Early mitochondrial abnormalities in hippocampal neurons cultured from Fmr1 premutation mouse model

Abstract

Premutation CGG repeat expansions (55–200 CGG repeats; preCGG) within the fragile X mental retardation 1 (FMR1) gene cause fragile X-associated tremor/ataxia syndrome (FXTAS) in humans. Defects in neuronal morphology, early migration, and electrophysiological activity have been described despite appreciable expression of FMRP in a preCGG knock-in (KI) mouse model. The triggers that initiate and promote preCGG neuronal dysfunction are not understood. The absence of FMRP in a Drosophila model of fragile X syndrome was shown to increase axonal transport of mitochondria. Here we show that dissociated hippocampal neuronal culture from preCGG KI mice (average 170 CGG repeats) express 42.6% of the FMRP levels and 3.8-fold higher Fmr1 mRNA than that measured in wild type neurons at 4 days in vitro. PreCGG hippocampal neurons show abnormalities in the number, mobility, and metabolic function of mitochondria at this early stage of differentiation. PreCGG hippocampal neurites contained significantly fewer mitochondria and greatly reduced mitochondria mobility. In addition, preCGG neurons had higher rates of basal oxygen consumption and proton leak. We conclude that deficits in mitochondrial trafficking and metabolic function occur despite the presence of appreciable FMRP expression and may contribute to the early pathophysiology in preCGG carriers and to the risk of developing clinical FXTAS.

Introduction

Fragile X-associated tremor/ataxia syndrome (FXTAS) is a late-onset neurodegenerative disorder that can occur in carriers of a trinucleotide (CGG) expansion, (between 55–200 repeats; premutation), within the 5’-non-coding region of the fragile-X mental retardation 1 (FMR1) gene (Goodlin-Jones et al. 2004, Hessl et al. 2005). Larger expansions (> 200 repeats; full mutation) generally result in hypermethylation of the FMR1 gene and subsequent transcriptional silencing. The consequent absence of the FMR1 protein (FMRP), results in Fragile X syndrome (FXS), which is the most common inherited form of cognitive disability, and syndromic form of autism (Jacquemont et al. 2007, Hagerman et al. 2011). Premutation alleles have estimated frequencies of 1:250-810 males and 1:130-250 females (Hagerman 2008, Rodriguez-Revenga et al. 2009). These premutation carriers can display a range of clinical features that include behavioral and cognitive abnormalities in children (Goodlin-Jones et al. 2004, Hessl et al. 2005, Farzin et al. 2006, Hagerman 2006), primary ovarian insufficiency (POI) in about 20% of women (Amiri et al., 2008), and a late-onset neurodegenerative disorder, fragile X-associated tremor/ataxia syndrome (FXTAS) (Amiri et al. 2008, Brouwer et al. 2009, Tassone et al. 2007) in approximately 40% of males (Jacquemont et al 2004). Core clinical features of FXTAS include progressive gait ataxia and intention tremor, often accompanied by cognitive decline and executive dysfunction, peripheral neuropathy, dysautonomia, and Parkinsonism (Amiri et al. 2008, Brouwer et al. 2009, Berry-Kravis et al. 2007, Bourgeois et al. 2009).

Premutation carriers display a form of gene dysregulation that is quite distinct from the gene silencing observed with FXS, and which is manifest by substantially increased levels of FMR1 mRNA, and normal or moderately decreased levels of FMRP. The extent of this altered expression is a function of the size of the CGG-repeat expansion within the premutation range, with larger CGG-repeat expansions associated with higher levels of mRNA and lower levels of protein (Tassone et al. 2000). The absence of POI and FXTAS in full mutation patients implies that FMRP deficiency per se is not responsible for these premutation disorders. Indeed, evidence from both human and animal studies implicates a direct toxic gain-of-function of premutation CGG (preCGG) alleles due to an increase in the CGG-repeat-containing FMR1 mRNA (Willemsen et al. 2003, Brouwer et al. 2007, Tassone et al. 2000, Sellier et al. 2010). Consistent with this hypothesis, characteristic intranuclear inclusions found in neuronal and glial cells of FXTAS cases (Greco et al. 2006, Greco et al. 2002) have been demonstrated to contain FMR1 mRNA but not FMRP (Tassone et al. 2004). Moreover, the expanded CGG repeat-RNA is sufficient to form the intranuclear inclusions in both established neural cell lines and primary neural progenitor cells (Arocena et al. 2005), as well as in Purkinje neurons (Hashem et al. 2009). However, as reduced FMRP levels have been observed in the premutation in both mouse and human; we cannot exclude the possibility that a moderate reduction in FMRP might play a role in modulating some of the premutation phenotypes (Primerano et al. 2002, Tassone et al. 2000, Allen et al. 2004, Kenneson et al. 2001, Peprah et al. 2010, Hunsaker et al. 2010, Hunsaker et al. 2011, Qin et al. 2011).

