FPGS rs1544105 polymorphism is associated with treatment outcome in pediatric B-cell precursor acute lymphoblastic leukemia

Background Folypolyglutamate synthase (FPGS) catalyzes the polyglutamation of folates and antifolates, such as methotrexate (MTX), to produce highly active metabolites. FPGS tag SNP rs1544105C > T is located in the gene promoter. The aim of the present study was to investigate the impact of rs1544105 polymorphism on the treatment outcome in pediatric B-cell precursor acute lymphoblastic leukemia (BCP-ALL). Methods This study enrolled 164 children with BCP-ALL. We genotyped the FPGS SNP rs1544105, and analyzed the associations between its genotypes and treatment outcome. We also examined FPGS mRNA levels by real-time PCR in 64 of the 164 children, and investigated the function of this polymorphism on gene expression. Results We found significantly poor relapse-free survival (RFS) (p = 0.010) and poor event-free survival (EFS) (p = 0.046) in carriers of CC genotype. Multivariable Cox regression analyses adjusted for possible confounding variables showed that, relative to the CT + TT genotypes, the CC genotype was an independent prognostic factor for poor RFS (hazard ratio [HR], 4.992.; 95% CI, 1.550-16.078; p = 0.007). No association was found between any toxicity and rs1544105 polymorphism. Quantitative PCR results showed that individuals with the T allele had lower levels of FPGS transcripts. Conclusions Our study indicates that FPGS rs1544105C > T polymorphism might influence FPGS expression and affect treatment outcome in BCP-ALL patients.


Background
Cure rates for childhood acute lymphoblastic leukemia (ALL) have exceeded 80%. However, for the 15-20% of children with newly diagnosed ALL who will ultimately relapse, traditional risk assessment remains inadequate [1]. Multiple studies have suggested that the outcome in childhood ALL may be affected by how rapidly and effectively an individual patient metabolizes certain chemotherapeutic agents [2]. Thus, if the germline polymorphisms in drug-metabolism genes were better defined, tailored drug therapy based on these factors might further improve outcome.
Methotrexate (MTX) is a major component in all contemporary treatment protocols for childhood ALL. High-dose MTX (HDMTX) is associated with an improved outcome in ALL patients. Despite its clinical success, MTX may lead to several toxicities in a considerable number of patients [3,4]. Folypoly-γ-glutamate synthetase (FPGS) is responsible for the activation of MTX to MTX polyglutamates (MTXPGs). Loss of FPGS activity is an established mechanism of resistance to MTX in vitro and in vivo [5][6][7]. FPGS activity is significantly correlated with FPGS mRNA levels [8]. Overexpression of FPGS has been associated with increased sensitivity of several glioma cell lines to MTX [9]. In addition, altered FPGS expression was found associated with outcome of colorectal cancer patients [10].
In childhood ALL, a strong correlation exists between FPGS expression, intracellular MTXPGs accumulation and treatment outcome [8,11].
Up to now, there have been a few reports on the clinical importance of polymorphisms in FPGS gene [12][13][14]. However, its clinical role in childhood ALL remains to be defined. FPGS rs1544105 C > T is a tag SNP through HapMap database (release #27) in the promoter of FPGS gene. It is reported that FPGS rs1544105 C allele was associated with MTX poor response [13,14]. Panetta et al. found that rs1544105 polymorphism showed a significant relation to FPGS activity in childhood ALL [15]. So we hypothesized that rs1544105 might have important clinical roles in childhood ALL.
In the present study, we investigated the role of FPGS rs1544105 polymorphism in treatment outcome in children with newly diagnosed B-cell precursor ALL (BCP-ALL) and its influence on FPGS expression.
FPGS rs1544105 T allele might cause decreased mRNA levels As rs1544105 polymorphism is located in FPGS promoter, this polymorphism might impact the expression of FPGS. Therefore, we examined FPGS mRNA levels by real-time PCR in 64 children who had ≥70% leukemic cells in diagnostic bone marrow samples and sufficient cDNA for analysis, and whose FPGS rs1544105 genotypes were determined. As shown in Figure 1, we found that the FPGS rs1544105 T variant was associated with a relatively lower mRNA level. Average expression of patients harboring the CC genotype was approximately 1.5 times higher than CT (p = 0.041) and TT (p = 0.031) carriers. No significant differences were detected between CT and TT groups. In addition, FPGS mRNA expression had no association with gene fusions (Additional file 1: Table S1).

