An inactivated nuclease-like domain in RecC with novel function: implications for evolution
© Rigden; licensee BioMed Central Ltd. 2005
Received: 15 April 2005
Accepted: 28 June 2005
Published: 28 June 2005
The PD-(D/E)xK superfamily, containing a wide variety of other exo- and endonucleases, is a notable example of general function conservation in the face of extreme sequence and structural variation. Almost all members employ a small number of shared conserved residues to bind catalytically essential metal ions and thereby effect DNA cleavage. The crystal structure of the RecBCD prokaryotic DNA repair machinery shows that RecB contains such a nuclease domain at its C-terminus. The RecC C-terminal region was reported as having a novel fold.
The RecC C-terminal region can be divided into an alpha/beta domain and a smaller alpha-helical bundle domain. Here we show that the alpha/beta domain is homologous to the RecB nuclease domain but lacks the features necessary for catalysis. Instead, the domain has a novel function within the nuclease superfamily – providing a hoop through which single-stranded DNA passes. Comparison with other structures of nuclease domains bound to DNA reveals strikingly different modes of ligand binding. The alpha-helical bundle domain contributes the pin which splits the DNA duplex.
The demonstrated homology of RecB and RecC shows how evolution acted to produce the present RecBCD complex through aggregation of new domains as well as functional divergence and structural redeployment of existing domains. Distantly homologous nuclease(-like) domains bind DNA in highly diverse manners.
The largest evolutionary superfamilies of proteins cover such a large range of sequence space that the relationships shared by members may not be apparent by standard means of sequence comparison, and hence are often only recognized after structural determinations. Such has frequently been the case for the PD-(D/E)xK superfamily of nucleases. Within the superfamily, structures were first obtained for four restriction enzymes, of such diverse sequences that they were initially assumed not to share homology (reviewed in ). Since then structures have confirmed distant and often unexpected homologies of those four with many other restriction enzymes, as well as exo- and endo-nucleases involved in such diverse cellular processes as DNA repair , transposition , Holliday junction resolution  and recombination .
The unifying catalytic site characteristic of the superfamily is the presence of one or more catalytically essential divalent cations [6, 7]. The conserved acidic residues of the PD-(D/E)xK motif, which can be separated by any number of residues, bind one metal cation while the conserved lysine residue is involved in positioning water suitably to attack the DNA backbone. In some lineages of the superfamily variation on this classical motif is apparent in the substitution of the second acidic residue by a catalytically essential His residue (2), or in the migration of the second acidic residue  or the lysine residue  to other parts of the fold. Irrespective of this variation, the catalytic site is placed at one edge of the core four or five-stranded β-sheet at the heart of the α /β domain structure [1, 6, 7]. While an overwhelming majority of the superfamily contain one of these catalytic site variants some interesting exceptions have been noticed. Thus, while clearly containing a PD-(D/E)xK superfamily-like domain structure , the tRNA splicing endoribonuclease EndA, has evolved an unrelated catalytic site on the opposite side of the fold to the conventional site . A catalytically inactive version of the fold has also been seen in the N-terminal domain of S. cerevisiae RPB5, an RNA polymerase subunit, where evidence suggests that it functions in protein-protein interactions .
Although extremely diverse in structure and sequence, modern sequence comparison methods have played their part in elucidating the full range of PD-(D/E)xK superfamily members [9, 13–15]. Nevertheless, structure determinations and structure-informed bioinformatics  will continue to be crucial in this most diverse of superfamilies. Some five years ago it was predicted that the nuclease activity associated with the C-terminus of RecB  resulted from the presence of a domain homologous to that of λ-exonuclease, despite RecB not possessing a PD-(D/E)xK motif [13, 14]. This prediction has been recently confirmed with the crystal structure determination of the structure of the RecBCD heterotrimer . This remarkable complex (see  and references therein) which functions to process double-stranded breaks in DNA, contains two distinct helicase activities, contributed by RecB and RecD. Also present is a catalytically inactive subunit, RecC. Among its proposed roles is recognition of the Chi DNA sequence . Remarkably, twin helicase(-like) motor domains (canonically named 1A and 2A) are present in all three subunits, although those in RecC are inactivated and only those in RecB and RecC contain α-helical insert domains in each motor domain (named 1B and 2B, respectively). As mentioned, the helicase domains of RecB are followed by a PD-(D/E)xK superfamily nuclease domain 3. In contrast, the C-terminal 'domain 3' of RecC was reported as being of novel fold .
