Structure–function analysis of pectate lyase Pel3 reveals essential facets of protein recognition by the bacterial type 2 secretion system

The type II secretion system (T2SS) transports fully folded proteins of various functions and structures through the outer membrane of Gram-negative bacteria. The molecular mechanisms of substrate recruitment by T2SS remain elusive but a prevailing view is that the secretion determinants could be of a structural nature. The phytopathogenic γ-proteobacteria, Pectobacterium carotovorum and Dickeya dadantii, secrete similar sets of homologous plant cell wall degrading enzymes, mainly pectinases, by similar T2SSs, called Out. However, the orthologous pectate lyases Pel3 and PelI from these bacteria, which share 67% of sequence identity, are not secreted by the counterpart T2SS of each bacterium, indicating a fine-tuned control of protein recruitment. To identify the related secretion determinants, we first performed a structural characterization and comparison of Pel3 with PelI using X-ray crystallography. Then, to assess the biological relevance of the observed structural variations, we conducted a loop-substitution analysis of Pel3 combined with secretion assays. We showed that there is not one element with a definite secondary structure but several distant and structurally flexible loop regions that are essential for the secretion of Pel3 and that these loop regions act together as a composite secretion signal. Interestingly, depending on the crystal contacts, one of these key secretion determinants undergoes disorder-to-order transitions that could reflect its transient structuration upon the contact with the appropriate T2SS components. We hypothesize that such T2SS-induced structuration of some intrinsically disordered zones of secretion substrates could be part of the recruitment mechanism used by T2SS. The type II secretion system (T2SS) transports fully folded proteins of various functions and structures through the outer membrane of Gram-negative bacteria. The molecular mechanisms of substrate recruitment by T2SS remain elusive but a prevailing view is that the secretion determinants could be of a structural nature. The phytopathogenic γ-proteobacteria, Pectobacterium carotovorum and Dickeya dadantii, secrete similar sets of homologous plant cell wall degrading enzymes, mainly pectinases, by similar T2SSs, called Out. However, the orthologous pectate lyases Pel3 and PelI from these bacteria, which share 67% of sequence identity, are not secreted by the counterpart T2SS of each bacterium, indicating a fine-tuned control of protein recruitment. To identify the related secretion determinants, we first performed a structural characterization and comparison of Pel3 with PelI using X-ray crystallography. Then, to assess the biological relevance of the observed structural variations, we conducted a loop-substitution analysis of Pel3 combined with secretion assays. We showed that there is not one element with a definite secondary structure but several distant and structurally flexible loop regions that are essential for the secretion of Pel3 and that these loop regions act together as a composite secretion signal. Interestingly, depending on the crystal contacts, one of these key secretion determinants undergoes disorder-to-order transitions that could reflect its transient structuration upon the contact with the appropriate T2SS components. We hypothesize that such T2SS-induced structuration of some intrinsically disordered zones of secretion substrates could be part of the recruitment mechanism used by T2SS. The Gram-negative bacteria possess a multilayer cell envelope composed of the inner membrane surrounding the cytoplasm and the outer membrane facing the external medium. The two membranes delimit an extracytoplasmic compartment, the periplasm, which contains a peptidoglycan layer (1Silhavy T.J. Kahne D. Walker S. The bacterial cell envelope.Cold Spring Harb. Perspect. Biol. 2010; 2a000414Crossref PubMed Scopus (1458) Google Scholar, 2Miller S.I. Salama N.R. The gram-negative bacterial periplasm: Size matters.PLoS Biol. 2018; 16e2004935Crossref PubMed Scopus (38) Google Scholar). To ensure the selective transport of proteins and other macromolecules across this complex cell envelope, bacteria have evolved several specialized cell machineries (3Costa T.R. Felisberto-Rodrigues C. Meir A. Prevost M.S. Redzej A. Trokter M. Waksman G. Secretion systems in gram-negative bacteria: Structural and mechanistic insights.Nat. Rev. Microbiol. 