Scabin was previously identified as a novel DNA-targeting mono-ADP-ribosyltransferase (mART) toxin from the plant pathogen 87.22 strain of Streptomyces scabies. Scabin is a member of the Pierisin-like subgroup of mART toxins, since it targets DNA. An in-depth characterization of both the glycohydrolase and transferase enzymatic activities of Scabin was conducted. Several protein variants were developed based on an initial Scabin·DNA molecular model. Consequently, three residues were deemed important for DNA-binding and transferase activity. Trp128 and Trp155 are important for binding the DNA substrate and participate in the reaction mechanism, whereas Tyr129 was shown to be important only for DNA binding, but was not involved in the reaction mechanism. Trp128 and Trp155 are both conserved within the Pierisin-like toxins, whereas Tyr129 is a unique substitution within the group. Scabin showed substrate specificity toward double-stranded DNA containing a single-base overhang, as a model for single-stranded nicked DNA. The crystal structure of Scabin bound to NADH — a competitive inhibitor of Scabin — was determined, providing important insights into the active-site structure and Michaelis–Menten complex of the enzyme. Based on these results, a novel DNA-binding motif is proposed for Scabin with substrate and the key residues that may participate in the Scabin·NAD(+) complex are highlighted.

Introduction

Pathogenic bacteria utilize virulence factors to recognize and gain access to host cells, provide defense against the host immune response, and to obtain nutrient stores from the host [1]. Mono-ADP-ribosyltransferase (mART) toxins are a family of enzymes that catalyze the transfer of an ADP-ribose group from NAD+ to a target macromolecule within the host cell [2,3]. In the absence of a target substrate, mART toxins may possess glycohydrolase (GH) activity, whereby a water molecule may act as the acceptor for the ADP-ribose moiety. Pathogens that utilize mART toxins can cause infection in a variety of organisms, including humans, insects, and plants. mART toxins can be further classified based on their target macromolecule. Diphtheria toxin (DT)-like members target eukaryotic elongation factor-2, whereas cholera toxin (CT)-like group targets a variety of protein substrates, including actin and the Gαs protein [46]. CT-like mART toxins possess a catalytic signature that is distinct from DT-like members [7]. The NAD+ active-site scaffold is formed by the Ser-Thr-Thr/Ser motif; a catalytic Arg forms hydrogen bonds with the phosphates of NAD+; and finally, the Gln/Glu-X-Glu catalytic motif — located in the ADP-ribosyl-turn-turn (ARTT) loop of mART toxins — is essential for catalysis [8]. Host–cell invasion by an mART toxin can lead to a change in the function of the macromolecular target, consequently ending in mutation or cell death [5,6].

Streptomyces scabies is a filamentous, soil-dwelling plant pathogen [9]. S. scabies is the causative agent of the common scab disease that affects taproot and tuberous vegetables, producing deep-pitted and corky lesions on the surface of the tuber, affecting the market value of the infected crop [10,11]. The common scab disease is of global economic importance, as there is currently no effective pesticide treatment once a field is contaminated with the pathogen. Recently, the use of biocontrol agents has been studied to help suppress S. scabies growth and reduce the progression of the common scab disease [12,13].

Scabin is a 200-residue, 22-kDa, single-domain enzyme produced and secreted by S. scabies [14,15]. Scabin was cloned, purified, and shown to possess both GH and ADP-ribosyltransferase activities. Scabin is one of only a few members of the mART toxin family that utilizes DNA as a target macromolecular substrate, and exhibits specificity toward genomic DNA from potato tubers (Solanum tuberosum). Previous bioinformatic analyses revealed that Scabin shares nearly 40% sequence identity with the Pierisin family of eukaryotic mART toxins [14]. The Pierisin family is distinguished by its unique target specificity where ADP-ribose is transferred to the guanine base in DNA, leading to host–cell apoptosis from Pieris rapae [16,17]. Scabin therefore represents the first mART toxin of bacterial origin that labels DNA as its target substrate.

Kinetic characterization of Scabin revealed a highly active enzyme (GH activity) when compared with other members of the mART family. Scabin exhibited sigmoidal kinetic behavior in the presence of the deoxyguanosine substrate, unlike the Michaelis–Menten behavior of most mART toxins. In our earlier report, we presented the first crystal structure of Scabin as a DNA-acting mART, as well as co-crystal structures of two good (lead) inhibitors of Scabin activity. A preliminary model was developed of the Scabin·NAD+·DNA complex to help guide future experiments, including mutagenesis of the active-site architecture [14].

Here, we investigate the role of several catalytic residues involved in DNA binding and enzyme function. We have determined the crystal structure of Scabin with NADH, which is a potent competitive inhibitor against the NAD+ substrate. We used this complex to shed important insights into the nature of the Scabin·NAD+ structure and to assist in the interpretation of the kinetic experiments involving Scabin catalytic variants. Based on the observed binding and kinetic data involving the wild-type (WT) and catalytic variants, we propose a DNA-binding mechanism for Scabin, representing the first evidence of a DNA-binding motif for bacterial mART toxins.

Experimental procedures

Materials

Unless otherwise stated, materials were purchased from Sigma–Aldrich and OriginPro 8.0 (OriginLab Corp., Northampton, U.S.A.) was used for data fitting and plotting.

Scabin expression and purification

Detailed expression and purification methods for Scabin toxin were conducted as described recently [14].

Circular dichroism spectroscopy

Circular dichroism (CD) spectra were acquired for Scabin WT and variants using a JASCO J-815 CD spectropolarimeter (250–190 nm scan, average of nine spectra). The proteins were at 0.16 mg/ml in a buffer containing 20 mM Tris–HCl (pH 8.2) and 50 mM NaF in a 1 mm path length rectangular quartz CD cuvette.

GH activity

Scabin (WT 50 nM; Q158A/E160A 2.3 µM; S117A 0.5 µM; W128Y 250 nM; Y129A 50 nM; W155A 50 nM; and W199A 50 nM) was incubated with increasing concentrations of ε-NAD+ (etheno-NAD+; 0–500 µM) in GH buffer containing 20 mM Tris–HCl (pH 7.9) and 50 mM NaCl. Reactions were performed in triplicate on a Cary Eclipse fluorescence spectrophotometer (Agilent Technologies, Inc., Mississauga, Canada) and monitored for 20 min. The slope of a fluorescent ε-AMP (etheno-AMP) standard curve (arbitrary units × μM−1) was used to calculate the steady rate of formation of ε-ADP-ribose (μM min−1) from the rate of the development of the fluorescence (arbitrary units min−1) during the data acquisition time. The calculated GH activity was plotted against initial ε-NAD+ concentration and fitted to a hyperbolic function (Michaelis–Menten model). Excitation and emission wavelengths were set to 305 and 405 nm, respectively, with bandpasses of 5 nm.

ADP-ribosyltransferase reaction with deoxyguanosine substrate

ADP-ribosyltransferase assays were performed as described under ‘GH activity’ with some changes. In brief, Scabin (WT 10 nM; Q158A/E160A 1.5 µM; S117A 0.1 µM; W128Y 50 nM; Y129A 10 nM; and W199A 10 nM) was incubated with 250 µM ε-NAD+ (saturating conditions) and increasing concentrations of deoxyguanosine (0–2 mM) in GH buffer containing 1% dimethyl sulfoxide. Data for Q158A/E160A and S117A were fit to the bi-dose–response model as found in OriginPro8 (eqn 1); all other data were fit to the sigmoidal dose–response model (eqn 2).

 
formula
1

where log(K0.5(1)) and log(K0.5(2)) are the log(K0.5) values of the high-affinity and low-affinity sites, respectively; p is the fraction of protein occupying the high-affinity site at saturation of the ligand, and h1 and h2 are the slopes (co-operativity) of the first and second transitions, respectively.

 
formula
2

NAD+ and NADH binding

The apparent binding constant, KD, for β-NAD+ was determined by monitoring quenching of intrinsic tryptophan fluorescence using a Cary Eclipse fluorescence spectrophotometer. In quartz UV cuvettes (0.5 × 0.5 cm), 1.25 µM of protein in 25 mM Tris–HCl (pH 8.2) and 200 mM NaCl was titrated in triplicate with increasing concentrations of β-NAD+ (0–1000 µM). Excitation and emission wavelengths were set to 295 and 340 nm, respectively, with bandpasses of 5 nm. Data were corrected for increasing reaction volume upon the addition of β-NAD+; a blank titration using N-acetyl-l-tryptophanamide (NATA) was used to correct for inner filter effects. Data were fit to a one-site binding model. Binding of WT Scabin with NADH was monitored, as described above, in the absence of β-NAD+. β-NAD+ will now be referred to as only NAD+, distinguished from ε-NAD+ used in kinetic assays.

IC50 and Ki determination for NADH

Inhibition assays with NADH were performed as described under ‘GH activity’, with some changes. In brief, 50 nM WT Scabin with 250 µM ε-NAD+ in GH buffer was titrated with increasing concentrations of NADH (0–100 µM). Reactions were performed in triplicate and data were fit to a dose–response curve to determine the IC50 value. The Ki value was determined using the experimental IC50 value and the Cheng–Prusoff equation [18], Ki = IC50/(1+[S]/KM), where [S] is the concentration of ε-NAD+ used (saturation at 250 µM) and KM (68 µM) is for the ε-NAD+ substrate.

Cyanine-3-DNA binding

Synthetic double-stranded (ds)DNA oligomers (oligomer 1: 5′-GGAAGAGAGAGAG AAAGAGAG-3′; oligomer 2: 5′-CTCTCTTTCTCTCTCTCTTCC-3′; oligomer 3: 5′-CCTCTCTTTCTCTCTC TCTTC-3′; oligomer 4: 5′-TTCTCTCTTTCTCTCTCTCTT-3′) with a 5′ cyanine-3 (Cy-3) tag on oligomer 1 were ordered from Sigma–Aldrich. Oligomers were mixed in equal molar amounts (oligomers 1 with 2 as a blunt-ended substrate; 1 with 3 for a one-base overhang; 1 with 4 for a two-base overhang) and annealed by heating to 90°C followed by cooling to 20°C at a rate of 1°C/min in a Techne TC-512 PCR (Burlington, NJ). In an ultra-micro quartz cuvette (3 mm × 3 mm), 5 µM Cy3-dsDNA in 25 mM Tris–HCl (pH 8.2) and 100 mM NaCl was titrated with increasing concentrations of WT Scabin. Additionally, 5 µM single-stranded Cy-3-tagged oligomer 1 was titrated with WT Scabin. Binding of Scabin variants W128Y, Y129A or W155A with blunt-ended substrate (double-stranded oligomers 1 and 2) was tested.

For reactions where Scabin was in complex with ADP-ribose, nicotinamide, or NADH, the protein was titrated into a quartz cuvette containing double-stranded Cy-3-tagged DNA; oligomers 1 and 3 were used to make a one-base overhang substrate. Scabin was first allowed to bind to 1.5 mM ADP-ribose, nicotinamide, or NADH to form a complex. Five micromolar double-stranded Cy-3-tagged DNA with one-base overhang and 1.5 mM ADP-ribose, nicotinamide, or NADH was titrated with the Scabin complex and the change in anisotropy was measured. The addition of nicotinamide, ADP-ribose, or NADH did not affect the relative fluorescence or the excitation and emission maxima of the Cy-3 fluorophore.

