Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Biochemistry. Author manuscript; available in PMC 2011 July 6.
Published in final edited form as:
PMCID: PMC2894275

Crystal Structure of Arginase from Plasmodium falciparum and Implications for L-Arginine Depletion in Malarial Infection


The 2.15 Å resolution crystal structure of arginase from Plasmodium falciparum, the parasite that causes cerebral malaria, is reported in complex with the boronic acid inhibitor 2(S)-amino-6-boronohexanoic acid (ABH; Kd = 11 μM). This is the first crystal structure of a parasitic arginase. Various protein constructs were explored to identify an optimally active enzyme form for inhibition and structural studies, and to probe the structure and function of two polypeptide insertions unique to malarial arginase: a 74-residue low complexity region contained in loop L2, and a 11-residue segment contained in loop L8. Structural studies indicate that the low complexity region is largely disordered and is oriented away from the trimer interface; its deletion does not significantly compromise enzyme activity. The loop L8 insertion is located at the trimer interface and makes several intra- and intermolecular interactions important for enzyme function. In addition, we also demonstrate that arg- P. berghei sporozoites show significantly decreased liver infectivity in vivo. Therefore, inhibition of malarial arginase may serve as a possible candidate for antimalarial therapy against liver-stage infection, and ABH may serve as a lead for the development of inhibitors.

Malaria, an infectious disease caused by a parasitic protozoan, is a serious health threat in developing countries due to its facile transmission by the Anopheles mosquito (1, 2). Malaria is currently considered to be eradicated in the US; however, ten species of Anopheles mosquitoes are indigenous to the US, suggesting that malaria acquired in endemic countries can be reintroduced in the US by human travel and propagated by mosquitoes. For example, there were 1,505 cases of malaria diagnosed in the US during 2007, including one transfusion-related case and one fatality (3). Worldwide statistics are significantly more severe, with 243 million cases of malaria being diagnosed and 863,000 deaths resulting in 2008 (primarily children in sub-Saharan Africa) (4). Five species of malaria specifically infect humans: Plasmodium falciparum, vivax, malariae, ovale, and knowlesi; Plasmodium falciparum is the most lethal (5). Since the parasite is capable of developing resistance to drug therapy, combination drug therapies are generally required to circumvent this problem (6, 7).

Malarial infection drastically alters host metabolism since the parasite co-opts essential nutrients for its own survival at the expense of a human host; moreover, the parasite introduces waste products and toxins into the circulatory system of the host (8). The pathogenic course of the disease is characterized by rapidly proliferating parasite cells, massive erythrocyte lysis, and ischemia (9). Clinical manifestations of malaria include hypoglycemia, lactic acidosis, hemolytic anemia, hemoglobinuria, and hypoargininemia (9).

Malaria patients often present with hypoargininemia (10, 11), and metabolomic studies of Plasmodium falciparum during its 48-h intraerythrocytic life cycle reveal nearly complete depletion of L-arginine levels (12). Consistent with this observation, P. falciparum infection does not affect the influx kinetics of L-arginine in erythrocytes (13). Low levels of L-arginine correlate with decreased immunity and nitric oxide (NO)1 production (14). The depletion of L-arginine in culture is achieved by an arginase from the malarial parasite, which catalyzes the hydrolysis of the side chain guanidinium group to form L-ornithine and urea (12). Increased arginase activity characterizes the alternative immune response, which down-regulates inflammation and tissue damage while up-regulating angiogenesis and tissue repair mechanisms (15). Such conditions favor parasite growth through suppressed T cell and inflammatory responses to pathogens and increased concentrations of L-ornithine-derived polyamines, which facilitate cellular proliferation (15).

Increased exogenous arginase activity also characterizes infections by the parasitic protozoan Leishmania major and the bacterium Helicobacter pylori (1618). Consequently, the arg- variants of these organisms do not deplete host L-arginine levels and are accordingly more susceptible to the NO-dependent immune response. Decreased L-arginine levels have been observed in malaria patients and supplemental L-arginine increases endothelial function in patients with moderate to severe malaria (11, 19). Arginase activity is also implicated in slowing the recovery of malaria patients during treatment (20). Considering that current patient outcomes within the first 48 h of therapy do not change between quinine and artesunate treatment, it is proposed that treating endothelial dysfunction by targeting P. falciparum arginase with inhibitors within this time frame may represent a possible adjuvant therapy (20).

The P. falciparum arginase gene has been cloned and characterized (21, 22). Arginase from P. falciparum shares 28% and 27% amino acid sequence identity with human arginases I and II (hAI and hAII), respectively. Like the human arginases, P. falciparum arginase (PFA) is a binuclear manganese metalloenzyme that exists as a trimer with optimal activity at basic pH (21). Unlike the human arginases, PFA contains a large, 74-residue low complexity region (LCR) inserted within loop L2 (Figure 1). Amino acid sequence alignments of arginases from P. falciparum, P. vivax, P. yoelii, P. knowlesi, and P. berghei reveal that this insert is as large as 100 residues in P. vivax and as small as 15 residues in P. berghei and shows no significant sequence conservation (22). Such LCRs are found in more than 90% of all proteins in the P. falciparum genome (2326). A second ~11 residue insert is also located in loop L8 (Figure 1) (22). Despite these polypeptide insertions, PFA still exists as an active trimer (21). The question remains as to what effects, if any, these polypeptide insertions have on arginase structure and function? Kinetic measurements on a PFA construct bearing a C-terminal streptavidin affinity tag initially suggested that malarial arginase is less catalytically efficient than the mammalian arginases, with kcat/KM = 7.4 × 103 M−1s−1 (21) (compared with kcat/KM = 2.0 × 105 M−1s−1 for rat arginase I (27) or 1.27 × 105 M−1s−1 for hAI (28)).

Figure 1
Topology diagram of human arginase I (hAI) and Plasmodium falciparum arginase (PFA). The positions of metal ligands are indicated by black spheres. Amino acid insertions within loops L2 and L8 of PFA relative to hAI are indicated by red dashed lines; ...

