Renata Voltolini Velho

Subunit interactions of the disease-related hexameric GlcNAc-1-phosphotransferase complex

The multimeric GlcNAc-1-phosphotransferase complex catalyzes the formation of mannose 6-phosphate recognition marker on lysosomal enzymes required for receptor-mediated targeting to lysosomes. GNPTAB and GNPTG encode the α/β-subunit precursor membrane proteins and the soluble γ-subunits, respectively. Performing extensive mutational analysis, we identified the binding regions of γ-subunits in a previously uncharacterized domain of α-subunits comprising residues 535-698, named GNPTG binding (GB) domain. Both the deletion of GB preventing γ-subunit binding and targeted deletion of GNPTG led to significant reduction in GlcNAc-1-phosphotransferase activity. We also identified cysteine 70 in α-subunits to be involved in covalent homodimerization of α-subunits which is, however, required neither for interaction with γ-subunits nor for catalytic activity of the enzyme complex. Finally, binding assays using various γ-subunit mutants revealed that residues 130-238 interact with glycosylated α-subunits suggesting a role for the mannose 6-phosphate receptor homology domain in α-subunit binding. These studies provide new insight into the assembly of the GlcNAc-1-phosphotransferase complex, and the functions of distinct domains of the α- and γ-subunits.

Identification of the interaction domains between α- and γ-subunits of GlcNAc-1-phosphotransferase

The disease‐associated hexameric N‐acetylglucosamine (GlcNAc)‐1‐phosphotransferase complex (α2β2γ2) catalyzes the formation of mannose 6‐phosphate residues on lysosomal enzymes required for efficient targeting to lysosomes. Using pull‐down experiments and mutant subunits, we identified a potential loop‐like region in the α‐subunits comprising residues 535–588 and 645–698 involved in the binding to γ‐subunits. The interaction is independent of the mannose 6‐phosphate receptor homology domain but requires the N‐terminal unstructured part of the γ‐subunit consisting of residues 26–69. These studies provide new insights into structural requirements for the assembly of the GlcNAc‐1‐phosphotransferase complex, and the functions of distinct domains of the α‐ and γ‐subunits.

Enigmatic in vivo GlcNAc-1-phosphotransferase (GNPTG) transcript correction to wild type in two mucolipidosis III gamma siblings homozygous for nonsense mutations.

Mucolipidosis (ML) III gamma is a rare autosomal-recessive disorder caused by pathogenic mutations in the GNPTG gene. GNPTG encodes the γ-subunit of GlcNAc-1-phosphotransferase that catalyzes mannose 6-phosphate targeting signal synthesis on soluble lysosomal enzymes. ML III gamma patients are characterized by missorting of lysosomal enzymes. In this report, we describe the probable occurrence of mRNA editing in two ML III gamma patients. Patients A and B (siblings) presented at the adult age with a typical clinical picture of ML III gamma, mainly compromising bone and joints, and high levels of lysosomal enzymes in plasma and low levels in fibroblasts. Both were found to be homozygous for c.-112C>G and c.328G>T (p.Glu110Ter) mutations in genomic DNA (gDNA) analysis of GNPTG. Analysis of complementary DNA (cDNA), however, showed normal genotypes for both patients. Low GNPTG mRNA expression was observed in both patients. The mRNA editing can explain the differences found in patients A and B regarding gDNA and cDNA analysis, and the mild clinical phenotype associated with homozygosity for a nonsense mutation. Our results suggest that mRNA editing can be more frequent than expected in monogenic disorders and that GNPTG analysis should be performed on gDNA.

GNPTAB missense mutations cause loss of GlcNAc-1-phosphotransferase activity in mucolipidosis type II through distinct mechanisms

Mucolipidoses (ML) II and III alpha/beta are lysosomal storage diseases caused by pathogenic mutations in GNPTAB encoding the α⁄β-subunit precursor of GlcNAc-1-phosphotransferase. To determine genotype-phenotype correlation and functional analysis of mutant GlcNAc-1-phosphotransferase, 13 Brazilian patients clinically and biochemical diagnosed for MLII or III alpha/beta were studied. By sequencing of genomic GNPTAB of the MLII and MLIII alpha/beta patients we identified six novel mutations: p.D76G, p.S385L, p.Q278Kfs*3, p.H588Qfs*27, p.N642Lfs*10 and p.Y1111*. Expression analysis by western blotting and immunofluorescence microscopy revealed that the mutant α⁄β-subunit precursor p.D76G is retained in the endoplasmic reticulum whereas the mutant p.S385L is correctly transported to the cis-Golgi apparatus and proteolytically processed. Both mutations lead to complete loss of GlcNAc-1-phosphotransferase activity, consistent with the severe clinical MLII phenotype of the patients. Our study expands the genotypic spectrum of MLII and provides novel insights into structural requirements to ensure GlcNAc-1-phosphotransferase activity.

