A polyalanine antibody for the diagnosis of oculopharyngeal muscular dystrophy and polyalanine-related diseases

Eighteen severe human diseases have so far been associated with trinucleotide repeat expansions coding for either polyalanine (encoded by a GCN repeat tract) or polyglutamine (encoded by a CAG repeat tract). Among them, oculopharyngeal muscular dystrophy (OPMD), spinocerebellar ataxia type-3 (SCA3), and Huntington’s disease (HD) are late-onset autosomal-dominant disorders characterized by the presence of intranuclear inclusions (INIs). We have previously identified the OPMD causative mutation as a small expansion (from 2 in normal to 7 in disease) of a GCG repeat tract in the gene. In addition, -1 ribosomal frameshifting has been reported to PABPN1 occur in expanded CAG repeat tracts in the (SCA3) and (HD) ATXN3 HTT genes, resulting in the translation of a hybrid CAG/GCA repeat tract and the production of a polyalanine-containing peptide. Previous studies on OPMD suggest that polyalanine-induced toxicity is very sensitive to the dosage and length of the alanine stretch. Here we report the characterization of a polyclonal antibody that selectively recognizes pathological expansions of polyalanine in PABPN1. Furthermore, our antibody also detects the presence of alanine proteins in INIs of SCA3 and HD patient samples.


