Skip to main content

Archived Comments for: Association of TGFβ1, TNFα, CCR2 and CCR5 gene polymorphisms in type-2 diabetes and renal insufficiency among Asian Indians

Back to article

  1. Comment on Prasad et al. 2007

    Bhaskar Lakkakula, Centre for cellular and Molecular biology

    9 November 2007

    Abstract is well written and reflecting the minute details of the entire article except the total no of samples.

    Introduction was mainly concentrated on candidate genes used in the present study and their implication in the predisposition of glomerulosclerosis and fibrosis in progressive diabetic kidney diseases. The authors never mentioned about diabetic retinopathy.

    Methods section clearly indicating that the authors recruited, 192 diabetic subjects with urinary albumin excretion rate (AER) > 200 mg/l as cases and 225 individuals with type-2 diabetes with Normoalbuminuric (AER<20mg/l) as controls. But at the end of this section the authors referring diabetic retinopathy.

    The authors selected 9 SNPs that were reported earlier

    TGFB1 (19q13.2)

    TGFB1 (-800 Ato G, rs1800468)

    TGFB1 (-509 C to T, rs1800469)

    TGFB1 Arg25Pro (G to C, rs1800471)

    TGFB1 Tyr81His (G to C, at exactly 167 bases downstream of Arg25Pro site)

    TGFB1 Thr 263Ile (C to T, rs1800472)

    TNF-a (6p21.1–21.3)

    TNF-{alpha} (G-308A, rs1800629)

    CCR2 (3p21)

    CCR2 Val64Ile (G to A, rs1799864)

    CCR5 (3p21)

    CCR5 promoter polymorphism 59029 (G to A, rs1799987)

    CCR5 D32 (32bp Ins/Del, rs333).

    The primer sequences reported in table 1 were not giving the exact size of the amplicon what they represented.

    Example 1.

    For TGFβ1 Arg25Pro the primer used was F: 5’TTC CCT CGA GGC CCT CCT A3′and R: 5′CAC AGC AGC GGT AGC AGC TG3’ to get 294 bp PCR product with the sequence of

    “TTCCCTCGAGGCCCTCCTAccttttgccgggagacccccagcccctgcaggggcggggcctccccaccacaccagccctgttcgcgctctcggcagtgccggggggcgccgcctcccccatgccgccctccgggctgcggctgctgccgCTGCTGCTACCGCTGCTGTGgctactggtgctgacgcctggccGgccggccgcgggactatccacctgcaagactatcgacatggagctggtgaagcggaagcgcatcgaggccatccgcggccagatcctgtccaagctgcggc” , but the primers used is not covering the target SNP (Electronic PCR).

    Restriction enzymes need to be checked and should be represented in italics. It seems they never tried to repeat the experiment with the alternative technique like DNA re-sequencing.

    In the Statistical analysis segment the authors mentioned that they used t-test and chi square tests separately. Only chi square test results were presented in the table 2, but t-test results are absent. Hardy Weinberg proportions were calculated for each SNP in about 200 healthy individuals from different states of India and made no comment on the results. Pairwise LD between multiple SNPs of a gene was calculated using EMLD software. No reference to the gene, SNPs, LD measures and their results in the entire article except one proclamation in the results as “Since the r2 value between G>A –800 and C>T –509 SNPs (D' = 0.99, r2 = 0.05) was not significant both of these were genotyped in the case-control population”. Power of sample size calculated using PAWE software was also having zilch reference.


    In the inclusion criteria, the authors clearly defined that the controls of the present study are only diabetic with out any kidney diseases. However the authors showing significant differences in diabetic retinopathy between CRI subjects and DM controls. Results presented in the table 2 are confusing because in the methods the authors mentioned that the present study comprises196 cases and 225 controls.

    Table 2. Represents the number of allele, genotype and their frequencies and chi square results. Sum of two alleles number should be exactly double of the sum of three genotypes number of that locus. The same should match with the no of sample included in the present study. Except for TGFβ1 G>A (-800), CCR2 C>T Val64Ile and CCR5 G>A (59029) SNPs the allele number is showing much deviation than the actual number. Similarly in cases deviation was observed (Over represented) for TGFβ1 G>A (-800) and TGFβ1 C>T (-509) loci. The genotype number was also deviated from the actual numbers at many loci other than CCR2 C>T Val64Ile and CCR5 G>A (59029). Anyhow the authors were taken care to keep the sum of two alleles as 1 and sum of three genotypes as 1.

    With these background shortcomings whatever the results they got in Chi Square and regression analysis will mislead the entire association study.

    Competing interests

    None declared