Bi-allelic variation of FANCJ (BRIP1) identified in premature ovarian insufficiency

Variation screening of FANCJ in the in-house WES database of POI

We conducted the LoF variant screening of FANCJ in the internal WES database of idiopathic POI, which comprises 1030 cases selected from the Reproductive Hospital Affiliated to Shandong University.4 The variants include frameshift deletions or insertions, nonsense, canonical splice site and start-loss. Primary or secondary amenorrhoea before the age of 40 with at least twice an independent serum FSH level above 25 IU/L was the inclusion criteria for idiopathic POI; chromosomal abnormalities, ovarian surgery history, radiotherapy, chemotherapy and autoimmune disorders were the exclusion criteria. The patient’s medical history and all clinical data were gathered at Shandong University’s Center for Reproductive Medicine.

DNA extraction, variant validation and whole-exome sequencing

Genomic DNA was extracted from the peripheral blood samples using the QlAamp DNA Blood Kit (Qiagen, Hilden, Germany). Sanger sequencing verified the variation, which SnapGene software then examined. The genomic DNA samples of the proband’s parents and sister were then captured using iGeneTech’s AIExome V1-CNV and sequenced using 150-bp paired-end reads on Illumina’s NovaSeq systems. Sequence reads were aligned using Burrows-Wheeler Aligner (BWA 0.7.17) MEM to the human reference genome, GRCh37/hg19.23

Cell culture, plasmid construction and transfection

HEK293 cells were grown in Dulbecco’s modified Eagle’s high glucose medium (Gibco, Grand Island, New York, USA) supplemented with 1% penicillin-streptomycin (Gibco) and 10% fetal bovine serum (Gibco) at 37°C. The YouBio firm in Hunan, China, created the FANCJ-EGFP plasmids. By introducing human FANCJ cDNA into the pEGFP-N1 vector (forward primer: 5'-CTACCGGACTCAGATCTCGAGCCACC ATGTCTTCAATGTGGTCTGAAT-3'; reverse primer: 5'-GTACCGTCGACTGCAATTCG CTTAAAACCAGGAAACATGC-3'), wild-type FANCJ-EGFP plasmids were created. Using the wild-type plasmid as the template, site-directed mutagenesis was used to introduce the mutation (c.1A>G) using the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Technologies) (forward primer: 5'-CTACCGGACTCAGATCTCGAGCCACCG TGTCTTCAATGTGGTCTGAAT-3'; reverse primer: 5'-GTACCGTCGACTGCAGAATTCG CTTAAAACCAGGAAACATGC-3'). HEK293 cells were transiently transfected with plasmids using Lipofectamine 3000 reagent (Invitrogen).

RNA extraction and quantitative real-time PCR

Total RNA was extracted using the RNA-Quick Purification Kit (ES Science, Shanghai, China). The NanoDrop One spectrophotometer (Thermo Fisher Scientific) was used to assess RNA concentration, purity and integrity. PrimeScript RT reagent kit with gDNA Eraser (Perfect Real Time) (Takara, Dalian, China) was used for cDNA synthesis. Using the SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biotechnology, Hunan, China), the resulting cDNA samples were submitted to quantitative PCR. FANCJ mRNA relative quantification was calculated using the 2-∆∆Ct method. Human GAPDH was used as the reference gene. For this assay, specific primer sets were employed: FANCJ (forward: CTTACCCGTCACAGCTTGCTA, reverse: CACTAAGAGATTGTTGCCATGCT) and GAPDH (forward: GGAGCGAGATCCCTCCAAAAT, reverse: GGCTGTTGTCATACTTCTCATGG).

Western blot

Following a 48-hour transfection of HEK239 cells, total proteins were extracted using Beyotime’s Radio Immunoprecipitation Assay (RIPA) lysis solution, which included a protease inhibitor cocktail (Roche Applied Science, Mannheim, Germany). Using the main antibodies anti-GFP-Tag (catalogue #M20004, RRID: AB_2619674) and anti-β-Actin (catalogue #HRP-60008, RRID: AB_2819183), equal amounts of proteins were subjected to western blotting analysis. ImageJ was used to examine the photos, which were taken with the BIO-RAD ChemiDoc MP (Hercules, California, USA).

Immunofluorescence

24-well plates were used to seed HEK239 cells, which were subsequently transfected transiently with plasmids. The cells were permeabilised with 0.3% Triton X-100, blocked with 10% bovine albumin for 1 hour, then fixed in 4% paraformaldehyde (Solarbio, Beijing, China) for 20 min at room temperature after 24 hours post-transfection. The coverslips were adhered to glass slides with an antifade medium containing 4',6-Diamidino-2-phenylindole (DAPI, Beyotime, Shanghai, China). Immunofluorescence images were obtained under an Olympus BX61 microscope (Tokyo, Japan).

Cycloheximide chase assay and protein degradation pathway analysis

Protein synthesis inhibitor cycloheximide (CHX) was used to inhibit the translational elongation of FANCJ. After 48 hours of transfection, HEK293 cells were incubated in CHX (10 µM, #HY-12320, Selleck, Shanghai, China) for 0, 2, 4 or 8 hours separately and collected for western blotting.

The ubiquitin-proteasome degradation pathway inhibitor MG132 (10 mM, #S2619, Selleck, Shanghai, China) and the lysosomal degradation pathway inhibitor chloroquine (CQ) (50 µM, #C6628, Sigma, St. Louis, Missouri, USA) were used to explore which protein degradation pathway was mutant FANCJ involved in. After 48 hours of transfection on 6-well plates, HEK293 cells were either co-cultured with CQ or MG132, or treated with CHX alone. After 4 hours, cells were harvested and analysed by western blotting.

