Severe congenital neutropenia resulting from G6PC3 deficiency with increased neutrophil CXCR4 expression and myelokathexis

Abstract

Mutations in more than 15 genes are now known to cause severe congenital neutropenia (SCN); however, the pathologic mechanisms of most genetic defects are not fully defined. Deficiency of G6PC3, a glucose-6-phosphatase, causes a rare multisystem syndrome with SCN first described in 2009. We identified a family with 2 children with homozygous G6PC3 G260R mutations, a loss of enzymatic function, and typical syndrome features with the exception that their bone marrow biopsy pathology revealed abundant neutrophils consistent with myelokathexis. This pathologic finding is a hallmark of another type of SCN, WHIM syndrome, which is caused by gain-of-function mutations in CXCR4, a chemokine receptor and known neutrophil bone marrow retention factor. We found markedly increased CXCR4 expression on neutrophils from both our G6PC3-deficient patients and G6pc3(-/-) mice. In both patients, granulocyte colony-stimulating factor treatment normalized CXCR4 expression and neutrophil counts. In G6pc3(-/-) mice, the specific CXCR4 antagonist AMD3100 rapidly reversed neutropenia. Thus, myelokathexis associated with abnormally high neutrophil CXCR4 expression may contribute to neutropenia in G6PC3 deficiency and responds well to granulocyte colony-stimulating factor.

Figure 1

Patient phenotypes: brother (left panels); sister (right panels). (A) Neutropenia. All available absolute neutrophil counts (ANC) values were plotted by date (nonlinear time scale). Neither patient received G-CSF or other hematopoietic growth factors during the period of observation plotted. Green bars represent clinically uninfected; red bars, active clinical infection; and black bars, unknown infection status. (B) Ectatic superficial veins on chest, abdomen, and extremities. (C) Cardiac defects. Echocardiography revealed a thickened and stenotic mitral valve (arrow) with reduced excursion in the brother and a catheter-implanted atrial septal defect occlusion device (arrow) in the sister. (D) Pulmonary infiltrates and bronchiectasis on chest computed tomography.

Figure 2

G6PC3 deficiency resulting from homozygous inheritance of G6PC3 mutation G260R in 2 siblings with a multisystem syndrome involving severe congenital neutropenia and recurrent infections. P1 indicates affected brother; and P2, affected sister. (A) Family pedigree. Circles represent female; squares, male; filled symbols, affected; and unfilled symbols, unaffected. Numbers indicate subject ages. (B) DNA sequencing. Chromatograms from family members (father refused). (C) G6PC3 G260R is a complete loss-of-function mutation. G6PC3 activity was measured in lymphoblastoid cell lines from 2 healthy blood donors (HD1 and HD2) and both patients.

Figure 3

Patient blood and bone marrow histopathology. (A-F) Human data for brother (A,C,E) and sister (B,D,F). Similar results were observed for both siblings. The patient samples shown were obtained before the patients had ever received hematopoietic growth factors and at a time when there was no evidence of active infection. (A-B) Patient peripheral blood smears. (A) Giant platelets (arrow) and vacuolization of monocytes. (B) Rare but morphologically normal neutrophils in the blood. (C-F) Patient bone marrow smears. (C-D) All stages of myeloid differentiation through an abundance of morphologically mature neutrophils are evident. (E) Vacuolization of 2 metamyelocytes. A majority of megakaryocytes were monocytoid (mean, 68%) and frequently (12%) demonstrated penetration by intact cells (emperipolesis) that appeared to be neutrophils (arrow). (G-H) Mouse data in the absence of G-CSF or AMD3100. (G) G6pc3^−/− mouse peripheral blood. Morphologically mature neutrophils were evident. (H) G6pc3^−/− mouse bone marrow. An abundance of mature neutrophils is evident.