2015; 13: 343-359Crossref PubMed Scopus (482) Google Scholar, 4Sherman D.J. Xie R. Taylor R.J. George A.H. Okuda S. Foster P.J. Needleman D.J. Kahne D. Lipopolysaccharide is transported to the cell surface by a membrane-to-membrane protein bridge.Science. 2018; 359: 798-801Crossref PubMed Scopus (59) Google Scholar, 5Ekiert D.C. Bhabha G. Isom G.L. Greenan G. Ovchinnikov S. Henderson I.R. Cox J.S. Vale R.D. Architectures of lipid transport systems for the bacterial outer membrane.Cell. 2017; 169: 273-285.e217Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar, 6Dong C. Beis K. Nesper J. Brunkan-Lamontagne A.L. Clarke B.R. Whitfield C. Naismith J.H. Wza the translocon for E. coli capsular polysaccharides defines a new class of membrane protein.Nature. 2006; 444: 226-229Crossref PubMed Scopus (249) Google Scholar). The type II secretion system (T2SS) is widespread among Gram-negative bacteria, and it is used to secrete fully folded proteins, usually lytic enzymes and toxins, from the periplasm into the external medium or host tissue (7Cianciotto N.P. White R.C. Expanding role of type II secretion in bacterial pathogenesis and beyond.Infect. Immun. 2017; 85e00014-17Crossref PubMed Scopus (80) Google Scholar, 8Gu S. Shevchik V.E. Shaw R. Pickersgill R.W. Garnett J.A. The role of intrinsic disorder and dynamics in the assembly and function of the type II secretion system.Biochim. Biophys. Acta. 2017; 1865: 1255-1266Crossref Scopus (12) Google Scholar, 9Korotkov K.V. Sandkvist M. Hol W.G. The type II secretion system: Biogenesis, molecular architecture and mechanism.Nat. Rev. Microbiol. 2012; 10: 336-351Crossref PubMed Scopus (284) Google Scholar, 10Thomassin J.L. Santos Moreno J. Guilvout I. Tran Van Nhieu G. Francetic O. The trans-envelope architecture and function of the type 2 secretion system: New insights raising new questions.Mol. Microbiol. 2017; 105: 211-226Crossref PubMed Scopus (26) Google Scholar). For instance, the phytopathogenic γ-proteobacteria, Pectobacterium carotovorum and Dickeya dadantii, cause soft rot disease in a variety of plants through the action of several pectinases secreted by the T2SS, called Out (11Hugouvieux-Cotte-Pattat N. Condemine G. Shevchik V.E. Bacterial pectate lyases, structural and functional diversity.Environ. Microbiol. Rep. 2014; 6: 427-440Crossref PubMed Scopus (80) Google Scholar, 12Hugouvieux-Cotte-Pattat N. Condemine G. Gueguen E. Shevchik V.E. Dickeya plant pathogens.in: eLS. John Wiley & Sons, Ltd, New York, NY2020: 1-10Crossref Google Scholar). The T2SS is a sophisticated transenvelope scaffold that is composed of at least 12 conserved core elements, generically called GspC to GspO (for General Secretory Pathway) or, more specifically, OutC to OutO for Pectobacterium and Dickeya. An inner membrane platform, formed by GspC, L, M, and F, interacts with the cytoplasmic ATPase GspE. GspE is thought to energize the assembly of the periplasmic pseudopilus, composed of GspG, H, I, J, and K, which participates in the translocation of folded exoproteins through the proteinaceous channel formed by the outer membrane secretin GspD (8Gu S. Shevchik V.E. Shaw R. Pickersgill R.W. Garnett J.A. The role of intrinsic disorder and dynamics in the assembly and function of the type II secretion system.Biochim. Biophys. Acta. 2017; 1865: 1255-1266Crossref Scopus (12) Google Scholar, 9Korotkov K.V. Sandkvist M. Hol W.G. The type II secretion system: Biogenesis, molecular architecture and mechanism.Nat. Rev. Microbiol. 2012; 10: 336-351Crossref PubMed Scopus (284) Google Scholar, 10Thomassin J.L. Santos Moreno J. Guilvout I. Tran Van Nhieu G. Francetic O. The trans-envelope architecture and function of the type 2 secretion system: New insights raising new questions.Mol. Microbiol. 2017; 105: 211-226Crossref PubMed Scopus (26) Google Scholar). The folded nature of proteins secreted by T2SS, together with an apparent absence of any common linear sequences, led to a widely held hypothesis that the T2SS secretion determinants could be of a structural nature (13Palomäki T. Pickersgill R. Riekki R. Romantschuk M. Saarilahti H.T. A putative three-dimensional targeting motif of polygalacturonase (PehA), a protein secreted through the type II (GSP) pathway in Erwinia carotovora.Mol. Microbiol. 