Data were collected in a ‘T-format’ configuration using a PTI Alphascan-2 spectrofluorometer (Photon Technologies, Inc., South Brunswick, NJ) equipped with a thermostatted cell holder, with the temperature held at 22°C. T-format detection simultaneously compares the fluorescence intensities of both the vertically (IVV) and horizontally (IVH) polarized emitted light, when excited with vertically polarized light. Fluorescence anisotropy is then calculated using the following equation:

 
formula
3

where the ‘G’ instrumental factor is measured as IHV/IHH, the vertically (IHV) and horizontally (IHH) polarized emitted light when excited with horizontally polarized light. Changes in anisotropy (r) of the Cy-3 fluorophore were measured for 20 s intervals, using excitation and emission wavelengths of 550 and 570 nm, respectively. Excitation and emission bandpasses were set to 4 nm.

For all titrations, the ratio of the quantum yield of Cy3-dsDNA in protein-bound (Qb) and free (Qf) states was calculated by exciting at 550 nm and scanning emission from 560 to 620 nm in 1 nm increments. Data were corrected for dilution upon the addition of protein. The concentration of bound DNA ([DNA]b) at each point of the titration was calculated using the following equation:

 
formula
4

where [E]t is the total concentration of enzyme, [DNA]t is the total concentration of DNA, and KD is the binding constant for Cy3-dsDNA. Anisotropy values were corrected for changes in the quantum yield of the Cy3 fluorophore upon binding to the protein using the following equation:

 
formula
5

where rf and rb are defined as the anisotropy of free and bound states of Cy3-dsDNA. By Ft we denote the total fluorescence at each point, defined by the following equation:

 
formula
6

Tryptophan absorbance and emission of single Trp variants

The absorbance of single Trp variants (W128Y, W155A, and W199A) was performed in a Cary50 UV–Vis spectrophotometer (Varian Instruments, Mississauga, Canada); 40 µM NATA was utilized as a control. Each variant was run in conjunction with WT ([WT] = 2.6 µM; [W128Y] = 3.2 µM; [W155A] = 2.2 µM; and [W199A] = 3.1 µM) in 25 mM Tris (pH 8.2) and 100 mM NaCl in an absorbance ultra-micro quartz cuvette (10 mm × 10 mm). The theoretical extinction coefficients for the Scabin WT and variants are: ε(WT)280 = 40 130; ε(W155A)280 = ε(W199A)280 = 34 630; ε(W128Y)280 = 36 120 (in M−1 cm−1). Concentrations were determined based on having an absorbance less than 0.05 at 295 nm, to avoid reabsorption issues (absorptive screening) at the excitation wavelength.

Emission of (n−1) Trp variants (W128Y, W155A, and W199A) was performed on a PTI-QuantaMaster spectrofluorometer with excitation wavelength of 295 nm and scanning emission from 310 to 450 nm in 1 nm increments and bandpasses set to 4 nm, and 40 µM NATA was utilized as a control. Each variant was run in conjunction with WT (at the above concentrations) in 25 mM Tris (pH 8.2) and 100 mM NaCl in a fluorescence ultra-micro quartz cuvette (3 mm × 3 mm). Quantum yield measurements were calculated using the following equation:

 
formula
7

where Qr and Q represent the quantum yields of reference (NATA, Qr = 0.14) and sample, respectively; Ir and I are the integrals of emission scan (310–450 nm) for reference and sample, respectively; and ODr and OD are the optical densities for reference and sample, respectively.

Spectral decomposition of the Scabin Trp emission fluorescence

A normalized emission spectrum of NATA centered at was used to obtain the set of parameters p that best fit an arbitrary 7° polynomial function of as the independent variable, . The parameters that defined the function f were then employed to develop four composite functions, , that correspond to the normalized emission spectra of four n−1 Scabin Trp variants, as follows:

 
formula
 
formula
 
formula
 
formula

with and representing the molar fluorescence and maximum emission wavelength length of each Trp probe, respectively. From a global fit of the four functions to the experimental emission spectra of the four proteins, with p as shared parameters and the concentration of each protein, the spectral properties ( and ) that describe each individual Trp were obtained.

Protein crystallography

Crystal conditions for Scabin were screened using the Natrix HT suite (Hampton Research, Aliso Viejo, CA) in 96-well sitting-drop screen trays. Drops were formed from 1 µl of 1.5 mg/ml protein and 1 µl of precipitant. Crystal hits were observed after 24 h in a condition containing 100 mM KCl, 50 mM sodium cacodylate trihydrate (pH 6.0), 16% PEG 1000 and 0.5 mM spermine. Scabin W128Y and W155A were crystallized in a previously determined condition by optimizing the concentration of PEG 400 [14]. Scabin W128Y was crystallized in 14% PEG 400 and 0.1 M MES (pH 6.5); Scabin W155A was crystallized in 15% PEG 400 and 0.1 M MES (pH 6.5).

The crystal condition was scaled up using 18-mm hanging drop trays, 2 µl of each protein (1.5 mg/ml) and precipitant, with 150 µl of the well solution. NADH was co-crystallized by pre-incubating 1.5 mg/ml of Scabin with 500 µM NADH for 15 min at room temperature. Equal volume of precipitant was added and allowed to incubate with the protein–ligand mixture for 10 min prior to setting up 4 µl drops. Scabin·NADH crystals were washed with mother liquor containing 500 µM NADH and 12% glycerol to act as a cryoprotectant and immediately flash-frozen in liquid nitrogen. Scabin variants were washed with stepwise increments of glycerol from 12, 13.5, and 15% glycerol and immediately flash-frozen in liquid nitrogen. X-ray diffraction data were collected at the Canadian Light Source in the Canadian Macromolecular Crystallography Facility (beamline 08ID-1).

Scabin structures

Collected data were processed in XDS [19]. Molecular replacement was performed using Phenix [20] with the Scabin-apo (PDB:5DAZ) structure as the model. Iterative cycles of model building were performed in COOT [21] and subsequent refinement in Phenix. The Scabin·NADH, Scabin W128Y, and W155A structures have been deposited in the Protein Data Bank database, with the codes 5TLB, 6APY, and 5UVQ, respectively. Table 1 gives the refinement statistics for these structures. The phase error was the following.

Table 1
Crystallographic data and refinement statistics for the Scabin–NADH complex, Scabin W128Y, and Scabin W155A.
Diffraction data Scabin–NADH Scabin W128Y Scabin W155A 
PDB ID 5TLB 6APY 5UVQ 
X-ray source CLSI-08-ID-1 CLSI-08-ID-1 CLSI-08-ID-1 
Wavelength (Å) 0.97949 0.97949 0.97949 
Resolution range (Å)1 42.33–1.7 (1.76–1.7) 37.34–1.5 (1.55–1.5) 37.86–1.6 (1.657–1.6) 
Space group CCC
Unit cell parameters (Å) a = 86.04, b = 62.26, c = 38.08,
α = 90.0, β = 100.3, γ = 90.0 
a = 88.42, b = 60.75, c = 37.87,
α = 90.00, β = 99.61, γ = 90.00 
a = 88.11, b = 60.80, c = 38.37,
α = 90.0, β = 99.355, γ = 90.0 
Redundancy 3.8 (3.7) 4.2 (4.1) 4.5 (4.4) 
Data completeness (%) 99.7 (99.8) 99.8 (99.8) 100.0 (99.9) 
CC1/2 0.999 (0.716) 0.999 (0.767) 0.999 (0.821) 
Rmeas (%) 6.0 (89.5) 5.0 (76.2) 5.5 (94.9) 
Average I/σ(I15.4 (1.8) 17.8 (2.2) 16.0 (1.7) 
Molecular replacement program Phenix Phenix Phenix 
Rwork (%)2 15.92 15.0 14.50 
Rfree (%)3 19.61 17.9 18.60 
No. of atoms in protein 1281 1308 1311 
No. of waters 139 130 129 
RMSD from ideal bond length (Å) 0.02 0.01 0.02 
Total reflections 82 203 (8084) 132 089 (12 973) 118 396 (11 385) 
Unique reflections 21 771 (2158) 31 670 (3152) 26 414 (2615) 
Reflections used in refinement 21 755 (2157) 31 656 (3149) 26 361 (2600) 
Reflections used for Rfree 1088 (108) 1583 (158) 1319 (131) 
Co-ordinate error (Å) 0.2 0.17 0.15 
RMSD from ideal bond angles (°) 1.32 0.91 1.25 
Ramachandran plot favored (%) 98 97 97 
Ramachandran plot outliers (%) 0.6 
Average B-factor 32.82 24.86 35.81 
B-factors (Å2) for Scabin 31.69 23.64 34.76 
B-factors (Å2) for water 41.92 37.06 46.49 
Diffraction data Scabin–NADH Scabin W128Y Scabin W155A 
PDB ID 5TLB 6APY 5UVQ 
X-ray source CLSI-08-ID-1 CLSI-08-ID-1 CLSI-08-ID-1 
Wavelength (Å) 0.97949 0.97949 0.97949 
Resolution range (Å)1 42.33–1.7 (1.76–1.7) 37.34–1.5 (1.55–1.5) 37.86–1.6 (1.657–1.6) 
Space group CCC
Unit cell parameters (Å) a = 86.04, b = 62.26, c = 38.08,
α = 90.0, β = 100.3, γ = 90.0 
a = 88.42, b = 60.75, c = 37.87,
α = 90.00, β = 99.61, γ = 90.00 
a = 88.11, b = 60.80, c = 38.37,
α = 90.0, β = 99.355, γ = 90.0 
Redundancy 3.8 (3.7) 4.2 (4.1) 4.5 (4.4) 
Data completeness (%) 99.7 (99.8) 99.8 (99.8) 100.0 (99.9) 
CC1/2 0.999 (0.716) 0.999 (0.767) 0.999 (0.821) 
Rmeas (%) 6.0 (89.5) 5.0 (76.2) 5.5 (94.9) 
Average I/σ(I15.4 (1.8) 17.8 (2.2) 16.0 (1.7) 
Molecular replacement program Phenix Phenix Phenix 
Rwork (%)2 15.92 15.0 14.50 
Rfree (%)3 19.61 17.9 18.60 
No. of atoms in protein 1281 1308 1311 
No. of waters 139 130 129 
RMSD from ideal bond length (Å) 0.02 0.01 0.02 
Total reflections 82 203 (8084) 132 089 (12 973) 118 396 (11 385) 
Unique reflections 21 771 (2158) 31 670 (3152) 26 414 (2615) 
Reflections used in refinement 21 755 (2157) 31 656 (3149) 26 361 (2600) 
Reflections used for Rfree 1088 (108) 1583 (158) 1319 (131) 
Co-ordinate error (Å) 0.2 0.17 0.15 
RMSD from ideal bond angles (°) 1.32 0.91 1.25 
Ramachandran plot favored (%) 98 97 97 
Ramachandran plot outliers (%) 0.6 
Average B-factor 32.82 24.86 35.81 
B-factors (Å2) for Scabin 31.69 23.64 34.76 
B-factors (Å2) for water 41.92 37.06 46.49 
1

Values in parenthesis are for the highest resolution shell.

