CARD15 mutations in familial granulomatosis syndromes: A study of the original Blau syndrome kindred and other families with large-vessel arteritis and cranial neuropathy




To analyze the CARD15 gene in families with heritable multi-organ granulomatoses, including the original Blau syndrome kindred as well as other families with related granulomatous conditions.


Linkage mapping was performed in 10 families. Observed recombination events were used to exclude regions centromeric or telomeric to 16q12.1, and the Blau gene critical region was refined to <3 cM, corresponding to a physical distance of 3.5 megabasepairs. Based on its known biochemical function, CARD15 was analyzed as a positional candidate for the Blau syndrome susceptibility gene, by direct DNA sequencing.


These studies resulted in the identification, in 5 of the families, of 2 sequence variants at position 334 of the gene product (R334W and R334Q). Affected family members from the original Blau syndrome kindred were heterozygous for the R334W missense mutation; mutations at the same position were also observed in several unrelated Blau syndrome families, some of whose phenotypes included large-vessel arteritis and cranial neuropathy. The missense mutations segregated with the disease phenotype in the families, and were not seen in 208 control alleles.


These findings demonstrate that CARD15 is an important susceptibility gene for Blau syndrome and for other familial granulomatoses that display phenotypic traits beyond those of classic Blau syndrome.

Blau syndrome (MIM no. 186580), an autosomal-dominant disorder of childhood, is a rare familial granulomatous disease (1). Affected individuals typically exhibit one or more of the following granulomatous inflammations, which are variable in terms of age at onset: acute anterior uveitis, arthritis (sometimes associated with camptodactyly), and skin rash. Recent studies on families with Blau syndrome have revealed the involvement of other organs in addition to the skin, joint, and eye, as well as other symptoms such as cranial neuropathies (2), fever (3–6), cerebral infarction (7), sarcoid-like hepatic granulomata (6), arteritis and/or malignant hypertension (3–5), and renal lesions (8). The Blau syndrome phenotype therefore may be more complex than previously suspected, and we have suggested that Blau syndrome and related disorders are a subset of the familial granulomatosis syndromes (9).

The phenotype of Blau syndrome shares some features with another multisystemic granulomatous disease, Crohn's disease (CD) (MIM no. 266600), which is a chronic disorder of the intestinal tract (10). CD displays phenotypic overlap with Blau syndrome with respect to organ involvement: while there is no evidence for bowel inflammation in Blau syndrome, CD does present with uveitis, arthritis, and skin rash in some patients (10).

We previously performed a genomewide search for the Blau syndrome susceptibility locus with the original, multi-generation kindred described by one of us (EB) in 1985 (1) and mapped a single risk locus, the Blau syndrome locus, to chromosome 16p12–q21 (11). Of the several putative susceptibility loci for CD, the Crohn's IBD1 locus on 16q between markers D16S409 and D16S419 (56.1–65.6 cM from p-telomere) (12–14) overlaps with that of Blau syndrome. In the present study, we further refined the Blau susceptibility interval by genotyping additional families. The exclusive expression of CARD15 (GenBank accession no. NM022162.1) in monocytes and its relevance to CD led us to examine this gene for mutations in families with Blau syndrome and related heritable granulomatoses.


Family studies and isolation of genomic DNA.

Since our previous genetic study of Blau syndrome (11), we identified 9 additional families for a total of 10 families with 131 participating members, 50 of whom were affected with Blau syndrome. Blood samples were obtained from family members and genomic DNA was extracted using the PureGene DNA Isolation Kit (Gentra Systems, Minneapolis, MN). The project to define the molecular basis for Blau syndrome was approved by the Institutional Review Board at Thomas Jefferson University.

Linkage analyses.

Linkage analyses were performed with the FASTLINK versions of the LINKAGE and SLINK package of programs as previously described (11). Briefly, pairwise logarithm of odds (LOD) scores for disease versus each individual marker were calculated using MLINK, and the maximum 2-point LOD scores (Zmax) were estimated by the ILINK program of the LINKAGE package. The LINKMAP program of the LINKAGE package was used in the multipoint analysis. The analyses were performed using the previously described genetic model that includes liability classes for age-related penetrance and phenocopy potential in Blau syndrome (11).