To better understand the mechanistic basis for the premutation disorders, two knock-in (KI) mouse models have been developed to study the developmental onset and progressive neuropathology resulting from premutation CGG expansions. The models were created either by replacing the native 9–10 CGG repeat allele in the homologous Fmr1 gene with CGG expansions that vary from 100 to >300 (Berman & Willemsen 2009) or by serially ligating CGG-CCG repeats in exon 1 of the endogenous mouse Fmr1 gene (Entezam et al. 2007). Similar to human premutation carriers, premutation mice with large CGG-repeat expansions exhibit elevated Fmr1 mRNA and variable reductions in FMRP (Willemsen et al. 2003, Brouwer et al. 2007). The premutation mouse models display progressive deficits in processing spatial and temporal information, cognitive deficits (Hunsaker et al. 2010), motor deficits (Hunsaker et al. 2011), and hyperactivity (Qin et al. 2011). Ubiquitin-positive intranuclear inclusions, which are neuropathological hallmarks of FXTAS, are also found in premutation mouse neurons and astrocytes (Willemsen et al. 2003, Wenzel et al. 2010). Early defects in neuronal morphology (Chen et al. 2010), migration (Cunningham et al. 2011) as well as aberrant spontaneous Ca²⁺ oscillations and clustered burst firing (Cao et al. 2012) have been observed in studies of the preCGG mouse. The functional abnormalities observed in vitro appear to be related, at least in part, from abnormal development of inhibitory (GABAergic) and excitatory (glutamatergic) neuronal networks (Cao et al. 2012, D'Hulst et al. 2009).

Mitochondria generate the metabolic energy required for neuronal growth and therefore their distribution and dynamics are essential for proper synaptic transmission (Hollenbeck & Saxton 2005, Li et al. 2004). Mitochondrial intracellular transport is a dynamic process, which is greatly influenced by Ca²⁺-dependent processes (Sheng & Cai 2012). Importantly, fibroblasts cultured from human preCGG carriers demonstrate decreased complex III and V activities, increased production of reactive oxygen species (ROS), and decreased ATP production via oxidative phosphorylation (Ross-Inta et al. 2010, Giulivi et al. 2011). Decreased levels of a number of mitochondrial proteins were also reported in fibroblast and brain samples from individuals with FXTAS (Ross-Inta et al. 2010, Giulivi et al. 2011). Recently, the absence of FMRP has been reported to negatively influence the numbers and increase the transport of mitochondria in axons in a Drosophila model of FXS (Yao et al. 2011). Accordingly, in the present study, we sought to determine whether neurons cultured from preCGG KI mice exhibit early alterations in mitochondrial density, transport dynamics, and metabolic function. Alterations in the preCGG KI neurons may be similar to those observed in the FXS model; or entirely distinct, reflecting part of the mRNA gain-of-function mechanism believed to underlie FXTAS.

We demonstrate that preCGG hippocampal neuronal neurites have significantly decreased mitochondrial density and mobility, as well as aberrant metabolic function. These defects contrast with findings in the FXS model, and are apparent as early as 4 days in vitro (DIV). Metabolic impairments include higher basal oxygen consumption, ATP production, and proton leakage compared to wild type (WT). We conclude that deficits in mitochondrial trafficking and metabolic function occur as a consequence of overexpression of Fmr1 mRNA, although the moderate reductions of FMRP may play a secondary role. We suggest that the observed mitochondrial dysregulation may contribute to the late-onset pathophysiology observed in premutation carriers and to the risk of developing clinical FXTAS.

