Structural study of Legionella pneumophila effector DotY (Lpg0294), a component of the Dot/Icm type IV secretion system

DotY is a component of the inner membrane-embedded coupling complex T4CC that delivers effectors to the type IV secretion system translocation apparatus. DotY is a highly mobile component of this subcomplex. The crystal structure of DotY on its own was determined and its fold and the connectivity of its secondary-structure elements were established.

Abstract

The bacterium Legionella pneumophila is a causative agent of Legionnaires’ disease. It utilizes the Dot/Icm type IV secretion system (T4SS) to deliver over 300 effector proteins into the host cell, leading to modification of cellular processes and creating a safe environment for bacterial proliferation. Dot/Icm is a multi-subunit molecular machine. The effectors are recognized by the inner membrane-embedded coupling complex (T4CC), which then delivers them to the translocation apparatus. This T4CC subcomplex is made up of DotL, DotM, DotN, IcmS, IcmW, LvgA, DotY and DotZ, and its structure was recently determined by cryo-EM. DotY is a highly mobile component of this subcomplex and its structure was only partially defined. DotY is a unique component of the T4SS that is only found in the Legionella genus. Here, the crystal structure of DotY on its own is presented and its fold and the connectivity of its secondary-structure elements are established. The protein is divided into three segments. The first and last segments form a four-helix bundle domain, while the middle segment forms an α/β domain that has a unique fold. The flexibility of the interdomain linkers allows the reorientation of the two domains between that observed in the crystal structure and that assumed within the T4CC subcomplex.

1. Introduction

Legionella pneumophila causes Legionnaires’ disease, which leads to a severe form of pneumonia. It is also an intracellular parasite of protozoa. Once inside the lungs, L. pneumophila colonizes macrophages by delivering a wide range of effector molecules to subvert the host functions for its own survival (Prashar & Terebiznik, 2015 ▸). The effectors are delivered to the host by a type IV secretion system (T4SS; Grishin et al., 2022 ▸; Kitao et al., 2022 ▸). The Legionella T4SS is denoted Dot/Icm [defect in organelle transport (Dot) or intracellular multiplication (Icm)] and belongs to group B of the type IV secretion systems. Legionella T4SS is a large macromolecular complex with ∼30 identified protein components, including adaptor proteins that are involved in effector recruitment. An inner membrane-embedded coupling complex (T4CC) is involved in effector recognition and delivers effectors to the T4SS translocation apparatus. This complex has previously been purified and contains eight proteins: DotL, DotM, DotN, IcmS, IcmW, LvgA, DotY and DotZ. Its structure has recently been determined by cryo-EM with an average resolution of 3.7 Å (Meir et al., 2020 ▸; PDB entry 6sz9). While seven proteins in this complex were well resolved, only the N-terminal segment (residues 5–74 of 230 residues) of DotY was visible, indicating flexibility of its C-terminal segment within the complex. Subsequently, cryo-EM data to a slightly better resolution (3.61 Å; PDB entry 7ovb; Macé et al., 2022 ▸) allowed nearly the entire length of DotY to be traced (residues 5–192). DotY binds to the abovementioned subcomplex only through interaction with DotZ. DotY is important for effector delivery to the host and modifies the dynamics of the effector-delivery complex IcmSW, which is bound to the C-terminus of DotL through a long flexible linker (Meir et al., 2020 ▸; Macé et al., 2022 ▸). A Legionella dotY deletion mutant (ΔdotY) showed a reduction in intracellular growth in the protozoon Acanth­amoeba castellanii when compared with the wild type (Macé et al., 2022 ▸). Growth was restored to the wild-type level when the ΔdotY strain was complemented with the wild-type dotY gene. Monitoring the secretion of several effectors (RalF, LegC8, Lem21, CegC3 and Lpg1663) showed that deletion of the dotY gene led to a significant reduction in translocation of these effectors into CHO cells.

