Studies on human eRF3-PABP interaction reveal the influence of eRF3a N-terminal glycin repeat on eRF3-PABP binding affinity and the lower affinity of eRF3a 12-GGC allele involved in cancer susceptibility

Abstract

The eukaryotic release factor 3 (eRF3) has been involved in the control of mRNA degradation through its association with the cytoplasmic Poly(A) Binding Protein, PABP. In mammals, eRF3 N-terminal domain contains two overlapping PAM2 motifs which specifically recognize the MLLE domain of PABP. In humans, eRF3a/GSPT1 gene contains a stable GGC repeat encoding a repeat of glycine residues in eRF3a N-terminus. There are five known eRF3a/GSPT1 alleles in the human population, encoding 7, 9, 10, 11 and 12 glycines. Several studies have reported that the presence of eRF3a 12-GGC allele is correlated with an increased risk of cancer development. Using surface plasmon resonance, we have studied the interaction of the various allelic forms of eRF3a with PABP alone or poly(A)-bound PABP. We found that the N-terminal glycine repeat of eRF3a influences eRF3a-PABP interaction and that eRF3a 12-GGC allele has a decreased binding affinity for PABP. Our comparative analysis on eRF3a alleles suggests that the presence of eRF3a 12-GGC allele could modify the coupling between translation termination and mRNA deadenylation.

Keywords: Cancer; GSPT1; GSPT2; PABP; eRF3; mRNA degradation; mRNA poly(A) tail; surface plasmon resonance; translation termination.

Figure 1.

SPR analysis of the interaction between eRF3 and PABP (2000–2200 RU) immobilized on a CM5 sensor chip surface. Typical experimental sensorgrams (top panels) and residual plots (difference between calculated and experimental data points, bottom panels) are shown for eRF3a 10-GGC allele (A), eRF3a 12-GGC allele (B), eRF3a 7-GGC allele (C), eRF3a 11-GGC allele (D), and eRF3b (E). The different concentrations of the injected analytes are indicated on the right of the sensorgrams (for experimental details see Materials and Methods). Data were treated and integrated using a simple Langmuir 1:1 model. For each analyte concentration the fitted curve is shown in black. (F) Comparison of eRF3a alleles and eRF3b equilibrium dissociation constants (K_D) listed in Table 1. Error bars indicate the standard error (SEM). Statistical significance (P-values from unpaired Student's t-test) of differences between eRF3a 12-GGC and other eRF3a alleles are indicated.

Figure 2.

Kinetic analyses of PABP binding to 5′ biotinylated OligoA₁₂₀ RNA (220 RU) immobilized onto a SA sensor chip surface. (A) Sensorgrams of the binding profiles using a Langmuir 1:1 binding model (top panel) and residuals plots (bottom panel) are shown. Concentrations of PABP injected are indicated on the right of the sensorgrams. Kinetic parameters (k_a, k_d and K_D) and the Chi2 value of the fitting are also indicated. (B) Plot of the response vs. PABP concentration used for the steady-state affinity fitting with the BIAevaluation software (see text for details).

Figure 3.

Sensorgrams of single-binding analysis of eRF3a and 7-GGC, 11-GGC, and 12-GGC alleles. eRF3a alleles were injected at 10 and 20 nM concentrations (indicated on the right of each sensorgram) over PABP (300 RU) bound to OligoA₁₂₀ RNA (30 RU) immobilized on a SA sensor chip.

Figure 4.

Kinetic analyses of eRF3a 10-GGC allele (A) and eRF3a 12-GGC allele (B) interacting with PABP (300 RU) bound to OligoA₁₂₀ RNA (30 RU) immobilized onto a SA sensor chip surface. Typical sensorgrams and fitting of the binding profiles to a Langmuir 1:1 binding model (top panel) and residuals plots (bottom panel) are shown. Concentrations of injected analytes are indicated on the right of the sensorgrams. For each analyte concentration the fitted curve is shown in black.