Early Evidence of Delayed Oligodendrocyte Maturation in The

RESEARCH ARTICLE Early evidence of delayed oligodendrocyte maturation in the mouse model of mucolipidosis type IV Molly Mepyans1,*, Livia Andrzejczuk2,*, Jahree Sosa2, Sierra Smith1, Shawn Herron1, Samantha DeRosa1, Susan A. Slaugenhaupt1, Albert Misko1, Yulia Grishchuk1,‡,§ and Kirill Kiselyov2,‡,§

ABSTRACT was reported in 1999 (Bargal et al., 2000; Raas-Rothschild et al., 1999; Mucolipidosis type IV (MLIV) is a lysosomal disease caused by Slaugenhaupt et al., 1999). MCOLN1 encodes the late endosomal/ mutations in the MCOLN1 gene that encodes the endolysosomal lysosomal ion channel TRPML1, which has a range of reported transient receptor potential channel mucolipin-1, or TRPML1. MLIV functions including zinc and iron homeostasis, regulation of lysosomal results in developmental delay, motor and cognitive impairments, and acidification and calcium release driving membrane fusion/fission vision loss. Brain abnormalities include thinning and malformation of events, as well as lysosomal biogenesis (Cheng et al., 2014; Dong et al., the corpus callosum, white-matter abnormalities, accumulation of 2008; Eichelsdoerfer et al., 2010; Kukic et al., 2013; Miao et al., 2015; undegraded intracellular ‘storage’ material and cerebellar atrophy in Samie et al., 2013; Soyombo et al., 2006; Venkatachalam et al., 2008; older patients. Identification of the early events in the MLIV course is key Wong et al., 2012; Xu and Ren, 2015; Xu et al., 2006; Zhang et al., to understanding the disease and deploying therapies. The Mcoln1−/− 2012). Abnormalities of membrane traffic, autophagy and mitochondria mouse model reproduces all major aspects of the human disease. We have been reported in various MLIV tissues and models (Coblentz et al., have previously reported hypomyelination in the MLIV mouse brain. 2014; Colletti et al., 2012; Jennings et al., 2006; Li et al., 2016; Miedel Here, we investigated the onset of hypomyelination and compared et al., 2008; Venkatachalam et al., 2008; Vergarajauregui et al., 2008; oligodendrocyte maturation between the cortex/forebrain and Wong et al., 2012; Xu et al., 2006; Zhang et al., 2016). cerebellum. We found significant delays in expression of mature MLIV patients typically present in the first year of life with oligodendrocyte markers Mag, Mbp and Mobp in the Mcoln1−/− cortex, psychomotor delay and visual impairment, which stems from a manifesting as early as 10 days after birth and persisting later in life. combination of corneal clouding, retinal dystrophy and optic atrophy Such delays were less pronounced in the cerebellum. Despite our (Abraham et al., 1985; Altarescu et al., 2002; Riedel et al., 1985; previous finding of diminished accumulation of the ferritin-bound iron in Schiffmann et al., 2014). Developmental gains can be appreciated the Mcoln1−/− brain, we report no significant changes in expression of during the first decade of life, although axial hypotonia, appendicular the cytosolic iron reporters, suggesting that iron-handling deficits in hypertonia and upper motor neuron weakness prevent ambulation and MLIV occur in the lysosomes and do not involve broad iron deficiency. limit fine motor function in the majority of patients. Although a static These data demonstrate very early deficits of oligodendrocyte course has been classically reported, progressive psychomotor decline maturation and critical regional differences in myelination between the has been documented. It begins in the second decade, eventually forebrain and cerebellum in the mouse model of MLIV. Furthermore, resulting in severe spastic quadriplegia (A.M., personal observations). they establish quantitative readouts of the MLIV impact on early brain In congruence with the clinical course, ancillary brain imaging has development, useful to gauge efficacy in pre-clinical trials. demonstrated stable white-matter abnormalities (corpus callosum hypoplasia/dysgenesis and white-matter lesions), with the emergence KEY WORDS: Mucolipidosis type IV, Lysosome, TRPML1, of subcortical volume loss and cerebellar atrophy in older patients. Mucolipin-1, Myelination, Oligodendrocyte Despite a 19-year-long effort in the MLIV natural history program (NCT00015782, NCT01067742), owing to the low number of patients INTRODUCTION and poor availability of the human brain tissue for research (single case Mucolipidosis type IV (MLIV) is an autosomal-recessive lysosomal tissues are now available at the University of Maryland Brain and storage disease caused by mutations in the MCOLN1 gene, located in Tissue Bank, Baltimore, MD, USA), the mechanism(s) of MLIV brain humans in the 19p13.2 locus (Acierno et al., 2001). The disease was disease is still not well understood. Pinpointing the early events in first described in 1974 (Berman et al., 1974) and the corresponding gene MLIV brain pathology is important to define the therapeutic window for intervention and set expectations for clinical trial outcomes. 1Center for Genomic Medicine and Department of Neurology, Massachusetts Some important insights regarding brain pathology in MLIV have General Hospital Research Institute and Harvard Medical School, Boston, been obtained from the MLIV mouse model [Mcoln1 knockout (KO) MA 02114, USA. 2Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA 15260, USA. mouse] developed by us (Grishchuk et al., 2014, 2016; Micsenyi *These authors contributed equally to this work et al., 2009; Venugopal et al., 2007). The mouse model shows all the ‡ These authors contributed equally to this work hallmarks of MLIV, including motor deficits, retinal degeneration §Authors for correspondence ([email protected]; [email protected]) and reduced secretion of chloric acid by stomach parietal cells, leading to elevated plasma gastrin (Schiffmann et al., 1998). These Y.G., 0000-0002-4684-4159; K.K., 0000-0001-6683-2895 hallmarks were also supported by the murine MLIV model generated

This is an Open Access article distributed under the terms of the Creative Commons Attribution by the Muallem group (Chandra et al., 2011). Similar to MLIV License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, patients, we observed intracellular storage formations, thinning of the distribution and reproduction in any medium provided that the original work is properly attributed. corpus callosum in the Mcoln1−/− mouse brain, micro- and Handling Editor: Steven J. Clapcote astrogliosis, and partial loss of Purkinje cells at the late stage of the

Received 27 January 2020; Accepted 16 June 2020 disease. Electron microscopy analysis shows reduced thickness of the Disease Models & Mechanisms

1 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 myelin sheaths consistent with central nervous system (CNS) hypomyelination (Grishchuk et al., 2014). Based on the existing natural history data in MLIV patients and our observations in the Mcoln1−/− mouse, MLIV can be thought of as having two distinct phases: the developmental phase, associated with early-onset delayed oligodendrocyte maturation and gliosis (Grishchuk et al., 2014; Weinstock et al., 2018), and the late degenerative phase, associated with partial loss of Purkinje cells in the MLIV mouse (Micsenyi et al., 2009) and mild cerebellar atrophy in some MLIV patients (Frei et al., 1998). Understanding the timing and mechanisms behind these distinct events is important for designing and testing therapies. Here, we investigate the role of TRPML1 in oligodendrocyte maturation and myelin deposition in the developing brain. Our data show early problems with oligodendrocyte maturation and myelination in higher-functioning brain regions such as the cerebral cortex. The role of TRPML1 in early brain development is further supported by the fact that its expression gradually rises in postnatal brain and expression is higher in the regions that display the most prominent deficits in myelination in TRPML1-deficient (Mcoln1−/−) mice (cerebral cortex). Finally, these data establish quantitative readouts of the TRPML1 impact on early brain development, which will be useful to gauge efficacy readouts in future therapy trials.

