Background
Sanfilippo syndrome or Mucopolysaccharidosis III (MPS III), is a group of autosomal recessive lysosomal storage disorders caused by mutations in genes encoding enzymes responsible for heparan sulfate degradation [
1]. There are four types of the disease, depending on the gene affected. They all present similar clinical symptoms, including severe central nervous system degeneration accompanied by mild somatic manifestations [
2]. Sanfilippo syndrome type C (MPS IIIC) is caused by mutations in the
HGSNAT gene. This gene codes for the acetyl-CoA: α-glucosaminide N-acetyltransferase (EC 2.3.1.78), a protein localized in the lysosomal membrane which catalyses the acetylation of the terminal glucosamine residues of heparan sulphate prior to their hydrolysis by α-
N-acetyl glucosaminidase [
3]. The
HGSNAT gene, identified by two independent groups in 2006 [
4],[
5], is located at the chromosome 8 (8p11.1) and contains 18 exons. The cDNA encodes a polypeptide of either 635 or 663 amino acids, since there is a controversy concerning the real initiation codon [
6],[
7]. To date, 64 mutations have been reported, 16 of which (25%) involve splicing alterations: 13 are described as splicing mutations and three as small deletions and duplications that affect the splicing process (HGMD®Professional Spring 2014.1 Release).
Splicing is an essential step for the expression of most of human genes, in which the 5’ and 3’ splice sites (ss), the branch point sequence and the polypyrimidine tract (both the last two within 50 nucleotides upstream of the 3’ ss) play a fundamental role. These sites present sequence variability throughout the human genome. The splicing process is conducted by the spliceosome, which is formed of five small nuclear RNAs (snRNAs) and more than 200 different proteins (reviewed in [
8]). The U1 snRNA, which presents a partially complementary sequence to the 5’ ss, is essential for the recognition of the 5’ ss consensus motif (CAG/GTRAGT, exon/intron, R = purine). The U1 snRNP is composed of a 164-nt long U1 snRNA and several protein factors. The 5’ region of the U1 snRNA is involved in the recognition of the 5’ ss, with the C8 nucleotide being the one that binds the first nucleotide (G) of the intron (reviewed in [
9]). Application of modified U1 snRNAs to improve recognition of mutated 5’ ss represents a new strategy for recovering the normal splicing process. They have been assayed as a therapeutic approach for different diseases and splicing mutations affecting different positions of the 5’ ss with variation in the efficacy of the treatment [
10]-[
19]. Recently, assays correcting an exogenous injected construct have been performed using modified U1 snRNAs in mice as a treatment for severe human factor VII deficiency [
20].
In some cases, alternative splicing caused by specific mutations can give rise to misfolded proteins, which may be prone to rapid intracellular degradation (reviewed in [
21]). Molecular chaperones are proteins that act on the correct folding of other polypeptides in cells. Pharmacological and chemical chaperones are small compounds that can be used, in a similar way, to avoid the misfolding of mutant proteins. They are principally potent enzyme inhibitors which interact specifically with their active sites to restore the correct folding and to increase stability [
22]. In the case of lysosomal storage disorders, once in the lysosome, the enzyme substrate replaces the chaperone, thereby completing restoration of enzyme activity [
23]. Aminosugars and iminosugars are the most common pharmacological chaperones used in enzyme enhancement therapy (EET) for lysosomal disorders. EET have been reported for several of these diseases including Fabry disease, G
M1-gangliosidosis, Morquio B disease, Pompe disease, Gaucher disease, Krabbe disease, Niemann-Pick A/B and C diseases; as well as for many other types of disorders such as retinitis pigmentosa, cystic fibrosis, Parkinson’s disease, Alzheimer’s disease and cancer (reviewed in Ref. [
21]). In the case of Sanfilippo syndrome type C, glucosamine, a competitive HGSNAT inhibitor, has been shown to increase residual enzyme activity in cultured skin fibroblasts from patients affected with a number of missense mutations [
24].