DotY is a 230-residue protein. Here, we report the crystal structure of DotY(8–219) at 2.43 Å resolution. The connectivity of the secondary-structural elements differs from that in the cryo-EM model, the long linker between the N-terminal and middle domains is highly flexible and the C-terminal helix joins the three helices of the N-terminal domain, forming a four-helix bundle. Comparison with the cryo-EM structure suggests that while the location of the helices in the second domain is in general agreement with the crystal structure, the connectivity and the sequence assignment to the secondary structures are substantially different, presumably due to the difficulty of interpretation of the lower resolution EM map, compounded by significant flexibility of this domain within the T4CC subcomplex (Macé et al., 2022 ▸; Meir et al., 2020 ▸).

2. Materials and methods

2.1. DNA cloning

The dotY gene (lpg0294; UniProt Q5ZYR7; residues 8–219) from L. pneumophila strain Philadelphia 1 was cloned into vector pMCSG7, a derivative of vector pET-21a adapted for ligation-independent cloning (Eschenfeldt et al., 2009 ▸). This plasmid was then transformed into Escherichia coli BL21 (DE3) cells. The expressed protein contained a TEV-cleavable 6×His tag at the N-terminus. For expression of the selenomethionine (SeMet) derivative, the cell pellet from a 100 ml overnight culture grown in LB medium was inoculated into a litre of M9 minimal medium. After shaking at 37°C until the OD₆₀₀ reached 0.6, a mixture of l-amino acids (100 mg lysine, phenylalanine and threonine; 50 mg of isoleucine, leucine and valine) and 60 mg selenomethionine were added to the culture. After 15 min, protein expression was induced with 0.5 mM isopropyl β-d-1-thiogalactopyranoside. The induced culture was grown overnight at 18°C and the cells were harvested by centrifugation at 9000g for 15 min.

2.2. Protein purification

The cell pellet was resuspended in lysis buffer (50 mM Tris–HCl pH 8.0, 10% glycerol, 0.1% Triton X). The cells were lysed in a cell disruptor (Constant Systems, United Kingdom) and cell debris was then removed by centrifugation at 31 000g for 30 min. The resulting supernatant was applied onto a 3 ml Cobalt TALON metal-affinity resin (Takara, California, USA). The column was washed with five column volumes of standard buffer (20 mM Tris pH 8.0, 50 mM NaCl). The His-tagged protein was eluted using 200 mM imidazole in standard buffer. The His tag was removed by incubating the elution fractions with 0.03 mg ml⁻¹ TEV protease overnight at 4°C. The treated sample was then loaded onto an ENrich SEC 650 10 × 300 Column (Bio-Rad, California, USA) equilibrated with crystallization buffer (15 mM Tris–HCl pH 8.0, 100 mM NaCl). Fractions containing DotY were pooled and concentrated to 62 mg ml⁻¹ using a Millipore centrifugal filter with a molecular-weight cutoff of 10 000 Da in preparation for crystallization trials. The concentration was measured with a Nanodrop UV spectrophotometer (Themo Scientific) using an extinction coefficient of 19 940 for DotY as calculated by ProtParam (Gasteiger et al., 2005 ▸).

2.3. Protein crystallization, data collection and structure determination

Initial crystals were obtained by screening in a 96-well plate format using a Gryphon crystallization robot (Art Robbins Instruments, Sunnyvale, California, USA). The best crystals were obtained using the hanging-drop method by mixing 1 µl protein solution at 15 mg ml⁻¹ and 1 µl reservoir solution consisting of 0.1 M sodium citrate pH 3.5, 25% PEG 8K, 0.3 M sodium acetate. The drop was suspended over 0.5 ml reservoir solution. The cryosolution consisted of 70% reservoir solution and 30% ethylene glycol. Data were collected on the 08B1-1 beamline at the Canadian Light Source (Table 1 ▸). The data were processed and scaled with XDS (Kabsch, 2010 ▸) using the AutoProcess script (Fodje et al., 2014 ▸). The structure was solved by the single-wavelength anomalous dispersion (SAD) method using AutoSol in the Phenix program suite (Liebschner et al., 2019 ▸). The initial model was refined using Phenix and missing parts were manually rebuilt using Coot (Emsley et al., 2010 ▸). There are four molecules in the asymmetric unit. In the final model, molecules A and B contain residues 5–219, with residues 75–82 being less well defined, while molecules C and D contain residues 5–74 and 88–219; residues 75–87 were disordered and could not be modeled. Data-collection and refinement statistics are shown in Table 1 ▸. The coordinates and structure factors were deposited in the Protein Data Bank as entry 7ups.

