Chem. Res. Toxicol. 2003, 16, 95-102
95
Selective Nitration of Tyr99 in Calmodulin as a Marker of Cellular Conditions of Oxidative Stress Heather S. Smallwood,† Nadezhda A. Galeva,‡ Ryan K. Bartlett,§ Ramona J. Bieber Urbauer,§ Todd D. Williams,‡ Jeffrey L. Urbauer,§ and Thomas C. Squier*,† Department of Molecular Biosciences and Mass Spectrometry Laboratory, University of Kansas, Lawrence, Kansas 66045, and Fundamental Sciences Division, Pacific Northwest National Laboratory, Richland, Washington 99352 Received June 6, 2002
We examined the possible role of methionines as oxidant scavengers that prevent the peroxynitrite-induced nitration of tyrosines within calmodulin (CaM). We used mass spectrometry to investigate the reactivity of peroxynitrite with CaM at physiological pH. The possible role of methionines in scavenging peroxynitrite (ONOO-) was assessed in wild-type CaM and following substitution of all nine methionines in CaM with leucines. We find that peroxynitrite selectively nitrates Tyr99 at physiological pH, resulting in the formation of between 0.05 and 0.25 mol of nitrotyrosine/mol of CaM when the added molar ratio of peroxynitrite per CaM was varied between 2.5 and 15. In wild-type CaM there is a corresponding oxidation of between 0.8 and 2.8 mol of methionine to form methionine sulfoxide. However, following sitedirected substitution of all nine methionines in wild-type CaM with leucines, the extent of nitration by peroxynitrite was unchanged. These results indicate that Tyr99 is readily nitrated by peroxynitrite and that methionine side chains do not function as an antioxidant in scavenging peroxynitrite. Thus, separate reactive species are involved in the oxidation of methionine and nitration of Tyr99 whose relative concentrations are determined by solution conditions. The sensitivity of Tyr99 in CaM to nitration suggests that CaM-dependent signaling pathways are sensitive to peroxynitrite formation and that nitration of CaM represents a cellular marker of peroxynitrite-induced changes in cellular function.
Introduction Peroxynitrite (ONOO-)1 is formed in vivo from superoxide (O2•-) and nitric oxide (NO•) and is a long-lived reactive species that can result in the oxidation or nitration of membrane lipids, nucleic acids, or proteins. In this respect, peroxynitrite has been implicated in the nitration of tyrosine side chains and associated alterations in the function of critical signal transduction proteins during biological aging and a host of pathologies (1, 2). For example, nitration by peroxynitrite has been demonstrated to activate tyrosine kinase mediated signal transduction cascades and may mimic adenylylation of tyrosyl residues or phosphotyrosine side chains in promoting the formation of activated protein complexes (37). Thus, nitration of tyrosines by peroxynitrite has the potential to profoundly alter cellular metabolism through specific signal transduction pathways. * To whom correspondence should be addressed. Telephone: (509) 376-2218. Fax: (509) 376-1494. E-mail:
[email protected]. † Fundamental Sciences Division. ‡ Mass Spectrometry Laboratory. § Department of Molecular Biosciences. 1 Abbreviations: buffer A, 100 mM HEPES (pH 7.5), 0.1 M KCl, 5 mM MgCl2, 0.1 mM EGTA, 0.44 mM CaCl2, 5 mM ATP, and 4 µM A23187; CaM, calmodulin; CaM-L9, expressed CaM in which all nine methionines are substituted with leucines; ESI-MS; electrospray ionization mass spectrometry; HEPES, β-(2-hydroxyethyl)piperazineN-2-ethansulfonic acid; IPTG, isopropyl β-D-1-thiogalactopyranoside; PM-Ca-ATPase, plasma membrane calcium pump; SERCA, sarcoplasmic or endoplasmic reticulum Ca-ATPase; UV, ultraviolet; OdNOO-, oxoperoxonitrate (1-) or peroxynitrite; ONOOCO2-, 1-carboxylato-2nitrosodioxidane.
