Disruption of the productive encounter complex results in dysregulation of DIAPH1 activity

The diaphanous-related formin, Diaphanous 1 (DIAPH1), is required for the assembly of Filamentous (F)-actin structures. DIAPH1 is an intracellular effector of the receptor for advanced glycation end products (RAGE) and contributes to RAGE signaling and effects such as increased cell migration upon RAGE stimulation. Mutations in DIAPH1, including those in the basic “RRKR” motif of its autoregulatory domain, diaphanous autoinhibitory domain (DAD), are implicated in hearing loss, macrothrombocytopenia, and cardiovascular diseases. The solution structure of the complex between the N-terminal inhibitory domain, DID, and the C-terminal DAD, resolved by NMR spectroscopy shows only transient interactions between DID and the basic motif of DAD, resembling those found in encounter complexes. Cross-linking studies placed the RRKR motif into the negatively charged cavity of DID. Neutralizing the cavity resulted in a 5-fold decrease in the binding affinity and 4-fold decrease in the association rate constant of DAD for DID, indicating that the RRKR interactions with DID form a productive encounter complex. A DIAPH1 mutant containing a neutralized RRKR binding cavity shows excessive colocalization with actin and is unresponsive to RAGE stimulation. This is the first demonstration of a specific alteration of the surfaces responsible for productive encounter complexation with implications for human pathology.

Although no structural information is available for human DIAPH1, structures of the murine homolog of DIAPH1, with 90.3% sequence identity to human DIAPH1 (14), provide a guide to understanding the disparate functions and regulation of DIAPH1.Crystal structures of complexes between the N-and C-terminal fragments of mouse DIAPH1, also called mDia1, present a glimpse into the autoinhibited state (11,(15)(16)(17).In DAD, the segment directly N terminal to the RRKR motif forms a helix and a turn, anchored at F1195, that binds to DID to maintain mouse DIAPH1 in an autoinhibited state.The RRKR motif, which is also found in many unrelated proteins, such as coronavirus spike protein (18), proplatelet-derived growth factor A (19), chloroplast signal recognition particle 54 (20), and nuclear valosin-containing protein-like 2 (21), is likely flexible because it is either absent (11,15) or poorly resolved (16,17) but is oriented proximal to negatively charged patches on DID (22,23) and contributes to the binding (11).It is not clear why mutations of this flexible region of DAD result in DIAPH1 activation (22,24).In addition, the T-helix in mouse DIAPH1 blocks the negatively charged patches on DID, thereby precluding transient interactions between DID and the RRKR motif of DAD (15,17).Crystal structures of the N-and C-terminal fragments of mouse DIAPH1 display different angles between DID and DD in mouse DIAPH1 dimers, suggesting flexibility in the IH that links DID and DD (11,15,23,25).
To understand the structural determinants of DIAPH1 regulation, DID-DAD interactions were characterized by using solution NMR spectroscopy, cross-linking with mass spectrometry (MS), fluorescence spectroscopy, and functional assays.Flexible regions of DID were identified in the IH.The results also revealed that binding of RRKR to DID is transient  M1199L acquires structure upon binding to DID (Fig. S1A).Peaks of the RRKR motif residues are labeled in red.C, 1 H- 15 N HSQC spectrum of [U- 15 N]-DID bound to DAD M1199L indicates that the complex is structured.Peaks labeled in blue correspond to conformer B. Blue peaks were used in thermodynamic calculations.DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; HSQC, heteronuclear single-quantum coherence.and forms a productive encounter complex (33,34) that does not compete with the T-helix.These observations allowed a consistent picture of DIAPH1 regulation to be proposed.

DID-DAD M1199L complex stoichiometry and affinity
The DID-DAD interaction was characterized by using a combination of solution NMR spectroscopy and cross-linking MS (Figs. 1 and 2).Residues 142 to 380 of DIAPH1, corresponding to DID, and residues 1194 to 1222, corresponding to DAD, were bacterially overexpressed for labeling and NMR spectroscopy (Fig. 1A).DAD was expressed as an insoluble fusion protein containing an N-terminal his-tag and modified tryptophan leader sequence (35,36), and the resulting gene product accumulated in inclusion bodies.DAD methionine 1199 was conservatively (37) changed to a leucine, DAD M1199L , to ensure that only a single cyanogen bromide cleavage site was present so that cleavage of the fusion protein would yield soluble DAD M1199L peptide.This substitution did not change the binding to DID (Figs.S1 and S2).
The stoichiometry of the interaction between DID and DAD M1199L was determined by titrating 100 μM [U- 15 N]-DAD M1199L with unlabeled DID (Figs. 1B and S1A).Chemical shifts (δ) and cross peak intensities (I) were monitored by using heteronuclear single-quantum coherence ( 1 H- 15 N HSQC) NMR spectroscopy (38) (Fig. S1A).Changes in δ and I were observed at substoichiometric amounts of unlabeled DID.Negligible perturbations occurred beyond a 1:1 mole ratio, indicating a 1:1 stoichiometry for the complex.The strength of the DID-DAD M1199L interaction was measured using tryptophan fluorescence spectroscopy and yielded a macroscopic binding constant of 270 ± 100 nM, in agreement with that reported for WT DAD (11) (Fig. S1B).

DAD M1199L peptide acquires structure when bound to DID
The 1 H-15 N HSQC, spectrum of [U- 15 N]-DAD M1199L , is shown in Figure 1B; 27 out of 29 residues were assigned.N nOes of DAD M1199L bound to DID showed that the structural region of DAD M1199L spans from G1197 to R1213 (Fig. S1A).B, steady state { 1 H}- 15 N nOe of the 13 C-terminal DID residues of IH showed the presence of two conformers: a structured conformer A (white squares) and an unstructured conformer B (red circles).C, van't Hoff analysis of selected DID C-terminal residues of IH showed an entropically driven equilibrium between conformers A and B. The temperature dependence of the equilibrium constant, K eq , resolved ΔH, ΔS, and ΔG for the unstructured to structured transition (Table S1).D, high-energy collision MS spectrum obtained for the DSG cross-linked product between K228 of DID and the N terminus of DAD M1199L  Changes in amide resonances of DAD M1199L upon complex formation with DID suggest that free DAD M1199L undergoes a structural transition when bound (Fig. S1A).Steady-state heteronuclear { 1 H}- 15 N nuclear Overhauser effect ({ 1 H}- 15 N nOe) (38) was used to identify structured regions in bound DAD M1199L (Fig. 2A).Negative { 1 H}- 15 N nOes indicate a high degree of local flexibility and fast motions on the picosecond to nanosecond time scale (39) consistent with unstructured regions.In contrast, positive { 1 H}- 15 N nOes reflect reduced flexibility indicative of structured regions (39).The { 1 H}- 15 N nOe for the bound DAD M1199L showed that the structured region spans from G1197 to R1213 ({ 1 H}-15 N nOes >0.5), while the N and C termini are flexible (Fig. 2A).Importantly, bound DAD M1199L exhibited increasingly negative { 1 H}- 15 N nOes from the positively charged RRKR motif (residues 1213-1216) to the C terminus.