2

|Σ||Fo| − |Fc||Σ||Fo|, where |Fo| and |Fc| are the observed and calculated structure factor amplitudes, respectively.

3

The Rfree value was calculated with a random 5% subset of all reflections excluded from refinement.

Stability of Scabin variants

The relative fold stability change of each variant with respect to WT Scabin, ΔΔG = ΔGf→ u(variant) − ΔGf → u (WT), with f: folded and u: unfolded, was calculated with the STRUM methodology [22] by submitting the apo-Scabin X-ray structure (PDB:5DAZ) to the server http://zhanglab.ccmb.med.umich.edu/STRUM/. The STRUM method predicts the fold stability change (ΔΔG) of variant proteins upon single-point mutations. A ΔΔG value below zero means that the mutation causes destabilization, whereas a value above zero means that the mutation induces stabilization in the protein.

Structure preparation for molecular mechanics calculations

Protein preparation, molecular mechanics (MM) calculations, and protein rendering were performed using the computational suite Molecular Operative Environment (MOE) release 2016.08 (Chemical Computing Group, Inc., Montreal, CA). The X-ray structure for the Scabin–NADH complex (PDB:5TLB) was protonated using the MOE Protonate3D module to assign the ionization states and tautomers of protein side-chains and to orient crystallographic water molecules (CWMs) at T = 300 K (pH 7.4) and ionic strength of 0.1 M, along with the GB-VI (Generalized Born-Volume Integral) solvation model and MMFF94 partial charges. The protonated structure was initially geometry-optimized by tethering all heavy atoms with a 100 kcal/mol force constant (0.25 Å buffer) and energy-minimized until an RMS gradient ≤0.001 kcal/mol/Å2. The force field employed was the Amber12:EHT (Extended Hückey Theory)/GB-VI, with AMBER12 parameters set (ff12) for protein and parameters calculated from the EHT for NADH, along with the GB-VI solvation model. The molecular surfaces are solvent-excluded surfaces obtained by rolling a probe sphere of 1.4 Å diameter (water radius) and colored by several schemes. The van der Waals (vdW) interaction surfaces correspond to zero-potential contours of the vdW potential, , between the considered set of atoms and a water O-atom as a mobile probe, using a standard 12-6 Lennard-Jones definition.

Modeling WT and variant Scabin complexed with NAD+

Using the Scabin·NADH preparation as a template, the various Scabin proteins (WT, Q158A/E160A, S117A, W68Y and W68A, W128Y, W155A and W199A) were modeled by ‘mutating in silico’ the selected residue to the target residue with an optimum side-chain conformation obtained from an all-atom backbone-independent MOE rotamer library. In each case, the NADH molecule was crafted in situ to NAD+. The Protonate3D protocol was performed while protecting the oxidized state of the ligand. Then, the new Scabin complexes were repacked by tethering the NAD+ molecule (10 kcal/mol, 0.25 Å buffer) and all backbone atoms and CWMs with atoms at >4.5 Å distance from the substituted residue or from the nicotinamide moiety. An energy minimization calculation was then conducted until an RMS gradient ≤0.001 kcal/mol/Å2 was achieved.

Determining percentage of free thiols in different Scabin variants

A method using monobromobimane (MBBr) was developed to quantitatively determine the presence of free thiols in cysteine residues of native Scabin and variants. A standard curve was generated with 40 µM MBBr and increasing concentrations of cysteine hydrochloride monohydrate (0–12.5 µM). The samples were incubated at room temperature for 20 min and fluorescence was measured using a FLUOstar Omega plate reader from BMG LABTECH, with excitation and emission wavelengths of 350 and 460 nm, respectively. The apo form of WT Scabin, along with six Scabin variants (W128Y, W155A, W199A, S117A, and Q158A/E160A), was subjected to reaction with MBBr, and the percentage of free thiols was determined for each respective variant. Protein (15 µM) was incubated with 40 µM MBBr at room temperature for 1.5 h and fluorescence was measured with the FLUOstar Omega plate reader, utilizing the same parameters established during determination of the standard curve.

Results

Stability of Trp variants

The primary sequence of Scabin reveals that there are four Trp residues: Trp68, Trp128, Trp155, and Trp199 [14] (Figure 1A). Trps 68 and 199 are outside of the catalytic core of Scabin and do not likely participate in substrate binding and catalysis, although Trp 68 in Scabin is conserved with Pierisins, but not for Mtx from Lysinibacillus sphericus. Notably, Trp128 and Trp155 are in the active-site core of Scabin, and may participate in transferase substrate interaction and enzyme activity. As reported previously, these two Trp residues are highly conserved in the Pierisin members of mARTs, and therefore may play a role in targeting the DNA substrate [14].

Primary sequence overview of Scabin and spectral data of Trp variants.

Figure 1.
Primary sequence overview of Scabin and spectral data of Trp variants.

(A) Sequence alignment of Scabin with select Pierisin-like toxins was generated using the T-Coffee Web server to align the sequences and ESPript to generate the alignment figure [41]. Key catalytic regions are indicated. Identical residues are shown in red, and similar residues are printed in red type. Highly conserved Trp residues are indicated with an asterisk below the alignment (Trps 68, 128, and 155). Bottom: Overview of the primary sequence, with the N-terminal leader peptide highlighted in red, illustrating Trp residue locations near catalytic motifs including the PN and ARTT loops. (B) CD spectra of Scabin W128Y (blue), W155A (red), and W199A (green) in 20 mM Tris, 50 mM NaF (pH 8.2) buffer. WT is shown in black (control). The concentration of each protein was held at 0.16 mg/ml with each spectrum representing the average of nine independent spectra. (C) Absorbance spectra for Scabin W128Y (blue), W155A (red), W199A (green) and WT (black); NATA control (orange) was prepared at 40 µM. Absorbance was measured from 245 to 350 nm, and the protein concentrations were adjusted to produce a similar A280 value for each spectrum. (D) Deconvolution of the intrinsic protein fluorescence for Scabin Trp (n−1) variants. Recorded emission spectra of the global fluorescence (brown dotted line) and fitted curved (yellow solid line) of WT, W128Y, W155A, and W199A Scabin variants, and the simulated fluorescence emission spectra from the constituent Trps, colored as follows: 68 (black), 128 (red), 155 (green), and 199 (blue). All plots were shown on the same scale for comparison. The reported concentrations correspond to estimated values (nominal ± 5%).

Figure 1.
Primary sequence overview of Scabin and spectral data of Trp variants.

(A) Sequence alignment of Scabin with select Pierisin-like toxins was generated using the T-Coffee Web server to align the sequences and ESPript to generate the alignment figure [41]. Key catalytic regions are indicated. Identical residues are shown in red, and similar residues are printed in red type. Highly conserved Trp residues are indicated with an asterisk below the alignment (Trps 68, 128, and 155). Bottom: Overview of the primary sequence, with the N-terminal leader peptide highlighted in red, illustrating Trp residue locations near catalytic motifs including the PN and ARTT loops. (B) CD spectra of Scabin W128Y (blue), W155A (red), and W199A (green) in 20 mM Tris, 50 mM NaF (pH 8.2) buffer. WT is shown in black (control). The concentration of each protein was held at 0.16 mg/ml with each spectrum representing the average of nine independent spectra. (C) Absorbance spectra for Scabin W128Y (blue), W155A (red), W199A (green) and WT (black); NATA control (orange) was prepared at 40 µM. Absorbance was measured from 245 to 350 nm, and the protein concentrations were adjusted to produce a similar A280 value for each spectrum. (D) Deconvolution of the intrinsic protein fluorescence for Scabin Trp (n−1) variants. Recorded emission spectra of the global fluorescence (brown dotted line) and fitted curved (yellow solid line) of WT, W128Y, W155A, and W199A Scabin variants, and the simulated fluorescence emission spectra from the constituent Trps, colored as follows: 68 (black), 128 (red), 155 (green), and 199 (blue). All plots were shown on the same scale for comparison. The reported concentrations correspond to estimated values (nominal ± 5%).

To understand the roles of the various Trp residues in the enzyme activity of Scabin, site-directed mutagenesis was conducted to replace them with either an Ala or Tyr (whichever produced a stable protein). It was not possible to obtain either W68A or W68Y variants; however, replacements at the three other Trp sites within Scabin were possible. Variant W199A was expressed in Escherichia coli and was a stable protein (see Supplementary Table S1 for the calculated ΔΔGfold → unfold for each Trp variant). Also, the Scabin Trp variants, W128Y and W155A, were prepared and purified and their crystal structures were determined (Table 1). The structural alignment for W128Y and W155A variants with WT Scabin reported a Cα-RMSD (root-mean-square deviation) of 0.2 and 0.3 Å, respectively. More specifically, when the Trp variants were aligned with the WT protein, the all-atom RMSD was 0.1 Å for Trp155 in the W128Y variant and was 0.8 Å for Trp128 in the W155A variant. These results indicate that minimal perturbation occurred upon Trp substitution of Trp 128 and 155 in Scabin, which was expected given the relatively exposed nature of the PN (phosphate-nicotinamide) and ARTT loops, respectively. Furthermore, CD spectroscopy was conducted to assess the effect of Trp substitutions on the folded integrity and stability of the variants in solution. Figure 1B shows the CD spectra of the Trp variants (W128Y, W155A, and W199A) and WT Scabin. Although some differences were noted, the overall effect of the Trps was not significantly perturbing.

Emission properties of Trp variants

The fluorescence emission spectra of the WT and Trp variants were all quite similar in their emission envelope and maxima (Figure 1C). However, differences are seen in the average fluorescence quantum yields, , with (Supplementary Table S1), which reflects variations of each residue in its electronic environment and/or exposure to the aqueous solution. In this regard, it is important to know whether Trp128 (in W155A) and Trp155 (in W128Y) ‘behave’ in the same way as in the WT protein. For this, taking advantage of having four global fluorescence emission spectra (three from the variants and one for WT Scabin), it was possible to spectrally decompose each global signal in the individual emission spectra of each Trp residue (Figure 1D). For a quantitative assessment of the contribution of each Trp in the corresponding n (for WT) or n−1 (for variants) composite signal, a global fit was performed for the emission spectra of the WT and variant proteins, using NATA emission as an elemental ‘waveform function’. This enabled the deconvolution of the recorded spectra for each of the participating Trp emitters (see Experimental procedures), where each was characterized by its own maximum wavelength, , and molar fluorescence, (Supplementary Table S1).