Here, we report the preparation of a PFA construct bearing an N-terminal histidine tag instead of a C-terminal streptavidin affinity tag to facilitate purification. Kinetic studies with this new construct show that PFA actually exhibits catalytic efficiency comparable to that of hAI. The 2.15 Å resolution crystal structure of PFA complexed with the boronic acid inhibitor 2(S)-amino-6-boronohexanoic acid (ABH) (29) reveals a comparable binding mode to that observed in the hAI-ABH complex (30) despite significantly weaker inhibitor binding affinity. Structural comparisons between mammalian, bacterial, and parasitic arginases reveal intriguing differences in the stabilization of oligomeric structure, and possible functions for the two polypeptide insertions in PFA are explored. Finally, the study of arg- P. berghei sporozoites confirms the importance of arginase activity in the mechanism of infection, suggesting that inhibition of malarial arginase may serve as a possible candidate for anti-malarial therapy against liver-stage infection.

Materials and Methods

Plasmodium falciparum arginase constructs

The previously reported P. falciparum arginase construct (21, 22) utilized the entire P. falciparum arginase gene with a C-terminal streptavidin affinity tag (PFA-Strep). Here, we report the preparation and analysis of three new arginase constructs: full-length PFA bearing a C-terminal histidine (His) tag (PFA-C6H); an N-terminal truncation beginning with K22 and bearing an N-terminal His-tag (dN-PFA); and a similar N-terminal truncation with residues N84-D157 of loop L2 deleted (dN-PFA-L2S). Additionally, the H381A variant and the L8 chimera (in which loop L8 is replaced with the shorter hAI sequence) were generated from the dN-PFA plasmid. Restriction enzymes and ligase were purchased from New England Biosciences. Taq Hifi Polymerase was purchased from Invitrogen. All cell lines were purchased from Stratagene.

The PFA-C6H plasmid was constructed by first amplifying the P. falciparum arginase coding sequence from parasite genomic DNA (3D7 strain) with primers designed to incorporate a 5′ NheI site in place of the start codon and a 5′ XhoI site in place of the stop codon (primer 1: GCT AGC TTG GAT ACT ATA GAA AGT TAC ATC; primer 2: CTC GAG CAC TAT ATC GTA TCC TAA CAC; restriction sites are underlined). This amplicon was cloned into the pET24a plasmid to give a recombinant gene with an N-terminal Met-Ala-Ser in place of the initial Met and a C-terminal Leu-Glu-6 × His tag.

To generate dN-PFA, the arginase gene was amplified from the C-terminal His-tagged construct PFA-C6H. NheI and HindIII restriction sites were incorporated so as to delete the first 21 amino acids (primer 1: GAG CGC TAG CAA AAA CGT TTC CAT TAT TGG TTC TCC; primer 2: CTC AAG CTT TTA CAC TAT ATC GTA TCC; restriction sites are underlined). Deletion of the corresponding N-terminal segment of hAII facilitated its crystallization (31), so we reasoned that a comparable deletion would facilitate the crystallization of PFA. Digested PCR product was ligated into PET28a (Novagen), utilizing the N-terminal His-tag and thrombin cleavage site, and transformed into XL-1 Blue cells for DNA isolation and sequencing. For the L2-loop deletion construct (dN-PFA-L2S), an AgeI restriction site was incorporated into the dN(1-21) construct L2-loop, resulting in the deletion of residues N84 to D157 (primer 1: CGG ACC GGT AAT ATA AGG AAT ATA AA; primer 2: CGG ACC GGT TTT TTT TTC TTG TTT CAT; restriction sites are underlined). Digested product was purified using a Qiagen extraction kit and transformed into XL-1 Blue cells for DNA isolation and sequencing. Sequencing was performed by the University of Pennsylvania DNA Sequencing Facility, and the identified positive clone was then transformed into E. coli BL21-CodonPlus™(DE3)-RIL cells for expression as previously published (21).

The H381A variant was generated using the QuikChange mutagenesis kit (Stratagene) with the dN-PFA template (primer 1: 5′- GTT GAT AAA AAA GTT gcT GGA GAT TCA TTG CC -3′, primer 2: 5′- GGC AAT GAA TCT CCA gcA ACT TTT TTA TCA AC -3′). The PCR product was transformed into XL1-Blue cells for DNA isolation and sequencing. The construct for the L8 chimera was generated utilizing TA overhangs generated by Platinum Taq (Invitrogen) polymerase. The dN-PFA plasmid without residues D374-T392 was elongated (primer 1: 5′-AAA ACA GGC AAG TTG TGT TTA GAA CTT ATC GCC-3′; primer 2: 5′- CCC AAG TGA TGG ATT ATA TTC TAC TAA ATC C -3′). The insert was purchased from IDT (primer 3: 5′-AG ACA CCA GAA GAA GTA ACT-3′; primer 4: 5′-GTT ACT TCT TCT GGT GTC TT-3′) and annealed separately before ligating with the PCR product. Ligation product was transformed into XL-1 Blue cells for DNA isolation and sequencing, and the identified positive clone was then transformed into E. coli BL21-CodonPlus™(DE3)-RIL cells for expression as previously described (21).

Protein expression and purification

Briefly, transformed or streaked cells were grown on Luria-Bertani (LB) agar supplemented with 50 μg/mL kanamycin. The LB media supplemented with 50 μg/mL kanamycin was inoculated with a single colony and grown for 8 h at 37 °C, 250 rpm. This starter culture was then transferred to 250 mL LB media supplemented with 50 μg/mL kanamycin and 34 μg/mL chloramphenicol and grown for 12–16 h. The 250 mL cell growth was used to inoculate 5 L of either LB or minimal media supplemented with 50 μg/mL kanamycin. Cells were induced with 1 mM IPTG in the presence of 50 μg/mL kanamycin, 10 μg/mL phenylmethylsulfonyl fluoride (PMSF), and 1 μg/mL of tosyl-arginine methyl ester (TAME) at 37 °C.

Cells were harvested by centrifugation at 6000g and resuspended in 50 mM Tris (pH 8.0), 500 mM NaCl, and 3 mM β-mercaptoethanol (BME) in the presence of protease inhibitors (10 μg/mL PMSF and 1 μg/mL of TAME). Cells were lysed by sonication and centrifuged at 30,000g for 1 h. The supernatant was loaded onto Ni-nitrilotriacetic acid resin and eluted with increasing imidazole concentrations (10 mM, 50 mM, 200 mM). Pulled protein fractions were dialyzed into 50 mM Tris (pH 8.0), 200 mM NaCl, 1 mM MnCl2, and 1 mM tris(2-carboxyethyl)phosphine (TCEP), and subsequently concentrated. Protein was loaded onto a Superdex 26/60 size exclusion column preequilibrated with dialysis buffer. The trimer protein peak was collected, concentrated, and stored in aliquots at 4 °C.