C terminus of DID IH has two conformations
The 1 H-15 N HSQC, spectrum of [U- 15 N]-DID in the presence of unlabeled DAD M1199L , is shown in Figure 1C.Two hundred twenty-five out of 234 (96%) residues were assigned.Importantly, the M1199L mutation resulted only in minor changes in the chemical shifts of DID, suggesting no major structural changes compared to the WT complex (Fig. S2).Cross peak doubling was observed for 13 C-terminal cross peaks of IH (Fig. 1C blue labels).{ 1 H}-15 N nOes acquired for [U- 15 N]-DID indicate that the C-terminal residues of IH conformer A exhibit restricted ps-ns dynamics suggesting that conformer A is structured (Fig. 2B).The C-terminal residues of IH conformer B exhibit increased ps-ns dynamics, suggesting that conformer B is unstructured (Fig. 2B).To assess the relative stability of these conformations, van't Hoff analyses of selected C-terminal residues of IH were performed by collecting 1 H-15 N HSQC spectra of [U- 15 N]-DID bound to DAD M1199L over the temperature range 280 to 310 K. Cross peaks G369, S373, Y374, and G378 were quantified and equilibrium constants for specific residues, K eq , were estimated as the ratio of the amide peak amplitudes of IH conformer A over IH conformer B (Fig. 2C).The negative slopes in Figure 2C showed that the unstructured to structured transition is endothermically unfavored and the y-intercepts in Figure 2C showed that the unstructured to structured transition is entropically favored.When combined, the IH conformer A trends to be thermodynamically unfavored over IH conformer B (Table.S1).The enhanced ps-ns dynamics of IH conformer B compared to conformer A, which results in a decrease in entropy, is likely compensated by the entropically favorable release of bound water, bound ions, or the structural rearrangement of other parts of DID during the B to A transition.

Solution structure of the DID-DAD M1199L complex
The solution structure of a 1:1 complex of DID-DAD M1199L , restricted to IH conformer A of DID, was resolved by using structural restraints derived from solution NMR and crosslinking MS (Figs. 1 and 2D, and Table S3).A standard suite of NMR experiments (38) was performed to assign backbone and side chain resonances of [U-15 N, 13 C]-DID and [U-15 N, 13 C]-DAD M1199L .To facilitate unambiguous assignment of DID residues, 1 H-15 N HSQC and 3D 1 H- 15 N NOESY (38) spectra of individual 15 N-labeled amino acids of DID bound to WT DAD were recorded at 305 K or 298 K. Intramolecular and intermolecular nOes from 3D 1 H-15 N NOESY spectra were collected for the DID-DAD M1199L complex and converted into upper limits for the proton-proton distances utilized in structural calculations.Distance constraints from MS analyses of DID cross linked to DAD M1199L were obtained by using amine-to-amine chemical cross-linkers.The crosslinked peptides were purified using SDS-PAGE and analyzed using tandem mass spectrometry (MS/MS).Cross-linkers of different lengths gave results for the same set of amines, indicating a flexible and transient structure of the N and C terminus, including the RRKR motif of DIAPH1.(Fig. S3 and Tables S3 and Fig. S4).For this reason, our structure calculations used cross-links that provided tighter distance constraints.A typical spectrum is shown in Figure 2D for the cross-linked product between the N terminus of DAD M1199L and K228 of DID.
Backbone chemical shifts of DID and DAD M1199L were used to predict acceptable ranges of backbone torsion angles per residue using Torsion Angle Likelihood Obtained from Shift and sequence similarity (TALOS)-N software (40).Structural calculations were performed by using Combined assignment and dYnamics Algorithm for NMR Applications (CYANA) (41), followed by refinement using Yet Another Scientific Artificial Reality Application (YASARA) (42) in combination with 901 DID intramolecular nOes, 94 DAD M1199L intramolecular nOes, 13 DID-DAD M1199L intermolecular nOes, (Table S2), two DID intramolecular cross-link distances, and two DID-DAD M1199L cross-link distances (Table S3, and Fig. S3).The torsion angles were combined with the distance restraints to yield an ensemble of 20 solution structures for the DID-DAD M1199L complex with a backbone RMSD of 0.3 Å for the helical regions (Fig. 3A), indicating high quality structural convergence.Notably, the loop between α2A and α2B, which is absent from the published structures, was well-resolved despite some conformational variability (Fig. 3A, dashed box).
DAD M1199L peptide residues V1198-S1208 formed an amphipathic α-helix that was shorter than previously resolved for mouse DIAPH1 (17,43) and contacted the concave hydrophobic surface comprised of the central B helices in DID (Fig. 3B).Residues G1209-R1213 turned perpendicular to the helix, anchoring at F1212 (F1195 in mouse DIAPH1), in agreement with published structures (11,(15)(16)(17), while the DIAPH1 and the productive encounter complex positively charged C terminus remained unstructured.The L1199 side chain had a dihedral angle of χ 1 = −160 , comparable to the −177 observed for the corresponding residue, M1182, in the crystal structure of the complex formed by the C-and N-terminal fragments of mouse DIAPH1 (16).L1199 also has similar χ 2 and χ 3 values and resides in a hydrophobic pocket contacting M225, I231, L262, and A265, in agreement with published structures (Fig. 3C) (11,(15)(16)(17) The results confirmed the NMR observations (Fig. S2) and demonstrated that the leucine substitution does not perturb the WT structure.
Encounter complexes are reaction intermediates that arise from an ensemble of conformations that can sample various orientations to properly align binding determinants (33,34,45).As postulated for other cases of encounter complexes (33,34,45), DAD residues R1213-R1216 are electrostatically steered (34,45) toward the negatively charged surface of DID.The RRKR motif does not exhibit NMR-visible interactions with DID and the entirety of this highly basic region is absent in the published structures.Cross-linking together with MS/ MS experiments showed that DAD M1199L K1215 is within 11.4 Å of DID K228 and 11.4 Å of DID K377 (Table S3).In addition, MS analysis yielded cross-links for the K228-K1215 pair obtained using a 21.7 Å linker (Table S4).Both DID lysine residues are too far apart to interact simultaneously with DAD M1199L K1215, indicating a high degree of flexibility for the C terminus of DAD M1199L in agreement with the NMR data.When either the DAD M1199L K1215/DID K377 or DAD M1199L K1215/DID K228 cross-link was used for structure calculations, the RRKR motif interacted with the acidic DID residues E326, E367, E371, and D375, in agreement with mutagenesis studies (23,25) (Figs.S4 and 3D).These residues constitute a negative patch on DID that interacts with the basic RRKR motif of DAD M1199L , suggesting that electrostatic steering may play a role in complex formation as previously described for productive encounter complexes (33,34,45,46).

T-helix does not impede DID-DAD M1199L interactions
Available structures of DRFs revealed a long helical structure preceding DAD M1199L , the T-helix, which also contacts DID (15,17).This helix was poorly resolved in the crystal structure suggesting flexibility (15,17).We used NMR spectroscopy to resolve this element and observed a well-dispersed 1 H-15 N HSQC spectrum of the T-helix, which indicated the presence of a folded structure (Fig. S5A and Table S5).Analyses of amide-amide cross-peaks in the 1 H-1 H NOESY spectrum (Fig. S5B) and chemical shift indices (Fig. S5C) revealed a transient helical structure in solution.Superimposing the DID-DAD M1199L structure from this work with the T-helix and DAD from mouse DIAPH1 (15) shows that the RRKR motif of DAD M1199L appears to sterically clash with the T-helix (Fig. S5D).Such an occlusion could prevent the formation of an encounter complex by masking the acidic patch on DID.Contribution of the RRKR motif to the overall binding affinity of DID for DAD M1199L can be estimated from binding studies 11 to be in the millimolar range, assuming that binding of the DAD helix and the RRKR motif to DID are cooperative.To probe this interaction, 120 μM of [U- 15 N]-DAD M1199L bound to unlabeled DID was titrated with 360 μM of purified T-helix and analyzed via NMR (Fig. S5E).The absence of chemical shift perturbations in [U- 15 N]-DAD M1199L indicates that there are no T-helix residues that compete with RRKR for binding to DID.

