Long-term outcomes in patients with Burkitt lymphoma older than 65 years: an analysis of the Texas Cancer Registry.

Burkitt lymphoma (BL) is an extremely aggressive but curable subtype of non-Hodgkin lymphoma. While younger patients have excellent outcomes in response to aggressive chemoimmunotherapy, the rarity of this disease in older patients and limitations caused by age, comorbidities, and performance status may negate survival advantages. This analysis assessed outcomes of older adults with BL through data provided by the Texas Cancer Registry (TCR). Patients ≥65 years with BL were assessed. Patients were dichotomized into 1997-2007 and 2008-2018. Median overall survival (OS) and disease-specific survival (DSS) were assessed using Kaplan-Meier methodology, and covariates including age, race, sex, stage, primary site, and poverty index were analyzed using Pearson Chi-squared analysis. Odds ratio (OR) with 95% confidence intervals (CI) was used to assess factors contributing to patients not offered systemic therapy. P value <0.05 was considered statistically significant. Non-BL mortality events were also categorized. There were 325 adults, 167 in 1997-2007 and 158 in 2008-2018; 106 (63.5%) and 121 (76.6%) received systemic therapy, a trend that increased with time (p = 0.010). Median OS for 1997-2007 and 2008-2018 was 5 months (95% CI 2.469, 7.531) and 9 months (95% CI 0.000, 19.154) (p = 0.013), and DSS was 72 months (95% CI 56.397, 87.603) (p = 0.604) and not reached, respectively. For patients that received systemic therapy, median OS was 8 months (95% CI 1.278, 14.722) and 26 months (95% CI 5.824, 46.176) (p = 0.072), respectively, and DSS was 79 months (95% CI: 56.416, 101.584) and not reached, respectively (p = 0.607). Age ≥75 years (HR 1.39 [95% CI 1.078, 1.791], p = 0.011) and non-Hispanic whites (HR 1.407 [95% CI 1.024, 1.935], p = 0.035) had poorer outcomes, and patients at the 20-100% poverty index (OR 0.387 [95% CI 0.163, 0.921], p = 0.032) and increasing age at diagnosis (OR 0.947 [95% CI 0.913, 0.983], p = 0.004) were less likely to receive systemic therapy. Of 259 (79.7%) deaths, 62 (23.9%) were non-BL deaths, and 6 (9.6%) of these were from a second cancer. This two-decade analysis of older Texas patients with BL indicates a significant improvement in OS over time. Although patients were more likely to receive systemic therapy over time, treatment disparities existed in patients residing in poverty-stricken regions of Texas and in advancing age. These statewide findings reflect an unmet national need to find a systemic therapeutic strategy that can be tolerated by and augment outcomes in the growing elderly population.

SEL1L-HRD1 endoplasmic reticulum-associated degradation controls STING-mediated innate immunity by limiting the size of the activable STING pool.

Stimulator of interferon genes (STING) orchestrates the production of proinflammatory cytokines in response to cytosolic double-stranded DNA; however, the pathophysiological significance and molecular mechanism underlying the folding and maturation of nascent STING protein at the endoplasmic reticulum (ER) remain unknown. Here we report that the SEL1L-HRD1 protein complex-the most conserved branch of ER-associated degradation (ERAD)-is a negative regulator of the STING innate immunity by ubiquitinating and targeting nascent STING protein for proteasomal degradation in the basal state. SEL1L or HRD1 deficiency in macrophages specifically amplifies STING signalling and immunity against viral infection and tumour growth. Mechanistically, nascent STING protein is a bona fide substrate of SEL1L-HRD1 in the basal state, uncoupled from ER stress or its sensor inositol-requiring enzyme 1α. Hence, our study not only establishes a key role of SEL1L-HRD1 ERAD in innate immunity by limiting the size of the activable STING pool, but identifies a regulatory mechanism and therapeutic approach to targeting STING.

