Female X-chromosome mosaicism for NOX2 deficiency presents unique inflammatory phenotype and improves outcome in polymicrobial sepsis

Abstract

Cellular X-chromosome mosaicism, which is unique to females, may be advantageous during pathophysiological challenges compared with the single X-chromosome machinery of males, and it may contribute to gender dimorphism in the inflammatory response. We tested the hypothesis of whether cellular mosaicism for the X-linked gp91phox (NOX2) deficiency, the catalytic component of the superoxide anion-generating NADPH oxidase complex, is advantageous during polymicrobial sepsis. Deficient, wild-type (WT), and heterozygous/mosaic mice were compared following polymicrobial sepsis initiated by cecal ligation and puncture. Compared with WT littermates, sepsis-induced mortality was improved in deficient mice, as well as in mosaic animals carrying both deficient and WT phagocyte subpopulations. In contrast, blood bacterial counts were greatest in deficient mice. Consistent with poor survival, WT mice also showed the most severe organ damage following sepsis. In mosaic animals, the deficient neutrophil subpopulations displayed increased organ recruitment and elevated CD11b membrane expression compared with WT neutrophil subpopulations within the same animal. The dynamics of sepsis-induced blood and organ cytokine content and WBC composition changes, including lymphocyte subsets in blood and bone marrow, showed differences among WT, deficient, and mosaic subjects, indicating that mosaic mice are not simply the average of the deficient and WT responses. Upon oxidative burst, interchange of oxidants between WT and deficient neutrophil subpopulations occurred in mosaic mice. This study suggests that mice mosaic for gp91phox expression have multiple advantages compared with WT and deficient mice during the septic course.

Fig 1. Mosaic neutrophil subpopulations in mice heterozygous for gp91phox deficiency

Part A: Bone marrow cell suspension or whole blood from mice heterozygous for gp91phox deficiency was processed according to the BD phospho-lyse-fix-Perm-III buffer protocol. Samples were incubated with anti-CD11b-PERCP as well as with either anti-gp91phox or the corresponding isotype-IgG followed by incubation with PE-conjugated secondary antibody as described in the materials and methods section. As shown, the well-defined CD11b positive cell population (left panels) separated into two cell populations by the level of gp91phox expression (middle panels). The histograms on the right indicate that the ratio of Xp and Xm expressing mosaic neutrophil subpopulations was approximately one. Part B: Bone marrow (left panels) or WBC (right panels) was preincubated with DHR for 20 min followed by the incubation with 1uM phorbol–myristate-acetate (PMA) for 15 min. CD11b positive myeloid cells were gated and analyzed for DHR fluorescence in deficient (top row), WT (middle row) and heterozygous mosaic animals (bottom row). As shown, in WT animals PMA resulted in a marked response whereas in deficient samples there was no increase in DHR fluorescence. In mosaic animals, the presence of two-cell population was evident and the oxidative burst response of mosaic subpopulations corresponded well with the deficient and WT responses, respectively. Representative findings from several experiments with similar observations are shown.

Fig 2. NOX2 mosaicism increases survival and improves bacterial clearance in polymicrobial sepsis

Part A: WT, gp91phox deficient and gp91phox mosaic animals were made septic by cecal ligation and puncture (CLP). Animals received fluid resuscitation postoperatively and then repeatedly at every 24h and were observed for mortality. Mosaic and deficient animals showed improved survival as compared to WT. *Statistically significant difference compared to WT (Long Rank Test, n=32–53 in individual groups, as indicated). Part B: In a separate set of experiments, animals were subjected to CLP and resuscitation and 24h later bacterial counts in blood were determined. Deficient animals showed the greatest bacterial counts following sepsis. *Statistically significant difference compared to mosaic animals (Mean ± S.E.M., n=12–17 in each group).

