Identification of the High Affinity Mn2+ Binding Site of Bacteriophage λ Phosphoprotein Phosphatase: Effects of Metal Ligand Mutations on Electron Paramagnetic Resonance Spectra and Phosphatase Activities†

Bacteriophage λ phosphoprotein phosphatase (λPP) has structural similarity to the mammalian Ser/Thr phosphoprotein phosphatases (PPPs) including the immunosuppressant drug target calcineurin. PPPs possess a conserved active site containing a dinuclear metal cluster, with metal ligands provided by a phosphoesterase motif plus two additional histidine residues at the C-terminus. Multiple sequence alignment of λPP with 28 eubacterial and archeal phosphoesterases identified active site residues from the phosphoesterase motif and in many cases 2 additional C-terminal His metal ligands. Most highly similar to λPP are E. coli PrpA and PrpB. Using the crystal structure of λPP [Voegtli, W. C., et al. (2000) Biochemistry 39, 15365-15374] as a structural and active site model for PPPs and related bacterial phosphoesterases, we have studied mutant forms of λPP reconstituted with Mn2+ by electron paramagnetic resonance (EPR) spectroscopy, Mn2+ binding analysis, and phosphatase kinetics. Analysis of Mn2+-bound active site mutant λPP proteins shows that H22N, N75H, and H186N mutations decrease phosphatase activity but still allow mononuclear Mn2+ and [(Mn)2] binding. The high affinity Mn2+ binding site is shown to consist of M2 site ligands H186 and Asn75, but not H22 from the M1 site which is ascribed as the lower affinity site. Reversible phosphorylation of proteins is essential for the regulation of metabolism in all organisms and involves the transfer of the phosphoryl group (-PO3) from ATP to Ser, Thr, Tyr, Asp, and His residues of target proteins and enzymes. The enzymes catalyzing the addition of the phosphoryl group, protein kinases, include mitogen activated protein (MAP)1 kinases (1-3), protein kinase C (4, 5), cyclic AMP-dependent protein kinases (6-8), and two component histidine kinases prevalent in bacteria (9, 10). Phosphoprotein phosphatases reverse the action of protein kinases by removing the phosphoryl group from target proteins, ultimately transferring it to water to form orthophosphate as a product. The two major classes of eukaryotic protein phosphatases are the protein tyrosine phosphatases (PTPs) and protein serine/threonine phosphatases. These phosphatases are members of three structurally and mechanistically distinct gene families. The PTPs have in common a HCX5R sequence that constitutes the active site, and a catalytic mechanism involving the formation of a phosphoenzyme intermediate in which the phosphoryl group is transferred transiently to this conserved cysteine residue (11). The protein serine/threonine phosphatases, represented by the PPP and PPM protein phosphatase families, dephosphorylate phosphoserine and phosphothreonine residues and are metalloenzymes catalyzing direct phosphoryl group transfer to water (12-14). A subgroup of the PTPs, the dualspecificity phosphatases, are capable of dephosphorylating serine, threonine, and tyrosine phosphoamino acid residues (15-18). Initially the serine/threonine phosphatases were divided into four classes depending upon their activities in the absence and presence of metal ions and inhibitors (19-22). The type 1 phosphatases, represented by protein phosphatase 1 (PP1), selectively dephosphorylate the â subunit of phosphorylase kinase and are inhibited by the phosphopeptide inhibitors 1 and 2. The class 2 phosphatases, which include protein phosphatases 2A (PP2A), 2B (PP2B or calcineurin), and 2C (PP2C), dephosphorylate the R subunit of phosphorylase kinase and are not inhibited by inhibitors 1 and 2. In addition, PP2B/calcineurin is inhibited by the immunosuppressant drugs cyclosporin A and FK506, while PP1 and PP2A are inhibited by nanomolar concentrations of okadaic acid. Protein phosphatase 2C (PP2C), a PPM class Ser/Thr phosphatase not related to the PPP class, is activated by Mg2+ and unaffected by these macrolide inhibitors. A more recent classification based on multiple sequence alignments places PP1, PP2A, and the catalytic subunit of calcineurin in the PPP family based on a conserved amino † This work was supported by a grant from the National Institutes of Health (GM46865). * To whom correspondence should be addressed at Mayo Clinic and Foundation, 200 First St. S.W., Rochester, MN 55905. Telephone: (507) 284-4743, Fax: (507) 284-8286, Email: [email protected]. 