Introduction
Expansion of trinucleotide repeated sequences within the coding regions of distinct genes has been established to cause a number of severe human diseases [for reviews, see 1-4.The expanded coding triplet sequences so far implicated in disease are either CAG repeats, which translate into polyglutamine tracts, or GCN repeats, which encode for polyalanine stretches.The former were shown to cause at least nine distinct adult-onset neurodegenerative conditions such as Huntington's disease (HD), spinal bulbar muscular atrophy (SBMA), spinocerebellar ataxia (SCA) types 1, 2, 3, 6, 7 and 17 and dentatorubral-pallidoluysian atrophy (DRPLA) 2,4 ; whereas polyalanine expansions have been implicated in oculopharyngeal muscular dystrophy (OPMD) and in numerous developmental disorders 1,3 .
The so-called "polyglutamine" diseases share a number of genetic and molecular events/features; among which are their mutation process (dynamic expansion of their respective CAG repeat), intergenerational repeat instability, anticipation, and a disease course that is progressive following a late onset (10 to 20 years) 5 .For these reasons, it has been proposed that expanded CAG repeat tract diseases also share, to some extent, a common pathogenic mechanism, whereas the phenotypic variability of each disease would reflect the intrinsic properties the protein in which the repeat resides and the cellular environment where the affected protein is expressed.
Mutant protein aggregation, often in the form of intranuclear inclusions (INIs), is a hallmark of these disorders and INIs were at first believed to be key contributors of the toxicity leading to the neurodegeneration associated with pathological repeat expansions.However, some evidence now suggests that the soluble form of these mutant proteins may be more toxic than their insoluble counterparts found in INIs 6 , and aggregation might actually protect cells from the toxic insults inherent to misfolded soluble forms of the mutant proteins [7][8][9] .Finally, for each of the polyglutamine diseases, the repeat tract expansion mutation affects specific populations of neuronal cells, despite ubiquitous expression of the mutant proteins [for a review see 4].This could either be due to native properties of each protein, or could be explained by novel interactions of the mutant species with other cellular factors, specific for each cell type.
Fibrillar INIs have also been described in OPMD.OPMD is mainly a disease of skeletal muscle, with some reports suggesting partial neurological involvement [10][11][12] .The disease is caused by the expansion of a short polyalanine repeat in the polyadenylate binding protein nuclear 1 (PABPN1) gene 13 .The INIs can be typically found in the nuclei of affected muscle fibers, but an OPMD transgenic mouse model with expression of PABPN1 with a 13 alanine tract has INIs in muscle cells as well as neuronal cells of the spinal cord and cerebellum, which implies that the polyalanine expansion within PABPN1 can also be toxic to nervous tissues 14 .This finding was confirmed in post-mortem cerebellar samples of an OPMD patient 14 .
A -1 base shift in reading frame within an expanded CAG repeat tract would lead to translation of protein polyalanine stretch from the GCA reading frame.Using (SCA3) as a model, we have previously postulated that (i) translational frameshifts in large CAG stretches result in a new reading frame with formation of a hybrid protein containing a mixed polyglutamine/polyalanine tract, (ii) the resultant polyalanine polymers aggregate, and (iii) polyalanine-containing peptides are toxic to cells.We have demonstrated the presence of -1 frameshifting events in cells cultured in vitro, in transgenic Drosophila lines, in mouse organotypic cultures, as well as in pontine neurons from SCA3 human brain autopsy material [15][16][17] .In cell culture, switching from polyglutamine to polyalanine seems to be CAG lengthdependent and to occur during translation 17 .More importantly, we have established a direct correlation between the -1 translational frameshifting events (which we will henceforth refer to as frameshifting) and cellular toxicity using a stably transfected cell model.In addition, treating cells with specific antibiotics that are known to either enhance (e.g.sparsomycin) or inhibit (e.g.anisomycin) frameshifting can modulate the frequency of frameshifting events and the associated toxicity.Sparsomycin favours frameshifting by slowing the peptidyl transfer, allowing time for transfer RNA (tRNA) realignment, whereas anisomycin inhibits the accommodation of the frameshifted tRNA to the codon in the -1 frame 17,18 .Finally, the substitution of the expanded CAG repeat in the ATXN3 cDNA by an expanded CAA repeat of similar length (which also encodes a polyglutamine stretch in the main frame but will not produce polyalaninecontaining peptides if a frameshift occurs) abolishes the toxicity of the transgene 16,17 .These findings suggest a pathogenic role for the -1 frameshifted protein species in SCA3, and possibly in other expanded CAG repeat tract diseases.
Frameshifting has recently been shown to occur within the CAG repeats of the hungtingtin gene (HTT) 19,20 , but a clear link has not been established between these events and toxic outcomes in vivo.The question thus remains as to the biological relevance of frameshifting within large CAG repeats of HD patients.
Polyalanine toxicity may underlie a number of severe human disorders.It would therefore be useful to develop a screening tool that would allow the detection of alanine polymers at a size above pathological threshold.A similar tool was developed for the polyglutamine expansion diseases in the form of an antibody directed against polyglutamine 21 , as well as for the expanded GGGGCC repeat in C9orf72 (amyotrophic lateral sclerosis and frontotemporal dementia; ALS/FTD) in the form of antibodies generated against the dipeptide products which were observed to arise from the pathological expansion of the GGGGCC hexonucleotide, antiC9RANT 22 .Here we report the characterization of a polyclonal polyalanine-targeting antibody, antibody 4340 (Ab4340), that selectively recognizes pathological expansions of the protein PABPN1 implicated in OPMD, as well as alaninecontaining INIs in SCA3 and HD patient samples. 1 mg of purified GST-polyala was injected subcutaneously in two rabbits (following the collection of a pre-immune blood samples to have a pre-immune antiserum).Four immune response boosts were subsequently made (4 weeks apart) using the same dose.Bleeds were collected every month for 6 months (8.5 ml/kg).Animals were to be euthanized if the titer was insufficient 6 months after the initial immunization.Through the entire process care was taken to make sure the animals suffering was minimal.Animals were observed 15 minutes post-injections, and also on a daily for responses at the injection sites and signs of overall distress.When blood samples were collected the animals were tranquilized by injecting xylazine (5mg/kg) and ketamine (20-35 mg/kg) intramuscularly.At the end of the antisera production, the animals were euthanized by exsanguination under general anaesthesia and the antisera were prepared for storage at -80°C.Animals were obtained from Charles River Laboratories and maintained (one per cage) at the Comparative Medicine and Animal Resources Center (McIntyre, McGill University).

Methods
The specific and further purification of the antisera against polyalanine was established using Western blots loaded with MBP-polyalanine fusion proteins transferred on PVDF membranes (Sigma #P2563).The use of this PVDF transfer and immobilization approach insured two aspects: First, the antisera were detecting the polyalanine regions of the fusion peptide used for immunization and not the GST region.Second, the region of the PVDF membranes corresponding to the polyalanine signal could be cut out to elute antibodies that specifically recognized the polyalanine.Aliquots of PVDF eluted polyalanine antibodies were used in the immunodetections of material expressing polyalanine (and control).