Methods

Animals

Experiments were conducted using the expanded CGG trinucleotide repeat (average 170 repeats) knock-in mouse model of the fragile X premutation. The generation of these mice has been described previously (Willemsen et al. 2003). Throughout, male hemizygous premutation and WT mice in the C57BL/6J are used for paired cultures of hippocampal neurons. Breeding a female mouse homozygous for the expanded allele with a male lacking the mutation derived male mice hemizygous for the premutation. Congenic WT male mice were bred with WT females in parallel timed pregnancies. Mice were housed in 12/12-h light-dark cycle with unrestricted access to food and water. All experiments were conducted in compliance with the NIH Guide for Care and Use of Laboratory Animals and were approved by the institutional Animal Care and Use Committee of the University of California at Davis.

Genotyping

DNA was extracted from mouse-tail snips as previously described (Chen et al., 2010). The number of CGG repeats were determined by PCR using the Expanded High Fidelity Plus PCR System (Roche Diagnostics) using forward and reverse primers previously reported (Chen et al., 2012). The DNA bands were separated using agarose gels and stained with ethidium bromide to identify their sizes.

Primary Hippocampal Cultures

Cultures of dissociated hippocampal neurons were prepared by dissection of hippocampi from P1 postnatal mice. Hippocampi were dissected into ice cold Hanks balanced salt solution (HBSS, Ca^2+/Mg²⁺ free, Invitrogen), and then incubated in HBSS containing trypsin (0.03%) at 37°C for 15 minutes. Hippocampal tissue was washed three times in warm HBSS, and then triturated with a fire-polished glass pipet. Undissociated tissue fragments were discarded and the remaining cell-containing supernatant was spun down (1100rpm for 3 minutes). The cells were resuspended and plated at a density of 10⁵ per glass bottom culture dish (MatTek Corporation) coated with poly-L-lysine (Peptides International) in Neurobasal medium (Invitrogen) containing NS21 supplement (Chen et al. 2008), 0.5mM glutamine, and 5% fetal bovine serum (FBS). For measurement of OCR, the neurons were plated onto XF24 cell culture microplates (Seahorse Bioscience, North Billerica, MA) at a density of 7x10⁴ per well. For measuring FMRP and FMR 1 levels, the neurons were plated on 6-well plates at a density of 2x10⁶ per well. Four hours after plating, the medium was replaced with serum-free Neurobasal medium containing NS21 supplement and 0.5mM glutamine. At 2 DIV, the medium was replaced with medium containing fluorodeoxyuridine (FUDR, 30µM) and uridine (60µM), to limit the growth of astrocytes. Cells were maintained at 37°C in a humidified environment of ambient air/ 5% CO₂.

Mitotracker Staining and Imaging

After hippocampal neurons were allowed to grow to 4 DIV, culture medium was removed and dishes were gently washed with warm HBSS. Neurons were then incubated in staining solution containing 100nM Mitotracker Red CMXRos (Invitrogen, Carlsbad, CA) in HBSS for 20 minutes at 37°C. HBSS containing dishes were maintained at 37°C while imaging using a micro-incubator (PDMI-2, Warner Instruments). Time-lapse images of neurons were acquired every second for 2 minutes with a 100x objective on an Olympus Ix71 inverted microscope (Olympus). The sequence of images was captured using EasyRatioPro software (Photon Technologies International, Birmingham, NJ). Images were analyzed using Image J (NIH) software to determine mitochondrial density in proximal and distal neurites, as well as number of mitochondria that were mobile or immobile within the entire neurite. Mobile mitochondria were scored as such if they could be clearly resolved traveling at least 1 µm within the 2 minute imaging session. Highly mobile mitochondria, which were much less frequently observed, were defined as such if they travelled a distance of 5µm or greater within the imaging session. All the images were taken and scored with the investigator blinded to the identity of genotype.

