Loss-of-function mutations in a human gene related to Chlamydomonas reinhardtii dynein IC78 result in primary ciliary dyskinesia

Abstract

Primary ciliary dyskinesia (PCD) is a group of heterogeneous disorders of unknown origin, usually inherited as an autosomal recessive trait. Its phenotype is characterized by axonemal abnormalities of respiratory cilia and sperm tails leading to bronchiectasis and sinusitis, which are sometimes associated with situs inversus (Kartagener syndrome) and male sterility. The main ciliary defect in PCD is an absence of dynein arms. We have isolated the first gene involved in PCD, using a candidate-gene approach developed on the basis of documented abnormalities of immotile strains of Chlamydomonas reinhardtii, which carry axonemal ultrastructural defects reminiscent of PCD. Taking advantage of the evolutionary conservation of genes encoding axonemal proteins, we have isolated a human sequence (DNAI1) related to IC78, a C. reinhardtii gene encoding a dynein intermediate chain in which mutations are associated with the absence of outer dynein arms. DNAI1 is highly expressed in trachea and testis and is composed of 20 exons located at 9p13-p21. Two loss-of-function mutations of DNAI1 have been identified in a patient with PCD characterized by immotile respiratory cilia lacking outer dynein arms. In addition, we excluded linkage between this gene and similar PCD phenotypes in five other affected families, providing a clear demonstration of locus heterogeneity. These data reveal the critical role of DNAI1 in the development of human axonemal structures and open up new means for identification of additional genes involved in related developmental defects.

Figure 1

Predicted amino acid sequence and genomic organization of the human DNAI1 gene. A, Comparison of deduced amino acid sequences of human DNAI1, sea urchin IC2, and Chlamydomonas IC78. Identical residues are denoted by a blackened background, whereas similar residues are shaded. The five WD repeats are within boxes. Gaps that were introduced to optimize the alignment are denoted by dashes. B, Intron-exon organization of the human DNAI1 gene. The 20 exons (E1–E20) encoding the DNAI1 protein are indicated by unshaded or shaded boxes (translated and untranslated regions, respectively). The locations of the ATG initiating codon, the stop codon, and the polyadenylation signal are shown. The number of the first codon in each exon is indicated; exons beginning with the second or third base of a codon are indicated by subscript 2 or 3, respectively. The exons are drawn to scale. Intron-exon boundaries are denoted by blackened triangles; the size (kb) of each intron (I1–I19) is indicated at the bottom.

Figure 2

The deduced amino acid sequence of human DNAI1 (arrow) integrated into an evolutionary tree of representative classes of dynein intermediate chains (IC): axonemal outer dynein IC and cytoplasmic dynein IC subfamilies. A parsimony tree for Chlamydomonas IC78 and IC69, sea urchin IC2 and IC3, rat IC74 (GenBank accession number X66845), rat dynein cytoplasmic intermediate chain–2A (IC2A [GenBank accession number U39044]), mouse Dnci1 (GenBank accession number AF063229) and Dnci2 (GenBank accession number AF063231), Drosophila Cdic (GenBank accession number AF070687), Dictyostelium cytoplasmic dynein intermediate chain (dicA [GenBank accession number U25116]), human DNCI1 (GenBank accession number AF063228), and the newly identified human DNAI1 was created by the PILEUP program.

Figure 3

Expression analysis of the human DNAI1 gene. Northern blot analysis was performed with the use of human multiple-tissue northern blots (Clontech) hybridized with the full-length coding sequence of DNAI1 (top). A 2.5-kb transcript was detected in trachea and testis tissues. An additional transcript of ∼4 kb was detected in trachea tissue only. Probing with β-actin used as a control (resulting in bands at 2.0 kb and 1.8 kb) is shown (bottom).

Figure 4

Chromosomal mapping of the human DNAI1 gene. A, FISH showing localization of hybridization signals to both chromosomes 9 on metaphase spread (arrowheads). B, Magnification of chromosome 9 that allows the precise mapping of the DNAI1 gene to the p13-p21 region on both chromatids. C, Ideogram of human chromosome 9 that shows the precise localization of the DNAI1 gene (red arrowhead).