RESULTS Mcoln1/TRPML1expression increases with age in the postnatal brain Despite the emerging role of TRPML1 in a range of cellular functions, including cationic lysosomal homeostasis, cell repair (Colletti et al., 2012; Li et al., 2016; Miedel et al., 2008; Venkatachalam et al., 2008; Vergarajauregui et al., 2008; Wong et al., 2012; Xu et al., 2006; Zhang et al., 2016) and, more recently, aging and cancer (Jung et al., 2019; Xu et al., 2019), the time course of Mcoln1/TRPML1 expression during the development of living tissue has not been tracked. To answer how Mcoln1/TRPML1 is regulated during brain development we analyzed Fig. 1. Increasing Mcoln1 expression in the brain during development. its expression using quantitative reverse transcription PCR (qRT-PCR) (A,B) qPCR analysis of Mcoln1 mRNA in the whole forebrain (A) and cerebellum +/+ −/− in 1-, 10-, 21- and 60-day-old mice. Actin B (Actb) was used as a (B) in Mcoln1 (WT) and Mcoln1 (KO) mice, showing a significant increase with age. Analysis of expression was performed using ΔΔCt method; Actb was housekeeping gene. Primer specificity was confirmed using used as a housekeeping gene. Statistical analysis was performed using two-way −/− Mcoln1 samples. Our data show a steady increase in Mcoln1 ANOVA followed by Tukey’s multiple comparison tests. n=3-5; ***P<0.0005 and expression levels in the forebrain from birth to 2 months of age (Fig. 1). **P<0.005. (C) Mcoln1 expression dynamics in the cortical and cerebellum Mcoln1 mRNA levels increased 4-fold (4.45±0.56, n=3 or 4, P=0.003) samples of P10-P60 mice. The symbols represent an average of three to four in the whole-cerebrum samples of wild-type (WT) animals between experiments. Three to four biological replicates; **P<0.005 calculated using one- +/+ −/− postnatal day (P)1 and P10 (Fig. 1A). Focusing specifically on the way ANOVA. (D) Tpcn1 mRNA expression in the Mcoln1 and Mcoln1 isolated cortical tissues, we found a 6-fold increase in Mcoln1 mRNA cerebral cortex. n=3-5; *P<0.05 calculated using one-way ANOVA. ns, P>0.05; error bars represent s.e.m. levels between P10 and P60 (6.31±1.10, n=3, statistically significant linear trend with P=0.0028 using one-way ANOVA; Fig. 1C). By We compared its expression profile with that of another endolysosomal contrast, although Mcoln1 mRNA levels also increased in the cation channel, TPC1 (encoded by Tpcn1) (Yamaguchi et al., 2011). cerebellar samples between P1 and P10 (Fig. 1B) and P10 and P60 Interestingly, Tpcn1 mRNA did not increase with age in the mouse (Fig. 1C), the increases were small and did not attain statistical cerebral cortex (Fig. 1D). This, to our knowledge, is the first analysis of significance or significant linear trend. We found a similar trend in the Tpcn1 expression in the developing brain. Therefore, Mcoln1 mRNA Mcoln1 expression data generated by the Functional Annotation of upregulation during development is specific to this gene rather than a Mammalian Genome (FANTOM5) project (Ha et al., 2019; Lizio more general increase in expression of the lysosomal channels. The et al., 2015; Noguchi et al., 2017; Vitezic et al., 2014) (Fig. S1A). difference between Mcoln1 and Tpcn1 expression implies that Although the design of the study does not allow direct comparison of functions of the corresponding ion channels in the developing brain transcriptional activation of Mcoln1 between visual cortex and are distinct. The rapid pace of Mcoln1 mRNA increase in the cerebellum, the stable expression of Mcoln1 in cerebellum in the developing brain suggests its importance in development. pre- and early-postnatal phase and increased expression in postnatal Interestingly, there was a significant increase in the Tpcn1 mRNA visual cortex are consistent with the data we report here. Our Mcoln1 levels in the Mcoln1−/− samples at the P60 time point (7.83±2.27-fold expression data are the first on the developmental dynamics of Mcoln1 increase, n=3, statistically significant linear trend with P=0.047 using transcription that allow comparison between different brain regions, one-way ANOVA). RNA sequencing analysis on isolated brain cell cortex and cerebellum. types shows enrichment of the Tpcn1 transcripts specifically in Mcoln1-encoded endolysosomal cation channel TRPML1 is astrocytes, macrophage/microglia and endothelial cells (Zhang et al., −/− involved in lysosomal function and signaling (Xu and Ren, 2015). 2014); therefore, the increase in its expression in Mcoln1 brain Disease Models & Mechanisms

2 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 could reflect astrocytosis and microgliosis, a known prominent feature maturation, we used three of the key myelin genes: myelin-associated of brain pathology in MLIV (Grishchuk et al., 2014; Micsenyi et al., glycoprotein precursor (Mag), myelin basic protein (Mbp)andmyelin- 2009; Weinstock et al., 2018) (Fig. S1B). associated oligodendrocyte basic protein (Mobp) (Han et al., 2013; Huang et al., 2013; Pérez et al., 2013; Robinson et al., 2014). In the Early myelination deficits due to loss of Mcoln1/TRPML1 are forebrain samples, the three markers showed linear increase between caused by delayed oligodendrocyte maturation P1 and P21, allowing us to compare the dynamics of oligodendrocyte We have previously shown hypomyelination in the mouse model of maturation during brain development (Fig. 2). In both WT and KO MLIV (Grishchuk et al., 2015), suggesting oligodendrocyte samples, Mobp mRNA showed a statistically significant linear trend involvement. The original data, however, did not clarify whether the (P=0.0021 and P=0.0438, respectively, calculated using one-way hypomyelination is a result of a lower number of oligodendrocytes due ANOVA, n=3 or 4; Fig. 2A). At P21, Mobp mRNA increase was to loss of precursors or caused by their defective maturation. To gain significantly higher in the Mcoln1+/+ relative to Mcoln1−/− samples more insight, we performed an expression time course of (P=0.037 using unpaired Student’s t-test, n=5-7). This trend persisted oligodendrocyte maturation markers in postnatal development for Mbp (Fig. 2B). Interestingly, while Mag mRNA levels showed starting at birth to 2 months of age. To assess oligodendrocyte linear increase between P1 and P21, the rate of Mag mRNA increase

Fig. 2. Delayed oligodendrocyte maturation in the Mcoln1−/− cerebral cortex. (A-D) qPCR analysis of cerebrum (P1) and cortex (P10, P21 and P60) of Mcoln1+/+ and Mcoln1−/− littermate mice, showing expression of the mature oligodendrocyte markers Mobp, Mbp and Mag (A-C) and oligodendrocyte precursor marker Ng2 (D). The analysis was performed using the ΔΔCt method; Actb was used as a housekeeping gene. Slopes were obtained using simple linear regression, all of which were significant except WT in D. **P<0.005 and *P<0.05 using two-way ANOVA followed by Tukey’s multiple comparison tests; n=3-5; ns, P>0.05; error bars represent s.e.m. Disease Models & Mechanisms

