Substrate recruitment via eIF2γ enhances catalytic efficiency of a holophosphatase that terminates the integrated stress response

Significance Protein phosphorylation activates important biological processes that are later deactivated by dephosphorylation. Phosphoserine/threonine dephosphorylation is catalyzed by holophosphatases comprising a catalytic subunit, specialized in hydrolytic phosphate removal and regulatory subunit(s) that select phosphoprotein substrates. Dynamics of (de)phosphorylation of phosphoserine 51 on the alpha subunit of eukaryotic translation initiation factors 2 (eIF2) regulates protein synthesis in stressed cells. Previous research has focused on mechanisms operating near the catalytic site of the eIF2-directed holophosphatase. Here, computational, crystallographic, biochemical, and cellular techniques uncover interactions between elements distant from the catalytic site of the eIF2 holophosphatase and its cognate multisubunit eIF2 phosphoprotein substrate. These interactions reveal physiologically important action-at-a-distance that facilitates phosphate removal from eIF2 to efficiently terminate signaling in a mammalian stress response.

Fig. S2 Flexibility of eIF2 trimer and the N-terminally extended PPP1R15A-containing holophosphatase hint at the difficulty in solving the structure of a dephosphorylation complex of eIF2 P .
(A) Two and three dimensional reconstructed images of complexes of eIF2 P and an extended holophosphatase comprised of human PPP1R15A  , PP1A and G-actin. The gren volume on the left represents the previously determined cryo-EM structure of PP1 D64A /G-actin/DNase I/R15A 553- 624 /eIF2α P -NTD (PDB 7NZM).Two low resolution cryo-EM volumes were obtained from the complex of PP1 H66K /G-actin/DNase I/R15A 325-636 /eIF2α P γ (blue and yellow surface contoured at a similar level) by ab-initio reconstruction in CryoSPARC: the blue is similar to the 7NZM (map correlation is 0.86) and the yellow volume is different and bigger than the former.Below are selected 2D classes generated from the particles classified for these two volumes.Despite the presence on the grid of particles of a size consistent with the eIF2 P dephosphorylation complex, we were unable to solve the structure at a meaningful resolution.
(B) MD simulations of the eIF2 trimer Shown is a simulation starting from the highest ranked AFM model of eIF2.The top panel display overlaid three distinct conformations of the eIF2 trimer throughout the simulation trajectory, underscoring the flexibility of the eIF2α-CTD, eIF2β, and eIF2γ, compared to the relatively stable eIF2α-NTD.The middle panels depict the RMSD plots for eIF2α-hinge (residue 184-190) and the composite eIF2α-CTD-eIF2β-eIF2γ.RMSD calculations were performed after aligning the entire trajectory on the eIF2α-NTD, because of its considerable stability (RMSD 2.3 ± 0.3 Å).The cartoon illustrates the potential for the bulky eIF2 lobe comprised of βγ and α-CTD subunits to interfere with the alignment of the flexibly-attached eIF2α P -NTD in the active site, by clashing with the holoenzyme.
(C) MD simulations of the extended holophosphatase MD simulations of the extended holophosphatase highlight the flexibility of the N-terminal extension of PPP1R15A (PPP1R15A  ). Three secific conformations of the extended holophosphatase are displayed over the course of the simulation trajectory.The components are coloured as indicated and the blue circle represents the PP1 active site.RMSD of PPP1R15A was calculated with respect to PP1 and G-actin proteins.RMSD values reported and computed were specifically for the Cɑ atoms.
Solid lines in plot panels represent the exponential moving average throughout the MD simulations for each respective replicate.The cartoon depicts the largely disordered N-terminal extended PPP1R15A inhibiting substrates engagement.
Fig. S3 AFM prediction and MD simulations of eIF2γ and PPP1R15A complexes.
(A) The pLDDT (predicted local distance difference test) and PAE (Predicted Aligned Error) plots for AFM predictions of the human eIF2 trimer in complex with the repeat-containing region of PPP1R15A  .
(B) The pLDDT and PAE plots for AFM prediction of complexes of human eIF2γ and individual repeats of PPP1R15A: R1 331-376 , R2 377-420 , R3  or R4  respectively. (C) Plos of time-dependent variation of distances between the atoms of human PPP1R15A L 338 or V 343 and the indicated residues of human eIFγ throughout 500 ns of an all-atom MD simulations performed using the AFM predicted complex structure of eIF2γ and repeat 1 (PPP1R15A  ).
Shown are three replicates of the simulation.
(D) The dynamic interactions between the acidic segment of PPP1R15A's repeat 1 region (PPP1R15A [348][349][350][351][352][353][354][355][356][357][358][359][360] ) and the positively charged surface of eIF2γ.The left panel displays the interface interaction captured at time=500 ns from the calculations of replicate 1.The right panel presents a stability analysis based on the count of contacts between PPP1R15A 348-360 and the positively charged surface of eIF2γ throughout the MD simulations.The cutoff radius for computing the contact was set at 6 Å.  (B) AFM model predicting the ability of human R15A-R3  to bind to the same conserved hydrophobic groove in yeast eIF2γ as in human eIF2γ. On the eft is an overlaid AFM prediction of human R15A in complex with human eIF2γ (grey) or yeast eIF2γ (colored by pLDDT).The RMSD over 426 pruned atom pairs is 0.48Å.The arrow points at the helix binding in the hydrophobic groove of eIF2γ.On the right are the pLDDT and PAE plots of AFM prediction of the human R15A-R3 421-467 and yeast eIF2γ complex.-------E

Fig. S4
Fig. S4 Phylogenetic conservation of PPP1R15 repeats and their binding site in eIF2γ.

(
A) Sequence alignment of repeat 3 of human PPP1R15A (aa.424-450) with homologues and orthologues.Conserved hydrophobic residues that contact eIF2γ across PPP1R15A, PPP1R15B in mammals and PPP1R15 in other vertebrates and in human herpesvirus 1 are marked by stars.

Fig
Fig. S6 PPP1R15A contacts with eIF2γ contribute to ISR termination in cells transfected with mammalian expression plasmids encoding the counterpart to the PPP1R15A 325-636 used in vitro.

Table S1 :
List and description of plasmids used in this study.

Table S2 :
X-ray data collection and refinement statistics.Values in parentheses are for the highest resolution shell. a Table S3 Key resource table