Structure and Function of the Acetylpolyamine Amidohydrolase from the Deep Earth Halophile Marinobacter subterrani

Abstract

Polyamines are small organic cations essential for cellular function in all kingdoms of life. Polyamine metabolism is regulated by enzyme-catalyzed acetylation-deacetylation cycles in similar fashion to the epigenetic regulation of histone function in eukaryotes. Bacterial polyamine deacetylases are particularly intriguing, since these enzymes share the fold and function of eukaryotic histone deacetylases. Recently, acetylpolyamine amidohydrolase from the deep earth halophile Marinobacter subterrani (msAPAH) was described. This Zn²⁺-dependent deacetylase shares 53% amino acid sequence identity with the acetylpolyamine amidohydrolase from Mycoplana ramosa (mrAPAH) and 22% amino acid sequence identity with the catalytic domain of histone deacetylase 10 from Danio rerio (zebrafish; zHDAC10), the eukaryotic polyamine deacetylase. The X-ray crystal structure of msAPAH, determined in complexes with 7 different inhibitors as well as the acetate coproduct, shows how the chemical strategy of Zn²⁺-dependent amide hydrolysis and the catalytic specificity for cationic polyamine substrates is conserved in a subterranean halophile. Structural comparisons with mrAPAH reveal that an array of aspartate and glutamate residues unique to msAPAH enable the binding of one or more Mg²⁺ ions in the active site and elsewhere on the protein surface. Notwithstanding these differences, activity assays with a panel of acetylpolyamine and acetyllysine substrates confirm that msAPAH is a broad-specificity polyamine deacetylase, much like mrAPAH. Broad substrate specificity contrasts with the narrow substrate specificity of zHDAC10, which is highly specific for N⁸-acetylspermidine hydrolysis. Notably, quaternary structural features govern the substrate specificity of msAPAH and mrAPAH, whereas tertiary structural features govern the substrate specificity of zHDAC10.

Graphical Abstract

Introduction

The arginase-deacetylase superfamily of metalloenzymes has been extensively studied in recent years, but much of the phylogenetic tree remains uncharacterized in terms of structure and function.^1–3 Rat arginase I was the first member of this superfamily to yield an X-ray crystal structure,⁴ and the subsequent crystal structure determination of the structurally homologous histone deacetylase-like protein from Aquifex aeolicus⁵ led to the designation of their common α/β-metallohydrolase topology as the arginase-deacetylase fold. Additional crystal structures of arginase-related metalloenzymes have since been reported, including proclavaminic acid amidino hydrolase,⁶ agmatinase,⁷ and formiminoglutamase.⁸ Among the deacetylases, crystal structures of class I and class II histone deacetylases (HDACs),^3,9–16 as well as acetylpolyamine amidohydrolase from Mycoplana ramosa (mrAPAH),^17,18 have been described.

Structural comparisons indicate that the arginases and deacetylases divergently evolved from a common ancestral α/β-metallohydrolase despite sharing low amino acid sequence identities generally ranging 10–15%.^19–21 The stoichiometry of and selectivity for catalytic metal ion(s) have also diverged: arginases generally require 2 Mn²⁺ ions for function,²² whereas deacetylases require a single Zn²⁺ ion (or possibly a single Fe²⁺ ion) for function.²³ The Zn²⁺ binding site in the deacetylases is conserved as the Mn²⁺_B binding site of the arginases.^19–21

Among the diversity of hydrolytic functions represented in the arginase-deacetylase phylogenetic tree, enzymes of polyamine biosynthesis are particularly intriguing. Arginase itself functions in polyamine biosynthesis to generate product L-ornithine, which undergoes decarboxylation to yield putrescine, a building block in the biosynthesis of spermidine and spermine.^24–27 Generally present at millimolar concentrations in the cell, polyamines serve myriad biological functions such as the stabilization of nucleic acid structure^28–30 and the regulation of transcription and translation.^31–33 Notably, polyamines undergo reversible acetylation in cellular trafficking and function, and enzymes that catalyze polyamine deacetylation are structurally and functionally related to HDACs.^17,18 Indeed, HDAC10 was recently discovered to be a highly specific N⁸-acetylspermidine deacetylase.^3,34 Although this eukaryotic cytosolic activity was discovered more than 40 years ago,^35,36 the responsible enzyme was not identified at the time.

In recent years, Marinobacter species have been identified in harsh marine and non-marine environments, including the deep subsurface based on their isolation in hydraulic fracturing effluent.^37–39 These species are well adapted to the high salinity and Fe²⁺ concentrations found in mines, wells, and other deep subsurface locations. The halophile Marinobacter subterrani was recently identified 714 m below the surface in the Soudan iron mine in Minnesota,⁴⁰ and genomic analysis reveals an acetylpolyamine amidohydrolase (msAPAH, UniProt A0A0J7JFD7) similar to mrAPAH (53% amino acid sequence identity). Both msAPAH and mrAPAH retain approximately 22% amino acid sequence identity with the polyamine deacetylase domain of HDAC10 from Danio rerio (zebrafish; zHDAC10).

Here, we report the X-ray crystal structure of msAPAH, and we show that msAPAH is a broad-specificity polyamine deacetylase (Figure 1a). To illustrate the structural basis of substrate specificity and the catalytic mechanism, crystal structures of msAPAH complexed with the acetate product as well as seven inhibitors (Figure 1b) are presented. These studies provide important insight regarding the relationship between eukaryotic histone deacetylases and prokaryotic polyamine deacetylases.

Figure 1.

