Advertisement

Structural and Mechanistic Insight into the Listeria monocytogenes Two-enzyme Lipoteichoic Acid Synthesis System*

  • Ivan Campeotto
    Affiliations
    Section of Microbiology and MRC Centre for Molecular Bacteriology and Infection, and Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • Matthew G. Percy
    Affiliations
    Section of Microbiology and MRC Centre for Molecular Bacteriology and Infection, and Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • James T. MacDonald
    Affiliations
    Centre for Structural Biology, Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • Andreas Förster
    Affiliations
    Centre for Structural Biology, Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • Paul S. Freemont
    Correspondence
    To whom correspondence may be addressed: Dept. of Medicine, Imperial College London, South Kensington Campus, Sir Ernst Chain Building, London SW7 2AZ, United Kingdom. Tel.: +44-207-594-5327
    Affiliations
    Centre for Structural Biology, Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • Angelika Gründling
    Correspondence
    To whom correspondence may be addressed: Section of Microbiology, South Kensington Campus, Flowers Building, Rm. 3.21, London SW7 2AZ, United Kingdom. Tel.: +44-207-594-5256
    Affiliations
    Section of Microbiology and MRC Centre for Molecular Bacteriology and Infection, and Imperial College London, London SW7 2AZ, United Kingdom
    Search for articles by this author
  • Author Footnotes
    * This work was supported by European Research Council Grant 260371 (to A. G.).
    This article contains supplemental Table S1.
    3 The abbreviations used are:LTAlipoteichoic acidPGPpolyphosphatidylglycerol phosphateDAGdiacylglycerolPGphosphatidylglycerolphosphatePtdphosphatidylGroPglycerolphosphatePDBProtein Data BankNBD7-nitrobenz-2-oxa-1,3-diazole-4-yl.
Open AccessPublished:August 15, 2014DOI:https://doi.org/10.1074/jbc.M114.590570
      Lipoteichoic acid (LTA) is an important cell wall component required for proper cell growth in many Gram-positive bacteria. In Listeria monocytogenes, two enzymes are required for the synthesis of this polyglycerolphosphate polymer. The LTA primase LtaPLm initiates LTA synthesis by transferring the first glycerolphosphate (GroP) subunit onto the glycolipid anchor and the LTA synthase LtaSLm extends the polymer by the repeated addition of GroP subunits to the tip of the growing chain. Here, we present the crystal structures of the enzymatic domains of LtaPLm and LtaSLm. Although the enzymes share the same fold, substantial differences in the cavity of the catalytic site and surface charge distribution contribute to enzyme specialization. The eLtaSLm structure was also determined in complex with GroP revealing a second GroP binding site. Mutational analysis confirmed an essential function for this binding site and allowed us to propose a model for the binding of the growing chain.

      Introduction

      Lipoteichoic acid (LTA)
      The abbreviations used are: LTA
      lipoteichoic acid
      PGP
      polyphosphatidylglycerol phosphate
      DAG
      diacylglycerol
      PG
      phosphatidylglycerolphosphate
      Ptd
      phosphatidyl
      GroP
      glycerolphosphate
      PDB
      Protein Data Bank
      NBD
      7-nitrobenz-2-oxa-1,3-diazole-4-yl.
      is an important cell wall component found in many Gram-positive bacteria, including human pathogens such as Staphylococcus aureus and Listeria monocytogenes. In its absence, bacteria are impaired in growth and show cell morphology and cell division defects (
      • Gründling A.
      • Schneewind O.
      Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus aureus.
      ,
      • Oku Y.
      • Kurokawa K.
      • Matsuo M.
      • Yamada S.
      • Lee B.L.
      • Sekimizu K.
      Pleiotropic roles of polyglycerolphosphate synthase of lipoteichoic acid in growth of Staphylococcus aureus cells.
      ,
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ). Therefore, enzymes involved in its synthesis are attractive targets for the design of new antimicrobials. This has been experimentally validated with the identification of a small molecule LTA synthesis inhibitor that prevented the growth of antibiotic-resistant Gram-positive bacteria as well as prolonging the survival of mice challenged with a lethal dose of S. aureus (
      • Richter S.G.
      • Elli D.
      • Kim H.K.
      • Hendrickx A.P.
      • Sorg J.A.
      • Schneewind O.
      • Missiakas D.
      Small molecule inhibitor of lipoteichoic acid synthesis is an antibiotic for Gram-positive bacteria.
      ).
      A common type of LTA consists of a linear 1,3-linked polyglycerolphosphate (PGP) polymer that is attached to the outside of the membrane via a glycolipid anchor (
      • Fisher W.
      ,
      • Percy M.G.
      • Gründling A.
      Lipoteichoic acid synthesis and function in Gram-positive bacteria.
      ). In L. monocytogenes, the glycolipids anchor is Gal(α1–2)-Glc(α1–3)-diacylglycerol (Gal-Glc-DAG) or Gal(α1–2)Ptd-6-Glc(α1–3)-DAG (Gal-Ptd-6Glc-DAG), in which the glucose moiety is lipidated with an additional phosphatidyl (Ptd) group (
      • Fisher W.
      ,
      • Hether N.W.
      • Jackson L.L.
      Lipoteichoic acid from Listeria monocytogenes.
      ,
      • Uchikawa K.
      • Sekikawa I.
      • Azuma I.
      Structural studies on lipoteichoic acids from four Listeria strains.
      ). The PGP backbone chain is polymerized by lipoteichoic acid synthase or LtaS-type enzymes (
      • Gründling A.
      • Schneewind O.
      Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus aureus.
      ). This class of enzyme uses the membrane lipid phosphatidylglycerolphosphate (PG) as a substrate, hydrolyzes the glycerolphosphate (GroP) head group of this lipid and subsequently adds it to the tip of the growing chain (
      • Koch H.U.
      • Haas R.
      • Fischer W.
      The role of lipoteichoic acid biosynthesis in membrane lipid metabolism of growing Staphylococcus aureus.
      ,
      • Taron D.J.
      • Childs 3rd, W.C.
      • Neuhaus F.C.
      Biosynthesis of d-alanyl-lipoteichoic acid: role of diglyceride kinase in the synthesis of phosphatidylglycerol for chain elongation.
      ). In S. aureus only one enzyme, namely LtaSSa, is required for LTA backbone synthesis. This enzyme initiates LTA synthesis by the transfer of the first GroP subunit onto the glycolipid anchor and subsequently polymerizes the backbone chain by the repeated addition of GroP subunits (
      • Gründling A.
      • Schneewind O.
      Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus aureus.
      ,
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ). In contrast, L. monocytogenes uses a two-enzyme system for LTA synthesis (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ). The lipoteichoic acid primase LtaPLm transfers the initial GroP subunits to the glycolipid anchor but is unable to extend the chain further. Chain polymerization is performed by the lipoteichoic acid synthase LtaSLm (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ).
      Regardless of whether LTA synthase or primase, LtaS-type enzymes, have the same overall architecture. They are composed of an N-terminal domain with five transmembrane helices, which is followed by an extracellular C-terminal domain (eLtaS) containing the catalytic site (recently reviewed in Ref.
      • Reichmann N.T.
      • Gründling A.
      Location, synthesis and function of glycolipids and polyglycerolphosphate lipoteichoic acid in Gram-positive bacteria of the phylum Firmicutes.
      ). For many organisms, including the human pathogens S. aureus, Staphylococcus epidermidis, L. monocytogenes, and Bacillus anthracis, it has been shown that LtaS is cleaved by an endogenous peptidase and a fraction of the extracellular eLtaS is released into the culture supernatant as well as partially retained within the cell wall fraction (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ,
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ,
      • Powers M.E.
      • Smith P.A.
      • Roberts T.C.
      • Fowler B.J.
      • King C.C.
      • Trauger S.A.
      • Siuzdak G.
      • Romesberg F.E.
      Type I signal peptidase and protein secretion in Staphylococcus epidermidis.
      ,
      • Antelmann H.
      • Williams R.C.
      • Miethke M.
      • Wipat A.
      • Albrecht D.
      • Harwood C.R.
      • Hecker M.
      The extracellular and cytoplasmic proteomes of the non-virulent Bacillus anthracis strain UM23C1–2.
      ,
      • Wörmann M.E.
      • Reichmann N.T.
      • Malone C.L.
      • Horswill A.R.
      • Gründling A.
      Proteolytic cleavage inactivates the Staphylococcus aureus lipoteichoic acid synthase.
      ). In vitro, the extracellular eLtaS has been shown to be sufficient for PG hydrolysis (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ,
      • Wörmann M.E.
      • Corrigan R.M.
      • Simpson P.J.
      • Matthews S.J.
      • Gründling A.
      Enzymatic activities and functional interdependencies of Bacillus subtilis lipoteichoic acid synthesis enzymes.
      ). However, expression of the extracellular enzymatic domain is not sufficient for LTA production in vivo and the full-length membrane embedded LtaS protein is required for polymer production (
      • Wörmann M.E.
      • Reichmann N.T.
      • Malone C.L.
      • Horswill A.R.
      • Gründling A.
      Proteolytic cleavage inactivates the Staphylococcus aureus lipoteichoic acid synthase.
      ).
      The structures of the extracellular enzymatic eLtaS domains of the S. aureus (PDB code 2W5Q) and B. subtilis (PDB code 2W8D) have been reported (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ,
      • Schirner K.
      • Marles-Wright J.
      • Lewis R.J.
      • Errington J.
      Distinct and essential morphogenic functions for wall- and lipo-teichoic acids in Bacillus subtilis.
      ). These previous studies showed that the enzymes are related to arylsulfatase family enzymes with the same α/β-barrel fold. A conserved metal binding site was revealed and its requirement for enzyme function confirmed experimentally (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ). In addition, a Thr amino acid within the active center was identified as the catalytic residue and its essential role was confirmed as an LtaSSa-T300A variant was enzymatically inactive both in vitro and in vivo (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ). The active site Thr was found to be phosphorylated in the B. subtilis and unmodified in the S. aureus structure, but the biological significance of this modification has not yet been determined. It was further hypothesized that the reaction proceeds through a covalent GroP-enzyme intermediate through the catalytic Thr (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ).
      To understand better the reaction mechanism and enzyme specificity of this class of proteins, we performed a structural analysis of the extracellular soluble domains of the two L. monocytogenes enzymes eLtaPLm and eLtaSLm. This analysis revealed a substantially smaller cavity around the catalytic center in the primase enzyme compared with the synthase enzyme. The eLtaSLm structure was also determined in complex with GroP. This led to the identification of a second GroP binding site in eLtaSLm that is essential for enzyme function. Detailed bioinformatics analyses revealed specific motifs that differentiate LtaS and LtaP enzymes and highlighted that primase-related enzymes are only present in a small subset of bacteria. Taken together the structural and functional data allowed us to propose a revised mechanism for LTA biosynthesis in Gram-positive bacteria.

