Article Text


A single base alteration in the CGG repeat region of FMR1: possible effects on gene expression and phenotype
  1. J Tarleton1,
  2. A Kenneson2,
  3. A K Taylor3,
  4. K Crandall1,
  5. R Fletcher4,
  6. R Casey1,
  7. P S Hart5,
  8. D Hatton6,
  9. G Fisch7,
  10. S T Warren8
  1. 1Fullerton Genetics Center, Mission St Joseph's Health System, 267 McDowell Street, Asheville, NC 28803, USA
  2. 2Department of Biochemistry, Emory University School of Medicine, Atlanta, GA, USA
  3. 3Kimball Genetics Inc, Denver, CO, USA
  4. 4North Carolina Department of Health and Human Services, Asheville, NC, USA
  5. 5Department of Pediatrics, Wake Forest University, Winston-Salem, NC, USA
  6. 6Frank Porter Graham Child Development Center, University of North Carolina, Chapel Hill, NC, USA
  7. 7General Clinical Research Center, Yale University. New Haven, CT, USA
  8. 8Howard Hughes Medical Institute and Departments of Genetics, Biochemistry, and Pediatrics, Emory University School of Medicine, Atlanta, GA, USA
  1. Correspondence to:
 Dr J Tarleton, Fullerton Genetics Center, Mission St Joseph's Health System, 267 McDowell Street, Asheville, NC 28803, USA;

Statistics from

Fragile X syndrome is characterised by mild to severe mental retardation, minor physical anomalies such as large ears, long facies with a high forehead, joint hyperextensibility, and macro-orchidism, behaviour such as hyperactivity and social avoidance, and speech and developmental delays.1 Trinucleotide repeat (CGG) expansion mutations account for nearly all cases of fragile X syndrome with only a few patients reported with deletions or point mutations in the FMR1 gene.2–4 Expansion beyond approximately 200 repeats (normal range 5-54 repeats), commonly referred to as a “full mutation”, usually results in hypermethylation of the FMR1 promoter region with a concomitant loss of FMR1 transcription and the protein product of the gene, FMRP.5 People with 55-200 repeats have a premutation, which is unstable and prone to expansion when transmitted by females to their offspring. Premutation carriers produce FMRP and typically have few manifestations of the syndrome.

It is the loss or severe reduction in FMRP that results in the fragile X syndrome.6–8 However, there are numerous reports of subjects with mild phenotypes related to modest FMRP production.9 Such subjects may have learning disabilities but are not retarded (IQ>70), and the milder phenotype appears related to FMRP production arising from transcription of unmethylated alleles. Typically, such patients are cellular mosaics having both full mutations and premutations (termed “full mutation/premutation mosaics”), or having full mutations with varying degrees of methylation (termed “methylation mosaics”). Using a detection technique comprising immunocytochemical staining and counting of lymphocytes expressing FMRP,7 Tassone et al8 identified a significant correlation between FMRP expression and full scale IQ in males with full mutation/premutation mosaicism and methylation mosaicism. These authors observed that the presence of >50% FMRP expressing lymphocytes correlated with non-retarded status (IQ>70). These “high functioning” males, however, displayed some learning disabilities and behavioural manifestations of fragile X syndrome. The findings of improved function in these patients show that the fragile X syndrome phenotype can be modified by the production of substantial amounts of FMRP. The recognition that substantial FMR1 expression ameliorates the phenotype suggests that there may be a general threshold of FMRP expression beyond which normal intellect results. Attempts to define this possible expression threshold are complicated by difficulties in precise quantification of intracellular FMRP, possible differences in FMRP expression between blood and other tissues, and background genetic and social effects on intellect.

We report here on a subject who has a point mutation in the CGG repeat region of FMR1 which reduces gene expression by approximately 24% in a lymphoblastoid cell line established from his peripheral blood. This young male has a history of speech and developmental delay and low-normal IQ measures (table 1). He was initially categorised as having fragile X syndrome based on his clinical presentation as a 2 year old with delays in speech and also on the results of molecular diagnosis for FMR1, which mistakenly indicated a deletion within the gene. Later investigations have shown that the proband does not have a FMR1 deletion but a single base alteration within the CGG repeat segment creating a new EagI restriction enzyme site and hence leading to an unexpected change in migration on Southern blot analysis (fig 1). To our knowledge, this is the first report of a point mutation within the CGG repeat of FMR1. This patient is instructive for two reasons. One, he may represent a “natural experiment” whereby a single base alteration within the CGG repeat sequence affects FMRP production and results in mild phenotypic effects. Two, the single base alteration creates a new restriction enzyme site and illustrates the potential for misinterpretation of molecular diagnostic tests for fragile X syndrome.

