New Results

Molecular Mechanism of PP2A/B55α Inhibition by IER5

Ruili Cao, Daniel TD Jones, Li Pan, Shumei Wang, Shaun Rawson, Jon C Aster, Stephen C Blacklow
doi: https://doi.org/10.1101/2023.08.29.555174

ABSTRACT

PP2A serine/threonine protein phosphatases are heterotrimeric complexes that have a wide range of essential physiologic functions. The B55α form of PP2A has critical roles in cell cycle regulation, mitotic exit, and the DNA damage response1–6. Its activity is modulated by additional regulatory proteins, such as ARPP197, FAM122A8, and IER59. However, the precise mechanisms underlying the modulation of PP2A activity by these proteins remain elusive. Here, we show that IER5 inhibits pTau dephosphorylation by PP2A/B55α in biochemical assays and report a cryoelectron microscopy structure of the PP2A/B55α-IER5 complex, which reveals that IER5 occludes a surface on B55α used for substrate recruitment10–12. Mutation of interface residues on IER5 interferes with recovery of B55α in co-immunoprecipitation assays and suppresses events in squamous carcinoma cells, such as KRT1 expression, that depend on inhibition of PP2A/B55α by IER59. These studies define the molecular basis for PP2A inhibition by IER5 and suggest a roadmap for selective pharmacologic modulation of PP2A/B55α complexes.

Main

PP2A serine/theonine protein phosphatases are assembled from a scaffolding subunit (A, A’), a catalytic subunit (C, C’) and a regulatory subunit derived from one of four different protein subfamiliies (B/B55, B’/B56, B’’/PR48-PR70, and B’’’/striatin)13. In addition to its role in regulating the cell cycle and the DNA damage response, the PP2A A-C subcomplex is also incorporated into the INTAC submodule of the integrator complex regulating gene transcription14,15.

IER5 is a member of the AP1-regulated immediate early response (IER) gene family16, encoding a protein of 327 amino acids. Expression of IER5 is induced in response to ionizing radiation, and IER5 is implicated in the cellular response to DNA damaging agents and heat shock17–19. In squamous cell carcinoma (SCC) cells, IER5 is a direct transcriptional target of activated Notch. Expression of IER5 in response to Notch activation in these SCC cells is required for induction of a cell differentiation program that arrests cell growth and stimulates expression of KRT1 and other keratinocyte-associated genes9.

IER5 executes its cellular function, at least in part, by binding to and modulating the activity of heterotrimeric B55α holoenzyme complexes of PP2A9,20–22. Though one proposed model for IER5 function proposes that it acts to direct selection of substrates such as S6K and HSF122, IER5 functions as a PP2A/B55α antagonist in SCC cells, where its suppression of PP2A/B55α activity is necessary for Notch-dependent induction of genes associated with keratinocyte differentiation 9. Here, we report the structure of a PP2A/B55α-IER5 complex, uncovering the molecular basis for its function as an inhibitor of PP2A/B55α, and identifying a strategy for selective modulation of PP2A/B55α activity.

Results

IER5 inhibits PP2A dephosphorylation of pTau. Previous work demonstrated that the 50-residue N-terminal IER domain of IER5, predicted to be a helical hairpin by alphafold2 (Extended Data Fig. 1), is necessary and sufficient for binding to B55α/PP2A, whereas the C-terminal region, predicted to be unstructured (Extended Data Fig. 1), does not interact9. We purified the PP2A/B55α holoenzyme both in isolation and in complex with the IER5 N-terminal domain (Extended Data Fig. 2), and compared the dephosphorylation activity of the two complexes using phosphorylated Tau (pTau) as substrate12 (Fig. 1a,b). The rate of pTau dephosphorylation was approximately 5-fold slower when PP2A/B55α was complexed with IER5 (Fig. 1c), showing that IER5 acts as an inhibitor of protein substrate dephosphorylation.

Extended Data Fig. 1:
Extended Data Fig. 1: Structural model of IER5 predicted by Alphafold244.

a, Schematic representation of the domain organization of IER5. The IER-homology region is shown in red. b, Cartoon representation of the alphafold2 model of IER5, colored by pIDDT value on a sliding scale from blue (pIDDT < 50) to red (pIDDT >90). c, Predicted aligned error plot for the IER5 model, colored on a sliding scale from dark green (0 Å) to white (>30 Å).

Extended Data Fig. 2:
Extended Data Fig. 2: Biochemical characterization of PP2A and PP2A-IER5 complexes.

a, Schematic illustration of the protein constructs used to prepare PP2A and PP2A-IER5 complexes. Baculoviruses expressing full-length A, B and C subunits of PP2A were used to produce both PP2A and PP2A-IER5 complexes, and a virus expressing the N-terminal 50 amino acids of IER5 was added for production of the PP2A-IER5 complex. b, Size exclusion chromatogram of the purified PP2A on a Superdex 200 column. An SDS-PAGE gel of the purified complex is shown to the right of the chromatogram. The small peak at approximately 20 mL elution volume corresponds to the FLAG peptide. c, Size exclusion chromatogram of the purified PP2A-IER5 complex on a Superdex 200 column. An SDS-PAGE gel of the purified complex is shown to the right of the chromatogram. Both peaks are PP2A-IER5 complexes; peak 2 was used in all biochemical and structural studies.