Site-1 protease and lysosomal homeostasis

The Golgi-resident site-1 protease (S1P) is a key regulator of cholesterol homeostasis and ER stress responses by converting latent transcription factors sterol regulatory element binding proteins (SREPBs) and activating transcription factor 6 (ATF6), as well as viral glycoproteins to their active forms. S1P is also essential for lysosome biogenesis via proteolytic activation of the hexameric GlcNAc-1-phosphotransferase complex required for modification of newly synthesized lysosomal enzymes with the lysosomal targeting signal, mannose 6-phosphate. In the absence of S1P, the catalytically inactive α/β-subunit precursor of GlcNAc-1-phosphotransferase fails to be activated and results in missorting of newly synthesized lysosomal enzymes, and lysosomal accumulation of non-degraded material, which are biochemical features of defective GlcNAc-1-phosphotransferase subunits and the associated pediatric lysosomal diseases mucolipidosis type II and III. The early embryonic death of S1P-deficient mice and the importance of various S1P-regulated biological processes, including lysosomal homeostasis, cautioned for clinical inhibition of S1P. This article is part of a Special Issue entitled: Proteolysis as a Regulatory Event in Pathophysiology edited by Stefan Rose-John.

The Lysosomal Protein Arylsulfatase B Is a Key Enzyme Involved in Skeletal Turnover: ARSB IS REQUIRED FOR BONE REMODELING

Skeletal pathologies are frequently observed in lysosomal storage disorders, yet the relevance of specific lysosomal enzymes in bone remodeling cell types is poorly defined. Two lysosomal enzymes, ie, cathepsin K (Ctsk) and Acp5 (also known as tartrate‐resistant acid phosphatase), have long been known as molecular marker proteins of differentiated osteoclasts. However, whereas the cysteine protease Ctsk is directly involved in the degradation of bone matrix proteins, the molecular function of Acp5 in osteoclasts is still unknown. Here we show that Acp5, in concert with Acp2 (lysosomal acid phosphatase), is required for dephosphorylation of the lysosomal mannose 6‐phosphate targeting signal to promote the activity of specific lysosomal enzymes. Using an unbiased approach we identified the glycosaminoglycan‐degrading enzyme arylsulfatase B (Arsb), mutated in mucopolysaccharidosis type VI (MPS‐VI), as an osteoclast marker, whose activity depends on dephosphorylation by Acp2 and Acp5. Similar to Acp2/Acp5–/– mice, Arsb‐deficient mice display lysosomal storage accumulation in osteoclasts, impaired osteoclast activity, and high trabecular bone mass. Of note, the most prominent lysosomal storage accumulation was observed in osteocytes from Arsb‐deficient mice, yet this pathology did not impair production of sclerostin (Sost) and Fgf23. Because the influence of enzyme replacement therapy (ERT) on bone remodeling in MPS‐VI is still unknown, we additionally treated Arsb‐deficient mice by weekly injection of recombinant human ARSB from 12 to 24 weeks of age. We found that the high bone mass phenotype of Arsb‐deficient mice and the underlying bone cell deficits were fully corrected by ERT in the trabecular compartment. Taken together, our results do not only show that the function of Acp5 in osteoclasts is linked to dephosphorylation and activation of lysosomal enzymes, they also provide an important proof‐of‐principle for the feasibility of ERT to correct bone cell pathologies in lysosomal storage disorders.

Next-generation sequencing corroborates a probable de novo GNPTG variation previously detected by Sanger sequencing

Letter to the Editor Mucolipidosis III (MLIII) is an autosomal recessive disease caused by pathogenic variations in theGNPTAB (MLIII alpha/beta) or GNPTG (MLIII gamma) genes. GNPTAB and GNPTG encode, respectively, the α/β and γ subunits of GlcNAc-1-phosphotransferase, the enzyme responsible for catalyzing the addition of a mannose-6-phosphate residue to lysosomal hydrolases, allowing their entry into lysosomes [1,2]. In 2014, our group published a paper describing an MLIII gamma patient, born to a non-consanguineous couple, who carried the variations c.244_247dupGAGT at exon 3 and 328G N T at exon 5, both detected by Sanger sequencing of GNPTG. Analysis of parental DNA using the same technique showed c.328G N T in the father and no alterations in the mother, despite using different pairs of primers and different tissues. Maternity was confirmed, and c.244_247dupGAGT was thus considered a de novomutation [3]. Next-generation sequencing (NGS) has proven more sensitive than Sanger sequencing for the detection of mosaicism when the variants are located in exons, as in the present case [4–10]. However, NGS has several limitations, such as incomplete acceptable coverage of coding regions due to regions rich in CpG islands or repetitive sequences [11]. Thus, we decided to analyze DNA samples obtained from the blood of the patient and her mother using the Ion 314TM Chip Kit v2 and sequence GNPTG with the Ion PGM Hi-Q Sequencing Kit (Thermo Fisher Scientific). The fragment of GNPTG where c.[244_247dupGAGT] is located was amplified 406 and 1047 times in the maternal and patient samples, respectively. No alteration was identified in the maternal sample, while both were identified in the patient, confirming our previous findings. As we failed to detect mosaicism using this technique, the possibility of a de novo event is even higher. Our findings corroborate the hypothesis that de novo variations may be more frequent than expected [12], and speak strongly in favor of always investigating the parents of a child presenting with an autosomal recessive disease in order to confirm heterozygous status.