DNAH11 variants and its association with congenital heart disease and heterotaxy syndrome

Congenital heart diseases (CHDs) are the most common types of birth defects, affecting approximately 1% of live births and remaining the leading cause of mortality. CHD patients often show a higher incidence of heterotaxy syndrome. However, the exact aetiology of CHD and heterotaxy syndrome remains unclear. In this study, targeted sequencing and Sanger sequencing were performed to analyze the exonic regions of 37 primary ciliary dysfunction (PCD)- related candidate genes in 42 CHD patients with heterotaxy syndrome. Variants affecting protein-coding regions were filtered according to databases of known variants and predicted in silico using functional prediction program. Thirty-four potential disease-causing heterozygous variants in 11 genes were identified in the 19 CHD patients with heterotaxy syndrome (45.2%, 19/42). The DNAH11 gene showed the highest mutation rate (16.7%; 14 of 84 alleles) among the CHD patients with heterotaxy. Fisher’s exact test revealed a significant association of DNAH11 variants with CHD and heterotaxy (P = 0.0001). In families, six different compound heterozygous variants of DNAH11 were validated in family 1-5031 (p.W802X/p.M282I), family 2-5045 (p.T3460K/p.G4425S), family 3-5065 (p.G447R/p.L1157R), family 4-5130 (p.I2262T/p.D3800H), family 5-5707 (p.S1823fs/p.F2759L/p.R4395X) and family 6-5062 (p.D3610V/p.I243V). These findings suggest that the DNAH11 variants are significantly associated with CHD and heterotaxy syndrome and that compound heterozygous DNAH11 variants may be the common genetic cause of the development of familial CHD and heterotaxy syndrome.

www.nature.com/scientificreports www.nature.com/scientificreports/ and 10 genes regulating vesicular trafficking (a pathway important for ciliogenesis and cell signaling), including a loss-of-function DNAH11 mutation known to cause PCD 10 .
DNAH11, located on chromosome 7p15.3, encodes a ciliary outer dynein arm protein (520 kDa) and is a member of the dynein heavy chain family, localizing exclusively to the proximal region of respiratory cilia 11 . DNAH11 mutations have been found to result in abnormal ciliary ultrastructure and hyperkinetic ciliary beating 12 . Gene editing of DNAH11 mutations can restore normal cilia motility in primary ciliary dyskinesia 13 . An overwhelming majority of previous studies related to DNAH11 have focused on PCD and situs inversus totalis, and few studies have concentrated on heterotaxy and CHD.
To date, over 40 PCD-related genes have been found in heterotaxic patients with PCD. However, whether PCD-related gene mutations are associated with CHD and heterotaxy syndrome remains unclear. In this study, we performed targeted sequencing and Sanger sequencing analysis of PCD-related genes in Chinese CHD patients with heterotaxy syndrome to explore the genetic aetiology of CHD and heterotaxy syndrome.

Results
Assessment of the ciliary movement status in CHD patients with heterotaxy syndrome. A cohort of 42 CHD patients with heterotaxy syndrome were recruited from unrelated families, including 14 females (33.3%) and 28 males (66.7%) with ages ranging from 0.1 to 20.1 years (mean ± SD: 5.6 ± 4.7 years). To assess the ciliary movement status, we examined the ciliary beat pattern using a slow-motion playback of the video sequence to generate tracings of the ciliary beat and found that 29 cases (69%, 29/42) showed CD (18 males and 11 females), and 13 cases (31%, 13/42) presented normal ciliary function (10 males and 3 females).
The clinical features of the study subjects are summarized in Table 1. For these genes, 758 regions of gene coding exons and their intron/exon junctions were sequenced. The coverage of the sequencing results was 90%-99%, with an average coverage of approximately 97%, and the average depth of sequencing was 100X. By filtering from 1000 Genomes Project and ExAC databases and using SIFT, PolyPhen2 and MutationTaster prediction programs, we focused on novel or rare coding variants (MAF < 0.01%) present in CHD patients with heterotaxy, excluding the common variants, synonymous variants and non-synonymous variants that are predicted to have no deleterious effect on protein function.
