Background
A deficiency of alpha-galactosidase A (GLA, E.C.3.2.1.22) leads to Fabry disease (FD), an X-linked lysosomal storage disorder. The impaired glycosphingolipid catabolism provokes progressive accumulation of glycosphingolipids, mainly globotriaosylceramide (Gb3), resulting in a multisystemic disease [
1]. Progressive accumulation of Gb3 leads to macro- and microangiopathic alterations with transient ischemic attack (TIA), stroke, myocardial infarction, life-threatening cardiac arrhythmia and end-stage renal disease resulting in a 10–15 years reduced life-span without adequate therapy [
2]. First symptoms of FD include neuropathic pain attacks due to small fiber neuropathy (SFN) [
2],[
3]. FD routine diagnosis is based on the described typical clinical picture, decreased enzymatic GLA activities and analysis of coding
GLA variants.
The non-coding
GLA -10T allele (rs2071225), located within the 5′-untranslated region (UTR), has been suggested to be associated with decreased GLA protein expression [
4], although the -10T allele co-segregates in a haplotype background with three additional intronic variants (IVS281_-77delCAGCC [rs5903184], IVS4-16A>G [rs2071397], and IVS6-22C>T [rs2071228]) [
5],[
6]. This haplotype has been reported in patients with SFN of unknown etiology as well as in patients with classical FD [
5],[
6]. Until now, the functional role of the -10T allele and the co-segregating intronic variants remains unclear [
6]. In contrast to mutations in coding regions affecting peptide sequences and possibly modifying protein structure and function, the consequences of intronic sequences are not predictable. As shown for the mid-intronic
GLA mutation IVS4+919A>G, intronic variations can affect the process of alternative splicing [
7],[
8]. Due to this mutation, a weak splice site can be converted, resulting in an increased recognition and the insertion of an intronic sequence into the GLA transcript leading to a cardiac phenotype of FD [
7],[
8].
In general, efficient splicing of pre-mRNAs depends on conserved intronic sequences. The efficiency of splicing can further be modified by splicing enhancers or suppressors, sequences located within exons and introns. As opposed to the well-defined consensus splice sites, these elements are not completely characterized. Thus, a prediction whether a genomic variation affects splicing is not possible yet and the impact has to be confirmed experimentally [
9].
In the current work, we retrospectively analyzed 15 -10T allele carrying patients from our database after presentation of a symptomatic index patient with a neurological phenotype. Our complementary approach included clinical data and detailed molecular functional analyses.
Discussion
Patients suffering from Fabry-typical manifestations without variants in the coding region represent a diagnostic/therapeutic dilemma as they are usually not treated with enzyme replacement therapy (ERT). In the current study, we describe a patient cohort suffering from neurological manifestations carrying the
GLA -10T haplotype, which has been suggested to be associated with FD in family studies and case reports [
4]–[
6]. To the best of our knowledge, the molecular functional basis of the -10T haplotype has not been analyzed yet.
Our main findings are: 1) Symptomatic carriers of the -10T haplotype suffer from neurological manifestations and show decreased GLA mRNA expression levels. 2) Intronic variants transmitted together with -10T have no impact on GLA mRNA splicing. 3) The -10T allele is functionally relevant, leading to a significant decrease of GLA transcriptional activity as well as T allele-specific protein-DNA binding. 4) GLA gene expression depends on the lysosomal “master regulator” TFEB.
Thirteen patients with FD-typical neurological manifestations such as stroke, TIA, WML and SFN with neuropathic pain presenting at our Fabry center, had no classical FD-typical genetic variants, but the -10T haplotype. Recently, Pisani et al. [
5] described a classical FD phenotype in a young female patient suffering from TIA associated with the -10T haplotype. In addition, Oliveira et al. [
4] reported that the -10T haplotype was associated with cryptogenic cerebrovascular small-vessel disease. This genetic combination was also observed with high frequency (16%) in patients with SFN of unknown etiology [
6]. These individual reports support our findings that the -10T haplotype might be associated with neurological manifestations.
Due to the retrospective study design skin biopsies, cerebral MRI and QST were not available for all patients of our study cohort. Skin biopsies of four patients out of five revealed SFN, with typical reduction of IENFD and focal axonal swelling of intra-epidermal axons. QST indicated functional impairment of A-delta- and C-fibers in 80% of the examined patients. All patients diagnosed with SFN by skin biopsy were also positive in QST analyses and further examination revealed no evidence for SFN of other etiologies. Transmission electron microscopy of two patients (#6 and #7) revealed lysosomal inclusions representing myelin-like figures, which could be precursors of FD-pathognomonic zebra bodies. A high proportion of the patients had reduced GLA mRNA expression levels, which were associated with reduced GLA enzyme activity in three patients, two hemizygous males and one 72 year old female patient. Notably, two of these patients presented WML or stroke as cerebrovascular manifestation. After exclusion of the functional impact of the co-segregated intronic variants, the observed neurological manifestations, which might be based on GLA mRNA expression reduction, may depend on the identified functional impact of the -10T promoter allele.
Our analysis confirms that the presence of the -10T allele caused altered transcription factor binding and reduced
GLA promoter transcriptional activity as a potential cause for the observed reduction of GLA expression in our patients. Of note, the -10T allele showed a dose-dependent reduction of GLA expression in our patients. The observed mRNA reduction translated into reduced enzymatic GLA activity in three patients when compared to the reference values and hemi-/homozygous carriers of the minor T allele in our cohort tended to have decreased enzymatic activities compared to heterozygous T allele carriers. Consistently, previous studies on -10T allele carriers also reported slightly reduced or normal GLA activities in plasma or leucocytes [
5],[
6]. Considering the observed clinical picture of our patients, it seems conceivable that neurological manifestations may occur in certain, mainly hemi- or homozygous -10T individuals, while GLA activities in plasma or leucocytes are still in the reference range.