Here we show that the C-terminal region ('domain 3') of RecC can actually be dissected into two domains, the first of which is clearly related to PD-(D/E)xK superfamily nuclease domains (hereafter called simply nuclease domains) and particularly to the corresponding domain of RecB. The nuclease-like domain of RecC is inactivated and therefore possesses not even the metal-ligating residues of the PD-(D/E)xK motif. Instead, it carries out a function not hitherto observed in the superfamily, providing an aperture through which one strand of newly split DNA duplex is fed. Comparisons show that nuclease(-like) domains are extraordinarily versatile in their mode of interaction with duplex DNA. Characteristics of the RecC nuclease-like domain show that RecB and RecC share a common ancestor and reveal how evolution has progressed by sequential addition of domains to the C-terminus, as well as by altering function of, and repositioning of, existing domains.
Results and discussion
An unsuspected nuclease-like domain in RecC
Interestingly, comparison of the nuclease domains of RecB, RecC and λ-exonuclease shows that the Rec subunits clearly share a more recent ancestor than the common ancestor of all three structures. As Figures 1 and 2 show, a single helix present in λ-exonuclease is replaced in both RecB and RecC by a three-helix α-helical bundle. This bundle is not present in the more distant relatives of λ-exonuclease highlighted by the CE server  such as archaeal Holliday junction resolvase, tRNA endonuclease and the PvuII restriction enzyme. Curiously, the degree of structural superposition that can be achieved between the RecB and RecC nuclease domains and λ-exonuclease suggests no closer relationship between the former pair. For example, 71 Cα atoms of RecB nuclease domain may be superimposed on their equivalents in RecC to produce an rms deviation of 2.19 Å. In comparison, 82 Cα atoms of RecB superimpose on equivalents of λ-exonuclease with a lower rms deviation of 1.71Å. However, the superimposable three-helix α-helical bundle shared only by RecB and RecC (Figures 1 and 2) show that they are more closely evolutionarily related to each other than to other homologous structures. The closer structural superposition of RecB and λ-exonuclease seems likely to arise from their shared nuclease activity, while RecC has evolved a different function.
Novel function of the nuclease-like domain in RecC
As mentioned, nuclease domains as represented in the present PDB are extremely diverse in sequence but share conserved residues that bind essential metal ions and are almost invariably catalytically active. The recognition of the third domain of RecC as an inactivated nuclease domain highlights a wholly unexpected new function for a non-catalytic but clearly nuclease-like domain. As shown in Fig 2e, the nuclease-like domain of RecC provides a hoop through which a single strand of the newly separated DNA duplex is passed. The hoop is the entrance to the 5' channel leading to RecD in the RecBCD complex . The pin responsible for separating the two DNA strands consists of a loop extending out of the α-helical bundle domain 4 of RecC.