      EXPERIMENTAL PROCEDURES

      Plasmid and Strain Construction

      Strains and primers used in this study are listed in TABLE 1, TABLE 2, respectively. Escherichia coli strains were grown in LB medium and L. monocytogenes strains in BHI medium. The cultures were grown at the indicated temperatures and the growth medium was supplemented with antibiotics as indicated in Table 1. Plasmids for the expression of eLtaSLm variants with T307A, S486A, N488A, and H489A single amino acid substitutions and the triple mutant S486A/N488A/H489A (AAA variant) were constructed by QuikChange mutagenesis using plasmid pProEX-eLtaSLm (strain ANG1449) as template and primer pairs ANG1649/ANG1650, ANG1651/ANG1652, ANG1653/ANG1654, ANG1655/ANG1656, and ANG1657/ANG1658. The resulting plasmids were initially transformed into E. coli strain XL1-Blue yielding strains ANG2935 to ANG2939 and subsequently transformed for protein expressing into the E. coli Rosetta strain yielding strains ANG2940 to ANG2944. Plasmid pPL3-lmo0927His6 (Strain ANG1401) allows for the expression of full-length LtaSLm with a C-terminal His tag from its native promoter in L. monocytogenes (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ). This vector was used as template for the construction of plasmids pPL3-lmo0927His6-T307A, pPL3-lmo0927His6-S286A, pPL3-lmo0927His6-N488A, pPL3-lmo0927His6-H489A, pPL3-lmo0927His6-AAA for the expression of the different LtaSLm variants in L. monocytogenes. The desired mutations were introduced by SOE PCR. More specifically, plasmid pPL3-lmo0927His6-T307A was constructed by amplifying the front and back of lmo0927 and introducing the desired point mutation using plasmid pPL3-lmo0927His6 as template and primer pairs ANG674/ANG1650 and ANG676/ANG1649 in two separate PCR reactions. The two fragments were subsequently fused in a second round of PCR using primers ANG674/ANG676. The resulting product was digested with PstI and SalI and ligated with vector pPL3 that has been cut with the same enzymes. Plasmids pPL3-lmo0927His6-S286A, pPL3-lmo0927His6-N488A, pPL3-lmo0927His6-H489A, and pPL3-lmo0927His6-AAA were constructed using the same strategy and primers ANG1652 to ANG1658 as listed in Table 2. The resulting plasmids were initially recovered in E. coli strain XL1-Blue yielding strains ANG2930 to ANG2934 and subsequently transformed along with plasmid pPL3-lmo0927His6 into E. coli strain SM10 yielding strains ANG1460 and ANG2946 to ANG2950. Next all plasmids were conjugated from SM10 into L. monocytogenes strain 10403SΔlmo0927 using a previously described method (
      • Lauer P.
      • Chow M.Y.
      • Loessner M.J.
      • Portnoy D.A.
      • Calendar R.
      Construction, characterization, and use of two Listeria monocytogenes site-specific phage integration vectors.
      ) but maintaining the L. monocytogenes 10403SΔlmo0927 strain at 30 °C throughout the procedure. This yielded L. monocytogenes strains ANG1454, and ANG2951 to ANG2955, which were also propagated at 30 °C. The sequences of all inserts were verified by automated fluorescence sequencing at the MRC Clinical Sciences Centre Genomics Core Laboratory, Imperial College London.
      TABLE 1Bacterial strains used in this study
      StrainRelevant featuresReference
      Escherichia coli strains
      XL1 BlueCloning strain, TetR – ANG127Stratagene
      SM10E. coli strain used for conjugations; KanR – ANG618
      • Simon R.
      • Priefer U.
      • Pühler A.
      A broad host range mobilization system for in vitro genetic engineering: transposon mutagenesis in Gram negative bacteria.
      DH-E898XL1 Blue pPL3; L. monocytogenes integration vector; CamR – ANG1276
      • Gründling A.
      • Burrack L.S.
      • Bouwer H.G.
      • Higgins D.E.
      Listeria monocytogenes regulates flagellar motility gene expression through MogR, a transcriptional repressor required for virulence.
      ANG1401XL1 Blue pPL3-lmo0927His6; Lmo0927 (LtaSLm) with C-terminal His-tag under native promoter control; CamR
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ANG1449DH5α pProEX-eLtaSLm; plasmid for expression of eLtaSLm; AmpR
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ANG1478Rosetta pProEX-eLtaPLm; strain for overexpression of eLtaPLm; AmpR
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ANG1479Rosetta pProEX-eLtaSLm; strain for overexpression of eLtaSLm; AmpR
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ANG2930XL1-Blue pPL3-lmo0927His6-T307A; Lmo0927-T307A with C-terminal His-tag under native promoter control; CamRThis study
      ANG2931XL1-Blue pPL3-lmo0927His6-S486A; Lmo0927-S486A with C-terminal His-tag under native promoter control; CamRThis study
      ANG2932XL1-Blue pPL3-lmo0927His6-N488A; Lmo0927-N488A with C-terminal His-tag under native promoter control; CamRThis study
      ANG2933XL1-Blue pPL3-lmo0927His6-H489A; Lmo0927-H489A with C-terminal His-tag under native promoter control; CamRThis study
      ANG2934XL1-Blue pPL3-lmo0927His6-AAA; Lmo0927-AAA with C-terminal His-tag under native promoter control; CamRThis study
      ANG2935XL1-Blue pProEX-eLtaSLm-T307A; plasmid for expression of eLtaSLm-T307A variant; AmpRThis study
      ANG2936XL1-Blue pProEX-eLtaSLm-S486A; plasmid for expression of eLtaSLm-S486A variant; AmpRThis study
      ANG2937XL1-Blue pProEX-eLtaSLm-N488A; plasmid for expression of eLtaSLm-N488A variant; AmpRThis study
      ANG2938XL1-Blue pProEX-eLtaSLm-H489A; plasmid for expression of eLtaSLm-H489A variant; AmpRThis study
      ANG2939XL1-Blue pProEX-eLtaSLm-AAA; plasmid for expression of eLtaSLm-AAA variant; AmpRThis study
      ANG2940Rosetta pProEX-eLtaSLm-T307A; strain for overexpression of eLtaSLm-T307A variant; AmpRThis study
      ANG2941Rosetta pProEX-eLtaSLm-S486A; strain for overexpression of eLtaSLm-S486A variant; AmpRThis study
      ANG2942Rosetta pProEX-eLtaSLm-N488A; strain for overexpression of eLtaSLm-N488A variant; AmpRThis study
      ANG2943Rosetta pProEX-eLtaSLm-H489A; strain for overexpression of eLtaSLm-H489A variant; AmpRThis study
      ANG2944Rosetta pProEX-eLtaSLm-AAA; strain for overexpression of eLtaSLm-AAA variant; AmpRThis study
      ANG1460SM10 pPL3-lmo0927His6; KanR, CamRThis study
      ANG2946SM10 pPL3-lmo0927His6-T307A; KanR, CamRThis study
      ANG2947SM10 pPL3-lmo0927His6-S486A; KanR, CamRThis study
      ANG2948SM10 pPL3-lmo0927His6-N488A; KanR, CamRThis study
      ANG2949SM10 pPL3-lmo0927His6-H489A; KanR, CamRThis study
      ANG2950SM10 pPL3-lmo0927His6-AAA; KanR, CamRThis study
      Listeria monocytogenes strains
      10403SStrepR – ANG1263
      • Bishop D.K.
      • Hinrichs D.J.
      Adoptive transfer of immunity to Listeria monocytogenes. The influence of in vitro stimulation on lymphocyte subset requirements.
      ANG138610403SΔlmo0927; StrepRThis study
      ANG141110403SΔlmo0927 pPL3; StrepR, CamRThis study
      ANG145410403SΔlmo0927 pPL3-lmo0927His6; StrepR, CamRThis study
      ANG295110403SΔlmo0927 pPL3-lmo0927His6-T307A; StrepR, CamRThis study
      ANG295210403SΔlmo0927 pPL3-lmo0927His6-S486A; StrepR, CamRThis study
      ANG295310403SΔlmo0927 pPL3-lmo0927His6-N488A; StrepR, CamRThis study
      ANG295410403SΔlmo0927 pPL3-lmo0927His6-H489A; StrepR, CamRThis study
      ANG295510403SΔlmo0927 pPL3-lmo0927His6-AAA; StrepR, CamRThis study
      TABLE 2Primers used in this study
      NumberNameSequence
      ANG6745-PstI-Lmo0927-withPAACTGCAGCTAGCAGACTTCCATTCCAAATGGTTC
      ANG6763-SalI-Lmo0927-C-HisACGCGTCGACTTAGTGATGGTGATGGTGATGaccTTTATCGGATGAATCAGTTGATTTTTTC
      ANG16495-Lmo0927-T307ACCACCAAACTGGACAAGGGAAAGCAGCTGACTCCGAAATGTTAC
      ANG16503-Lmo0927-T307AGTAACATTTCGGAGTCAGCTGCTTTCCCTTGTCCAGTTTGGTGG
      ANG16515-Lmo0927-S486AGTACGGTGACCATTATGGTATTGCCGACAACCATGAAGAAGCAATG
      ANG16523-Lmo0927-S486ACATTGCTTCTTCATGGTTGTCGGCAATACCATAATGGTCACCGTAC
      ANG16535-Lmo0927-N488AGACCATTATGGTATTTCCGACGCCCATGAAGAAGCAATGACAAAAATTC
      ANG16543-Lmo0927-N488AGAATTTTTGTCATTGCTTCTTCATGGGCGTCGGAAATACCATAATGGTC
      ANG16555-Lmo0927-H489ACCATTATGGTATTTCCGACAACGCTGAAGAAGCAATGACAAAAATTCTTG
      ANG16563-Lmo0927-H489ACAAGAATTTTTGTCATTGCTTCTTCAGCGTTGTCGGAAATACCATAATGG
      ANG16575-Lmo0927-AAAGTACGGTGACCATTATGGTATTGCCGACGCCGCTGAAGAAGCAATGACAAAAATTCTTG
      ANG16583-Lmo0927-AAACAAGAATTTTTGTCATTGCTTCTTCAGCGGCGTCGGCAATACCATAATGGTCACCGTAC

      Protein Expression and Purification

      Strains ANG1478 Rosetta pProEX-eLtaPLm (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ) and ANG1479 Rosetta pProEX-eLtaSLm (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ) were used for the expression and purification of N terminally His-tagged eLtaPLm and eLtaSLm proteins, respectively. Protein induction and nickel affinity purification were performed as previously described (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ,
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ). The proteins were further purified by size exclusion chromatography using a Superdex S200 16/60 column (GE Healthcare) and a 50 mm Tris-HCl, pH 7.5, 200 mm NaCl, 5% glycerol buffer system for eLtaSLm and the different alanine substitution variants or 20 mm Tris-HCl, pH 7.5, for eLtaPLm. Protein-containing fractions spanning the main peak were pooled and concentrated to ∼10 mg/ml using 10-kDa molecular mass cut-off Amicon filtration devices (Millipore), if not otherwise stated. These proteins were subsequently used in structural studies. eLtaS variants with T307A, S486A, N488A, and H489A single amino acid substitutions and a S486A/N488A/H489A (AAA variant) triple mutant were expressed in E. coli strains ANG2940 to ANG2944 (Table 1). L. monocytogenes strain 10403S pPL3-LtaSLm-His6 (ANG1424) (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ) was used for the expression and purification of eLtaSLm from the native host.

