Interethnic Differences in Single and Haplotype Structures of Folylpolyglutamate Synthase and Gamma-glutamyl Hydrolase Variants and Their Influence on Disease Susceptibility to Acute Lymphoblastic Leukemia in the Indian Population: An Exploratory Study

Keywords

Antifolates - folate - folylpolyglutamate synthase - gamma-glutamyl hydrolase - haplotypes - polymorphism

Introduction

Acute lymphoblastic leukemia (ALL) is a hematologic malignancy characterized by the production of immature leukocytes. The estimated number of new cases of ALL in the United States in 2018 is 5960 and there are 1470 predicted deaths from ALL this year.[1] In India, the lymphoid leukemia cases are expected to be 18,449 by the year 2020.[2] Both genetic and nongenetic factors play a role in ALL; however, despite many studies, the etiology of ALL is still poorly understood. Folate deficiency has been associated with the increased risk of some cancers,[3],[4] and lower folate levels were found to be associated with ALL in the Indian population.[5] Folates and antifolates are small molecules that are metabolized intracellularly into their more potent polyglutamate derivatives. Folylpolyglutamate synthase (FPGS) and gamma-glutamyl hydrolase (GGH) are genes located on chromosomes 9 and 8, respectively, that are essential for the intracellular accumulation of folate and antifolate polyglutamates.[6] Mutations in FPGS and GGH genes might affect the activity of these enzymes, altering intracellular levels of polyglutamates [Table 1].[7],[8],[9] Variants in FPGS and GGH are also relevant in the context of the efficacy and safety of antifolate-based therapy.[10],[11],[12] Genetic variants associated with disease among the populations of other countries may not be associated with those in India [13],[14] because Indians are genetically diverse and may differ from other populations.[15],[16],[17] To date, very few studies are available regarding the influence of FPGS and GGH variants and their haplotypes on the risk of ALL in the global populace. Therefore, we aimed to establish the frequency of FPGS and GGH variants in healthy volunteers to provide a normative frequency which can be used to compare with those of patients with cancer risk.

Table 1

Genetic variants investigated in the study

Gene	rsid	Nucleotide change	Type of variant	Chromosome number: position	Effect on enzyme	Consequences on folate and MTX levels
UTR – Untranslated region; MTX – Methotrexate; FPGS – Folylpolyglutamate synthase; GGH – Gamma-glutamyl hydrolase
FPGS	rs1544105	2572 G>A	Intron	9:127800446	Decreased transcripts	Decreased[7]
	rs10106	1944 A>G	3’UTR	9:127813796	Not known	-
GGH	rs3758149	-401 C>T	5’ UTR	8:63039169	Increased expression	Decreased[8]
	rs11545078	452 C>T	Missense	8:63026205	Decreased activity	Increased[9]

Materials And Methods

The present study was approved by the JIPMER Institute's Ethics Committee (IEC; Number: JIP/IEC/SC/2/2012/28). Written informed consent was obtained from all study participants, and in the case of children, consent was obtained from their legally accepted representatives.

Study population

A case–control study consisting of 220 unrelated healthy volunteers (controls) and 151 patients (cases) with ALL of either sex was conducted. All the participants were residing in South India for at least three consecutive generations and spoke one of the Dravidian languages (Tamil, Telugu, Malayalam, or Kannada) as their mother tongue. The mean ages (± standard deviation) of the cases and controls were 15.5 (±10.5) and 24.5 (±4.8) years, respectively. There were 99 (65.6%) males and 52 (34.4%) females in the patient group and 118 (53.6%) males and 102 (46.4%) females in the control group. We could not recruit age-matched healthy children due to difficulty in obtaining consent from patients' parents/legal guardians that resulted in the difference in mean age between cases and controls. Details of sample collection, DNA extraction, and quantification have been described previously.[18] FPGS (rs10106 and rs1544105) and GGH (rs11545078 and rs3758149) TaqMan assays were procured, to detect considered variants, from Applied Biosystems (Foster City, CA; USA). Genotyping was done by allelic discrimination using real-time polymerase chain reaction (Applied Biosystems-7300) according to the manufacturer's instructions. Genotyping was done in duplicates in 30% of the randomly selected samples and were found to be in 100% concordance. Genotype details of the studied SNPs in other ethnic populations were obtained from the 1000 Genomes Project, phase-3, which include five major populations: Africans (AFR), Americans (AMR), East Asians (EAS), Europeans (EUR), and South Asians (SAS). We have also considered subpopulations of SAS such as Gujarati Indians from Houston (GIH), Punjabis from Lahore, Pakistan (PJL), Bengalis from Bangladesh (BEB), Sri Lankan Tamils from the UK (STU), and Indian Telugu from the UK (ITU) for the comparison of genotype frequencies.

