Metabolic abnormalities in G6PC3-deficient human neutrophils result in severe functional defects

Abstract

Severe congenital neutropenia type 4 (SCN-4) is an autosomal recessive condition in which mutations in the G6PC3 gene encoding for the catalytic 3 subunit of glucose-6-phosphatase-β result in neutropenia, neutrophil dysfunction, and other syndromic features. We report a child with SCN-4 caused by compound heterozygous mutations in G6PC3, a previously identified missense mutation in exon 6 (c.758G>A[p.R235H]), and a novel missense mutation in exon 2 (c.325G>A[p.G109S]). The patient had recurrent bacterial infections, inflammatory bowel disease, neutropenia, and intermittent thrombocytopenia. Administration of granulocyte colony-stimulating factor (G-CSF) resolved the neutropenia and allowed for detailed evaluation of human neutrophil function. Random and directed migration by the patient's neutrophils was severely diminished. Associated with this were defects in CD11b expression and F-actin assembly. Bactericidal activity at bacteria/neutrophil ratios >1:1 was also diminished and was associated with attenuated ingestion. Superoxide anion generation was <25% of control values, but phox proteins appeared quantitatively normal. Extensive metabolomics analysis at steady state and upon incubation with stable isotope-labeled tracers (U-13C-glucose, 13C,15N-glutamine, and U-13C-fructose) demonstrated dramatic impairments in early glycolysis (hexose phosphate levels), hexosemonophosphate shunt (required for the generation of the NADPH), and the total adenylate pool, which could explain the dramatic cell dysfunction displayed by the patient's neutrophils. Preliminary experiments with fructose supplementation to bypass the enzyme block demonstrated that the metabolic profile could be reversed, but was not sustained long enough for functional improvement. In human deficiency of G6PC3, metabolic defects resulting from the enzyme deficiency account for diverse neutrophil functional defects and present a major risk of infection.

Graphical abstract

Figure 1.

Diagnostic micrographs of the patient’s bone marrow biopsy specimen and aspirates and genetic mutation. (A) Wright-Giemsa–stained low-power view of diagnostic patient bone marrow biopsy specimen before initiation of G-CSF administration showing hypercellularity with left-shifted myeloid maturation. Original magnification ×400. (B) Wright-Giemsa stained high-power view of bone marrow aspirate demonstrating vacuolization of myeloid precursors and pyknotic nuclei (2% myeloblasts, 12% promyelocytes, 26% myelocytes, 26% metamyelocytes, 8% bands, 9% segmented neutrophils). Original magnification ×1000. (C) High-power view of bone marrow aspirate after chronic G-CSF administration showing improved myeloid maturation (2% promyelocytes, 10% myelocytes, 11% metamyelocytes, 20% bands, 36% segmented neutrophils). Original magnification ×600. (D) Biallelic patient gene mutation showing compound heterozygous mutations in the G6PC3 gene. The paternal mutation at the terminus of exon 2 is a novel pathogenic variant that has not been described previously.

Figure 2.

Cell motility, CD11b expression, and F-actin content and assembly. (A) Chemotaxis measured using the under-agarose technique showing distance migrated for each stimulus (mean ± SEM) for the control and patient’s neutrophils, respectively, in 3 to 4 experiments, using independent neutrophil isolation. *P < .05, by 2-tailed Student t test,. (B) Chemotaxis measured using the transmembrane migration technique in response to fMLF. Fluorescence-labeled patient cells and control cell were detected as they crossed the fluorescence-blocking PET membrane into the lower wells in the system. Fluorescence moving into the lower wells is expressed as a percentage of the fluorescence detected in 10⁶ labeled cells and is plotted over time. Directed migration was decreased in the patient’s cells; nondirected migration with buffer was also decreased in the patient’s cells (data not shown). (C) Flow cytometry histogram showing a representative example of CD11b expression for control and patient neutrophils after incubation with buffer, PMA, and fMLF. (D) Cell surface levels of CD11b determined by flow cytometry and presented as mean log fluorescence after incubation with buffer, PMA, and fMLF for the control (red) and patient’s (blue) neutrophils. Bars and brackets represent the mean ± SEM of results in 3 to 5 experiments on independent neutrophil isolations. (E) F-actin assembly was measured as in “Methods” and plotted as mean log fluorescence for buffer, PMA, and fMLF for the control (red) and the patient (blue). The bars and brackets are ±SEM of results in 3 experiments using independent neutrophil isolations. (D-E) *P < .05, patient vs control in a given treatment condition, and #P < .05, given stimulus vs buffer, by 2-tailed Student t test.