(a) The polyamine deacetylase reaction catalyzed by msAPAH illustrated for substrate acetylputrescine. (b) Inhibitors studied in complex with msAPAH: 1, 5-[(3-aminopropyl)amino]pentylboronic acid; 2, 7-[(3-aminopropyl)amino]-1,1,1-trifluoroheptan-2-one; 3, 5-[(3-aminopropyl)amino]pentane-1-thiol; 4, 6-amino-N-hydroxyhexanamide; 5, 8-amino-N-hydroxyoctanamide; 6, 6-[(3-aminopropyl)amino]-N-hydroxyhexanamide; 7, 4-(dimethylamino)-N-[7-hydroxyamino)-7-oxoheptyl]benzamide (also known as M344). Inhibitor syntheses have been described.^18,41

Materials and Methods

Reagents.

In general, all chemicals were purchased from Fisher Scientific, Millipore Sigma, or Hampton Research and used without further purification. Inhibitory polyamine analogues 1–6 were synthesized as previously described.^18,41 Inhibitor 7, 4-(dimethylamino)-N-[7-hydroxyamino)-7-oxoheptyl]benzamide (also known as M344), was purchased from Cayman Chemical.

Protein Preparation.

Full-length msAPAH was recombinantly expressed using a pGEX-6P-1 vector by Genscript. This construct utilizes a BAMHI/Xhol cloning site and contains ampicillin bacterial resistance. An N-terminal GST tag is attached, followed by a PreScission protease cleavage site prior to the start of the protein sequence.

Protein was expressed using Escherichia coli One Shot BL21(DE3) cells (Invitrogen) and grown in 2xYT medium in the presence of 100 μg/mL ampicillin. Cells were grown at 37° C and 250 RPM in an Innova 40 incubator shaker until OD₆₀₀ reached approximately 0.80. The temperature was then reduced to 18° C until OD₆₀₀ reached 1.0. At this point, cells were supplemented with 200 μM isopropyl β-L-1-thiogalactopyranoside (IPTG; Gold Biotechnology) and were grown for an additional 18 h at 250 RPM. Cells were then centrifuged for 20 min at 5,000 RPM using a Sorvall LYNX 6000 centrifuge. Cell pellets were stored at −80° C until needed.

The cell pellet was thawed and re-suspended in 100 mL of Buffer A [50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) sodium salt (pH 7.5), 250 mM NaCl, 1 mM tris(2-carboxyethyl)phosphine (TCEP), 10% glycerol, and 10 μM ZnCl₂]. Additionally, 2 protease inhibitor tablets, 0.1 mg/mL lysozyme, and 50 μg/mL DNAse were added to the solution. The cells were then lysed by sonication. Cell lysate was centrifuged for 1 h at 15,000 RPM using a Sorval LYNX 6000 centrifuge. Cell lysate was then applied to a 5-mL pre-equilibrated GSTrap HP column.

The GST-msAPAH fusion protein bound to the GSTrap HP column and was eluted using Buffer B [50 mM HEPES sodium salt (pH 8.0), 250 mM NaCl, 1 mM TCEP, 10% glycerol,10 μM ZnCl₂, and 10 mM reduced glutathione]. Protein containing fractions were collected and digested overnight using 5 mg/mL recombinant PreScission protease in 4 L Buffer A through dialysis (RC dialysis tubing, 6–8 kD molecular weight cut-off).

After protease cleavage for 12 h, the protein digest was applied to the 5-mL GSTrap HP column. The GST tag remained bound to the column and the msAPAH protein flowed through the column. The msAPAH-containing fractions were concentrated to 5 mL using a 15-mL centrifugal filter unit with a molecular weight cut-off of 10 kDa. The protein was then filtered using a 0.22-μM Millex-GV filter unit prior to being loaded onto a HiLoad Superdex 26/600 200 pg column. The column was pre-equilibrated with 360 mL of Buffer C [50 mM HEPES sodium salt (pH 7.5), 200 mM KCl, 1 mM TCEP, and 5% glycerol]. The 5-mL sample of msAPAH was injected at a rate of 1 mL/min, and 5-mL fractions were collected. Pure msAPAH protein was confirmed by SDS-PAGE, concentrated to approximately 10 mg/mL, and stored at −80° C.

Enzyme Kinetics.

To measure the rate of acetylpolyamine hydrolysis catalyzed by msAPAH, the generation of the acetate coproduct was measured using a colorimetric assay kit (Sigma-Aldrich, catalogue number MAK086). Briefly, the reaction mixture (50 μL) was prepared as described in the kit protocol: 37 μL Acetate Assay Buffer, 2 μL Acetate Enzyme Mix, 2 μL ATP, 2 μL Acetate Substrate Mix, 2 μL Probe, and 5 μL of 10 μM msAPAH. Separately, 50 μL substrate solutions were prepared at concentrations ranging 0–5,000 μM in Acetate Assay Buffer.

To initiate the reaction, 50 μL of the reaction mixture was combined with 50 μL of substrate solution in a Greiner 96 well flat-bottom transparent plate, which was immediately placed in a Tecan Infinite M1000Pro plate reader and monitored at 450 nm. The pH of the assay solution was measured to be 7.75. Measurements were made every 10 s for a total of 40 min, with shaking every 5 s. The plate was covered at all times due to the photosensitivity of the probe.

To calculate msAPAH activity, absorbance measurements at 450 nm were converted into product formation using an acetate standard curve according to the kit protocol. Data were analyzed using GraphPad Prism version 7.00 for MAC OS X (GraphPad software, La Jolla California USA, www.graphpad.com). Assays were performed in triplicate at 25° C. Nonlinear regression fits to the Michaelis-Menten equation were used to determine steady-state parameters.

Inhibitory Activity Measurements.

Inhibitory potencies of compounds 1–7 against msAPAH were determined using the colorimetric acetate assay described above. Briefly, the reaction mixture (50 μL) was prepared as described above, except that the protein was omitted: 42 μL Acetate Assay Buffer, 2 μL Acetate Enzyme Mix, 2 μL ATP, 2 μL Acetate Substrate Mix, and 2 μL Probe. Separately, 20 μL of 2 μM msAPAH protein was prepared with 10 μL inhibitor (prepared in Buffer C) with final inhibitor concentrations ranging from 0–2,000 μM. The inhibitor was incubated with enzyme for 15 minutes prior to the addition of substrate.

