Myosin VB is critical for progenitor cell identity and function in the intestine

Summary

Microvillus inclusion disease (MVID) is a congenital diarrheal disorder, caused by inactivating mutations in myosin Vb (MYO5B). MYO5B-deficient mice and cell lines have demonstrated the importance of MYO5B in brush border development; however, the previous models lacked specificity to test intestinal stem cell functions. In the present study, we investigated the effects of progressive MYO5B deficiency originating in intestinal crypt cells utilizing mouse models. Our transcriptomic and multiplex immunostaining datasets demonstrate that MYO5B is critical for intestinal stem cell function. MYO5B-deficient crypts acquire a hyperproliferative phenotype with incomplete cell differentiation in vivo and an elevated organoid formation rate compared to control crypts. An evident disruption in mitochondrial structure and fatty acid metabolism likely underlies these crypt phenotypes. Consistent with mouse models, MVID patient biopsies demonstrate abnormal expansion of the proliferative zone along with villus blunting. These data reveal the direct role of MYO5B in intestinal epithelial progenitor cell functions.

Graphical abstract

Highlights

• •Two mouse strains were generated to address MYO5B function in intestinal progenitor cells
• •The loss of MYO5B alters stem cell identity, instead promoting a TA cell state
• •MYO5B is critical for intestinal crypt cell proliferation, differentiation, and cell metabolism.

In this study, Kaji and colleagues reveal that the loss of MYO5B in intestinal epithelial progenitor cells drives a progressive wave of hyperproliferation along the crypt-villus axis at the expense of proper cell maturation and differentiation. Utilizing two crypt-targeted mouse models of MYO5B deficiency, this study establishes that MYO5B is required for intestinal progenitor cell identity, mitochondrial metabolism, and differentiation function.

Introduction

The small intestine is the primary site of nutrient absorption in the body, where nutrients are taken up by the lumen-facing epithelium by a variety of brush border-resident nutrient transporters. The intestinal epithelium is composed of crypt-villus units, where the crypt houses undifferentiated, proliferative progenitor cells that give rise to differentiated absorptive (enterocyte) and secretory (tuft, enteroendocrine [EEC], Paneth, and goblet) cell lineages (Barker et al., 2008). The differentiated epithelial cells carry out critical small intestinal functions, including barrier formation, luminal sensing, and nutrient absorption. Proper intestinal progenitor functions, proliferation, and differentiation, are necessary to maintain epithelial turnover and overall intestinal function. Related literature has illustrated that progenitor cell metabolism and niche signals provided by Paneth cells and subepithelial cells of the lamina propria are critical to maintain a balance of epithelial progenitor cell proliferation vs. differentiation (Quintero and Samuelson, 2025; Sailaja et al., 2016; Shay and Yilmaz, 2025; Wang et al., 2021).

Microvillus inclusion disease (MVID) is an autosomal recessive disorder that typically presents with severe watery diarrhea within the first week of life (Bowman et al., 2022; Cutz et al., 1989; Davidson et al., 1978). MVID primarily affects the small intestinal epithelium, with mistrafficking of apical membrane components in enterocytes leading to nutrient malabsorption and diarrhea (Thiagarajah et al., 2018). MVID is caused by inactivating mutations in the motor protein myosin VB (MYO5B), which is highly expressed throughout the intestinal epithelium (Erickson et al., 2008; Muller et al., 2008). Several point mutations in MYO5B have been documented in MVID patients, such as MYO5B(G519R) and MYO5B(P660L), and have been recently shown to differentially alter MYO5B motor function and Rab11a cellular localization (Bowman et al., 2025; Burman et al., 2023). MVID is fatal without treatment, which is currently limited to chronic total parenteral nutrition (TPN) and small bowel transplant (Kaji et al., 2024). To identify alternative, less-invasive treatment options for MVID, the function of MYO5B in the intestinal epithelium has been explored.

Prior to 2015, MVID research exclusively relied on patient biopsies and immortalized cell lines. Recently, several mammalian models of MVID have been generated, including MYO5B-mutated pigs and mice, to better understand the phenotype of MYO5B deficiency in the small intestine (Engevik et al., 2020; Schneeberger et al., 2015; Weis et al., 2016). Animal models recapitulate MVID hallmarks seen in patients, including villus blunting and defects of apical transporter trafficking. Recently, we demonstrated that Vil1-Cre^ERT2; Myo5b^fl/fl (Vil1^ΔMYO5B) mice have a disruption in differentiated secretory cell numbers and an elongation of the proliferative crypts (Kaji et al., 2021). Alterations in epithelial cell fatty acid metabolism and Wnt:Notch signaling balance were identified by bulk transcriptomics of jejunal tissues and enteroids, likely correlated to crypt dysfunction caused by epithelial MYO5B deficiency (Kaji et al., 2021; Momoh et al., 2024). Although these initial studies were critical in establishing a role for MYO5B in intestinal crypt function, these progenitor cell functional and metabolic deficits may reflect secondary effects of villus blunting and nutrient malabsorption characteristic of the MVID-like phenotype rather than direct effects of MYO5B loss. Therefore, the crypt-specific phenotype of MYO5B-deficiency remained unresolved.

Results

Lrig1^ΔMYO5B mice are a model of progressive MYO5B deficiency originating in the intestinal crypts

To evaluate the crypt-specific effects of MYO5B deficiency, we crossbred mice harboring a tamoxifen-inducible, intestinal epithelial progenitor-specific Cre and YFP lineage marker, Lrig1-Cre^ERT2; Rosa26-LSL-YFP (Powell et al., 2012), with Myo5b^fl/fl mice (Weis et al., 2016) to generate Lrig1-Cre^ERT2; Rosa26-LSL-YFP; Myo5b^fl/fl (Lrig1^ΔMYO5B) mice (Figure 1A). Lrig1^ΔMYO5B mice began losing significant weight compared to controls on day 4 following tamoxifen administration, averaging a loss of 18.4% of their starting body weight on day 5 (Figure 1B). This drastic weight loss was associated with MVID-like watery diarrhea at day 5, where the small intestine contained clear, watery luminal contents and the colon was devoid of solid feces (Figure 1C).

Figure 1.

Lrig1^ΔMYO5B mice are a model of progressive MYO5B deficiency originating in the small intestinal crypt

(A) Experimental timeline of tamoxifen administration (TMX) on day 0 to induce the Lrig1-Cre recombinase and the collection of intestinal tissues on days 3 and 5. MYO5B loss and corresponding YFP+ lineage cells are depicted in green.

(B) Changes in mouse body weight following TMX administration. Daily results are presented as mean ± SEM; N = 5–9 mice per group. ^∗∗∗∗p < 0.0001 by two-way ANOVA with Tukey’s test. Asterisk colors depict significances between Lrig1^ΔMYO5B and M5B f/f (black) or Lrig1-CreER^T2; M5B fl/+ (gray).

(C) Gastrointestinal tract from day 5 control and Lrig1^ΔMYO5B mice. Scale bars, 1 cm.