Analyses of CARD15.

All 12 exons of the CARD15 gene were sequenced from genomic DNA of probands from each Blau syndrome family and several unaffected and unrelated, racially matched controls. Primers for amplification by polymerase chain reaction (PCR) were designed to include at least 100 bp of the flanking intronic sequence. Sequencing was performed by automated dideoxynucleotide terminator chemistry on an ABI 377 sequencer. Sequence variants were confirmed by sequencing the products with putative variants with the antisense primer. The primers for amplification of a region of exon 4 that harbored the sequence variants were as follows: sense (nod2/976) 5′-GAT-GCG-GAC-ACT-GTG-CTG-GT-3′, antisense (nod2/1227) 5′-GAC-ACG-GTC-AGG-GTG-GTC-AA-3′ (the numbers 976 and 1227 denote the complementary DNA [cDNA] positions of the primers). The primers generated a PCR product of 252 bp.


No observable cytogenetic rearrangements, either within or outside of the broad Blau syndrome locus previously identified (11), were found in members of the Blau syndrome pedigree. In the absence of observable cytogenetic abnormalities, efforts to refine the Blau syndrome locus relied on the addition of more Blau syndrome families to our original studies. Since our initial study (11), we identified an additional 9 families with the Blau syndrome (Figure 1). All families displayed the classic Blau syndrome phenotype; however, families 03 and 07 also displayed the additional phenotypes of large-vessel arteritis and cranial neuropathy, respectively. Thirty di-, tri- or tetranucleotide-repeat markers in the Blau susceptibility region were genotyped; haplotype analyses of all 10 families identified several informative recombinations that assisted us in refining the locus of the Blau gene on 16q. The recombinations refined the location of the Blau gene to a critical region of <3 cM (∼3.5 megabasepairs [Mbp]), which was flanked by the centromeric marker D16S409 (48.9 Mbp; Build 29 of the National Center for Biotechnology Information [NCBI] Map Viewer) and the telomeric marker D16S770 (52.4 Mbp; Build 29 of NCBI Map Viewer). No common haplotype was observed in affected individuals from the 10 families studied.

Figure 1.

Pedigrees of 10 Blau syndrome families used in genotyping analyses. Blau01 represents the original Blau syndrome kindred (1). Member 8 of Blau03 died of large-vessel arteritis, and members 4 and 5 of Blau07 have cranial neuropathies.

On the basis of pairwise LOD scores from 2-point linkage analysis, the marker most closely linked to Blau syndrome, at θ = 0.037, appeared to be D16S3137 (54.6 Mbp; Build 29 of NCBI Map Viewer), with an LOD score of 5.67. There was also strong evidence for linkage of Blau syndrome to nearby markers D16S3034 and D16S390, with LOD scores of 4.55 and 3.90, respectively. In order to obtain information on the regional localization of the Blau syndrome and its relationship to flanking marker loci, a multipoint linkage analysis was performed. The results were consistent with the 2-point data (data not shown).

Gene content of the Blau syndrome interval.

The refined Blau syndrome interval contained >80 expressed sequence tags and several known genes, including cerebellin 1 precursor (CBLN1), adenylate cyclase 7 (ADCY7), sal (Drosophila)–like 1 (SALL1), and human CARD15 (15). CARD15 (located at position 51.6 Mbp on Build 29 of NCBI Map Viewer) is a gene that appears to be strongly associated with CD in families displaying linkage to chromosome 16q (16, 17). Furthermore, CARD15 is reported to be exclusively expressed in monocytes (15), a cell type that is a focal component in the granulomas seen in patients with Blau syndrome.

Analyses of CARD15 sequence variants.

Based on its putative biologic function and expression pattern, and association in some CD families (16, 17), we undertook analyses of CARD15. Sequence analysis of all 12 exons of the gene resulted in the detection of 2 sequence variants in 5 of the 10 Blau syndrome families studied. These variants included a C-to-T transition at cDNA position 1105 that substituted an R for a W at amino acid position 334 in the original Blau syndrome family and 1 other family (Blau03), and a C-to-A transversion at the same codon (cDNA position 1106) to substitute an R for a Q in 3 other families (Table 1). The variants were contained in the nucleotide-binding domain (NBD) of the gene product. All variants were confirmed by antisense sequencing. The variants segregated with the phenotype in the Blau syndrome families and were not detected in 208 control alleles derived from unrelated but racially matched individuals (data not shown).