Kinetic assays

dN-PFA, dN-PFA-L2S, and PFA-C6H were assayed for catalytic activity using a colorimetric method (32) with slight modifications. The reaction of urea with α-isonitroso propiophenone was measured at a wavelength of 550 nm using the Envision plate reader (courtesy of the Dr. Scott Diamond laboratory, Institute of Medicine and Engineering, University of Pennsylvania). Product formation was determined by means of urea standard curves. No background signal from substrate reacting with α-isonitroso propiophenone was observed when testing up to 400 mM L-arginine. At concentrations above 100 mM L-arginine, a decrease was observed for the dye developing reaction; therefore, standard curves containing stoichiometric ratios of urea and L-arginine were used for reactions above 100 mM L-ariginine. The arginase reaction was performed in 50 mM Tris (pH 8.0), 0.5 mM TCEP, 1 mM MnCl2, 0.05–1 μM protein, and 1–200 mM L-arginine for 5–10 min at 37 °C. Assay mixture and protein were preincubated at 37 °C for >1 min before initiating the reaction. The assay mixture (20 μL) was stopped with a sulfuric-phosphoric acid/α-isonitroso propiophenone mixture (140 μL). Reaction points were developed in a thermocycler (90 °C, 1 hr), followed by incubation at room temperature (21 °C, 15 min). The dN-PFA expressed in minimal media exhibited increased activity and was use for kcat calculations as measured by linear rates at 200 mM L-arginine with 50 nM protein. Kinetic parameters were determined with the software Graphpad Prism (2008). Inhibition studies were performed with 25 mM L-arginine at 37 °C for 10 min to obtain significant absorbance signals to plot IC50 data. The Ki value for ABH was calculated using the Cheng-Prusoff equation (33). Results are recorded in Table 1.

Table 1
Kinetic parameters for P. falciparum arginase (PFA) constructs

Isothermal titration calorimetry of dN-PFA-ABH complexation

Experiments were conducted on a ITC200 isothermal calorimeter from MicroCal, Inc. in the laboratory of Dr. William DeGrado at the University of Pennsylvania. Enzyme (50 μM) was dialyzed against 50 mM Tris (pH 8.0), 200 mM NaCl, 1 mM MnCl2, and 1 mM TCEP; 0.65 mM ABH was dissolved in dialysis buffer. The sample cell (0.3 mL) was overfilled and the reference cell was filled with distilled water. The inhibitor ABH was titrated into the sample cell with 20 sequential aliquots (2 μL each; an initial 0.2 μL injection was made, but not used in data analysis). A control experiment titrating the inhibitor into buffer was subtracted out, and data analysis was performed using ORIGIN 7.0 software.


The inhibitor ABH was synthesized as previously described (29). The construct dN-PFA was used successfully in crystallization experiments. Briefly, a 2 μL sitting drop of 5–10 mg/mL dN-PFA in 50 mM Tris (pH 8.0), 200 mM NaCl, 1 mM MnCl2, 1 mM TCEP, and 5 mM ABH was mixed with a drop of 1.4 M sodium/potassium phosphate (pH 8.2) and equilibrated against a 500 μL reservoir of 40–50% 2-methyl-2,4-pentanediol (MPD) at room temperature. Small crystals appeared overnight and grew to typical dimensions of 50 × 50 × 50 μm3. Crystals were harvested and cryoprotected in 30% Jeffamine ED-2001 and 0.1 M HEPES (pH 7.0), and then flash cooled in liquid nitrogen. Diffraction data were measured on beamline 24-ID-E at the Advanced Photon Source (APS, Argonne, IL). Crystal parameters and data collection statistics are recorded in Table 2.

Table 2
Data collection and refinement statistics

Data were indexed and merged using HKL2000 (34). Molecular replacement calculations were performed with PHASER (35) using the atomic coordinates of hAII less inhibitor and solvent atoms (PDB accession code 1PQ3) (31) as a search probe for rotation and translation function calculations. Iterative cycles of refinement and model building were performed using PHENIX (36) and COOT (37), respectively, in order to improve each structure as guided by Rfree. The N-terminal His-tag is completely disordered, and electron density is observable beginning at residue K22 of the dN-PFA protein sequence. Residues G72 - N153 of the L2-loop insert are disordered and excluded from the final model. Two cis-peptide linkages are found in the refined structure: residues G190 and G191 immediately after β3, and residues G382 and D383 within the loop L8 insert. All refinement statistics are recorded in Table 2. Buried surface area calculations were performed using the PISA server ( (38).

Mice and Plasmodium Liver Infection

C57BL/6 mice were housed in the Instituto de Medicina Molecular (IMM) facilities and experiments were performed under IMM Animal Care Committee approval following National and EU guidelines. The arg- (12) and wild-type P. berghei sporozoites were obtained by dissection of Anopheles stephensi infected mosquitoes bred at the IMM insectarium. Mice were infected by intravenous inoculation of 3 × 104 sporozoites and parasite liver load was quantified 40 h post-infection by quantitative RT-PCR, as previously described (39). Relative amount of P. berghei ANKA 18S rRNA was calculated against the hypoxanthine guanine phosphoribosyltransferase (hprt) housekeeping gene. The PbA 18S rRNA and hprt specific primer sequences were: 5′-CGG CTT AAT TTG ACT CAA CAC G-3′ and 5′-TTA GCA TGC CAG AGT CTC GTT C-3′ for PbA 18S rRNA, and 5′ – TGC TCGAGA TGT GAT GAA GG – 3′ and 5′ – TCC CCT GTT GAC TGG TCA TT – 3′ for mouse hprt. External standardization was performed using plasmids encoding the full-length of P. berghei 18S rRNA and hprt cDNA cloned in TOPO TA (Invitrogen).