Double mutant DID E326G/E327A alters the affinity and kinetics of DAD M1199L binding
To biochemically test if the acidic patch on DID (Fig. 3D) is involved in electrostatic steering (34,45), mutations E326G/ E327A, DID E326G/E327A , were introduced.These mutations change the electrostatic character of the patch converting it from negatively charged to mostly neutral (Fig. 4A).The 1 H-15 N HSQC spectrum of [U-15 N]-DID E326G/E327A complexed with DAD M1199L exhibited a well-dispersed amide envelope and minor chemical shift changes (Fig. S6, A and B) as compared to that of the WT [U- 15 N]-DID-DAD M1199L complex, indicating that the tertiary structure of DID E326G/E327A is comparable to that of the WT.
To test if the DID negative patch is critical for DAD binding, tryptophan fluorescence titrations were performed with 200 nM DID E326G/E327A and DAD M1199L concentrations ranging from 10 to 90 μM (Fig. S1C).The K D resolved for DID E326G/E327A was 1.5 ± 0.5 μM, which is about five times higher than that for WT DID (Fig. S1B).This result suggests that the interaction between the RRKR motif of DAD M1199L   and the negatively charged patch of DID is specific and increases the overall affinity of DAD M1199L for DID.
Stopped flow fluorescence titrations were used to evaluate the effect of the E326G/E327A mutation on the DID-DAD M1199L association kinetics.Two hundred nanomolar (200 nM) DID and DID E326G/E327A were titrated with 1 to 20 μM DAD M1199L to create pseudo first order kinetics.The data were fit to a single exponential to estimate the observable rate constants, k obs (Fig. 4B).A k on of 52 ± 1.5 (μM s) −1 and a k off of 19 ± 17 s −1 were resolved for the DID-DAD M1199L interaction and a k on of 13 ± 1 (μM s) −1 and a k off of 27 ± 14 s −1 for the DID E326G/E327A -DAD M1199L interaction.The 4-fold decrease in k on suggests that the decreased binding affinity of DAD M1199L to DID E326G/E327A is primarily due to a decrease in the rate of association, in agreement with an electrostatic steering mechanism.

DAD M1199L also forms futile encounter complexes by nonspecific binding to DID
The model used to fit the tryptophan fluorescence data (Fig. S1) could not discriminate between specific and nonspecific interactions.To identify residues involved in nonspecific interactions, [U- 15 N]-DID was titrated with DAD M1199L and changes in the 1 H-15 N HSQC spectra were monitored (Fig. 5A).With a large molar excess of DAD M1199L , nonspecific binding interactions could be inferred from changes in DID chemical shifts.We observed minor chemical shift perturbations of DID residues (Fig. 5B), most of which were clustered in at least four noncontiguous areas on the DID surface (Fig. 5C).Thus, our chemical shift mapping reveals nonspecific binding sites for DAD M1199L that could give rise to so-called futile encounter complexes (34).

DIAPH1 and the productive encounter complex
Mutations in the negative patch of DID result in dysregulation of DIAPH1 activity Autoinhibited DIAPH1 is free and cytosolic; however, upon activation it binds to the ends of actin fibers (3,4).Deletion of the DAD sequence results in activation of DIAPH1 and increased actin binding (11,22,23).Importantly, deletion of the RRKR motif in DIAPH1 also results in elongation of microvilli with subsequent increased DIAPH1-actin colocalization (24).We hypothesized that mutations E326G/E327A in DID, which disrupt the binding of RRKR and thus the formation of the productive encounter complex, would also lead to increased DIAPH1-actin colocalization.
To test this hypothesis, the DAD deletion mutant (12) (DIAPH1 ΔDAD ), which was used as a positive control, and double mutant DIAPH1 E326G/E327A , were introduced into HEK293T cells.All constructs were expressed with an N-terminal GFP tag to monitor the location of the proteins by fluorescent microscopy.F-actin was tagged with phalloidin-Alexa Fluor 568 to measure the colocalization of DIAPH1-GFP constructs with F-actin (Fig. 6A).Manders coefficients, M1 and M2 (47), were used to quantify the colocalization of Factin with DIAPH1-GFP constructs and DIAPH1-GFP constructs with F-actin, respectively (Fig. 6B).As expected, both DIAPH1 ΔDAD -GFP and DIAPH1 E326G/E327A -GFP exhibited significant increases in colocalization with F-actin compared to that of the WT.
To further test the functional importance of the RRKR motif, we examined RAGE signal transduction (7)(8)(9).DIAPH1 is an intracellular effector of RAGE; stimulation of RAGE by RAGE ligand AGEs, such as carboxymethyllysine human serum albumin (CML-HSA) results in a DIAPH1-dependent increase in migration of vascular smooth muscle cells (SMCs) (7)(8)(9).Human SMCs were transfected with plasmidsexpressing yellow fluorescent protein (YFP), DIAPH1-YFP, DIAPH1 ΔDAD -YFP, and DIAPH1 E326G/E327A -YFP, and the cells were treated with 1 μM of CML-HSA (Fig. 6C).SMCs transfected with DIAPH1-YFP exhibited an insignificant increase in migration upon stimulation compared to those transfected with YFP, suggesting that endogenous DIAPH1, which is present in SMCs, is sufficient to support RAGE-dependent cell migration.SMCs transfected with the positive control DIA-PH1 ΔDAD -YFP exhibited a statistically significant reduction in cell migration upon stimulation, suggesting that the deletion of DAD disrupts RAGE signal transduction.SMCs transfected with DIAPH1 E326G/E327A -YFP also exhibited a statistically significant reduction of cell migration upon stimulation, although this reduction trended lower than that of the positive control.

