Conformational flexibility is a key determinant of the lytic activity of the pore forming protein, Cytolysin A

Bacterial pore-forming toxins (PFTs) bind and oligomerize on mammalian cell membranes forming nanopores, that cause cell lysis to promote a wide range of bacterial infections. Cytolysin A (ClyA), an alpha(α)-PFT, is known to undergo one of the largest conformational changes during its transition from a water soluble monomeric form to the membrane embedded dodecameric nanopore assembly. Despite extensive work on the structure and assembly of ClyA, a complete molecular picture of the interplay between the protein segments and membrane lipids driving this transformation remains elusive. In this study, we combine experiments and all-atom molecular dynamics (MD) simulations of ClyA and its mutants to unravel the role of the two key membrane interacting motifs, namely, the β-tongue and N-terminus helix, in facilitating this critical transition. Erythrocyte turbidity and vesicle leakage assays reveal a loss of activity for β-tongue mutant (Y178F), and delayed kinetics for the N-terminus mutants (Y27A and Y27F). All atom, thermal unfolding molecular dynamics simulations of the monomer carried out at 310, 350 and 400 K reveal a distinct reduction in the flexibility in both the β-tongue and N-terminal regions of the mutants when compared with the wild type. This decreased loss of conformational flexbility correlates positively with the reduced lytic and leakage activity observed in experiments, indicating that the tendency to lose secondary structure in the β-tongue region is an important step in the conformational transition bistability of the ClyA protein. Simulations with the membrane inserted oligomeric arcs representing the pore state reveal a greater destabilization tendency among the β-tongue mutant as inferred from secondary structure and N-terminal positioning. Our combined experimental and simulation study, reveals that conformational flexibility is indispensable for the outward movement of the β-tongue and the tendency to induce disorder in the β-tongue is an important step in the transition to the membrane mediated helix-turn-helix motif integral to ClyA pore formation. This observed loss of secondary structure is akin to the structural transitions observed in intrinsically disordered proteins (IDPs) to support protein function. Our finding suggest that inherent flexibility in the protein could play a wider and hitherto unrecognized role in the membrane mediated conformational transitions of PFTs in general. Author summary Bacterial pore-forming toxins (PFTs) bind and oligomerize on mammalian cell membranes to form bilayer spanning nanopores. Unregulated pore formation disrupts ionic balance that compromises the permeability of the cell, leading to cell death and infection. PFTs display remarkable structural plasticity that allow these proteins to interconvert between the two physiochemically distinct water soluble and membrane bound forms. However the molecular mechanism for this robust interconversion is poorly understood. For example, upon membrane binding, cytolysin A (ClyA), an α-PFT, shows one of the largest conformational changes in protein structure among the PFT family of proteins. In order to understand this transtition we characterized several point mutations in ClyA using experiments and molecular dynamics (MD) simulations to understand the role of two essential ClyA motifs (β tongue and N-terminus) implicated in the conformational changes responsible for oligomerization and conferring stability to the pore state. Our study reveals that the innate conformational flexibility of the β tongue results in a disordered intermediate state that facilitate the complete transition to the pore state. This tendency to disorder is compromised to varying degrees in ClyA mutants, correlating with the loss of lytic activity. Our results suggest that the finely tuned conformational flexibility in the membrane motifs of ClyA are critical to its function, revealing a broader paradigm that could be at play during membrane associated secondary structure transitions of proteins in general.