      Purification of Native eLtaSLm from L. monocytogenes Culture Supernatant and Mass Spectrometry Analysis

      The L. monocytogenes strain 10403S pPL3-LtaSLm-His6 (ANG1424) (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ), which contains a plasmid for the expression of the C terminally His-tagged LtaSLm variant, was used for the purification of the secreted eLtaSLm fragment directly from Listeria culture supernatant. This strain was grown overnight in 6 liters of BHI medium. The bacterial cells were pelleted by centrifugation for 10 min at 7,000 × g and the cleared culture supernatant was filtered and loaded into a nickel-nitrilotriacetic acid column for protein purification, as previously reported (
      • Wörmann M.E.
      • Reichmann N.T.
      • Malone C.L.
      • Horswill A.R.
      • Gründling A.
      Proteolytic cleavage inactivates the Staphylococcus aureus lipoteichoic acid synthase.
      ). The elution fractions containing the C terminally His-tagged eLtaSLm protein were pooled together and concentrated to a final volume of ∼50 µl at 0.5 mg/ml using a 10-kDa molecular mass cut-off Centricon. The sample was mixed with an equal volume of protein loading buffer and 5 µg of protein separated on a 12% SDS-PAGE gel alongside 100 µg of eLtaSLm protein produced and purified from E. coli strain ANG1479. Protein bands were visualized by Coomassie staining. The eLtaSLm protein bands were excised from the gel, digested with chymotrypsin, and subjected to mass spectrometry analysis at the TAPLIN Mass spectrometry facility (Harvard Medical School, Boston, MA). The expected active site threonine containing peptide FHQTGQGKTADSEM (T catalytic threonine) has a calculated mass of 1536.6 Da when unmodified or 1616.6 Da with a phosphorylated threonine residue.

      Protein Crystallization and Structure Determination

      The solubility of eLtaPLm was 120 mg/ml in 20 mm Tris-HCl, pH 7.5, buffer and most crystallization drops remained clear in the initial screens. To decrease the solubility, the protein was subjected to Lys-methylation (
      • Walter T.S.
      • Meier C.
      • Assenberg R.
      • Au K.F.
      • Ren J.
      • Verma A.
      • Nettleship J.E.
      • Owens R.J.
      • Stuart D.I.
      • Grimes J.M.
      Lysine methylation as a routine rescue strategy for protein crystallization.
      ). Crystals appeared after 7–10 days at 4 °C in 100 mm sodium cacodylate buffer, pH 5.4, 100 mm MgCl2, 33% PEG2000 at a protein concentration of 40 mg/ml. Crystals were flash cooled in liquid nitrogen without additional cryoprotection. Non-methylated protein alone failed to produce crystals under these conditions. However, macro-seeding or micro-seeding using the methylated protein promoted crystallization of the non-methylated protein. Therefore a methylated seed stock, stored at stored at 4 °C, was routinely used for seeding. Data were collected at the SOLEIL synchrotron at the PROXIMA1 beamline (Saint-Aubin, France) from a single crystal at 100 K. The crystal belonged to space group P1 with unit cell parameters a = 53.20 Å, b = 53.70 Å, c = 85.07Å; α = 71.57°, β = 87.89°, γ = 65.12°. A mini-κ goniometer was used to obtain high completeness in all resolution shells. Data were indexed with XDS (
      • Kabsch W.
      Xds.
      ) and reduced with SCALA (
      • Kabsch W.
      Xds.
      ,
      • Evans P.
      Scaling and assessment of data quality.
      ) to 1.75-Å resolution. The Rfree set was generated randomly in UNIQUE (
      • Winn M.D.
      • Ballard C.C.
      • Cowtan K.D.
      • Dodson E.J.
      • Emsley P.
      • Evans P.R.
      • Keegan R.M.
      • Krissinel E.B.
      • Leslie A.G.
      • McCoy A.
      • McNicholas S.J.
      • Murshudov G.N.
      • Pannu N.S.
      • Potterton E.A.
      • Powell H.R.
      • Read R.J.
      • Vagin A.
      • Wilson K.S.
      Overview of the CCP4 suite and current developments.
      ). The structure was solved by molecular replacement using PHASER as implemented in PHENIX AutoMR (
      • Adams P.D.
      • Afonine P.V.
      • Bunkóczi G.
      • Chen V.B.
      • Davis I.W.
      • Echols N.
      • Headd J.J.
      • Hung L.W.
      • Kapral G.J.
      • Grosse-Kunstleve R.W.
      • McCoy A.J.
      • Moriarty N.W.
      • Oeffner R.
      • Read R.J.
      • Richardson D.C.
      • Richardson J.S.
      • Terwilliger T.C.
      • Zwart P.H.
      PHENIX: a comprehensive Python-based system for macromolecular structure solution.
      ) using, after side chain pruning and ligands removal in SCULPTOR (
      • Adams P.D.
      • Afonine P.V.
      • Bunkóczi G.
      • Chen V.B.
      • Davis I.W.
      • Echols N.
      • Headd J.J.
      • Hung L.W.
      • Kapral G.J.
      • Grosse-Kunstleve R.W.
      • McCoy A.J.
      • Moriarty N.W.
      • Oeffner R.
      • Read R.J.
      • Richardson D.C.
      • Richardson J.S.
      • Terwilliger T.C.
      • Zwart P.H.
      PHENIX: a comprehensive Python-based system for macromolecular structure solution.
      ), the S. aureus eLtaS structure as model (PDB 2W5Q). Initial refinement and model building were performed in PHENIX AutoBuild and completed by cycles of reiterated manual building in COOT (
      • Emsley P.
      • Lohkamp B.
      • Scott W.G.
      • Cowtan K.
      Features and development of COOT.
      ) and refinement in REFMAC (
      • Murshudov G.N.
      • Skubák P.
      • Lebedev A.A.
      • Pannu N.S.
      • Steiner R.A.
      • Nicholls R.A.
      • Winn M.D.
      • Long F.
      • Vagin A.A.
      REFMAC5 for the refinement of macromolecular crystal structures.
      ). Structure validation was performed using MOLPROBITY (
      • Davis I.W.
      • Leaver-Fay A.
      • Chen V.B.
      • Block J.N.
      • Kapral G.J.
      • Wang X.
      • Murray L.W.
      • Arendall 3rd, W.B.
      • Snoeyink J.
      • Richardson J.S.
      • Richardson D.C.
      MolProbity: all-atom contacts and structure validation for proteins and nucleic acids.
      ).
      Crystals of eLtaSLm grew in 5–7 days at 20 °C in 0.64 m sodium acetate, pH 4.6, 4% PEG3350, 100 mm MgCl2 and were cryo-protected with 25% PEG400 before flash-cooling in liquid N2. A micro-seeding technique was employed to improve the crystal size (
      • Bergfors T.
      Seeds to crystals.
      ) and crystallization trials were repeated in the same buffer conditions but lowering the protein concentration to 5 mg/ml. For the GroP co-crystallization experiments, the protein was incubated for 10 min at room temperature with a final concentration of 50 mm GroP. The crystals obtained from the co-crystallization were further soaked for 5 min in crystallization buffer supplemented with 25% PEG400 and 50 mm GroP before flash cooling in liquid N2. Data collection of the apo-eLtaSLm was performed at the Diamond Light Source synchrotron, beamline I24 (Didcot, Oxford, UK), from a single crystal at 100 K. The apo-structure of eLtaSLm belonged to the space group P41212 with unit cell dimensions of a = b = 119.76 Å, c = 473.91 Å; α = β = γ = 90.0°. The data were indexed, scaled, and Rfree was generated randomly in UNIQUE (
      • Winn M.D.
      • Ballard C.C.
      • Cowtan K.D.
      • Dodson E.J.
      • Emsley P.
      • Evans P.R.
      • Keegan R.M.
      • Krissinel E.B.
      • Leslie A.G.
      • McCoy A.
      • McNicholas S.J.
      • Murshudov G.N.
      • Pannu N.S.
      • Potterton E.A.
      • Powell H.R.
      • Read R.J.
      • Vagin A.
      • Wilson K.S.
      Overview of the CCP4 suite and current developments.
      ). The structure was solved by molecular replacement using BALBES (
      • Long F.
      • Vagin A.A.
      • Young P.
      • Murshudov G.N.
      BALBES: a molecular-replacement pipeline.
      ) and the B. subtilis eLtaSBs structure (PDB code 2W8D) as a starting model. Rigid body and restrained refinement produced a drop of Rfactor and Rfree from 42 and 43% to 25 and 31%, respectively. The structure was refined and validated as described above for eLtaPLm.
      The data collection of the eLtaSLm-GroP complex was performed at the Diamond Light Source synchrotron beamline I04-1 (Didcot, Oxford, UK) from a single crystal at 100 K. The crystals belonged to the space group P212121 with unit cell dimensions of a = 119.25 Å, b = 119.63 Å, c = 472.66 Å; α = β = γ = 90.0°. Indexing was performed in XDS and data merging was performed in SCALA and TRUNCATE (
      • Winn M.D.
      • Ballard C.C.
      • Cowtan K.D.
      • Dodson E.J.
      • Emsley P.
      • Evans P.R.
      • Keegan R.M.
      • Krissinel E.B.
      • Leslie A.G.
      • McCoy A.
      • McNicholas S.J.
      • Murshudov G.N.
      • Pannu N.S.
      • Potterton E.A.
      • Powell H.R.
      • Read R.J.
      • Vagin A.
      • Wilson K.S.
      Overview of the CCP4 suite and current developments.
      ) H- and L-test analysis in TRUNCATE highlighted the presence of pseudo-meroheydral twinning. The Rfree set was generated randomly in UNIQUE and the structure was solved by molecular replacement in PHASER using apo-eLtaSLm as a model. Ten cycles of rigid body refinement (10.0–6.0 Å) followed by 10 cycles of restrained refinement in REFMAC gave an R value of 23.6% and Rfree of 25.0%. Twin refinement in REFMAC highlighted a twin fraction of 9% with twinning operator k, h, -l. Therefore the twin option was kept for the whole refinement process, which was iterated with manual building in COOT. The final step of the refinement with rotamer optimization was performed in PHENIX, which did not detect any twinning. Composite omit maps were calculated in PHENIX and used to orient the terminal OH group of GroP. Structure validation was performed using MOLPROBITY. Ligand coordinate and dictionary files were generated and regularized in JLIGAND (
      • Lebedev A.A.
      • Young P.
      • Isupov M.N.
      • Moroz O.V.
      • Vagin A.A.
      • Murshudov G.N.
      JLigand: a graphical tool for the CCP4 template-restraint library.
      ). Anomalous maps were generated using the SFTOOLS (
      • Winn M.D.
      • Ballard C.C.
      • Cowtan K.D.
      • Dodson E.J.
      • Emsley P.
      • Evans P.R.
      • Keegan R.M.
      • Krissinel E.B.
      • Leslie A.G.
      • McCoy A.
      • McNicholas S.J.
      • Murshudov G.N.
      • Pannu N.S.
      • Potterton E.A.
      • Powell H.R.
      • Read R.J.
      • Vagin A.
      • Wilson K.S.
      Overview of the CCP4 suite and current developments.
      ) and visualized in PYMOL. The statistics for all data sets are shown in Table 3.
      TABLE 3Dataset statistics
      eLtaP apoeLtaS apoeLtaS-GroP
      SynchrotronSoleilDiamondDiamond
      BeamlineProxima1I24I04–1
      Space groupP1P41212P212121
      a, b, c (Å)53.20119.76119.25
      α, β, γ (o)53.70119.76119.63
      85.04473.91472.66
      71.6390.0090.00
      76.7890.0090.00
      65.1290.0090.00
      Resolution (Å)47.96 (1.84-1.75)106.88 (3.16-3.00)48.71 (2.32-2.20)
      Rmerge
      Rmerge = ΣhΣl |Ihl − 〈Ih〉|/ΣhΣl 〈Ih〉, where Ihl is the Ith observation of reflection h and 〈Ih〉.
      0.069 (0.501)0.117 (0.560)0.089 (0.459)
      Rpim
      Rpim as described in Ref. 43.
      (all I+ and I)
      0.040 (0.293)0.062 (0.318)0.052 (0.265)
      I〉/SD 〈I10.7 (2.5)8.6 (2.3)11.5 (2.7)
      Completeness (%)94.8 (92.6)90.7 (84.6)98.2 (95.7)
      Redundancy3.8 (3.8)4.0 (3.5)3.8 (3.8)
      No. reflections293,738 (41,866)257,649 (29941)1,285,177 (180393)
      No. unique77,013 (11,049)63,877 (8,525)335,456 (47,358)
      Rfactor
      Rfactor = Σh‖Fobs(h)| − |Fcal(h)‖/Σh|Fobs(h), where Fobs(H) − Fcal(h) are the observed and calculated structure factors for reflection h, respectively.
      0.178 (0.296)0.222 (0.319)0.178 (0.208)
      Rfree
      Rfree factor was calculated same as Rfactor using 5% reflections, which were selected randomly and omitted from refinement.
      0.207 (0.330)0.260 (0.376)0.214 (0.237)
      No. atoms7,11916,74540,238
      Protein6,60416,74037,131
      Water4812,997
      Ligands345110
      Average B-factors (Å2)31.450.933.1
      Protein30.950.933.0 (30.6)
      B-factor calculated excluding the disordered monomer K.
      Waters37.033.7
      Ligands44.745.538.6
      Root mean square deviations
      Bond lengths (Å)0.0090.0060.007
      Bond angles (°)1.230.931.07
      Ramachandran most favored (%)979798
      Ramachandran additional allowed (%)332
      Ramachandran outliers (%)000
      PDB code4UOP4UOO4UOR
      a Rmerge = ΣhΣl |Ihl − 〈Ih〉|/ΣhΣlIh〉, where Ihl is the Ith observation of reflection h and 〈Ih〉.
      b Rpim as described in Ref.
      • Karplus P.A.
      • Diederichs K.
      Linking crystallographic model and data quality.
      .
      c Rfactor = ΣhFobs(h)| − |Fcal(h)‖/Σh|Fobs(h), where Fobs(H)Fcal(h) are the observed and calculated structure factors for reflection h, respectively.
      d Rfree factor was calculated same as Rfactor using 5% reflections, which were selected randomly and omitted from refinement.
      e B-factor calculated excluding the disordered monomer K.

