Gene cloning and mutagenesis
A set of genes was identified within a single operon from Legionella pneumophila Philadelphia 1, specifically HEPNLpg (lpg2920) and MNTLpg (lpg2921), with an overlap of 8 base pairs observed between the genes of HEPNLpg and MNTLpg. The genes encoding HEPNLpg, MNTLpg, and the HEPNLpg-MNTLpg complex were synthesized by Bionics (Seoul, Korea) and amplified by polymerase chain reaction (PCR) using the following primers (Supplementary Table 3). The PCR products were digested by Nde I and Xho I and ligated separately into pET-28a (+), with N-terminal His-tags. The recombinant plasmids were subsequently transformed into the E. coli strain XL10-Gold for amplification.
To identify the residues that are essential for the activity of HEPNLpg and MNTLpg, as well as the tetrameric interface for HEPNLpg, several mutations were introduced using the EZchangTM Site-Directed Mutagenesis Kit (Enzynomics, Daejon, Korea) according to the manufacturer’s protocol. The mutations in HEPNLpg were Q64AR98A, Q64AH103A, Q64AQ44A, Q64AE47A, Q64AQ44AE47A, Q64A and R73A, while the mutations in MNTLpg were G36AS37T, D48E, D50E, and D48ED50E. The primers used for mutagenesis are described in Supplementary Table 3.
Protein expression and purification
The recombinant HEPNLpg WT protein and its mutants were overexpressed in E. coli Rosetta2 (DE3) cells using Luria–Bertani (LB) medium supplemented with kanamycin. E. coli Rosetta2 (DE3) cells harboring HEPNLpg WT were grown until the OD600 reached 0.5–0.6, after which the proteins were induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 4 h at 37 °C. Due to the high toxicity, the cells harboring HEPNLpg Q64A and R73A were grown to the stationary phase (OD600 of ~ 1–1.2) and induced by 0.5 mM IPTG for only 2 h; to ensure consistent induction conditions, all HEPNLpg mutants potentially related to the active site (Q64AR98A, Q64AH103A, Q64AQ44A, Q64AE47A, and Q64AQ44AE47A) were induced in the same manner as HEPNLpg Q64A. The cells were harvested by centrifugation at 6,400 × g for 10 min and frozen at −80 °C. The harvested cells were resuspended in buffer A (20 mM Tris-HCl, pH 7.9, and 500 mM NaCl) containing 10% (v/v) glycerol and lysed by ultrasonication. After centrifugation at 28,300 × g for 1 h at 10 °C, the supernatant containing the HEPNLpg protein with an N-terminal His-tag was applied to a Ni2+-NTA column (Qiagen, Hilden, Germany), preequilibrated with buffer A and washed with buffer A containing 50 mM imidazole. The target proteins were eluted with an imidazole gradient (100 mM–500 mM). The next purification step involved size-exclusion chromatography (SEC) using a Hiload 16/60 Superdex 200 prep-grade column (GE Healthcare, Chicago, IL, USA) preequilibrated with buffer containing 20 mM HEPES (pH 7.5) and 200 mM NaCl. For the standalone MNTLpg and its mutants, the entire process of expression and purification steps was identical to that of HEPNLpg. To isolate NMPylated HEPNLpg, cells harboring the plasmid encoding the HEPNLpg-MNTLpg complex were grown until the OD600 reached 0.4–0.6 and induced with 0.2 mM IPTG for 20 h at 15 °C. After the cells were lysed by ultrasonication, the supernatant was loaded on a Ni2+-NTA column and washed with buffer A containing 50 mM imidazole. The NMPylated HEPNLpg proteins were eluted with an imidazole gradient ranging from 300 mM to 500 mM. In this scenario, most of the standalone MNTLpg was eluted in the wash buffer. The second purification step involving SEC was conducted in the same manner as that used for HEPNLpg. To obtain the HEPNLpg-MNTLpg complex, the purified HEPNLpg WT protein was incubated with purified MNTLpg for 30 min at 20 °C. The sample was then subjected to a Hiload 16/60 Superdex 200 prep-grade column of SEC for further purification. The purity of the target protein was confirmed by SDS‒PAGE.
