Inherited Disorders of Lysosomal Membrane Transporters

Abstract

Disorders caused by defects in lysosomal membrane transporters form a distinct subgroup of lysosomal storage disorders (LSDs). To date, defects in only 10 lysosomal membrane transporters have been associated with inherited disorders. The clinical presentations of these diseases resemble the phenotypes of other LSDs; they are heterogeneous and often present in children with neurodegenerative manifestations. However, for pathomechanistic and therapeutic studies, lysosomal membrane transport defects should be distinguished from LSDs caused by defective hydrolytic enzymes. The involved proteins differ in function, localization, and lysosomal targeting, and the diseases themselves differ in their stored material and therapeutic approaches. We provide an overview of the small group of disorders of lysosomal membrane transporters, emphasizing discovery, pathomechanism, clinical features, diagnostic methods and therapeutic aspects. We discuss common aspects of lysosomal membrane transporter defects that can provide the basis for preclinical research into these disorders.

1. INTRODUCTION

Lysosomal storage diseases (LSDs) are a group of over 80 inherited monogenic disorders associated with lysosomal dysfunction [1]. Most LSDs result from defects in genes encoding enzymes involved in the lysosomal hydrolysis of macromolecules, while defects in lysosomal membrane transport proteins comprise only a small portion (~12%) of LSDs. At least 40 lysosomal membrane transport proteins exist [2]; all are essential to cell metabolism, largely for their salvage functions. Many of these proteins are solute carrier transporters (SLC) for amino acids, di- or tri-peptides, sugars, inorganic ions and vitamins [2–4], and some have not yet been molecularly characterized or investigated for a role in health and disease [4–6]. About 50% of these transporters are associated with human inherited diseases [2,7]. Here we review the dominant and recessive inherited human disorders associated with membrane transporters and resulting in lysosomal storage of metabolites (Figure 1, Table 1). We exclude lysosomal membrane transporters whose expression has been associated with non-inherited human disorders (e.g., SLC1A2 associated with ALS and schizophrenia [8]), and whose disease mechanism and localization is not solely lysosomal (e.g., SLCA714 associated with retinitis pigmentosa [9]), some of these disorders are listed elsewhere [4]. We also exclude lysosome-associated membrane proteins that have no demonstrated transport activity themselves, such as OSTM1 and TCIRG1 associated with osteopetrosis [10,11], ABCD4 associated with cobalamin F disease [12], and LAMP-2 associated with Danon disease [13].

Figure 1: Simplified overview of lysosomal transmembrane transport proteins involved in human inherited disorders.

Features of the 10 transmembrane proteins discussed in this review are indicated per category: A, Lysosomal catabolite export proteins. B, Lysosomal membrane ion transport proteins. C, Lysosomal export of endocytosed component proteins. D, Defective transmembrane importer/enzyme protein. E, Lysosomal transport proteins with other or unknown function. Topology cartoons are not to scale, position and amino acid sequence of lysosomal sorting signals are indicated with arrows, location of frequent mutations are marked with * (deletions and splice site mutations are not indicated). See text for details and references.

Table 1:

Overview of inherited lysosomal storage diseases caused by defective lysosomal membrane transport proteins

Disease Name	Disease Gene Human Locus (year)1	Alternative Names	MIM#	GenelD mRNA2 (# exons; # splice var) Protein ID2 (# aa;#TM)	Gene Variants3 Reported Cases4	Lysosomal Membrane Transport	Approved (experimental) Therapy
Defective Lysosomal Catabolite Export
Cystinosis	CTNS 17p13.2 (1998)	CYSTINOSIN SLC66A4 PQLC4	219800 (AR) 219900 (AR) 219750 (AR)	ID:1497 NM 001031681 (13 ex; 7 var) NP_001026851 (400 aa; 7 TM)	>159 variants ~500 cases	H+/cystine symport	cysteamine
Free Sialic Acid Storage Disease (FSASD)	SLC17A5 6q13 (1999)	SIALIN	604369 (AR) 269920 (AR)	ID:26503 NM 012434(11 ex; 1 var) NP 036566 (495 aa; 12 TM)	40 variants ~200 cases	H+/sialic acid (acidic hexoses) symport	-
H Syndrome	SLC29A3 10q22.1 (2008)	ENT3	602782 (AR)	ID:55315 NM 018344 (6 ex; 3 var) NP 060814 (475 aa; 11TM)	28 variants ~130 cases	pH-dependent nucleoside export	-
Defective Lysosomal Membrane Ion Transport
Mucolipidosis type IV (MLIV)	MCOLN1 19p13.2 (2000)	Mucolipin-1 MLN1 TRPML1	252650 (AR)	ID: 57192 NM 020533 (14 ex; 1 var) NP 065394 (580 aa; 6 TM)	33 variants ~120 cases	Cation channel (Na+, K+, Ca2+; Fe2+; H+ transport)	-
Osteopetrosis, AR type 4 (OPTB4)	CLCN7 16p13.3 (2001)	CLC7 0PTB4	611490 (AR)	ID: 1186 NM 001287 (25 ex; 2 var) NP 001278 (805 aa; 18TM5)	60 variants (AR) <50 cases	2Cf/1H+ antiport	HSCT IFN-γ
Osteopetrosis, AD type 2 (OPTA2)	CLCN7 16p13.3 (2001)	CLC7 0PTA2	166600 (AD)	ID: 1186 NM 001287 (25 ex; 2 var) NP 001278 (805 aa; 18TM5)	41 variants (AD) >250 cases	2Cf/1H+ antiport	(IFN-γ)
HOD6	CLCN7 16p13.3 (2019)	HOD	618541 (AD)	ID: 1186 NM 001287 (25 ex; 2 var) NP 001278 (805 aa; 18TM5)	1 variant (AD) 2 cases	2Cf/1H+ antiport	(chloroquine)
Defective Lysosomal Export of an Endocytosed Component
Niemann-Pick Disease type C1 (NPC1)	NPC1 18q 11.2 (1997)	SLC65A1 NPC POGZ	257220 (AR)	ID:4864 NM 000271 (25 ex; 1 var) NP 000262 (1278 aa; 13 TM)	~500 variants >500 cases	(predicted) cholesterol transport	miglustat7 (HPBCD, arimoclomol)
Cobalamin F Disease (cblF)	LMBRD1 6q13 (2009)	LMBD1 NESI	277380 (AR)	ID:55788 NM 018368 (16 ex; 4 var) NP 060838 (540 aa; 9 TM)	9 variants ~16 cases	(predicted) cobalamin export	hydroxy- cobalamin
Defective Lysosomal Transmembrane Importer/Enzyme Protein
Mucopolysaccharidosis type MIC	HGSNAT 8p11.21 -p11.1 (2006)	MPS3C HGNAT TMEM76	252930 (AR)	ID: 138050 NM 152419 (18 ex; 4 var) NP 689632 (635 aa; 11 TM)	74 variants <100 cases	acetyl-CoA import	-
Defective Lysosomal Transmembrane Transporters of Other or Unknown Function
Ceroid Lipofuscinosis Neuronal 3	CLN3 16p12.1 (1995)	BATTENIN BTN1, BTS JNCL	204200 (AR)	ID: 1201 NM 001042432 (16 ex; 6 var) NP_001035897 (438 aa; 6 TM)	77 variants >450 cases	(predicted) metabolites	(mycophenolate mofetil; AAV9- CLN3 gene therapy)
Ceroid Lipofuscinosis Neuronal 7	MFSD8 4q28.2 (2007)	CLN7 CCMD	610951 (AR)	ID: 256471 NM 152778 (13 ex; 1 var) NP 689991 (518 aa; 12 TM)	51 variants 104 cases	metabolites	-

Abbreviations: aa, amino acids; AD, autosomal dominant inheritance; AR, autosomal recessive inheritance; ex, exons; HSCT, hematopoietic stem cell transplantation; HPBCD, hydroxypropyl beta cyclodextrin; H-syndrome, Histiocytosis-Lymphadenopathy Plus syndrome; HOD, Hypopigmentation, organomegaly, and delayed myelination and development; IFN-γ, interferon gamma; MIM, Online Mendelian Inheritance in Man; FSASD, Free Sialic Acid Storage Disease; TM, transmembrane domains; var, splice variants;

¹Year the disease gene was first reported.

²Genbank accession numbers of the mRNA and protein ID representing the longest isoform (often transcript variant 1). As of January 2020.

³Retrieved from Human Genome Mutation Database (HGMD;), January 2020.

⁴Estimated from literature reports, January 2020.

⁵Some CLCN7 helices only partially span the lysosomal membrane.

⁶Hypopigmentation, organomegaly, delayed myelination and development.

⁷Miglustat therapy for NPC is approved in several countries, not including the United States.

The clinical presentations of LSDs due to defective membrane transport proteins (Table 2) and those due to defective hydrolytic enzymes highly resemble each other; they are heterogeneous and often present in children with neurodegenerative manifestations. However, these 2 groups of LSDs should be distinguished for pathomechanistic and therapeutic studies. Important differences between LSDs due to defective membrane transport proteins and those due to defective hydrolytic enzymes, involve protein localization, function, lysosomal targeting signals, stored materials, and therapeutic approaches. A major difference is the targeting-sorting mechanism [14]. Most soluble lysosomal enzymes employ mannose or mannose-6-phosphate receptors, whose presence on the plasma membrane allowed the development of enzyme replacement therapy (ERT) for some lysosomal storage disorders [15]. ERT is not effective for membrane-bound (not soluble) lysosomal transport proteins, which are not dependent on mannose-6-phosphate-mediated lysosomal targeting. Instead, lysosomal membrane proteins depend upon lysosomal targeting motifs (i,e., tyrosine- or dileucine-based) in their cytosolic tails [16,17]. These motifs interact with adaptor protein (AP) complexes that recruit coat proteins (i.e., clathrin) to form coated vesicles that are trafficked through the endosomal system to reach and fuse with lysosomes [16]. Lysosomal targeting motifs have been identified for most known lysosomal membrane transporters, including cystinosin [18], SLC17A5 [19], ENT3 [20], and others [7,10,21–24] (Fig 1). No therapeutic strategies have been designed for disorders of lysosomal membrane transporters based on lysosomal membrane targeting motifs, but these motifs have been utilized (i.e., mutated) to redirect lysosomal transporters to the cell surface to facilitate research-based transport activity measurements [2].