Transgenic Drosophila lines
Stocks used in this study were previously described 16 .Full-length ATXN3 cDNAs bearing wt CAG

Drosophila immunohistochemistry
Adult flies were decapitated (3 days post eclosion), with heads immediately placed in Tissue-Tek (Sakura) and on dry ice to freeze.Ten micron sections were obtained by cryosectioning on a Leica CM3050S cryostat, dried for 30 min at room temperature, and then fixed in 4% paraformaldehyde (PFA) for 15 min.Permeabilization, blocking, and incubation in primary antibodies [mouse monoclonal anti-SCA3 antibody (1:1,000, Chemicon), rabbit polyclonal antibody 4340 (1:500)] was performed as described above.Sections were then incubated with the appropriate fluorescent secondary antibodies for 1 hour (anti-mouse or anti-rabbit fluorescent tagged secondary antibodies, 1:500, Alexafluor) and mounted with Mowiol.Visualization of immunofluorescence staining was carried out on a Leica CTR6000 fluorescence microscope.

Results
Generation of a polyclonal antibody sensitive to polyalanine at the pathological threshold in OPMD We generated an antibody (4340) against a 19-mer peptide composed of 18 alanines followed by a glycine.In order to evaluate the usefulness of this antibody, it was critical to determine the number of alanine repeats it could detect.Using OPMD as the disease model and Western blot immunodetection as a first assay, our analyses revealed that the antibody was able to produce a strong signal from whole protein lysates prepared from HeLa cells that transiently expressed a vector encoding a GFP-tagged hPABPN1 cDNA bearing alanine repeat lengths of 13, 17, 30, and 40 (Figure 1a).In contrast, only a weak signal was observed from lysates prepared from cells expressing the same cDNA if it encoded a 10-alanine repeat, and no signal could be observed from lysates prepared from cells that were either expressing a cDNA with no polyalanine tract (0-alanine) or that were untransfected (Figure 1a).To test whether the signals detected were the putative GFP-hPABPN1-alanine proteins, we probed the same samples with an antibody against GFP, and observed corresponding bands at ~75 kDa (Figure 1b).This suggests that the ~75 kDa bands detected by both antibodies correspond to the same protein, whereas the ~55 kDa bands detected by our antibody alone appear to be an unspecific contaminating signal.
HeLa cells that were transfected with the same expression vectors which were used for the Western blot analyses were also used to test the sensitivity of Ab4340 to polyalanine tracts through an in vitro immunofluorescence assay.The fusion of an N-terminal GFP-tag to each construct made it possible to visualize protein expression using fluorescence microscopy.Intranuclear expression with a strong GFP signal was observed across all constructs (Figure 1c to 1h).Using the 4340 antibody, we were able to specifically target the alanine-containing proteins and detect their expression in cells transfected with the expression vectors that encoded repeat lengths of 10, 13,17, 30, and 40 alanines (Figure 1d to 1h).No alanine signal was detected following the expression of the 0-alanine construct (Figure 1c).The alaninecontaining protein appears in aggregates, colocalizing with the GFP-expressing INIs (Figure 1d to 1h).These findings suggest that Ab4340 is more sensitive in detecting alanine expansions using an immunofluorescence assay (immunocytochemistry) than Western blot immunodetection.