Long term outcomes in older patients with primary central nervous system lymphoma: an analysis of the Texas Cancer Registry.

Primary central nervous system lymphoma (PCNSL) is an aggressive subtype of non-Hodgkin lymphoma that carries a poor prognosis in the elderly. The aim of this study is to investigate treatment patterns and survival trends in patients ≥ 65 years with PCNSL through data provided by the Texas Cancer Registry. Adults ≥ 65 years diagnosed with PCNSL and followed between 1995-2017 were identified and separated into three eras: 1995-2003, 2004-2012, and 2013-2017. Baseline covariates compared included patient demographics and treatments administered. Pearson's chi-squared test and Cox proportional hazard models compared covariates; overall survival (OS) and disease-specific survival (DSS) were assessed via Kaplan-Meier methodology. There were 375 patients; 104 (27.7%) in 1995-2003, 146 (38.9%) in 2004-2012, and 125 (33.3%) in 2013-2017. There were 50 (48.1%), 55 (37.7%), and 31 (24.8%) in 1995-2003, 2004-2012, and 2013-2017, respectively, that did not receive treatment. At last follow up, 101 (97.1%), 130 (89.0%), and 94 (75.2%) in each era died, of which 89 (85.6%), 112 (76.7%), and 70 (56.0%) were attributed to PCNSL. Median OS per era was eight (95% confidence interval [CI] 5.06-10.93), six (95% CI, 2.30-9.69), and five months (95% CI, 2.26-7.73) (p = 0.638). DSS per era was nine (95% CI: 0.00, 26.53), 10 (95% CI: 5.14, 14.86), and 19 (95% CI, 0.00-45.49) (p = 0.931) months. Spinal cord as primary disease site (HR: 0.668 [95% CI, 0.45-0.99], p = 0.049), and chemotherapy (HR 0.532 [95% CI, 0.42-0.673], p = < 0.001) or chemotherapy + radiation (HR, 0.233 [95% CI, 0.11-0.48] p < 0.001) had better outcomes compared to no therapy or radiation therapy alone. Survival in older patients ≥ 65 with PCNSL has not improved per our analysis of the TCR from 1995-2017 despite increasing trends of treatment utilization. Strategies to augment recruitment of older individuals in trials are needed in order to determine who would derive treatment benefit and minimize treatment toxicities.

Loss of AID exacerbates the malignant progression of CLL.

Activation-induced cytidine deaminase (AID) has been implicated as both a positive and a negative factor in the progression of B cell chronic lymphocytic leukemia (CLL), but the role that it plays in the development and progression of this disease is still unclear. We generated an AID knockout CLL mouse model, AID-/-/Eµ-TCL1, and found that these mice die significantly earlier than their AID-proficient counterparts. AID-deficient CLL cells exhibit a higher ER stress response compared to Eµ-TCL1 controls, particularly through activation of the IRE1/XBP1s pathway. The increased production of secretory IgM in AID-deficient CLL cells contributes to their elevated expression levels of XBP1s, while secretory IgM-deficient CLL cells express less XBP1s. This increase in XBP1s in turn leads AID-deficient CLL cells to exhibit higher levels of B cell receptor signaling, supporting leukemic growth and survival. Further, AID-/-/Eµ-TCL1 CLL cells downregulate the tumor suppressive SMAD1/S1PR2 pathway and have altered homing to non-lymphoid organs. Notably, CLL cells from patients with IgHV-unmutated disease express higher levels of XBP1s mRNA compared to those from patients with IgHV-mutated CLL. Our studies thus reveal novel mechanisms by which the loss of AID leads to worsened CLL and may explain why unmutated CLL is more aggressive than mutated CLL.

IRE-1-Targeting Caged Prodrug with Endoplasmic Reticulum Stress-Inducing and XBP-1S-Inhibiting Activities for Cancer Therapy.

