21 20 D I A M O N D L I G H T S O U R C E A N N U A L R E V I E W 2 0 2 1 / 2 2 D I A M O N D L I G H T S O U R C E A N N U A L R E V I E W 2 0 2 1 / 2 2 Macromolecular Crystallography Group Beamlines I03 (and Beamline B21 fromthe Soft CondensedMatter Group) Protein Periscopes regulate bacterial interactions Related publication title: Whelan F., Lafita A., Gilburt J., Dégut C., Griffiths S. C., Jenkins H. T., St John A. N., Paci E., Moir J.W. B., PlevinM. J., Baumann C. G., Bateman A., & Potts J. R. Periscope Proteins are variable-length regulators of bacterial cell surface interactions. Proc Natl Acad Sci 118, (2021). DOI: 10.1073/pnas.2101349118 Publication keywords : Protein structure; SHIRT; Cell adhesion; Bacteria; Grampositive; Streptococcus ; Immune evasion B acteria can adapt to changing environments by altering their cell surface. Inhumanpathogens, this variability has been implicated in immune evasion. Researchers sought to understand one potential mechanism (length variation of proteins on the bacterial surface) through studies of repetitive bacterial proteins. Sgo0707 is a repetitive protein found on the surface of the bacteria Streptococcus gordonii. The teamused X-ray diffraction on Diamond Light Source’s Macromolecular Crystallography (MX) beamline (I03) to study the atomic structure of repeats from Sgo0707. They also investigated the shape of longer regions of Sgo0707 in solution, approximating the structure of the protein on the surface of bacteria, using small-angle X-ray scatteringmethods on the High Throughput SAXS beamline (B21). High-resolutionstructures of Sgo0707 repeats revealedanovel structure the researchersnamed‘SHIRT’. Studies suggested that SHIRT repeats adopt a rod-like conformation that would project the functional domain of Sgo0707 away from the bacterial cell surface. The researchers identified that Sgo0707 andmany related proteins exhibit apparent repeat number variability and called these ‘Periscope Proteins’ to reflect their rod-like shape and role in projecting a functional domain away from the bacterial cell surface. Their identification of this large class of Periscope Proteins suggests thatmany bacteria can use protein length variation as away of adapting their surface to a changing environment. Such understanding is important in a range of settings, such as studies of bacterial colonisation of tissues and surfaces, including implantedmedical devices. Mechanisms of dynamic bacterial surface variation are a key component of bacterial survival and adaptation in a range of challenging and complex environments 1 . Such mechanisms include changes in capsular polysaccharide or changes in expression of protein adhesins. One mechanism, to date under- recognised, is length variation of repetitive bacterial cell surface proteins. Gram-positive bacteria, such as staphylococcal and streptococcal species, produce cell surface-anchored repetitive proteins such as SasG and Rib and previous studies suggest variation in the number of repeats 2 . In addition to sequence repeats, such proteins present an N-terminal domain, predicted to be involved in host colonisation, to the extracellular environment (Fig. 1). We previously characterised repeats from SasG 3 and Rib 4 and showed that they form stable folded domains that, when arrayed in tandem, form highly elongated structures. In this study we sought to characterise the repetitive region of a third bacterial surface protein, Sgo0707 from Streptococcus gordonii . With 82-100% identity between adjacent domains, the identification of the structured boundaries of repeat domains is a difficult task, especially with limited PDB homologues for guidance; biophysical characterisation of an initially predicted folded repeat (‘ΔN-Sgo_R2’) indicated awell-folded domain. However, when we solved the structure at a resolution of 0.95 Å via the use of X-ray crystallography and a b initio molecular replacement (MR) using an idealised β-strand, this construct appeared to be N-terminally truncated (Fig. 2; PDB 7AVJ). It was clear that N-terminal extension and C-terminal truncation would complete the domain fold and we were able to solve the structure of a completed domain at a resolution of 0.82 Å (Fig. 2; PDB 7AVK). This Sgo0707 domain contained no definition in the Pfam database and formed a novel fold, and thus we named the domain ‘SHIRT’ (Streptococcal High Identity Repeats in Tandem; Pfam entry PF18655). The SHIRT domain contains a mixed α/β fold, comprising seven β-strands constituting two main β-sheets, with a single short intra-domain α-helix. Using our newly-identified SHIRT domain definitions, we sought to structurally characterise a tandem SHIRT domain construct (Sgo_R3-R4); the structure revealed a short (Pro-Ala-Pro) linker between adjacent domains, as well as a very limited inter-domain interface (Fig. 2; PDB 7AVH). This arrangement was validated using thermal unfolding experiments, which showed that Sgo_R3-R4 was no more thermally stable than the single domain in isolation (Fig. 2, T m = 76-77 ℃ ). Notably, the extended nature of the tandem SHIRT structure with a short linker that would be predicted to retain rigidity and thus limit domain- domain flexibility, suggested that addition of extra repeats would result in an elongated solution conformation. To test this hypothesis, we collected Small- Angle X-Ray Scattering (SAXS) data for both the tandem and 7-repeat domain (Sgo_R2-R8) SHIRT architectures on beamline B21. This SAXS solution data revealed that for both constructs an overall rod-like solution conformation is maintained, with a conserved cross-sectional radius (Fig. 2) consistent with the tandem repeat crystal structure. DNA sequence analyses suggested that Sgo0707, Rib and SasG (and many other bacterial surface proteins) are produced with varying numbers of repeats. If the repeat region forms a rod, as in our examples of Sgo0707, Rib and SasG, this would result in the functional ( e.g. host colonisation) domain being projected differing distances from the bacterial cell surface. Using a memorable analogy, we called this diverse class of cell surface-anchored bacterial proteins ‘Periscope Proteins’ 2 . We propose that selection pressure, combinedwith stochastic repeat number variation enabled by high DNA repeat identity, results in enrichment of bacteria expressing short or long versions of the protein (dependent on the nature of the selection pressure) 2 . Enabled by access to Diamond Light Source, we have shown that length variation in Periscope Proteins appears to be a widespread mechanism through which pathogenic bacteria can modulate interactions with their host survival niche. Such modulated interactions could play an important role in processes during infection including host colonisation, biofilm formation and immune system evasion. References: 1. van der Woude, M. W. et al. Phase and antigenic variation in bacteria. Clinical Microbiology Reviews 17 , 581–611 (2004). DOI: 10.1128/ CMR.17.3.581-611.2004 2. Whelan, F. et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions. Proceedings of the National Academy of Sciences 118, (2021). DOI: 10.1073/pnas.2101349118 3. Gruszka, D. T. et al. Cooperative folding of intrinsically disordered domains drives assembly of a strong elongated protein. Nature Communications 6 , 7271 (2015). DOI: 10.1038/ncomms8271 4. Whelan, F. et al. Defining the remarkable structural malleability of a bacterial surface protein Rib domain implicated in infection. Proceedings of the National Academy of Sciences 116 , 26540–26548 (2019). DOI: 10.1073/pnas.1911776116 Funding acknowledgement: JRP, SCG and FW were funded by the British Heart Foundation (FS/12/36/29588 and PG/17/19/32862). Corresponding authors: Dr Samuel Griffiths, Evotec (U.K.),
[email protected] Dr FionaWhelan, University of Adelaide,
[email protected] Repeat region A domain Host receptor 'Periscope protein' Host cell surface Bacterium Figure 1: Periscope Proteins are often C-terminally attached to the bacterial cell wall, have an N-terminal folded (A) domain involved in host colonisation and have a highly repetitive region. ΔN-Sgo_R2 T m 60 ℃ ‘SHIRT’ T m 76 ℃ Sgo_R3-R4 T m 77 ℃ Bacterial surface Host surface D max 107 Å R g c 6.4 Å D max 370 Å R g c 7.0 Å 7 SHIRT domains A Figure 2: Structures of N-terminally truncated ( ΔN-Sgo_R2; PDB 7AVJ), complete (SHIRT; PDB 7AVK) and tandemly arrayed (Sgo_R3-R4; PDB 7AVH) SHIRT domains with corresponding melting temperatures (T m ). A model of 7 arrayed SHIRT domains and structure- and SAXS-determined parameters for the two lengths of tandem array (2 and 7) are also shown. C-terminal attachment to the bacterial cell wall, and the unmodelled N-terminal SHIRT domain and A domain are indicated.