Combined ascorbate and glutathione deficiency leads to decreased cytochrome b5 expression and impaired reduction of sulfamethoxazole hydroxylamine

Abstract

Sulfonamide antimicrobials such as sulfamethoxazole (SMX) have been associated with drug hypersensitivity reactions, particularly in patients with AIDS. A reactive oxidative metabolite, sulfamethoxazole-nitroso (SMX-NO), forms drug-tissue adducts that elicit a T-cell response. Antioxidants such as ascorbic acid (AA) and glutathione (GSH) reduce SMX-NO to the less reactive hydroxylamine metabolite (SMX-HA), which is further reduced to the non-immunogenic parent compound by cytochrome b (5) (b5) and its reductase (b5R). We hypothesized that deficiencies in AA and GSH would enhance drug-tissue adduct formation and immunogenicity toward SMX-NO and that these antioxidant deficiencies might also impair the activity of the b5/b5R pathway. We tested these hypotheses in guinea pigs fed either a normal or AA-restricted diet, followed by buthionine sulfoximine treatment (250 mg/kg SC daily, or vehicle); and SMX-NO (1 mg/kg IP 4 days per week, or vehicle), for 2 weeks. Guinea pigs did not show any biochemical or histopathologic evidence of SMX-NO-related toxicity. Combined AA and GSH deficiency in this model did not significantly increase tissue-drug adduct formation, or splenocyte proliferation in response to SMX-NO. However, combined antioxidant deficiency was associated with decreased mRNA and protein expression of cytochrome b (5), as well as significant decreases in SMX-HA reduction in SMX-NO-treated pigs. These results suggest that SMX-HA detoxification may be down-regulated in combined AA and GSH deficiency. This mechanism could contribute to the higher risk of SMX hypersensitivity in patients with AIDS with antioxidant depletion.

Fig 1

Biotransformation of sulfamethoxazole (SMX) in humans. The parent drug is predominantly cleared by N-acetylation, leading to a non-immunogenic product. SMX is alternatively oxidized to SMX-HA, which spontaneously generates the nitroso metabolite (SMX-NO). SMX-NO forms adducts with tissue proteins that can lead to T cell activation. Ascorbate, and the thiols glutathione or cysteine, can non-enzymatically reduce SMX-NO to the hydroxylamine. SMX-HA is further detoxified by cytochrome b₅ and its reductase (Kurian et al. 2004).

Fig 2

Summary of the different treatment groups used in the present experiments. The following treatment groups were compared for different end points: to establish a model of antioxidant depletion, group 1 vs. 2, and group 3 vs. 4 (ANOVA with Bonferroni's post hoc test); to compare drug adducts and T cell proliferation, group 3 vs. 4 (unpaired t test); to evaluate clinical toxicity (e.g. serum ALT activity): groups 1, 2, 3, and 4 (ANOVA with Bonferroni's post hoc test); and to evaluate the effect of antioxidant depletion on b5 and b5R expression, groups 1 vs. 2 (unpaired t test).

Fig 3

Scatter plot of hepatic ascorbic acid (AA) concentrations in guinea pigs provided replete or deficient levels of AA intake for 4 weeks. For the last two weeks of the study, deficient pigs were also treated with buthionine sulfoximine (BSO; 250 mg/kg/day SC) to inhibit GSH synthesis. Each data point represents one pig at week 4; horizontal lines indicate the group means. * Indicates significantly different from both replete groups (P < 0.001).

Fig 4

Hepatic and red blood cell (RBC) GSH concentrations in guinea pigs depleted of both GSH and AA. Upper panel: Guinea pigs with AA restriction plus BSO treatment (groups 2 and 4) had significantly decreased hepatic GSH concentrations at week 4 compared to replete pigs. * P < 0.05 compared to replete pigs; ** P < 0.01 compared to replete pigs given SMX-NO. Lower panel: Pigs with antioxidant depletion also had significantly decreased RBC GSH concentrations compared to replete pigs. * P < 0.001 compared to replete pigs; ** P < 0.01 compared to replete pigs given SMX-NO.

Fig 5

Quantification of SMX-tissue adducts in the (A) serum (175 kD band represented), (B) peritoneum, and (C) spleen of AA and GSH replete (group 3) and deficient (group 4) guinea pigs treated with SMX-NO. Densitometry readings were normalized by subtracting readings obtained with rabbit pre-immune sera. No significant differences in adduct formation could be demonstrated.

Fig 5

Quantification of SMX-tissue adducts in the (A) serum (175 kD band represented), (B) peritoneum, and (C) spleen of AA and GSH replete (group 3) and deficient (group 4) guinea pigs treated with SMX-NO. Densitometry readings were normalized by subtracting readings obtained with rabbit pre-immune sera. No significant differences in adduct formation could be demonstrated.

Fig 6

Proliferative response of splenocytes from AA and GSH deficient (group 4) and replete (group 3) guinea pigs treated with SMX-NO. Cells were cultured with SMX-NO (100 μM) in vitro for 72 h; proliferation was measured by incorporation of [³H] thymidine. Data are reported as stimulation index relative to incubation with media alone in vitro. Results represent the mean from four animals in each group, with incubations carried out in triplicate. No response was seen to SMX (1 mM in vitro) over vehicle controls in any pig (last 2 columns on graph).

Fig 7

Cytochrome b₅ (b5) and b₅ reductase (b5R) protein expression in liver microsomes of guinea pigs, measured using immunoblotting. All densitometry readings were normalized to ß-actin as a loading control. (Upper panel): b5 expression was significantly decreased in combined AA and GSH deficient guinea pigs (group 2) compared to anti-oxidant replete guinea pigs (group 1; P = 0.01). (Lower panel): Trend toward reduction in b5R expression in deficient guinea pig liver microsomes (group 2) compared to replete guinea pigs (group 1; P = 0.10).

Fig 8

SMX-hydroxylamine (SMX-HA) reduction activities in guinea pig liver microsomes, after 4 weeks of differential ascorbate intake and two weeks of treatment with BSO and SMX-NO (or respective vehicles). (Upper panel): Hepatic SMX-HA reduction was significantly decreased in antioxidant deficient versus replete pigs treated with SMX-NO (* P = 0.029 versus replete/SMX-NO group). (Lower panel): Hepatic glutathione concentrations at week 4 were modestly but significantly correlated with reduction of (SMX-HA) to SMX in guinea pig liver microsomes. All 4 groups of guinea pigs are included (n = 18).