Measurements of oxygen consumption rate

A Seahorse Bioscience XF24 Extracellular Flux Analyzer (Seahorse Bioscience, North Billerica, MA) was used to measure the rate change of dissolved O₂ in medium immediately surrounding the neurons cultured in a XF24 cell culture microplate (Seahorse Bioscience, North Billerica, MA). After growing for 4 days, the growth medium was removed and replaced with 675 µl of assay medium pre-warmed to 37°C, composed of Dulbecco’s Modified Eagle’s Medium without bicarbonate and phenol red (Sigma, St. Louis, MO, Catalog# D5030) supplemented with 31 mM NaCl, 25 mM glucose, 1mM sodium pyruvate and 2 mM glutamax (pH 7.4). Measurements of oxygen consumption rate (OCR) were performed after equilibration in assay medium for about 30 min. Briefly, the Seahorse analyzer uses a cartridge with 24 optical fluorescent O₂ and pH sensors that are embedded in a sterile disposable cartridge, 1 for each well. Before each rate measurement, the plungers mix assay media in each well to allow the oxygen partial pressure to reach equilibrium. For measurements of the rates, the plungers gently descend into the wells, forming a chamber that entraps the cells in an approximately 7 µl volume. The O₂ concentration is periodically measured and OCR is obtained from the slopes of concentration change vs. time. After the rate measurements, the plungers ascend and the plate is gently agitated to re-equilibrate the medium. OCR is reported in the unit of picomoles/ minute /µg protein. Baseline rates are measured three times. The testing chemicals are preloaded in the reagent delivery chambers of the sensor cartridge and then sequentially pneumatically injected into the wells to reach the desired final working concentration. The non-mitochondria OCR (rotenone insensitive) were subtracted and the averages of three baseline rates and the test rates were used for data analysis.

Western blot

The sample preparation for western blot was performed as described previously (Cao et al. 2007). Equal amounts (20 µg) of samples were loaded onto a 10% SDS-PAGE gel and transferred to a nitrocellulose membrane by electroblotting. The membranes were blocked with 5% non-skimmed milk in PBS buffer+0.1% Tween-20 for 1.5-2 h at room temperature. After blocking, membranes were incubated overnight at 4°C in primary antibody dilution (anti-FMRP, 1:20,000 and anti-β-actin, 1:20,000). The blots were washed and incubated with the IRDye (800CW or 700CW)-labeled secondary antibody (1:10,000) for 1 h at room temperature. After washing with 0.1% Tween in PBS for 5 times, the membrane was scanned with the LI-COR Odyssey Infrared Imaging System (LI-COR Biotechnology, Lincoln, NE, USA). The densitometry was performed using LI-COR Odyssey Infrared Imaging System application software 2.1.

Quantitative measurements of Fmr1 mRNA levels

Total RNA from primary hippocampal cultures was isolated by standard method (Trizol, Ambion Inc., Austin, TX, USA). Precise estimates of Fmr1 mRNA levels in total RNA were obtained by real time PCR. Details of the method and its application to the study of Fmr1 mRNAs are described as previously (Tassone et al. 2000). The reference gene was β-glucoronidase (GUS). Primers and probes were mouse specifics (ABI Assay on Demand, Foster City , CA). The analysis was repeated for 3 different RNA concentrations, in duplicate, and incorporated standards for each determination to compensate for any changes in reaction efficiency.

Data analysis

Graphing and statistical analysis were performed using GraphPad Prism software (Version 5.0, GraphPad Software Inc., San Diego, CA). Statistical significance between different groups was calculated using Student’s t test or by an ANOVA and, where appropriate, a Dunnett's Multiple Comparison Test. The p values below 0.05% were considered significant.

Results

Western blotting with a chicken monoclonal antibody (Iwahashi et al. 2009) detects FMRP in the lysate of neuronal cultures with the major band at 72 kDa (Fig. 1A), a band absent in brain lysates generated from the FMRP knock out mouse, a model of FXS (data not shown). When normalized to the intensity of β-actin, hippocampal neurons cultured from preCGG mice (mean expansion, 170 CGG repeats) express 3.8-fold higher Fmr1 mRNA levels than observed in 4 DIV WT neurons, whereas FMRP levels are moderately reduced (43±1% of WT levels) (Fig. 1B and 1C)

Figure 1. Premutation cultures express higher levels of Fmr1 mRNAs with decreased FMRP proteins compared with WT paired cultures.