Discussion
The solution structure of human DID-DAD M1199L elucidated features of autoinhibited human DIAPH1 that were unresolved in previous crystal structures of the mouse analog to reveal important aspects of DIAPH1 dynamics and its role in regulation.Part of the IH of DIAPH1, residues 367 to 380, exhibits an equilibrium between structured and unstructured conformers.The structured conformation is helical and similar to that detected in crystalline mouse DIAPH1 DID (11, 15-17, 23, 25,43,44).Thermodynamic data show that the structured form is likely favored but still retains enough flexibility observed in variability between DID and DD domains in crystalline forms of mouse DIAPH1 (11,15,23,25).Flexibility in the regions between the DID and the rest of DIAPH1 contributes to the formation of the autoinhibited state by increasing the probability of DID, encountering DAD in the context of full-length DIAPH1.
The lack of structural information on the DAD C terminus complicates efforts to understand atomic-level interactions between the basic RRKR motif and DID implicated in human pathologies (24,29,30,32).Previous studies have shown that the RRKR motif is not essential for DID-DAD binding but does contribute to the binding affinity (11,22,31).Removal of the RRKR motif decreases DID-DAD binding affinity by 5-fold (11), resulting in gain-of-function (24).Our NMR and crosslinking studies show that the RRKR motif is flexible in the DID-DAD M1199L complex (Tables S3 and S4).How could such a strong effector of K D be flexible?Findings from crosslinking and stopped flow experiments in this work reveal transient, yet specific, interactions involving RRKR, thus providing insight into a possible structural mechanism of autoinhibitory regulation.
Earlier structural studies revealed that Rho GTPase binding to DIAPH1 elicits a conformational change that displaces the DAD helix from DID (12,25,48).These studies showed that  (15,17,43,72).Active DIAPH1 forms a productive encounter complex via electrostatic steering, dotted line, between the RRKR motif (blue star) and the acidic patch of DID (red star).The interaction increases the probability of DID-DAD helix coupled folding and binding and formation of the autoinhibited conformation.Active DIAPH1 lacking the basic RRKR motif, such as in DIAPH-(R1213X) (22,31), cannot create a productive encounter complex resulting in an equilibrium shift toward the active complex.Note that deletion of DAD, such as in DIAPH1 ΔDAD , would preclude the autoinhibition.CC, coiled-coil; CML-HAS, carboxymethyllysine human serum albumin; DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; IH, interdomain helix; RAGE, receptor for advanced glycation end products; YFP, yellow fluorescent protein.
the binding causes only partial activation of DIAPH1, suggesting that other regulatory steps are involved in DIAPH1 activity.Our study further elaborates on this point by at least partially implicating the formation of an endogenous encounter complex in the regulation of DIAPH1 activity.
Previous mutagenesis studies showed that mutating DID residues E367R, E371R, or D375R (23), which are located in the vicinity of the RRKR-binding site (Figs.3D and S4), resulted in a 2-to 3-fold decrease in DID-DAD binding affinity, but E273K and R278E did not (25).Our findings identify E326 and E327 as critical for binding (Figs. 3D, 4 and S1).The E326 and E327 acidic patch is located greater than 30 Å away from the Rho GTPase binding site and represents a new surface for DIAPH1 regulation (Fig. S4B) (25).Interruption of the electrostatic interaction between the basic RRKR motif and these acidic residues decreased DID-DAD M1199L binding affinity by a factor of five.Due to the long-range nature of electrostatic interactions, the combined effect of E326G/ E327A mutations with those previously studied ( 23) is expected to be cumulative.Importantly, the E326G/E327A mutations that disrupt the RRKR binding surface also increased colocalization of actin and DIAPH1 in vivo (Fig. 6, A and B), consistent with previous observations that deletion of RRKR leads to DIAPH1 colocalization with actin in filopodia (24).
DIAPH1 is critically involved in RAGE signaling: SMC migration is increased when RAGE is activated by its ligand CML-HSA and this increase depends on RAGE-DIAPH1 signal transduction (8-10).Perhaps counterintuitively, the E326G/E327A mutation resulted in significant reduction of SMC migration.Note that efficient cellular migration relies on regulation of actin polymerization/depolymerization and, thus, dysregulation leads to reduced migration (49)(50)(51).This suggests that DIAPH1 E326G/E327A was not fully regulated by autoinhibition (Fig. 6C).An even larger reduction was observed in the SMCs expressing DIAPH1 ΔDAD where the regulatory DID-DAD interaction is completely removed; DIAPH1 ΔDAD was used as a positive control for dysregulated constitutively active DIAPH1.
Because the K D for DID-DAD M1199L increased when DID E326G/E327A mutations were introduced and the interaction remained transient, it is likely that DID E326G/E327A mutations affect primarily the association of DAD M1199L to DID.Kinetic experiments and NMR titrations suggest that RRKR interacts with the acidic residues on DID and this interaction is not obstructed by the T-helix (15,17).Altogether, these observations paint a dynamic picture.The conformational exchange between structured and unstructured in the IH segment of DID allows the DID to search more conformational space.This, in turn, allows the basic RRKR motif to interact with the acidic patch of DID facilitating the wobbling and searching (52), folding and binding (53) of the DAD helix.This interaction is absent in RRKR deletion mutants leading to shifting equilibrium to the active form of DIAPH1 (Fig. 6D) (22,31).Subsequent exchange of RRKR with the basic portion of the Thelix may then free the C terminus of DAD.This agrees with the current understanding that the RRKR motif interacts with actin and facilitates actin fiber nucleation (22,54).
Importantly, the mechanism proposed here describes a productive encounter complex between DID and DAD required for autoinhibition, which can be disrupted by mutations in DID or deletions in DAD leading to gain-of-function activity.This suggests that the disease states characterized by gain of function, such as autosomal dominant hearing loss (24) and macrothrombocytopenia (32), may arise from enhanced actin assembly mediated by DIAPH1 dysregulation.

DID cloning, expression, and purification
Expression plasmid pET-28a(+)-DID, which confers kanamycin resistance and expresses an N-terminal his-tagged DID, residues 142 to 380 from human DIAPH1 (UniProt O60610), was purchased from Genscript.pET28a(+)-DID was transformed into Escherichia coli strain BL21(DE3) and grown overnight at 37 C in 100 ml of LB containing 35 μg/ml of kanamycin (LB-Kn).The culture was transferred to 1 L of LB-Kn and incubated at 37 C until the A 550 reached 0.90.To overexpress DID without stable isotope labeling, IPTG was added to a final concentration of 1 mM and the cells were incubated for 4 h at 37 C.For isotopic labeling, cells were harvested by centrifugation at 4000g and resuspended in minimal (M9) medium, 5 mM Na 2 HPO 4 , 2.5 mM KH 2 PO 4 , pH 7, 1 mM NaCl, 2 mM MgSO 4 , 0.1 mM CaCl 2 , and 1 mg/l thiamine hydrochloride, supplemented with 35 μg/ml of kanamycin (M9-Kn).For [U-15 N] labeling, 1 g/l of 15 Nammonium chloride (Sigma-Aldrich) was used as the sole nitrogen source.For [U-15 N, 13 C] labeling 2 g/l of 13 C-glucose (Sigma-Aldrich) was used as the sole carbon source.Overexpression of labeled DID was induced with 1 mM IPTG and proceeded for 4 h at 37 C. Cells were centrifuged at 4000g, resuspended in lysis buffer, 20 mM Hepes, pH 8, 1 M NaCl, 12.5% w/v sucrose, 4 M urea, and 10 mM β-mercaptoethanol, containing an EDTA-free protease inhibitor tablet (Roche) and sonicated with a Model 250 Digital Sonifier (Branson) at 40% amplitude using 2.5-min cycles of 0.3 s pulses interrupted by 1 s pauses.The lysate was centrifuged for 40 min at 20,000g and 4 C.The supernatant was loaded into a nickelnitriloacetic acid agarose affinity column (Qiagen) preequilibrated with lysis buffer.The partially denatured protein bound to the column was refolded by washing with five column volumes of refolding buffer, 10 mM Hepes, pH 8, 300 mM NaCl, and 10 mM imidazole.The column was washed with five column volumes of wash buffer, 10 mM Hepes pH 8, 300 mM NaCl, and 20 mM imidazole, and the refolded DID was eluted with elution buffer, 10 mM Hepes, pH 8, 300 mM NaCl, and 250 mM imidazole.EDTA was added from a 0.5 M pH 8 stock to a final concentration of 10 mM, and the eluant was dialyzed into ion exchange chromatography buffer A, 20 mM NaH 2 PO 4 /Na 2 HPO 4 , pH 8.10, 100 mM NaCl, prior to loading onto a Hi-Trap Q-XL anion exchange column (Cytiva).A linear gradient of buffer B, 20 mM sodium phosphate, pH 8.10, 1 M NaCl, was used to elute the protein at a conductivity of $40 mS/cm.The purity of the DID protein (molecular weight of 29,078 Da) was found to be >95% on a 12% SDS-PAGE, and the concentration was determined by absorbance using a molar extinction coefficient of 10,220 M −1 cm −1 at 280 nm.DID E326G/E327A was overexpressed using the plasmid pET-28a(+)-DID E326G/E327A (GenScript), which confers kanamycin resistance and expresses an N-terminal his-tagged, mutated DID E326G/E327A .Purification was performed using the protocol identical to that of DID.
Residue-selective labeling of DID was accomplished by transferring the bacterial cells grown in 1 L of LB-Kn to 1 L of M9-Kn medium containing a mixture of 19 unlabeled amino acids plus 0.5 g/l of the 15 N-labeled amino acid of interest (55).Overexpression was carried out for 4 h at 21 C after adding IPTG to 1 mM.The cells were harvested and selectively labeled DID protein was purified using the methodology described above for [U- 15 N]-labeled DID.