Crystallization and diffraction data collection
Crystallization was performed using the sitting-drop vapor diffusion method at 20 °C. The initial screening was conducted by mixing equal volumes (0.5 µl each) of protein solution and reservoir solution, and the conditions were subsequently further optimized. The protein concentrations of HEPNLpg WT, HEPNLpg Q64A, and NMPylated HEPNLpg used for crystallization were 13 mg/ml; the best crystal of HEPNLpg WT was grown in the presence of 20% (w/v) PEG 3350 and 0.2 M ammonium tartrate dibasic (pH 6.5). The crystal buffer for HEPNLpg Q64A was 20% (w/v) PEG 3350 and 0.2 M lithium acetate dehydrate. For NMPylated HEPNLpg, the crystal buffer consisted of 25% (w/v) PEG 3350, 0.1 M Bis-Tris and 0.2 M ammonium sulfate (pH 5.5). The MNTLpg protein was concentrated to 3 mg/ml and crystallized in a buffer containing 20% (w/v) PEG 6000, 0.1 M citric acid/sodium hydroxide and 1 M lithium chloride (pH 4.0). A crystal of the HEPNLpg-MNTLpg complex was obtained at a concentration of 12 mg/ml in the presence of 20% (w/v) PEG 3350 and 0.2 M ammonium citrate tribasic (pH 7.0). Prior to data collection, all of the optimized crystals were soaked in a cryoprotectant consisting of the reservoir solution with 20% (v/v) glycerol. The diffraction data were collected using a Quantum Q270r CCD detector (ADSC, Poway, CA, USA) at beamline 7 A and a PILATUS3 6 M CCD detector (Dectris, Baden-Daettwil, Switzerland) at beamline 11 C of the Pohang Light Source, Republic of Korea, as well as an EIGER X 4 M (x2) detector (Dectris) at beamline BL-1A of the Photon Factory, Japan.
The raw data were processed and scaled using the HKL2000 program package30 and XDS program package31. The detailed data statistics are summarized in Supplementary Table 4.
Structure determination and refinement
The HEPNLpg WT and MNTLpg structures were solved by the molecular replacement (MR) method using the program Phaser-MR32 with a prediction model from Alphafold233 as an initial template, and resolutions of ~1.8 Å and ~ 2.8 Å were obtained, respectively. The refined model of HEPNLpg WT was subsequently used to solve the phase problem for NMPylated HEPNLpg and HEPNLpg Q64A using the same MR method at resolutions of ~1.7 Å and ~1.59 Å, respectively. The structure of the HEPNLpg-MNTLpg complex was solved through Phaser-MR, utilizing refined HEPNLpg WT and MNTLpg as templates at a resolution of ~2.4 Å. Five percent of the data were randomly reserved as the test set to calculate Rfree for the entire dataset34. The model was manually modified using the program Coot35 and automatically refined by Refmac in the CCP4 Program suite36 and phenix. refine in PHENIX37. Water molecules and ligands were introduced using the program Coot35. All the final models were subjected to stereochemical analysis using MOLPROBITY38. PISA39 and protein interaction calculator (PIC)40 were used to calculate the interface area and associated interactions. PyMOL (PyMOL Molecular Graphics System, version 2.1; Schrödinger, LLC., Cambridge, MA, USA) was used to visualize and generate figures.
Analytical ultracentrifugation (AUC)
Analytical ultracentrifugation was conducted using an Optima AUC (Beckman Coulter, Inc., Brea, CA, USA) equipped with an absorbance detector set to 280 nm. Sedimentation velocity experiments were performed at a rotor speed of 45,000 rpm with a 60Ti rotor. The experiments utilized a 2-sector EPON centerpiece, which was loaded with ~0.3 mg/ml (10 µM) HEPNLpg (WT, Q64A, R73A) in a final buffer composed of 20 mM HEPES (pH 7.5) and 200 mM NaCl. The reference solution used was the final buffer alone. Data analysis was carried out using the SEDFIT software, version 16.1c.