Fig. 1
Fig. 1 Tau dephosphorylation assay.

a, SYPRO(R) Ruby stained SDS-PAGE gel of 100 nM PP2A complex purified alone (left) and with IER5 (right). b, Anti-pTau (S396) Western blot comparing the pTau phosphatase activity of PP2A purified in the absence (left) and presence (right) of IER5. c, Densitometry analysis of bands observed in (b). Data are presented as mean ± s.d. of three biological replicates. Data are normalized to time point t=0. The Welch t-test statistic was used to compare the means between each PP2A and PP2A-IER5 time point (*P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001).

Structure of a PP2A/B55α-IER5 complex. To elucidate the molecular basis for IER5-dependent inhibition of PP2A/B55α substrate dephosphorylation, we determined a structure of the complex using single particle electron cryomicroscopy (cryo-EM; Fig. 2 and Extended Data Figs. 3-4). During data processing, we observed that PP2A-IER5 complexes existed in both monomeric and dimeric assemblies (Extended Data Fig. 3). We utilized both assemblies during refinement to minimize anisotropy, and produced a final map of the monomeric assembly with a resolution of 3.09 Å. Using a local mask, we also refined a focused map around B55α, IER5 and part of the scaffolding subunit to 3.13 Å. An alphafold model of the PP2A/B55α-IER5 complex created with alphafold multimer23 was then docked into the combined and focused cryo-EM maps before refinement to produce the final models (see methods).

Extended Data Fig. 3:
Extended Data Fig. 3: Cryo-EM image processing workflow.

Cryo-EM processing scheme for PP2a-IER5 reconstruction. Milestone maps are shown to display progression of processing (the discarded, “junk” class is shown as pink, the monomer as orange, the dimer as blue and the combined map as purple). The masks used for local refinement are light blue.

Extended Data Fig. 4:
Extended Data Fig. 4: Cryo-EM data quality.

a, Representative micrograph of PP2A-IER5 in vitreous ice visualized by cryo-EM on a Titan Krios microscope equipped with a Gatan K3 detector. Scale bar indicates 500 Å. b, GS-FSC curves with default CryoSPARC masks. c, Orientation distribution of particles (from CryoSPARC) used in preparing the combined map of PP2A-IER5 for model building. d, Local resolution map of the combined map generated by CryoSPARC. e, Local resolution map of the focused map generated by CryoSPARC. f, Cryo-EM focused map for the portion of the model corresponding to IER5.

Fig. 2
Fig. 2 Cryo-EM structure of a PP2A-IER5 complex and interface analysis.

a, Cryo-EM map of the PP2A-IER5 complex. The scaffolding A subunit is green, the regulatory B55α subunit is magenta, the catalytic C subunit is wheat and IER5 is cyan. b, Cartoon rendering of the modelled structure, colored as in (a), with manganese atoms of the catalytic subunit rendered as grey spheres. c, Open book view depicting the interface between B55α and IER5. IER5 is cyan, and B55α is blue. Residues at the contact interface in the complex (center), on B55α (left), and on IER5 (right) are colored in a darker shade. d, Electrostatic surface representation of the complex (center), of B55α alone (left), and IER5 alone (right), colored on a sliding scale from blue to red (blue: basic, red: acidic). The catalytic and scaffolding subunits have been removed for clarity in (c) and (d).

IER5 adopts a helical hairpin conformation in the complex, and exclusively contacts the B55α subunit. Complexation of PP2A/B55α with IER5 increases the curvature of the scaffolding subunit, resulting in a 22 Å displacement of its C-terminus upon binding and closer approximation of the B55α and catalytic subunits (Extended Data Fig. 5). It is also possible to model the C-terminal end of the catalytic subunit in the complex, which extends across a gap between the regulatory and catalytic subunits to engage the third blade of the B55α β-propeller. Prior work suggests that LCMT1 methylation of the C-terminal end of the PP2A catalytic subunit facilitates holoenzyme assembly24,25, but our local resolution at this site is not sufficient to unambiguously identify this modification.

Extended Data Fig. 5:
Extended Data Fig. 5: Comparison of PP2A structures in free and IER5-bound states.

Cartoon representation showing superposition of the structure of free PP2A (PDB:3dw8; gray) on the structure of PP2A-IER5 (colored as in Fig 2a). Superposition was performed with the B55α subunit. The 22 Å displacement of the C-terminal end of the scaffolding subunit is indicated with a double-headed arrow.

Interface between IER5 and PP2A. The IER5-B55α interface buries a total surface area of 2367

Å2 (Fig. 2c). This interface includes a predominantly hydrophobic contact surface toward the N-and C-termini of the helical hairpin, and a site with electrostatic complementarity between a basic region of IER5 centered on the loop connecting the helices and a series of acidic residues on B55α (Fig. 2d). Among the residues at the hydrophobic interface at the top of IER5 are I10, L35, and V39, which pack into pockets created by F280, F281, I284, F343 and Y337 of B55α (Fig. 3a). Adjacent to the hydrophobic interface there is also a hydrogen bond between the phenolic hydroxyl of Y337 and the S14 side chain hydroxyl group (Fig. 3a), which also lies within hydrogen bonding distance of the backbone carbonyl of I10. At the interface around the IER5 loop between helices, R25 of IER5 protrudes into a negatively charged groove created by E91 and E93 of B55α. (Fig. 3b), engaging in a salt bridge with the carboxylate group of E93. K32 of IER5 engages in a charged hydrogen bond with D340 of B55α and makes aliphatic contact with F343. There is also a charged hydrogen bond between H31 of IER5 and S287 of B55α. K17 and K29 of IER5 also contribute to the positively charged electrostatic surface on IER5 that complements the acidic B55α surface (Fig. 2d).