2002; 43: 585-596Crossref PubMed Scopus (26) Google Scholar, 14Filloux A. Secretion signal and protein targeting in bacteria: A biological puzzle.J. Bacteriol. 2010; 192: 3847-3849Crossref PubMed Scopus (27) Google Scholar, 15Francetić O. Pugsley A.P. Towards the identification of type II secretion signals in a nonacylated variant of pullulanase from Klebsiella oxytoca.J. Bacteriol. 2005; 187: 7045-7055Crossref PubMed Scopus (41) Google Scholar, 16Hirst T.R. Holmgren J. Conformation of protein secreted across bacterial outer membranes: A study of enterotoxin translocation from Vibrio cholerae.Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7418-7422Crossref PubMed Scopus (109) Google Scholar). The T2SS is very versatile, and depending on the bacteria, it allows the secretion of up to 20 sequence-unrelated and structurally dissimilar proteins (11Hugouvieux-Cotte-Pattat N. Condemine G. Shevchik V.E. Bacterial pectate lyases, structural and functional diversity.Environ. Microbiol. Rep. 2014; 6: 427-440Crossref PubMed Scopus (80) Google Scholar, 17Sikora A.E. Zielke R.A. Lawrence D.A. Andrews P.C. Sandkvist M. Proteomic analysis of the Vibrio cholerae type II secretome reveals new proteins, including three related serine proteases.J. Biol. Chem. 2011; 286: 16555-16566Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar, 18White R.C. Cianciotto N.P. Assessing the impact, genomics and evolution of type II secretion across a large, medically important genus: The Legionella type II secretion paradigm.Microb. Genom. 2019; 5e000273Google Scholar, 19Filloux A. Protein secretion systems in Pseudomonas aeruginosa: An essay on diversity, evolution, and function.Front. Microbiol. 2011; 2: 155Crossref PubMed Scopus (88) Google Scholar). On the other hand, the same T2SS restricts the secretion of very similar orthologous proteins from other species. For example, the T2SS of D. dadantii secretes more than 15 different proteins, but it can discriminate between its own pectate lyase PelI and the Pel3 from P. carotovorum, although they share 67% of sequence identity (20Bouley J. Condemine G. Shevchik V.E. The PDZ domain of OutC and the N-terminal region of OutD determine the secretion specificity of the type II out pathway of Erwinia chrysanthemi.J. Mol. Biol. 2001; 308: 205-219Crossref PubMed Scopus (76) Google Scholar). Recently, we have exploited this species-specific secretion to study the molecular mechanisms of substrate recognition by T2SS (21Pineau C. Guschinskaya N. Robert X. Gouet P. Ballut L. Shevchik V.E. Substrate recognition by the bacterial type II secretion system: More than a simple interaction.Mol. Microbiol. 2014; 94: 126-140Crossref PubMed Scopus (25) Google Scholar). We have shown that PelI interacts with two T2SS components, the inner membrane GspC and the outer membrane GspD (21Pineau C. Guschinskaya N. Robert X. Gouet P. Ballut L. Shevchik V.E. Substrate recognition by the bacterial type II secretion system: More than a simple interaction.Mol. Microbiol. 2014; 94: 126-140Crossref PubMed Scopus (25) Google Scholar). In addition, we have found that an exposed 9 residue-long region, loop 3 of PelI, acts as a specific secretion signal that controls protein recruitment by the T2SS. The interaction of this loop with the dedicated domains of GspC and GspD is essential for the T2SS to discriminate between the cognate substrate PelI and heterologous Pel3. Furthermore, these data suggest that some other zones of PelI could also be involved in protein recruitment by the T2SS, indicating that this process is more complex than simply the recognition of a single loop (21Pineau C. Guschinskaya N. Robert X. Gouet P. Ballut L. Shevchik V.E. Substrate recognition by the bacterial type II secretion system: More than a simple interaction.Mol. Microbiol. 