It is, therefore, critical to understand biological defense mechanisms that have evolved to mitigate the loss of cellular function associated with oxidative stress and the generation of peroxynitrite (8). In this latter respect, methionine side chains have been suggested to be important antioxidant defense mechanisms due to (i) their reactivity with a large variety of oxidants and (ii) the ability of methionine sulfoxide reductases endogenous in all cells to reduce oxidized methionines and restore their antioxidant capacity (9, 10). In support of this hypothesis, methionines surrounding the active site in glutamine synthetase are selectively oxidized under in vitro conditions and are suggested to minimize oxidant-induced lossof-function in vivo (11). Similarly, the calcium signaling protein CaM contains nine methionines and two tyrosines (Tyr99 and Tyr138), which are substrates of insulindependent signaling (12). Thus, protection of tyrosine from nitration by one or all methionines within this central signaling molecule would maintain normal signaling cascades. In support of this suggestion, methionines in CaM are selectively oxidized in senescent brain (13). Moreover, in vitro exposure to peroxynitrite had previously been suggested to result in methionine oxidation without nitration of either tyrosine side chain in CaM (14), suggesting that methionines may serve as oxidant scavengers. To investigate the possible role of methionine side chains in modulating the sensitivity of tyrosines in CaM to nitration by peroxynitrite, we have used site-directed mutagenesis combined with mass spectrometry to exam-
10.1021/tx025566a CCC: $25.00 © 2003 American Chemical Society Published on Web 01/01/2003
96
Chem. Res. Toxicol., Vol. 16, No. 1, 2003
Smallwood et al.
Figure 1. Sequence of wild-type CaM and CaM-L9 and location of tryptic cleavage sites. Symbols represent amino acids that are identical (:) or involve the replace of methionine in wild-type CaM to leucine in CaM-L9 (X). Primary tryptic cleavage sites (1) are denoted, and the resulting peptides are designated by the letter T followed by a number relative to the amino terminus. Secondary cleavage sites (2) that occur within the primary designated fragment are indicated.
ine the site-specific sensitivities of tyrosine residues in CaM to nitration by peroxynitrite. Specifically, we have investigated (i) the possible role of methionine side chains in minimizing the nitration of tyrosine side chains and (ii) the sensitivity of CaM to nitration at physiological pH. To investigate the possible role of methionines, we have engineered a mutant CaM in which all nine methionines in wild-type CaM are replaced with leucines (CaM-L9) (Figure 1). Our results suggest that at pH 7.0 the presence of methionine residues does not affect the extent of tyrosine nitration. Furthermore, tyrosine nitration is site-specific, suggesting the local environment plays a role in mediating nitration. These results indicate that under physiological conditions involving the generation of peroxynitrite that nitration of Tyr99 in CaM is expected.
Experimental Procedures Materials. Sequence-grade trypsin was obtained from Promega (Madison, WI). Phenyl Sepharose CL-4B was purchased from Pharmacia (Piscataway, NJ). A micro-BCA reagent assay kit was obtained from Pierce (Rockford, IL). Restriction endonucleases, primers, and dNTPs for PCR were purchased from Gibco-Invitrogen Corporation (Carlsbad, CA). The vectors pBluescript II SK and Pfu DNA polymerase were purchased from Stratagene (La Jolla, CA). The vector pET15b and the Escherichia coli BL21(DE3) competent cells were purchased from Novagen (Madison, WI). The E. coli competent cell strain DH10B was purchased from Invitrogen/Life Technologies (Carlsbad, CA). Peroxynitrite was prepared essentially as previously described (14, 15) and generously provided by both Professors Christian Schoneich at the University of Kansas and Joseph Beckman at Oregon State University. Calmodulin Mutagenesis, Expression, and Purification. The coding region for chicken CaM [accession number MCCH (PIR database) or P02593 (SWISS-PROT database)] was excised from the plasmid pCaMPL provided by Professor Samuel George (Duke University), and subcloned into the mutagenesis and expression vector pALTER-Ex1, as previously described (16). cDNA encoding wild-type CaM was mutated to replace all nine methionine residues with leucines (CaM-L9). The mutated gene for CaM-L9 was produced by sequentially mutating all nine methionine residues to leucine using the PCR based mutagenesis method outlined previously (17, 18). The intact mutant CaM gene for CaM-L9 was subcloned into pET-15b, and transformed into the E. coli BL21(DE3) cell strain for protein overexpression. The bacteria were grown in minimal media and protein produc-