Activation of the IRE-1/XBP-1s pathway supports tumor progression. Here, we report a novel prodrug, TC-D-F07, in which a thiol-reactive dinitrobenzenesulfonyl (Dns) cage was installed onto the C8 hydroxyl of the covalent IRE-1 inhibitor D-F07. The electron-withdrawing Dns group in TC-D-F07 stabilizes the neighboring 1,3-dioxane acetal, allowing for stimulus-mediated control of its inhibitory activity. TC-D-F07 exhibits high sensitivity to intracellular thiols. Because tumor cells exhibit higher concentrations of glutathione and cysteine, treatment with TC-D-F07 results in more sustained levels of D-F07 in transformed versus normal cells. In addition, we show that a dinitrophenyl cysteine adduct resulting from cleavage of the Dns group induces endoplasmic reticulum (ER) stress, causing tumor cells to increase the expression of XBP-1s. The accumulated levels of D-F07 and its gradual decomposition into the active IRE-1 inhibitor eventually deprive tumor cells of XBP-1s, leading to more severe apoptosis than those treated with its uncaged analogue.

XBP-1s Promotes B Cell Pathogenicity in Chronic GVHD by Restraining the Activity of Regulated IRE-1α-Dependent Decay.

Allogeneic hematopoietic cell transplantation (allo-HCT) is an effective therapeutic procedure to treat hematological malignancies. However, the benefit of allo-HCT is limited by a major complication, chronic graft-versus-host disease (cGVHD). Since transmembrane and secretory proteins are generated and modified in the endoplasmic reticulum (ER), the ER stress response is of great importance to secretory cells including B cells. By using conditional knock-out (KO) of XBP-1, IRE-1α or both specifically on B cells, we demonstrated that the IRE-1α/XBP-1 pathway, one of the major ER stress response mediators, plays a critical role in B cell pathogenicity on the induction of cGVHD in murine models of allo-HCT. Endoribonuclease activity of IRE-1α activates XBP-1 signaling by converting unspliced XBP-1 (XBP-1u) mRNA into spliced XBP-1 (XBP-1s) mRNA but also cleaves other ER-associated mRNAs through regulated IRE-1α-dependent decay (RIDD). Further, ablation of XBP-1s production leads to unleashed activation of RIDD. Therefore, we hypothesized that RIDD plays an important role in B cells during cGVHD development. In this study, we found that the reduced pathogenicity of XBP-1 deficient B cells in cGVHD was reversed by RIDD restriction in IRE-1α kinase domain KO mice. Restraining RIDD activity per se in B cells resulted in an increased severity of cGVHD. Besides, inhibition of RIDD activity compromised B cell differentiation and led to dysregulated expression of MHC II and costimulatory molecules such as CD86, CD40, and ICOSL in B cells. Furthermore, restraining the RIDD activity without affecting XBP-1 splicing increased B cell ability to induce cGVHD after allo-HCT. These results suggest that RIDD is an important mediator for reducing cGVHD pathogenesis through targeting XBP-1s.

Targeting the IRE1α/XBP1s pathway suppresses CARM1-expressing ovarian cancer.

CARM1 is often overexpressed in human cancers including in ovarian cancer. However, therapeutic approaches based on CARM1 expression remain to be an unmet need. Cancer cells exploit adaptive responses such as the endoplasmic reticulum (ER) stress response for their survival through activating pathways such as the IRE1α/XBP1s pathway. Here, we report that CARM1-expressing ovarian cancer cells are selectively sensitive to inhibition of the IRE1α/XBP1s pathway. CARM1 regulates XBP1s target gene expression and directly interacts with XBP1s during ER stress response. Inhibition of the IRE1α/XBP1s pathway was effective against ovarian cancer in a CARM1-dependent manner both in vitro and in vivo in orthotopic and patient-derived xenograft models. In addition, IRE1α inhibitor B-I09 synergizes with immune checkpoint blockade anti-PD1 antibody in an immunocompetent CARM1-expressing ovarian cancer model. Our data show that pharmacological inhibition of the IRE1α/XBP1s pathway alone or in combination with immune checkpoint blockade represents a therapeutic strategy for CARM1-expressing cancers.