2014; 94: 126-140Crossref PubMed Scopus (25) Google Scholar). Previous studies have largely benefited from the high-resolution structures available for the pectate lyase PelI and for the periplasmic domains of the GspC and GspD components (22Korotkov K.V. Johnson T.L. Jobling M.G. Pruneda J. Pardon E. Heroux A. Turley S. Steyaert J. Holmes R.K. Sandkvist M. Hol W.G. Structural and functional studies on the interaction of GspC and GspD in the type II secretion system.PLoS Pathog. 2011; 7e1002228Crossref PubMed Scopus (67) Google Scholar, 23Gu S. Kelly G. Wang X. Frenkiel T. Shevchik V.E. Pickersgill R.W. Solution structure of homology region (HR) domain of type II secretion system.J. Biol. Chem. 2012; 287: 9072-9080Abstract Full Text Full Text PDF PubMed Scopus (18) Google Scholar, 24Creze C. Castang S. Derivery E. Haser R. Hugouvieux-Cotte-Pattat N. Shevchik V.E. Gouet P. The crystal structure of pectate lyase peli from soft rot pathogen Erwinia chrysanthemi in complex with its substrate.J. Biol. Chem. 2008; 283: 18260-18268Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 25Korotkov K.V. Pardon E. Steyaert J. Hol W.G. Crystal structure of the N-terminal domain of the secretin GspD from ETEC determined with the assistance of a nanobody.Structure. 2009; 17: 255-265Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar). However, the structure of Pel3 from P. carotovorum was still unresolved. It can be expected that the orthologous Pel3 and PelI would share a similar overall topology, but fine specific structural features may also exist since Pel3 is not recognized by the T2SS of D. dadantii. Therefore, we undertook a structural characterization of Pel3 to identify such potential secretion determinants and to carry out a rational design and construction of the Pel3 variants that could be secreted by D. dadantii. We reveal that Pel3 shares the same general topology as PelI, consisting of an N-terminal domain of a fibronectin type III fold (Fn3) linked to a catalytic domain (CD) that adopts a parallel β-helix topology. Whereas the core structure of both Fn3 and CD is well conserved in Pel3 and PelI, significant differences were observed for several exposed loop regions, indicating that these are putative secretion determinants. To test this hypothesis, we systematically substituted such divergent zones of Pel3 with those from PelI and then assessed the secretion of the generated hybrids in D. dadantii. We have demonstrated that, in addition to loop 3 of Fn3, several other loop regions are essential for secretion. Some of them are spatially close to the loop 3 and could together constitute a composite secretion determinant, whereas others are more distant and may act as independent secretion signals. These data suggest that the proteins secreted by T2SS pass through a multifaceted control that monitors the adequacy of several secretion determinants. Remarkably, structural analysis of Pel3 reveals that the key secretion signal, loop 3 of Fn3, is present in the crystals in different conformations. Such conformational transitions could reflect a transient structuration of loop 3 when in contact with an appropriate T2SS component. We hypothesize that such T2SS-induced structuration of some intrinsically disordered zones of the protein to be secreted is part of the recruitment mechanism used by T2SS. We crystallized the full-length Pel3 in two types of monoclinic crystals with one or two monomers in the asymmetric unit, named Pel31m and Pel32m, and solved the structures at 1.8 and 2.1 Å resolution, respectively (Table S1). The structures are very similar, with the same fold and domain arrangement, as testified by an r.m.s. deviation of 1.6 Å on all Cα pairs. Pel3 adopts a compact pear-like overall shape composed of two unequal domains, a small fibronectin type III domain (Fn3) (residues 1–109) and a large catalytic domain that has a β-helix fold (residues 120–347) (Fig. 1). The two domains are linked by a decapeptide segment (residues 110–119), which is observed in the electron density map of Pel31m, and have the same respective orientation in Pel31m and Pel32m. When superimposed onto the structure of the orthologous PelI from D. dadantii, the same domain composition and overall shape are observed, with an r.m.s. deviation of 2 Å on all Cα pairs (Fig. 1A). The pectate lyases (EC 4.2.2.2 and EC 4.2.2.9) catalyze the cleavage of polymeric α-1-4-linked polygalacturonic acid within the pectin component of the plant cell wall, leaving an unsaturated