Introduction 20
Pore-forming proteins (PFPs) are a special class of proteins expressed as water soluble 21 monomers across a wide variety of living organisms, with the unique ability to 22 compromise the permeability of cell membranes by the formation of transmembrane 23 oligomeric pore-like assemblies [9]. Among them, pore-forming toxins (PFTs) represent 24 a subset of pathogenic PFPs, produced by virulent bacteria to mediate and transmit 25 infections. The large family of PFTs is broadly classified based on the secondary 26 structure of the membrane-inserted segments as α-PFTs (bundle of α-helices) and 27 β-PFTs (mainly β-barrels) [11]. Cytolysin A (ClyA; also known as hemolysin E 28 (HlyE)), is one of the best studied members of α-pore-forming toxin family produced by 29 Escherichia coli, Salmonella typhi, and Shigella flexneri [9]. Remarkably, ClyA is known 30 to undergo one of the largest conformation transitions in the PFT family of proteins, 31 during its conversion from the water soluble monomer (Fig 1A) form to the membrane 32 inserted protomer (Fig 1B) state as observed in the final dodecameric pore structure. 33 The two membrane binding motifs of ClyA play a critical role in driving this transition 34 as highlighted by major alterations in their structure. The highly hydrophobic 'β-tongue' 35 (β-hairpin) motif rearranges into a helix-turn-helix motif in its membrane embedded 36 form. On the other hand, the N-terminus helix detaches from the core helical bundle 37 with a ∼126°outward movement to reposition itself to form the membrane inserted 38 hydrophilic transmembrane pore [34]. Although, there have been several experimental 39 and computational studies on ClyA since the crystal structure of the dodecameric pore 40 complex was first elucidated [33,40], several aspects of the pore forming mechanism are 41 incompletely understood. Whether ClyA follows a pre-pore mechanism, where the hydrophobic β-tongue and its subsequent conversion to the helix-turn-helix motif in 49 the membrane. Structure based molecular dynamics (MD) simulations [22] illustrate 50 that membrane binding is dominated by the insertion of the β tongue followed by 51 insertion of the N-terminus to form the membrane inserted protomeric state (Fig 1B). 52 Although in a second less sampled pathway, the N-terminus was found to initially  Although cholesterol is known to enhance the pore forming ability of ClyA, the 67 molecular basis for this enhancement has only recently been elucidated. Using MD 68 simulations and single particle tracking experiments, cholesterol was found to enhance 69 binding of the N-terminus and distinct cholesterol binding sites were found in the 70 pockets formed between adjacent β-tongues in the membrane inserted pore assembly [41, 71 22]. Additionally the presence of a cholesterol recognition and consensus (CRAC) 72 motif [19] was identified in the N-terminus of ClyA [41,22]. Thus cholesterol enhances 73 the kinetics of pore formation, imparting greater stability to the membrane inserted 74 pore assembly. In contrast to the β-tongue, the amphiphatic N-terminal helices are 75 directly involved in forming the hydrophilic membrane inserted ClyA transmembrane 76 pore. N-terminal truncations by Ludwig et al. [31] showed considerably weakened 77 hemolytic activity and single channel conductance in β-tongue deletion mutants were 78 attributed to short lived pores momentarily stabilized by the N-terminus [18]. In the MD simulations of the pore complex reveal the interplay between these two seemingly 84 disparate motifs in the ClyA pore assembly [13]. In addition to the β-tongue and 85 N -terminus which are an integral part of the membrane inserted pore complex of ClyA, 86 the C-terminus which remains in the extracellular regions distal from the membrane 87 interface is also found to be important for pore formation. Truncations in the 88 C-terminus, unexpectedly shows loss of activity upon deletion [42,31,5]. Thermal 89 unfolding molecular dynamics simulations draw a subtle connection between the 90 C-terminus and the movement of the N-terminus from the bundle of α-helices in the 91 ClyA monomer. Although mutagenesis experiments with pore forming toxins enable 92 linking protein structure to function, the molecular basis for the ensuing loss of activity 93 or compromised pore formation is only indirectly inferred. Furthermore, the specific 94 step in the pore forming pathway influenced by a mutation is usually undetermined.

95
In this manuscript we present a combined point mutagenesis experimental study on 96 the ClyA monomer with all-atom MD simulations to elicit molecular details for the loss 97 in lytic activity observed in human erythrocyte turbidity and small unilamellar vesicle October 20, 2021 4/23 to determine the lysis half-life, t lysis . In Eq 1, C 0 , C R,sat , C R,slope are the starting versus time,t) was fitted to an exponential decay function, to determine the leakage rates (t leak ) by different ClyA mutants. In Eq 2, C L,slope , Crystal structure of ClyA monomer (PDB ID 1QOY) [48] was used as an initial 168 structure where the missing residues (299-303) were modeled using Modeller 9.9 as a 169 loop while keeping the remaining atom fixed (see Fig 1A). VMD software [24] was used 170 to perform four point mutations (Y178F, Y27A, Y27F, and D74A) in the monomer.