DAD and DAD M1199L cloning, expression, and purification
The 29-residue DAD peptide, DETGVMDSLLEALQS-GAAFRRKRGPRQAN, corresponding to WT residues 1194 to 1222 (UniProt O60610), was purchased from GenScript.Stock solutions were prepared by dissolving the peptide in 10 mM potassium phosphate, pH 6.8 containing 15 mM hexamethylphosphoramide, HMPA.Concentrations were calculated using the dissolved weight of the lyophilized peptide.
The DNA sequence for the M1199L mutant peptide, DAD M1199L , was purchased from GenScript and inserted into pTM-7 to express an N-terminal his-tagged TrpL-DAD M1199L fusion protein (35).Fusing the DAD M1199L peptide to a hydrophobic protein (TrpL) directed the gene product into inclusion bodies.The methionine residues in TrpL were mutated to leucine residues (36) to allow for a single cyanogen bromide cleavage site between TrpL and DAD M1199L .The plasmid was transformed into E. coli BL21(DE3) to overexpress both labeled and unlabeled peptide as described for the DID constructs.
To purify DAD M1199L , the bacterial cell pellet was resuspended in 50 mM Tris-HCl, pH 7.2, 1% w/v Triton X-100, and 1 mM EDTA, sonicated as described above and centrifuged at 10,000g for 30 min at 4 C.The pellet was washed with a series of buffers, 50 mM Tris-HCl, pH 7.2, containing 2% w/v Triton X-100 and 1 mM EDTA, followed by 25 mM Tris-HCl, pH 7.2 containing 1 M NaCl and 0.5 mM EDTA, and 50 mM Tris-HCl, pH 7.2.The cells were sonicated and centrifuged between washes.The final pellet was dissolved in denaturing buffer, 50 mM sodium phosphate, pH 8.3 containing 6 M guanidinium chloride, and clarified by centrifugation at 10,000g.The supernatant was incubated with nickelnitriloacetic acid agarose beads overnight for batch binding.The beads were packed into a column and washed once with denaturing buffer.The column was washed with five volumes of 50 mM sodium phosphate, pH 7 containing 6 M urea, followed by 50 mM sodium phosphate, pH 6.4, and 6 M urea.Elution was carried out using 50 mM sodium phosphate, pH 3.6, and 6 M urea.Fractions containing TrpL-DAD M1199L were pooled, dialyzed into water, and lyophilized.Cleavage was performed by dissolving the TrpL-DAD M1199L fusion protein in 70% formic acid with a 100 M excess of cyanogen bromide and incubating for 1.5 h at 25 C.The cleaved products were dried in vacuo, loaded onto a C18 column (Agilent ZORBAX 300SB-C18) for HPLC.Cleaved products were resolved using a gradient of 0% to 90% acetonitrile in 0.1% trifluoroacetic acid.DAD M1199L eluted first and was collected, lyophilized, and stored at −20 C. DAD M1199L stock solutions were prepared by dissolving the peptide in 10 mM potassium phosphate buffer, pH 6.8 containing 15 mM HMPA.The concentration of DAD M1199L stock solutions was determined by integrating HPLC chromatogram traces at 260 nm, the approximate wavelength of maximal absorption of phenylalanine residues relative to reference WT DAD peptide solutions.

T-helix expression and purification
The T-helix protein, KRRETEEKMRRAKLAKEKAE-KERLEKQQ, corresponding to residues 1154 to 1181 (UniProt O60610) was purchased from GenScript as a lyophilized powder.Samples were resuspended in 90 mM sodium phosphate, pH 7.2, 75 mM sodium chloride, 0.01% sodium azide, and 10%v/v D 2 O at 291 K for structure determination and used immediately for NMR experiments or stored at −20 C until further use.
The cross-linked products were resolved by using 8% SDS-PAGE.Bands corresponding to cross-linked DID/DAD M1199L were visualized by staining with Coomassie Blue R-250 (Fig. S3B).The bands of interest were excised from the gels and cut into millimeter-length pieces.The gel pieces were destained in 50% v/v methanol 5% v/v acetic acid, dehydrated by soaking in acetonitrile, and lyophilized.The gel pieces were treated with 10 mM DTT for 15 min at 37 C to break disulfide bonds, then with 50 mM iodoacetamide for 30 min at 25 C to alkylate the protein sulfhydryl groups.Excess iodoacetamide was removed by washing the gels with 100 mM NH 4 HCO 3 , followed by dehydrating in acetonitrile.After three rounds of washing and dehydrating, the gel pieces were completely dried by lyophilization.In-gel tryptic digestion was performed by incubating the lyophilized gel pieces in 100 mM NH 4 HCO 3 with 20 ng/μl trypsin/Lys-C (Promega) on ice for 30 min, after which excess liquid was removed and the gel pieces were incubated overnight at 37 C. Trypsin/Lys-C was chosen for DIAPH1 and the productive encounter complex high fidelity and is expected to only partially cleave the RRKR motif to dicationic or tricationic peptides; tetra-cationic peptides are trypsin digested 100 times faster than dicationic and tricationic regions (56), and thus the full RRKR motif is not expected to remain after tryptic digestion.The gel pieces were washed with 5% v/v formic acid, and peptides were extracted by washing with 5% v/v formic acid and 50% v/v acetonitrile.The extracts were concentrated by lyophilization and stored at −20 C until use in LC-MS/MS analysis.

Liquid chromatography and MS
MS was performed at the RNA Epitranscriptomics & Proteomics Resource located at the Life Sciences Research Building, University at Albany, supported by the SUNY Research Foundation.LC-MS/MS was performed on an integrated micro LC-Orbitrap Velos system (Thermo Fisher Scientific), comprising a 3-pump Waters Cap LC microscale chromatography system (Waters Corp) with an autosampler, a stream-select module configured for precolumn plus analytical capillary column, and an Orbitrap Velos mass spectrometer fitted with an H-ESI probe operated under xcalibur 2.2 control.
Samples were resuspended in 0.1% v/v formic acid and 3% v/v acetonitrile.The peptides were separated on an ACE C18 capillary column (15 cm × 500 μm internal diameter) packed with ACE C18-300 particles (5 μm resin, Advanced Chromatography Technologies).The C18 column was connected in-line with the mass spectrometer.Peptides were eluted at a flow rate of 20 μl/min with a 40 min gradient of 5 to 80% acetonitrile in 0.1% formic acid into quartz emitters and analyzed by electrospray ionization MS, using a Orbitrap Velos mass spectrometer (Thermo Fisher Scientific) with an emitter voltage of 4.5 kV, a sheath gas flow of 10 and a capillary temperature of 275 C. The Orbitrap was operated in a data-dependent acquisition mode in which multiple charged ions with abundance >6000 cps were selected for MS/MS fragmentation.Full scan mass spectra, 350 to 2000 m/z, with the resolution setting of 30,000 at 200 m/z, were detected in the Orbitrap analyzer after accumulation of 10 6 ions.Peptides with multiple charges were selected automatically and fragmented in the collision cell through high-energy collision dissociation with 34% normalized collision energy at a resolution of 7500.A lock mass of diisooctyl phthalate, m/z 391.28428, was used for mass calibration.Data analysis was performed using pLink 2.3.9 (http://pfind.org/software/pLink/)software (57).Raw data from the LC-MS/MS experiments were deposited in the jPOSTrepo repository (58) under the accession codes JPST002156 and PXD042130.

Cross-link MS analysis
Raw data from MS were processed using pLink 2.3.9 software (57) using the.RAW files and the following settings: flow type, high-energy collision dissociation; enzyme, trypsin, up to five missed sites; precursor mass tolerance, 20 ppm; fragment mass tolerance, 0.02 Da; peptide length, minimum three amino acids and maximum 60 amino acids per chain; peptide mass, minimum 300 and maximum 6000 Da; carbamidomethylation of cysteine, fixed; oxidation of methionine, variable; filter tolerance, 10 ppm; and false discovery rate <5% at the peptidespectrum match level.Since only bacterially expressed or chemically synthesized and highly purified proteins were used in this study, the database searched consisted of the DID and DAD sequences and the masses used for cross-linked DSG, BS3, and BS(PEG) 5 were 96.02113Da, 138.06808Da and 302.13656Da, respectively.The results were annotated using pLabel (http://pfind.org/software/pLabel/index.html)(59), requiring a mass deviation <0.5 Da.