Table 1. Data-collection and structure-refinement statistics.

	SeMet DotY	DotY
Data-collection statistics
Space group	C2	C2
a, b, c (Å)	153.6, 75.6, 101.1	153.2, 75.6, 101.2
α, β, γ (°)	90, 127.9, 90	90, 127.9, 90
Wavelength (Å)	0.97935	0.97935
Resolution (Å)	43.57–2.43 (2.55–2.43)	43.55–2.43 (2.55–2.43)
Total reflections	415536 (29228)	413796 (29059)
Unique reflections	64550 (6735)†	33201 (3535)‡
R meas	0.23 (2.24)	0.172 (1.46)
CC1/2	99.1 (20.2)	99.8 (50.4)
Completeness (%)	94.8 (73.6)	95.5 (76.3)
Multiplicity	6.4	6.4
〈I/σ(I)〉	8.12 (0.64)	9.35 (1.76)
SAD phasing
No. of sites	24 of 24
FOM	0.282
Refinement statistics
R cryst/R free		0.2371/0.2622
Unique reflections		33176 (3535)
Wilson B (Å2)		49.5
No. of atoms
Total		6495
Protein		6285
Water		194
Solute		16
Average B factors (Å2)
Overall		52.8
Protein		53.1
Water		43.6
Solute		51.2
R.m.s. deviations
Bond lengths (°)		0.0028
Bond angles (Å)		0.57
Ramachandran statistics
Favored (%)		96.96
Outliers (%)		0.61
Clashscore		5.9
PDB code		7ups

^†Friedel pairs recorded separately.

^‡Friedel pairs merged.

3. Results and discussion

The structure of DotY folds into two domains (Fig. 1 ▸ a). The N-terminal domain is a four-helix bundle containing the three N-terminal α-helices (α1–α3) augmented by the C-terminal α-helix (α7). This domain is connected by a long and flexible loop to the C-terminal domain, which contains two nearly perpendicular three-stranded β-sheets (β2–β1–β6 and β3–β4–β5) and three α-helices (α4, α5 and α6) that are packed into the space between the β-sheets. The third α-helix in this domain (α6) is interrupted in the middle by a short loop that breaks this helix into two shorter helices (Fig. 1 ▸ a). The second domain is connected by a short loop (amino acids 199–205) to the C-terminal α-helix (α7). This loop is well ordered in the crystal structure (Fig. 1 ▸ a). There are four copies of DotY in the asymmetric unit and they differ slightly in the relative orientations of the two domains (Fig. 1 ▸ b); however, the individual domains superimpose with a root-mean-square deviation of less than 0.4 Å (Fig. 1 ▸ c). The linker between the N- and C-terminal domains (amino acids 73–81) is highly flexible and could only be traced at the backbone level in two independent molecules, while it is disordered in the other two molecules. Moreover, in the latter two molecules the first α-helix (α4) in the C-terminal domain (amino acids 88–100) is distorted and has weak electron density. The N-terminal helix bundle packs against the first, more extensive β-sheet of the second domain and contacts it through side chains within the last three turns of the third, longest helix (Fig. 1 ▸ a).

Figure 1.

Structure of DotY. (a) Cartoon representation of the DotY molecule colored in a rainbow from blue at the N-terminus to red at the C-­terminus. The secondary-structure elements are numbered. Helix α6 is unwound in the middle. The flexible loop is colored gray and indicated by an arrow. (b) Superposition of four independent molecules based on the second domain. Each molecule is colored a different color. Small differences in the orientation of the N-terminal domain relative to the second domain are visible. (c) Independent superposition of the N-­terminal domains and second domains shows that they are nearly identical in all four independent molecules.