C4–C5 bond at the newly formed nonreducing end (24Creze C. Castang S. Derivery E. Haser R. Hugouvieux-Cotte-Pattat N. Shevchik V.E. Gouet P. The crystal structure of pectate lyase peli from soft rot pathogen Erwinia chrysanthemi in complex with its substrate.J. Biol. Chem. 2008; 283: 18260-18268Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar). Pectate lyases have been classified into five structurally and phylogenetically unrelated families of polysaccharide lyases (PL-1, 2, 3, 9, and 10) (www.cazy.org, (26Lombard V. Golaconda Ramulu H. Drula E. Coutinho P.M. Henrissat B. The carbohydrate-active enzymes database (CAZy) in 2013.Nucleic Acids Res. 2014; 42: D490-495Crossref PubMed Scopus (3219) Google Scholar)). The catalytic domain of Pel3 adopts the general β-helix fold typical of the structurally characterized proteins from the PL-3 family: it is made up of eight right-handed coils stacked on top of one another. Each coil consists of three consecutive strand-turn motifs, termed PBn.m-Tn.m, where n is 1 to 3 and m is 1 to 8 (Fig. 1). Five disulfide bonds are well conserved between PelI and Pel3 (Fig. S1): three of them (124/137, #1, 180/185, #3, and 312/337, #5) reinforce the stability of the β-solenoid fold while the two other disulfide bonds (146/196, #2 and 257/260, #4) fasten the positions of the extended loops. The hydrophobic core of the Pel3 β-helix is stabilized by a series of hydrophobic interactions between the inward-pointing side chains of several aliphatic residues, Ile, Val, and Leu. These residues are organized into two regular ladders, extending along the whole length of the β-helix, at the β-strands PB1 and PB3, respectively (Fig. S2). These aliphatic interstrand stackings are well conserved among the PL-3 family members, while in some proteins, Phe is present instead of aliphatic residues (Fig. 2). The Pel3 catalytic site is nearly identical to that of PelI and similar to those of two other structurally characterized PL-3 members from Bacillus sp. KSM-P15 and from Caldicellulosiruptor bescii (PDB entries 1EE6 and 3T9G, respectively) (27Alahuhta M. Chandrayan P. Kataeva I. Adams M.W. Himmel M.E. Lunin V.V. A 1.5 Å resolution X-ray structure of the catalytic module of Caldicellulosiruptor bescii family 3 pectate lyase.Acta Crystallogr.Sect. F Struct. Biol. Cryst. Commun. 2011; 67: 1498-1500Crossref PubMed Scopus (12) Google Scholar, 28Akita M. Suzuki A. Kobayashi T. Ito S. Yamane T. The first structure of pectate lyase belonging to polysaccharide lyase family 3.Acta Crystallogr. D Biol. Crystallogr. 2001; 57: 1786-1792Crossref PubMed Scopus (47) Google Scholar). It carries the same invariant residues with Lys227 acting as the catalytic base and Lys252 and Arg255 implicated in the binding of the substrate (Fig. 2 and Fig. S3). In the catalytic site of Pel31m, a sulfate ion was detected, which forms salt bridges with Lys227, Lys252, and Arg255 mimicking the hydroxyl groups of the natural substrate, polygalacturonic acid (Fig. S4). In addition, a structural calcium ion was present in the monomer B of Pel32m where it coordinates the main-chain carbonyl O of Ile195 and Pro221, the side-chain Oδ atom of Asp194 and Asn216, and two water molecules, which complete the pentagonal bipyramidal geometry (Fig. S3A). Superimposition of monomers A and B of Pel32m indicated that the Ca2+ ion seems to be necessary for a proper fold of loop T3.4. Indeed, in the monomer B, T3.4 forms a lid over the bound Ca2+, whereas this turn is highly destabilized, in the monomer A, in the absence of calcium (Fig. S5). Interestingly, this arrangement is specific to Pel3 and PelI as the loop forming T3.4 is absent from other pectate lyases of the PL-3 family (Fig. 2 and Fig. S3). Superimposition of the catalytic domains of Pel3 and PelI reveals some subtle structural differences that are essentially restricted to the extended loops formed by T3.1, T3.2, T3.4, and T1.8 (Fig. 3A). These zones have a low degree of sequence conservation and show variable arrangements in Pel3 and PelI. Notably, these loops are also differently arranged in the various crystal forms of Pel3, indicating their high intrinsic flexibility (Fig. S6). Such properties could be consistent with the formation of specific secretion patterns for cognate T2SS. The