(D) Duodenal histology with Alcian blue-PAS (AB-PAS) staining and MYO5B immunostaining (inverted). Immunostaining illustrates a decrease in targeted epithelial MYO5B expression, and unclassified non-epithelial signal is evident in MYO5B-deficient lamina propria tissues. Example crypts outlined in yellow dashes. Scale bars, 50 μm.

(E) Quantification of crypt and villus lengths from AB-PAS-stained mouse tissues. Mean ± SEM; N = 5–10 mice per group. ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by two-way ANOVA with Dunnett’s test vs. control. Asterisk colors indicate significances in villus (gray) or crypt (yellow) values.

(F) Immunostaining for SGLT1, YFP, and ACTG1 in the mouse duodenum. Scale bars, 50 μm.

(G) Control (adult and pediatric) and pediatric MVID patient small intestinal tissues with multiplexed immunostaining for PCNA (proliferative marker), ACTG1 or Ezrin (brush border), CTNNB1 (basolateral membrane), DEF5A (Paneth cell), and nuclei. Scale bars, 50 μm.

See also Figure S1.

Immunostaining of intestinal tissues revealed a crypt-specific loss of MYO5B at day 3, while MYO5B expression was maintained in the villi (Figures 1D and S1A). Corresponding to the loss of MYO5B, day 3 Lrig1^ΔMYO5B small intestinal crypts were significantly elongated compared to controls (Figures 1D, 1E, and S1A). Importantly, the day 3 Lrig1^ΔMYO5B mice did not experience weight loss in comparison to day-matched controls, suggesting that this time point permits investigation of progenitor cell-specific MYO5B deficiency independent from systemic malnutrition.

Following the expected rate of epithelial turnover, the wave of MYO5B deletion originating in the crypt-resident Lrig1-expressing progenitor cells reaches the tips of villi at day 5 (Figures 1D and S1A). Despite the uniform loss of MYO5B expression, YFP lineage labeling remained sporadic (Figures 1F and S1B), which is consistent with previous reports of inefficient recombination at the Rosa26-LSL-YFP locus (Johnson et al., 2022). The day 5 Lrig1^ΔMYO5B intestine presented with villus blunting and abnormal accumulations of periodic acid Schiff (PAS)+ vesicles in enterocytes, similar to MVID patient biopsies (Figures 1D and 1E) (Phillips et al., 2000; Ruemmele et al., 2006). The crypt length was significantly increased in the day 5 Lrig1^ΔMYO5B small intestine compared to controls (Figures 1D and 1E).

To confirm a progressive loss of MYO5B function originating in the crypts, small intestinal sections from each post-tamoxifen day were immunostained for apical nutrient transporters, which were previously indicated to be trafficked by MYO5B (Engevik et al., 2018; Burman et al., 2023). Sodium-dependent glucose cotransporter 1 (SGLT1) was progressively diminished and mislocalized away from the apical surface in the proximal small intestine following Cre induction (Figure 1F). Similarly, an ileal-specific apical sodium-dependent bile acid transporter (ASBT) was mislocalized into the cytoplasm in the day 5 Lrig1^ΔMYO5B distal small intestine (Figure S1B). The mistrafficking of SGLT1 and ASBT is indicative of a progressive loss of MYO5B function throughout the small intestine of Lrig1^ΔMYO5B mice, resulting in a general loss of nutrient-absorbing transporters from the apical surface and diarrhea. These data illustrate Lrig1^ΔMYO5B as a mouse model of progressive MYO5B loss originating in crypt progenitor cells that presents an MVID-like phenotype at 5 days following Cre recombination.

To determine if the crypt expansion phenotype in Lrig1^ΔMYO5B mice mirrors the intestinal histology of MVID patients, we conducted multiplexed immunofluorescence (MxIF) staining on the intestinal biopsies from two MVID patients, harboring different mutations in MYO5B, along with adult and pediatric control tissues. As expected, tissues from both MVID patients revealed characteristic villus blunting compared to controls (Figure 1G). The proliferative crypts, illustrated by immunostaining for PCNA (PCNA+), of both MVID patient biopsies appeared elongated compared to control tissues. Continuous structure of crypt-villus axes was limited in these small biopsies, preventing precise quantification of crypt proliferative cells, and age-dependent changes in intestinal morphology need to be considered. However, a previous report demonstrating a significant increase in proliferative Kiel 67 (Ki67)+ cells in the intestinal epithelium of MVID patients (genetic information unknown) (Groisman et al., 2000) corroborates the present observations induced by MYO5B-inactivating mutations. These data suggest that functional loss of MYO5B in multiple MVID patients and progenitor cell-specific Cre-expressing mouse models gives rise to hyperproliferative, elongated small intestinal crypts.

MYO5B is necessary for the transcriptional identity of intestinal stem cells

To explore the cellular phenotype of the MYO5B-deficient crypts and the cells derived from MYO5B-deficient progenitors, we conducted single-cell RNA sequencing (scRNA-seq) of the jejunum of day 3 and day 5 Lrig1^ΔMYO5B mice and littermate controls. Knowing the contribution of lamina propria cells to maintain the epithelial progenitor niche, single cells from both the epithelium and lamina propria were included in these analyses. The scRNA-seq dataset was clustered into the epithelial subtypes of the villus (goblet, tuft, enterocyte, EEC) and crypt (stem, transient amplifying [TA], absorptive progenitor, secretory progenitor, and Paneth) cells, as well as non-epithelial cell types (fibroblast, myofibroblasts, B cell, T cell, macrophage, endothelial, pericyte, glial, and immune progenitor) across 71,052 cells (Figures 2A and S2A). The scRNA-seq data revealed that both the Cre driver (Lrig1) and the flox-targeted (Myo5b) genes are highly expressed in control intestinal stem cells (ISCs), supporting that the recombination is expected to primarily occur in the stem cell population (Figure S2B).

Figure 2.

MYO5B is critical for transcriptional identity of intestinal stem cells

(A) UMAP of scRNA-seq of control, day 3 Lrig1^ΔMYO5B, and day 5 Lrig1^ΔMYO5B mouse jejunum. Upper UMAP includes all samples (control and Lrig1^ΔMYO5B), and lower UMAPs only include specified sample types. N = 2 mice per group.

(B) UMAP of scRNA-seq data color coded by sample type.

(C) Heatmap depicting relative transcription levels of intestinal epithelial progenitor genes in combined progenitor clusters (stem, TA, absorptive progenitor, and secretory progenitor).

(D) Heatmap depicting relative transcription levels of fetal stem gene signature in combined progenitor clusters across individual mouse samples.

See also Figure S2.

