Background
Ankylosing spondylitis (AS) is a severe, debilitating and incurable autoimmune disease affecting multi millions of people in the world [
1]. AS is the major subtype of spondyloarthritis, a spectrum of inter-related rheumatic diseases that also includes reactive arthritis (ReA), psoriatic arthritis (PsA), juvenile spondyloarthritis (JSpA), enteropathic arthritis (spondylitis/arthritis associated with inflammatory bowel disease), and undifferentiated spondyloarthritis (USpA) (
https://www.spondylitis.org) [
2,
3].
The symptoms of AS often appear gradually and progress from stiffness and chronic dull pain in the lower back to severe pain felt from the sacroiliac joint. In severe cases, AS can cause a complete fusion (ankylosis) of the spine. Many AS patients suffer severe loss of mobility and as a consequence lose working capabilities.
The average AS prevalence varies among different populations, with 0.24% in Europe, 0.17% in Asia, 0.3% in North America, 0.1% in Latin America and 0.7% in Africa. Based on these ratios, it is estimated that the number of AS cases in Europe and Asia alone could reach 1.30–1.56 million and 4.63–4.98 million, respectively [
1].
AS is highly inheritable. The recurrence risks in monozygous twins and first-degree relatives are 63 and 8.2%, respectively [
4,
5]. Most AS is presented in sporadic cases, probably reflecting the oligogenic nature of this disease. The human leukocyte antigen B27 (
HLA-B*27) genotype was found strongly associated with AS [
6‐
8]. Specifically
HLA-B*27 carriers could have a 20-fold increase in the risk of developing spondylarthropathy-related diseases [
9], which is exemplified by the fact that most AS patients are
HLA-B*27 positive in the general population. However the presence of
HLA-B*27 genotype is not sufficient for AS pathogenesis, as only 1–5%
HLA-B*27 carriers eventually develop AS [
8,
10,
11].
Recently large-scale genome-wide association studies on patients with European ancestry and of the Han Chinese have identified at least 31 non
-HLA-B genetic loci associated with AS [
11‐
17]. Among these loci,
IL23R,
2p15,
ERAP1 exhibit the most significant association [
12,
15‐
17]. Nevertheless these loci, together with
HLA-B*27, could explain only 24.4% of the heritability of AS [
15]. Therefore the major genetic causes of AS remain to be identified. Ideally, AS cases from consanguinity inheritance or large pedigrees could provide a simpler inheritance pattern compared to sporadic cases, which might facilitate the identification of more elusive risk genes of AS.
To further understand the genetics of AS, we employed a combination of genome-wide linkage analysis and next-generation sequencing (NGS) in a three-generation Han Chinese pedigree with five AS patients. The analysis revealed a missense variant affecting a conserved amino acid in the novel gene ANKDD1B that segregates with the disease. We further identified a distinct ANKDD1B missense variant in a male by surveying a group of sporadic AS patients using exome sequencing. These findings suggest that ANKDD1B variants might be related with the pathogenesis of AS.
Discussion
In this study, we combined genome-wide linkage analysis and exome sequencing to identify the ANKDD1B gene as a potential locus related ankylosing spondylitis in a Chinese AS pedigree.
AS is an oligogenic disease with over 75% of the heritability unexplained [
11]. Of the 32 identified loci associated with AS [
11,
15],
HLA-B*27 is the most significant risk factor, followed by
IL23R,
2p15 and
ERAP1 [
15]. Among the seven linkage regions we identified in AS9 pedigree (Fig.
2b), one contains the
HLA-B locus and two others are closely linked to the
2p15 and
ERAP1 loci, respectively (Fig.
2c). This correlation does not appear to be a random coincidence. Instead, it suggests that these linkage regions more or less correspond to the risk loci of AS and could be useful for distinguishing other genetic variants related to AS, e.g., those variants derived from exome sequencing. Indeed, among the six candidate variants that we identified from exome sequencing of the AS9 pedigree, only
ANKDD1B is contained within a linkage region on Chr. 5 (Fig.
2c). Furthermore, a different
ANKDD1B variant (R102L) was identified in a sporadic AS patient, emphasizing the potential correlation between
ANKDD1B variants and AS in the Chinese patients.
The genetic location of
ANKDD1B is
5q13.3 (
www.ensembl.org), which is 0.1 cM away from
5q14.3, a previously identified risk locus in Asian populations [
16]. The significance of
5q14.3 locus was not replicated in the study of European populations [
15]. Therefore it is unclear whether
5q14.3 represents an Asian-specific locus or the association of this locus with AS [
16] was in fact caused by its closeness with
ANKDD1B. The linkage region on Chr. 5 identified in this study is also close to but does not contain the risk locus
ERAP1 (Fig.
2c). Whether the haplotype of
ERAP1 affects the identification of this linkage region is unknown. The linkage region that we identified on Chr. 2 is close to another major AS risk locus,
2p15 [
15] (Fig.
2c). Currently we could not determine whether the identification of this linkage region is caused by its closeness to
2p15 or it represents a different risk locus for AS. The other linkage regions (Table
2, in blue) do not contain any of the AS risk loci identified so far [
11].
ANKDD1B encodes a novel protein containing 10 tandem ankyrin repeats and a death domain and is conserved from zebrafish to human (Additional file
7: Figure S1). Ankyrin repeats are found in proteins involved in inflammatory response, transcription, cell-cycle regulation, cytoskeleton integrity development, cell-cell signaling and protein transport [
20]. Death domains are usually regulators of inflammation, innate immune response and cell death through their interactions with TNF (tumor necrosis factor) receptors and Toll-like receptors [
21‐
23].
A protein expression database (
www.proteinatlas.org) indicates that the ANKDD1B protein is expressed in muscles, distinct cells of the lymph node and tonsil, a chronic myeloid leukemia cell line (K-562), a multiple myeloma cell line (LP-1) and an ovarian cystadenocarcinoma cell line (EFO-21). In addition the
ANKDD1B transcript was detected in multiple human tissues including the lymph node and spleen (
www.genecards.org). In mouse, the
ANKDD1B transcript is expressed in higher levels in multiple B- and T-cell lines and in various parts of the nervous system (
www.biogps.org). The site of expression and the domain structure of ANKDD1B suggest a function in the immune system.
The complex oligogenic nature of AS hinders the genetic analysis of AS pathogenesis. A previous analysis of AS recurrence in relation to genetic distances estimated that multiplicative interactions involving ~five genetic loci could partially explain the occurrence of AS in the general population [
5]. In our pedigree and the sporadic male patient, the presence of the
ANKDD1B variant and
HLA-B*27 are the most significant factors predicting whether an individual develops AS (Table
1), while the roles of three other GWAS-derived significant loci (
IL23R,
2p15,
ERAP1) could not be evaluated, probably due to the limited size of our cohort.
In the AS9 pedigree, all four male carriers are
HLA-B*27 positive and developed AS (Fig.
2a and Table
1). Four of the five female carriers are
HLA-B*27 positive, while only one developed AS (Fig.
2a and Table
1). This phenomenon is consistent with AS epidemiology in the general population, in which males are three times more frequently affected than females [
2].
Besides the AS9 pedigree described in our study, three other Han Chinese AS pedigrees were also described recently [
24], in which the
2q36.1-36.3 locus was found to be associated with disease transmission in addition to
HLA-B*27. This locus has not been associated with AS in other association studies [
11]. It is possible that a combination of genome-wide linkage analysis and genome sequencing would narrow down the potential AS-related genetic variants in these pedigrees.