|
Fig. 1. Clinical representation of the three patients. A Patient 1 at the age of six months: a part of the face b side view showing a pronounced neck fold, a flat back of the head and a dysmorphic auricle c side view of the right leg with wrinkles and d side view with a view of the back area with pronounced skin fold formation. B Patient 2 at the age of five months: a front view showing deep-set eyes, a round face and temporal indentations; b side view, which also shows a flat back of the head and a slightly dysplastic auricle. Due to the fixation of the tracheostoma, the pronounced neck fold is not visible; c view of the back with wrinkles; d view of the back of patient 2 at the age of two and a half years. A remarkable regression of the skin folds was observed. C Patient 1 at the age of 6 years and 7 months: a front-view and b side-view showing dysmorphic features. c Contractures on both hands and fingers are shown (d, e) view of the teeth. D front-view (a) and side-view (b) of patient 3 at the age of 12 years and 6 months. c, d facial appearance of patient 3 at the age of 14 years and 8 months. Contractures of both hands and fingers are observed, as well as wide-spaced teeth
|
|
Fig. 2. Scheme of validated human FBRSL1 isoforms and localization of the detected mutations. The scheme was created using the ExonâIntron-Graphic Makeravailable at https://wormweb.org/exonintron by Nikhil Bhatla (2012) (version 4)
|
|
Fig. 3. Western blot analysis of endogenous and transfected FBRSL1 isoforms. a Scheme representation of FBRSL1 isoforms detectable with an N-terminal antibody. The scheme was created using a domain architecture software (https://prosite.expasy.org/mydomains). Isoform 1, consisting of the AUTS2 domain, has a predicted molecular weight of 110Â kDa, while isoform 3.1 has a predicted molecular weight of 66Â kDa and isoform 3.2 of 55Â kDa. Isoforms 3.1. and 3.2 lack the AUTS2 domain and contain a Ftsk DNA translocase domain with an unknown function. b Plasmids containing isoforms 1, 3.1 and 3.2 in fusion to an HA-taq were detected with either an HA-antibody (HA) or with the N-terminal FBRSL1 antibody (N-terminal) in comparison to the endogenous FBRSL1 expression of HEK293 cells. The approximately estimated sizes of 110Â kDa, 66Â kDa and 55Â kDa of the three different isoforms were detected.
|
|
Fig. 5. Fbrsl1 loss-of-function causes craniofacial defects in Xenopus laevis development. a Scheme of the 5′ region of Xenopus laevis fbrsl1 with indicated exons, introns and the fbrsl1 E1/I1 splice-blocking Morpholino target side and possible outcome (mRNA with intron 1 inclusion) after splicing. Xenopus laevis fbrsl1 consists of 19 exons. The Morpholino sequence is given in the red dashed square. Arrows under exon 1 and exon 7 indicate the locations of the forward and reverse primer used for RT-PCR. b RT-PCR analysis of temporal fbrsl1 expression in the oocyte and different developmental stages of Xenopus embryos. RT-PCR analysis of histone H4 serves as loading control. Ma: Marker, O: Oocyte. c Stage 40 wild-type and control Morpholino (10 ng) injected embryos developed normal craniofacial structures and eyes. Injection of 10 ng fbrsl1 MO results in severe craniofacial defects and a reduction of the eye on the fbrsl1 MO injected side (marked by *). d Anti-Collagen Type II immunofluorescence staining of stage 44 Xenopus embryos showing cartilage defects in embryos injected with fbrsl1 MO but not in wild-type or Co MO injected embryos. M: Meckel’s, Q: quadrate, C: ceratohyal, BH: basihyal, BA: branchial arches. Scale bar represents 500 µm. e Anti-Ncam immunofluorescence staining of stage 40 Xenopus embryos shows normal brain development in control embryos, but reduced Ncam expression in embryos injected with fbrsl1 MO. f, g, h The graphs summarize craniofacial, cartilage and brain defects of at least three independent experiments; number of embryos (n, above each bar) and standard errors of the mean are given. ***p < 0.001 in a Student’s t-test and a one-way ANOVA test
|
|
Fig. 6. Neuronal migration is disturbed in fbrsl1 depleted embryos. a Anti-Ncam immunofluorescence staining of stage 40 Xenopus laevis embryos indicate normal neuronal migration of cranial nerves and motor neurons in wild-type and 10 ng Co MO injected embryos, but disturbed neuronal migration in embryos injected with 10 ng fbrsl1 MO (arrow). b The graph summarizes three independent experiments, number of embryos (n, above each bar)) and standard errors of the mean are given. ***p < 0.001 in a Student’s t-test and a one-way ANOVA test
|
|
Fig. 7. The short human N-terminal FBRSL1 isoform 3.1 can rescue craniofacial malformations induced by Fbrsl1 depletion in Xenopus laevis.