Construction of haplotypes and linkage disequilibrium

To construct haplotype blocks and to obtain their corresponding frequencies, the genotype data of two loci per gene (FPGS, rs10106 and rs1544105 and GGH, rs3758149 and rs11545078) were used. Details of the variants are mentioned in [Table 1].

A total of 218 and 211 samples of healthy volunteers were used for haplotype analysis of GGH and FPGS variants, respectively.

Haploview software v4.2 (Broad, Cambridge, MA, USA)[19] was used to estimate the pairwise Linkage Disequilibrium (LD) pattern and haplotype frequencies. All markers/SNPs with minor allele frequencies <0.05 were excluded, and the minimum haplotype frequency was set at 1%. Strong LD between a pair of markers is indicated by D' values from 0.7 to 1, whereas moderate LD is indicated by D' values from 0.2 to 0.7 and D' values from 0 to 0.2 indicate no linkage disequilibrium.

Statistical analysis

The observed genotype frequencies were tested for Hardy–Weinberg equilibrium (HWE) using the Chi-squared test. Fisher's exact test was used to test the differences in genotypes between ALL patients and healthy volunteers (controls), and odds ratios with 95% confidence interval were obtained. Comparison between genotype and allele frequencies of South Indians (SIs) with the 1000 Genomes Project data was made using the Chi-squared test. GraphPad InStat 3 (GraphPad Software Inc., San Diego, CA, USA) and SPSS software (version 16, SPSS Inc.; Chicago, IL, USA) were used for statistical analysis. The threshold for statistical significance was set at P < 0.05.

Results

Comparison of genotype distribution of FPGS and GGH variants between patients with acute lymphoblastic leukemia and healthy individuals

The observed genotype frequencies of FPGS and GGH variants in healthy individuals and patients with ALL were found to be in HWE (P > 0.05). Among the studied variants, FPGS rs1544105'AA' genotype carriers were found to be at risk of developing ALL [Table 2].

Table 2

Distribution of genotypes and allele frequencies of folylpolyglutamate synthase (rs10106 and rs1544105) and gamma-glutamyl hydrolase (rs3758149 and rs11545078) polymorphisms in patients with acute lymphoblastic leukemia and normal healthy individuals

Genotypes and Alleles	Patients with ALL	Healthy volunteers	P value	OR (95% CI)
*P<0.05. OR – Odds ratio; CI – Confidence interval; N – Total number of patients considered to study the respective genetic variant; n (%) – Number of patients possessing respective particular genotype; ALL – Acute lymphoblastic leukemia
FPGS 1944 A>G (rs10106)	N=145; n (%)	N=212; n (%)
AA	49 (33.8)	82 (38.7)		1.00 (reference)
AG	70 (48.3)	103 (48.57)	0.63	1.13 (0.71-1.81)
GG	26 (17.9)	27 (12.6)	0.18	1.61 (0.84-3.07)
A	168 (57.9)	267 (63)		1.00 (reference)
G	122 (42.1)	157 (37)	0.81	1.2 (0.91-1.67)
FPGS 2572 G>A (rs1544105)	N=149; n (%)	N=219; n (%)
Genetic models
Codominant model
GG	44 (29.5)	83 (37.9)		1.00 (reference)
GA	74 (49.7)	109 (49.8)	0.34	1.16 (0.86-1.57)
AA	31 (20.8)	27 (12.3)	0.02*	2.16 (1.15-4.07)
Recessive model GG + GA versus AA	118 (79.2)	192 (87.7)		1.00 (reference)
	31 (20.8)	27 (12.3)	0.04*	1.40 (1.06-1.85)
G	162 (54.4)	275 (62.8)		1.00 (reference)
A	136 (45.6)	163 (37.2)	0.02*	1.41 (1.05-1.91)
GGH -401 C>T (rs3758149)	N=151; n (%)	N=220; n (%)
CC	74 (49.0)	108 (49.1)
CT	67 (44.4)	93 (42.3)	0.82	1.00 (0.79-1.32)
TT	10 (6.6)	19 (8.6)	0.68	0.84 (0.49-1.44)
C	215 (71.2)	309 (70.2)		1.00 (reference)
T	87 (28.8)	131 (29.8)	0.80	0.97 (0.80-1.17)
GGH 452 C>T (rs11545078)	N=151; n (%)	N=218; n (%)
CC	116 (76.8)	151 (69.3)		1.00 (reference)
CT	34 (22.5)	61 (28)	0.41	0.39 (0.06-2.50)
TT	1 (0.7)	6 (2.7)	0.24	0.32 (0.05-2.03)
C	266 (88.0)	363 (83.25)		1.00 (reference)
T	36 (12)	73 (16.74)	0.07	0.78 (0.58-1.03)

Comparison of haplotype structures of studied FPGS and GGH variants between cases and controls

Haplotype structures (HS) of FPGS and GGH variants were compared between cases and controls and were not found to be significantly different. There was however a trend observed with the GGH HS3 haplotype (carrying the variant allele 'T' of both GGH-401 and 452) towards the protection against risk of ALL [Table 3], but it was not statistically significant (P = 0.06).

