Fragile X syndrome, the most common inherited cause of intellectual disability, results from mutation of the FMR1 gene on the X chromosome that disrupts expression of the fragile X mental retardation protein (FMRP) [1],[2]. Although there are very rare cases in which the fragile X syndrome is due to point mutation or deletion of the FMR1 gene [3]-[7], the most common fragile X mutation eliminates FMRP expression through expansion of a CGG repeat in the 5'-untranslated region of FMR1 to more than 200 triplets (the full mutation). Although the fragile X syndrome results directly from the absence of functional FMRP, diagnosis of this syndrome is based on allele size and detection of the full mutation expansion in genomic DNA.

FMR1 alleles are sorted into four size categories based on their stability upon transmission: normal (up to 44 CGG repeats): intermediate (45-54 repeats); premutation (55-200 repeats); and full mutation (>200 repeats). In the North American population the full-mutation allele (>200 repeats) which is exclusively maternally inherited, has an approximate prevalence of 1 in 4,000 while the premutation is much more common with a prevalence of 1 in 151 females and 1 in 468 males [8]. The premutation can expand to the full mutation when transmitted from mother to offspring and the likelihood of this expansion is dependent on the length and structure of the allele. The majority of premutation alleles are less than 90 triplet repeats in length. FMRP expression is reduced as triplet repeat length increases from the normal range and is eliminated by a process analogous to X inactivation when the repeat exceeds approximately 200 copies [9],[10].

Since males carry only a single X chromosome, those with a full mutation allele develop the fragile X syndrome due to the absence of FMRP. Approximately 40% of males with the full mutation have some somatic cells with smaller alleles from triplet repeat contractions during early embryogenesis. These alleles express FMRP but only rarely does this mosaic expression ameliorate the syndrome. Contractions are also presumably responsible for the premutation size alleles in the sperm of full mutation males [11]. Large premutation alleles (approximately 150-200 repeats) are associated with intellectual impairment due presumably to reduced FMRP expression.

Females carry two X chromosomes and two copies of the FMR1 gene, only one of which is expressed in any particular somatic cell after X inactivation during early embryogenesis. When FMRP expression is reduced because one FMR1 allele in a female carries the full mutation, the degree of impairment can range from undetectable to severe. This variability is due to variation in random X inactivation and the resulting mosaic distribution of somatic cells in which FMRP is expressed. Since the full mutation must be maternally inherited, homozygous full mutation females do not occur. As in males, mosaicism for smaller alleles occurs in females but it is more rarely observed.

We recently reported the development of a simple, accurate, and inexpensive capture immunoassay that determines the level of FMRP in dried blood spots (DBS) as well as in lymphocytes and other tissues [12]. Our initial study of FMRP in DBS from 215 individuals with normal, premutation, and full-mutation FMR1 alleles was designed to characterize the FMRP expression of different FMR1 genotypes. In samples from normal individuals we found a broad distribution of FMRP. The level of the protein declined with age from infants to preteens. It leveled off in teenage years and remained unchanged through adulthood, with no difference between males and females. We used these DBS samples to evaluate how effectively this assay distinguished affected from unaffected individuals. While the assay readily identified affected (full-mutation) males with sensitivity and specificity approaching 100%, we needed to test a larger set of random population samples to establish the FMRP variability detected by the assay. Since residual DBS from state-mandated newborn screening for metabolic and genetic diseases are available for research and represent a uniform age, we decided to use 2,000 randomly selected anonymous fresh newborn DBS to characterize the variability of FMRP expression in the newborn population which is a potential target for screening with this assay. Considering the estimated prevalence of fragile X, it was relatively unlikely that we would find any affected individuals (i.e. those with virtually no FMRP) among the 1,000 male and 1,000 females sampled. We were primarily interested, however, in the variability of FMRP expression, especially the low extreme of normal expression.

To evaluate the effect of long term storage of newborn DBS on the qFMRP assay, we conducted a retrospective study of 74 aged newborn DBS that had been stored for up to 7 years. These DBS were from a different newborn screening program and included samples from 6 affected (full mutation) males.

Newborn DBS disks (3-mm-diameter; ~7.1 mm²) were obtained from the New York State Department of Health's Wadsworth Center, Albany, NY, USA and the New South Wales Newborn Screening Program, Sydney, Australia. Studies performed with both sets of disks were reviewed and approved by the Institutional Review Boards of the Institute for Basic Research in Developmental Disabilities (IBR) and the DBS source institutions.

The distribution of FMRP levels in the sample of 2,000 newborn DBS from the NY State collection (Figure 1) parallels the profile of 134 non-newborn DBS samples from individuals of normal phenotype and normal size CGG repeat alleles that were analyzed previously [12]. The mean FMRP level was considerably higher than expected--approximately seven-fold higher than in the previous study of 134 normal individuals. While the higher white blood cell level in newborns [17] is almost certainly a contributor to this increase, only about half of the increase in the average FMRP level is explained by the predicted increase in white blood cell count. This suggests that newborns may have increased FMRP expression in white blood cells. Whatever the explanation, the increase in the average FMRP level will enhance the specificity and sensitivity of the qFMRP assay for screening newborn DBS.

