Logo of ajhgLink to Publisher's site
Am J Hum Genet. 2000 Jan; 66(1): 6–15.
Published online 2000 Jan 10. doi:  10.1086/302720
PMCID: PMC1288349

Elevated Levels of FMR1 mRNA in Carrier Males: A New Mechanism of Involvement in the Fragile-X Syndrome


Fragile-X syndrome is a trinucleotide-repeat–expansion disorder in which the clinical phenotype is believed to result from transcriptional silencing of the fragile-X mental retardation 1 (FMR1) gene as the number of CGG repeats exceeds ∼200. For premutation alleles (∼55–200 repeats), no abnormalities in FMR1-gene expression have been described, despite growing evidence of clinical involvement in premutation carriers. To address this (apparent) paradox, we have determined, for 16 carrier males (55–192 repeats), the relative levels of leukocyte FMR1 mRNA, by use of automated fluorescence-detection reverse transcriptase–PCR, and the percent of lymphocytes that are immunoreactive for FMR1 protein (FMRP). For some alleles with >100 repeats, there was a reduction in the number of FMRP-positive cells. Unexpectedly, FMR1 mRNA levels were elevated at least fivefold within this same range. No significant increase in FMR1 mRNA stability was observed in a lymphoblastoid cell line (160 repeats) derived from one of the carrier males, suggesting that the increased message levels are due to an increased rate of transcription. Current results support a mechanism of involvement in premutation carriers, in which reduced translational efficiency is at least partially compensated through increased transcriptional activity. Thus, diminished translational efficiency may be important throughout much of the premutation range, with a mechanistic switch occurring in the full-mutation range as the FMR1 gene is silenced.


In fragile-X syndrome (MIM 309550), clinical involvement is thought to be the result of reduced levels of the normal protein product of the fragile-X mental retardation 1 (FMR1) gene. In the standard model of FMR1 expression, protein levels are reduced as a direct consequence of transcriptional silencing of the FMR1 gene, which occurs when it becomes fully expanded (>∼200 CGG repeats; full mutation) and methylated (Pieretti et al. 1991). In accord with this model, nearly all males with methylated, full-mutation alleles have mental retardation and have little (<10% of normal) or no FMR1 protein (FMRP) in the peripheral lymphocytes (Willemsen et al. 1997; Tassone et al. 1999). Modest FMRP levels can occur in individuals who have fully expanded alleles with a partial or complete lack of methylation, or they can occur as a consequence of somatic mosaicism, in which some cells harbor alleles with <200 repeats. Individuals with higher FMRP levels usually present with a milder clinical phenotype, with learning disabilities occurring in the absence of mental retardation (Smeets et al. 1995; de Vries et al. 1996; Hagerman 1996; Tassone et al. 1999).

For those individuals with CGG-repeat numbers in the range of ∼55–200, the term “premutation” has been coined, to reflect both the propensity for allele expansion in subsequent generations and the absence of direct clinical involvement (Fu et al. 1991; Oberlè et al. 1991). The findings of most clinical investigations of individuals—mainly females—with premutation alleles have demonstrated that the individuals have normal intellectual abilities (Mazzocco et al. 1993; Reiss et al. 1993; Rousseau et al. 1994). However, a limited number of females with the premutation do have mild physical and/or emotional problems (Franke et al. 1998; Riddle et al. 1998). Males with the premutation have also been characterized, and the results of several studies have demonstrated that at least some males have cognitive impairments (Loesch et al. 1987, 1994; Dorn et al. 1994; Rousseau et al. 1994; Smits et al. 1994; Hagerman et al. 1996; Steyaert et al. 1996). Findings of clinical involvement in the premutation range suggest the need to reexamine the standard model—and the concept of the premutation—at the molecular level.

Expression of the FMR1 gene has not been investigated, in a systematic fashion, for alleles in the premutation size range. FMR1 mRNA and FMRP levels were reported to be normal in lymphocytes and in cultured lymphoblastoid and fibroblast cell lines from individual carriers of a premutation (Pieretti et al. 1991; Devys et al. 1993; Feng et al. 1995a, 1995b; Hmadcha et al. 1998). Barring the possibility of occult expansion in other tissues, these latter observations present an apparent paradox for the standard model, since clinical involvement in some males and females with premutation alleles leads to the expectation of reduced FMR1 mRNA and/or FMRP levels. To examine this issue, we have employed a quantitative-fluorescence reverse transcriptase–PCR (RT-PCR) method (Livak et al. 1995; Heid et al. 1996), to obtain precise estimates of FMR1 mRNA levels in peripheral blood leukocytes. This method employs a dual-labeled, fluorogenic hybridization probe, to provide accurate and reproducible quantification of mRNA levels. Quite unexpectedly, for premutation alleles in the 100–200-repeat range, FMR1 message levels are approximately fivefold higher than the levels found in normal individuals. These elevated mRNA levels are present even in the face of lowered percentages of FMRP-positive (FMRP[+]) lymphocytes in this repeat range. Our observations suggest that mechanisms other than reduced transcription (e.g., blocks in nuclear export or translation) are responsible for the FMRP deficit and, ultimately, for clinical involvement in the premutation range.

