Main

The common feature of AA is an empty bone marrow and pancytopenia in the peripheral blood (1, 2). Although the causes for aplastic anemia are multiple, with the majority of cases remaining idiopathic, the pathogenic mechanism leading to the disease might be limited to two main pathways: environmental factors (3, 4) and genetic susceptibility (5, 6). Environmental factors include viral infections (7, 8), exposure to radiation (9), drugs (10), and chemicals (11–13). They may cause DNA damage on the level of the hematopoietic progenitors (14) or lead to an activation of autologous immunologic mechanism (15), also resulting in damage on the level of the hematopoietic progenitors (16–18). However, a history of a massive exposure to environmental pollutants will rarely be described in the history of a Caucasian child suffering from AA. Individuals vary in their ability to metabolize several DNA-damaging agents because of polymorphism of detoxifying enzymes. Variability in the expression or efficiency of these enzymes could explain interindividual risks for development of aplastic anemia in the presence of the above-mentioned environmental agents. Three of the most important biotransforming enzymes, CYP isoenzymes, mEh, and GST, show genetic polymorphisms. p450-cyp2E1 metabolizes a wide variety of drugs and chemicals, with preference for small molecules such as aldehydes, ketones, alcohols, and aromatic compounds (19). A number of polymorphisms have been described in the p450-cyp2E1 gene (20), with the Pst I/ Rsa I polymorphism in the 5′ flanking region being best defined (21). The C2/C2 variant leads to an enhanced transcription of the (C2) p450-cyp2E1 gene, which increases enzyme activity compared with the C1/C1 variant (21).

Among the phase II enzymes, mEh is involved in the first-pass metabolism of highly reactive epoxide intermediates and oxygen radicals. Two polymorphisms of mEH with a broad substrate spectrum (22, 23) have been found to be associated with susceptibility for several malignant diseases (24–26). The exon 3 point mutation T to C results in a tyrosine to histidine shift at position 113 (His 113), with a 50% reduced enzyme activity (slow allele), and the A to G transition leading to a histidine to arginine shift at position 139 (Arg 139), with a 25% faster enzyme activity (fast allele) (27, 28).

GST represent a group of xenobiotic detoxifying enzymes (29, 30), with GSTM (31–34), -T (35, 36), and -P (37) being polymorphs. A null genotype is characterized by the absence of the respective GST protein (38). The GSTM1 null genotype has been shown to be associated with an increased risk of smoke-induced cancers (39–41), breast cancers (42), and hepatocellular carcinoma (43, 44). The GSTT1 null genotype has been shown to be associated with an increased risk for MDS (44–46). However, further studies on MDS have revealed conflicting results (44, 47–49).

In this study, we analyzed three different enzymes in pediatric AA patients. p450-cyp2E1, mEh, and GST (M1 and T1) were chosen with regard to their substrate specificity, including hematotoxic and mutagenic DNA/damaging agents as aromatic compounds, radicals, and reactive oxygen.

PATIENTS AND METHODS

Patients.

The majority of patients suffering from AA were recruited between 1996 and 1999 through the pediatric cooperative multicenter study SAA-94, involving 37 centers in Germany and Austria. Some were tested independently from the study, as their diagnosis was before the beginning of the study. The diagnosis for AA was performed following the Camitta criteria and confirmed by bone marrow biopsy. Most of the patients suffered from VSAA (polymorphonuclear cells <200/μL) or SAA (polymorphonuclear cells <500/μL). Three patients had marked chromosomal abnormalities (+19, −7, +8) without morphologic signs of acute myeloid leukemia (AML) or MDS. Patients were ineligible if they showed clinical signs of paroxysmal nocturnal hematuria, Fanconi anemia, MDS, or congenital bone marrow failure disorders, or had undergone radiation, chemotherapy, or pretreatment with antilymphocyte globulin, cyclosporine A, or granulocyte colony-stimulating factor. The patients' ages ranged between 1 and 16 y, with a median age of 9.6 y; 38 girls and 53 boys were analyzed. Details concerning the patients of the SAA-94 study, the treatment, and results are described by Fuhrer et al.(50). The research protocol was approved by the ethics committee of the Ludwig-Maximilians University, Munich, and the Heinrich Heine University, Duesseldorf. Nonrandom controls were obtained from pediatric stem cell donors: four were siblings of patients suffering from solid tumors (one neuroblastoma, two non-Hodgkin's lymphoma, and one Hodgkin's lymphoma), four were siblings of patients suffering from X-linked severe combined immunodeficiency, one was a sibling of a patient suffering from adrenoleukodystropy, and six were siblings of patients diagnosed for relapsed acute leukemia. There was no increased incidence of cancer in the medical history of the families. The other donors were healthy Caucasian volunteers, whose samples were obtained from the blood bank. As all results obtained from controls' samples fit perfectly to the results from larger patient collectives, we believe that they are representative. The study has been approved by the ethics committee of the University of Munich as part of the SAA-94 study and by the ethics committee of the Heinrich Heine University, Duesseldorf. Informed consent was signed by the parents or the patients.

