Crystal structure of the nucleoid-associated protein Fis (PA4853) from Pseudomonas aeruginosa

The crystal structure of Pseudomonas aeruginosa Fis is composed of an N-terminal flexible loop and a C-terminal helix–turn–helix motif.

Abstract

Factor for inversion stimulation (Fis) is a versatile bacterial nucleoid-associated protein that can directly bind and bend DNA to influence DNA topology. It also plays crucial roles in regulating bacterial virulence factors and in optimizing bacterial adaptation to various environments. Fis from Pseudomonas aeruginosa (PA4853, referred to as PaFis) has recently been found to be required for virulence by regulating the expression of type III secretion system (T3SS) genes. PaFis can specifically bind to the promoter region of exsA, which functions as a T3SS master regulator, to regulate its expression and plays an essential role in transcription elongation from exsB to exsA. Here, the crystal structure of PaFis, which is composed of a four-helix bundle and forms a homodimer, is reported. PaFis shows remarkable structural similarities to the well studied Escherichia coli Fis (EcFis), including an N-terminal flexible loop and a C-terminal helix–turn–helix (HTH) motif. However, the critical residues for Hin-catalyzed DNA inversion in the N-terminal loop of EcFis are not conserved in PaFis and further studies are required to investigate its exact role. A gel-electrophoresis mobility-shift assay showed that PaFis can efficiently bind to the promoter region of exsA. Structure-based mutagenesis revealed that several conserved basic residues in the HTH motif play essential roles in DNA binding. These structural and biochemical studies may help in understanding the role of PaFis in the regulation of T3SS expression and in virulence.

1. Introduction  

Factor for inversion stimulation (Fis) is an abundant bacterial nucleoid-associated protein. It can directly bind and bend gene-promoter DNA to influence DNA topology and regulate gene expression (Browning et al., 2010 ▸). Moreover, it plays a versatile role in the regulation of various physiological processes, such as DNA replication (Filutowicz et al., 1992 ▸; Flåtten & Skarstad, 2013 ▸), DNA supercoiling and DNA topology (Schneider et al., 1997 ▸, 1999 ▸; Muskhelishvili & Travers, 2003 ▸), the transcription of rRNA and tRNA operons (Nilsson et al., 1992 ▸; Nilsson & Emilsson, 1994 ▸; Emilsson & Nilsson, 1995 ▸; Appleman et al., 1998 ▸) and the site-specific recombination of bacteriophage lambda (Ball & Johnson, 1991a ▸,b ▸).

Multiple structure–function studies have focused on Escherichia coli Fis (referred to as EcFis) and have shown that it has a four-helix bundle structure and functions as a homodimer (Kostrewa et al., 1991 ▸; Osuna et al., 1991 ▸; Stella et al., 2010 ▸; Shao et al., 2008 ▸; Hancock et al., 2013 ▸, 2016 ▸). The disordered or flexible N-terminal region is required for Hin-mediated inversion. The C-terminal region forms a helix–turn–helix (HTH) DNA-binding motif that is required for DNA binding and bending. The HTH region is strongly conserved, suggesting that the DNA-binding properties of Fis are conserved. Several DNA-bound structures suggest that EcFis initially recognizes DNA targets with intrinsically narrow minor grooves using the separation between the HTH motifs in the EcFis dimer as a ruler (Stella et al., 2010 ▸; Hancock et al., 2013 ▸, 2016 ▸).

Several studies have demonstrated that Fis may regulate the expression of multiple virulence factors in various bacterial pathogens, such as pathogenic E. coli (Falconi et al., 2001 ▸; Goldberg et al., 2001 ▸), Salmonella (Schechter et al., 2003 ▸; Kelly et al., 2004 ▸), Shigella flexneri (Falconi et al., 2001 ▸), Dickeya dadantii (Lautier & Nasser, 2007 ▸) and Yersinia pseudo­tuberculosis (Green et al., 2016 ▸). A recent study showed that Pseudomonas aeruginosa Fis (PA4853, referred to as PaFis) was required for the virulence of P. aeruginosa in a murine acute pneumonia model (Deng et al., 2017 ▸). PaFis was found to specifically bind to the −10 box of the P_exsA promoter and directly controls mRNA transcription elongation initiated by the P_exsC promoter. ExsA is the master transcriptional activator for the type III secretion system (T3SS) genes in P. aeruginosa, including exsC. Therefore, PaFis is required for the P_exsC-dependent transcription of exsA, which in turn is required for the full activation of the T3SS in P. aeruginosa. In this study, we determined the crystal structure of PaFis and found that it is composed of a four-helix bundle. PaFis showed remarkable structural similarity to EcFis. However, we found that the amino-acid sequence and the conformation of the N-terminal loop differ remarkably when compared with EcFis. PaFis can efficiently bind to the promoter region of exsA, and structure-based mutagenesis revealed the importance of several conserved residues in DNA binding. Our structure–function studies may help in understanding the role of PaFis in the regulation of T3SS expression and in virulence.