Surface plasmon resonance (SPR)
The homotetramerization affinities of HEPNLpg WT, HEPNLpg Q64A, and HEPNLpg R73A were performed using SPR kinetics experiments. All experiments were conducted at 25 °C using a Biacore T200 system (GE Healthcare). Amine coupling was employed for immobilization, utilizing a kit containing 0.1 M N-hydroxysuccinimide and 0.4 M 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride on a CM5 sensor chip. HBS-P buffer (10 mM HEPES, pH 7.5, 150 mM NaCl, and 0.005% Tween 20) was used according to the manufacturer’s protocol (GE Healthcare). For homotetramerization affinity measurements, each HEPNLpg WT, HEPNLpg Q64A, and HEPNLpg R73A, dissolved in 10 mM sodium acetate at pH 5.5, was injected and immobilized on the flow cell 2 (sample cell) at 10 μg/mL as ligands until the immobilization level reached approximately 600 response units (RU). To prevent redundant self-assembly of HEPNLpg during immobilization, sufficient HBS-P buffer was applied to achieve the desired immobilization level. The remaining activated carboxyl groups on the CM5 sensor chip surface were deactivated with 1 M ethanolamine at pH 8.5 for 450 s. The control was settled identically with flow cell 1 (reference cell) without immobilized proteins to subtract the response from each sample dataset. The corresponding proteins of immobilized ligands (HEPNLpg WT, HEPNLpg Q64A, and HEPNLpg R73A) were diluted in HBS-P buffer at concentrations of 0.39 μM, 0.78 μM, 1.56 μM, 3.13 μM, and 6.25 μM. They were injected as analytes at a rate of 10 μL/min for 120 s overflow cells 1 and 2, followed by dissociation for 300 s in multi-cycle reactions. As the response units adequately returned to baseline following dissociation, no regeneration process was performed. SPR response data were fit to a steady-state affinity model to determine the maximal response (Rmax) using Biacore T200 evaluation software 3.0 (GE Healthcare).
Molecular Dynamics (MD) simulation
The Maestro software from the Schrödinger suite was employed for the molecular dynamics simulations. Initially, the protein models for HEPNLpg WT and HEPNLpg Q64A, were prepared by the protein preparation wizard with OPLS4 force field41. Desmond was used to carry out the molecular dynamics simulations. For explicit solvent simulations, periodic boundary conditions with orthorhombic boxes buffered at 10 × 10 × 10 Å distances were applied. The system was solvated using the TIP3P water model and supplemented with 150 mM NaCl after being neutralized with sodium or chloride ions to maintain an electrical balance. The solvated system containing protein underwent energy minimization and relaxation for 100 ps using the minimization step in Desmond with the OPS2005 forcefield. The simulations were conducted in the NPT ensemble (isothermal and isobaric simulations) using the Martyna-Tobias-Klein method for isotropic pressure, maintained at 1 atm and the Nose-Hoover thermostat algorithm for constant temperature at 300 K42,43. A total of 200 ns simulations were performed, with trajectories saved at 200 ps intervals. The trajectories were analyzed using the simulation interaction diagram. MD simulations input model and output files and MD simulations reliability and reproducibility checklist are available as Supplementary Data 1.