      One-dimensional 1H NMR Analysis of eLtaPLm

      10 mg of eLtaPLm in 1 ml of 20 mm Tris-HCl, pH 7.5 buffer, was used for the one-dimensional 1H NMR analysis. 10% D2O was added to the protein sample and the spectra were recorded at 800 MHz at 37 °C before and after the addition of 10 mm EDTA final concentration.

      Modeling of the GroP Trimer in the Catalytic Site of eLtaSLm

      The coordinate and restraint files of the GroP trimer in its energy minimized form were generated with JLIGAND (
      • Lebedev A.A.
      • Young P.
      • Isupov M.N.
      • Moroz O.V.
      • Vagin A.A.
      • Murshudov G.N.
      JLigand: a graphical tool for the CCP4 template-restraint library.
      ). Superposition of the coordinates of the GroP trimer with the eLtaS-GroP complexes was performed in PYMOL.

      Enzyme Activity Assay

      The activity of wild-type eLtaSLm and eLtaSLm variants T307A, S486A, N488A, H489A, and S486A/N488A/H489A was determined as previously reported (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ). Briefly, 4 µg of the fluorescently labeled NBD-PG lipid substrate was incubated for 3 h at 37 °C with 30 µg of enzyme in 10 mm sodium succinate buffer, pH 6.0, adjusted to an ionic strength of 50 with NaCl and 10 mm MnCl2. The lipid reaction products were subsequently extracted with chloroform and methanol, separated by thin layer chromatography, and the signal of the NBD-DAG hydrolysis product quantified as previously described (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ). Each TLC plate contained a negative no-enzyme control lane to determine the background signal, as well as a wild-type eLtaSLm enzyme reaction, which was for normalization purposes set to 100%. The activity of the different variants was calculated as percentage of activity compared with the wild-type control reaction. Four independent experiments with two different protein purifications were performed and the average percentage of activity and standard deviation were plotted.

      LTA and Protein Detection by Western Blot

      The different L. monocytogenes strains were grown overnight at 30 °C in BHI medium. Sample analysis for the detection of LTA or the His-tagged LtaS variants by Western blot was performed as previously described (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ).

      Listeria Growth Curves and Microscopy Analysis

      The indicated L. monocytogenes strains were grown overnight at 30 °C in BHI medium. The next day, the cultures were back diluted to an A600 of 0.05, incubated at 37 °C with shaking, and growth was monitored by determining A600 readings at timed intervals. For microscopy analysis, the different L. monocytogenes strain was propagated for at least 6 h at 37 °C in BHI medium. Subsequently culture aliquots were adjusted to an A600 of 0.5 and analyzed by phase-contrast microscopy using a Nikon Eclipse TS100 microscope with a ×20 objective. Images were recorded using a Sony HDR-CX11 high-definition camcorder mounted onto the microscope. Two independent microscopy experiments and three independent growth curves were performed and representative results are shown.

      Bioinformatics and Sequence Analysis

      Sequences homologous to the full-length LtaS and LtaP sequences were retrieved from the RefSeq microbial non-redundant database (
      • Pruitt K.D.
      • Tatusova T.
      • Klimke W.
      • Maglott D.R.
      NCBI Reference Sequences: current status, policy and new initiatives.
      ) using PSI-BLAST (
      • Altschul S.F.
      • Madden T.L.
      • Schäffer A.A.
      • Zhang J.
      • Zhang Z.
      • Miller W.
      • Lipman D.J.
      Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.
      ) with an E-value cutoff 1e-40. Sequences were filtered to have an alignment length of at least 400 residues, an identity of at least 28.7%, and similarity of 48.5% to either LtaS or LtaP. These cutoff values were chosen as they are the sequence identity and similarity between LtaS and LtaP. Sequences with a higher similarity to LtaP than LtaS were assigned to a primase-like sequence list (50 sequences), whereas sequences with a higher similarity to LtaS were assigned to a synthase-like list (1038 sequences). The LtaP and LtaS sequences were separately aligned using MUSCLE (
      • Edgar R.C.
      MUSCLE: multiple sequence alignment with high accuracy and high throughput.
      ) and then combined using MUSCLE profile-profile alignment. The phylogenetic tree using the combined alignment (having removed any columns not aligned to either LtaS or LtaP) was generated using the program PROML from PHYLIP version 2.3 (
      • Felsenstein J.
      PHYLIP-Phylogeny interference package (version 3.2).
      ) and plotted using the R package APE (
      • Paradis E.
      • Claude J.
      • Strimmer K.
      APE: analyses of phylogenetics and evolution in R language.
      ). All logo plots were produced using WebLogo (
      • Crooks G.E.
      • Hon G.
      • Chandonia J.M.
      • Brenner S.E.
      WebLogo: a sequence logo generator.
      ). For the PSICOV (
      • Jones D.T.
      • Buchan D.W.
      • Cozzetto D.
      • Pontil M.
      PSICOV: precise structural contact prediction using sparse inverse covariance estimation on large multiple sequence alignments.
      ) amino acid covariation analysis a new larger alignment was produced of LtaSLm homologous retrieved from the non-redundant database using PSI-BLAST and an E-value cutoff of 10−10. These sequences were individually aligned to the LtasSLm sequence using the BLOSUM62 matrix and Smith-Waterman algorithm, insertions were removed and pairwise alignments were combined to produce a multiple sequence alignment. Redundant sequences and sequences covering less than 60% of the LtaSLm sequence were removed resulting in 6943 final sequences. This final alignment was subsequently analyzed using the residue contact prediction program PSICOV (
      • Jones D.T.
      • Buchan D.W.
      • Cozzetto D.
      • Pontil M.
      PSICOV: precise structural contact prediction using sparse inverse covariance estimation on large multiple sequence alignments.
      ).