FPGS rs1544105 polymorphism and treatment outcome
Given that rs1544105 polymorphism might influence FPGS mRNA levels, and altered FPGS expression was found associated with outcome of colorectal cancer patients [9], we analyzed the correlation between FPGS rs1544105 genotypes and outcomes in 164 patients. Figure 2 showed the effects of rs1544105 polymorphism on treatment outcome.
There was a significantly higher relapse rate in CC group compared with CT + TT carriers (CC: 28.6%, CT + TT: 8.0%; χ 2 : p = 0.034). A poor RFS was found in the CC group (log rank: p = 0.010), as well as a poor EFS (log rank: p = 0.046). Multivariate survival analysis showed that the rs1544105 CC genotype was an independent prognosis factor for poor RFS (HR, 4.992; 95% CI, 1.550-16.078; p = 0.007) compared with the CT + TT genotypes (Table 3).
FPGS rs1544105 polymorphism and MTX-related toxicity FPGS is responsible for the activation of MTX to MTXPGs and higher MTXPGs level is associated with increased toxicity [4]. Given that rs1544105 polymorphism might influence FPGS mRNA levels, so we next analyze the association between FPGS rs1544105 genotypes and MTXrelated toxicities during consolidation therapy. As the HR group was treated with HD-MTX and other chemotherapeutic drugs, for example Dexamethasone, Vincristine, Cyclophosphamide, Cytarabine, L-asparaginase, which may influenced the assessment of MTX toxicity, only 139 patients enrolled in SR and MR treatment groups were evaluable for the toxicity analysis. As shown in Table 4, results showed that there was no correlation between any toxicity and rs1544105 genotypes, calculated by χ 2 test.

Discussion
In the present study, we investigated whether FPGS rs1544105C > T polymorphism was associated with treatment outcome among childhood BCP-ALL patients. We found that CC genotype is significantly associated with poor RFS (p = 0.010) and EFS (p = 0.046) compared with the CT and TT genotypes. Furthermore, in multivariate Cox regression analysis, we found that the FPGS rs1544105 is an independent prognostic factor for RFS. Patients carrying the rs1544105 CC genotype displayed a higher risk of relapse than CT + TT genotypes (HR, 4.992; 95% CI, 1.550-16.078; p = 0.007), indicating that rs1544105 polymorphism is a potential predictor for survival in childhood BCP-ALL. Two previous studies reported that FPGS rs1544105 C allele was associated with MTX poor response in rheumatoid arthritis (RA) patients in north Indians [13,14], which were in line with our finding. To our knowledge, this is the first report that FPGS rs1544105 is associated with outcome in pediatric ALL, which extended our understanding of the role of FPGS in MTX response. Previous observations suggested that FPGS downregulation leads to resistance to MTX and novel antifolates in some cancer cell lines [5,6,16,17]. As rs1544105 T allele carriers had lower FPGS mRNA level, it seems that they  higher intracellular folate levels compared with CC genotype carriers, and consequently, increase in polyglutamylation of intracellular folate cofactors may diminish or even abolish the effects of decrease in polyglutamylation of antifolates. This might lead to enhanced chemosensitivity and better survival in T allele carriers. The differences of folate concentrations between T allele carriers and CC genotype carriers should be determined in the future. Treatment-related toxicity can not only be life threatening, but is also one of the main reasons for interruption or discontinuation of chemotherapy, which may increase relapse risk [8,21]. Two previous studies reported that FPGS rs1544105 C allele was associated with MTX poor response in rheumatoid arthritis (RA) patients in north Indians [13,14]. In our study, no significant association was revealed between this polymorphism and any toxicity. The differences may result from our limited samples.
SNPs in gene regulatory region that alter the expression or function of proteins targeted by drugs can contribute significantly to variation in the responses of individuals. The rs1544105 is located in the promoter of FPGS gene, so we hypothesized that this polymorphism may affect treatment outcome by influencing FPGS mRNA expression. Among the164 genotyped patients, we analyzed the mRNA levels of 64 patients. We found that rs1544105 genotypes were correlated with mRNA levels of FPGS. The T allele carriers (CT + TT) had lower mRNA expression compared with CC genotype. Panetta et al. reported that FPGS activity of rs1544105 CC carriers was 2.6 times higher than CT and TT carriers in a St Jude ALL cohort [15]. As FPGS mRNA expression correlates with activity [8], their results were in agreement with ours. In order to understand the mechanism underlying the altering expression, we investigated the sequence for putative transcription factors binding sites using TRAP [22]. The studied SNP rs1544105 was located within a predicted binding site for cAMP response element-binding protein (CREB) transcription factors, and to modulate the binding affinity of CREB, with higher affinities predicted for the C-allele. CREB is a transcription factor that regulates gene expression principally through activation of the cyclic AMP (cAMP)-dependent cell signal transduction pathways by binding to the cAMP response element (CRE) region at the promoters. CREB is reported to be overexpressed in childhood ALL and be critical in leukemogenesis [23]. Further experiments are needed to reveal the detailed molecular mechanisms.
No statistical association was found between rs1544105 polymorphism and any clinical indices. It was reported that patients with TEL-AML1 and E2A-PBX1 fusion genes had significantly lower level of FPGS mRNA expression in childhood ALL. And TEL-AML1 and E2A-PBX1 can downregulate FPGS promoter activity and lead to lower mRNA expression and enzymatic activity [24,25]. As rs1544105 influences FPGS expression, we supposed that this polymorphism might correlate with TEL-AML1and E2A-PBX1 fusions. However, we found no association between rs1544105 and TEL-AML1 (p = 0.071) and E2A-PBX1 (p = 0.601) fusions. Neither did we found any association between FPGS mRNA level and TEL-AML1 (p = 0.070) and E2A-PBX1 (p = 0.198) fusions. The difference may result from our limited samples. But our data suggested that TEL-AML1 + patients had lower FPGS mRNA expression than TEL-AML1ones (Additional file 1: Table  S1), which was in agreement with previous results.