Material and Methods

Isolation and Analysis of Genomic DNA

Genomic DNA was isolated from peripheral blood leukocytes that were derived from 3–5 ml of blood obtained from carrier males or normal controls and from lymphoblastoid cell lines. This research was performed with protocols approved by the local institutional review board and with signed, informed consent. DNA isolation employed Puregene kits (Gentra). Southern blot analyses were performed for all carrier males, as described elsewhere (Taylor et al. 1994). For each analysis, 5 μg DNA were digested with the use of EcoRI and NruI; this was followed by electrophoresis in 1% agarose/Tris-acetate gels and by transfer to nylon membranes. Blots were hybridized with the FMR1-specific probe StB12.3 (Oberlè et al. 1991). Standard PCR analysis was also used to determine the number of CGG repeats in carrier males, by use of primers “1” and “3,” as described elsewhere (Brown et al. 1993). Of the 16 males with CGG-expansion (premutation) alleles represented in the present investigation, 9 possess alleles within the 55–100-repeat range (n = 55, 58, 60, 66, 73, 81, 84, 85, and 94), and 7 possess alleles within the 100–200-repeat range (CGG)n, n = 113, 126, 130, 133, 180+, 183, and 190); the plus sign (+) designates a smear of unmethylated alleles extending from 180 to 280 repeats, on the basis of the results of a Southern blot. The aforementioned groups were compared with eight control individuals with alleles within the normal repeat range (n = 25, 28, 29, 30, 35, 39, 44, and 54).

Total-RNA Isolation and cDNA Synthesis

Total cellular RNA was prepared from 3–5 ml of blood, by use of standard methods (Purescript kits [Gentra]; Trizol [BRL]), and was quantified by total absorbance at 260 nm. RT reactions were performed in 100-μl aliquots containing 1 × PCR buffer (20 mM Tris-HCl, pH 8.4, 50 mM KCl) (Gibco/BRL), 5.5 mM MgCl2, 1 mM each dNTP, 5 μM random sequence deoxyoligonucleotide hexamers (Gibco/BRL), 0.4 U RNAse inhibitor (Gibco/BRL), and 2.5 U Moloney murine leukemia-virus RT (Gibco/BRL). At least three concentrations of total RNA (500 ng, 250 ng, and 125 ng, per 100-μl reaction) were used for each sample, to ensure linearity of the RT-PCR response. The RT temperature profile was as follows: 25°C for 10 min, 48°C for 40 min, 95°C for 5 min, and final cooling to 4°C. As a control for genomic contamination, 500 ng total RNA were treated as described above, with the exception that the RT was omitted.

Quantitative Fluorescence PCR for Determination of Relative FMR1 mRNA Levels

Fluorescence PCR reactions were performed with the use of the 7700 Sequence Detector (PE Biosystems) (Livak et al. 1995; Heid et al. 1996). The instrument determines relative abundances of mRNA, by use of real-time fluorescence detection of dual-labeled (TaqMan) probes, 5′-FAM (6-carboxyfluorescein)-deoxyoligonucleotide-TAMRA (6-carboxytetramethylrhodamine)-3′ (PE Biosystems), which are complementary to the PCR amplicon. A passive reference dye, ROX (PE Biosystems User Bulletin), is used to normalize all fluorescence signal intensities. During PCR amplification, the fluorescence intensity of the reporter dye (fFAM) is normalized to the fluorescence intensity of the reference dye (fROX) (RnfFAM/fROX). The baseline value for Rn is determined during the first 10–15 cycles (Heid et al. 1996), as is the standard deviation (SD) of RnRn). The change in Rn, ΔRn(≡ Rn-Rn,baseline), is directly proportional to the number of copies of the amplicon (Gibson et al. 1996). PCR primers were synthesized from β-cyanoethylphosphoramidites, by use of a Millipore Expedite automated synthesizer, model 8909. TaqMan probes were obtained from Applied Biosystems/PE or Synthetic Genetics. Cross-comparisons of quantitative-PCR results were performed with the use of at least two of three 7700 Sequence Detectors, one of which was located at the University of California, San Francisco Cancer Center, and two of which were located at the University of Colorado School of Medicine. There was no instance of a significant difference in estimated mRNA level among the instruments.

In the present investigation, the FMR1 amplicon is a 122-bp product that spans the junction between exons 3 and 4 of the FMR1 gene (GenBank accession number L29074), thus eliminating false signals resulting from genomic contamination. The following FMR1 primer/probe set was used: forward (F) primer (F-FMR1), 5′-GCA GAT TCC ATT TCA TGA TGT CA-3′; reverse (R) primer (R-FMR1), 5′-ACC ACC AAC AGC AAG GCT CT-3'; and TaqMan probe, 5′-(FAM)-TGA TGA AGT TGA GGT GTA TTC CAG AGC AAA TGA-(TAMRA)-3′. The relative abundance of FMR1 mRNA, as well as variations encountered during sample preparation, was assessed by use of a reference (81-bp) amplicon derived from β-glucoronidase (GUS) mRNA (GenBank accession number NM000181). GUS mRNA represents a convenient reference, since we have observed that normal peripheral blood leukocytes maintain comparable levels of FMR1 and GUS mRNAs. The GUS amplicon spans the junction between exons 11 and 12 of the GUS gene and utilizes the following primer/probe set: F-GUS, 5′-CTC ATT TGG AAT TTT GCC GAT T-3′; R-GUS, 5′-CCG AGT GAA GAT CCC CTT TTT A-3′; and TaqMan (sense) probe, 5′-(FAM)-TGA ACA GTC ACC GAC GAG AGT GCT GG-(TAMRA)-3′.

