REVIEW published: 22 January 2021 doi: 10.3389/fncel.2020.627887 Megalencephalic Leukoencephalopathy: Insights Into Pathophysiology and Perspectives for Therapy Assumpció Bosch 1,2,3 * and Raúl Estévez 4,5 * 1 Department of Biochemistry and Molecular Biology, Institute of Neurosciences, Univ. Autònoma de Barcelona, Barcelona, Spain, 2 Unitat Mixta UAB-VHIR, Vall d’Hebron Institut de Recerca (VHIR), Barcelona, Spain, 3 Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Instituto de Salud Carlos III, Madrid, Spain, 4 Departament de Ciències Fisiològiques, IDIBELL—Institute of Neurosciences, Universitat de Barcelona, Barcelona, Spain, 5 Centro de Investigación Biomédica en Red sobre Enfermedades Raras (CIBERER), Instituto de Salud Carlos III, Madrid, Spain Megalencephalic leukoencephalopathy with subcortical cysts (MLC) is a rare genetic Edited by: disorder belonging to the group of vacuolating leukodystrophies. It is characterized Matthias Klugmann, University of New South Wales, by megalencephaly, loss of motor functions, epilepsy, and mild mental decline. In Australia brain biopsies of MLC patients, vacuoles were observed in myelin and in astrocytes Reviewed by: surrounding blood vessels. It is mainly caused by recessive mutations in MLC1 and Arturo Ortega, HEPACAM (also called GLIALCAM) genes. These disease variants are called MLC1 and Instituto Politécnico Nacional de México (CINVESTAV), Mexico MLC2A with both types of patients sharing the same clinical phenotype. Besides, Rogier Min, dominant mutations in HEPACAM were also identified in a subtype of MLC patients Amsterdam University Medical (MLC2B) with a remitting phenotype. MLC1 and GlialCAM proteins form a complex Center, Netherlands mainly expressed in brain astrocytes at the gliovascular interface and in Bergmann glia at *Correspondence: Assumpció Bosch the cerebellum. Both proteins regulate several ion channels and transporters involved in assumpcio.bosch@uab.cat the control of ion and water fluxes in glial cells, either directly influencing their location and Raúl Estévez restevez@ub.edu function, or indirectly regulating associated signal transduction pathways. However, the MLC1/GLIALCAM complex function and the related pathological mechanisms leading to Specialty section: This article was submitted to MLC are still unknown. It has been hypothesized that, in MLC, the role of glial cells in Cellular Neuropathology, brain ion homeostasis is altered in both physiological and inflammatory conditions. There a section of the journal is no therapy for MLC patients, only supportive treatment. As MLC2B patients show an Frontiers in Cellular Neuroscience MLC reversible phenotype, we speculated that the phenotype of MLC1 and MLC2A Received: 10 November 2020 patients could also be mitigated by the re-introduction of the correct gene even at later Accepted: 30 December 2020 stages. To prove this hypothesis, we injected in the cerebellar subarachnoid space of Published: 22 January 2021 Mlc1 knockout mice an adeno-associated virus (AAV) coding for human MLC1 under the Citation: Bosch A and Estévez R control of the glial-fibrillary acidic protein promoter. MLC1 expression in the cerebellum (2021) Megalencephalic extremely reduced myelin vacuolation at all ages in a dose-dependent manner. This study Leukoencephalopathy: Insights Into Pathophysiology and Perspectives could be considered as the first preclinical approach for MLC. We also suggest other for Therapy. potential therapeutic strategies in this review. Front. Cell. Neurosci. 14:627887. doi: 10.3389/fncel.2020.627887 Keywords: myelin abnormalities, ion channel, water homeostasis, chloride, cell-cell adhesion Frontiers in Cellular Neuroscience | www.frontiersin.org 1 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy INTRODUCTION between different cells (trans interactions; Elorza-Vidal et al., 2020). Most missense mutations found in MLC2A or MLC2B Leukodystrophies are human diseases affecting the central patients are located in the extracellular domain of the protein, nervous system (CNS) myelin. Although individual and affect protein localization at cell-cell junctions (López- leukodystrophies are rare, epidemiological data are Hernández et al., 2011a; Arnedo et al., 2014a,b; Elorza-Vidal indicating their high incidence. Within the leukodystrophies, et al., 2020). A recessive mutation in the signaling peptide of Megalencephalic leukoencephalopathy with subcortical cysts GlialCAM (p.L23H) abolishes protein expression in transfected (MLC; MIM 604004) is a vacuolating white matter disorder cells in vitro, suggesting that the lack of GlialCAM expression of infantile-onset (van der Knaap et al., 2012). Macrocephaly can also cause the disease (Arnedo et al., 2014b). Based on recent is observed in all affected individuals and develops during the biochemical and structural data, it seems that missense dominant first year of life. Most patients show delayed mental or motor MLC2B mutations specifically affect GlialCAM cis and trans development and seizures (Hamilton et al., 2018). Magnetic interaction surfaces (Elorza-Vidal et al., 2020). resonance imaging (MRI) is crucial to diagnose the disease in GlialCAM exerts two roles on MLC1: on one hand, it childhood, showing diffuse signal abnormality, swelling of the acts as an ER chaperone (Capdevila-Nortes et al., 2013) and, cerebral white matter, and the presence of subcortical cysts, on the other hand, it confines MLC1 at astrocyte-astrocyte which are mostly present in the anterior temporal areas and junctions (López-Hernández et al., 2011b). Thus, most MLC2A often in the frontoparietal region (van der Knaap et al., 1995). and MLC2B HEPACAM mutations affect GlialCAM targeting From this initial presentation, which is common to all and, consequently, also MLC1 location to cell-cell junctions. As MLC patients, two different phenotypes have been described: GlialCAM improves MLC1 folding, co-expression of GlialCAM a classical phenotype, which is found in the majority of with mutated MLC1 increases levels of MLC1 protein and its patients, and an improving phenotype, which shows a bettering membrane expression (Capdevila-Nortes et al., 2013). clinical course (van der Knaap et al., 2010). In the MLC classical phenotype, there is a gradual onset of ataxia, spasticity, EFFECT OF MLC MUTATIONS INTO and sometimes extrapyramidal findings; and late-onset of GLIALCAM AND MLC1 CELL BIOLOGY mild mental deterioration. In the improving phenotype, motor function ameliorates or normalizes; some patients Different cellular models have been used to express wild-type have stable intellectual disability, with or without autism MLC1 and GlialCAM or containing MLC mutations, such as (Hamilton et al., 2018). Xenopus oocytes (Teijido et al., 2004), Spodoptera frugiperda The classical MLC can be caused by two biallelic mutations (Sf9) cells (Ridder et al., 2011) and other cell lines such in the MLC1 or HEPACAM genes, as the disease is inherited as fibroblast HeLa (López-Hernández et al., 2011b), COS-1 in a recessive manner (Leegwater et al., 2001; López-Hernández (Hwang et al., 2019), HEK293 or glioblastoma U373 (Ambrosini et al., 2011a). Classical MLC patients with mutations in MLC1 et al., 2008) and human U251 astrocytoma (Lanciotti et al., are diagnosed as MLC1 type, whereas those with mutations 2020). Other studies have been performed in primary astrocyte in HEPACAM are named MLC2A. The improving phenotype, cultures from mice, rats, and humans (Duarri et al., 2011; on the other hand, is caused by heterozygous mutations in Ridder et al., 2011; Brignone et al., 2014). Finally, studies HEPACAM (López-Hernández et al., 2011a). These patients have also been performed in lymphoblast cell lines and are diagnosed as MLC2B type, and the disease is inherited human monocytes or monocyte-derived macrophages from dominantly. Mutations in MLC1 and HEPACAM are found in control and MLC patients (Duarri et al., 2008; Petrini et al., 76% and 22% of the patients, respectively. Pathogenic variants in 2013). Studies with human astrocytes derived from iPS MLC1 or HEPACAM have not been identified in the remaining cells would also be useful to gather more information on 2% of individuals showing clinical features of MLC, suggesting the disease. the existence of unknown additional genes of the disease From a pathological point of view, these studies in cell lines (van der Knaap et al., 1993). have defined the biochemical mechanism of how different MLC MLC1 encodes for an integral oligomeric membrane protein mutations affect different parameters such as protein stability, of unknown function expressed only in astrocytes, being one protein