Biophysical Mechanism of Protein Export by Bacterial Type III Secretion System

Abstract

Type III secretion system (T3SS) is a protein injection nano-machine consisting of syringe and needle-like structure spanning bacterial inner and outer membranes. Bacteria insert the tip of T3SS needle to host cell membranes, and deliver effector proteins directly into host cells via T3SS to prime the host cell environment for infection. Thus inhibition of T3SS would be a potent strategy for suppressing bacterial infection. We previously demonstrated that T3SS needle rotates by proton-motive force (PMF) with the same mechanism as two evolutionally related rotary protein motors, flagellum and ATP synthase (FASEB J., 27, 2013, Ohgita et al.). Inhibition of needle rotation resulted in suppression of effector secretion, indicating the requirement of needle rotation for effector export. Simulation analysis of protein export by the T3SS needle suggests the importance of a hydrophobic helical groove formed by the conserved aromatic residue in the needle components. Based on these results, we have proposed a novel model of protein export by the T3SS needle, in which effector proteins are exported by PMF-dependent needle rotation oppositely to the hydrophobic helical groove in the needle. Quantitative examinations of the correlation between the speeds of T3SS rotation and the amount of effector export support this model. In this review, we summarize our current understanding of T3SS, and discuss our novel model of the protein export mechanism of T3SS based on the needle rotation.

1. Type III Secretion System as the Target for Novel Antibiotics

Type III secretion system (T3SS) is a protein export nano-machine that spans the bacterial inner and outer membranes consisting of a syringe and needle-like structure. The system comprises a sorting platform and an export gate in cytoplasm, a basal body across two membranes, and a needle protruding from the cell surface.^1–4) Pathogenic bacteria insert the T3SS needle tip into host membranes and inject effector proteins into host cytosol. In this process, effector proteins are first loaded to a sorting platform by cytosolic chaperones (e.g. Pseudomonas low calcium response locus protein H; PcrH), and are unfolded by ATPases in the platform (e.g. Pseudomonas secretion component protein N; PscN).^5,6) The unfolded effector proteins are then passed to an export gate and transported into host cytosol through the needle (Fig. 1). Various effector proteins are delivered by T3SS, although distinct secretion signals are not found.^7,8) The injected effector proteins prime the host cell environment for bacterial infection by modulating the function of host regulatory molecules (e.g. small guanosine 5′-triphosphate (GTP)-binding proteins, mitogen-activated protein kinases, and phosphoinositides).^9,10) Therefore T3SS is a potent target for novel antibiotics that suppress bacterial infection without killing bacteria. Previous studies of T3SS have mainly focused on the genetic regulation of effector genes,^2,11) the loading mechanism of effector proteins into T3SS by cytosolic chaperones,^12,13) and functions of each effector protein.^9,10) In contrast, little is known about the export mechanism of effector proteins by T3SS. Uncovering this mechanism will facilitate the development of T3SS inhibitors that suppress the secretion of all effector proteins at once.

Fig. 1. Process of Effector Export by T3SS

Bacterial effector proteins are directly exported into host cells via T3SS. This process comprises 4 steps: 1) expression of effector genes in bacterium; 2) loading of effector proteins into sorting platform of T3SS; 3) export of loaded effector proteins in T3SS needle; and 4) change of host cell environment by injected effector proteins.

2. Simulation Experiment on Export Mechanism of Effector Proteins by T3SS Needle

In 2010, Rathinavelan et al.¹⁴⁾ reported an interesting simulation analysis about the motion of effector proteins during transport by the T3SS needle. They built a T3SS needle model based on the crystal structure of the needle subunit protein MxiH in Shigella flexneri, and analyzed the motion of effector proteins in the needle by steered molecular dynamics simulations. These experiments suggested that Trp10 residue in MxiH forms a helical groove inside the needle channel, and the effector proteins are exported helically along the groove. This aromatic residue is highly conserved in the needle components, and its mutation to hydrophilic residue remarkably reduces the type III secretion activity. These results indicated the importance of the hydrophobic helical groove inside the T3SS needle as a rail for effector export. In this simulation, an external force for pulling effector proteins to the outside of the needle was artificially added. However, in nature, how is such force generated?

3. Similarities of T3SS with Bacterial Flagellum and ATP Synthase

In 1998, Kubori et al.¹⁵⁾ first succeeded in observing the structure of T3SS in Salmonella typhimurium by electron microscopy. The obtained structure was highly similar to that in bacterial flagellum except for the exterior region, in which T3SS has thin and short needle (approximately 100 nm in length and 10 nm in width) whereas flagellum has thick and long filament (µm-ordered length and approximately 20 nm in width).¹⁶⁾ High sequence homology of some components indicates the evolutional relationship of these two machineries.^17–19) Recently, structural similarities of cytosolic components of T3SS with F₀F₁-ATP synthase have also been indicated: the ATPase unit FliI/SctN, the central stalk FliJ/SctO, and the negative regulator FliH/SctL of T3SS resemble the α/β subunits, the γ subunit, and the b/δ subunits of the ATP synthase, respectively.^20–22) Flagellum rotates its filament with proton-motive force (PMF) derived from the concentration gradient of protons across the inner membrane.²³⁾ The γ subunit of F₀F₁-ATP synthase also rotates depending on PMF generated by mitochondrial electron transport system, and the rotation promotes ATP synthesis on α/β subunits.^24,25) Since T3SS is known to require PMF for effector secretion,²⁶⁾ we hypothesized that effector proteins are exported by the T3SS needle similarly to PMF-dependent needle rotation.