Results and Discussion

Catalytic activity of PFA constructs

The catalytic efficiency (kcat/KM) of dN-PFA increases with increasing pH, and kinetic parameters are comparable to those measured for human arginase I (Table 1). Importantly, the dN-PFA construct is the most catalytically active form of PFA reported to date. While the KM values measured for the PFA constructs reported in Table 1 are in the same range as KM values measured for mammalian arginases (KM = 1.4 mM for rat arginase I, pH 9.0 (40); KM = 1.5 mM for hAI, pH 9.0 (41); KM = 0.55 mM for hAII, pH 9 (42)), the KM value of 12 mM measured for the PFA-C6H construct at pH 8.0, as well as the KM value of 13 mM measured for the C-terminal streptavidin affinity tag construct reported by Müller and colleagues at pH 8.0 (PFA-Strep) (21), are notably higher. Therefore, it appears that a C-terminal affinity tag on PFA is detrimental for substrate binding and catalytic activity.

The tightest binding inhibitor of rat and human arginases known to date is the boronic acid analogue of L-arginine, ABH (29), which inhibits hAI with Kd = 5 nM (30). Indeed, ABH is an orally bioavailable drug lead that shows promising in vivo activity for the treatment of cardiovascular diseases such as erectile dysfunction and atherosclerosis in animal models (31, 43). However, the inhibition of dN-PFA by ABH at physiological pH is significantly weaker, with Ki = 11 ± 2 μM assuming competitive inhibition (Table 1). This affinity was confirmed by isothermal titration calorimetry (pH 8.0) with Kd= 11 ± 2 μM (Figure 2). The origin of weaker binding to dN-PFA is not evident in the crystal structure of the enzyme-inhibitor complex, since ABH appears to make identical interactions in the active sites of hAI and dN-PFA (vide infra).

Figure 2
Isothermal titration calorimetry of the PFA-ABH complex in 50 mM Tris (pH 8.0), 200 mM NaCl, 1 mM TCEP, and 1 mM MnCl2 at 25 °C. Shown are the raw data obtained by titration of 50 μM arginase with 20 × 2 μL injections of ...

To assess the functions of the two amino acid sequence inserts in PFA, two deletion constructs were generated. Approximately 70 residues (N84–D157) of the 74-residue LCR were deleted from the L2 loop to generate the dN-PFA-L2S construct. A survey of arginase crystal structures in the Protein Data Bank suggested that this deletion construct would be able to fold properly. Similarly, residues D374-T392 of loop L8 were replaced with the corresponding sequence from human arginase I, yielding the dN-PFA-L8 chimera. While these constructs do not express well, soluble protein expression is confirmed by Western blotting analysis using an anti-His antibody (data not shown). Due to the significant decrease in protein purity and expression, accurate protein concentrations were difficult to estimate. However, a KM value of 4.0 ± 0.3 mM was determined for the L2S construct with Vmax = 2.3 ± 0.1 μM s−1, and a KM value of 20 ± 4 mM was determined for the L8 chimera with Vmax = 2.6 ± 0.5 μM s−1. The KM value for the L2S construct is in accord with that measured for the dN-PFA construct, indicating that the LCR does not significantly impact substrate affinity. The L8 chimera shows moderately decreased substrate affinity; moreover, this construct shows substrate inhibition with Ki = 130 ± 50 mM (data not shown).

Parasitic LCR inserts are expected to exist primarily as unstructured domains made up of hydrophilic and flexible amino acids such as Asn, Lys, Glu, and Asp (26), and this is the case for the 74-residue LCR in loop L2 of PFA. In PFA, the LCR insert does not affect substrate binding, and the protein readily crystallizes despite its presence (vide infra). While the possible functions of LCR inserts are not fully understood, they are thought to be external, nonglobular domains that do not necessarily affect protein function. However, examples are known in which LCR inserts affect enzyme activity and quaternary structure, including dihydrofolate reductasethymidylate synthase, subtilisin-like protease-1, and S-adenosylmethionine decarboxylase/ornithine decarboxylase of P. falciparum (4446).

Crystal Structure

Since the dN-PFA construct was utilized for the X-ray crystal structure determination, the acronym “PFA” refers specifically to this construct in the remainder of this paper. PFA crystallizes as a loosely-associated dimer of trimers related by a crystallographic two-fold axis. The observed quaternary structure is identical to that of the hexameric arginases from Thermus thermophilus and Bacillus caldovelox (47, 48), except that there is less than 5% buried surface area between the symmetry-related trimers of PFA. Dynamic light scattering experiments indicate that PFA is a trimer in solution (data not shown).

The PFA-ABH and hAI-ABH monomers superimpose with an r.m.s. deviation of 0.93 Å for 245 Cα residues. Significant structural differences are observed near the two sequence insertions in PFA (Figure 3a). While the 11-residue insert in loop L8 is well ordered, the 74 residue LCR insert in loop L2 is largely disordered: only the flanking residues D70-N71 and C154-N158 are modeled into visible electron density, and C154 is refined with half occupancy. The LCR is distant from the active site and is oriented away from the trimer interface (Figure 3b). Thus, the LCR does not significantly impact enzyme structure or function.

Figure 3
(a) Superposition of the hAI monomer (yellow, PDB accession code 2AEB (30)) with the PFA monomer (green), both in complex with the inhibitor 2(S)-amino-6-boronohexanoic acid (ABH). Mn2+ ions are shown as purple spheres, and the inhibitor ABH is colored ...

The structure of the binuclear manganese cluster in the PFA-ABH complex is very similar to that observed in the rat and human arginase I complexes (30, 49). Both metal ions exhibit distorted octahedral coordination geometry (Figure 4a). The electron density for the inhibitor ABH indicates that the boronic acid moiety undergoes nucleophilic attack by the metal-bridging hydroxide ion to yield a tetrahedral boronate anion. This binding mode mimics the tetrahedral intermediate and its flanking transition states in the hydrolysis of L-arginine (50).

Figure 4
(a) Simulated annealing omit map of 2(S)-amino-6-boronohexanoic acid (ABH) contoured at 5σ. The binding mode of ABH is similar to that observed for binding to mammalian arginases (30, 59). Intriguingly, H381B from an adjacent monomer (green) extends ...