Targeting the IRE1α/XBP1 Endoplasmic Reticulum Stress Response Pathway in ARID1A-Mutant Ovarian Cancers.

The SWI/SNF chromatin-remodeling complex is frequently altered in human cancers. For example, the SWI/SNF component ARID1A is mutated in more than 50% of ovarian clear cell carcinomas (OCCC), for which effective treatments are lacking. Here, we report that ARID1A transcriptionally represses the IRE1α-XBP1 axis of the endoplasmic reticulum (ER) stress response, which confers sensitivity to inhibition of the IRE1α-XBP1 pathway in ARID1A-mutant OCCC. ARID1A mutational status correlated with response to inhibition of the IRE1α-XBP1 pathway. In a conditional Arid1aflox/flox/Pik3caH1047R genetic mouse model, Xbp1 knockout significantly improved survival of mice bearing OCCCs. Furthermore, the IRE1α inhibitor B-I09 suppressed the growth of ARID1A-inactivated OCCCs in vivo in orthotopic xenograft, patient-derived xenograft, and the genetic mouse models. Finally, B-I09 synergized with inhibition of HDAC6, a known regulator of the ER stress response, in suppressing the growth of ARID1A-inactivated OCCCs. These studies define the IRE1α-XBP1 axis of the ER stress response as a targetable vulnerability for ARID1A-mutant OCCCs, revealing a promising therapeutic approach for treating ARID1A-mutant ovarian cancers. SIGNIFICANCE: These findings indicate that pharmacological inhibition of the IRE1α-XBP1 pathway alone or in combination with HDAC6 inhibition represents an urgently needed therapeutic strategy for ARID1A-mutant ovarian cancers.

Deletion of mFICD AMPylase alters cytokine secretion and affects visual short-term learning in vivo.

Fic domain-containing AMP transferases (fic AMPylases) are conserved enzymes that catalyze the covalent transfer of AMP to proteins. This posttranslational modification regulates the function of several proteins, including the ER-resident chaperone Grp78/BiP. Here we introduce a mouse FICD (mFICD) AMPylase knockout mouse model to study fic AMPylase function in vertebrates. We find that mFICD deficiency is well tolerated in unstressed mice. We also show that mFICD-deficient mouse embryonic fibroblasts are depleted of AMPylated proteins. mFICD deletion alters protein synthesis and secretion in splenocytes, including that of IgM, an antibody secreted early during infections, and the proinflammatory cytokine IL-1ß, without affecting the unfolded protein response. Finally, we demonstrate that visual nonspatial short-term learning is stronger in old mFICD-/- mice than in wild-type controls while other measures of cognition, memory, and learning are unaffected. Together, our results suggest a role for mFICD in adaptive immunity and neuronal plasticity in vivo.

Distinct mechanisms govern populations of myeloid-derived suppressor cells in chronic viral infection and cancer.

Myeloid-derived suppressor cells (MDSCs) are major negative regulators of immune responses in cancer and chronic infections. It remains unclear if regulation of MDSC activity in different conditions is controlled by similar mechanisms. We compared MDSCs in mice with cancer and lymphocytic choriomeningitis virus (LCMV) infection. Chronic LCMV infection caused the development of monocytic MDSCs (M-MDSCs) but did not induce polymorphonuclear MDSCs (PMN-MDSCs). In contrast, both MDSC populations were present in cancer models. An acquisition of immune-suppressive activity by PMN-MDSCs in cancer was controlled by IRE1α and ATF6 pathways of the endoplasmic reticulum (ER) stress response. Abrogation of PMN-MDSC activity by blockade of the ER stress response resulted in an increase in tumor-specific immune response and reduced tumor progression. In contrast, the ER stress response was dispensable for suppressive activity of M-MDSCs in cancer and LCMV infection. Acquisition of immune-suppressive activity by M-MDSCs in spleens was mediated by IFN-γ signaling. However, it was dispensable for suppressive activity of M-MDSCs in tumor tissues. Suppressive activity of M-MDSCs in tumors was retained due to the effect of IL-6 present at high concentrations in the tumor site. These results demonstrate disease- and population-specific mechanisms of MDSC accumulation and the need for targeting different pathways to achieve inactivation of these cells.