The spectral deconvolution was performed under the assumption that each Trp residue exhibits the same spectral properties regardless of the presence of the other (n−1) Trp residues, i.e. each Trp has the same fluorescence emission either in the WT Scabin or in any of the variants. The validity of the assumption and therefore the correctness of the deconvolution outcome (Supplementary Table S1) are supported by the following: (i) the deconvolution showed a ‘blue’ (solvent protected) environment surrounding Trp68 and Trp199, in comparison with the ‘red’ (solvent-exposed) environment of Trp128 and Trp155, which is compatible with their relative exposure observed in the X-ray structures; likewise, (ii) Trp199 and Trp68 have the lowest and highest molar fluorescence, , respectively, while Trp128 and Trp155 are similar to one another; and (iii) the calculated average of the molar fluorescence of each composite spectrum, , follows the same pattern among the variants as observed in the experimental average quantum yield, i.e. .

In summary, the low Cα-RMSD among the X-ray structures as well as the consistency in the CD spectra and in the fluorescence emission spectra suggest that Trp128 and Trp155 have similar average conformations and surrounding environments in the Scabin WT and variants under study. This establishes that a change in catalytic properties of a variant is due to the loss of the role of that residue and not due to structural changes inherent in the variant protein in aqueous solution.

Catalytic signature residues in Scabin GH activity

The kinetic and binding parameters for WT Scabin with the substrate ε-NAD+, a fluorescent analog of NAD+ (GH activity), are presented in Table 2. We previously developed a fluorescence-based assay using ε-NAD+ to report reliable kinetic data on the NAD+ substrate [23]. As reported recently [14], the WT enzyme bound NAD+ with low micromolar affinity (70 µM) and showed a similar KM value (68 µM), with kcat and catalytic efficiency (kcat/KM) values of 94 ± 2 min−1 and 1.4 × 106 M−1 min−1, respectively (for ε-NAD+). This indicates that Scabin is a highly active NAD+ GH enzyme when compared with most mART toxins [2]. Crystal structures of only two variants, W128Y and W155A, were obtained for this study (Table 1). However, to assist with analysis of changes between variant and WT proteins, in silico mutagenesis was performed for variants that could not be crystallized (see Experimental procedures). These variants in complex with NAD+ were prepared based on the WT Scabin·NAD+ complex model (see later).

Table 2
Kinetic parameters of Scabin WT and variants for GH activity (ε-NAD+) and NAD+ binding

Data represent the mean ± SD of n = 3 replicates.

Protein KM (μM) kcat (min−1KD (μM) kcat/KM (M−1 min−1
WT 68 ± 3 94 ± 2 70 ± 3 1.4 × 106 
Q158A/E160A 53 ± 9 0.3 ± 0.02 86 ± 7 5.8 × 103 
S117A 88 ± 14 1.3 ± 0.16 23 ± 3
435 ± 10 
1.5 × 104 
W128Y 17 ± 3 1.0 ± 0.01 10 ± 4 5.9 × 104 
W155A 55 ± 10 10 ± 0.4 43 ± 24 1.8 × 105 
Protein KM (μM) kcat (min−1KD (μM) kcat/KM (M−1 min−1
WT 68 ± 3 94 ± 2 70 ± 3 1.4 × 106 
Q158A/E160A 53 ± 9 0.3 ± 0.02 86 ± 7 5.8 × 103 
S117A 88 ± 14 1.3 ± 0.16 23 ± 3
435 ± 10 
1.5 × 104 
W128Y 17 ± 3 1.0 ± 0.01 10 ± 4 5.9 × 104 
W155A 55 ± 10 10 ± 0.4 43 ± 24 1.8 × 105 

Replacement of the catalytic Q–X–E motif with Ala (Q158A/E160A) reduced kcat by 300-fold (Table 2). By modeling the Q158A/E160A–NAD+ complex in silico, it was revealed that the smaller Ala residues probably allow a better overall vdW fit within the NAD+ pocket (Figure 2A). However, the reduced H-bond capability of the Q158A/E160A variant compared with the WT Scabin, combined with the smaller contact surface for the variant (Figure 2B), resulted in a significant reduction in the calculated interaction energy with NAD+ (Eint of −99.43 and −104.25 kcal/mol for Q158A/E160A and Q–X–E, respectively), which agrees with the reduced binding affinity observed for the Q158A/E160A variant (Table 2).

Analysis of WT vs. variant Scabin structures.

Figure 2.
Analysis of WT vs. variant Scabin structures.

Active conformations and interactions of catalytic residues of WT Scabin and variants complexed with NAD+in silico. (A) Effect of the Q158A/E160A variant on the vdW contacts with NAD+. (B) Effect of the Q158A/E160A variant on the hydrogen bond interactions with bound NAD+. (C) Effect of the S117A substitution and the direct interaction of (Ser/Ala)117 with NAD+. (D) Effect of the S117A substitution and the Ser117-mediated Tyr106−Glu160 interaction. (E) Effect of W128Y substitution on the interaction of the PN loop with NAD+. (F) The vdW interaction with Gln158. (G) Comparative effect of W155A and W128Y on Gln158. Combined rendering of (Trp/Tyr)128, (Trp/Ala)155, and WT Gln158 configurations. (H) Effect of the W155A substitution on Gln158 based on the combined rendering of (Trp/Ala) 155 and WT Gln158 configurations. In all panels, vdW surfaces were colored according to their electrostatic potential as calculated for the WT protein, with NAD+ in cyan and the C-atoms for the variants in gray. WT residues are shown in black, and C-atoms and substitution residues are shown in green.

Figure 2.
Analysis of WT vs. variant Scabin structures.

Active conformations and interactions of catalytic residues of WT Scabin and variants complexed with NAD+in silico. (A) Effect of the Q158A/E160A variant on the vdW contacts with NAD+. (B) Effect of the Q158A/E160A variant on the hydrogen bond interactions with bound NAD+. (C) Effect of the S117A substitution and the direct interaction of (Ser/Ala)117 with NAD+. (D) Effect of the S117A substitution and the Ser117-mediated Tyr106−Glu160 interaction. (E) Effect of W128Y substitution on the interaction of the PN loop with NAD+. (F) The vdW interaction with Gln158. (G) Comparative effect of W155A and W128Y on Gln158. Combined rendering of (Trp/Tyr)128, (Trp/Ala)155, and WT Gln158 configurations. (H) Effect of the W155A substitution on Gln158 based on the combined rendering of (Trp/Ala) 155 and WT Gln158 configurations. In all panels, vdW surfaces were colored according to their electrostatic potential as calculated for the WT protein, with NAD+ in cyan and the C-atoms for the variants in gray. WT residues are shown in black, and C-atoms and substitution residues are shown in green.

The S–T–S/T motif is known for its role in binding and positioning the NAD+ substrate in CT-like mART toxins. The S117A variant showed an increase in and a large reduction in the kcat value compared with WT Scabin (Table 2). In the in silico structure of the S117A·NAD+ complex, the side-chain of Ala117 preserves the weak CH-bond with the pyridinium ring (Figure 2C), while maintaining the magnitude of the steric interaction. Nevertheless, the higher of the variant is compatible with the lower total interaction energy of NAD+ in the S117A variant (Eint = −102.9 kcal/mol) in comparison with the WT (Eint = −104.3 kcal/mol). S117A also lacks potentially critical H-bond contacts with Tyr106 (Figure 2D), which might affect the stability of helix α2. Furthermore, the lower kcat for the GH activity of the S117A variant in comparison with the WT (Table 2) might be explained by the absence of the stabilizing effect of the Ser117 hydroxyl on the catalytic Glu160 (Figure 2D).

In kinetic terms, the combined action of the catalytic residues, Gln158 and Glu160, can be related by the ratio kcat(WT)/kcat(Q158A/E160A) of ∼300-fold (Table 2; i.e. setting the residual GH activity of the double Q158A/E160A variant as the baseline activity). Accordingly, the effect of individual residue substitution within Scabin on the GH activity was evaluated by their respective kcat ratio with respect to the Q158A/E160A variant and assessed by the influence on the conformation/stability of the catalytic residues and on their interaction with NAD+ (Figure 2A,B). Notably, the W128Y variant eliminates an important H-bond with the NAD+ phosphate (Figure 2E) and potentially reduces the conformational stability of Gln158 (Figure 2F), which may explain its significantly reduced kcat for GH activity (Table 2). Moreover, if W128Y obliterates the catalytic role of Gln158, and likewise S117A on Glu160 (Figure 2B), the individual participation of each catalytic residue can be assessed by the ratio kcat(W128Y)/kcat(Q158A/E160A) of ∼3-fold for Glu160, and by the ratio kcat(S117A)/kcat(Q158A/E160A) of ∼4-fold for Gln158 (Table 2). This result implies a synergistic effect in the action of both these catalytic residues in the WT toxin.

Similarly, Trp128 and Trp155 sandwich Gln158 in their active conformations (Figure 2G). Crystal structures of W128Y and W155A (Table 1) were used in conjunction with in silico mutagenesis to study the structural changes of these variants. Notably, the smaller impact of the W155A substitution [assessed by the ratio kcat(WT)/kcat(W155A) of ∼10-fold] when compared with the W128Y substitution [assessed by the ratio kcat(WT)/kcat(W128Y) of ∼100-fold] can be accounted for by a reduced interaction with Gln158, which in turn affects the stability of the Gln158 catalytic residue (Figure 2H). In other words, the major interaction surface of Trp155 with Gln158 corresponds to the methylene side-chain of the former, which is largely preserved by the W155A substitution.

Finally, the Y129A variant was tested for catalytic activity. Tyr129 was targeted due to its potential role in DNA binding and transferase activity, as predicted in our previous Scabin·DNA model [14]. From the multiple sequence alignment in Figure 1A, Tyr129 is not a conserved residue among the Pierisin members (Thr in Pierisin-1) and is therefore unique to Scabin. Y129A exhibited a of 86 μM, and a kcat and catalytic efficiency of 60 min−1 and 7.0 × 105 M−1 min−1, respectively. Overall, the changes in GH activity of the Y129A variant compared with WT activity were only minor (Table 2), which is consistent with the negligible contact of Y129 with either NAD+ (model) or NADH (X-ray structure)-bound ligands.

ADP-ribosyltransferase activity

The kinetic parameters for WT Scabin transferase activity with a model nucleoside substrate, deoxyguanosine, are given in Table 3. The WT enzyme produced a K0.5 value of 302 µM, and a kcat and catalytic efficiency of 83 min−1 and 2.8 × 105 M−1 min−1, respectively [14]. The relatively similar kcat(transferase) (83 min−1) with the (70 min−1) values (Table 2) reveal that the hydrolytic and transfer steps are kinetically concerted, without a rate-limiting step, at least for the ADP-ribosylation of the mononucleoside substrate. The kinetic plots for the Q158A/E160A and S117A variants were non-Michaelis–Menten and yielded two K0.5 values for each (Table 3). These K0.5 values consisted of a lower and higher value compared with the unique K0.5 value of the WT. The Q158A/E160A variant gave a kcat of 0.5 min−1, which is 166-fold less than WT Scabin (300-fold less for GH activity). The S117A variant showed a 36-fold loss in transferase activity compared with the WT enzyme (72-fold loss for the GH activity).