The molecular recognition of the α-amino and α-carboxylate groups of ABH is mediated by several direct and water-mediated interactions with active site residues, most of which are also observed in the hAI-ABH complex (30) (Figure 4b). Water-mediated hydrogen bonds are particularly notable. Two water molecules interact with the α-amino group of ABH. One water molecule is within hydrogen bonding distance to E277 and the backbone C=O of G234, and the other is within hydrogen bonding distance to the carboxylate side chains of D274 and D272, and the backbone C=O group of D272. This water molecule also makes a van der Waals contact with H381 (vide infra). Three water molecules are observed within hydrogen bonding distance to the α-carboxylate moiety of ABH and mediate interactions with S227, N231, H233, and H381 from an adjacent monomer.

Several hydrogen bonded salt links in the bacterial and mammalian arginases are known to be important for stabilization of the trimeric quaternary structure (27, 41, 51, 52), and some of these salt links are conserved in PFA. Specifically, at the center of the trimer interface, the hAI salt link network involving D204A, E256A, and R255B is conserved as E295A, E347A, and R346B in PFA (subscripts “A” and “B” indicate adjacent monomers) (Figure 5). The aspartate→glutamate substitution results in a conformational change that nonetheless maintains the hydrogen bond interaction with the central arginine residue. However, in hAI, D204A makes a syn-oriented hydrogen bond with R255B, whereas in PFA, E295 makes an anti-oriented hydrogen bond with R346B.

Figure 5
The arginase trimer contains several salt links at the trimer interface. The hAI subunit carbon atoms are red, blue, and green; oxygen and nitrogen atoms are red and blue, respectively. The PFA trimer subunits are colored similarly, except that carbon ...

A second salt link network observed in mammalian arginases and the arginase from T. thermophilus is not conserved within PFA. This network consists of D204A, E262B, and R308B in hAI and is linked to the conserved salt link network described above through residue D204A, which accepts hydrogen bonds from both R255B and R308B of an adjacent monomer (Figure 5). Interestingly, arginase from B. caldovelox does not contain the corresponding R308 residue, but a free L-arginine molecule is reported to bind and make interactions similar to those of R308 in the mammalian arginases (47).

The 11-residue insertion in loop L8 influences the structure and function of PFA. As previously noted in the structure determination of rat arginase I (53), the C-terminal polypeptide mediates more than 50% of intermonomer contact surface area. The enlarged L8 loop of PFA precedes helix H2 and the C-terminal polypeptide and augments intermonomer contacts (Figures 3b, ,6).6). Interestingly, K340A from loop L7 protrudes into a pocket in the adjacent monomer defined by residues from helices G and H2 and loop L8 (Y345B, D377B, K378B, D383B, K393B, K396B, and E400B). Of these residues, only Y345 is conserved as Y254 in hAI, so it is clear that the complementary surface area at the subunit interface has evolved to accommodate the loop L8 insertion. Parenthetically, we note that electron density for the side chain of K378 was not as well defined as for other residues in this pocket. However, the main chain atoms of K378 were built into well defined electron density, and the side chain was modeled with a conformation such that the Nζ atom was located near that of K340.

Figure 6
Unique inter- and intra-monomer interactions involving loop L8 and helices G and H2 of PFA. PFA monomers are green and yellow, and the L8 loop is red. The superimposed hAI structure is grey. Red dashed lines indicate hydrogen bonds and cyan dashed lines ...

Another intermonomer interaction is observed for loop L8 of PFA, in that H381A is within hydrogen bonding distance to residue D272B in the active site; H381A also makes a solvent-mediated interaction with the carboxylate group of ABHB (Figures 3 and and4).4). Notably, H381A is 3.4 Å from the Cβ atom of the inhibitor molecule ABHB, interacting through van der Waals interactions. A water molecule interacts with the α-amino group of ABHB, D272B, D274B, and is also located within hydrogen bonding distance to H381A. Intermonomer interactions with the bound inhibitor are unique to the active site of PFA. Since H381 can make an intermonomer van der Waals contact with a bound inhibitor, it can also make a similar interaction with a bound substrate molecule. The influence of H381 on enzyme activity was probed by preparing the H381A variant, which exhibits modestly diminished kinetic parameters in comparison with dN-PFA; the binding affinity of ABH is also modestly compromised (Table 1). These results suggest that H381 contributes to substrate binding and catalysis.

Parenthetically, we note that a new intramonomer hydrogen bond network is observed in PFA involving R404A, E401A, and Q349A (Figure 6). Although these residues are located near the trimer interface, none make intermonomer interactions. However, the R404A mutation results in a partially active monomer (22), suggesting that the R404-E401-Q349 hydrogen bond network helps maintain the structure of the monomer near the subunit interface to facilitate trimer assembly.

Altered infectivity of arg- parasites

Previous work has suggested a possible connection between the plasmodial arginase and immune evasion by the parasite (12). However, disruption of the parasitic arginase in the rodent malaria model Plasmodium berghei does not compromise the in vivo viability in arginase knockout parasites, i.e., proliferation in blood-stage infection is not affected (12). This previous study used inoculations containing 106 parasitized red blood cells to infect BALB/c mice. Here, we have compared the infectivity of arg- and wild-type P. berghei by inoculation of mice using sporozoites dissected from mosquito salivary glands. As opposed to direct infection with blood-stage parasites, sporozoites must first infect liver cells to produce merozoites that invade red blood cells and produce the manifestations of clinical malaria. Using qRT-PCR of P. berghei 18S rRNA to measure the degree of infectivity, we find a significant reduction (p < 0.01) in infectivity in the arg- strain 40 h post-infection (Figure 7), suggesting that liver-stage infection is compromised by parasite arginase deficiency. Wild-type Plasmodium sporozoites and developing liver forms express arginase (arg), which functionally degrades L-arginine. We hypothesize that during infection with wild-type parasites, L-arginine levels are lowered and as such the production of nitric oxide (NO) is likely reduced. This may allow the parasite to evade an NO-dependent immune response in the host. Indeed, the apparent reduction in liver-stage infection observed in arg- P. berghei parasites is consistent with this hypothesis, although an increase in the production of L-ornithine by the parasitic arginase may also be important for parasite survival. The decreased liver-stage infectivity of arg- parasites clearly warrants further investigation. Arginase inhibitors, perhaps in synergistic combination with inhibitors of parasite polyamine biosynthesis (54), may prove to be effective against liver-stage infection and may possibly serve as prophylactic agents in addition to their potential use as adjuvant therapies.