Total RNA extraction from L. pneumophila and in vitro total RNA digestion assay
To isolate total RNA from L. pneumophila Philadelphia 1, the bacterial strain was obtained from the Korean Collection for Type Cultures (KCTC). The cells were cultured in buffered yeast extract (BYE) broth, which consists of 10 g yeast extract, 1 g α-ketoglutarate, 10 g ACES buffer, 0.4 g L-cysteine HCl, and 0.25 g iron (III) pyrophosphate per liter of purified water, with the pH adjusted to 6.9. When L. pneumophila cells were grown until the OD600 reached 1–1.1, the cells were harvested by centrifugation at 18,000 × g for 2 min. Total RNA, including rRNA and tRNA, was isolated from L. pneumophila using the TRI reagent RNA isolation method. To further isolate the small-sized RNA (5S rRNA and tRNA) from the purified total RNA, a FARB Mini Column (FAVORGEN, Ping Tung, Taiwan) was utilized. The 5S rRNA and tRNA were separated into the flow-through by centrifugation at 18,000 × g for 30 sec. They were further purified using 75% ethanol and finally dissolved in RNase-free water.
To determine the RNase activity of dimeric HEPNLpg toxin, an in vitro total RNA digestion assay was conducted with dimeric HEPNLpg using total RNA from L. pneumophila at 37 °C for 30 min. The 10 μl reaction mixture contained 1 μM RNA with a dimeric HEPNLpg concentration range of 10 nM–160 nM. The assays were conducted in buffer containing 20 mM HEPES (pH 7.5) and 200 mM NaCl. The reaction mixture was subjected to electrophoresis on a 1.8% agarose gel with 0.5x TBE for 25 min.
To verify whether dimeric HEPNLpg can cleave rRNA in an intact ribosome, the purified 70S ribosome was obtained from New England Biolabs (Ipswich, MA, USA). The ribosome was dissolved in a storage buffer (20 mM HEPES-KOH, pH 7.6, 10 mM Mg(OAc)2, 30 mM KCl, and 7 mM beta-mercaptoethanol) to maintain its integrity. For the in vitro rRNA digestion assay, 1 μM ribosome was used, following the same method as the total RNA digestion assay. To improve the resolution for detecting small-sized RNA, a 12% polyacrylamide gel was used for electrophoresis to analyze the impact of HEPNLpg on 5S rRNA and tRNA. The electrophoresis was conducted using 0.5x TBE at 100 V for 120 min. All electrophoresis results were visualized using a PrintGraph 2 M.
In vitro RNase assay
To conduct a quantitative analysis of the RNase activity of apo HEPNLpg, a fluorescence quenching assay was conducted on dimeric HEPNLpg and its variants using an RNase Alert Kit (IDT, Coralville, IA, USA). A synthetic RNA oligonucleotide covalently linked to fluorescein at one end and a dark quencher at the other end exhibited fluorescence quenching. When the RNA was digested by an RNase, fluorescein was liberated from the quencher. The released fluorescein exhibited green fluorescence (490 nm excitation, 520 nm emission), which was measured via RFU on a SpectraMax M5 plate reader (Molecular Devices, San Jose, CA, USA). The assays were performed in a buffer comprising 20 mM HEPES (pH 7.5) and 200 mM NaCl at 37 °C. To compare the RNase activity of dimeric HEPNLpg with that of its variants, 20 μl of reaction mixture containing 2 μM protein and 0.2 μM RNA was used. The fluorescence endpoints were detected after a 1 h reaction. To assess the potential impact of contamination from other RNases in the in vitro RNase assays, we employed the Ribolock RNase inhibitor (henceforth, Ribolock) (Thermo Scientific, Waltham, MA, USA), which effectively inhibits the activity of various RNases. 1 U Ribolock was added to reaction mixture containing HEPNLpg samples (including dimeric HEPNLpg or its mutants) and RNA substrates. Fluorescence kinetics were detected at 30-s intervals throughout a 1 h reaction. To measure the Michaelis–Menten kinetics of dimeric HEPNLpg and its variants, 20 μl of reaction mixture containing 0.35 μM protein and an RNA concentration ranging from 0.1–1.4 μM were used. The initial velocities and kinetic parameters Km and kcat were analyzed using Prism software (version 10.0.1 for macOS; GraphPad Software, Boston, Massachusetts, USA). All pipette tips and tubes used in the in vitro RNase assay were autoclaved, and the laboratory bench surfaces and pipettes were treated with RNaseZapTM RNase Decontamination Solution (Invitrogen, Carlsbad, CA, USA) to prevent RNase contamination during the assays.