Considering the prevalence of fragile X, it is not surprising that none of the 1,000 samples from males and 1,000 from females lacked FMRP or had an FMRP level low enough to signal the presence of the fragile X syndrome. Although there were clearly no samples in this set of 2,000 that were from individuals who would develop the fragile X syndrome, we examined the FMR1 CGG region in the DBS samples with FMRP lower than two SDs below the mean to see what alleles were present at the low extreme of the distribution. Thirteen of the 14 samples in this group had normal CGG repeat alleles, which confirmed that these FMRP levels were within the normal range and that they contained reduced levels of FMRP due to factors other than CGG repeat length-for example, due to the variability in the number of white blood cells in newborns [18].

In contrast one sample (sample 2 in Table 1) appeared to be at the low extreme of the normal FMRP distribution due to reduced FMRP expression from a 161/167-repeat premutation allele. FMR1 alleles of this size are rare and have reduced levels of FMRP [9],[19]. Methylation analysis (Figure 2) indicated skewed inactivation of approximately 90% of the X chromosome with the normal 30-repeat allele. Thus, FMRP expression was primarily from the premutation allele with its reduced FMRP expression which is probably the explanation for the low-normal amount of FMRP. The assignment of this one sample out of 2,000 to the low normal range is extremely unlikely to have occurred by chance and demonstrates the potential discriminatory power of this assay.

The aged DBS from New South Wales newborn screening collection were very different from the fresh 2,000 DBS from the NY State collection. The former newborn DBS had been archived for an extended period and were analyzed for FMRP only after some of the males represented were later found to have the fragile X syndrome. The extended storage time reduced the level of FMRP that could be detected (Figure 3) which shifted the distribution of FMRP levels toward zero (Figure 4). The mean normal concentration was approximately half that of the normal adult population we had previously analyzed [12] and less than a third of the mean of newborn DBS that had been stored for only five weeks (Table 1). The results shown in Table 2 indicate that extended DBS storage, especially for longer than 48 months, could lead to some overlap between normal and full mutation. Despite the reduction of detectable FMRP due to prolonged storage, this retrospective analysis was a highly effective screen that identified all specimens from affected (full mutation) males. However, prolonged storage of newborn DBS (more than 47 months) could lead to an increase in false positive samples. Thus, the qFMRP assay will have limited utility for DBS stored for four years or more. Future retrospective studies with aged newborn DBS should be performed with samples that have been stored less than 47 months, and the levels of FMRP should be compared to those of aged normal DBS having the same prolonged storage time.

It is likely that the qFMRP assay would be able to predict cognitive impairment in full mutation females and distinguish between those that had an IQ <70 and high functioning full mutation as well as premutation females. Studies of FMR1 alleles have shown that the methylation of specific sites is predictive of intellectual impairment in full mutation females [20],[21]. This methylation is presumably an indirect measure of FMRP expression and thus a direct qFMRP analysis is likely to be as predictive as an assay of FMR1 methylation.

Fragile X does not have a distinctive phenotype in infants and young children and the average age at which the syndrome is diagnosed in males is 36 months in the USA and 54 months in Australia [22]-[24]. While there is currently no specific treatment for fragile X, early diagnosis of the syndrome would allow early therapeutic intervention for affected children and timely genetic counseling for their parents. Early diagnosis would become even more critical if pharmacological therapies for fragile X that are currently in phase 2 or 3 clinical trials prove effective.

Diagnosis of the fragile X syndrome is currently based on analysis of the CGG repeat in genomic DNA and DNA molecular tests have been used in pilot newborn screening for fragile X [25],[26]. However, DNA molecular tests that determine CGG allele size identify premutation carriers as well as full mutation individuals. Premutation carriers are at risk for an adult onset disorder, fragile X associated tremor/ataxia syndrome (FXTAS) [27]. Since there is currently no treatment for FXTAS, the identification of newborns with the premutation complicates the ethics of fragile X screening by DNA analysis. There are currently no fragile X newborn screening programs.

Our data show for the first time that it is feasible to measure FMRP in 3-mm-diameter newborn DBS disks which are used in mandatory screening for metabolic and hereditary diseases. The accurate measurement of FMRP in these small diameter disks is enhanced by the relatively high levels of the protein present in neonates, which could in part be due the high leukocyte count [17],[18].

The level of FMRP in newborn is variable and is, on average, seven times higher than that detected in adults. Even though the levels of FMRP detected in DBS decrease with storage time, the qFMRP assay allowed us to identify all affected males from the set of aged DBS from normal and fragile X individuals. Our data suggests that the qFMRP assay could serve as the initial step in a fragile X newborn screening program. In a second screening step, characterization of CGG size and/or methylation status of the FMR1 alleles associated with DBS at the low extreme of the FMRP distribution would indicate the presence of the fragile X syndrome. The correlation between highly reduced or absent FMRP expression and the fragile X syndrome is firmly established for males. This correlation is likely to apply to females as well but further studies will be necessary to firmly establish this link. Fragile X screening by qFMRP has distinct advantages over techniques that detect a CGG expansion since it avoids ethical issues associated with identification of asymptomatic premutation carrier infants and is consistent with current newborn screening technology and cost parameters.