Sondorp et al. generated patient-derived organoids from 13 patients with PTC and 3 with RAIRD, and found that the marker expression of PTC and RAIRD organoids was more like that of the tumour tissue of origin [47]. Furthermore, NIS was not expressed in RAIRD organoids, which may explain why postoperative iodine 131 treatment in patients with PTC is ineffective, leading to patient relapse [47]. DNA sequencing of thyroid cancer organoids combined with drug screening can be applied to individualized treatment in thyroid cancer patients. Chen et al. [48] created organoid cultures of tumour tissues from PTC patients and carried out whole-exome sequencing (WES) in PTC organoids and parental tumour tissues to identify significant mutant genes. The results showed that PTC-derived organoids harboured the same mutated genes as the parental tumours and remained unchanged in long-term in vitro culture [48]. The authors then applied antitumour drugs to evaluate the drug sensitivity of PTC organoids, and found that there was a correlation between the sensitivity of PTC organoids and their mutation spectrum [48]. Although most of the PTC-derived organoids with BRAF gene mutation were sensitive to the BRAFV600E inhibitors vemurafenib and dabrafenib, a few were resistant to these drugs, despite harbouring BRAF gene mutation [48]. This observation highlights the value of the combination of WES and organoid drug screening for identifying tumours that are insensitive to targeted drugs [48]. Gene editing can also be applied to thyroid cancer organoids. For instance, Saito et al. introduced the cDNA of NRASQ61R, the main oncogenic driver of thyroid cancer, into thyroid organoids established from p53 knockout (KO) mice, and then transplanted the resulting NRASQ61R/p53KO thyroid organoids into mice to generate poorly differentiated thyroid cancer (PDTC) [25]. This study not only enriches the thyroid cancer biobank but also reduces the ethical risk and provides a reliable model for thyroid cancer research.
The above-mentioned studies indicate that thyroid carcinoid organoids cultured in vitro retain the mutational profile and histopathological characteristics of the primary tumour [47,48]. Accordingly, they can represent living tumours, thus allowing the establishment of cancer models in vitro for relevant studies (Fig. 2B). Additionally, thyroid organoids can produce PDTCs in mice through gene editing and transplantation [25], indicating that tumour organoids can be xenografted and that tumour samples can be enriched for laboratory-based analysis. Finally, thyroid cancer organoids, combined with DNA sequencing, can be used to predict the sensitivity of drugs in the human body [48], suggesting their potential as an effective tool for the individualized treatment of patients with thyroid tumours. Several studies have been undertaken on organoid culture and drug screening for gastrointestinal, liver, lung, and breast tumours [49–52]. For example, Vlachogiannis et al. tested the effects of anticancer agents on ex vivo PDOs, and compared the responses of patients in clinical with PDO-based mouse xenograft models to anticancer agents, they concluded that PDOs can recapitulate drug response of patients and have the potential to be executed in personalized medicine programs [49]. Research on thyroid cancer organoids is currently mainly focused on PTC, while studies on poorly differentiated and undifferentiated thyroid cancer organoids are lacking. More research on this kind of thyroid cancer and its organoids is needed to better understand thyroid cancer pathogenesis and identify effective intervention targets. Organoids will likely become a new tool for physiological and pathological research of thyroid cancer.
Discussion
Unlike traditional 2D culture and animal models, organoid culture technology can lead to the establishment of an ideal model that highly recapitulates the structure, function, and immunophenotype of the thyroid gland and thyroid cancer in vitro. Accordingly, organoids can play an important role in tumour research and regenerative medicine. However, the technique is not fully mature, and many limitations in thyroid organoid culture remain. First, there is no uniform criterion regarding the substances required for the culture of thyroid organoids, and the factors used for culture vary slightly in a study-dependent manner. The organoid formation rate is also relatively low. Sondorp et al. reported that the formation efficiency of PTC-derived organoids is stable at about 7%, while that of RAIRD organoids is approximately 5% [47]. Secondly, organoid formation should ideally mimic the development of the whole organ; however, differences in the timing of the action of the various growth factors involved and the concentrations of these growth factors make it difficult to produce organoids that perfectly match the relevant organs in vivo [53]. There are also minor differences in gene expression patterns between established organoids and the corresponding tissues in vivo [54]. Finally, generated organoids lack vascularization [55]. Increasing organoid volume is accompanied by increasing levels of hypoxia and metabolic waste, which kills the cells; consequently, the organoids currently cultured in vitro cannot fully represent the real organs. In conclusion, the culture and application of thyroid organoids require further development. Current research is mainly concentrated on thyroid- and PTC-derived organoids, and studies on organoids relating to highly invasive, high-mortality, poorly differentiated, and undifferentiated cancers are lacking. Future research on thyroid organoids must focus on improving culture systems and developing more comprehensive in vitro models, thereby promoting the development of precision medicine.
Funding
This study was supported by the Excellent Youth Project of the Fourth Affiliated Hospital of Harbin Medical University (Grant No. HYDSYYXQN202006, principal investigator Jinlu Zhao), the youth project of Science and Technology Innovation project of Heilongjiang Academy of traditional Chinese Medicine (Grant No. ZHY19-080, principal investigator Jinlu Zhao).