Table 3
Kinetic parameters of Scabin WT and variants for transferase activity with deoxyguanosine substrate

Data represent the mean ± SD of n = 3 replicates. Abbreviations: Not Det, no detectable transferase activity was observed.

Construct K0.5 (μM) kcat (min−1kcat/K0.5 (M−1 min−1
WT 302 ± 12 83 ± 5 2.8 × 105 
Q158A/E160A 101 ± 12 0.5 ± 0.04 5.0 × 103 
357 ± 26 1.4 × 103 
S117A 165 ± 20 2.3 ± 0.1 1.4 × 104 
634 ± 14 3.6 × 103 
Y129A 365 ± 28 254 ± 14 6.9 × 105 
W128Y (KM)4 1399 ± 167 23 ± 0.4 1.7 × 104 
W155A Not Det Not Det Not Det 
W199A (KM)1 641 ± 82 178 ± 7 2.8 × 105 
Construct K0.5 (μM) kcat (min−1kcat/K0.5 (M−1 min−1
WT 302 ± 12 83 ± 5 2.8 × 105 
Q158A/E160A 101 ± 12 0.5 ± 0.04 5.0 × 103 
357 ± 26 1.4 × 103 
S117A 165 ± 20 2.3 ± 0.1 1.4 × 104 
634 ± 14 3.6 × 103 
Y129A 365 ± 28 254 ± 14 6.9 × 105 
W128Y (KM)4 1399 ± 167 23 ± 0.4 1.7 × 104 
W155A Not Det Not Det Not Det 
W199A (KM)1 641 ± 82 178 ± 7 2.8 × 105 
1

Data displayed a hyperbolic response and thus are described by the Michaelis–Menten constant (KM) and catalytic efficiency of kcat/KM.

The Y129A variant exhibited a K0.5 of 365 µM, and a kcat and catalytic efficiency of 254 min−1 and 6.9 × 105 M−1 min−1, respectively. Overall, the kinetic properties of the Y129A variant were better than the WT enzyme, indicating that Tyr129 is not involved in the transferase reaction with the nucleoside substrate.

Notably, the Trp variants showed Michaelis–Menten kinetic behavior (transferase) with the deoxyguanosine substrate. The W128Y enzyme showed a higher KM (1399 µM; 4.6-fold increase) with a 3.6-fold loss in turnover number (23 min−1) than the WT enzyme (Table 3). Interestingly, the W155A enzyme completely lost its transferase activity, confirming its important role in substrate interaction and catalysis. Additionally, Ala replacement at Trp199 produced a mixed effect on the enzyme catalytic activity — impaired by 2-fold for the K0.5 value (641 µM), but an increased kcat value (178 min−1). Contrary to the direct effect produced by the Trp128 and Trp155 substitutions (W128Y and W155A) in the active conformation of ligand(s) and pocket residues, the effect of W199A may be allosteric in origin.

NADH binding and inhibition of Scabin GH activity

It was found that NADH is a strong competitive inhibitor against the ε-NAD+ substrate; the inhibition curve with an IC50 value of 7.2 µM (Ki = 1.5 µM, estimated using the Cheng–Prusoff equation; see Experimental procedures) is shown in Figure 3A. The binding of NADH to Scabin as monitored by Trp fluorescence quenching is shown in Figure 3B. NADH binds Scabin with an affinity of 12 µM (Figure 3B). The ∼6-fold higher binding affinity of Scabin to NADH (non-hydrolyzable) in comparison with the of ∼70 μM for the natural substrate, NAD+ (ΔΔGbind ∼0.9 kcal/mol), must reside in a stronger vdW interaction with the two out-of-plane H-atoms at the pyridine C4 position (not shown). This may be in addition to an attenuation of unfavorable electrostatic interactions with adjacent polar residues due to the lack of the charged pyridine N1-atom in NADH. Thus, based on the structural and chemical similarities and in the similar affinities for the Scabin active site, the toxin-NADH structure appears like that of NAD+, making NADH a model compound for crystallographic studies of the ES complex when the GH activity of the enzyme is too high for trapping non-hydrolyzed NAD+ substrate.

NADH and DNA binding.

Figure 3.
NADH and DNA binding.

(A) IC50 curve for NADH on the GH activity of WT Scabin. Fifty nM WT Scabin with 250 µM ε-NAD+ in GH buffer was titrated with increasing concentrations of NADH (0–100 µM). (B) NADH-binding curve for WT Scabin. Scabin (1.25 μM) was titrated with increasing concentrations of NADH (0–1000 µM) in 25 mM Tris (pH 8.2) and 200 mM NaCl. (C) Binding curves for Scabin variants with blunt-ended dsDNA (see Experimental procedures). W155A (— · — ·), W128Y (… … …), Y129A (solid square symbol) or WT (——) was titrated into 5 µM Cy3-tagged dsDNA. (D) Binding curves for Scabin with Cy3-tagged dsDNA oligomer containing +1 overhang on either termini (see Experimental procedures). Scabin was pre-incubated with 1.5 mM of either nicotinamide (○), ADP-ribose (•), NADH (□), or buffer only control (▴). dsDNA was prepared at 5 µM in 25 mM Tris–HCl (pH 8.2) and 100 mM NaCl. In the case where ligand was used, 1.5 mM of ligand (nicotinamide, ADP-ribose, or NADH) was added to DNA prior to titration of the complex. The change in anisotropy (Δr) was measured for 20 s intervals; bandpasses were set to 4 nm, with excitation and emission wavelengths of 550 and 570 nm, respectively. Data are represented as n = 3 ± SD and were collected using a QuantaMaster fluorimeter.

Figure 3.
NADH and DNA binding.

(A) IC50 curve for NADH on the GH activity of WT Scabin. Fifty nM WT Scabin with 250 µM ε-NAD+ in GH buffer was titrated with increasing concentrations of NADH (0–100 µM). (B) NADH-binding curve for WT Scabin. Scabin (1.25 μM) was titrated with increasing concentrations of NADH (0–1000 µM) in 25 mM Tris (pH 8.2) and 200 mM NaCl. (C) Binding curves for Scabin variants with blunt-ended dsDNA (see Experimental procedures). W155A (— · — ·), W128Y (… … …), Y129A (solid square symbol) or WT (——) was titrated into 5 µM Cy3-tagged dsDNA. (D) Binding curves for Scabin with Cy3-tagged dsDNA oligomer containing +1 overhang on either termini (see Experimental procedures). Scabin was pre-incubated with 1.5 mM of either nicotinamide (○), ADP-ribose (•), NADH (□), or buffer only control (▴). dsDNA was prepared at 5 µM in 25 mM Tris–HCl (pH 8.2) and 100 mM NaCl. In the case where ligand was used, 1.5 mM of ligand (nicotinamide, ADP-ribose, or NADH) was added to DNA prior to titration of the complex. The change in anisotropy (Δr) was measured for 20 s intervals; bandpasses were set to 4 nm, with excitation and emission wavelengths of 550 and 570 nm, respectively. Data are represented as n = 3 ± SD and were collected using a QuantaMaster fluorimeter.

Scabin crystal structure with NADH

The structure of recombinant Scabin was previously reported at 1.50 Å resolution in the apo (substrate-free) form and in complex with active-site inhibitors [14]. Herein, we solved the structure of Scabin with NADH, as a substrate analog of NAD+, providing a reasonable model for the Scabin·NAD+ Michaelis complex. Table 1 presents the refinement statistics for the Scabin·NADH complex.

Figure 4A shows the structure of Scabin with NADH bound in the active site along with the key catalytic residues (Arg77, Ser117, Gln158 and Glu160) and two important disulfide bonds (C42–C72; C176–C190, shown in red). The NADH electron density map is shown as an inset to Figure 4A, demonstrating that this substrate analog was well resolved in the crystal structure. This competitive inhibitor/ligand was surprisingly stable when bound to Scabin and was not oxidized to NAD+ during the crystallization process. The disulfide bridge between Cys42 and Cys72 exhibited some unrefined electron density, suggesting an alternate, reduced conformation in the NADH-bound state; this unrefined electron density was not observed in the apo form. Notably, we cannot discount the possibility that NADH (E°′ = −0.28 V) could be acting as a reducing agent for this disulfide bridge (E°′ = −0.22 V). Using AreaIMol, the solvent accessible surface area was calculated for each atom of the Scabin·NADH complex [24]. The sulfur atoms of the C42–C72 bridge are only slightly more solvent-exposed than the C176–C190 sulfur atoms, with the latter not exhibiting unrefined electron density. We tested the hypothesis that Scabin could exist in both oxidized and reduced forms by measuring the percentage of free thiols. WT Scabin was found to have 9.0% free thiols, notably in the absence of NADH or other substrates (Supplementary Table S2).

Scabin binds competitive inhibitor/substrate analog NADH.

Figure 4.
Scabin binds competitive inhibitor/substrate analog NADH.

(A) Cartoon representation of Scabin (gray) bound to NADH (cyan; represented as sticks). Conserved residues are shown as sticks: Arg77 (purple), Ser117 (blue), and the catalytic Gln158-X-Glu160 (green). The two disulfide bridges (Cys42–Cys72 and Cys176–Cys190) are shown as sticks in red. Inset is the electron density map (2mFo − nFc) around the NADH ligand, contoured at 1σ. (B) Induced fit of the Scabin active site for pocket residues upon NADH binding. Depiction of the pocket residues that change conformation between the apo (dark gray C-atoms, PDB:5DAZ) and the complexed (green C-atoms, PDB:5TLB) forms of Scabin (gray molecular surface) with the NADH ligand (black C-atoms and molecular surface colored by electrostatic potential). (C) Non-bonded interactions between NADH and Scabin. Active conformations of the ScabinNADH complex showing (top) the network of direct H-bonds between pocket residues (black C-atoms) and NADH (cyan C-atoms); and (bottom) the steric interaction surface (EvdW = 0 kcal/mol) spectrally colored according to the electrostatic potential around the NADH molecule (green C-atoms). (D) Scabin active-site in-pocket water molecules. A 2D diagram depicting the presence of five CWMs that either bridge the NADH molecule to pocket residues, or stabilize the active conformation of the ligand. (E) Role of CWMs in stabilizing NADH and the ScabinNADH complex. Changes in the configuration of in-pocket water molecules between the apo form (yellow molecules, PDB:5DAZ) and bound form (brown molecules, PDB:5TLB) upon NADH binding (cyan C-atoms). (F) Induced fit of Trp128 and HOH171. Reconfiguration of Trp128 and HOH171 between the apo form (green C-atoms, PDB:5DAZ) and the bound form (yellow C-atoms, PDB:5TLB) upon NADH binding. The rotation of the Trp128 indole group increases contacts with NADH (purple shells) and displaces HOH171 to intramolecularly bridge the NADH ligand. Hydrogen atoms shown were modeled with Phenix.

Figure 4.
Scabin binds competitive inhibitor/substrate analog NADH.

