Genotype-Phenotype Correlation Reanalysis in 83 Chinese Cases with OCRL Mutations

Background Both Lowe syndrome and Dent-2 disease are caused by variants in the OCRL gene. However, the reason why patients with similar OCRL gene mutations presented with different phenotypes remains uncertain. Methods Children with hemizygous pathogenic or likely pathogenic variants in OCRL were compiled from published and unpublished consecutive cases from China. Furthermore, a Chi-square test was employed to analyze the correlation of the location and types of mutations on the phenotype of children with Lowe syndrome or Dent-2 disease. Results Among the total 83 patients, 70.8% (34/48) cases of Lowe syndrome presented with truncating mutations, while only 31.4% (11/35) cases of Dent-2 disease presented with truncating mutation (Χ2 = 12.662; P < 0.001). Meanwhile, the majority of mutations in Dent-2 disease are located in Exon 2–12 (21/35, 60.0%), while the majority of mutations in Lowe syndrome are located in Exon 13–23 (39/48, 81.3%; Χ2 = 14.922; P < 0.001). Conclusions Truncating mutations of the OCRL gene were more common in patients with Lowe syndrome than in Dent-2 disease, while mutation is more likely located at exon 2–12 in Dent-2 disease than that in Lowe syndrome. The type and location of mutation are important indicators for the phenotypes in patients with OCRL mutation. This is a large cohort study analyzing the genotype-phenotype correlation in patients with Lowe syndrome and Dent-2 disease in China. Our data may improve the interpretation of new OCRL variants and genetic counseling. Furthermore, a large international study would be necessary to illustrate the genotype-phenotype correlation in patients with OCRL mutations.


Background
Lowe syndrome, or oculo-cerebro-renal syndrome, is a rare X-linked multisystemic disorder characterized by the triad of congenital cataracts, intellectual disability, and proximal renal tubular dysfunction with slowly progressive renal failure [1]. Lowe syndrome is caused by variants in the OCRL gene on chromosome Xq25-26. In the meantime, mutations in the OCRL also can lead to a type 2 Dent disease (Dent-2 disease) that can present as isolated proximal renal tubulopathy, characterized by low-molecular-weight proteinuria (LMWP), hypercalciuria, and progressive renal insufficiency in the absence of extrarenal pathologies [2]. Except for rickets noted in some patients, no extrarenal manifestations of the disease have been reported in Dent-2 disease.
To date, a hypothesis of phenotype-genotype correlation exists between Dent-2 disease and Lowe syndrome [3]. Disease-causing mutations occur throughout the OCRL gene in patients with Lowe syndrome, but mainly in exons 9-15, which encode the catalytic domain of the protein. In the meantime, disease-causing mutations from patients with Dent-2 disease occur mainly in exons 1-8, which mainly encode the PH domain of the protein. is continuum was not only observed between patients harboring different OCRL mutations but also occurred between patients harboring the same mutation [4]. Understanding how mutations in OCRL give rise to two clinical entities with differing extrarenal manifestations represents an opportunity to identify molecular pathways that could be targeted to develop treatments for these conditions [4].
In the present study, 48 consecutive Chinese children with Lowe syndrome and 35 with Dent-2 disease from published or new data were collected and reanalyzed to further understand the phenotype-genotype correlation of the OCRL gene.

2.
1. e Aim, Design, and Setting of the Study. Both Lowe syndrome and Dent-2 disease are caused by variants in the OCRL gene. However, the reason why patients with similar OCRL gene mutations presented with different phenotypes remains uncertain. e aim of the present study is to analyze the correlation between the genotype of OCRL gene mutations and the phenotype of children with Lowe syndrome or Dent-2 disease.