Fluorescence titrations
Tryptophan fluorescence spectroscopy titration experiments were performed using a Fluorolog-3 Spectrofluorometer (Horiba).Excitation and emission wavelengths were 280 nm and 352 nm, respectively, with a slit width of 5 nm.Temperature was maintained at 20 C and measurements were performed in triplicate.Protein samples were prepared in binding buffer, 10 mM 3-(N-morpholino)propanesulfonic acid, pH 7.5, 300 mM NaCl, and 0.01% NaN 3 , at 200 nM for WT DID and DID E326G/E327A while DAD M1199L concentrations ranged from 1 nM to 15 μM.The dissociation constant, K D , was calculated using the "one sitetotal, accounting for ligand depletion" regression model in the GraphPad Prism (https://www.graphpad.com/)version 9.0 software (GraphPad) using where [DID-DAD M1199L ] is the amount of bound complex, K D is the dissociation constant, and α is the nonspecific-binding coefficient (60).Through algebraic manipulations, the following quadratic is defined where F is the fluorescence at a specific DAD M1199L concentration, F o is the fluorescence intensity of the blank, β is a scaling factor in the units of μM, [DAD M1199L ] is the total ligand concentration and α = 0.00001.

Stopped flow kinetics
Stopped flow fluorescence measurements were performed at 20 C using a Stopped-Flow Module SFM-400 (Bio-Logic) equipped with an MPS-60 motor power supply, an ALX-250 arc lamp and a PMS-250 photo multiplier system.The tryptophan residue of DID was excited at 285 nm, and the fluorescence emission was measured at 305 nm with 700 V applied to the photo multiplier tube.Two hundred nanomolar (200 nM) DID, in binding buffer, was titrated from 1.5 μM to 20 μM with DAD M1199L .The measured voltage was plotted as a function of time and fit to a single exponential using a pseudo first-order kinetic model to estimate k obs , where k obs = k on [DAD M1199L ] + k off .Linear regression of k obs versus [DAD M1199L ] using Prism v9 software (GraphPad) yielded estimates for k on (slope) and k off (y-intercept), where k on is the rate of complex formation and k off is the rate of complex dissociation.
Three dimensional HNCA, HNCACB, CBCACONH and HNCO as well as 2D 1 H- 15 N HSQC (38), spectra were recorded at 305 K to assign backbone residues of 250 μM [U-15 N, 13 C]-DID in the presence of a molar excess of WT DAD.To facilitate assignment of DID residues, 2D 1 H- 15 N HSQC and 3D 15 N NOESY spectra of DID bound to WT DAD with DID containing selectively 15 N-labeled amino acids were recorded at 305 K or 298 K to facilitate DID resonance assignment.Selective labeling was accomplished for Ala, Asn, Met, Tyr, Cys, Arg, Gly, His, Ile, Leu, Lys, Phe, Thr, Trp, and Val.
To determine temperature dependence of the transition between DID conformers A and B, 2D 1 H-15 N HSQC spectra were recorded with [U- 15 N]-DID bound to DAD M1199L at 280, 283,285,288,290,292,295,298,300,303,305,308, and 310 K.The peak amplitudes of DID C-terminus residues G369, S373, Y374, and G378 were quantified.Equilibrium constants (K eq ) were calculated for the transition between conformer A and B as the ratio of the cross-peak amplitudes of conformer A over conformer B. A linear fit of the van't Hoff plot, ln(K eq ) versus 1/T (K −1 ) was performed using GraphPad Prism 9.0 software.The linear regression analysis yielded thermodynamic parameters ΔG, ΔH, and ΔS for the conformer transition.
3D 15 N-NOESY spectra were collected for [U- 15 N]-DID in the presence of excess unlabeled DAD M1199L and for [U- 15 N]-DAD M1199L in the presence of unlabeled DID.To facilitate the identification of intermolecular nOes, 3D 15 N-NOESY spectra were collected for [U- 15 N]-DID in the presence of [U- 13 C]-DAD M1199L in the presence and absence of a decoupling pulse on the 13 C channel during 1 H detection in the Bruker pulse program noesyhsqcf3gp193d.
Steady state { 1 H}-15 N nOe spectra (61) were collected at 305 K using 450 μM [U- 15 N] DID with 650 μM unlabeled DAD M1199L and 100 μM [U- 15 N]-DAD M1199L with 200 μM unlabeled DID.Measurements of { 1 H}-15 N nOes for free [U-15 N]-DAD M1199L were performed at 298 K to reduce exchange broadening.Interleaved spectra were recorded with and without proton saturation through a high-power 120 pulse applied every 5 ms within the 5 s recycle delay.Quantification of { 1 H}- 15 N nOes was performed using the equation { 1 H}-15 N nOe = I saturated /I unsat, where I saturated is the peak amplitude from spectra collected with 1 H saturation, and I unsat corresponds to the condition without saturation.Error estimates of the { 1 H}- 15 N nOe, δ nOe , were determined from the noise levels in saturated, δ saturated , and unsaturated spectra, δ unsat , using the equation NMR structure calculations CYANA 26 version 3.98.5 was used to calculate the 3D structure of DID in complex with DAD M1199L .The input consisted of a chemical shift list obtained from the resonance assignment, a sequence file of a single polypeptide chain in which the DID sequence was connected to DAD M1199L via linker residues (LL5 in the standard CYANA library), an unassigned 3D 15 N-NOESY peak list for DID, an unassigned 3D 15 N-NOESY peak list for DAD M1199L , upper distance limits from cross-linking MS data (Table S3 footnote b), upper distance limits for manually assigned intramolecular and intermolecular nOes set at 5 Å, and distance restraints for helical regions identified in crystal structure PDB 2F31.Hydrogen bond N-O distances in acceptor-donor pairs were relaxed by 0.5 Å and N-C distances in helical regions were relaxed by 0.5 Å.
NOESY peaks were assigned automatically and converted into distance constraints using the standard CYANA macro with seven cycles of nOe assignment and simulated annealing in torsion angle space.The I-PINE NMR server (http://pine.nmrfam.wisc.edu)classified Pro 272 of DID as cis and all other DID prolines as trans based on the average Cβ chemical shifts (62), therefore Pro 272 was designated as cPro in the sequence file.The CISPROCHECK routine of CYANA 3.98.5 classified the isomerization state of Pro 1218 in DAD M1199L as trans based on the average value and SD for the difference between the chemical shifts of Cβ and Cγ. 29 Backbone Φ and Ψ dihedral angle constraints were determined using the TALOS-N webserver, (http://spin.niddk.nih.gov/bax/software/TALOS-N)(40).The experimental restraints used for the structure calculation are summarized in Tables S2 and S3.
CYANA-generated structures were subjected to minimization in explicit water using YASARA Structure version 20.12.24 (63).Distance restraints from manual and automatic assignment of nOes and cross-linking MS as well as the torsion angle list used in CYANA simulated annealing were used as input in the refinement procedure.In YASARA Structure version 20.12.24, refinement was performed using the Amber14 force field and the nmr_refinesolvent macro.Coordinates of the refined ensemble of 20 DID-DAD M1199L structures were deposited in the Protein Data Bank with entry code 8FG1 (https://deposit-1.wwpdb.org/).

Chemical shift index analysis
DID, DAD M1199L , and T-helix secondary structures were predicted using chemical shift index, CSI, analyses on the CSI 3.0 web server http://csi3.wishartlab.com.The consensus secondary structures were defined in terms of probabilities for a particular segment in DID/DAD M1199L or T-helix to fold into α-helix, random coil, or β-sheet.

Iso-surface map visualization
Iso-surface maps of the DID-DAD M1199L and DID E326G/ E327A -DAD M1199L complexes were constructed using an optimized DID-DAD M1199L structure from our refinements.Using the Adaptive Poisson-Boltzmann Solver (64) (https:// server.poissonboltzmann.org/), the electronic surface properties of the proteins were calculated using the AMBER force field.Results were visualized using the Visual Molecular Dynamics (65) (https://www.ks.uiuc.edu/Research/vmd/)software with blue surfaces set to +10 eV and red surfaces set to −10 eV.