(A) Cartoon representation of Scabin (gray) bound to NADH (cyan; represented as sticks). Conserved residues are shown as sticks: Arg77 (purple), Ser117 (blue), and the catalytic Gln158-X-Glu160 (green). The two disulfide bridges (Cys42–Cys72 and Cys176–Cys190) are shown as sticks in red. Inset is the electron density map (2mFo − nFc) around the NADH ligand, contoured at 1σ. (B) Induced fit of the Scabin active site for pocket residues upon NADH binding. Depiction of the pocket residues that change conformation between the apo (dark gray C-atoms, PDB:5DAZ) and the complexed (green C-atoms, PDB:5TLB) forms of Scabin (gray molecular surface) with the NADH ligand (black C-atoms and molecular surface colored by electrostatic potential). (C) Non-bonded interactions between NADH and Scabin. Active conformations of the ScabinNADH complex showing (top) the network of direct H-bonds between pocket residues (black C-atoms) and NADH (cyan C-atoms); and (bottom) the steric interaction surface (EvdW = 0 kcal/mol) spectrally colored according to the electrostatic potential around the NADH molecule (green C-atoms). (D) Scabin active-site in-pocket water molecules. A 2D diagram depicting the presence of five CWMs that either bridge the NADH molecule to pocket residues, or stabilize the active conformation of the ligand. (E) Role of CWMs in stabilizing NADH and the ScabinNADH complex. Changes in the configuration of in-pocket water molecules between the apo form (yellow molecules, PDB:5DAZ) and bound form (brown molecules, PDB:5TLB) upon NADH binding (cyan C-atoms). (F) Induced fit of Trp128 and HOH171. Reconfiguration of Trp128 and HOH171 between the apo form (green C-atoms, PDB:5DAZ) and the bound form (yellow C-atoms, PDB:5TLB) upon NADH binding. The rotation of the Trp128 indole group increases contacts with NADH (purple shells) and displaces HOH171 to intramolecularly bridge the NADH ligand. Hydrogen atoms shown were modeled with Phenix.

Upon NADH binding, the overall substrate pocket architecture of the enzyme is preserved with respect to the apo form (PDB:5DAZ), according to the small Cα-RMSD of 0.27 Å (for 24 residues). This fluctuation in the pocket backbone atoms might have its origin in the intrinsic dynamics of the protein since it has the same magnitude as in the entire protein (Cα-RMSD = 0.27 Å for 165 residues). However, the side-chains of pocket residues are conformationally shifted (RMSD = 0.94 Å, 24 residues), as observed mainly for Arg81, Lys94, Asn110, Trp128 and the catalytic Gln158 (RMSD = 1.76 Å for all atoms of five residues) (Figure 4B). This is also observed in the water-co-ordinated structure (see later). For the catalytic Gln158, two alternate conformations of this side-chain were observed, one that is identical with the apo form and one that is distinct.

A significant conformational change upon NADH binding involves the Trp128 side-chain; it appears to have rotated nearly 180°, shifting the nitrogen of the indole ring 4 Å. The shifts in side-chain location upon NADH binding without changing backbone orientation could explain the relatively high GH catalytic efficiency (1.4 × 106 M−1 min−1) (Table 2). For example, upon NAD+ binding, the ARTT loop of C3 toxins usually displays large shifts in conformation, signifying an ‘in’ and ‘out’ phase of the loop; reported catalytic efficiencies are orders of magnitudes lower than Scabin for GH activity (C3larvin = 11 M−1 min−1; C3cer = 2.1  × 105 M−1 min−1), which suggests that the large structural changes that C3 toxins employ during catalysis add to their inefficiency as enzymes [2528].

NADH is co-ordinated in the Scabin active site by a network of hydrogen bonds (Figure 4C, top) and steric contacts (Figure 4C, bottom), with some similarities and differences in the pattern of interactions described for the bound NAD+ substrate in other mART toxins [2931]. The nicotinamide amide group is anchored by two reciprocal H-bonds with the backbone of Ser78 (at β1), which remarkably is a unique substitution for this position into the CT and DT groups (usually at Gly, but is a Trp in the Pierisin-like and MTX toxins), and is only found in Sco5461, a DNA-targeting mART from Streptomyces coelicolor [32].

Residues of the S117–T118–T119 motif, Trp128 (in the PN loop), and Leu124, together form a flat hydrophobic pocket for the nicotinamide moiety. However, there is room for two ‘out-of-plane’ H-atoms at C4 of the pyridine ring due to the trans location of Thr118, which is expected to be similar in conformation for bound NAD+. The catalytic Q158-X-E160 motif (at B5-ARTT) interacts with the N-ribose, while both conserved Arg77 and Asn110 residues contact the PO2–O–PO2 linker, assisted by Lys94 via a salt bridge (Figure 4C). At the adenine side of NADH, Lys94 and Arg81 stack with the adenine moiety, and Ser80 H-bonds the A-ribose.

Ten CWMs observed in the apo form are displaced upon NADH binding to Scabin. Other in-pocket CWMs are used to stabilize the ligand or to bridge it to the protein (Figure 4D). HOH39 (apo-Scabin numbering) shifts and rotates to intramolecularly bridge the NADH ligand which helps stabilize the binding pose (Figure 4E). Indeed, bound NADH has its highest strain energy at the adenine moiety. In addition, the A-phosphate is bridged to Asn110 by the displacement of HOH181 and to Ser117 via the rotation of HOH41 (Figure 4E).

A case that deserves special attention corresponds to HOH171 and Trp128 (Figure 4F). Trp128 adopts a different conformation from that of apo-Scabin. The PN-loop region is highly disordered, exhibiting poor electron density throughout. Analyzing the CWMs of both the apo and bound crystal structures was required to determine the significance of this change. Trp128 weakly interacts with neighboring residues in the apo form (Eint = −1.5 kcal/mol), and it rotates upon NADH binding. This results in an increase in the steric contact with NADH (Eint = −6.3 kcal/mol, purple shells in Figure 4F), inducing a shift in HOH171 leading to a greater stabilization by co-ordination with NADH. In summary, the active conformation of Trp128 is compatible with the bonded location of HOH171 and with the active pose of NADH, providing evidence to support its conformation represented in the crystal structure.

Scabin·DNA binding

The ability of WT Scabin to bind the DNA substrate was measured with a synthetic dsDNA tagged at the 5′-termini with a Cy-3 fluorophore. Affinity of W128Y, Y129A, and W155A variants for blunt-ended dsDNA was measured by employing a fluorescence anisotropy binding assay, as described in Experimental procedures (Figure 3C). WT Scabin produced a KD of 51 µM with the dsDNA substrate, whereas the W128Y variant exhibited 9-fold weaker affinity for dsDNA with a KD of 458 µM (Table 4). Interestingly, Y129A showed the most significant loss in dsDNA binding; although some binding was detected, the data could not be accurately fitted to our model since we could not accurately predict saturation. The KD for the W155A variant was determined as 167 µM, a 3-fold weaker affinity than that of WT. Despite showing no transferase activity using deoxyguanosine as a substrate, W155A was still able to bind dsDNA.

Table 4
Binding constants of Scabin WT and variants with blunt-end dsDNA

Data represent the mean ± SD of n = 3 replicates.

Scabin protein KD (μM) 
WT 51 ± 4 
W128Y 458 ± 119 
Y129A Not Det 
W155A 167 ± 21 
Scabin protein KD (μM) 
WT 51 ± 4 
W128Y 458 ± 119 
Y129A Not Det 
W155A 167 ± 21 

Abbreviations: Not detected, no significant binding was observed.

In relation to the conformation of the dsDNA substrate, Scabin exhibited an increase in affinity for dsDNA with a single-base overhang on either terminus, in comparison with the blunt-ended dsDNA (Figure 3D and Table 5). The addition of a second-base overhang did not show a significant increase in affinity from the single-base overhang substrate, indicating a preference for binding dsDNA with a single-base overhang. Single-stranded DNA was also tested, but no change in affinity compared with blunt-ended dsDNA was observed (data not shown).

Table 5
Binding constants of Scabin WT for different synthetic dsDNA

Data represent the mean ± SD of n = 3 replicates.

dsDNA KD (μM) KD (μM) with 1.5 mM nicotinamide KD (μM) with 1.5 mM ADP-ribose KD (μM) with 1.5 mM NADH 
Blunt end 51 ± 4 − − − 
+1 overhang 15 ± 2 29 ± 1 21 ± 1 ND 
dsDNA KD (μM) KD (μM) with 1.5 mM nicotinamide KD (μM) with 1.5 mM ADP-ribose KD (μM) with 1.5 mM NADH 
Blunt end 51 ± 4 − − − 
+1 overhang 15 ± 2 29 ± 1 21 ± 1 ND 

Abbreviations: ND, no detectable binding was observed.

The ability of Scabin to bind dsDNA with an overhang was also tested in the presence of various substrate fragments or analogs (ADP-ribose, nicotinamide, or NADH). The KD of Scabin with dsDNA, when in complex with nicotinamide or ADP-ribose, was determined to be 21 and 29 µM, respectively (Figure 3D and Table 5). However, no detectable binding was observed when Scabin was in complex with NADH, which may find its explanation in the lack of a pyridinium cation within NADH compared with the natural NAD+ substrate. Consequently, there may be dynamic differences between the Scabin·NAD+ and Scabin·NADH complexes regarding the binding of the DNA substrate. For instance, the charged nature of the pyridine moiety in the bound NAD+ substrate may affect relevant residues (Trp128 and Trp155) in a binding conformation not achievable by the bound NADH inhibitor. Accordingly, the conformation of Trp128 is different between the apo and NADH-bound forms; and likewise, it could also be different for the NAD+-bound form. If so, the conformation of the DNA-related pair Trp128–Tyr129 might indeed depend upon the ligation state.

Proposed binding mode for DNA with Scabin

Overall, the in-depth kinetic study of Scabin revealed three key residues involved in transferase substrate binding and catalysis. Trp128 and Tyr129 are both required for efficient DNA binding; however, they are not required for ADP-ribosylation of the small, deoxyguanosine substrate. Trp128 and Tyr129 are both located in the PN loop, a region that is proposed to interact with the transferase substrate for some mART toxins [33]. Trp155 — located in the ARTT loop — is absolutely required for transferase activity, as shown by the lack of any detectable transferase activity for the W155A variant.

Based on these results, a mode of DNA substrate binding for Scabin is proposed, which represents the first insight on a recognition motif for DNA-targeting bacterial mART toxins (Figure 5). Trp128 and Tyr129 interact with adjacent nucleotides to the guanine nucleophile, allowing Trp155 to dock the target guanine base. The target guanine then moves into position and is recognized by Gln158, like the proposed recognition of Asn by Gln212 of C3exoenzyme [34]. Gln 158 allows for the guanine nucleophile to be placed near the glycosidic bond of NAD+, ready for ADP-ribosylation. Glu160 stabilizes the oxacarbenium ion intermediate, whereby the C1 of N-ribose undergoes nucleophilic attack by the N2 exocyclic amine of the guanine base [34]. The mechanism ends with the formation of an ADP-ribosylated guanine base within DNA. Additionally, Scabin probably binds single-stranded breaks in dsDNA, similarly to poly-ADP-ribosylpolymerase-1 (PARP-1) [35]. However, Scabin possesses no sequence or structural similarities to PARP-1 beyond the classic catalytic motifs present in ADP-ribosyltransferases. Therefore, Scabin is clearly a unique enzyme among members of the bacterial mART family.

