Alterations in cellular metabolic pathway and epithelial cell maturation induced by MYO5B defects are partially reversible by LPAR5 activation

Abstract

Functional loss of the motor protein myosin Vb (MYO5B) induces various defects in intestinal epithelial function and causes a congenital diarrheal disorder, namely, microvillus inclusion disease (MVID). Utilizing the MVID model mice Vil1-Cre^ERT2;Myo5b^flox/flox (MYO5BΔIEC) and Vil1-Cre^ERT2;Myo5b^flox/G519R [MYO5B(G519R)], we previously reported that functional MYO5B loss disrupts progenitor cell differentiation and enterocyte maturation that result in villus blunting and deadly malabsorption symptoms. In this study, we determined that both absence and a point mutation of MYO5B impair lipid metabolism and alter mitochondrial structure, which may underlie the progenitor cell malfunction observed in the MVID intestine. Along with a decrease in fatty acid oxidation, the lipogenesis pathway was enhanced in the MYO5BΔIEC small intestine. Consistent with these observations in vivo, RNA sequencing of enteroids generated from the two MVID mouse strains showed similar downregulation of energy metabolic enzymes, including mitochondrial oxidative phosphorylation genes. In our previous studies, we reported that lysophosphatidic acid (LPA) signaling ameliorated epithelial cell defects in MYO5BΔIEC tissues and enteroids. The present study demonstrated that the highly soluble LPA receptor (LPAR)5-preferred agonist Compound-1 improved sodium transporter localization and absorptive function and tuft cell differentiation in patient-modeled MVID animals that carry independent mutations in MYO5B. Body weight loss in male MYO5B(G519R) mice was ameliorated by Compound-1. These observations suggest that Compound-1 treatment has a trophic effect on the intestine with MYO5B functional loss through epithelial cell-autonomous pathways that can accelerate the differentiation of progenitor cells and the maturation of enterocytes. Targeting LPAR5 may represent an effective therapeutic approach for the treatment of MVID symptoms induced by different point mutations in MYO5B.NEW & NOTEWORTHY This study demonstrates the importance of MYO5B for cellular lipid metabolism and mitochondria in intestinal epithelial cells, previously unexplored functions of MYO5B. The alterations may underlie the progenitor cell malfunction observed in microvillus inclusion disease (MVID) intestines. To examine the therapeutic potential of progenitor-targeted treatments, the effects of the LPAR5-preferred agonist Compound-1 were investigated utilizing several MVID model mice and enteroids. Our observations suggest that Compound-1 may provide a therapeutic approach for treating MVID.

Keywords: enteroid; lysophosphatidic acid receptor; microvillus inclusion disease; mitochondria; mouse model.

Graphical abstract

Figure 1.

MYO5B loss negatively alters mitochondrial structures in the jejunal epithelial cells. A: TEM micrographs demonstrate the organized microvilli and electron-dense mitochondria (pink arrows) in subapical areas of healthy control mouse intestine. In MYO5BΔIEC mouse intestine, disorganized microvilli coincide with the accumulation of subapical autophagic vesicles (yellow arrows) and swollen mitochondria (pink arrows). B: immunostaining for a mitochondrial complex-I protein, NDUFB8, demonstrates the sporadic mitochondrial distribution in the MYO5BΔIEC mouse intestine compared with control tissues. Scale bars = 50 µm and 10 µm. MYO5B, myosin Vb; TEM, transmission electron microscopy.

Figure 2.

Disrupted mitochondria in MVID patient biopsy. Duodenal biopsies of healthy control and a patient with MVID who possesses a compound heterozygous mutation in MYO5B (7). Control tissues harbor dense mitochondria including organized cristae (pink arrows). In the tissues of the patient with MVID patient, a disorganized microvillus structure of enterocytes is associated with abnormal formation of mitochondria (pink arrows) and multivesicular bodies (yellow arrows). MVID, microvillus inclusion disease; MYO5B, myosin Vb.

Figure 3.

Jejunal tissue sections from fasted healthy mice. Four control mice were fasted for 48 h, and their epithelial phenotypes are compared with those in MYO5B-deficient mice. A: representative images of immunostaining for complex-I protein, NDUFB8. V = villus and C = crypt regions. B: Periodic Acid-Schiff (PAS)-staining shows established brush borders. C: immunostaining for a ketogenic enzyme, HMGCS2. D and E: DCLK1+ tuft cell frequency and apical localization of SGLT1 and CFTR are similar to those of fed control tissues. Scale bars = 50 µm and 10 µm (insets). MYO5B, myosin Vb.

Figure 4.