tion was induced with IPTG. Both wild-type and mutant CaM was purified by chromatography on phenyl Sepharose CL-4B, essentially as previously described (19). PM-Ca-ATPase Purification and Functional Assay. Erythrocyte ghost membranes containing the PM-Ca-ATPase were obtained from porcine blood, and stored at -70 °C (20). The CaM-dependent ATPase activity of the PM-Ca-ATPase was measured as described by Lanzetta and co-workers for measuring phosphate release (21). The ghost membrane protein concentration was determined by the Biuret method (22), using BSA as a standard. The concentration of CaM standard was determined using the published extinction coefficient (e277nm ) 3029 M-1 cm-1) for calcium-saturated CaM (19, 23) or with the micro-BCA assay using CaM as a protein standard. ATPase activity was measured at 37 °C in a solution containing approximately 2 nM PM-Ca-ATPase (i.e., 0.05 mg mL-1 porcine erythrocyte ghost membranes) in 100 mM HEPES (pH 7.5), 0.1 M KCl, 5 mM MgCl2, 0.1 mM EGTA, 0.44 mM CaCl2, 5 mM ATP, and 4 µM A23187 (buffer A). The free calcium concentration was calculated to be 100 µM (24). Calculation of Free CaM Concentrations. The concentrations of free CaM were obtained from the following relationship:
[CaM]free ) [CaM]total -
(V - Vmin)
× (Vmax - Vmin) [PM - Ca - ATPase] (1)
where Vmax is the maximal CaM-dependent ATPase activity, V is the observed ATPase activity at a defined concentration of CaM, [CaM]free is the concentration of CaM free in solution, [CaM]total is the total concentration of CaM added to the solution, and [PM-Ca-ATPase] is the total binding capacity of the erythrocyte ghosts for CaM, which was measured to be 40 pmol of CaM bound/mg of porcine erythrocyte ghost (25). Oxidation of CaM. Final concentrations of peroxynitrite were quantified by using its absorbance spectrum (λmax ) 302 nm), where 302 ) 1670 M-1 cm-1 (14, 26). Addition of variable amounts of peroxynitrite was made to calcium-saturated CaM (60 µM) by vortexing a droplet of peroxynitrite suspended above the sample. Unless otherwise specified, sample buffer contained 200 mM KH2PO4 (pH 7.0) and formation of nitrotyrosine was monitored spectrophotometrically, where 381 ) 2200 M-1 cm-1 (27). Confirmation of nitrotyrosine was determined from the appearance of an absorbance peak at 430 nm following the addition of NaOH, where 430 ) 4400 M-1 cm-1 at pH 10.0 (28). Mass Spectrometric Analysis. Following exposure to peroxynitrite, electrospray ionization mass spectrometry (ESI-MS) was used to identify the distribution of posttranslational modifications to CaM, adapting strategies previously described
Methionine Oxidation and Tyrosine Nitration
Figure 2. Activation of the PM-Ca-ATPase by CaM. CaMdependence of the activation of the PM-Ca-ATPase by wild-type CaM (b) and CaM-L9 (O) CaM. ATPase activity was measured at 37 °C in buffer A containing 0.05 mg/mL porcine erythrocyte ghosts. Average errors in ATPase measurements were 8% of the indicated values. (13, 29). The degree of oxidative modification was first assessed by measuring whole protein molecular weight, than localization was determined from tryptic digests of samples. The ESI mass spectra in positive ion mode were acquired on a Q-Tof 2 quadrupole, time-of-flight hybrid instrument (Micromass Ltd, Manchester, U.K.). The analysis of intact CaM was done by desalting 20 µg of protein on a trapping column (1.5 cm × 1 mm i.d.) hand packed with Zorbax SB-C18, 5µm (Agilent Technologies, Wilmington, DE.), washing with 1% acetic acid, then eluting directly into the source with 90% MeOH, 0.5% formic acid. Instrument parameters found that offered the best compromise of sensitivity and minimizing the water loss peak (13) were 60 V on the cone and 12 V on the collision cell with no Ar. Quantitation of the oxidation and nitration of specific sites in CaM was determined following proteolytic digestion of CaM and analysis by HPLC/MS. CaM (30 µM) was digested with 0.6 µM sequence-grade trypsin in 2% (NH4)2CO3 at pH 8.0 for 9 h at 37 °C. The resulting peptide was separated by reversedphase (C18) capillary chromatography (5 cm × 320 mm i.d. column with Zorbax C18SB 300, Micro Tech Scientific, Sunnyvale, CA) using a methanol/H2O gradient (0.08% formic acid) developed at a flow rate of 10 µL min-1 (Ultra Plus II chromatograph, Micro Tech Scientific, Sunnyvale, CA). Spectra from digests were acquired with the Q-Tof 2 operated at maximum resolution (10 000 RP fwhh), Ar in the collsion cell, 5 V on the cell and a cone voltage of 35 V.