N-terminal domain of Pel3 has a seven-stranded Fn3 fold very similar to that of the Fn3 domain of PelI, with an r.m.s. deviation of 1.0 Å on all Cα pairs. It is composed of two antiparallel β-sheets packed against each other (Fig. 1). Except for Pel3, PelI, and a few close homologs from γ-proteobacteria, other characterized representatives of the PL-3 family do not possess an Fn3-like domain. Some PL-3 members carry, instead, a carbohydrate-binding module (Firmicutes and Fungi) or a ricin B-like lectin domain (Actinobacteria), while the Nematode's pectate lyases usually consist of the catalytic PL-3 domain only (Table S2). The Fn3 domain does not affect the catalytic activity of PelI and its biological function remains unclear (24Creze C. Castang S. Derivery E. Haser R. Hugouvieux-Cotte-Pattat N. Shevchik V.E. Gouet P. The crystal structure of pectate lyase peli from soft rot pathogen Erwinia chrysanthemi in complex with its substrate.J. Biol. Chem. 2008; 283: 18260-18268Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar). Fn3-like modules have also been identified in some other carbohydrate-active enzymes suggesting its possible implication in plant cell wall degradation (29Little E. Bork P. Doolittle R.F. Tracing the spread of fibronectin type III domains in bacterial glycohydrolases.J. Mol. Evol. 1994; 39: 631-643Crossref PubMed Scopus (90) Google Scholar, 30Kataeva I.A. Seidel 3rd, R.D. Shah A. West L.T. Li X.L. Ljungdahl L.G. The fibronectin type 3-like repeat from the Clostridium thermocellum cellobiohydrolase CbhA promotes hydrolysis of cellulose by modifying its surface.Appl. Environ. Microbiol. 2002; 68: 4292-4300Crossref PubMed Scopus (132) Google Scholar, 31Sidar A. Albuquerque E.D. Voshol G.P. Ram A.F.J. Vijgenboom E. Punt P.J. Carbohydrate binding modules: Diversity of domain architecture in amylases and cellulases from filamentous microorganisms.Front. Bioeng. Biotechnol. 2020; 8: 871Crossref PubMed Scopus (8) Google Scholar). The Fn3 topology is, however, very common in eukaryotes, and it has been found in about 2% of all animal proteins (32Bork P. Doolittle R.F. Proposed acquisition of an animal protein domain by bacteria.Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8990-8994Crossref PubMed Scopus (190) Google Scholar). Consistent with these observations, a search for structural homologs of Fn3 Pel3, using the Dali server (33Holm L. Laakso L.M. Dali server update.Nucleic Acids Res. 2016; 44: W351-355Crossref PubMed Scopus (561) Google Scholar), shows that the highest scoring hits are the Fn3 domains from eukaryotic proteins (Fig. S7 and Table S3). In contrast, a few characterized Fn3 domains from bacterial carbohydrate-active enzymes are structurally divergent from Fn3 Pel3. Regardless of the level of structural similarity of the considered Fn3 domains, their sequence identity is low (less than 15% for the best hits) and only a few aromatic and/or hydrophobic core-forming residues are conserved across these Fn3 domains (Fig. S7). In addition, the loop/turn regions vary widely, even between the orthologous Fn3 domains of Pel3 and PelI where the sequence identity is very poor (Fig. 1B and Fig. S7). This could be consistent with the specific functions of these zones. For instance, in some eukaryotic Fn3 domains, loop regions have been shown to constitute binding sites for cognate protein partners (34Kavran J.M. Ward M.D. Oladosu O.O. Mulepati S. Leahy D.J. All mammalian Hedgehog proteins interact with cell adhesion molecule, down-regulated by oncogenes (CDO) and brother of CDO (BOC) in a conserved manner.J. Biol. Chem. 2010; 285: 24584-24590Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar, 35McLellan J.S. Zheng X. Hauk G. Ghirlando R. Beachy P.A. Leahy D.J. The mode of Hedgehog binding to Ihog homologues is not conserved across different phyla.Nature. 