In this work we focus on
HGSNAT mutations that affect the splicing process. Three previously described splicing mutations [
5],[
25],[
26] were the object of the study with different modified U1 snRNAs to improve recognition of the donor ss and enhance the correct splicing process. The studies were performed using cells transfected with minigene constructs bearing the specific mutation as well as cultured patients’ skin fibroblasts. Mutations include the most frequent change in Moroccan patients (c.234 + 1G > A), a mutation found in Spanish patients (c.633 + 1G > A) and a mutation found in a French patient (c.1542 + 4dupA). Furthermore, EET approach, was tested for a splicing mutation c.372-2A > G. This mutation is the most prevalent in Spanish and Portuguese patients [
25],[
27] and affects the acceptor site at the end of the 4
th intron of the
HGSNAT gene [
27], thereby altering the splicing process. The use of a downstream alternative cryptic ss generates an mRNA with an in-frame deletion that codes for a protein with the loss of four amino acids (p. [L125_R128del]). Here, we show the effect of this mutation, which reduces enzyme activity, and the recovery of a part of that activity through treatment with glucosamine as a chaperone.
Our results show that, depending on the context of the mutated donor site, modified U1 snRNAs can be a promising therapeutic tool. The use of glucosamine as a chaperone improved enzyme activity, suggesting a therapeutic effect of this compound for Sanfilippo C patients.
Methods
Mutation analysis of the HGSNATgene
This study included five MPS IIIC patients: three previously described, two Spanish and one Moroccan [
25]; and two recently diagnosed, one French and one Portuguese, carrying mutations already reported by us [
25],[
26] (Table
1). Studies were approved by the authors’ Institutional Ethics Committee and conducted under the Declaration of Helsinki. Patients were encoded to protect their confidentiality. Genetic analysis was performed using control and patients’ fibroblast cell lines as the source of RNA and genomic DNA whenever necessary. Total RNA was extracted using High Pure RNA Isolation Kit (Roche, Basel, Switzerland) and converted into cDNA using High-Capacity cDNA Reverse Transcription Kit (Life Technologies, Carlsbad, CA) following the manufacturers’ instructions. The RT-PCR amplifications were performed using the primers described in Additional file
1: Table S1 and information regarding specific conditions for each cDNA amplification and sequence analysis is available upon request.
Table 1
Genotype and origin of MPS IIIC affected patients
SFCP
| Portugal | c.234 + 1G > A | c.234 + 1G > A | This study |
SFC3
| Morocco | c.234 + 1G > A | c.234 + 1G > A | |
SFC6
| Spain | c.633 + 1G > A | c.1334T > C | |
SFC7
| Spain | c.372-2A > G | c.372-2A > G | |
SFC13
| France | c.1542 + 4dupA | c.1150C > T | This study |
Minigene cloning and U1 snRNA expression constructs
For the
in vitro splicing approaches, wild type (WT) and mutant minigenes were constructed for each mutation under study. A gene fragment including exon 2 and the flanking intronic regions was amplified from the DNA of the c.234 + 1G > A patient’s fibroblast cells using the primers described in Additional file
1: Table S1, and cloned into the TOPO vector (Life Technologies). The insert was excised with EcoRI, purified using Wizard® SV Gel and PCR clean-up system (Promega, Madison, WI), and subsequently cloned into the pSPL3 vector (Exon Trapping System, Life Technologies; kindly provided by Dr. B. Andresen) using the Rapid Ligation Kit (Roche Applied Science, Mannheim, Germany). Theclone containing the desired mutant insert in the correct orientation was identified by restriction enzyme analysis and DNA sequencing. The exon 6 (c.633 + 1G > A) and exon 15 (c.1542 + 4dupA), together with their intronic flanking sequences, were also cloned in the pSPL3 vector. These mutated minigenes were ordered from GenScript (Piscataway, NJ). In all cases, the WT minigenes were obtained by site-directed mutagenesis using the QuikChange II XL site-directed Kit (Agilent Technologies, Santa Clara, CA) following the manufacturer’s instructions and nucleotide changes were confirmed by sequencing analysis. The primers used are listed in Additional file
1: Table S1.