Results Nonessential Role of Methionines in the Activation of the PM-Ca-ATPase. Earlier results suggest that methionine side chains in CaM are nonessential with respect to the activation of the plasma membrane CaATPase (30). Consistent with this suggestion, the ability of CaM to fully activate the PM-Ca-ATPase is retained following replacement of all nine methionine residues in CaM with leucine (CaM-L9) (Figure 2). In fact, CaM-L9 addition to the PM Ca-ATPase results in a 29 ( 6% increase in the level of enzyme activation, suggesting that the substitution of all nine methionines with leucines enhances those interactions associated with enzyme activation. There is a small decrease in the apparent binding affinity between CaM-L9 and the PM-Ca-ATPase in comparison to wild-type CaM; half-maximal activation of the PM-Ca-ATPase, respectively, requires 7 ( 2 and 26 ( 7 nM of wild-type CaM and CaM-L9. These latter results are consistent with earlier suggestions that the
Chem. Res. Toxicol., Vol. 16, No. 1, 2003 97
Figure 3. Electrospary ionization mass spectra of wild-type CaM. Spectra correspond to untreated CaM (A) or following exposure to peroxynitrite at a molar ratio of 2.5:1 (B), 5:1 (C), 15:1 (D), and 25:1 (E). Spectra were obtained following deconvolution of multiply charged ions. Experimentally, 20 µg of CaM in 0.1 mM EGTA and 10 mM (NH4)2CO3 (pH 8.6) was trapped, desalted, and then directly infused (on-line) into a Q-Tof 2 mass spectrometer.
flexible methionine side chains promote the high-affinity association between CaM and target proteins (33). The lack of correlation between high-affinity binding and the extent of enzyme activation indicates that in this case there is a distinction between interactions necessary for binding and those necessary for enzyme activation. However, irrespective of the underlying mechanism of enzyme activation, these results indicate that the structure of wild-type CaM and CaM-L9 are virtually the same, permitting an accurate assessment of the role of endogenous methionines on the peroxynitrite-induced modification of tyrosine side chains. Oxidation of CaM by Peroxynitrite. ESI-MS permits an identification of the distribution of oxidative modifications to CaM (13, 34). The observed mass of expressed CaM is 16 706.2 ( 0.7 Da (n ) 4; Figure 3), which is consistent with the theoretical average mass value of 16 706.4 calculated from the sequence. A smaller peak is observed at 16 690 Da, which corresponds to a dehydration product generated in the mass spectrometer (34). An additional peak observed at 16 745 corresponds to a potassium salt of CaM. In vitro exposure of CaM to peroxynitrite results in the formation of higher-mass products that differ by 16 amu. This is consistent with earlier suggestions that the presence of peroxynitrite results in the selective oxidation of methionines to their corresponding sulfoxides (14). The distribution of oxidative products in CaM is similar to that previously reported following in vitro oxidation by H2O2, where the extent of methionine oxidation correlated with the surface accessibility of individual methionine side chains in the native protein structure (34, 35). Thus, oxidation with peroxynitrite also appears to involve diffusion-mediated reactions that are modulated by surface accessibility. However, in contrast to oxidation by hydrogen peroxide, additional high mass peaks are observed above 16 850 Da (corresponding to the mass of CaM with all nine methionines oxidized to their corresponding sulfoxide), suggesting that peroxynitrite induces the formation of additional posttranslational modifications other than methionine sulfoxide. In support of this, the character-
98
Chem. Res. Toxicol., Vol. 16, No. 1, 2003
Smallwood et al.