Table 1. Mutations in CARD15 in Blau syndrome families
  • *

    Classic Blau = arthritis, skin rash, uveitis.

Blau01R334WClassic Blau*
Blau03R334WClassic Blau + large-vessel arteritis
Blau05R334QClassic Blau
Blau07R334QClassic Blau + cranial neuropathy
Blau09R334QClassic Blau


Blau syndrome offers a unique opportunity to identify a gene causing an autoimmune disease, since it is one of only a few examples with dominant inheritance. Our previous study localized the Blau syndrome gene to 16p12–q21 (11), a region shared by loci contributing to other autoimmune diseases with less well-defined inheritance. In our current study with additional Blau syndrome families, we were able to refine the Blau syndrome susceptibility interval to a region on the long arm of chromosome 16 that included a likely candidate gene: CARD15. We identified 2 different mutations in CARD15 in families with Blau syndrome. Of interest, only 5 of the 10 Blau syndrome families analyzed had detectable CARD15 mutations despite the fact that all families displayed the classic Blau syndrome symptoms of arthritis, uveitis, and skin rash. There may be other mutations in CARD15 that were not detected in our analyses. A lack of significant quantities of high-molecular-weight DNA precluded our ability to detect a major deletion of CARD15 via Southern blot analysis. Additionally, promoter regions of the gene were not examined, and a lack of RNA precluded the study of transcript levels of CARD15 in patients versus controls.

During the course of our analyses of CARD15, a report by Miceli-Richard et al described the detection of the same 2 variants that we observed, in 3 small European families with Blau syndrome (18). These families expressed classic Blau syndrome, without additional granulomatous lesions in other organ systems. The combination of 5 families with variants at position R334 that we have observed, and the mutations in this position observed in 3 families described by Miceli-Richard et al, would suggest that this position may be a “hot spot” for mutations in patients with some familial granulomatoses. Interestingly, in the same study, Miceli-Richard et al examined 2 families with early-onset sarcoidosis showing granulomatous infiltration of kidney and lung alveoli; they failed to detect mutations in CARD15 and concluded that mutations in CARD15 are involved in Blau syndrome but not in early-onset sarcoidosis. Taken together, our observations and those of Miceli-Richard et al would suggest that mutations in CARD15 may provide an important diagnostic criterion in the nosology of the familial granulomatoses, in light of the complex literature regarding the classification of these syndromes.

With regard to the observation that mutations in CARD15 are found in both Blau syndrome and CD, these 2 distinct diseases share key pathologic features, including granulomatous inflammation and increased T cell activation. In CD, disease-associated polymorphisms are located in the leucine-rich repeat (LRR) regions of the protein, a domain that is involved in the recognition of bacterial lipopolysaccharides. In contrast, the R334 mutations in Blau syndrome localize to the NBD in a region between the P loop and the B box. R at position 334 appears to be an important residue that is conserved in Nod1 and Nod2 orthologs in the mouse and Japanese puffer fish (Fugu rubripes). The NBD in NOD1 and in CARD15 regulates dimerization of these proteins, a function that is critical for activation of nuclear factor κB. Thus, it is likely that mutations in the NBD and LRR of CARD15 result in proteins whose aberrant biologic functions are associated with distinct clinical manifestations. The phenotypic differences between CD and Blau syndrome may also be explained by modifying loci, especially with respect to the complex genetics of CD.

In conclusion, we have demonstrated dominant CARD15 mutations in 5 families with Blau syndrome, including the original kindred described in 1985 (1) and 4 other families, 2 of which displayed extended phenotypes such as large-vessel arteritis and cranial neuropathy. All observed mutations reside at amino acid 334. These observations support our contention that Blau syndrome is a systemic disease; furthermore, the genotype/phenotype correlations presented here and elsewhere suggest that mutations in CARD15 may provide an important diagnostic criterion in the nosology of the familial granulomatoses.


The authors thank the family members for their cooperation, and Rodney McGrath for technical assistance.