The reaction was initiated by the addition of 20 μL of 4,000 μM acetylcadaverine substrate. Immediately after substrate addition, 50 μL of reaction mixture was added to the 50 μL of mixture containing the msAPAH protein, acetylcadaverine substrate, and inhibitor. The reaction was analyzed using a Tecan Infinite M1000Pro plate reader as described above. Assays were run at 25°C, with the absorbance being measured at 450 nm for a total of 100 min.

The IC₅₀ values were calculated from raw data using Prism 7. Data were first transformed using the function x = log(x). Data was then normalized by setting the 0% identity definition to be the smallest mean in each dataset, while the 100% identity definition was set to be the largest mean in each dataset. Finally, the data were fit to a nonlinear regression curve of log(inhibitor) vs. normalized response. All measurements were run in duplicate. Goodness of Fit statistics ranged from 88%−97%. IC₅₀ measurements are recorded in Figure S1.

Isothermal Titration Calorimetry (ITC).

Given the somewhat modest inhibitory potencies for some of the inhibitors studied, we additionally performed direct measurements of enzyme-inhibitor complexation using ITC. Dissociation constants (K_d) for inhibitors 1–7 to msAPAH were measured using a MicroCal iTC 200 isothermal titration calorimeter (GE Healthcare). For each measurement, 300 μM of inhibitor was titrated into 30 μM msAPAH in 50 mM HEPES (pH 7.5), 150 mM NaCl, 1 mM TCEP-HCl, and 10% glycerol (v/v) at 25 °C. In the case of inhibitor 6, 600 μM inhibitor was titrated into 60 μM of msAPAH. For inhibitor 7, both inhibitor and msAPAH solutions included 5% (v/v) DMSO. Nineteen 2-μL injections were made over 60 min. Enthalpogram data were visualized and analyzed using Origin software (OriginLab, Northampton, MA). Enthalpograms for inhibitors 1, 2, 4, and 7 could not be fit to binding curves. Regardless, the direct calorimetric measurement of enzyme-inhibitor affinity for inhibitors 5 and 6 yielded micromolar dissociation constants while inhibitor 3 yielded a nanomolar dissociation constant (Figure S2).

Crystallization.

All msAPAH-inhibitor complexes were crystallized using the sitting-drop vapor diffusion method. For cocrystallization of all msAPAH complexes, a 350-nL drop of protein solution [10 mg/mL msAPAH in Buffer C, and 2 mM inhibitor in phosphate-buffered saline (PBS)] was added to a 350-nL drop of precipitant solution as outlined below, and equilibrated against 80 μL of precipitant solution in the well reservoir. All crystals appeared within approximately two days at 4° C, except those of the msAPAH-3 complex which grew at 21° C. Ethylene glycol (15% v/v) was added to the mother liquor as a cryoprotectant prior to flash cooling crystal for X-ray diffraction data collection.

For cocrystallization of the msAPAH-acetate complex, the precipitant was 0.2 M calcium acetate hydrate, 20% w/v PEG 3350, which yielded thick cubic crystals. For cocrystallization of the msAPAH-1 complex, the precipitant was 0.2 M magnesium chloride hexahydrate, 0.1 M HEPES (pH 7.5), 25% w/v PEG 3350, which yielded long rod-like crystals. For cocrystallization of the msAPAH-2 complex, the precipitant was 0.2 M magnesium acetate tetrahydrate, 20% w/v PEG 3350, which yielded thick plate-like crystals. For cocrystallization of the msAPAH-3 complex, the precipitant was 20% v/v 2-propanol, 0.1 M MES monohydrate (pH 6.0), 20% w/v polyethylene glycol monomethyl ether 2000, which yielded long rod-like crystals. For cocrystallization of the msAPAH-4 complex, the precipitant was 0.2 M magnesium acetate tetrahydrate, 20% w/v PEG 3350, which yielded thick plate-like crystals. For cocrystallization of the msAPAH-5 complex, the precipitant consisted of 0.2 M magnesium acetate tetrahydrate, 20% w/v PEG 3350, which yielded thick plate-like crystals. For cocrystallization of the msAPAH-6 complex, the precipitant was 0.2 M calcium chloride dihydrate, 20% w/v PEG 3350, which yielded thick plate-like crystals. For cocrystallization of the msAPAH-7 complex, the precipitant was 0.2 M calcium chloride dihydrate, 20% w/v PEG 3350, which yielded thick plate-like crystals.

Data Collection and Structure Determination.

X-ray diffraction data were collected at Northeastern Collaborative Access Team beamline 24-ID-E, Advanced Photon Source, Argonne National Laboratory, from crystals of msAPAH complexed with 1 and 3. X-ray diffraction data were collected at Northeastern Collaborative Access Team beamline 24-ID-C, Advanced Photon Source, Argonne National Laboratory, from crystals of msAPAH complexed with 2, 5, and 7. X-ray diffraction data were collected at beamline 12–2, Stanford Synchrotron Radiation Laboratory, Stanford University, from crystals of msAPAH complexed with 4 and 6. X-ray diffraction data were collected at Frontier Macromolecular Crystallography beamline 17-ID-2 (FMX), National Synchrotron Light Source II, Brookhaven National Laboratory, from crystals of msAPAH complexed with acetate. Data were indexed and integrated using iMosflm⁴² and scaled using Aimless in the CCP4 program suite.⁴³ Data reduction statistics are recorded in Table 1.

Table 1:

Crystallographic data collection and refinement statistics.