For each blood sample, quantitative-PCR reactions were performed in duplicate for each starting total RNA concentration, for both FMR1 and GUS amplicons (a total of 12 reactions). Control reactions (FMR1 and GUS) were run in parallel (an additional 12 reactions), with the use of a standard lymphoblastoid line known as “RP” (RPM 7666; normal lymphoblasts [ATCC CCL114]). The control reactions are necessary for correction of potential variations in buffer composition and/or pipetting errors and for correction of relative efficiencies of RT reactions. PCR amplifications were performed in 50-μl reaction volumes containing 1 × PCR buffer (Applied Biosystems/PE), 5.5 mM MgCl2, 0.1 μM each primer and probe, 200 μM each dNTP, and 0.025 U Taq Gold (Applied Biosystems/PE)/μl. Cycle parameters were as follows: 95°C for 12 min, then (95°C for 15 s, 60°C for 1 min) × 50 cycles, and then a final temperature of 25°C.

Data Analysis

The number of cycles required for FAM fluorescence to reach a threshold level that is ∼10 SDs (of fluctuations in background fluorescence, σRn) above the mean background fluorescence (≡Ct value) was determined for each PCR reaction (FMR1, GUS, and RNA concentration), by use of 7700 automated software. Plots of Ct versus total RNA were performed for each sample, to verify appropriate log-linear behavior (2-Ct ∝ total RNA). For a perfect log-linear response, each cycle represents one doubling of the number of amplicon DNA molecules; thus, a ΔCt of 1 reflects a twofold difference in RNA concentration. All relative FMR1 mRNA levels were determined from the following relation: FMR1 mRNArel = 2-[ΔCt(sample) - ΔCt(RP)] = 2-ΔΔCt, where ΔCt(sample) [ =Ct(FMR1) − Ct(GUS)] is the difference, in threshold cycle number, between FMR1 and GUS, for a given sample and total-RNA concentration; where ΔCt(RP) is the same difference for the control cell line; and where ΔΔCt is the cell-line–corrected difference. This approach will be discussed in more detail below. RT-PCR amplifications were also examined, by means of PAGE (12%), to verify the presence of the correct amplicon. After ethidium bromide staining, single bands were visible only at the expected positions for the FMR1 and GUS amplicons. Furthermore, control experiments done with the use of an FMR1 deletion (Gu et al. 1994) extending from ∼80 kb upstream of the FMR1 gene through exon 7 (thus deleting the entire 5′ regulatory region and the [CGG]n repeat as well as the targets of the PCR primers) did not produce detectable FMR1 amplicon after 50 cycles of PCR. All analyses of Ct data, including nonpaired t-test and regression analyses, were performed on SigmaPlot (SPSS, Inc.).


The percentage of lymphocytes expressing detectable FMRP was determined by immunocytochemistry of blood smears (Willemsen et al. 1995, 1997; Tassone et al. 1999), with use of the anti-FMRP monoclonal antibody from hybridoma clone 1C3-a (Devys et al. 1993). The method is an indirect alkaline phosphatase–staining technique, in which cells fixed with 3% paraformaldehyde and permeabilized with methanol are incubated with FMRP-specific antibody, followed by incubation with goat anti-mouse immunoglobulin conjugated with biotin (DAKO) and a final incubation with streptavidin-biotinylated alkaline phosphatase (DAKO). Visualization of FMRP is accomplished with the use of a fuchsin-substrate chromagen system (DAKO). For each blood sample, 200 lymphocytes are counted, and FMRP(+) lymphocytes are scored for the presence of a red-staining cytoplasmic ring that indicates the presence of FMRP.

Cell Culture and Inhibition of Transcription

Lymphoblastoid cell lines were grown in RPMI 1640 supplemented with 12% FCS and glutamine. For determination of FMR1 mRNA stabilities, actinomycin D (10 μg/ml) was used to block transcription (Baeyens and Cornett, 1993; Feng et al. 1995a). Lymphoblastoid cell lines that were established, by means of Epstein-Barr virus transformation, from normal males and carrier males were plated at a density of 1 ×106cells/ml in 50 ml media. After 24 h, actinomycin D was added, and aliquots of suspended cells were removed at designated times. Cells were washed in PBS, and total RNA was isolated as described above. Samples were subjected to fluorescence RT-PCR, which was performed as described above.


Automated Fluorescence-Detection RT-PCR as a Quantitative, Reproducible Method for Measurement of FMR1 mRNA Levels

We have used an automated, fluorescence-detection RT-PCR assay to quantify the levels of FMR1 mRNA in carrier males with CGG-expansion alleles lying within the range of ∼55–200 repeats (premutation alleles). The basis of this method is the accurate determination of the increase in FMR1-specific amplicon DNA concentration during the early (log-phase) cycles of the PCR process. In the fluorescence-based assay, the increase in DNA concentration is monitored by the use of a complementary, dual-fluorophore–labeled oligonucleotide probe (TaqMan probe), with the extent of fluorescence being directly proportional to DNA copy number. An overview of the analysis, for a carrier male with 192 CGG repeats, is presented in figure 1.