Fig 3. Sepsis-induced cell composition changes in blood and BM

From control or CLP-subjected mice, the total numbers of major WBC subtypes as well as percent distribution of cells were determined. Bars represent absolute cell numbers calculated from cell counts and percent distributions, whereas numbers within bars depict values as percent of total cells. BM lymphoid line was identified by double positive staining for CD45 and CD19, neutrophils for CD45, CD11b whereas macrophages by triple staining for CD45, CD11b and CD115. Cell composition in circulating blood was determined by CD11b straining for neutrophils and CD19 for B-cells. Major distribution of helper and cytotoxic T cells were determined by CD3/CD4 or CD3/CD8 dual staining, respectively. *Statistically significant difference between septic and control within the same genotype. Mean ± S.E.M., n=7–8 animals in each group.

Fig 4. Sepsis skews neutrophil ratios towards deficient subpopulation in blood and spleen in mosaic subjects

Blood was sampled 2 days before CLP and tested for WT/deficient-ratio of CD11b neutrophils by gp91phox expression or lack of. Subsequently, blood and spleen was harvested from the same animals 24h pot-CLP and the analyses repeated. Part A shows a typical flow analysis from an animal. Part B indicates the WT/Deficient-ratios of individual animals. (Arrows connect values derived from the same mouse). Sepsis caused a statistically significant decrease in the WT/deficient ratio of blood neutrophils (p=0.004, Pre-CLP blood versus Post-CLP blood). After CLP, the WT/Deficient-ratios in blood and spleen were similar in the same subject after CLP (Post-CLP Blood versus Post-CLP spleen). Because splenocytes cannot be obtained for analyses before CLP we also depict WT/Deficient neutrophil ratios in spleen and blood from an independent sample of controls (Unrelated subjects).

Fig 5. Increased neutrophil tissue infiltration following sepsis in mosaic animals

Myeloperoxidase (MPO) activity in organs represents a marker for phagocyte infiltration. 24h after CLP animals were sacrificed and MPO activity in lung, spleen, liver and kidney homogenates were determined as described in the materials and methods section. *Statistically significant difference (p<0.05) compared to group as indicated by lines. Mean ± S.E.M., n=6–8 animals in each group.

Fig 6. Elevated neutrophil CD11b expression in the deficient subpopulations of mosaic animals

Myeloid cells from bone marrow of WT and deficient animals (A) or mosaic mice (B) were stained for CD11b under PMA-stimulated conditions (AB). Deficient neutrophils or deficient subsets from mosaics showed elevated CD11b expression as compared to WT. In a separate set of experiments neutrophil CD11b membrane expression was also determined in naïve and septic animals from BM, spleen and blood. *Statistically significant difference (p<0.05) as compared to WT. Mean ± S.E.M., n=6–7 animals in each group.

Fig 7. Cytokine levels indicate an altered inflammatory course in mosaic animals as compared to either WT or deficient mice

Blood was collected at 6h and 24h after CLP and also from lung and spleen 24h after CLP. Cytokine content was determined using ELISA as described in the materials and methods section. Mean ± S.E.M., n=5–8 animals in each group and time. *Statistically significant difference (p<0.05) as indicated by lines.

Fig 8. Worsened tissue dysfunction in WT subjects following sepsis

Blood was collected 24h after CLP and markers of organ dysfunction were measured as described in the material and methods section. Mean ± S.E.M., n=6–8 animals of each genotype in the septic group and n=3 in controls. Control levels are shown with the dotted line with numeric values in brackets. *Statistically significant difference as compared to control (p<0.1).

Fig 9. Cross talk through ROS between mosaic neutrophil subpopulations

Part A indicates basal DHR staining in resting neutrophils from WT, deficient and mosaic animals after the administration of vehicle. Part B depicts PMA-induced responses in WT cells alone (marked increase), deficient cells alone (no response) and also in the “artificially mixed” 1:1-ratio of the same deficient and WT cells. PMA stimulation of “artificially mixed” sample resulted in a detectable right shift in fluorescence in deficient cells indicating that oxidants released from WT cells reached the intracellular site of deficient cells. Part C compares the response of a “naturally mixed” mosaic specimen to that of WT and deficient cells, respectively. The right shift by the deficient subpopulation in “naturally mixed” mosaic samples was similar to that observed in the “artificially mixed” sample (Compare parts B and C). The histograms are representative findings from several experiments with similar observations.