3 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 was significantly lower than that of Mobp and Mbp mRNA and did not expression of Ng2 increased in the Mcoln1−/− cortices compared with differ significantly between Mcoln1+/+ relative to Mcoln1−/− samples Mcoln1+/+ cortices (Fig. 2D). Increase in Ng2 expression might reflect (Fig. 2C), suggesting that Mcoln1 loss actuates specific changes in the attempt to overcome the downstream myelination deficits during oligodendrocyte maturation rather than the loss of entire cell the active phase of brain myelination in postnatal development. populations. The relative abundance of Mbp and Mobp mRNA Therefore, we conclude that brain hypomyelination in MLIV is caused decreased between P21 and P60 (Fig. 2A,B), whereas levels of Mag by delayed oligodendrocyte maturation and reduced myelin production mRNA continued to grow (Fig. 2C). These data are compatible with rather than the death of the oligodendrocyte precursors. other published results (Ehmann et al., 2013). The observation that the Next, we used immunohistochemistry to analyze developmental expression of all three maturation markers was lower in the Mcoln1−/− myelination of the corpus callosum in 10-day-old Mcoln1+/+ and cortices compared with Mcoln1+/+ cortices demonstrates early Mcoln1−/− mice using antibodies for the mature oligodendrocyte myelination delays (Fig. 2). marker myelin proteolipid protein (PLP1) (Wang et al., 2006), one of To answer whether the observed delays in the acquisition of the the most abundant proteins in the myelin, and the neuronal mature oligodendrocyte gene expression in the cortex are the result of neurofilament marker SMI312 (Vasung et al., 2011) (Fig. 3). delayed oligodendrocyte maturation or caused by loss of Previously, we reported that thinning of the corpus callosum in 2- oligodendrocytes precursors, we analyzed the expression of the and 7-month-old Mcoln1−/− mice, while the thickness of the underlying oligodendrocyte precursor marker neural/glial antigen 2 (Ng2;also neuronal fibers in the corpus callosum remained intact, was caused by known as Cspg4) (Hill et al., 2014). Interestingly, we found that defective myelination (Grishchuk et al., 2015). Similar to these

Fig. 3. Reduced myelination of the corpus callosum in the MLIV mouse model is not accompanied by significant reduction in neuronal fiber thickness. (A) Representative images of Mcoln1+/+ and Mcoln1−/− P10 brain sections stained with the mature oligodendrocyte/ myelin marker PLP1 (red) and neuronal filament marker SMI312 (green). Corpus callosum and cortical areas are shown. Arrows show oligodendrocyte-deficient lesions in the Mcoln1−/− brain. Scale bar: 1 μm. (B) Analysis of the corpus callosum area identified via either PLP1 or SMI312 staining, showing reduction of the PLP1- positive but not SMA312-positive area and significant reduction of the PLP1- to SMI312- positive area ratio in Mcoln1−/− mice, indicating deficient myelination but normal underlying structure in the forming corpus callosum in the MLIV mouse. n=4 (Mcoln1+/+) and n=6 (Mcoln1−/−); ***P<0.001 using unpaired Student’s t-test; n.s, P>0.05; error bars represent s.e.m. (C) Enlarged representative images of the Mcoln1+/+ and Mcoln1−/− P10 brain sections immunostained for PLP1, showing dysmorphic myelinating fibers in the Mcoln1−/− brain. Scale bar: 200 nm. Disease Models & Mechanisms

4 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 observations made in adult mice, we here detect no significant decrease Transcriptional analysis of the iron-regulated genes failed in the SMI312-positive area of the corpus callosum in P10 pups, to reveal signs of iron depletion in the whole-cortical representing the normally forming tract of the callosal neuronal fibers, homogenates but observed a significant decrease in myelin deposition by mature Iron handling by the lysosomes has been shown to be impaired in oligodendrocytes, demonstrated by reduced PLP1-positive area alone Mcoln1/TRPML1-deficient cells and brain tissues (Coblentz et al., and decreased ratio of the PLP1-positive area to the corresponding 2014; Dong et al., 2008; Grishchuk et al., 2015). Despite the existing SMI312-positive area (Fig. 3A,B). Furthermore, we observed the evidence for iron involvement, the mechanism by which iron irregular distribution of PLP1-positive cells and forming myelinated mishandling impacts Mcoln1/TRPML1-deficient cells is unclear. fibers in the Mcoln1−/− cortex, the islands of tissue that are devoid of the Iron uptake in many cell types is facilitated through receptor- PLP1 staining (Fig. 3C and arrows in A), further supporting defective mediated endocytosis and therefore its late events are handled by the maturation of oligodendrocyte precursors in this model. endolysosomes (Todorich et al., 2009, 2011; Valerio, 2007). TRPML1 may be involved in this by transferring ferrous iron from Delayed maturation of oligodendrocytes in the Mcoln1−/− the lysosomal lumen to the cytoplasm. It is, therefore, possible that brain is region specific Mcoln1/TRPML1 loss traps ferrous iron in the lysosomes, depriving To answer whether the impact of Mcoln1/TRPML1 loss on the cytoplasm of iron and causing lysosomal oxidative stress. myelination is different in the cerebellum, we performed Importantly, cytoplasmic iron is key to oligodendrocyte maturation quantitative PCR (qPCR) analysis on the cerebellar samples from and myelin production as it is a co-factor in ATP production, which is P1-P60 mice. We found that the overall differences in expression of highly important due to extremely high energy requirements of the mature oligodendrocyte maturation markers in Mcoln1−/− and oligodendrocytes during myelination, and in the activity of key control Mcoln1+/+ cerebella were not as pronounced as they were in enzymes of lipid catabolism and biosynthesis required for myelin the cortical samples (Fig. 4A). This qPCR analysis was also production (Todorich et al., 2009). Iron retention in the lysosome due supported by immunohistochemistry, using which we detected no to loss of Mcoln1/TRPML1 might therefore directly affect difference in the percentage of PLP1-positive area, nor in the counts oligodendrocyte maturation by depriving the reactions of iron. of oligodendrocytes labeled by APC-CC1 (Fig. 4B,C), in Mcoln1−/− To answer whether Mcoln1/TRPML1 loss is associated with the cerebella compared with Mcoln1+/+ cerebella. Therefore, the impact depletion of the cytoplasmic iron, we analyzed the expression of the of Mcoln1/TRPML1 loss on oligodendrocyte maturation and iron-regulated genes, such as transferrin receptor (Tfrc), hephaestin myelination is brain-region specific; this should be taken into (Heph), ferritin heavy chain 1 (Fth1) and ferroportin (Fpn; also consideration when discussing the impact of MLIV and its treatment. known as Slc40a1), by qPCR. All these genes are regulated by

Fig. 4. Cerebellar myelination is preserved in MLIV mice. (A) qPCR analysis of P1, P10, P21 and P60 cerebella from Mcoln1+/+ (WT) and Mcoln1−/− (KO) littermate mice, showing expression of the mature oligodendrocyte markers Mobp, Mbp and Mag. The analysis was performed using the ΔΔCt method; Actb was used as a housekeeping gene. ns, P>0.05 using two-way ANOVA followed by Tukey’s multiple comparison test; n=3-5; error bars represent s.e.m. (B) Representative images of the Mcoln1+/+ and Mcoln1−/− P10 brain sections stained with the mature oligodendrocyte/myelin marker PLP1 (red), oligodendrocyte marker APC-CC1 (green) and nuclear counterstain Nuc-Blue (blue). Scale bars: 500 nm (left), 100 nm (right). (C) Statistical analysis of the APC- CC1-positive cell counts per section (left) and PLP1-positive area (right) in the cerebellar sections show no significant changes between Mcoln1+/+ and Mcoln1−/− − − mice. n.s, P>0.05 using unpaired Student’s t-test; n=4 (Mcoln1+/+) and n=5 (Mcoln1 / ). Disease Models & Mechanisms