Figure  1
Outline of analysis of FMR1 mRNA levels, by use of automated, fluorescence-detection RT-PCR, with peripheral blood leukocytes obtained from a male premutation carrier with 192 CGG repeats. A, Logarithm of the relative reporter (FAM) fluorescence intensities ...

In the present study, we used FMR1- and GUS-based probes that possess a FAM reporter at the 5′ ends and a TAMRA quencher at the 3′ ends. During the extension phase of each PCR cycle, the FAM reporter is cleaved from the probe, as a result of the 5′ exonuclease activity of the polymerase, thereby resulting in an incremental increase in fluorescence intensity. The increases in fluorescence ΔRn (see the Material and Methods section, above) for each cycle are monitored in real time, by use of a 7700 Sequence Detector (PE Biosystems).

Relative mRNA levels are determined by a comparison of the cycle numbers for which the relative fluorescence signal, ΔRn, exceeds the threshold value (threshold cycle number is designated “Ct” (see the Material and Methods section above). The precise value of the threshold fluorescence is not critical, provided that it lies in the log-phase region. For example, for the carrier male (192 CGG repeats) represented in figure 1A, the ΔCt value of −1.750 [= Ct(FMR1carrier)-Ct(GUScarrier)], determined at ΔRn=0.01, differs, by only ∼1%, from the value determined at a 10-fold-higher fluorescence threshold level (ΔCt = −1.734, at ΔRn=0.1); note, however, that the threshold level must be held constant for any given comparison. For all of the samples analyzed in the current investigation, Ct values fall within the range of ∼25–30. Within this range, all samples display the expected log-linear dependence of Ct on total cellular RNA concentration.

Although blood samples are routinely processed <24 h after the blood draw, delays are sometimes encountered during transportation of the sample. We have investigated the influence of delayed RNA extraction, by isolating RNA, at various times, from aliquots of blood (stored at 4°C) obtained from a carrier female. For extractions performed 0, 2, 5, and 7 d postdraw, no reduction in either FMR1 or GUS mRNAs was observed. In addition, we generally see no significant variation in FMR1 mRNA levels for either normal individuals or individuals with premutations who have had blood drawn on more than one occasion.

Sets of replicate RT-PCR experiments for each blood sample are always accompanied by equivalent sets of reference measurements for the lymphoblastoid cell line (RPM 7666 [RP]; see the Material and Methods section above). The ΔCt values determined for each blood sample (ΔCt,sample) are normalized to the corresponding ΔCt values for the cell line (ΔCt,RP); thus, ΔΔCt,sample = ΔCt,sample − ΔCt,RP. This latter correction takes into account batch-to-batch variations in Taq polymerase, probe and/or primer efficiency, and variations in RT efficiency. This approach thus allows one to directly compare data generated in different experiments and reduces experimental uncertainty in relative mRNA levels. Failure to correct for such variation in RT-PCR experiments would severely degrade the precision of experimental comparisons of mRNA concentration. A second procedure—namely, an analysis of the concentration dependence of the Ct and ΔCt values (fig. 1B)—is also performed for each sample and cell line control. Strong variation in ΔCt may indicate problems with pipetting; such experiments can be repeated.

Significant Elevation of FMR1 mRNA Levels in Carrier Males

We have determined the relative FMR1 mRNA levels for 16 carrier males, by means of the fluorescence-based RT-PCR approach. Nine males possess alleles with a range of 55–100 repeats (designated as “low premutation” [lp]), and 7 possess alleles with a range of ∼100–200 repeats (designated as “high premutation” [hp]) (for allele sizes, see the Material and Methods section above). Longer exposures (3–5 d) of Southern blots were performed, to eliminate the possible presence of methylated alleles. Note that one member of this latter group (180+; see the Material and Methods section above) possesses a smear of allele sizes that extends from 180 repeats into the low full-mutation range (⩽280 repeats). Because all of his alleles were unmethylated, which is typical of a premutation, his mRNA data were included. Allele sizes were determined by conventional PCR analysis, which was augmented in some instances by Southern gel analysis. These groups were compared with a group of eight normal control individuals. The relative FMR1 mRNA levels for these three groups are presented in table 1; an analysis of the significance of the differences reported for these three groups is presented in table 2. The relative FMR1 mRNA levels for each individual are presented in figure 2.

Figure  2
Plot of FMR1 mRNA levels (relative to GUS mRNA [rel]), as a function of the CGG repeat number (n). nl, n<55, lp = 55 ⩽ n<100, and hp = 100 ⩽ n ⩽ 200. Horizontal lines represent the group (unweighted) means from ...
Table 1
Relative FMR1 mRNA Levels in Peripheral Blood Leukocytes of Males with Premutation Alleles
Table 2
Analysis of Significance for the Differences in FMR1 mRNA Levels among Three Allele Ranges (Nonpaired t-Test)