Concerning the highest mutation rate in the DNAH11 gene, we further analyzed the association of DNAH11 mutations with CHD and heterotaxy syndrome. In this study, there were 14 mutations in the DNAH11 genes in 7 of 42 patients, including 11 missense, 1 frameshift and 2 stop-gain mutations. Eight of these variants were novel and not present in the ExAC or 1000 Genomes Project databases, and 6 were low-frequency variants (MAF < 0.01%). Conservation analysis was processed via UCSC Genome Browser hg19.
In our previous study, whole genome sequencing was performed in 98 CHD patients without heterotaxy, and the methodology used in the previous study is strictly the same as that used in the present investigation. The CHD subtypes of these 98 CHD patients are listed in Table 5.
By consulting the previous exome database in our laboratory, we found no disease-causing mutations in the DNAH11 gene among 98 CHD cases. Moreover, one DNAH11 mutation (c.A9584G:p.N3195S) was found in 1 Characteristic CHD/heterotaxy + CD a CHD/heterotaxy + no CD b CHD/heterotaxy (Total)  www.nature.com/scientificreports www.nature.com/scientificreports/ case (Patient 2073) among 3 CHD patients with heterotaxy. Combined with the results showing 7 of 42 patients with DNAH11 mutations in this study, the prevalence of DNAH11 mutations was higher in CHD patients with heterotaxy (8 of 45 cases) than in CHD patients (0 of 98 cases). The 98 CHD cases were considered controls because these cases exhibited only CHD, and the association of DNAH11 mutations with CHD and heterotaxy was significant (8 of 45 CHD patients with heterotaxy vs. 0 of 98 controls, P = 0.0001 by Fisher's exact test). These findings suggest a significant association of DNAH11 mutations with the risk of CHD and heterotaxy syndrome ( Table 6). DNAH11 compound heterozygous mutations in CHD families with heterotaxy. Disease-causing DNAH11 mutations are inherited by autosomal recessive inheritance. Because there were no homozygous mutations in DNAH11 in this study, we focused on compound heterozygous mutations in the DNAH11 gene, which Characteristic CHD/heterotaxy + CD A (n = 29) CHD/heterotaxy + no CD B (n = 13) P a Gene mutation rate (%) 18 (62.1%) 1 (7.7%) 0.0018  www.nature.com/scientificreports www.nature.com/scientificreports/ may be the main cause of the development of CHD/heterotaxy. The 14 disease-causing heterozygous mutations in DNAH11 were distributed among 7 CHD patients with heterotaxy, with 6 patients having two or more DNAH11 mutations. These DNAH11 mutations were further confirmed to be present in the available DNA of parents and other family members of the patients by Sanger sequencing. Interestingly, six different compound heterozygous variants in DNAH11 were validated respectively in six different families, including family 1-5031 (p.W802X/p. M282I), family 2-5045 (p.T3460K/p.G4425S), family 3-5065 (p.G447R/p.L1157R), family 4-5130 (p.I2262T/p. D3800H), family 5-5707 (p.S1823fs/p.F2759L/p.R4395X) and family 6-5062 (p.D3610V/p.I243V) ( Table 7).
In family 1 (Fig. 1A), there were four members. The proband (#5031), who carried 2 heterozygous mutations (c.G2406A:p.W802X and c.G846C:p.M282I), was male and diagnosed with CHD and heterotaxy, including   www.nature.com/scientificreports www.nature.com/scientificreports/ isolated right heart, complete atrioventricular canal (CAVC), double outlet right ventricle (DORV) and atrial septal defect (ASD), and showed abnormal ciliary function. His parents and younger brother were all without clinical manifestations, but their ciliary movements were abnormal as evidenced by uncoordinated ciliary waves. In this family, the heterozygous variant c.G2406A (p.W802X) was a novo stop-gain mutation located on exon 14 of DNAH11. This variant was present in the proband and his young brother and inherited from their mother. Functional analysis showed the variant p.W802X was predicted to be damaging by SIFT software and conserved among human, rhesus and dog species. The other heterozygous variant c.G846C (p.M282I) located in exon 4 in the proband was transmitted from his father and was not observed in his young brother and mother. This mutation was reported in the Exome Aggregation Consortium (ExAc) (0.00006547) and 1000 Genomes Project   www.nature.com/scientificreports www.nature.com/scientificreports/ (0.000199681) databases. The p.M282I change was also predicted to be damaging by SIFT software and conserved among human, rhesus and mouse species (Fig. 1B).