2. Materials and methods  

2.1. Macromolecule production  

The full-length PaFis gene (PA4853) was amplified by PCR from P. aeruginosa PAO1 genomic DNA and was cloned into the pET-28a-SUMO vector (with a cleavable His-SUMO tag at the N-terminus). The recombinant plasmid was transformed into the E. coli BL21(DE3) expression strain (Invitrogen). Site-directed mutagenesis of PaFis was performed by a PCR-based technique according to the QuikChange site-directed mutagenesis strategy (Stratagene) following the manufacturer’s instructions. The mutant genes were sequenced and found to contain only the desired mutations. The mutation primers used in the study are listed in Supplementary Table S1.

The bacterial cells were grown to mid-log phase (OD_600 nm = ∼0.8) in LB medium at 310 K in the presence of 50 µg ml⁻¹ kanamycin. Induction of the culture was then carried out with 0.3 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) at 293 K. The cells were pelleted after 20 h by centrifugation at 6500g for 10 min at 277 K. The cell pellet was resuspended in buffer A [20 mM Tris, 300 mM NaCl, 5%(v/v) glycerol, 1 mM phenylmethylsulfonyl fluoride pH 8.0] and lysed by ultrasonification on ice. The cell debris and membranes were pelleted by centrifugation at 36 000g for 45 min at 277 K. The soluble N-terminally hexahistidine-tagged PaFis was purified by affinity chromatography with nickel–nitrilotriacetic acid resin (Bio-Rad) and the His-SUMO tag was removed by overnight hydrolysis with the protease ULP1. Untagged proteins were removed with buffer A containing 50 mM imidazole. Recombinant PaFis was then eluted with buffer A containing 250 mM imidazole. The protein was further purified by gel filtration on a Superdex 200 column (GE Healthcare) equilibrated in buffer B [20 mM Tris, 300 mM NaCl, 5%(v/v) glycerol pH 8.0] using an ÄKTApurifier system (Amersham). Protein concentrations were determined using the Bio-Rad Protein Assay Kit II. Macromolecule-production information is summarized in Table 1 ▸. The production and purification of the protein mutants were the same as described above.

Table 1. Macromolecule-production information.

Source organism	P. aeruginosa PAO1
DNA source	Genomic DNA
Forward primer†	5′-CGCGGATCCCTGCGTGATAGCGTGGA-3′
Reverse primer†	5′-CCGCTCGAGTTACAGCAGATCGTATTGCTTC-3′
Cloning vector	pET-28a-SUMO
Expression vector	pET-28a-SUMO
Expression host	E. coli BL21(DE3)
Complete amino-acid sequence of the construct produced‡	MGSSHHHHHHHMMSDSEVNQEAKPEVKPEVKPETHINLKVSDGSSEIFFKIKKTTPLRRLMEAFAKRQGKEMDSLRFLYDGIRIQADQTPEDLDMEDNDIIEAHREQIGGS MTTETLVSGTTPVSDNANLKQHLTTPTQEGQTLRDSVEKALHNYFAHLEGQPVTDVYNMVLCEVEAPLLETVMNHVKGNQTKASELLGLNRGTLRKKLKQYDLL

^†Restriction-enzyme sites are underlined.

^‡The PaFis sequence is underlined and the ULP1 cleavage site is in bold.