When comparing the DotY molecule from our crystal structure with the model obtained by cryo-EM of the complex that also contains DotL, DotM, DotN, IcmS, IcmW, LvgA and DotZ (PDB entry 7ovb; Macé et al., 2022 ▸), we noticed that while there is an overall similarity between the two models, there are some significant differences in the connectivity of the secondary structures. Firstly, in the cryo-EM model the C-terminal helix α7 has been released from the N-terminal domain and is not visible in the cryo-EM map, indicating its high flexibility. Secondly, the connectivity within the C-terminal domain is very different in the two models (Fig. 2 ▸). While there are three α-helices in this domain in both models and their locations are similar, they have been assigned to different regions within the protein sequence. Moreover, the two β-sheets present in the crystal structure were modeled as long extended segments in the cryo-EM structure and overlap only partially between the two models (Fig. 2 ▸ b). We believe that the lower resolution of the cryo-EM model and the high mobility of the middle segment of DotY within the context of the larger complex made the assignment of secondary structures to the primary sequence and the connectivity of secondary structures in the cryo-EM model rather difficult.

Figure 2.

Comparison of the second domain of the X-ray (a) and cryo-EM (b) models. The structures are in cartoon representation and are colored in a rainbow; the order of helices in the model is indicated, as is the location of the C-terminus of the domain. While their overall shape is similar, the color distribution is different and is indicative of different connectivity. The α6 helix in the X-ray model is interrupted by a short loop, which is well defined in the electron density.

The N-terminal helix-bundle domain displays a commonly observed fold. On the other hand, a DALI search performed for the second domain identified only low Z-value matches (Z < 5), of which the most similar is the ethanolamine-utilization microcompartment shell subunit from E. coli (PDB entry 3ia0; Tanaka et al., 2010 ▸), which forms hexamers. The topological similarity extends over only 68 residues within the segment 101–198, with a root-mean-square deviation of ∼3 Å (Fig. 3 ▸). The lack of a close structural homolog along the entire length of this compact domain indicates that it has a unique fold.

Figure 3.

Superposition of the C-terminal domain of DotY (colored in a rainbow from blue to red) with the best structural homolog, the ethanolamine-utilization microcompartment shell subunit from E. coli (PDB entry 3ia0, dark gray). There is a partial overlap of two of the three helices of DotY and several β-strands retaining sequential connectivity. The loops emanating from this core are significantly different in these two proteins.

The N-terminal and middle segments are connected by a long and flexible linker, which is partially disordered to a somewhat different extent in each of the four independent molecules in the crystal. The flexibility of this linker is likely to be important for its function within the T4SS secretion system and leads to the possibility of large-scale interdomain movements during the secretion process. Comparison of the crystal and cryo-EM structures shows a very different disposition of the middle domain relative to the N-terminal helix bundle. The first two helices of the N-terminal domain form the main contacts with the DotZ component of the entire T4CC sub­complex (Macé et al., 2022 ▸). The orientation of the middle domain relative to the N-terminal domain observed in the crystal structure is incompatible with binding to DotZ in the secretion-system complex as it would lead to significant steric hindrance, colliding with parts of DotZ and DotN (Fig. 4 ▸). This observation confirms the functional significance of a long and flexible tether connecting the two domains of DotY. In the crystal structure, the C-terminal helix associates with three N-terminal helices to form a four-helix bundle and binds to the opposite side that is involved in binding to DotZ. This C-terminal helix was not visible in the cryo-EM structure. This suggests that its association with the N-terminal helical domain is disrupted in the T4CC complex, likely as a result of a reorientation of the second domain. Within the context of the T4CC complex, the C-terminal helix could therefore play a role in effector docking to this large complex.

Figure 4.

Superposition of the X-ray (yellow, green and red) and cryo-EM (blue and sky blue) models showing the different orientation of the two domains in DotY on its own and within the T4CC complex. The orientation of the C-terminal domain observed in the crystal structure is incompatible with the complex as it collides with other proteins in the complex. The flexibility of the linker between the two domains is crucial for DotY to be able to bind DotZ in the T4CC complex.

Supplementary Material

PDB reference: DotY, 7ups

Funding Statement

MC acknowledges support from the Canada Research Chairs program and Canadian Institutes of Health Research grant MOP 48370.