MYO5B loss and LPA treatment alters cellular energy metabolic pathways in the mouse small intestine. A: transcriptional analysis in intestinal epithelial cells indicates significantly altered energy metabolic pathways in MYO5B-deficient (MYO5BΔIEC) mice compared with control mice. LPA treatment on MYO5BΔIEC mice significantly increased the transcription of the rate-limiting ketogenic enzyme Hmgcs2 compared with vehicle-treated group. B: imaging mass spectrometry was performed in fresh tissues of jejunum from three mice of each genotype, and representative images are shown as heatmaps. Optical images show morphology of horizontal sections of intestine with luminal space in the center. Derivatives of long-chain fatty acids and cholesterol are accumulated in MYO5BΔIEC mouse tissues, particularly in the villi. C: immunostaining for HMGCS2 and Ki67 in mouse jejunum. Daily LPA treatment increases the expression of HMGCS2 in the jejunal epithelial cells of MYO5BΔIEC mice. Scale bars = 50 µm. D: relative mRNA expression of LPA receptor (Lpar) subtypes in intestinal epithelial cells isolated from control and MYO5BΔIEC mice. Mean ± SD and each datapoint represents each mouse. LPA, lysophosphatidic acid; MYO5B, myosin Vb. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by ANOVA with Tukey’s multiple-comparisons test. n = 3 mice per group.

Figure 5.

LPAR5 agonist, Compound-1, recapitulates the trophic effect of natural LPA. A: compound-1 structure and experimental design of mouse treatment. B: immunostaining for PCNA as a proliferative crypt marker, and ACTG1 for epithelial actin. C: villus/crypt ratio in jejunum of MYO5BΔIEC and MYO5B(G519R) mice. Bars indicate median values and each datapoint represents an average value of 10 regions in each mouse (N = 3–5 mice). D: immunostaining for NHE3 and LAMP1, a lysosomal marker, in the MYO5BΔIEC mouse jejunum. In vehicle-treated mice, NHE3 is diffuse in cytoplasm of epithelial cells or localized in lysosomes. Compound-1-treated mice demonstrate the separation of NHE3 from lysosomes and the reestablishment of the apical localization of NHE3. E: colocalization analysis of NHE3 and LAMP-1. Compound-1 significantly reduced the NHE3 accumulation in lysosomes. Mean ± SD and each datapoint represents the average of five regions in each mouse (N = 3–4 mice). Scale bar = 50 µm. LPAR5, lysophosphatidic acid receptor 5; MYO5B, myosin Vb. **P < 0.01, ***P < 0.001 by two-way ANOVA with least significant difference (LSD) test.

Figure 6.

Histology and proliferative cell population in crypts. A: H&E-stained jejunal and colonic tissue sections and immunostaining for proliferation markers. MVID model mice demonstrate expanded crypts and blunted villi in the jejunum and the dilation of colonic crypts. Compound-1 treatment improves villus morphology of MYO5BΔIEC mice, but does not remarkably alter MYO5B(G519R) villi. Mitotic marker, PHH3, and a proliferating marker, Ki67, in jejunal tissues demonstrate elongated crypt area in both MVID models. B: quantification of mitotic cells in the jejunum. PHH3+ nuclei per crypt were counted in 20 regions of each mouse section. Each datapoint on the graph indicates the averaged value of each mouse. N = 3–5 mice per group. No significant difference was detected by ANOVA. Scale bars = 50 µm. H&E, hematoxylin-eosin; MYO5B, myosin Vb; MVID, microvillus inclusion disease.

Figure 7.

Membrane SGLT1 localization is enhanced by Compound-1 treatment in MVID model mice. A: immunostaining for SGLT1 and LAMP1 in jejunum of MYO5BΔIEC and MYO5B(G519R) mice. B: colocalization analysis of SGLT1 and LAMP-1. *P < 0.05 by Mann–Whitney test. C: representative images of cell segmentation analysis in control tissue sections. D: histograms representing SGLT1 intensity ratios of all identified epithelial cells from jejunum of four mice in each group. Red dash lines indicate the mean values. Compound-1 treatment shifts the mean ratio in MVID model mice toward membrane localization. MYO5B, myosin Vb; MVID, microvillus inclusion disease.

Figure 8.

Functional distribution of mitochondria is partially rescued by Compound-1 treatment in MVID model mice. Jejunal sections of MYO5BΔIEC and MYO5B(G519R) mice were immunostained for NDUFB8 and villin after daily Compound-1 treatments. Control mouse tissues showed uniform dense brush border in villi and abundant subapical mitochondria close to the brush border. Following Compound-1 treatment, both MVID model mice demonstrated improved enterocytes that possess thick brush borders and intense mitochondrial signal in the subapical area (white squares). However, in other parts of the villi, enterocytes continued to show thin brush borders and diminished mitochondrial signal in the subapical area (yellow squares). Scale bars = 50 µm. MYO5B, myosin Vb; MVID, microvillus inclusion disease.

Figure 9.

