ABSTRACT
Primary ciliary dyskinesia (PCD, OMIM 242650) is an autosomal recessive disorder affecting cilia motility with an estimated incidence of 1 in 20,000 live births (1). Impaired or absent motility of respiratory tract cilia leads to recurrent sino-pulmonary infections and permanent lung damage (bronchiectasis) (2). Abnormalities of the related flagella structure of sperm lead to male subfertility (3). Other ciliated tissues such as those of the female reproductive tract and the brain ependyma can also be affected and are associated with reduced fertility, ectopic pregnancy, and hydrocephalus in some patients (4,5).Patients also exhibit abnormal left-right body axis (LRA) determination, and laterality defects are observed in 50% of PCD patients-commonly complete mirror-image reversal of internal body organs (situs inversus). Patients with ciliary defects and situs inversus are classified as having Kartagener syndrome (KS, OMIM 244400) (1). Studies in mice have suggested a link between movement 109
of monocilia at the embryonic node during development and left-right pattern formation, although such an association has yet to be proven in PCD patients (6).PCD is a genetically heterogeneous disorder, and this is reflected by the variation in ciliary ultrastructural defects and clinical presentations that have been documented in patients (7,8). The most common defect is a reduced number or complete absence of the inner and/or outer axonemal dynein arms. Defects of other ciliary structural components, such as the radial spokes and central pair microtubules, as well as elongation and incorrect orientation of the cilia have also been documented (9,10). CANDIDATE GENES FOR PCD
Cilia have an evolutionarily conserved structure and are composed of over 250 different polypeptides (11,12). The genes encoding structural components of the ciliary axoneme are obvious candidates to underlie different ultrastructural defects seen in PCD patients. These include dynein protein components of the dynein arms (13), the radial spoke head proteins (11), and other microtubule-associated proteins, such as kinesins (14). Much of the knowledge of the structural components of the human ciliary axoneme, and the genes encoding their proteins, comes from studies of motility mutants of Chlamydomonas reinhardtii(15) and the subsequent identification of homologous genes in humans. To date, 11 dynein genes and 12 kinesin and kinesin-related genes have been assigned chromosomal localizations in humans (http:/www.gene.ucl.ac.uk/nomenclature).Another source of human candidate disease genes comes from the study of animal models of PCD and of genes involved in the LRA determination pathway (16). Several rodent mutants have been investigated that display features of the PCD disease phenotype. The hophpy mouse mutant has defects of the ciliary axoneme in respiratory and reproductive tissue and the sperm flagellae (17,18). The hfh4 mouse mutant has a complete absence of respiratory cilia in conjunction with randomized LRA determination (19,20). The kif3B (21) and iv mouse (22) have randomized LRA determination and immotile nodal monocilia but display normal functional respiratory cilia. The WIC-Hyd rat has immotile respiratory and ependymal cilia in conjunction with situs inversus and hydrocephalus (23). APPROACHES USED TO IDENTIFY GENES INVOLVED IN PCD
A candidate gene approach can be used to try and identify the genes mutated in PCD. A multitude of putative candidate genes for PCD are distributed across the human genome, and this technique has been used successfully to identify the first
known PCD gene. Sequencing of the human DNAI1 gene located on human 9pl3-p21, a homolog of the Chlamydamonas IC78 gene, which is essential for axonemal outer dynein arm assembly (24), identified two separate loss-of-function mutations in a PCD patient with an outer dynein arm defect (25). Another candidate gene, analyzed in a collaborative effort including our laboratory, is hfti4, a winged-helix/forkhead transcription factor with expression restricted to ciliated cells (26). PCD patients were screened that, like the hfh4 mouse mutant, had no respiratory cilia, or who were consistent with linkage to the region of 17q that HFH4 mapped, but no sequence abnormalities were observed. Positional Cloning Approach Using Genetic Linkage Analysis
Linkage analysis allows mapping of disease genes without prior knowledge of the functional components involved in the disorder. It is based on detection of the cosegregation of chromosome DNA markers with a disease trait in families. Linkage is significant if a lod score-defined as the logarithm of the odds ratio that the disease and marker loci are linked rather than unlinked-of >3 is achieved. A number of linkage studies in PCD have been reported. In a study by Witt et al. (27), linkage analysis was performed on chromosome 7, which was of interest because (1) the hophpy mouse gene maps to a region syntenic to human 7q33-q34; (2) several dynein related genes have been positioned on the p and q arms of chromosome 7; and (3) uniparental disomy of chromosome 7 has been observed in an individual with KS and cystic fibrosis (28). A weak suggestion of linkage was obtained at the P-dynein