It is evident, from the data presented in table 2, that FMR1 mRNA levels are elevated in both subgroups of carrier males, relative to those in the normal control population. Moreover, there is a substantial (fivefold) elevation in the upper portion (hp) of the premutation range relative to the lower portion (lp). We recognize that the boundary between hp and lp (100 CGG repeats) is somewhat arbitrary in the absence of an underlying mechanistic basis; however, it does serve to underscore the growth in relative mRNA levels as one approaches the upper portion of the premutation range. Moreover, the current division yields a more conservative estimate for the increases in FMR1 mRNA than do regression analyses (see the Discussion section, below). Also significant is the fact that, in our limited group, we have not observed any hp males with FMR1 mRNA levels in the normal range. No FMR1 mRNA was detected, after 50 PCR cycles, for a male with an FMR1 deletion (Gu et al. 1994) that encompasses the 5′ portion of the FMR1 gene. Finally, the absence of any reduction in GUS mRNA levels in hp samples relative to the levels in nl samples (data not shown) rules out general suppression of GUS in carrier leukocytes.

mRNA Decay Experiments as Evidence That the Elevated FMR1 mRNA Levels Are Not a Result of Increased mRNA Stability

To determine whether the increased FMR1 mRNA levels are due to increased mRNA stability in carrier males, we have estimated the half-lives (t 1/2 values) for the decay of both FMR1 and GUS mRNAs from an EBV-transformed lymphoblastoid line MM, derived from the carrier male with 183 CGG repeats. DNA analysis of the transformed cell line indicated a small reduction in repeat number (to 160 repeats). A second line (AG09391, “AG”; NIA Cell Repository) was used as a normal control. After treatment of cells in culture with the RNA polymerase II inhibitor actinomycin D (10 μg/ml), mRNA levels were measured, over time, by use of the fluorescence-based RT-PCR method. The results of this analysis are presented in figure 3 and in table 3. The absence of any significant increase in FMR1 half-life in the premutation cell line, coupled with the approximately eightfold increase in mRNA level, indicates that increased stability cannot account for the higher mRNA level. This observation implies that the increase in FMR1 mRNA levels in the carrier males is the result of an increased rate of mRNA production.

Figure  3
Plots of the increase in Ct values for FMR1 and GUS mRNAs, as a function of time after treatment of cell lines with 10 μg actinomycin D/ml. FMR1 (blackened symbols) and GUS (unblackened symbols) probes indicate results for both a normal control ...
Table 3
Linear-Regression Analysis of Decay of FMR1 and GUS mRNAs in Lymphoblastoid Cell Lines after Treatment with Actinomycin D[Note]

Immunocytochemical Staining as Revealing a Reduced Percentage of FMRP(+) Lymphocytes in Carrier Males with hp Alleles (100–200 CGG Repeats)

Production of FMRP is generally thought to be normal in the premutation range (Devys et al. 1993; Verheij et al. 1993; Feng et al. 1995a; Kaufmann et al. 1999; Tassone et al. 1999), although we have recently observed reduced fractions of FMRP(+) cells in three males with premutations with hp alleles (Tassone et al., in press). In two of the males, reduced FMRP production was also observed by western blot analysis (Hagerman et al. 1996; D. Nelson, personal communication). We have used the immunocytochemical staining method (Willemsen et al. 1995, 1997) to determine the percentage of FMRP(+) lymphocytes from several additional hp carrier males. On the basis of previous estimates of the fraction of FMRP(+) lymphocytes from 33 males with nonexpanded alleles (89%±9%) (Willemsen et al. 1997; also see Tassone et al. 1999), males with allele sizes that have a range of 100–200 repeats (fig. 4) appear to have reduced FMRP-protein expression. It should be noted that positive staining for FMRP does not directly report cellular levels of FMRP; rather, it reflects the presence of a threshold level of FMRP in each lymphocyte. Therefore, the results presented in figure 4 are interpreted simply as a reflection of reduced FMRP production in hp-carrier males.

Figure  4
Plot of %FMRP(+) lymphocytes, as a function of CGG-repeat number. The standard error associated with each point is ∼4%. The horizontal dotted line represents 1 SD below the mean value of the %FMRP(+) lymphocytes for males with nonexpanded alleles ...


In the present study, we have demonstrated that males who are fragile-X carriers maintain substantially higher levels of FMR1 message in peripheral blood leukocytes than do normal control individuals. Moreover, the magnitude of the increase appears to be approximately fivefold greater for larger alleles (∼100–200 CGG repeats). Although our sample size is relatively limited, the increase is highly significant (table 2); we have not observed any males with alleles in that size range whose FMR1 mRNA levels are not increased. Therefore, the current data establish the presence of an FMR1-specific, (molecular) phenotypic abnormality in males with alleles in the premutation size range.

There have been several previous reports of FMR1 mRNA levels in the premutation range (Pieretti et al. 1991; Devys et al. 1993; Feng et al. 1995a, 1995b; Hmadcha et al. 1998). Pieretti et al. (1991) reported normal RNA levels in 10 females with premutations; however, no allele sizes were reported. Devys et al. (1993) reported normal mRNA levels in three lymphoblastoid lines with CGG repeats in the 90–100 range. Similar results were reported, by Feng et al. (1995a), for several lymphoblastoid lines with CGG repeats in the 84–104 range and, by Hmadcha et al. (1998), for a premutation male (leukocytes) with 101 repeats. These latter results are qualitatively consistent with the current results for the lp group, which show only a modest increase in mRNA level. Feng et al. (1995b) presented RNAse protection results for subcloned fibroblast cell lines from a single male with mosaicism. The relative FMR1 mRNA levels were not reported for individual alleles within the premutation range, although the average mRNA level for all expanded alleles was stated to be only slightly (albeit not significantly) elevated, compared with the levels seen for normal alleles. It should be noted, however, that the average was taken over both premutation (lp and hp) and full-mutation alleles. Thus, it is difficult to make an allele-specific comparison of the results of the present study and those of the study by Feng et al. (1995b). Finally, it should be noted that the results of previous studies did not report controls for concentration dependence; such experiments are difficult to perform with the use of RNAse protection or gel-based RT-PCR methods. The ability to perform multiple control experiments represents a singular advantage of the automated method—one that we have found to be necessary for the accurate quantification of relative FMR1 mRNA levels.