Table 1. LC-MS Identification of Tryptic Fragments Isolated from Wild-Type CaM Following Oxidation by ONOO[M + 2H]2+ observed
tryptic fragment
sequence
T1 T2a T2b T2 T3 T3 (SO) T4 T4 (SO)1 T4 (SO)2 T4 (SO)3 T5 + T6 + T7 T5 + T6 + T7 (SO) T6 + T7 T6 + T7 (SO) T6 + T7 + T8 T6 + T7 + T8 (SO) T7 T7 + T8 T9 T9 (NY) T9b T9b (NY) T10a T10a (SO) T10b T10b (SO) T11 T11 (SO) T11 (SO)2 T11 (SO)2 + NY
Ala1 - Lys13 Glu14 - Lys21 Asp22 - Lys30 Glu14 - Lys30 Glu31 - Arg37 Glu31 - Met(O) - Arg37 Ser38 - Arg74 Ser38 - Met(O) - Arg74 Ser38 - Met(O)2 - Arg74 Ser38 - Met(O)3 - Arg74 Lys75 - Arg86 Lys75 - Met(O) - Arg86 Met76 - Arg86 Met76 - Met(O) - Arg86 Met76 - Arg90 Met76 - Met(O) - Arg90 Asp78 - Arg86 Asp78 - Arg90 Val91 - Arg106 Val91 - NY - Arg106 Asp95 - Arg106 Asp95 - NY - Arg106 His107 - Lys115 His107 - Met(O) - Lys115 Leu116 - Arg126 Leu116 - Met(O) - Arg126 Glu127 - Lys148 Glu127 - Met(O) - Lys148 Glu127 - Met(O)2 - Lys148 Glu127 - Met(O)2 - NY - Lys148
[M + 3H]3+ observed
[M + 4H]4+ observed
a
[M + H]+ experimental
[M + H]+ theoretical
1521.83 956.53 907.51 1844.96 805.47 821.45 4070.10 4086.14 4102.14 4118.10 1480.76 1496.78 1352.63 1368.68 1855.94 1871.96 1093.53 1596.80 1754.96 1799.93 1265.71 1310.65 1028.55 1044.55 1349.71 1365.67 2490.23 2506.22 2522.18 2567.18
1521.74 956.47 907.44 1844.89 805.42 821.42 4069.85 4085.84 4101.84 4117.83 1480.70 1496.69 1352.60 1368.60 1855.85 1871.84 1093.46 1596.71 1754.87 1799.86 1265.61 1310.60 1028.52 1044.51 1349.63 1365.62 2490.08 2506.08 2522.07 2567.06
761.42 478.77 454.26 615.66 403.24 411.23 1018.28 1022.29 1026.29 1030.28 494.26 499.60 451.55 456.90 619.32 624.66 547.27 532.94 585.66 600.65 633.36 655.83 514.78 522.78 675.36 683.34 830.75 836.08 841.40 856.40
a Tryptic peptides use common nomenclature (35, 36) and are listed relative to the amino-terminus together with respective sequence. Posttranslational modifications involving the formation of methionine sulfoxide (SO) or nitrotyrosine (NY) are indicated. Cleavage sites within the parent tryptic fragment are indicated with a letter designation. The experimental masses [M + H]+ are monoisotopic after charge state correction for comparison with predicted mass of peptides.
Table 2. Quantification of Methionine Sulfoxide and Nitrotyrosine Formationa wild-type CaM [ONOO-]/[CaM] 2.5 5 15 25
CaM-L9
Met(O)
nitrotyrosine
36
51, 71, 72
76
109
124
144, 145
99
138
99
138
0.02 n.d. 0.09 0.33
0.09 0.26 0.87 1.25
0.03 0.09 0.30 0.53
0.01 0.04 0.22 0.39
0.05 0.11 0.40 0.63
0.62 0.77 0.94 0.83
0.10 0.14 0.27 0.22