Fig. 3:
Fig. 3: Intermolecular contacts between B55α and IER5.

a, b, Close-up views of the interface between B55α and IER5. IER5 is rendered as a cyan cartoon and side-chains in contact with B55α are rendered as sticks. B55α is rendered as a magenta cartoon under a transparent surface colored by electrostatic surface charge on a sliding scale from blue to red (blue: basic, red: acidic), with side-chains in contact with IER5 rendered as yellow sticks. The “top” of the interface, which is predominantly hydrophobic, is shown in (a), and the base of the interface, which is dominated by charge-charge interactions, is shown in (b). c, Sites of control mutations analyzed in co-immunoprecipitation and functional assays. The labels for IER5 residues are red at the hydrophobic interface (a), blue at the charged interface (b), and black at the sites of the control mutations in (c). d, Western blot analysis of anti-HA immunoprecipitants prepared from I5 cells expressing wild-type or mutant IER5 tagged with FLAG and HA. Labels of the mutants above the blot are color coded according to the labeling scheme for IER5 in (a-c).

To test the importance of the hydrophobic and electrostatic IER5 surfaces in complex formation with B55α, we introduced single point mutations into full-length IER5 at interface residues or control mutations at sites distant from the contact interface (L15 and N20 of IER5; Fig. 3c) and tested their effects on B55α binding in a co-immunoprecipitation assay in IER5 knockout cells (Fig. 3d). Wild-type IER5, the IER5 L15R mutant, and the N20A mutant all strongly coimmunoprecipitated B55α. In contrast, the interface mutations either completely prevented (S14A, K17E, L30E), or greatly reduced (I10A, R25E, K29E, H31A, K32E, L35A, V39A) co-immunoprecipitation of B55α (Fig. 3c), confirming that the interface seen in the cryo-EM structure is required for binding of IER5 to B55α in cells.

Interface-disrupting mutations block IER5-dependent induction of KRT1 transcription. In the SCC cell line SC2, IER5 is needed for Notch-dependent induction of KRT1 expression (Fig. 4a), a requirement that is relieved when B55α is knocked out and restored when wild-type IER5 is reintroduced into IER5 knockout, B55α wild-type SC2 cells9 (Fig. 4b). Reintroduction of the L15R or N20A control mutants of IER5 also rescues expression of KRT1 in response to Notch activation (Fig. 4b). In contrast, the I10A, S14A, K17E, and L30E interface disrupting mutations all greatly reduce KRT1 expression, as determined by RT-qPCR (Fig. 4b). These data show that interfering with the binding of IER5 to B55α leads to a loss of IER5 function in SC2 cells.

Fig. 4:
Fig. 4: Effects of IER5 mutations on downstream signaling responses.

a, Experimental design for controlled activation of Notch1 by GSI washout. The readout for IER5 dependent is the induction of KRT1 expression (figure design adapted from ref. 9). b, RT-qPCR analysis of KRT1 RNA abundance measured 72 h after GSI washout in I5 cells expressing wild-type or mutant IER5. Transcript abundance was normalized against GAPDH. All data points were obtained in biological sextuplicate. Error bars represent standard deviations of the mean. *, p<0.05; **, p<0.01, ***, p<0.001, ****, p<0.0001 (Welch t-test).

Discussion

The work reported here shows that the N-terminal helical hairpin of IER5 acts as an inhibitor of PP2A/B55α phosphatase activity by occluding the substrate-binding platform of the B55α regulatory subunit. In complex with PP2A/B55α, IER5 masks an extensive region on the surface of B55α that faces the catalytic subunit but does not directly contact the catalytic subunit itself.

Substrate recruitment to B55α relies primarily on short linear motifs (SLiMs), which may adopt an alpha helical conformation when bound10. Mapping of residues on B55α that reportedly participate in substrate recruitment shows substantial overlap with the IER5 binding site10–12 (Extended Data Fig. 6). The mode of IER5 binding differs, however, from that predicted for the inhibitors FAM122A and ARPP19, which have both been proposed to engage the B55α subunit with a short helical motif and bridge to the catalytic subunit using a discontinuous unstructured segment26.

Extended Data Fig. 6:
Extended Data Fig. 6: Overlap of the IER5 binding site with residues implicated in binding of different substrates.

The binding epitope of IER5 (a) or residues implicated in binding of pTau (b), p107 (c) and PRC1 (d) inferred from mutational studies10–12 are shown on the B55α surface using blue, magenta, red and yellow, respectively.

In our studies with purified complexes, we observed incomplete inhibition of pTau dephosphorylation (Fig. 1). This finding may stem from limitations in our assay for inhibition of PP2A/B55α. Because we were unable to purify IER5 for addition in trans to purified PP2A/B55α, it was necessary to directly compare the activity of free PP2A/B55α with the IER5-complexed form of PP2A/B55α, purified separately. It was thus possible that some of the IER5 dissociated from the complex during storage or during the course of the reaction. We also cannot exclude the possibility that phosphorylated serines or negatively charged glutamates present in a conserved region near the C-terminus of IER5 (Extended Data Fig. 7), which is predicted to be natively unstructured, are used to occupy the active site of the C-subunit and further potentiate inhibition. Alternatively, as others have proposed, it is also possible that the exposed face of IER5 does not preclude some access of pTau to the PP2A active site, and is even able, under certain circumstances, to bind to a different cohort of proteins than the B55α substrate-recruitment platform, and thereby alter the substrate selectivity of the phosphatase when complexed with IER522.

