Balkan Journal of Medical Genetics

It is difficult to consider this group of patients as a true “population” since they were referred for a variety of reasons, usually including developmental delay and/or various dysmorphologies (Table 1). Our original expectation was that this group of patients might contain an increased proportion of individuals with unusually high M_f or otherwise distinctive GPA somatic mutation. We have observed such rare “outliers” with unusually high mutation frequencies [24] in all of our normal populations, including newborns [4]. Instead, we observed an unusual phenotype for the group as a whole, with an overall normal total GPA M_f, but a skewing of mutant subclasses towards allele loss and duplication at the expense of simple allele loss. It is easier to discuss these two observations separately, before considering them together.

Increased Frequency of Allele Loss and Duplication Mutants. Only two types of populations have shown an effect on the frequency of allele loss and duplication variants in the GPA assay: a Chinese population exposed to benzene [25] and patients with BS [13,14], WS [15,16] and FA [12,18]. In the DNA repair-deficient patients, however, unlike in our group of patients, there were also concomitant increases in the frequency of allele loss variants. Our pediatric patients may therefore have undergone greater than normal genotoxic exposure, specifically of a type, such as organic solvents, that impacts primarily on the loss and duplication endpoint. Alternatively, they may be particularly susceptible to this type of mutation. The ~1.5-fold increased M_f may not seem particularly remarkable, but increases of this magnitude have been observed in WS carriers [14], FA heterozygotes [12,18] and in sporadic cancer patients [26].

On the other hand, loss and duplication M_f are generally higher than allele loss M_f over a lifetime [22], with the difference increasing with age, perhaps because the dosage compensation inherent in their mechanisms improves their viability. Our patients may simply have an exaggerated separation of these two types of mutations, either due to a subtle change in some regulating mechanism or simply due to the small size of the sample group. Nine of the 11 patients had higher loss and duplication M_f than their allele loss M_f, It is interesting to note that the biggest differences between these two endpoints were, in general, more associated with patients with neurological dysfunctions rather than physical dysmorphologies.

Decreased Frequency of Allele Loss and Duplication Mutants. Possible explanations for the relatively low values observed for allele loss M_f in our patients are that they have an unusually low amount of genotoxic exposure, or that they are somewhat resistant to the effects of normal environmental genotoxicity (note that these possibilities are not, in general, consistent with the increased frequency of allele loss and duplication mutants discussed above). Ionizing radiation [27-30] and most genotoxic chemicals [31-33] seem to act exclusively on the allele loss class of mutants in the GPA assay.

Another possibility is that the patient M_f values are not unusual, and that normal controls of the same ages would also show lower than expected allele loss M_f. The only other widely applied method for measuring gene-specific human somatic mutation is the hypoxanthine-quanine phosphoribosyltranserase (HPRT) assay [3,34]. Newborn cord blood HPRT M_f values are 10-fold lower than expected from an age regression from adult values [35,36]. This has been interpreted as the protective effect of the maternal placental barrier, which, being lost at birth, allows for a sudden increase in point mutations caused by environmental genotoxicants [36]. There is no evidence for a similar effect at the GPA locus, but this may be due more to our newborn population than to the pediatric controls. As seen in Fig. 1, our “normal” pediatric controls are not evenly distributed about the age regression from adult values; they actually skew towards lower values as well. It may be important that our newborn GPA values are derived from a study of an indigent population at a university hospital, and that the HPRT M_f established concurrently in this population was unusually high [4].

Altered Mutational Spectrum in the Absence of an Increased Mutation Frequency. It would be best if a single explanation could account for both effects on GPA M_f. We have previously argued that, since induced mutation occurs against a background of endogenous mutagenesis, a shift in mutational spectrum must always be associated with an increase in mutation frequency [37]. Otherwise, a compensatory preventive mechanism must occur concurrently with mutation induction, in order for the overall M_f to remain unchanged. There are three examples of significant shifts in mutation spectrum in the literature without concomitant increases in M_f, all involving mutation at the HPRT locus: in newborn cord bloods of babies whose mothers were exposed to environmental tobacco smoke [38,39], in patients with FA [12,40,41], and in mouse cells bioengineered to be deficient in the error-prone replication enzyme DNA polymerase z [42]. In the first of these, a meta-analysis [43] showed that the lack of an increased induced M_f was peculiar to the study reporting the shift in mutational spectrum [38,39], whereas evidence from other studies showed induced mutation does occur in newborns of mother exposed to tobacco smoke [44,45]; moreover, while the increases observed in the frequencies of certain types of mutations VDJ (variable-diversity-joining)-recombinase associated deletions, point mutations] were often significant on their own, observed decreases in the frequencies of other types of mutations (random deletions) were never significant. In the second example, the lack of an increase in HPRT M_f has been attributed to the innate inviability of FA cells [46]; the authors suggested that the mutagenicity and cytotoxicity of genotoxic exposures are too similar to separate in these cells. Fanconi’s anemia patients have a significant (50-fold) increase in allele loss mutation at the GPA locus, however [11,18]. In the third example, the induction of DNA damage in polymerase z deficient cells is far more lethal than in normal cells, and the enzyme deficiency produces a shift in the mutational spectrum of viable mutations that can be cloned and analyzed.

When these interpretations are applied to the altered pattern of GPA M_f observed in our patients two explanations are possible. The first is that we have analyzed a small group of patients, and that the changes we have seen may not be observed in larger studies. The second possibility is that cells from some of our patients are simply more vulnerable to cytotoxicity. This could account for the clinical observations and the unusual laboratory findings. The sensitivity could arise by a variety of mechanisms, differentially affecting different cell types. This explanation also suggests that the subtle somatic mutational phenotype we have demonstrated within this populations may only be an indication of much more substantial effects that are being masked by excessive cell death. Such an explanation could also be invoked for patients with chromosomal abnormalities, since they also often manifest greater cellular inviability, generally or in specific cell-types.