Differentiation can be made between OPMD and control patient samples
The results from our Western blot immunodetections based on HeLa cells transiently expressing hPABPN1 cDNA with different polyalanine tracts demonstrate that the sensitivity of Ab4340 coincides with the pathological threshold known to cause OPMD.To determine whether or not it could be used to discriminate between samples obtained from OPMD patients and control individuals, we performed another series of Western blot immunodetections for which the protein lysates were prepared from lymphoblastoid cell lines (LCLs).Furthermore, we used our 4340 antibody in immunohistochemistry assays of cerebellar sections from OPMD patients and controls.
Western blots probed with Ab4340 reveal a strong signal at ~60 kDa in nuclear lysates prepared from OPMD patient material (Figure 2a, lanes 3-6), whereas no bands were detected in nuclear lysates prepared from unaffected individuals (Figure 2a, lanes 1 and 2).These same lysates were probed with an antibody directed against PABPN1, and a corresponding band at ~60 kDa was observed (Figure 2b).This indicates that the ~60 kDa bands detected by the two antibodies are the same predicted PABPN1-alanine protein.In contrast to the results obtained from HeLa cells, no unspecific contaminant signal was observed from patient lymphoblastoid cell lines.
Immunohistochemistry detections made using Ab4340 and an antibody directed against ubiquitin revealed strongly stained intranuclear structures in cerebellar neurons of the OPMD patient (Figure 2c and 2d).The ubiquitin-detecting antibody also revealed intranuclear signals in sections prepared using tissue sections of a control individual (Figure 2e); however, when Ab4340 was used on similar sections no intranuclear signal was observed (Figure 2f).Alanine-containing proteins are detected in a transgenic Drosophila model of SCA3, and lymphoblastoid cells of SCA3 and HD patients To test whether our antibody could detect polyalanine-containing proteins in polyglutamine diseases that a propensity to present frameshifting, we investigated SCA3 and HD.Using expCAG 92 and isogenic control flies from our previously reported transgenic ATXN3 Drosophila model 16 , we made immunohistochemical detections with both our 4340 antibody and one directed against ataxin-3.Alanine-containing proteins (red) were observed exclusively within the eyes of expCAG 92 flies (Figure 3a).In these same flies, ataxin-3 containing aggregates (green) were present throughout the eye, confirming transgene expression (Figure 3a).No ataxin-3 containing proteins were detected in the isogenic control flies (Figure 3b).
Immunocytochemical detections were also made with LCLs derived from SCA3 patients, HD patients, and control individuals.Ab4340 detected alanine-containing protein aggregates in LCLs from both SCA3 (Figure 3c, arrows) and HD (Figure 3d, arrows) patients, whereas no aggregates were observed in the control individual's LCLs (Figure 3e).When comparing the number of cells presenting aggregates among the SCA3 and HD patients, their occurrence were observed more frequently in HD patient LCLs.

Discussion
Ab4340 was assessed for its ability to selectively detect alaninecontaining proteins in disease models of OPMD, SCA3, and HD, while confirming that unaffected control individuals would not present significant levels of these same polyalanine peptides.We chose to test the antibody's sensitivity using OPMD as a model since the protein underlying this pathology contained an expanded polyalanine tract, and this disease shares a number of similarities with polyglutamine expansion diseases: late-onset, autosomal-dominant, repeat expansion effects age of onset and severity, and the presence of aggregated proteins (INIs) 12,13 .Importantly, of the nine severe human diseases that have been associated with expansions of the polyalanine tract, PABPN1 is the only gene that does not encode for a transcription factor that acts during early development phases 1 .The results of Western blots prepared using lysates of HeLa cells expressing GFPtagged hPABPN1 showed that the signal generated by Ab4340 was substantially stronger in lysates of cells where the length of the polyalanine tract was longer than what is found in the unaffected population (10 alanines) and within the pathological threshold (11 to 17 alanines).While fluorescent immunohistochemistry detections of these same HeLa cells did not show a corresponding profile (increased signal in cells expressing a pathological length polyalanine tract) the antibody could discriminate between biological materials of OPMD patients and control individuals; the antibody did so by both Western blots and immunohistochemistry detections.The discrepancy seen with the transient expression assays made using HeLa cells may be due to the combination of the strong cytomegalovirus promoter used and the high sensitivity of the confocal microscopy which could detect lower amounts of fluorescence-tagged proteins 24 .Another explanation for this discrepancy may be the structural conformation of complexes formed during the aggregation of polyalanine expanded proteins.In vitro studies have shown that polyalanine proteins transition from α-helical monomers to macromolecular β-sheets as the number of alanine residues increase (7 to 15), whereas in vivo these same polyalanine proteins adopt mainly β-sheet confirmations [25][26][27] .Thus, the affinity of our antibody could be directed toward the α-helical/β-sheet transition complex of 10-alanine repeats found predominantly in HeLa cells transiently expressing them.
In support of earlier reports where we established the occurrences of frameshifting in SCA3 using cell culture, cerebellar and cortical organotypic slice culture, transgenic Drosophila, and patient tissue samples [15][16][17] , we detected the expression of alanine-containing proteins in the expCAG 92 Drosophila line, as well as alanine-containing protein aggregates in the LCLs of SCA3 patients.Moreover the antibody could also detect alanine-positive aggregates in LCLs of HD patients; the morphology of these aggregates was similar to what was observed with SCA3.This result with HD LCLs is in agreement with the detection of frameshifted products in human huntingtin (HTT) stable transfectant cells, an HTT transgenic mouse model, and HD patient tissue samples 19,20 .
It is important to note that Ab4340 did not detect alaninecontaining proteins in any samples obtained from control individuals, and this is noteworthy as there are currently over 100 known human proteins to comprise a polyalanine tract of seven alanines or greater 28 .Since the majority of these polyalaninecontaining roteins are DNA binding transcription regulators, which often bind transcription factors, it is likely that they share a similar low level of expression that is below the detection threshold level of our antibody 28,29 .
In summary, our experiments with Ab4340 demonstrate that it is a valuable tool for the detection of alanine-containing proteins in OPMD, SCA3, and HD.This antibody could be used to screen other "orphan" neurodegenerative or developmental diseases for the presence of expanded alanine tracts which may help uncover new polyalanine diseases.It could also help to further characterize the subcellular localization of proteins containing such polyalanine tracts.

