TO THE EDITOR:
The human proteins rabenosyn-5 and VPS45 form a complex that plays a key role in early endocytosis.1-4 Pathogenic variants in VPS45 cause severe congenital neutropenia (SCN) with impaired neutrophil function, reticulin fibrosis of the bone marrow, and extramedullary hematopoiesis (Online Mendelian Inheritance in Man [OMIM]: 615285).5-8 Patients with a specific VPS45 variant (p.Glu238Lys) also have intellectual disability and bilateral optic nerve hypoplasia.5,6 To date, the only evidence of a potential role for RBSN in human disease is the report of a homozygous missense variant (p.Gly425Arg) in a patient with intellectual disability, seizures, microcephaly, osteopenia, mild reticulin fibrosis of the bone marrow, and transient neutropenia.9
We evaluated 3 siblings born to consanguineous parents for a phenotype of SCN, anemia, thrombocytopenia, reticulin fibrosis of the bone marrow, dysmorphic facial features, osteopenia, hypertriglyceridemia, hepatomegaly, microphthalmia, and optic nerve hypoplasia (Figure 1A-E). Bone marrow examination revealed trilineage hematopoiesis without maturation arrest (supplemental Figure 1, available on the Blood Web site). Myelofibrosis was documented as early as age 4 weeks and was progressive and severe (supplemental Figures 1 and 2). The proband (patient VI.2) had complete 46,XY male-to-female sex reversal and died at age 20 months after multiple infections. At autopsy, she was found to have extensive extramedullary hematopoiesis. The other 2 affected siblings (patients VI.3, and VI.4) are currently ages 9 and 5 years. They underwent unrelated-donor bone marrow or stem cell transplantation at 8 and 6.5 months of age, respectively. Posttransplantation bone marrow examinations have shown a reduction in myelofibrosis (Figure 1F), and their circulating blood counts have remained normal (supplemental Figure 3). Both have severe intellectual disability and are nonambulatory and nonverbal. Clinical presentation is detailed in Table 1, supplemental Figures 1-3, and supplemental Table 1.
The parents and a fourth sibling (patient VI.5 who is currently 7 months of age) have no neurologic or ophthalmologic abnormalities. The fourth sibling had transient neonatal thrombocytopenia and anemia and has persistent, mild neutropenia (supplemental Figure 4). The mother has had chronic moderate thrombocytopenia, an elevated immature platelet fraction, and peripheral blood smear with macrothrombocytes among variably sized platelets. These findings have been interpreted as consistent with immune thrombocytopenia. Her complete blood count is otherwise normal. The father’s complete blood count is normal.
This study was approved by the Institutional Review Board for Human Subject Research at Baylor College of Medicine. The parents gave written informed consent before evaluation, counseling, and testing.
Whole-exome sequencing on the proband (patient VI.2) revealed a variant (NM_022340.3:c.289G>C; NP_001289307.1:p.Gly97Arg) in RBSN, the gene encoding rabenosyn-5, which was homozygous in the 3 affected children, heterozygous in each parent and patient VI.5, and predicted to affect messenger RNA splicing and protein function (supplemental Figures 5 and 6A).
To assess splicing, we sequenced complementary DNA libraries from the proband (patient VI.2) and the father. Normally spliced transcripts were not observed in the proband and represented 45% of transcripts in the father. The most common abnormal transcripts had skipping of exon 5 or an alternative exon 5a (supplemental Figure 7). High-throughput RNA sequencing confirmed these observations and revealed abnormally spliced transcripts that retained exon 5 and its neighboring introns (supplemental Figure 6B).
Immunoblotting showed an absence of intact rabenosyn-5 in the affected patients (supplemental Figure 6C). Immunofluorescence staining revealed larger, clustered EEA1-positive endosomes in patient fibroblasts compared with controls (supplemental Figure 6D). These findings suggest that the absence of intact rabenosyn-5 affects the structure and function of early endosomes.
Transferrin uptake and recycling assays revealed quantitative and qualitative differences in cells from the affected patients compared with those from a healthy control. After 30 minutes of uptake, transferrin was localized to a more clustered perinuclear region in cells from the affected patient (supplemental Figure 6E). In addition, total transferrin accumulation was higher and the washout phase was delayed in cells from the affected patient (supplemental Figure 6F). These findings are consistent with a decreased rate of recycling of the transferrin receptor and increased steady-state intracellular accumulation of the ligand in cells harboring the RBSN variant. The perinuclear clustering and the increased transferrin accumulation were reversed in cells transfected with a version of the human RBSN gene that does not have the variant (supplemental Figure 8).