2.2. Crystallization  

The purified PaFis was concentrated to ∼15 mg ml⁻¹ using a Millipore Amicon Ultra apparatus. The initial crystallization conditions were obtained by the utilization of several sparse-matrix screens (Emerald Biosystems, USA) with the sitting-drop vapor-diffusion method at room temperature after 2–3 days. Crystal quality was optimized by adjusting the concentrations of the precipitant and buffer. The best crystals were both obtained in a solution consisting of 0.1 M Na HEPES pH 7.5, 20% PEG MME 2000 at 293 K. Crystallization information is summarized in Table 2 ▸.

Table 2. Crystallization.

Method	Sitting-drop vapor diffusion
Plate type	48-well
Temperature (K)	293
Protein concentration (mg ml−1)	15
Buffer composition of protein solution	20 mM Tris HCl pH 8.0, 100 mM NaCl
Composition of reservoir solution	0.1 M Na HEPES pH 7.5, 20% PEG MME 2000
Volume and ratio of drop	1 µl, 1:1
Volume of reservoir (µl)	100

2.3. Data collection and processing, structure solution and refinement  

Diffraction data were collected from a single crystal at the BL18U beamline station of Shanghai Synchrotron Radiation Facility (SSRF) using a PILATUS 6M detector at a wavelength of 0.9788 Å. The total oscillation was 360° with 0.5° per image, and the exposure time was 0.1 s per image. Before data collection, the crystals were soaked in reservoir solution supplemented with 20%(v/v) glycerol for a few seconds and then flash-cooled in liquid nitrogen. All data were processed using HKL-2000 (Otwinowski & Minor, 1997 ▸). The initial phases were calculated with Phaser using EcFis (PDB entry 1f36; Safo et al., 1997 ▸) as the search model. The structure was refined with phenix.refine (Afonine et al., 2012 ▸) and manually corrected in Coot (Emsley et al., 2010 ▸). The quality of the final models was validated with MolProbity (Chen et al., 2010 ▸). The atomic coordinates and structure factors of PaFis have been deposited in the PDB with accession code 6m10. Summaries of the data-collection and final refinement statistics are listed in Tables 3 ▸ and 4 ▸, respectively. PyMOL (http://www.pymol.org) was used to prepare structural figures.

Table 3. Data collection and processing.

Values in parentheses are for the outer shell.

Diffraction source	BL18U1, SSRF
Wavelength (Å)	0.9788
Temperature (K)	100
Detector	PILATUS 6M
Crystal-to-detector distance (mm)	400
Rotation range per image (°)	0.5
Total rotation range (°)	360
Exposure time per image (s)	0.1
Space group	C2221
a, b, c (Å)	44.5, 194.0, 92.0
α, β, γ (°)	90, 90, 90
Mosaicity (°)	0.42
Resolution range (Å)	60.00–2.99 (3.05–2.99)
Total No. of reflections	91678 (2941)
No. of unique reflections	8292 (346)
Completeness (%)	97.6 (84.4)
Multiplicity	11.1 (8.5)
〈I/σ(I)〉	13.23 (1.46)
R merge	0.193 (0.842)
R meas	0.202 (0.890)
CC1/2	0.993 (0.974)
Overall B factor from Wilson plot (Å2)	68.5

Table 4. Structure solution and refinement.

Values in parentheses are for the outer shell.

Resolution range (Å)	48.51–2.99 (3.17–2.99)
Completeness (%)	97.1 (87.0)
No. of reflections, working set	7427 (1108)
No. of reflections, test set	826 (121)
Final R cryst	0.2612 (0.3363)
Final R free	0.2986 (0.3737)
No. of non-H atoms
Protein	2562
Ion	0
Ligand	0
Water	0
Total	2562
R.m.s. deviations
Bonds (Å)	0.006
Angles (°)	1.00
Average B factors (Å2)
Protein	64.0
Ion	0
Ligand	0
Water	0
Ramachandran plot
Most favored (%)	91.46
Allowed (%)	7.59
Outliers (%)	0.95

2.4. Gel-electrophoresis mobility-shift assay (EMSA)  

A DNA fragment (5′-ATCCTGATTATTATAAAATCGACTCCG-3′) derived from the exsA promoter region was chemically synthesized by Beijing AuGCT Biotechnology Co. Ltd, People’s Republic of China. DNA samples (2 µM final concentration) were annealed in buffer consisting of 10 mM NaCl, 20 mM Tris pH 8.0. PaFis–DNA complexes were prepared by adding PaFis at final concentrations of 0, 10, 20 and 40 µM and incubating for 30 min at room temperature. For each sample, free DNA and complexes were separated on a 5% acrylamide native gel run for 40 min at 80 V at 293 K and were visualized by staining with ethidium bromide (Thermo Fisher Scientific, USA).