To visually compare the identity of control vs. Lrig1^ΔMYO5B jejunal cells, a uniform manifold approximation and projection (UMAP) was generated with individual cells color coded by sample type (control, day 3 or day 5 Lrig1^ΔMYO5B; Figure 2B). Stratifying the scRNA-seq by sample type revealed the altered crypt cell composition, particularly, the stem cell cluster composed almost entirely of control cells (Figures 2A, 2B, and S2A). Compared to an average of 15.9% of total control cells defined as stem cells, there were only 0.8% (day 3) and 0.3% (day 5) of Lrig1^ΔMYO5B cells (Figure S2C). This remarkable loss of stem cell identity corresponded to a decrease in several canonical ISC markers, including Lgr5, Olfm4, and Ascl2 (Figure 2C). Although not all fetal stem markers were uniformly upregulated in MYO5B-deficient progenitor cells, several fetal reversion signature genes (i.e., Akr1b8, Anxa3, Ccn2/Ctgf, Ddit4l, Il33, Ly6a, Ly6c1, Ly6f, S100a6, Sprr1a, Sptssb, and Syt14) (Mustata et al., 2013) were increased in day 3 and/or day 5 Lrig1^ΔMYO5B progenitor cells (Figure 2D). However, other targets were decreased (i.e., Clu and Acsm3) and many markers were inconsistently altered in MYO5B-deficient progenitor cells. Rather than stem cells, the proliferative epithelial cells of Lrig1^ΔMYO5B mice were identified as TA and partially lineage-committed absorptive or secretory progenitor cells (Figures 2A–2C and S2C; Table S1).

Corresponding to these alterations in progenitor cell identity, the transcription of ISC niche-related genes (i.e., R-spondins, Wnt ligands and modulators, bone morphogenetic protein [BMPs], and growth factors) was remodeled in fibroblast and myofibroblast clusters, which could underlie crypt dysfunction in day 3 and 5 Lrig1^ΔMYO5B mice (Figure S2D). Taken together, this dataset suggests that (1) the proliferative cells of the MYO5B-deficient crypts have skewed transcriptional cell identity toward that of TA or absorptive/secretory progenitor cells and (2) progenitor cell niche factors provided by fibroblasts and myofibroblasts are indirectly altered by the MYO5B-deficient crypt cells.

An increase in intestinal epithelial proliferation corresponds to the wave of progressive MYO5B loss

We next assessed the effect of MYO5B deficiency on one of the primary functions of the intestinal crypts: epithelial cell renewal by proliferation. Corresponding to MYO5B loss in the crypts, the proliferative PCNA+ zone was elongated in the small intestine of the day 3 Lrig1^ΔMYO5B mice compared to control tissues (Figures 3A, 3B, and S3A). On day 5, the proliferative zone was further expanded, where PCNA+ epithelial cells were even present in the villi (Figures 3A and 3B). Consistent with the lengthening of the proliferative zone, there was a significant decrease in non-proliferative (PCNA–) length in the Lrig1^ΔMYO5B tissues compared to that of control mice. Additionally, the traveling distance of 5-Ethynyl-2'-deoxyuridine positive (EdU+) epithelial cells labeled 24 h prior to the tissue collection was significantly increased at both day 3 and 5 time points in the Lrig1^ΔMYO5B mouse duodenum compared to control tissues (Figures 3C and 3D).

Figure 3.

An increase in intestinal epithelial proliferation corresponds to the wave of progressive MYO5B loss

(A) Immunostaining for PCNA and ACTG1 in the mouse duodenum. Scale bar, 50 μm.

(B) Quantification of proliferative (PCNA+) and post-proliferative (PCNA−) epithelial length in the mouse duodenum. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by two-way ANOVA with Dunnett’s test vs. control. Asterisk colors indicate significances in PCNA+ (magenta) and PCNA– (black) values. Mean ± SEM; N = 3–7 mice per group.

(C) Fluorescence labeling for EdU (injected 24 h prior to the tissue collection) and nuclei in the mouse duodenum. Scale bars, 50 μm.

(D) Quantification of EdU+ vs. EdU− epithelial length in the mouse duodenum. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗∗p < 0.0001 by two-way ANOVA with Dunnett’s test vs. control. Asterisk colors indicate significances in EdU+ (yellow) and EdU– (blue) values. Mean ± SEM; N = 3–7 mice per group.

(E) Epithelial cell cycle stage proportions determined by scRNA-seq. Gene transcriptions related to G1/G0 (non-proliferative), S (proliferative), and G2M (proliferative) phases are analyzed in total epithelial clusters.

(F) UMAP of scRNA-seq data labeled by cell cycle stages (G1/G0 [blue], S [orange], and G2M [green]).

(G) Significantly upregulated pathways related to proliferation in day 3 Lrig1^ΔMYO5B epithelial progenitor cells by GSEA. Differentially regulated pathways were queried using WikiPathways_2024_Mouse gene set, comparing day 3 Lrig1^ΔMYO5B vs. control progenitor cells.

(H) Immunostaining for KIF11 (kinesin family member 11), PCNA, and CTNNB1 in the mouse jejunum. Arrows indicate mitotic KIF11+ epithelial cells. Cell division angle (relative to the basement membrane) is measured in each KIF11+ cell, and the ratio of abnormal cell division is quantified per crypt per section (N = 13–15 fields from 3–4 mice per group). Scale bars, 25 μm. Datapoint shapes represent values from individual mice. ^∗p < 0.05 by Kruskal-Wallis test with Dunn’s test. Error bars represent SD.

See also Figure S3.

Mirroring the histological findings, our scRNA-seq dataset illustrated an increase in the proportion of proliferative S and G2M phases of the Lrig1^ΔMYO5B epithelial cells compared to controls (Figures 3E and S3B). As revealed by a color-coded UMAP by each cell cycle phase, these proliferative cells were predominantly immature cell types in both control and Lrig1^ΔMYO5B samples (Figures 2A and 3F). At both days 3 and 5, the gene set enrichment analysis on epithelial progenitor cells demonstrated a significant upregulation of proliferation-related pathways in Lrig1^ΔMYO5B cells compared to controls (Figures 3G and S3C). These immunostaining and transcriptomic data highlight a hyperproliferative phenotype of the small intestinal crypts directly induced by MYO5B deficiency. These findings in the Lrig1^ΔMYO5B mouse model recapitulate the increased proliferation observed in the MVID patient biopsies (Figure 1G) (Groisman et al., 2000).

MYO5B deficiency perturbs mitotic spindle organization in progenitor cells

A recent study demonstrated the direct interaction of RAB11 proteins with mitotic spindle machinery, as the intestinal crypt cells of mice lacking both RAB11A and RAB11B were elongated and had perturbed mitotic spindle formation demonstrated by kinesin family member 11 (KIF11) immunostaining (Joseph et al., 2023). As both RAB11 proteins have documented interactions with MYO5B (Lapierre et al., 2001), we investigated mitotic spindle formation in the Lrig1^ΔMYO5B small intestinal crypts. In the control crypts, mitotic cells contained bipolar spindles as revealed by symmetric localization of KIF11 (Figure 3H). However, day 3 and 5 Lrig1^ΔMYO5B crypts contained significantly more abnormal spindles, predominantly monopolar, compared to controls. Mitotic spindle angle did not significantly differ between genotypes. These tissue analyses suggest that MYO5B-deficient progenitors have mitotic spindle defects correlated to crypt elongation, similar to that of RAB11A & B double-knockout mice.

