Chronic H. pylori infection not only results in severe acid-gland atrophy and drives SPEM development (39, 40), but also leverages inflammation-related metaplastic changes to expand its niche through spatial colonization. In this process, the initial event is the colonization of H. pylori, altering the distribution of gastric microbiota and the abundance of proteins. By influencing microbial metabolism, H. pylori secretes effector proteins and toxins, including CagA and VacA (41, 42), disrupting gastric cell junctions and the polarization of apical-basal cells. Further activation of pro-inflammatory and carcinogenic signaling pathways leads to impaired integrity of gastric epithelium, disrupted cell differentiation and disturbed self-renewal, ultimately resulting in inflammatory response (43). Chronic damage to the gastric mucosa can induce the recruitment, homing, and proliferation of bone marrow mesenchymal stem cells (BMSCs) under inflammatory stimulation, further becoming a prerequisite for dysplasia and cancer (44, 45). Inflammation-mediated immune cell infiltration is an important mechanism and broad background for the occurrence of SPEM (46).

SPEM-induced mucosal remodeling creates more suitable conditions for H. pylori colonization. H. pylori mainly colonizes by binding its two adhesins Bab and Sab to glycosylation receptors Lewis B (Leb) and sialylated Lewis X (sLex) in host epithelial cells. Compared to intestinal metaplasia, SPEM can specifically enhance sLex expression, which to some extent determines the colonization and diffusion of H. pylori towards the proximal part of the gastric body and deeper glandular layers. Following the onset of SPEM, H. pylori can indirectly promote the spread of SPEM throughout the stomach by inducing this metaplastic repair response via chronic inflammation, or directly promote the spread of SPEM throughout the stomach by secreting toxins such as CagA that affect epithelial differentiation (47). In low gastric acid conditions, H. pylori’s tendency to target SPEM increases, further enhancing its colonization, adhesion, and invasion abilities, resulting in persistent inflammation and the further establishment of a tumor microenvironment.

SPEM caused by H. pylori infection presents as chronic lesions, most of which are irreversible. H. pylori utilizes its motility, chemotaxis, toxin production, and other mechanisms to adapt to the acidic conditions of the gastric lumen (48), thereby evading immune recognition and causing persistent inflammation. Continuous damage and repair induce abnormal cell proliferation, directly increasing DNA replication pressure at the genetic level (49), promoting double strand breaks (DSBs) (50), and overloading the DNA damage response (DDR) pathway (51). This response is normally triggered by signalling damage, repairs cells in an error-free manner and causes apoptosis and senescence. This prevents malignant progression in the presence of incorrectly repaired DNA, which can lead to tumour formation when the DDR signalling pathway is inefficient due to overload and DNA repair is compromised. Eradication of H. pylori reverses DDR activation, but not cellular senescence. An increased number of senescent cells in the area of metaplasia acts as a carcinogen and promotes disease progression further (52). This could explain why H. pylori eradication is ineffective in preventing gastric cancer in the presence of precancerous lesions such as IM (53, 54).

Current animal models of SPEM are mostly induced and established using high doses of drugs such as tamoxifen, DMP-777, and L-635 (55, 56), which directly cause an acute decrease in the number of gastric lining cells and the development of SPEM within 3–14 days through drug toxicity effects. Among them, tamoxifen belongs to the selective estrogen receptor modulators, which are mostly used in breast cancer chemotherapy or as inducers involved in gene editing. A single dose of tamoxifen ≥ 3 mg/20 g body weight given to normal mice for 3 days resulted in apoptosis of more than 90% of the parietal cells and metaplasia of the chief cells (57). Comparative studies utilising wild-type and gastrin-deficient animal models have demonstrated that SPEM-induced oxyntic gland atrophy occurs through gastrin-independent mechanisms (56). Notably, findings from Manning et al. (58) confirmed this observation, revealing that tamoxifen acts as a protonophore by disrupting the proton gradient across the acid-secretory membrane of parietal cells. This disruption to acid-base homeostasis induces parietal cell apoptosis, a process thought to involve muscarinic receptor-dependent pathways. DMP-777 also exhibits protonophore activity but does not affect the activity of H⁺/K⁺-ATPase (59). Additionally, as a neutrophil elastase inhibitor, DMP-777 induces the development of SPEM by reducing the degradation of extracellular matrix components such as elastin and collagen. This preservation of matrix integrity thereby prevents the onset of inflammation. L-635 is a prototype carrier analog of DMP-777. Due to its lack of elastase-inhibitory activity, treatment of mice with L-635 can establish a parietal cell depletion model accompanied by acute inflammatory responses. This intervention further promotes the transition of the SPEM lineage from relatively inert metaplasia to a proliferative metaplastic phenotype, ultimately leading to the development of dysplasia (6). The feature is similar to the chronic SPEM model established in cats infected with H. pylori.