To express WT U1 snRNA (U1-WT), we used the vector pG3U1, which includes the sequence coding for human U1 ([
28]; kindly provided by Dr. F. Pagani). The different U1 vectors adapted to the donor ss of
HGSNAT exon 2, exon 6 and exon 15 were obtained by site-directed mutagenesis using the QuikChange II XL site-directed Kit (Agilent Technologies). For each construct (U1 suppressor 1 to 9) the PCR reaction was performed with the specific primers shown in Additional file
1: Table S1. The desired mutations were confirmed by sequence analysis.
Cell culture and U1 transfection experiments
To perform the splicing assays, COS-7 cells and fibroblasts were grown in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich, St. Louis, MO) supplemented with 10% fetal bovine serum (Life Technologies) and 1% PenStrep (Life Technologies) at 37°C with 5% CO2. For co-transfection experiments, COS-7 cells at 90% of confluence were transfected in 6 or 12-well plates with either 1 or 2 μg of each WT or mutant minigene and 1 to 4 μg of each different mutation adapted U1 snRNA vector, using Lipofectamine 2000 (Life Technologies) according to the manufacturer’s protocol. When required, the amount of DNA was adjusted with the pSPL3 empty vector. For splicing analysis of the endogenous HGSNAT transcripts, healthy control and patients’ fibroblasts at 90% of confluence were transfected in 6-well plates with 1, 2.5 or 3.5 μg of each modified U1 snRNA vector using either Lipofectamine 2000 or LTX (Life Technologies) according to the manufacturer’s instructions. To estimate transfection efficiency, healthy control and patients’ cells were transfected with a control plasmid encoding either GFP or RFP and fluorescent cells were monitored by microscopy.
RT-PCR transcript analysis after transfection of modified U1 snRNAs
Cells were harvested 24 or 48 h after transfection. Total RNA was extracted using High Pure RNA Isolation Kit (Roche Applied Science) and converted into cDNA using High-Capacity cDNA Reverse Transcription Kit (Life Technologies). The RT-PCR splicing analysis for minigene transfections was performed using the typical pSPL3 primers SD6 and SA2 (Additional file
1: Table S1). For the endogenous experiments, RT-PCRs were performed using the following primers: HGSNAT-Exon 2F/HGSNAT-Exon 3R for mutation c.234 + 1G > A, HGSNAT-Exon 5F/HGSNAT-Exon 13R for c.633 + 1G >A/c.1334T > C and HGSNAT-Exon 12F/HGSNAT-Exon 16R for the c.1542 + 4dupA/c.1150C > T mutation. In the case of the c.234 + 1G > A mutation, the forward primer was designed to anneal in the middle of exon 2 to amplify only the transcripts in which the correct splicing process was recovered. For the other mutations, the last nucleotide of one of the primers corresponded to the point mutation of the other allele (but with the WT nucleotide) to favour that only the cDNA from the splice-mutation allele was amplified. The RT-PCR products were sequenced to confirm their identity. All the primers used are listed in Additional file
1: Table S1 and information regarding the amplification conditions is available upon request.
Expression of recombinant human mutant and wild type HGSNATin COS-7 cells
COS-7 cells cultured in Eagle’s minimal essential medium supplemented with 10% fetal bovine serum to 70% confluence were transfected with the pcTAP-HGSNAT plasmid and mutant pcDNA-HGSNAT-L125_R128del plasmid as previously described [
24]. 48 h after transfection the cells were harvested and analysed for N-acetyltransferase enzymatic activity or by Western blot.
Twenty-four h after transfection with pcTAP-HGSNAT or mutant pcDNA-HGSNAT-L125_R128del plasmids, COS-7 cells were grown for 72 h in Eagle’s minimal essential medium supplemented with 10% fetal bovine serum containing 10 mM glucosamine in 6-well plates. Then the cells were kept for another 24 h in the normal culture medium without glucosamine. The cells were harvested, lysed by freeze-thaw in 750 μl of water and assayed for HGSNAT activity. Three independent experiments (each with 2 cell plates) were performed on 3 different occasions.