levels at the plasma membrane, and protein localization. of the most specific proteins for this cellular type (Teijido Thus, it is clear that most MLC1 missense mutations cause et al., 2004, 2007; Boor et al., 2005; Ambrosini et al., 2008). protein degradation. This has been verified in Xenopus oocytes Mutations found in MLC1 patients are distributed all over (Duarri et al., 2008), different cell lines (Duarri et al., 2008; the protein and they cause protein instability, which leads Brignone et al., 2014), primary astrocyte cultures (Duarri et al., to its degradation at the endoplasmic reticulum (ER) or in 2008), monocytes from patients (Duarri et al., 2008; Petrini lysosomes (Duarri et al., 2008; Lanciotti et al., 2010; Petrini et al., 2013), and even in a brain biopsy of an MLC1 patient et al., 2013). HEPACAM encodes for a single transmembrane (López-Hernández et al., 2011b; Figure 1A). Therefore, cells or type I protein (Moh et al., 2005), which works as an adhesion animals deficient in MLC1 constitute a model to understand molecule called HepaCAM or GlialCAM, the latter reflecting the pathophysiology of MLC. Animal models containing a its higher expression in glial cells (Barrallo-Gimeno et al., missense mutation in MLC1 could be used to screen for chemical 2015). GlialCAM can form homophilic interactions with other chaperones that could increase MLC1 protein stability and GlialCAM molecules within the same cell (cis interactions) and consequently, its levels at the plasma membrane. Frontiers in Cellular Neuroscience | www.frontiersin.org 2 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy with the reversibility of the phenotype observed in MLC2B patients in adult stages, a decreased GlialCAM/MLC1 function may not be so pathogenic in adults as it is in infants (Hamilton et al., 2018). Very few HEPACAM mutations seem to induce a gain of function, a phenotype that can only be observed in astrocytes from Mlc1 knockout mice in depolarizing conditions (Arnedo et al., 2014b). Further work is needed to understand the molecular mechanism of the pathogenesis of these specific mutations. INSIGHTS INTO GLIALCAM/MLC1 FUNCTION AND PATHOPHYSIOLOGY OF MLC It has nearly been 20 years since the discovery of the first gene implicated with the disease, MLC1 (Leegwater et al., 2001). However, we still do not have a clear picture of the function this protein exerts, and hence, the pathophysiology of the disease remains obscure (Estévez et al., 2018). First insights into the putative role of MLC1 came from protein sequence analysis (Leegwater et al., 2001). Thus, it was suggested that MLC1 could work as an ion channel or a transporter. However, no ion or transport activity directly mediated by MLC1 has been demonstrated so far, even when co-expressed with GlialCAM (Teijido et al., 2004). Based on the clinical phenotype of the patients, it was then speculated that FIGURE 1 | Effect of MLC1 (A) and GLIALCAM (B) mutations on the water/ion homeostasis should be negatively affected by the lack biochemistry and cell biology of their coding proteins. (A) Top panel: of MLC1 (van der Knaap et al., 2012). 2D-scheme of the MLC1 protein showing the residues that are mutated in To get more insights into the role of GlialCAM/MLC1, Megalencephalic leukoencephalopathy with subcortical cysts (MLC) patients. Below panel: mutation S69L found in a Spanish MLC1 patient abolished several research groups aimed to identify first the MLC1 and MLC1 protein expression. Reprinted by permission from Oxford University later GlialCAM interacting proteins. Different approaches were Press: Human Molecular Genetics, Molecular mechanisms of MLC1 and addressed such as yeast-to-hybrid, using MLC1 N-terminal GLIALCAM mutations in MLC, López-Hernández et al. (2011b). (B) 3D-Model domain (Brignone et al., 2011) or the full-length MLC1 of GlialCAM homo-oligomers forming cis and trans interactions. Reprinted by (unpublished results from our group), immunopurification of permission from Oxford University Press: Human Molecular Genetics, Structural basis for the dominant or recessive character of GLIALCAM MLC1 or GlialCAM associated complexes (López-Hernández mutations found in leukodystrophies, Elorza-Vidal et al. (2020). Expression of et al., 2011a; Jeworutzki et al., 2012; Sugio et al., 2017), GlialCAM containing a recessive (R92Q) or a dominant (R92W) mutation in co-fractionation (sWGA and DEAE chromatography, ouabain primary astrocytes causes mistargeting. chromatography; Ambrosini et al., 2008; Brignone et al., 2011, 2014), co-localization (by confocal or electronic microscopy immunogold) and by testing possible candidates (Lanciotti et al., On the other hand, most mutations in HEPACAM cause 2012). All these studies resulted in a list of proteins that might be protein mislocalization, as shown in cell lines and cultured involved in MLC (Table 1). astrocytes (Figure 1B). Biochemical studies also suggest that Classically, two approaches were followed to understand dominant MLC2B mutations affect the trafficking of the the role played by MLC and MLC-interacting proteins. One wild-type protein, possibly by interfering with the formation method compared the effects of overexpressing wild-type MLC1 of GlialCAM complexes in the same cell (cis-interactions) or (with or without GlialCAM) with the effects of overexpressing between different cells (trans-interactions; Elorza-Vidal et al., MLC1-containing mutations. However, an important caveat has 2020). These in vitro studies were validated in vivo with the to be stated in this method. Overexpression of Mlc1 is very toxic generation of a Glialcam knockin (KI) mouse containing a in mice (Sugio et al., 2017), leading to a more drastic vacuolation dominant mutation, which showed that GlialCAM protein also phenotype than in the Mlc1 knockout mice. Also, overexpression displays a trafficking defect in heterozygous mice (Hoegg-Beiler of a protein folding mutant may cause alterations in the ER and et al., 2014). With this model, it has been speculated that in lysosomes leading to additional effects (Duarri et al., 2008). dominant MLC2B mutations may lead to a minor reduction On the other hand, other groups explored the effects of partially of GlialCAM/MLC1 function, which might be pathogenic or totally depleting MLC proteins. Also, some studies have in humans only at the early stages of life. In agreement functionally analyzed the activity of possible interacting proteins Frontiers in Cellular Neuroscience | www.frontiersin.org 3 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy TABLE 1 | Proteins that have been related to the Megalencephalic leukoencephalopathy with subcortical cysts (MLC) proteins. Protein related Genetical, cell biology and Overexpression Reduction of expression Some comments with MLC biochemical evidence GlialCAM 1. Mutations in GLIALCAM Overexpression of GlialCAM 1. KO of Glialcam affect There is no doubt that GlialCAM is a cause MLC (López-Hernández stabilize MLC1 wild-type and MLC1 expression and localization obligate subunit of MLC1 in vivo. et al., 2011a). mutants (Capdevila-Nortes et al., (Hoegg-Beiler et al., 2014; Bugiani However, MLC1 overexpressed 2. GlialCAM interact with MLC1 2013). et al., 2017). in vitro arrive at the PM in vitro: co-IP, FRET, split-TEV 2. KO of Mlc1 affect GlialCAM independently of GlialCAM. (López-Hernández et al., localization (Hoegg-Beiler et al., 2011b). 2014; Bugiani et al., 2017). 3. Co-localization and co-IP in vivo (López-Hernández et al., 2011a). 4. GlialCAM change MLC1 localization (López-Hernández et al., 2011b). ClC-2 1. ClC-2 interact with GlialCAM Overexpression of ClC-2 and 1. KO of Mlc1 and Glialcam change The association between and MLC1 in vitro: coIP, GlialCAM change functional ClC-2 localization and functional GlialCAM/MLC1 and ClC-2 is not split-TEV, stabilization at the PM properties of ClC-2 (Jeworutzki properties (Hoegg-Beiler et al., constitutive (Sirisi et al., 2017). (Jeworutzki et al., 2012; et al., 2012, 2014). 2014; Bugiani et al., 2017). ClC-2 KO and/or mutations of Gaitán-Peñas et al., 2017). 2. In KO of ClC-2 there is a slight CLCN2 have a different phenotype 2. Co-localization and co-IP increase of MLC1 (Hoegg-Beiler than MLC (Depienne et al., 2013). in vivo (Jeworutzki et al., 2012). et al., 2014). ClC-2 dysfunction might contribute 3. ClC-2 localization change by to MLC but it is not the whole story GlialCAM (Jeworutzki et al., (Hoegg-Beiler et al., 2014). 