An interdomain helix in IRE1α mediates the conformational change required for the sensor's activation.

The unfolded protein response plays an evolutionarily conserved role in homeostasis, and its dysregulation often leads to human disease, including diabetes and cancer. IRE1α is a major transducer that conveys endoplasmic reticulum stress via biochemical signals, yet major gaps persist in our understanding of how the detection of stress is converted to one of several molecular outcomes. It is known that, upon sensing unfolded proteins via its endoplasmic reticulum luminal domain, IRE1α dimerizes and then oligomerizes (often visualized as clustering). Once assembled, the kinase domain trans-autophosphorylates a neighboring IRE1α, inducing a conformational change that activates the RNase effector domain. However, the full details of how the signal is transmitted are not known. Here, we describe a previously unrecognized role for helix αK, located between the kinase and RNase domains of IRE1α, in conveying this critical conformational change. Using constructs containing mutations within this interdomain helix, we show that distinct substitutions affect oligomerization, kinase activity, and the RNase activity of IRE1α differentially. Furthermore, using both biochemical and computational methods, we found that different residues at position 827 specify distinct conformations at distal sites of the protein, such as in the RNase domain. Of importance, an RNase-inactive mutant, L827P, can still dimerize with wildtype monomers, but this mutation inactivates the wildtype molecule and renders leukemic cells more susceptible to stress. We surmise that helix αK is a conduit for the activation of IRE1α in response to stress.

Molecular Evaluation of Endoplasmic Reticulum Homeostasis Meets Humoral Immunity.

The biosynthesis of about one third of the human proteome, including membrane receptors and secreted proteins, occurs in the endoplasmic reticulum (ER). Conditions that perturb ER homeostasis activate the unfolded protein response (UPR). An 'optimistic' UPR output aims at restoring homeostasis by reinforcement of machineries that guarantee efficiency and fidelity of protein biogenesis in the ER. Yet, once the UPR 'deems' that ER homeostatic readjustment fails, it transitions to a 'pessimistic' output, which, depending on the cell type, will result in apoptosis. In this article, we discuss emerging concepts on how the UPR 'evaluates' ER stress, how the UPR is repurposed, in particular in B cells, and how UPR-driven counter-selection of cells undergoing homeostatic failure serves organismal homeostasis and humoral immunity.

Endoplasmic reticulum-associated degradation is required for nephrin maturation and kidney glomerular filtration function.

Podocytes are key to the glomerular filtration barrier by forming a slit diaphragm between interdigitating foot processes; however, the molecular details and functional importance of protein folding and degradation in the ER remain unknown. Here, we show that the SEL1L-HRD1 protein complex of ER-associated degradation (ERAD) is required for slit diaphragm formation and glomerular filtration function. SEL1L-HRD1 ERAD is highly expressed in podocytes of both mouse and human kidneys. Mice with podocyte-specific Sel1L deficiency develop podocytopathy and severe congenital nephrotic syndrome with an impaired slit diaphragm shortly after weaning and die prematurely, with a median lifespan of approximately 3 months. We show mechanistically that nephrin, a type 1 membrane protein causally linked to congenital nephrotic syndrome, is an endogenous ERAD substrate. ERAD deficiency attenuated the maturation of nascent nephrin, leading to its retention in the ER. We also show that various autosomal-recessive nephrin disease mutants were highly unstable and broken down by SEL1L-HRD1 ERAD, which attenuated the pathogenicity of the mutants toward the WT allele. This study uncovers a critical role of SEL1L-HRD1 ERAD in glomerular filtration barrier function and provides insights into the pathogenesis associated with autosomal-recessive disease mutants.