MYO5B-deficient progenitor cells have disrupted mitochondrial structure and cell metabolism, independent from systemic malnutrition

Given the importance of mitochondrial metabolism in proper intestinal progenitor cell identity and function (Wang et al., 2021), we next investigated the effects of MYO5B deficiency on crypt cell metabolism. We focused on the day 3 Lrig1^ΔMYO5B mice, as MYO5B was specifically lost in progenitor cells and the MYO5B-intact villus enterocytes maintained proper nutrient absorption. Electron microscopy (EM) was conducted to evaluate crypt mitochondrial structure. Unlike control progenitor cells, which contain mitochondria with dense, organized cristae, MYO5B-deficient stem and TA cells exclusively contained swollen mitochondria with disrupted cristae (Figures 4A and 4B). On the other hand, day 3 Lrig1^ΔMYO5B Paneth cells contained normal mitochondria similar to those in controls. At this time point, Paneth cells maintained MYO5B expression assessed by tissue immunostaining (Figure S4A). These cell-specific and genotype-specific differences highlight the specificity of mitochondrial structural defects to MYO5B-deficient progenitor cells.

Figure 4.

Progenitor cells lacking MYO5B have disrupted mitochondrial structure and cell metabolism, independent from systemic malnutrition

(A) Representative EM images of jejunal crypt bases of control and day 3 Lrig1^ΔMYO5B mice. Stem cells (highlighted in blue) are in direct contact with Paneth cells (purple; contain dark subapical granules). Arrows indicate mitochondria, colored by cell type. Scale bars: 5 μm in upper two panels and 20 nm in insets.

(B) Proportions of normal vs. abnormal mitochondria in the EM images. Abnormal mitochondria are identified with a swollen appearance and disrupted cristae structure. Stem cells are identified as crypt base epithelial cells between Paneth cells, and TA cells as crypt epithelial lacking direct contact with Paneth cells. N = 1 mouse per genotype; 5 crypts per mouse; all mitochondria in 15–26 cells of each cell type are analyzed.

(C) Fatty acid oxidation assay in the mouse crypt and villus cells isolated from jejunum. Each datapoint indicates the value of individual mouse. Mean ± SEM; N = 4 mice/genotype. ^∗p < 0.05 by two-way ANOVA with Tukey’s test.

(D) Heatmap depicting relative transcription levels of downregulated FAO genes in combined epithelial progenitor clusters.

(E) Heatmap depicting relative transcription levels of downregulated OXPHOS genes in combined epithelial progenitor clusters.

(F) Uptake of fluorescent glucose (2NBDG) and fatty acid (12C BODIPY) analogs by control and day 3 Lrig1^ΔMYO5B jejunal crypts. Scale bars, 50 μm. N = 3 mice per group.

See also Figure S4.

To determine the functional differences in progenitor cell metabolism, we conducted a fatty acid oxidation (FAO) enzymatic assay on isolated crypt and villus epithelial cells, respectively, from control and day 3 Lrig1^ΔMYO5B mouse jejunum. Villus FAO activity did not differ between control and Lrig1^ΔMYO5B cell extracts. However, corresponding to crypt-specific MYO5B deficiency, Lrig1^ΔMYO5B crypts demonstrated significantly less FAO compared to control crypts (Figure 4C). The transcriptional pattern of progenitor cell clusters illustrated a decrease in the expression of many FAO-related genes (Figures 4D and S4B). Numerous oxidative phosphorylation (OXPHOS)-related genes were also decreased in Lrig1^ΔMYO5B progenitor cells (Figures 4E and S4B). These data suggest a possible mechanism for malfunction in MYO5B-deficient intestinal progenitor cells.

Given the previously indicated role of MYO5B in the trafficking of intestinal nutrient transporters, we questioned if the metabolic alterations in MYO5B-deficient progenitor cells could result from a lack of proper nutrient import. However, the uptake of fluorescence-labeled glucose and fatty acid analogs were comparable between MYO5B-deficient and control jejunal crypts (Figure 4F). These results suggest that the metabolic deficits in Lrig1^ΔMYO5B progenitor cells are unlikely due to a lack of metabolic fuel.

MYO5B-deficient progenitor cells give rise to altered cell lineages

To compare cell lineage outcome from control and MYO5B-deficient progenitor cells, MxIF and subsequent cell type quantification was conducted on jejunal Swiss rolls from day 5 Cre-harboring control (Lrig1-CreER^T2; Rosa-YFP; Myo5b^fl/+) and Lrig1^ΔMYO5B mice (Figure 5A). These jejunal sections underwent alternate phases of immunostaining and bleaching, totaling 16 stained markers, including those of proliferation, intestinal cell subtype identity, YFP lineage marker, and tissue structures (Figures 5B and S5A). QuPath analyses were employed to differentiate epithelium vs. stroma and to quantify cell proportions expressing the various markers (Figure 5A).

Figure 5.

The MYO5B-deficient, but not starved, progenitor cells give rise to altered differentiated cell lineages

(A) MxIF analyses strategy. Jejunal Swiss rolls from control, day 5 Lrig1^ΔMYO5B, and 48-h-fasted healthy mice are immunostained for 15 markers and nuclei. (Left column) Nuclei in whole Swiss rolls. (Middle column) MxIF overlay (color labels in B). (Right column) QuPath overlay. Non-epithelial cells are shown in gray, and epithelial cells are outlined with different colors based on their immunoreactivity for YFP and Ki67. The fasted mice lack ROSA-YFP allele and EdU administration. Scale bars, 50 μm.

(B) Individual channels from composite MxIF image from control tissues. Scale bar, 50 μm. Autofluorescence and post-bleaching carryover fluorescence are subtracted from each stained image.

(C) Quantification of epithelial cell types from MxIF of control, day 5 Lrig1^ΔMYO5B, and 48-h-fasted mouse jejunum. DCLK1+: tuft cells, CHGA+: EEC, LYZ1+: Paneth, TFF3+: Goblet, non-secretory or proliferating epithelial cell: enterocyte. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by one-way ANOVA with Tukey’s test. Mean ± SEM; N = 4–5 mice per group.

(D) Total epithelial cell type proportions as identified by MxIF staining and QuPath analyses. N = 4–5 mice per group.

(E and F) Quantification of total (E) and proliferative (F) 24-h EdU+ epithelial cells from MxIF of control and day 5 Lrig1^ΔMYO5B mouse jejunum. ^∗p < 0.05 by Mann-Whitney test. Mean ± SEM; N = 5 mice per genotype.

See also Figure S5.

Quantitative analyses of MxIF tissues revealed a significant decrease in tuft cells (doublecortin-like kinase-1, DCLK1+), EECs (chromogranin-A, CHGA+), and enterocytes (non-proliferative and non-secretory) in day 5 Lrig1^ΔMYO5B tissues (Figures 5C and 5D). A significant increase in proliferative (Ki67+, pHH3+, or 24 h EdU+) epithelial, but not stromal, cells was identified in the MYO5B-deficient tissues (Figures 5C–5E, S5B, and S5C). The cell fate of YFP+ lineage cells was similar to that of the overall epithelium (Figures 5D and S5D). A larger proportion of 24-h EdU+ cells also expressed Ki67+ in day 5 Lrig1^ΔMYO5B tissues, suggesting that the daughter cells of MYO5B-deficient progenitors remain immature rather than progressing to Ki67− differentiated lineages as seen in controls (Figure 5F). Together, these data suggest that MYO5B is necessary for proper progenitor cell differentiation in the small intestine.