(A) Representative western blot in paired cultures of WT and preCGG hippocampal neurons at 4 DIV. (B) Quantification of FMRP expression level relative to β-actin in paired WT and preCGG neuronal cultures at 4 DIV. Data were pooled from two independent cultures. (C) Fmr 1 mRNA comparison between WT and preCGG paired neuronal cultures at 4 DIV. Data were pooled from two independent cultures days performed in duplicate. **, p<0.01, preCGG vs. WT.

Because the distribution of mitochondria along the dendritic processes of neurons is essential for maintaining proper cellular function (Hollenbeck & Saxton 2005, Li et al. 2004), we measured the density and dynamics of mitochondria in neurites of hippocampal neurons. The 4 DIV time point was specifically chosen to permit quantification of mitochondrial density and dynamics at an early stage of neuronal developmental. Densities were evaluated in both the proximal (within 25µm of the cell soma) and distal (farther than 25µm from the cell soma) portions of the neurite. The density of mitochondria was significantly reduced in preCGG proximal neurites (3.67±0.32 preCGG vs. 4.88±0.29 WT, #mitochondria/10µm, p=0.01). The distal neurites showed a trend toward fewer mitochondria as well, but the reduction was not statistically significant (p=0.338) (Fig. 2). The movement of mitochondria was also investigated by the acquisition of time-lapse images, which allowed for the tracking of individual mitochondria over the course of the imaging session (Fig. 3 A and B). The mitochondria that maintained their position throughout the imaging session were referred to as “immobile”, whereas mitochondria that traveled at least 1 µm within the 2 minute imaging session were classified as “mobile”; mitochondria that traveled a distance of 5 µm or greater within the imaging session were termed “highly mobile”. The “highly mobile” mitochondria were also evaluated for the direction of their movement. Direction was scored as anterograde (ANTR), retrograde (RETR), or moving in both directions (Both).

Figure 2. Decreased mitochondrial density in premutation hippocampal neurites.

(A,B) Representative images of WT and preCGG 4 DIV hippocampal neurons labeled with Mitotracker dye. Below whole cell images are higher magnification images of proximal (within 25µm of soma) and distal (farther than 25µm from soma) neurites. (C) Number of mitochondria was decreased in preCGG proximal neurites by 25% (3.67±0.32 preCGG vs. 4.88±0.29 WT, p=0.01), WT n=98 neurites from 53 cells, preCGG n=53 neurites from 33 cells. (D) The number of mitochondria in distal neurites showed a trend toward fewer mitochondria in preCGG neurites, but this was not statistically significant (1.68±0.25 preCGG vs. 2.12±0.32 WT, p=0.338). WT n=66 neurites from 49 cells, preCGG n=41 neurites from 24 cells. Scale bars represent 5µm.

Figure 3. Decreased mitochondrial motility in premutation neurons.

(A, B) Representative images of WT and preCGG 4 DIV hippocampal neurons labeled with Mitotracker dye. Below whole cell images are higher magnification time lapse images of mitochondrial movement in neurites. Labeled examples are shown of immobile (white arrowhead) as well as anterogradely (green arrowhead) and retrogradely (blue arrowhead) moving mitochondria. (C) Number of mobile mitochondria was decreased in preCGG neurons by 48% compared to WT (1.21±0.15 preCGG vs. 2.36±0.15 WT, p<0.01). (D) Number of immobile mitochondria was not significantly different between preCGG and WT neurons (1.91±0.17 preCGG vs. 1.76±0.12 WT, p=0.47). (E) Number of highly mobile mitochondria, was decreased in preCGG neurons by 66% compared to WT (0.16±0.04 preCGG vs.0.47±0.06 WT, p<0.01) (F) The direction of travel of highly mobile mitochondria was not significantly different between preCGG and WT neurons. Both showed similar percentage of mitochondria moving anterograde (ANTR), retrograde (RETR), or both directions (Both). ANTR: 0.20±0.09 preCGG vs. 0.23±0.04 WT, p=0.693. RETR: 0.74±0.09 preCGG vs. 0.63±0.05 WT, p=0.276. BOTH: 0.06±0.03 preCGG vs. 0.13±0.03 WT, p=0.239. WT n=98 neurites from 53 cells, preCGG n=53 neurites from 33 cells. Scale bars represent 5µm.