The current observations are particularly germane to the phenotypic status of males with premutations, since recognition of clinical involvement in this range is inconstant. Variation in reported prevalences of clinical involvement is due, in part, to the definition of involvement. In some studies, the definition is limited to cognitive impairment, whereas, in others, both cognitive and behavioral abnormalities are recognized (Dorn et al. 1994; Loesch et al. 1994; Hagerman et al. 1996; Franke et al. 1998; reviewed in Hagerman 1999). Furthermore, for at least some apparent carriers, cognitive and/or behavioral involvement may be the result of conditions unrelated to fragile-X syndrome (Hagerman 1999; Tassone et al., in press). Therefore, the observation of increased FMR1 mRNA levels in males with premutations, apparent even for alleles with <100 repeats, establishes the presence of an abnormal, FMR1-specific molecular phenotype in the premutation range. The present observations provide a possible molecular foundation for the clinical observations in the premutation range, although the mechanistic linkage is not yet understood.

It is also interesting to speculate as to whether the elevated mRNA levels may give rise to a clinical phenotype that is unique to the premutation range. Although the current work has focused on carrier males, there is evidence, in a subgroup of carrier females, of a unique phenotype that includes premature menopause and a higher rate of twinning, which are features unobserved in females with the full mutation (Turner et al. 1994; Murray et al. 1998; Allingham-Hawkins et al. 1999). This issue is currently under investigation; however, the levels of FMR1 mRNA in premutation females are likely to be influenced by the activation ratio.

An immunocytochemical approach has been used to identify the percentage of cell populations in which reduced levels of FMRP are expressed (Willemsen et al. 1995, 1997). Although this approach is not particularly sensitive to mild reductions in the FMRP level, we do observe, for the larger alleles within the premutation range, significant reductions in the percentages of FMRP(+) lymphocytes (fig. 4). In view of the elevated mRNA levels in this range, reduced FMRP expression is likely to reflect a translation defect that is beginning even in the premutation range. A translation defect has been reported in fibroblast cell lines (derived from one male with mosaicism) in the full-mutation range (Feng et al. 1995b).

The mechanistic basis for the elevated mRNA levels in carrier males is not understood at present. At the most basic level, such elevations could be the result of either increased transcriptional activity of the FMR1 gene or greater stability of the message itself (reduced rate of loss) or both. Elevated levels of mRNA have been reported for the myotonic dystrophy protein kinase gene (DMPK) of myotonic dystrophy, another trinucleotide-repeat–expansion disease that involves a CTG expansion in the 3′ UTR region (Sabourin et al. 1993), although this issue is controversial (Waring and Korneluk 1998, pp. 131–146). We have examined the question of mRNA stability, by measurement of the rates of decay of FMR1 mRNA in a lymphoblastoid cell line (160 CGG repeats) derived from a carrier male (183 CGG repeats) and from a normal control. That the decay half-lives for the normal and premutation lines are nearly equal (table 3) indicates that altered (increased) stability cannot account for the increased steady-state level of the FMR1 mRNA.

The absence of a significant increase in stability of the FMR1 mRNA suggests that transcriptional activity of the FMR1 gene is increased in carrier males. This possibility is particularly intriguing in light of the reduced number of FMRP(+) lymphocytes in at least some carrier males with large premutation alleles (see fig. 4). If defective translation of the FMR1 mRNA is occurring for premutation alleles, then an early manifestation of reduced translational efficiency may be feedback induction of the FMR1 gene, in an attempt to overcome a mild protein deficit. Thus, normal or near-normal protein levels in the high-premutation range may only be maintained through abnormally high levels of message, although no direct evidence for such a model currently exists.

In the presentation of our results, we have introduced a somewhat arbitrary division—at 100 CGG repeats—between two groups of carrier males. This distinction was introduced simply to illustrate the trend in message level. We have also performed first-order through third-order regression analyses of all data in figure 2. These three analyses yielded similar outcomes—namely, (a) relative mRNA levels that, for alleles of 200 repeats, are eightfold higher than the levels for alleles of 30 repeats and (b) coefficients of determination that are comparable (r21st=.76; r22d=.76; r23d=.78). Thus, the current division yields a more conservative, model-free estimate for the degree of elevation, and it is, in any case, more appropriate in the absence of any underlying mechanism.

Finally, we wish to emphasize the utility of the fluorescence-based RT-PCR approach. Real-time measurements obtained throughout the PCR process enable one to define a range of cycle numbers in which the response of the fluorescent signal is linearly proportional to the original concentration of RNA. This feature greatly improves both precision and accuracy of the determination of relative mRNA levels. In addition, the current investigation involved >1,000 individual PCR reactions for full quantification of relative mRNA levels (concentration-dependence, secondary corrections, repetitions, etc.); such numbers of PCR reactions are facilitated by the relative ease of the automated measurements.