Table 3

Frequency of haplotype structures of folylpolyglutamate synthase and gamma-glutamyl hydrolase variants in patients with acute lymphoblastic leukemia and healthy volunteers

HS	rs1544105 G>A Allele 1	rs10106 A>G Allele 2	Cases (N=145)	Controls (N=211)	P value
P < 0.05. Chi-square test was used. HS – Haplotype structures; - – Frequency is either absent or <1%; FPGS – Folylpolyglutamate synthase; GGH – Gamma-glutamyl hydrolase; ALL – Acute lymphoblastic leukemia
FPGS
HS1	A	G	41.4	35.8	0.13
HS2	G	A	54.0	61.6	0.07
HS3	G	G	-	1.20	0.51
HS4	A	A	3.10	1.40	0.12
HS	rs1544105 C>T	rs10106 C7gt;T	Cases (N=151)	Controls (N=218)	P value
GGH
HS1	C	C	70.8	69.7	0.73
HS2	C	T	17.3	13.6	0.17
HS3	T	T	11.5	16.5	0.06
HS4	T	C	-	-

Comparison of frequency of studied variants in South Indian population with data from the 1000 Genomes Project and other studies

Allele frequencies of FPGS and GGH variants in our healthy volunteers were compared with those of five superpopulations found in the 1000 Genomes Project phase 3-data and with other studies. Both FPGS rs10106'G' and rs1544105'A' alleles in the SI population (37%) were significantly lower when compared to AFR, AMR, EAS, PJL, and Thai populations,[20] but were similar to EUR and subpopulations of SAS (except PJL) [Table 4].[21],[22],[23],[24],[25],[26],[27],[28],[29]

Table 4

Frequency of folylpolyglutamate synthase and gamma-glutamyl hydrolase variants in South Indians, major populations of 1000 Genome Project phase-3, and other ethnic groups