      RESULTS

      Apo-structures of eLtaPLm and LtaSLm

      To identify differences between LTA synthase and primase enzymes, the soluble extracellular enzymatic domains eLtaPLm and eLtaSLm were overexpressed and purified from E. coli and their crystal structures were determined at 1.75- and 3.0-Å resolution, respectively. Although both enzymes were monomers in solution, as assessed by size exclusion chromatography, eLtaPLm crystallized with two molecules in the asymmetric unit and eLtaSLm with five molecules in the asymmetric unit (Table 3). The overall structures of eLtaSLm and eLtaPLm are very similar (root mean square deviation = 1.4 Å for Cα atoms). Both comprise an α/β core and a C-terminal part of four anti-parallel β-strands and a long α-helix (Fig. 1). As predicted, both enzymes are similar to eLtaSSa (PDB code 2W5Q) and eLtaSBs (PDB code 2W8D) with a root mean square deviation on Cα atoms of 1.7 Å for eLtaPLm and 0.9 Å for eLtaSLm. Although the electrostatic surface potentials of eLtaSLm and eLtaPLm are similar around the α/β core at the N-terminal end, there are substantial differences in cavity size and surface charge distribution around the catalytic centers (Fig. 2, A and B).
      Figure thumbnail gr1
      FIGURE 1.Crystal structures of apo-eLtaPLm and apo-eLtaSLm. Schematic representation of the crystal structures of eLtaPLm (A) and eLtaSLm (B) with close ups of the catalytic Thr residues (left panel) and loop regions (right panel). The structural differences in the loop regions between eLtaPLm and eLtaSLm are highlighted in red and black, respectively. The 2FoFc electron density map (1.0 root mean square deviation) of Thr-279 in eLtaPLm is highlighted in blue and the omit FoFc map (3.0 root mean square deviation) of the phospho-Thr-307 in eLtaSLm is highlighted in green.
      Figure thumbnail gr2
      FIGURE 2.A and B, comparison of the catalytic pocket of eLtaPLm and eLtaSLm. Surface potential representation (blue, positive; red, negative; white, hydrophobic) of the area around the catalytic site of eLtaPLm (A) and eLtaSLm (B). The catalytic pocket of eLtaPLm is restricted through the highlighted amino acids Met-396 and Ile-458 and therefore significantly smaller and more hydrophobic than in eLtaSLm. C and D, surface potential representation of eLtaPLm (C) and eLtaSLm (D) structures with insertion loop regions boxed and a hydrophobic groove (white) stretching from loop 2 to the active site in eLtaPLm as indicated. E and F, web logo motif for the insertion loop 1 (E) and insertion loop 2 (F) region of the top 1090 LtaS-type sequences shown on top and the web logo motif for the 51 LtaP-type sequences shown below, using amino acid numbering for LtaPLm. The dimension of the letters in WebLogos are directly proportional to the degree of conservation of the given residue.
      A structure/sequence comparison of the two enzymes highlighted two sequence insertions in LtaP that form two extended loops (residues 544–552, loop 1; residues 561–570, loop 2), which interact with the long helix α18 (FIGURE 1., FIGURE 2.). There is no sequence conservation in loop 1 and loop 2 between eLtaPLm and eLtaSLm except for the salt bridge formed by residues Asp-600 and Arg-545, which correspond to Asp-616 and Arg-576 in the synthase enzyme. The insertion loop 2 in eLtaPLm forms a negatively charged protrusion, which is repositioned through Phe-566 on α18 by ∼2 Å compared with eLtaSLm. This also leads to the formation of a surface groove, which extends to the catalytic site (Fig. 2C). In eLtaSLm, this surface groove is constricted by Lys-306, which form a hydrogen bond with Tyr-483 (Fig. 2D). The specific loop 1 and loop 2 sequence insertions are conserved within primase homologues (Fig. 2, E and F) suggesting that the resulting surface features are specific for the function of primase enzymes.

      The Catalytic Threonine Is Unmodified in the Natural Host

      The catalytic residue of LtaS-type enzymes is a highly conserved Thr residue that in the B. subtilis eLtaSBs structure is phosphorylated but unmodified in the S. aureus eLtaSSa structure (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ,
      • Schirner K.
      • Marles-Wright J.
      • Lewis R.J.
      • Errington J.
      Distinct and essential morphogenic functions for wall- and lipo-teichoic acids in Bacillus subtilis.
      ). In this study, we found that Thr-307 in eLtaSLm is phosphorylated, whereas the corresponding residue Thr-279 in eLtaPLm is unmodified (Fig. 1). To gain insight into the physiological relevance of this modification, a C terminally His-tagged LtaSLm variant was expressed in L. monocytogenes and the cleaved eLtaSLm domain was purified from the culture supernatant. The purified protein was digested with chymotrypsin, and peptide fragments were analyzed by electron spray mass spectrometry. This analysis showed that for eLtaSLm expressed in E. coli the catalytic Thr is mostly phosphorylated (73%), whereas only 2% of the protein purified from the natural host is phosphorylated (Fig. 3). These data suggest that phosphorylation of the catalytic Thr is not physiological but is likely a result of expression in a heterologous host. However, as shown below this modification is likely a mimic of an enzyme-substrate intermediate.
      Figure thumbnail gr3
      FIGURE 3.Phosphorylation state of active site Thr as determined by mass spectrometry. A, eLtaSLm was purified from the E. coli cytoplasm or B, directly from the supernatant of a L. monocytogenes culture, separated on an SDS-PAGE gel, and subjected to a chymotrypsin digest and mass spectrometry analysis. The mass spectrometry traces corresponding to the active site containing peptide are shown for eLtaSLm purified from E. coli (A) or L. monocytogenes (B). The expected active site threonine containing peptide FHQTGQGKTADSEM (T, catalytic threonine) has a calculated mass of 1536.6 Da when unmodified or 1616.6 Da with a phosphorylated Thr residue. The fraction of protein with a phosphorylated active site Thr was estimated based on the intensity of the mass spectrometry signal and is indicated in % in each panel.

      Preferential Binding of Mn2+ to the Conserved Metal Binding Site

      LtaS-type proteins are metal-dependent enzymes and the highest in vitro enzyme activity is observed in the presence of Mn2+ (
      • Karatsa-Dodgson M.
      • Wörmann M.E.
      • Gründling A.
      In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
      ,
      • Wörmann M.E.
      • Corrigan R.M.
      • Simpson P.J.
      • Matthews S.J.
      • Gründling A.
      Enzymatic activities and functional interdependencies of Bacillus subtilis lipoteichoic acid synthesis enzymes.
      ). Our data show that the metal binding site is identical in the LtaSLm and LtaPLm structures. In previous LtaS crystal structures both Mn2+ and Mg2+ were identified in the metal binding site near the catalytic threonine, facilitating phosphatidylglycerol hydrolysis (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ,
      • Schirner K.
      • Marles-Wright J.
      • Lewis R.J.
      • Errington J.
      Distinct and essential morphogenic functions for wall- and lipo-teichoic acids in Bacillus subtilis.
      ). As the crystallization buffer for both Listeria proteins contained a high MgCl2 concentration, it is likely that Mg2+ is present in the active center in our structures. To determine the metal preference of the enzymes, crystallization trials were set up in the absence of any added metal ion. Although the eLtaPLm protein did not crystallize under these conditions, one-dimensional 1H NMR experiments showed an increase in peak sharpness upon addition of EDTA, suggesting the presence of a paramagnetic ion such as Mn2+ (Fig. 4A). Although the eLtaSLm crystals grown in the absence of any added metal ion diffracted only to 6.4 Å, anomalous difference maps showed a strong anomalous peak consistent with the presence of a bound Mn2+ ion after expression and purification (Fig. 4B). Together our data provide evidence for preferential Mn2+ binding of both eLtaPLm and eLtaSLm, in the absence of added metals consistent with previous biochemical activity measurements.
      Figure thumbnail gr4
      FIGURE 4.Characteristics of enzyme-bound metal. A, one-dimensional 1H NMR spectra of eLtaPLm in the absence or presence of EDTA one-dimensional NMR spectra of purified eLtaPLm protein recorded on a 800 MHz magnet at 37 °C before and after addition of 10 mm EDTA. B, anomalous electron density map of eLtaSLm. eLtaSLm crystals were grown in the absence of any added metal ions and data collected close to the Mn2+ edge (1.28 Å). The DANO SigDANO electron density map (shown in purple) confirms the presence of a Mn2+ ion.

      Identification of GroP Binding Sites in eLtaSLm

      LtaS-type enzymes belong to the arylsulfatase group of enzymes and the reaction mechanism of other members of this class of enzymes proceeds through the formation of a covalent enzyme-substrate intermediate. In the case of sulfatases, a post-translationally modified cysteine residue, a hydroxyformylglycine, is sulfated during catalysis (
      • Bond C.S.
      • Clements P.R.
      • Ashby S.J.
      • Collyer C.A.
      • Harrop S.J.
      • Hopwood J.J.
      • M J.G.
      Structure of a human lysosomal sulphatase.
      ). We previously speculated that LtaS-type enzymes also form a covalent GroP-Thr intermediate as part of the reaction mechanism (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ). Although we show here that the phosphorylation of the active site Thr residue observed in the eLtaSLm structure does not occur in the native host (Fig. 3), its presence in E. coli could, however, mimic such a covalent enzyme substrate intermediate. To provide additional experimental evidence for the formation of a covalent GroP-Thr intermediate, we performed co-crystallization and crystal soaking experiments with the eLtaSLm and PG lipid substrates with short chain fatty acids. However, co-crystallization experiments failed to produce crystals and crystal-soaking experiments abolished the diffraction power of the crystals. Next, co-crystallization and soaking experiments were performed with GroP, the hydrolysis product of the lipid substrate PG, and the structure was solved from crystals containing 11 molecules in the asymmetric unit.
      Using this approach, extra electron density was observed in each monomer within the catalytic site (Fig. 5). Similar as in the apo-structure, it was possible to build a phosphate group into a density extending from Thr-307 (Fig. 5). The phosphate oxygen binds to two structurally conserved water molecules, Trp-360, His-422, and a Mg2+ ion that is in turn further coordinated by Glu-263, Asp-481, and His-482 (Fig. 6, A and B). Additional difference electron density was observed in each monomer at the entrance of the catalytic pocket, into which a GroP molecule could be built (Fig. 5). In all chains, the phosphate group of the GroP molecule in this second site formed hydrogen bonds with residues Ser-486, Asn-488, and His-489 (Fig. 6, A and B). In eight molecules in the asymmetric unit an additional hydrogen bond was observed between the terminal hydroxyl group of GroP and a water molecule (W1), which in turn forms a hydrogen bond with Tyr-483 (Fig. 6B). In a previous study, the co-crystal structure of the S. aureus active site variant eLtaSSa-T300A with a GroP molecule within the active center was determined (PDB code 2W5R) (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ). The overlay of the catalytic sites of the GroP-eLtaSSa-T300A and the GroP-eLtaSLm structures revealed that the GroP molecule within the active center (referred to as GroP1) superposed with the phosphothreonine and the conserved water molecules W2 and W3 in eLtaSLm (Fig. 6C). Therefore, the phosphorylated Thr likely mimics a covalent GroP-Thr intermediate. The distance between the phosphorylated Thr and the terminal hydroxyl group of the GroP2 molecule bound at the entrance of the catalytic pocket is ∼6.3 Å, which is compatible with the length of one intervening GroP molecule. To test whether an additional GroP molecule could fit into this space, a GroP trimer model was generated in silico and fitted into the eLtaSLm structure using the experimental electron densities of the phosphothreonine and GroP as a guide (Fig. 6D). Our modeling showed that a GroP could fit in the intervening space suggesting that the growing PGP LTA chain could be bound in a similar manner during the catalytic cycle of eLtaSLm. The nature of the surface potential of the oligo-GroP binding groove further supports this conclusion (Fig. 6E). A series of ordered water molecules spans the catalytic site of eLtaSLm from residue His-353 to the trapped GroP2 molecule. The positions of these water molecules are conserved across all 11 monomers within a crystallographic unit and trace the position of the modeled GroP trimer (Fig. 6).
      Figure thumbnail gr5
      FIGURE 5.GroP/eLtaSLm co-crystal structure. A, ribbon representation of the active site of eLtaSLm with bound phosphate and a GroP molecule in a second site, as observed in chains B, C, D, F, H, I, J, and K. The phosphate group is covalently linked to the catalytic Thr-307 and GroP is bound at the entrance of the catalytic site. The omit FoFc electron density map (3.0 root mean square deviation) of the phospho-threonine (T307) and GroP are shown in green. B, electrostatic surface potential representation (blue, positive; red, negative) around the catalytic site showing the charge distribution surrounding the phospho-Thr-307 and the GroP molecule. C, ribbon representation of the active site of eLtaSLm with bound phosphate and a GroP molecule in a second site (as observed in chains A, E, and G) and D, the corresponding surface potential representation.
      Figure thumbnail gr6
      FIGURE 6.Detailed view of the active center with metal and ligand interactions. A, structural details of the eLtaSLm active site showing the phospho-Thr intermediate (TPO307) and the metal binding site (as seen in chains A, E, and G). Hydrogen bonds are indicated as gray lines for the metal binding site and as black lines for the ligands. The water molecules that are conserved in all 11 monomers are represented as spheres in pink. Residues His-353, Arg-362, Trp-360, His-422, and two water molecules are involved in the binding of the phosphate group and phospho-Thr-307, Glu-263, Asp-481, and His-482 coordinate the Mg2+ ion. The GroP molecule, which is 6.5 Å removed from the catalytic residue, forms hydrogen bonds with Ser-486, Asn-488, and His-489. B, structural details of the eLtaSLm active site showing the phospho-Thr intermediate (TPO307) and the metal binding site (as observed in chains B, C, D, F, H, I, J, and K). The GroP molecule forms hydrogen bonds with Ser-486, Asn-488, and His-489 like in panel A but in addition it also binds to a water molecule (W1), which in turn forms a hydrogen bond with Tyr-483. C, superposition between the catalytic site of eLtaSLm and the GroP molecules trapped in the eLtaSSa-T300A structure (PDB code 2W5R). D, model of the eLtaSLm active site bound to a GroP trimer. The GroP timer was produced and minimized with JLIGAND and superposed in PYMOL on the experimental crystal structure of the GroP-eLtaSLm co-crystal structure. E, electrostatic potential representation of the eLtaSLm active site with the modeled GroP trimer. For clarity, the image is rotated by +30° around the y axis compared with panel D.

