No contribution of GSTM1 and GSTT1 null genotypes to the risk of neutropenia due to benzene exposure in Southeastern Brazil

Exposure to benzene has been associated with haematological diseases such as neutropenia (NEB) and acute myeloid leukaemia (AML). We tested whether the null genotypes of the GSTM1 and GSTT1 genes, involved in benzene inactivation, altered the risk for NEB in southeastern Brazil. Genomic DNA from 55 NEB patients and 330 controls was analysed by multiplex-polymerase chain reaction. The frequency of the GSTM1, GSTT1 and combined null genotypes was similar in patients and controls (GSTM1, 27.3% vs. 38.8%, p = 0.16; GSTT1, 25.5% vs. 19.7%, p = 0.24; GSTM1/GSTT1, 12.7% vs. 6.7%, p = 0.26; respectively). The distribution of genotype classes in NEB patients was similar to normal controls, suggesting that GSTM1 and GSTT1 null genotypes make no specific contribution to the risk of NEB. As the GSTM1 and GSTT1 null genotypes were previously associated with increased risk for AML in Brazil and elsewhere, we hypothesise that different thresholds of chemical exposure relative to distinct GSTM1 and GSTT1 genotypes may determine whether AML or NEB manifests in benzene exposed individuals from southeastern Brazil. Although indicative, our results still require support by prospective and large scale epidemiological studies, with rigorous assessment of daily chemical exposures and control of the possible contribution of other polymorphic genes involved in benzene metabolism.

The past decade has seen an increasing interest in associations of haematological diseases and occupational exposure to chemical agents. In this context, neutropenia due to exposure to benzene (NEB) has generally been characterised by slight to moderate decrease in neutrophil count (Lorand et al., 1984;Ruiz et al., 1994;Queiroz et al., 1997;Augusto et al., 1999). In contrast, a marked reduction in erythrocyte, neutrophil and platelet counts has been seen in acute myeloid leukaemia (AML), a more severe form of occupational disease (Vaughan et al., 2005). Glutathione Stransferases (GST) modulate the effects of exposure to chemical agents linked to NEB and AML (Hayes et al., 2005). The GSTM1 and GSTT1 genes have null variant genotypes, which have been associated with AML susceptibility (Ye and Song, 2005).
Occupation-related diseases have been described as a serious health problem in southeastern Brazil (Lorand et al., 1984;Ruiz et al., 1994;Queiroz et al., 1997;Augusto et al., 1999), and associations of the GST null genotypes and increased risks for AML were previously reported by us in individuals from this area of the country (Arruda et al., 2001). Moreover, to the best of our knowledge, no studies regarding the associations of the GSTM1 and GSTT1 genotypes and NEB risk have been reported. For these reasons, the identification of genotypes of the GSTM1 and GST1 genes in NEB patients from southeastern Brazil was considered necessary in order to investigate their influence, if any, as a risk factor for NEB.
We analysed 55 consecutive NEB patients (34 men, 21 women; 29 Caucasians, 26 African-Brazilians; mean age 35 ± 13 years) seen at the hospital of the State University of Campinas, from January 2002 to December 2005. Neutropenia was defined as a neutrophil count below 1.8 x 10 9 /L for Caucasians and 1.5x10 9 /L for African-Brazilians (Watts, 1999;Dale, 2001). NEB patients presented a consistent history of daily benzene occupational exposure (they were painters, mechanics, shoemakers, or workers of construction companies, petrochemical industry and petrol stations) for a period of over 12 months. Hypocellularity of the granulocytic lineage in bone marrow and exclusion of other causes of neutropenia, such as infectious, autoimmune, haematological, thyroid and nutritional deficiency or drug-related neutropenia, as previously recommended by our group (Lima et al., 2006), were also required for NEB diagnosis. The control group consisted of 330 blood donors (247 men, 83 women; 174 Caucasians, 156 African-Brazilians; mean age 51 ± 3 years) without a consistent history of benzene exposure. They were recruited from the same university hospital in order to provide a representative group of the general population that seeks medical assistance in the region. All procedures were carried out according to the principles of the institutional guidelines and all patients and controls provided written informed consent.
Genomic DNA from peripheral blood of patients and controls was analysed by multiplex-polymerase chain reaction for identification of GST genotypes (Arruda et al., 2001). The GST genotypes were analysed after electrophoresis on 2.0% agarose gels (Figure 1), using the b-globin gene as internal control.
Differences between groups were analysed by means of chi-squared or Fishers exact tests. For analysing the associations with NEB, univariate and multivariate analyses were used throughout, in order to obtain odds ratio (OR), adjusted or not for age, gender and ethnic origin, and their corresponding 95% confidence intervals (CI). The statistical package Epi Info was used to perform all these analyses.
The frequencies of the GSTM1, GSTT1 and combined GSTM1 and GSTT1 null genotypes were similar in patients and controls. Patients with the distinct genotypes of the GSTM1 and GSTT1 genes exhibited similar distribution to normal controls, suggesting that GST genotypes make no significant contribution to NEB, under the chemical exposures encountered in this study (Table 1).
In Brazil, workers are exposed predominantly to solvents, such as benzopyrene, hexachlorobenzene, ethylene oxide, dichloromethane and epoxybutanes, which are metabolised by the GSTM1 and GSTT1 enzymes (Ruiz et al., 1994;Queiroz et al., 1997;Hayes et al., 2005). The chemicals have been consistently associated with both AML and NEB in Brazil (Lorand et al., 1984;Ruiz et al., 1994;Queiroz et al., 1997;Augusto et al., 1999) and in other parts of the world (Cronkite et al., 1989;Ye and Song, 2005;Vaughan et al., 2005). These data supported the association of both GST null genotypes with increased risks for AML previously found by our group (Arruda et al., 2001). On this basis, the GST null genotypes were also expected to be associated with increased NEB risk. Surprisingly we found similar frequencies of the GST genotypes in our NEB patients and controls.
Unfortunately, there was no available data concerning the levels of benzene exposure of the NEB patients in this study and of the AML patients in our previous study (Arruda et al., 2001). We assumed similar exposures to benzene for patients in either group. Taking these results together, we hypothesise that different thresholds of chemical exposure relative to distinct GST genotypes may determine whether NEB or AML arises in chemical hazard exposed individuals from southeastern Brazil. Thus, those highly exposed to chemicals and homozygous for the null GST alleles may develop AML, since this seems to be more dependent on the GST pathway of carcinogen metabolism, whilst those individuals with less exposition to chemicals may be less dependent on carcinogen inactivation by the GST isoenzymes, and therefore more prone to present the benign form of occupational disease, NEB, without mediation of GST genotypes. Although indicative, these results must, however, be confirmed by prospective studies with 710 Lima et al.  larger samples of NEB and AML patients and controls, with rigorous assessment of daily chemical exposures, and control of the influence of other polymorphic genes involved in benzene metabolism (Aydin-Sayitoglu et al., 2006).