Extended Data Fig. 7:
Extended Data Fig. 7: Conservation analysis of IER5.

Cartoon representation of the alphafold model of IER544, colored by conservation according to the Consurf server45 on a sliding scale from teal (not conserved) to maroon (most conserved).

The immediate early response genes also include a family member called IER3 (also known as IEX-1), which is closely related to IER5. IER3 inhibits the activity of the PP2A/B56 isoform by promoting the dissociation of B56 from the active enzyme27,28. Of interest, IER3 also appears to be a direct target of Notch signaling in SCC cells9. The induction of different immediate early response proteins by Notch and cell stresses such as radiation-induced DNA damage may serve to induce cell cycle arrest and differentiation of squamous cells through the coordinated inhibition of multiple PP2A holoenzyme species.

Lastly, there is interest in development of PP2A modulators for diseases ranging from neurodegeneration to cancer29,30. The IER5 contact site could serve as a target surface for developing protein-protein interaction (PPI) inhibitors that selectively block substrate recruitment to PP2A/B55α complexes, or conversely, for molecular glue modulators that can direct dephosphorylation of specific proteins, analogous to protein-degrading IMiDs31.

Funding

This work was supported by NIH award 1R35 CA220340 (to SCB), the Ludwig Center at Harvard (JCA), and the Blavatnik Institute of Harvard Medical School.

Author Contributions

JCA and SCB conceived the project and acquired funding. RC purified PP2A/B55α-IER5 complexes, acquired biochemical data, prepared samples for cryo-EM data collection, and collected cryo-EM images. DTDJ processed and analyzed cryo-EM data with input from SR and RC, and DTDJ built the structural models. LP performed co-immunoprecipitation studies and KRT1 expression analyses. All authors participated in data analysis and interpretation. SCB, RC, DTDJ and JCA wrote and edited the manuscript with input from all authors. All authors agreed on the final manuscript.

Competing Interests

SCB is on the board of directors of the non-profit Institute for Protein Innovation and the Revson Foundation, is on the scientific advisory board for and receives funding from Erasca, Inc. for an unrelated project, is an advisor to MPM Capital, and is a consultant for IFM, Scorpion Therapeutics, Odyssey Therapeutics, Droia Ventures, and Ayala Pharmaceuticals for unrelated projects. JCA is a consultant for Ayala Pharmaceuticals, Cellestia, Inc., SpringWorks Therapeutics, and Remix Therapeutics. The other authors declare that they have no competing interests.

Methods

Plasmid construction

cDNAs encoding PPP2R1A, PPP2CA and PPP2R2A (B55α) assembled in a pAC-derived baculovirus expression vector (pAC8RedNK) were gifts from the Fischer lab (Dana Farber Cancer Institute). PPP2R1A was engineered to include an N-terminal His6 tag followed by a Tobacco etch virus (TEV) cleavage site, the cDNA for PPP2R2A (B55α) had no affinity tag, and PPP2CA had an N-terminal Flag tag followed by a TEV cleavage site. The IER5 (1-50) fragment was subcloned into pAC8RedNK with an N-terminal Strep tag II followed by a TEV cleavage site. The full-length Tau protein was cloned into the bacterial expression vector ptd68, incorporating an N-terminal His6-SUMO tag. Insert sequences were confirmed by Sanger sequencing.

Protein Expression

The PP2A-B55α heterotrimer and the PP2A-IER5 complex were expressed in Sf9 cells. 1.5 µg of the pAC8RedNK vector and 0.5 µg of linearized baculoviral DNA were concurrently transfected into 1×106 Sf9 cells using 6 µl of FuGene HD (Promega) in ESF 921 Insect Cell Culture medium (Expression Systems). After 5 to 7 days of incubation, the supernatant was collected to harvest the baculovirus. The virus was subsequently amplified for 2 to 3 cycles using Sf9 cells at a concentration of 2×106. The collected viruses were then used to infect Hi5 cells for protein expression. For expression of the PP2A heterotrimer, a viral stoichiometric ratio of 1:1:1 was used for Hi5 cell infection. For the PP2A-IER5 complex, a ratio of 1:1:1:1.5 (IER5) was used. Cells were shaken at 27 °C for 72 hours before harvesting by centrifugation. Cell pellets were collected and stored at −80 °C until purification.

Recombinant Tau protein was expressed in E. coli BL21 (DE3) cells. Protein expression was induced at a culture optical density (OD) of 0.8 by addition of 0.2 mM isopropyl-1-thio-D-galactopyranoside (IPTG) and the culture was maintained at 16°C overnight. Cell pellets were collected and stored at −80 °C until purification.