Figure 3.

Bactericidal activity of control and patient neutrophils, generation of superoxide anion, and western blot of oxidase proteins. (A) Neutrophil-mediated killing of S aureus in the presence of normal human serum at a bacteria/neutrophil ratio of 1:1. Control (red) and patient (blue) results, plotted as viability at the sampling time points, are the average results of 2 experiments and cell isolations. (B) Bactericidal activity for control (red) and patient (blue) neutrophils at bacteria/neutrophil ratios >1:1 plotted as viability at the sampling time points. Error bars represent mean ± SEM of results in 3 experiments completed on separate cell isolations. *P < .05, by 2-tailed paired Student t test; #P < .05, by 1-tailed paired Student t test. (C) Representative superoxide anion generation assays for control and patient neutrophils in response to fMLF (left 2 panels) and PMA (right 2 panels), as described in “Methods”; assay results with addition of SOD are included. Luminescence as relative light units (RLU) is plotted over time. Patient neutrophils exhibit a marked decrease compared with control. (D) Superoxide anion generation in response to fMLF and PMA was measured as SOD-inhibitable luminescence, as described in “Methods.” Total luminescence for the assay is plotted for PMA and fMLF for control and patient. Bars and brackets represent mean ± SEM of results in 3 experiments performed on 3 separate occasions on independently isolated neutrophils. *P < .05 and **P < .005, by 2-tailed Student t test. (E) Western blots for p22phox, p67phox, p47phox, and gp91phox in patient and control neutrophils. The blots are representative of results obtained using neutrophils from 2 independent isolations. A second western blot (control 2) is shown for comparison purposes. Molecular weight markers are on the left of the blot, with phox protein designations on the right or in the middle.

Figure 4.

Metabolite levels in resting patient vs control neutrophils. (A) Metabolites decreased in the patient’s neutrophils relative to control cells. (B) Metabolites elevated in the patient’s neutrophils. Results of 3 separate measurements. Differences significant at P < .001. (C) Changes in adenylate pool (sum of ADP, AMP, adenosine, and adenine) between control and patient neutrophils. Data are plotted as means ± standard deviation. **P < .005.

Figure 5.

Schematic overview of [¹³C₆]glucose stable isotope tracings. (A) Tracing using [¹³C₆]glucose including glycolytic and HMS pathways. (B) Tracing using [¹³C₆]glucose in control and patient neutrophils before (time 0) and 5 minutes after incubation with 1 µM fMLF. Isotopologue intensities plotted correspond to labeling indicated in panel A (eg, ¹³C₆ for glucose).

Figure 6.

Schematic overview of [¹³C₅,¹⁵N₂]glutamine stable isotope tracings. (A) Tracing using [¹³C₅,¹⁵N₂]glutamine includes glutaminolytic and Krebs cycle pathways before (time 0) and 5 minutes after incubation with 1 µM fMLF. (B) Tracing using [¹³C₅,¹⁵N₂]glutamine in control and patient neutrophils before and after treatment with fMLF. The isotopologue intensities plotted correspond to labeling indicated in panel A (eg, ¹³C₅,¹⁵N₂ for Gln).

Figure 7.

Metabolic and function results in the presence of fructose incubation. (A) Levels of lactate, RP, and GSH in control and patient neutrophils following a 10-minute incubation with [U-¹³C]fructose; samples are before (time 0) and 5 minutes after incubation with 1 µM fMLF. (B) Superoxide anion production by neutrophils preincubated with or without 5 mM fructose and stimulated with PMA. SOD-inhibitable total luminescence is presented as described in “Methods.” (C) Chemotaxis in response to fMLF after preincubation with and without 5 mM fructose. Percentage of fluorescently labeled cells migrating across the polyethylene terephthalate (PET) membrane into the lower wells with fMLF stimulus over time is presented.