2012). Na+ /K+ -ATPase 1. β1 subunit interact with Astrocytes from mice with 1. No change in activity were In vitro studies expressing these MLC1 N terminus by Yeast MLC1 OE show reduced observed in Mlc1 KO mice (Sugio proteins are needed to prove their two-hybrid and pull down Na+ /K+ -ATPase activity, but et al., 2017). interaction. More evidence in vivo in vitro (Brignone et al., 2011). increased ouabain binding (Sugio 2. No change in localization were that Na+ /K+ -ATPase activity is However, β1 was not co-IP et al., 2017). observed in MLC KO mice (Dubey regulated by MLC1 is necessary. in vivo (Sugio et al., 2017). et al., 2015; Bugiani et al., 2017). 2. α2 and α3 subunits were co-IP in vivo (Sugio et al., 2017). 3. Some colocalization in astrocytes (Brignone et al., 2011). 4. Ouabain chromatography purified MLC1 (Brignone et al., 2011). 5. Other proteins copurifying with Na+ /K+ -ATPase also are found in MLC1 IP (i.e., GLAST, GLT-1; Sugio et al., 2017). Other ATPases: 1. They have been identified by In the case of V-ATPase: OE of No studies have been In vitro studies expressing these PMCA1, 2, 3; coIP from brain tissue or by MLC1 influences endosomal pH performed in KO animals. proteins are needed to prove their SERCA1, 2; pulldown from astrocytes (Brignone et al., 2014). interaction. More evidence in vivo V-ATPase expressing His-tagged MLC1 that ATPase activity is regulated by (Brignone et al., 2014; Sugio MLC1 is necessary. et al., 2017). 2. Some co-localization have been demonstrated by IF in astrocytes with V-ATPase (Brignone et al., 2014). VRAC channel 1. No coIP in vivo (Elorza-Vidal OE MLC1 but not GlialCAM 1. Mlc1 KO or Mlc1/Glialcam It is not clear how VRAC activity is (LRRC8A) et al., 2018). increases VRAC activity and depletion reduces VRAC activity regulated by MLC1. It may be an 2. No colocalization in tissue improves RVD (regulatory volume and RVD in astrocytes indirect or a compensatory and in primary astrocytes decrease; Ridder et al., 2011; (Capdevila-Nortes et al., 2013; mechanism. More experimental (Elorza-Vidal et al., 2018). Lanciotti et al., 2012; Dubey et al., 2015; Elorza-Vidal evidence is needed. For instance, Capdevila-Nortes et al., 2013). et al., 2018). VRAC current should be measured in brain slices from MLC KO models. (Continued) Frontiers in Cellular Neuroscience | www.frontiersin.org 4 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy TABLE 1 | Continued Protein related Genetical, cell biology and Overexpression Reduction of expression Some comments with MLC biochemical evidence 2. Lymphoblast MLC cell lines show reduced VRAC activity (although MLC1 mRNA expression is nearly negligible in WT cells; Ridder et al., 2011). Cx43 1. coIP in vivo with GlialCAM 1. OE MLC1 in U251 cell line 1. No experiments have been Experiments in vivo in KO MLC (our unpublished results) and reduces ERK1/2 activity which performed in KO animals. models are needed. It seems in vitro in U373G cells (Wu phosphorylates 2. Cx43 is misslocalized in that MLC1 may influence et al., 2016). Cx43 increasing Cx43 in gap Glialcam KO astrocytes (our different proteins indirectly by 2. Partial co-localization with junctions (Lanciotti et al., 2020). unpublished results). influencing signaling through MLC1 and GlialCAM in primary 2. OE GlialCAM in U373G cell ERK1/2 (VRAC, Cx43; Lanciotti astrocytes (Duarri et al., 2011). line increase Cx43 stability (Wu et al., 2016). et al., 2016). TRPV4 1. It has been identified by 1. Astrocytes OE MLC1 activate 1. No experiments have been It is not clear how pulldown from astrocytes or calcium influx in response to performed in KO animals or TRPV4 activity is regulated by astrocytoma cells expressing hyposmosis and 4αPDD more astrocytes. MLC1. It may be an indirect or His-tagged MLC1 (Lanciotti efficiently (Lanciotti et al., 2012). 2. In macrophages from a compensatory mechanism. et al., 2012). 2. MLC1 also favors recycling of patients Ca2+ responses are More experimental evidence is 2. Ouabain chromatography TRPV4 (Lanciotti et al., 2012). also affected (Petrini et al., needed. purified also TRPV4 (Lanciotti 2013). et al., 2012). Kir4.1, AQP4 1. They have been identified by In vitro experiments in Xenopus In KO animals or in brain MLC In vitro studies expressing these co-fractionation, ouabain oocytes show that MLC1 does biposies there is a redistribution proteins are needed to prove chromatography or pulldown, not influence Kir4.1 channel of AQP4 and Kir4.1, which has their interaction. More evidence mainly in cell culture activity (Teijido et al., 2004). been interpreted as a in vivo that their activity is (Brignone et al., 2011). compensatory mechanism regulated by MLC1 is 2. AQP4 association seem to (Dubey et al., 2015; Wu et al., necessary. increase in hypoosmotic 2016; Bugiani et al., 2017). conditions (Lanciotti et al., 2012). 3. In vivo there is no colocalization by EM with these proteins in tissue (Duarri et al., 2011). Other proteins not 1. They include: syntrophin, No studies have been No changes in localization have 1. Some of these proteins may related with ionic dystrophin, caveolin-1, ZO-1. performed in OE mice or in been reported in Mlc1 or help anchor MLC1 and homeostasis 2. ZO-1 has been identified by transfected cells. Glialcam KO animals (Dubey GlialCAM at specific locations, coIP from brain. Co-localization et al., 2015; Bugiani et al., so it is logical that its has been found by EM in tissue 2017). localization is not affected in KO and IF in primary astrocytes animals. (Duarri et al., 2011). 2. In vivo, it is not clear that 3. GlialCAM, MLC1 and other MLC1 forms part of the DGC related proteins (i.e., TRPV4) complex, since there is no are found in caveolae (Lanciotti colocalization between proteins et al., 2010). of the DGC complex and MLC 4. Syntrophin and Dystrophin proteins. are found by co-fractionation and pulldown (Ambrosini et al., 2008). There is no colocalization with MLC1 by EM (Duarri et al., 2011). KO of Dystrophin, α-Dystrobrevin and Utrophin does not affect MLC1 localization (Duarri et al., 2011). We indicate the biochemical evidence, the functional effects after overexpression or inhibition, and a conclusion of the effects. in animal models of the disease (knockouts or overexpression) affected might be directly interacting with GlialCAM/MLC1 and/or cells obtained from patients. (i.e ClC-2, Cx43, Na+ /K+ -ATPase), whereas others might In general, we can conclude that many different transporters be regulated in an indirect manner (i.e., VRAC, TRPV4). or ion channels seem to be affected in MLC. Some of the proteins An interesting new hypothesis is that GlialCAM/MLC1 may Frontiers in Cellular Neuroscience | www.frontiersin.org 5 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy modulate signal transduction events, and therefore, they could simultaneously influence different proteins (Lanciotti et al., 2016; Elorza-Vidal et al., 2018; Brignone et al., 2019). However, it is difficult to conclude whether the functional changes observed in these proteins are a consequence of a compensatory mechanism or a direct effect of MLC proteins. Possibly, those effects that show a different sign of change in overexpressing vs. reduced expression and that directly interact with GlialCAM/MLC proteins are probably a direct consequence of GlialCAM/MLC1 functional role. We indicate the results obtained for different proteins in Table 1. An interesting issue is that some interactions might be constitutive, such as those between GlialCAM and MLC1 (Sirisi et al., 2014), whereas others might be regulated, such as the interaction between GlialCAM/MLC1 and ClC-2 (Sirisi et al., 2017). Hence, GlialCAM/MLC1 could be forming a protein scaffold, where different proteins may interact in a regulated manner with this protein FIGURE 2 | The vacuolating phenotype of GlialCAM and MLC1 in mice and network resulting in activity changes. As the interaction zebrafish resembles the MLC phenotype observed in humans. (A–C) MRI with ClC-2 was dependent on extracellular potassium, images of a human MLC1 patient (A), Mlc1 knockout mice (B), and mlc1 we proposed that GlialCAM/MLC1 might regulate brain knockout zebrafish (C). Red arrows indicate areas with increased water. homeostasis by influencing the activity and the interaction Reprinted by permission from Oxford University Press: Human Molecular with different transporters/ion channels that could be Genetics, Molecular mechanisms of MLC1 and GLIALCAM mutations in megalencephalic leukoencephalopathy with subcortical cysts, involved in controlling brain potassium levels (Estévez et al., López-Hernández et al. (2011b) and by permission from Oxford University 2018). Apart from myelin vacuoles, defects in potassium Press: Human Molecular Genetics, Megalencephalic leukoencephalopathy clearance and increased excitability were described in MLC with subcortical