      The Second GroP Binding Site in eLtaSLm Is Essential for Enzyme Function

      To test the functional requirement of the second GroP binding site, we mutated residues Ser-486, Asn-488, and His-489 to alanines individually or in combination and tested the mutant enzymes for their ability to produce LTA (Fig. 7). The different variants were expressed as C-terminal His tag fusion proteins in the L. monocytogenes strain 10403SΔltaS, which contains a deletion of the native ltaS gene. As negative controls, an empty vector or a vector for the expression of the catalytic site variant T307A (pPL3-ltaST307A-His6) were introduced into 10403SΔltaS and as positive control a vector for expression of wild-type LtaS (pPL3-ltaSHis6). Expression of all LtaS variants was confirmed by Western blot. As previously reported for WT LtaSLm (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ), all GroP binding site variants were cleaved and the eLtaS fragment was detected in the culture supernatant as well as in the cell wall-associated fraction (Fig. 7A). The active site T307A variant remained unprocessed and the full-length protein was observed in the cell wall-associated fraction (Fig. 7A). In a previous study, a similar accumulation of the full-length protein was observed in S. aureus for the catalytic site variant (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ), suggesting that an enzyme/substrate intermediate is required to position the enzyme for efficient processing. However, it should also be noted that the protein processing step does not serve as an enzyme activation step; to the contrary, based on experiments performed in S. aureus it has been proposed that the LtaS cleavage step serves as a mechanism to inactivate the enzyme (
      • Wörmann M.E.
      • Reichmann N.T.
      • Malone C.L.
      • Horswill A.R.
      • Gründling A.
      Proteolytic cleavage inactivates the Staphylococcus aureus lipoteichoic acid synthase.
      ).
      Figure thumbnail gr7
      FIGURE 7.In vivo and in vitro activity of eLtaSLm GroP binding site variants and bioinformatics analysis of conserved binding residues. A, detection of LtaSLm protein and LTA by Western blot. Samples of wild-type L. monocytogenes 10403S (WT) and the 10403SΔltaS-derived strains containing an empty pPL3 vector or expressing the indicated LtaSLm variants as C-terminal His tag fusion proteins were prepared for Western blot analysis. The LtaS protein was detected in the supernatant and cell wall-associated fractions using a His tag-specific antibody and LTA in the cell wall-associated fraction using a polyglycerolphosphate-specific antibody. B, in vitro enzyme activity assay with purified WT eLtaSLm and the different eLtaS variants. Enzyme reactions were set up using the fluorescently labeled lipid NBD-PG as substrate. The reaction products were separated on TLC plates and the NBD-DAG product was quantified. Four independent experiments were performed and the enzyme activity of the eLtaSLm protein (labeled WT in the graph) was set to 100% in each experiment. The relative activity of the different variants compared with WT eLtaSLm was calculated and the average value and S.D. plotted. C, sequence logo motif of active site, metal binding, active site GroP (GroP1) and second GroP (GroP2) binding site residues. The 51 LtaP-like sequences (top panels) and the 1039 LtaS-type sequences (bottom panels) were aligned and logo motifs for selected amino acid regions are shown. Active site residue (*), GroP1 (■), GroP2 (−), and metal binding residues (●) are indicated and amino acid numbering for the respective L. monocytogenes protein is shown.
      As expected, LTA production was restored to wild-type levels in the positive control strain 10403SΔltaS pPL3-ltaSHis6, whereas no LTA-specific signal was detected when extracts from the negative control strains were analyzed by Western blot using a polyglycerolphosphate-specific monoclonal LTA antibody (Fig. 7A). Expression of the S486A/N488A/H489A variant (LtaSAAA) did not restore LTA production, revealing an essential function of the second GroP binding site for LTA production. Analysis of the single amino acid variants showed that residues Ser-486 and His-489, but not Asn-488 are important for the LTA polymerization step (Fig. 7A).
      For successful LTA production, PG substrate hydrolysis and the GroP transfer reaction must take place. To determine whether the second GroP binding site is required specifically for PG hydrolysis, the WT and different eLtaS variants were produced in E. coli, purified, and used for in vitro enzyme reactions with fluorescently labeled NBD-PG lipid as substrate. As expected, mutating the catalytic Thr-307 residue abolished enzyme activity (Fig. 7B). The S486A and N488A variants retained the ability to hydrolyze PG, but the activity dropped by ∼50% compared with wild-type eLtaSLm. The H489A and S486A/N488A/H489A (AAA) variants showed a marked decrease in activity to around 20% of WT (Fig. 7B). These data show that the second GroP binding site, in particular residue His-489, is also important for the PG hydrolysis step. The S486A variant, however, is of particular interest as this variant retains significant PG hydrolysis activity, whereas the PGP polymerase activity is nearly abolished. We would suggest that this is due to the inability of this variant to interact with the growing PGP chain and therefore, similar to what is observed naturally in the LTA primase enzyme, the two reactions are decoupled in this variant.
      In a previous study, it has been shown that strain 10403SΔltaS has growth and morphological defects when propagated at 37 °C (
      • Webb A.J.
      • Karatsa-Dodgson M.
      • Gründling A.
      Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
      ). To investigate if expression of any of the LtaS variants allows for sufficient LTA production to restore these defects, growth and microscopy analysis was performed with the complementation strains. As expected the ltaS deletion strain displayed the expected growth defect and a filamentation phenotype, which could be complemented by introducing a wild-type ltaS allele (Fig. 8). For the other complementation strains, only expression of the ltaST307A allele did not restore the growth (Fig. 8A) and morphological defects (Fig. 8B). These results suggest that even if no signal for LTA is detected by Western blot, limited LTA synthesis must take place in these strains, which is sufficient to support normal growth and cell division.
      Figure thumbnail gr8
      FIGURE 8.Growth and microscopy analysis of wild-type L. monocytogenes, mutant, and complementation strains. A, growth curves. The wild-type L. monocytogenes strain 10403S (WT) and 10403SΔltaS-derived strains containing an empty pPL3 vector or a pPL3 vector with the indicated ltaS allele were grown at 37 °C in BHI medium, A600 readings determined at timed intervals and plotted. B, microscopy analysis. The same strains as used for growth curves in panel A were analyzed by phase-contrast microscopy following growth at 37 °C.

      Bioinformatics Analysis and Structure Guided Identification of LtaP and LtaS Enzyme Family Motifs

      To obtain an overview of the distribution of LtaP and LtaS-type enzymes among Gram-positive bacteria and to investigate the conservation of the structural features identified in this study, bioinformatics analyses were performed. To this end, homologues to full-length LtaS and LtaP sequences were retrieved and filtered to those with an alignment length of more than 400 residues yielding 1088 sequences. This was done to remove proteins that do not contain an N-terminal membrane domain and are therefore unlikely involved in LTA production. Of the 1088 retrieved sequences, only 50 showed greater homology to LtaP than to LtaS (supplemental Table S1). Primase family enzymes are present in the different Listeria species and similar to L. monocytogenes these species also contain an LtaS-type enzyme. This analysis highlighted that a two-enzyme LTA synthesis system with highly divergent enzymes as seen in Listeria sp. is not widely distributed among bacteria (Fig. 9 and supplemental Table S1). For instance, Bacillus sp. also contain multiple enzymes, but they are more closely related to one another than to the two enzymes found in Listeria sp (Fig. 9 and supplemental Table S1). This could indicate that either a gene duplication event took place more recently in Bacillus sp. or that the divergent primase-like enzyme was only retained in a few species such as Listeria, Thermotoga, and Paenibacillus sp. Primase-like enzymes also appear to be present in a few specific bacterial strains such as Planococcus donghaensis MPA1U2, Brevibacillus laterosporus LMG, and B. cereus cytotoxis NVH 391–98 (Fig. 9 and supplemental Table S1). The latter strain was isolated from a fatal case of enteritis. It is therefore plausible that the gene coding for the primase enzyme was acquired through horizontal gene transfer from a Listeria strain by co-inhabiting the same ecological niche. It is also of note that the Thermotoga sp., B. laterosporus LMG, and several of the Paenibacillus sp. do not contain an LtaS-type enzyme and hence are unlikely to produce an actual LTA polymer.
      Figure thumbnail gr9
      FIGURE 9.Unrooted phylogenetic tree of representative LtaS and LtaP-type enzymes. 1088 LtaS and LtaP sequence homologues were retrieved as described under “Experimental Procedures.” An unrooted phylogenetic tree was generated for representative LtaS and LtaP-type enzymes. Thirty of 50 LtaP-like protein sequences are shown in red and 28 of the remaining 1038 LtaS-like sequences are shown in green if the same bacterial stain also contains an LtaP-like enzyme or in black if the bacterial strain only contains LtaS-like enzymes. For clarity, the majority of the LtaS-type sequences, which would fall onto the right side of the tree, are not shown. The L. monocytogenes 10403S proteins analyzed in this study are indicated with dots. The scale bar indicates the branch length unit of the tree as inferred using the program PROML and is the expected fraction of amino acids changed. A complete list of the organisms and RefSeq accession numbers can be found in using the same color-coding with primase-like sequences shown in red and synthase-like sequences shown in green or black.
      As shown above, we have identified a second GroP binding site in LtaSLm and confirmed its importance for LTA production experimentally. Next, we analyzed distribution of binding site residues Ser-486, Asn-488, and His-489 across LTA synthesis enzymes. Separate alignments were produced for the 1038 LtaS-type sequences and the 50 LtaP-type sequences. Subsequently, a logo motif was created to visualize the conservation of amino acids across the whole enzyme family (data not shown). As expected, the active site threonine, as well as the metal binding residues, were highly conserved and present in both LtaP and LtaS-type enzymes (Fig. 7C). In addition, conserved residues in the active site, which are required for binding of the GroP molecule within the active center, could also be identified in both enzyme types (Fig. 7C). The second GroP binding site residues corresponding to Ser-486 and His-488 in LtaSLm were also conserved, however, only found in LtaS-type but not in primase-like enzymes (Fig. 7C). Based on our functional data, which showed that residues Ser-486 and His-488 are required for LTA production, we suggest that the absence of these residues is an important factor contributing to the inability of the LtaP enzyme to produce a PGP polymer.