171
Based on previous studies [16,49,30], the highest temperature of 400 K was used to (E) Turbidity assay to determine the activity of ClyA mutants is performed by measuring the optical density (OP) of rabbit erythrocytes. Solid lines represent Boltzmann sigmoid fits to the data (see S4 Table). t 1/2 is shown in inset figure. (F) Vesicle dye leakage kinetics of ClyA mutants for 70% POPC and 30% cholesterol membranes. Solid lines represent single exponential fits to the leakage data (see S5  Table). were modelled using Modeller 9.9 [20]. A complete detail of membrane inserted trimer 195 arc simulations is given in the supplementary information (see S2 Table). Similar  Table S3 Table. The variation in 213 the distance, r between the residues in the monomer were modeled by a harmonic 214 oscillator to evaluate the spring constant [12] from the probability distributions of the 215 distance. Carrying out ensemble averages in the canonical ensemble for a harmonic 216 oscillator with a spring constant, k one obtains, where k B is the Boltzmann constant, T is the temperature, and < r 2 > − < r > 2 is the 218 variance in the distance, r between the two residues. Residue-wise average change in the 219 local helicity was evaluated using the difference in the secondary structure probabilities 220 using, 221 P = P simulation − P crystal structure (4) where, P simulation is the probability of secondary structure formation for the monomer 222 in solvent obtained in the MD simulation and P crystal structure is the probability of 223 occurrence of the corresponding secondary structure in the crystal structure. protein structural stability, membrane binding and haemolytic activity (Fig 1D).

230
Activity of proteins with point mutations in the β-tongue and N-terminus helix motifs 231 were compared to the wild-type or a salt-bridge defective variant of ClyA. Using 232 erythrocyte turbidity (Fig 1E), we find that the disruption of a stable salt bridge 233 between D74-K240/N142 residues by an Asp-Ala substitution at position 74 (D74A) did 234 not perturb ClyA cell lysis activity significantly. On the other hand, β-tongue (Y178F) 235 and N-terminal (Y27A and Y27F) mutants disrupt ClyA activity to different degrees.

236
The β-tongue mutant displays a ≥ 30-fold increase in t 1/2 for cell lysis. The N-terminus 237 October 20, 2021 7/23 mutants also display a marked but smaller reduction (t 1/2 increase by ∼ 3 fold) in lysis 238 activity with Y27A displaying faster kinetics (but a delayed onset of lysis) when 239 compared with Y27F. Vesicle leakage experiments ( Fig 1F)  activity in these mutants is arising from their inability to disrupt the lipid membranes. 244 Binding of the N-terminus mutants to RBCs ( Fig 1G) were lower (∼ 3 fold) as 245 compared to β-tongue and salt bridge mutant suggesting that the N-terminus also has a 246 significant role in direct binding in addition to being in the protomer state.  Since the ClyA structural transition to the protomer form is triggered by the 277 opening of the β-tongue [33,22], we evaluated the secondary structure changes in this 278 region (Fig 2F to Fig 2I). The αD (164-179) helix of the protein in the WT (Fig 2F) 279 progressively loses helicity over the course of the simulation. Simultaneously, the central 280 segment of the β-tongue converts into a turn motif, implying that the secondary 281 structure of the native ClyA monomer is susceptible to disruption (Fig 2F and S3 Fig). 282 The innate flexibility in the β-tongue is crucial to form the helix-turn-helix motif upon 283 the interaction with the membrane and the secondary structure analysis in this region 284 reveals interesting differences across the various mutants. In contrast to the WT, we did 285 not observe any significant disruption of the secondary structures in Y178F, and Y27A 286 mutants (Figs 2G and S3 Fig respectively) suggesting that these mutations impair the 287 flexibility of the β-tongue. The Y27F mutant displays a loss of helicity in the αD motif 288 residues that form part of the αD helix and the induced secondary structure changes are 292 greater when compared with other mutants, however less than those observed in the 293 WT ( Fig 2D). The secondary structure changes between 600-800 ns of the simulation 294 ( Fig 2I) clearly reveal that the tendency to disorder is greatest in the WT followed by 295 D74A, with the other mutants showing greater resistance to disorder. The delayed loss 296 of structure in Y27F correlates well with its marginal loss of pore forming activity. It is 297 likely that the Y27A mutant might also display a similar loss of structure at longer 298 timescales, however we did not pursue this aspect further. 299 We next focus on the structure of the β-tongue and evaluate various properties over 300 200-800 ns of the simulation trajectory to quantify the extent of flexibility in this region 301 (Fig 3). The greatest increase in the both the magnitude as well as the standard 302 deviations in RMSD are observed for the WT and D74A (Fig 3A) as the temperature is 303 increased to 400 K, with the mutants showing relatively smaller changes.  To further characterize the β-tongue fluctuations we monitored some critical 305 distances and angles as illustrated in Fig 3B. G180 and G201 residues act as a hinge for 306 the β-tongue serving as pivots during the release of the buried β-tongue [33]. Consistent 307 with this mechanism, a G180V mutant has been found to compromise haemolytic 308 activity [5] of ClyA. Interestingly, the distance distribution between these two residues, 309 d 180−201 displays the largest spread for the WT and D74A mutant (Fig 3C). Upon  (Table 1).