Table 2:

Affected organs in disorders of lysosomal transmembrane transporters

			Affected Organs
#	Disease Name	Disease Gene	Brain	Eye	Hepato-splenomegaly	Other organs
1	Cystinosis	CTNS	−	+	−	kidney, muscle, skin, skeletal, thyroid
2	Sialic Acid Storage Disease	SLC17A5	+	−	+	coarse facies, heart, muscle
3	H Syndrome	SLC29A3 (ENT3)	−	−	+	diabetes mellitus, ear, gastrointestinal system, hair, heart, hematologic, hypogonadism, joints, kidney, lymph nodes, skeletal, skin
4	Cobalamin F Disease	LMBRD1	+	−	−	heart, hematologic, kidney, liver, skeletal, skin
5	Mucolipidosis type IV	MCOLN1	+	+	−	gastrointestinal system, hematologic
6.1	Osteopetrosis, AR type 4	CLCN7	+	+	−	hematologic, skeletal
6.2	Osteopetrosis, AD type 2	CLCN7	+/−	+/−	−	skeletal
6.3	HOD	CLCN7	+	−	+	muscle, skin
7	Mucopolysaccharidosis type MIC	HGSNAT	+	+	+	coarse facies, ears, hernias, joints, lung, skeletal
8	Niemann-Pick Disease type C	NPC1	+	+	+	lung
9	Ceroid Lipofuscinosis Neuronal-3	CLN3	+	+	−	heart, muscle
10	Ceroid Lipofuscinosis Neuronal-7	MFSD8	+	−	−	muscle

The subgroup of inherited LSDs caused by defects in lysosomal membrane transporters can be categorized according to their lysosomal metabolite transport function, specifically catabolite export (cystinosis, sialic acid storage disease, H syndrome), ion transport (mucolipidosis type IV, osteopetrosis recessive type 4 and dominant type 2, hypopigmentation, organomegaly, delayed myelination and development (HOD)), export of an endocytosed component (cobalamin F disease, Niemann-Pick C1), metabolite import (mucopolysaccharidosis type IIIC), and membrane transport of other or unknown function (neuronal ceroid lipofuscinoses types-3 and −7) (Table 1). Here we provide a current overview of pathomechanism, clinical features, diagnostic aspects and therapeutic approaches of this distinct group of disorders of lysosomal membrane transporters.

2. LYSOSOMAL DISORDERS OF CATABOLITE EXPORT

2.1. Cystinosis

Background:

Cystinosis is an autosomal recessive inherited multisystem lysosomal storage disease caused by bi-allelic mutations in the CTNS gene (chromosome 17p13.2; Gene ID 1497), encoding the lysosomal membrane protein cystinosin [18,25–29]. This seven-transmembrane domain protein is a H+-cystine symporter responsible for transport of the amino acid cystine out of the lysosomal compartment [30,31] (Fig 1A). Deficient cystinosin leads to intra-lysosomal cystine accumulation and crystal formation in the acidic lysosome of almost all cells, including corneas, conjunctivae, kidneys, liver, spleen, muscle, brain, thyroid, intestines, rectal mucosa, lymph nodes, macrophages, and bone marrow [28,32]. Clinical symptoms of cystinosis vary; but kidney involvement is characteristic and occurs early. In fact, cystinosis is the most common hereditary cause of renal Fanconi syndrome in children.

The incidence of cystinosis is estimated to be 0.5–1 per 100,000 live births [33]. Higher incidence rates occur in populations with founder mutations, such as in Brittany, France (1 per 26,000 live births) [34] or in the Saguenay-Lac-St-Jean, Quebec, Canada (1 per 62,500 live births) [35]. Since 1998, 159 CTNS mutations were reported in subjects with cystinosis (Human Gene Mutation Database (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). A CTNS 57-kb deletion is detected in up to 76% of affected alleles of northern European inheritance [25,29]. The predicted number of cystinosis cases is ~500 in the US and ~2000 worldwide; to date, more than 500 cystinosis cases are reported in the literature.

Clinical Phenotypes:

Although cystinosis subjects present with a spectrum of severity, the disease has been categorized into 3 forms. The infantile nephropathic form (MIM #219800) is the most severe and most frequent (95% of cases) and is associated with severely truncating (or null) CTNS mutations. The onset of features occurs several months after birth, progressing to renal Fanconi syndrome, growth failure, photophobia, glomerular dysfunction, and other organ involvement due to cystine accumulation in almost all cells. When untreated, children with nephropathic cystinosis develop end-stage renal failure before age 10 years [32,36]. The juvenile nephropathic form (MIM #219900), also called intermediate or adolescent cystinosis, associates with least one milder (missense, non-truncating) CTNS mutation, and has an onset around age 10–12 years, with proteinuria due to glomerular damage, a mild renal Fanconi syndrome and other symptoms similar to but milder than those in the infantile form. In untreated subjects, glomerular failure usually occurs between ages 15 and 28 years [37]. The ocular, non-nephrotic form of cystinosis (MIM #219750), previously called the adult form, is associated with mild CTNS mutations and is characterized clinically only by photophobia due to corneal cystine crystal accumulation, and usually presents in adulthood [38].

Diagnosis:

The diagnosis of cystinosis has historically been made by measuring elevated cystine content in peripheral blood leukocytes [39,40], or prenatally in chorionic villi or amniocytes [41]. Detection of corneal cystine crystals by slit lamp examination may also establish the diagnosis [42]. Identifying bi-allelic pathogenic mutations in the CTNS gene ultimately confirms the cystinosis diagnosis [43]. With the availability of cysteamine therapy, early diagnosis and clinical treatment is pivotal for optimal outcomes.

Pathomechanism:

The exact pathophysiology of cystinosis remains unknown, but cell damage due to cystine accumulation is believed to trigger certain pathologic mechanisms including (glutathione-depletion related) oxidative stress [44–49], increased apoptosis [45,50,51], mitochondrial dysfunction [48,49], or inflammation [52,53]. Alternative functions of cystinosin have been proposed [54,55], including a role in autophagy [48,56], lysosomal biogenesis/intracellular vesicle trafficking [57–60] and mTOR signaling [61–63]; these may contribute to the pathology of cystinosis.

Therapy:

Remarkably, before identification of the gene defect in cystinosis, two interventions were already developed for nephropathic cystinosis. First, renal transplantation was reported in 1977 to be effective in nephropathic cystinosis [28,64], and after transplantation, the disease does not recur in the graft, but continues to progress in other organs and may cause complications. Second, cysteamine (β-mercaptoethylamine) therapy, was reported in 1976 to dramatically deplete cystinotic cells of cystine [65]. Cysteamine reacts with intralysosomal stored cystine and forms the mixed disulfide cysteamine-cysteine, which exits the lysosome via the cationic amino acid transporter PQLC2 [66]. Subsequent clinical trials and studies demonstrated safety and efficacy of oral cysteamine in delaying renal failure, enhancing growth, and preventing late complications of the disease [67,68]. Topical cysteamine eye drops alleviate cystine corneal deposits [69]. Other therapies have been proposed and tested on animal or cell models, including activation of transcription factor EB (TFEB) [62] and cell-based (gene) therapies [70–72].

2.2. Free Sialic Acid Storage Disease

Background:

Free Sialic acid Storage Disease (FSASD) is an autosomal recessive inherited multisystem neurodegenerative lysosomal storage disease caused by bi-allelic mutations in the SLC17A5 gene (chromosome 6q13; Gene ID 26503), encoding the lysosomal transmembrane protein SLC17A5 (also called SIALIN) [73]. SLC17A5, a twelve transmembrane domain protein, is a lysosomal proton-coupled acidic sugar exporter, known to transport sialic acid (N-acetylneuraminic acid, Neu5Ac) and glucuronic acid (GlcUA) [19,74] (Fig 1A). SLC17A5 defects lead to intra-lysosomal free sialic acid and glucuronic acid accumulation. Lysosomes of FSASD subjects are enlarged upon electron microscopic examination, and abnormally high (~ 10–100x normal) levels of free (i.e., unconjugated) sialic acid are excreted in urine. Progressive cerebellar atrophy and hypomyelination are often detected on brain magnetic resonance imaging (MRI) [73]. The clinical manifestations of FSASD include variable severity of organomegaly, coarse facial features and neurodegenerative symptoms characterized by muscular hypotonia, cerebellar ataxia, and cognitive impairment [73,75–77]. SLC17A5 was found to also act as a non-lysosomal glutamic acid-aspartate cotransporter in synaptic vesicles in the brain; this function may contribute to the neurodegenerative phenotype in FSASD [78].

The worldwide prevalence of FSASD is ~ <1 per 1,000,000 individuals (Orphanet; http://www.orpha.net/). Higher prevalence rates of 1–9/1,000,000 occur in the Salla region in Finland [79], due to a founder SLC17A5 variant: p.Arg39Cys [73,80], hence the name “Salla disease” for forms of FSASD [81]. The same founder variant was also reported in an Old Order Mennonite population [82]. Since 1999, 40 SLC17A5 mutations were reported (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php) in subjects with FSASD [73,80,83]. The number of reported cases is ~200, with at least ~150 of those having the milder “Salla” phenotype [77].

Clinical Phenotypes:

Three forms of FSASD exist [77]. The infantile sialic acid storage disease (ISSD; MIM #269920), associated with truncating (or null) SLC17A5 mutations, is the most severe form and exhibits severe global developmental delay, coarse facial features, hepatosplenomegaly, cardiomegaly and often death in early childhood [76,84]. The intermediate form exhibits moderate to severe global developmental delay, hypotonia, and hypomyelination with or without coarse facial features, and is associated with at least one mild (non-truncating) SLC17A5 variant [75,84]. The Salla disease (MIM #604369) adult phenotype is characterized by normal appearance, mild cognitive dysfunction, and spasticity; it is associated with non-truncating SLC17A5 variants [80,83].