This work was supported by funding (to P.J.H. and R.J.H.) from the Boory Family Fund and the Mulvey Family Fund, a FRAXA Research Foundation Fellowship (to F.T.), National Institutes of Health grants GM35305 (to P.J.H.) and HD36071 (to R.J.H.), and Maternal Child Health Bureau grant MCJ-089413 (to R.J.H.). The authors wish to thank Janine Mills for her help with oligonucleotide preparation.

Electronic-Database Information

Accession numbers and URLs for data in this article are as follows:

GenBank, http://www.ncbi.nlm.nih.gov/Genbank/GenbankOverview.html (for FMR1 gene [accession number L29074] and reference gene derived from GUS mRNA [accession number NM000181]
Online Mendelian Inheritance in Man (OMIM), http://www.ncbi.nlm.nih.gov/Omim/ (for fragile-X syndrome [MIM 309550])


Allingham-Hawkins DJ, Babul-Hirji R, Chitayat D, Holden JJ, Yang KT, Lee C, Hudson R, et al (1999) Fragile X premutation is a significant risk factor for premature ovarian failure: the International Collaborative POF in Fragile X study—preliminary data. Am J Med Genet 83:322–325 [PMC free article] [PubMed]
Baeyens DA, Cornett LE (1993) Transcriptional and posttranscriptional regulation of hepatic beta 2-adrenergic receptor gene expression during development. J Cell Physiol 157:70–76 [PubMed]
Brown WT, Houck GE Jr, Jeziorowska A, Levinson FN, Ding X, Dobkin C, Zhong N, et al (1993) Rapid fragile X carrier screening and prenatal diagnosis using a nonradioactive PCR test. JAMA 270:1569–1575 [PubMed]
de Vries BBA, Jansen CCAM, Duits AA, Verheij C, Willemsen R, van Hemel JO, van den Ouweland AMW, et al (1996) Variable FMR1 gene methylation of large expansions leads to variable phenotype in three males from one fragile X family. J Med Genet 33:1007–1010 [PMC free article] [PubMed]
Devys D, Lutz Y, Rouyer N, Bellocq J-P, Mandel J-L (1993) The FMR-1 protein is cytoplasmic, most abundant in neurons and appears normal in carriers of a fragile X premutation. Nat Genet 4:335–340 [PubMed]
Dorn MB, Mazzocco MMM, Hagerman RJ (1994) Behavioral and psychiatric disorders in adult male carriers of fragile X. J Am Acad Child Adolesc Psychiatry 33:256–264 [PubMed]
Feng Y, Lakkis L, Devys D, Warren ST (1995a) Quantitative comparison of FMR1 gene expression in normal and premutation alleles. Am J Hum Genet 56:106–113 [PMC free article] [PubMed]
Feng Y, Zhang F, Lokey LK, Chastain JL, Lakkis L, Eberhart D, Warren ST (1995b) Translational suppression by trinucleotide repeat expansion at FMR1. Science 268:731–734 [PubMed]
Franke P, Leboyer M, Gansicke M, Weiffenbach O, Biancalana V, Cornillet-Lefebre P, Croquette MF, et al (1998) Genotype-phenotype relationship in female carriers of the premutation and full mutation of FMR-1. Psychiatry Res 80:113–127 [PubMed]
Fu YH, Kuhl DP, Pizzuti A, Pieretti M, Sutcliffe JS, Richards S, Verkerk AJ, et al (1991) Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox. Cell 67:1047–1058 [PubMed]
Gibson UE, Heid CA, Williams PM (1996) A novel method for real time quantitative RT-PCR. Genome Res 6:995–1001 [PubMed]
Gu Yanghong, Lugenbeel KA, Vockley JG, Grody WW, Nelson DL (1994) A de novo deletion in FMR1 in a patient with developmental delay. Hum Mol Genet 3:1705–1706 [PubMed]
Hagerman RJ (1996) Physical and behavioral phenotype. In: Hagerman RJ, Cronister AC (eds) Fragile X syndrome: diagnosis, treatment, and research, 2d ed. Johns Hopkins University Press, Baltimore, pp 3–87
——— (1999) Fragile X syndrome. In: Neurodevelopmental disorders: diagnosis and treatment. Oxford University Press, Oxford, pp 61–132
Hagerman RJ, Staley LW, O'Conner R, Lugenbeel K, Nelson D, McLean SD, Taylor AK (1996) Learning-disabled males with a fragile X CGG expansion in the upper premutation size range. Pediatrics 97:122–126 [PubMed]
Heid CA, Stevens J, Livak KJ, Williams PM (1996) Real time quantitative PCR. Genome Res 6:986–994 [PubMed]
Hmadcha A, De Diego Y, Pintado E (1998) Assessment of FMR1 expression by reverse transcriptase-polymerase chain reaction of KH domains. J Lab Clin Med 131:170–173 [PubMed]
Kaufmann WE, Abrams MT, Chen W, Reiss AL (1999) Genotype, molecular phenotype, and cognitive phenotype: correlations in fragile X syndrome. Am J Med Genet 83:286–295 [PubMed]
Livak KJ, Flood SJ, Marmaro J, Giusti W, Deetz K (1995) Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods Appl 4:357–362 [PubMed]
Loesch DZ, Hay DA, Mulley J (1994) Transmitting males and carrier females in fragile X—revisited. Am J Med Genet 51:392–399 [PubMed]