Population	FPGS 1944 A>G (rs10106)				FPGS 2572 G>A (rs1544105)				GGH -401 C>T (rs3758149)				GGH 452 C>T (rs11545078)
	N	OFH	A	G	N	OFH	A	G	N	OFH	A	G	N	OFH	A	G
*The values are significant (P<0.05) when compared to those of South Indian population. OFH – Observed frequency of heterozygosity; N’ – Number of individuals in that particular group. ‘‑’ not studied. AFR – Africans; AMR – Americans; EAS – East Asians; EUR – Europeans; SAS– South Asians; BEB – Bengalis from Bangladesh; GIH – Gujarati Indians from Houston; ITU – Telugu from the UK; PJL – Punjabis from Lahore, Pakistan; STU – Sri Lankan Tamils from the UK; FPGS – Folylpolyglutamate synthase; GGH – Gamma‑glutamyl hydrolase
SI (present study)	212	49.0	63.0	37.0	219	49.8	62.8	37.2	220	42.3	70.2	29.8	218	28	83.25	16.7
AFR	661	49.2	49.5	50.5*	661	46.1	28.0	62.0*	661	28.0	83.3	16.7*	661	10.0	94.4	5.60*
AMR	347	49.9	53.2	46.8*	347	49.6	51.9	48.1*	347	34.6	77.2	22.8*	347	8.10	96.0	4.00*
EAS	504	42.5	31.0	69.0*	504	42.9	31.0	69.0*	504	33.5	78.1	21.9*	504	16.3	91.3	8.70*
EUR	503	48.3	60.9	39.1	503	49.5	60.3	39.7	503	38.2	72.2	27.8	503	16.5	90.8	9.20*
SAS subpopulation BEB
BEB	86.0	46.5	62.8	37.2	86.0	46.5	62.8	37.2	86.0	41.9	68.6	31.4	86.0	27.9	82.6	17.4
GIH	103	56.3	58.3	41.7	103	56.3	58.3	41.7	103	43.7	70.4	29.6	103	23.3	82.5	17.5
ITU	102	40.2	70.1	29.9	102	41.2	69.6	30.4	102	29.4	74.5	25.5	102	18.6	85.8	14.2
PJL	96.0	51.0	51.6	48.4*	96.0	51.0	50.5	49.5*	96.0	39.6	72.9	27.1	96.0	20.8	89.6	10.4
STU	102	49.0	58.8	41.2	102	46.1	58.3	41.7	102	44.1	70.1	29.9	102	27.5	85.3	14.7
Puerto Rican[21]	940	48.5	50.3	49.7*	-	-	-	-	966	37.7	73.5	26.5	899	53.4	72.6	27.4
Dutch'221	360	-	57.2	42.8	-	-	-	-	-	-	-	-	360	-	91.3	8.70*
Chinese[11]	-	-	-	-	91.0	37.4	34.1	65.9*	91.0	29.7	80.8	19.2*	-	-	-	-
North Indian[23]	-	-	-	-	77.0	-	69.0	31.0	77.0	-	75.0	25.0	77.0		81.0	19.0
Thai[20]	95.0	32.0	28.0	72.0*	98.0	29.6	21.2	71.8*		-	-	-	-	-	-	-
Thai[10]	-	-	-	-	-	-	-	-	100	39.0	76.5	23.5	-	-	-	-
Japanese[24]	-	-	-	-	-	-	-	-	-	-	-	-	269	10.4	94.4	5.60*
Chinese[25]	-	-	-	-	-	-	-	-	-	-	-	-	132	12.1	90.9	9.10*
Chinese[26]	-	-	-	-	-	-	-	-	-	-	-	-	82.0	16.9	87.0	13.0*
Brazilian[27]	-	-	-	-	-	-	-	-	-	-	-	-	200	-	93.0	7.00*
Mexican[12]	-	-	-	-	-	-	-	-	140	21.4	85.7	14.3*	140	3.60	98.2	1.80*
West Indian[28]	-	-	-	-	-	-	-	-	144	49.0	39.0	61.0*	-	-	-	-
Singapore Chinese^	462	41.8	29.5	70.5*	-	-	-	-	472	32.2	79.0	21.0*	474	18.6	89.2	10.8*

The frequency of the FPGS rs10106'G' allele in the SI population was also lower when compared to Puerto Rican,[21] Dutch,[22] and Singapore Chinese populations,[29] whereas rs1544105'A' allele frequency was lower when compared to a Chinese population (65.9%).[11] There was also a significant difference in the distribution of genotype and allele frequencies of GGH variants (GGH-401 (rs3758149) and 452 C>T (rs11545078)) between the SI population and other ethnicities, except for subpopulations of SAS such as BEB, GIH, ITU, PJL, and STU [Table 4]. Frequency of the GGH-401'T' allele in the SI population (29.8%) was significantly lower when compared to West Indians (61%),[28] but it was similar to EUR, North Indian,[23] and Thai populations.[10] In the present study, the frequency of GGH-452'T' allele (16.7%) in SI individuals was significantly higher when compared to Dutch,[22] Japanese,[24] Chinese,[25],[26] Brazilian,[27] and Mexican populations,[12] but similar to that found in Puerto Rican [21] and North Indian populations [23] [Table 4].

Comparison of frequency of studied haplotype structures in South Indian population with the 1000 Genomes Project data

FPGS rs10106 and rs1544105 allele frequencies were found to be in complete LD (D' =1) in BEB and PJL populations. A high LD pattern was observed between the considered FPGS variants in SI (0.94) and other populations. Similarly, GGH rs11545078 and rs3758149 (D' =0.97) alleles in SI, AFR, EUR, and GIH were also found to be in high LD. A complete linkage (D' =1) between GGH rs11545078 and rs3758149 alleles was observed in AMR, BEB, EAS, PJL, STU, and ITU populations. There was a difference in the distribution of HS of FPGS and GGH alleles between SI and other populations, except for the SAS population [Table 5].

Table 5

Comparison of haplotype frequencies of folylpolyglutamate synthase and gamma-glutamyl hydrolase variants in South Indian population with the superpopulations of 1000 Genomes Project phase-3