Spot-plating assays
Cultures of E. coli Rosetta2 (DE3) cells harboring the pET 28a vector with each HEPNLpg variant were grown overnight at 37 °C in liquid LB media supplemented with kanamycin. For spot assays, the overnight cultures were diluted 1:100 and grown at 37 °C until the OD600 reached ~0.5. The cells were harvested by centrifugation at 5000 rpm for 5 min, washed in phosphate-buffered saline (PBS), and serially diluted (10−2–10−6) before being spotted onto LB agar plates containing the indicated amount of inducer or repressor. Gene expression from plasmids carrying the pET promoter was repressed by 1% glucose and induced by a final concentration of 0.2 mM IPTG. LB agar plates were incubated at 37 °C for approximately 16 hr.
Size-exclusion chromatography with multiangle light scattering (SEC-MALS)
SEC-MALS experiments were performed using an FPLC system (GE Healthcare) connected to a Wyatt MiniDAWN TREOS MALS instrument and an Optilab rEX differential refractometer (Wyatt, Santa Barbara, CA, USA). A Superdex 200 10/300 GL (GE Healthcare) gel-filtration column preequilibrated with buffer containing 20 mM HEPES (pH 7.5) and 200 mM NaCl was used. Purified MNTLpg was injected at a concentration of 3 mg/ml. The sample was injected at a flow rate of 0.4 mL/min. The data were analyzed using the Zimm model for fitting static light-scattering data and graphed using the EASI graph with an RI peak in the ASTRA VI (Wyatt).
In vitro NMPylation assay
In vitro NMPylation assays were performed to investigate the function of MNTLpg. The experiments were divided into two steps, namely, storage and reaction procedures. The storage procedures were dependent on the purpose of the experiments and were performed in storage buffer (20 mM HEPES, pH 7.5, and 200 mM NaCl). First, to identify the types of nucleotides that MNTLpg transfers to HEPNLpg, 40 μM MNTLpg was incubated at 20 °C for 30 min with NTP (ATP, CTP, and GTP) at a concentration range of 10 μM–160 μM. Second, to determine the NMPylation active site of MNTLpg, 40 μM MNTLpg or its mutants (G36AS37T, D48E, D50E, and D48ED50E) was incubated with 160 μM ATP at 20 °C for 30 min. Third, to assess the impact of metal ions on the NMPylation ability of MNTLpg, EDTA-treated MNTLpg was used. EDTA-treated MNTLpg (40 μM) was incubated with 160 μM ATP and 1 mM metal ions (Mg2+, Zn2+, Ca2+, Co2+, and Ni2+) at 20 °C for 30 min. After the storage procedures, the subsequent reaction steps were conducted in reaction buffer (20 mM HEPES, pH 7.5, 200 mM NaCl, 5 mM DTT, and 10 mM MgCl2). 80 μM HEPNLpg was added to the reaction buffer and incubated at 37 °C for an additional 1 hr. In the case of the EDTA-treated MNTLpg sample, MgCl2 was excluded from the reaction buffer. SDS‒PAGE loading buffer was added to stop the reaction, and the results were analyzed via SDS‒PAGE.