Construction of plasmids for colocalization experiments
Preparation of plasmid constructs expressing human DIAPH1 and DIAPH1 ΔDAD with C-terminal monomeric enhanced YFP fluorescent tags, DIAPH1-YFP and DIA-PH1 ΔDAD -YFP, was described previously (10,66).Constitutively active DIAPH1 ΔDAD was obtained by deleting the Cterminal DAD, which corresponds to amino acids 1194 to 1272 of DIAPH1.YFP fluorescent tags were excised from the AgeI and NotI sites of pCAG-YFP (Addgene) and GFP was inserted resulting in plasmids pDIAPH1-GFP and pDIAPH1 ΔDAD -GFP.To obtain DIAPH1 constructs with Nterminal YFP tags, YFP was excised at the AgeI and NotI sites, and the gap was closed using a short linker duplex, 5 0 -CCGGTGTAATAGTGC-3 0 , 5 0 -GGCCGCGCACTATTACA-3 0 , containing two stop codons.YFP was amplified with two oligonucleotide primers 5 0 -TTAAGCGCTATGGTGAGCA AGGGCGAGGAG-3 0 and 5 0 -TTACTCGAGACCGCCGCTA CCGCCCTTGTACAGCTCGTCCATGCCGAG-3 0 , digested, and cloned into the AfeI and XhoI sites in upstream of the DIAPH1 sequence resulting in plasmid pYFP-DIAPH1.Selfcomplementary oligonucleotides 5 0 -GCTCTCATCACACCA GCGGGCGCCCTTGACTTCCGAGTTCACA-3 0 and 5 0 -TGT GAACTCGGAAGTCAAGGGCGCCCGCTGGTGTGATGAG AGC-3 0 were used with the QuickChange Lightning site-directed Mutagenesis Kit (Agilent Technologies) to introduce mutations E326G/E327A into the DIAPH1 sequence resulting in plasmids pYFP-DIAPH1 E326G/E327A and pDIAPH1 E326G/E327A -GFP.HEK 293T cell culture, transfection, confocal microscopy HEK 293T cells were grown in Dulbecco's modified Eagle's medium supplemented with 4 mM L-glutamine, 5.5 mM Dglucose, 1 mM sodium pyruvate, 10% fetal bovine serum (FBS), (HyClone), 100 units/ml penicillin, and 100 mg/ml of streptomycin.Fifteen millimeter (15 mm) diameter round glass coverslips (Ted Pella) were coated with Collagen I and seeded with 5 × 10 4 cells each in the wells of 24-well plate.After incubating overnight at 37 C, the Dulbecco's modified Eagle's medium was exchanged with Opti-minimal essential medium/ 5% FBS and the cells were transfected with plasmids using FuGENE 6 reagent according to the manufacturer's recommendations.After 48 h of growth, cells were fixed by replacing the medium with 4% formaldehyde in PBS for 20 min at room temperature and washed three times with PBS.Cells were incubated with Alexa Fluor 568 Phalloidin (Thermo Fisher Scientific) solution in PBS, prepared in line with the manufacturer's guidelines, for 45 min at 37 C, washed three times with PBS, and mounted on slides using Fluoromount G (Electron Microscopy Sciences).
Images were acquired by using a Zeiss LSM 710 confocal imaging system (Carl Zeiss Microscopy) equipped with an X63, 1.4 NA oil immersion objective.GFP was excited with a 488 nm argon laser line at 1% power and Alexa Fluor 568 with a 561 nm line at 0.8% power.Image stacks were recorded at 290 nm steps along the z-axis in two-channel mode, using a 488/561 nm beam splitter with a 492 to 550 nm emission window for the GFP channel and a 592 to 676 nm window for the Alexa Fluor 568 channel.A total of eight Z-stacks were acquired for DIAPH1 ΔDAD -GFP/F-Actin cells, and nine each for DIAPH1 E326G/E327A -GFP/F-Actin and DIAPH1-GFP/F-Actin cells.
3D images were processed using FluoRender 2.27 software (67) (https://www.sci.utah.edu/software/fluorender.html).ImageJ plugin JACoP 2.0 (68) (https://imagej.nih.gov/ij/plugins/track/jacop2.html)was used to assess the degree of colocalization of F-Actin with DIAPH1 or its mutants.Z-stacks were processed at default settings in 3D mode, and Manders coefficients M1 and M2, which measures the amount of colocalization for each image were recorded.Colocalization data was processed with Prism 9.0.2 (https://www.graphpad.com/)software (GraphPad).Column heights reported on the graph represent mean values, error bars represent the SD.The statistical significance of multiple comparisons was estimated by using the one-way ANOVA subroutine and Fisher's least significant difference test with a single pooled variance displayed in GraphPad style, where the absence of bars represents p > 0.05, * is p ≤ 0.05, ** is p ≤ 0.01, and *** is p ≤ 0.001.
Lipofectamine LTX reagent (Thermo Fisher Scientific) was used to perform transfection according to manufacturer's protocol.The protocol was optimized for 2.5 μg/ml of plasmid DNA in reduced serum Opti-minimal essential medium for efficient transfection.Human primary aortic SMCs were assessed for migration in response to 10 μg/ml of RAGE ligand, CML-HSA (8) with a wounding assay as previously described (69).Briefly, cells were grown to confluence in 24-well plates and serum-starved overnight.The following morning, the monolayer was wounded using a p10 pipette tip and fresh medium containing RAGE ligand was added for 4 h.Cells were maintained at 37 C and 5% CO 2 .Images were taken at 4 h intervals.Each image was measured, and an ingrowth area of effective migrating cells was calculated.The data was subjected to a Grubbs test with α = 0.1, and the resulting data was processed with Prism v 9.0.2 software (GraphPad).Column heights reported on the graph represent mean values, error bars represent the SD.The statistical significance of multiple comparisons was estimated by using the one-way ANOVA subroutine.

Figure 1 .
Figure 1.DID and DAD M1199L domains of DIAPH1 form a complex.A, schematic representation of the domain structure of DIAPH1.Below are amino acid sequences of the IH and C-terminal region from different organisms: DIAPH1, human (UniProt entry O60610), mouse DIAPH1, mouse (UniProt entry O08808), Drome, fruit fly (UniProt entry P48608).Numbering system corresponds to DIAPH1.Residues that are different from DIAPH1 are highlighted in black, helical regions are indicated by the coiled ribbon above the sequences and RRKR motif is in red.B, 1 H-15 N HSQC spectrum of [U-15 N]-DAD M1199L bound to DID shows well-dispersed backbone amide proton and nitrogen resonances, indicating that DAD M1199L acquires structure upon binding to DID (Fig. S1A).Peaks of the RRKR motif residues are labeled in red.C, 1 H-15 N HSQC spectrum of [U-15 N]-DID bound to DAD M1199L indicates that the complex is structured.Peaks labeled in blue correspond to conformer B. Blue peaks were used in thermodynamic calculations.DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; HSQC, heteronuclear single-quantum coherence.