Scabin–NAD+–DNA model.

Figure 5.
Scabin–NAD+–DNA model.

A 2D schematic model of key interacting residues with the DNA and NAD+ substrates. Arg77, Asn110, and Glu160 are depicted to interact with NAD+. Trp128 and Tyr129 interact with an adjacent nucleotide (depicted as dA: adenosine) to the reactive guanine base (dG), which interacts with Gln158 and docks with Trp155. Proposed π-stacking interactions are depicted as gray dashes; hydrogen bonds are shown as black dashes. The reactive center at the C1 of N-ribose is indicated by the red arrow. The NH2 nucleophile is highlighted by the red circle.

Figure 5.
Scabin–NAD+–DNA model.

A 2D schematic model of key interacting residues with the DNA and NAD+ substrates. Arg77, Asn110, and Glu160 are depicted to interact with NAD+. Trp128 and Tyr129 interact with an adjacent nucleotide (depicted as dA: adenosine) to the reactive guanine base (dG), which interacts with Gln158 and docks with Trp155. Proposed π-stacking interactions are depicted as gray dashes; hydrogen bonds are shown as black dashes. The reactive center at the C1 of N-ribose is indicated by the red arrow. The NH2 nucleophile is highlighted by the red circle.

Discussion

The present study reports a novel kinetic, structure–function analysis of Scabin, a putative DNA-acting mART toxin [14]. Scabin is produced by S. scabies as a secreted protein that may participate as a virulence factor in the development of the common scab disease in tuberous crops. Previously, we showed that Scabin was capable of utilizing various guanine-containing molecules as substrates, including mononucleosides, nucleotides, and both single-stranded and dsDNA [14]. Scabin exhibited co-operative transferase activity toward deoxyguanosine, for which we suggested a second binding site. Understanding the kinetic mechanism of Scabin with DNA as a true transferase substrate is therefore of interest in future studies. We tested Scabin transferase activity toward genomic DNA and found that the enzyme exhibited high activity levels toward gDNA from potato, with very little to no detectable activity toward gDNA from the producing organism, S. scabies. Interestingly, Scabin did not exhibit activity toward gDNA unless it had been previously treated with DNase to create nicks in the duplex structure [14].

The observed co-operativity of Scabin for the deoxyguanosine model substrate may be due to oligomerization of Scabin in the presence of the transferase substrate, or it may be explained by a hysteric enzyme model. One-site binding of monomeric enzymes that display co-operativity is described as ‘hysteric enzymes’, a term developed by Frieden [19]. Hysteric enzymes are characterized by their apparent time-lag in response to a rapid change in ligand concentration, often due to structural changes and sampling of different conformations in the enzyme. These alternate conformations can either (1) be induced by binding a substrate, termed mnemonic model [36,37], or (2) be inherently present in equilibrium, where one conformation is favored as ligand concentration increases, termed ligand-induced slow-transition model (LIST) [38,39]. The mnemonic model states that the enzyme exhibits two conformations, a low- and high-affinity state; in the absence of substrate, the enzyme strongly favors the low-affinity state, producing the apparent ‘lag’ in activity as the protein transitions to the high-affinity state after binding the substrate [36,37]. As substrate is added, the enzyme transitions to the high-affinity state and slowly returns to the low-affinity state when the product is released. As substrate concentration increases, the conformation equilibrium is pushed to favor the high-affinity state and the enzyme will remain in the high-affinity conformation as it releases product when the substrate is in excess. In contrast, the LIST model states that there is already an equilibrium between the low- and high-affinity states in the absence of substrate, where each conformation can complete a separate catalytic cycle; as concentration increases, the high-affinity state becomes more favored [38,39].

Notably, the transferase activity of WT Scabin is marginally higher than the enzyme's GH activity. However, when compared with other mART toxins, Scabin is a better transferase enzyme. The transferase activity is 17- and 8300-fold higher compared with that of C3larvin (kcat(RhoA) = 5 min−1 [27]) and Vis toxin (kcat(agmatine) = 0.01 min−1 [31]), respectively.

The Scabin–NADH complex represents the first substrate (analog)-bound crystal structure of a DNA-targeting bacterial mART toxin. This structure reveals a partially reduced conformation of the N-terminal disulfide bridge (Cys42–Cys72), which raises the possibility of alternate conformations of the enzyme in solution that might be important for function. The presence of alternate conformations would explain the biphasic nature of the enzymatic activity observed for some Scabin variants. Notably, the cellular compartment for activity of this toxin would be a reducing environment, and thus studying the significance of these non-conserved disulfide bridges is of importance.

Three residues in the Scabin active-site pocket have been identified as important for DNA binding and transferase activity, representing the first residues involved in DNA substrate recognition that have been identified within the bacterial mART toxin family. Trp128 and Tyr129 are both important for binding and docking the DNA substrate, whereas Trp155 is absolutely required for transferase activity with the deoxyguanosine nucleophile. Using these data, we proposed a DNA-binding mechanism for Scabin, based on our kinetic and structural data.

Oda et al. [40] recently solved the crystal structures of the E165Q variant of the N-terminal catalytic domain (res 1–233) of Pierisin-1 from the cabbage butterfly P. rapae with and without β-NAD+ substrate bound in the active site. Pierisin-1 exclusively modifies duplex DNA, resulting in apoptosis of various cancer cells. They found that the catalytic domain has a defined, positively charged region on the protein surface. It was also shown that Pierisin-1 only binds dsDNA and that the PN loop and a ‘basic cleft’ region are important for DNA binding. The catalytic core structure of Scabin is like most Pierisins, MTX (L. sphericus) and other ADP-ribosyltransferases, in general. However, the S. scabies genome possesses a separate Scabin B domain (gene 27 781) flanking the catalytic A-domain. The B-domain features a YVTN β-propeller domain that harbors a twin arginine translocation signal profile, a leucine-rich repeat region, and an NHL repeat profile. The role and function of the Scabin B-domain is not currently known. Unlike Pierisin-1, Scabin has neither a ricin-B-like domain nor an autoinhibitory linker, suggesting that enzyme regulation/control is different for Scabin and Pierisin-1. Additionally, the arrangement of α-helices and loops around the active-site core region is different between Scabin and Pierisin-1. The key residues for DNA binding are not conserved between Pierisin-1 and Scabin. Furthermore, Scabin binds ssDNA with similar affinity to dsDNA, indicating that the binding mode and preference for DNA and mononucleotide substrates are different from Scabin and Pierisin-1.

Future studies of Scabin will focus on characterizing the Scabin–DNA complex, and determining if there is a specific target gene within the potato host cell. If a specific gene target can be identified, Scabin may represent the first bacterial mART toxin that can modulate gene expression of the host cell as a mode of cytotoxicity. A further study of Scabin features and mechanisms will provide important new insights into the pathogenesis associated with DNA-targeting mART toxins and may prove helpful in understanding tuberous crop diseases such as the common scab in potatoes.

Database depositions

Uniprot accession no. for Scabin sequence, C9Z6T8; PDB ID for Scabin·NADH, 5TLB; Scabin W155A, 5UVQ; and Scabin W128Y, 6APY.

Abbreviations

     
  • ε-NAD+

    etheno-NAD+

  •  
  • ARTT

    ADP-ribosyl-turn-turn

  •  
  • CD

    circular dichroism

  •  
  • CT

    cholera toxin

  •  
  • CWM

    crystallographic water molecules

  •  
  • Cy-3

    cyanine-3

  •  
  • DT

    diphtheria toxin

  •  
  • GB-VI

    Generalized Born-Volume Integral

  •  
  • GH

    glycohydrolase

  •  
  • LIST

    ligand-induced slow-transition model

  •  
  • mART

    mono-ADP-ribosyltransferase

  •  
  • MBBr

    monobromobimane

  •  
  • MM

    molecular mechanics

  •  
  • NATA

    N-acetyl-l-tryptophanamide

  •  
  • PARP-1

    poly-ADP-ribosylpolymerase-1

  •  
  • PN

    phosphate-nicotinamide

  •  
  • RMSD

    root-mean-square deviation

  •  
  • SD

    standard deviation

  •  
  • vdW

    van der Waals

  •  
  • WT

    wild type

Author Contribution

A.R.M. conceived the project. B.L., M.R.L., S.C. and T.L. performed the experiments. B.L., M.R.L. and A.R.M. wrote the paper.

Funding

This work was supported by the Natural Sciences and Engineering Research Council of Canada (to A.R.M.). B.L. was supported by a Natural Sciences and Engineering Research Council graduate scholarship. X-ray data for this research were collected at Beamline 08ID-1 at the Canadian Light Source, which is supported by the Natural Sciences and Engineering Research Council of Canada, the National Research Council Canada, the Canadian Institutes of Health Research, the Province of Saskatchewan, Western Economic Diversification Canada and the University of Saskatchewan.