Results
Figure 1.It would be useful to add anti-PABPN1 and loading control.It is unclear whether the antibody recognizes Ala13 and Ala17 better than Ala30 and Ala40, or that is due to loading differences.This point is also noticeable in the immunofluorescence: the higher polyAla signal is found in the highest overexpressing cells, regardless of alanine expansion length.Moreover, a strong signal is found in the 10Ala expressing cells.The authors suggests that the polyAla antibody recognizes GFP-containing INIs, but the assay to confirm that (1M KCl or 1% Triton) was not employed.Nearly every anti-PABPN1 antibody detects the protein in immunofluorescence better than in Western blot, because the signal is concentrated to the nucleus (small volume), this is not specific to the polyAla antibody.Figure 2. Loading control is missing in 2A and 2B.The size of PABPN1 in a Western blot is 50-52 kDa.According to Fig. 2A and2b it is >60 kDa.Another protein ~140kDa is recognized by the polyAla antibody, can you please comment on it?Immunostaining in cerebellar neurons: should we assume that samples were taken from three OPMD patients and three controls (are those healthy controls, what is the age, gender?)This information must be added for both controls and OPMD patients.If the data is from only one patient, it should be clearly indicated, and the conclusions must be tuned down.The authors claim that staining is of intranuclear structures, but the nucleus is not clearly visible.A nuclear staining must be included.Figure 3. a-b.There is quite some auto fluorescence in the red channel in the a panel, and the overlap with the GFP signal is not specific.It is unclear why there is no auto fluorescence from the in the b panel?From the low magnification image of human samples it is clearly appreciated that there is more signal in cells from patients compared with control.But the foci the authors pointed to in the higher magnification are also seen in control, in the low magnification image.Also in that presentation, the descriptions of patients and control must be included.If the authors wish to specify intranuclear staining, a nuclear staining must be added.The structure of foci in those patients'cells, seems very different from those in OPMD.Can the authors comment on it?insoluble intranuclear inclusions?A control staining against wtPABPN1 (using Abcam ab75855 for example) would also be useful to compare both staining.
There is a mistake in the abstract: the authors wrote "from 2 in normal to 7 in disease" when it should read "from 2 to 7 in disease".Besides, it should be "from 1 to 8 in disease" with the addition of references Jouan Can J Neurol Sci.2014 (publication from the authors) and Richard et al et al J Neuromuscul Dis.2015 .Reference 12 is not correctly cited.The reference should be mentioned for the first sentence of the paragraph and not the second one regarding neurological involvement.

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Not applicable
Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
I have read this submission.I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Figure 2 .
Figure 2. Testing the ability of Ab4340 to differentiate between OPMD patient and control individual samples.Western blot immunodetections of alanine (a) and PABPN1 (b) from nuclear extracts prepared from LCLs.Lanes 1 and 2: extracts from control individuals; and Lanes 3-6: extracts from OPMD patients.The 4340 antibody cleanly detected alanine-containing proteins exclusively from OPMD patient extracts [(a), Lanes 3-6], despite strong detection of PABPN1 in all patient extracts (b).Immunohistochemical detection of ubiquitin (c,e) and polyalanine (d,f) containing proteins in cerebellar neurons of an OPMD patient (c,d) and control individual (e,f).Both antibodies immunostained intranuclear structures in the OPMD patient's sample (c,d), whereas only ubiquitin immunostaining was achieved in the control patient's sample (e).Scale bar, 2.5 µm.