In family 2, the proband (#5045) was male and diagnosed with CHD and heterotaxy, showing abnormal ciliary movement. His parents were both heterozygous carriers and showed normal phenotypes. The heterozygous variants (c.C10379A:p.T3460K/c.G13273A:p.G4425S) of the DNAH11 gene were confirmed in family 2 by Sanger sequencing (Fig. 2A). Of the two heterozygous variants, c.C10379A (p.T3460K) was a missense variant type located in exon 64 and was inherited from the proband's mother and presented in the ExAc (0.0001879) and 1000 Genomes Project (0.000199681) databases. The amino acid p.T3460K alteration is predicted to be damaging by SIFT, PolyPhen2 and MutationTaster and is highly conserved among mammalian species but not in chickens. The other variant, c.G13273A (p.G4425S), was a novel missense variant type located in exon 81 and which was transmitted from the proband's carrier father. The p.G4425S change was predicted to be damaging by the SIFT, PolyPhen2 and MutationTaster software programs and is highly conserved among mammalian species (Fig. 2B).
The family 3 proband (#5065) had severe CHD and heterotaxy, with abnormal ciliary movement. His parents were normal and healthy. This family carried two heterozygous missense mutations (c.G1339A:p.G447R/c. T3470G:p.L1157R) in the proband (Fig. 3A). The c.G1339A (p.G447R) in exon 7 was a novel variant and absent in his parents, indicating that this variant was de novo. The altered amino acid p.G447R was conserved among the human, rhesus, dog and elephant species. Functional analysis of this change was predicted to be benign by prediction software. The other c.T3470G (p.L1157R) in exon 18 was a missense mutation type and presented in the ExAc (0.000149) and 1000 Genomes Project (0.000199) databases. The mutation c.T3470G was inherited from his mother, resulting in the substitution of the 1157 amino acid Leu (L) with Arg (R) (p.L1157R) and was predicted to be a damaging change in the protein. The altered amino acid p.L1157R is highly conserved among many species. Although the p.G447R variant was predicted to be not damaging, considering the proband's mother normal phenotype, we concluded that the combined effects of the two heterozygous variants (p.G447R/p.L1157R) may be an important factor causing CHD/heterotaxy disease (Fig. 3B).
The family 4 proband (#5130), who was female and diagnosed with CHD, heterotaxy and CD, carried two heterozygous missense variants (c.T6785C:p.I2262T/c.G11398C:p.D3800H). We obtained the proband and her mother's blood sample for validating the detected variants by Sanger sequencing. The blood sample of the proband's father was not obtained. However, the parents both had normal phenotypes according to the medical history record (Fig. 4A). The c.T6785C (p.I2262T) variant was a novel heterozygous type and was located in exon 41 of DNAH11, which was inherited from the proband's mother. Functional analysis indicated that the p.I2262T variant was predicted to be damaging by the SIFT, PolyPhen2 and Mutation Taster software programs and was highly conserved among many species. The variant c.G11398C (p.D3800H) in exon 70 was also a novel heterozygous variant and was absent from the ExAc and 1000 Genomes Project databases. Functional analysis predicted that the variant p.D3800H was damaging (by MutationTaster software) and was conserved among different www.nature.com/scientificreports www.nature.com/scientificreports/ species (Fig. 4B). We could not collect a blood sample from the proband's father; therefore the inherited origin of the c.G11398C (p.D3800H) variant could not be determined, and we proposed that the compound heterozygous variants (p.I2262T/p.D3800H) may be associated with the development of the proband's disease.
In family 5, patient #5707, diagnosed with CHD, heterotaxy and CD, had three variants in DNAH11, one frameshift variant (c.5470dupC:p.S1823fs), one missense variant (c.T8275C:p.F2759L) and one stop-gain variant (C13183T:p.R4395X). We did not obtain a blood sample from the patient's parents, so we only confirmed these variants in the DNA sample of patient #5707 by Sanger sequencing. One frameshift variant (c.5470dupC:p. S1823fs) and two variants (c.T8275C:p.F2759L and C13183T:p.R4395X) were validated (Fig. 5A). The variant (c.5470dupC:p.S1823fs) was a novel frameshift insertion located in exon 32 and affected the protein function, which was conserved among different species. The c.T8275C variant was a missense variant located in exon 50 and was presented in the ExAc (0.0003) and 1000 Genomes Project (0.000599042) databases. The p.F2759L change was predicted to be damaging by the PolyPhen2 and MutationTaster software programs and was conserved among different species. The variant (C13183T:p.R4395X) in exon 81 was a novel stop-gain type that influenced protein function. The p.R4395X variant was highly conserved among different species (Fig. 5B).
