Mechanistic convergence across initiation sites for RAN translation in fragile X associated tremor ataxia syndrome

Abstract

Repeat associated non-AUG (RAN) translation of CGG repeats in the 5'UTR of FMR1 produces toxic proteins that contribute to fragile X-associated tremor/ataxia syndrome (FXTAS) pathogenesis. The most abundant RAN product, FMRpolyG, initiates predominantly at an ACG upstream of the repeat. Accurate FMRpolyG measurements in FXTAS patients are lacking. We used data-dependent acquisition and parallel reaction monitoring (PRM) mass spectrometry coupled with stable isotope labeled standard peptides to identify signature FMRpolyG fragments in patient samples. Following immunoprecipitation, PRM detected FMRpolyG signature peptides in transfected cells, and FXTAS tissues and cells, but not in controls. We identified two amino-terminal peptides: an ACG-initiated Ac-MEAPLPGGVR and a GUG-initiated Ac-TEAPLPGGVR, as well as evidence for RAN translation initiation within the CGG repeat itself in two reading frames. Initiation at all sites increased following cellular stress, decreased following eIF1 overexpression and was eIF4A and M7G cap-dependent. These data demonstrate that FMRpolyG is quantifiable in human samples and FMR1 RAN translation initiates via similar mechanisms for near-cognate codons and within the repeat through processes dependent on available initiation factors and cellular environment.

Figure 1

Predicted peptide sequences and monitored signature peptides for RAN translated product, FMRpolyG. (A) Schematic of FMRpolyG peptides produced via alternative translation initiation on FMR1. (top) FMR1 schematic showing location of the FMRpolyG uORF (pink), FMRP ORF (green) and sequence corresponding to overlap of these two ORFs (yellow). The mRNA sequence 5′ to the CGG repeat is shown on the right. Three potential FMRpolyG translation initiation sites (TIS) on FMR1 are indicated: two previously identified (purple) and one novel initiation site (blue). Grey highlights indicate position of each codon relative to the CGG repeat. Grey arrows point to specific mutant reporters used later in this study. (bottom) Predicted trypsin-digested signature peptides for FMRpolyG. Underlined sequences are those identified in this study. The grey M in brackets indicates a potential product of iMet excision. (B–D) Liquid chromatography MS (LC–MS/MS) spectra of FMRpolyG signature peptides, CAM-CGAPMALSTRHL (B), SPPLGGGLPALAGLK (C) and Ac-MEAPLPGGVR (D) from HEK293Ts transfected with FMRpolyG100-3XFlag. Observed b- and y-ions are indicated.

Figure 2

PRM-based quantification of FMRpolyG. Schematic for PRM-SIS quantification. Samples were digested with trypsin and then spiked with known concentrations of a signature SIS peptide. The amino acid sequence for our signature peptide of interest was detected by MS, and the endogenous peptide concentration was quantified relative to the SIS peptide. (B) Quantification of FMRpolyG signature peptides, Ac-MEAPLPGGVR and SPPLGGGLPALAGLK, in HEK293T cell lysates transfected with untagged FMRpolyG100 for 24 h and either processed directly or following IP with NTF1 antibody. Bars represent mean ± SD. N = 2. (C) Quantification of the absolute abundance of endogenous FMRpolyG peptide Ac-MEAPLPGGVR from human FXTAS and control brains by PRM-SIS.

Figure 3

CGG RAN translation initiation can occur at a 5′ GUG codon. (A) Representative LC–MS/MS chromatogram of FMRpolyG peptide, Ac-TEAPLPGGVR from NTF1 immunoprecipitated HEK293Ts transfected with FMRpolyG100. (B) Quantification of Ac-TEAPLPGGVR absolute abundance in HEK293Ts transfected with FMRpolyG100. Bars represent mean ± SD. N = 2. (C) The proportion of two different initiation fragments quantified via PRM-SIS from IPed lysates of HEK293Ts transfected with the indicated reporters. (D) WB of FMRpolyG from HEK293Ts transfected with the indicated reporters. GAPDH serves as a loading control. (E) Quantification of signature peptide absolute abundance in IPed lysates of HEK293Ts transfected with the indicated reporters. ‘Detected’ indicates observed, but not quantified (See Supplementary Material, Fig. S3C).