heavy-chain locus on 7pl5 in PCD families with no situs inversus. Witt et al. also reported linkage of 51 KS families to chromosome 15p with a lod score of 4.23 (29). Linkage analysis at the HLA complex on chromosome 6 has been performed due to the location there of a p-tubulin gene, an axoneme structural component (30). Affected individuals from two unrelated PCD families showed an association with the HLA-DR7; DQW2 haplotype. A genome-wide linkage screen was performed in 31 nuclear PCD families by Blouin et al. (31), but this identified no single major locus, confirming locus heterogeneity. Under the assumption that 40% of the individuals were linked to a single locus, slightly positive lod scores were obtained on chromosomes 3p, 4p, 5p, 8q, 15q, 16p, 17q, and 19q. Homozygosity Mapping as a Genetic Analysis Technique
Linkage analysis in a genetically heterogeneous disorder such as PCD, where more than one locus may be mutated in a particular ultrastructural phenotype and/or group of families from similar geographical region, can prove problematic due to the large number of families required for analysis. Identification of PCD loci by linkage analysis has been greatly enhanced by homozygosity mapping,
an approach that allows recessively inherited traits to be mapped in families of consanguineous unions (usually first or second cousin marriages) (32). It is based on the premise that a child of a consanguineous marriage affected with an autosomal recessive disorder will have inherited the same ancestral disease chromosome from each parent and will therefore contain a region of the chromosome spanning the disease locus that is homozygous by descent (HBD). An affected child from a consanguineous marriage would be expected to have several regions that are HBD since 1/16 genome is HBD in a first cousin marriage (33), but these regions will vary from child to child. By searching for a region HBD that is shared among all affected individuals in a number of families, a common region of the chromosome containing a locus for the shared disease may be positioned (see Fig. 1). One affected offspring from a first cousin marriage has been calculated to yield a lod score of 1.2, and this means that just three affected individuals are needed to obtain a lod score of >3 (33). Homozygosity mapping can also be applied to nuclear pedigrees from small, inbred populations, due to the high level of inbreeding and the likelihood that a single common mutation arose in the founder population (33). MAPPING OF A LOCUS FOR PCD TO 19q13.4 BY HOMOZYGOSITY MAPPING
Homozygosity mapping identified a locus for PCD in five Arabic families, four of which were consanguineous due to first cousin marriages. The resource consisted of 24 individuals, with 12 affected individuals, 5 of whom had situs inversus. Family information and linkage data are reported in Meeks et al. (34). A genome-wide screen identified a region of excess homozygosity among affected individuals of the four consanguineous families on chromosome 19ql3.4 (34). Haplotype analysis is summarized in Figure 2. A region of shared homozygosity between affected individuals of Families 66/67 and 89 limits the region to between D19S572 and D19S218 (Fig. 2).GENEHUNTER linkage analysis (35) was performed on the four consanguineous plus one nuclear family (072) (34), assuming recessive inheritance at a values (proportion of linked families) of 0.35, 0.65, and 1 (Fig. 3). The highest multipoint lod score allowing for heterogeniety (HLOD), of 5 at a = 0.65, was obtained between D19S572 and D19S890. These data provide statistically significant evidence for a locus for PCD on 19ql3.4 and confirm locus heterogeneity. CONCLUSIONS
The identification of a locus for PCD on chromosome 19ql3.4 illustrates the power of homozygosity mapping as an approach for linkage analysis of rare, recessive diseases in consanguineous families or very inbred populations, even
in the presence of genetic heterogeneity. Genetic loci can be identified in small family groups selected according to phenotypic trait and/or ethnic origin.Homozygosity mapping uses relatively simple and well-established methodology and is unbiased if the genome is analyzed as a whole when searching for regions that are HBD. In contrast, candidate gene analysis is limited by current understanding of the disease pathology and the number of candidate genes that have been characterized. Conclusive mutation analysis of potential candidates can also be problematic if the full genomic structure of the gene is unavailable.Work to isolate the 19q 13.4 PCD disease gene is currently in progress. Refinement of the genetic map and subsequent transcript identification at this and other PCD loci will be greatly expediated as a result of the accumulation of genetic information from the human genome and the genomes of numerous other species (36). DNA sequences available in public databases as a result of the Human Genome Project initiative will allow more effective genetic linkage and positional candidate gene analysis. Identification of the genes involved in PCD should enhance understanding of the molecular mechanisms involved in cilia function and LRA determination as well as allowing investigation into therapeutic methods.