2008; 455: 979-983Crossref PubMed Scopus (106) Google Scholar, 36Healey E.G. Bishop B. Elegheert J. Bell C.H. Padilla-Parra S. Siebold C. Repulsive guidance molecule is a structural bridge between neogenin and bone morphogenetic protein.Nat. Struct. Mol. Biol. 2015; 22: 458-465Crossref PubMed Scopus (62) Google Scholar). Comparison of the three monomer structures of Pel3 observed in the different crystals shows that loop 3 of Fn3, located between strands βC and βC', adopts distinct conformations depending on the intra- and intermolecular contacts. Notably, in the Pel32m monomer A, no electronic density corresponding to loop 3 could be detected, indicating that it is completely disordered (Fig. 4A). However, monomer B presents a well-structured loop with a 310-helix (Fig. 4B). The helix seems to be stabilized by Q59, which is sandwiched between two tryptophan residues, W89 from the same monomer B and W89∗ from monomer A. A similar conformation with a 310-helix is observed in Pel31m, where loop 3 is stabilized, by two hydrogen bonds, with another monomer and by a polar-π interaction between Q59 and W89 in the same monomer (Fig. 4C and Fig. S8B). Interestingly, a similar cation–π interaction is observed in the crystal of PelI, where R60 interacts with W87 and the loop 3 Fn3 is neatly ordered in a 310-helix (Fig. S8A). In PelI, the 310-helix is additionally stabilized by a salt bridge between D58 and K328∗, as well by a hydrogen bond between N56 and Y325∗. Since the loop 3 of Fn3 has been shown to interact with the cognate T2SS (21Pineau C. Guschinskaya N. Robert X. Gouet P. Ballut L. Shevchik V.E. Substrate recognition by the bacterial type II secretion system: More than a simple interaction.Mol. Microbiol. 2014; 94: 126-140Crossref PubMed Scopus (25) Google Scholar), it is tempting to hypothesize that such transient interactions with the T2SS components could play a similar structuring role during the Pel3/PelI recruitment by the secretion system. The two domains of Pel3 form a closed structure with a buried surface area of 880 Å2 stabilized by a series of hydrogen bonds and ionic interactions (Fig. S9). The Fn3 side of the interdomain interface includes several residues of strands βA, βB, and βE. On the CD side of the interface, the residues Asn166, Asn189, and Asn243 form a "Velcro"-like motif that offers a series of ionic and polar groups stabilizing the Fn3/CD interface. These residues form the turns T2 at coils 2, 3, and 5 of the β-helix, and they are well conserved among the PL-3 family members as a part of an Asn stacking (Fig. 2 and Fig. S2). In Pel3 and PelI, the Fn3 domain covers these residues of the CD until the protein is secreted from the periplasm by the T2SS. Once secreted in planta, the Fn3 domain of Pel3/PelI is cleaved off from the catalytic domain by the bacterial proteases (37Shevchik V.E. Boccara M. Vedel R. Hugouvieux-Cotte-Pattat N. Processing of the pectate lyase PelI by extracellular proteases of Erwinia chrysanthemi 3937.Mol. Microbiol. 1998; 29: 1459-1469Crossref PubMed Scopus (34) Google Scholar). A similar compact organization is observed for PelI, with an interdomain buried contact surface of 750 Å2 (Fig. S9). The interdomain interface has a similar orientation in Pel3 and PelI and most interacting residue pairs are conserved, though some specific contacts are more prominent in PelI. For instance, a salt bridge K29/E190 is present in the PelI interface but absent in Pel3 (Fig. S9). These subtle differences in organization of the interface cause some displacements of the Fn3 and CD domains relative to each other, in the Pel3 and PelI core structures. Indeed, when superimposing the catalytic domains of Pel3 and PelI, their respective Fn3 domains and, more specifically, the loops 3 and 5 are obviously shifted (Fig. 3A). Since the loop 3 of Fn3 has been previously identified as a key secretion signal, these differences could, in turn, affect recognition of the protein by the T2SS. Comparison of Pel3 and PelI revealed that their core structures composed of β-strands are highly conserved, whereas the arrangement of several extended loop regions of both Fn3 and CD varies significantly (Fig. 3A). To examine whether these regions could act as specific secretion signals, first, we systematically substituted these zones of Fn3 Pel3 with those from PelI and then assessed secretion of the generated hybrids in D. dadantii. As was expected from the previous study (21Pineau C. Guschinskaya N. Robert X. Gouet P. Ballut L. Shevchik V.E. Substrate recognition by the bacterial type II secretion system: More than a simple interaction.Mol. Microbiol. 2014; 94: 126-140Crossref PubMed Scopus (25) Google Scholar), the h16 hybrid, carrying the