2.2.
e Characteristics of Participants. Between January 2010 and July 2020, 5 probands with a clinical diagnosis of Lowe syndrome and 4 probands with a clinical diagnosis of Dent-2 disease were recruited at the Children's Hospital of Zhejiang University School of Medicine.
Appropriate informed consent was obtained from all patients and their families. ey were recruited according to the classical criteria for Lowe syndrome or Dent disease, respectively. e patients who presented with full oculo (congenital cataract, and congenital glaucoma), cerebro (hypotonia, developmental delay, and mental retardation), and renal symptoms were diagnosed with Lowe syndrome, while the patients who presented with isolated renal tubulopathy were identified with Dent-2 disease. e pediatric patients who met at least two of the following criteria for Dent-2 disease and detected OCRL gene mutations were eventually included in this cohort [5,6]. (1) LMWP: early renal injury index indicates that the lowmolecular-weight protein in urine is at least five times higher and is mainly LMWP. e low-molecular-weight proteins used for monitoring mainly include retinol-binding protein and α1-MG. β2-Microglobulin or urine protein electrophoresis suggests that low-molecular-weight protein accounts for more than 50%. (2) Hypercalciuria: 24-hour urinary calcium > 4 mg/kg (> 1 mmol/kg) or increased random urinary calcium/urine creatinine ratio (random urinary calcium/urine creatinine ratios vary strongly with age; specific criteria refer to a previous report) [7]; (3) one of the following conditions: microscopic hematuria, kidney stones, nephrocalcinosis, hypophosphatemia, or renal insufficiency. e exclusion criteria included the following: (a) the families who refused the genetic test or participant registration and those who (b) failed the quality control (mean depth coverage of less than × 20 or target coverage region < 90%).
e clinical features and family history of 8 probands in this cohort are presented in Tables 1 and 2.
Estimated glomerular filtration rates (eGFR) were calculated by the Schwartz equation [8] which was applied for patients <18 years. One hundred healthy children were selected as control subjects.
e study protocol was approved by the Ethics Committee of the Children's Hospital of Zhejiang University School of Medicine. All study participants or their guardians provided written informed consent before enrollment.
Furthermore, we conducted a systematic search in 3 English databases and 4 Chinese databases: MEDLINE, Cochrane Library, EMBASE database, CNKI, Wanfang database, the Chinese Scientific Journal database, and the Chinese biomedical literature service system. Studies were identified using keyword search terms: "Lowe syndrome," "oculo-cerebro-renal syndrome" and "Dent disease." e search was limited from January 2010 to July 2020 among all the above databases.
Studies were eligible for inclusion if (i) published in Chinese or English with the cases from China. (ii) Study population: children with diagnosis under 18 years. (iii) Genetic sequencing with OCRL mutations.

Mutation Analysis (As Mentioned before)
2.3.1. DNA Extraction. Genomic DNA was extracted from 5 mL of the peripheral blood of patients by a QIAamp Blood DNA Mini Kit (Qiagen, Milano, Italy) according to the manufacturer's instructions [6,9]. DNA concentrations were measured by a NanoDrop spectrophotometer ( ermo Scientific, Waltham, MA, USA). DNA samples were then stored at −20°C until use.

Whole-Exome
Sequencing. Exome sequencing was performed in two pools to optimize the results. Samples were pooled based on the clinical features of the patients. An array capture was used to enrich the relevant human genes (SeqCap EZ Human Exome Library v2.0, Roche ® , Basel, Switzerland), and these genes were sequenced on the Illumina HiSeq 2000 platform (Illumina, Inc, USA).

Data Filtering.
e following initial steps were performed to prioritize the high-quality variants: (i) variants within intergenic, intronic, and untranslated regions (UTRs) and synonymous mutations were excluded from downstream analysis; (ii) variants with a quality score < 20 were excluded; (iii) only conservation scores (phyloP) >3 were considered upon a comparison of humans and 43 other vertebrates. After the initial selection, the remaining genes were filtered by function. PolyPhen-2 software was used to predict the possible impact of variants (http://genetics.bwh. harvard.edu/pph2/). e final set of selected variants was visually inspected using the Integrative Genomics Viewer. irteen polymorphic variants previously described in       Genetics Research public databases were investigated and compared with the variations found in the current exome. e selected mutations investigated in this study were not found in previous exome sequences (http://evs.gs.washington.edu/EVS/).

Sanger Sequencing Validation.
To confirm the NGS data, Sanger sequencing was employed. DNA from all diagnosed children and their parents were subjected to a polymerase chain reaction (PCR), and polyacrylamide gel electrophoresis was used to determine the size of the amplification products. Products were purified using the QIAquick PCR Purification Kit (Qiagen, Milano, Italy) and sequenced with both forward and reverse primers using the ABI BigDye Terminator Cycle Sequencing Kit v. 3.1 on an ABIPRISM 3730XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). e results were aligned with reference sequences, and mutations were identified using Sequencher DNA Sequence Analysis Software (http://www.genecodes. com). All primers were designed using the online tool Primer3 (http://sourceforge.net/projects/primer3/).

Patients with Lowe Syndrome or Dent-2 Disease.
Between January 2010 and July 2020, 9 consecutive patients (5 with Lowe syndrome and 4 with Dent-2 disease) were recruited and sequenced at the Children's Hospital of Zhejiang University School of Medicine.
Furthermore, 43 consecutive patients with Lowe syndrome and 31 consecutive patients with Dent-2 disease were collected from published data from January 2010 to July 2020. erefore, the clinical manifestation and mutation profile of the OCRL gene from total of 48 patients with Lowe syndrome and 35 patients with Dent-2 disease were summarized in Table 3 and Figure 1 (Lowe syndrome), and Table 1 and Figure 2 (Dent-2 disease), respectively.
To compare the difference in mutation type between Lowe syndrome and Dent-2 disease, all types of mutations were summarized as truncating mutation (nonsense, splicing defect, and incomplete insertion or deletion resulting frameshift and truncated protein) and nontruncating mutation (missense mutation, small in-frame insertion or deletion). e results of mutation type are summarized in Table 2 from patients with Lowe syndrome  and Dent-2 disease. Compared with the results from the literature [1,2], the results of mutation location (Exon 1-7 for Dent-2 disease vs. Exon 8-23 for Lowe syndrome) are summarized in Table 4 from patients with Lowe syndrome and Dent-2 disease.
Furthermore, the results of mutation location (exon 2-12 for Dent-2 disease vs. exon 13-23 for Lowe syndrome) are summarized in Table 5 from patients with Lowe syndrome and Dent-2 disease.