The crypt phenotype of MYO5B-deficient progenitor cells is distinct from that of starvation

MYO5B loss-induced differentiation defects could be attributed to systemic malnutrition when approximately 20% body weight decrease is present. To identify the effects of MYO5B deficiency vs. starvation on progenitor cell function, 48-h-fasted littermate control (Myo5b^fl/fl) mice were included in MxIF analyses (Figures 5A and S5A). Unlike the day 5 Lrig1^ΔMYO5B tissues, differentiated and proliferative cell proportions were maintained in the 48-h-fasted tissues compared to controls (Figures 5C and 5D). These MxIF analyses suggest that the disrupted crypt phenotype seen in the day 5 Lrig1^ΔMYO5B tissues is directly due to the loss of MYO5B, rather than a secondary effect of nutrient malabsorption or subsequent starvation.

Deficits in MYO5B-deficient progenitor cell maturation correspond with cell morphology alterations and increased epithelial nuclear YAP

We next assessed the cellular phenotypes of the differentiated cell lineages derived from MYO5B-deficient progenitors. In our scRNA-seq dataset of day 5 Lrig1^ΔMYO5B jejunum, the transcriptional identity was skewed toward immature cell markers and away from differentiated cell signatures, most noticeably tuft cell, EEC subtypes, and mature enterocytes, in comparison to controls (Figure 6A). Consistent with MxIF data, the tuft cell cluster was nearly absent in day 5 Lrig1^ΔMYO5B (Figures 5C, 6A, S2C, and 6A).

Figure 6.

MYO5B-deficient progenitors give rise to immature cell lineages with maintained nuclear YAP expression

(A) Heatmap depicting relative transcription levels of major epithelial cell lineage markers in combined all epithelial cell clusters. Imm ENT: immature enterocyte.

(B) Immunostaining for Paneth (LYZ) and goblet (TFF3) cell markers. Green arrows highlight abnormal Paneth cells in upper crypts; orange arrows indicate abnormal cells co-expressing both markers. Insets depict normal goblet and Paneth cells from control (blue boxes), a double-positive cell from day 3 Lrig1^ΔMYO5B (orange box), and a goblet cell with abnormal shape from day 5 Lrig1^ΔMYO5B (magenta box). Scale bars: 50 μm in main panels and 10 μm in insets.

(C) Immunostaining for YAP, PCNA, CTNNB1, and nuclei in the day 5 Lrig1^ΔMYO5B and control mouse jejunum. Scale bars, 50 μm.

(D) Heatmaps of gene expression related to YAP signaling in epithelial progenitor cell clusters (upper) and all epithelial cell clusters (lower) from scRNA-seq.

See also Figure S6.

Corresponding to the increase in secretory progenitor cell clusters (primarily distinguished by goblet, Paneth, and low proliferative gene signatures), some upper crypt cells were double positive for goblet (TFF3) and Paneth (LYZ) cell markers in the Lrig1^ΔMYO5B jejunum (Figure 6B). LYZ+/TFF3− Paneth cells typically reside in the crypt base of the control small intestine; however, numerous mislocalized LYZ+ cells were present in the upper crypts of the MYO5B-deficient intestine. Villus goblet cells in the day 5 Lrig1^ΔMYO5B frequently lacked their characteristic goblet morphology observed in controls (Figure 6B, insets). This altered cell shape, along with reduced dark PAS+ staining in goblet cells (Figure 1D), suggests alterations in intracellular mucus contents derived from MYO5B-deficient progenitors. Despite the presence of single-marker secretory cells (TFF3+ or LYZ+) in both control and Lrig1^ΔMYO5B epithelia, the corresponding Paneth and goblet cell clusters were unexpectedly small in our scRNA-seq dataset (Figures S2B and S6A). This discrepancy suggests that our scRNA-seq protocol may have selectively enriched for crypt and enterocyte populations at the expense of certain secretory lineages.

MYO5B-deficient enterocytes presented with various MVID-characteristic structural differences, including an immature, disorganized brush border structure and an accumulation of mistrafficked subapical vesicles (Figure S6B). On the other hand, tuft and EEC cell morphology did not significantly differ between Lrig1^ΔMYO5B mice and controls (Figures S6C and S6D). Together, these data suggest that MYO5B deficiency promotes an immature epithelial phenotype and differentially affects secretory cell structures.

As YAP-associated signaling has been shown to be critical for proper epithelial maturation and intestinal progenitor niche signaling (Camargo et al., 2007; Deng et al., 2022; Zhou et al., 2011), we compared YAP protein expression and transcripts in the day 5 jejunum. In control tissues, nuclear staining for YAP was restricted to the crypt region of the epithelium (Figure 6C). Corresponding to a hyperproliferative, immature cell phenotype, nuclear YAP was observed throughout the elongated crypt and even in the villi of the day 5 Lrig1^ΔMYO5B intestine (Figure 6C). Likewise, the expression of several, but not all, TEADs and YAP target genes (i.e., Areg, Ccn1, Ccn2/Ctgf, and Tnfrsf12a) was elevated in the MYO5B-deficient epithelium (Figure 6D).

Small intestinal stem cell-specific MYO5B loss, under Olfm4-Cre driver, mirrors crypt dysfunction of Lrig1^ΔMYO5B mice

The tdTOM lineage reporter is more consistently expressed in Cre-derived intestinal cells compared to YFP (Johnson et al., 2022). Additionally, Olfm4 expression overlaps with the canonical ISC marker Lgr5 and Olfm4-driven Cre exhibits more uniform penetrance throughout the small intestine than the mosaic pattern observed in Lgr5-directed Cre models (Munoz et al., 2012; Schuijers et al., 2014). Therefore, we crossbred our Myo5b^fl/fl mice with the mice harboring Olfm4-Cre^ERT2; Rosa26-LSL-tdTOM allele (Bohin et al., 2020a; Schuijers et al., 2014) to generate Olfm4-Cre^ERT2; Rosa26-LSL-tdTOM; Myo5b^fl/fl (Olfm4^ΔMYO5B) (Figure 7A). Compared to the average 28% YFP+ epithelial cells in the day 5 Lrig1Cre+ control jejunum, nearly all jejunal epithelial cells (average 91%) were tdTOM+ at 7 days following tamoxifen administration in the Olfm4^ΔMYO5B mice (Figure S7A).

Figure 7.

MYO5B loss in the small intestinal stem cells in Olfm4^ΔMYO5B mice mirror the crypt dysfunction seen in Lrig1^ΔMYO5B mice

(A) Experimental timeline of TMX administration on day 0 to induce Olfm4-mediated Cre recombinase and isolation of intestinal tissue on day 5 or 7. MYO5B loss and corresponding tdTOM+ lineage cells are depicted in red. Organoids were generated from control and tdTOM+MYO5B− day 5 crypts.