Figure 4

GUG initiation is preferentially utilized in response to cellular stress. (A) Representative WB of FMRpolyG from HEK293Ts transfected with FMRpolyG100-3XFlag followed by DMSO or TG treatment. GAPDH serves as a loading control. N = 3. (B) Quantification of FMRpolyG from the WB in (A). Bars represent mean ± SD, N = 3. Unpaired t test, ^*P < 0.05. (C) The absolute abundance of SPPLGGGLPALAGLK by PRM-SIS in lysates of above samples. Bars represent mean ± SD, N = 3. Paired t test, ^*P < 0.05. (D) Quantification of the absolute abundance of Ac-TEAPLPGGVR and Ac-MEAPLPGGVR by PRM-SIS in lysates of samples in (A). Bars represent mean ± SD, N = 3. Paired t test, ^*P < 0.05. (E) Ratios of the two N-terminal peptides in TG versus DMSO-treated HEK293Ts. Bars represent mean ± SD, N = 3. Paired t test, ^*P < 0.05.

Figure 5

FMRpolyG initiation codon choice is influenced by eIF1 and eIF2A availability. (A) Representative WB of FMRpolyG from HAP1 WT and eIF2A KO cells transfected with FMRpolyG100-3XFlag. GAPDH serves as a loading control. N = 4. (B) Quantification of FMRpolyG from the WB in (A). Bars represent mean ± SD, N = 4. Unpaired t test, ^***P < 0.001. (C) Quantification of SPPLGGGLPALAGLK absolute abundance by PRM-SIS in Flag-IPed lysates of HAP1 cells transfected with FMRpolyG100-3XFlag. N = 4. Paired t test, ns: not significant. (D) Quantification of Ac-TEAPLPGGVR and Ac-MEAPLPGGVR absolute abundance by PRM-SIS in Flag-IPed lysates of HAP1 cells transfected with FMRpolyG100-3XFlag. N = 4. Paired t test, ^*P < 0.05, ns: not significant. (E) Ratio of the two N-terminal peptides in HAP1 WT and eIF2A KO cells transfected with FMRpolyG100-3XFlag. Bars represent mean ± SD, N = 4. Paired t test, ^**P < 0.01. (F) Representative WB of FMRpolyG (top), eIF5 (middle) and eIF1 (bottom), from HEK293Ts co-transfected with FMRpolyG100-3XFlag and control, eIF1 or eIF5 vectors. GAPDH serves as a loading control N = 3. (G) Quantification of FMRpolyG from the WB in (F). Bars represent mean ± SD, N = 3. Unpaired t test, ^*P < 0.05, ^**P < 0.01. (H and I) Quantification of SPPLGGGLPALAGLK (H) and Ac-TEAPLPGGVR and Ac-MEAPLPGGVR (I) absolute abundance by PRM-SIS in Flag-IPed lysates of HEK293Ts co-transfected with FMRpolyG100-3XFlag and control, eIF1 or eIF5 vectors. N = 3. Paired t test,^*P < 0.05, ns: not significant. (J) Ratio of the two N-terminal peptides in HEK293Ts co-transfected with FMRpolyG100-3XFlag and control, eIF1 or eIF5 vectors. Bars represent mean ± SD, N = 3. Paired t test, ^***P < 0.001.

Figure 6

CGG RAN translation can initiate within the repeat. (A) Representative WB of FMRpolyG from HEK293Ts transfected with the indicated reporters followed by DMSO or TG treatment. GAPDH serves as a loading control. Lysates from WT FMRpolyG transfected cells were diluted 1:10 to account for signaling differences between WT FMRpolyG and stop@-12 reporters. See Supplementary Material, Figure S4A for an undiluted version. (B) Quantification of stop@-12 from the WB in (A). Bars represent mean ± SD, N = 3. Unpaired t test, ^*P < 0.05. (C) Absolute amounts of FMRpolyG signature peptides by PRM-SIS from IPed lysates of HEK293Ts transfected with the indicated reporters and treatment from (A). (D and E) Representative WB and quantification of FMRpolyG from HEK293Ts co-transfected with the stop@-12 reporter and control or eIF1 vector. GAPDH serves as a loading control. Bars represent mean ± SD, N = 3. Unpaired t test, ^**P < 0.01. (F and G) WB of FMRpolyG-Flag from HEK293Ts transfected with FMrpolyG-3xF (F) or stop@-12 contructs (G) and treated with hippuristanol for 10 h. Tubulin serves as a loading control. N = 3. (H) Quantification of FMRpolyG-Flag from the WBs in (F) and (G). Bars represent mean ± SD. N = 3. One-way ANOVA and unpaired t test. ns: not significant, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. (I) Quantification of nanoluciferase signal from HEK293Ts transfected with indicated reporters. Bars represent mean ± SD, N = 9. Multiple T-test comparison. ^****P < 0.0001.