Confluent primary cultures of skin fibroblasts of the MPS IIIC patient homozygous for the c.372-2A > G mutation and healthy control fibroblasts (n = 5) were grown in 6-well plates for 72 h in Dulbecco’s minimal essential medium supplemented with 10% fetal bovine serum containing 10 mM glucosamine. Then the cells were kept for another 24 h in the medium without glucosamine, harvested, lysed in 500 μl of water and assayed for HGSNAT activity. Three independent experiments (each with 2 cell plates) were performed on 3 different occasions.
Enzyme assay
HGSNAT N-acetyltransferase enzymatic activity was measured using the fluorogenic substrate 4-methylumbelliferyl β-D-glucosaminide (4MU-β-GlcN; Moscerdam, Oegstgeest, Netherlands) as previously described [
24]. The protein concentration was measured according to the method of Bradford using a commercially available reagent (BioRad, Hercules, CA).
Statistical analysis
Statistical analysis of the data has been performed by two-tailed unpaired t-test using the Prism GraphPad software.
Western blotting
Cell lysates from 3 independent transfections (20 μg of total protein each) were analysed by Western blot as previously described [
24] using rabbit polyclonal antibodies raised against human HGSNAT Q52-N156 peptide (Sigma-Aldrich HPA029578, dilution 1:5000, incubation overnight at 4°C). Detection was performed with an anti-rabbit IgG antibodies-HRP conjugate (ref. 7074S, Cell Signalling, Beverly, MA), and the enhanced chemiluminescence reagent (ref. 32106, Thermo Scientific, Waltham, MA).
Discussion
In this study we describe two therapeutic approaches specific for several splicing mutations in the
HGSNAT gene that lead to defects in mRNA processing in five Sanfilippo C patients, each carrying at least one splicing mutation. To our knowledge, this is the first attempt to treat Sanfilippo syndrome type C splicing mutations with modified U1 snRNAs. The chaperone treatment, the second therapeutic strategy examined in the present work, was previously tested
in vitro for several missense Sanfilippo C syndrome mutations with promising results [
24] and it is applied here for the treatment of a mutant protein lacking four amino acids.
For patients carrying mutations in the donor ss (SFCP, SFC3, SFC6 and SFC13), modified U1 snRNAs have been tested as a therapeutic tool to recover the normal splicing process. This approach has previously been tested for different mutations in several disorders showing different efficiencies at rescuing the normal transcripts [
10]-[
19]. A total of 9 different U1s have been developed, as well as the U1-WT. None of them affected the normal splicing process in WT minigenes or healthy control fibroblasts when overexpressed.
Few assays to correct a +1 5’ ss mutation have been reported. In some of those, the efficiency of modified U1 snRNAs for splicing mutations in canonical positions +1 and +2 has been shown to be inexistent [
10],[
13]. However, Hartmann
et al. [
19] showed a partial correction of a +1 mutation in a case in which some degree of normal splicing was conserved in the mutant allele and the alternative donor site with a “gt” dinucleotide in positions +5 and +6 presented a low score according to different predictors. In this report, we present three cases with splicing mutations at position +1, two homozygous patients with the c.234 + 1G > A mutation (SFCP and SFC3) and one heterozygous patient with the c.633 + 1G > A mutation (SFC6). When we analysed the treatment efficiency of different modified U1 snRNAs in minigene constructions carrying these mutations, we were not able to detect restoration of the normal splicing process. Instead, we detected alternative splicing patterns due to the use of highly conserved “gt” nucleotides at positions +5 and +6 as a donor site, giving rise to a transcript that includes 4 intronic nucleotides. This result was obtained using three of the four modified U1 snRNAs for the c.234 + 1G > A mutation and the total complementary U1 snRNA for the c.633 + 1G > A mutation. Due to the presence of these “gt” dinucleotides at position +5 and +6 in many introns, it is important to sequence putative rescued transcripts to check whether this alternative site was used.