HS	Allele 1	Allele 2	Frequency in SI (%) (n=211)	Frequency in AFR (%) (n=661)	Frequency in AMR (%) (n=347)	Frequency in EAS (%) (n=504)	Frequency in EUR (%) (n=503)	Frequency in BEB (%) (n=86)	Frequency in GIH (%) (n=103)	Frequency in ITU (%) (n=102)	Frequency in PJL (%) (n=96)	Frequency in STU (%) (n=102)
*The values are significant (P<0.05) when compared to those of South Indian population (SI). -: Indicate either absent or <1%. FPGS – Folylpolyglutamate synthase; GGH – Gamma‑glutamyl hydrolase; AFR – Africans; AMR – Americans; EAS – East Asians; EUR – Europeans; SAS – South Asians; BEB – Bengalis from Bangladesh; GIH – Gujarati Indians from Houston; ITU – Telugu from the UK; PJL – Punjabis from Lahore, Pakistan; STU – Sri Lankan Tamils from the UK; HS – Haplotype structures
FPGS	rs1544105 G>A	rs10106 A>G
HS1	A	G	35.8	48.6*	45.6*	68*	38.1	37.2	41.3	29.4	48.4*	40.7
HS2	G	A	61.6	36.2*	50.7*	30*	59.3	62.8	57.8	69.1	50.5*	57.8
HS3	G	G	1.20	1.80	1.20	1.00	1.00	-	-	-	-	-
HS4	A	A	1.40	13.3*	2.50	1.00	1.60	-	-	1.00	1.00	1.00
GGH	rs11545078 C>T	rs3758149 C>T
HS1	C	C	69.7	83.0*	77.2*	78.1*	72.0	68.6	69.8	74.5	72.9	70.1
HS2	C	T	13.6	11.4	18.7*	13.2	18.7*	14.0	12.7	11.3	16.7	15.2
HS3	T	T	16.5	5.30*	4.00	8.70*	9.10*	17.4	16.9	14.2	10.4	14.7
HS4	T	C	3.00	3.00	-	-	-	-	5.00	-	-	-

Discussion

In the present study, the genotype frequencies of FPGS (rs10106 and rs1544105) and GGH (rs11545078 and rs3758149) variants have been established in SIs. Our study is the first to report that the rs1544105'A' allele confers a potential risk of susceptibility to ALL disease in Indians. We have also found a significant intra- and interethnic differences in the allelic distribution of studied FPGS and GGH variants.

All studied FPGS and GGH variants were found to be in HWE, indicating the absence of inbreeding or population stratification.[30] The frequencies of FPGS rs10106'G' and rs1544105'A' alleles were found to be 37% and 37.2%, respectively, whereas the frequencies of GGH-401'T' and GGH 452'T' alleles were observed to be 29.8% and 16.7%, respectively, in our study population. We observed the frequency of FPGS rs1544105'A' allele to be higher (45.6%) in patients with ALL compared to healthy volunteers (37.2%), making it a potential susceptibility factor for the development of ALL. FPGS rs1544105 G>A was predicted to modulate the affinity of the cyclic adenosine monophosphate response element-binding protein (CREB) transcription factor. CREB is reported to be overexpressed in childhood ALL and plays an important role in leukemogenesis.[31] The molecular mechanisms involved in the role of CREB in the pathogenesis of ALL ought to be explored in the future. The FPGS rs1544105'A' allele was associated with decreased FPGS mRNA levels compared to the 'G' allele in a Chinese population.[7] This might have led to decreased intracellular folate polyglutamates. Folate deficiency can increase the risk of cancer through altered methylation and uracil misincorporation during DNA synthesis. However, the differences in intracellular folate concentrations between 'A' allele carriers and 'G' allele carriers in the future should be measured to validate the above findings. Our study results are similar to Huang et al.,[32] but contradictory to the report by Piwkham et al. where 'AG' genotypes of FPGS rs1544105 and rs10106 were found to be associated with the risk of ALL in the Thai population.[20]

In the present study, the GGH452'T' allele was not significantly associated with the risk of ALL [Table 2], and our results are in accordance with the previous studies conducted on Mexican [12] and Chinese populations.[25]

Furthermore, in the present study, the GGH-401C>T polymorphism was also not associated with the susceptibility to ALL. Our study results are similar to the findings by Koomdee et al. in a Thai population,[10] but not in line with a study done on Mexican population where the -401'T' allele was associated with the risk of ALL (P = 0.001).[12]

These contradictory results could be due to differences in the frequency distribution of alleles (GGH-401'T' allele in SI [29.8%] vs. Thai population [23.5%], P = 0.2 and SI vs. Mexicans [14.3%], P < 0.05) [Table 5], and also the involvement of other enzymes in folate metabolism and differences in gene-environment interactions. Therefore, the effect of the GGH-401C>T polymorphism on the risk of ALL ought to be further studied along with other variants in the genes encoding folate-metabolizing enzymes. Comparison of HS of GGH and FPGS variants between healthy volunteers and patients with ALL did not show a significant difference [Table 2].

