Orthophosphate binding at the dimer interface of Corynebacterium callunae starch phosphorylase: mutational analysis of its role for activity and stability of the enzyme
© Mueller and Nidetzky; licensee BioMed Central Ltd. 2010
Received: 9 September 2009
Accepted: 29 January 2010
Published: 29 January 2010
Orthophosphate recognition at allosteric binding sites is a key feature for the regulation of enzyme activity in mammalian glycogen phosphorylases. Protein residues co-ordinating orthophosphate in three binding sites distributed across the dimer interface of a non-regulated bacterial starch phosphorylase (from Corynebacterium callunae) were individually replaced by Ala to interrogate their unknown function for activity and stability of this enzyme.
While the mutations affected neither content of pyridoxal 5'-phosphate cofactor nor specific activity in phosphorylase preparations as isolated, they disrupted (Thr28→Ala, Arg141→Ala) or decreased (Lys31→Ala, Ser174→Ala) the unusually strong protective effect of orthophosphate (10 or 100 mM) against inactivation at 45°C and subunit dissociation enforced by imidazole, as compared to wild-type enzyme. Loss of stability in the mutated phosphorylases appeared to be largely due to weakened affinity for orthophosphate binding. Binding of sulphate mimicking the crystallographically observed "non-covalent phosphorylation" of the phosphorylase at the dimer interface did not have an allosteric effect on the enzyme activity.
The phosphate sites at the subunit-subunit interface of C. callunae starch phosphorylase appear to be cooperatively functional in conferring extra kinetic stability to the native dimer structure of the active enzyme. The molecular strategy exploited for quaternary structure stabilization is to our knowledge novel among dimeric proteins. It can be distinguished clearly from the co-solute effect of orthophosphate on protein thermostability resulting from (relatively weak) interactions of the ligand with protein surface residues.
α-(1,4)-D-Glucan phosphorylases (GlgP) promote degradation of glycogen, starch or maltodextrins by catalyzing glucosyl transfer from the non-reducing end of the glucosidic substrate to orthophosphate. They often serve a physiological function in fuelling the energy metabolism of the cell with α-D-glucose 1-phosphate (G1P) . Although categorized as glycosyltransferases , GlgPs are special among enzymes of this class in that their activity is absolutely dependent on a pyridoxal 5'-phosphate (PLP) cofactor [3–5]. The PLP forms a Schiff-base linkage with ε-NH2 of an invariant Lys in the active site. The 5'-phosphate moiety is the cofactor group participating in catalysis [3, 4, 6]. All known GlgP enzymes are naturally active as dimers of two identical PLP-containing subunits [7–9]. Dimeric structure formation results in marked stabilization of the otherwise chemically labile protein-cofactor bond such that PLP is not detectably dissociable from native phosphorylase dimers [7, 10, 11]. GlgP enzymes in which activity is under control of covalent phosphorylation and/or allosteric effectors respond to regulatory signals through extensive rearrangements of their intersubunit contacts [5, 12, 13]. The dimer interface therefore is a key element of GlgP structure and function. While the overall pattern of subunit-subunit interactions is conserved in GlgPs, the molecular details vary among individual enzymes [5, 7–9, 13].
Extensive use of intersubunit binding of orthophosphate to stabilize the native dimer in Cc GlgP has not been described for other GlgP enzymes and appears to generally lack precedence in oligomeric proteins. We therefore used mutational analysis to determine the role of individual phosphate site residues for orthophosphate-dependent stability and activity in Cc GlgP.
Site-directed mutagenesis, enzyme production and purification
The plasmid pQE 30-GlgP harbouring the gene encoding wild-type Cc GlgP fused to an N-terminal metal affinity peptide (RGSHHHHHHGSA)  was used as template for site-directed mutagenesis. Mutations were introduced by employing a modified two-stage PCR protocol  in which the following pairs of oligonucleotide primers (Invitrogen) were used with mismatched codons underlined. T28A: 5'-ACCTCGCTGCT GATCGCAAG-3' (forward primer), 5'-AGAACTTGCGATCAGC AGCGAG-3' (reverse primer); K31A: 5'-CTACTGATCGCGCG TTCTGGACTG-3' (forward primer), 5'-CAGTCCAGAACGC GCGATCAGTAG-3' (reverse primer); R141A: 5'-TGGTCTGCTCTACGCC TTCGGTC-3' (forward primer), 5'- AGACCGAAGGC GTA-GAGCAGACC-3' (reverse primer); S174A: 5'-TCGTGCAGCC GACCAGTTG-3' (forward primer), 5'-TGGTCGGC TGCACGACGAATAG-3' (reverse primer). Plasmid vectors harbouring sequence-proven inserts (VBC Genomics) were transformed into E. coli JM109, and recipient strains were grown for recombinant protein production as reported previously . Protein purification followed a published protocol  except that no heat treatment was used. Purity of the obtained protein preparations was assessed by SDS PAGE. Isolated enzymes were stored at 4°C at a concentration of 4.0 - 12 mg/ml in 50 mM potassium phosphate buffer, pH 7.0.