1 Abbreviations: AA, atomic absorption spectroscopy; Amp, ampicillin; BSA, bovine serum albumin; DTT, dithiothreitol; EDTA, ethylenediaminetetraacetic acid; EPR, electron paramagnetic resonance; IPTG, isopropylthio-â-D-galactopyranoside; λPP, bacteriophage λ phosphoprotein phosphatase; MAP, mitogen activated protein; PP1, protein phosphatase 1; PP2A, protein phosphatase 2A; PP2B, calcineurin (protein phosphatase 2B, phosphatase-3); PP2C, protein phosphatase 2C; PPM, Ser/Thr phosphatase subfamily including PP2C homologues; PPP, Ser/Thr phosphatase subfamily including PP1, PP2A, calcineurin, and λPP; pNP, p-nitrophenol; pNPP, p-nitrophenyl phosphate; PTP, protein tyrosine phosphatase family; SDS, sodium dodecyl sulfate; WT, wild type. 8918 Biochemistry 2001, 40, 8918-8929 10.1021/bi010637a CCC: $20.00 © 2001 American Chemical Society Published on Web 07/06/2001 acid sequence termed the phosphoesterase motif, DXH(X)nGDXXD(X)mGNHD/E (23-25). This primary sequence motif defines a âRâRâ secondary structural motif that forms a scaffold for a dinuclear metal ion active site cofactor, with the residues denoted in boldface serving as ligands to both metal ions. One of these metal ions, referred to as the M1 metal ion, is coordinated by a carboxylate and imidazole group from the first aspartate and histidine residue in the phosphoesterase motif noted above. The second metal ion, termed M2, is coordinated by the amido group of the conserved asparagine residue of the phosphoesterase motif and two histidine residues that are positioned beyond the C-terminus of this motif. A conserved aspartate residue coordinates to both M1 and M2 metal ions in a μ-1,1 fashion. Metal binding studies indicate that of the two Mn2+ ions that can be bound by λPP, one is bound tighter than the other but the relationship of these to the structural M1 and M2 sites has not yet been revealed (26). The phosphoprotein phosphatase from bacteriophage λ (λPP) shares amino acid sequence and biochemical similarity with the catalytic cores of calcineurin A, PP1, and PP2A within the phosphoesterase motif. Furthermore, PP1, λPP, and calcineurin A share homology from a common ancestor as seen by comparison of their crystal structures (27-31). The folds of their catalytic core domains around the âRâRâ central motif are highly similar, and the active site protein metal ligands and their positions within the metal coordination sphere are almost identical. Two recently identified phosphatases from E. coli, PrpA and PrpB, show high sequence similarity to λPP and are involved in the sensing of misfolded proteins in the periplasm (32). Several cyanobacterial and archeal PPP type phosphatases also having sequence and likely structural similarity to calcineurin A, PP1, and λPP have been identified recently by the bacterial genome sequencing projects. The specific roles of these bacterial PPPs in regulating metabolism are beginning to be characterized (33-39). Several PPP type homologues having various metabolic regulation functions have been identified in yeast (40). The biological role of λPP in bacteriophage λ, if any, is unknown. In this work we compare the phosphatase activities and spectroscopic (EPR) properties of wild-type and metal ligand mutants of λPP to further characterize the biochemical properties and mechanistic roles of metal ligands to the dinuclear metal center. The identity of the high affinity