Diagnosis:

FSASD should be considered in individuals with a clinical pattern of global developmental delay or cognitive impairment particularly affecting speech development and regression in combination with coarse facies, organomegaly, truncal muscular hypotonia, ataxia, spasticity, bone anomalies, failure to thrive, and short stature [77,83]. Detecting elevated free sialic acid in urine and/or cerebrospinal fluid can confirm the suspicion of FSASD. The diagnosis was historically confirmed by demonstrating lysosomal (rather than cytoplasmic) localization of elevated free sialic acid in cultured cells [77,84], but is now commonly confirmed by identifying bi-allelic pathogenic variants in SLC17A5 [73,77,83]. A considerable diagnostic is common for FSASD [83] due to the rarity of the disorder, specialized (sialic acid detection) assays unavailable for routine screening, and non-specific clinical features (developmental delay, ataxia, infantile hypomyelination) creating an extensive differential diagnosis [77].

Pathomechanism:

The exact pathophysiology of FSASD remains unknown. SLC17A5 sialic acid transport function, SLC17A5 intracellular localization, and amount of stored free sialic acid has been directly correlated with disease severity and survival [19,83,85]. Nevertheless, how accumulated intra-lysosomal free sialic acid causes disease pathology remains poorly understood, and the contribution of additional stored compounds (i.e., glucuronic acid, gluconic acid) [19,74,86] or alternative transport functions of SLC17A5 (i.e., glutamic acid–aspartate cotransport in brain synaptic vesicles [78], plasma-membrane nitrate transport in salivary gland acinar cells [87]) is unknown. Sialic acids are typically found as the terminal sugars on glycoconjugates [88,89], but it is unknown if SLC17A5 defects affect the process of sialylation. Aberrant sialylation of brain glycoconjugates may contribute to the FSASD,central nervous system hypomyelination, including hyposialylation of myelin- associated glycoprotein (MAG) [19] or polysialic acid-neural cell adhesion molecule (PSA-NCAM) [90,91]. Interestingly, the SLC17A5 knock-out mouse model exhibited decreased mature myelinating oligodendrocytes due to apoptosis and downregulated PSA-NCAM [92].

Therapy:

There is no approved therapy for FSASD. The medical and psychosocial management of subjects is symptomatic and supportive. The fact that survival of individuals with FSASD appears to correlate with the amount of stored free sialic acid [83], suggests that lowering intralysosomal sialic acid may be a therapeutic avenue to pursue. The majority of cases have at least one p.Arg39Cys mutated allele, therapeutic targeting of this variant is also appealing. A multidisciplinary collaborative effort for development of therapeutics for FSASD, involving the NIH, academic clinical scientists, and a patient advocacy group (Salla Treatment and Research (STAR)) [93], was encouraged by a recent similar effort for Niemann-Pick C disease that successfully overcame the scientific, clinical and financial challenges facing the development of new drug treatments [94].

2.3. H syndrome

Background:

H syndrome (also named histiocytosis-lymphadenopathy plus syndrome; MIM #602782) is an autosomal recessive multisystem genodermatosis with systemic macrophage histiocytosis caused by bi-allelic mutations in the SLC29A3 gene (chromosome 10q22.1; Gene ID 55315), encoding the equilibrative nucleoside transporter 3 (ENT3) [20,95,96]. ENT3, an eleven transmembrane domain protein, is a pH-dependent lysosomal transporter that controls the flow of nucleosides, including adenine, adenosine and uridine, from lysosomes to the cytoplasm [20,96] (Fig 1A). ENT3 deficiency result in lysosomal nucleoside buildup, elevated intra-lysosomal pH, defective apoptotic cell clearance and altered macrophage function [97]. The term H syndrome refers to the major clinical findings of hyperpigmentation, hypertrichosis, hearing loss, hepatosplenomegaly, hematological illness, hypogonadism, hallux flexion contractures, heart anomalies, low height, and occasional hyperglycemia/diabetes mellitus [98,99].

The prevalence of H syndrome is ~<1 per 1,000,000 worldwide (Orphanet; http://www.orpha.net/). Since 2008, 28 SLC29A3 mutations were reported in subjects with H syndrome (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). The 2 most common mutations (p.Gly427Ser and p.Gly437Arg) are found in individuals of Arab-Palestinian ancestry [98]. To date, approximately 130 individuals with H syndrome have been reported worldwide.

Clinical Phenotype:

H syndrome usually becomes clinically apparent in childhood [98], after which it runs a variable but progressive course; early-death has been reported [99]. The hallmark of the disorder is cutaneous hyperpigmentation, induration, and hypertrichosis. Sensorineural hearing loss, flexion contractions of the fingers and toes, short stature and hepatosplenomegaly are also common features [99]. Sporadic features are cardiac anomalies, lymphadenopathy, insulin dependent diabetes mellitus, hypogonadism, angiopathy, and genital masses. Exophthalmos (with normal thyroid function), malabsorption (due to pancreatic exocrine insufficiency), renal anomalies, and lytic bone lesions, osteosclerosis and hematological abnormalities are also reported [98]. There is no genotype-phenotype correlation for SLC29A3 mutations with respect to severity or type of the clinical features [98]. Different allelic disorders include: pigmented hypertrichosis with insulin-dependent diabetes mellitus (PHID) syndrome, Faisalabad histiocytosis (FHC), familial histiocytosis syndrome (FHS), familial sinus histiocytosis with massive lymphadenopathy (FSHML; also named familial Rosai-Dorfman disease), and some types of dysosteosclerosis. However, since H syndrome encompasses most of these clinical features, and all share identical mutations in SLC29A3, these disorders are now considered the same entity [96,99–101].

Diagnosis:

A diagnosis of H syndrome is suspected by the pathognomonic cutaneous features. Additional findings such as hearing loss, fixed flexion contractures of fingers and toes, short stature and hepatosplenomegaly should also raise suspicion [99], as well as presence of dilated lateral scleral vessels, corneal arcus and shallow orbits on ophthalmic examination, especially when seen in children [102]. Pediatric rheumatologists should be aware of H syndrome and its association with arthritis, as these specialists often encounter H syndrome subjects prior to diagnosis [103]. Laboratory tests may show elevated erythrocyte sedimentation rates, mild microcytic anemia, and elevated liver enzymes. Histopathological examination of involved skin shows a dense dermal and subcutaneous infiltrate, composed mainly of CD68⁺ histiocytes, which is later replaced by fibrosis [100]. The diagnosis is confirmed by identification of bi-allelic pathogenic variants in SLC29A3 [95].

Pathomechanism:

ENT3 mediates pH dependent efflux of nucleosides, the end products of lysosomal nucleic acid degradation, across the lysosomal membrane [104]. The exact pathomechanism of the combination of clinical features resulting from defective ENT3 remains unknown. However, because of the wide spread biochemical roles of nucleosides, any disruption in metabolism and trafficking of nucleosides could result in various abnormal phenotypes [20,96,101]. While ENT3 was initially shown to localize to lysosomes though a N-terminal lysosomal targeting motif [20], it was later shown to also localize to mitochondrial membranes mediated by a N-terminal mitochondrial targeting motif, where it may transport nucleosides into mitochondria for the formation or repair of mitochondrial DNA and RNA; the mitochondrial ENT3 localization may also underly the mitochondrial toxicity of several nucleoside analogue drugs [105].

Tissue macrophages (histiocytes), which have the highest ENT3 expression among immune cells, phagocytose and degrade dying cells and their nucleic acids in their lysosomes [97]. ENT3 knockout mice exhibit intra-lysosomal nucleoside accumulation, leading to increased lysosomal pH, ineffective apoptotic cell clearance and increased macrophage colony-stimulating factor (M-CSF) signaling, contributing to elevated numbers of macrophages and histiocytosis [97]. Mouse studies suggest a cellular and molecular basis for the development of histiocytosis in ENT3 deficiency and also, potentially, in other LSDs that involve histiocytosis [97]. Another mouse study demonstrated that ENT3 mobilizes the lysosomal adenosine pool to regulate AMP-activated protein kinase (AMPK)/mammalian target of rapamycin (mTOR)/Unc-51 like Autophagy Activating Kinase 1 (ULK) signaling, the disruption of which leads to widespread autophagic insufficiency [106]. This same study reported adult stem cell exhaustion and impaired differentiation capacity due to ENT3 deficiency [106].

Therapy:

There is no approved therapy for H-syndrome, and given the complexity of the syndrome, there is a lack of consensus regarding treatment. Management of symptoms is mainly supportive. Treatment of the inflammatory nature of H syndrome, has been largely unsuccessful [98,107]. Oral steroids may temporarily improve cutaneous changes in some patients but are inappropriate for long term use due to side effects [103]. Tocilizumab appeared effective in some cases [103]. A combination therapy with biologics and other immune suppressants was suggested [103]. Improved understanding of the pathophysiology of H syndrome may aid additional treatment strategies, such as targeting the M-CSF signaling to ameliorate histiocytosis [97], pharmacological activation of AMPK [106] or stem cell therapy [106].

3. LYSOSOMAL DISORDERS OF ION TRANSPORT

3.1. Mucolipidosis type IV

Background:

Mucolipidosis IV (MLIV; MIM #252650) is an autosomal recessive neurodevelopmental lysosomal storage disorder caused by bi-allelic mutations in MCOLN1 (mucolipin 1; chromosome 19p13.2; Gene ID 57192), encoding the mucolipin-1 protein [108,109]. Mucolipin-1, also called TRPML1, is a six transmembrane domain lysosomal protein belonging to the transient receptor potential (TRP) gene family [109], with cationic channel function regulated by Ca²⁺ ions in cells and cell-free systems [110] (Fig 1B). Mucolipin-1-deficiency results in accumulation of heterogenous lipids and proteins in cytoplasmic vacuoles derived from lysosomes [111]. Typical clinical features of MLIV include severe psychomotor delay and ophthalmologic abnormalities starting in infancy, most patients survive into adulthood [112,113].

There are approximately 120 MLIV cases reported, at least 70% of them of Ashkenazi-Jewish descent (Orphanet; http://www.orpha.net/) [108,113,114]. Since 2002, 33 MCOLN1 pathogenic variants were reported in subjects with MLIV (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). Two MCOLN1 variants, a splice site variant (c.406–2A>G) and a 6.4-kb deletion, are frequent in the Ashkenazi Jewish population, [108,115].