DNA isolation.

DNA was prepared from whole blood or bone marrow specimens using the DNAeasy kit (QIAGEN, Hilden, Germany).

p450-cyp2E1 polymorphism.

p450-cyp2E1 alleles were determined in 91 AA patients and 235 controls by their melting curve differences in a light cycler (Roche Molecular Biochemicals, Mannheim, Germany). Some controls were analyzed parallel to patients' samples. Other control data were established in the same laboratory by the same method in a separate research project. Approximately 100 ng of genomic DNA were used for each PCR reaction. The following primers were used: forward primer: 5′ CCCGTGAGCCAGTCGAGT 3′; reverse primer: 5′ CAGACCCTCTTCCACCTTCTAT 3′ (5 μM each); Rsa I sensor: 5′ CATAAAGATTCATTGTTAATATAAAAGTACAAAATTX 3′ (1.5 μM); Rsa anchor: 5′-LCRes 540 CAACCTATGAATTAAGAACTTATATATATTGCCAG ph-3′ (1.5 μM). The PCR reaction contained 1 μL of DNA, 1 μL of each primer, 1.2 μL 25 mM MgCl2, 1 μL of the manufacturer's PCR mix [including Taq polymerase and nucleoside 5′-triphosphate (NTP); Roche Molecular Biochemicals]. PCR was performed in 35 cycles at 45°C and 72°C and the melting curves were generated by the equipment.

mEh polymorphisms.

The two distinct mEh polymorphisms were analyzed in 75 patients and 96 controls by the PCR/restriction length polymorphism (RFLP) technique. Due to technical problems and, thus, high consumption of DNA, not all 91 probes could be analyzed. For amplification of the polymorphic exon 3, 1 μM of the primers 5′ CCTTGTGCTCTGTCCTTCCCATCCC and 3′ AATCCTAGTCTTGAAGTGACGGT +, 0.15 μM dNTP, 0.15 μM Taq polymerase, 1 mM 10× buffer, and 5 μL Q - solution (QIAGEN) were added to 2.5 μM DNA. Thirty-five PCR cycles were run for 10 s at 94°C, 30 s at 45°C, and 45 s at 72°C. The PCR products were digested with Asp I for 1 h at 37°C. A restriction enzyme site is present in the wild type only, giving a 209-bp fragment, whereas the mutant is not digested and shows 231-bp band. The exon 4 polymorphism site was amplified with 5′ GGGGTGCCAGAGCCTGACCGT and 3′ AACACGGGGCCCACCCTTGGC as described and digested with Rsa I for 1 h at 37°C. The enzyme cuts the mutant allele, giving a 174-bp fragment, whereas the wild type shows a larger 295-bp band. The characteristic restriction products for exon 3 and exon 4 polymorphisms are depicted in Figure 1.

Figure 1
figure 1

Genotype analysis of mEh. Exon 3 (left panel): (a) fragment patterns expected in individuals homozygous or heterozygous for mEh polymorphism. The restriction digest using Asp I reveals 209-bp and 20-bp fragments in homozygous wild-type individuals and a 231-bp fragment in homozygous mutant individuals. Heterozygous individuals display all bands. (b) Typical electrophoresis gel showing the mEh-specific DNA fragments. Only the relevant 231- and 209-bp fragments are shown. Exon 4 (right panel): (a) fragment patterns expected in individuals homozygous or heterozygous for mEh polymorphism. The restriction digestion using Rsa I reveals a 295-bp fragment in homozygous wild-type individuals and 174-bp and 62-bp fragments in homozygous mutant individuals. Heterozygous individuals display all bands. (b) Typical electrophoresis gel envisioning the mEH-specific DNA fragments. Only the relevant 295- and 174-bp fragments are shown.