cysts protein 1 regulates glial surface localization of knockout mice and MLC patients (Dubey et al., 2018; GLIALCAM from fish to humans, Sirisi et al. (2014). (D) Histology studies of cerebellar brain white matter comparing wild-type (left) with Mlc1 KO mice Hamilton et al., 2018). indicate the presence of vacuoles in the KO mice. (E) EM images of vacuoles in Mlc1 KO mice. ANIMAL MODELS OF MLC: ZEBRAFISH AND MICE where myelination is more active and the onset of vacuolization. To understand the pathophysiology of MLC, several animal However, it takes much more time to be seen in mice (Hoegg- models have been generated. In mice, three different knockout Beiler et al., 2014; Dubey et al., 2015), and even longer in zebrafish lines of Mlc1 were produced (Hoegg-Beiler et al., 2014; Dubey (Sirisi et al., 2014). In line with these differences, expression et al., 2015; Sugio et al., 2017), two different knockouts for of MLC1 and GlialCAM is also higher in humans in the first Glialcam (Hoegg-Beiler et al., 2014; Bugiani et al., 2017), years of life and then is reduced and maintained (Dubey et al., several KI mice for Glialcam containing dominant or recessive 2015; Bugiani et al., 2017), whereas expression of both proteins mutations affecting GlialCAM targeting (Hoegg-Beiler et al., increases slowly in mice, reaching its peak at adult stage (Teijido 2014; Shi et al., 2019), double KO for Mlc1 and Glialcam (Pérez- et al., 2007; Gilbert et al., 2019). Rius et al., 2019), and a transgenic mouse that overexpress Mlc1 Furthermore, there are also regional differences in vacuolated in astrocytes (Sugio et al., 2017). It will be interesting to use areas. It is important to mention that in humans the edema Mlc1 floxed mice to generate conditional KO specifically in is mostly found in the subcortical white matter. While mice astrocytes during development, or KI mice for Mlc1 containing have a much more limited subcortical white matter, vacuoles are missense mutations as well as to obtain tissue-specific knockouts found in the cerebellum of KO mice, although MLC1 is broadly for GlialCAM. In zebrafish, an mlc1 KO line has been expressed at the endfeet of astrocytes in the whole brain (Hoegg- generated (Sirisi et al., 2014), with a mutation that abolishes Beiler et al., 2014). Thus, it can be speculated that MLC1 and mlc1 expression in vitro and in vivo; a glialcama KO and a double GlialCAM physiological roles might be more important in the KO for mlc1 and glialcama (Pérez-Rius et al., 2019). cerebellum than in the rest of the brain or that they might play All knockout animals share with human MLC patients an additional roles in cerebellar Bergmann glia. Several observations increase in the brain water content (Figure 2), which can be support these hypotheses: (1) in mice, expression of both proteins easily detected by MRI or histological techniques. As in biopsies in the cerebellum is higher than in the rest of the brain (Teijido of MLC patients (van der Knaap et al., 1996), vacuoles observed et al., 2007; López-Hernández et al., 2011a); (2) knockout of Mlc1 in animal models using electron microscopy are intramyelinic reduces total expression of Glialcam and ClC-2 in the cerebellum, (Figure 2). However, some differences exist between human and whereas no changes are observed in the cerebrum (Hoegg-Beiler animal models. MRI defects are observed in humans during the et al., 2014). Similarly, knockout of Glialcam decreases expression first years of life, thus there is a correlation between the period of ClC-2 in the cerebellum, but not in the rest of the brain. These Frontiers in Cellular Neuroscience | www.frontiersin.org 6 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy results are in agreement with the formation of a ternary complex GENE THERAPY FOR MLC between GlialCAM, MLC1, and ClC-2 that stabilizes ClC-2 in the cerebellum (Sirisi et al., 2017); (3) in a brain biopsy of an There are several major choices when developing a gene therapy MLC1 patient, lack of MLC1 causes GlialCAM mislocalization strategy for a disease like MLC, involving the CNS, which in Bergmann glia (Sirisi et al., 2014), but not in the rest of the can influence the efficiency of the treatment. This includes brain (López-Hernández et al., 2011b); and (4) overexpression the type of vector, its route of administration, the diffusion of Mlc1 causes a reduction in cerebellar size and Bergmann of the therapeutic protein if can be secreted, or the need for glia ectopia by an unknown mechanism in a development- a cell-specific promoter. Among the available gene delivery dependent manner (Kikuchihara et al., 2018). Thus, we postulate vectors, adeno-associated virus (AAV) vectors have several that GlialCAM/MLC1 may have more importance or additional features that make them appropriate for brain gene transfer and roles in the cerebellum, and this is perhaps more relevant in mice. one of the most promising for human trials: (i) the capacity Different from MLC patients, the MLC mouse model for generating long-term gene expression, as long as 15 years in containing a deletion of the first two exons of Mlc1 (Hoegg- non-human primate CNS (Sehara et al., 2017) mostly as episomal Beiler et al., 2014) does not display any cognitive or motor form, thus, avoiding the risk for insertional mutagenesis reported deficits and it is considered a model for the early stages of with integrative vectors; (ii) the absence of toxicity associated the disease. In contrast, the Mlc1 KO where the full Mlc1 gene with wild type virus (AAV vectors are classified as Biosafety was replaced by GFP showed some motor defects and higher Level 1), combined with the absence of wild type viral genes; susceptibility to develop epileptic seizures after kainate insults and (iii) the ability to easily produce pure high-titer viruses in (Dubey et al., 2018). It remains to be determined whether these the laboratory. differences are due to the different targeting strategies and/or to Current technology allows us to efficiently target a localized genetic background. brain region, for instance, the substantia nigra in Parkinson’s One interesting result observed in animal models is the disease, in which case direct intraparenchymal injection is mutual dependence between GlialCAM and MLC1. Thus, needed. This brings us to another important decision: the in mice and zebrafish, lack of Mlc1 causes GlialCAM appropriate route of administration has to be approached for mislocalization but not the reduced expression (Hoegg- each disease. Direct administration of the vector to a reduced Beiler et al., 2014; Sirisi et al., 2014) and, in mice, lack of and specific area of the brain through intracranial injection is GlialCAM causes Mlc1 reduced expression and mislocalization only suitable when the therapeutic protein needs to be localized (Hoegg-Beiler et al., 2014). These results are in agreement with in a particular structure, limiting the diffusion of the vector. similar clinical phenotypes among MLC2A and MLC1 patients Intracranial administration is extremely invasive and there is (Hamilton et al., 2018). an upper limitation of dose and volume to avoid toxicity and Animal models have also been very useful in defining the inflammation, although it has the advantage of using lower vector biochemical relationships between different proteins related to volumes, thus facilitating vector production. GlialCAM/MLC1 cell biology. A clear example is a biochemical When the therapeutic protein can be secreted to the and functional interaction between GlialCAM/MLC1 and the extracellular space, cross-correction from a reduced number of chloride channel ClC-2. Lack of Mlc1 or Glialcam causes cells producing the protein can be enough to achieve therapeutic ClC-2 mislocalization (Hoegg-Beiler et al., 2014), which is effects in most cells, independently of the transduced cell completely abolished in Bergmann glia and strongly diminished type. This is the case of some lysosomal storage diseases like at the astrocytic endfeet in the whole brain (Dubey et al., mucopolysaccharidoses, where the lysosomal enzyme is partially 2015; Bugiani et al., 2017). Furthermore, in vivo measurements secreted and recaptured through the mannose-6-phosphate of ClC-2 activity in cerebellar slices appeared linear in receptor. However, in the case of MLC disease, as MLC1 and oligodendrocytes from wild-type mice, and inwardly rectifying GlialCAM proteins are located in the cellular membrane of from Glialcam or Mlc1 knockout mice (Hoegg-Beiler et al., astrocytes, we might need to transduce as many astrocytes 2014). Other possible functional and biochemical interactions as possible. such as