(B) Changes in mouse body weight following the TMX administration. Daily results are presented as mean ± SEM. N = 5–9 mice per group. ^∗∗∗∗p < 0.0001 by two-way ANOVA with Tukey’s test. Asterisk colors depict difference between Lrig1^ΔMYO5B and M5B f/f (black) or Olfm4-Cre; M5B fl/+ (gray). Lrig1^ΔMYO5B value is copied from Figure 1B as a reference.

(C) Gastrointestinal tract from day 7 control and Olfm4^ΔMYO5B mice. Scale bars, 1cm.

(D) Morphological changes in the small intestine of day 7 Olfm4^ΔMYO5B mice. Duodenal histology with example crypts outlined in yellow dashes. Quantification of crypt and villus lengths in AB-PAS-stained duodenum and jejunum tissues from day 7 control and Olfm4^ΔMYO5B mice. ^∗∗∗∗p < 0.0001 by two-way ANOVA with Sidak’s test vs. control. Mean ± SEM; N = 5 mice per group. Asterisk colors depict significances in villus (gray) and crypt (yellow) values.

(E) Jejunal Swiss rolls from control and day 7 Olfm4^ΔMYO5B mice were immunostained for 13 markers using MxIF. (Left column) Composite images of MxIF (color labels in Figure S7C). (Right column) QuPath overlay from the corresponding MxIF images. Non-epithelial cells are shown in gray, and epithelial cells are colored based on immunoreactivity for 24-h EdU and PCNA. Scale bars, 50 μm.

(F) Quantification of epithelial cell types from MxIF of control and day 7 Olfm4^ΔMYO5B jejunum. DCLK1+: tuft cells, CHGA+: EEC, LYZ1+: Paneth, TFF3+: goblet, non-secretory or proliferating epithelial cell: enterocyte. ^∗∗p < 0.01 by Mann-Whitney test. Mean ± SEM; N = 5 mice per group.

(G) Quantification of total and proliferative 24-h EdU+ epithelial cells from MxIF of control and day 7 Olfm4^ΔMYO5B jejunum. ^∗p < 0.05 by Mann-Whitney test. Mean ± SEM; N = 4–5 mice per group.

(H) Immunostaining against PCNA, 24-h EdU, ACTG1, and nuclei in the mouse duodenum. Scale bars, 50 μm. PCNA+ vs. PCNA−, and EdU+ vs. EdU− epithelial lengths are measured. ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by two-way ANOVA with Dunnett’s test vs. Control. Mean ± SEM; N = 3–5 mice per group. Asterisk colors depict significances in PCNA+ (magenta) and PCNA– (black) values in upper graph, or EdU+ (blue) and EdU– (black) values in lower graph.

(I) Quantification of organoid formation of single cells isolated from day 5 control or OLFM4^ΔMYO5B mouse crypts. Organoids were imaged and quantified after 7 days in culture. Max projections of whole-dome z-stack images are presented. Scale bars, 500 μm. ^∗p < 0.05 by Mann-Whitney test. Mean ± SEM; N = 3–4 mice per group, with symbol shapes representing separate experimental trials.

See also Figure S7.

Weight loss curves of the Olfm4^ΔMYO5B mice mirrored that of Lrig1^ΔMYO5B mice but were delayed by 2 days (Figure 7B). Interestingly, day 7 Olfm4^ΔMYO5B mice contained solid feces (Figure 7C), despite the drastic weight loss and MVID-like tissue morphology in the small intestine (Figure 7D). This can be attributed to the lack of Olfm4 and corresponding Cre expression in the mouse colon as previously reported for Olfm4-GFP-Cre^ERT2 mice (Schuijers et al., 2014; van der Flier et al., 2009). Indeed, MYO5B immunostaining was retained in the Olfm4^ΔMYO5B colon but lost in the small intestine following the Cre induction (Figure S7B). The small intestinal MYO5B deficiency corresponded to the expression of the lineage marker, tdTOM (Figure S7B). Together, Olfm4^ΔMYO5B mice developed a progressive MVID-like phenotype similar to that of Lrig1^ΔMYO5B; however, this phenotype is restricted to the small intestine. These observations suggest that MYO5B loss in the small intestine is sufficient to cause systemic malnutrition and weight loss.

To validate our findings of intestinal progenitor cell dysfunction, immunostaining and MxIF analyses were conducted on day 7 Olfm4^ΔMYO5B and control small intestinal tissues (Figures 7E and S7C). MxIF staining and analyses demonstrated further similarities between the two mouse models of progenitor cell-mediated MYO5B-deficiency; several differentiated cell populations (tuft, EEC, enterocyte) were significantly decreased, and proliferative epithelial cells were significantly increased compared to controls (Figure 7F). Additionally, significantly more 24-h EdU+ cells were proliferative (PCNA+) in the day 7 Olfm4^ΔMYO5B jejunum compared to control tissues (Figure 7G). Although tdTOM immunostaining was not successful following the bleaching process in MxIF, our quantification of tdTOM+ cells suggest that nearly all epithelial cells were derived from Cre-expressing, MYO5B-deficient progenitors at day 7 (Figure S7A).

Corresponding to significantly elongated crypt length, the epithelial proliferative zone and 24-h EdU+ length were increased in the Olfm4^ΔMYO5B small intestine compared to controls (Figures 7D and 7H). Proliferative stromal cell and pHH3+ epithelial cell numbers did not significantly differ between controls and Olfm4^ΔMYO5B tissues (Figure S7D). The progenitor cell dysfunction observed in the Lrig1^ΔMYO5B mice was validated by the second, stem cell-mediated Olfm4^ΔMYO5B mouse model, further highlighting the importance of MYO5B for proper progenitor cell function.

To evaluate the epithelial cell-autonomous effects of MYO5B deficiency on progenitor cell proliferation, jejunal organoids (enteroids) were generated from day 5 OLFM4^ΔMYO5B and control mouse crypts. At the day 5 time point, crypts were lineage-labeled with tdTOM and yielded tdTOM+ organoids (Figure S7E). MYO5B-deficient enteroids were maintained through serial passaging with similar morphology to that of control passage-matched enteroids, suggesting a stable proliferative capacity of MYO5B-deficient progenitor cells. Additionally, we quantified organoid formation rate from single crypt cells to assess proliferative cell function. Despite a decrease in canonical ISC transcripts, MYO5B-deficient crypt cells formed significantly more organoids compared to control crypt cells (Figure 7I). This enhanced organoid growth recapitulates the elongated crypt phenotype observed in multiple MYO5B-deficient tissues and indicates that this hyperproliferative trait is epithelial intrinsic.