Clinical Phenotype:

MLIV is a neurodevelopmental disorder that is also neurodegenerative in about 15% of subjects. The phenotype can be either typical (~95% of subjects) or atypical (~5% of subjects) [112,113]. The first signs usually appear during the first year of life and clinical progression is usually slow [116]. Severe cognitive impairment, eye abnormalities (corneal clouding, blindness can occur), and neurological features (spasticity, hypotonia, and the inability to walk independently) are hallmarks of typical MLIV; symptoms can vary widely among patients [112–114,116–118]. Achlorhydria, which results in increased blood gastrin levels, and iron deficiency anemia are also frequent manifestations [113].;

Diagnosis:

Diagnosing MLIV is difficult, since MLIV subjects have normal levels of mucopolysaccharides, oligosaccharides and lysosomal hydrolases in blood. Although lysosomal inclusions are found in most MLIV tissues, including liver, kidney, skin, and spleen, organomegaly is not a feature of MLIV. Diagnosis may be suspected following detection of autofluorescence in cultured MLIV fibroblasts [112,117,118]. Measurement of elevated blood gastrin levels is a useful, cost-effective diagnostic tool [116] and MRI showing a thin corpus callosum may prompt consideration of MLIV [117]. The diagnosis is ultimately confirmed by identification of bi-allelic mutations in the MCOLN1 gene. MLIV is underdiagnosed, as its common presentation resembling cerebral palsy-like encephalopathy often leads to misdiagnosis, and individuals with atypical milder variants often escape diagnosis.

Pathomechanism:

Mucolipin-1, localized to late endosomal and lysosomal membranes, is a cationic membrane channel, permeable to Ca²⁺, Fe²⁺, Zn²⁺, Na⁺, K⁺, and H⁺, and is modulated by changes in Ca²⁺ concentration [112,119–121]. Mucolipin-1 appears to be involved in many intracellular membrane functions [114], including (Ca²⁺ion-induced) endosomal/lysosomal trafficking [21,122], autophagy [123], lysosomal exocytosis dysregulation [124], [119]mammalian target of rapamycin complex 1 (mTORC1) - transcription factor EB (TFEB) signaling [114,125,126], and heavy metal dyshomeostasis with ROS formation [127–129]. All these functions likely contribute to the spectrum of MLIV pathologies [114] and help explain the heterogenous storage phenotype and organelle dysfunction observed in MLIV tissues and may provide research targets for therapeutic intervention. More research is required to adequately explain the tissue-specific pathologies of the disease.

Therapy:

To date, there is no specific therapy for MLIV. Treatment is mainly symptomatic [112]. Corneal transplantation has not been successful for MLIV because the donor corneal epithelium is eventually replaced by the abnormal host epithelium [118]. Experimental therapies for MLIV are still in early research stages and include the immunomodulator fingolimod to promote astrocyte homeostasis [130], N-butyl-deoxynojirimycin (miglustat) inhibiting glycosphingolipid synthesis [131] bone marrow transplantation [132], drugs targeting the autophagy pathway [133], and a small (chaperoning) molecule for specific MCOLN1 point mutations [134].

3.2. CLCN7-Related Disorders

The CLCN7 gene (chromosome 16p13.3; Gene ID 1186) encodes the chloride voltage-gated channel 7 (CLCN7) membrane protein. CLCN7 is considered an 18-transmembrane domain protein, although some helices only partially span the membrane. The protein is a 2Cl⁻/1H⁺ antiporter regulated by a voltage-gating mechanism, expressed on the ruffled border of bone reabsorbing osteoclasts and on the membranes of late endosomes and lysosomes [10,135] (Fig 1B). Variants in CLCN7 underly three distinct disorders [136], described below. Loss of function defects in CLCN7 result from either recessive (in OPTB4, section 3.2.1.) or dominant (in OPTA2, section 3.2.2.) inherited variants in CLCN7 and lead to osteopetrosis due to decreased bone resorption [135,137,138]. A de novo gain of function variant was described to underlay hypopigmentation, organomegaly, and delayed myelination and development (HOD, section 3.2.3.), with no signs of osteopetrosis [139].

3.2.1. Osteopetrosis, Autosomal Recessive type 4 (OPTB4)

Background:

Osteopetrosis autosomal recessive type 4 (OPTB4; also named osteopetrosis infantile malignant 2; or autosomal recessive osteopetrosis (ARO); MIM #611490), is an osteoclast-rich osteopetrosis caused by bi-allelic mutations in CLCN7 [135,137]. Individuals with OPTB4 display a very severe form of osteopetrosis (increased bone mass and fragility) and hematological failures, in some cases associated with primary neurodegeneration with cerebral atrophy and visual impairment [137].

The worldwide incidence of autosomal recessive osteopetrosis was estimated to be 1 per 250,000 live births [136], only a very small fraction of these are due to bi-allelic CLCN7 defects. Since 2001, ~60 CLCN7 mutations were reported in subjects with OPTB4 (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). There are no apparent ethnic founder variants causing OPTB4 and fewer than 50 OPTB4 cases reported to date.

Clinical Phenotype:

OPTB4 is usually diagnosed at birth or early in infancy due to generalized osteosclerosis, severe hematological deficits, primary neurodegeneration resembling lysosomal storage disease, and cerebral and retinal atrophy [136,137]. OPTB4 is often lethal within the first 10 years of life, due to central nervous system involvement [140], despite improvement of hematological symptoms after hematopoietic stem cell transplantation (HSCT) [136]. The central nervous system involvement can include severe cerebral atrophy, severe spasticity, axial hypotonia and peripheral hypertonia. Visual impairment is found in almost all patients. There are a few OPTB4 cases with milder phenotypes (sometimes reported as intermediate autosomal recessive osteopetrosis (IAO)), which carry at least one mild (missense) CLCN7 variant, suggesting a genotype-phenotype correlation in recessive CLCN7-related disorders [136].

Diagnosis:

OPTB4 is suspected in individuals with characteristic radiographic findings, including generalized osteosclerosis, club-shaped long bones, osteosclerosis of the skull base, bone-within-bine appearance. Involvement of the central nervous system, visual impairment, hearing loss and hematological findings contribute to the suspected diagnosis [136]. Identification of bi-allelic pathogenic variants in CLCN7 establishes the diagnosis.

Pathomechanism:

In osteoclasts, multinucleated giant cells responsible for bone resorption during bone growth and tissue renewal, CLCN7 provides the chloride conductance necessary for efficient proton pumping across the osteoclast ruffled membrane; this achieves the low pH required for bone resorption [135,140]. CLCN7 deficiency leads to decreased or absent Cl⁻ channel function, impairing bone resorption and increasing bone mass (osteopetrosis) and, causing a disequilibrium of bone turnover, deformities, dental abnormalities, impaired mineral homeostasis, and structural bone fragility causing fractures [10,135,140]. The reduction of the bone marrow cavity of CLCN7-deficient individuals may contribute to the hematological features, including anemia, pancytopenia, frequent infections and hepatosplenomegaly [135–137]. An overly dense cranial nerve foramina at the skull base may underlie impaired neurologic functions, progressive blindness, deafness and nerve palsies [135,136].

CLCN7 cooperates with the lysosomal V-type H⁺-ATPase to sense, set and maintain an acid organellar lumen [141]; its role in intraluminal Cl⁻ accumulation may also regulate lysosomal enzymes and appears to drive some endosomal transport processes [142,143]. Impaired endosomal and lysosomal vesicle trafficking may also result in defective osteoclast ruffled-border formation and, hence, the inability to resorb bone and mineralize cartilage.

CLCN7 forms a tight complex with the heavily glycosylated OSTM1 protein, which appears to stabilize CLCN7 and is required for CLCN7 channel function; OSTM1 is dependent on CLCN7 for lysosomal localization [10]. Bi-allelic OSTM1 mutations underlay OPTB5 (MIM #259720) [10,136], with similar, to more severe, clinical features as CLCN7-deficient OPTB4.

Therapy:

Symptomatic treatment of OPTB4 includes: calcium supplementation for hypocalcemic convulsions; treatment of fractures by an experienced orthopedist; antibiotics for leukocytopenia; erythrocyte or platelet transfusions as needed; immunoglobulins for hypogammaglobulinemia; surgical decompression of the optic nerve; and specialized dental care [136]. Hematopoietic stem cell transplantation (HSCT) can be curative for most symptoms except the neurological features; the progressive severe neurologic decline continues in OPTB4 subjects after HSCT. Therefore, HSCT is not usually performed in OPTB4 subjects with central nervous system deficits, due to the little benefit they would gain [136,144,145]. Interferon gamma (IFN-γ) therapy is approved for use in severe recessive osteopetrosis [146], without recent reports on the clinical outcome [145]. A gene therapy-based strategy is under investigation, using genetically modified CD34⁺ cells, isolated from peripheral blood [147].

3.2.2. Osteopetrosis, Autosomal Dominant type 2 (OPTA2)

Background:

Osteopetrosis autosomal dominant type 2 (OPTA2; also called Albers-Schönberg disease or autosomal dominant osteopetrosis type II (ADOII); MIM #166600), is an osteoclast-rich osteopetrosis caused by heterozygous mutations in CLCN7 [138]. Dominant OPTA2 is allelic to recessive inherited OPTB4, and features of the CLCN7 protein and its deficiency are discussed above (OPTB4 section 3.2.1.). Individuals with OPTA2 display a milder phenotype than OPTB4 individuals, including a mild form of osteosclerosis with high risk of bone fractures; neurologic, eye or hematologic involvement is rare in OPTA2 [138,148].

The worldwide incidence of OPTA2 is estimated as high as 1 per 20,000 live births (Orphanet; http://www.orpha.net/) [136,149]. The disease is likely underdiagnosed in milder cases. Since 2001, ~41 CLCN7 variants, mostly missense, were reported in OPTA2 subjects (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). There are more than 250 OPTA2 cases reported to date, with ~70% of them carrying a CLCN7 missense variant.