the VRAC channel (containing LRRC8A), Na+ /K+ - Indeed, many neurological disorders need a global ATPase, Kir4.1, AQP4 or TRPV4 have not been demonstrated distribution of the therapeutic protein to the whole brain in vivo, since its localization is preserved in MLC knockout or at least to large CNS structures. Moreover, some brain mice (Hoegg-Beiler et al., 2014; Dubey et al., 2015; Bugiani areas are more refractory to transduction by global routes of et al., 2017). Hence, these proteins might be indirectly administration, including the cerebellum, and need a specific regulated by unknown mechanisms or changes in their activity approach (Bosch et al., 2000). For the Mlc1 KO mouse, where could be a consequence of a compensatory mechanism. most histological abnormalities are present in the cerebellum, Thus, we can conclude that many different processes might we explored different delivery routes involving intravenous (IV), be affected by the lack of GlialCAM and MLC1. How CSF direct administration at lumbar or at the cerebellar area and GlialCAM and MLC1 regulate all these processes needs to directly into the cerebellar parenchyma, into the white mater or be elucidated to define novel therapeutic strategies. In this in the molecular layer (Figure 3). Ideally, IV administration of respect, we explored the strategy of gene therapy as an the vector would be the route of choice if a single injection could approach to restore normal MLC1 protein levels in vivo, as reach the whole brain, without using more interfering surgeries. detailed below. However, most vectors are not able to cross the blood-brain Frontiers in Cellular Neuroscience | www.frontiersin.org 7 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy FIGURE 3 | Representation of the different routes of administration to deliver viruses into the cerebellum. Lumbar intrathecal (IT), Intracerebroventricular (ICV), Intracerebellar (ICB), white matter (ICB WM), ICB molecular layer (ICB ML), and ICB subarachnoid space (ICB SB). barrier (BBB). Recently, the capacity of transducing the CNS achieved transduction in most of the cerebellar parenchyma, with through the circulation was shown for AAV9 and some derivates the highest efficacy in the Bergmann glia, where MLC1 is highly (like PHP.B; Foust et al., 2009; Deverman et al., 2016). However, expressed (Figure 4). Given the small volume in the cerebellar PHP.B efficiency was demonstrated in C57/bl6 mouse strain intrathecal cavity, the best diffusion was achieved by loading (Deverman et al., 2016) but not in Balb/c mice (Hordeaux et al., the virus at a very slow rate (Huda et al., 2014), which in turn 2018), and more importantly, the receptor used by this modified may decrease the amount of virus drained to the periphery AAV capsid is not found in non-human primates (Hordeaux (Wang et al., 2018). et al., 2019), which relegates the use of this vector to mouse It is well known that AAV vectors inherently transduce preclinical studies. Intravenous AAV9 administration, however, neuronal cells, so when neurons are not the target cell and to can reach global areas in the cerebrum where it transduces avoid undesirable secondary effects, it is important to restrict astrocytes and neurons, as well as in the spinal cord, and is currently being applied to some clinical trials, particularly for very young children, in which the BBB is still not well closed. Encouraging results for spinal muscular atrophy and Mucopolysaccharidosis type IIIA have been reported so far (NCT02122952 and NCT04088734 in www.clinicaltrials.gov, respectively; Mendell et al., 2017). But, to achieve therapeutic levels of the transgene into the CNS when reaching it from the circulation, high titers of viruses need to be infused into the patients (>1014 vg/kg), with the risk of adverse secondary effects, some of which may be related to liver and dorsal root ganglia toxicity, as were seen in NHP and piglets (Hinderer et al., 2018). Moreover, the scale-up of vector production for human clinical trials is also challenging due to the high virus load needed for each patient. Distribution of the vector through the cerebrospinal fluid (CSF) has many advantages, as allows for a more restricted expression, compared to the IV; but better global CNS transduction, compared to intraparenchymal administration; and it requires lower amounts of viruses, thus increasing biosafety and lowering production constraints (Pagès et al., 2019). Moreover, particularly for lumbar puncture, it can be performed in an outpatient setting, as is much less invasive than intracranial administration. In the MLC mouse model, FIGURE 4 | Biodistribution of AAVrh10 coding for GFP under the regulation the best results were obtained by delivering the vector into the of Glial-derived fibrillary acidic protein recombinant promoter (GFAP). CSF by occipital subarachnoid administration, achieving a high Cerebellar administration of the adeno-associated virus (AAV) vector in the percentage of transduction of the cerebellum without damaging subarachnoid space efficiently transduces the whole cerebellum. Scale bar, the brain parenchyma, allowing for a more localized and 400 µm. Reprinted by permission from Springer Nature: Springer, efficient action and avoiding non-specific effects due to systemic Neurotherapeutics, Cerebellar Astrocyte transduction as gene therapy for megalencephalic leukoencephalopathy, Sánchez et al. (2020). expression (Sánchez et al., 2020). By this route, we efficiently Frontiers in Cellular Neuroscience | www.frontiersin.org 8 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy FIGURE 5 | Gene therapy for MLC corrects cerebellar leukodystrophy and MLC1, GlialCAM, and ClC-2 protein complex. Mlc1 KO mice show myelin vacuolization in the cerebellum, as a result of lack of MLC1 and mislocalization of GlialCAM and ClC-2. AAV-mediated gene therapy restores MLC1 and allows the formation of the MLC1-GlialCAM-ClC-2 protein triad, which is enough to correct the vacuolating phenotype of this model. expression to the target cell type. This is possible by combining (Sánchez et al., 2020). However, wild-type animals injected with a particular AAV serotype with the control of the transgene MLC1 did not show any deleterious effect, contrary to what was expression by recombinant promoters that recruit specific host reported in the MLC1 over-expressing transgenic animal (Sugio cell-derived transcription factors (Lawlor et al., 2009). To design et al., 2017), reinforcing the safety of the strategy and indicating a gene therapy strategy for MLC, and taking into account that only supra-physiological levels of MLC1 may be toxic. that MLC1 is exclusively expressed by astrocytes in the brain, Indeed, western blot analysis detected considerably lower levels we choose an AAVrh10 vector (Sánchez et al., 2020), as it of MLC1 in treated KO than in wild type animals, suggesting that seems to transduce a larger number of glial cells in the adult a low amount of protein could still be therapeutic, as expected for brain, compared to AAV9 that showed a higher and specific a recessive disease. tropism for neurons (Petrosyan et al., 2014). AAVrh10 was We tested two different approaches, first a preventive strategy, used in combination with a glial-derived fibrillary acidic treating the animals before the beginning of the pathology protein recombinant promoter (GFAP), with high selectivity and analyzing them at 8 months of age, when the cerebellar for astrocytes in the mature mouse brain (Brenner et al., vacuolation is evident. In the Mlc1 KO mouse, increased 1994). Moreover, using a non-human primate serotype such as MLC1 expression in Bergmann glia corrected GlialCAM and AAVrh10, we may increase the possibilities to avoid preexisting ClC-2 location in this area, confirming that the ternary complex neutralizing antibodies in patients due to previous natural formed by the three proteins is more stable at the plasma infections by human parvovirus, present in nature (Thwaite et al., membrane, as previously seen in vitro and in vivo (Sirisi 2015). Indeed, a high percentage of the human population is et al., 2017). Furthermore, myelin vacuoles were significantly seropositive for AAV2 (Calcedo et al., 2009) and although there reduced in number and size (Figure 5). More importantly, may be some cross-recognition between antibodies raised against when the animals were treated at the beginning of the different serotypes, this may decrease using viruses from other symptomatology, we not only prevented the onset of the species (Thwaite et al., 2015). This is particularly important in cerebellar abnormalities but also significantly reduced the adult