We are pleased by Dr. Blank's interest in our paper and offer the following in response to points raised in his letter. We have previously shown that the G238C transversion alone is associated with a 100-fold reduction in TPM catalytic activity in a heterologous expression system [1] and that this is the only mutation in some patients with TPM deficiency [1]. Likewise, we have shown that the G460A and the A719G mutations are the only mutations in the TPM complementary DNA (cDNA) of other patients with TPM deficiency, and that the cDNA with only these mutations has more than a 200-fold reduction in activity in a heterologous expression system [2]. The latter was also found by another group [3]. We have more recently established that the proteins encoded by these two mutant alleles undergo enhanced proteolysis [4]. Thus, alleles with these mutations are clearly associated with loss of TPM activity, based on cDNA sequencing [1-3], heterologous expression [1-4], and the concordance of phenotype and genotype.

Dr. Blank is correct that we provided no information on the presence or absence of other silent mutations or other "conservative substitutions that lack clinical significance"; we chose to focus our efforts on those mutations that are of clinical importance.

We are puzzled by Dr. Blank's impression that there was a "smaller-than-expected proportion of homozygotes with TPM deficiency," when in fact the opposite was the case in our study sample. The prevalence of TPM deficiency is approximately 1 in 300 white persons [5]; thus, about 10.7% of the population would be heterozygous according to Hardy-Weinberg equilibrium, a percentage consistent with our findings. Furthermore, our study sample included six homozygotes with TPM deficiency; the expected number is one. The over-representation occurred because we included all available TPM-deficient patients referred to our center for consultation. As expected, all six of these patients had two mutant TPM alleles detected by our PCR-based assays. Finally, no evidence currently shows that TPM deficiency or heterozygosity is associated with any aberrant phenotype unless these patients are treated with thiopurine medications. The natural or environmental substrates for TPM have not been identified, however, and it remains to be unequivocally determined whether this genetic polymorphism is associated with a clinical phenotype in the absence of thiopurine therapy. Studies of these and other important areas should be facilitated by the genotyping methods reported in our paper.