Hemochromatosis-related mutation detection.

To the Editor: In her recent letter to BLOOD, Dr Lynas reports the development of a ‘‘cheaper and more rapid polymerase chain reaction (PCR) restriction fragment length polymorphism (RFLP) technique’’1 for detection of mutations in HFE,2 the candidate gene for hemochromatosis. The PCR-RFLP method discussed by Dr Lynas uses Rsa I and Bcl I to cleave products amplified using primers designed by Feder et al.2 However, many authors, including Jouanolle et al3 in November 1996, Roberts et al4 in February 1997, Martinez et al5 in the March 1997 issue of this journal, and Carella et al6 and Merryweather-Clarke et al7 in April 1997, have already published PCR-RFLP methods incorporating the use of Rsa I for detection of the C282Y mutation, and Roberts et al4 and Martinez et al5 have already published the application of Bcl I for detection of the H63D mutation. Dr Lynas1 considers this technique to be superior to other RFLP-PCR methods because these two enzymes are cheaper than SnaBI (used by Jazwinska et al8 for C282Y RFLP-PCR) and Mbo I (used by Jouanolle et al,3 Carella et al,6 Merryweather-Clarke et al,7 and Jazwinska et al8 for H63D RFLP-PCR). In reply, we point out that our reasons for using the more frequent 4-bp cutter Mbo I instead of the 6-bp cutter Bcl I are particularly pertinent for the diagnostic applications referred to by Dr Lynas and afforded greater accuracy in a worldwide study of the HFE H63D/C282Y genotype of 5956 chromosomes.7 We redesigned one of the primers used to amplify the DNA fragment encoding the H63D mutation (H63DF primer, 58ACA TGG TTAAGG CCT GTT GC; H63DR primer, 58CTT GCT GTG GTT GTG ATT TTC C)7 to produce a 294-bp fragment instead of the 208-bp fragment amplified by Feder et al.2 The extra 86 bp incorporated in this fragment contain an additional Mbo I recognition site, which provides an internal control for restriction digestion.7,9 Following agarose gel electrophoresis, it is therefore possible to distinguish between a partially or nondigested product of 294 bp, which may be of either genotype, and a digested product containing the H63D mutation, which yields fragments of 57 bp and 237 bp on Mbo I digestion. (The wild-type PCR product yields Mbo I digestion products of 138 bp, 99 bp, and 57 bp.) This internal control against misdiagnosis due to partial or nondigestion is not possible in the method discussed by Dr Lynas, as the fragment amplified from an H63D allele remains uncleaved and is thus indistinguishable from a nondigested fragment from either allele.1 Dr Lynas places considerable emphasis on cost. We suggest that instead of using the QIAmp kit for extracting DNA from blood (Qiagen, Crawley, UK; £1362 for 1,000 extractions from 200 μL of blood), the more economical Nucleon Kit (Amersham Little Chalfont, UK; £95 for sufficient reagents for 2,500 extractions from 200 μL of blood) may be suitable. We have found the Nucleon Kit perfectly adequate for producing DNA for PCR,10 a procedure taking only 30 minutes. Another rapid method of DNA extraction11 enables one to process 20 samples in 1 hour at a total consumables cost of 5 pence per sample. We have also found that the H63D PCR works using whole blood as a template if it is first subjected to 3 cycles of 95°C for 5 minutes followed by 30°C for 5 seconds, as suggested by Rees et al.12 Another cheap and rapid method for C282Y typing is mutagenically separated (MS)PCR.10 This allele-specific technique is ideally suited to routine analysis of blood samples in a diagnostic laboratory, as whole blood may be used as the template in a single PCR reaction per sample and restriction enzymes are not required. Other methods more suited to automation include heteroduplex analysis13 and PCR-SSP.14 The availability of many different methods for the diagnosis of HFE mutations C282Y and H63D means that each will have its own particular advantage in different situations and that diagnostic labs can confirm diagnoses by the simultaneous use of independent methods. However, while we congratulate Dr Lynas on the adaptation of PCR-RFLP for use in her laboratory1 and applaud her focus toward the diagnosis of hemochromatosis patients, we contest her claim to have developed a new method. Finally, there has been a plethora of nonpeer-reviewed material on the subject of HFE, and we welcome the interest this stimulates in the field of hemochromatosis. However, we encourage editors and authors alike to take more care over both the nomenclature used (the gene is correctly called HFE,15 not HLA-H) and the emphasis given to nonpeer-reviewed literature over peer-reviewed publications. Perhaps this would help to avoid the confusion and duplication of methods observed in this case.