Structural characterization of the Imm52 family protein TsiT in Pseudomonas aeruginosa

Abstract

The bacterial type VI secretion system (T6SS) utilizes many toxic effectors to gain advantage over interbacterial competition and eukaryotic host infection. Meanwhile, the cognate immunity proteins of these effectors are employed to protect themselves from the virulence. TseT and TsiT form an effector‐immunity (E‐I) protein pair secreted by T6SS of Pseudomonas aeruginosa. TseT is toxic for other bacteria, whereas TsiT can suppress the virulence of TseT. Here, we report the crystal structure of TsiT at 1.6 Å resolution. TsiT is a typical α + β class protein and belongs to a novel Imm52 protein family of the polymorphic toxin system. Apart from TsiT, only one structure of the Imm52 family proteins is present in the Protein Data Bank (PDB), but that structure is not characterized and shares low sequence identity with TsiT. We characterized the basic features of TsiT structure and identified conserved residues of the Imm52 family proteins according to homology comparison. Our work provided structural information of a new protein family and should aid future functional studies.

Short abstract

PDB Code(s): 6JDP

Introduction

As a versatile weapon, the bacterial Type VI secretion system (T6SS) secretes many effector‐immunity (E‐I) protein pairs to gain advantage in competing with bacterial rivals, and invading their hosts.1, 2

Several E‐I pairs have been studied in the last few years, such as the Tse1‐3 family effectors, and their corresponding cognate immunity proteins Tsi1‐3.3, 4, 5, 6 Tse1‐3 and Tsi1‐3 proteins play important part in various aspects of interbacterial competition, whereas some “trans‐kingdom” effectors,7 such as Tle5 and Tle5B proteins in Pseudomonas aeruginosa, possess the ability of invading both prokaryotic and eukaryotic cells.7, 8

Recently, a new T6SS effector was identified and characterized in P. aeruginosa.9 The effector protein was called TseT for type VI effector TOX‐REase and displayed antibacterial activity.10 PA3908, the gene next to TseT's encoding gene PA3907, encodes a type VI immunity protein TsiT, which can suppress the virulence of TseT.9

Here, we present the crystal structure of P. aeruginosa TsiT protein at 1.6 Å resolution. TsiT is a typical α + β class protein with a central antiparallel β‐sheet surrounded by helices and loops. Our structure information, together with structural comparison with its homolog, provide novel insights into a new family of T6SS immunity proteins.

Results

Overall structure of TsiT

TsiT falls into the α + β class proteins, containing 9 α‐helices and an 8‐stranded antiparallel central β‐sheet. Overall structure and topology diagram of TsiT are shown in Figure 1(a,b). N‐terminal domain of TsiT consists of β1 which is located in the center of the whole structure, followed by a long loop with 5 α‐helices α1–α5. The antiparallel β‐strands form the middle layer of the structure. The C‐terminal domain is composed of a loop containing a long helix α9. The main forces that stable the whole structure are the well aligned hydrogen bonds between the antiparallel β‐strands. Besides, the hydrophobic forces among α2, α6, and central β‐sheet also greatly stabilize the overall structure.

Figure 1.

Structure bases of Pseudomonas aeruginosa TsiT. (A) Top and side view of TsiT in cartoon representation. The secondary‐structure elements referred to in the text are labeled. (B) Topology diagram of TsiT. β‐strands are shown as arrows and α‐helices are shown as columns. (C) Electrostatic potential surfaces of TsiT viewing from top and bottom.

The solvent‐accessible surface area of the TsiT structure is 12206 Ų. The electrostatic potential surfaces of TsiT are shown in Figure 1(c). The majority part of TsiT's surface is electronegative, especially for the side where two long helices α2 and α6 are present (Fig. 1(c), left). On the opposite side, several electro‐positive patches are found formed by the N‐terminal loop and the short turn between α5‐β2 and β3‐β4 (Fig. 1(c), right).

