Tyrosine nitration on calmodulin enhances calcium-dependent association and activation of nitric-oxide synthase

Abstract

Production of reactive oxygen species caused by dysregulated endothelial nitric-oxide synthase (eNOS) activity is linked to vascular dysfunction. eNOS is a major target protein of the primary calcium-sensing protein calmodulin. Calmodulin is often modified by the main biomarker of nitroxidative stress, 3-nitrotyrosine (nitroTyr). Despite nitroTyr being an abundant post-translational modification on calmodulin, the mechanistic role of this modification in altering calmodulin function and eNOS activation has not been investigated. Here, using genetic code expansion to site-specifically nitrate calmodulin at its two tyrosine residues, we assessed the effects of these alterations on calcium binding by calmodulin and on binding and activation of eNOS. We found that nitroTyr–calmodulin retains affinity for eNOS under resting physiological calcium concentrations. Results from in vitro eNOS assays with calmodulin nitrated at Tyr-99 revealed that this nitration reduces nitric-oxide production and increases eNOS decoupling compared with WT calmodulin. In contrast, calmodulin nitrated at Tyr-138 produced more nitric oxide and did so more efficiently than WT calmodulin. These results indicate that the nitroTyr post-translational modification, like tyrosine phosphorylation, can impact calmodulin sensitivity for calcium and reveal Tyr site-specific gain or loss of functions for calmodulin-induced eNOS activation.

Introduction

Fundamental biological processes like cell proliferation, gene transcription, cell death, exocytosis, and metabolism are predicated on the tight regulation of cytosolic Ca²⁺ concentration (1). At the heart of Ca²⁺ regulation is the Ca²⁺-sensing protein, calmodulin (CaM),³ which transduces intracellular Ca²⁺ signals into biological responses. CaM accomplishes this by serving as a Ca²⁺-binding regulatory subunit of a wide array of enzymes, structural proteins, and membrane transporters. Cellular CaM fluctuates between its Ca²⁺-bound form (Ca²⁺–CaM) and Ca²⁺ and target protein–bound forms (Ca²⁺–CaM–target), depending on the combination of intracellular Ca²⁺ levels and target affinity. Cytosolic Ca²⁺ levels are tightly regulated between ∼0.1 μm under resting conditions and ∼10 μm after stimulation (2). CaM is an abundant protein with intracellular concentrations up to 10 μm; however, because of the collective concentration of its many target proteins, CaM is the limiting agent in Ca²⁺ sensing (3, 4).

CaM target proteins in high concentration or that bind CaM with the highest affinity will scavenge the CaM from the cell. At maximal intracellular calcium levels, 25% of the Ca²⁺–CaM pool in endothelial cells is associated with endothelial nitric-oxide synthase (eNOS) (3). This amount of CaM bound by tight association with eNOS removes sufficient CaM from the intracellular pool to affect other target proteins as seen with the intracellular calcium regulatory channel, plasma membrane Ca²⁺-ATPase (3). Post-translational modifications (PTMs) on CaM serve as an additional level of control over CaM–target affinity and CaM activity necessary for the intracellular Ca²⁺ regulatory balance. CaM has two tyrosines, Tyr-99 and Tyr-138, which are phosphorylated at different stoichiometries to regulate target protein association (3, 6). Tyrosine phosphorylation of CaM has been shown to impact Ca²⁺-dependent CaM interaction with nitric-oxide synthase (NOS). PhosphoCaM-99 binds four times more tightly to neuronal NOS as compared with WT–CaM (4).