Despite these results, and taking into account that minigenes only included partial intronic sequences which could lack some splicing regulatory sites and that they were assayed in non-human cells, modified U1 snRNAs were tested on patients’ fibroblasts. Partial rescue (almost 50%) of the normal splicing for the c.234 + 1G > A mutation was observed in patients SFCP and SFC3 overexpressing the U1-sup4. In the case of the c.633 + 1G > A mutation (patient SFC6), no rescue was observed after the modified U1 snRNAs overexpression. Analysis of the ss sequences using the Human Splicing Finder predictor [
29], indicated that the alternative site of patient SFC6 had a high score (68.17/100), while that of patients SFCP and SFC3 had a null score. This could explain the difference in the rescue results between these two “+1” mutations: the mutant allele of patient SFC6 would efficiently use the alternative site and, thus, would not be rescued.
The molecular reason why the modified U1 promotes, in general, a new splicing process using the +5 + 6 “gt” donor site when its sequence perfectly matches the mutated +1 + 2 site remains unclear. One explanation could be the involvement of U6 snRNA complementarity, in addition to that of U1, in the choice of alternative ss in proximity to the normal one, which has been previously described [
30],[
31].
The rescue of the splicing mutation shown here is one of the few positive results for a “+1” mutation using a modified U1 snRNA and the first one performed on an allele that did not produce any of the correctly spliced mRNAs when untreated. It is important to note that a few natural U2-type 5’ ss that present “au” instead of the normal “gu” at the first two nucleotides of the intron have been described [
32],[
33]. These introns present “ac” instead of “ag” at the last two nucleotides. Thus, one could think that mutated “au” sites could promote an alternative splicing using an “ac” acceptor site. However, for this +1 mutation the “au-ac” alternative splicing type was not observed. On the contrary, we found that the “au” 5’ ss correctly paired with the normal “ag” and not with a possible cryptic “ac” 3’ ss, restoring the normal splicing of intron 2.
Further studies are needed to improve the efficacy and to test the toxicity and side effects of U1 overexpression in order to confirm the feasibility of the use of this modified U1 snRNAs in therapy. Overexpression of U1 snRNA vectors introduced with adeno-associated virus into mice liver has been shown to be toxic at high viral doses but safe at low doses, suggesting the viability of this treatment when low amounts of U1 snRNA viral particles are injected [
20].
In the case of the c.1542 + 4dupA mutation (patient SFC13), it is important to note that the canonical “gt” site is not affected by the mutation. However, in this case there is also another “gt” dinucleotide in positions +6 and +7 (one nucleotide downstream due to the duplication). The minigene analysis showed that the splicing that uses the alternative donor site occurs, even without treatment, together with that causing exon 15 skipping. This alternative splicing is slightly enhanced with U1-sup7 and greatly enhanced with U1-sup9. Many different ss predictors were used in order to estimate the score for each donor site in this case (Additional file
4: Table S2). Clearly, in the absence of mutation, the scores of both sites are similar; but when the mutation is present, the normal site presents a lower score while the alternative site increases its score. This is consistent with the fact that in the presence of the mutation the alternative “gt” site is the only one used (see Figure
2I). As it was previously discussed, it remains unclear why the modified U1 snRNAs, designed to perfectly match the mutated site, enhance the splicing using the alternative site. When the modified U1 snRNAs were tested on patient’s fibroblasts, no rescue was detected: either for the normal mRNA or for the one including the five intronic nucleotides. The latter was detected without any treatment, indicating that it takes place in normal conditions in the patient’s cells. In any case, the alternative transcript would not restore the enzyme activity since it does not keep the frame, so it could not have any therapeutic use. Many different mutations in the +3 to +6 positions have been partially or totally rescued using U1 snRNAs [
10]-[
13],[
15],[
17],[
18], while one +5 mutation was not corrected [
15]. Our current results point to the importance of the presence of an additional “gt” dinucleotide in the region which, depending on the sequence context, may be used as an alternative donor site. In previous reports where mutations have been partially or totally corrected and the “gt” dinucleotide was present at positions +5 and +6 [
15],[
19], the alternative site presented a low score in accordance with different predictors, in contrast to the cases described here.