Clinical Phenotype:

OPTA2 usually presents in late childhood or adolescence with a progressive segmentary osteosclerosis, mainly at the vertebral endplates, iliac wings, and skull base. Bone phenotypes may include severe pain, fractures, scoliosis, hip osteoarthritis, and osteomyelitis of the mandible or septic osteitis or osteoarthritis elsewhere [136,145,148]. Many OPTA2 subjects undergo several orthopedic procedures, complicated by the hardness and the brittleness of their skeletons [148]. Visual loss, bone marrow failure, neurological complications and decreased life expectancy are relatively rare in OPTA2 [136,138,148]. Reduced penetrance (60–70%) of a CLCN7 mutation in OPTA2 results in a wide disease spectrum, [136,148]. No statistically significant OPTA2 genotype-phenotype correlation could be made [148].

Diagnosis:

OPTA2 is suspected based upon characteristic radiographic findings, including osteosclerosis of the spine (a “sandwich vertebra” or “rugger jersey” appearance), finding of endobones (bone-within-bone, mainly in the iliac wings), Erlenmeyer flask-shaped femoral metaphyses, mild osteosclerosis of the skull base, and transverse bands of osteosclerosis in long bones [136,148]. Elevated serum levels of tartrate-resistant acid phosphatase and the BB isoenzyme of creatine kinase may assist in OPTA2 diagnosis and predicting the clinical severity of disease [150]. Identification of dominant inherited, heterozygous pathogenic variant in CLCN7 ultimately establishes the OPTA2 diagnosis.

Pathomechanism:

Pathogenic heterozygous variants in OPTA2 (mostly missense) have an apparent dominant negative effect on CLCN7 chloride channel function, either by affecting dimer formation or voltage gating, or causing CLCN7 mislocalization [10,136,138]. Of note, carriers of recessive OPTAB4 CLCN7 variants are unaffected and appear to have CLCN7 loss of function mutations [138]. The pathomechanism of a defective CLCN7 protein is discussed above (OPTB4 section 3.2.1.).

Therapy:

There is no approved therapy for OPTA2. Supportive therapies include orthopedic treatment for fractures and arthritis with attention to potential post-surgical complications, fractures near joints may require total joint arthroplasty. Analgesics to reduce pain and antibiotics to treat osteomyelitis, good routine dental care and avoidance of activities with high fracture risk are recommended [136,145]. Since there are no data to determine whether HSCT is therapeutic for OPTA2 [145], the risks of performing HSCT may outweigh those of the disease itself. Other OPTA2 therapies are under investigation. Small interfering RNA to silence the heterozygous mutated CLCN7 allele was effective in human OPTA2 cells and in an OPTA2 mouse model [151]. IFN-γ improved the osteopetric phenotype of OPTA2 mice [152], and a clinical trial with IFN-γ in patients was recently completed (ClinicalTrials.gov Identifier NCT02584608).

3.2.3. Hypopigmentation, Organomegaly, and Delayed Myelination and Development (HOD)

Background:

Hypopigmentation, organomegaly, and delayed myelination and development (HOD; OMOM #618541) is caused by a de novo missense variant in CLCN7 [139]. HOD was reported in only 2 unrelated children, a 22-month old Caucasian girl and a 14-month old Ghanaian boy, both carrying the same de novo heterozygous CLCN7 variant, p.Tyr715Cys [139]. Neither of the subjects had osteopetrosis, prevalent in subjects with other CLCN7 variants [136], and p.Tyr715Cys was deemed a gain-of-function variant [139].

Clinical features:

Both reported HOD subjects presented with hypopigmentation of skin and hair but normally pigmented irides. They had organomegaly, involving liver, spleen and kidneys. Both infants had delayed myelination on brain MRI with developmental delay in gross and fine motor skills. Neither subject had osteopetrosis or other radiographic bone abnormalities [139]. Biopsy findings from skin and other organs were consistent with a lysosomal storage disorder [105].

Diagnosis:

HOD should be suspected in young children presenting with hypopigmentation, organomegaly, and delayed myelination. Identification of more affected individuals may improve diagnostic parameters. Identification of a heterozygous p.Tyr715Cys variant or another confirmed gain of function variant in CLCN7 establishes the diagnosis.

Pathomechanism:

The CLCN7 p.Tyr715Cys variant involves a highly conserved amino acid in the C-terminal cytoplasmic domain. An orthologous CLCN7 p.Tyr715Cys knock-in mouse model recapitulated the human HOD phenotype and revealed no osteopetrosis [139]. Expression of CLCN7 p.Tyr715Cys in Xenopus oocytes for current measurements [10] resulted in increased (gain-of-function) chloride current reflecting an increased chloride ion transport into lysosomes [139]. CLCN7 gain-of-function was supported by increased lysosomal acidification in HOD subjects‟ fibroblasts, in which the lysosomal pH was approximately 0.2 units lower than that of control cells, underlying the vacuolar formation in HOD fibroblasts [139]. This phenotype was recapitulated when overexpressing CLCN7 p.Tyr715Cys in control fibroblasts [139]. The CLCN7 p.Tyr715Cys gain of function explains the absence of osteopetrotic symptoms in HOD; common in OPTB4 or OPTA2 cases with CLCN7 loss of function variants [136].

Therapy:

Treatment of HOD fibroblasts with the alkalinizing drug chloroquine normalized the pH and reduced the abundance of large vacuoles [139], supporting the hypothesis that CLCN7 plays a critical role in lysosomal acidification [137,141,153] and encouraging therapeutic studies of chloroquine in HOD. Chloroquine is now being pursued as a therapeutic intervention for HOD.

4. LYSOSOMAL DISORDERS OF EXPORT OF AN ENDOCYTOSED COMPOUND

4.1. Niemann-Pick Disease type C1

Background:

Niemann–Pick disease type C1 (NPC1; MIM # 257220) is an autosomal recessive disorder of the lysosomal cholesterol export machinery due to bi-allelic mutations in the NPC1 gene (chromosome 18q11.2; GeneID 4864) encoding the NPC1 protein [154–156]. NPC1, a 13 transmembrane domain protein, is thought to act in lysosomal export of cholesterol that reaches the late endosomal and lysosomal lumen through endocytosis [157,158] (Fig 1C). NPC1 defects lead to lysosomal storage of cholesterol and other lipophilic compounds, resulting in diverse symptoms affecting brain, motor control, liver and spleen, and resulting in premature death [155,156].

The incidence of NPC is ~1 per 100,000–150,000 live births, but many cases go misdiagnosed or undiagnosed, obscuring the true frequency in the general population (Orphanet; http://www.orpha.net/) [155,156]. Since 1997, ~500 NPC1 gene variants were reported in subjects with NPC1 (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php); ~70% are missense variants. Ethnic isolates have been described in French Acadians in Nova Scotia (p.Gly992Trp) [159], in individuals of Hispanic descent in parts of Colorado and New Mexico (p.Ile1061Thr), in the United Kingdom and in France [160]. NPC1 genotype-phenotype studies, have reported a few missense variants that result in milder phenotypes than other missense variants [155,160–162]. There are more than 500 NPC1 cases reported.

Clinical Phenotype:

NPC1 is a highly variable disease in both age of onset and clinical features, even among affected family members. It ranges from a fatal disorder within the neonatal period to a late onset, chronic progressive disorder that may remain undiagnosed well into adulthood [155,156]. Most cases are detected during childhood and progress to cause life-threatening complications by the second or third decade of life [155,161,162]. Depending on the age of onset, variable clinical features may occur, including prolonged neonatal jaundice or cholestasis, isolated splenomegaly, and progressive, often severe neurological symptoms such as cognitive decline, cerebellar ataxia, vertical supranuclear gaze palsy (VSPG), dysarthria, dysphagia, dystonia, seizures, gelastic cataplexy, and psychiatric disorders [155,161,162]. Death often occurs in the late second or third decade from aspiration pneumonia. Late onset of symptoms can lead to longer life spans but it is extremely rare for any person with NPC1 to reach age 40 [155].

Diagnosis:

NPC1 is suspected based on age of onset and characteristic clinical symptoms, summarized in a “Suspicion Index Tool” designed to aid physicians to make the diagnosis of NPC1 [163,164]. Strong indicators for a NPC1 diagnosis include vertical gaze palsy, enlarged liver, enlarged spleen, or jaundice in young children. Subsequent specialized biochemical assays in fibroblasts, including detecting impaired cholesterol esterification [165], positive filipin staining [156], or blood-based testing for biomarkers (oxysterols, lysosphingolipids, bile acid metabolites), can confirm the diagnosis [166,167]. Genetic testing, identifying pathogenic biallelic variants in NPC1 ultimately confirms the diagnosis [154,161,162]. Of note, most individuals with NPC have mutations in NPC1; fewer than 20 individuals are reported with NPC2 [155], caused by pathogenic variants in the NPC2 gene resulting in a similar clinical, cellular and biochemical phenotype as NPC1 [155,168].

Pathomechanism:

The NPC1 and NPC2 proteins, both equipped with sterol-binding domains, cooperate to remove lipoprotein-derived free cholesterol from late endosomes and lysosomes. Soluble NPC2 binds cholesterol in the ensodsomal/lysosomal lumen and directly transfers it to the transmembrane protein NPC1 [169]. NPC1 is then thought to transport cholesterol into and/or across the membrane [157,158]. In NPC, free cholesterol accumulates in lysosomes and may lead to a deficiency in membrane cholesterol [157,158]. Given the critical role of cholesterol in maintaining membrane function, NPC1 deficiency is reported to trigger many other (membrane) processes, including lipid redistribution, dysfunctional endocytosis, lysosomal apoptosis, dysfunctional endosomal trafficking, impaired autophagy, impaired mitochondria, and cell death [155,156,158,161]. The excessive accumulation of cholesterol within the lysosomal lumen is thought to underly storage of various other lipids in NPC1, such as gangliosides and glycosphingolipids, which may result from aberrant lysosomal acid lipase and other lysosomal enzyme activities due to altered luminal pH and/or stored lipids or chemical interactions among stored intra-lysosomal lipids [155–157].