Protein ubiquitination assay

After 48 hours of transfection with HA-ubiquitin plasmids (YouBio, Hunan, China) and FANCJ-EGFP plasmids, HEK293 cells were exposed to MG132 for 4 hours. Then, cells were lysed in NP40 lysis buffer (Invitrogen, Carlsbad, California, USA) enhanced with EDTA-free protease inhibitor (Abcam, Cambridge, UK). After centrifuging supernatants, 500 µg of protein was immunoprecipitated using 25 µl of GFP-Trap Magnetic Agarose (Chromotek, catalogue #gtma) for 3 hours at room temperature. The IP Lysis Buffer and SDS-PAGE Sample Loading Buffer (Beyotime, Shanghai, China) were added to the agarose beads and boiled for 10 min at 100°C. Western blot analysis was used to separate the proteins using the primary antibodies anti-GFP-Tag and anti-HA-Tag (Catalogue #YN5572, Immunoway).

Statistical analysis

Every single piece of data came from a minimum of three separate experiments. In the figures, the data were displayed as bar plots of the group mean, with error bars representing the SD. Student’s t-test was used to compare the relative amounts of mRNA and protein expression between groups after the Kolmogorov-Smirnov test was used to determine the normalcy of the data. It was deemed statistically significant when p<0.05.

Bi-allelic defects of FANCJ identified in one patient with POI

Through the variation screening, a rare homozygous start-loss variation of FANCJ (NM_032043.3:c.1A>G, p.Met1Val) was identified in one patient with POI, which was predicted to disturb the protein translation due to altered initiator Methionine codon. The allele frequency (rs764585550) was below 0.001 in the public databases GnomAD (http://gnomad.broadinstitute.org/), ExAC (http://exac.broadinstitute.org/) and 1000Genomes (http://www.internationalgenome.org/1000-genomes-browsers). The mutation was classified as likely pathogenic by the American College of Medical Genetics and Genomics guideline.24 There was no other variant in known POI-causing genes in the patient by WES analysis.

Sanger sequencing was employed to verify the family members’ variations (figure 1), which revealed that the proband’s mother (I-2) carried the heterozygous variant c.1A>G, while her father (I-1) and unaffected sister (II-2) did not have this variant. To illustrate the origin of the bi-allelic variant in the POI patient, we performed WES on her family members. Surprisingly, the results showed that I-1 was not the biological father of the proband. Therefore, the bi-allelic variant might be inherited from her mother and biological father or a de novo mutation.

Clinical characteristics of the variant carriers

The patient with the bi-allelic variation of FANCJ went through three stages of menarche: menarche at age 18, an irregular menstrual cycle at age 28 and amenorrhoea at age 31. She was diagnosed with POI according to twice elevated serum FSH level (above 25 IU/L) and diminished oestradiol level at an interval of 4–6 weeks. Moreover, her serum level of anti-Müllerian hormone was barely detectable (<0.0060 ng/mL). Gynaecological ultrasound showed normal morphology of the uterus, while bilateral ovaries were invisible. No abnormalities were observed in the uterine adnexal bladder and surrounding tissues. She has a normal chromosome karyotype. No history of autoimmune disease, hypothalamus, pituitary or adrenal disease, ovarian surgery or radio/chemotherapy due to tumours has been recorded.

The patient’s mother had regular menstruation before 40 years of age, and menopause occurred over the age of 50. She gave birth to two daughters and a son. The patient’s sister performed regular menstruation at the time of the investigation.

Protein stability of FANCJ was impaired by the start-loss variation

To elucidate whether the start-loss variation affected the protein expression of FANCJ, transfecting HEK293 cells with either wild-type or mutant FANCJ-EGFP plasmids was done separately. As a negative control, the empty vector pEGFP-N1 was used. The results demonstrated that the relative mRNA levels of overexpressed FANCJ were similar between the wild-type and mutant groups (figure 2A), and the FANCJ-EGFP protein at 168 kDa was detectable in both cells by western blot (figure 2B). However, compared with wild-type, the dosage of mutant FANCJ-EGFP protein was much lower, indicating the variant might affect the translation efficiency or protein stability. Moreover, the nuclear localisation sequence of FANCJ is located at the N-terminus of the protein,16 thus we tested the GFP signal of FANCJ-EGFP and found that both mutant and wild-type FANCJ-EGFP proteins were localised in the nuclei (figure 2C), indicating the variation has no effect on the transportation of FANCJ into nuclei.

Then, the CHX chase assay was performed to illustrate the impact of the variant on the protein stability of FANCJ. In contrast to the wild-type FANCJ, the mutant protein degraded significantly faster, suggesting decreased protein stability (figure 3A). Further analysis revealed that both mutant and wild-type FANCJ proteins exhibited a slow degradation rate following inhibition of the ubiquitin-proteasome degradation pathway by the use of MG132, but the differences in mutant group protein levels reached statistical significance (figure 3B). On the contrary, after inhibiting the lysosomal degradation pathway by using CQ, both mutant and wild-type FANCJ proteins presented a similar dosage to those treated with CHX isolated (figure 3B). These findings suggested that the accelerated degradation of mutant FANCJ protein might be predominantly mediated by the ubiquitin-proteasome pathway.

To verify the hypothesis above, we further performed the protein ubiquitination assay. The result showed that both wild-type and mutant FANCJ proteins exhibited ubiquitin labelling, while a notably higher intensity was observed in the mutant FANCJ protein compared with the wild-type (figure 3C), suggesting the mutant FANCJ protein was more extensively ubiquitinated. This observation confirmed that the FANCJ mutation sped up the degradation of FANCJ protein by using the ubiquitin-proteasome pathway.