a Injection of 10 ng fbrsl1 MO caused a reduction of craniofacial structures and the eye, while wild-type and 10 ng Co MO injected embryos developed normally. b–e Co-injection of the human FBRSL1 isoform 3.1 significantly rescues craniofacial malformations. In contrast, co-injection of the human FBRSL1 isoform 1 or the mutated human isoform 3.1-p.Gln163* with 10 ng fbrsl1 MO does not rescue craniofacial malformations caused by Fbrsl1 depletion. Embryos injected with 300 pg of the human FBRSL1 isoforms are shown in (b). c–e Graphs summarizing the percentage of craniofacial defects of at least three independent experiments after co-injection of increasing concentrations (100, 200 and 300 pg) of the indicated human FBRSL1 isoforms with fbrsl1 MO. ± s.e.m. and numbers of embryos are indicated (n, above each bar). Scale bar: 500 µm. ***p < 0.001 in a Student’s t-test and one-way ANOVA with Dunnett’s multiple comparisons test
|
|
Supplementary Figure 1: Exome data analysis with Varbank (https://varbank.ccg.uni-koeln.de) and results of Sanger sequencing of patient 2 as well as her parents and results of RT-PCR analysis
A. For patient 2, only two reads were observed in Varbank for region 12:133085800-133085880 (GRCh37/hg19), each of which showed a 23-bp deletion (12:133085843-033085866), while for the parents only one read without deletion was detected. Sanger sequencing of genomic DNA for this region confirmed the wild-type sequence in the healthy parents and indicated a heterozygous status (frameshift) for the 23-bp deletion in the patient.
B. Gel electrophoresis of the RT-PCR analysis on RNA isolated from lymphocytes of the affected child 2 and her parents. H2O was used as negative control. Gel extraction was used to sequence the detected bands, solely. Three bands were detected in the affected child, while only one band, correlating to the size of the wild-type band, was detected in the parents.
C. After Sanger sequencing, the upper band (one star) was identified as a duplex from a wild-type product and a deleted product. The middle band (2 stars) corresponds to the wild-type sequence. The lower band (3 stars) contains the 23-bp deletion.
|
|
Supplementary Figure 2: RNA-Analysis revealed that all three muations escaped the mechanism of NMD. Reads mapped to gene FBRSL1 were visualised using Integrated Genome Viewer (IGV) version 2.8.2, with the samples ordered by family (patient1 with parents 1, patient 2 with parents 2 and patient 3 with mother 3). The reads were scaled by family, such that the read counts of FBRSL1 for each child are directly comparable to the read counts of FBRSL1 in the parents. The annotation (bottom panel) displays two tracks, the first being of FBRSL1 annotated by ENSEMBL hg38 version 97, and the second containing custom annotation of FBRSL1 including exon 3
|
|
Supplementary Figure 3: Expression analysis performed by RT-PCR on human fetal and adult tissues using cDNA panels (Clontech). In addition RT-PCR analysis was performed on RNA isolated from human lymphocytes and HeLa and HEK293 cells. For isoform 1, a ubiquitous expression pattern was observed, as well as for isoform 3.2, while isoform 3.1 shows a clear expression in fetal tissues, with partial lack of expression in the adult tissues
|
|
Supplementary Figure 4: Fbrsl1 MO blocks the exon 1/intron 1 splice site of fbrsl1.
A. Scheme of a part of the Xenopus laevis fbrsl1 with the fbrsl1 splice-blocking Morpholino binding site. The fbrsl1 splice-blocking Morpholino targets the exon 1/intron 1 splice site. The locations of the forward and reverse primer are indicated.
B. RT-PCR using the indicated primer pair results in the amplification of a ~ 900 bp band from cDNA isolated from 10 ng fbrsl1 MO injected embryos but not from cDNA isolated from 10 ng Co MO injected embryos.
C. Sequence alignment of the amplified band confirmed inclusion of intron 1. Exon 1 (marked in yellow) and a part of intron 1 were detected. Red stars * mark the location of in-frame stop codons. Similar results were obtained from three independent experiments.
|
|
Supplementary Figure 5: Whole-mount in situ hybridization of fbrsl1 mRNA in stage 31 wild-type Xenopus embryos. Fbrsl1 is expressed in the head of tailbud Xenopus embryos. br: brain, cn: cranial nerves, ba: branchial arches.
|
|
fbrsl1 (fibrosin-like 1) gene expression in Xenopus laevis embryo, NF stage 31, lateral view, anterior left, dorsal up.
|
|
Fig. 4. Immunofluorescence analysis performed on HEK293 cells (a) and human fibroblasts (b). The N-terminal antibody detected isoforms 1, 3.1 and 3.2 in the cytoplasm and nucleus. Interestingly, an association with centrosomes (white arrow) and kinetochores was detected. Staining with a C-terminal antibody, detecting the full-length isoform 1 and additional hypothetic short C-terminal isoforms, showed a mainly nuclear pattern without a co-localization with the mitotic spindle, centrosomes or kinetochores. α-Tubulin was used for cytoskeletal staining and nuclei were stained using DAPI. Images were obtained using a confocal laser microscope with × 600 magnification, and an additional software magnification as indicated in the respectively images
|