Mn2+ binding site is revealed and discussed in the context of metal activation and catalytic mechanism of the PPP family of metallophosphatases. EXPERIMENTAL PROCEDURES Materials. Dithiothreitol, MnCl2, p-nitrophenyl phosphate (pNPP), DEAE-Sephadex CL-6B, phenyl-Sepharose, Coomassie Brilliant Blue-R250, and BSA were purchased from Sigma (St. Loius, MO). YM10 Diaflo ultrafiltration membranes and centricon microconcentrators were purchased from Amicon, Inc. (Beverly, MA). NAP-25 gel filtration columns and phenyl-Sepharose resin were purchased from Pharmacia Biotech (Piscataway, NJ). The plasmid pT7-7 was obtained from Dr. Stanley Tabor (41). All chemicals were ACS grade, and glass still distilled water was used throughout. Bacterial growth media were purchased from Fisher Scientific Co. (Hanover Park, IL). Where required, buffers and glassware were treated with Chelex 100 resin (BioRad, Hercules, CA) to minimize metal contamination. Multiple Sequence Alignment of λPP-like Microbial Metallophosphoesterases. Microbial metallophosphoesterase gene and/or amino acid sequences with high similarly to λPP, as identified by BLAST searches (42, 43), were retrieved from the National Center for Biotechnology Information (NCBI) finished and unfinished microbial genome database (http:// www.ncbi.nlm.nih.gov:80/Microb_blast/unfinishedgenome.html). Pairwise alignments of protein sequences were computed according to amino acid sequence similarity using CLUSTALX v1.8 (44-46), using the default gap opening and extension parameters (10 and 0.1, respectively) and the Gonnet 250 protein weight matrix. The output was trimmed at the Nand C-termini such that the termini of the alignment contain regions of high similarity. The alignment was performed again using the trimmed alignment as the input, improving the quality of the alignment by removing bias from dissimilar sequences at the Nand C-termini of the individual sequences. Polypeptides similar to λPP were included from the following organisms (with corresponding three letter abbreviated names): Aac, Actinobacillus actinomycetemcomitans; Afu, Archaeoglobus fulgidus; Asp, Anebena sp.; Ban, Bacillus anthracis; Bps, Burkholderia pseudomallei; Bsu, Bacillus subtilis; Ccr, Caulobacter crescentus; Dra, Deinococcus radiodurans; Eco, Escherichia coli; Efa, Enterococcus faecalis; Fis, FerVidobacterium islandicum; Hin, Haemophilus influenzae; Kae, Klebsiella aerogenes; lam, bacteriophage λ; MaeU, Microcystis aeruginosa UTEX 2063; MaeP, Microcystis aeruginosa PCC 7820; Mth, Methanosarcina thermophila; Nco, Nostoc commune; Pab, Pyrodictium abyssi; Ppu, Pseudomonas putida; Sco, Streptomyces coelicolor; Spn, Streptococcus pneumoniae; Sso, Sulfolobus solfataricus; Ssp, Synechocystis sp.; Sty, Salmonella typhimurium; Tma, Thermatogota maritima. Site-Directed Mutagenesis of λPP. Single and double point mutations encoding single amino acid substitutions in the λPP gene were created using the MORPH II kit (Eppendorf-5 Prime, Inc., Boulder, CO). Oligonucleotide primers representing the λPP coding strand sequence containing base substitutions to bring about the appropriate amino acid substitution were synthesized at the Mayo Clinic Molecular Biology Core Facility. The primers used in the mutagenesis reaction, with mutated codons underlined, are as follows: Mutant primers were phosphorylated using T4 polynucleotide kinase, and the mutagenesis reactions were carried out according to the manufacturer’s instructions using the wildtype λPP gene cloned into the plasmid T7-7 (25) as a template. Following the mutagenesis reaction, mutant plasmid DNA products were used to transform E. coli BMH 71-18 H22N: 5′-TTGGCGATCTGAACGGATGCTACA-3′ N75H: 5′-CTGTACGTGGACACCATGAGCAAA-3′ H92N: 5′-GAAACGTTAATAACTGGCTGCTTA-3′ H117N: 5′-AAGCTCTTGCCAATAAAGCAGATG-3′ H139N: 5′-ATGTTATCTGCAACGCCGATTATC-3′ H186N: 5′-TCATCTTTGGTAATACGCCAGCAG-3′ High and Low Affinity Mn2+ Sites in λPP Biochemistry, Vol. 40, No. 30, 2001 8919