5 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 cytoplastic iron through Hif1α, Hif2α and other transcription factors leading to brain hypomyelination, which is a characteristic of both the involved in metal handling and oxidative stress (Acikyol et al., MLIV mouse model and the human disease. A dramatically significant 2013; Li et al., 2013). qPCR showed no distinct difference between aspect of MLIV uncovered by these studies is the extremely early onset the WT and Mcoln1−/− samples (Fig. 5). We, therefore, conclude of the oligodendrocyte maturation delay, manifesting already by P10. that whole-tissue transcriptomic analysis reports no brain-wide Importantly, these early myelination deficits are not reflecting or cytoplasmic iron depletion. Owing to the relatively low abundance causing underlying structural abnormalities of neuronal fibers or their of oligodendroglial mRNA in the whole-tissue samples, however, loss. This is a key piece of information for designing MLIV treatments. we cannot rule out whether dysregulation of these genes in We have previously reported reduced myelination in Mcoln1−/− Mcoln1−/− animals takes place only in oligodendrocytes and may be mice via expression analysis of oligodendrocyte markers MAG, masked by the signal from more-abundant cell types. MBP, MOBP and platelet-derived growth factor receptor (PDGFR; also known as PDGFRB) in whole-brain homogenates at 10 days, DISCUSSION 2 months and 7 months of age (Grishchuk et al., 2015). We also In the present study, we show that loss of the lysosomal channel showed lower protein levels of the mature oligodendrocyte marker TRPML1 is associated with delayed oligodendrocyte maturation PLP1 in the Mcoln1−/− brain and reduced thickness of the

Fig. 5. Cytoplasmic iron markers in the MLIV mouse forebrain. (A-D) qPCR analysis of mRNA for cytoplasmic iron markers Tfrc (A), Heph (B), Fth1 (C) and Fpn (D) in the cerebrum (P1) and cortex (P10-P60) samples of WT (Mcoln1+/+) and Mcoln1−/− (KO) mice. The analysis was performed using the ΔΔCt method; Actb was used as a housekeeping gene. *P<0.05 using two-way ANOVA followed by Tukey multiple comparison test; n=3-5; ns, P>0.05; error bars represent s.e.m. Disease Models & Mechanisms

6 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230

PLP1-immunostained corpus callosum. Importantly, the 16 months of age have not been reported, largely due to the first neurofilament marker SMI312 remained unchanged in the early symptoms presenting around 6 months of age and diagnosis rarely symptomatic Mcoln1−/− mice (2 months of age), showing normally occurring before the first year. The delayed manifestation of motoric formed neuronal fibers in the corpus callosum. In the present study, dysfunction in MLIV patients might reflect selective myelination we tracked and compared oligodendrocyte maturation markers in the deficit in higher-cortical regions. This is also supported by the fact forebrain/cortex and cerebellum in Mcoln1−/− mice at 1, 10, 21 and that in the first months of life (prior to MLIV symptom onset) motor 60 days after birth. We found dramatic differences between function is mostly attributable to primitive brain regions, where, genotypes and brain regions. More specifically, while we report based on our mouse data, myelination is not affected by the loss of significant myelination deficits in the anterior regions (cortex, corpus Mcoln1/TRPML1. callosum) of the Mcoln1−/− brain, myelination in the posterior brain Our understanding of leukodystrophies, genetically determined (cerebellum) was normal. This regional and temporal pattern of disorders that selectively affect white matter in the CNS, has rapidly myelination deficits in the Mcoln1−/− mouse, along with the time evolved with the recent application of brain imaging techniques paired course of Mcoln1/TRPML1 expression in postnatal brain with next-generation sequencing (van der Knaap and Bugiani, 2017). development, provides new insights into the role of the TRPML1 Surprisingly, mutations in myelin- or oligodendrocyte-specific channel in the myelination process. It is well recognized that proteins are only responsible for a portion of clinically discernible myelination progresses from caudal to rostral pathways and in the leukodystrophies, while genes important for housekeeping processes order of increasing complexity of brain function (Gerber et al., 2010; in multiple cell linages are largely responsible for the remainder. As a Purger et al., 2016), starting from the areas responsible for basic result, the new extended classification of the leukodystrophies now homeostatic functions (pons, cerebellum) and proceeding to the areas includes more than 50 genetic syndromes with mutations in myelin- or controlling more complex tasks (frontal cortex). In the mouse, oligodendrocyte-specific genes (hypomyelinating and demyelinating deposition of myelin in the cerebellum starts in the deepest regions leukodystrophies; leukodystrophies with myelin vacuolization), and earlier than P6 and is largely accomplished by P23 (Purger et al., other white-matter components including axons (leuko-axonopathies), 2016). In the corpus callosum, the first myelinating cells are noticed glia (astrocytopathies and microgliopathies) and blood vessels on P9. In TRPML1-deficient brains, the regions that normally acquire (leukovasculopathies). We believe that the prominent and early myelination later in development are affected the most. Interestingly, involvement of developmental hypomyelination strongly advocates we show that, in parallel with brain myelination, expression of for MLIV to be included to this list. TRPML1 steadily increases in the postnatal brain and is more At the cellular level, TRPML1 has been shown to be involved in prominent in the cortex at the final stage of the myelination process, various cell functions including membrane fission and fusion, as well indicating its need for myelination of the higher-hierarchical regions. as the redistribution of ions between the lysosomes and the cytoplasm In corroboration with our findings in Mcoln1−/− KO mice, (Cheng et al., 2014; Dong et al., 2008; Eichelsdoerfer et al., 2010; specific abnormalities in cerebral white matter have been observed Kukic et al., 2013; Miao et al., 2015; Samie et al., 2013; Soyombo in MLIV patients. In a cross-sectional cohort analysis of 15 MLIV et al., 2006; Venkatachalam et al., 2008; Wong et al., 2012; Xu and patients ranging in age from 16 months to 22 years (Frei et al., Ren, 2015; Xu et al., 2006; Zhang et al., 2012). The latter effects were 1998), hypoplasia of the corpus callosum and elevated T2 signal in or can be attributed to TRPML1 iron conductance, or its role in the the subcortical white matter were reported in all patients, while traffic of iron or iron-bound proteins in the endocytic pathway. cortical atrophy was only noted in two of the oldest (15 and Impairment of TRPML1-dependent iron uptake is an attractive 22 years). A subsequent prospective study of brain magnetic hypothesis for the oligodendrocyte maturation deficits in the MLIV resonance imaging (MRI) changes in five MLIV patients, ranging brain, as cytoplasmic iron is a key co-factor of the myelin synthesis from 7 to 18 years of age, found increasing cortical gray matter and pathway (Todorich et al., 2009). Indeed, iron deficiency and decreasing subcortical white-matter volumes across a 2- to 3-year abnormal cellular iron absorption are associated with follow-up period (Schiffmann et al., 2014). Although additional hypomyelination (Morath and Mayer-Pröschel, 2001; Todorich longitudinal data in individual patients are still needed to support et al., 2009, 2011). Iron handling deficits have been reported by us conclusions, the available data in humans illustrate an early in the brain of the Mcoln1−/− mouse model (Grishchuk et al., 2015). developmental defect followed by a late degenerative process At that time, we were unable to provide a definitive answer regarding specifically affecting white-matter tracts. The possibility that axonal the impact of iron mishandling in the Mcoln1−/− brain. Here, we degeneration could accompany white matter loss in MLIV patients extended our studies using a battery of iron-reporter genes but found was raised by the finding of decreased N-acetylaspartate (NAA)/ no evidence of the cytoplasmic iron deficit in the Mcoln1−/− cortical creatine or choline ratios in multiple cortical regions (Bonavita et al., tissue homogenates. We, therefore, conclude that iron mishandling 2003; Cambron et al., 2012). Decreased NAA levels, however, can caused by Mcoln1/TRPML1 loss is either (1) limited to a specific be seen in the setting of neuronal mitochondrial dysfunction without subset of cells (such as mature oligodendrocytes) and being masked neuronal degeneration. Taken together with our current findings, we in the whole-tissue preparations, (2) taking place at certain propose that the neurological deficits observed in the Mcoln1 KO developmental stages, and/or (3) only manifesting as a lysosomal mice and MLIV patients mainly arise from abnormalities in CNS iron buildup without significantly affecting the cytoplasmic iron white matter. uptake and therefore unable to be detected via analysis of expression Our findings in Mcoln1−/− mice have important implications for of the iron-related genes. To this end, isolated Mcoln1−/− understanding the disease pathogenesis in MLIV patients, for whom oligodendrocyte cultures and co-cultures with neurons and glia will data on the earliest stages of myelination are not available. In regards be instrumental in delineating the mechanism by directly measuring to myelination, P10 in mice is roughly equivalent to the full-term cytosolic iron reporters, lysosomal iron content, oxidative stress and human neonate (Hagberg et al., 2002; Semple et al., 2013). While their impact on myelination, and will be the focus of our future study. corpus callosum hypoplasia/dysgenesis and subcortical white-matter In conclusion, we present evidence of the role of the TRPML1 abnormalities have consistently been reported in older individuals lysosomal channel in developmental myelination. Genetic ablation of