Figure 5

Electron micrograph of cross-sections of respiratory cilia (original magnification ×90,000). A, Absence of outer dynein arms (arrow) is observed on all the peripheral doublets of the ciliary sections obtained from PCD patient II-1. B, Normal ciliary ultrastructure from a control subject.

Figure 6

The exonic insertion identified on the maternal DNAI1 allele of patient II-1 from family 1. A, SSCP analysis of the PCR products of exon 5 from three members (I-1, I-2, and II-1) of this family and from a control subject (N), showing bandshifts (arrows) both in patient II-1, who is denoted by a blackened square, and in his mother (I-2). Lane C corresponds to a negative control (PCR without genomic DNA). B, Nucleotide sequence of the mutant (top) and normal (bottom) DNAI1 alleles. The insertion of four nucleotides in exon 5 is denoted by the area outlined by a solid line. The recognition sequence for the VspI restriction endonuclease is denoted by the area outlined by a dashed line. C, Segregation analysis of the 4-bp insertion within the family. The presence of the mutation in the genomic DNA from all three family members was assessed by VspI digestion of PCR products generated with primers P8 and P9. The affected child (II-1) and his mother (I-2) yield a banding pattern consistent with the presence of both the normal (276-bp) and the mutant (80-bp and 200-bp) alleles, whereas both his father (I-1) and a control subject (N) display only normal alleles. Lane C corresponds to a negative control (PCR without genomic DNA). The size marker (M) is a 1-kb ladder from Gibco BRL.

Figure 7

The splice mutation identified on the paternal DNAI1 allele of patient II-1 from family 1. A, SSCP analysis of the PCR products of exon 1, which were obtained from three members of this family and from a control subject (N), show bandshifts (arrows) both in patient II-1 (denoted by the blackened square) and in his father (I-1). Lane C corresponds to a negative control (PCR without genomic DNA). B, Nucleotide sequence of the mutant (top) and normal (bottom) DNAI1 alleles. The 1-bp insertion (T) at nucleotide +3 of the intronic sequence following exon 1 is outlined by a solid line. The recognition sequence for the HpaI restriction endonuclease is denoted by the area outlined by a dashed line, and the position of the exon-intron boundary is indicated above the sequence. C, Segregation analysis of the 1-bp insertion within the family. The presence of the mutation in the genomic DNA from all three family members was assessed by HpaI digestion of PCR products generated with primers P6 and P11. The affected child (II-1) and his father (I-1) yield a banding pattern consistent with the presence of both normal (159-bp) and mutant (97-bp and 63-bp) alleles, whereas both his mother (I-2) and a control subject (N) display only the normal alleles. Lane C corresponds to a negative control sample (PCR without genomic DNA). The size marker (M) is a 1-kb ladder from Gibco BRL. D, Functional consequences of this 1-bp insertion on DNAI1 RNA splicing. RT-PCR amplifications of total RNA obtained from nasal cells from a control subject (lane N) and from the patient (lane II-1), with the use of primers P6 and P7. The RNA sample from the control subject generates a 290-bp product, whereas the RNA sample from patient II-1 generates two products, one of normal size (290 bp) and one arising from abnormal splicing (422 bp). Lane C contains a control without RNA sample. The size marker (lane M) is a 1-kb ladder from Gibco BRL. E, Schematic representations of the splicing mechanisms leading to normal (top) and abnormal (bottom) DNAI1 transcripts, from the corresponding genomic DNA fragments (middle). The 1-bp insertion at the third nucleotide of intron 1 is indicated below the genomic sequence (arrowhead). Exonic sequences (E1–E4) and intronic sequences are denoted by gray boxes and thick solid lines, respectively. In the intronic sequence following exon 1, lowercase letters indicate the 5′ splice consensus site normally used in primary DNAI1 transcripts; letter D denotes the cryptic splice-donor site used in the presence of the 1-bp insertion. The sizes of the normal (top) and mutant (bottom) RT-PCR products generated with primers P6 and P7 (arrows) are indicated. In the presence of the 1-bp insertion (asterisk), the 5′ part of the intron (open box) following exon 1 is retained in the mutant cDNA (bottom).