4. PMF-Dependent Rotation of T3SS Needle

Whether T3SS rotates was debatable. Büttner²⁷⁾ denied the possibility of T3SS rotation because T3SS does not have a stator unit, whereas flagellum has the corresponding MotAB. To examine the possibility of T3SS rotation, we attempted to observe the motion of T3SS needle. According to the system to observe the motion of biological nanomachines,^25,28,29) we constructed an observation system for T3SS needle motion as shown in Fig. 2A.³⁰⁾ In this system, PcrV, the tip component of T3SS of Pseudomonas aeruginosa PAO1 strain, was genetically modified with Strep tag II peptide, and streptavidin-coated fluorescent microbead was specifically attached to the tip via Strep tag II-streptavidin interaction. As a result, the motion of T3SS needle was observable as that of the attached microbead (Fig. 2B). When PMF was dissipated by the protonophore, carbonyl cyanide m-chlorophenylhydrazone, the rotation was largely inhibited, indicating the requirement of PMF for T3SS rotation. In addition, when the rotation was inhibited by viscous polymers (e.g. polyethylene glycol 8000 or alginate), the effector secretion was also suppressed.^30,31) These results strongly indicate a correlation between PMF-dependent T3SS rotation and effector transport by T3SS.

Fig. 2. A) Observation System for T3SS Needle Rotation

Fluorescent microbead was specifically attached to the tip of T3SS needle via strep tagII-streptavidin interaction to detect the motion of the needle. B) Motion of the bead attached to the needle under the condition of effector secretion. To induce effector secretion, rotation of T3SS needle was observed under conditions of Ca²⁺ depletion. Each frame represents every 1 s. Scale bar represents 1 µm.

5. Model of Effector Export Based on T3SS Needle Rotation

Based on the importance of hydrophobic helical groove inside the needle and the correlation between the PMF-dependent needle rotation and the exported amount of effector proteins by T3SS, we have proposed a novel model of effector protein export by T3SS (Fig. 3A). In this model, the helical groove in the needle functions as a rail for one-directional export of effector proteins set to the rail when T3SS needle rotates oppositely to the helical rail.

Fig. 3. A) Model of Effector Export by T3SS Based on Needle Rotation

In this model, effector is first set to the hydrophobic helical groove in the needle, and exported by PMF-dependent needle rotation oppositely to the groove. B) Time–course of effector protein export (ExoS, ExoT, and ExoY) by competitive ELISA. Almost same kinetics of export of these effector proteins suggest that primary structures are not important for the export of effector proteins by T3SS. The estimated rotation speed (approximately 3.2 rpm) was in good agreement with the actual observation (approximately 3 rpm) shown in Fig. 2B.

To verify this model, we analyzed correlations between the kinetics of effector export and T3SS rotation. The speed of effector export by T3SS was estimated using genetically modified effector proteins with an epitope tag 3FLAG to quantify the tag in supernatant by a competitive enzyme-linked immunosorbent assay (ELISA) with FLAG-tagged bacterial alkaline phosphatase. Almost same kinetics of export of three PAO1 effector proteins, ExoS, ExoT, and ExoY were observed as 1.20, 1.13, and 0.935 nM of amino acid min⁻¹, respectively, suggesting that primary structure of effector proteins is not important for the export speed by T3SS (Fig. 3B). Since these measurements were performed at OD₆₀₀ of 0.3 (corresponding to a bacterial density of 3 × 10⁸ colony-forming units mL⁻¹),³²⁾ the average speed of effector export in T3SS was estimated 2184 amino acids min⁻¹ under the assumption that each bacterium has only one T3SS. In addition, we assumed that the length of the needle is 345 Å and the number of turns in helical groove is 15 based on the needle model previously reported.¹⁴⁾ These considerations indicate that the feed of effector protein in the needle would be 23 Ǻ per revolution.

According to the calculations mentioned above, the export speed of ExoS (453 amino acids) was estimated approximately 5 molecules min⁻¹, and the linearized length of ExoS calculated as 1586 Ǻ using the average length of amino acid of 3.5 Å. As a result, the rotation speed of T3SS needle was calculated approximately 420 rpm from the observed secretion speed of ExoS. Finally, the speed of microbead was calculated approximately 3.2 rpm under the assumption that the rotation speed is inversely proportional to the rotation radius (75 Ǻ of needle width versus 1 µm of bead width). This estimated rotation speed of microbead is in good agreement with the actual observation (approximately 3 rpm),³⁰⁾ strongly supporting our model of effector export shown in Fig. 3A.