Data Collection	msAPAH-1	msAPAH-2	msAPAH-3	msAPAH-4	msAPAH-5	msAPAH-6	msAPAH-7	msAPAH-Acetate
Space Group	P21	P21	P21	P1	P21	P21	P21	P21
Unit Cell Dimensionsa
a,b,c (Å)	65.6 163.5 65.7	47.4 82.4 90.7	45.9 120.7 65.4	66.3 67.0 167.5	52.3 119.1 66.2	65.7 163.2 65.7	65.6 163.0 65.6	52.4 121.1 66.6
α,β,γ (°)	90 94 90	90 98 90	90 109 90	90 90 93	90 109 90	90 94 90	90 95 90	90 109 90
Rmergeb	0.161 (0.456)	0.180 (0.597)	0.163 (1.018)	0.205 (0.474)	0.172 (0.742)	0.092 (0.531)	0.105 (0.370)	0.104 (0.826)
Rpimc	0.102 (0.312)	0.180 (0.595)	0.106 (0.675)	0.158 (0.403)	0.172 (0.742)	0.085 (0.470)	0.100 (0.353)	0.066 (0.531)
CC1/2d	0.958 (0.706)	0.952 (0.525)	0.991 (0.657)	0.830 (0.588)	0.969 (0.537)	0.992 (0.657)	0.986 (0.811)	0.997 (0.703)
Redundancy	3.6 (3.6)	3.4 (3.5)	6.2 (6.1)	1.9 (1.9)	3.4 (3.4)	3.2 (3.0)	3.4 (3.1)	6.7 (6.6)
Completeness (%)	98.9 (99.7)	90.2 (91.4)	99.1 (99.2)	83.2 (86.5)	86.7 (88.7)	95.0 (95.6)	90.8 (81.8)	99.9 (99.7)
I/σ	4.7 (2.8)	10.2 (6.1)	6.7 (2.1)	4.4 (2.2)	5.6 (2.1)	7.3 (1.9)	6.8 (2.5)	10.2 (2.1)
Refinement
Resolution (Å)	65.44–1.65 (1.68–1.65)	89.79–2.00 (2.07–2.00)	61.77–1.65 (1.71–1.65)	66.97–2.03 (2.10–2.03)	59.53–1.54 (1.60–1.55)	46.27–1.75 (1.81–1.75)	60.71–1.70 (1.76–1.70)	27.85–1.64 (1.69–1.64)
No. Reflections	125734 (12518)	41498 (4163)	79658 (7965)	154543 (15944)	95821 (7699)	131376 (12851)	136205 (13035)	95961 (9560)
Rwork/Rfreee	0.193/0.225 (0.216/0.264)	0.161/0.212 (0.158/0.227)	0.179/0.208 (0.254/0.284)	0.220/0.247 (0.278/0.315)	0.244/0.280 (0.415/0.415)	0.175/0.212 (0.242/0.293)	0.162/0.199 (0.202/0.247)	0.168/0.194 (0.277/0.333)
No. Atomsf
Protein	10580	5332	5325	21126	5343	10689	10673	5386
Ligand	69	44	54	112	33	85	105	25
Solvent	598	354	294	648	254	774	953	502
Average B Factors Å2
Protein	12	11	15	10	12	17	13	19
Ligand	15	21	19	8	16	20	24	19
Solvent	15	14	19	9	14	21	20	26
R.m.s. Deviations
Bond Lengths (Å)	0.007	0.007	0.008	0.004	0.006	0.006	0.006	0.006
Bond Angles (°)	0.9	0.9	0.9	0.7	0.8	0.8	0.8	0.9
Ramachandran Plotg
Favored	96.0	97.0	96.0	97.0	97.0	96.0	96.0	97.0
Allowed	4.00	3.00	4.00	3.00	3.00	4.00	4.00	3.00
Outliers	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
PDB code	6PHT	6PHZ	6PHR	6PIC	6PID	6PIA	6PI1	6PI8

^aValues in parentheses refer to the highest-resolution shell indicated.

^bR_merge = ∑_hkl∑_i|I_i,_hkl − ⟨I⟩_hkl|/∑_hkl∑_iI_i,hkl, where ⟨I⟩_hkl is the average intensity calculated for reflection hkl from replicate measurements.

^cR_p.i.m.= (∑_hkl(1/(N-1))^1/2∑_i|I_i,hkl − ⟨I⟩_hkl|)/∑_hkl∑_i I_i,_hkl, where ⟨I⟩_hkl is the average intensity calculated for reflection hkl from replicate measurements and N is the number of reflections.

^dPearson correlation coefficient between random half-datasets.

^eR_work = ∑||F_o| − |F_c||/∑|F_o| for reflections contained in the working set. |F_o| and |F_c| are the observed and calculated structure factor amplitudes, respectively. R_free is calculated using the same expression for reflections contained in the test set held aside during refinement.

^fPer asymmetric unit.

^gCalculated with PROCHECK.

Molecular replacement using the atomic coordinates of mrAPAH [Protein Data Bank (PDB) entry 4ZUM]¹⁸ was used as a search model to phase the initial electron density map of the first msAPAH structure. Subsequent msAPAH structures were determined in similar fashion using either mrAPAH or msAPAH as search probes for molecular replacement. Autobuild⁴⁴ was used to register the msAPAH sequence in the initial electron density map. The graphics program COOT⁴⁵ was used to manually adjust residue conformations and structure refinement was performed using PHENIX.⁴⁶ The quality of each structure was assessed using MolProbity⁴⁷ and PROCHECK.⁴⁸ Refinement statistics are recorded in Table 1.

Results

Catalytic activity.