Full size image

GSTM1 and GSTT1 polymorphisms.

To analyze the polymorphism of GSTM1 and -T1 genes, we developed a multiplex PCR detecting GSTM1 and -T1 as well as an internal β-globin control in a one-tube reaction. Samples from 78 patients and 122 controls were analyzed. Due to technical problems and high consumption of DNA, only 78 of 91 probes were analyzed. Briefly, GSTM1 was amplified using primers corresponding to the 3′-coding region of the human GST: 5′ GAACTCCCTGAAAAGCTAAAG and 5′ GTTGGGAAATATACGGTG. Primers coding for the human GSTT1 were 5′ TTCCTTACTGGTCCTCACATCTC and 5′ TCACCGGATCATGGCCAGCA. The housekeeping gene β-globin was amplified using the primers 5′ CAACTTCATCCACGTTCACC and 5′ GAAGAGCCAAGGACAGGTAC. The PCR reaction was carried out by heating to 94°C for 4 min, followed by 34 cycles of 1 min at 94°C, 60°C, and 72°C. A final run at 72°C for 7 min was included. To exclude systematic errors, 14 samples were analyzed repetitively. An example is given in Figure 2.

Figure 2
figure 2

Genotype analysis of GSTM1 and GSTT1. The frequency of GSTT1 and GSTM1 null genotype was analyzed in pediatric patients with AA and in healthy controls. The results of a multiplex PCR amplifying GSTM1, GSTT1 and the β-globin housekeeping gene in a one-tube reaction are shown. The 270-bp lane shows the β-globin control, the ??-bp lane the GSTT1 gene, and the ??-bp lane the GSTM1 gene. Lane 1 shows the DNA size-standard; in lane 2, the β-globin product and the GSTM1 gene is shown (GSTT1 could not be detected). In lane 3, only β-globin was detected and both GSTT1 and GSTM1 were lacking. In lane 4, the β-globin and GSTT1 gene are shown, and GSTM1 is absent. In lane 5, both analyzed GST genes are detected.

Full size image

Sensitivity of the PCR reaction was 8.6 ng/μL DNA, as tested by stepwise dilution of GSTT1- and GSTM1-positive DNA. The PCR reaction was performed in the presence of 1 ng β-globin DNA as a standard control.

Statistics.

Statistical analyses were performed using SPSS statistical software (version 7.5, SPSS Inc., Chicago, IL, U.S.A.). Significance was measured by Fisher's exact test (two-tailed), χ2 test (two-tailed), and the Bonferoni method. Odds ratios (OR) were calculated and given within the 95% confidence interval (CI). Unconditional multivariate analysis was performed separately for each genotype. Age, gender, severity of acquired aplastic anemia, and therapeutic response were analyzed as co-variables for each locus and in combinations. Statistical analyses were performed with the help of the Institute for Statistics at the Heinrich Heine University, Duesseldorf.

RESULTS

p450-cyp2E1 polymorphism.

The frequency of the p450-cyp2E1 alleles is given in Table 1. The prevalence for homozygosity of the rarer C2 was comparable in AA, with 0 of 91 (0%) analyzed samples and 1 of 235 (0.4%) controls with (OR, 1.6; 95% CI, 0.7–4.1; p > 0.28; NS). Heterozygous C1/C2 were found in 8 of 91 (9%) patients with AA and 12 of 235 (5.1%) controls (NS). The wild-type C1/C1 alleles were comparable in 83 of 91 (91%) patients with AA and 222 of 235 (94.5%) controls (NS). The respective prevalence in healthy controls was comparable to data from other studies that analyzed Caucasian patients (20, 21, 43). These data exhibit no association of p450-cyp2E1 polymorphism with a risk for AA in children.

Table 1 Distribution of p450-cyp2E1 polymorphism in children suffering from acquired aplastic anemia
Full size table

mEH polymorphism.

The frequency of mEh polymorphism is given in Table 2. The prevalence of the respective alleles in our controls was comparable to other studies (49). Forty-one of 75 (55%) patients showed the tyrosine (Tyr) 113 polymorphism in exon 3 and 52 of 96 controls (53%) showed the Tyr 113 polymorphism in exon 3 (NS). The slow allele [histidine (His) 113] variant of exon 3 was found in 15 of 75 (20%) patient samples and in 15 of 96 (16%) control samples (NS). The slow allele (His 139) variant of exon 4 was found in 47 of 75 (63%) patient samples and in 65 of 96 (68%) control samples. The fast allele [arginine (Arg) 139] of exon 4 was found in 4 of 75 (5%) patient samples and in 2 of 96 (2%) control samples (NS). The respective prevalence for each analyzed allele in healthy controls was comparable to data from other studies that analyzed Caucasian patients (45, 49, 51). Transfer of the genotypes into putative phenotypes revealed no significant difference between the two alleles (Table 3).