3. Results and discussion  

3.1. Overall structure of PaFis  

The structure of PaFis was solved by molecular replacement in space group C222₁, using EcFis (PDB entry 1f36), with a sequence identity of 54%, as the search model, and was refined to final R and R _free values of 0.26 and 0.29, respectively, at 2.99 Å resolution. The asymmetric unit contains four PaFis molecules (Supplementary Fig. S1): two of the molecules form a homodimer, while the other two molecules also form dimers with crystallographic symmetry-related molecules by crystal packing. The N-terminal residues Met1–Gln21 (Fig. 1 ▸ a), which may be flexible in solution, were not observed in the electron-density map and are not included in the current model.

Figure 1.

Overview of the PaFis (PA4853) structure from P. aeruginosa PAO1. (a) Structure-based sequence alignment of PaFis with representative homologs from various species performed using ClustalX (version 1.81) and ESPript3. They include Pseudomonas aeruginosa (P. aeruginosa), Escherichia coli (E. coli), Shigella flexneri (S. flexneri), Yersinia pestis (Y. pestis) and Salmonella typhimurium (S. typhimurium). (b) Overall crystal structure of the PaFis homodimer at 2.99 Å resolution. The two protomers are colored green and magenta. (c) Electron-density map (2F _o − F _c; blue) of the N-terminal region contoured at the 1.5σ level.

PaFis is composed of a four-α-helical bundle with a hydrophobic core (Fig. 1 ▸ b). The C-terminal helices α2, α3 and α4 form a typical helix–turn–helix (HTH) motif for potential DNA binding. The N-terminal flexible loop (His22–Leu33) can be unambiguously built with clear electron density despite the moderate resolution (Fig. 1 ▸ c). A compact homodimer is formed between two monomers by making extensive contacts (hydrogen bonds and hydrophobic interactions). The dimerization interface is formed by their respective α1 and α2 helices (Fig. 1 ▸ b). The buried surface area in the dimer interface is 1605 Ų, which is 22.7% of the total surface area per monomer (7069 Ų).

A DALI search (http://ekhidna.biocenter.helsinki.fi/dali_server) for globally similar proteins was performed against the Protein Data Bank (PDB). PaFis shows significant structural similarities to both apo and DNA-bound EcFis (Figs. 2 ▸ a–2c). In addition, the HTH motifs of the DNA-binding domains (DBDs) of nitrogen regulatory protein C (NtrC) family members also have notable similarities to the HTH motif in PaFis.

Figure 2.

Structural comparisons of PaFis with its homologs. (a) Representative PaFis homologs returned by the DALI web server. (b, c) Structural superposition of PaFis with apo wild-type EcFis (PDB entry 4fis; Yuan et al., 1991 ▸) and the DNA-binding domain (DBD) of NtrC4 (PDB entry 4fth; Vidangos et al., 2014 ▸).

3.2. The sequences of the N-terminal loop in PaFis are not conserved  

Sequence alignment of PaFis with representative homologs showed that most of the N-terminal residues are variable in PaFis compared with its homologs, including that from E. coli, in which most residues in the N-terminus are highly conserved (Fig. 1 ▸ a). Notably, the N-terminal residues between residues 10 and at least 34 that are required for controlling DNA inversion in EcFis (Koch et al., 1991 ▸; Osuna et al., 1991 ▸) are not conserved in PaFis. For example, the residues that are critical for Hin-catalyzed DNA inversion in EcFis, including Leu11, Val16, Asn17, Asp20, Val22 and Pro26, correspond to Thr11, Ala17, Asn18, Gln21, Leu23 and Thr32 in PaFis, respectively (Fig. 1 ▸ a). Therefore, further functional studies are required to investigate the exact role of the N-terminal region of PaFis in the regulation of DNA inversion.