Structural comparisons show that a series of three structural adaptations have been required in RecC in order to achieve this novel ssDNA-hoop function. These involve three regions of sequence marked on Figures 2 and 3. Region 1 comprises a long linker sequence between the extended structure that starts the domain and the three helix α-helical bundle subdomain. This linker region is very poor in regular secondary structure and adopts dramatically different conformations in the two domains. Significant sequence identity between RecB and RecC seems absent in the region. In RecB this linker lies along the surface of the remainder of the domain. In dramatic contrast, region 1 in RecC has few contacts with the rest of the domain (although it contacts other parts of RecC – see below) and forms most of the rim of the hoop through which ssDNA is passed. Region 2 is the connection between the two strands forming an antiparallel β-sheet. In E. coli RecC the connection is a minimal β-turn and connections in other RecC sequences are also very short (Figure 3). In contrast, Region 2 in RecB is usually much larger, tracing out, in the E. coli RecB structure, an 11-residue α-helix and a substantial stretch lacking regular secondary structure. Structure comparison shows the reason for the short connectors in RecC (Figure 2); larger connectors occupy the same space as the fourth domain of RecC. Thus, a larger connector would be incompatible with a RecC-style pin domain. Region 3, providing the connector between a β-strand and an α-helix, is again larger in RecB than in RecC and again contains an α-helix in RecB. Here the reason for the shorter connector in RecC is even more fundamental; were it to have the longer connector of RecB, the aperture whereby ssDNA passes through the RecC nuclease-like domain would be sterically obstructed.
DNA interactions with nuclease and nuclease-like domains
Unfortunately, no structure of λ-exonuclease in complex with DNA is yet available. However, other enzymes sharing the same fold, including many type II restriction enzymes, have been crystallized in complex with DNA. Therefore, DNA-bound structures were sought for the enzymes identified as closest structural neighbours for λ-exonuclease by the CE server . This analysis pinpointed the restriction enzyme PvuII (PDB code 1pvi; ) and the vsr exonuclease (PDB code 1odg; ) involved in repair of bacterial G:T mismatches. Further analysis (not shown) showed that the mode of binding of DNA to Pvu II was, in fact, typical of many restriction enzymes, irrespective of dimeric vs tetrameric quaternary state and of differing modes of dimerization.
Homology of RecB and RecC
The observation of inactivated helicase-like domains in RecC was not considered reason enough to propose the existence of homology between RecB and RecC extending over their whole length . Indeed, both sequence and structural comparisons at first suggest that RecB more closely resembles other helicases than it does RecC. For example, in the results of PSI-BLAST  starting with E. coli RecB, PcrA, another helicase that contains large helical-insert domains in each helicase domain , appears as a significant hit (e = 6 × 10-9) in the results of the first iteration. In contrast, using an e-value cut-off of 0.0001 four iterations are required before RecC sequences, including that of E. coli RecC, appear among the significant hits. While the BLAST alignments centred on the helicase(-like) domains the C-terminal nuclease(-like) domains were sometimes matched, although PSI-BLAST runs of the nuclease domain of RecB failed to hit the nuclease-like domain of RecC, and vice versa. Similarly, structural comparisons show that both helicase domains and both α-helical insert domains of RecB are more similar to their counterparts in PcrA than to the corresponding RecC domains (not shown). Nevertheless, the clear homology of the RecB and RecC nuclease(-like) domains, evident in their common three α-helical bundle (see above) strongly suggests that RecB and RecC share a more recent common ancestor than they have in common with other extant helicases. How then to explain the apparently closer relationship of RecB with PcrA than with RecC? As was proposed for the nuclease(-like) domains (see above) it seems like the dramatic functional differences between corresponding RecB and RecC domains are responsible. As discussed above, the RecC nuclease-like domain is significantly shorter than the RecB nuclease domain in two key regions, each associated with its new role as provider of an ssDNA hoop. Thus, it seems plausible that the maintenance of helicase activity by the helicase domains of PcrA and RecB is responsible for their apparently closer relationship, the structural changes accompanying evolution of the helicase-like domains in RecC for new roles having obscured their more recent shared ancestry with RecB.
There is an interesting parallel to be drawn between RecBCD and AddAB (also known as RexAB), a different DNA repair system found in Gram positive bacteria where RecBCD is lacking (see  for a review). AddA and AddB also appear homologous and each possesses helicase and nuclease motifs. Within AddAB, it is AddB that recognises the Chi sequence and therefore is the counterpart of RecC in RecBCD. Most interestingly, however, both the nuclease domains of AddA and AddB appear to be active . The AddAB system may therefore resemble an evolutionarily intermediate stage, through which the RecBCD machine passed before inactivation of the RecC nuclease domain and recruitment of RecD.