Discussion

Despite the extensive characterization of MYO5B function and the knockout phenotype in intestinal enterocytes and cell lines, the role of MYO5B in the intestinal crypts remained unexplored. In the current study, we overcame the previous limitations and provided progenitor cell specificity by generating two mouse models with MYO5B deficiency originating in epithelial progenitor cells of the crypts. The day 3 time point in Lrig1^ΔMYO5B mice revealed a crypt-specific deficiency of MYO5B expression, without a diarrheal phenotype or significant weight loss as the villi retained MYO5B. Two days later, Cre-induced progenitors gave rise to MYO5B-deficient daughter cells in the villi and corresponding watery diarrhea. This MVID-like phenotype was mirrored histologically, as blunted villi and internalized apical transporters were evident throughout the small intestine of day 5 Lrig1^ΔMYO5B mice. This progressive MVID-like phenotype originating in crypt progenitors was corroborated with the second mouse strain, harboring an ISC-specific Cre driver, Olfm4^ΔMYO5B mice. Our study presents several new insights into the crypt phenotype of MYO5B deficiency, including alterations in progenitor cell identity, differentiation, and cellular metabolism. Importantly, these deficits in MYO5B-deficient intestinal crypt cells are distinct from that of starvation, as 48-h-fasted healthy mice did not share the hyperproliferative phenotype or differentiation deficits observed in the day 5 Lrig1^ΔMYO5B or day 7 Olfm4^ΔMYO5B mice, which recapitulate MVID patient tissues.

Our scRNA-seq analysis highlighted alterations in progenitor cell identity in Lrig1^ΔMYO5B mice. Unexpectedly, our scRNA-seq data revealed epithelial cell population differed from what is typically observed by histology. Specifically, epithelial progenitor cell populations were enriched, while goblet and Paneth cells were under-represented, even in control tissues. This crypt enrichment is consistent with previously published intestinal scRNA-seq datasets (Haber et al., 2017) and allowed for in-depth characterization of intestinal epithelial progenitor cell identity, highlighting crypt cell alterations that are not detectable by histological assessment alone. There was an overall decrease in the transcription levels of canonical ISC genes and thereby, transcriptionally defined stem cell numbers in MYO5B-deficient tissues. Rather than stem cells, the elongated MYO5B-deficient crypts were predominantly populated by TA cells and abnormal progenitor cells with early lineage commitment. The loss of canonical stem cells alone is unlikely to be a direct cause of TA expansion, as previous studies using a Lgr5+ cell ablation mouse model lack a crypt expansion phenotype (Tan et al., 2021; Tian et al., 2011). Although Lgr5+ cell loss has been shown to abolish organoid formation capacity (Tan et al., 2021), our data suggested that MYO5B-deficient crypts have enhanced organoid-formation capacity despite a decrease in canonical stem cell transcripts. MYO5B loss in progenitors possibly induces a pool of alternative stem cells, possessing organoid-forming function.

The decrease in stem cell numbers and canonical ISC markers is reminiscent of intestinal crypt injury models, also referred to as fetal reversion (Viragova et al., 2024). Many stem markers are YAP signaling targets, implicated in intestinal proliferation and regeneration (Gregorieff et al., 2015; Imajo et al., 2015; Pikkupeura et al., 2023; Yui et al., 2018). Our immunostaining results revealed an increase in active, nuclear YAP expression in the MYO5B-deficient epithelium. Furthermore, an MYO5B trafficking partner, RAB11A, has been implicated in YAP subcellular localization, with RAB11a-deficient models demonstrating an increase in nuclear YAP (D'Agostino et al., 2019; Goswami et al., 2021). This nuclear localization of YAP is necessary for the association of YAP with TEAD transcription factors to promote YAP-associated signaling and is characteristic of proliferative, immature epithelial cells (Camargo et al., 2007; Huang et al., 2005; Imajo et al., 2015; Zhao et al., 2008). A lack of inactive, phosphorylated YAP was demonstrated in MVID patient tissues, correlated with their immature enterocyte phenotype (Kravtsov et al., 2016). In the present Lrig1^ΔMYO5B mice, several fetal stem markers were elevated in the epithelial progenitors and regeneration-promoting Igf1 and Nrg1 in fibroblasts (Bohin et al., 2020b; Lemmetyinen et al., 2023). These data suggest that active YAP signaling may be associated with the Wnt/Lgr5-independent proliferation and incomplete differentiation of MYO5B-deficient epithelial cells, possibly through a disruption in MYO5B/RAB11A-mediated YAP trafficking.

The fibroblasts underlying Lrig1^ΔMYO5B progenitor cells presented with remarkable alterations in various crypt niche-supporting factors, including R-spondins, Wnt ligands, Wnt inhibitors, BMPs, and growth factors. Although fibroblasts appear to have some Lrig1 expression, the effect on fibroblasts is likely indirect as the fibroblasts lack Myo5b expression. Until now, the involvement of fibroblasts in the crypt phenotype of epithelial progenitor-specific MYO5B deficiency has been unexplored. Future work can utilize the first scRNA-seq dataset generated in this study to further characterize the alterations in epithelial-stroma communication in MVID.

In agreement with the Vil1^ΔMYO5B models (Kaji et al., 2020; Schneeberger et al., 2015; Weis et al., 2016), the MYO5B-deficient small intestinal crypts are elongated in both Lrig1^ΔMYO5B and Olfm4^ΔMYO5B mice compared to controls. The hyperproliferative, MYO5B-deficient crypts had an increase in abnormal mitotic spindle formation, typically with monopolar spindles, compared to control crypt cells with bipolar mitotic spindles. Despite an increase in epithelial cells in proliferative cell cycle phases, our scRNA-seq data demonstrate similar cell cycle distributions between MYO5B-deficient and control epithelial progenitor cells. Moreover, a greater fraction of EdU+ cells remained proliferative 24 h post-labeling in MYO5B-deficient tissues compared to controls. These data suggest that, despite having mitotic spindle alterations, MYO5B progenitor cells can progress through the cell cycle but produce more proliferative daughter cells compared to controls. This phenotype likely contributes to the association of MYO5B loss with carcinogenesis in recent literature (Carton-Garcia et al., 2022; Dooley et al., 2025; Tomić et al., 2020).

Our recent report suggested the alterations in mitochondrial structure and cell metabolism in the Vil1^ΔMYO5B mice and MVID patient tissues (Momoh et al., 2024). In the present study, we determined direct alterations in progenitor cell metabolism at the crypt-specific MYO5B-deficient time point: decreases in FAO- and OXPHOS-related gene expression, an increase in disorganized and swollen mitochondria, and a crypt-specific decrease in FAO enzymatic activity, all independent of MVID-characteristic diarrheal phenotype due to MYO5B loss in the villi. Previous studies using mice lacking FAO-related proteins (i.e., HNF4αγ) demonstrate crypt elongation (Chen et al., 2020). The mechanism of how MYO5B loss alters cellular metabolism remains unclear. The uptake of either glucose or fatty acid analog did not noticeably differ between control and Lrig1^ΔMYO5B crypt cells, likely depending on MYO5B-independent basolateral membrane transporters, excluding a possibility of mitochondrial starvation. The mitochondrial alterations in MYO5B-deficient progenitor cells are likely indirect and downstream of the disruption of previously characterized interactions of myosin V with mitochondrial inactivation protein, BCL-2-modifying factor (BMF), and/or with RAB proteins that show direct interactions with mitochondria (Hausmann et al., 2011; Joseph et al., 2025; Landry et al., 2014; Puthalakath et al., 2001). Future work should investigate possible differences in the import of other nutrients to fuel proper cell metabolism, such as amino acids, and explore alterations in RAB-mitochondrial interactions in the MYO5B-deficient crypts.

