During the current reporting period we have focused on 2 new diseases identified through whole-exome sequencing. We discovered the first, denoted the deficiency of adenosine deaminase 2 (DADA2), during the previous reporting period. During the past year we completed our initial characterization of DADA2 and published our findings in the New England Journal of Medicine. The second disorder, caused by mutations in TRNT1, was discovered in our laboratory during the last six months, and is now under active investigation. Both projects are summarized below. Deficiency of ADA2, a New Autoinflammatory Disease Beginning about 10 years ago, we saw a total of 5 patients at the NIH Clinical Center with intermittent fevers, recurrent lacunar strokes, elevated acute phase reactants, livedoid rash, hepatosplenomegaly, and hypogammaglobulinemia. In 4 patients skin biopsies of livedoid lesions showed an inflammatory infiltrate of neutrophils, macrophages, and T lymphocytes;in one patient necrotizing vasculitis was present in the deep dermis. Magnetic resonance imaging showed evidence for small vessel acute brain ischemia and old ischemic strokes in the deep nuclei of the brain. Several stroke events were hemorrhagic or underwent hemorrhagic transformation, although the interpretation is clouded by the concomitant use of anti-platelet agents and/or warfarin. Four patients had hepatosplenomegaly, and one had documented portal hypertension. IgM levels were consistently low in all 5. We performed whole-exome sequencing in the first 3 patients and their unaffected parents, filtering the data under both recessive and de novo dominant models. A single common candidate gene was identified only under the recessive model. We found compound heterozygous predicted deleterious mutations in CECR1 (chromosome 22), encoding adenosine deaminase type 2 (ADA2). Two of the patients shared the p.Tyr453Cys mutation;one patient had a 28 kb genomic deletion that included the 5 prime UTR and exon 1 of CECR1. Among the 3 patients we identified a total of 5 CECR1 mutations. We then sequenced CECR1 in the 2 remaining NIH patients and a patient from the UK with a similar phenotype. All 3 were compound heterozygous for CECR1 mutations, including 3 novel mutations. We subsequently performed CECR1 candidate gene sequencing on 3 Turkish patients, 2 brothers who carried the diagnosis of polyarteritis nodosum (PAN) and a third patient with necrotizing small vessel vasculitis. All 3 were homozygous for the p.Gly47Arg mutation. Among these 9 patients we identified 9 predicted deleterious ADA2 mutations, including 2 null mutations. Mutations in the related protein ADA1 cause severe combined immunodeficiency disease (SCID) because of the failure to catalyze the conversion of adenosine (and deoxyadenosine) to inosine (and deoxyinosine), and the consequent accumulation of toxic metabolites. ADA2 also catalyzes this reaction, but with a lower affinity for adenosine. Using enzyme assays that distinguish ADA1 from ADA2 activity, we found that patients had nearly absent ADA2 activity, but normal ADA1 activity, in the blood. Direct assays of adenosine/deoxyadenosine metabolites were negative. Assays of serum cytokines, peripheral T cell function, and peripheral B cells demonstrated modest defects in the B cell compartment. Although there is no clear CECR1 ortholog in the mouse, there are 2 paralogs in the zebrafish. Morpholino knockdown of one of the zebrafish paralogs (cecr1b) caused intracranial hemorrhages and neutropenia, phenotypes that were rescued by wild type but not mutant human CECR1. These data, in conjunction with data from the literature indicating that ADA2 may be the prototype for a family of adenosine deaminase-derived growth factors, prompted a closer examination of the role of ADA2 in endothelial and leukocyte development in patients. We found that ADA2 is not produced in human endothelial cells. From the literature it is known that ADA2 is expressed in myeloid cells. Immunohistochemistry of skin and brain biopsies from patients demonstrates both endothelial damage and activation;staining for IL-1beta, TNF-alpha, and iNOS indicates inflammation. Using shRNA to silence the expression of ADA2 in myeloid U937 cells, and studies of patient monocytes stimulated with defined growth factors, we found that ADA2 deficiency is associated with a skewing of monocytes to the more inflammatory M1 subpopulation. When shRNA-treated U937 cells, or untreated patient monocytes, were cocultured with endothelial cell layers, they induced damage, relative to control U937 cells or healthy control monocytes. Thus, loss-of-function mutations in CECR1 cause a spectrum of vascular and inflammatory phenotypes ranging from early-onset recurrent stroke to systemic vasculopathy and/or vasculitis. Together, zebrafish and patient data indicate that ADA2 deficiency may diminish endothelial integrity while polarizing macrophage/monocyte subsets towards pro-inflammatory cells, establishing a vicious circle of vasculopathy and inflammation. Since the publication of our initial paper, we have continued to monitor the natural history of this disorder, and have identified 8 additional patients. There are now a total of 15 documented missense CECR1 mutations and 1 genomic deletion. Based on encouraging data from Israeli PAN patients with CECR1 mutations, some of our patients are now receiving TNF inhibitors. We are also evaluating the pharmacokinetics of ADA2 blood levels in patients who have received infusions of fresh frozen plasma, we are beginning to screen other vasculitis patients for CECR1 mutations, and we are establishing zebrafish cecr1 knockout lines for mechanistic studies. TRNT1-Associated Periodic Fever Syndrome Through whole-exome sequencing and candidate gene screening, we identified 5 children from 4 unrelated families with unexplained autoinflammatory disease and shared mutations in one common gene. All patients carried missense recessive mutations in TRNT1 (tRNA nucleotidyl transferase, CCA-adding, 1), encoded on chromosome 3. Two affected sisters from a Saudi Arabian consanguineous family were homozygous for a p.His215Arg missense mutation, while the other 3 children were compound heterozygous for a missense mutation, p.Ile223Thr or p.Arg99Trp, and one shared mutation, p.Asp163Val. p.His215Arg was not found in any public database nor in 1061 Arab control DNA samples. Among the 3 Caucasian mutations, p.Arg99Trp was novel whereas p.Ile223Thr and p.Asp163Val were found at a very low allele frequency (<0.001) in the NHLBI exome database. All mutations affect highly conserved amino acid residues and are predicted to be damaging to protein function. The TRNT1 protein adds CCA to the 3 prime end of all transfer RNAs both in the cytoplasm and mitochondria, which is critical for tRNA aminoacylation, protein synthesis, and tRNA degradation. Given the essential nature of this enzyme, disease-associated mutations are likely to be hypomorphs. All patients had recurrent episodes of fever and gastrointestinal symptoms with multisystem features that included developmental delay, nystagmus, hypotonia, optic nerve atrophy, sensorineural hearing loss, dysmorphic features, musculoskeletal symptoms, and B cell immunodeficiency. Flow cytometric studies suggested a maturation defect of the B cell lineage in the bone marrow. Preliminary data from cytokine analyses in two patients have shown elevated levels of the proinflammatory cytokines interleukin 6 and type 1 interferon. Knockdown of the zebrafish TRNT1 homolog caused hydrocephaly, defects in tail development, anemia, and a reduction in the number of hair cells present in the lateral line, which functions similarly to the human inner ear. Current studies focus on the relationship between defective tRNA processing and inflammation.
Showing the most recent 10 out of 33 publications