Therapy:

No curative treatment for NPC1 exists. Individualized supportive treatments include symptomatic therapy for seizures, dystonia, and cataplexy, a nocturnal sedative to help disordered sleep, physical therapy to maintain mobility, speech therapy to optimize swallowing function, and a gastronomy tube to meet caloric needs.

Several therapies for NPC1 have been tested in clinical studies. Liver transplantation corrected the hepatic dysfunction but did not ameliorate the neurologic disease [155]. Although a trial of cholesterol-lowering agents cholestyramine, lovastatin, and nicotinic acid showed that the amount of free cholesterol in the liver of individuals with NPC was reduced, it did not ameliorate the neurologic progression [155]. Preclinical and clinical studies showed that miglustat, an inhibitor of glycosphingolipid synthesis, may slow the progression of neurological symptoms associated with NPC [170,171]. Miglustat is approved for the management of neurologic symptoms of NPC1 in several countries, not including the United States [155,156]. The drug hydroxypropyl beta cyclodextrin (HPBCD), presumed to solubilize lipophilic molecules such as cholesterol, was shown to reduce lysosomal cholesterol accumulation in NPC1 cells [172] and demonstrated efficacy in NPC mouse models [173]. For more efficient delivery of HPBCD to the brain [174]intrathecal administration is currently under investigation in in NPC1 subjects (ClinicalTrials.gov, Identifier NCT01747135 and NCT03471143), and showed initial success [175,176] The small molecule arimoclomol increases expression of heat shock proteins (HSPs), in particular HSP70, which improves the binding of sphingolipid-degrading lysosomal enzymes to their essential cofactor and attenuated the phenotype in NPC1 fibroblasts [177] and neurological features in NPC1 mice [178]. Arimoclomol is currently under investigation in a Phase 2/3 study in NPC1 subjects (ClinicalTrials.gov Identifier NCT02612129).

Other investigational NPC1 therapies are in preclinical stages. Cell studies showed that overexpression of Rab 9 corrects the NPC1 phenotype, suggesting the existence of alternate pathways for mobilizating of endosomal cargoes, which may be potential targets for small-molecule therapies [155,179]. Histone deacetylases (HDACs), an enzyme family involved in regulating gene transcription, are overexpressed in NPC [180]; HDAC inhibitors achieved increased NPC1 expression and a marked reduction of stored cholesterol in NPC1 fibroblasts [180,181], experiments in mice are ongoing [182].

4.2. Cobalamin F Disease

Background:

Cobalamin F (cblF) disease (also named methylmalonic aciduria with homocystinuria, type F; MIM #277380) is an autosomal recessive inborn error of vitamin B12 (cobalamin) metabolism caused by bi-allelic mutations in LMBRD1 (lipocalin membrane receptor domain containing 1; chromosome 6q13; Gene ID 55788), encoding the LMBD1 protein [183,184]. LMBD1, a nine transmembrane domain protein, is presumed to be the lysosomal membrane transporter for cobalamin, thus associating cblF disease with defective cobalamin export from the lysosome [183,185] (Fig 1C). Indeed, free cobalamin accumulates in cblF disease lysosomes, however, these lysosomes appear to have normal ultrastructural morphology, which may be due to the low abundance of cobalamin as compared to sugars or amino acids [184,186]. Typical features of cblF disease include poor growth, developmental delay, macrocytic anemia and seizures [183,184,186,187]. Of note, the origin of the cobalamin “F type” comes from cell fusion experiments that established nine complementation groups for intracellular cobalamin metabolism; in the F type complementation group (including cblF), cobalamin is not converted into enzyme cofactors, resulting in combined accumulation of homocysteine and methylmalonic acid, whose metabolism requires cobalamin as cofactor [188].

The prevalence of cblF is <1 per 1,000,000 individuals worldwide (Orphanet; http://www.orpha.net/). Since 2009, 9 LMBRD1 mutations (all truncating) were reported in subjects with cblF disease (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). A 1-bp deletion (c.1056delG; p.Asn353Ilefs*18) is common in subjects of European ancestry [183]. There are only 16 cblF cases reported to date [183,189].

Clinical Phenotype:

cblF disease often presents in the first year of life with variable features, including poor postnatal growth, feeding difficulties, development delay, megaloblastic anemia, hypotonia, stomatitis and skin rashes. Additional features such as congenital heart malformations, cleft palate, intraventricular hemorrhage, neutropenia, thrombocytopenia, renal and hepatic features may also occur [183,187,189]. Early death has been reported [183]. Since all reported LMBRD1 mutations are truncating and the resulting phenotypes vary widely, there is no apparent genotype/phenotype correlation [183].

Diagnosis:

A cobalamin metabolism defect should be suspected in any child that presents with unexplained recurrent infections, failure to thrive, developmental delay, and /or megaloblastic anemia. Biochemical and/or genetic testing can further assist in the diagnosis. cblF disease should be suspected in individuals with low plasma vitamin B12 levels with an abnormal Schilling test, elevated plasma total homocysteine, and elevated plasma and urine methylmalonic acid. The cblF diagnosis was historically confirmed by complementation analysis in skin fibroblasts using cell lines of the various complementation groups in the cobalamin pathway. Currently, the diagnosis is confirmed by identification of biallelic pathogenic mutations in LMBRD1. An early diagnosis of cblF is valuable since injectable hydroxocobalamin improves survival and may reduce, stabilize or prevent clinical manifestations. Newborn screening based on a genomic testing approach was postulated to benefit not only neonates with cblF [187,190], but also those with other complex clinical presentations [191], based upon the availability of therapies.

Pathomechanism:

LMBD1 is presumed to mediate the efflux of cobalamin across the lysosomal membrane, resulting in accumulation of intralysosomal free cobalamin and a lack of accessible cytoplasmic cobalamin to create the co-factors methylcobalamin and adenosylcobalamin, required for the methylation of homocysteine to methionine in the cytosol and for the conversion of methylmalonyl coenzyme A to succinyl coenzyme A in mitochondria, respectively [188]. The lack of these cofactors results in homocystinuria and methylmalonic aciduria, detectable in plasma and urine. Cobalamin metabolism disorders exhibit similarities in phenotypes, including macrocytic anemia, failure to thrive, and severe neurological disturbances if left untreated by high doses of cobalamin, indicating that defects in (intermediates of) this pathway underly these clinical symptoms. A proposed pathomechanism for the neurological features in cblF deficiency is low adenosylmethionine (caused by an impaired formation of methionine via methionine synthase), which is the rate limiting substrate for transmethylation, important for the maintenance of myelination and neurotransmitter formation (e.g., catecholamines) [187,192]. Of interest, LMBD1 was reported to translocate the ABCD4 (ATP-binding cassette (ABC) subfamily D, member 4) protein, from the ER to lysosomes [193]. ABCD4 contains transmembrane domains and ATPase function, and is presumed to mediate lysosomal release of cobalamin into the cytosol [194], further confirmed by association of ABCD4 defects with cobalamin J disease (cblJ) [12].

Therapy:

Injectable (intramuscular) hydroxycobalamin (intramuscular injection), with or without oral betaine and folic acid, has proven beneficial for survival and may reduce, halt or prevent the primary clinical manifestations of cblF disease [187,189,190]. Early diagnosis and treatment are pivotal for favorable treatment outcomes. Improved metabolic control and correction of hematologic problems is often achieved with hydroxycobalamin treatment, but most cases continue to have signs of motor and language delay, intellectual deficit and abnormal ophthalmologic findings [187,189,190]. Individuals with cblF are advised to avoid situations that result in catabolism, such as prolonged fasting and dehydration, and remain on a weight-appropriate dose of hydroxycobalamin.

5. LYSOSOMAL DISORDERS OF METABOLITE IMPORT

5.1. Mucopolysaccharidosis type IIIC

Background:

Mucopolysaccharidosis type IIIC (MPS IIIC; also named Sanfilippo syndrome; MIM #252930) is an autosomal recessive inherited lysosomal storage disease due to impaired lysosomal degradation of the glycosaminoglycan heparan sulfate, due to bi-allelic mutations in HGSNAT (heparan acetyl-CoA:alpha-glucosaminide N-acetyltransferase; chromosome 8p11.2-p11.1; GeneID 138050), encoding the HGSNAT protein [195,196]. HGSNAT, a lysosomal eleven transmembrane domain protein, is presumed to act both as a transport protein, importing cytosolic acetyl-coenzyme A into lysosomes, and an enzyme that uses cytosolic acetyl-coenzyme A for the acetylation of luminal α-glucosamine residues from heparin and heparan sulfate chains, a step required for their subsequent cleavage by α-N-acetylglucosaminidase [197–200] (Fig 1D), although the exact transmembrane acetylation mechanism of HGSNAT is still debated [201]. Individuals with MPS IIIC present with severe progressive neurodegenerative decline, including neuropsychiatric signs (chaotic behavior, sleep disturbances) and psychomotor deterioration with mild somatic signs and premature death in the 2nd to 4th decade [1,202].

The incidence of MPS IIIC is ~1 per 1000,000 live births worldwide (Orphanet; http://www.orpha.net/) [203]. Specific incidences in live births for MPS IIIC were reported as 0.07 per 100,000 live births in Australia [204], 0.12 per 100,000 in Portugal [205] and 0.21 per 100,000 in the Netherlands [206]. Since 2006, there are 74 HGSNAT variants (~60% missense) reported in MPSIIIC subjects (January 2020; HGMD; http://www.hgmd.cf.ac.uk/ac/index.php). Two variants, p.Arg344Cys and p.Ser518Phe, are frequent among Dutch cases and p.Arg384* is a common variant across populations [207]. Fewer than 100 MPS IIIC cases are reported to date.

Clinical features:

The clinical features of all four types of MPS III (MPS IIIA, B, C, D) are similar [1], although MPS IIIA may have a more severe course [208]. Often there are no symptoms until the second year of life, when subjects experience a highly variable progressive mental deterioration, leading to severe dementia. Neuropsychiatric problems occur within the first decade, including hyperactivity, aggressive behavior, seizures and sleep disorders. Motor problems generally develop in the second decade. Subjects can also develop deafness or loss of vision (retinitis pigmentosa). It is assumed that death, often due to respiratory complications, may occur before age 20, although survival into the fifties has been reported [208]. Only a limited genotype-phenotype correlation could be established [207]. Therefore, prediction of the clinical phenotype at diagnosis remains largely impossible. Interestingly, some cases with higher residual enzyme activity (>15–20% of normal) are described, presenting at age >60 years with retinitis pigmentosa and without other MPS IIIC symptoms [203].