Comparison with homologs

The structural similarity search using DALI web server returned several proteins that share structural homology with TsiT.11 The structure shares the highest similarity with TsiT, that is, a hypothetical protein (BPSL2088) from Burkholderia pseudomallei K96243 (PDB: 4RHO chain B, Z‐score: 19.4). The structural alignment between the two proteins gave an root‐mean‐square deviation (RMSD) of 2.8 Å for 216 aligned residues (Fig. 2(b)). However, the sequence identity between them is only 24% and we did not manage to determine the TsiT structure by molecular replacement in the first place with BPSL2088 structure as the search model.

Figure 2.

Comparison of TsiT and its homologs. (A) Sequence alignment of TsiT (PA3908), BPSL2088 (4RHO), and other Imm52 family proteins. (B) Structural superposition of TsiT (cyan) and BPSL2088 (green). (C) Superposition of conserved residues between TsiT (cyan) and BPSL2088 (green).

According to the protein families (Pfam) database,12 TsiT and BPSL2088 both belong to the immunity protein 52 family (Imm52, PF15579), which is characterized with an α + β fold and conserved tryptophan and phenylalanine residues, and a GT motif.10, 12 The detailed sequence alignment of Imm52 family proteins were shown in Figure 2(a). We found that apart from the conserved residues mentioned earlier, additional residues such as tyrosine, histidine, and leucine were also conserved. After comparing the conserved region of PF15579 that covers 135–234 residues of each protein, we found that 90 out of 100 residues were aligned with an RMSD of 1.5 Å. This indicates that the conserved region of Imm52 family proteins are indeed similar structurally. A comparison of conserved residues between them is shown in Figure 2(c). All conserved residues between the two proteins aligned well except the H217 residue located at the rear of TsiT's α9, as α9 of TsiT was somehow bent whereas the corresponding helix in BPSL2088 remained straight.

BPSL2087, the gene immediately neighboring BPSL2088, also encodes a protein containing a Tox‐REase‐5 domain. Therefore, it is obvious that BPSL2087‐BPSL2088 is also an E‐I pair that belongs to the bacterial polymorphic toxin system.

Discussion

In 2012, by using comparative genomics, a group had identified a great number of toxins and antitoxins which were classified into the polymorphic toxin system,10 in which Tox‐REase‐5, Imm52, and many other protein families were included. However, very few of these protein families were closely investigated.

No structural information of the TseT homologs is available at present, and we could not get soluble TseT protein alone when expressing in Escherichia coli, even with glutathione s‐transferase (GST) or maltose binding protein (MBP) fusion tags. However, when expressing TseT and TsiT together using a Deut vector, we managed to acquire a large amount of soluble TseT‐TsiT complex even in 800 mM NaCl. According to size exclusion chromatography, the TseT‐TsiT complex forms a 1:1 heterodimer in solution (data now shown). This suggests that TsiT could strongly bind to TseT and block the way of its function. The isoelectric point of TseT is 8.45, which suggests that the surface of TseT is mainly electro‐positive and may interact closely with the electro‐negative surface of TsiT.

Our study has provided novel information on the structural basis of the Imm52 family proteins. However, much more needs to be done to uncover the detailed mechanisms of TseT's toxicity and how TsiT inhibits the virulence of its cognate effector.

Materials and Methods

Cloning and expression

The gene encoding TsiT was amplified from the P. aeruginosa PAO1 genomic DNA. The PCR product was cloned into the pET‐28at vector with an N‐terminal his‐tag followed by a Tobacco etch virus (TEV) cleavage site. The plasmid containing the target gene was transformed into E. coli BL21 (DE3) strain. Large cultures were induced at OD₆₀₀ = 0.6 with a final concentration of 0.4 mM isopropylb‐D‐1‐thiogalactopyranoside (IPTG) and left to express at 289 K for 20 h.

