Structural and biochemical analyses of selectivity determinants in chimeric Streptococcus Class A sortase enzymes

Abstract

Sequence variation in related proteins is an important characteristic that modulates activity and selectivity. An example of a protein family with a large degree of sequence variation is that of bacterial sortases, which are cysteine transpeptidases on the surface of gram-positive bacteria. Class A sortases are responsible for attachment of diverse proteins to the cell wall to facilitate environmental adaption and interaction. These enzymes are also used in protein engineering applications for sortase-mediated ligations (SML) or sortagging of protein targets. We previously investigated SrtA from Streptococcus pneumoniae, identifying a number of putative β7-β8 loop-mediated interactions that affected in vitro enzyme function. We identified residues that contributed to the ability of S. pneumoniae SrtA to recognize several amino acids at the P1' position of the substrate motif, underlined in LPXTG, in contrast to the strict P1' Gly recognition of SrtA from Staphylococcus aureus. However, motivated by the lack of a structural model for the active, monomeric form of S. pneumoniae SrtA, here, we expanded our studies to other Streptococcus SrtA proteins. We solved the first monomeric structure of S. agalactiae SrtA which includes the C-terminus, and three others of β7-β8 loop chimeras from S. pyogenes and S. agalactiae SrtA. These structures and accompanying biochemical data support our previously identified β7-β8 loop-mediated interactions and provide additional insight into their role in Class A sortase substrate selectivity. A greater understanding of individual SrtA sequence and structural determinants of target selectivity may also facilitate the design or discovery of improved sortagging tools.

Keywords: enzymes; protein biochemistry; protein engineering; sortases; structural biology; target selectivity.

FIGURE 1

Differences between Staphylococcus and Streptococcus SrtA proteins. The SrtA proteins are in cartoon representation, with the conserved eight‐stranded antiparallel β‐sheet that defines the sortase fold colored as labeled. S. aureus SrtA (saSrtA, PDB ID 2KID) is bound to the peptidomimetic, LPAT*, in black sticks and colored by heteroatom (O = red, N = blue, S = yellow), (a). The arrows indicate differences with S. pyogenes SrtA (spySrtA, 3FN5) (b). Specifically, Ca²⁺ is required for saSrtA activity, the β7–β8 loop contains an additional five residues and contains a Trp (W194) that dramatically affects activity, and the spySrtA protein is 24 residues longer than saSrtA

FIGURE 2

Biochemical characteristics of SrtA enzymes from S. agalactiae and S. pyogenes. (a) Sequence alignment of the extracellular regions of the S. pneumoniae SrtA (spSrtA), S. agalactiae SrtA (sagSrtA) and S. pyogenes SrtA (spySrtA) proteins. Sequences were aligned using T‐coffee and visualized with Boxshade. The β7–β8 loop residues are indicated with a red box. (b) Comparison of substrate selectivity for wild‐type spSrtA, spySrtA, sagSrtA₂₃₈ (the crystallized construct reported in PDB ID 3RCC), and spySrtA₂₄₇. Substrate cleavage was monitored via an increase in fluorescence at 420 nm from reactions of the fluorophore‐quencher probes Abz‐LPATGG‐K(Dnp), Abz‐LPATAG‐K(Dnp), and Abz‐LPATSG‐K(Dnp) (represented as LPATG, LPATA, and LPATS) in the presence of excess hydroxylamine. Bar graphs represent mean normalized fluorescence (± standard deviation) from at least three independent experiments at the 2 h reaction timepoint, as compared to saSrtA and the peptide LPATG. The spSrtA data was previously published and is shown for comparison. Averaged assay values and standard deviations for spySrtA, sagSrtA₂₃₈, and sagSrtA₂₄₇ are in Table S1

FIGURE 3

Enzyme assays of β7–β8 loop chimeras of spySrtA, sagSrtA, and spSrtA. Comparison of substrate selectivity for β7–β8 loop chimeras of (a) spySrtA, (b) sagSrtA, and (c) spSrtA proteins. Assays were run and data was collected as in Figure 2. Averaged assay values and standard deviations for spySrtA and sagSrtA variants are in Table S1. The spSrtA data was previously published and is shown for comparison