313
Another critical residue in the β-tongue, F190 is located at the tip of the β-tongue 314 forming hydrophobic interactions with the terminal residues of four helices, αA2, αB,  (Fig 3B). In order to quantify this movement, we calculated the angular distribution θ 324 October 20, 2021 10/23 between the β-tongue and N-terminus with F190 as the vertex (Fig 3B). While the θ 325 distribution for the WT ClyA is bimodal and dominated by values around 100°, the 326 other mutants mostly sample conformations where θ ∼155° (Fig 3E). The smaller angles 327 sampled by the WT signify an increased propensity for the N-terminus to flip out from 328 the groove formed by the four helices which contains the β-tongue (Fig. 1A) during the 329 formation of the membrane inserted state in the protomer (Fig. 1B). We however did 330 not see a similar variation for D74A in this particular situation, however we point out 331 that Y27A also sampled lower values of θ.

332
To further quantify the extent of secondary structure changes effected by the 333 mutations, we evaluated the residue-wise average change in the local helicity across the 334 different mutants using Eq. 4. ∆P varies between -1 to 1 indicative of the loss and the 335 gain of secondary structure respectively (Fig 4A). In addition, the changes that occur 336 between the monomer to protomer are also illustrated in the first row of Fig 4A. The this trend is only observed for the WT and D74A mutant correlating positively with the 348 higher activity observed in the experiments. We did not observe a similar trend for the 349 other mutants pointing to a reduced tendency to lose helical content and disorder, which 350 appears to correlate with their lowered pore forming ability. The N-terminus (αA1) 351 inserts into the membrane by flipping out from the helical bundle in the monomer 352 ( Fig 1A). αA1 is connected to αA2 by a small turn (34-37) which converts into a helix 353 upon membrane insertion. At 400 K we see a high propensity of helix formation around 354 residue number 30 in the WT and D74A, again indicative of the increased propensity to 355 induce secondary structural changes during the formation of the protomer (Fig 4).

356
Interestingly, all the mutants show a small increase in helicity for residues 176-179 which 357 form part of the terminal end of αD which fuses with the β-tongue to form the extended 358 αC helix (Fig. 1). Snapshots of the β-tongue residues in Fig. 4   simulations suggest that the transition of the β-tongue into the helix-turn-helix motif of 399 protomer is possibly compromised due to these mutations (Fig 4).

400
While the RMSD fluctuations in N-terminus are conspicuously low in both the 401 β-tongue mutant (Fig. 5B), their N-terminus retains its secondary structure similar to 402 the WT over the course of the simulation. We probed this discrepancy by quantifying 403 the tilt angle of the N-terminus that determines the extent of membrane insertion of 404 this motif. All the activity compromised mutants (and especially Y178F) display a 405 significantly larger tilt angle compared to the WT (Fig 5D) indicating that the 406 propensity of the N-terminus to traverse the membrane is compromised. Such mutants 407 will show a greater tendency for pore closure. Previous study from the group has shown 408 the critical role played by cholesterol in stabilizing the membrane inserted ClyA 409 oligomer by binding to the pockets formed by β-tongues from adjacent protomers [41]. 410 We observe higher cholesterol occupancy in the β-tongue pockets of Y178F (Fig 6).