Protein purification and crystallization

The cells containing TsiT protein were pelleted and resuspended in buffer containing 25 mM HEPES pH 7.6, 500 mM NaCl, and 1 mM phenylmethylsulfonyl fluoride (PMSF). After sonication, the lysate was centrifuged at 14,000 g for 40 min. The supernatant containing the soluble protein was purified via affinity chromatography with nickel–nitrilotriacetic acid resin (Bio‐Rad, Hercules, CA, USA). The fractions were diluted with buffer containing 25 mM HEPES pH 7.6 and 50 mM NaCl and purified by ion exchange chromatography (Hitrap Q HP, GE Healthcare, Marlborough, MA, USA). The TsiT protein was further purified by gel‐filtration chromatography (Superdex 200, GE Healthcare, Marlborough, MA, USA) with buffer containing 25 mM HEPES pH 7.6 and 150 mM NaCl. The purified protein was collected and ultrafiltered to 18 mg/mL. The selenomethionine (SeMet)‐labeled TsiT protein was also expressed and purified using the same procedure as native TsiT.

Crystallization screening of TsiT was carried out at 293 K using the sitting‐drop vapor‐diffusion technique. The best crystals of Tli5 were obtained within 5 days under the condition of 1.26 M sodium phosphate monobasic monohydrate and 0.14 M potassium phosphate dibasic. The SeMet‐labeled TsiT crystals were grown under the same condition.

Data collection, structure determination, and refinement

Both native and SeMet‐labeled TsiT crystal data sets were collected at 100 K on the station BL18U of the Shanghai Synchrotron Radiation Facility (SSRF). All the data were processed with HKL‐3000.13 The crystal structure of TsiT was determined by the single wavelength anomalous dispersion method (SAD). The selenium atoms were located with the program Shelxd and then used to calculate the initial phases in Shelxe.14 The initial phases were used for automatic model building with the program RESOLVE.15 Coot16 and PHENIX.REFINE17 were used for manually model building and refinement, respectively. The qualities of the final model were checked with the program Mol‐Probity.18 Data collection and refinement statistics were given in Table 1. The program PyMOL (the PyMOL Molecular Graphics System, Version 1.2 Schrödinger, LLC., New York, NY, USA) was used to prepare structural figures.

Table 1.

Data Collection and Refinement Statistics for TsiT

Data collection	Native	SeMet
Wavelength (Å)	0.9793	0.9789
Space group	R 3	R 3
Unit‐cell parameters	a = b = 86.96, c = 81.108, α = β = 90°, γ = 120°	a = b = 87.387, c = 81.262, α = β = 90°, γ = 120°
Resolution (Å) a	1.60 (1.60–1.63)	2.40 (2.40–2.44)
Unique reflections	30124 (3014)	9046 (604)
Completeness (%)	100 (100)	98.6 (89.1)
Redundancy	4.3 (4.2)	8.1 (4.2)
Mean I/ ơ (I)	19.417 (1.222)	31.699 (1.836)
Molecules in asymmetric unit	1	1
R merge (%)	5.2 (75.2)	10.4 (42.8)
Refinement
Resolution range (Å)	1.60–21.7
R work/R free (%)	18.66/22.00
Average B factor (Å2)
Overall	34.64
Protein	34.17
Water	40.08
Other	45.08
Number of nonhydrogen atoms
Protein	1949
Water	157
Other	6
Ramachandran plot (%)
Most favored	97.49
Allowed	2.09
Outlier	0.42
r.m.s. deviations
Bond lengths (Å)	0.006
Bond angles (°)	0.93

^aValues in parenthesis for the highest resolution shell.

Sequence alignment

Amino acid sequence alignment was performed by Clustal Omega.19 The figure of sequence alignment was generated by ESpript.20

Electrostatic potential and surface area calculations

Electrostatic potentials were calculated using the PDB2PQR server with the AMBER force field and rendered using Pymol in conjunction with the adaptive poisson‐boltzmann solver (APBS) plugin.21, 22 The solvent‐accessible surface area was calculated using the PDBePISA webserver.23

PDB Accession Numbers

Atomic coordinates and structure factors of TsiT were deposited in the Protein Data Bank under the accession number 6JDP.

Conflict of Interest

The authors declare that they have no conflicts of interest with the contents of this article.