Saliva is one of the most indispensable substances for maintaining oropharyngeal health. The various components of saliva, including anti-microbial peptides, mucus, and amylase, can protect the soft and hard oral tissues from invaders, promote tissue recovery, and help digestion [1-4].

Xerostomia, also known as dry mouth syndrome, is a pathologic condition of hyposalivation due to salivary gland (SG) dysfunction. This condition causes great discomfort in speaking, chewing, swallowing, and wearing dentures, thereby significantly lowering the quality of life of affected patients. Several autoimmune diseases (e.g., Sjögren syndrome [SS]) and medical interventions (e.g., radiotherapy for head and neck cancer) often lead to xerostomia; however, its etiology has not been fully elucidated [5-7]. Up to 10% of the general population experiences dry mouth, while the senior population seems to be more susceptible to xerostomia. Histologically, the number of acinar cells decreases while the amount of fat and fibrous tissue increases in the SG with aging. The prevalence of the disease is much higher in women than in men. Moreover, according to recent reports, dry mouth syndrome is frequently diagnosed in menopausal women, implying a correlation between salivary function and sex hormones, such as estrogen [8,9]. It is well known that the decrease in estrogen levels after menopause correlates with various complications such as facial flushing, headache, skin changes, osteoporosis, and xerostomia. Importantly, more than 25% of elderly women complain of discomfort due to xerostomia after menopause [10-12]. However, there is no cure available and the current treatment of xerostomia is limited to non-specific allopathic medicine with several side effects. Therefore, safe and effective therapeutic approaches should be developed.

Recent advances in 3-dimensional (3D) culture technology allow pluripotent and multipotent stem cells to exhibit their remarkable self-organizing properties, and the resulting organoids reflect key structural and functional properties of mature organs such as the kidney, lung and gut. Organoid technology can therefore be used to model human organ development “in a dish” with properties that closely resemble the real organ in its structure and function [13,14]. In addition, since organoids preserve tissue heterogeneity compared to single-cell cultures, various pathologic conditions and disease features can be simulated more efficiently than conventional approaches [15]. In addition, mature organoids can be used as transplantable tissue in the field of cell-based regenerative medicine [16,17]. In these aspects, understanding and treating xerostomia will necessitate the optimization of the culture method for salivary gland organoids (SGOs) and utilization of SGOs as a disease-modeling system for xerostomia. Therefore, this review summarized the cutting-edge technologies for the generation and maturation of organoids using SG stem cells, as well as their application for understanding the development of SG and modeling SG hypofunction to treat the disease.

In the field of developmental biology, an ex vivo organotypic slice culture method has been widely applied to study SG organogenesis [18-21]. In this format, SG epithelial tissue, mesenchyme, and extracellular matrix (ECM) reconstitute 3D structure that reflects several aspects of developmental hallmarks including branching morphogenesis and acinar cell maturation as observed in vivo. With the combination of various chemical treatments and mesenchyme-free culture conditions, organotypic slice culture is useful to define essential signaling pathways and regulatory niche factors contributing to SG development; however, long-term maintenance and subculture, as well as experimental manipulation, are excessively limited because this technique requires naïve tissue slices. Organotypic slice culture is also generally optimized with embryonic- and neonate-derived samples, while culturing adult tissue is still challenging. For these reasons, SGOs can be an excellent alternative to solve the current limitations of the organotypic 3D culture method [22]. Several attempts to culture salispheres (SSp) and SGOs have been conducted since the mid-2000s. SGOs can be derived from adult stem/progenitor cells (ASPCs), embryonic SG germ tissue, and pluripotent stem cells (PSCs) [23].

In general, the self-renewal of SGSCs should be maintained during long-term passaging and the organoid forming efficiency, doubling time and the proliferative capacity of SGOs are investigated to evaluate the SGSC population. In addition, cellular complexity and structural similarity to the naïve organ are important to determine the organoid quality, especially for differentiated organoids. Therefore, matured SGOs should contain both polarity-organized ductal and functional acinar cells, as observed in the epithelial region of SGs. Various functional tests, including a calcium (Ca²⁺) flux assay, amylase secretion assay, and lumen swelling assay, followed by cholinergic stimulation, have been applied to assess the functionality of mature SGOs.