patients, as they have been exposed to various infections vacuolating phenotype in the transduced tissues. In fact, no throughout their life. differences were seen between the preventive and the therapeutic A dose-escalation study demonstrated a correlation between group at 8 months of age. MLC1 expression and the degree of correction of the white Based on the remitting edema phenotype of MLC2B patients, matter vacuolation in the cerebellum of treated KO mice who carry one dominant mutation in the HEPACAM gene, we Frontiers in Cellular Neuroscience | www.frontiersin.org 9 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy hypothesized that in vacuolating leukoencephalopathies such as answered. Administration of AAVs coding for GlialCAM under MLC, or as it was demonstrated for similar pathologies such as the regulation of an astrocyte or an oligodendrocyte-specific Canavan’s or Pelizaeus-Merzbacher-like disease (Georgiou et al., promoter may give us the clue for this question. Other AAV 2017; Gessler et al., 2017), the therapeutic correction could be serotypes or capsid modifications to increase vector tropism achieved even in advanced stages of the disease. Importantly, in for glial cells (Powell et al., 2016), using tandem promoters for this type of disease, there is no marked neuronal degeneration, both cell types (Colella et al., 2019) or detargeting the neuronal and possibly the clinical phenotype could be reversible. Indeed, tropism of AAVs by using microRNAs may also increase the the results obtained from animals treated at 15 months of age, efficiency of the strategy (Colin et al., 2009). almost 1 year after the development of the vacuolating pathology, Importantly, to follow the progression of the disease in demonstrated a similar level of correction to animals treated as gene therapy clinical trials we need to select appropriate asymptomatic (Sánchez et al., 2020). This is relevant for therapy, readouts. Alternatively, the development of larger animal models as indicates that we may have a wide therapeutic window for deficient for GlialCAM or MLC1 may facilitate direct invasive these patients and they might be treated at any age, while in most procedures and to follow physiological in vivo approaches such neurodegenerative diseases, successful treatment is conditioned as structural and functional imaging of small brain structures, to very early intervention, thus, subjected to their perinatal which is not possible in a mouse brain, and may open the detection, only if they have older affected siblings, or through possibility to develop new in vivo imaging biomarkers. Moreover, perinatal screening, which unfortunately is still not the case for larger animal models may facilitate CSF collection, and will many inherited diseases. allow finding richer behavioral phenotypes which will make This study is only a proof of concept of the feasibility of more complex behavioral testing possible. Besides, it will allow gene therapy for MLC. Therefore, the translatability of this assessing for a larger biodistribution of the vector in the brain and approach needs to be further confirmed as human patients possibly identifying new parameters to follow in patients during display pathology through the whole brain. However, some gene therapy treatment. of the most severe symptoms such as mobility problems might be partially related to cerebellar pathology, so, it CONCLUSIONS AND PERSPECTIVES needs to be tested whether expression of MLC only in the cerebellum is enough to ameliorate these symptoms. The first gene for MLC (MLC1) was identified 20 years Nevertheless, other phenotypes such as autism or cognitive ago. Although many discoveries have been made in this deterioration might be more difficult to correct since long journey, the exact function of the MLC1 protein is not these phenotypes may be the consequence of alteration of known and there are no therapeutic interventions for MLC developmental processes at very early stages. Whether the patients. We expect that novel results will provide wider mouse model may reveal higher susceptibility to develop perspectives of the pathophysiology of the disease, which epileptic seizures after kainate insults (Dubey et al., 2018) will help to clarify MLC protein’s role in glial cell biology. and if this may be corrected after gene therapy, remains to Hopefully, this knowledge will be accompanied by advanced be evaluated. Anyway, extrapolating from the tiny mouse therapeutic solutions. cerebellum to the same structure in young children is challenging. So, probably we will need to deliver the virus AUTHOR CONTRIBUTIONS through multiple injections, although other strategies like convection-enhanced administration should be assayed in Both authors participated in the writing and revision of the larger animals such as non-human primates, dogs, or pigs manuscript. All authors contributed to the article and approved (Lonser et al., 2015). the submitted version. There are still several important concerns to be studied in this disease before thinking in a human clinical trial. It FUNDING would be interesting to see if gene therapy in the Glialcam KO mouse model can mimic the results obtained in the This work was supported in part by the European Mlc1 KO mouse. Whether GlialCAM needs to be expressed Leukodystrophies Association (ELA) Research Foundation in both oligodendrocytes and astrocytes to recover normal (ELA 2018-0512) to AB; the Spanish Ministerio de Ciencia e histology, or which of these cell types has a major role in Innovación (MICINN; RTI2018-093493-B-I00) to RE. RE is a the development of the disease are questions that need to be recipient of an ICREA Academia prize. REFERENCES with subcortical cysts in two patients with GLIALCAM mutations. Neurogenetics 15, 41–48. doi: 10.1007/s10048-013-0381-x Ambrosini, E., Serafini, B., Lanciotti, A., Tosini, F., Scialpi, F., Psaila, R., et al. Arnedo, T., López-Hernández, T., Jeworutzki, E., Capdevila-Nortes, X., Sirisi, S., (2008). Biochemical characterization of MLC1 protein in astrocytes and its Pusch, M., et al. (2014b). Functional analyses of mutations in HEPACAM association with the dystrophin-glycoprotein complex. Mol. Cell. Neurosci. 37, causing megalencephalic leukoencephalopathy. Hum. Mutat. 35, 1175–1178. 480–493. doi: 10.1016/j.mcn.2007.11.003 doi: 10.1002/humu.22622 Arnedo, T., Aiello, C., Jeworutzki, E., Dentici, M. L., Uziel, G., Simonati, A., et al. Barrallo-Gimeno, A., Gradogna, A., Zanardi, I., Pusch, M., and (2014a). Expanding the spectrum of megalencephalic leukoencephalopathy Estévez, R. (2015). Regulatory-auxiliary subunits of CLC chloride Frontiers in Cellular Neuroscience | www.frontiersin.org 10 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy channel-transport proteins. J. Physiol. 593, 4111–4127. doi: 10.1113/ Dubey, M., Bugiani, M., Ridder, M. C., Postma, N. L., Brouwers, E., Polder, E., JP270057 et al. (2015). Mice with megalencephalic leukoencephalopathy with cysts: a Boor, P. K., De Groot, K., Waisfisz, Q., Kamphorst, W., Oudejans, C. B., developmental angle. Ann. Neurol. 77, 114–131. doi: 10.1002/ana.24307 Powers, J. M., et al. (2005). MLC1: a novel protein in distal astroglial Elorza-Vidal, X., Sirisi, S., Gáitan-Peñas, H., Pérez-Rius, C., Alonso-Gardón, M., processes. J. Neuropathol. Exp. Neurol. 64, 412–419. doi: 10.1093/jnen/ Armand-Ugón, M., et al. (2018). GlialCAM/MLC1 modulates LRRC8/VRAC 64.5.412 currents in an indirect manner: implications for megalencephalic Bosch, A., Perret, E., Desmaris, N., Trono, D., and Heard, J. M. (2000). Reversal leukoencephalopathy. Neurobiol. Dis. 119, 88–99. doi: 10.1016/j.nbd.2018. of pathology in the entire brain of mucopolysaccharidosis type VII mice 07.031 after lentivirus-mediated gene transfer. Hum. Gene Ther. 11, 1139–1150. Elorza-Vidal, X., Xicoy-Espaulella, E., Pla-Casillanis, A., Alonso-Gardón, M., doi: 10.1089/10430340050015194 Gáitan-Peñas, H., Engel-Pizcueta, C., et al. (2020). Structural basis for Brenner, M., Kisseberth, W. C., Su, Y., Besnard, F., and Messing, A. (1994). GFAP the dominant or recessive character of GLIALCAM mutations found in promoter directs astrocyte-specific expression in transgenic mice. J. Neurosci. leukodystrophies. Hum. Mol. Genet. 29, 1107–1120. doi: 10.1093/hmg/ 14, 1030–1037. doi: 10.1523/JNEUROSCI.14-03-01030.1994 ddaa009 Brignone, M. S., Lanciotti, A., Macioce, P., Macchia, G., Gaetani, M., Aloisi, F., Estévez, R., Elorza-Vidal, X., Gáitan-Peñas, H., Pérez-Rius, C., Armand-Ugón, M., et al. (2011). The beta1 subunit of the Na,K-ATPase pump interacts with Alonso-Gardón, M., et al. (2018). Megalencephalic leukoencephalopathy with megalencephalic leucoencephalopathy with subcortical cysts protein 1 (MLC1) subcortical cysts: a personal biochemical retrospective. Eur. J. Med. Genet. 