Protein purification

For the PP2A heterotrimer, cells were resuspended in lysis buffer containing 20mM Tris-HCl, pH 7.6, 200mM NaCl, 2mM Tris-(2-carboxyethyl) phosphine (TCEP), 0.1% (v/v) Triton X-100, protease inhibitor cocktail (Sigma) and Benzonase (EMD Millipore). Cells were lysed by sonication and centrifuged at 50,000 g for 1 h. The soluble fraction was passed over an anti-FLAG M2 affinity resin. The resin was washed with 10 column volumes (CVs) of wash buffer (20 mM Tris-HCl, pH 7.6, 200 mM NaCl, 2 mM TCEP), and the protein was then eluted using wash buffer supplemented with 0.2 mg/ml of FLAG peptide. The elution fractions were collected, concentrated, and further purified using size-exclusion chromatography (SEC) on a Superdex S200 10/300 column, which was pre-equilibrated with buffer (20 mM Tris-HCl, pH 7.6, 200 mM NaCl, 2 mM TCEP). Protein purity was assessed by SDS-PAGE using a Coomassie blue stain. Peak fractions were pooled for biochemical studies.

For the PP2A-IER5 complex, lysis and ultracentrifugation was performed as above. The affinity purification step was performed using Strep-Tactin XT Superflow resin (IBA), followed by elution with wash buffer supplemented with 50 mM biotin. After elution from the column the fractions were concentrated and further purified using size exclusion chromatography on a Superdex S200 10/300 column as above. Protein purity was assessed by SDS-PAGE using a Coomassie blue stain. Peak fractions were pooled for biochemical studies and for Cryo-EM data collection.

For preparation of Tau protein, bacterial cells were resuspended in lysis buffer containing 20 mM Tris-HCl, pH 8, 200 mM NaCl, 2 mM Tris-(2-carboxyethyl) phosphine (TCEP), protease inhibitor cocktail (Sigma), 20 mM Imidazole. Cells were lysed by sonication and centrifuged to remove debris. After centrifugation, the supernatant containing the recombinant His-SUMO-Tau was applied to Ni(NTA) affinity resin, and the resin was washed using the lysis buffer. The His-SUMO tag was removed overnight by on-column cleavage with ULP1 protease. Protease-liberated Tau was recovered from the column flow-through and was further purified by size exclusion chromatography using a Superdex 75 10/300 GL (GE Healthcare) column, in buffer consisting of 20 mM Tris-HCl pH 8, 200 mM NaCl, and 2mM TCEP. Peak fractions were pooled for phosphorylation using GSK3β. The concentration of purified Tau was determined by UV absorbance at 280 nm.

Phosphorylation of Tau protein and purification of pTau

Purified Tau was incubated with GSK3β (SinoBiological) in a 100:1 ratio, using a buffer of 25 mM HEPES pH 7.5, containing 100 mM NaCl, 10 mM MgCl2, 10 mM ATP, and 2 mM TCEP. The reaction was allowed to proceed for 19 hours at 37 °C. Verification that the reaction had gone to completion was performed by SDS-PAGE with Coomassie blue staining. The pTau was further purified by size exclusion chromatography using a Superdex 75 10/300 GL (GE Healthcare) column in buffer consisting of 20 mM Tris-HCl pH 8, 200 mM NaCl, and 2mM TCEP. The concentration of purified pTau was determined by UV absorbance at 280 nm. Fractions were collected and frozen in −80 prior to use.

Cryo-EM grid preparation and data collection

Samples were frozen on Cryo-EM grids using a Vitrobot Mark IV (Thermo Fisher Scientific) instrument for vitrification. 3.5 µl of freshly purified PP2A-IER5 complex, at a concentration of 2 mg/ml, was deposited onto glow-discharged C-flat holey carbon grids (R1.2/1.3, 400 mesh copper, Electron Microscopy Sciences). These grids were blotted for 6 seconds with a blot force of 15 at 100% humidity and 22 degrees C before being rapidly submerged into liquid ethane that had been cooled by liquid nitrogen for plunge freezing.

Images were acquired on a Titan Krios microscope equipped with a BioQuantum K3 Imaging Filter (slit width 25 eV) and a K3 direct electron detector (Gatan), operating at an acceleration voltage of 300 kV. Images were recorded at a defocus range of −0.8 to −2.0 μm with a nominal magnification of 105 kx, resulting in a pixel size of 0.825 Å. Each image was dose fractionated into 51 movie frames with a total exposure time of 2.8 s, resulting in a total dose of ∼53.7 electrons per Å2. SerialEM was used for data collection.

Structure Determination

Data were processed using CryoSPARC32, as summarized in Extended Data Fig. 3. A total of 5383 micrographs were subjected to patch motion correction and patch CTF estimation. 3.95 million particles were identified and extracted using a box size of 360 pixels using the CryoSPARC blob picker. 2D classification was used to remove poorly classifying particles, reducing the total number of particles to 1.17 million. From the projected 2D class averages, a mixture of monomeric and dimeric PP2A-IER5 assemblies was evident. 100,000 particles were split and used to generate three preliminary maps to represent the different species present in the data. The three maps were subjected to heterogeneous refinement using all identified particles, resulting in three classes designated as monomeric (∼47 %), dimeric (∼25 %) and junk (∼27 %) (e.g., scaffolding subunit only) classes. The monomeric and dimeric classes were then processed independently. Non-uniform refinement, local motion correction, and then another round of non-uniform refinement was applied to the monomer class, yielding a map with a nominal resolution of 3.67 Å, with reasonable anisotropy (cryoEF = 0.7133). The monomer class displayed no clear IER5 density at this stage of processing. Non-uniform refinement, local motion correction, and then another round of non-uniform refinement was applied to the dimeric class, resulting in a resolution of 3.67 Å for the dimer class map. Alignment of the dimeric map along an apparent C2 symmetry axis followed by non-uniform refinement under a C2 symmetry constraint improved the map to 3.44 Å resolution. A further round of local motion correction and two rounds of sequential non-uniform refinement produced a map of 3.17 Å resolution. To resolve symmetry breaking features, symmetry expansion was performed along the C2 symmetry axis. Masked local refinement with a soft mask around a single copy from an asymmetric unit, produced a map at 3.04 Å resolution. Despite this reasonable global resolution, the map displayed strong directional anisotropy (cryoEF = 0.4633).