Figure 3 .
Figure 3. Detection of polyalanine in a transgenic Drosophila model of SCA3, and lymphoblastoid cells of an SCA3 and HD patient.(a,b) Double-labelling immunofluorescence detection of alanine (red) and GFP (green) in an expCAG 92 transgenic fly (a) and an isogenic control (b), showing polyalanine-and ubiquitin-labeled aggregates exclusive to the transgenic line (a).Scale bar, 25 µm.(c-e) Immunocytochemical detection of polyalanine containing proteins in lymphoblastoid cells of an SCA3 patient (c), HD patient (d), and control individual (e).The 4340 antibody immunostained intranuclear inclusions in both the SCA3 (c, arrows) and HD (d, arrows) patient cells, whereas no intranuclear staining was present in the control patient lymphoblast cell line (e).Scale bar, 2.5 µm.

References 1 .
Vest KE, Apponi LH, Banerjee A, Pavlath GK, Corbett AH: An Antibody to Detect Alanine-Expanded PABPN1: A New Tool to Study Oculopharyngeal Muscular Dystrophy.L, Rocheford D, Szuto A, Carney E, David K, Dion PA, Rouleau GA: An 18 alanine repeat in a severe form of oculopharyngeal muscular dystrophy. .2014; (4): 508-11 Can J Neurol Sci 41 PubMed Abstract 3. Richard P, Trollet C, Gidaro T, Demay L, Brochier G, Malfatti E, Tom FM, Fardeau M, Lafor P, Romero N, Martin-N ML, Sol G, Ferrer-Monasterio X, Saint-Guily JL, Eymard B: PABPN1 (GCN)11 as a Dominant Allele in Oculopharyngeal Muscular Dystrophy -Consequences in Clinical Diagnosis and Genetic Counselling. .2015; (2): 175-180 | J Neuromuscul Dis 2 PubMed Abstract Publisher Full Text Is the work clearly and accurately presented and does it cite the current literature?Partly Is the study design appropriate and is the work technically sound?Yes Genetic Services, Inc.).Flies bearing transgenic constructs in a homozygous state were maintained at 25°C.Adult males were crossed to virgin gmr-GAL4 flies to obtain lines expressing transgenes in developing eyes ( wt CAG 14 /gmr-GAL4, exp CAG 92 /gmr-GAL4, exp CAA 96 / gmr-GAL4, STOP-CAG 94 /gmr-GAL4 and STOP-CAA 94 /gmr-GAL4 genotypes).To obtain isogenic control flies, w 1118 male flies were crossed with virgin gmr-GAL4.Cell culture and transfectionsEpstein-Barr Virus (EBV) immortalized lymphoblastoid cell lines (LCL) were prepared from patient and control individuals in the laboratory of Dr Rouleau.All subjects signed an informed consent, which is part of the experimental protocol approved by the Research Ethical Board (REB) of McGill University Health Research Centre (#NEU-14-051, June 11 2015).LCL and HeLa cells were cultured at 37°C in a humid atmosphere enriched with 5% CO 23, exp CAG 92, exp CAA 96 , STOP-CAG 94 or STOP-CAA 94 repeats were subcloned in pUAST (some vectors have a STOP codon upstream of the repeat).Epitope tags were added to each reading frame: Myc for main-frame, HA for −1 frame and His for +1 frame.Vectors sequenced before injection into w 1118 Drosophila eggs; a step followed by selection of positive transformants, mapping and balancing (2 .HeLa cells (ATCC #ATCC® CCL-2 TM ) cat were grown in Dulbecco's Modified Eagle Medium (Gibco), supplemented with 10% fetal bovine serum (Gibco) and 1% Penicillin/Streptomycine/Glutamine (Gibco), while the lymphoblastoid cells were grown in Iscove's Modified Dulbecco's Medium (Gibco) supplemented with 10% fetal bovine serum (Gibco), 1% Penicillin/Streptomycine/Glutamine (Gibco) and Fungizone antimycotic (Gibco).For transient transfections, HeLa cells were transfected at 70% confluency for 48 hours with GFP-tagged hPABPN1 plasmid DNA containing various length alanine expansions (0, 10, 13, 17, 30, and 40) using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions.These constructs were graciously provided by Dr. Bernard Brais (McGill University), and previously described23.monoclonal anti-GFP antibody (1:5,000; Clontech #632381); or the rabbit polyclonal antibody 4340 (1:500-3,000) developed by the laboratory of Dr McPherson.Membranes were then washed three times for 10 min in PBS-T, incubated for 2 hours at room temperature with the appropriate horseradish peroxidase (HRP) conjugated secondary antibody [donkey anti-mouse IgG antibody