Several attempts have been conducted to fulfill these criteria (Table 2) [29,35-43]. Specifically, the regulation of signaling pathways associated with SG organogenesis is one of the common strategies to improve basic culture conditions. For instance, sustained activation of the Wnt-β-catenin axis significantly consolidates the self-renewal and organoid-forming efficiency of SGSCs [29]. Treatment of the Rho kinase inhibitor, Y27632, can increase the population of c-kit⁺ and CK5⁺ SGSC-like cells in SGOs under expansion conditions, while it accelerates the differentiation of pro-acinar Mist⁺ cells into aquaporin 5-positive (AQP5⁺) acinar cells during the Fgf2-mediated maturation process [35]. In addition, chemical inhibition of activin receptor-like kinase signaling or treatment of morphogenic proteins such as the Fgf family can induce the development of acini-containing budding structure in SGOs [30,36,37]. In addition to the branching morphogenesis, Kim et al. [38] focused on recapitulating the lumen formation process during the maturation of SGOs. According to that study, the Fgf2/Fgf10 combination initiates the budding and branching elongation process, and the addition of retinoic acid (RA) enhances lumen formation within this duct-like compartment. The authors further revealed that activation of RA signaling upregulates the vasoactive intestinal protein pathway, one of the essential mediators in ductal growth and contiguous lumen formation during the development, and directly stimulates CK7⁺ luminal cell differentiation.

Early studies have documented the importance of the ECM in regulating the morphogenesis of developing SGs [44]. Thus, the ECM and ECM-mimicking substances in SGO culture not only function as a physical scaffold, but also provide biological interactions facilitating the generation and maturation of SGOs. Matrigel is the most commonly used hydrogel at present, but potential benefits of other natural- and synthetic biomaterials have been explored for organoid culture [38,44]. The natural ECM obtained from naïve SG tissue can be utilized as hydrogel or scaffold for SG-derived cell culture after decellularization [39,45]. Srinivasan et al. [40] fabricated hyaluronic acid (HA) hydrogels with the basement membrane peptides from laminin and perlecan to provide in vivo-like enriched cues and applied them in SGO culture. They found that SGOs embedded in peptide-modified HA hydrogel showed enhanced progenitor potential with increased differentiation capacity towards acinar cell lineage compared to those cultured with naïve HA. In another study, researchers generated various concentrations of chitosan in the form of soluble or coating material and investigated whether chitosan-based biomaterials would be applicable to SGO culture [41]. It was noted that both types of chitosan affected the morphology of SGOs significantly by suppressing spontaneous cavitation in a dose-dependent manner. Chitosan seems to affect the cellular adhesion and polarity establishment during the SGO organization process, contributing to the maintenance of SGOs in the primitive state.

Meanwhile, bioengineering techniques have been applied to aid the 3D organization of dissociated SG cells. Lim and colleagues have studied the benefits of hydrogel-micropatterned nanofibrous microwells in the self-assembly of human PG-derived epithelial cells (hPECs). They developed polyethylene glycol hydrogel-micropatterned polycaprolactone nanofibrous microwells to culture hPECs and found that microwell-cultured SG aggregates exhibit a higher level of acini phenotypic markers such as AQP5 and α-amylase with enhanced tight junction and structural protein expression [42]. This system is applicable to culture clonally expanded SGSCs, in which it contributes to the 3D organization of SGSCs followed by the generation of functional SGOs [43]. Considering the unstable quality due to batch-to-batch variation and xenogenic features of Matrigel, well-defined synthetic biomaterials would be an excellent alternative to Matrigel for the therapeutic application of SGOs in regenerative medicine.

Increasingly many attempts have recently been conducted to apply organoid technology as an alternative for the conventional modeling systems represented by cell line culture and animal models in the field of biomedical research. Since the derivation of SGOs from SGSCs recapitulates several key aspects of the developmental process, including self-organization, budding formation, and duct elongation, SGOs have emerged as a model system to understand SG development and homeostasis. Several studies have been conducted to elucidate the putative SGSC population by evaluating the organoid forming capacity of SGSC candidates expressing single- or multiple SC markers as described above. Researchers also take advantage of SGOs to verify hypotheses extrapolated from organotypic cell culture and animal studies. For instance, the importance of the fibroblast growth factor family members in acini cell development and the role of the RA signaling pathway on lumen formation have been confirmed in the process of SGO differentiation. Orhon et al. [46] indicated the role of autophagy in regulating the self-renewal and regenerative response of SGSCs upon injury signals using SGOs. They treated SGOs with the autophagy blockers bafilomycin A1 and chloroquine and found that the organoid-forming capacity of SGSCs dropped significantly due to the impaired autophagy. This phenomenon was reproduced in an Atg5-deficient SGO line established from Atg5 knockout model mice. Of note, when the endogenous autophagy flux was activated in SGSCs by Tat-beclin 1 peptide treatment, the self-renewal efficiency of SGOs improved without affecting differentiation capacity. Based on these findings, the authors revealed that basal autophagy levels in SGSCs might be regulated depending on their status (quiescence or active) and suggested the autophagy pathway as a key modulator in homeostasis and regeneration of slowly self-renewing organs.