Discussion
Mutations affecting the OCRL gene were primarily associated with Lowe syndrome, and subsequently with Dent-2 disease. More than 140 pathogenic mutations in OCRL have been described so far and reported throughout the entire gene from exon 2 to exon 23. eir phenotype is influenced significantly by the genotype (the mutation type and location of the disease-causing gene and related surrogate gene).
In the present study, 48 series of cases of Lowe syndrome and 35 cases of Dent-2 disease were recruited. Among them, 34 in 48 cases of Lowe syndrome presented with truncating mutations, while 11 in 35 cases of Dent-2 disease presented with truncating mutation, demonstrating that truncating mutations of OCRL gene were more likely seen in patients with Lowe syndrome than Dent-2 disease.
In 2009, from 6 cases of Dent-2 disease reported by Shrimpton et al. [2], all missense mutations fall in the phosphatidylinositol phosphate 5-phosphatase domain of the OCRL protein, while all the other mutations, nonsense, and frameshift fall in the first 7 exons of the gene. ey concluded that this distribution suggests that two different classes of mutations underlie the two diseases (Lowe syndrome and Dent-2 disease).
In 2011, from a large cohort study recruited 130 Lowe families and 6 Dent-2 disease cases by Hichri et al. [1], the specific mapping of the frameshift and nonsense mutations, exclusively identified in exons 1-7 and exons 8-23, respectively, for Dent-2 disease and Lowe syndrome together with the possible use of alternative initiation codons might be related to their clinical expression, that is, Lowe syndrome or Dent-2 disease.
Recently, a large international study [3] of OCRL variants widened the range of exons leading to Dent-2 disease phenotype demonstrating that mutations in Lowe syndrome are located among exons 8 and 24, while exons 4-15 are affected in Dent-2 disease. ere are also reported cases in which OCRL mutations affecting exons at the 3′ side of exon 15 led to the Dent-2 disease phenotype [10].
To further demonstrate the correlation between mutation location and phenotype of Lowe syndrome or Dent-2 disease, Chi-square testing was employed to analyze the difference in mutation location between Lowe syndrome and      Alternatively, if reconsideration of the "cutoff" value of mutation location to exon 2-7 for Dent-2 disease, and exon 8-23 for Lowe syndrome, as suggested by Hichri et al. [1] and Shrimpton et al. [2], then the results demonstrated that 8 in 35 cases with Dent-2 disease are located in exon 2-7 (22.9%), while 47/48 cases with Lowe syndrome are located in exon 8-23 (97.9%), and a significant difference was found between two groups (Χ 2 � 9.035, P � 0.003).
Furthermore, after reanalysis of the results, the distribution of the mutations in Figures 1 and 2 in the present study suggests that (1) pathogenic variants in exon 2-9 are clearly more associated with Dent-2 disease; (2) pathogenic variants in exon 16-23 preferentially cause Lowe syndrome; (3) variants in exon 10-15 can cause one or the other condition in a similar ratio.
OCRL is a multidomain protein of 110kDa. In addition to its 5-phosphatase catalytic domain, it contains a pleckstrin homology (PH) domain, an ASPM, SPD-2, Hydin (ASH) domain, characteristic of proteins that localize to centrosomes and primary cilia, and a RhoGAP-like domain, which mediates the interaction of OCRL with Cdc42 and Rac1. According to the results of the present study and others, it is speculated that mutations in the N-terminal domain of OCRL, where most Dent-2 disease mutations are located, would allow the expression of splicing variants that retain some biological activity. For this reason, this apparent difference between the two diseases might be explained by the partial preservation of other noncatalytic functions of the protein in patients with Dent-2 disease or by the presence of other modifier genes.

Conclusions
is is a large cohort study to reanalyze the genotypephenotype correlation in patients with Lowe syndrome and Dent-2 disease in China. e results demonstrated truncating mutations of the OCRL gene were mostly seen in patients with Lowe syndrome than in Dent-2 disease, while mutation of the OCRL gene is more likely located at exon 2-12 in Dent-2 disease than that in Lowe syndrome. Our data may improve the interpretation of new OCRL variants and genetic counseling.
Data Availability e full protein sequence of OCRL has been submitted to SWISS-MODEL (http://swissmodel.expasy.org/) servers.

Ethical Approval
e study protocol was approved by the Ethics Committee of the Children's Hospital of Zhejiang University School of Medicine (2020-12510).

Consent
All study participants or their guardians provided written informed consent before enrollment. Disclosure e funding sustained the design of the study, collection, and analysis of data.