In summary, the improved domain dissection of RecC presented here and its ramifications enhance our understanding of the evolutionary processes responsible for the remarkable DNA processing machinery that is the RecBCD complex . It is now even more apparent that relatively straightforward addition of modular functionality has been accompanied by quite dramatic functional evolution of homologous domains.
Protein structures were retrieved from the Protein Databank (PDB; ). Protein structural superpositions were obtained at the CE  and DALI  servers and by using the program LSQMAN . Structural relationships were also explored in the SCOP database . Protein structure visualization employed O  and PyMOL , the latter also being used for production of figures. Iterative database searches were carried out using PSI-BLAST . Sequences were retrieved from the COG  entries for RecB (COG1074) and RecC (COG1330). Maximally diverse representatives were chosen using JALVIEW  which was also used for general sequence manipulation. Protein sequence alignment was carried out using MUSCLE  and T-COFFEE . Formatting of sequence alignments was done with ESPRIPT  using default options for colouring of sequence conservation.
I am grateful for the helpful remarks of one of the anonymous referees regarding AddAB.
- Aggarwal AK: Structure and function of restriction endonucleases. Curr Opin Struct Biol 1995, 5: 11–19. 10.1016/0959-440X(95)80004-KView ArticlePubMedGoogle Scholar
- Tsutakawa SE, Muto T, Kawate T, Jingami H, Kunishima N, Ariyoshi M, Kohda D, Nakagawa M, Morikawa K: Crystallographic and functional studies of very short patch repair endonuclease. Mol Cell 1999, 3: 621–628. 10.1016/S1097-2765(00)80355-XView ArticlePubMedGoogle Scholar
- Hickman AB, Li Y, Mathew SV, May EW, Craig NL, Dyda F: Unexpected structural diversity in DNA recombination: the restriction endonuclease connection. Mol Cell 2000, 5: 1025–1034. 10.1016/S1097-2765(00)80267-1View ArticlePubMedGoogle Scholar
- Bond CS, Kvaratskhelia M, Richard D, White MF, Hunter WN: Structure of Hjc, a Holliday junction resolvase, from Sulfolobus solfataricus. Proc Natl Acad Sci U S A 2001, 98: 5509–5514. 10.1073/pnas.091613398PubMed CentralView ArticlePubMedGoogle Scholar
- Kovall R, Matthews BW: Toroidal structure of lambda-exonuclease. Science 1997, 277: 1824–1827. 10.1126/science.277.5333.1824View ArticlePubMedGoogle Scholar
- Kovall RA, Matthews BW: Type II restriction endonucleases: structural, functional and evolutionary relationships. Curr Opin Chem Biol 1999, 3: 578–583. 10.1016/S1367-5931(99)00012-5View ArticlePubMedGoogle Scholar
- Pingoud A, Jeltsch A: Structure and function of type II restriction endonucleases. Nucleic Acids Res 2001, 29: 3705–3727. 10.1093/nar/29.18.3705PubMed CentralView ArticlePubMedGoogle Scholar
- Skirgaila R, Grazulis S, Bozic D, Huber R, Siksnys V: Structure-based redesign of the catalytic/metal binding site of Cfr10I restriction endonuclease reveals importance of spatial rather than sequence conservation of active centre residues. J Mol Biol 1998, 279: 473–481. 10.1006/jmbi.1998.1803View ArticlePubMedGoogle Scholar
- Feder M, Bujnicki JM: Identification of a new family of putative PD-(D/E)XK nucleases with unusual phylogenomic distribution and a new type of the active site. BMC Genomics 2005, 6: 21. 10.1186/1471-2164-6-21PubMed CentralView ArticlePubMedGoogle Scholar