Masked non-uniform refinement using a single copy of an asymmetric unit of the symmetry expanded dimer map, using the monomeric particles as input, resulted in a 3.82 Å map with improved signal for IER5. We then combined all the aligned monomer and symmetry expanded dimer particles and performed a masked local refinement that resulted in a combined map of 3.09 Å resolution with improved directional isotropy (cryoEF = 0.68). We finally performed a local refinement on the combined map with a mask that enveloped all IER5, B55α, and the scaffolding subunit HEAT repeats 1-9 (residues 1-355), resulting in a focused map of 3.13 Å resolution (cryoEF = 0.71).

Model refinement and atomic model building

Initial models were generated using alphafold. The models were fitted to the map using MOLREP34, and the initial model-to-map fit was improved using ISOLDE35. The model was refined using REFMAC Servalcat36, preliminarily using restraints generated from PDB 3DW8 using ProSMART37, followed by sequential rounds of jelly body refinement. The model was refined in Phenix using Phenix Real-Space Refine38 before manual inspection and refinement in Coot39. The final models were evaluated using MolProbity40. Statistics of the map reconstruction and model refinement are presented in Extended Data Table 1. Structural biology applications used in this project (except CryoSPARC) were compiled and configured by SBGrid41. Molecular graphics and analyses were performed with UCSF ChimeraX42 (developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from National Institutes of Health R01-GM129325 and the Office of Cyber Infrastructure and Computational Biology, National Institute of Allergy and Infectious Diseases), or PyMOL (Schrödinger).

View this table:
Table 1 Cryo-EM data collec3on, refinement and valida3on sta3s3cs.

Cell culture

Cells were grown under 5% CO2 at 37°C in media supplemented with streptomycin/penicillin. IER5 knock-out cell line I59 was derived from SC2 cells, which are engineered to contain a cDNA encoding a mutated truncated form of NOTCH1, ΔEGF-L1596H, that is regulatable with a ψ-secretase inhibitor (GSI) 9. I5 and its derivatives overexpressing wide-type and mutant IER5 were cultured in Keratinocyte medium as described43 in the presence of GSI (1μM compound E) to maintain Notch in the off-state. Timed activation of Notch was triggered by GSI washout as described 9.

Expression Constructs, Viral Production and Infection of Cells

The expression vector MIEG3-IER5-FH was constructed by inserting human IER5 tagged with HA and 3XFLAG into MIEG3 vector (Pan et al), which is a murine stem cell virus (MSCV)-based bicistronic retroviral construct expressing EGFP. MIEG3 expression vectors containing IER5 mutations were generated by QuickChange II Site-Directed Mutagenesis Kit (Agilent).

Retrovirus was prepared by transfecting Phoenix-gp cells with the MIEG3 vector and expression vectors for Gag/Pol and GALV. Viral supernatant was collected 48h posttransfection, centrifuged, and filtered through a 0.45 µm filter (Corning). For infection of target cells, 1 ml of virus was mixed with cells and protamine sulfate in a 6-well plate, and the plate was then centrifuged at 2,250 rpm for 90 min at room temperature. GFP-expressing cells were then isolated by cell-sorting 24 hr after transduction as described 9.

Quantitative RT-PCR

After 72 hours of GSI washout, cells were resuspended in Trizol (Life Technologies), and total RNA was prepared with RNeasy Mini kit (Qiagen). cDNA was synthesized with the High Capacity cDNA Reverse Transcripition Kit (Applied Biosystems). PCR was performed using the PowerUP SYBR Green Master Mix (Applied Biosystems) with the QuantStudio 3 Real-Time PCR System (Applied Biosystems). Primers used for KRT1 and GAPDH are: forward 5’-GGACAGCTCCTTAGCATCTTATC-3’, reverse 5’-GGAGTTTAAGACCTCTCCACAAA-3’ and forward 5’-GAAGGTGAAGGTCGGAGTCAAC-3’, reverse 5’-TGGAAGATGGTGATGGGATTTC-3’, respectively.

Immunoprecipitation and Western Blotting

For immunoprecipitation assays, cells in 10-cm dishes were washed in cold PBS and lysed in 1 ml of Pierce IP Lysis Buffer (Thermo Scientific) supplemented with protease inhibitors (Sigma). Cell lysates with equal amounts of protein were incubated with 25µl of washed Pierce Anti-HA Magnetic Beads (Thermo Scientific) overnight at 4°C with mixing. The beads were then washed three times with TBS-T and once with ultrapure water. The HA-tagged IER5 and its associated proteins were eluted with Pierce HA peptide (Themo Scientific). The eluates and input proteins were loaded on 3-8% SDS-polyacrylamide gels and resolved by electrophoresis. Proteins were incubated at 4 C overnight following transfer to nitrocellulose membranes with the following primary antibodies: anti-PP2A B55 (100C1) and anti-HA (C29F4) (both from Cell Signaling Technology); anti-IER5 (HPA029894) and anti-Flag (F3165) (both from Sigma). Secondary antibody was either goat anti-rabbit (7074) or horse anti-mouse (7076) IgG conjugated with horseradish peroxidase (Cell Signaling Technology). Staining was developed with SuperSignal West Dura Extended Duration Substrate (Thermo Scientific) for 2min at room temperature and documented by exposure to x-ray film.