Our observations suggest that RBSN loss of function causes a severe syndromic form of congenital myelofibrosis and confirm that rabenosyn-5 plays a vital role in human development and hematopoiesis. The mechanisms by which RBSN loss of function leads to hematologic and developmental abnormalities are unknown. As previously suggested for VPS45, the neutropenia may result from apoptosis induced by the toxic effects of impaired endosomal trafficking.5,7 It is important to note, however, that other congenital disorders associated with increased neutrophil apoptosis, such as pathogenic variants in ELANE, do not result in congenital myelofibrosis. Thus, additional factors must be at play. One hypothesis is that the myelofibrosis results from impaired trafficking of α-granules and their release from megakaryocytes, resulting in a proinflammatory, profibrotic response. In addition, given that RBSN loss of function is predicted to result in a reduction of β1 integrin expression at the cell surface, as observed for VPS45 deficiency,5 it is tempting to speculate that the optic pathway and other central nervous system effects are the result of abnormal axonal transport of integrin transmembrane proteins during development.
The phenotypic overlap between our patients and those with mutations affecting VPS455-8 suggests the existence of a distinct disorder of early endocytosis caused by pathogenic variants in the genes encoding either member of the rabenosyn-5/VPS45 complex. SCN without maturation arrest, reticulin fibrosis of the bone marrow, and extramedullary hematopoiesis seem to be the hallmarks of the disease. Anemia and thrombocytopenia in our patients were variable and exacerbated by infection and progressive hepatosplenomegaly (supplemental Table 1; supplemental Figure 3). Long-term survival without substantial hematologic or infectious complications was possible after bone marrow or stem cell transplantation, suggesting that the hematopoietic defect is intrinsic to hematopoietic stem and progenitor cells, although a transferable factor from the transplanted cells to the bone marrow cannot be excluded. The neurologic and ophthalmologic abnormalities in the surviving patients have remained stable. Other concordant features in our patients include facial dysmorphism, abnormal bone development, and hypertriglyceridemia. The only sibling homozygous for the RBSN variant who had a 46,XY chromosome complement also had a disorder of sex development. This seems to be part of the syndrome, and we hypothesize that it could be the result of ineffective endocytosis of sex hormones in utero, as has been observed in animal models.10 Confirmation will require additional cases of 46,XY individuals with loss-of-function RBSN variants.
The mother’s hematologic abnormality is consistent with immune thrombocytopenia. This could explain the transient neonatal thrombocytopenia in the heterozygous sibling (patient VI.5), but it would not explain his transient anemia and persistent, mild neutropenia. It is certainly possible that the hematologic manifestations in these 2 individuals represent a milder phenotype caused by the RBSN variant in heterozygous state.
The structural and phenotypic effects of the variant we report are considerably more severe than those of the previously reported RBSN variant (p.Gly425Arg).9 In that patient, intact rabenosyn-5 was detected by immunoblot, the subcellular localization of the protein by immunofluorescence staining was normal, and transferrin endocytosis and recycling assays showed a significantly enhanced rate of recycling suggestive of a gain-of-function effect. However, the phenotypic overlap with our patients suggests that the p.Gly425Arg variant is also pathogenic, leading to a milder phenotype within the same spectrum. Therefore, pathogenic RBSN variants, in homozygous or heterozygous state, should also be considered in cases of unexplained transient neutropenia, particularly when neurologic abnormalities and/or myelofibrosis are also present.
The online version of this article contains a data supplement.
Acknowledgments
The authors thank the family who participated in the study, the health care providers who referred them to us for evaluation, and Cynthia Dunbar for critical review of the manuscript.
This work was supported by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (NIH) (L.M.F.) and the Intramural Research Program of the National Human Genome Research Institute at NIH (O.A.S.).
Authorship
Contribution: O.A.S., M.N.B., S.Y., C.S., G.Z., P.P.H., M.G., D.E., D.M.M., R.A.G., A.A.B., S.C., and L.M.F. designed and performed the research; P.L.M., O.A.S., S.B.-S., L.P., R.A.L., A.A.B., D.A.S., and L.M.F. collected data; P.L.M., O.A.S., M.N.B., S.Y., R.A.L., C.S., A.N.M., M.T.E., G.Z., P.P.H., M.G., D.E., D.M.M., R.A.G., A.A.B., D.A.S., S.C., and L.M.F. analyzed and interpreted data; S.C. performed statistical analysis; P.L.M., O.A.S., R.A.L., A.A.B., D.A.S., S.C., and L.M.F. wrote the manuscript; and P.L.M., O.A.S., and L.M.F. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Luis M. Franco, Laboratory of Immune System Biology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, 9000 Rockville Pike, Building 4, Room 138, Bethesda, MD 20892; e-mail: luis.franco@nih.gov.
References
Author notes
P.L.M. and O.A.S. contributed equally to this study.