In family 6 ( Fig. 6A), there were three members. The proband (#5062), who carried 2 heterozygous mutations (c.A10829T:p.D3610V and c.A727G:p.I243V), was female and diagnosed with CHD and heterotaxy, including isolated right heart, PA, levo-transposition of the great arteries (L-TGA) and ASD, and showed abnormal ciliary function. His parents both showed normal phenotypes. The c.A10829T (p.D3610V) variant was a novel heterozygous type located in exon 66 of DNAH11 and was inherited from the proband's mother. Functional analysis showed the variant p.D3610V was predicted to be damaging by SIFT software and conserved among human, rhesus and dog species. The other heterozygous variant c.A727G (p.I243V) was a novel missense mutation type located in exon 4 of DNAH11 and was transmitted from the proband's father. The p.I243V change was also predicted to be damaging by SIFT software and conserved among human, rhesus and mouse species (Fig. 6B).
Based on these confirmed variants and the medical history of the families, the findings suggest that these DNAH11 compound heterozygote variants are responsible for the development of CHD/heterotaxy syndrome.

Discussion
Human LR asymmetry plays an important role in normal organogenesis and provides the developmental basis for correct heart looping 2 . LR asymmetry disorders in early embryonic development may result in a series of congenital birth defects, such as heterotaxy syndrome 14 . Although many genes have been reported to be associated with LR asymmetry disorders 15 , the exact aetiological mechanism of CHD and heterotaxy remain unknown. Moreover, studies have found that cilia participate in the formation of the left and right asymmetric mode by regular swinging to form a nodal flow during the embryonic period 16 . PCD and CD have been shown to be associated with CHD and   www.nature.com/scientificreports www.nature.com/scientificreports/ heterotaxy syndrome. PCD is considered a monogenic heterogeneous recessive disorder, and CHD and heterotaxy syndrome are multiple, complex inherited diseases. There may be a link between PCD and CHD/HTX.
In this study, we found 34 potential disease-causing heterozygous variants in 11 genes present in 19 CHD/ HTX patients, accounting for 45.2% of 42 CHD/HTX patients. We compared the gene mutation rate in the CHD/HTX patients with CD and CHD/HTX patients without CD and found that CHD/HTX patients with CD had a significantly higher gene mutation rate than CHD/HTX patients without CD. The results suggest that PCD-related gene mutations are significantly associated with CHD with heterotaxy and CD.
DNAH11 is localized to the proximal region of respiratory cilia and is known as a PCD-related gene. Approximately 6% of PCDs are caused by DNAH11 mutations 17 . Although cilia are required for LR body-axis determination and second heart field (SHF) Hedgehog (Hh) signalling, Burnicka et al. observed that DNAH11 mutations did not disrupt SHF Hh signalling and caused AVSDs only concurrently with heterotaxy, a LR axis abnormality 18 .
DNAH11 showed the highest mutation rate in this study, followed by HYDIN and DNAH5. We concluded that DNAH11 mutations may be an important risk factor involved in the development of CHD/HTX syndrome. We reanalyzed the exome database in our previous study, including 98 CHD cases and 3 CHD/HTX cases, and found no mutation in the DNAH11 gene among 98 CHD cases and 1 DNAH11 mutation in 1 of 3 CHD patients with heterotaxy. We considered the 98 CHD cases as controls, fisher's exact test revealed that DNAH11 mutations can significantly increase the risk of developing CHD/HTX syndrome, indicating a significant association of DNAH11 mutations with CHD and heterotaxy syndrome.
It is known that DNAH11 mutations are inherited by autosomal recessive inheritance pattern. Bartoloni L et al. found a homozygous nonsense mutation (R2852X) in DNAH11 in a patient with situs inversus totalis 19 . DNAH11 compound heterozygotes (p.R2250*/p.Q3604*) were observed in two monochorionic biamniotic male twins with PCD 13 . Nader Nakhleh et al. found DNAH11 compound heterozygotes (p.Q1507P/p.E3133K) in a patient with heterotaxy; these two mutations were predicted to be damaging and involved in hyperkinetic ciliary beats 4,20 .