Figure 7
P. berghei sporozoite infection shows a dependency on plasmodial arginase expression. Sporozoites from wild-type (WT) and arg- (12) P. berghei strains were dissected from mosquito salivary glands and used to infect C57BL/6 mice. Data for WT and arg- ...

Concluding Remarks

It is well documented that arginase activity can regulate NO biosynthesis and thereby facilitate immune evasion (18, 55, 56). The arginase can be that of the host, e.g., in certain cancer tumor cells (57) or circulating myeloid-derived suppressor cells in cancer patients (56), or it can be that of the invading pathogen, e.g., H. pylori arginase (58). In the current work, disruption of P. falciparum arginase compromises in vivo viability in arginase knockout sporozoites, possibly by decreasing the ability of the parasite to modify the surrounding environment to its benefit and promote survival, and possibly by exposing the less abundant and metabolically active liver-stage parasite to NO-mediated killing. Interestingly, exogenous arginase activity in L. major or H. pylori infection similarly depletes host L-arginine levels and results in impaired T cell responses, and arginase inhibitors or supplemental L-arginine attenuates infection and parasite growth (1618).

Analysis of P. falciparum arginase may explain decreased substrate and inhibitor potencies compared with human arginase I. Specifically, novel interactions are observed at the trimer interface due to the conserved 11-residue loop L8 insertion, and H381 of this loop extends from one monomer into the active site of an adjacent monomer in the trimer. The 74-residue LCR domain is disordered in the crystal structure and its deletion does not affect substrate affinity. Thus, the three-dimensional structure of PFA promises to guide the future design and development of inhibitors that may be useful adjuvants in the treatment of malarial infections.


This work is based upon research conducted at the Northeastern Collaborative Access Team beamlines of the Advanced Photon Source, which are supported by award RR-15301 from the National Center for Research Resources at the National Institutes of Health. Use of the Advanced Photon Source is supported by the U.S. Department of Energy, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. We thank Dr. Scott Diamond for his assistance with kinetic colorimetric measurements, and Dr. William DeGrado and Emma Magavern for assistance with the isothermal titration calorimetry measurements.

This work was supported by National Institutes of Health grant GM49758 (D.W.C.), an NIH Director's New Innovator award (1DP2OD001315-01 to M.L.), the Burroughs Wellcome Fund (M.L.), a National Science Foundation Graduate Research Fellowship (K.L.O.), the Howard Hughes Medical Institute (M.M.M.), and the Fundação para a Ciência e Tecnologia (FCT, Portugal) (S.P. and M.M.M.).


The atomic coordinates of the P. falciparum arginase-ABH complex have been deposited in the Protein Data Bank ( with accession code 3MMR.

1Abbreviations: ABH, 2(S)-amino-6-boronohexanoic acid; BME, β-mercaptoethanol; hAI and hAII, human arginases I and II; dN-PFA, PFA beginning at residue K22 bearing an N-terminal His-tag; L2S, PFA construct containing a 70 residue deletion in the L2 loop; LCR, low complexity region; MPD, 2-methyl-2,4-pentanediol; NO, nitric oxide; PFA, Plasmodium falciparum arginase; PMSF, phenylmethylsulfonyl fluoride; TAME, tosyl-arginine methyl ester; TCEP, tris(2-carboxyethyl)-phosphine hydrochloride.