Acknowledgments

We thank Tom Keeling for contributions to the cloning of the Scabin gene and building various constructs and for technical support. We are also grateful to Kayla Heney for technical support and preparation of general chemicals and supplies.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Rahme
,
L.G.
,
Ausubel
,
F.M.
,
Cao
,
H.
,
Drenkard
,
E.
,
Goumnerov
,
B.C.
,
Lau
,
G.W.
et al. 
(
2000
)
Plants and animals share functionally common bacterial virulence factors
.
Proc. Natl Acad. Sci. U.S.A.
97
,
8815
8821
2
Yates
,
S.P.
,
Jørgensen
,
R.
,
Andersen
,
G.R.
and
Merrill
,
A.R.
(
2006
)
Stealth and mimicry by deadly bacterial toxins
.
Trends Biochem. Sci.
31
,
123
133
3
Holbourn
,
K.P.
,
Shone
,
C.C.
and
Acharya
,
K.R.
(
2006
)
A family of killer toxins. Exploring the mechanism of ADP-ribosylating toxins
.
FEBS J.
273
,
4579
4593
4
Lang
,
A.E.
,
Schmidt
,
G.
,
Schlosser
,
A.
,
Hey
,
T.D.
,
Larrinua
,
I.M.
,
Sheets
,
J.J.
et al. 
(
2010
)
Photorhabdus luminescens toxins ADP-ribosylate actin and RhoA to force actin clustering
.
Science
327
,
1139
1142
5
Collier
,
R.J.
(
2001
)
Understanding the mode of action of diphtheria toxin: a perspective on progress during the 20th century
.
Toxicon
39
,
1793
1803
6
Aktories
,
K.
,
Lang
,
A.E.
,
Schwan
,
C.
and
Mannherz
,
H.G.
(
2011
)
Actin as target for modification by bacterial protein toxins
.
FEBS J.
278
,
4526
4543
7
Simon
,
N.C.
,
Aktories
,
K.
and
Barbieri
,
J.T.
(
2014
)
Novel bacterial ADP-ribosylating toxins: structure and function
.
Nat. Rev. Microbiol.
12
,
599
611
8
Masignani
,
V.
,
Balducci
,
E.
,
Serruto
,
D.
,
Veggi
,
D.
,
Aricò
,
B.
,
Comanducci
,
M.
et al. 
(
2004
)
In silico identification of novel bacterial ADP-ribosyltransferases
.
Int. J. Med. Microbiol.
293
,
471
478
9
Lerat
,
S.
,
Simao-Beaunoir
,
A.-M.
and
Beaulieu
,
C.
(
2009
)
Genetic and physiological determinants of Streptomyces scabies pathogenicity
.
Mol. Plant. Pathol.
10
,
579
585
10
Goyer
,
C.
and
Beaulieu
,
C.
(
1997
)
Host range of Streptomycete strains causing common scab
.
Plant Disease
81
,
901
904
11
Williams
,
S.T.
,
Locci
,
R.
,
Beswick
,
A.
,
Kurtboke
,
D.I.
,
Kuznetsov
,
V.D.
,
Le Monnier
,
F.J.
et al. 
(
1993
)
Detection and identification of novel actinomycetes
.
Res. Microbiol.
144
,
653
656
12
Wanner
,
L.A.
,
Kirk
,
W.W.
and
Qu
,
X.S.
(
2014
)
Field efficacy of nonpathogenic Streptomyces species against potato common scab
.
J. Appl. Microbiol.
116
,
123
133
13
St-Onge
,
R.
,
Gadkar
,
V.J.
,
Arseneault
,
T.
,
Goyer
,
C.
and
Filion
,
M.
(
2011
)
The ability of Pseudomonas sp. LBUM 223 to produce phenazine-1-carboxylic acid affects the growth of Streptomyces scabies, the expression of thaxtomin biosynthesis genes and the biological control potential against common scab of potato
.
FEMS Microbiol. Ecol.
75
,
173
183
14
Lyons
,
B.
,
Ravulapalli
,
R.
,
Lanoue
,
J.
,
Lugo
,
M.R.
,
Dutta
,
D.
,
Carlin
,
S.
et al. 
(
2016
)
Scabin, a novel DNA-acting ADP-ribosyltransferase from Streptomyces scabies
.
J. Biol. Chem.
291
,
11198
11215
15
Joshi
,
M.V.
,
Mann
,
S.G.
,
Antelmann
,
H.
,
Widdick
,
D.A.
,
Fyans
,
J.K.
,
Chandra
,
G.
et al. 
(
2010
)
The twin arginine protein transport pathway exports multiple virulence proteins in the plant pathogen Streptomyces scabies
.
Mol. Microbiol.
77
,
252
271
16
Watanabe
,
M.
,
Kono
,
T.
,
Matsushima-Hibiya
,
Y.
,
Kanazawa
,
T.
,
Nishisaka
,
N.
,
Kishimoto
,
T.
et al. 
(
1999
)
Molecular cloning of an apoptosis-inducing protein, pierisin, from cabbage butterfly: possible involvement of ADP-ribosylation in its activity
.
Proc. Natl Acad. Sci. U.S.A.
96
,
10608
10613
17
Takamura-Enya
,
T.
,
Watanabe
,
M.
,
Totsuka
,
Y.
,
Kanazawa
,
T.
,
Matsushima-Hibiya
,
Y.
,
Koyama
,
K.
et al. 
(
2001
)
Mono(ADP-ribosyl)ation of 2′-deoxyguanosine residue in DNA by an apoptosis-inducing protein, pierisin-1, from cabbage butterfly
.
Proc. Natl Acad. Sci. U.S.A.
98
,
12414
12419
18
Cheng
,
Y.-C.
and
Prusoff
,
W.H.
(
1973
)
Relationship between the inhibition constant (KI) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction
.
Biochem. Pharmacol.
22
,
3099
3108
19
Frieden
,
C.
(
1970
)
Kinetic aspects of regulation of metabolic processes. The hysteretic enzyme concept
.
J. Biol. Chem.
245
,
5788
5799
PMID:
[PubMed]
20
Adams
,
P.D.
,
Afonine
,
P.V.
,
Bunkóczi
,
G.
,
Chen
,
V.B.
,
Davis
,
I.W.
,
Echols
,
N.
et al. 
(
2010
)
PHENIX: a comprehensive Python-based system for macromolecular structure solution
.
Acta Crystallogr. D Biol. Crystallogr.
66
,
213
221
21
Emsley
,
P.
and
Cowtan
,
K.
(
2004
)
Coot: model-building tools for molecular graphics
.
Acta Crystallogr. D Biol. Crystallogr.
60
,
2126
2132
22
Quan
,
L.
,
Lv
,
Q.
and
Zhang
,
Y.
(
2016
)
STRUM: structure-based prediction of protein stability changes upon single-point mutation
.
Bioinformatics
32
,
2936
2946
23
Armstrong
,
S.
and
Merrill
,
A.R.
(
2001
)
Application of a fluorometric assay for characterization of the catalytic competency of a domain III fragment of Pseudomonas aeruginosa exotoxin A
.
Anal. Biochem.
292
,
26
33
24
Winn
,
M.D.
,
Ballard
,
C.C.
,
Cowtan
,
K.D.
,
Dodson
,
E.J.
,
Emsley
,
P.
,
Evans
,
P.R.
et al. 
(
2011
)
Overview of the CCP4 suite and current developments
.
Acta Crystallogr. D Biol. Crystallogr.
67
,
235
242
25
Ménétrey
,
J.
,
Flatau
,
G.
,
Boquet
,
P.
,
Menez
,
A.
and
Stura
,
E.A.
(
2008
)
Structural basis for the NAD-hydrolysis mechanism and the ARTT-loop plasticity of C3 exoenzymes
.
Protein Sci.
17
,
878
886
26
Visschedyk
,
D.
,
Rochon
,
A.
,
Tempel
,
W.
,
Dimov
,
S.
,
Park
,
H.-W.
and
Merrill
,
A.R.
(
2012
)
Certhrax toxin, an anthrax-related ADP-ribosyltransferase from Bacillus cereus
.
J. Biol. Chem.
287
,
41089
41102
27
Krska
,
D.
,
Ravulapalli
,
R.
,
Fieldhouse
,
R.J.
,
Lugo
,
M.R.
and
Merrill
,
A.R.
(
2015
)
C3larvin toxin, an ADP-ribosyltransferase from Paenibacillus larvae
.
J. Biol. Chem.
290
,
1639
1653
28
Lugo
,
M.R.
,
Ravulapalli
,
R.
,
Dutta
,
D.
and
Merrill
,
A.R.
(
2016
)
Structural variability of C3larvin toxin. Intrinsic dynamics of the alpha/beta fold of the C3-like group of mono-ADP-ribosyltransferase toxins
.
J. Biomol. Struct. Dyn.
34
,
2537
2560
29
Evans
,
H.R.
,
Sutton
,
J.M.
,
Holloway
,
D.E.
,
Ayriss
,
J.
,
Shone
,
C.C.
and
Acharya
,
K.R.
(
2003
)
The crystal structure of C3stau2 from Staphylococcus aureus and its complex with NAD
.
J Biol. Chem.
278
,
45924
45930
30
Fieldhouse
,
R.J.
,
Jorgensen
,
R.
,
Lugo
,
M.R.
and
Merrill
,
A.R.
(
2012
)
The 1.8 Å cholix toxin crystal structure in complex with NAD+ and evidence for a new kinetic model
.
J. Biol. Chem.
287
,
21176
21188
31
Ravulapalli
,
R.
,
Lugo
,
M.R.
,
Pfoh
,
R.
,
Visschedyk
,
D.
,
Poole
,
A.
,
Fieldhouse
,
R.J.
et al. 
(
2015
)
Characterization of Vis toxin, a novel ADP-ribosyltransferase from Vibrio splendidus
.
Biochemistry
54
,
5920
5936
32
Nakano
,
T.
,
Matsushima-Hibiya
,
Y.
,
Yamamoto
,
M.
,
Takahashi-Nakaguchi
,
A.
,
Fukuda
,
H.
,
Ono
,
M.
et al. 
(
2013
)
ADP-ribosylation of guanosine by SCO5461 protein secreted from Streptomyces coelicolor
.
Toxicon
63
,
55
63
33
Sun
,
J.
,
Maresso
,
A.W.
,
Kim
,
J.-J.P.
and
Barbieri
,
J.T.
(
2004
)
How bacterial ADP-ribosylating toxins recognize substrates
.
Nat. Struct. Mol. Biol.
11
,
868
876
34
Han
,
S.
,
Arvai
,
A.S.
,
Clancy
,
S.B.
and
Tainer
,
J.A.
(
2001
)
Crystal structure and novel recognition motif of rho ADP-ribosylating C3 exoenzyme from Clostridium botulinum: structural insights for recognition specificity and catalysis
.
J. Mol. Biol.
305
,
95
107
35
Eustermann
,
S.
,
Wu
,
W.-F.
,
Langelier
,
M.-F.
,
Yang
,
J.-C.
,
Easton
,
L.E.
,
Riccio
,
A.A.
et al. 
(
2015
)
Structural basis of detection and signaling of DNA single-strand breaks by human PARP-1
.
Mol Cell
60
,
742
754
36
Ricard
,
J.
,
Meunier
,
J.-C.
and
Buc
,
J.
(
1974
)
Regulatory behavior of monomeric enzymes. 1. The mnemonical enzyme concept
.
Eur. J. Biochem.
49
,
195
208
37
Rabin
,
B.R.
(
1967
)
Co-operative effects in enzyme catalysis: a possible kinetic model based on substrate-induced conformation isomerization
.
Biochem. J.
102
,
22C
23C
38
Ainslie
, Jr,
G.R.
,
Shill
,
J.P.
and
Neet
,
K.E.
(
1972
)
Transients and cooperativity. A slow transition model for relating transients and cooperative kinetics of enzymes
.
J. Biol. Chem.
247
,
7088
7096
PMID:
[PubMed]
39
Cardenas
,
M.L.
,
Rabajille
,
E.
and
Niemeyer
,
H.
(
1984
)
Suppression of kinetic cooperativity of hexokinase D (glucokinase) by competitive inhibitors. A slow transition model
.
Eur. J. Biochem.
145
,
163
171
40
Oda
,
T.
,
Hirabayashi
,
H.
,
Shikauchi
,
G.
,
Takamura
,
R.
,
Hiraga
,
K.
,
Minami
,
H.
et al. 
(
2017
)
Structural basis of autoinhibition and activation of the DNA-targeting ADP-ribosyltransferase pierisin-1
.
J Biol Chem.
292
,
15445
15455
41
Gouet
,
P.
,
Courcelle
,
E.
,
Stuart
,
D.I.
and
Metoz
,
F.
(
1999
)
ESPript: analysis of multiple sequence alignments in PostScript
.
Bioinformatics
15
,
305
308

Supplementary data