Biochemical characterization of mutated CcGlgP
Phosphorylase activity was measured with a continuous coupled enzyme assay described elsewhere . The Bio-Rad dye binding assay referenced against BSA was used for determination of protein concentrations. The PLP content of isolated protein preparations was quantitated using a reported colorimetric method .
Activity loss at elevated temperature. Protein solutions (22 - 53 μg/ml) were prepared in 50 mM triethanolamine buffer, pH 7.0 or 6.6, containing 24 mM KCl, 10.0 mM or 100 mM K2HPO4. Incubations were carried out in 1.5 ml tubes at 45°C. Samples (10 μl) were taken after 15 sec and then in regular intervals, depending on the stability of the enzyme used. They were diluted immediately into the continuous coupled assay of phosphorylase activity. Test for reversibility of inactivation involved cooling of the sample to room temperature followed by a 1 h-long incubation in the presence of 50 mM potassium phosphate, pH 7.0, and then activity measurement.
Inactivation by imidazole. A buffer (pH 7.0) containing 0.4 M imidazole and 0.1 M L-cysteine hydrochloride was used. K2HPO4 or (NH4)2SO4 was optionally added in a concentration of 5.0 mM. The protein was diluted to a final concentration of 22 - 53 μg/ml in the above-described buffer, and incubations were carried out at 30°C. Samples were taken at the times indicated and residual activity was measured using the continuous assay. Restoration of activity in partially imidazole-denatured preparations of the wild-type enzyme was examined in the presence of 200 mM K2HPO4 (pH 7.0) and 500 μM PLP.
Gel filtration analysis
Size exclusion chromatography was performed using a BioLogic Duo-Flow System (model 2128; Bio-Rad, Hercules, U.S.A.) equipped with a HiLoad 16/60 Superdex 200 prep grade column (GE Healthcare). The column was equilibrated with 50 mM potassium phosphate buffer, pH 7.0, containing 0.15 M NaCl. It was operated with the same buffer using a flow rate of 1 ml/min. Gel Filtration Standard from Bio-Rad was employed for calibration. The applied sample (2 ml) typically contained = 0.8 mg of native or partially denatured protein.
Effect of sulphate on α-glucan-synthesizing activity of wild-type and mutated forms of Cc GlgP
Initial rate measurements were performed in the direction of α-(1,4)-D-glucan synthesis at 30°C using a concentration of 2.5, 25 and 50 nM for the respective enzyme subunit. Reactions were carried out in 50 mM triethanolamine buffer, pH 7.0, with and without 5.0 - 40 mM (NH4)2SO4 present. Maltodextrin DE19 (AGENAMALT 20.235, Agrana, Austria) or soluble starch was used as acceptor substrate (20 g/l). The concentration of orthophosphate (Pi) released from G1P (5.0, 20, and 50 mM) was measured in a minimum of three samples taken after 4 to 120 min using an assay described elsewhere . The rate was calculated from the linear relationship of [Pi] against time.
Results and Discussion
Selection of residues for site-directed substitution and properties of mutated enzymes
Biochemical properties of wild-type and mutated forms of Cc GlgP.