Figure 2 .
Figure 2. DAD M1199L and interhelical domain of DIAPH1 are flexible.A, steady state { 1 H}-15 N nOes of DAD M1199L bound to DID showed that the structural region of DAD M1199L spans from G1197 to R1213 (Fig.S1A).B, steady state { 1 H}-15 N nOe of the13 C-terminal DID residues of IH showed the presence of two conformers: a structured conformer A (white squares) and an unstructured conformer B (red circles).C, van't Hoff analysis of selected DID C-terminal residues of IH showed an entropically driven equilibrium between conformers A and B. The temperature dependence of the equilibrium constant, K eq , resolved ΔH, ΔS, and ΔG for the unstructured to structured transition (TableS1).D, high-energy collision MS spectrum obtained for the DSG cross-linked product between K228 of DID and the N terminus of DADM1199L at 1119.036 m/z.For clarity, only the molecular ion, and the y and b peptidic fragmentations are labeled.The cross-linked peptides are shown in black and red along with the cross-linker DSG.The experimental mass of 3356.6906Da is in good agreement with the theoretical mass of 3356.6830Da calculated from putative elemental composition.Unassigned peaks are from water or ammonia loss, and/or nonpeptidic bond fragmentation.MS data was deposited to https://repository.jpostdb.org/under acquisition codes JPST002156 and PXD042130.DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; DSG, disuccinimidyl glutarate.
Figure 2. DAD M1199L and interhelical domain of DIAPH1 are flexible.A, steady state { 1 H}-15 N nOes of DAD M1199L bound to DID showed that the structural region of DAD M1199L spans from G1197 to R1213 (Fig.S1A).B, steady state { 1 H}-15 N nOe of the13 C-terminal DID residues of IH showed the presence of two conformers: a structured conformer A (white squares) and an unstructured conformer B (red circles).C, van't Hoff analysis of selected DID C-terminal residues of IH showed an entropically driven equilibrium between conformers A and B. The temperature dependence of the equilibrium constant, K eq , resolved ΔH, ΔS, and ΔG for the unstructured to structured transition (TableS1).D, high-energy collision MS spectrum obtained for the DSG cross-linked product between K228 of DID and the N terminus of DADM1199L at 1119.036 m/z.For clarity, only the molecular ion, and the y and b peptidic fragmentations are labeled.The cross-linked peptides are shown in black and red along with the cross-linker DSG.The experimental mass of 3356.6906Da is in good agreement with the theoretical mass of 3356.6830Da calculated from putative elemental composition.Unassigned peaks are from water or ammonia loss, and/or nonpeptidic bond fragmentation.MS data was deposited to https://repository.jpostdb.org/under acquisition codes JPST002156 and PXD042130.DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; DSG, disuccinimidyl glutarate.

Figure 3 .
Figure 3. Solution structure of human DIAPH1 DID-DAD M1199L complex reveals the binding mode of the RRKR motif.A, overlay of the backbone traces of the 20 solved DID-DAD M1199L solution structures (PDB code 8FG1) refined using YASARA (42).The ARR region of DID is in orange, the IH is in yellow, and the DAD M1199L peptide is colored by pKa where blue is basic, gray is neutral, and red is acidic.The dashed box outlines residues absent from published structures.B, tube diagram of a representative DID-DAD M1199L structure showing the superhelical orientation of the armadillo repeats of DID and the DAD M1199L peptide (blue).The dashed box outlines in panel C.The solid box outlines the RRKR motif in panel D. C, close up of the DID van der Waals surface of residues within 5 Å of L1199.White is hydrogen, red is oxygen, blue is nitrogen, gray is carbon, and orange is sulfur.M1182 of mouse DIAPH1, which corresponds to M1199 of DIAPH1, and L1199 side chains are shown as balls and sticks.The DAD M1199L peptide backbone (this work) is shown as a thin blue ribbon.M1182 is from PDB 2F31.D, DID-DAD M1199L complex, rotated 180 horizontally and downward 90 relative to 3B, showing putative interactions between the basic RRKR motif and acidic residues of DID.Residues E273, R278, E367, E371, and D375 were subjects of previous mutagenesis studies (23, 25).All models were created in Discovery Studio.ARR, armadillo repeat region; DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; IH, interdomain helix.

Figure 4 .
Figure 4.The acidic patch of DID is required for strong DAD M1199L binding.A, iso-potential surface of DID (left) and DID E326G/E327A (right) bound to DAD M1199L where blue and red represents +10 eV and −10 eV, respectively.The basic RRKR motif of DAD M1199L is in green and the rest of DAD M1199L is in yellow.The E326G and E327A mutations reduce a large negatively charged surface on DID as evidenced by the relocation of the basic side chain (green) above the −10 eV iso-contour surface (right).The orientation is analogous to Figure 3D.B, stopped flow fluorescence titrations for the association kinetics of DID-DAD M1199L (left) and DID E326G/E327A -DAD M1199L (right).Surface models of DID and residues of DAD M1199L were made by using Visual Molecular Dynamics software (65).All graphs were made using GraphPad Prism (graphpad.com).DAD, diaphanous autoinhibitory domain; DID, diaphanous inhibitory domain.

Figure 5 .
Figure 5. Nonspecific binding of DAD M1199L results in the formation of futile encounter complexes.A, superimposed 1 H-15 N-HSQC spectra of 100 μM [U-15 N] DID in the presence of a 1× (red), 10× (blue), and 18× (green) molar excess of DAD M1199L .B, chemical shift changes, Ω, in the 1 H-15 N-HSQC spectra of [U-15 N]-DID in the presence of a 1× and 18× molar excess of DAD M1199L .Ω = √((ΔH) 2 + (ΔN/2) 2 ), where ΔH and ΔN are changes in chemical shift in the hydrogen and nitrogen dimensions, respectively.C, semitransparent space filling model of DID showing significant chemical shift changes (red) upon nonspecific DAD M1199L binding.The cut-off value for significant changes, Ω = 0.03, defines the largest 5% of the Ω values.The core structure of DID is tinted brown and the interdomain helix is yellow.The orientation of the DAD M1199L peptide in the encounter complex is shown in blue.DAD, diaphanous autoinhibitory domain; DID, diaphanous inhibitory domain; HSQC, heteronuclear single-quantum coherence.

Figure 6 .
Figure 6.Formation of a productive encounter complex increases DIAPH1-actin colocalization and regulates RAGE ligand-induced cellular migration in human vascular smooth muscle cells.A, fluorescence microscopy of HEK293T cells with GFP-tagged WT DIAPH1 (top), DIAPH1 ΔDAD (middle), and DIAPH1 E326G/E327A (bottom) in green, and actin labeled with Alexa Fluor 568 Phalloidin in red.Yellow indicates colocalization.B, manders coefficients, M1 and M2, quantifying colocalization between F-actin and DIAPH1 constructs and DIAPH1 constructs and F-actin, respectively.Larger values correspond to a higher degree of colocalization.Column heights reported on the graph represent mean values, error bars represent the SD.C, percent migration of human primary aortic vascular smooth muscle cells, SMCs, when stimulated with CML-HSA as compared to those treated with serum-free medium (SFM).Cells expressing YFP and DIAPH1-YFP were used as a reference and compared with those expressing DIAPH1 ΔDAD -YFP and DIAPH1 E326G/E327A -YFP.Stars indicate statistical significance: p-value <0.05 (*), p-value <0.01 (**), and p-value <0.001 (***).D, proposed model for the equilibrium of WT DIAPH1 dimers between the active, encounter, and autoinhibited complexes.Double headed arrow indicates IH flexibility, and unstructured DAD is shown in the circle.Colored domains match that of Figure 1A.In mouse DIAPH1, DD, CC, and FH2 domains facilitate dimerization(15,17,43,72). Active DIAPH1 forms a productive encounter complex via electrostatic steering, dotted line, between the RRKR motif (blue star) and the acidic patch of DID (red star).The interaction increases the probability of DID-DAD helix coupled folding and binding and formation of the autoinhibited conformation.Active DIAPH1 lacking the basic RRKR motif, such as in DIAPH-(R1213X)(22,31), cannot create a productive encounter complex resulting in an equilibrium shift toward the active complex.Note that deletion of DAD, such as in DIAPH1 ΔDAD , would preclude the autoinhibition.CC, coiled-coil; CML-HAS, carboxymethyllysine human serum albumin; DAD, diaphanous autoinhibitory domain; DIAPH1, Diaphanous 1; DID, diaphanous inhibitory domain; IH, interdomain helix; RAGE, receptor for advanced glycation end products; YFP, yellow fluorescent protein. 15