In addition to H. pylori infection and drug-induced damage, bile reflux and immune dysregulation also contribute to the development of SPEM (60). During bile reflux, bile phosphatidylcholine is hydrolyzed into lysophosphatidylcholine by phospholipase A2, resulting in degradation of the phospholipid layer in gastric mucosal epithelial cells. This disruption of the gastric mucosal barrier facilitates the back-diffusion of H⁺ into the submucosa (61), thereby altering the microenvironment critical for parietal cell survival. Furthermore, bile acids inhibit nitric oxide synthase (NOS) activity and impair cellular H⁺/Na⁺ exchange (62), while modifying DNA methylation patterns (63, 64). These cumulative effects induce DNA damage, apoptosis, and mutagenesis in gastric mucosal cells, ultimately driving mucosal atrophy and metaplasia. Immune-mediated SPEM predominantly emerges during the pathological progression of autoimmune gastritis. Autoimmune gastritis is an autoimmune disorder characterized by autoimmune attacks against parietal cells and intrinsic factor, with SPEM representing one of the most frequent metaplasia subtypes observed in this condition (65). CD4⁺ T cells trigger an autoimmune response targeting the H⁺/K⁺-ATPase on the secretory canalicular membranes of parietal cells. This immune-mediated destruction leads to parietal cell loss and atrophy of the gastric glands, representing a core mechanism of autoimmune-related parietal cell depletion. A more specific mechanism may involve Fas ligand-induced activation of the extrinsic apoptosis pathway (66, 67) ( Figure 1 ).

Chief cells are functional cells located at the base of gastric fundic glands, responsible for secreting pepsinogen granules. They express stem cell molecular markers muscle intestine and stomach expression 1 (Mist1), tumor necrosis factor receptor superfamily member 19 (Troy), and leucine-rich repeat-containing G-protein coupled receptor 5 (Lgr5), and can act as reserve stem cells in vivo ( 68, 69). These cells normally maintain homeostasis of themselves and other gastric gland cells through slow turnover (70). Once injured, chief cells change their transcriptional signature to increase repair of damaged tissue (71). A reduction in parietal and chief cell numbers is observed in most SPEM animal models, with the highly plastic chief cells being recognized as the cellular origin of metaplasia (6). Caldwell et al. (72) developed novel chief cell-specific GIF-trTA allele mouse models for lineage tracing. In this study, GFP tracing markers were initially detected exclusively in the basal gland regions housing chief cells, showing co-expression only with chief cell markers (GIF) and SPEM markers (CD44v9, TFF2, GSII). This co-expression pattern significantly diminished after 12 months, while GIF-negative GFP-labeled cells became scattered in gastric corpus glands, correlating with Ki67 (proliferative cells), UEAI (surface cells), GSII (mucous neck cells), and H/K-ATPase (parietal cells) positive populations. These findings indicate that during the longer survival cycle, subsets of mature chief cells exhibit reserve stem cell properties to transdifferentiate into various lineages, resulting in self-depletion, thus providing direct evidence for chief cells’ regenerative potential in gastric glands. Conventional lineage tracing approaches are fundamentally limited by the temporal imprecision of Cre-loxP systems, including both promoter activation delays and reporter expression lag. These technical constraints restrict detection to initial progenitor populations and terminal differentiated states, precluding discrimination between transdifferentiation and dedifferentiation pathways. In contrast, single-cell sequencing technologies enable high-resolution trajectory reconstruction by computationally ordering cells based on transcriptional profiles, thereby revealing both linear differentiation cascades and branched fate decisions.A Monocle pseudotime trajectory analysis of representative chief cell genes (73) has for the first time delineated their potential differentiation pathways at single-cell resolution, demonstrating that chief cells can transdifferentiate into mucous neck cells before progressing to SPEM, strongly supporting the chief cell origin hypothesis of SPEM.