Of note, plasma lyso-Gb3 levels of our patients were within the reference range. Lyso-Gb3 is proposed as a potential biomarker for classical FD phenotypes. The level of plasma lyso-Gb3 in affected patients is often higher than in mildly-affected patients [
29],[
30]. Notably, it has been emphasized that atypical FD variants are often not associated with increased lyso-Gb3 levels, although biopsies of affected organs revealed lamellar inclusion bodies characteristic for FD [
31],[
32]. A FD screening program in young patients with cryptogenic ischemic stroke identified a patient with the -10T haplotype [
33]. The plasma lyso-Gb3 level of this patient was moderately increased in the first measurement and normal at retest, suggesting variability of the lyso-Gb3 concentration. These observations suggest that lyso-Gb3 is of limited use in FD variants.
The neurological system of FD patients is highly sensitive to reduced GLA enzyme activity, leading to neuropathic pain as one of the first symptoms in childhood [
34],[
35]. The potential and possibly mild long-term effect of the -10T allele on GLA expression might therefore manifest predominantly in the central and peripheral nervous system, which seems to be more susceptible to differences in GLA expression. Lately, other non-classical FD mutations have been described, resulting also in predominantly neurological manifestations [
36],[
37]. Furthermore, our observations suggest that the observed neurological phenotype might be the result of GLA expression dysregulation in the presence of the -10T allele and an additional yet unknown neurological factor. This would explain Fabry-like disease manifestations even if GLA enzyme activities are measured within the low normal range. Life-time exposure to dysregulated GLA expression may thus lead to neurological manifestations.
A clinical approach with ERT in the index patient (agalsidase-beta, 1.0 mg/kg BW i.v. every other week, Fabrazyme, Genzyme) was started in 2006 and cardiac, renal and neurological examinations as well as biochemical analyses were performed in yearly follow-ups. ERT led to clinical stabilization of the FD manifestations, significant reduction of neuropathic pain, increase of daily activity and quality of life. Of note, within the first year of therapy the index patient showed significant pain reduction under ERT with standard dose of agalsidase-beta (1.0 mg/kg BW i.v. every other week, Fabrazyme, Genzyme). General disease stabilization and increased physical activity was also observed. During the worldwide agalsidase-beta shortage and subsequently dose reduction (0.5 mg/kg BW i.v. every other week, Fabrazyme, Genzyme) pain increased and daily physical activity was dramatically reduced due to physical weakness. This observation is underlined by Weidemann et al. [
38] who reported a significant increase of pain attacks and crisis under enzyme reduction in a multi-center study for FD patients. The latter symptoms were completely restored in the index patient some months after re-switch to a regular dose of agalsidase-beta.
With respect to our findings, we suggest that patients carrying the -10T haplotype suffering from typical neuropathic pain and after exclusion of concurrent etiologies should be treated with established symptomatic medication (e.g. anticonvulsants blocking calcium channels, anticonvulsants and local anesthetics blocking sodium channels, opioids, etc.). In patients with therapy-resistant neuropathic pain ERT could be an option as shown in our index patient. Additionally, WML are no primary indication for ERT, but progression of lesion load should be discussed as ERT-indication, since ERT has been shown to ameliorate endothelial dysfunction in Fabry patients [
39]–[
42].
Of note, the frequency of the -10T allele is reported to be 7% within Caucasian populations [
43]. Between 07/2011 and 12/2013, 49 patients, which were diagnosed as classical Fabry patients by genetic testing, presented at the Fabry center in Muenster. In the same period, 12 symptomatic patients were identified as -10T allele carriers without any further FD-causing mutation. This observation indicates an estimated incidence of ~1:11,500 symptomatic -10T allele carriers among all -10T allele carriers.
Acknowledgments
We thank the patients whose participations make this work possible. We thank Arndt Rolfs (University of Rostock, Germany) for lyso-Gb3 measurements. The technical assistance of Samira Schiwek and Birgit Orlowski is gratefully acknowledged. EA.hy926 cells were a kind gift of Cora-Jean S. Edgell (University of North Carolina, USA). IHKE cells were a kind gift of Eberhardt Schlatter (University Hospital Muenster, Germany). The pcDNA3.1(+) TFEB was generated using a p3xFLAG-CMV-10, a kind gift of Andrea Ballabio (University Naples, Italy). The exon trapping vector pSPL3 was kindly provided by Jörg Gromoll (University Hospital Muenster, Germany). We acknowledge support by the Deutsche Forschungsgemeinschaft and Open Access Publication Fund of the University of Muenster. EB was supported by a Heisenberg professorship from the Deutsche Forschungsgemeinschaft (Br1589/8-2). Parts of this manuscript were supported by Shire Human Genetic Therapies. The funder had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
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
All authors have contributed to the article by participating in the conception and design (MS, ML, BS, SMB, EB), acquisition of data (MS, ML, TD, KG, BS, CT, SS, DM, IK, JW, EB) or analysis and interpretation of data (MS, ML, TD, BS, SMB, EB), drafting the article (MS, ML, TD, BS, SMB, EB) or revising it critically for important intellectual content (KG, CT, SS, DM, IK, JW). All authors read and approved the final version of the manuscript.