Table 1

Clinical assessment for the proband

Figure 1

Southern blot and DNA sequence analysis of DNA from the proband. (A) Southern blot analysis. Lane 1 = patient; lane 2 = patient's mother; lanes 3, 5, and 6 = male controls (patients with normal alleles with <54 repeats); lane 4 = female control with both FMR1 alleles <54 repeats; lane 7 = full mutation control patient with complete methylation. (B) DNA sequence of FMR1 repeat region from the patient. The trinucleotide repeat segment containing 31 repeats is underlined. The G to C transversion is noted by an asterisk and the newly created EagI site is capitalised.


The proband was first evaluated for developmental and speech delay in late 1993 at 30 months of age. He had a five word vocabulary at this time and often used these words inappropriately. A physical examination indicated height and weight on the 10th centile, head circumference >95th centile, prominent forehead, and speech delay. A follow up examination at 4½12 years showed height on the 45th centile, weight on the 25th centile, head circumference >95th centile, long face, prominent forehead, speech delay, but no hyperactivity. Molecular genetic testing for FMR1 was ordered based on the clinical findings.

Initially, the FMR1 alteration was identified as a deletion of approximately 400 bp when a DNA fragment of ∼2.4 kb was observed on Southern blot analysis (using probe StB12.3) instead of the expected 2.8 kb fragment (fig 1A). Routinely, Southern blot analysis using FMR1 specific DNA probes near the CGG repeat and double restriction enzyme digestion with EcoRI and EagI (or other methylation sensitive restriction enzymes) generates a 2.8 kb fragment in males having trinucleotide repeats within the normal range (<54 repeats).10 Polymerase chain reaction (PCR) analysis specific for the FMR1 CGG repeat segment11 showed that the proband had 31 repeats (his mother was homozygous for alleles with 31 repeats), suggesting the “deletion” did not involve the FMR1 region containing the CGG repeat and the nearby flanking sequence. Based on the Southern blot results and clinical presentation of delays typical of fragile X syndrome in young boys, the proband was expected to have fragile X syndrome. Subjects with FMR1 deletions resulting in the loss of gene expression invariably have fragile X syndrome indistinguishable from those who have the common trinucleotide repeat expansion (full mutation).12 The PCR result was unusual since it indicated no disruption of the region in FMR1 exon 1 containing the trinucleotide repeat, a known “hotspot” for many deletion events.13 After several years of follow up evaluation, however, it is clear that the proband has significantly higher cognitive abilities than most males with fragile X syndrome (table 1).

The proband's development has been frequently evaluated since 1994, and he was enrolled in the Carolina Fragile X Project, a long term study to follow the development of young males with fragile X syndrome. Investigators in the Project frequently noted that he was an outlier when compared to other boys of a similar age with fragile X syndrome. More recent cognitive-behavioural assessments at the ages of 6 years 10 months and 8 years 11 months found borderline to low-normal cognitive skills and average adaptive behaviour levels for a child of his age. Unlike same aged males with fragile X syndrome who show declines in IQ and adaptive behaviour levels,14 the proband's test-retest IQ and adaptive behaviour scores were quite stable (table 1).

As a part of the Project's evaluation, a blood sample was obtained from the proband in 1998 to evaluate his production of FMRP (performed at Kimball Genetics Inc). Fragile X DNA testing was repeated along with FMRP testing, but Southern blot analysis at this clinical testing laboratory showed a normal result with a 2.8 kb fragment and no 2.4 kb fragment, as seen in the previous testing. After consultation between the laboratories involved in testing the proband, it was realised that the respective Southern blot analysis protocols differed slightly. While EagI was used as the methylation sensitive restriction enzyme in the original study, another methylation sensitive enzyme, NruI, was used in the second study. This realisation led to DNA sequencing of the region around the trinucleotide repeat segment. A single base transversion (G to C) was identified embedded within the CGG repeat segment (fig 1B). This alteration created a new EagI site 356 bp downstream from the EagI site normally detected, and led directly to production of the 2.4 kb “deletion” fragment originally detected in the proband. Thus, his “deletion” is, in fact, not a deletion but a single base alteration. The 2.4 kb fragment has been found (using EagI in the Southern blot procedure) in both the proband's mother (fig 1) and older sister (data not shown), neither of whom have any characteristics of fragile X syndrome.

The effect of the G to C alteration on FMRP production was assessed in the patient. Using FMRP detection by immunocytochemistry of blood smears,7 he was found to express FMRP from 85% of lymphocytes inspected, which is within the range found in people without fragile X syndrome. This confirms that he is producing detectable levels of FMRP in a normal percentage of cells. Quantitation of the total amount of FMRP is not possible by immunocytochemistry, however. In order to investigate whether total FMRP is affected by the G to C alteration, FMRP quantitation assays were performed on an Epstein-Barr virus transformed B cell line established from a peripheral blood sample obtained from the patient. The amount of FMRP produced by his cell line and two control cell lines was determined in a slot-blot based assay relative to known amounts of purified FMRP15 and using an anti-FMRP antibody.16, 17 The FMRP level was normalised to that of a control protein, eIF-4e, which was determined using a similar method with purified eIF-4e18 and an anti-eIF-4e antibody (Transduction Labs). The molar ratio of FMRP:eIF-4e for the two normal cell lines (29 repeats each) was 2.34 (SE=0.194, n=11). The molar ratio for the patient's cell line was 1.78 (SE=0.146, n=11) representing a reduction of 24%. This reduction is statistically significant by a paired t test (t=5.6455, p=0.0002). Although findings of reduced FMRP in a cell line may not reflect an in vivo reduction, the results were obtained reproducibly on several occasions with the same cell line. After long term culturing of the cell line, the FMRP production eventually returned to levels comparable with controls, presumably as an artefact of culture selection for direct mutation reversion or a suppressor mutation. The phenotypic consequences of a minor reduction in FMRP are unclear without additional studies of males with the G to C transversion. The proband's extended family history does show several maternal great uncles with possible learning deficiencies, but we have been unable to study these people despite repeated requests.