(Frei et al., 1998), brain MRI data for MLIV patients younger than TRPML1 leads to deficient oligodendrocyte maturation, resulting in Disease Models & Mechanisms

7 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230 hypomyelination of the cortex and corpus callosum already noticeable were compared between genotypes using GraphPad Prism 7.04 software at P10. However, mechanisms of how the loss of TRPML1 leads to (GraphPad, La Jolla, CA, USA) using unpaired Student’s t-test. Gaussian such deficits remain unknown. Brain myelination in development is a distribution was tested using the Shapiro–Wilk normality test. complex and highly regulated process that depends on both cell- intrinsic (oligodendroglial) and -extrinsic (neuronal, astrocytic and RNA extraction and qRT-PCR microglial) signaling and cues. Indeed, reduced myelination in the For qRT-PCR experiments, total RNA was extracted from whole cerebra or − − dissected cerebella from P1 pups. Whole-cerebella and cortical samples from early-postnatal Mcoln1 / cortex is accompanied by the observations the brains of 10-day-, 21-day- and 2-month-old mice were used. Cortical of robust activation of astrocytes, resulting in elevated cytokine/ samples were dissected to include motor and somatosensory areas and did not chemokine release and neuroinflammation (Weinstock et al., 2018). contain corpus callosum tissue. RNA was extracted using TRizol (Invitrogen, Therefore, better understanding of the role of TRPML1 in Carlsbad, CA, USA), according to the manufacturer’s protocol. All oligodendrocyte maturation and myelin deposition will require new experimental groups were designed to contain at least four animals per genetic in vivo (using the Cre-LoxP system for cell-type-specific genotype and time point; lower numbers in some conditions were due to low manipulation of Mcoln1 expression) and in vitro oligodendrocyte mRNA yields that prevented them being included in qRT-PCR experiments. and other brain cell co-culture models and this will be the scope of For complementary DNA (cDNA) synthesis, MuLV Reverse Transcriptase our future investigations. Elucidating the involvement of other (Applied Biosystems, Foster City, CA, USA) was used with 2 μg total RNA brain cell types, neurons and glia, affected by the loss of TRPML1 and oligo(dT)18 (IDT, Coralville, IA, USA) as a primer. qRT-PCR was carried out using 1:75 dilutions of cDNA, 2× SYBR Green/ROX Master Mix in the development of the oligodendrocyte lineage cells is important, (Fermentas, Glen Burnie, MD, USA) and 4 μM primer mix per 10 μl reaction. given the complex picture of brain pathology and dysfunction in For gene expression analysis, the following primers were used (IDT, this disease. Coralville, IA, USA): Actb, forward 5′-GCTCCGGCATGTGCAAAG-3′ and reverse 5′-CATCACACCCTGGTGCCTA-3′; Fpn,forward5′-ATCCTCT- MATERIALS AND METHODS GCGGAATCATCCT-3′ and reverse 5′-CAGACAGTAAGGACCCATCCA- Animals 3′; Fth1,forward5′-TGATCCCCACTTATGTGACTTCAT-3′ and reverse Mcoln1 KO mice were maintained and genotyped as previously described 5′-ACGTGGTCACCCAGTTCTTT-3′; Heph,forward5′-ACTAAAGAG- − (Venugopal et al., 2007). The Mcoln1+/ breeders for this study were ATTGGAAAAGCAGTGA-3′ and reverse 5′-TCATGCCCAGCATCTTC- obtained by backcrossing onto a C57Bl6J background for more than ten ACA-3′; Mag,forward5′-CGGGATTGTCACTGAGAGC-3′ and reverse 5′- generations. Mcoln1+/+ littermates were used as controls. Experiments were AGGTCCAGTTCTGGGGATTC-3′; Mbp,forward5′-GTGCCACATGTA- performed according to the institutional and National Institutes of Health CAAGGACT-3′ and reverse 5′-TGGGTTTTCATCTTGGGTCC-3′; Mobp, guidelines and approved by the Massachusetts General Hospital forward 5′-CACCCTTCACCTTCCTCAAC-3′ and reverse 5′-TTCTGGT- Institutional Animal Care and Use Committee. AAAAGCAACCGCT-3′; Ng2,forward5′-CATTTCTTCCGAGTGGTGG- C-3′ and reverse 5′-CACAGACTCTGGACAGACGG-3′; Tfrc,forward5′- Immunohistochemistry and image analysis GAGGCGCTTCCTAGTACTCC-3′ and reverse 5′-TGGTTCCCCACCAA- To obtain brain tissue for histological examination 1-, 10-, 21- and 60-day-old ACAAGT-3′. Mcoln1−/− and control mice were sacrificed using a carbon dioxide chamber. To avoid amplification of genomic DNA and ensure amplification of Immediately after, they were transcardially perfused with ice-cold phosphate- cDNA, all primers were designed to span exons and negative reverse buffered saline (PBS). Brains were post-fixed in 4% paraformaldehyde in transcription reactions (without reverse transcriptase) were performed as a PBS for 48 h, washed with PBS, cryoprotected in 30% sucrose in PBS for control. Samples were amplified in triplicate on a 7300 Real-Time System 24 h, split into cerebrum and cerebellum, frozen in isopentane and stored at (Applied Biosystems, Foster City, CA, USA) using the following program: −80°C. Brains were bisected along the midline and one hemisphere was 2 min at 50°C, 10 min at 95°C, and 40 cycles at 95°C for 15 s followed by examined histologically. Coronal sections of the cerebrum and sagittal 1 min at 60°C. In addition, a dissociation curve step was run to corroborate sections of the cerebellum were cut at 10 μm using a cryostat. Sections were that amplification with specific primers resulting in one product only. The mounted onto glass SuperFrost plus slides. These sections were stored at ΔΔCt method was used to calculate relative gene expression, where Ct −20°C before any staining procedures. For PLP1 and SMI312 staining, corresponds to the cycle threshold. Ct values for a representative set of genes sections were microwaved in citrate buffer (10 mM citric acid, 0.05% Tween are plotted in Fig. S2. ΔCt values were calculated as the difference between Ct 20, pH 6.0) at low power for 18 min for antigen retrieval, blocked in 0.5% values from the target gene and the housekeeping gene Actb.Experiments Triton X-100 and 5% normal goat serum (NGS) in PBS and incubated with were performed double blind using numerically coded samples; the sample primary antibodies diluted in 1% NGS overnight. The following primary identity was disclosed at the final stage of data analysis. Data are presented as antibodies were used: PLP1 rabbit polyclonal (1:500; ab28486) and SMI312 fold change. (pan axonal neurofilament) mouse monoclonal (1:1000; ab24574), both from Abcam (Cambridge, MA, USA). Sections were incubated with secondary Acknowledgements antibodies in 1% NGS for 1.5 h at room temperature. The following We thank Dr Franca Cambi for fruitful discussions. secondary antibodies were used: donkey anti-mouse AlexaFluor 488 and donkey anti-rabbit AlexaFluor 555 (1:500; Invitrogen, Eugene, OR, USA). Competing interests For APC-CC1 staining no antigen retrieval was performed. After blocking, The authors declare no competing or financial interests. sections were incubated with anti-APC-CC1 antibody (mouse monoclonal; 1:100; OP80, Calbiochem, Billerica, MA, USA), followed by goat anti- Author contributions mouse AlexaFluor 488 (1:500; Invitrogen). Sections were counterstained with Conceptualization: S.A.S., Y.G., K.K.; Methodology: S.H., Y.G., K.K.; Formal NucBlue nuclear stain (Life Technologies, Eugene, OR, USA). analysis: L.A., Y.G., K.K.; Investigation: M.M., L.A., J.S., S.S., S.H., S.D., Y.G., K.K.; Images were acquired on DM8i Leica Inverted Epifluorescence Data curation: Y.G., K.K.; Writing - original draft: A.M., Y.G., K.K.; Writing - review & editing: A.M., Y.G., K.K.; Supervision: S.A.S., Y.G., K.K.; Funding acquisition: Microscope with Adaptive Focus (Leica Microsystems, Buffalo Grove, S.A.S., Y.G., K.K. IL, USA) with Hamamatsu Flash 4.0 camera and advanced acquisition software package MetaMorph 4.2 (Molecular Devices, San Jose, CA, USA) Funding using an automated stitching function. The exposure time was kept constant This work was supported by the National Institutes of Health (R01NS096755). for all sections (WT and Mcoln1−/− sections within the same immunohistochemistry experiment). Image analysis was performed using Supplementary information Fiji software (National Institutes of Health, Bethesda, MD, USA). Areas of Supplementary information available online at interest were selected in each section. Area and mean pixel intensity values https://dmm.biologists.org/lookup/doi/10.1242/dmm.044230.supplemental Disease Models & Mechanisms

