Correspondence: “Association of CD40 gene polymorphisms and immune thrombocytopenic purpura in the adult Egyptian population”

Abstract

Full text

To the editor:

We read the paper titled “Association of CD40 gene polymorphisms and immune thrombocytopenic purpura in the adult Egyptian population” published by Ellithy et al. [1], and we would like to explain some of the methodological information provided by the authors, which may impact the interpretation of the reported results.

The authors genotyped the CD40 rs1883832 (NM_001250.6:c.−1 T > C) polymorphism in the participants using polymerase chain reaction–restriction fragment length polymorphism (PCR–RFLP) with NcoI restriction endonuclease. This enzyme identifies the CCATGG sequence and cleaves the two cytosines. The presence of the CD40 rs1883832 T allele alters the sequence to CTATGG, preventing the NcoI enzyme from performing this cut.

The genotyping criteria for CD40 rs1883832 appear to be incorrect according to the methodological information provided by Ellithy et al. [1]. The NcoI cut of the amplified PCR products was reported for the CD40 rs1883832 T allele in the gel electrophoresis pattern. The presence of two electrophoretic bands of 130 bp and 373 bp was described for the TT genotype, whereas the presence of a single 503 bp band was described for the wild-type (CC) genotype. As such, the genotyping findings and allele frequencies of CD40 rs1883832 (−1 T > C) reported by Ellithy et al. [1] appear to be incorrect. Accordingly, the analysis and interpretation of ITP predisposition and the response to treatment in association with the CD40 rs1883832 (−1 T > C) genotype were affected.

Other researchers have genotyped CD40 rs1883832 (−1 T > C) using the PCR–RFLP technique and have consistently demonstrated that the NcoI enzyme digests the amplified DNA fragments containing the C allele. These results were validated via direct sequencing [2–5].

Figure 1 in Ellithy et al. [1] contains an error: the change in the CD40 rs1883832 nucleotide was incorrectly labeled as G > A.

We think that the authors should carefully review their genotyping methods and validate their results considering the above remarks.