A strong deficit in the dynamics of mitochondrial movement was apparent in the neurites from preCGG neurons compared to WT neurons. Although the number of immobile mitochondria was not significantly different between the two genotypes, the number of mobile mitochondria in preCGG neurons was significantly reduced (1.21±0.15 preCGG vs. 2.36±0.15 WT, #mitochondria/10µm, p<0.01) (Fig. 3C and D). The discrepancy in mobility between preCGG and WT neurons was especially apparent in the number of mitochondria that traveled a great distance. Highly mobile mitochondria numbers were significantly reduced (0.16±0.04 preCGG vs.0.47±0.06 WT, #mitochondria/10µm, p<0.01) in the neurites of preCGG neurons compared to those of WT (Fig. 3E). Similar results were obtained when the data were normalized relative to total number of mitochondria (mobile+immobile) for each of the proximal and distal dendrites (Supporting Fig. 1). The percentage of highly mobile mitochondria that moved in an ANTR, RETR, or in Both directions, was not significantly different between genotypes (Fig. 3F). Highly mobile mitochondria in both preCGG and WT neurons more commonly moved in the retrograde direction, with a smaller percentage of mitochondria moving ANTR or in Both directions (Fig. 3A, B and F).

The oxygen consumption rate (OCR) of mitochondria from intact 4 DIV hippocampal neurons was measured in real time, using a Seahorse Bioscience XF24 extracellular flux analyzer (Seahorse Bioscience, North Billerica, MA). Hippocampal neurons from premutation mice displayed a 23.0±2.9% higher (p<0.01) basal OCR compared to their WT counterpart (Fig. 4A and B). The oligomycin-sensitive OCR, which is related to ATP production, was 17.1±2.8% higher (p<0.01) in preCGG neurons than that measured with WT neurons, suggesting that preCGG neurons produced more ATP than WT. The oligomycin-insensitive OCR, which reflects proton leakage, was 43.2 ±4.8% higher (p<0.01) than that of WT neurons. The maximal OCR in both genotypes was also evaluated in the presence of the uncoupling agent FCCP (1µM). FCCP stimulated OCRs were comparable to respective basal levels in both genotypes. This suggests that in our dissociated neuronal system, the spare respiratory capacity is small. This is consistent with the previous report in which FCCP (1µM) treatment led to little stimulation in a hippocampal slice preparation (Schuh et al. 2011). However, the maximal OCR was higher in preCGG compared to WT neurons (17.1±5.5% higher than WT, p<0.05).

Figure 4. Bioenergetics in WT and preCGG neurons.

(A) Time-response relationships for oxygen consumption before and after addition of oligomycin (1µM), FCCP (1µM) and rotenone (1µM) in both genotypes. The non-mitochondrial OCR determined by addition of rotenone (1µM) was subtracted. (B) Quantification of basal, ATP production, proton leak, and maximal OCR in WT and preCGG 4DIV hippocampal neurons. In preCGG neurons, the basal OCR was 23% higher (16.71±0.40 pmole/min/µg protein, n=36) compared to WT neurons (13.58±0.68 pmole/min/µg protein, n=30, p<0.01). ATP production in preCGG neurons was 17% higher (12.27±0.29 pmole/min/µg protein, n=30) than observed in WT neurons (10.48±0.48 pmole/min/µg protein, n=36, p<0.01). Proton leak was 43% higher (4.44±0.15 pmole/min/µg protein, n=30) compared to WT neurons (3.1±0.28 pmole/min/µg protein, n=36, p<0.01). Maximal OCR was 17% higher in preCGG neurons (14.67±0.69 pmole/min/µg protein, n=30) than observed for WT neurons (12.53±0.58 pmole/min/µg protein, n=36, p<0.05).