      DISCUSSION

      Model for the Enzyme Reaction Mechanism and LTA Chain Extension of LtaS-type Enzymes

      Our new data presented in this study combined with previous results allow us to speculate how the LTA synthesis proceeds. We suggest that the reaction is initiated by nucleophilic attack of Thr-307 to PG resulting in the breakage of the phosphoester bond yielding one molecule of DAG and a covalent GroP-Thr intermediate (Fig. 10). LtaS belongs to the alkaline phosphatase superfamily and arylsulfatase family, in which Ser and Thr residues are often phosphorylated to be activated (
      • Krogh A.
      • Larsson B.
      • von Heijne G.
      • Sonnhammer E.L.
      Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes.
      ). For this reason it has been postulated that phosphorylation of the catalytic Thr as observed in the B. subtilis LtaS structure is required for initiation of the reaction (
      • Lu D.
      • Wörmann M.E.
      • Zhang X.
      • Schneewind O.
      • Gründling A.
      • Freemont P.S.
      Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
      ,
      • Schirner K.
      • Marles-Wright J.
      • Lewis R.J.
      • Errington J.
      Distinct and essential morphogenic functions for wall- and lipo-teichoic acids in Bacillus subtilis.
      ). However, we show in the current study that this is not the case for eLtaSLm. Although the active site threonine residue is phosphorylated in the eLtaSLm structure (Fig. 1), mass spectrometry analysis showed that this phosphorylation is likely an artifact caused by the purification of the protein from E. coli extracts as only a very small fraction of the protein obtained from the natural host L. monocygenes is phosphorylated (Fig. 3). The threonine phosphorylation is more likely to mimic the covalent GroP-Thr intermediate.
      Figure thumbnail gr10
      FIGURE 10.Proposed reaction mechanism of LtaSLm. The active site threonine is polarized by the Mn2+ ion allowing for a nucleophilic attack of Thr-307 to PG (1) generating the Thr-glycerolphosphate intermediate with the elimination of a DAG molecule (
      • Oku Y.
      • Kurokawa K.
      • Matsuo M.
      • Yamada S.
      • Lee B.L.
      • Sekimizu K.
      Pleiotropic roles of polyglycerolphosphate synthase of lipoteichoic acid in growth of Staphylococcus aureus cells.
      ). In our model the penultimate GroP molecule of the incoming GroP chain (GroPn) would be held in place within the second GroP binding site. The hydroxyl group of the terminal GroP unit will be deprotonated by a base (amino acid residue or water) (2) allowing for a nucleophilic attack on the Thr-GroP intermediate to occur (3). The product of the reaction, the LTA chain extended by one GroP unit, is released and the cycle completed through the deprotonation of the base of the reaction and the catalytic Thr-307 is repolarized by the metal ion (4).
      Next, the covalent GroP-Thr intermediate (GroP donor molecule) has to be attached to the incoming LTA chain (GroP acceptor molecule). In this study, we identified a second GroP binding site in the L. monocytogenes LtaS enzyme, which consists of residues S486A, N488A, and H489A. A reanalysis of the previously published S. aureus and B. subtilis eLtaS revealed that this binding site is identical in all three enzymes. It can be speculated that the tip of the LTA chain is bound in a similar manner to the GroP molecule within this second binding site. However, for a transfer reaction to occur, the enzyme would need to undergo a significant conformational change in order for the terminal hydroxyl group to reach the 6.3 Å removed charged active site threonine. Therefore we hypothesize that the trapped GroP molecule represents more likely the penultimate GroP subunit of a growing LTA chain (FIGURE 6., FIGURE 10.). Residues Lys-306 and Tyr-483 were located close to the active center, and could assist the binding of a terminal GroP subunit of an incoming chain by coordinating its phosphate group (Fig. 6). No electron density is observed for the side chain of Lys-306 in both the Listeria and Staphylococcus eLtaS enzymes, suggesting that the lysine is flexible and therefore could be used for stabilizing the phosphate group of an incoming terminal GroP (Fig. 6). It is of note that both Lys-306 and Tyr-483 are conserved residues among LtaS-type enzymes. In LtaP-type enzymes, where there is no requirement for binding of incoming GroP chains, these residues are replaced with Asn-278 and a range of amino acids at position 457 (Figs. 7C).
      For the polymerization reaction to occur the proton of the terminal hydroxyl group of the incoming LTA chain must be displaced. No obvious candidate residues can be identified in the vicinity of this terminal GroP or near the bound GroP2. Previous findings showing that the full-length enzyme is required in vivo for LTA production highlights a crucial function of the membrane domain for enzyme function (
      • Wörmann M.E.
      • Corrigan R.M.
      • Simpson P.J.
      • Matthews S.J.
      • Gründling A.
      Enzymatic activities and functional interdependencies of Bacillus subtilis lipoteichoic acid synthesis enzymes.
      ). One hypothesis is that a residue(s) within the transmembrane domain of the full-length LtaS enzyme could act as a base to remove a proton from the hydroxyl group of the acceptor GroP chain. Based on topology predictions, LtaSLm has five transmembrane helices and two extracellular loops, which span residues 35 to 48 (extracellular loop 1) and residues 98 to 105 (extracellular loop 2). Strikingly Asp-101 and Phe-102 within the second loop are highly conserved among LtaS-type enzymes but not in LtaP (data not shown) suggesting a possible functional role for these residues; in particular Asp-101 could act as a base required for the polymerization reaction. Once the terminal hydroxyl group is deprotonated it can act as a nucleophile to attack the phosphoester of the bound GroP-Thr assisted by the bound metal (Fig. 10).
      To date, no structural information is available for the membrane portion of any of the LTA synthesis enzymes. Previously it has been reported that hybrid proteins, in which the membrane and extracellular domains of two functional proteins are swapped, are non-functional suggesting a specific interaction between the transmembrane and extracellular enzymatic domains (
      • Wörmann M.E.
      • Corrigan R.M.
      • Simpson P.J.
      • Matthews S.J.
      • Gründling A.
      Enzymatic activities and functional interdependencies of Bacillus subtilis lipoteichoic acid synthesis enzymes.
      ). If a direct interaction between the two domains is crucial for enzyme function, one might expect interacting amino acids to co-vary within the two domains of LtaS enzymes. To explore this, a new larger alignment was made using 6943 sequences from the non-redundant database. Residue contacts were predicted using PSICOV and plotted alongside experimentally confirmed contacting amino acids based on the eLtaSLm structure (Fig. 11). Using this analysis, several residues within the transmembrane region were predicted to be in contact with amino acid residues within the extracellular domain (primarily located in proximity of the active site or at the back of the molecule), supporting the notion of a physical interaction between the transmembrane and extracellular domain.
      Figure thumbnail gr11
      FIGURE 11.Experimental and PSICOV-predicted contacts in LtaSLm. PSICOV-predicted residue contacts using 6943 sequences are shown in black and experimentally confirmed contacting amino acids (Cβ-Cβ distance <8 Å) based on the LtaSLm structure within the extracellular domain are shown in red. The eight predicted transmembrane domain extracellular domain contacts are labeled based on the LtaSLm amino acid numbering. Inset shows eLtaSLm structure with contacting amino acids in the extracellular domain highlighted in orange.
      The LtaPLm and LtaSLm structures determined as part of this study provide information on the molecular basis for the restricted enzyme activity and inability of the LtaPLm enzyme to polymerize LTA chains. Specifically, our work revealed that LtaPLm has a smaller active site cavity, lacks a second GroP binding site, and that two conserved loop insertions results in subtle alterations to surface cavities. These data allowed us to propose a model on how the incoming LTA chain could bind during the chain extension step. Supported by bioinformatics analyses, we further suggest that a crucial catalytic residue for activating the GroP acceptor chain might be located within the transmembrane domain. To confirm this and to understand the functional significance of highly conserved amino acids within the extracellular loops or the conserved aspartic acid residues with the fourth transmembrane helix will require further studies and in particular a structural investigation on the full-length enzyme.
      LTA synthesis enzymes are currently being actively pursued as target proteins for the development of novel antibiotics and recently, the first LtaS enzyme inhibitor was identified (
      • Richter S.G.
      • Elli D.
      • Kim H.K.
      • Hendrickx A.P.
      • Sorg J.A.
      • Schneewind O.
      • Missiakas D.
      Small molecule inhibitor of lipoteichoic acid synthesis is an antibiotic for Gram-positive bacteria.
      ). Based on our findings, we would suggest that future structure-based design of LTA synthesis enzyme inhibitors should be extended to include the second GroP binding site. We envisage that targeting this binding site may offer a better chance of obtaining LtaS-specific inhibitors and decrease the possibility of obtaining compounds that are cross-reactive toward members of the same protein family such as mammalian alkaline phosphatases. Expanding the chemical landscape search to a larger enzyme area might increase the chances of discovering new enzyme-specific inhibitors, which could be used to treat infections caused by important Gram-positive human pathogens.

      Acknowledgments

      We acknowledge the staff at beamline PROXIMA1 (SOLEIL synchrotron, Paris, France) and at beamlines I24 and I04-1 (DIAMOND synchrotron, Dicot, Oxford, United Kingdom) for assistance during data collection. We also thank Dr. Andrey Lebedev (STFC Rutherford Appleton Laboratory, Didcot, Oxford, UK) and Dr. Arwen Pearson (Hamburg Centre for Ultrafast Imaging, University of Hamburg, CFEL, Germany) for discussions about twinning and order-disorder structures.