The loss of parietal cells has been identified by multiple research groups as a critical precursor to chief cell transdifferentiation following gastric mucosal injury (14, 74). In gastric parietal cell-specific solute carrier family 26 member 9 (Slc26a9) knockout mouse models (Slc26a9fl/fl/Atp4b-Cre) (75), pyroptosis-driven parietal cell loss and SPEM development can be observed. Similar findings have been replicated in other animal models. However, it remains undetermined whether SPEM arises through proliferative differentiation or direct transdifferentiation of chief cells post-parietal cell depletion. Subsequently, the Nam research team (6) developed Mist1CreER/+/Rosa26RLacZ mice using the CRISPR/Cas9 gene-editing system for novel lineage tracing studies. In this model, tamoxifen-induced Cre was knocked into the chief cell-specific Mist1 locus, leading to β-galactosidase expression upon chief cell maturation. Observation of this reporter gene revealed that SPEM cells exhibiting β-galactosidase activity emerged in the gastric fundic mucosa across three parietal cell-depletion models: DMP-777 (non-inflammatory), L-635 (acute inflammation), and H. pylori infection (chronic inflammation). Notably, enhanced β-galactosidase activity and accelerated SPEM progression were observed in inflammatory models. These findings suggest that under conditions of significant parietal cell reduction, SPEM originates predominantly from mature chief cell transdifferentiation, and that the combination of parietal cell loss and inflammation synergistically promotes SPEM development. In another model using 5-fluorouracil to block mitotic division in gastric mucosal cells (8), tamoxifen was still able to induce parietal cell loss and SPEM formation in mice with suppressed proliferation. It was inferred that some SPEM formation may be related to direct reprogramming of existing basal chief cells, not dependent on mitotic genetic mechanisms.

The aforementioned research demonstrates from two aspects that SPEM occurrence is temporally and spatially related to chief cell transdifferentiation induced by parietal cell loss (1): Gastric epithelial metaplasia consistently emerges following parietal cell atrophy. Autoimmune gastritis (76), chronic H. pylori infection (22), acute drug induction (55, 56), and novel parietal cell-specific knockout models (72, 75) collectively confirm that mature parietal cells are crucial regulators of gastric epithelial differentiation. After oxyntic gland atrophy, the phenotypic characteristics of emerging SPEM cells resembling chief cells suggest their origin from parietal cell loss-induced chief cell alterations (77) (2). Subsequent lineage tracing (6) and immunohistochemical evidence (8) have robustly demonstrated chief cells’ multidirectional differentiation potential through longitudinal observation and multi-model validation. Critically, these investigations established that chief cells can directly transdifferentiate into SPEM cells independent of proliferative activity, challenging the conventional paradigm of mitosis-dependent metaplasia. Similar mechanisms are observed in tumor epithelial-mesenchymal transition (EMT), where gastric epithelial cells lose polarity and acquire stem-like properties through coordinated epigenomic and genomic changes, entering mesenchymal states with enhanced migratory/invasive capacities. Although enhanced proliferative activity may coexist with EMT in tumor cells, these processes function synergistically (78, 79). This provides insights for understanding SPEM-associated chief cell transdifferentiation, revealing that SPEM essentially represents cellular plasticity akin to EMT (80). The collective significance of the aforementioned studies lies in establishing the central role of chief cell plasticity in the context of parietal cell loss in SPEM formation. However, they have not yet precisely unraveled the dynamic process of cell fate transition. To uncover more intuitive and profound regulatory mechanisms, it is essential to establish a more accurate functional research system for SPEM, which requires further development of animal models with permanently induced parietal cell atrophy, the advancement of lineage tracing technologies with higher spatiotemporal resolution, and the application of multi-omics analyses.

Previous studies predominantly focused on parietal cell loss as the initiating factor driving chief cell transdifferentiation. However, research involving targeted ablation of parietal cells failed to induce SPEM, challenging the prevailing view. This evidence indicates that parietal cell loss serves as a permissive condition rather than a driver for SPEM development (81). Emerging evidence from chief cell-specific gene knockout models and pharmacogenetic approaches supports the perspective that active depletion of chief cells, even in the presence of intact parietal cells, can lead to SPEM and subsequent gastric epithelial malignant transformation (69, 82, 83).