Acknowledgements

We thank the Cryo-EM Center for Structural Biology at Harvard Medical School for help and advice with data collection, and members of the Blacklow lab and Alan Brown for helpful discussions.

References

  1. 1.↵
    Grallert, A. et al. A PP1-PP2A phosphatase relay controls mito8c progression. Nature 517, 94–98 (2015).
    OpenUrlCrossRefPubMed
  2. 2.↵
    Krasinska, L. et al. Protein phosphatase 2A controls the order and dynamics of cell-cycle transi8ons. Mol Cell 44, 437–450 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  3. 3.
    Swartz, S. Z. et al. Selec8ve dephosphoryla8on by PP2A-B55 directs the meiosis I-meiosis II transi8on in oocytes. Elife 10 (2021).
  4. 4.
    Kalev, P. et al. Loss of PPP2R2A inhibits homologous recombina8on DNA repair and predicts tumor sensi8vity to PARP inhibi8on. Cancer Res 72, 6414–6424 (2012).
    OpenUrlAbstract/FREE Full Text
  5. 5.
    Wurzenberger, C. & Gerlich, D. W. Phosphatases: providing safe passage through mito8c exit. Nature Reviews Molecular Cell Biology 12, 469–482 (2011).
    OpenUrlCrossRefPubMed
  6. 6.↵
    Schmitz, M. H. A. et al. Live-cell imaging RNAi screen iden8fies PP2A-B55alpha and impor8n-beta1 as key mito8c exit regulators in human cells. Nat Cell Biol 12, 886–893 (2010).
    OpenUrlCrossRefPubMedWeb of Science
  7. 7.↵
    Gharbi-Ayachi, A. et al. The substrate of Greatwall kinase, Arpp19, controls mitosis by inhibi8ng protein phosphatase 2A. Science 330, 1673–1677 (2010).
  8. 8.↵
    Li, F. et al. CHK1 Inhibitor Blocks Phosphoryla8on of FAM122A and Promotes Replica8on Stress. Mol Cell 80, 410–422 e416 (2020).
    OpenUrl
  9. 9.↵
    Pan, L. et al. IER5, a DNA damage response gene, is required for Notch-mediated induc8on of squamous cell differen8a8on. Elife 9 (2020).
  10. 10.↵
    Fowle, H. et al. PP2A/B55alpha substrate recruitment as defined by the re8noblastoma-related protein p107. Elife 10 (2021).
  11. 11.
    Cundell, M. J. et al. A PP2A-B55 recogni8on signal controls substrate dephosphoryla8on kine8cs during mito8c exit. J Cell Biol 214, 539–554 (2016).
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    Xu, Y., Chen, Y., Zhang, P., Jeffrey, P. D. & Shi, Y. Structure of a Protein Phosphatase 2A Holoenzyme: Insights into B55-Mediated Tau Dephosphoryla8on. Mol Cell 31, 873–885 (2008).
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    Amin, P. et al. PP2A-B55: substrates and regulators in the control of cellular func8ons. Oncogene 41, 1–14 (2022).
    OpenUrlCrossRefPubMed
  14. 14.↵
    Huang, K. L. et al. Integrator Recruits Protein Phosphatase 2A to Prevent Pause Release and Facilitate Transcrip8on Termina8on. Mol Cell 80, 345–358 e349 (2020).
    OpenUrl
  15. 15.↵
    Zheng, H. et al. Iden8fica8on of Integrator-PP2A complex (INTAC), an RNA polymerase II phosphatase. Science 370 (2020).
  16. 16.↵
    Williams, M. et al. Ier5, a novel member of the slow-kine8cs immediate-early genes. Genomics 55, 327–334 (1999).
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    Ding, K. K. et al. Induced expression of the IER5 gene by gamma-ray irradia8on and its involvement in cell cycle checkpoint control and survival. Radiat Environ Biophys 48, 205–213 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.
    Ishikawa, Y. & Sakurai, H. Heat-induced expression of the immediate-early gene IER5 and its involvement in the prolifera8on of heat-shocked cells. FEBS J 282, 332–340 (2015).
    OpenUrlCrossRef
  19. 19.↵
    Kis, E. et al. Microarray analysis of radia8on response genes in primary human fibroblasts. Int J Radiat Oncol Biol Phys 66, 1506–1514 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  20. 20.↵
    Asano, Y. et al. IER5 generates a novel hypo-phosphorylated ac8ve form of HSF1 and contributes to tumorigenesis. Sci Rep 6, 19174 (2016).
  21. 21.
    Ishikawa, Y., Kawabata, S. & Sakurai, H. HSF1 transcrip8onal ac8vity is modulated by IER5 and PP2A/B55. FEBS LeE 589, 1150–1155 (2015).
    OpenUrl
  22. 22.↵
    Kawabata, S., Ishita, Y., Ishikawa, Y. & Sakurai, H. Immediate-early response 5 (IER5) interacts with protein phosphatase 2A and regulates the phosphoryla8on of ribosomal protein S6 kinase and heat shock factor 1. FEBS LeE 589, 3679–3685 (2015).
    OpenUrl
  23. 23.↵
    Evans, R., et al. Protein complex predic8on with AlphaFold-Mul8mer. bioRxiv (2022).
  24. 24.↵
    Stanevich, V. et al. The structural basis for 8ght control of PP2A methyla8on and func8on by LCMT-1. Mol Cell 41, 331–342 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  25. 25.↵