STING negatively regulates allogeneic T-cell responses by constraining antigen-presenting cell function.

Stimulator of interferon genes (STING)-mediated innate immune activation plays a key role in tumor- and self-DNA-elicited antitumor immunity and autoimmunity. However, STING can also suppress tumor immunity and autoimmunity. STING signaling in host nonhematopoietic cells was reported to either protect against or promote graft-versus-host disease (GVHD), a major complication of allogeneic hematopoietic cell transplantation (allo-HCT). Host hematopoietic antigen-presenting cells (APCs) play key roles in donor T-cell priming during GVHD initiation. However, how STING regulates host hematopoietic APCs after allo-HCT remains unknown. We utilized murine models of allo-HCT to assess the role of STING in hematopoietic APCs. STING-deficient recipients developed more severe GVHD after major histocompatibility complex-mismatched allo-HCT. Using bone marrow chimeras, we found that STING deficiency in host hematopoietic cells was primarily responsible for exacerbating the disease. Furthermore, STING on host CD11c+ cells played a dominant role in suppressing allogeneic T-cell responses. Mechanistically, STING deficiency resulted in increased survival, activation, and function of APCs, including macrophages and dendritic cells. Consistently, constitutive activation of STING attenuated the survival, activation, and function of APCs isolated from STING V154M knock-in mice. STING-deficient APCs augmented donor T-cell expansion, chemokine receptor expression, and migration into intestinal tissues, resulting in accelerated/exacerbated GVHD. Using pharmacologic approaches, we demonstrated that systemic administration of a STING agonist (bis-(3'-5')-cyclic dimeric guanosine monophosphate) to recipient mice before transplantation significantly reduced GVHD mortality. In conclusion, we revealed a novel role of STING in APC activity that dictates T-cell allogeneic responses and validated STING as a potential therapeutic target for controlling GVHD after allo-HCT.

STING regulates BCR signaling in normal and malignant B cells.

STING is an endoplasmic reticulum (ER)-resident protein critical for sensing cytoplasmic DNA and promoting the production of type I interferons; however, the role of STING in B cell receptor (BCR) signaling remains unclear. We generated STING V154M knock-in mice and showed that B cells carrying constitutively activated STING specifically degraded membrane-bound IgM, Igα, and Igß via SEL1L/HRD1-mediated ER-associated degradation (ERAD). B cells with activated STING were thus less capable of responding to BCR activation by phosphorylating Igα and Syk than those without activated STING. When immunized with T-independent antigens, STING V154M mice produced significantly fewer antigen-specific plasma cells and antibodies than immunized wild-type (WT) mice. We further generated B cell-specific STINGKO mice and showed that STINGKO B cells indeed responded to activation by transducing stronger BCR signals than their STING-proficient counterparts. When B cell-specific STINGKO mice were T-independently immunized, they produced significantly more antigen-specific plasma cells and antibodies than immunized STINGWT mice. Since both human and mouse IGHV-unmutated malignant chronic lymphocytic leukemia (CLL) cells downregulated the expression of STING, we explored whether STING downregulation could contribute to the well-established robust BCR signaling phenotype in malignant CLL cells. We generated a STING-deficient CLL mouse model and showed that STING-deficient CLL cells were indeed more responsive to BCR activation than their STING-proficient counterparts. These results revealed a novel B cell-intrinsic role of STING in negatively regulating BCR signaling in both normal and malignant B cells.

Development of Tumor-Targeting IRE-1 Inhibitors for B-cell Cancer Therapy.