Runt-related transcription factor 3 (RUNX3), a pivotal developmental regulator, integrates microenvironmental signals to modulate cell cycle progression and cell fate determination (84). In the BALB/c background RUNX3^-/- adult mouse model established by Ito et al. (83), researchers observed loss of chief cells and SPEM development without inflammatory intervention, while parietal cells remained intact. This model subsequently developed gastric adenocarcinoma, indicating that mere parietal cell atrophy is insufficient to fully induce metaplasia and that SPEM may arise through mechanisms beyond parietal cell injury or death. However, prior studies have established that RUNX3 dysregulation disrupts cellular differentiation and proliferative homeostasis, driving intestinal-type gastric carcinogenesis (85, 86). Thus, the above findings only confirm that RUNX3 deficiency induces SPEM independently of parietal cells, with chief cell loss being a critical event in this process. Since the RUNX3^-/- model targets the entire gastric epithelium (not chief cell-specific), it cannot directly support the notion that primary chief cell depletion initiates SPEM. Subsequently, Liu et al. (82) demonstrated that CHIA is highly expressed in chief cells, and its loss in chief cells constitutes a pivotal event driving SPEM development and gastric cancer progression independently of parietal cell loss. To validate whether primary chief cell depletion directly induces SPEM, the team established a chief cell-selective CHIA-deficient mouse model (CHIA^rox/rox-GIF-Dre) using Dre-Rox technology. In this model with intact parietal cells but depleted chief cells, they observed upregulated expression of SPEM markers (Wfdc2 and CD44v9) and loss of the chief cell marker MIST1, accompanied by reduced expression of the stemness marker Lgr5. The parietal cell marker ATP4b showed only minor changes. Electron microscopy further revealed relocalization of the mucous neck cell marker SRY-box transcription factor 9 (SOX9) from the isthmus-neck region to the gastric corpus, with co-localization of SPEM markers TFF2 and CD44v9. These findings suggest that in the presence of parietal cells, chief cells may transdifferentiate into mucous neck cells, thereby inducing SPEM. However, the Barker group (69) proposed a different view on which kind of cells the exhausted chief cells transdifferentiate into. The team selectively ablated Lgr5⁺ chief cells (accounting for 40% of all chief cells) using diphtheria toxin in the Lgr5-DTR-EGFP mouse model, revealing that damage to this subset drives gastric epithelial regeneration. Furthermore, upon tamoxifen-induced Kras (G12D) mutation (Lgr5-2A-CreER-T2^tg/tg/LSL-Kras(G12D)^tg/+), these cells directly progressed to gastric cancer. This evidence indicates that Lgr5⁺ chief cells are prone to malignant transformation and can serve as the cellular origin of gastric cancer. Thus, the study proposes that under physiological conditions, Lgr5⁺ chief cells function as terminally differentiated cells; however, upon gastric tissue injury, they acquire stem-like properties, participating in SPEM formation and initiating malignant transformation.

Previous studies have revealed that gastric chief cells are dynamically regulated by p57 and can coordinate tissue homeostasis and damage repair in response to microenvironmental changes (87). Current evidence (69, 82, 83) further reveals a dual-pathway role for chief cells in SPEM pathogenesis: SPEM may arise partially through secondary responses to parietal cell loss; partially through gene mutations induced by active chief cell depletion. Subsets of chief cells marked by distinct molecular signatures may exhibit fate heterogeneity during transdifferentiation: some differentiate into mucous neck cells, while others acquire stem-like properties—yet both pathways drive SPEM development. Nevertheless, additional evidence is required to validate the chief cell hypothesis. Current research disproportionately focuses on the entire gastric corpus gland rather than specifically on its basal region (the chief cell zone). Future studies involving targeted ablation of mucous neck cells and isthmus stem cell depletion hold crucial implications for delineating the temporal dynamics of SPEM progression and elucidating its underlying mechanisms.

It is generally accepted that multipotent stem cells in the gastric isthmus are rapid-cycling stem cells that govern cellular differentiation in gastric glands and mediate tissue repair (88, 89). Stem cells generate lineage-specific progenitor cells that differentiate into pit cells at the gland apex, maintain stem cells and parietal cells in the isthmus, and form neck mucous cells in the neck region. These neck mucous cells subsequently migrate along the basal region and progressively differentiate into chief cells (90, 91).Over the past three decades, the consensus kinetic model of gastric epithelial and chief cell dynamics proposes that upon activation, isthmus-derived stem cells traverse specific migratory pathways along the neck region. Through intermediate transitional stages involving pre-neck mucous cells and pre-chief cells, they ultimately undergo transdifferentiation into mature chief cells (92–94). This conceptual framework has established the theoretical basis for the stem cell/pre-metaplastic phenotype origin hypothesis of SPEM.