Recombinant msAPAH exhibits broad specificity for short and long acetylpolyamine substrates, with a moderate preference for putrescine or N⁸-acetylspermidine (depending on the concentration of KCl) based on catalytic efficiency (k_cat/K_M) (Figure 2, Table 2). Even so, catalytic efficiency is modest; at best, k_cat/K_M = 470 M⁻¹s⁻¹ for N⁸-acetylspermidine in the presence of 10 mM KCl. We were unable to determine steady-state kinetic parameters in higher concentrations of KCl. No significant catalytic activity was observed for the hydrolysis of Aly-NH₂ or the MSH2-based peptide substrates Ac-Ala-Aly, Ac-Gly-Ala-Aly, Ac-Gly-Ala-Aly-Asn-Leu-Gln-NH₂, or Ac-Aly-Asn-Leu-NH₂ (Aly = acetyllysine) using the liquid chromatography-mass spectrometry assay employed for the assay of HDAC6 activity.¹⁶ Thus, while msAPAH is not a particularly efficient polyamine deacetylase, it is clearly not a lysine deacetylase.

Figure 2.

Michaelis-Menten plot showing the steady-state kinetics of msAPAH-catalyzed acetylpolyamine hydrolysis.

Table 2:

Steady-state kinetic parameters for acetylpolyamine hydrolysis by msAPAH

Substrate	kcat (s−1)	KM (μM)	kcat/KM (M−1s−1)
acetylcadaverine	0.026 ± 0.004	800 ± 300	32 ± 6
acetylputrescine	0.0078 ± 0.0003	44 ± 10	180 ± 30
acetylspermine	0.007 ± 0.002	1300 ± 600	5 ± 1
N1-acetylspermidine	0.0038 ± 0.0006	410 ± 200	9 ± 4
N8-acetylspermidine (0 mM KCl and 10 mM KCl)	0.0092 ± 0.0009 0.008 ± 0.003	280 ± 80 17 ± 5	32 ± 20 470 ± 40

All measurements represent mean ± average of absolute deviations from the mean in a given set of data. All measurements were run in triplicate at 25° C.

Structure of the msAPAH–1 Complex.

The 1.65 Å-resolution crystal structure of msAPAH complexed with the boronic acid analogue of N⁸-acetylspermidine (1, IC₅₀ = 390 μM) contains 4 monomers in the asymmetric unit. The structure of msAPAH reveals the arginase-deacetylase fold, as previously observed for the broad-specificity polyamine deacetylase mrAPAH.¹⁷ Like mrAPAH, msAPAH is a dimer with identical quaternary structure (Figure 3). There are no major conformational changes between the monomers or dimers of msAPAH and mrAPAH (PDB 4ZUM): the root-mean-square (rms) deviation is 0.53 Å for 268 Cα atoms (monomer A) and 0.75 Å for 566 Cα atoms for the dimer, respectively.

Figure 3.

Crystal structure of the msAPAH-1 complex (top) illustrating the dimer interface between monomer A (blue) and monomer B (red), looking down the two-fold symmetry axis. There are no major conformational differences between monomers and least-squares superposition yields an rms deviation of 0.15 Å for 300 Cα atoms. The cutaway view of the active site (bottom) shows how the dimer interface constricts the approach to the catalytic Zn²⁺ ions (gray spheres). Only long and slender acetylpolyamine substrates, and not bulky acetyllysine-containing peptide substrates, can bind in this constricted active site.

Electron density corresponding to the boronic acid moiety of 1 clearly indicates trigonal planar geometry (Figure 4), with the hydroxyl groups coordinated to the catalytic Zn²⁺ ion with asymmetric bidentate coordination geometry (average Zn²⁺---O1 distance = 2.1 Å and Zn²⁺---O2 distance = 2.4 Å). Notably, inhibitor binding displaces the Zn²⁺-bound solvent molecule expected for the unliganded enzyme. Surprisingly, the electrophilic boronic acid does not undergo nucleophilic attack to form a tetrahedral boronate anion, which would mimic the tetrahedral transition state for acetylpolyamine hydrolysis. Such chemistry is typically observed, for example, in the binding of boronic acids in the arginase active site,^49–52 but not exclusively so.⁵³ Boronic acid hydroxyl group O1 hydrogen bonds with Y323, and hydroxyl group O2 hydrogen bonds with H158 and H159. These catalytic residues are unique to metal-dependent deacetylases. In catalysis, Y323 assists the Zn²⁺ ion in polarizing the amide carbonyl group of the substrate for nucleophilic attack by a Zn²⁺-bound solvent molecule, and tandem histidine residues H158-H159 serve general base-general acid functions.

Figure 4.

Polder omit map of the msAPAH-1 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH monomer D), dark gray (msAPAH monomer B) or wheat (inhibitor), B = light green, N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

Additional hydrogen bond interactions stabilize the binding of 1 in the msAPAH active site. The primary amino group of 1 forms water-mediated hydrogen bonds with the side chain of E17, the backbone carbonyl of L18, and the side chain of D19. The binding conformation of 1 is slightly different in monomer B, such that there is a direct hydrogen bond with the side chain of Y168. The secondary amino group of 1 forms water-mediated hydrogen bonds with the side chains of D117, Y168, and the primary amino group of 1.

Although the quaternary structure of msAPAH is identical to that of mrAPAH, there are several key differences in residues that define the active site cleft at the dimer interface of msAPAH. Residue Y19 in mrAPAH corresponds to D19 in msAPAH, a residue that is critical for accepting hydrogen bonds from the polyamine substrates. Residue F27 in mrAPAH corresponds to H27 in msAPAH and is located at the surface of the active site pocket. The most noticeable difference at the dimer interface is the substitution of residues Y83, S91 and E106 in mrAPAH by K83, F91 and D106 in msAPAH. This difference enables the binding of a Mg²⁺ ion at the dimer interface of msAPAH. This Mg²⁺ ion is observed in all msAPAH crystal structures described herein and is coordinated by the carboxylate side chains of D104 and D106; the Mg²⁺ coordination polyhedron is completed by water molecules. Additional residues important for dimer assembly of mrAPAH include F92, which engages in offset π stacking interactions with F92 in the adjacent monomer. In msAPAH, this interaction is maintained by W92-W92 at the dimer interface.