In addition to proliferation, MYO5B deficiency also perturbed the other primary function of intestinal progenitor cells: differentiation. Our data revealed a decrease in the generation of tuft, EEC, and mature enterocytes derived from MYO5B-deficient progenitors in both Lrig1^ΔMYO5B and Olfm4^ΔMYO5B mice. MYO5B-deficient progenitors showed evident alterations in the transcriptional signatures and cell morphology of Paneth and goblet cells, indicating that MYO5B is critical for the differentiation of both secretory and absorptive cell lineages. Alterations in cell differentiation have been implicated in FAO-deficient mouse models (Chen et al., 2020; Mihaylova et al., 2018; Stine et al., 2019), suggesting that MYO5B loss-induced alterations in cell metabolism could contribute to these cell differentiation deficits.

Importantly, our MxIF analyses establish a distinction between MYO5B deficiency and systemic malnutrition. Unlike the day 5 Lrig1^ΔMYO5B or day 7 Olfm4^ΔMYO5B mice, 48-h-fasted MYO5B-intact tissues contained comparable cell populations to the control mouse small intestine. These observations corroborate other fasted mouse models, where short-term fasting leads to a decrease in intestinal epithelial proliferation (Mihaylova et al., 2018; Shay and Yilmaz, 2025), rather than the hyperproliferation seen in the MYO5B-deficient intestine. Similarly, the intestinal biopsies of several MVID patients under TPN demonstrate an elongated crypt phenotype, despite adequate systemic nutrition (Groisman et al., 2000). These data illustrate that the abnormal progenitor phenotype of our MVID mouse models is specific to MYO5B deficiency, and not a secondary effect of systemic malnutrition.

In conclusion, this study demonstrates the importance of MYO5B in epithelial progenitor cell identity, metabolism, and function using mouse models of MYO5B deficiency originating in the crypts of small intestine. These data suggest that future MVID therapeutics should focus on promoting proper intestinal progenitor function to correct the alterations in MVID-characteristic small intestinal epithelial architecture.

Methods

Mouse experiments

All animal work was conducted under the protocol approved by the Vanderbilt University Animal Care and Use Committee (M2000104). Mice were fed standard chow ad libitum and on a 12-h light/dark cycle. Lrig1-Cre^ERT2; Rosa26-LSL-YFP (Powell et al., 2012) and Olfm4-Cre^ERT2; Rosa26-LSL-tdTOM mice (Bohin et al., 2020a; Schuijers et al., 2014) were crossbred with Myo5b^fl/fl mice (Weis et al., 2016) to generate the Lrig1^ΔMYO5B and Olfm4^ΔMYO5B mice, respectively. The littermates of each colony were used as controls (Myo5b^fl/fl, Lrig1-Cre^ERT2; Rosa26-LSL-YFP; Myo5b^fl/+, or Olfm4-Cre^ERT2; Rosa26-LSL-tdTOM; Myo5b^fl/+). All strains were maintained on a C57BL/6 background. There were no obvious phenotypic differences between Cre-containing vs. Cre-negative controls prior to the tamoxifen treatment. Both male and female mice were included in experimentation. Cre recombinase was induced by a single injection of tamoxifen (2 mg; Sigma) administered intraperitoneally at 8–12 weeks of age. EdU (50 mg/kg) was administered intraperitoneally 24 h prior to tissue collection in some mice for proliferative cell tracing. For fasting experiments, tamoxifen-treated control mice (Myo5b^fl/fl) were placed in clean cages without food (with access to water) 48 h before euthanasia. The cages with clean bedding were replaced every 12 h to prevent coprophagy. Mice were euthanized, and intestinal tissues were isolated at different time points up to 5 days (Lrig1^ΔMYO5B) or 7 days (Olfm4^ΔMYO5B) following the tamoxifen administration. The intestinal region analyzed per experiment is specified in the figure legends. The jejunum, the major site of nutrient absorption, was primarily examined (Balen et al., 2008; Overduin et al., 2023). Mouse duodenal data were also collected, as patient biopsies are duodenal, and jejunal and duodenal data are presented separately where indicated. Ileal data were collected in select experiments to demonstrate that MYO5B expression and trafficking function were lost throughout the small intestine.

Human tissues

The formalin-fixed paraffin-embedded (FFPE) sections of duodenal biopsy were obtained from a Navajo MVID patient, possessing a homozygous mutation at MYO5B(P660L) (Patient 1), at 3 weeks of age in the Phoenix Children’s Hospital under IRB approval #10-019. The sections of duodenal biopsies from an MVID patient at 6.5 months of age, carrying a point mutation at MYO5B(G519R) (Burman et al., 2023) (Patient 2), and pediatric (13 years old) tissues free of abnormal pathology tissues were obtained from the PediCODE Consortium at the Vanderbilt under IRB#110835. Archived FFPE sections from healthy adult intestine were obtained from the Vanderbilt University Medical Center (IRB#171845).

Resource availability

Lead contact

Further information and requests for resources should be directed to the lead contact, Izumi Kaji (izumi.kaji@vumc.org).

Materials availability

This study did not generate new unique reagents.

Data and code availability

The accession number for the scRNA-seq data reported in this paper is GEO: GSE300032. Original code is available in Zenodo.

Acknowledgments

The Lrig1-CreERT2; Rosa26-LSL-YFP mouse was a kind gift from Drs. Won Jae Huh and Robert J. Coffey at the Vanderbilt University School of Medicine. We thank the Vanderbilt core facilities involved in this study: the Digital Histology Shared Resource (DHSR) for whole-slide scanning and MxIF analysis, the Translational Pathology Shared Resource (supported by the National Cancer Institute/NIH Cancer Center support grant P30 CA68485-19) for tissue processing and sectioning, Dr. Evan Krystofiak at the Cell Imaging Shared Resource for EM imaging, and the Vanderbilt Technologies for Advanced Genomics (VANTAGE) for next-generation sequencing. Core Services were performed through Vanderbilt University Medical Center’s Digestive Disease Research Center supported by NIH grant P30 DK058404. We also thank Jennifer Peek for QuPath software expertise, Alan J. Simmons and Dr. Paige N. Spencer for single-cell dissociation expertise, and Dr. Michaela Quintero for providing an organoid formation assay protocol. Cartoons are created with BioRender.com. This study was supported by the NIH R01 DK128190 and RC2 DK118640 to I.K., R01 DK103831 to K.S.L., and NSF GRFP nos. 193797963 and 2444112 to A.B. and no. 2444112 to M.E.B. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Author contributions

A.B. designed and performed and analyzed most experiments with assistance from M.M., F.A., and C.R. and wrote the first draft of the manuscript; M.E.B. conducted transcriptomic analysis and visualization; Y.Y. supported data analysis; K.S.L. performed bioinformatics validation; L.C.S. and M.D.S. generated essential materials; J.T.R. performed data curation and formal analysis; and I.K. conceived and supervised the research and finalized the draft of the manuscript.

Declaration of interests

The authors declare no competing interests.

Published: February 19, 2026