(1) Snow RW, Guerra CA, Noor AM, Myint HY, Hay SI. The global distribution of clinical episodes of Plasmodium falciparum malaria. Nature. 2005;434:214–217. [PMC free article] [PubMed]
(2) Greenwood BM, Fidock DA, Kyle DE, Kappe SHI, Alonso PL, Collins FH, Duffy PE. Malaria: progress, perils, and prospects for eradication. J. Clin. Invest. 2008;118:1266–1276. [PMC free article] [PubMed]
(3) C.D.C. Malaria Surveillance - United States. Morbidity and Mortality Weekly Report. 2007;58:1–10.
(4) W.H.O. World Malaria Report. 2009. pp. 1–163.
(5) Cox-Singh J, Davis TME, Lee K-S, Shamsul SSG, Matusop A, Ratnam S, Rahman HA, Conway DJ, Singh B. Plasmodium knowlesi malaria in humans is widely distributed and potentially life threatening. Clin. Infect. Dis. 2008;46:165–171. [PMC free article] [PubMed]
(6) Winstanley PA, Ward SA, Snow RW. Clinical status and implications of antimalarial drug resistance. Microbes Infect. 2002;4:157–164. [PubMed]
(7) Kelly JX, Smilkstein MJ, Brun R, Wittlin S, Cooper RA, Lane KD, Janowsky A, Johnson RA, Dodean RA, Winter R, Hinrichs DJ, Riscoe MK. Discovery of dual function acridones as a new antimalarial chemotype. Nature. 2009;459:270–273. [PubMed]
(8) Planche T, Dzeing A, Ngou-Milama E, Kombila M, Stacpoole PW. Metabolic complications of severe malaria. Curr. Top. Microbiol. Immunol. 2005;295:105–136. [PubMed]
(9) Mackintosh CL, Beeson JG, Marsh K. Clinical features and pathogenesis of severe malaria. Trends Parasitol. 2004;20:597–603. [PubMed]
(10) Gramaglia I, Sobolewski P, Meays D, Contreras R, Nolan JP, Frangos JA, Intaglietta M, van der Heyde HC. Low nitric oxide bioavailability contributes to the genesis of experimental cerebral malaria. Nat. Med. 2006;12:1417–1422. [PubMed]
(11) Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, McNeil YR, Darcy CJ, Granger DL, Weinberg JB, Lopansri BK, Price RN, Duffull SB, Celermajer DS, Anstey NM. Impaired nitric oxide bioavailability and L-arginine-reversible endothelial dysfunction in adults with falciparum malaria. J. Exp. Med. 2007;204:2693–2704. [PMC free article] [PubMed]
(12) Olszewski KL, Morrisey JM, Wilinski D, Burns JM, Vaidya AB, Rabinowitz JD, Llinás M. Host-parasite interactions revealed by Plasmodium falciparum metabolomics. Cell Host Microbe. 2009;5:191–199. [PMC free article] [PubMed]
(13) Elford BC, Haynes JD, Chulay JD, Wilson RJM. Selective stage-specific changes in the permeability to small hydrophilic solutes of human erythrocytes infected with Plasmodium falciparum. Mol. Biochem. Parasitol. 1985;16:43–60. [PubMed]
(14) Modolell M, Corraliza IM, Link F, Soler G, Eichmann K. Reciprocal regulation of the nitric oxide synthase/arginase balance in mouse bone marrow-derived macrophages by TH1 and TH2 cytokines. Eur. J. Immunol. 1995;25:1101–1104. [PubMed]
(15) Peranzoni E, Marigo I, Dolcetti L, Ugel S, Sonda N, Taschin E, Mantelli B, Bronte V, Zanovello P. Role of arginine metabolism in immunity and immunopathology. Immunobiol. 2007;212:795–812. [PubMed]
(16) Muleme HM, Reguera RM, Berard A, Azinwi R, Jia P, Okwor IB, Beverley S, Uzonna JE. Infection with arginase-deficient Leishmania major reveals a parasite number-dependent and cytokine-independent regulation of host cellular arginase activity and disease pathogensis. J. Immunol. 2009;183:8068–8076. [PMC free article] [PubMed]
(17) Modolell M, Choi B-S, Ryan RO, Hancock M, Titus RG, Abebe T, Hailu A, Müller I, Rogers ME, Bangham CRM, Munder M, Kropf P. Local suppression of T cell responses by arginase-induced L-arginine depletion in nonhealing leishmaniasis. PLoS Negl. Trop. Dis. 2009;3:e480. [PMC free article] [PubMed]
(18) Chaturvedi R, Asim M, Lewis ND, Algood HMS, Cover TL, Kim PY, Wilson KT. L-Arginine availability regulates inducible nitric oxide synthase-dependent host defense against Helicobacter pylori. Infect. Immun. 2007;75:4305–4315. [PMC free article] [PubMed]
(19) Weinberg JB, Lopansri BK, Mwaikambo E, Granger DL. Arginine, nitric oxide, carbon monoxide, and endothelial function in severe malaria. Curr. Opin. Infect. Dis. 2008;21:468–475. [PMC free article] [PubMed]
(20) Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, McNeil YR, Darcy CJ, Granger DL, Weinberg JB, Lopansri BK, Price RN, Duffull SB, Celermajer DS, Anstey NM. Recovery of endothelial function in severe falciparum malaria: relationship with improvement in plasma L-arginine and blood lactate concentrations. J. Infect. Dis. 2008;198:602–608. [PMC free article] [PubMed]
(21) Müller IB, Walter RD, Wrenger C. Structural metal dependency of the arginase from the human malaria parasite Plasmodium falciparum. Biol. Chem. 2005;386:117–126. [PubMed]
(22) Wells GA, Müller IB, Wrenger C, Louw AI. The activity of Plasmodium falciparum arginase is mediated by a novel inter-monomer salt-bridge between Glu295-Arg404. FEBS J. 2009;276:3517–3530. [PubMed]
(23) Aravind L, Iyer LM, Wellems TE, Miller LH. Plasmodium biology: genomic gleanings. Cell. 2003;115:771–785. [PubMed]
(24) DePristo MA, Zilversmit MM, Hartl DL. On the abundance, amino acid composition, and evolutionary dynamics of low-complexity regions in proteins. Gene. 2006;378:19–30. [PubMed]
(25) Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, Paulsen IT, James K, Eisen JA, Rutherford K, Salzberg SL, Craig A, Kyes S, Chan M-S, Nene V, Shallom SJ, Suh B, Peterson J, Angiuoli S, Pertea M, Allen J, Selengut J, Haft D, Mather MW, Vaidya AB, Martin DMA, Fairlamb AH, Fraunholz MJ, Roos DS, Ralph SA, McFadden GI, Cummings LM, Subramanian GM, Mungall C, Venter JC, Carucci DJ, Hoffman SL, Newbold C, Davis RW, Fraser CM, Barrell B. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002;419:498–511. [PMC free article] [PubMed]
(26) Pizzi E, Frontali C. Low-complexity regions in Plasmodium falciparum proteins. Genome Res. 2001;11:218–229. [PubMed]
(27) Lavulo LT, Sossong TM, Jr., Brigham-Burke MR, Doyle ML, Cox JD, Christianson DW, Ash DE. Subunit-subunit interactions in trimeric arginase. Generation of active monomers by mutation of a single amino acid. J. Biol. Chem. 2001;276:14242–14248. [PubMed]
(28) Alarcón R, Orellana MS, Neira B, Uribe E, García JR, Carvajal N. Mutational analysis of substrate recognition by human arginase type I -agmatinase activity of the N130D variant. FEBS J. 2006;273:5625–5631. [PubMed]