    Tolstykh, T., Lee, J., Vafai, S. & Stock, J. B. Carboxyl methyla8on regulates phosphoprotein phosphatase 2A by controlling the associa8on of regulatory B subunits. EMBO J 19, 5682–5691 (2000).
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    Wasserman, J. S., et al. FAM122A ensures cell cycle interphase progression and checkpoint control as a SLiM-dependent substrate-compe88ve inhibitor to the B55⍺/PP2A phosphatase. bioRxiv (2023).
  27. 27.↵
    Letourneux, C., Rocher, G. & Porteu, F. B56-containing PP2A dephosphorylate ERK and their ac8vity is controlled by the early gene IEX-1 and ERK. EMBO J 25, 727–738 (2006).
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    Rocher, G., Letourneux, C., Lenormand, P. & Porteu, F. Inhibi8on of B56-containing protein phosphatase 2As by the early response gene IEX-1 leads to control of Akt ac8vity. J Biol Chem 282, 5468–5477 (2007).
    OpenUrlAbstract/FREE Full Text
  29. 29.↵
    Clark, A. R. & Ohlmeyer, M. Protein phosphatase 2A as a therapeu8c target in inflamma8on and neurodegenera8on. Pharmacol Ther. 201, 181–201 (2019).
    OpenUrlCrossRefPubMed
  30. 30.↵
    Ronk, H., Rosenblum, J. S., Kung, T. & Zhuang, Z. Targe8ng PP2A for cancer therapeu8c modula8on. Cancer Biol Med 19, 1428–1439 (2022).
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    Fischer, E. S. et al. Structure of the DDB1-CRBN E3 ubiqui8n ligase in complex with thalidomide. Nature 512, 49–53 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  32. 32.↵
    Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determina8on. Nat Methods 14, 290–296 (2017).
    OpenUrlCrossRefPubMed
  33. 33.
    Naydenova, K. & Russo, C. J. Measuring the effects of par8cle orienta8on to improve the efficiency of electron cryomicroscopy. Nat Commun 8, 629 (2017).
  34. 34.↵
    Vagin, A. & Teplyakov, A. MOLREP: an Automated Program for Molecular Replacement. Journal of Applied Crystallography 30, 1022–1025 (1997).
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    Croll, T. I. ISOLDE: a physically realis8c environment for model building into low-resolu8on electron-density maps. Acta Crystallogr D Struct Biol 74, 519–530 (2018).
    OpenUrl
  36. 36.↵
    Yamashita, K., Palmer, C. M., Burnley, T. & Murshudov, G. N. Cryo-EM single-par8cle structure refinement and map calcula8on using Servalcat. Acta Crystallogr D Struct Biol 77, 1282–1291 (2021).
    OpenUrl
  37. 37.↵
    Nicholls, R. A., Fischer, M., McNicholas, S. & Murshudov, G. N. Conforma8on-independent structural comparison of macromolecules with ProSMART. Acta Crystallogr D Biol Crystallogr 70, 2487–2499 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  38. 38.↵
    Afonine, P. V. et al. Real-space refinement in PHENIX for cryo-EM and crystallography. Acta Crystallogr D Struct Biol 74, 531–544 (2018).
    OpenUrl
  39. 39.↵
    Casanal, A., Lohkamp, B. & Emsley, P. Current developments in Coot for macromolecular model building of Electron Cryo-microscopy and Crystallographic Data. Protein Sci 29, 1069–1078 (2020).
    OpenUrl
  40. 40.↵
    Williams, C. J. et al. MolProbity: More and beker reference data for improved all-atom structure valida8on. Protein Sci 27, 293–315 (2018).
    OpenUrlCrossRefPubMed
  41. 41.↵
    Morin, A. et al. Collabora8on gets the most out of solware. Elife 2, e01456 (2013).
    OpenUrlCrossRefPubMed
  42. 42.↵
    Pekersen, E. F. et al. UCSF ChimeraX: Structure visualiza8on for researchers, educators, and developers. Protein Sci 30, 70–82 (2021).
    OpenUrlCrossRef
  43. 43.↵
    Purdie, K. J., Pourreyron, C. & South, A. P. Isola8on and culture of squamous cell carcinoma lines. Methods Mol Biol 731, 151–159 (2011).
    OpenUrlCrossRefPubMed
  44. 44.↵
    Jumper, J. et al. Highly accurate protein structure predic8on with AlphaFold. Nature 596, 583–589 (2021).
    OpenUrlCrossRefPubMed
  45. 45.↵
    Ashkenazy, H. et al. ConSurf 2016: an improved methodology to es8mate and visualize evolu8onary conserva8on in macromolecules. Nucleic Acids Res 44, W344–350 (2016).
    OpenUrlCrossRefPubMed
Back to top
Posted August 29, 2023.
Download PDF
Email
Molecular Mechanism of PP2A/B55α Inhibition by IER5
Ruili Cao, Daniel TD Jones, Li Pan, Shumei Wang, Shaun Rawson, Jon C Aster, Stephen C Blacklow
bioRxiv 2023.08.29.555174; doi: https://doi.org/10.1101/2023.08.29.555174
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area