8 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230

References novel cerebellar developmental transcriptional regulators with motif activity Abraham, F. A., Brand, N., Blumenthal, M. and Merin, S. (1985). Retinal function analysis. BMC Genomics 20, 718. doi:10.1186/s12864-019-6063-9 in mucolipidosis IV. Ophthalmologica 191, 210-214. doi:10.1159/000309589 Hagberg, H., Ichord, R., Palmer, C., Yager, J. Y. and Vannucci, S. J. (2002). Acierno, J. S., Jr, Kennedy, J. C., Falardeau, J. L., Leyne, M., Bromley, M. C., Animal models of developmental brain injury: relevance to human disease. a Colman, M. W., Sun, M., Bove, C., Ashworth, L. K., Chadwick, L. H. et al. summary of the panel discussion from the third Hershey conference on (2001). A physical and transcript map of the MCOLN1 gene region on human developmental cerebral blood flow and metabolism. Dev. Neurosci. 24, chromosome 19p13.3-p13.2. Genomics 73, 203-210. doi:10.1006/geno.2001. 364-366. doi:10.1159/000069040 6526 Han, H., Myllykoski, M., Ruskamo, S., Wang, C. and Kursula, P. (2013). Myelin- Acikyol, B., Graham, R. M., Trinder, D., House, M. J., Olynyk, J. K., Scott, R. J., specific proteins: a structurally diverse group of membrane-interacting molecules. Milward, E. A. and Johnstone, D. M. (2013). Brain transcriptome perturbations in Biofactors 39, 233-241. doi:10.1002/biof.1076 the transferrin receptor 2 mutant mouse support the case for brain changes in iron Hill, R. A., Patel, K. D., Goncalves, C. M., Grutzendler, J. and Nishiyama, A. loading disorders, including effects relating to long-term depression and long-term (2014). Modulation of oligodendrocyte generation during a critical temporal window potentiation. Neuroscience 235, 119-128. doi:10.1016/j.neuroscience.2013.01.014 after NG2 cell division. Nat. Neurosci. 17, 1518-1527. doi:10.1038/nn.3815 Altarescu, G., Sun, M., Moore, D. F., Smith, J. A., Wiggs, E. A., Solomon, B. I., Huang, H., Zhao, X.-F., Zheng, K. and Qiu, M. (2013). Regulation of the timing of Patronas, N. J., Frei, K. P., Gupta, S., Kaneski, C. R. et al. (2002). The oligodendrocyte differentiation: mechanisms and perspectives. Neurosci. Bull. 29, neurogenetics of mucolipidosis type IV. Neurology 59, 306-313. doi:10.1212/ 155-164. doi:10.1007/s12264-013-1314-2 WNL.59.3.306 Jennings, J. J., Jr, Zhu, J.-H., Rbaibi, Y., Luo, X., Chu, C. T. and Kiselyov, K. Bargal, R., Avidan, N., Ben-Asher, E., Olender, Z., Zeigler, M., Frumkin, A., (2006). Mitochondrial aberrations in mucolipidosis Type IV. J. Biol. Chem. 281, Raas-Rothschild, A., Glusman, G., Lancet, D. and Bach, G. (2000). 39041-39050. doi:10.1074/jbc.M607982200 Identification of the gene causing mucolipidosis type IV. Nat. Genet. 26, Jung, J., Cho, K. J., Naji, A. K., Clemons, K. N., Wong, C. O., Villanueva, M., 118-123. doi:10.1038/79095 Gregory, S., Karagas, N. E., Tan, L., Liang, H. et al. (2019). HRAS-driven cancer Berman, E. R., Livni, N., Shapira, E., Merin, S. and Levij, I. S. (1974). Congenital cells are vulnerable to TRPML1 inhibition. EMBO Rep. 20, e46685. doi:10.15252/ corneal clouding with abnormal systemic storage bodies: a new variant of embr.201846685 mucolipidosis. J. Pediatr. 84, 519-526. doi:10.1016/S0022-3476(74)80671-2 Kukic, I., Lee, J. K., Coblentz, J., Kelleher, S. L. and Kiselyov, K. (2013). Zinc- Bonavita, S., Virta, A., Jeffries, N., Goldin, E., Tedeschi, G. and Schiffmann, R. dependent lysosomal enlargement in TRPML1-deficient cells involves MTF-1 (2003). Diffuse neuroaxonal involvement in mucolipidosis IV as assessed by transcription factor and ZnT4 (Slc30a4) transporter. Biochem. J. 451, 155-163. proton magnetic resonance spectroscopic imaging. J. Child Neurol. 18, 443-449. doi:10.1042/BJ20121506 doi:10.1177/08830738030180070701 Li, Y., Guan, Q., Chen, Y., Han, H., Liu, W. and Nie, Z. (2013). Transferrin receptor Cambron, M., D’Haeseleer, M., Laureys, G., Clinckers, R., Debruyne, J. and De and ferritin-H are developmentally regulated in oligodendrocyte lineage cells. Keyser, J. (2012). White-matter astrocytes, axonal energy metabolism, and Neural. Regen. Res. 8, 6-12. axonal degeneration in multiple sclerosis. J. Cereb. Blood Flow Metab. 32, Li, X., Rydzewski, N., Hider, A., Zhang, X., Yang, J., Wang, W., Gao, Q., Cheng, 413-424. doi:10.1038/jcbfm.2011.193 X. and Xu, H. (2016). A molecular mechanism to regulate lysosome motility for Chandra, M., Zhou, H., Li, Q., Muallem, S., Hofmann, S. L. and Soyombo, A. A. lysosome positioning and tubulation. Nat. Cell Biol. 18, 404-417. doi:10.1038/ (2011). A role for the Ca2+ channel TRPML1 in gastric acid secretion, based on ncb3324 analysis of knockout mice. Gastroenterology 140, 857-867. doi:10.1053/j.gastro. Lizio, M., Harshbarger, J., Shimoji, H., Severin, J., Kasukawa, T., Sahin, S., 2010.11.040 Abugessaisa, I., Fukuda, S., Hori, F., Ishikawa-Kato, S. et al. (2015). Gateways Cheng, X., Zhang, X., Gao, Q., Ali Samie, M., Azar, M., Tsang, W. L., Dong, L., to the FANTOM5 promoter level mammalian expression atlas. Genome Biol. 16, Sahoo, N., Li, X., Zhuo, Y. et al. (2014). The intracellular Ca(2)(+) channel 22. doi:10.1186/s13059-014-0560-6 MCOLN1 is required for sarcolemma repair to prevent muscular dystrophy. Nat. Miao, Y., Li, G., Zhang, X., Xu, H. and Abraham, S. N. (2015). A TRP channel Med. 20, 1187-1192. doi:10.1038/nm.3611 senses lysosome neutralization by pathogens to trigger their expulsion. Cell 161, Coblentz, J., St Croix, C. and Kiselyov, K. (2014). Loss of TRPML1 promotes 1306-1319. doi:10.1016/j.cell.2015.05.009 production of reactive oxygen species: is oxidative damage a factor in Micsenyi, M. C., Dobrenis, K., Stephney, G., Pickel, J., Vanier, M. T., mucolipidosis type IV? Biochem. J. 457, 361-368. doi:10.1042/BJ20130647 Slaugenhaupt, S. A. and Walkley, S. U. (2009). Neuropathology of the Colletti, G. A., Miedel, M. T., Quinn, J., Andharia, N., Weisz, O. A. and Kiselyov, Mcoln1(−/−) knockout mouse model of mucolipidosis type IV. J. Neuropathol. K. (2012). Loss of lysosomal ion channel transient receptor potential channel Exp. Neurol. 