Diagnosis:

Suspicion of MPS III, based on clinical symptoms, can be confirmed for all types of MPS III by detecting increased concentration of heparan sulphate in the urine. Enzymatic assays in leukocytes and/or fibroblasts historically confirmed the diagnosis and allowed for discrimination among the different subtypes of the disease [209]. The diagnosis of MPS IIIC is ultimately confirmed by identification of biallelic pathogenic mutations in HGSNAT.

Pathomechanism:

Negligible HGSNAT residual activity (< 3% of normal) was detected in MPS IIIC patients‟ cells, including those with missense variants [203,210]. HGSNAT missense variants were reported to result in protein misfolding and retention in the endoplasmic reticulum followed by proteasome degradation, instead of being targeted to the lysosome [203,210,211]. Storage of heparan sulfate, especially in the brain, likely underlies the severe neuropsychiatric features of MPS III disorders. MPS IIIC mouse models showed storage of heparan sulfate in microglial cells, suggested to trigger a downstream cascade of reactions including neuroinflammation and neuronal death [203]. Mouse neurons showed disorganized mitochondrial networks and accumulation of auto-fluorescent material resembling that detected in ceroid lipofuscinosis [203].

Therapy:

There is no specific treatment for MPS IIIC. Therapies are mainly supportive rather than curative and require multidisciplinary management. The behavioral issues do usually not respond to medication. Delivery to the brain and the membrane location of the HGSNAT protein exclude MPS IIIC from effective lysosomal enzyme replacement therapy and underly low effectiveness of bone marrow transplantation or blood-forming (haemopoietic) stem cell transplants. In addition, early intervention is essential for MPS IIIC; by the time the disease has been diagnosed clinically, it may be too late for some treatments to be effective. Some progress is being made on gene or cell-based therapies in MPS IIIC mouse models [203]. Since ~50% of MPS IIIC cases have missense mutations, pharmacological chaperone therapies are also being explored [203,211].

6. OTHER LYSOSOMAL DISORDERS OF MEMBRANE TRANSPORT

6.1. Ceroid Lipofuscinosis Neuronal 3 (CLN3)

Background:

Neuronal ceroid lipofuscinosis 3 (CLN3; also called Batten disease, juvenile ceroid lipofuscinosis, or Spielmeyer-Sjogren disease; MIM #204200) is an autosomal recessive severe neurodegenerative disorder caused by bi-allelic mutations in the CLN3 gene (ceroid lipofuscinosis, neuronal 3; chromosome 16p12.1; Gene ID 1201), encoding the lysosomal transmembrane protein CLN3, also called battenin (BTN1, BTS or JNCL) [212–215]. CLN3 is a 6 transmembrane protein, mainly residing in endosomal and lysosomal membranes of most cells, its precise function remains unknown [213,215,216] (Fig 1D). CLN3 disease is characterized by and named after „ceroid lipofuscin‟, an autofluorescent lysosomal storage material in cells. The disease clinically presents with retinal degeneration in the first decade of life followed by progressive neurological deterioration and death in the second to third decade of life [212–214].

CLN3 is the most common form of ceroid lipofuscinosis and is one of the most common neurodegenerative disorders affecting children. The prevalence of CLN3 is estimated to be 1 per 1,000,000 and the incidence was estimated to be 3 per 1,00,000 in the USA, and 1 per 250,000 in Europe (https://rarediseases.org/rare-diseases/batten-disease/). It occurs with higher frequency in families of northern European Scandinavian descent [217]. Since 1995, 78 CLN3 mutations were reported in CLN3 subjects (January 2020; HGMD: http://www.hgmd.cf.ac.uk/ac/index.php; NCL Resource: https://www.ucl.ac.uk/ncl/cln3.shtml). A ~1-kb CLN3 deletion is common [218] and occurs on at least one allele among >85% of CLN3 subjects; it is believed to be a founder mutation of European ancestry [212–214,216,218]. More than 450 affected individuals are reported to date (NCL Resource: https://www.ucl.ac.uk/ncl/cln3.shtml).

Clinical Phenotype:

The onset of CLN3 is typically juvenile, between the ages of 4 and 10 years. Progressive visual loss due to retinal degeneration is usually the first sign, which can progress to severe visual impairment within one to two years and to blindness by adolescence [212,217,219]. Around the age of 9–18 years, children with CLN3 start to manifest loss of motor coordination, mental decline, and seizures. Behavioral problems, extrapyramidal signs, and sleep disturbance may also occur in the second decade [217,220]. Over time, affected individuals lose the ability to walk or sit independently and require wheelchair assistance. In rare cases, CLN3 subjects develop cardiac problems in adolescence or early adulthood, including heart rhythm abnormalities and an increase in the size of the heart muscle (hypertrophic cardiomyopathy). Most people with CLN3 disease live into early adulthood [212,214,217]. Individuals with a missense CLN3 variant have a milder phenotype, which still includes visual failure, but a much milder course of neurodegeneration and a longer life expectancy [213,216].

Diagnosis:

A CLN3 diagnosis should be suspected in young subjects who initially developed normally but then present with a progressive retinopathy (usually age 4–10) and later (usually age 9–18) develop an unexplained progressive neurological disorder with dementia, epilepsy, and motor deterioration [213,214]. The diagnosis of CLN3 is suggested by the presence of vacuolated lymphocytes in peripheral blood detectable by light microscopy of blood smears. These vacuoles may look empty or contain a solid autofluorescent material (ceroid lipofuscin) that displays typical rectilinear „fingerprint profiles‟ on ultrastructural studies. Similar vacuolization can be detected in skin biopsies, skeletal muscle biopsies and in ganglionic neurons [213,214,216]. Identification of bi-allelic mutations in the CLN3 gene ultimately confirms the diagnosis [212,214].

Pathomechanism:

While CLN3 protein function remains greatly unknown, there is evidence that CLN3 is a lysosomal membrane transporter, including its 6 transmembrane domains, presence of lysosomal targeting motifs, AP1 and AP3 mediated lysosomal targeting, endo-lysosomal membrane localization, similarity to the SLC29 family of equilibrative nucleoside membrane transporters (ENTs) and similarity to the major facilitator superfamily (MFS) of small-solute membrane transporters, and lysosomal accumulation of compounds upon CLN3 deficiency [24,213,216]. Various functional studies proposed roles for CLN3 in lysosomal acidification, lysosomal arginine import, membrane fusion, vesicular transport, cytoskeletal-linked functions, autophagy, apoptosis, and post-Golgi trafficking [213,214,216]. Biochemical analysis of the ceroid lipofuscin lysosomal storage material of CLN3-deficient cells showed presence of dolichols, dolichol phosphates, alcohol-insoluble lipids, iron, and proteins [221], implying a role in lipid and protein turnover. The predominant accumulating protein was subunit c of the mitochondrial ATP synthase F₀, [216,221], implicating a role of CLN3 in lysosomal turnover of mitochondria [222]. Interactions among different CLN proteins were described [216,223], which may explain the overlapping pathological/clinical spectra of CLN subtypes. Apart from endo-lysosomal membranes, CLN3 was also found located in synaptosomal fractions of neuronal cells, possibly playing a role the neurodegeneration of CLN3 subjects [216].

Therapy:

No approved curative therapy for CLN3 exists. Current treatment is symptomatic and palliative only. Seizures, malnutrition, gastroesophageal reflux, pneumonia, sialorrhea, depression and anxiety, spasticity, and dystonia can be adequately managed with medications [214]. To treat the neuroinflammation and autoimmunity in CLN3 [224], immunosuppressive treatment with prednisolone showed limited efficacy [225,226] but trials with short term mycophenolate mofetil administration did not show a benefit (ClinicalTrials.gov Identifier: NCT01399047) [227]. Other CLN3 therapeutic approaches are under pre-clinical investigation and several have shown efficacy in CLN3-deficient mice and/or human CLN3 fibroblasts [228]. These include the drugs EGIS-8332 and Talampanel (LY300164) which target AMPA receptors [229,230], the drug MK2206 or the disaccharide trehalose, both dephosphorylating transcription factor EB (TFEB), stimulating its nuclear translocation [231], and phosphodiester-4 inhibitors to increase cAMP production [232]. Gene therapy using an intravenous AAV9-CLN3 or AAVrh.10-CLN3 vector partially corrected the neurological phenotype in CLN3-deficient mice [233]. A gene therapy clinical trial in CLN3 subjects using intracranial injections of AAV9-CLN3 is ongoing (ClinicalTrials.gov Identifier: NCT03770572).

6.2. Ceroid Lipofuscinosis Neuronal 7 (CLN7)

Background:

Neuronal ceroid lipofuscinosis 7 (CLN7; (Turkish variant) late infantile ceroid lipofuscinosis; MIM #601951) is an autosomal recessive inherited neurodegenerative disorder caused by bi-allelic mutations in the MFSD8 gene (major facilitator superfamily domain containing 8; chromosome 4q28.2; Gene ID 256471), encoding the lysosomal transmembrane protein MFSD8, also called CLN7 [23,234]. MFSD8 is a 12 transmembrane protein residing in lysosomes, with a membrane transport function of unknown etiology [23] (Fig 1D). Late-infantile onset CLN7 presents with progressive symptoms including neurological deterioration with seizures and ataxia, visual impairment, and premature death typically in teenage years [23,234,235].

The prevalence of all forms of CLN are estimated at 1 per 1000,000 individuals worldwide [214]. A very small portion of these are CLN7 cases, whose prevalence is unreported. Since 2007, 51 MFSD8 mutations were reported in subjects with CLN7 disease (January 2020; HGMD: http://www.hgmd.cf.ac.uk/ac/index.php). An MFSD8 missense founder variant, p.Lys294Thr, occurs in the Roma population [234]. To date, ~104 CLN7 cases have been reported (https://www.ucl.ac.uk/ncl-disease/ncl-resource-gateway-batten-disease/).

