SCID is defined by profound deficiency of T cell numbers and of T and B cell function, leading to opportunistic infections and early death unless treated[9
]. We describe the first case of human SCID due to CORO1A
deficiency. Our patient developed progressively worsening infections that necessitated HCT, which has been curative. Although the majority of SCID infants are diagnosed in the first year of life, delayed presentations have been described, some of which involve leaky genetic defects, such as hypomorphic mutations allowing residual activity of the ADA enzyme[10
]. In contrast, our patient carries null alleles of CORO1A
. Thus, her delayed diagnosis reflects either meticulous management of infections, preservation of some functions in a very small pool of peripheral T cells despite Coronin-1A deficiency, or both.
Coronins regulate the actin cytoskeleton through antagonizing actin polymerization and promoting actin severing[3
], in contrast to the role of Wiskott Aldrich syndrome protein (WASP) that promotes actin polymerization[2
]. While WASP has already shown how actin can be linked to hematopoietic cell dysfunction, CORO1A
deficiency now highlights the role of actin cytoskeleton regulation in T cell homeostasis and therefore SCID. Mutations causing SCID have previously been defined in genes that disrupt pathways mediating antigen receptor rearrangement (RAG1/2
); the transmission of signals delivered by antigen, cofactor or cytokine binding (CD3
component genes, CD45
); or purine metabolism (ADA
). Our patient’s mutations in CORO1A
now define defective actin regulation as a new pathway in which genetic lesions causing SCID can be found.
Distinctive features of this patient’s immunodeficiency are consistent with findings in mouse models of Coronin-1A deficiency. Unlike other coronin family members, Coronin-1A is expressed primarily in the hematopoeitic system[3
], and in Coronin-1A deficient mice, only the T cell compartment was affected[4
]. Additional coronin family members, such as Coronin-1B or Coronin-1C, may compensate for Coronin-1A deficiency in other leukocytes. This functional compensation appears to occur similarly in humans because our patient had a T-B+NK+ SCID phenotype. The T cells recoverable in small numbers from Coronin-1A deficient mice have an intrinsic migration defect and diminished, but not absent signaling[4
]; similarly our patient’s few T cells had residual function, including proliferative responses to PHA and ConA, and enough B cell helper function to produce normal serum antibody concentrations and low, but measurable antibody titers to pneumococcal protein conjugates (). Pre-HCT T cells were not available to measure T cell receptor diversity. However T cell function was insufficient to provide normal immune protection from infectious agents such as vaccine strain varicella virus, necessitating eventual HCT therapy. This T cell specific defect suggests that Coronin-1A could be a future therapeutic target for selective modulation of T cell immune responses.
Dysregulation of the actin cytoskeleton impairs survival and egress of mature single positive CD4+CD8− or CD4−CD8+ T cells from the murine thymus[4
]. In contrast to most human SCID patients in whom a small or undetectable thymus is a helpful diagnostic feature[14
], our Coronin-1A deficient patient had a thymic image on a pre-transplant CT scan, consistent with thymic accumulation of T cells that were unable to exit into the peripheral compartment[5
]. Interestingly, an intact thymus has also been reported in SCID patients with CD3δ deficiency, associated with impaired differentiation beyond the somewhat earlier double negative CD4−CD8− stage of thymocyte maturation[15
The patient’s de novo 600 kb deletion on her maternal copy of chromosome 16p11.2 is flanked by 146 kb segmental duplications with >99% identity (annotated as duplication cluster 3576 [16
]). Genomic regions flanked by such duplications are prone to copy number variation (CNV) through non-allelic homologous recombination[17
]. Indeed, deletion as well as duplication of this same interval on 16p11.2 is well recognized and has been associated with autism spectrum disorder and neurodevelopmental disorders including ADHD[18
]. Thus, in addition to contributing to CORO1A
deficiency, our patient’s 16p11.2 deletion also predisposed her to ADHD. CORO1A
is expressed in microglia[21
], but the particular gene or genes within the deleted region that are related to neurological phenotypes of 16p11.2 CNV are not yet known. Reduced Coroin-1A is not likely to be the basis for neurocognitive difficulties in 16p11.2 CNV cases since its complete deficiency was not accompanied by a severe neurocognitive phenotype in our patient.
CNV is an increasingly appreciated form of genomic diversity with reports estimating that up to 12% of human genes are located in CNV-prone regions[22
]. Both de novo and inherited CNVs are recognized in many human conditions[23
]. Low copy number repeats flank the DiGeorge syndrome region at 22q11.2[27
] and Alu-mediated deletions have resulted in Bruton’s agammaaglobulinemia[16
] and ADA deficiency[31
]. CNV at 16p11.2 has been estimated to occur in 1.5% of subjects with developmental delay, 1% of individuals with autism, and 0.1% in patients with psychiatric or language disorders, but less than 0.01% in the general population[18
]. Given the propensity for CNV at 16p11.2, we anticipate that future cases T-B+NK+ SCID will involve CORO1A
defects, including deletions. The flow cytometric assay described here is an efficient means to evaluate patients who have SCID or combined immunodeficiency of unknown genotype, or atypically severe varicella, for Coronin-1A protein expression. Moreover, molecular investigation with high-resolution genomic arrays may reveal additional CNVs that underlie other undefined cases of primary immunodeficiency.