The IRE-1 kinase/RNase splices the mRNA of the XBP-1 gene, resulting in the spliced XBP-1 (XBP-1s) mRNA that encodes the functional XBP-1s transcription factor that is critically important for the growth and survival of B-cell leukemia, lymphoma, and multiple myeloma (MM). Several inhibitors targeting the expression of XBP-1s have been reported; however, the cytotoxicity exerted by each inhibitor against cancer cells is highly variable. To design better therapeutic strategies for B-cell cancer, we systematically compared the ability of these compounds to inhibit the RNase activity of IRE-1 in vitro and to suppress the expression of XBP-1s in mouse and human MM cell lines. Tricyclic chromenone-based inhibitors B-I09 and D-F07, prodrugs harboring an aldehyde-masking group, emerged as the most reliable inhibitors for potent suppression of XBP-1s expression in MM cells. The cytotoxicity of B-I09 and D-F07 against MM as well as chronic lymphocytic leukemia and mantle cell lymphoma could be further enhanced by combination with inhibitors of the PI3K/AKT pathway. Because chemical modifications of the salicylaldehyde hydroxy group could be used to tune 1,3-dioxane prodrug stability, we installed reactive oxygen species-sensitive structural cage groups onto these inhibitors to achieve stimuli-responsive activities and improve tumor-targeting efficiency.

Sel1L-Hrd1 ER-associated degradation maintains ß cell identity via TGF-ß signaling.

ß Cell apoptosis and dedifferentiation are 2 hotly debated mechanisms underlying ß cell loss in type 2 diabetes; however, the molecular drivers underlying such events remain largely unclear. Here, we performed a side-by-side comparison of mice carrying ß cell-specific deletion of ER-associated degradation (ERAD) and autophagy. We reported that, while autophagy was necessary for ß cell survival, the highly conserved Sel1L-Hrd1 ERAD protein complex was required for the maintenance of ß cell maturation and identity. Using single-cell RNA-Seq, we demonstrated that Sel1L deficiency was not associated with ß cell loss, but rather loss of ß cell identity. Sel1L-Hrd1 ERAD controlled ß cell identity via TGF-ß signaling, in part by mediating the degradation of TGF-ß receptor 1. Inhibition of TGF-ß signaling in Sel1L-deficient ß cells augmented the expression of ß cell maturation markers and increased the total insulin content. Our data revealed distinct pathogenic effects of 2 major proteolytic pathways in ß cells, providing a framework for therapies targeting distinct mechanisms of protein quality control.

Mitochondrial lipid droplet formation as a detoxification mechanism to sequester and degrade excessive urothelial membranes.

The apical surface of the terminally differentiated mammalian urothelial umbrella cell is mechanically stable and highly impermeable, in part due to its coverage by urothelial plaques consisting of 2D crystals of uroplakin particles. The mechanism for regulating the uroplakin/plaque level is unclear. We found that genetic ablation of the highly tissue-specific sorting nexin Snx31, which localizes to plaques lining the multivesicular bodies (MVBs) in urothelial umbrella cells, abolishes MVBs suggesting that Snx31 plays a role in stabilizing the MVB-associated plaques by allowing them to achieve a greater curvature. Strikingly, Snx31 ablation also induces a massive accumulation of uroplakin-containing mitochondria-derived lipid droplets (LDs), which mediate uroplakin degradation via autophagy/lipophagy, leading to the loss of apical and fusiform vesicle plaques. These results suggest that MVBs play an active role in suppressing the excessive/wasteful endocytic degradation of uroplakins. Failure of this suppression mechanism triggers the formation of mitochondrial LDs so that excessive uroplakin membranes can be sequestered and degraded. Because mitochondrial LD formation, which occurs at a low level in normal urothelium, can also be induced by disturbance in uroplakin polymerization due to individual uroplakin knockout and by arsenite, a bladder carcinogen, this pathway may represent an inducible, versatile urothelial detoxification mechanism.