In addition to the Mg²⁺ ion bound in the active site at the dimer interface, an additional Mg²⁺ ion is coordinated by E288 in monomer A. The coordination sphere of this Mg²⁺ ion is completed by 4 water molecules, yielding a pentacoordinate metal ion. This Mg²⁺ ion is absent in monomers B, C, and D.

Structure of the msAPAH–2 Complex.

The 2.00 Å-resolution crystal structure of msAPAH complexed with trifluoroketone analogue of N⁸-acetylspermidine (2, IC₅₀ = 350 μM) contains 2 monomers in the asymmetric unit. The trifluoroketone moiety exists predominantly as the gem-diol in solution;⁵⁴ accordingly, it binds as the gem-diol or perhaps the ionized gem-diolate in the active site of msAPAH (Figure 5). This binding mode mimics the binding of the tetrahedral intermediate and its flanking transition states in the hydrolysis of an acetylpolyamine substrate. The gem-diol(ate) coordinates to the catalytic Zn²⁺ ion with nearly symmetric bidentate coordination geometry, with average Zn²⁺---O1 and Zn²⁺---O2 distances of 2.2 Å and 2.3 Å, respectively. The gem-diol O1 group also hydrogen bonds with Y323, which adopts the “in” conformation in monomer B. This hydrogen bond is absent in monomer A, where Y323 adopts the “out” conformation. The gem-diol O2 group hydrogen bonds with H158 and H159. The binding of trifluoroketone inhibitors as gem-diols has similarly been observed in the active sites of other deacetylases.^3,16,18

Figure 5.

Polder omit map of the msAPAH-2 complex, monomer B (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH monomer B), dark gray (msAPAH monomer A) or wheat (inhibitor), F = magenta, S = yellow, N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

Each of the C-F bonds of the trifluoro moiety of 2 engages in hydrogen bond interactions with protein atoms. One fluorine atom accepts a hydrogen bond from Y323, and a second fluorine atom accepts a hydrogen bond from the backbone NH group of G321. A third fluorine atom accepts a hydrogen bond from C169; this fluorine atom also makes a van der Waals contact with the backbone carbonyl of G167.

The primary amino group of 2 donates hydrogen bonds to the side chain of E17, the side chain of D19, and the backbone carbonyl oxygen of L18. The secondary amino group makes water-mediated hydrogen bond interactions with the side chains of D117 (monomer B) or Y168 (monomer A).

In addition to the Mg²⁺ ion bound in the active site, magnesium hexahydrate (Mg²⁺(OH₂)₆) binds near D275 in monomer A, and the side chain carboxylate of D275 makes hydrogen bond interactions with metal-bound water molecules. In monomer B, E288 coordinates to a Mg²⁺ ion along with 5 water molecules, yielding octahedral metal coordination geometry.

Structure of the msAPAH-Acetate Complex.

The 1.64 Å-resolution crystal structure of msAPAH complexed with the acetate coproduct contains 2 monomers in the asymmetric unit. Absent the polyamine coproduct bound in the active site, this structure represents a partially dissociated product complex. Acetate coordinates to the active site Zn²⁺ ion with nearly symmetric bidentate geometry (average Zn²⁺---O1 and Zn²⁺---O2 distances = 2.2 Å and 2.4 Å, respectively) (Figure 6). The carboxylate O1 atom additionally accepts a hydrogen bond from Y323, and the carboxylate O2 atom accepts hydrogen bonds from H158 and H159.

Figure 6.

Polder omit map of the msAPAH-acetate complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH monomer A), dark gray (msAPAH monomer B), wheat (inhibitor), or orange (ethylene glycol), N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

In addition to the Mg²⁺ ion bound in the active site, an additional Mg²⁺ ion is coordinated by E288, D324, D326, and 3 water molecules in monomer A. However, this Mg²⁺ ion is absent in monomer B.

Structure of the msAPAH–3 Complex.

The 1.65 Å-resolution crystal structure of msAPAH complexed with the thiol analogue of N⁸-acetylspermidine (3, IC₅₀ = 160 μM, K_d = 0.2 μM) contains 2 monomers in the asymmetric unit. The thiol group is presumably ionized to the negatively charged thiolate anion as it coordinates to the active site Zn²⁺ ion, with an average Zn²⁺---S distance of 2.4 Å (Figure 7). With Zn²⁺---S–C angles of 116° and 100°, and Zn²⁺---S–C–C dihedral angles of −7° in monomer A and 59° in monomer B, respectively, thiolate-metal coordination parameters deviate somewhat from ideal values for thiolate-metal interactions first outlined for the side chain of cysteine (Zn²⁺---S–C angle of 90° and Zn²⁺---S–C–C dihedral angle of ± 90° or 180°).⁵⁵ The phenolic hydroxyl group of Y323 additionally donates a hydrogen bond to the Zn²⁺-bound thiolate.

Figure 7.

Polder omit map of the msAPAH-3 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH, monomer B), dark gray (msAPAH, monomer A), or wheat (inhibitor and MES), S = yellow, N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

As observed in the msAPAH-2 complex, the primary amino group of 3 hydrogen bonds to E17, D19, and the backbone carbonyl of L18. Also similar are the hydrogen bonds between the secondary amino group of 3 and D117 and Y168.

Unique to this structure is the binding of a buffer molecule, 2-(N-morpholino)-ethanesulfonic acid (MES), in the active site tunnel (Figure 7). The sulfonate group of MES accepts a hydrogen bond from the backbone NH group of F225 and a Mg²⁺-bound water molecule. It also makes a hydrogen bond with H197.

Structure of the msAPAH–4 Complex.