Orthophosphate-dependent stability at elevated temperature
Comparison of half-life times (t1/2) of recombinant wild-type Cc GlgP and site-directed enzyme variants in thermal denaturation experiments and in the presence of 0.4 M imidazole at 30°C
t (1/2) [min]
inactivation at 45°C
inactivation by imidazole at 30°C
10.0 mM Pi
100 mM Pi
5.0 mM Pi
2.8 × 102
3.5 × 103
3.3 × 102
1.7 × 102
3.5 × 103
9.9 × 102
1.4 × 103
Decreased affinity for orthophosphate binding or intrinsically lowered stability of the protein-orthophosphate complex could explain the loss in orthophosphate-dependent stability of the Cc GlgP mutants. To distinguish between these possibilities, we determined t1/2for wild-type and mutated phosphorylases in the presence of 100 mM Pi (Table 2). T28A and R141A were stabilized by a factor of 150 and 187, respectively, as compared to the corresponding t1/2 determined at 10.0 mM Pi. Enhancement of t1/2 resulting from the increase in Pi concentration was 21- and 30-fold in K31A and S174A, respectively, and can be compared to a 13-fold effect on t1/2for the wild-type phosphorylase. Because differences in stability among the individual phosphorylases seen at 10.0 mM Pi were, to a very substantial extent, removed at 100 mM Pi, we believe that it was mainly the Pi binding affinity of the respective site (not the mechanism of stabilization) that was influenced by the chosen single point mutations. We did not test higher concentrations of orthophosphate than 100 mM because under these conditions, it is exceedingly difficult to distinguish the stabilization resulting from specific binding at a defined phosphate site from another stabilization due to non-specific protein-orthophosphate interactions . Part of the enhancement of t1/2 for the wild-type phosphorylase upon increasing the orthophosphate concentration from 10.0 to 100 mM could already reflect non-specific stabilization.
The results in Table 2 show that structural modification of either phosphate site (e.g. R141A and T28A) can result in a nearly complete loss of orthophosphate-dependent stability at 10.0 mM Pi. This finding suggests that the CAP-sites (where Thr28 is located) and the P-site (where Arg141 is located) do not function independently one from another, be it that orthophosphate binding at the two sites is truly cooperative or occupancy of both sites is a critical requirement for dimer stability. In the case that orthophosphate binding at each binding site made an independent contribution to the kinetic stability of Cc GlgP (measured as t1/2 at 45°C), one would expect that site-directed mutagenesis of one binding site causes only partial disruption of orthophosphate-dependent stability, which is contrary to observations for T28A and R141A.
Orthophosphate-dependent stability in the presence of imidazole
With the exception that all mutated phosphorylases, however especially T28A, were less stable than the wild-type enzyme under conditions where 5.0 mM orthophosphate was lacking, data confirm the overall trend seen in inactivation experiments at 45°C that the mutations decreased the stabilizing effect of orthophosphate in the wild-type enzyme. Generally, orthophosphate was much less stabilizing to denaturation by imidazole than denaturation by heat (45°C). However, we must consider that, while both methods of denaturation promote dissociation of subunits in the phosphorylase dimer [11, 15], their effects on the protein structure are probably not identical. The disruptive effect of the mutations on orthophosphate-dependent stability was smaller when using imidazole as compared to 45°C as trigger of denaturation.
Allosteric effect of orthophosphate binding on enzyme activity?
Initial rates of α-(1,4)-glucan synthesis (Vs) catalyzed by native Cc GlgP were recorded in the absence and presence of (NH4)2SO4. Previous work has shown that sulphate is similarly efficient as orthophosphate in stabilizing the dimer structure of Cc GlgP [11, 14], validating the use of sulphate as an orthophosphate surrogate in kinetic experiments. Note the added orthophosphate would have interfered with the assay applied for determination of Vs.
Figure 4B shows a comparison of the effect of 10 mM sulphate on Vs for wild-type and mutated forms of Cc GlgP measured at a protein concentration of 2.5 nM. The concentration of αG1P was 50 mM. The 2.4-fold apparent activation of the wild-type enzyme under these conditions was retained in T28A whereas it was almost completely lost in K31A and R141A. The strong (~5-fold) enhancement of activity of S174A at the low protein concentration was attenuated to a 1.7-fold "activation" at a higher protein concentration of 50 nM. Addition of sulphate partly eliminated differences in specific activity between the individual enzymes observed under conditions where the oxyanion was lacking. However, the concentration of sulphate required to raise the specific activity of T28A to the level of the wild-type enzyme was higher than 10 mM, and full complementation of the mutated phosphorylase was obtained at 40 mM oxyanion. These results agree with the notion (Table 2) that orthophosphate/sulphate binding affinity at the CAP site was strongly decreased as result of individual substitutions of Thr28 by Ala. They also concur with the proposed mechanism of action of sulphate where stimulation of activity is apparent and derives from a stabilized dimer structure. A possible allosteric effect of interfacial orthophosphate/sulphate binding on the enzyme activity is therefore not supported.