Loesch DZ, Hay DA, Sutherland GR, Halliday J, Judge C, Webb GC (1987) Phenotypic variation in male-transmitted fragile X: genetic inferences. Am J Med Genet 27:401–417 [PubMed]
Mazzocco MM, Pennington BF, Hagerman RJ (1993) The neurocognitive phenotype of female carriers of fragile X: additional evidence for specificity. J Dev Behav Pediatr 14:328–335 [PubMed]
Murray A, Webb J, Grimley S, Conway G, Jacobs P (1998) Studies of FRAXA and FRAXE in women with premature ovarian failure. J Med Genet 35:637–640 [PMC free article] [PubMed]
Oberlè I, Rousseau F, Heitz D, Kretz C, Devys D, Hanauer A, Bouè J, et al (1991) Instability of a 550–base pair DNA segment and abnormal methylation in fragile X syndrome. Science 252:1097–1102 [PubMed]
Pieretti M, Zhang FP, Fu YH, Warren ST, Oostra BA, Caskey CT, Nelson DL (1991) Absence of expression of the FMR-1 gene in fragile X syndrome. Cell 66:817–822 [PubMed]
Reiss AL, Freund L, Abrams MT, Boehm C, Kazazian H (1993) Neurobehavioral effects of the fragile X premutation in adult women: a controlled study. Am J Hum Genet 52:884–894 [PMC free article] [PubMed]
Riddle JE, Cheema A, Sobesky WE, Gardner SC, Taylor AK, Pennington BF, Hagerman RJ (1998) Phenotypic involvement in females with the FMR1 gene mutation. Am J Ment Retard 102:590–601 [PubMed]
Rousseau F, Heitz D, Tarleton J, MacPherson J, Malmgren H, Dahl N, Barnicoat A, et al (1994) A multicenter study on genotype-phenotype correlations in the fragile X syndrome, using direct diagnosis with probe StB12.3: the first 2,253 cases. Am J Hum Genet 55:225–237 [PMC free article] [PubMed]
Sabourin LA, Mahadevan MS, Narang M, Lee DSC, Surh LC, Korneluk RG (1993) Effect of the myotonic dystrophy (DM) mutation on mRNA levels of the DM gene. Nat Genet 4:233–238 [PubMed]
Smeets HJM, Smits APT, Verheij CE, Theelen JPG, Willemsen R, van de Burgt I, Hoogeveen AT, et al (1995) Normal phenotype in two brothers with a full FMR1 mutation. Hum Mol Genet 4:2103–2108 [PubMed]
Smits A, Smeets D, Hamel B, Dreesen J, de Haan A, van Oost B (1994) Prediction of mental status in carriers of the fragile X mutation using CGG repeat length. Am J Med Genet 51:497–500 [PubMed]
Steyaert J, Borghgraef M, Legius E, Fryns J-P (1996) Molecular-intelligence correlations in young fragile X males with a mild CGG repeat expansion in the FMR1 gene. Am J Med Genet 64:274–277 [PubMed]
Tassone F, Hagerman RJ, Ilke DN, Dyer PN, Lampe M, Willemsen R, Oostra BA, et al (1999) FMRP expression as a potential prognostic indicator in fragile X syndrome. Am J Med Genet 84:250–261 [PubMed]
Tassone F, Hagerman RJ, Taylor AK, Mills JB, Wood S, Gane LW, Hagerman PJ. Clinical involvement and protein expression in individuals with the FMR1 premutation. Am J Med Genet (in press) [PubMed]
Taylor AK, Safanda JF, Fall MZ, Quince C, Lang KA, Hull CE, Carpenter I, et al (1994) Molecular predictors of cognitive involvement in female carriers of fragile X syndrome. JAMA 271:507–514 [PubMed]
Turner G, Robinson H, Wake S, Martin N (1994) Dizygous twinning and premature menopause in fragile X syndrome. Lancet 344:1500 [PubMed]
Verheij C, Bakker CE, de Graaff E, Keulemans J, Willemsen R, Verkerk AJ, Galjaard H, et al (1993) Characterization and localization of the FMR-1 gene product associated with fragile X syndrome. Nature 363:722–724 [PubMed]
Waring JD, Korneluk RG (1998) Genetic studies of the myotonic dystrophy CTG repeat. In: Wells RD, Warren ST (eds) Genetic instabilities and hereditary neurological diseases. Academic Press, San Diego
Willemsen R, Mohkamsing S, de Vries BBA, Devys D, van den Ouweland A, Mandel J-L, Galjaard H, et al (1995) Rapid antibody test for fragile X syndrome. Lancet 345:1147–1148 [PubMed]
Willemsen R, Smits A, Mohkamsing S, van Beerendonk H, de Haan A, de Vries B, van den Ouweland A, et al (1997) Rapid antibody test for diagnosing fragile X syndrome: a validation of the technique. Hum Genet 99:308–311 [PubMed]

Articles from American Journal of Human Genetics are provided here courtesy of American Society of Human Genetics
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence and PMC links.
  • MedGen
    Related information in MedGen
  • Nucleotide
    Primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • OMIM
    Genome Survey Sequence (GSS) nucleotide records reported in the current articles.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...