The 2.03 Å-resolution crystal structure of msAPAH complexed with hydroxamate inhibitor 4 (IC₅₀ = 410 μM) contains 8 monomers in the asymmetric unit. Inhibitor 4 is a hydroxamate analogue of N-acetylputrescine. Intermolecular interactions of the hydroxamate group provide useful structural inferences on catalysis (Figure 8). Specifically, just as the hydroxamate carbonyl group coordinates to Zn²⁺ and accepts a hydrogen bond from Y323, so too would the scissile carbonyl group of N-acetylputrescine in the precatalytic enzyme-substrate complex. Both Zn²⁺ coordination and Y323 hydrogen bond interactions are required to activate the amide carbonyl of N-acetylputrescine for nucleophilic attack by a Zn²⁺-bound solvent molecule in catalysis.

Figure 8.

Polder omit map of the msAPAH-4 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH, monomer A), dark gray (msAPAH, monomer C, symmetry mate), or wheat (inhibitor), N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

The primary amino group of inhibitor 4 donates hydrogen bonds to D117 and Y168 (Figure 8), in a manner similar to the interactions of the secondary amino group of inhibitor 3 (Figure 7). The primary amino group of 4 also engages in water-mediated hydrogen bonds with E17, D19, and the backbone carbonyl of L18.

Structure of the msAPAH–5 Complex.

The 1.55 Å-resolution crystal structure of msAPAH complexed with hydroxamate inhibitor 5 (IC₅₀ = 380 μM, K_d = 29 μM) contains 2 monomers in the asymmetric unit. Hydroxamate-metal coordination interactions and hydrogen bond interactions (Figure 9) are identical to those observed for the binding of hydroxamate inhibitor 4. While inhibitor 5 is not isosteric with any naturally occurring polyamine substrate, its intermolecular interactions reveal that alternative hydrogen bond interactions can be achieved for the primary amino group of a long, slender polyamine analogue in the active site of msAPAH. Inhibitor 5 is two methylene groups longer than inhibitor 4 and accordingly extends further out of the active site tunnel. The additional length allows for the primary amino group to engage in direct hydrogen bonds rather than water-mediated hydrogen bonds with E17, D19, and the backbone carbonyl of L18. The D19 hydrogen bond only occurs in monomer A while a hydrogen bond to Y168 occurs in monomer B.

Figure 9.

Polder omit map of the msAPAH-5 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH, monomer A), dark gray (msAPAH, monomer B), or wheat (inhibitor), N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

In addition to the Mg²⁺ ion bound in the active site, an additional Mg²⁺ ion is coordinated by E288, D324, D326, and 3 water molecules in monomer A. However, this Mg²⁺ ion is absent in monomer B.

Structure of the msAPAH–6 Complex.

The 1.75 Å-resolution crystal structure of the complex between msAPAH and hydroxamate inhibitor 6 (IC₅₀ = 470 μM, K_d = 11 μM) contains 4 monomers in the asymmetric unit. Hydroxamate-metal coordination interactions and hydrogen bond interactions (Figure 10) are identical to those observed for other hydroxamate inhibitors described above. Inhibitor 6 is the hydroxamate analogue of the polyamine substrate N⁸-acetylspermidine. The primary amino group of 6 donates hydrogen bonds to E17, D19, and the backbone carbonyl of L18. The secondary amino group of 6 donates hydrogen bonds to D117 in monomers A and B, and Y168 in all four monomers.

Figure 10.

Polder omit map of the msAPAH-6 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH, monomer A), dark gray (msAPAH, monomer B), or wheat (inhibitor), N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

In addition to the Mg²⁺ ion bound in the active site, an additional Mg²⁺ ion is coordinated by E288 and 5 water molecules in monomer B; D324 and D326 form hydrogen bonds with metal-bound water molecules. However, this Mg²⁺ ion is absent in monomers A, C, and D.

Structure of the msAPAH–7 Complex.

The 1.70 Å-resolution crystal structure of msAPAH complexed with inhibitor 7 (IC₅₀ = 160 μM) contains 4 monomers in the asymmetric unit. Inhibitor 7 is also known as M344 and is characterized by a m-dimethylaminobenzamide “capping group”. Hydroxamate-metal coordination interactions and hydrogen bond interactions (Figure 11) are identical to those observed for other hydroxamate inhibitors described above. The aromatic capping group is nestled in the dimer interface and is oriented away from the Mg²⁺ ion. The amide group forms water-mediated hydrogen bonds with H197, the backbone amide of F225, the backbone carbonyl of I291, and a Mg²⁺-bound water molecule. The amide group also forms water-mediated hydrogen bonds with E17, D117, and Y168.

Figure 11.

Polder omit map of the msAPAH-7 complex (contoured at 3.0σ). Atoms are color-coded as follows: C = light blue (msAPAH, monomer B), dark gray (msAPAH, monomer A), or wheat (inhibitor), N = blue, O = red, Zn²⁺ = gray sphere, Mg²⁺ = large, dark red sphere, and solvent = small red spheres. Metal coordination and hydrogen bond interactions are indicated by solid and dashed black lines, respectively.

In addition to the Mg²⁺ ion bound in the active site, an additional Mg²⁺ ion is coordinated by E288 and 5 water molecules in monomer B. However, this Mg²⁺ ion is absent in monomers A, C, and D.

Discussion

Crystal structures of msAPAH–inhibitor complexes reveal key insight on substrate recognition and catalysis by a bacterial polyamine deacetylase. Common to all metal-dependent deacetylases in the arginase-deacetylase superfamily, catalytically important residues in the msAPAH active site include tandem histidine residues H158 and H159. We speculate that H158 serves as a general base in catalysis by deprotonating a Zn²⁺-bound water molecule for nucleophilic attack at the scissile amide carbonyl of the substrate, which in turn is polarized by coordination to Zn²⁺ and also by a hydrogen bond with Y323. This mechanistic proposal is consistent with intermolecular interactions observed for the binding of trifluoroketone transition state analogue 2 (Figure 5) as well as the acetate coproduct (Figure 6). Based on the observation of two conformations for Y323 in the msAPAH–2 complex – the “in” conformation oriented into the active site, and the “out” conformation oriented toward bulk solvent – it appears likely that Y323 undergoes a substrate-induced conformational change in catalysis. Conformational flexibility of the catalytic tyrosine has also been observed in crystal structures of mrAPAH,^17,18 as well as molecular dynamics simulations of HDAC3 and HDAC8.^56–58 This flexibility appears to be enhanced by the location of Y323 in a glycine-rich loop.⁵⁸

Molecular recognition of cationic polyamine substrates by msAPAH and mrAPAH is facilitated by substantial negative electrostatic potential on the protein surface surrounding the mouth of the active site (Figure 12). The negatively charged protein surface provides electrostatic attraction to facilitate the binding of positively charged polyamine substrates. Once bound in the active site of msAPAH, the backbone carbonyl of L18 and the side chains of E17, D19, D117 and Y168 provide direct or water-mediated hydrogen bond interactions to the positively charged amino groups of the substrate, as observed in the array of crystal structures presented in Figures 4–11.