Observation of genotype distribution of studied variants in other ethnicities revealed that FPGS rs10106'A' allele frequency was highest in ITU (70.1%) followed by SI populations (63%), and the 'G' allele was found at a higher frequency in the Thai population (72%)[20] followed by EAS (69%). A significant difference existed in the distribution of FPGS variant alleles between SI and other populations from the 1000 Genomes Project, except Europeans and SAS [Table 5]. The allelic frequencies of FPGS variants in the SI population showed greatest similarity to genetically closer populations such as SAS, except the PJL population. BEB and GIH populations had a high occurrence of the GGH- 401'T' allele (31.4%) and the GGH 452'T' allele (17.5%), respectively. In a Mexican population, the frequencies of both GGH-401'C' (85.7%) and 452'C' (98.2%) alleles were found to be higher compared to the present study.[12] GGH-401'T' allele frequency in the SI population significantly differed from the frequency in a West Indian population.[28] The significant differences in FPGS variants between SI and PJL populations and GGH variants between SI and West Indians suggest that populations with similar geographical background may not be considered together because they may possess significant differences in their genetic loci.

Haplotype analysis revealed a high LD pattern between studied FPGS (rs10106 and rs1544105) variants (D' >0.95). FPGS HS2 was the most frequent haplotype in SI, followed by HS1, HS4, and HS3, in descending order. HS3 frequency was found to be <3% in all populations. The frequency of HS4 carrying variant allele of rs1544105'A' was high in AFR (13.3%) whereas it was either absent or occurred at <3% frequency in other populations. Therefore, the influence of these polymorphisms on disease susceptibility and drug response might vary in AFR, relative to other populations. There was a significant difference in the distribution of the HS of FPGS alleles between SI and other populations. GGH rs11545078 and rs3758149 alleles were also found to be in strong LD in SI population. HS1 of GGH was the predominant haplotype in all populations, followed by HS3 and HS2 in SI, BEB, GIH, and ITU. HS2 is the second most frequent haplotype in AMR, EAS, EUR, PJL, and STU, followed by HS3. HS4 was found at very low frequencies in SI, AFR, and GIH populations and was absent in other populations. The HS3 haplotype (16.5%) of GGH carrying variant allele was higher in SI population compared to other populations, except BEB and GIH. Significant differences in HS between SI and other populations could be due to differences in allele frequencies, suggesting interethnic variations in the susceptibility to disease and response to treatment. Limitation of our study might be a lack of data on folate levels that could have strengthened our findings. Therefore, folate levels need to be measured at the time of disease diagnosis in the future. Other variants in the genes encoding folate-metabolizing enzymes also need to be explored, to find a reliable biomarker for susceptibility to ALL disease.

Clinical relevance of FPGS and GGH variants in acute lymphoblastic leukemia

FPGS and GGH enzymes are involved in both folate and antifolate metabolism. Therefore, changes in the activities of these enzymes due to genetic variants may the influence the levels of antifolates and there by affect the treatment response also. Patients with FPGS rs1544105'CC' genotype had lower relapse-free survival (P = 0.01) and event-free survival (P = 0.04), but did not develop MTX toxicity.[7] Higher FPGS activity was associated with accumulation of long-chain MTXPGs and better overall survival in patients with ALL.[33] The GGH 452'TT+CT' genotype was associated with increased risk of hepatotoxicity and mucositis, but not hematological toxicity, in a Chinese population.[25] In European children, the GGH 452'T' allele was associated with thrombocytopenia, but neither GGH polymorphisms nor haplotypes were associated with MTX response and survival.[34] GGH -401C>T and 'TT' genotype carriers were at increased risk of developing leukopenia and thrombocytopenia after high-dose methotrexate in a Thai population.[10] In a Mexican population, the GGH-401C>T polymorphism was found to increase the risk of relapse significantly whereas GGH 452C>T polymorphism did not affect ALL outcome.[12] In Chinese patients, a higher serum MTX concentration/dose ratio and a higher concentration of MTX above the therapeutic threshold (>40 μM) were observed in GGH rs3758149'CT' or 'TT' genotype carriers when compared to 'CC' genotype carriers after high-dose MTX therapy. However, FPGS polymorphism was not associated with serum MTX levels.[11] The above-observed differences in clinical outcome of ALL between various ethnicities could partly be explained by the differences in the distribution of GGH and FPGS variant alleles. Therefore, the impact of each SNP on the susceptibility and outcome of diseases might vary among different populations.