- Bujnicki JM, Rychlewski L: Unusual evolutionary history of tRNA splicing the endonuclease EndA: relationship to the LAGLIDADG and PD-(D/E)XK deoxyribonucleases. Protein Sci 2001, 10: 656–660. 10.1110/ps.37101PubMed CentralView ArticlePubMedGoogle Scholar
- Li H, Trotta CR, Abelson J: Crystal structure and evolution of a transfer RNA splicing enzyme. Science 1998, 280: 279–284. 10.1126/science.280.5361.279View ArticlePubMedGoogle Scholar
- Todone F, Weinzierl RO, Brick P, Onesti S: Crystal structure of RPB5, a universal eukaryotic RNA polymerase subunit and transcription factor interaction target. Proc Natl Acad Sci U S A 2000, 97: 6306–6310. 10.1073/pnas.97.12.6306PubMed CentralView ArticlePubMedGoogle Scholar
- Daiyasu H, Komori K, Sakae S, Ishino Y, Toh H: Hjc resolvase is a distantly related member of the type II restriction endonuclease family. Nucleic Acids Res 2000, 28: 4540–4543. 10.1093/nar/28.22.4540PubMed CentralView ArticlePubMedGoogle Scholar
- Aravind L, Makarova KS, Koonin EV: Holliday junction resolvases and related nucleases: identification of new families, phyletic distribution and evolutionary trajectories. Nucleic Acids Res 2000, 28: 3417–3432. 10.1093/nar/28.18.3417PubMed CentralView ArticlePubMedGoogle Scholar
- Bujnicki JM, Rychlewski L: Grouping together highly diverged PD-(D/E)XK nucleases and identification of novel superfamily members using structure-guided alignment of sequence profiles. J Mol Microbiol Biotechnol 2001, 3: 69–72.PubMedGoogle Scholar
- Bujnicki JM: Crystallographic and bioinformatic studies on restriction endonucleases: inference of evolutionary relationships in the "midnight zone" of homology. Curr Protein Pept Sci 2003, 4: 327–337. 10.2174/1389203033487072View ArticlePubMedGoogle Scholar
- Yu M, Souaya J, Julin DA: The 30-kDa C-terminal domain of the RecB protein is critical for the nuclease activity, but not the helicase activity, of the RecBCD enzyme from Escherichia coli. Proc Natl Acad Sci USA 1998, 95: 981–986. 10.1073/pnas.95.3.981PubMed CentralView ArticlePubMedGoogle Scholar
- Singleton MR, Dillingham MS, Gaudier M, Kowalczykowski SC, Wigley DB: Crystal structure of RecBCD enzyme reveals a machine for processing DNA breaks. Nature 2004, 432: 187–193. 10.1038/nature02988View ArticlePubMedGoogle Scholar
- Alexandrov N, Shindyalov I: PDP: protein domain parser. Bioinformatics 2003, 19: 429–430. 10.1093/bioinformatics/btg006View ArticlePubMedGoogle Scholar
- Holm L, Sander C: Dali: a network tool for protein structure comparison. Trends Biochem Sci 1995, 20: 478–480. 10.1016/S0968-0004(00)89105-7View ArticlePubMedGoogle Scholar
- Shindyalov IN, Bourne PE: Protein structure alignment by incremental combinatorial extension (CE) of the optimal path. Protein Eng 1998, 11: 739–747. 10.1093/protein/11.9.739View ArticlePubMedGoogle Scholar
- Cheng X, Balendiran K, Schildkraut I, Anderson JE: Structure of PvuII endonuclease with cognate DNA. EMBO J 1994, 13: 3927–3935.PubMed CentralPubMedGoogle Scholar
- Bunting KA, Roe SM, Headley A, Brown T, Savva R, Pearl LH: Crystal structure of the Escherichia coli dcm very-short-patch DNA repair endonuclease bound to its reaction product-site in a DNA superhelix. Nucleic Acids Res 2003, 31: 1633–1639. 10.1093/nar/gkg273PubMed CentralView ArticlePubMedGoogle Scholar
- Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997, 25: 3389–3402. 10.1093/nar/25.17.3389PubMed CentralView ArticlePubMedGoogle Scholar
- Subramanya HS, Bird LE, Brannigan JA, Wigley DB: Crystal structure of a DExx box DNA helicase. Nature 1996, 384: 379–383. 10.1038/384379a0View ArticlePubMedGoogle Scholar
- Chedin F, Kowalczykowski SC: A novel family of regulated helicases/nucleases from Gram-positive bacteria: insights into the initiation of DNA recombination. Mol Microbiol 2002, 43: 823–834. 10.1046/j.1365-2958.2002.02785.xView ArticlePubMedGoogle Scholar
- Quiberoni A, Biswas I, El Karoui M, Rezaiki L, Tailliez P, Gruss A: In vivo evidence for two active nuclease motifs in the double-strand break repair enzyme RexAB of Lactococcus lactis. J Bacteriol 2001, 183: 4071–4078. 10.1128/JB.183.13.4071-4078.2001PubMed CentralView ArticlePubMedGoogle Scholar
- Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE: The Protein Data Bank. Nucleic Acids Res 2000, 28: 235–242. 10.1093/nar/28.1.235PubMed CentralView ArticlePubMedGoogle Scholar
- Kleywegt GJ: Use of non-crystallographic symmetry in protein structure refinement. Acta Cryst 1996, D52: 842–857.Google Scholar
- Murzin AG, Brenner SE, Hubbard T, Chothia C: SCOP: a structural classification of proteins database for the investigation of sequences and structures. J Mol Biol 1995, 247: 536–540. [http://www.sciencedirect.com/science?_ob=IssueURL&_method=viewMINPage&_tockey=%23TOC%236899%239999%23997529995%239999%23MIN%23Volume_247,_Issue_4&_acct=C000044499&_version=1&_urlVersion=0&_userid=822084&md5=a0c105250d976d9d488358b33d8562fd] 10.1006/jmbi.1995.0159PubMedGoogle Scholar
- Jones TA, Zou JY, Cowan SW, Kjeldgaard M: Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Cryst 1991, A47: 110–119.View ArticleGoogle Scholar
- PyMOL Home Page[http://pymol.sourceforge.net]
- Tatusov RL, Natale DA, Garkavtsev IV, Tatusova TA, Shankavaram UT, Rao BS, Kiryutin B, Galperin MY, Fedorova ND, Koonin EV: The COG database: new developments in phylogenetic classification of proteins from complete genomes. Nucleic Acids Res 2001, 29: 22–28. 10.1093/nar/29.1.22PubMed CentralView ArticlePubMedGoogle Scholar
- Clamp M, Cuff J, Searle SM, Barton GJ: The Jalview Java alignment editor. Bioinformatics 2004, 20: 426–427. 10.1093/bioinformatics/btg430View ArticlePubMedGoogle Scholar
- Edgar RC: MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 2004, 32: 1792–1797. 10.1093/nar/gkh340PubMed CentralView ArticlePubMedGoogle Scholar
- Notredame C, Higgins DG, Heringa J: T-Coffee: A novel method for fast and accurate multiple sequence alignment. J Mol Biol 2004, 302: 205–217. 10.1006/jmbi.2000.4042View ArticleGoogle Scholar
- Gouet P, Courcelle E, Stuart DI, Metoz F: ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 1999, 15: 305–308. 10.1093/bioinformatics/15.4.305View ArticlePubMedGoogle Scholar
- Velankar SS, Soultanas P, Dillingham MS, Subramanya HS, Wigley DB: Crystal structures of complexes of PcrA DNA helicase with a DNA substrate indicate an inchworm mechanism. Cell 1999, 97: 75–84. 10.1016/S0092-8674(00)80716-3View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.