FIGURE 4

Structural characteristics of sagSrtA₂₄₇. (a) SagSrtA₂₄₇ adopts the conserved sortase fold. The protein is in gray cartoon, with β‐strands numbered and colored as labeled. (b) C‐terminal residues in sagSrtA₂₄₇ that were in the construct previously crystallized, but were not resolved in the structure, make intra‐protein hydrophobic interactions. Unresolved residues from PDB ID 3RCC are in yellow, with other sagSrtA₂₄₇ residues in gray. Residues involved in the interaction are labeled and their side chains are shown as sticks. The electron density in this region is also shown, the 2F _o − F _c map is rendered at 1σ. (c) C‐terminal residues not included in the previously crystallized construct make several interactions in sagSrtA₂₄₇. All distances are labeled and residues involved in contacts (including main chain of residues 238–244) are shown as sticks and colored by heteroatom (C = yellow, O = red, N = blue). Side chain atoms are shown when involved in the interaction, otherwise they are omitted for visual clarity. (d) The hydrophobic interaction involving I245 is shown, all residues are colored and the electron density is as in (b)

FIGURE 5

Structural characteristics of sagSrtA and spySrtA β7–β8 loop chimeras. In all, relevant residues are colored by structure as labeled and when represented, side chains are shown as sticks and colored by heteroatom (O = red, N = blue). Other residues are shown as gray cartoon. (a) The intraloop hydrogen bond is conserved between spySrtA (PDB ID 3FN5) and spySrtA_faecalis. (b) The β7–β8⁺³ Pro in lmSrtA (5HU4) occupies the same position as the β4–β5⁺³ Phe of spySrtA_{monocytogenes}, suggesting why the β7–β8 loop in spySrtA_{monocytogenes} is not well ordered and this variant is not as active. Furthermore, the intraloop hydrogen bond is not maintained in spySrtA_{monocytogenes}, as compared to lmSrtA. The black arrows indicate the residues involved in this interaction in lmSrtA. (c) The β7–β8⁺³ Pro in spySrtA_{monocytogenes} is translated up, as compared to that of lmSrtA in (b). The black arrows show the relevant residues for the intraloop hydrogen bond that is not conserved, as compared to (b). (d–f) Comparison of the ΔN188 sagSrtA_aureus β7–β8 loop with sagSrtA₂₄₇ (d) and saSrtA (2KID, e–f). In (f), the LPAT* peptidomimetic is shown as black sticks and colored by heteroatom. The different positions of W194 are labeled

FIGURE 6

Residues in the β7–β8 loop of Streptococcus proteins that regulate enzyme function. (a) Although the peptide‐binding pockets of sagSrtA and spySrtA are well conserved, there are four non‐conservative amino acid differences, as labeled. Both proteins are shown in cartoon representation, with side chains as sticks and colored by heteroatom. The β4–β5, β6–β7, and β7–β8 loops are also labeled. (b) The side chain of β7–β8⁻¹ E213 interacts with that of β6⁻² K183 in sagSrtA₂₄₇. This is an interaction that negatively affects enzyme activity for spSrtA, as previously reported, and which spySrtA does not share. (c) Mutation of the β7–β8⁺³ residue in sagSrtA from Pro to Ile, as in spySrtA, increases activity. However, the converse mutation in spySrtA, from Ile to Pro, has little to no effect. Assays were run and data was collected as in Figures 2 and 3. Averaged assay values and standard deviations are in Table S1. (d) WebLogo analysis of 37 Streptococcus SrtA proteins from the UniProt database. All sequences are in Table S2. (e) The absolutely conserved β7–β8⁺⁵ Ala in the sequences in (d) is stereochemically located in the same position as the W194 residue in saSrtA (2KID). A hydrogen bond between the A211 carbonyl and guanidinium group of the catalytic R214 is labeled