Previous studies on SPEM predominantly relied on lineage tracing, identifying cellular changes through expression domains of molecular markers, which partially elucidated spatial distribution patterns of local cell populations. Although the expansion of TFF2/GIF double-positive cells and lineage tracing using GIF/Mist1 support chief cell origins, some scholars contend that the transdifferentiation hypothesis remains flawed (1): Earlier studies assumed Mist1, Troy, and Lgr5 as chief cell-specific markers, yet Mist1 and Troy also label isthmus stem cells (95, 96), while Troy is expressed in parietal cells (68). Although the Lgr5-2A-CreERT construct demonstrates relative chief cell specificity (69), evidence shows Lgr5 mRNA expression in the isthmus post-injury, inducing genetic recombination in Lgr5-EGFP-IRES-CreERT mice (9) (2); Lineage tracing models primarily employ tamoxifen-induced Cre activation, yet TAM concurrently induces parietal cell death, epithelial cell apoptosis, and impairs stem cell activity (57, 97); (3) Troy-CreERT knock-in mouse studies proposed chief cells possess “reserve” stem cell properties (68), but observed behaviors were later confounded by Troy haploinsufficiency (95); (4) Most models utilize drug-induced acute SPEM, which differs from chronic genetically-reprogrammed SPEM in morphology and molecular mechanisms, failing to represent authentic metaplasia (98).

Hayakawa, a leading proponent of skepticism, addressed these limitations by developing an alternative lineage tracing system (99). Bypassing conventional Cre-based constructs, his team crossed TetO-Cre mice with Gpr30-rtTA mice to specifically label GPR30+ chief cells, coupled with fluorescent reporter gene (R26-TdTomato) tracking, thereby proposing GPR30 - a G protein-coupled estrogen receptor - as a novel chief cell marker.The researchers induced gastric metaplasia through high-dose tamoxifen, DMP-777, and H. pylori infection, while establishing Kras(G12D) and HRAS(G12V) mutant mouse models. Using dichloroacetate (DCA) to inhibit PDK activity, they investigated cellular competition in gastric mucosal renewal and homeostasis. Key findings revealed: GPR30+ chief cells were depleted upon Kras activation without triggering metaplasia/dysplasia, suggesting Kras mutation selectively impairs chief cells while neck-derived cell lineages expand via compensatory mechanisms.Mechanistically, metaplastic stimuli eliminate chief cells through GPR30/PDK-dependent cellular competition. Genetic ablation of GPR30 or pharmacological PDK inhibition preserved chief cell populations and attenuated neck lineage expansion. Metabolic reprogramming via PDK activation under metaplastic stress creates metabolic vulnerability in chief cells, ultimately excluding them as SPEM precursors. Instead, compensatory proliferation of neck progenitors emerges as the probable cellular origin of SPEM ( Figure 2 ).

This study circumvented the limitations of the Cre-loxP system, enabling faster induction of specific labeling while reducing the toxic interference of tamoxifen, thereby revealing the critical role of PDK-dependent cell clearance. However, the key conclusions critically hinge on the assertion that GPR30 exclusively labels all chief cells, whereas current understanding of GPR30’s structural expression patterns remains inconsistent (100). This suggests that whether chief cells are completely labeled remains uncertain. Furthermore, some limitations of the study remain unresolved, including the lack of validation in chronic models and the undefined clinical relevance to humans. The dual-pathway regeneration strategy in gastric epithelial repair can be understood as follows: during homeostasis, isthmus stem cells replenish multiple epithelial lineages (pit, neck, and parietal cells), representing a homeostatic repair process that is slow yet precise. In contrast, chief cells in the basal region primarily maintain their homeostasis through self-replication (7). Upon acute gastric mucosal injury, the chief cell transdifferentiation pathway is activated, enabling rapid but less stable repair that predisposes to carcinogenesis (70). This paradigm has led some researchers to postulate dual stem cell niches in gastric glands - isthmic and basal compartments - potentially reconciling the enduring controversy regarding SPEM origins (chief cell vs stem cell hypotheses) and addressing ongoing debates about stem cell marker specificity (95) ( Table 1 ).