Conclusion

In our study, the FPGS rs1544105'AA' genotype was found to be associated with the susceptibility to ALL in SI population. Genotype and haplotype distributions of FPGS (rs10106 and rs1544105) and GGH (rs3758149 and rs11545078) variants in the SI population significantly differed from those of other ethnicities. Our data emphasized that each ethnicity has unique allele frequencies of studied FPGS and GGH variants. Thus, knowledge of genotype frequency distribution within a population can be useful to tailor drug therapy by optimizing drug doses and identifying potential risk groups which may develop toxicity.

Compliance with ethical standards

All procedures performed in our study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Acknowledgments

We would like to thank Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India, for providing an intramural fund to conduct the study.

• References

• 1 Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin 2016; 66: 7-30
• 2 Takiar R, Nadayil D, Nandakumar A. Projections of number of cancer cases in India (2010-2020) by cancer groups. Asian Pac J Cancer Prev 2010; 11: 1045-9
• 3 Duthie sj. Folic acid deficiency and cancer: Mechanisms of DNA instability. Br Med Bull 1999; 55: 578-92
• 4 Zhou x, Meng y. Association between serum folate level and cervical cancer: A meta-analysis. Arch Gynecol Obstet 2016; 293: 871-7
• 5 Sadananda Adiga MN, Chandy S, Ramaswamy G, Appaji L, Aruna Kumari BS, Krishnamoorthy L. et al. Association between plasma homocysteine and riboflavin status in acute lymphoblastic leukemia in children. Indian J Clin Biochem 2009; 24: 257-61
• 6 Longo GS, Gorlick R, Tong WP, Lin S, Steinherz P, Bertino JR. et al. Gamma-glutamyl hydrolase and folylpolyglutamate synthetase activities predict polyglutamylation of methotrexate in acute leukemias. Oncol Res 1997; 9: 259-63
• 7 Liu SG, Gao C, Zhang RD, Jiao Y, Cui L, Li WJ. et al. FPGS rs1544105 polymorphism is associated with treatment outcome in pediatric B-cell precursor acute lymphoblastic leukemia. Cancer Cell Int 2013; 13: 17
• 8 Chave KJ, Ryan TJ, Chmura SE, Galivan J. Identification of single nucleotide polymorphisms in the human gamma-glutamyl hydrolase gene and characterization of promoter polymorphisms. Gene 2003; 319: 167-75
• 9 Cheng Q, Wu B, Kager L, Panetta JC, Zheng J, Pui CH. et al. A substrate specific functional polymorphism of human gamma-glutamyl hydrolase alters catalytic activity and methotrexate polyglutamate accumulation in acute lymphoblastic leukaemia cells. Pharmacogenetics 2004; 14: 557-67
• 10 Koomdee N, Hongeng S, Apibal S, Pakakasama S. Association between polymorphisms of dihydrofolate reductase and gamma glutamyl hydrolase genes and toxicity of high dose methotrexate in children with acute lymphoblastic leukemia Asian Pac J Cancer Prev 2012; 13: 3461-4
• 11 Wang SM, Sun LL, Zeng WX, Wu WS, Zhang GL. Influence of genetic polymorphisms of FPGS, GGH, and MTHFR on serum methotrexate levels in chinese children with acute lymphoblastic leukemia. Cancer Chemother Pharmacol 2014; 74: 283-9
• 12 Organista-Nava J, Gómez-Gómez Y, Saavedra-Herrera MV, Rivera-Ramírez AB, Terán-Porcayo MA, Alarcón-Romero Ldel C. et al Polymorphisms of the gamma-glutamyl hydrolase gene and risk of relapse to acute lymphoblastic leukemia in Mexico. Leuk Res 2010; 34: 728-32
• 13 Rani DS, Carlus SJ, Poongothai J, Jyothi A, Pavani K, Gupta NJ. et al. CAG repeat variation in the mtDNA polymerase gamma is not associated with oligoasthenozoospermia. Int J Androl 2009; 32: 647-55
• 14 Mehrotra S, Oommen J, Mishra A, Sudharshan M, Tiwary P, Jamieson SE. et al. No evidence for association between SLC11A1 and visceral leishmaniasis in India. BMC Med Genet 2011; 12: 71
• 15 Indian Genome Variation Consortium. Genetic landscape of the people of India: A canvas for disease gene exploration. J Genet 2008; 87: 3-20
• 16 Reich D, Thangaraj K, Patterson N, Price AL, Singh L. Reconstructing Indian population history. Nature 2009; 461: 489-94