Figure 4

Neutrophil CXCR4 expression in human and mouse G6PC3 deficiency. (A-B) Flow cytometry (FACS). FACS was performed on freshly isolated neutrophils from (A) P1 (brother) and P2 (sister) or (B) freshly isolated bone marrow cells gated for size and CD11b⁺Gr-1^high expression from G6pc3^−/− mice, and compared with patterns found in healthy human donors and wild-type mice. Results are shown as representative histograms for patient 2 and for a knockout mouse in the left panels and are summarized as the percentage of CXCR4⁺ cells and as the MFI in the right panels of A (human) and B (mouse). Human data are from a single experiment performed before the patients were administered G-CSF. Mouse data are the summary of 3 separate experiments with 2 knockout and 2 wild-type mice in each experiment. (C) Effect of G-CSF treatment on neutrophil CXCR4 and ANC. Freshly isolated peripheral blood neutrophils were analyzed by FACS before and after starting G-CSF. ANC and CXCR4 expression on each dose were measured once for each subject after approximately 2 months on that dose. Data are summarized as the average of both subjects for each parameter at each dose and are presented on the same axes as percentage of maximum MFI (where maximum is defined as the MFI of the pre–G-CSF sample; left axis) and as the average ANC (right axis). (D) Effect of acute CXCR4 blockade on mouse ANC. G6pc3 knockout and littermate control mice (G6pc3^+/+, n = 4 and G6pc3^+/−, n = 6) were injected subcutaneously with 200 μg of the specific CXCR4 antagonist AMD3100. Absolute blood neutrophil counts per microliter were determined immediately before and 2.5 hours after injection. Data are from 2 experiments and are presented as the mean ± SEM.

Figure 5

Increased neutrophil apoptosis in patients with G6PC3 deficiency resulting from homozygous inheritance of G6PC3 G260R. Neutrophils from both patients and 2 healthy donors (HD) were cultured at 37°C for 2 hours (left panels) or 4 hours (right panels) in medium alone or medium with the addition of tumor necrosis factor (TNFα) or thapsigargin. Flow cytometry was then performed. Apoptotic cells are defined as annexin V⁺. Dead cells are defined as propidium iodide–positive (PI)⁺. Data are from a single experiment performed after the patients were treated for 7 months with G-CSF.

Figure 6

Defective neutrophil superoxide production in patients with G6PC3 deficiency resulting from homozygous inheritance of G6PC3 G260R. (A) DHR assay. (Left panels) Neutrophils in orange by side scatter (SSC) and forward scatter (FSC) characteristics. Fluorescence of DHR-loaded neutrophils after PMA stimulation for each subject is graphed in the corresponding panels to the right. Prestimulation histograms for each subject showed MFIs less than 200. HD indicates healthy donor; and P1 and P2, affected brother and sister, respectively. The number in the top left corner of each panel on the right indicates the percentage of gated cells that failed to respond to PMA. (B) Real-time assay of superoxide production. PMA was used as the stimulus. Data are from a single experiment with 2 replicates at each time point plotted as a single curve for healthy donors versus patients. (C) Endpoint superoxide production assay. Production of superoxide was measured 15 minutes after the addition of the indicated agonists. Data are shown for a healthy donor versus the average of the 2 patients together. Data are summarized from a single experiment in which each subject was tested in duplicate. (B-C) Data were obtained after the patients were treated for 5 months with G-CSF.

Figure 7

Neutrophil functional capacity in G6PC3 deficiency. (A) Defective microbicidal activity in neutrophils from patients with G6PC3 deficiency resulting from homozygous inheritance of G6PC3 G260R. Staphylococcus aureus was incubated with polymorphonuclear leukocytes from a healthy donor (HD) and the 2 patients in an 8:1 target/effector ratio. Bacterial killing kinetics is plotted for each donor as a percentage of colony-forming units observed for control S aureus incubated in the absence of neutrophils. Data were obtained after the patients were treated for 5 months with G-CSF and are from a single experiment in which each condition was tested in duplicate. (B) Purified mouse bone marrow neutrophils from G6pc3 knockout and wild-type littermate control mice chemotax to CXCL12 equally well (P > .05 at all time points). Data are summarized from 2 experiments with 2 mice in each group.