68, 125-135. doi:10.1097/NEN.0b013e3181942cf0 mucolipin-1 (TRPML1) leads to cathepsin B-dependent apoptosis. J. Biol. Chem. Miedel, M. T., Rbaibi, Y., Guerriero, C. J., Colletti, G., Weixel, K. M., Weisz, O. A. 287, 8082-8091. doi:10.1074/jbc.M111.285536 and Kiselyov, K. (2008). Membrane traffic and turnover in TRP-ML1-deficient Dong, X.-P., Cheng, X., Mills, E., Delling, M., Wang, F., Kurz, T. and Xu, H. (2008). cells: a revised model for mucolipidosis type IV pathogenesis. J. Exp. Med. 205, The type IV mucolipidosis-associated protein TRPML1 is an endolysosomal iron 1477-1490. doi:10.1084/jem.20072194 release channel. Nature 455, 992-996. doi:10.1038/nature07311 Morath, D. J. and Mayer-Pröschel, M. (2001). Iron modulates the differentiation of a Ehmann, H., Hartwich, H., Salzig, C., Hartmann, N., Clément-Ziza, M., Ushakov, distinct population of glial precursor cells into oligodendrocytes. Dev. Biol. 237, K., Avraham, K. B., Bininda-Emonds, O. R., Hartmann, A. K., Lang, P. et al. 232-243. doi:10.1006/dbio.2001.0352 (2013). Time-dependent gene expression analysis of the developing superior Noguchi, S., Arakawa, T., Fukuda, S., Furuno, M., Hasegawa, A., Hori, F., olivary complex. J. Biol. Chem. 288, 25865-25879. doi:10.1074/jbc.M113.490508 Ishikawa-Kato, S., Kaida, K., Kaiho, A., Kanamori-Katayama, M. et al. (2017). Eichelsdoerfer, J. L., Evans, J. A., Slaugenhaupt, S. A. and Cuajungco, M. P. FANTOM5 CAGE profiles of human and mouse samples. Sci. Data 4, 170112. (2010). Zinc dyshomeostasis is linked with the loss of mucolipidosis IV-associated doi:10.1038/sdata.2017.112 TRPML1 ion channel. J. Biol. Chem. 285, 34304-34308. doi:10.1074/jbc.C110. Pérez, M. J., Fernandez, N. and Pasquini, J. M. (2013). Oligodendrocyte 165480 differentiation and signaling after transferrin internalization: a mechanism of Frei, K. P., Patronas, N. J., Crutchfield, K. E., Altarescu, G. and Schiffmann, R. action. Exp. Neurol. 248, 262-274. doi:10.1016/j.expneurol.2013.06.014 (1998). Mucolipidosis type IV: characteristic MRI findings. Neurology 51, 565-569. Purger, D., Gibson, E. M. and Monje, M. (2016). Myelin plasticity in the central doi:10.1212/WNL.51.2.565 nervous system. Neuropharmacology 110, 563-573. doi:10.1016/j.neuropharm. Gerber, S., Tasdizen, T., Thomas Fletcher, P., Joshi, S., Whitaker, R. and 2015.08.001 Alzheimers Disease Neuroimaging, I. (2010). Manifold modeling for brain Raas-Rothschild, A., Bargal, R., DellaPergola, S., Zeigler, M. and Bach, G. population analysis. Med. Image Anal. 14, 643-653. doi:10.1016/j.media.2010.05. (1999). Mucolipidosis type IV: the origin of the disease in the Ashkenazi Jewish 008 population. Eur. J. Hum. Genet. 7, 496-498. doi:10.1038/sj.ejhg.5200277 Grishchuk, Y., Sri, S., Rudinskiy, N., Ma, W., Stember, K. G., Cottle, M. W., Sapp, Riedel, K. G., Zwaan, J., Kenyon, K. R., Kolodny, E. H., Hanninen, L. and Albert, E., Difiglia, M., Muzikansky, A., Betensky, R. A. et al. (2014). Behavioral D. M. (1985). Ocular abnormalities in mucolipidosis IV. Am. J. Ophthalmol. 99, deficits, early gliosis, dysmyelination and synaptic dysfunction in a mouse model 125-136. doi:10.1016/0002-9394(85)90220-X of mucolipidosis IV. Acta Neuropathol. Commun. 2, 133. doi:10.1186/s40478- Robinson, A. P., Rodgers, J. M., Goings, G. E. and Miller, S. D. (2014). 014-0133-7 Characterization of oligodendroglial populations in mouse demyelinating disease Grishchuk, Y., Peña, K. A., Coblentz, J., King, V. E., Humphrey, D. M., Wang, using flow cytometry: clues for MS pathogenesis. PLoS ONE 9, e107649. doi:10. S. L., Kiselyov, K. I. and Slaugenhaupt, S. A. (2015). Impaired myelination and 1371/journal.pone.0107649 reduced brain ferric iron in the mouse model of mucolipidosis IV. Dis. Model. Samie, M., Wang, X., Zhang, X., Goschka, A., Li, X., Cheng, X., Gregg, E., Azar, Mech. 8, 1591-1601. doi:10.1242/dmm.021154 M., Zhuo, Y., Garrity, A. G. et al. (2013). A TRP channel in the lysosome Grishchuk, Y., Stember, K. G., Matsunaga, A., Olivares, A. M., Cruz, N. M., King, regulates large particle phagocytosis via focal exocytosis. Dev. Cell 26, 511-524. V. E., Humphrey, D. M., Wang, S. L., Muzikansky, A., Betensky, R. A. et al. doi:10.1016/j.devcel.2013.08.003 (2016). Retinal dystrophy and optic nerve pathology in the mouse model of Schiffmann, R., Dwyer, N. K., Lubensky, I. A., Tsokos, M., Sutliff, V. E., Latimer, Mucolipidosis IV. Am. J. Pathol. 186, 199-209. doi:10.1016/j.ajpath.2015.09.017 J. S., Frei, K. P., Brady, R. O., Barton, N. W., Blanchette-Mackie, E. J. et al. Ha, T. J., Zhang, P. G. Y., Robert, R., Yeung, J., Swanson, D. J., Mathelier, A., (1998). Constitutive achlorhydria in mucolipidosis type IV. Proc. Natl. Acad. Sci.