It is tempting to speculate regarding the effect on the patient of possible reduced FMRP production. A clear correlation exists between FMRP production and the intellectual/behavioural phenotype in patients with FMR1 mosaicism. Yet the amount of FMRP production that is enough for rescue or modification of the phenotype has not been definitively established owing to the impossibility of FMRP quantitation in the brain, the limited number of fragile X male patients expressing substantial amounts of FMRP, and the difficulty of isolating the effect of a single gene defect on performance. The uncovering of a point mutation possibly affecting FMR1 expression may provide an example of a quantitative phenotypic effect by minor FMRP reduction, especially if other patients with identical or similar mutations in the repeat segment confirm the effect of the mutation described here. The patient functions in the low average intellectual range, yet has specific deficiencies in some areas, for example, short term memory, which lower his overall scores. Although we have only speculative thoughts regarding a mechanism for his reduced FMRP production, possible effects on FMRP production may occur related to mRNA stability or translatability, or to the association of the 5`-(CGG)(n)-3` binding protein (also known as CGGBP1 or p20). This latter protein binds non-methylated, but not methylated, CGG repeats and appears to modulate the activity of the FMR1 promoter.19 Perhaps subtle effects on FMRP production approximating the 24% found in the patient's cell line lead to minor deficiencies in learning.

Despite nearly a decade of FMR1 molecular analysis with many laboratories world wide using the double restriction enzyme digestion protocol with EcoRI and EagI, no other subjects have been reported with a G to C transversion creating a new EagI site. Although precise numbers of subjects tested for possible FMR1 repeat expansion using this Southern blot protocol have not been tabulated, the protocol is used extensively and the number of tested subjects is likely to number in the tens of thousands or greater. Inspection of the CGG repeat segment shows many sites where a single base alteration may create a new EagI site (CGGCCG), and several possible new sites in surrounding DNA for other commonly used methylation sensitive restriction enzymes, BssHII (GCGCGC), NruI (TCGCGA), or SacII (CCGCGG) (fig 2). Based on the large number of tested subjects, the repeat segment is not frequently prone to mutations which create a new site. However, as this case has instructed, it is prudent to confirm potential FMR1 “deletions” by using a second methylation sensitive enzyme from those listed above or by restriction mapping of the deletion endpoints.

Figure 2

FMR1 repeat segment and flanking DNA. (A) Diagram of a portion of the FMR1 gene showing commonly used restriction enzymes for Southern blot analysis, fragment sizes, the StB12.3 probe, and the location of the repeat segment. The clustered methylation sensitive restriction enzyme sites appear underlined in (B). (The second BssHII site contains overlapping restriction sites.) (B) DNA sequence of FMR1 promoter (GenBank Accession L29074) and repeat regions illustrating numerous potential single base alterations which may create new EagI (E), BssHII (B), SacII (S), or NruI (N) restriction enzyme sites. Bases followed by * indicate potential new sites (* marks bases at any position within the six base target recognition) and letters over the asterisks indicate which restriction enzyme site could be created by a single base change. Within the CGG repeat segment, single base changes similar to that found in the proband can create only EagI or SacII sites (of the four listed enzymes). However, these potential new sites can be created by point mutation at any consecutive CGGCGG (denoted as SE when either a SacII or EagI site can be created by a single base change in the same position). The arrow indicates the location of the proband's mutation and the FMRP translation start site is denoted.

Patients with point mutations in the promoter or 5` untranslated region of FMR1 containing the CGG repeat may be underascertained since the usual molecular diagnostic approaches of assaying for repeat expansion do not identify subjects with small alterations. Mila et al20 recently identified two patients with mental retardation (one reportedly with a phenotype highly suggestive of fragile X syndrome) having alterations in the FMR1 promoter. If similarly affected patients with alterations in the FMR1 promoter or repeat segment are identified, consideration should be given to follow up DNA sequence investigations in patients with phenotypic characteristics of fragile X syndrome yet no repeat expansion. In addition, it is possible that some subjects diagnosed with a “deletion” in FMR1 may actually have a point mutation in a fashion similar to our patient.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.