Discussion

Males with FMR1 premutation have reduced hippocampal activation during memory recall tasks, presumably because of dysfunction in the posterior hippocampus, which also correlated with psychological symptoms (Koldewyn et al., 2008). Anxiety-related problems are also common both prior to and after the onset of FXTAS, and appear to be related to hippocampal changes (Adams et al., 2009). High levels of FMR1 mRNA have been demonstrated in the hippocampus of human patients (Tassone et al., 2004) where ubiquitin positive inclusions appear in 10–40% of the hippocampal neurons (Greco et al., 2002; Greco et al., 2006), findings modeled by the knock-in premutation mouse used in the present study (Wenzel et al., 2011).

In this study, we have shown abnormalities in both mitochondrial trafficking and mitochondrial bioenergetics in hippocampal neurons from a mouse model of FXTAS as early as 4 DIV. The density and intracellular trafficking of mitochondria in neurites were significantly reduced in 4 DIV preCGG mouse hippocampal neurons. The preCGG neurons also displayed higher basal OCR, proton leakage, and higher ATP production. The findings presented here are of particular interest considering the increasing evidence linking mitochondrial defects and neurodegenerative disorders (Pickrell & Moraes 2010). These mitochondrial deficits in preCGG neurons occur in the presence of elevated Fmr1 mRNA and moderately reduced FMRP levels, in accord with previous findings in brain lysates and neuronal cultures from KI mice expressing preCGG repeats (Brouwer et al. 2007, Brouwer et al. 2008a, Brouwer et al. 2008b), and with findings from human premutation carriers and FXTAS patients (Tassone et al. 2000, Tassone et al. 2004). Therefore, the presence of both significantly elevated Fmr1 mRNA and intermediate levels of FMRP reflects the human preCGG molecular phenotype (Tassone et al. 2000, Tassone et al. 2004), thus distinguishing preCGG hippocampal neurons from those of FXS models. These abnormalities in the density and mobility of mitochondria may contribute to the altered pattern of neuronal complexity reported previously using the same in vitro preCGG model, where the dendritic complexity in preCGG hippocampal neurons was decreased as early as 7 DIV (Chen et al. 2010). Moreover, such mitochondrial deficits may contribute to migration defects (Cunningham et al. 2011), reduced dendritic branching, and altered spine length observed in vivo in the preCGG mouse neocortex (Berman 2012).

Recently, it was reported in a Drosophila model of FXS (Yao et al. 2011) that the numbers of axonal mitochondria were inversely correlated with FMRP level. Observed increases in the number of mitochondria were caused specifically by the loss of FMRP, and neuronal over-expression of FMRP led to lowered mitochondrial numbers. These observations stand in direct contrast to the observed decreases in mitochondrial number and dynamics in the preCGG KI mouse; suggesting that Fmr1 mRNA toxicity resulting from the ˜4-fold increased RNA levels, not the moderately reduced FMRP levels, is likely responsible for the current mitochondrial phenotype in premutation mice. This conclusion is supported by the recent observation that the variation of FMRP among individuals in the general population (normal FMR1 alleles) is greater than four-fold, despite the absence of any clinical features of fragile X premutation-associated disorders (Lessard et al. 2011, Iwahashi et al. 2009).

In summary, we demonstrate that preCGG hippocampal neurons show abnormalities in the number, mobility, and metabolic function of mitochondria. Premutation hippocampal neurons displayed higher basal oxygen consumption, ATP production, as well as higher proton leakage. The deficits in mitochondrial trafficking and metabolic function may contribute to pathophysiology in premutation carriers and may constitute a risk factor of developing clinical FXTAS.

Abbreviations

FXTAS

Fragile X-associated tremor/ataxia syndrome

FMR1

Fragile-X mental retardation 1 gene

FMRP

Fragile-X mental retardation protein

OCR

Oxygen consumption rate

DIV

Days in vitro