Wasserman, W. W., Im, S., Itoh, M., Kawaji, H. et al. (2019). Identification of USA 95, 1207-1212. doi:10.1073/pnas.95.3.1207 Disease Models & Mechanisms

9 RESEARCH ARTICLE Disease Models & Mechanisms (2020) 13, dmm044230. doi:10.1242/dmm.044230

Schiffmann, R., Mayfield, J., Swift, C. and Nestrasil, I. (2014). Quantitative promoter regions of the genes implicated in Rett Syndrome. BMC Genomics 15, neuroimaging in mucolipidosis type IV. Mol. Genet. Metab. 111, 147-151. doi:10. 1177. doi:10.1186/1471-2164-15-1177 1016/j.ymgme.2013.11.007 Wang, S.-Z., Dulin, J., Wu, H., Hurlock, E., Lee, S. E., Jansson, K. and Lu, Q. R. Semple, B. D., Blomgren, K., Gimlin, K., Ferriero, D. M. and Noble-Haeusslein, (2006). An oligodendrocyte-specific zinc-finger transcription regulator cooperates L. J. (2013). Brain development in rodents and humans: Identifying benchmarks with Olig2 to promote oligodendrocyte differentiation. Development 133, of maturation and vulnerability to injury across species. Prog. Neurobiol. 106-107, 3389-3398. doi:10.1242/dev.02522 1-16. doi:10.1016/j.pneurobio.2013.04.001 Weinstock, L. D., Furness, A. M., Herron, S. S., Smith, S. S., Sankar, S. B., Slaugenhaupt, S. A., Acierno, J. S., Jr, Helbling, L. A., Bove, C., Goldin, E., DeRosa, S. G., Gao, D., Mepyans, M. E., Scotto Rosato, A., Medina, D. L. et al. Bach, G., Schiffmann, R. and Gusella, J. F. (1999). Mapping of the (2018). Fingolimod phosphate inhibits astrocyte inflammatory activity in mucolipidosis type IV gene to chromosome 19p and definition of founder mucolipidosis IV. Hum. Mol. Genet. 27, 2725-2738. doi:10.1093/hmg/ddy182 haplotypes. Am. J. Hum. Genet. 65, 773-778. doi:10.1086/302549 Wong, C.-O., Li, R., Montell, C. and Venkatachalam, K. (2012). Drosophila Soyombo, A. A., Tjon-Kon-Sang, S., Rbaibi, Y., Bashllari, E., Bisceglia, J., TRPML is required for TORC1 activation. Curr. Biol. 22, 1616-1621. doi:10.1016/j. Muallem, S. and Kiselyov, K. (2006). TRP-ML1 regulates lysosomal pH and acidic lysosomal lipid hydrolytic activity. J. Biol. Chem. 281, 7294-7301. doi:10. cub.2012.06.055 1074/jbc.M508211200 Xu, H. and Ren, D. (2015). Lysosomal physiology. Annu. Rev. Physiol. 77, 57-80. Todorich, B., Pasquini, J. M., Garcia, C. I., Paez, P. M. and Connor, J. R. (2009). doi:10.1146/annurev-physiol-021014-071649 Oligodendrocytes and myelination: the role of iron. Glia 57, 467-478. doi:10.1002/ Xu, S.-Z., Muraki, K., Zeng, F., Li, J., Sukumar, P., Shah, S., Dedman, A. M., glia.20784 Flemming, P. K., McHugh, D., Naylor, J. et al. (2006). A sphingosine-1- Todorich, B., Zhang, X. and Connor, J. R. (2011). H-ferritin is the major source of phosphate-activated calcium channel controlling vascular smooth muscle cell iron for oligodendrocytes. Glia 59, 927-935. doi:10.1002/glia.21164 motility. Circ. Res. 98, 1381-1389. doi:10.1161/01.RES.0000225284.36490.a2 Valerio, L. G. (2007). Mammalian iron metabolism. Toxicol. Mech. Methods 17, Xu, M., Almasi, S., Yang, Y., Yan, C., Sterea, A. M., Rizvi Syeda, A. K., Shen, B., 497-517. doi:10.1080/15376510701556690 Richard Derek, C., Huang, P., Gujar, S. et al. (2019). The lysosomal TRPML1 van der Knaap, M. S. and Bugiani, M. (2017). Leukodystrophies: a proposed channel regulates triple negative breast cancer development by promoting classification system based on pathological changes and pathogenetic mTORC1 and purinergic signaling pathways. Cell Calcium 79, 80-88. doi:10. mechanisms. Acta Neuropathol. 134, 351-382. doi:10.1007/s00401-017-1739-1 1016/j.ceca.2019.02.010 Vasung, L., Jovanov-Milošević, N., Pletikos, M., Mori, S., Judaš, M. and Yamaguchi, S., Jha, A., Li, Q., Soyombo, A. A., Dickinson, G. D., Churamani, D., Kostović,I.(2011). Prominent periventricular fiber system related to ganglionic Brailoiu, E., Patel, S. and Muallem, S. (2011). Transient receptor potential eminence and striatum in the human fetal cerebrum. Brain Struct. Funct. 215, mucolipin 1 (TRPML1) and two-pore channels are functionally independent 237-253. doi:10.1007/s00429-010-0279-4 organellar ion channels. J. Biol. Chem. 286, 22934-22942. doi:10.1074/jbc.M110. Venkatachalam, K., Long, A. A., Elsaesser, R., Nikolaeva, D., Broadie, K. and 210930 Montell, C. (2008). Motor deficit in a Drosophila model of mucolipidosis type IV Zhang, X., Li, X. and Xu, H. (2012). Phosphoinositide isoforms determine due to defective clearance of apoptotic cells. Cell 135, 838-851. doi:10.1016/j.cell. compartment-specific ion channel activity. Proc. Natl. Acad. Sci. USA 109, 2008.09.041 11384-11389. doi:10.1073/pnas.1202194109 Venugopal, B., Browning, M. F., Curcio-Morelli, C., Varro, A., Michaud, N., Zhang, Y., Chen, K., Sloan, S. A., Bennett, M. L., Scholze, A. R., O’Keeffe, S., Nanthakumar, N., Walkley, S. U., Pickel, J. and Slaugenhaupt, S. A. (2007). Neurologic, gastric, and opthalmologic pathologies in a murine model of Phatnani, H. P., Guarnieri, P., Caneda, C., Ruderisch, N. et al. (2014). An RNA- mucolipidosis type IV. Am. J. Hum. Genet. 81, 1070-1083. doi:10.1086/521954 sequencing transcriptome and splicing database of glia, neurons, and vascular Vergarajauregui, S., Connelly, P. S., Daniels, M. P. and Puertollano, R. (2008). cells of the cerebral cortex. J. Neurosci. 34, 11929-11947. doi:10.1523/ Autophagic dysfunction in mucolipidosis type IV patients. Hum. Mol. Genet. 17, JNEUROSCI.1860-14.2014 2723-2737. doi:10.1093/hmg/ddn174 Zhang, X., Cheng, X., Yu, L., Yang, J., Calvo, R., Patnaik, S., Hu, X., Gao, Q., Vitezic, M., Bertin, N., Andersson, R., Lipovich, L., Kawaji, H., Lassmann, T., Yang, M., Lawas, M. et al. (2016). MCOLN1 is a ROS sensor in lysosomes that Sandelin, A., Heutink, P., Goldowitz, D., Ha, T. et al. (2014). CAGE-defined regulates autophagy. Nat. Commun. 7, 12109. doi:10.1038/ncomms12109 Disease Models & Mechanisms