The CAP- and P-sites for orthophosphate binding at the subunit-subunit interface of Cc GlgP appear to be cooperatively functional in conferring extra kinetic stability to the native dimer structure of the active enzyme. The molecular strategy exploited for quaternary structure stabilization is to our knowledge novel among dimeric proteins. It can be distinguished clearly from the co-solute effect of orthophosphate on protein thermostability resulting from (relatively weak) interactions of the ligand with protein surface residues, often lysines [23, 24]. However, Treponema denticola cystalisin is an interesting example of a PLP-containing enzyme that utilizes hydrogen bonding between the 5'-phosphate of the cofactor and a tyrosine from the respective other subunit to stabilize the functional protein homodimer . We have shown here that Thr28 at the CAP site of Cc GlgP is of key importance for orthophosphate-dependent stability of the enzyme. Therefore, although an allosteric effect of oxyanion binding on enzyme activity was not clearly supported by the data, it was nevertheless interesting that an unregulated phosphorylase has accommodated a functional phosphate site in a protein region where the principle of phosphate group recognition was exploited by nature to evolve the regulatory sites in today's eukaryotic α-(1,4)-D-glucan phosphorylases.
Financial support was obtained from the FWF (P18138-B09 to B.N.). The contribution of Dr. R. Grießler in an early phase of the project is gratefully acknowledged. Dr. A. Schwarz is thanked for support and discussion.
- Schinzel R, Nidetzky B: Bacterial α-glucan phosphorylases. FEMS Microbiol Lett. 1999, 171: 73-79.PubMedGoogle Scholar
- Coutinho PM, Deleury E, Davies GJ, Henrissat B: An evolving hierarchical family classification for glycosyltransferases. J Mol Biol. 2003, 328: 307-317. 10.1016/S0022-2836(03)00307-3.PubMedView ArticleGoogle Scholar
- Palm D, Klein HW, Schinzel R, Buehner M, Helmreich EJ: The role of pyridoxal 5'-phosphate in glycogen phosphorylase catalysis. Biochemistry. 1990, 29: 1099-1107. 10.1021/bi00457a001.PubMedView ArticleGoogle Scholar
- Withers SG, Madsen NB, Sprang SR, Fletterick RJ: Catalytic site of glycogen phosphorylase: structural changes during activation and mechanistic implications. Biochemistry. 1982, 21: 5372-5382. 10.1021/bi00264a039.PubMedView ArticleGoogle Scholar
- Johnson LN: Glycogen phosphorylase: control by phosphorylation and allosteric effectors. FASEB J. 1992, 6: 2274-2282.PubMedGoogle Scholar
- Chang YC, McCalmont T, Graves DJ: Functions of the 5'-phosphoryl group of pyridoxal 5'-phosphate in phosphorylase: a study using pyridoxal-reconstituted enzyme as a model system. Biochemistry. 1983, 22: 4987-4993. 10.1021/bi00290a017.PubMedView ArticleGoogle Scholar
- Watson KA, Schinzel R, Palm D, Johnson LN: The crystal structure of Escherichia coli maltodextrin phosphorylase provides an explanation for the activity without control in this basic archetype of a phosphorylase. EMBO J. 1997, 16: 1-14. 10.1093/emboj/16.1.1.PubMedPubMed CentralView ArticleGoogle Scholar
- Newgard CB, Hwang PK, Fletterick RJ: The family of glycogen phosphorylases: structure and function. Crit Rev Biochem Mol Biol. 1989, 24: 69-99. 10.3109/10409238909082552.PubMedView ArticleGoogle Scholar
- Buchbinder JL, Rath VL, Fletterick RJ: Structural relationships among regulated and unregulated phosphorylases. Annu Rev. 2001, 30: 191-209.Google Scholar
- Shaltiel S, Hedrick JL, Fischer EH: On the role of pyridoxal 5'-phosphate in phosphorylase. II. Resolution of rabbit muscle phosphorylase b. Biochemistry. 1966, 5: 2108-2116. 10.1021/bi00870a044.PubMedView ArticleGoogle Scholar
- Griessler R, Psik B, Schwarz A, Nidetzky B: Relationships between structure, function and stability for pyridoxal 5'-phosphate-dependent starch phosphorylase from Corynebacterium callunae as revealed by reversible cofactor dissociation studies. Eur J Biochem. 2004, 271: 3319-3329. 10.1111/j.1432-1033.2004.04265.x.PubMedView ArticleGoogle Scholar