Finally, we attempted to rescue the c.372-2G > A mutation (present in SFC7), using molecular chaperone. This mutation gives rise to two different splicing processes in fibroblasts: one causing exon 4 skipping; the other using an alternative acceptor site, 12 nucleotides downstream of the normal site and producing a transcript with an in-frame deletion. We showed that this alternative splicing produces a protein lacking 4 amino acids p. [L125_R128del] that has some residual activity but the majority of the protein does not reach the lysosome, remaining in the ER due to its misfolding.
In order to restore the correct protein folding, cells transfected with the plasmid carrying the mutant protein and the patient’s fibroblasts were treated with glucosamine chaperone, which resulted in a significant increase in the residual HGSNAT activity. This indicates the feasibility of this therapeutic approach for patients carrying this splicing mutation. Different chaperones assays for lysosomal disorders have been performed before and some have been tested in humans, showing their safety and potential as a therapeutic tool (reviewed in Ref. [
21]).
Conclusions
In conclusion, the results of two therapeutic approaches for different splicing mutations varied depending on the nature of the mutation. For the treatment of donor ss mutations, U1 snRNAs could represent a feasible option. This would depend on the presence of alternative donor sites close to the normal site that could interfere with the correction process and, thus, with the success of the therapy. This is important since many introns present a “gt” in positions +5 and +6. In the present study, we have shown that it is possible to partially recover the normal splicing process for +1 mutations, which was reported only once before. Additionally, a chaperone treatment using glucosamine for a mutant protein with a loss of 4 amino acids, caused by an acceptor ss mutation, has been shown to result in a significant increase in the enzyme activity. These promising results encourage further research into the therapeutic use of U1 snRNAs and chaperones to treat Sanfilippo syndrome type C patients.
Acknowledgements
The authors would like to thank Dr. Lúcia Lacerda from Centro de Genética Médica Dr. Jacinto Magalhães - Centro Hospitalar do Porto, Portugal for providing the SFCP patient fibroblasts sample and Helena Ribeiro from the same institution for the technical support. We thank also Dr. Mónica Sousa and Dr. Elsa Logarinho research groups from Instituto de Biologia Molecular e Celular (IBMC), Porto for the collaboration in the electroporation studies. We would like equally to thank Xavier Roca from the School of Biological Sciences, Nanyang Technological University, Singapore, for advice on the U1s constructs and to the Institut de Bioquímica Clínica, Barcelona, for their collaboration. The authors are also grateful for the support of the Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER), which is an initiative of the ISCIII. This study was partially funded by a grant from the Spanish Ministry of Science and Innovation (SAF2011-25431) and from the Catalan Government (2009SGR971). We are also grateful for the permanent support, including financial aid, from 'patient-support' associations, such as Jonah's Just Begun-Foundation to Cure Sanfilippo Inc. (USA), Association Sanfilippo Sud (France), Fundación Stop Sanfilippo (Spain), Asociación MPS España (Spain). LM was supported by a grant (SFRH/BD/64592/2009) from the Fundação para a Ciência e Tecnologia IP (FCT) /POPH/FSE, Portugal, IC by a grant from the University of Barcelona (APIF), Spain and AVP by an operating grant MOP111068 from Canadian Institutes of Health Research.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LM and IC were involved in the conception and design of the study, performed the experimental work of the U1 snRNA part, the analysis and interpretation of the data, and participated in the drafting and revising of the manuscript. LD and YC were involved in the conception, design and experimental work, as well as in the interpretation of the data and the drafting of the manuscript regarding to the glucosamine part. SA, DG, LV and AVP supervised all the research contributing critically to the design of the work and data interpretation as well as in the revision of the manuscript. MJP, PJ, LRD and BP are collaborators with experience in the field and supervised the research. All authors read and approved the final manuscript.