As in the structure of PaFis, apo wild-type EcFis (PDB entry 4fis; Yuan et al., 1991 ▸) has a very flexible N-terminus, residues 1–25 of which cannot be observed in the crystal structure. The N-terminus of EcFis is properly folded into two β-hairpins in its mutants, including K36E (PDB entry 1f36; Safo et al., 1997 ▸), R71Y (PDB entry 1etq; Cheng et al., 2000 ▸) and R71L (PDB entry 1eto; Cheng et al., 2000 ▸), and in the wild type bound to DNA (PDB entries 3iv5, 5e3o and 5e3n; Stella et al., 2010 ▸; Hancock et al., 2016 ▸). The two β-hairpin arms protrude over 20 Å from the protein core (Figs. 3 ▸ a and 3 ▸ b), and function to activate DNA invertases such as Hin (Koch et al., 1991 ▸; Osuna et al., 1991 ▸; Safo et al., 1997 ▸). In apo PaFis no such β-arm can be observed in the N-terminus, probably as a result of its flexibility.

Figure 3.

Structural comparison of PaFis with DNA-bound EcFis. (a) Structural superposition of the PaFis homodimer with DNA-bound EcFis (PDB entry 3iv5). The two protomers of EcFis are colored light gray and dark gray, respectively, and PaFis is colored as in Fig. 1 ▸(b). (b) Close-up view of the N-terminal regions of PaFis and EcFis. (c) Close-up view of the DNA-binding regions of PaFis and EcFis. (d) EMSA assays of the DNA-binding capacity of PaFis mutants.

3.3. Several conserved Arg/Lys residues play essential roles in DNA binding  

Structural superposition of PaFis with DNA-bound EcFis showed that the α4 helices of PaFis could insert into the major groove of DNA without a steric clash (Fig. 3 ▸ c). In EcFis, several residues located in the HTH region (such as Arg85, Thr87 and Lys90) are commonly required for specific DNA recognition of different sequences, while several other residues (such as Asn73, Asn84, Arg89, Lys91 and Lys93) make variable contributions to the binding affinities of different sequences (Feldman-Cohen et al., 2006 ▸). These key residues are highly conserved in PaFis. For example, the residues Arg85, Arg89, Lys90 and Lys91 in EcFis that are important for sequence-specific binding correspond to Arg91, Arg95, Lys96 and Lys97 in PaFis. In order to evaluate the role of these residues in DNA binding by PaFis, they were mutated to alanines and the mutants were tested for DNA binding using an electrophoretic mobility-shift assay (EMSA). The EMSA results showed that wild-type PaFis bound and shifted the DNA fragment derived from the exsA promoter, as expected (Fig. 3 ▸ d). When increasing amounts (10–40 µM) of protein were co-incubated with DNA, no obviously shifted bands were observed for any of the four mutants. These results show the four conserved basic residues are critical for DNA binding to the exsA promoter by PaFis.

As a global regulator, EcFis has been found to promiscuously and dynamically bind DNA in vitro (Skoko et al., 2006 ▸). Despite the remarkably diverse DNA sequences that bind to EcFis with high affinities, there is a 15 bp AT-rich core element bounded by G/C and C/G base pairs (Hancock et al., 2016 ▸). There is a similar AT-rich motif (5′-ATTATTATAAAAT-3′) in the exsA promoter region bound by PaFis (Fig. 3 ▸ d), and mutation of the AT-rich motif has previously been found to abrogate binding by PaFis (Deng et al., 2017 ▸). The results suggest that the DNA-binding properties of PaFis and EcFis are conserved.

4. Conclusion  

In this work, we determined the crystal structure of PaFis, the second structure of a bacterial Fis. Although the N-terminal loop in PaFis is flexible like that in EcFis, its sequence is unlike that of EcFis and other homologs. Moreover, the key residues required for controlling DNA inversion in EcFis are not conserved in PaFis. Considering these notable differences, further studies will be required to elucidate the exact role of the N-terminal region in PaFis.

Supplementary Material

PDB reference: Pseudomonas aeruginosa Fis, 6m10

Funding Statement

This work was funded by National Basic Research Program of China (973 Program) grant 2017YFA0504900. National Natural Science Foundation of China grants U1732113, 31670059, and 31570744). Chinese Academy of Sciences, Strategic Priority Research Program grant XDB08030103.