SGOs possess excellent potential for modeling SG-associated diseases and elucidating their pathological mechanisms. Since xerostomia is the most common condition associated with the SG hypofunction, researchers have tried to recapitulate this condition in SGOs from different perspectives. A recent study demonstrated the modeling of radiotherapy-associated xerostomia by evaluating the radiation response of SGOs [47]. Maimets et al. [48] compared the characteristics of young- and old mouse-derived SGOs mainly focusing on the self-renewal of SGSCs. Although aged SMGs contain a higher number of CD24^hi/CD29^hi cells than young SMGs, SGSCs from old mice showed a lower organoid-forming capacity than that of young cells, indicating that stemness and self-renewal ability of SGSCs might be declined with aging. Meanwhile, hyposalivation is one of the major symptoms of the autoimmune disorder SS, although the causal relationship remains unclear. To explore the characteristics of SGSCs in primary SS, SGOs were established from PG biopsy samples of SS-affected patients and compared with healthy SGOs [49]. It was noted that both SGO generation efficiency and the yield of EpCAM^high SGSCs were significantly lower in SS samples compared to the control. RNA sequencing analysis of SGOs revealed that gene sets associated with cell cycle progression, cell division, and senescence phenotype were significantly upregulated in SS-SGOs; indeed, the average telomere length in control SGSCs was greater than that of SS-derived SGSCs, implying that SS-SGSCs might be subjected to replicative senescence-induced injuries. The authors further found that a combination of pro-inflammatory cytokines (interferon-α, tumor necrosis factor, and interleukin-6) induced an accelerated proliferation in normal SGOs, leading to premature aging. Since the organoid system is more physiologically relevant to naïve tissue than to established cell lines, SGOs can be applied to study cancer biology and Takada et al. [50] reported the first approach utilizing SGOs as a salivary adenoid cystic carcinoma (ACC) model. In this study, several ACC organoid lines were generated then orthotopically transplanted into NGS mice to obtain a patient-derived tumor xenograft. Interestingly, cancer SGOs maintain the genetic and histological phenotypes of their original tumor tissues and their reactivity to chemotherapeutic agents can be evaluated by in vitro drug sensitivity assays. Therefore, these studies suggest the value of SGOs not only for investigating the mechanism of uncurable SG-associated disorders, but also for establishing a novel therapeutic strategy.

Finally, organoids are regarded as promising tools in the field of cell therapeutics and regenerative medicine [22,51]. Since tissue-specific stem cells are stably maintained and expanded within organoids, continuously cultured SGOs can provide SGSCs with high potential, and transplantation of these SGO-derived stem cells can improve radiation- or ligation-mediated SG damage in vivo [24,25,28,31]. Furthermore, not only isolated SGSCs but also SGOs themselves are capable of tissue restoration. Several attempts have been conducted to test the therapeutic impact of reconstituted SGO-mesenchyme-biomaterial complex [4,32,52]. After the orthotopic engraftment of bioengineered SG structures in vivo, they were integrated into the recipient tissue properly with a differentiated structure. Notably, transplanted artificial SGs could produce saliva, leading to the amelioration of dry mouth symptoms of the SG-deficient mice model, demonstrating the possibility of organ replacement therapy through the combination of biotechnology and organoid technology.

Despite the remarkable achievements in the establishment of SGOs, current technology has several limitations. Compared to other well-defined epithelial organoid systems, culture conditions for SGOs have not been optimized yet [14]. Of note, although the present expansion media can support SGSC enrichment in vitro, the organoid differentiation protocol for the induction of both duct and acini compartments in SGOs still needs to be improved. Naïve SG tissue consists of an organized ductal system and produces multiple secretory proteins through the cooperative action of exocrine acini cells and contractile myoepithelial cells [53]. Furthermore, humans have 3 major SGs—the PG, SMG, and sublingual gland—with multiple minor SGs, and each gland has distinctive histological characteristics and secretory profiles. Similar to other exocrine organs, the production of saliva is tightly regulated by the autonomous nervous system via interactive crosstalk between receptor signaling [54-56]. In these aspects, Lee et al. [32] successfully generated region-specific SGOs with unique cellular composition, although their secretory functions were not determined. Therefore, it would be important to check the expression pattern of receptors (cholinergic and purinergic receptors) and to test the reactivity of differentiated SGOs to various stimuli for functional evaluation. In addition, current organoids mainly consist of epithelium-derived cells and ECM; thus, the establishment of a co-culture system with other components including mesenchyme, vascular structure, immune cells, and microbiota would provide an advanced SG platform to study the mechanisms underlying SG organogenesis, regeneration, and disease development. Taken together, improved SGOs that overcome current limitations might critically contribute to the development of stem cell-based therapeutics for disease modeling and regeneration.