- Browner MF, Fletterick RJ: Phosphorylase: a biological transducer. Trends Biochem Sci. 1992, 17: 66-71. 10.1016/0968-0004(92)90504-3.PubMedView ArticleGoogle Scholar
- Rath VL, Fletterick RJ: Parallel evolution in two homologues of phosphorylase. Nat Struct Biol. 1994, 1: 681-690. 10.1038/nsb1094-681.PubMedView ArticleGoogle Scholar
- Griessler R, D'Auria S, Tanfani F, Nidetzky B: Thermal denaturation pathway of starch phosphorylase from Corynebacterium callunae: oxyanion binding provides the glue that efficiently stabilizes the dimer structure of the protein. Protein Sci. 2000, 9: 1149-1161. 10.1110/ps.9.6.1149.PubMedPubMed CentralView ArticleGoogle Scholar
- Griessler R, Schwarz A, Mucha J, Nidetzky B: Tracking interactions that stabilize the dimer structure of starch phosphorylase from Corynebacterium callunae. Roles of Arg234 and Arg242 revealed by sequence analysis and site-directed mutagenesis. Eur J Biochem. 2003, 270: 2126-2136. 10.1046/j.1432-1033.2003.03562.x.PubMedView ArticleGoogle Scholar
- Nidetzky B, Griessler R, Pierfederici FM, Psik B, Scire A, Tanfani F: Mutagenesis of the dimer interface region of Corynebacterium callunae starch phosphorylase perturbs the phosphate-dependent conformational relay that enhances oligomeric stability of the enzyme. J Biochem (Tokyo). 2003, 134: 599-606.View ArticleGoogle Scholar
- Schwarz A, Brecker L, Nidetzky B: Probing the active site of Corynebacterium callunae starch phosphorylase through the characterization of wild-type and His334→Gly mutant enzymes. FEBS J. 2007, 274: 5105-5115. 10.1111/j.1742-4658.2007.06030.x.PubMedView ArticleGoogle Scholar
- Schwarz A, Nidetzky B: Asp-196→Ala mutant of Leuconostoc mesenteroides sucrose phosphorylase exhibits altered stereochemical course and kinetic mechanism of glucosyl transfer to and from phosphate. FEBS Lett. 2006, 580: 3905-3910. 10.1016/j.febslet.2006.06.020.PubMedView ArticleGoogle Scholar
- Schwarz A, Pierfederici FM, Nidetzky B: Catalytic mechanism of α-retaining glucosyl transfer by Corynebacterium callunae starch phosphorylase: the role of histidine-334 examined through kinetic characterization of site-directed mutants. Biochem J. 2005, 387: 437-445. 10.1042/BJ20041593.PubMedPubMed CentralView ArticleGoogle Scholar
- Wada H, Snell EE: The enzymatic oxidation of pyridoxine and pyridoxamine phosphates. J Biol Chem. 1961, 236: 2089-2095.PubMedGoogle Scholar
- Drueckes P, Schinzel R, Palm D: Photometric microtiter assay of inorganic phosphate in the presence of acid-labile organic phosphates. Anal Biochem. 1995, 230: 173-177. 10.1006/abio.1995.1453.PubMedView ArticleGoogle Scholar
- Weinhäusel A, Griessler R, Krebs A, Zipper P, Haltrich D, Kulbe KD, Nidetzky B: α-1,4-D-Glucan phosphorylase of gram-positive Corynebacterium callunae: isolation, biochemical properties and molecular shape of the enzyme from solution X-ray scattering. Biochem J. 1997, 326: 773-783.PubMedPubMed CentralView ArticleGoogle Scholar
- Fayos R, Pons M, Millet O: On the origin of the thermostabilization of proteins induced by sodium phosphate. Am Chem Soc. 2005, 127: 9690-9691. 10.1021/ja051352e.View ArticleGoogle Scholar
- Tadeo X, Pons M, Millet O: Influence of Hofmeister anions on protein stability as studied by thermal denaturation and chemical shift perturbation. Biochemistry. 2007, 46: 917-923. 10.1021/bi0613426.PubMedView ArticleGoogle Scholar
- Montioli R, Cellini B, Bertoldi M, Paiardini A, Voltattorni CB: An engineered folded PLP-bound monomer of Treponema denticola cystalysin reveals the effect of the dimeric structure on the catalytic properties of the enzyme. Proteins. 2008, 74: 304-317. 10.1002/prot.22160.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.