61, in brain astrocytes: new insights into MLC pathogenesis. Hum. Mol. Genet. 20, 50–60. doi: 10.1016/j.ejmg.2017.10.013 90–103. doi: 10.1093/hmg/ddq435 Foust, K. D., Nurre, E., Montgomery, C. L., Hernandez, A., Chan, C. M., Brignone, M. S., Lanciotti, A., Serafini, B., Mallozzi, C., Sbriccoli, M., Veroni, C., and Kaspar, B. K. (2009). Intravascular AAV9 preferentially targets neonatal et al. (2019). Megalencephalic leukoencephalopathy with subcortical neurons and adult astrocytes. Nat. Biotechnol. 27, 59–65. doi: 10.1038/ cysts protein-1 (MLC1) counteracts astrocyte activation in response to nbt.1515 inflammatory signals. Mol. Neurobiol. 56, 8237–8254. doi: 10.1007/s12035-019- Gaitán-Peñas, H., Apaja, P. M., Arnedo, T., Castellanos, A., Elorza-Vidal, X., 01657-y Soto, D., et al. (2017). Leukoencephalopathy-causing CLCN2 mutations are Brignone, M. S., Lanciotti, A., Visentin, S., De Nuccio, C., Molinari, P., associated with impaired Cl− channel function and trafficking. J. Physiol. 595, Camerini, S., et al. (2014). Megalencephalic leukoencephalopathy with 6993–7008. doi: 10.1113/JP275087 subcortical cysts protein-1 modulates endosomal pH and protein trafficking Georgiou, E., Sidiropoulou, K., Richter, J., Papaneophytou, C., Sargiannidou, I., in astrocytes: relevance to MLC disease pathogenesis. Neurobiol. Dis. 66, 1–18. Kagiava, A., et al. (2017). Gene therapy targeting oligodendrocytes provides doi: 10.1016/j.nbd.2014.02.003 therapeutic benefit in a leukodystrophy model. Brain 140, 599–616. Bugiani, M., Dubey, M., Breur, M., Postma, N. L., Dekker, M. P., Ter Braak, T., doi: 10.1093/brain/aww351 et al. (2017). Megalencephalic leukoencephalopathy with cysts: the Glialcam- Gessler, D. J., Li, D., Xu, H., Su, Q., Sanmiguel, J., Tuncer, S., et al. (2017). null mouse model. Ann. Clin. Transl. Neurol. 4, 450–465. doi: 10.1002/ Redirecting N-acetylaspartate metabolism in the central nervous system acn3.405 normalizes myelination and rescues Canavan disease. JCI Insight 2:e90807. Calcedo, R., Vandenberghe, L. H., Gao, G., Lin, J., and Wilson, J. M. (2009). doi: 10.1172/jci.insight.90807 Worldwide epidemiology of neutralizing antibodies to adeno-associated Gilbert, A., Vidal, X. E., Estévez, R., Cohen-Salmon, M., and Boulay, A.-C. (2019). viruses. J. Infect. Dis. 199, 381–390. doi: 10.1086/595830 Postnatal development of the astrocyte perivascular MLC1/GlialCAM complex Capdevila-Nortes, X., López-Hernández, T., Apaja, P. M., López de Heredia, M., defines a temporal window for the gliovascular unit maturation. Brain Struct. Sirisi, S., Callejo, G., et al. (2013). Insights into MLC pathogenesis: Funct. 224, 1267–1278. doi: 10.1007/s00429-019-01832-w GlialCAM is an MLC1 chaperone required for proper activation of volume- Hamilton, E. M. C., Tekturk, P., Cialdella, F., van Rappard, D. F., Wolf, N. I., regulated anion currents. Hum. Mol. Genet. 22, 4405–4416. doi: 10.1093/hmg/ Yalcinkaya, C., et al. (2018). Megalencephalic leukoencephalopathy with ddt290 subcortical cysts: characterization of disease variants. Neurology 90, Colella, P., Sellier, P., Costa Verdera, H., Puzzo, F., Van Wittenberghe, L., e1395–e1403. doi: 10.1212/WNL.0000000000005334 Guerchet, N., et al. (2019). AAV gene transfer with tandem promoter design Hinderer, C., Katz, N., Buza, E. L., Dyer, C., Goode, T., Bell, P., et al. (2018). Severe prevents anti-transgene immunity and provides persistent efficacy in neonate toxicity in nonhuman primates and piglets following high-dose intravenous pompe mice. Mol. Ther. Methods Clin. Dev. 12, 85–101. doi: 10.1016/j.omtm. administration of an adeno-associated virus vector expressing human SMN. 2018.11.002 Hum. Gene Ther. 29, 285–298. doi: 10.1089/hum.2018.015 Colin, A., Faideau, M., Dufour, N., Auregan, G., Hassig, R., Andrieu, T., et al. Hoegg-Beiler, M. B., Sirisi, S., Orozco, I. J., Ferrer, I., Hohensee, S., Auberson, M., (2009). Engineered lentiviral vector targeting astrocytes in vivo. Glia 57, et al. (2014). Disrupting MLC1 and GlialCAM and ClC-2 interactions in 667–679. doi: 10.1002/glia.20795 leukodystrophy entails glial chloride channel dysfunction. Nat. Commun. Depienne, C., Bugiani, M., Dupuits, C., Galanaud, D., Touitou, V., Postma, N., 5:3475. doi: 10.1038/ncomms4475 et al. (2013). Brain white matter oedema due to ClC-2 chloride channel Hordeaux, J., Wang, Q., Katz, N., Buza, E. L., Bell, P., and Wilson, J. M. (2018). deficiency: an observational analytical study. Lancet Neurol. 12, 659–668. The neurotropic properties of AAV-PHP.B are limited to C57BL/6J mice. Mol. doi: 10.1016/S1474-4422(13)70053-X Ther. 26, 664–668. doi: 10.1016/j.ymthe.2018.01.018 Deverman, B. E., Pravdo, P. L., Simpson, B. P., Kumar, S. R., Chan, K. Y., Hordeaux, J., Yuan, Y., Clark, P. M., Wang, Q., Martino, R. A., Sims, J. J., et al. Banerjee, A., et al. (2016). Cre-dependent selection yields AAV variants for (2019). The GPI-linked protein LY6A drives AAV-PHP.B transport across the widespread gene transfer to the adult brain. Nat. Biotechnol. 34, 204–209. blood-brain barrier. Mol. Ther. 27, 912–921. doi: 10.1016/j.ymthe.2019.02.013 doi: 10.1038/nbt.3440 Huda, F., Konno, A., Matsuzaki, Y., Goenawan, H., Miyake, K., Shimada, T., et al. Duarri, A., Lopez De Heredia, M., Capdevila-Nortes, X., Ridder, M. C., (2014). Distinct transduction profiles in the CNS via three injection routes of Montolio, M., López-Hernández, T., et al. (2011). Knockdown of MLC1 in AAV9 and the application to generation of a neurodegenerative mouse model. primary astrocytes causes cell vacuolation: a MLC disease cell model. Neurobiol. Mol. Ther. Methods Clin. Dev. 1:14032. doi: 10.1038/mtm.2014.32 Dis. 43, 228–238. doi: 10.1016/j.nbd.2011.03.015 Hwang, J., Vu, H. M., Kim, M.-S., and Lim, H.-H. (2019). Plasma membrane Duarri, A., Teijido, O., López-Hernández, T., Scheper, G. C., Barriere, H., Boor, I., localization of MLC1 regulates cellular morphology and motility. Mol. Brain et al. (2008). Molecular pathogenesis of megalencephalic leukoencephalopathy 12:116. doi: 10.1186/s13041-019-0540-6 with subcortical cysts: mutations in MLC1 cause folding defects. Hum. Mol. Jeworutzki, E., Lagostena, L., Elorza-Vidal, X., López-Hernández, T., Estévez, R., Genet. 17, 3728–3739. doi: 10.1093/hmg/ddn269 and Pusch, M. (2014). GlialCAM, a CLC-2 Cl− channel subunit, activates the Dubey, M., Brouwers, E., Hamilton, E. M. C., Stiedl, O., Bugiani, M., Koch, H., slow gate of CLC chloride channels. Biophys. J. 107, 1105–1116. doi: 10.1016/j. et al. (2018). Seizures and disturbed brain potassium dynamics in the bpj.2014.07.040 leukodystrophy megalencephalic leukoencephalopathy with subcortical cysts. Jeworutzki, E., López-Hernández, T., Capdevila-Nortes, X., Sirisi, S., Bengtsson, L., Ann. Neurol. 83, 636–649. doi: 10.1002/ana.25190 Montolio, M., et al. (2012). GlialCAM, a protein defective in a leukodystrophy, Frontiers in Cellular Neuroscience | www.frontiersin.org 11 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy serves as a ClC-2 Cl− channel auxiliary subunit. Neuron 73, 951–961. of megalencephalic leukoencephalopathy with subcortical cysts. Mol. Cell. doi: 10.1016/j.neuron.2011.12.039 Neurosci. 56, 307–321. doi: 10.1016/j.mcn.2013.07.001 Kikuchihara, S., Sugio, S., Tanaka, K. F., Watanabe, T., Kano, M., Yamazaki, Y., Petrosyan, H. A., Alessi, V., Singh, V., Hunanyan, A. S., Levine, J. M., and et al. (2018). Ectopic positioning of Bergmann glia and impaired cerebellar Arvanian, V. L. (2014). Transduction efficiency of neurons and glial cells by wiring in Mlc1-over-expressing mice. J. Neurochem. 147, 344–360. AAV-1, -5, -9, -rh10, and -hu11 serotypes in rat spinal cord following contusion doi: 10.1111/jnc.14486 injury. Gene Ther. 21, 991–1000. doi: 10.1038/gt.2014.74 Lanciotti, A., Brignone, M. S., Belfiore, M., Columba-Cabezas, S., Mallozzi, C., Powell, S. K., Khan, N., Parker, C. L., Samulski, R. J., Matsushima, G., Gray, S. J., Vincentini, O., et al. (2020). Megalencephalic leukoencephalopathy with et al. (2016). Characterization of a novel adeno-associated viral vector with subcortical cysts disease-linked MLC1 protein favors gap-junction intercellular preferential oligodendrocyte tropism. Gene Ther. 23, 807–814. doi: 10.1038/gt. communication by regulating connexin 43 trafficking in astrocytes. Cells 2016.62 9:1425. doi: 10.3390/cells9061425 Ridder, M. C., Boor, I., Lodder, J. C., Postma, N. L., Capdevila-Nortes, X., Lanciotti, A., Brignone, M. S., Camerini, S., Serafini, B., Macchia, G., Raggi, C., Duarri, A., et al. (2011). Megalencephalic leucoencephalopathy with cysts: et al. (2010). MLC1 trafficking and membrane expression in astrocytes: role defect in chloride currents and cell volume regulation. Brain 134, 3342–3354. of caveolin-1 and phosphorylation. Neurobiol. Dis. 37, 581–595. doi: 10.1016/j. doi: 10.1093/brain/awr255 nbd.2009.11.008 Sánchez, A., García-Lareu, B., Puig, M., Prat, E., Ruberte, J., Chillón, M., et al. Lanciotti, A., Brignone, M. S., Molinari, P., Visentin, S., De Nuccio, C., (2020). Cerebellar astrocyte transduction as gene therapy for megalencephalic Macchia, G., et al. (2012). Megalencephalic leukoencephalopathy with leukoencephalopathy. Neurotherapeutics doi: 10.1007/s13311-020 subcortical cysts protein 1 functionally cooperates with the TRPV4 cation -00865-y [Online ahead of print]. channel to activate the response of astrocytes to osmotic stress: Sehara, Y., Fujimoto, K.