DNAH11 homozygous mutations were not observed in our study. We focused on recessive DNAH11 mutations in the families and found six different compound heterozygous variants in six families and concluded that these compound heterozygous variants may be the main factors causing CHD/heterotaxy syndrome.
Recently, a study found that heterotaxy patients with heterozygous DNAH6 mutations also had heterozygous mutations in DNAH5 and DNAHI1 genes, which experimentally showed that the trans-heterozygous interactions of DNAH6 with DNAI1 or DNAH5 may contribute to heterotaxy syndrome 21 .
In our study, in addition to these DNAH11 compound heterozygotes, we also found that the patients with the DNAH11 heterozygous mutations possessed other heterozygous mutations in known PCD genes, such as HYDIN and DNAH5. Patients 5033, who possessed two different heterozygous variants in the DNAH11 and HYDIN genes www.nature.com/scientificreports www.nature.com/scientificreports/ (Table 2), presented primary CD and heterotaxy/CHD, indicating that interactions between trans-heterozygous variants of DNAH11 and HYDIN may be involved in heterotaxy/CHD and PCD.
Although over 40 PCD pathogenic genes were revealed, our study found compound heterozygous variants in only DNAH11 among the CHD patients with heterotaxy, which is subject to the limited number of patients recruited in this study. We speculated that larger studies may uncover comprehensive genetic pathogenic factors related to cilia among these patient populations. However, we can still conclude that pathogenic DNAH11 mutations are an important cause of heterotaxy with CHD. Additionally, we did not conduct deeper functional studies of these DNAH11 heterozygous variants for pathogenic confirmation and further interpretation of the genotypic and phenotypic mechanisms, which should be the key aspects of future works.
These findings expand the spectrum of DNAH11 gene mutations causing the development of HTX/CHD and provide an important clue for understanding the genetic mechanism of HTX/CHD syndrome.

Methods
Patient cohorts. Blood samples from patients or their family members were collected in accordance with the Declaration of Helsinki, and the Ethics Committees of Children's Hospital of Fudan University (CHFU) approved this study. Written informed consent was obtained from the parents and guardians of all the probands. Informed consent has been obtained for the blood samples taken from the available family members.
Forty-two participants with CHD and heterotaxy syndrome were recruited from the CHFU, Shanghai, China, including 28 males and 14 females, with ages ranging from 0.1 to 12.1 years. Family medical history was obtained at the cardiovascular centre of the CHFU, and medical records were reviewed to confirm the disease diagnosis. Blood samples from probands and their available family members were obtained for further genetic analysis.
Nasal tissue sampling and ciliary motion analysis. Nasal tissues were collected from patients with Rhino-Probe (Arlington Scientific, Springville, UT) curettage of the inferior nasal turbinate. Exclusion criteria included severe bleeding diathesis or conditions such as haemophilia or hereditary haemorrhagic telangiectasia syndrome. The nasal tissues were suspended in L-15 medium (Invitrogen, CA) for videomicroscopy using a Leica inverted microscope (DMIRE2) with a 67× oil objective under differential interference contrast optics. Movies were recorded at 200 frames/s at room temperature using a 680 PROSILICA GE camera (Allied Vision, PA), and digital recordings were evaluated by a blinded panel of coinvestigators. Abnormal ciliary motion was described as follows: immotile (I), discordance (D), wave (W), restricted (R) and no cilia (None).
Blood DNA extraction. A QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) was used to extract blood genomic DNA from probands and available family members according to the manufacturer's instructions. The concentration and purity of genomic DNA were measured by absorbance at 260 and 280 nm by using a NanoDropTM 1000 Spectrophotometer (Thermo Scientific, Wilmington, USA). The extracted genomic DNA were stored at −80 °C until the samples were ready for further analysis. For each subject, base calls were detected with Torrent Suite software. Raw sequencing data were aligned against the human reference genome GRCh37/hg19 (NCBI) using NextGENe software. Single-nucleotide variations (SNVs) were aligned based on the following criteria: 1) the variant was detected on both strands of the sequence reads; 2) the minimum coverage of reads was no less than 10×; 3) the variant reads represented more than 20% of the sequence reads.