6. Conclusion

Using T3SS, pathogenic bacteria export effector proteins for efficient infection into host cells across the membrane barriers. Although recent studies are uncovering the molecular mechanism of T3SS, there are still many unanswered questions. One such question is how effector proteins are exported by T3SS. Recently, we have proposed a novel model of export of effector proteins by T3SS based on our observation of T3SS needle rotation. We demonstrated good correlation between the needle rotation and the exported amount of effector proteins, strongly supporting the effector export mechanism by T3SS needle rotation. Our results provide new insights into the development of novel inhibitors of bacterial infection to target T3SS.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Numbers JP14536616 and JP16763638 (T.O.). T.O. is indebted to Drs. Kentaro Kogure, Naomasa Gotoh, and Naoki Hayashi for valuable advice.

Conflict of Interest

The authors declare no conflict of interests.

References

• 1) Cornelis G. R., Nat. Rev. Microbiol., 4, 811–825 (2006).
• 2) Hauser A. R., Nat. Rev. Microbiol., 7, 654–665 (2009).
• 3) Portaliou A. G., Tsolis K. C., Loos M. S., Zorzini V., Economou A., Trends Biochem. Sci., 41, 175–189 (2016).
• 4) Deng W., Marshall N. C., Rowland J. L., McCoy J. M., Worrall L. J., Santos A. S., Strynadka N. C. J., Finlay B. B., Nat. Rev. Microbiol., 15, 323–337 (2017).
• 5) Izoré T., Job V., Dessen A., Structure, 19, 603–612 (2011).
• 6) Notti R. Q., Stebbins C. E., Microbiol. Spectr., 4, VMBF-0004-2015 (2016).
• 7) Löwer M., Schneider G., PLoS ONE, 4, e5917 (2009).
• 8) da Cunha M., Milho C., Almeida F., Pais S. V., Borges V., Maurício R., Borrego M. J., Gomes J. P., Mota L. J., BMC Microbiol., 14, 40 (2014).
• 9) Engel J., Balachandran P., Curr. Opin. Microbiol., 12, 61–66 (2009).
• 10) Ogawa M., Handa Y., Ashida H., Suzuki M., Sasakawa C., Nat. Rev. Microbiol., 6, 11–16 (2008).
• 11) Yahr T. L., Wolfgang M. C., Mol. Microbiol., 62, 631–640 (2006).
• 12) Lara-Tejero M., Kato J., Wagner S., Liu X., Galán J. E., Science, 331, 1188–1191 (2011).
• 13) Deane J. E., Abrusci P., Johnson S., Lea S. M., Cell. Mol. Life Sci., 67, 1065–1075 (2010).
• 14) Rathinavelan T., Zhang L., Picking W. L., Weis D. D., De Guzman R. N., Im W., Biophys. J., 98, 452–461 (2010).
• 15) Kubori T., Matsushima Y., Nakamura D., Uralil J., Lara-Tejero M., Sukhan A., Galán J. E., Aizawa S. I., Science, 280, 602–605 (1998).
• 16) Aizawa S. I., Dean G. E., Jones C. J., Macnab R. M., Yamaguchi S., J. Bacteriol., 161, 836–849 (1985).
• 17) Gophna U., Ron E. Z., Graur D., Gene, 312, 151–163 (2003).
• 18) Erhardt M., Namba K., Hughes K. T., Cold Spring Harb. Perspect. Biol., 2, a000299 (2010).
• 19) Bergeron J. R., Peer J., 4, e1718 (2016).
• 20) Imada K., Minamino T., Tahara A., Namba K., Proc. Natl. Acad. Sci. U.S.A., 104, 485–490 (2007).
• 21) Diepold A., Armitage J. P., Philos. Trans. R. Soc. Lond. B Biol. Sci., 370, 20150020 (2015).
• 22) Imada K., Minamino T., Uchida Y., Kinoshita M., Namba K., Proc. Natl. Acad. Sci. U.S.A., 113, 3633–3638 (2016).
• 23) Gabel C. V., Berg H. C., Proc. Natl. Acad. Sci. U.S.A., 100, 8748–8751 (2003).
• 24) Noji H., Ueno H., McMillan D. G. G., Biophys. Rev., 9, 103–118 (2017).
• 25) Noji H., Yasuda R., Yoshida M., Kinosita K. Jr., Nature (London), 386, 299–302 (1997).
• 26) Wilharm G., Lehmann V., Krauss K., Lehnert B., Richter S., Ruckdeschel K., Heesemann J., Trülzsch K., Infect. Immun., 72, 4004–4009 (2004).
• 27) Büttner D., Microbiol. Mol. Biol. Rev., 76, 262–310 (2012).
• 28) Silverman M., Simon M., Nature (London), 249, 73–74 (1974).
• 29) Ryu W. S., Berry R. M., Berg H. C., Nature (London), 403, 444–447 (2000).
• 30) Ohgita T., Hayashi N., Hama S., Tsuchiya H., Gotoh N., Kogure K., FASEB J., 27, 2862–2872 (2013).
• 31) Ohgita T., Hayashi N., Gotoh N., Kogure K., Open Biol., 3, 130133 (2013).
• 32) Li M., Holmes V., Zhou Y., Ni H., Sanzari J. K., Kennedy A. R., Weissman D., PLOS ONE, 9, e85665 (2014).