(29) Baggio R, Elbaum D, Kanyo ZF, Carroll PJ, Cavalli RC, Ash DE, Christianson DW. Inhibition of Mn2+2-arginase by borate leads to the design of a transition state analogue inhibitor, 2(S)-amino-6-boronohexanoic acid. J. Am. Chem. Soc. 1997;119:8107–8108.
(30) Di Costanzo L, Sabio G, Mora A, Rodriguez PC, Ochoa AC, Centeno F, Christianson DW. Crystal structure of human arginase I at 1.29-Å resolution and exploration of inhibition in the immune response. Proc. Natl. Acad. Sci. U. S. A. 2005;102:13058–13063. [PubMed]
(31) Cama E, Colleluori DM, Emig FA, Shin H, Kim SW, Kim NN, Traish AM, Ash DE, Christianson DW. Human arginase II: crystal structure and physiological role in male and female sexual arousal. Biochemistry. 2003;42:8445–8451. [PubMed]
(32) Archibald RM. Colorimetric determination of urea. J. Biol. Chem. 1945;157:507–518.
(33) Cheng Y-C, Prusoff WH. Relationship between the inhibition constant (KI) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 1973;22:3099–3108. [PubMed]
(34) Otwinowski Z, Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997;276:307–326.
(35) McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J. Appl. Cryst. 2007;40:658–674. [PubMed]
(36) Adams PD, Grosse-Kunstleve RW, Hung L-W, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK, Terwilliger TC. PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr. D. 2002;58:1948–1954. [PubMed]
(37) Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D. 2004;60:2126–2132. [PubMed]
(38) Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 2007;372:774–797. [PubMed]
(39) Bruña-Romero O, Hafalla JCR, González-Aseguinolaza G, Sano G, Tsuji M, Zavala F. Detection of malaria liver-stages in mice infected through the bite of a single Anopheles mosquito using a highly sensitive real-time PCR. Int. J. Parasitol. 2001;31:1499–1502. [PubMed]
(40) Cavalli RC, Burke CJ, Soprano DR, Kawamoto S, Ash DE. Mutagenesis of rat liver arginase expressed in Escherichia coli: role of conserved histidines. Biochemistry. 1994;33:10652–10657. [PubMed]
(41) García D, Uribe E, Lobos M, Orellana MS, Carvajal N. Studies on the functional significance of a C-terminal S-shaped motif in human arginase type I: Essentiality for cooperative effects. Arch. Biochem. Biophys. 2009;481:16–20. [PubMed]
(42) Colleluori DM, Morris SM, Jr., Ash DE. Expression, purification, and characterization of human type II arginase. Arch. Biochem. Biophys. 2001;389:135–143. [PubMed]
(43) Ryoo S, Gupta G, Benjo A, Lim HK, Camara A, Sikka G, Lim HK, Sohi J, Santhanam L, Soucy K, Tuday E, Baraban E, Ilies M, Gerstenblith G, Nyhan D, Shoukas A, Christianson DW, Alp NJ, Champion HC, Huso D, Berkowitz DE. Endothelial arginase II: a novel target for the treatment of atherosclerosis. Circ. Res. 2008;102:923–932. [PubMed]
(44) Birkholtz L-M, Wrenger C, Joubert F, Wells GA, Walter RD, Louw AI. Parasite-specific inserts in the bifunctional S-adenosylmethionine decarboxylase/ornithine decarboxylase of Plasmodium falciparum modulate catalytic activities and domain interactions. Biochem. J. 2004;377:439–448. [PubMed]
(45) Jean L, Withers-Martinez C, Hackett F, Blackman MJ. Unique insertions within Plasmodium falciparum subtilisin-like protease-1 are crucial for enzyme maturation and activity. Mol. Biochem. Parasitol. 2005;144:187–197. [PubMed]
(46) Yuvaniyama J, Chitnumsub P, Kamchonwongpaisan S, Vanichtanankul J, Sirawaraporn W, Taylor P, Walkinshaw MD, Yuthavong Y. Insights into antifolate resistance from malarial DHFR-TS structures. Nat. Struct. Biol. 2003;10:357–365. [PubMed]
(47) Bewley MC, Jeffrey PD, Patchett ML, Kanyo ZF, Baker EN. Crystal structures of Bacillus caldovelox arginase in complex with substrate and inhibitors reveal new insights into activation, inhibition and catalysis in the arginase superfamily. Structure. 1999;7:435–448. [PubMed]
(48) Kumarevel TS, Karthe P, Kuramitsu S, Yokoyama S, RSGI Crystal structure of the arginase from Thermus thermophilus. 2009. PDB accession code: 2EF4.
(49) Cox JD, Kim NN, Traish AM, Christianson DW. Arginaseboronic acid complex highlights a physiological role in erectile function. Nat. Struct. Biol. 1999;6:1043–1047. [PubMed]
(50) Christianson DW. Arginase: structure, mechanism, and physiological role in male and female sexual arousal. Acc. Chem. Res. 2005;38:191–201. [PubMed]
(51) Ash DE, Scolnick LR, Kanyo ZF, Vockley JG, Cederbaum SD, Christianson DW. Molecular basis of hyperargininemia: structure-function consequences of mutations in human liver arginase. Mol. Genet. Metab. 1998;64:243–249. [PubMed]
(52) Sabio G, Mora A, Rangel MA, Quesada A, Marcos CF, Alonso JC, Soler G, Centeno F. Glu-256 is a main structural determinant for oligomerisation of human arginase I. FEBS Lett. 2001;501:161–165. [PubMed]
(53) Kanyo ZF, Scolnick LR, Ash DE, Christianson DW. Structure of a unique binuclear manganese cluster in arginase. Nature. 1996;383:554–557. [PubMed]
(54) van Brummelen AC, Olszewski KL, Wilinski D, Llinás M, Louw AI, Birkholtz LM. Co-inhibition of Plasmodium falciparum S-adenosylmethionine decarboxylase/ornithine decarboxylase reveals perturbation-specific compensatory mechanisms by transcriptome, proteome, and metabolome analyses. J. Biol. Chem. 2009;284:4635–4646. [PubMed]
(55) Ochoa AC, Zea AH, Hernandez C, Rodriguez PC. Arginase, prostaglandins, and myeloid-derived suppressor cells in renal cell carcinoma. Clin. Cancer Res. 2007;13:721s–726s. [PubMed]
(56) Zea AH, Rodriguez PC, Atkins MB, Hernandez C, Signoretti S, Zabaleta J, McDermott D, Quiceno D, Youmans A, O'Neill A, Mier J, Ochoa AC. Arginase-producing myeloid suppressor cells in renal cell carcinoma patients: a mechanism of tumor evasion. Cancer Res. 2005;65:3044–3048. [PubMed]
(57) Singh R, Pervin S, Karimi A, Cederbaum S, Chaudhuri G. Arginase activity in human breast cancer cell lines: N(ω)-hydroxy-L-arginine selectively inhibits cell proliferation and induces apoptosis in MDA-MB-468 cells. Cancer Res. 2000;60:3305–3312. [PubMed]
(58) Gobert AP, McGee DJ, Akhtar M, Mendz GL, Newton JC, Cheng Y, Mobley HLT, Wilson KT. Helicobacter pylori arginase inhibits nitric oxide production by eukaryotic cells: a strategy for bacterial survival. Proc. Natl. Acad. Sci. U. S. A. 2001;98:13844–13849. [PubMed]
(59) Cama E, Pethe S, Boucher JL, Han S, Emig FA, Ash DE, Viola RE, Mansuy D, Christianson DW. Inhibitor coordination interactions in the binuclear manganese cluster of arginase. Biochemistry. 2004;43:8987–8999. [PubMed]