CaM is also susceptible to oxidative post-translational modifications (ox-PTMs) by reactive oxygen species and reactive nitrogen and oxygen species, such as H₂O₂, NO, and peroxynitrite (5–8). Proteomic analysis has shown that reactive oxygen species and reactive nitrogen oxygen species induce a variety of ox-PTMs on CaM in vivo, resulting in an accumulation of significant amounts of tyrosine nitration and methionine sulfoxidation (9–12). Previous studies have indicated that nitration of CaM at tyrosine 99 is a biomarker of oxidative stress and that nitration of CaM at tyrosine 138 is subject to a denitrase activity in activated macrophages (5, 11). Up to 30% of the cellular CaM pool is nitrated on critical regulatory tyrosine residues following macrophage activation (5); however, the effect of tyrosine nitration on CaM function has not been determined. However, given the impact of phosphorylation of these tyrosines on the CaM–NOS interaction, it is perhaps not surprising to think that CaM tyrosine nitration will alter CaM–eNOS affinity.

To evaluate the impact of ox-PTMs on CaM, previous work used standard site-directed mutagenesis with unreactive canonical amino acids to control sites of chemical oxidation. This approach, although the best available at the time, is problematic because altering amino acids to prevent ox-PTM formation can have unknown effects at these regulatory hot spots (11). Genetic code expansion (GCE) provides the means to assess the effects of site-specific protein tyrosine nitration through co-translational installation of nitroTyr into the protein of interest. This methodology has been used recently to show that tyrosine nitration can cause a toxic gain of function for key regulatory sites in heat shock protein 90 (Hsp90) and apolipoprotein A-I, as well as a loss of function in manganese superoxide dismutase (13–15).

The abundance of tyrosine nitration on CaM led us to characterize the effect this ox-PTM has on altering CaM function with site-specifically modified nitroTyr-containing CaM generated with GCE. We focused on CaM target eNOS because it utilizes a significant fraction of the cellular CaM pool and because of the regulatory role of tyrosine phosphorylation in the system. eNOS is a multidomain enzyme that catalyzes the conversion of l-arginine to the biological signaling molecule nitric oxide to regulate vascular tone and angiogenesis (16). In addition, oxidative stress has been shown to alter nitric-oxide levels and compromise the regulation of vascular tone and angiogenesis (12, 18).

Using homogeneously nitrated CaM, we found that tyrosine nitration increases the affinity of CaM for its key target protein, eNOS, at reduced calcium levels, and interestingly, site-specific tyrosine nitration also increases its ability to produce the product, NO. This study provides the first evidence that tyrosine nitration of CaM can provide a gain of function to the Ca²⁺-dependent activation of eNOS signaling.

Results

Production of site-specific nitroTyr–CaM

We sought to characterize the functional effects of site-specifically nitrating the tyrosine residues on CaM caused by nitroTyr–CaM abundance in cells. We installed nitroTyr co-translationally into CaM at positions 99 and 138 using GCE to generate homogeneous nitroTyr–CaM (Fig. 1). When nitroTyr was withheld from the media, no full-length protein was purified, indicating the fidelity of the engineered aaRS/tRNA pair for nitroTyr (Fig. 1B). The fidelity of site-specific nitroTyr incorporation was verified by MS, with a mass increase of 44 Da corresponding to the addition of a single nitro group (Fig. 1D). The presence of nitroTyr was further confirmed by anti-nitroTyr Western blotting (Fig. 1C). These cumulative results validate the specific incorporation of nitroTyr into CaM.

Figure 1.

Incorporation of nitroTyr into CaM expressed in E. coli. A, the human CaM construct with C-terminal His₆ tag bearing either the native Tyr codon or an amber stop codon at amino acid position 99 or position 138 was co-transformed with a plasmid expressing an Methanocaldococcus jannaschii tyrosyl-tRNA synthetase/tRNA_CUA pair engineered to efficiently incorporate nitroTyr. B, WT– and nitroTyr–CaM was expressed in autoinduction medium with or without nitroTyr supplementation, purified using the C-terminal His₆ tag, and analyzed by 15% SDS-PAGE gel. C, nitroTyr incorporation into CaM was determined here by Western blotting with a primary antibody against nitroTyr. D, electrospray ionization mass spectrometry confirms the quantitative incorporation of nitroTyr into CaM because all measured pure protein masses match their expected molecular masses.