Liquid chromatography-mass spectrometry (LC-MS) analysis
To identify the NMP products that were cleaved from NMPylated HEPNLpg, LC-MS was performed. Purified NMPylated HEPNLpg was treated with phosphodiesterase I (PDE I) at 37 °C for 1 hr. The small molecules were then collected by 3 kDa MWCO membranes for further analysis. In this experiment, a buffer without small molecules was used as the control, and both the control and small molecule samples were analyzed in duplicate. LC-MS was performed using electrospray ionization mass spectrometry (ESI-MS) with an integrated HPLC/ESI-MS system (InfinityLab LC/MSD iQ, Agilent Technologies, G6160A). The sample was injected into an InfinityLab Poroshell 120 EC-C18 column (2.1 × 50 mm, 2.7 μm) at a flow rate of 0.3 ml/min. Elution was performed using a linear gradient of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile) as follows: 0–1 min, 1%–5% B; 1–2 min, 5%–10% B; 2–3 min, 10%–5% B; and 3–4 min, 5%–1% B. The ionization capillary voltage was adjusted to 3500 V, while the fragmentor was set to 110 V. The raw data obtained from LC-MS were processed using OpenLab CDS (version 2.8; Agilent Technologies, Santa Clara, CA, USA). The software was employed to identify the peaks based on retention times and m/z values. Any ambiguous peaks were further validated manually by analyzing their fragmentation patterns. For data analysis, signal intensities were normalized to the internal standard, and baseline corrections were applied. LC-MS raw data are available with this paper as Supplementary Data 2.
High-resolution LC-MS/MS
The molecular weights of the protein samples (apo HEPNLpg and NMPylated HEPNLpg) were analyzed by MS using a triple time of flight (TOF) 5600+ system (AB Sciex, Framingham, MA, USA), which was equipped with an Ultimate 3000 HPLC (Thermo Scientific, Waltham, MA, USA). The respective proteins were obtained following the procedure outlined in the previous section. Each sample was injected once into an INNO-P C4 column (2.0 mm × 50 mm, 5 μm) at a flow rate of 0.4 ml/min. The elution gradient for the separation of proteins was as follows: 0–0.5 min, 15% B; 0.5–11 min, 15%–100% B; 11–15 min, 100% B; 15–15.5 min, 100%–15% B; and 15.5–20 min, 15% B. The mobile phase consisted of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). Data were collected in both MS1 and MS2 modes, using a full scan in positive ionization mode. Ionization was performed via electrospray ionization (ESI). The ion source pressures for both source 1 and source 2 were set at 50 psi, while the curtain gas pressure was maintained at 25 psi. The desolvation temperature was kept at 500 °C, and the ion spray voltage was set to 5.5 kV. Nitrogen was used as the carrier gas, and the declustering potential (DP) was set at 60 V. High-resolution mass spectra of the proteins were acquired over an m/z range of 200–4000. The results were analyzed using the Sciex OS program (Sciex, Framingham, MA, USA). High-resolution LC-MS/MS raw data for both apo HEPNLpg and NMPylated HEPNLpg are available in this paper as Supplementary Data 3.
Monitoring the stoichiometry of HEPNLpg and MNTLpg
To monitor the stoichiometry of HEPNLpg and MNTLpg, native PAGE was performed. MNTLpg was incubated with tetrameric HEPNLpg or dimeric HEPNLpg with different ratios at 20 °C for 30 min. The samples were subsequently subjected to electrophoresis on a 15% polyacrylamide gel at 120 V for 100 min in running buffer (30 mM Tris, pH 8.8, and 200 mM glycine). The results were visualized using Coomassie blue staining.
To quantify the stoichiometry of HEPNLpg and MNTLpg in solution, size exclusion chromatography (SEC) experiments were performed using an FPLC system (GE Healthcare). A Superdex 200 10/300 GL gel-filtration column (GE Healthcare) preequilibrated with a buffer containing 20 mM HEPES (pH 7.5) and 200 mM NaCl was used. The representative protein samples were injected at a flow rate of 0.4 mL/min. Standard components were injected using the same buffer and flow rate.
Statistics and reproducibility
In vitro total RNA digestion assays in Fig. 3b, In vivo expression tests in Fig. 6a, In vitro AMPylation assays in Fig. 6c–e were repeated two or three times independently with similar results.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.