A syndrome with congenital neutropenia and mutations in G6PC3

Abstract

Background: The main features of severe congenital neutropenia are the onset of severe bacterial infections early in life, a paucity of mature neutrophils, and an increased risk of leukemia. In many patients, the genetic causes of severe congenital neutropenia are unknown.

Methods: We performed genomewide genotyping and linkage analysis on two consanguineous pedigrees with a total of five children affected with severe congenital neutropenia. Candidate genes from the linkage interval were sequenced. Functional assays and reconstitution experiments were carried out.

Results: All index patients were susceptible to bacterial infections and had very few mature neutrophils in the bone marrow; structural heart defects, urogenital abnormalities, and venous angiectasia on the trunk and extremities were additional features. Linkage analysis of the two index families yielded a combined multipoint lod score of 5.74 on a linkage interval on chromosome 17q21. Sequencing of G6PC3, the candidate gene encoding glucose-6-phosphatase, catalytic subunit 3, revealed a homozygous missense mutation in exon 6 that abolished the enzymatic activity of glucose-6-phosphatase in all affected children in the two families. The patients' neutrophils and fibroblasts had increased susceptibility to apoptosis. The myeloid cells showed evidence of increased endoplasmic reticulum stress and increased activity of glycogen synthase kinase 3beta (GSK-3beta). We identified seven additional, unrelated patients who had severe congenital neutropenia with syndromic features and distinct biallelic mutations in G6PC3.

Conclusions: Defective function of glucose-6-phosphatase, catalytic subunit 3, underlies a severe congenital neutropenia syndrome associated with cardiac and urogenital malformations.

Figure 1. Clinical and hematological phenotype

Panel A shows severe paucity of mature neutrophils in bone marrow smears from patient P1. In contrast, bone marrow smears from a healthy individual (Panel B) show normal myeloid differentiation. Panel C illustrates the phenotype of unusual visibility/angiectasia of subcutaneous veins, which was found in all patients from the two pedigrees SCN-I and SCN-II. Panel D shows an atrial septal defect, type II, in patient P3.

Figure 2. Haplotype on Chromosome 17q, mutation analysis of G6PC3 and assessment of enzymatic activity of G6PC3^R253H

Panel A indicates the pedigrees from 2 unrelated SCN families (SCN-I and SCN-II). Affected patients are marked with filled symbols, whereas unaffected family members are represented by open symbols. P1–P5 refer to the patients described in more detail in Table 1. The linkage interval is shown in gray for each family. The preferred linkage interval was defined under the assumption that the genetic defect in SCN-I and SCN-II was the same, and ranged from markers D17S930 to D17S950, which contained a total of 36 genes including G6PC3. Panel B shows the missense mutation (c. G758A, p. R253H) found in an index patient from SCN-I as compared to the sequence in a healthy control. Panel C represents the enzymatic activity of wild type and G6PC3^R253H with [¹⁴C]G6P as substrate compared to an empty vector control measured in microsomes isolated from stably transfected yeast cells. The presence of 0.1% saponin, a mild detergent preserving the membrane integrity, allows substrate penetration of the microsomes. In the presence of 0.5% triton, membranes are disrupted completely. The difference between the values represents the membrane dependent activity. The results are presented as nmol of glucose produced per minute per gram of yeast cells. Error bars indicate the standard deviation of three independent microsomal preparations. * Significantly different with p = 0.012.

Figure 3. Increased susceptibility to apoptosis in myeloid cells and fibroblast cell lines

Panel A shows a FACS plot indicating that peripheral blood neutrophils from patient P5 had increased spontaneous apoptosis as compared to a healthy sibling or a rh-G-CSF treated healthy control (HD) (x axis, Annexin-V; y axis, propidium iodide; logarithmic scales in both axes). Similarly, TNF-alpha induced apoptosis was more pronounced in patients. Panel B shows an analysis of enhanced apoptosis in patient fibroblasts after exposure to dithiothreitol (DTT) as compared to normal donors. Panel C shows a FACS plot of in vitro differentiated myeloid progenitor cells transduced with a control vector (upper row) or a retroviral vector encoding wildtype G6PC3 (lower row). Cells were treated with tunicamycin and apoptosis was assessed using Annexin-V (x axis) and propidium iodide staining (y axis). G6PC3-transduced patient cells from patient P2 show a decreased susceptibility to apoptosis 16 hours after induction with tunicamycin, as compared to control vector-transduced cells. Data are representative of two independent experiments.

Figure 4. Pathophysiologic consequences of G6PC3 deficiency

Panel A shows the aberrantly enlarged rough ER in myeloid progenitor cells as compared to a healthy individual (Panel B). Panel C indicates increased ER stress in G6PC3-deficient patients, as documented by increased expression levels of Bip mRNA in purified promyelocytes. Panel D represents a Western Blot revealing a decrease in the enzymatically active, dephosphorylated form of GSK3β in patient P5 after induction with tunicamycin. Panel E shows increased phosphorylation of Mcl1 and increased levels of Bip in neutrophils from a G6PC3-deficient patient (P5) at different timepoints after induction with tunicamycin. Panel F shows 2-deoxyglucose (2DG)-induced dephosphorylation of GSK3β in 2 healthy donors. Panel G indicates that healthy neutrophils, but not CD3-positive T cells, are susceptible to apoptosis upon pharmacological glucose-depletion.