nitroTyr–CaM exhibits a gain-of-function interaction with target protein eNOS

To determine whether nitration of CaM changed its Ca²⁺-dependent ability to bind target proteins, we supplemented EA.hy926 human endothelial cell lysate with the site-specific nitroTyr–CaM–V5 tag at 20% of the endogenous CaM concentration. Immunoprecipitation of CaM co-precipitated a clearly evident protein with an apparent molecular mass of 130 kDa, consistent with the CaM target eNOS (Fig. 2). Because eNOS was co-immunoprecipitated by both WT– and nitroTyr–CaM–V5 tag, this indicates that nitroTyr–CaM–V5 interacts with eNOS in the presence of other intracellular targets of CaM.

Figure 2.

WT–CaM, CaM–nitroTyr-99, and CaM–nitroTyr-138 interact with eNOS protein from HEK293–eNOS and EA.hy926 cell lines. Western blotting analysis of immunoprecipitates of V5-tagged WT–, nitroTyr-99–, and nitroTyr-138–CaM from lysates of HEK293 cells stably expressing eNOS (A) and EA.hy926 human endothelial cells (B) is shown. This confirms that WT– and nitroTyr–CaM interact with eNOS in the context of the cellular milieu containing other CaM target proteins and endogenous CaM. IB, immunoblotting; IP, immunoprecipitation; nY, nitroTyr.

To determine the extent of the impact of nitration of CaM on its Ca²⁺-dependent ability to bind eNOS, we employed the use of biolayer interferometry (BLI), outlined in Fig. 3. This method allows for uniform attachment of a target protein–binding region to a BLI optical biosensor tip and measurement of CaM binding in solutions containing 20–225 nm buffered Ca²⁺.⁴³ To determine whether CaM nitration alters its affinity for eNOS, we used the synthesized 20-amino acid peptide CaM-binding domain of eNOS with an N-terminal PEG-biotin to aid in solubility and allow for defined surface attachment to BLI streptavidin biosensor tips. The association and dissociation of CaM proteins to immobilized eNOS peptide was fit to a 1:1 interaction model that provides the K_d under different Ca²⁺ concentrations (Table 1 and Fig. 3) (20, 21).

Figure 3.

Outline of biolayer interferometry experiments to determine CaM–eNOS interactions at varying free Ca²⁺ concentrations. A, the peptide constituting the eNOS CaM-binding domain was synthesized with an N-terminal biotin-PEG that was attached to a streptavidin functionalized BLI tip (light blue lines). The eNOS peptide functionalized tip was then moved to a well containing a particular CaM species (WT–CaM or nitroTyr–CaM, shown in purple) and a buffered Ca²⁺ solution (Ca²⁺ shown as green circles). CaM associated (shown as the red curve) with the eNOS peptide on the surface of the tip. White light was transmitted through the BLI tip, and binding events resulted in a shift in the interference of white light, which is measured as a response in nanometers. Following association, the BLI tip is moved to a well containing the same buffered Ca²⁺ solution but without CaM present. The dissociation of CaM–eNOS peptide (shown as the blue curve) is monitored again as a response in nm over time. Fitting the data for both the association constant and dissociation constant gives the K_d of the interaction at each Ca²⁺ concentration. B, representative traces for WT–, nitroTyr-99–, and nitroTyr-138–CaM at 20, 50, and 100 nm free Ca²⁺ are shown.

Consistent with values reported in the literature, WT–CaM binding of eNOS peptide affinity ranges from ∼1 nm to undetectable levels as [Ca²⁺]_free is decreased from saturating levels (2 mm) to physiologically resting levels (20–50 nm) (17). Both nitroTyr–CaM species exhibit similar eNOS–peptide binding affinity to WT–CaM at saturating calcium levels. Strikingly, both of the nitroTyr–CaM species retain ∼5 nm affinity at physiologically resting [Ca²⁺]_free, compared with undetectable affinity for WT–CaM. Because of the retention of high affinity for eNOS at resting [Ca²⁺]_free, nitroTyr–CaM has the potential to constitutively activate eNOS in the absence of calcium signal, resulting in a gain-of-function alteration regulating eNOS activity.