411
This could in part be driven due to the increased hydrophobicity or enhanced π − π 412 stacking interactions with cholesterol. comparison to the WT (Fig 1G), both show delayed activity (Fig 1E and Fig 1F). Since 420 the measurement of the binding is at equlibrium, the lower binding might be due to 421 faster unbinding kinetics and slower binding kinetics. While the cholesterol occupancy 422 in the β-tongue region (Fig 6A, lower panel) is comparable to the WT, Y27F shows role in cholesterol interaction [41]. It also forms a salt bridge with A183 in the β-tongue 425 purportedly providing stability to the inner ring of the membrane inserted α helices in 426 the pore complex. N-terminus mutants show a larger tilt angle for the N-terminus helix 427 than the WT (Fig 5D), therefore, its translocation across the membrane is affected 428 (though not as significantly as in case of Y178F). Both Y27 mutations show increased 429 cholesterol interaction in the N-terminus (Fig 6) which can lead to local membrane 430 heterogeneities slowing down pore formation consistent with the delayed kinetics 431 observed in the experiments. Moreover, N-termini in these mutants are less flexible with 432 a moderate increase in helicity when compared with the WT (Fig 5C). This further  [196][197][198][199]. The loss of secondary structure in this region is 467 directly associated with the formation of the helix-turn-helix turn motif between αC and 468 αF in the membrane inserted pore structure (Fig 1D). Thus the complete conversion of 469 the β-tongue motif is accompanied by an extended loss of secondary structure as noted 470 earlier. Simultaneously, increased helicity is observed at residues 200-206 in the loop 471 region between αE and αF (Fig 7B) which eventually converts to the extended helix αF 472 in the pore structure. An small increase in helicity is also observed at residues 177-178 473 located toward the end of the αD helix in the monomer which eventually fuses to form 474 the end of the αC helix extending up to residue 181 in the protomer. The residues Y178 475 where the β-tongue mutation is carried out are located in this region which lie at the 476 interface between αD and the β-tongue. Interestingly Y178F shows similar regions as 477 the WT that undergo a loss in helicity, albeit to a significantly lowered degree. These 478 changes are observed in both αD and αE with an increase in helicity at residues 200-205. 479 The qualitatively similar trends observed in the structural changes between Y178F 480 and the WT are reflected in the weak and delayed (> 40 min) erythrocyte lysis for this 481 mutant (Fig 1E) indicating that Y178F can form pores albeit with a significantly 482 reduced efficacy when compared with either the WT or D74A. Y178F shows an increase 483 in helix formation propensity at residues 177 and 178 when compared with the WT.

484
Thus the loss of secondary structure and tendency to disorder in this region, serving as 485 a precursor to assist helix formation in the protomer, is also absent and helicity is 486 retained in this region. 487 We next discuss the nature of the different mutations with regard to their helix 488 forming propensities [36] as well as tendency for membrane insertion based on their 489 hydrophobicity scale [32]. For Y178F the change in the helix propensity scale is marginal 490 with Y having a scale of 0.53. However the hydropathy scale changes by a factor of two 491 from Y to F. This is reflected in the increased membrane binding for Y178F which is 492 comparable to D74 (Fig 1G). The lowered flexibility in the β-tongue hinge region is also 493 a key factor in the outward release of the β-tongue essential for pore formation. Concomitant with the structural changes that occur in the β-tongue region, the 495 repositioning of the N-terminus involves the fusion of the short helix αA2 with αB, with 496 the conversion of the interconnecting loop (47-55) to a helix ( Fig 7A). The greatest 497 propensity in this region (47)(48)(49)(50)(51)(52)(53)(54)(55) to convert to a helix is observed in the WT, along with 498 a marginal increase in the helical content between residues 27-33. This region of the 499 N-terminus converts to form part of the membrane inserted αA1 helix in the pore state, 500 and we see a propensity for local structuring to assist in this transformation. Here it 501 appears that this local structuring assists in the movement of the N-terminus away form 502 the helical bundle consisting of αG and αF where inter-helical contacts are present. The 503 interaction between αG and the N-terminus has been studied in previous C-terminus 504 mutational and thermal unfolding studies [42]. This tendency to structure in this region 505 (27-33) and from a helix in the loop (47-55) is seen across all the mutants to varying 506 degrees indicating a common pathway that occurs during the transformation from the 507 monomer to a protomer. Since the initial membrane binding pathway is dominated by 508 the insertion of the β-tongue followed by N-terminal insertion [33,12] we posit that repositioning of the N -terminus and the concomitant helix-helix fusion to from the 511 extended αB helix (Fig 7A). Our simulations also illustrate that the N-terminus of 512 Y27A has a higher propensity to swing out of the 4-helix bundle resulting in breaking 513 the hydrophobic interaction of αA F50 residue with the β-tongue tip residue F190 514 thereby lowering the barrier for the outward movement of β-tongue. Here again the 515 spring constants for d 180−201 (Fig 3C) are larger (k = 14.2 × 10 5 ) for Y27A when 516 compared with Y27F suggestive of a larger barrier for the outward movement of the 517 β-tongue for Y27A. The membrane insertion step occurs with slower kinetics for the 518 Y27F mutant where lysis is gradual and extending over a longer time duration. In 519 contrast, although lysis is delayed for Y27A, it occurs over a smaller time window when 520 compared with Y27F ( Fig 1E) indicative of robust pore formation. The increased 521 hydropathy [32] (-1 to -2 kcal/mol) associated with the Y27F mutant could potentially 522 increase membrane sampling and it is likely to influence the distribution of oligomers 523 present in the membrane due to enhanced membrane insertion events. We speculate 524 that the extended lysis that occurs for Y27F is suggestive of a pre-pore pathway with a 525 larger population of smaller arcs present on the membrane when compared with Y27A 526 where the reduced hydropathy, (-1 to 0 kcal/mol) might lead to the formation of larger 527 oligomeric assemblies that are required to form stable pores [2,40]. Efficient pore 528 formation for Y27A is also revealed in the vesicle leakage data (Fig 1F).