Figure 12.

Electrostatic surfaces of msAPAH, mrAPAH, and zHDAC10. Arrows indicate the mouth of each enzyme active site.

Unique to msAPAH is the binding of a Mg²⁺ ion in the substrate binding cleft in the structures of all enzyme-inhibitor complexes. This metal ion is coordinated by D104 and D106 of the alternate subunit of the dimer, and water molecules usually complete an octahedral coordination complex. The binding of this Mg²⁺ ion further constricts the active site and likely contributes to the catalytic preference for acetylpolyamine substrates. The Mg²⁺ ligand D106 is not conserved in mrAPAH and Mg²⁺ binding has not been observed in this related bacterial deacetylase. The binding of additional Mg²⁺ ions as observed in all msAPAH structures may reflect the functional adaptation of this deacetylase as M. subterrani evolved in the high-salt subterranean environment.

A characteristic feature of halophilic proteins is an increase in the number of acidic residues on the protein surface.^59–62 Accordingly, msAPAH has increased aspartate and glutamate content (15% total) compared with mrAPAH (12% total). Increased negative charged accounts for the lower pI of msAPAH compared with that of mrAPAH: based on pI values calculated from their amino acid sequences, the pI of msAPAH is 5.0 and the pI of mrAPAH is 5.3. Additional aspartate residues present in msAPAH also account for additional Mg²⁺ binding sites observed in the crystal structures of complexes with different inhibitors.

The structural features described above likely contribute to the substrate specificity of mrAPAH and msAPAH. Our previous steady-state kinetic studies utilizing a liquid chromatography-mass spectrometry assay for the direct measurement of dansylated polyamine products³ show that mrAPAH exhibits broad substrate specificity and catalyzes the hydrolysis of N-acetylcadaverine, N-acetylputrescine, N¹-acetylspermidine, N⁸-acetylspermidine, and N-acetylspermine with turnover numbers ranging 0.20–2.5 s⁻¹ and catalytic efficiencies ranging from 2.5 × 10³ M⁻¹s⁻¹ to 6.8 × 10⁴ M⁻¹s⁻¹. These results are consistent with previous observations of broad substrate specificity based on specific activity measurements using native and recombinant mrAPAH.^63,64 Similarly, msAPAH exhibits broad substrate specificity but lower catalytic activity using an assay that quantitates the generation of coproduct acetate: turnover numbers range 0.0038–0.026 s⁻¹ and catalytic efficiencies range from 5 M⁻¹s⁻¹ to 470 M⁻¹s⁻¹. We cannot rule out the possibility that msAPAH assay conditions do not fully reflect the cytosolic conditions of the deep earth halophile from which msAPAH derives, nor can we rule out the possibility that the biological substrate of msAPAH in vivo is an as-yet unidentified acetylated small molecule. The biological substrate is unlikely to be acetyllysine, since a peptide or protein substrate would not fit in the constricted approach to the active site (Figure 3).

Recently, eukaryotic HDAC10 was discovered to be a highly specific cytosolic polyamine deacetylase that utilizes N⁸-acetylspermidine as a substrate.³ Although this activity was first observed in partially purified cell extracts more than 40 years ago, the responsible enzyme had not been identified.^35,36 It is interesting to contrast the structural basis of catalytic specificity in the bacterial polyamine deacetylases msAPAH and mrAPAH with that of HDAC10.³⁴ This specificity is rooted in structural and electrostatic features in each enzyme active site. First, each enzyme active site is sterically constricted so as to favor the binding of long, slender polyamine substrates over bulky, acetyllysine-containing peptide substrates. However, the active sites of the bacterial enzymes are constricted by quaternary structure (Figure 3), whereas the active site of the eukaryotic enzyme is constricted by tertiary structure: a 3₁₀ helix unique to HDAC10 and no other HDAC isozymes.^3,34 Second, each enzyme is characterized by significant negative electrostatic potential on the protein surface flanking the active site (Figure 12), as well as specific aspartate and/or glutamate residues in the active site cleft that engage in electrostatic interactions with the positively charged amino groups of polyamine substrates. As noted by Hai and colleagues,³ polyamine substrate specificity in HDAC10 and APAH enzymes appears to have evolved convergently.

In closing, it is interesting to consider the biology and ecology of the halophile Marinobacter subterrani from which msAPAH derives. As noted in the introduction, these species are well adapted to the high Fe²⁺ concentrations found in deep subsurface locations.⁴⁰ Given that HDAC enzymes can utilize Fe²⁺ for catalytic activity slightly exceeding that measured with Zn²⁺, and the implication that metal-dependent deacetylases could utilize Fe²⁺, Zn²⁺, or perhaps a mixture of both in vivo,²³ it is interesting to speculate that msAPAH could potentially function as an Fe²⁺-dependent polyamine deacetylase in the iron-rich environment of the deep terrestrial biosphere. Future structure-function studies of msAPAH will shed further light on this possibility.

Supplementary Material

Supporting Information

Figure S1: IC₅₀ plots; Figure S2: ITC enthalpograms.