Table 2 Distribution of mEH exon 3 and exon 4 polymorphisms in children with acquired aplastic anemia
Full size table
Table 3 Putative functional distribution of mEh in children suffering from acquired aplastic anemia
Full size table

GST polymorphism.

The frequencies of GSTT1 and GSTM1 alleles for all patients and their respective controls are given in Table 4. The incidence of GSTT1 was significantly increased in AA cases compared with controls. Among the patients, 34 of 78 (39%) children were negative for the GSTT1 gene compared with 24 of 122 (20%) controls. We found a significantly increased risk for AA in children with a GSTT1 null genotype (OR, 2.8, 95% CI, 1.5–5.7; p < 0.004).

Table 4 GSTM1 and GSTT1 genotypes and their estimated association with the risk for developing acquired aplastic anemia
Full size table

The GSTM1 null genotype occurred in 34 of 78 (44%) patients and in 66 of 122 (54%) controls (NS). The GSTM1 null genotype was not associated with an increased risk for AA. Nine of 78 patients (12%) and 3 of 122 controls (8%) were found to be negative for both GST alleles (NS). The prevalence in healthy controls was comparable to data from other studies that analyzed Caucasian patients.

Interaction among the genetic polymorphisms.

To assess the combined effects of these polymorphisms, interactions between the GSTT1 null genotype and the other polymorphisms were analyzed by multivariate analysis. This analysis did not show a significant increase in the risk for developing AA for either of the tested interactions with the GSTT1 null genotype (data not shown).

Variables such as age or gender, the phenotype of the aplastic anemia (NSAA, SAA, and VSAA), response to immunosuppressive therapy, frequency of relapse, duration of the disease (data not shown), and development of a malignant disease (data not shown) were tested for correlation with either of the analyzed genotypes. However, none of these parameters showed a significant correlation.

DISCUSSION

Immunologic mechanisms and/or oxidative DNA damage are widespread means of induction in the pathogenesis of AA (41). Bone marrow protection from xenobiotics depends on an intact detoxification pathway. Individual differences in expression of biotransforming enzymes may lead to susceptibility to toxic agents. We analyzed genetic polymorphisms of biotransforming enzymes and correlated the data with the risk for AA. We found a significantly higher frequency of the GSTT1 null genotype among children with AA, whereas p450-cyp2E1, mEh, and GSTM1 were not correlated. However, the statistics show that GST may play a secondary role in the pathogenesis of AA in childhood. Viewing the literature, many factors have been associated with aplastic anemia, and the GSTT1 null genotype may be one risk factor among others. GSTT1 determines the ability to conjugate reactive components (44) such as monohalomethanes (52) and ethylene oxide (53, 54) and metabolism of low-molecular-weight-halogenated xenobiotics or reactive epoxides. GSTT1 is expressed in erythrocytes and lymphocytes (51, 52) and hence acts in the hematopoietic system. The importance of GSTT1 in the protection of hematopoietic cells from environmental pollutants has been proven in a population exposed to 1,3-butadiene—the in vitro sister chromatid exchange in human lymphocytes in the presence of 1,3-butadiene was 16-fold higher in cells from individuals lacking the GSTT1 gene expression (55, 56). With regard to benzene detoxification, which is highly important in context with AA, GSTT1 is a major protection factor of individual susceptibility to benzene-induced chromosomal damage (51) and induces the urinary excretion of benzene metabolites with higher excretion of t,tMA (57).

Proving a direct causative coincidence between exogenous pollutants and marrow failure is difficult because the onset of marrow failure cannot be dated accurately. The delay in marrow disease after exposure to pollutants may be variable, and individuals with a genetic susceptibility due to an enzyme deficiency may be at greater risk than individuals with normal enzyme levels in the presence of environmental factors, even at low exposures. The GSTT1 null genotype may modulate the metabolism of exogenous pollutants or endogenous DNA-damaging processes. A genetic susceptibility toward certain pollutants might determine the individual risk against common levels of mutagenic pollutants.