Differential effect of Compound-1 treatment on body weight loss and epithelial transporter function. A: change in body weight on day 4 post tamoxifen compared with the original weight on day 0. Each datapoint indicates the value of each mouse. Among genotype and sex groups, only MYO5B(G519R) male mice show a significant improvement in body weight by Compound-1 treatment. *P < 0.05, **P < 0.01, ***P < 0.001 by two-way ANOVA with Tukey’s test. B: SGLT1 activity in Ussing chambered jejunum. Change in short-circuit current (I_sc) was measured in response to an SGLT1 inhibitor, phlorizin. Compound-1 treatment significantly increased SGLT1 function both in MYO5BΔIEC and MYO5B(G519R) mice. Circle datapoints indicate female mouse tissues and triangle datapoints indicate male mouse tissues. *P < 0.05, ****P < 0.0001 by two-way ANOVA with Tukey’s test. C–E: secretory responses in the jejunum. Chloride and water secretion was sequentially stimulated with carbachol (CCh; C) and forskolin (FK; D). CFTR dependency in secretory state was measured by a CFTR inhibitor (R)-BPO (E). There is no statistical difference between vehicle vs. Compound-1 treatment in each genotype. *P < 0.05, **P < 0.01 by two-way ANOVA with Tukey’s test.

Figure 10.

Lack of effect of Compound-1 on MUC13 localization in MVID model mice. Immunostaining for MUC13 and ACTG1 in jejunal tissues of control, MYO5BΔIEC, and MYO5B(G519R) mice treated with Compound-1 or vehicle. In control jejunum and distal colon, MUC13 (red) is localized to the apical tips of microvilli above actin filaments (green). Both MYO5B loss-of-function mouse tissues demonstrate strong MUC13 immunostaining in cytoplasm of jejunal and colonic epithelial cells. The mislocalization of MUC13 is not remarkably altered by Compound-1 treatments in MYO5BΔIEC or MYO5B(G519R) mice. Scale bars = 20 µm. MYO5B, myosin Vb; MVID, microvillus inclusion disease.

Figure 11.

LPAR5 activation ameliorates tuft cell differentiation in MVID models. A and B: immunostaining for DCLK1 in the jejunum of MYO5BΔIEC (A) and MYO5B(G519R) (B) mice treated with Compound-1 or vehicle. DCLK1+ tuft cells (magenta) do not express the proliferative marker, Ki67 (green). Counterstaining for ACTG1 and nuclei is shown in black as an inverted color. Scale bars = 50 µm. Tuft cell number per mucosal cell number is determined in whole jejunal Swiss rolls by digital image analysis. Each datapoint indicates a value from each mouse. **P < 0.01 by Kruskal–Wallis test with Dunn’s multiple comparisons. C: enteroids were generated from jejunal crypts of wild-type (WT) and genetically engineered pig with the point mutation MYO5B(P663L) and expanded in Human Organoid Growth Medium. Whole well images were analyzed to determine organoid forming efficacy and perimeters in wild-type (WT) and MVID models. Scale = 100 µm. **P < 0.01 by t test. *P < 0.05 by two-way ANOVA. D: whole mount immunostaining for tuft cell markers, pY1798-Girdin and POU2F3, and phalloidin staining for F-actin. E: differentiation medium was supplemented with Compound-1 (100 nM) or vehicle for 5 days. MYO5B(P663L) enteroids have less tuft cells compared with WT, and Compound-1 treatment significantly increased tuft cell differentiation. Each datapoint indicates each enteroid. MYO5B, myosin Vb; MVID, microvillus inclusion disease. ***P < 0.001, ****P < 0.0001 by ANOVA with Dunnett’s multiple-comparisons test.

Figure 12.

Transcription signatures in MYO5B-inactivated enteroids show the defects of cellular energy metabolism. A: summary of differential expression genes in mouse enteroids. Four individual cultures from four mice of each group are analyzed; iKO represents MYO5BΔIEC, iKO_C1 is Compound-1-treated MYO5BΔIEC, G519R is MYO5B(G519R), and Control is healthy littermates. B: selected transcription signatures in enteroids generated from MYO5BΔIEC (iKO) and MYO5B(G519R) mice. C: mitochondrial gene transcriptions including most OXPHOS metabolic genes were significantly decreased in both iKO and MYO5B(G519R) enteroids. D: MYO5B-deficient enteroids were treated with Compound-1 (100 nM) or vehicle for 2 days. Group 1 genes are significantly increased by MYO5B loss and decreased by Compound-1 treatment. Only Nox1 is decreased by MYO5B loss and reversed by Compound-1, whereas Group 3 genes are increased by MYO5B loss and further increased by Compound-1. N = 4 mice per each genotype. iKO, induced knockout; MYO5B, myosin Vb; OXPHOS, oxidative phosphorylation.

Figure 13.

Differential expression of genes in MYO5BΔIEC enteroids (A) and epithelial cells isolated from jejunal tissues (B) compared with each control sample. Mutual genes of interest that are up- or downregulated in MYO5B-deficient epithelial tissues and enteroids were marked as black dots. MYO5B, myosin Vb.