      Supplementary Material

      REFERENCES

        • Gründling A.
        • Schneewind O.
        Synthesis of glycerol phosphate lipoteichoic acid in Staphylococcus aureus.
        Proc. Natl. Acad. Sci. U.S.A. 2007; 104: 8478-8483
        • Oku Y.
        • Kurokawa K.
        • Matsuo M.
        • Yamada S.
        • Lee B.L.
        • Sekimizu K.
        Pleiotropic roles of polyglycerolphosphate synthase of lipoteichoic acid in growth of Staphylococcus aureus cells.
        J. Bacteriol. 2009; 191: 141-151
        • Webb A.J.
        • Karatsa-Dodgson M.
        • Gründling A.
        Two-enzyme systems for glycolipid and polyglycerolphosphate lipoteichoic acid synthesis in Listeria monocytogenes.
        Mol. Microbiol. 2009; 74: 299-314
        • Richter S.G.
        • Elli D.
        • Kim H.K.
        • Hendrickx A.P.
        • Sorg J.A.
        • Schneewind O.
        • Missiakas D.
        Small molecule inhibitor of lipoteichoic acid synthesis is an antibiotic for Gram-positive bacteria.
        Proc. Natl. Acad. Sci. U.S.A. 2013; 110: 3531-3536
        • Fisher W.
        Bacterial Phosphoglycolipids and Lipoteichoic Acids. Plenum Press, New York1990
        • Percy M.G.
        • Gründling A.
        Lipoteichoic acid synthesis and function in Gram-positive bacteria.
        Annu. Rev. Microbiol. 2014; 68: 81-100
        • Hether N.W.
        • Jackson L.L.
        Lipoteichoic acid from Listeria monocytogenes.
        J. Bacteriol. 1983; 156: 809-817
        • Uchikawa K.
        • Sekikawa I.
        • Azuma I.
        Structural studies on lipoteichoic acids from four Listeria strains.
        J. Bacteriol. 1986; 168: 115-122
        • Koch H.U.
        • Haas R.
        • Fischer W.
        The role of lipoteichoic acid biosynthesis in membrane lipid metabolism of growing Staphylococcus aureus.
        Eur. J. Biochem. 1984; 138: 357-363
        • Taron D.J.
        • Childs 3rd, W.C.
        • Neuhaus F.C.
        Biosynthesis of d-alanyl-lipoteichoic acid: role of diglyceride kinase in the synthesis of phosphatidylglycerol for chain elongation.
        J. Bacteriol. 1983; 154: 1110-1116
        • Karatsa-Dodgson M.
        • Wörmann M.E.
        • Gründling A.
        In vitro analysis of the Staphylococcus aureus lipoteichoic acid synthase enzyme using fluorescently labeled lipids.
        J. Bacteriol. 2010; 192: 5341-5349
        • Reichmann N.T.
        • Gründling A.
        Location, synthesis and function of glycolipids and polyglycerolphosphate lipoteichoic acid in Gram-positive bacteria of the phylum Firmicutes.
        FEMS Microbiol. Lett. 2011; 319: 97-105
        • Lu D.
        • Wörmann M.E.
        • Zhang X.
        • Schneewind O.
        • Gründling A.
        • Freemont P.S.
        Structure-based mechanism of lipoteichoic acid synthesis by Staphylococcus aureus LtaS.
        Proc. Natl. Acad. Sci. U.S.A. 2009; 106: 1584-1589
        • Powers M.E.
        • Smith P.A.
        • Roberts T.C.
        • Fowler B.J.
        • King C.C.
        • Trauger S.A.
        • Siuzdak G.
        • Romesberg F.E.
        Type I signal peptidase and protein secretion in Staphylococcus epidermidis.
        J. Bacteriol. 2011; 193: 340-348
        • Antelmann H.
        • Williams R.C.
        • Miethke M.
        • Wipat A.
        • Albrecht D.
        • Harwood C.R.
        • Hecker M.
        The extracellular and cytoplasmic proteomes of the non-virulent Bacillus anthracis strain UM23C1–2.
        Proteomics. 2005; 5: 3684-3695
        • Wörmann M.E.
        • Reichmann N.T.
        • Malone C.L.
        • Horswill A.R.
        • Gründling A.
        Proteolytic cleavage inactivates the Staphylococcus aureus lipoteichoic acid synthase.
        J. Bacteriol. 2011; 193: 5279-5291
        • Wörmann M.E.
        • Corrigan R.M.
        • Simpson P.J.
        • Matthews S.J.
        • Gründling A.
        Enzymatic activities and functional interdependencies of Bacillus subtilis lipoteichoic acid synthesis enzymes.
        Mol. Microbiol. 2011; 79: 566-583
        • Schirner K.
        • Marles-Wright J.
        • Lewis R.J.
        • Errington J.
        Distinct and essential morphogenic functions for wall- and lipo-teichoic acids in Bacillus subtilis.
        EMBO J. 2009; 28: 830-842
        • Lauer P.
        • Chow M.Y.
        • Loessner M.J.
        • Portnoy D.A.
        • Calendar R.
        Construction, characterization, and use of two Listeria monocytogenes site-specific phage integration vectors.
        J. Bacteriol. 2002; 184: 4177-4186
        • Walter T.S.
        • Meier C.
        • Assenberg R.
        • Au K.F.
        • Ren J.
        • Verma A.
        • Nettleship J.E.
        • Owens R.J.
        • Stuart D.I.
        • Grimes J.M.
        Lysine methylation as a routine rescue strategy for protein crystallization.
        Structure. 2006; 14: 1617-1622
        • Kabsch W.
        Xds.
        Acta Crystallogr. D Biol. Crystallogr. 2010; 66: 125-132
        • Evans P.
        Scaling and assessment of data quality.
        Acta Crystallogr. D Biol. Crystallogr. 2006; 62: 72-82
        • Winn M.D.
        • Ballard C.C.
        • Cowtan K.D.
        • Dodson E.J.
        • Emsley P.
        • Evans P.R.
        • Keegan R.M.
        • Krissinel E.B.
        • Leslie A.G.
        • McCoy A.
        • McNicholas S.J.
        • Murshudov G.N.
        • Pannu N.S.
        • Potterton E.A.
        • Powell H.R.
        • Read R.J.
        • Vagin A.
        • Wilson K.S.
        Overview of the CCP4 suite and current developments.
        Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 235-242
        • Adams P.D.
        • Afonine P.V.
        • Bunkóczi G.
        • Chen V.B.
        • Davis I.W.
        • Echols N.
        • Headd J.J.
        • Hung L.W.
        • Kapral G.J.
        • Grosse-Kunstleve R.W.
        • McCoy A.J.
        • Moriarty N.W.
        • Oeffner R.
        • Read R.J.
        • Richardson D.C.
        • Richardson J.S.
        • Terwilliger T.C.
        • Zwart P.H.
        PHENIX: a comprehensive Python-based system for macromolecular structure solution.
        Acta Crystallogr. D Biol. Crystallogr. 2010; 66: 213-221
        • Emsley P.
        • Lohkamp B.
        • Scott W.G.
        • Cowtan K.
        Features and development of COOT.
        Acta Crystallogr. D Biol. Crystallogr. 2010; 66: 486-501
        • Murshudov G.N.
        • Skubák P.
        • Lebedev A.A.
        • Pannu N.S.
        • Steiner R.A.
        • Nicholls R.A.
        • Winn M.D.
        • Long F.
        • Vagin A.A.
        REFMAC5 for the refinement of macromolecular crystal structures.
        Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 355-367
        • Davis I.W.
        • Leaver-Fay A.
        • Chen V.B.
        • Block J.N.
        • Kapral G.J.
        • Wang X.
        • Murray L.W.
        • Arendall 3rd, W.B.
        • Snoeyink J.
        • Richardson J.S.
        • Richardson D.C.
        MolProbity: all-atom contacts and structure validation for proteins and nucleic acids.
        Nucleic Acids Res. 2007; 35: W375-W383
        • Bergfors T.
        Seeds to crystals.
        J. Struct. Biol. 2003; 142: 66-76
        • Long F.
        • Vagin A.A.
        • Young P.
        • Murshudov G.N.
        BALBES: a molecular-replacement pipeline.
        Acta Crystallogr. D Biol. Crystallogr. 2008; 64: 125-132
        • Lebedev A.A.
        • Young P.
        • Isupov M.N.
        • Moroz O.V.
        • Vagin A.A.
        • Murshudov G.N.
        JLigand: a graphical tool for the CCP4 template-restraint library.
        Acta Crystallogr. D Biol. Crystallogr. 2012; 68: 431-440
        • Pruitt K.D.
        • Tatusova T.
        • Klimke W.
        • Maglott D.R.
        NCBI Reference Sequences: current status, policy and new initiatives.
        Nucleic Acids Res. 2009; 37: D32-D36
        • Altschul S.F.
        • Madden T.L.
        • Schäffer A.A.
        • Zhang J.
        • Zhang Z.
        • Miller W.
        • Lipman D.J.
        Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.
        Nucleic Acids Res. 1997; 25: 3389-3402
        • Edgar R.C.
        MUSCLE: multiple sequence alignment with high accuracy and high throughput.
        Nucleic Acids Res. 2004; 32: 1792-1797
        • Felsenstein J.
        PHYLIP-Phylogeny interference package (version 3.2).
        Cladistics. 1989; 5: 164-166
        • Paradis E.
        • Claude J.
        • Strimmer K.
        APE: analyses of phylogenetics and evolution in R language.
        Bioinformatics. 2004; 20: 289-290
        • Crooks G.E.
        • Hon G.
        • Chandonia J.M.
        • Brenner S.E.
        WebLogo: a sequence logo generator.
        Genome Res. 2004; 14: 1188-1190
        • Jones D.T.
        • Buchan D.W.
        • Cozzetto D.
        • Pontil M.
        PSICOV: precise structural contact prediction using sparse inverse covariance estimation on large multiple sequence alignments.
        Bioinformatics. 2012; 28: 184-190
        • Bond C.S.
        • Clements P.R.
        • Ashby S.J.
        • Collyer C.A.
        • Harrop S.J.
        • Hopwood J.J.
        • M J.G.
        Structure of a human lysosomal sulphatase.
        Structure. 1997; 5: 227-289
        • Krogh A.
        • Larsson B.
        • von Heijne G.
        • Sonnhammer E.L.
        Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes.
        J. Mol. Biol. 2001; 305: 567-580
        • Simon R.
        • Priefer U.
        • Pühler A.
        A broad host range mobilization system for in vitro genetic engineering: transposon mutagenesis in Gram negative bacteria.
        Nat. Biotechnol. 1983; 1: 784-791
        • Gründling A.
        • Burrack L.S.
        • Bouwer H.G.
        • Higgins D.E.
        Listeria monocytogenes regulates flagellar motility gene expression through MogR, a transcriptional repressor required for virulence.
        Proc. Natl. Acad. Sci. U.S.A. 2004; 101: 12318-12323
        • Bishop D.K.
        • Hinrichs D.J.
        Adoptive transfer of immunity to Listeria monocytogenes. The influence of in vitro stimulation on lymphocyte subset requirements.
        J. Immunol. 1987; 139: 2005-2009
        • Karplus P.A.
        • Diederichs K.
        Linking crystallographic model and data quality.
        Science. 2012; 336: 1030-1033