Conclusions
FPGS rs1544105C > T polymorphism may influence the expression of FPGS and be associated with treatment outcome in childhood BCP-ALL. Larger prospective studies should be undertaken to confirm our recent results. In addition, the underlying mechanisms are still unclear and needs further study.

Study population
The criterion for patient inclusion in this study was newly  Chinese Children's Leukemia Group (CCLG)-ALL 2008 protocol were collected in our study. All subjects were unrelated ethnic Chinese (aged 1.0-15.0 years, median 4.0 years). The common clinical characteristics of 164 patients were shown in Table 1. The median follow-up time was 48.0 months (range, 2.0 to 59.0) with 13 patients suffered bone marrow relapse, 1 suffered isolated central nervous system (CNS) relapse, 2 suffered isolated testis relapse, 4 died of severe infections during induction, 7 lost follow-up, and 137 cases were in continued complete remission (CCR). Patients were stratified into standard-risk, medium-risk and high-risk treatment groups (SR, MR and HR) according to age, white blood cell count (WBC) in peripheral blood, immunophenotype, fusion gene, karyotype, CNS status, prednisone response, day 15 and day 33 BM remission status, minimal residue disease (MRD) on day 33 and at day 78, etc. (Additional file 2: Table S2). Sixty-two (37.8%) patients were assigned to SR group, 81 (49.4%) to MR group, and 21 (12.8%) to HR group (Table 1). Each group of patients received different treatment regimens. The details of treatment during the consolidation period are outlined in Additional file 3: Table S3. Doses of MTX differed by different treatment branch: the standard-risk (SR) group received 2 g/m 2 , while the medium-risk (MR) and highrisk (HR) groups received 5 g/m 2 . In addition, Non-HR (SR and MR) patients received oral 6-mercaptopurine (6-MP) and HR patients received other drugs concurrently. Leucovorin rescue was given at 42 h, 48 h and 54 h from the start of MTX infusion. Of the 164 patients in this study, 139 were enrolled into the analysis of association between FPGS SNP rs1544105 polymorphism and MTXrelated toxicity during consolidation therapy. There were a total of 25 patients not analyzed for MTX-related toxicity, including 4 patients who died before consolidation therapy, and 21 HR patients who were treated with HD-MTX and other chemotherapeutic drugs, for example Dexamethasone, Vincristine, Cyclophosphamide, Cytarabine, L-asparaginase, which may influenced the assessment of MTX toxicity.
Bone marrow (BM) samples of patients were collected at diagnosis. Monouclear cell separation was carried out as previously described [4,26,27] and cells were immediately stored at −70°C until use. This study was approved by the BCH Institutional Ethics Committee and informed consent according to the Declaration of Helsinki was signed by the guardians of the patients.