• 17 Tamang R, Singh L, Thangaraj K. Complex genetic origin of Indian populations and its implications. J Biosci 2012; 37: 911-9
• 18 Kodidela S, Pradhan SC, Dubashi B, Basu D. NeonaInfluence of dihydrofolate reductase gene polymorphisms rs408626 (-317A>G) and rs442767 (-680C>A) on the outcome of methotrexate-based maintenance therapy in South Indian patients with acute lymphoblastic leukemiatal. Eur J Clin Pharmacol 2015; 71: 1349-58
• 19 Barrett JC, Fry B, Maller J, Daly MJ. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263-5
• 20 Piwkham D, Siriboonpiputtana T, Beuten J, Pakakasama S, Gelfond JA, Paisooksantivatana K. et al. Mutation screening and association study of the folylpolyglutamate synthetase (FPGS) gene with susceptibility to childhood acute lymphoblastic leukemia. Asian Pac J Cancer Prev 2015; 16: 4727-32
• 21 DeVos L, Chanson A, Liu Z, Ciappio ED, Parnell LD, Mason JB. et al. Associations between single nucleotide polymorphisms in folate uptake and metabolizing genes with blood folate, homocysteine, and DNA uracil concentrations. Am J Clin Nutr 2008; 88: 1149-58
• 22 Sinha S, van der Straaten RJ, Wessels JA, de Vries-Bouwstra JK, Goekoop-Ruiterman YP, Allaart CF, Bogaartz J. et al. Exploratory analysis of four polymorphisms in human GGH and FPGS genes and their effect in methotrexate-treated rheumatoid arthritis patients. Pharmacogenomics 2007; 8: 141-50
• 23 Sharma S, Das M, Kumar A, Marwaha Marwaha, Shankar S, Aneja R. et al. nteraction of genes from influx-metabolism-efflux pathway and their influence on methotrexate efficacy in rheumatoid arthritis patients among Indians. Pharmacogenet Genomics 2008; 18: 1049-9
• 24 Hayashi H, Fujimaki C, Inoue K, Suzuki T, Itoh K. et al. Genetic polymorphism of C452T (T127I) in human gamma-glutamyl hydrolase in a Japanese population. Biol Pharm Bull 2007; 30: 839-41
• 25 Chen X, Wen F, Yue L, Li C. Genetic polymorphism of γ-glutamyl hydrolase in Chinese acute leukemia children and identification of a novel double nonsynonymous mutation. Pediatr Hematol Oncol 2012; 29: 303-12
• 26 Zhang HH, Yue LJ, Chen XW, Zhao W, Hu CY, Zheng MM. et al. Analysis of a 452C/T single nucleotide polymorphism in γ-glutamyl hydrolase gene in children with acute leukemia. Zhonghua Yi Xue Yi Chuan Xue Za Zhi 2012; 29: 352-5
• 27 Chiabai MA, Lins TC, Pogue R, Pereira RW. Population analysis of pharmacogenetic polymorphisms related to acute lymphoblastic leukemia drug treatment. Dis Markers 2012; 32: 247-53
• 28 Ghodke Y, Chopra A, Shintre P, Puranik A, Joshi K, Patwardhan B. et al. Profiling single nucleotide polymorphisms (SNPs) across intracellular folate metabolic pathway in healthy Indians. Indian J Med Res 2011; 133: 274-9
• 29 Oppeneer SJ, Ross JA, Koh WP, Yuan JM, Robien K. Genetic variation in folylpolyglutamate synthase and gamma-glutamyl hydrolase and plasma homocysteine levels in the Singapore Chinese Health Study. Mol Genet Metab 2012; 105: 73-8
• 30 Wigginton JE, Cutler DJ, Abecasis GR. A note on exact tests of hardy-weinberg equilibrium. Am J Hum Genet 2005; 76: 887-93
• 31 Pigazzi M, Ricotti E, Germano G, Faggian D, Aricò M, Basso G. et al CAMP response element binding protein (CREB) overexpression CREB has been described as critical for leukemia progression. Haematologica 2007; 92: 1435-7
• 32 Huang Z, Tong HF, Li Y, Qian JC, Wang JX, Wang Z. et al. Effect of the polymorphism of folylpolyglutamate synthetase on treatment of high-dose methotrexate in pediatric patients with acute lymphocytic leukemia. Med Sci Monit 2016; 22: 4967-73
• 33 Wojtuszkiewicz A, Peters GJ, van Woerden NL, Dubbelman B, Escherich G, Schmiegelow K. et al. Methotrexate resistance in relation to treatment outcome in childhood acute lymphoblastic leukemia. J Hematol Oncol 2015; 8: 61
• 34 Garcia-Bournissen F, Moghrabi A, Krajinovic M. Therapeutic responses in childhood acute lymphoblastic leukemia (ALL) and haplotypes of gamma glutamyl hydrolase (GGH) gene. Leuk Res 2007; 31: 1023-5