Clinical Phenotype:

The CLN7 disease onset typically begins between ages 2 and 7. The most common initial symptoms are seizures and developmental regression; ataxia or visual impairment can also occur early [234,235]. Affected children may later develop muscle twitches (myoclonus), difficulty coordinating movements (ataxia), speech impairment, and vision loss. Mental functioning and motor skills decline with age. Individuals with CLN7 disease typically do not survive past their teens [214,235]. A CLN7 genotype-phenotype correlation appears to exist: some milder gene variants result in atypical disease progression including isolated retinopathy [23,234,235], or isolated macular dystrophy with central cone involvement but without neurologic features [236]. Similar to other CLN subtypes, CLN7 blood smears show vacuolated lymphocytes on light microscopy, and ultrastructural studies of CLN7 cells show rectilinear and fingerprint profiles with occasional curvilinear profiles [214,234].

Diagnosis:

The age of onset, clinical presentation, disease course, and ultrastructural morphology of storage material of individuals with CLN7 disease is indistinguishable from that of other late infantile CLN forms (CLN2, CLN5, CLN6, or CLN8) [214,234]. Therefore, for an accurate diagnosis of CLN7, identification of bi-allelic pathogenic variants in the MSFD8 gene is the most practical method [23]. Of note, biallelic MSFD8 variants including one mild (missense) should be considered in the differential diagnosis of nonsyndromic macular dystrophy with central cone involvement but without neurological symptoms [236].

Pathomechanism:

While MSFD8 protein function remains elusive, there is evidence that it is a lysosomal membrane transporter. The protein has 12 transmembrane domains, an N-terminal dileucine lysosomal targeting motif, a sugar (and other metabolites) transporter domain, homology to the major facilitator superfamily (MSF) of transporter proteins, lysosomal membrane localization, and lysosomal accumulation of compounds resulting from MSFD8 deficiency [23,214]. The substrates that MFSD8 transports are unknown. Although CLN7 is ubiquitously expressed, high expression levels have been seen in specific neuronal cell types and retinal cells, i.e., the most severely affected tissues. Murine MFSD8 localizes to the photoreceptor presynaptic terminals in the outer plexiform layer, which may help explain the ocular phenotype in CLN7 subjects [235]. The neurodegeneration in a CLN7 mouse model was attributed to dysfunctional lysosomes and impaired autophagy in brain tissue [237].

Therapy:

No approved curative therapy for CLN7 exists. Current treatment is symptomatic and palliative only, similar to therapy for CLN3 above (Section 6.1.). No research studies of CLN7-specific therapies are reported.

7. DISCUSSION AND FUTURE DIRECTIONS

Since the discovery of the first gene associated with a human lysosomal membrane transporter disorder in 1998, the CTNS gene underlying cystinosis [25], only 9 other such genes have been identified (Table 1). This is a remarkably small number, considering that at least 40 lysosomal membrane transport proteins exist [2]. One explanation is that not all lysosomal membrane transporters are characterized at the molecular or functional levels [4–6], other reasons include the lack of a specific phenotype (Table 2) or specific diagnostic tests to identify subjects with these disorders, and embryonic lethality or functional redundancy between transporters may also obscure identification of other human lysosomal membrane transporter defects. The increased interest of cell biologists in lysosomal biology, rapidly improving experimental and diagnostic tools, new animal models, and increased funding for rare disease research and therapeutics, will undoubtedly help fill this gap. This is exemplified by the recent discovery by the NIH Undiagnosed Diseases Network of a de novo CLCN7 variant underlying hypopigmentation, organomegaly, and delayed myelination and development (HOD) in two unrelated infants [139].

Only a few small molecule therapies are approved for specific disorders of lysosomal membrane transporters, including cysteamine for cystinosis [67,68], miglustat for NPC1 [155,156], IFN-γ for severe cases of OPTB4 [145], and hydroxycobalamin for clbF [187] (Table 1). Some therapies are in clinical trials, including IFN-γ for OPTA2 (ClinicalTrials.gov Identifier NCT02584608), chloroquine for HOD [139], HPBCD for NPC1 (NCT01747135 and NCT03471143), arimoclomol for NPC1 (NCT02612129), and mycophenolate mofetil for CLN3 (NCT01399047) (Table 1). Most small molecule drugs under preclinical or clinical investigation work toward minimizing disease symptoms rather than curing the disease. These drugs often require frequent (lifelong) administration, and treatment of neurological symptoms remains challenging because difficulty of drug delivery to the brain.

Since lysosomal membrane transporter disorders are monogenic diseases, gene therapy approaches have the potential to correct underlying genetic defects, offering a cure rather than simply managing symptoms. Moreover, successful gene therapy may require only a single dose to confer lifelong improvement. Gene therapy approaches that cross the blood brain barrier are emerging [238–241]. Since protein replacement therapies for transmembrane proteins are more complex than for enzymes or other soluble proteins [242], investments in gene therapy approaches will benefit therapeutic development for disorders of lysosomal transmembrane proteins. So far, gene therapy for disorders of lysosomal membrane transporters has only reached the clinical stage for CLN3 subjects using intracranial injections of AAV9-CLN3 (ClinicalTrials.gov Identifier: NCT03770572) [228]. Gene-based therapy for other disorders, including for cystinosis [70] and OPTA2 [151], is progressing.

Most therapies for inherited disorders of lysosomal membrane proteins are based on either reduction of intralysosomal metabolites (e.g., cysteamine treatment of cystinosis) or extralysosomal provision of stored metabolites (e.g., hydroxycobalamin for cobalamin F disease); they are specific for each individual disorder and expensive to develop for only a small patient cohort (2–500 cases per disorder; Table 1). However, as a group, disorders of lysosomal membrane transporters comprise at least 2,500 reported cases, which may be an incentive for development of therapies that target common features of this group of membrane transporters, i.e., cell-based therapies, chaperone-based therapies, activation of transcription factor EB (TFEB) and utilizing endosomal transport mechanisms.

Cell-based therapies for inherited disorders of lysosomal membrane transporters are under investigation. Hematopoietic stem cell transplantation (HSCT) was explored for several of these disorders [71,72,106,132,145,147,203] and, although it may eliminate some symptoms (e.g., in OPTB4 [144,145]), HSCT is not curative for the neurological features. Therefore, HSCT is not commonly performed in these disorders, since the HSCT risks may outweigh those of the disorder itself [145]. Addressing the neurological symptoms of disorders of lysosomal membrane transporters has proven difficult and requires therapeutics that cross the blood-brain barrier. An alternative treatment involves intrathecal injections, which is currently under investigation with HPBCD for NPC1 patients (NCT01747135 and NCT03471143).

Chaperone-based therapies may be effective for specific point mutations that result in membrane protein misfolding and degradation, mislocalization, or impaired channel activity. A recent example is the small molecule arimoclomol that increases expression of the human heat shock protein 70 (HSP70), which improves the binding of sphingolipid-degrading lysosomal enzymes to their essential cofactor and attenuates the phenotype in NPC1 [177], including the neurological symptoms in NPC1 mice [178]. Arimoclomol is currently investigated in a Phase 2/3 study in NPC1 subjects (ClinicalTrials.gov Identifier NCT02612129). Similarly, small molecule chaperones were reported to restore channel function and rescue disease phenotype for specific MCOLN1 point mutations in MLIV cells [134], and chaperones are being explored for MPS IIIC, since ~50% of affected individuals harbor HGSNAT missense mutations [203,211]. It is worth investigating if the known chaperones for specific lysosomal membrane transporter defects are also effective for variants in other lysosomal transporters.

Activation of the transcription factor EB (TFEB) has emerged as a therapeutic mechanism for LSDs [125]. Increasing TFEB expression and/or nuclear translocation results in upregulated lysosomal biogenesis and function, including the autophagy and exocytosis pathways; this may help deplete LSD-lysosomes of their stored components and/or renew cells and lysosomes. Signaling of the lysosomal membrane-associated mTORC1 kinase complex, involved in lysosomal nutrient sensing and TFEB activation [243], is affected in MLIV because MCOLN1 (TRPML1) Ca²⁺ transport is pivotal for mTORC1-mediated TFEB activation [125,133], but it also appears to be affected in cystinosis [61,63]. Activation of TFEB, using the tryrosine kinase inhibitor genistein, rescued lysosomal abnormalities in cystinotic kidney cells [62]. The drugs MK2206 or the disaccharide trehalose were both shown to diminish Akt kinase activity, preventing phosphorylation of TFEB resulting in its translocation into the nucleus, which triggered enhanced clearance of proteolipid aggregates, reduced neuropathology and prolonged survival of CLN3 mice [231]. These findings open new perspectives for clinical application of TFEB-mediated enhancement to other lysosomal membrane transporter disorders.

Utilizing endosomal transport mechanisms along the endocytic pathway may uncover potential novel therapeutic targets for disorders of lysosomal membrane transporters. One such approach involved identifying impaired Rab function in NPC1 fibroblasts, followed by overexpressing Rab 9, leading to clearance of the stored endo-lysosomal lipids [155,179]. Other potential approaches could utilize the dileucine and tyrosine-based lysosomal targeting motifs in the cytosolic tails of lysosomal membrane transporters [16,17], or address adaptor-complex mediated lysosomal targeting pathways [16], or enhance drug delivery through utilizing endosomal trafficking and sorting mechanisms [244].

In sum, we reviewed the current knowledge of inherited disorders of lysosomal membrane transporters, to enhance awareness, diagnosis, pathomechanisms and (preclinical) research initiatives for this unique group of disorders.

HIGHLIGHTS.

• Membrane transporter defects form a distinct group of lysosomal storage disorders

• Inherited disorders of lysosomal membrane transporters are underdiagnosed

• Lysosomal membrane transporters can be categorized by their transport function

• Lysosomal membrane transporters are targeted through distinct sorting motifs

• Enzyme replacement therapy is ineffective for lysosomal membrane transporter defects

• Therapies for lysosomal membrane transporter disorders as a group are being explored