DNA extraction and SNP genotyping
Genomic DNA was extracted using a Genomic DNA Isolation Kit (U-gene, Anhui, China). Genotypes of the FPGS SNP rs1544105 were determined by PCR-RFLP methods [14]. FPGS rs1544105 primer sequences were forward: 5′-GTGCCTCCTTCACACACAG-3′and reverse: 5′-CCCAGAGTCCTTATTCTTAGCC-3′. PCR products were digested with HpyCH4IV (New England BilLabs, Inc, Beverly, MA) and separated on 10% polyacrylamide gel. A 10% random sample (n = 17) of the cases was verified by directly DNA sequencing of the PCR products.

RNA extraction and FPGS expression analysis
The criterion for patient inclusion in expression analysis was ≥70% leukemic cells in diagnostic bone marrow samples and sufficient cDNA for analysis. Sixty-four patients among the 164 BCP-ALL genotyped in this study met this criterion and were included for a quantitative RNA expression assay. Total RNA from mononuclear cells was isolated using Trizol reagent (Invitrogen, Paisley, UK). The mRNAs were reverse-transcribed into cDNAs using random hexamers and moloney murine leukemia virus reverse transcriptase according to the manufacturer's instructions (Promega, Madison, USA). The mRNA expression level of FPGS was measured by real-time quantitative reverse transcription-PCR using the ABI PRISM 7500 Sequence Detection System (Applied Biosystems, Singapore) based on the SYBR Green method. The ABL gene was used as the internal control, as suggested by the Europe Against Cancer program [28]. The primer sequences used for FPGS PCR were forward: 5′-CCCCGAGGTTCGAGTCTTG-3′ and reverse: 5′-GTCACTGTGAAGTTCTGTTGGTCTG-3′; and the primer sequences used for ABL were forward: 5′-TGGA GATAACACTCTAAGCATAACTAAAGGT-3′ and reverse: 5′-GATGTAGTTGCTTGGGACCCA-3′. The PCR reactions were performed in a total volume of 25 μl, including 12.5 μl of 2 × SYBR® Green PCR Master Mix (Applied Biosystems), 2 μl of cDNA, and 1 μl of primer mix (10 μM each). The PCR program was as follows: 95°C for 10 minutes, 40 cycles of 95°C for 15 seconds, 60°C for 1 minute, and a final step for generation of a dissociation curve to assess the specificity of the amplified products. Relative quantification analysis was performed using the comparative CT method (2 -△△CT ). Reactions were in triplicate in three independent experiments.

Toxicity
Toxicities were graded according to the NCI Common Toxicity Criteria version 1.0. Hematological toxicity (anemia, thrombocytopenia, neutropenia) and non-hematological toxicity (skin toxicity, oral mucositis, and hepatic toxicity which means transaminase levels > 2 times as high as normal limit) were included. The highest grade of toxicity observed in each patient during the consolidation therapy period was recorded. No patients died from toxicity.

Statistical analysis
The allele frequencies of polymorphisms were tested for Hardy-Weinberg equilibrium using the Chi-square (χ 2 ) test. Toxicities were represented by the value 1or 0, indicating whether an adverse event did or not occur during the MTX course. χ 2 test was used to determine the relationship between rs1544105 genotypes with toxicity, outcome and other clinicobiological characteristics. Relapse was defined as the reappearance of leukemic cells in BM (>5% blasts) and/or the reappearance of clinical evidence of the disease. Relapse-free survival (RFS) was defined as the time from CR to the date of relapse, censored at date of last time of contact or death in remission. The duration of event-free survival (EFS) was defined from the date of diagnosis to the date of relapse or death, or the last contact with patients in continuous hematologic CR. The probability of RFS and EFS were estimated using the Kaplan-Meier method and univariate associations between rs1544105 genotypes were compared by log-rank tests. Multivariate Cox regression models adjust for potential confounding variables were used to analyze the prognostic significance of rs1544105 genotypes. All statistical analyses were 2 tailed and performed with SPSS 16.0 for Windows. P value less than 0.05 were considered significant.

Additional files
Additional file 1: Table S1. The association between FPGS mRNA expression and gene fusions (N = 64).