529
A common theme that emerges in our study is the correlation between the observed 530 increased lytic ability and the enhanced disorder or "plasticity" observed in the 531 monomer thermal unfolding simulations. This specific feature, wherein conformational 532 transitions are assisted by the presence of a disordered state is gaining acceptance and 533 extensively investigated in intrinsically disordered proteins(IDPs) [43,8,10,44,4,46,1]. 534 The inherent flexibility in IDPs play a central role in allowing them to transition from a 535 disordered to ordered state upon binding to a target molecule. One such example is the 536 flexible region of N tail that regulates the coupled folding and binding with X domain to 537 form the measles virus where N tail shows a conformational transition to an α-helix [45]. 538 Unstructured N-terminus of lipoproteins works as an anchor to support protein 539 function [17] and the intrinsically disordered state of α-synuclein converts to an α helix 540 upon membrane binding [29]. The influence of vesicle size and composition on 541 alpha-synuclein structure and stability [25]. Our findings suggest that the pre-existing 542 functional disorder in IDPs is more universal and appears to be exploited by the 543 bistable ClyA protein for optimal functionality during the conformational transition and 544 subsequent oligomerization that occurs from the water soluble monomeric state to the 545 membrane inserted protomer.

546
A related aspect is the path followed by proteins undergoing a secondary structure 547 transition. Enhanced sampling methods to compute the minimum free energy path of 548 small proteins that transform from an α-helix to a β-sheet [35] reveal the presence of a 549 partially unfolded intermediate which corresponds to a metastable state in the free 550 energy landscape during the transition. The comparison of entropic and enthalphic 551 contributions on the path indicates that the entropy of the unfolded metastable state is 552 high, whereas the enthalpy at a minima, and the enthalpic barrier of the transition is is 553 offset by the entropic contribution along the folding pathway.

554
Our study further shows that mutations that compromise this inherent tendency to 555 disorder and lose secondary structure can drastically limit and abrogate the pore 556 forming pathway and subsequent lytic ability of ClyA. Indeed, recent single molecule 557 FRET experiments [15] to examine the unfolding pathway of ClyA reveal the presence 558 of partially folded intermediates with unfolded C and N-termini and earlier ensemble 559 experiments in detergent point to the formation of a molten globule intermediate during 560 the pore assembly pathway [7].

562
Our study illustrates that the inherent flexibility in the α PFT, ClyA is essential for a 563 robust conformational transition to the membrane inserted protomeric state. Thermal 564 unfolding molecular dynamics simulations reveal a direct link between the loss of 565 plasticity in the key membrane interrogating motifs with reduced lytic ability as 566 revealed in both erythrocyte lysis and vesicle leakage experiments with ClyA. The 567 predominantly hydrophobic β-tongue which remains buried in a four helix bundle in the 568 monomer state and the N-terminus are the primary membrane interacting motifs that 569 undergo large conformational changes during pore formation. Mutations in the vicinity 570 of the β-tongue mitigate secondary structure disruptions in this region when compared 571 with the disorder induced in wild type and other mutants which show high lytic ability. 572 N-terminal mutants also result in increased rigidity in the β-tongue residues illustrating 573 a strong allosteric effect between these motifs. Both the N-terminal mutants 574 investigated in this study show delayed lytic ability in contrast to the β-tongue mutant 575 where activity is completely abrogated or significantly reduced. Our study not only 576 reinforces the view that the β-tongue is a central driver in the pore forming pathway of 577 ClyA, but illustrates that the flexibility of essential motifs that interact with the 578 membrane are a key to their functionality. The notion that protein flexibility or 579 tendency to disorder is essential for structural transformations is similar to folding 580 phenomenon observed in IDPs. Although relatively unexplored in connection with the 581 pore forming pathways for PFTs, our study suggests that inherent disorder could be a 582 more general phenomenon invoked across a wider class of PFTs which form large