-I., Ikeguchi, K., Katakai, Y., Ono, F., Takino, N., et al. dysregulation by pathological mutations. Hum. Mol. Genet. 21, 2166–2180. (2017). Persistent expression of dopamine-synthesizing enzymes 15 years after doi: 10.1093/hmg/dds032 gene transfer in a primate model of Parkinson’s disease. Hum. Gene Ther. Clin. Lanciotti, A., Brignone, M. S., Visentin, S., De Nuccio, C., Catacuzzeno, L., Dev. 28, 74–79. doi: 10.1089/humc.2017.010 Mallozzi, C., et al. (2016). Megalencephalic leukoencephalopathy with Shi, Z., Yan, H.-F., Cao, B.-B., Guo, M.-M., Xie, H., Gao, K., et al. subcortical cysts protein-1 regulates epidermal growth factor receptor (2019). Identification in Chinese patients with GLIALCAM mutations signaling in astrocytes. Hum. Mol. Genet. 25, 1543–1558. doi: 10.1093/hmg/ of megalencephalic leukoencephalopathy with subcortical cysts and brain ddw032 pathological study on Glialcam knock-in mouse models. World J. Pediatr. 15, Lawlor, P. A., Bland, R. J., Mouravlev, A., Young, D., and During, M. J. 454–464. doi: 10.1007/s12519-019-00284-w (2009). Efficient gene delivery and selective transduction of glial cells in the Sirisi, S., Elorza-Vidal, X., Arnedo, T., Armand-Ugón, M., Callejo, G., Capdevila- mammalian brain by AAV serotypes isolated from nonhuman primates. Mol. Nortes, X., et al. (2017). Depolarization causes the formation of a ternary Ther. 17, 1692–1702. doi: 10.1038/mt.2009.170 complex between GlialCAM, MLC1, and ClC-2 in astrocytes: implications Leegwater, P. A. J., Yuan, B. Q., van der Steen, J., Mulders, J., Konst, A. A. M., in megalencephalic leukoencephalopathy. Hum. Mol. Genet. 26, 2436–2450. Boor, P. K. I., et al. (2001). Mutations of MLC1 (KIAA0027), encoding doi: 10.1093/hmg/ddx134 a putative membrane protein, cause megalencephalic leukoencephalopathy Sirisi, S., Folgueira, M., López-Hernández, T., Minieri, L., Pérez-Rius, C., with subcortical cysts. Am. J. Hum. Genet. 68, 831–838. doi: 10.1086/ Gáitan-Peñas, H., et al. (2014). Megalencephalic leukoencephalopathy with 319519 subcortical cysts protein 1 regulates glial surface localization of GLIALCAM Lonser, R. R., Sarntinoranont, M., Morrison, P. F., and Oldfield, E. H. (2015). from fish to humans. Hum. Mol. Genet. 23, 5069–5086. doi: 10.1093/hmg/ Convection-enhanced delivery to the central nervous system. J. Neurosurg. 122, ddu231 697–706. doi: 10.3171/2014.10.JNS14229 Sugio, S., Tohyama, K., Oku, S., Fujiyoshi, K., Yoshimura, T., Hikishima, K., López-Hernández, T., Ridder, M. C., Montolio, M., Capdevila-Nortes, X., et al. (2017). Astrocyte-mediated infantile-onset leukoencephalopathy mouse Polder, E., Sirisi, S., et al. (2011a). Mutant GlialCAM causes megalencephalic model. Glia 65, 150–168. doi: 10.1002/glia.23084 leukoencephalopathy with subcortical cysts, benign familial macrocephaly and Teijido, O., Casaroli-Marano, R., Kharkovets, T., Aguado, F., Zorzano, A., macrocephaly with retardation and autism. Am. J. Hum. Genet. 88, 422–432. Palacin, M., et al. (2007). Expression patterns of MLC1 protein in the central doi: 10.1016/j.ajhg.2011.02.009 and peripheral nervous systems. Neurobiol. Dis. 26, 532–545. doi: 10.1016/j. López-Hernández, T., Sirisi, S., Capdevila-Nortes, X., Montolio, M., Fernández- nbd.2007.01.016 Dueáas, V., Scheper, G. C., et al. (2011b). Molecular mechanisms of Teijido, O., Martínez, A., Pusch, M., Zorzano, A., Soriano, E., Del Rio, J. A., et al. MLC1 and GLIALCAM mutations in megalencephalic leukoencephalopathy (2004). Localization and functional analyses of the MLC1 protein involved in with subcortical cysts. Hum. Mol. Genet. 20, 3266–3277. doi: 10.1093/hmg/ megalencephalic leukoencephalopathy with subcortical cysts. Hum. Mol. Genet. ddr238 13, 2581–2594. doi: 10.1093/hmg/ddh291 Mendell, J. R., Al-Zaidy, S., Shell, R., Arnold, W. D., Rodino-Klapac, L. R., Thwaite, R., Pagès, G., Chillón, M., and Bosch, A. (2015). Prior, T. W., et al. (2017). Single-dose gene-replacement therapy AAVrh.10 immunogenicity in mice and humans. Relevance of antibody cross- for spinal muscular atrophy. N. Engl. J. Med. 377, 1713–1722. reactivity in human gene therapy. Gene Ther. 22, 196–201. doi: 10.1038/gt. doi: 10.1056/NEJMoa1706198 2014.103 Moh, M. C., Zhang, C., Luo, C., Lee, L. H., and Shen, S. (2005). Structural and van der Knaap, M. S., Abbink, T. E. M., and Min, R. (1993). ‘‘Megalencephalic functional analyses of a novel Ig-like cell adhesion molecule, hepaCAM, in leukoencephalopathy with subcortical cysts,’’ in GeneReviews((R)), eds the human breast carcinoma MCF7 cells. J. Biol. Chem. 280, 27366–27374. M. P. Adam, H. H. Ardinger, R. A. Pagon, S. E. Wallace, L. J. H. Bean, doi: 10.1074/jbc.M500852200 K. Stephens, et al. (Seattle, WA: University of Washington), 1993–2020. Pagès, G., Giménez-Llort, L., García-Lareu, B., Ariza, L., Navarro, M., van der Knaap, M. S., Barth, P. G., Stroink, H., Van Nieuwenhuizen, O., Casas, C., et al. (2019). Intrathecal AAVrh10 corrects biochemical and Arts, W. F., Hoogenraad, F., et al. (1995). Leukoencephalopathy with swelling histological hallmarks of mucopolysaccharidosis VII mice and improves and a discrepantly mild clinical course in eight children. Ann. Neurol. 37, behavior and survival. Hum. Mol. Genet. 28, 3610–3624. doi: 10.1093/hmg/ 324–334. doi: 10.1002/ana.410370308 ddz220 van der Knaap, M. S., Barth, P. G., Vrensen, G. F., and Valk, J. (1996). Pérez-Rius, C., Folgueira, M., Elorza-Vidal, X., Alia, A., Hoegg-Beiler, M. B., Histopathology of an infantile-onset spongiform leukoencephalopathy Eeza, M. N.H., et al. (2019). Comparison of zebrafish and mice with a discrepantly mild clinical course. Acta Neuropathol. 92, 206–212. knockouts for megalencephalic leukoencephalopathy proteins indicates doi: 10.1007/s004010050510 that GlialCAM/MLC1 forms a functional unit. Orphanet J. Rare Dis. 14:268. van der Knaap, M. S., Boor, I., and Estévez, R. (2012). Megalencephalic doi: 10.1186/s13023-019-1248-5 leukoencephalopathy with subcortical cysts: chronic white matter oedema due Petrini, S., Minnone, G., Coccetti, M., Frank, C., Aiello, C., Cutarelli, A., to a defect in brain ion and water homoeostasis. Lancet Neurol. 11, 973–985. et al. (2013). Monocytes and macrophages as biomarkers for the diagnosis doi: 10.1016/S1474-4422(12)70192-8 Frontiers in Cellular Neuroscience | www.frontiersin.org 12 January 2021 | Volume 14 | Article 627887 Bosch and Estévez MLC: Pathophysiology and Therapy van der Knaap, M. S., Lai, V., Köhler, W., Salih, M. A., Fonseca, M.- Conflict of Interest: The authors declare that the research was conducted in the J., Benke, T. A., et al. (2010). Megalencephalic leukoencephalopathy with absence of any commercial or financial relationships that could be construed as a cysts without MLC1 defect two phenotypes. Ann. Neurol. 67, 834–837. potential conflict of interest. doi: 10.1002/ana.21980 Wang, D., Li, J., Tran, K., Burt, D. R., Zhong, L., and Gao, G. (2018). Slow infusion Copyright © 2021 Bosch and Estévez. This is an open-access article distributed of recombinant adeno-associated viruses into the mouse cerebrospinal fluid under the terms of the Creative Commons Attribution License (CC BY). The use, space. Hum. Gene Ther. Methods 29, 75–85. doi: 10.1089/hgtb.2017.250 distribution or reproduction in other forums is permitted, provided the original Wu, M., Moh, M. C., and Schwarz, H. (2016). HepaCAM associates with connexin author(s) and the copyright owner(s) are credited and that the original publication 43 and enhances its localization in cellular junctions. Sci. Rep. 6, 36218. in this journal is cited, in accordance with accepted academic practice. No use, doi: 10.1038/srep36218 distribution or reproduction is permitted which does not comply with these terms. Frontiers in Cellular Neuroscience | www.frontiersin.org 13 January 2021 | Volume 14 | Article 627887