nitroTyr–CaM exhibits a gain-of-function activation of eNOS function

Based on the BLI data, we can conclude that nitration of CaM leads to CaM–eNOS association at resting intracellular calcium concentrations; however, this change in affinity only has a regulatory gain of function if these nitration sites on CaM do not abolish eNOS enzyme activity. Electron transfer between eNOS domains depends on the reversible binding of CaM, which is governed through changes in the intracellular Ca²⁺ concentration. To determine whether the increased binding affinity of nitroTyr–CaM altered eNOS activity, we measured the steady-state NO synthesis activities of WT-eNOS in presence of WT–CaM and nitroTyr–CaMs (Table 2). The coincident rates of NADPH oxidation during the assays were also measured to determine the efficiency of NO production. The cytochrome c reduction assay also provides a means to assess correct electron transfer through the enzyme. Generally, these assays are conducted under saturating calcium levels (2 mm free Ca²⁺ concentration). At these Ca²⁺ concentrations, all CaM species bind eNOS with the same affinity (Table 1), so any differences in NO production seen are independent of CaM–eNOS affinity. It is not feasible to perform eNOS assays at very low Ca²⁺ concentrations because of slow eNOS turnover and detection limits of the oxyhemoglobin, NADPH, and cytochrome c assays. At saturating Ca²⁺ levels, the eNOS was ∼1.3 times more active with nitroTyr–CaM-138 than eNOS with WT–CaM. NitroTyr-CaM-99 had an inhibitory effect and lowered NO synthesis by ∼40%. We also compared how these CaMs impact the NADPH oxidation rates of WT-eNOS during NO synthesis from l-Arg. In general, the rate of NADPH oxidation followed the rate of NO synthesis, consistent with the coupling of these processes. Remarkably, the coupling in WT-eNOS with nitroTyr–CaM-138 (3.0 NADPH per NO) was more efficient than in WT-eNOS with WT–CaM (4.4 NADPH per NO). This means that that the eNOS works more efficiently with nitroTyr–CaM-138 and enables greater NO synthesis without an increased production of other reactive oxygen species.

The steady-state cytochrome c reductase activity is a useful way to measure the electron flux passing through the NOS FMN subdomain. Because of the closed conformation of the NOS reductase domain, cytochrome c activity is suppressed in the absence of CaM, and cytochrome c activity increases in the presence of CaM. A typical 5-fold increase in cytochrome c activity was induced by the binding of WT–CaM to WT-eNOS in our assays (18). We found that nitroTyr–CaM-138 has a 10% increase in activity as compared with WT–CaM and that nitroTyr–CaM-99 has a 10% decrease in activity as compared with WT–CaM. Cytochrome c reductase activity data also indicate that the eNOS works more efficiently with nitroTyr–CaM-138 than WT–CaM. When supplemented with either of the nitroTyr–CaM species as opposed to WT–CaM, eNOS retained more activity at 225 nm [Ca²⁺]_free (Table 3). This indicates that the gain of function seen with nitroTyr–CaM–eNOS binding is conserved with the full-length protein.

nitroTyr–CaM interacts with and stimulates the function of eNOS in eNOS-expressing HEK293 cell lysate

To recapitulate the effects seen in vitro with the pure CaM–eNOS system in a system more closely resembling the intracellular medium, we used lysate from eNOS-producing HEK293 cells, allowing us to still control both the concentration of supplemented CaM and [Ca²⁺]_free and verified that all CaM forms were able to stimulate eNOS activity in HEK293–eNOS lysate. As expected for eNOS we see a calcium concentration-dependent (Fig. 4) and time-dependent (data not shown) increase in activity. This activity is responsive to the NOS inhibitor l-NAME (data not shown), indicating that NO synthesis is a result of eNOS activity. Because of the presence of endogenous CaM, there was a significant eNOS activity without the addition of exogenous CaM (Fig. 4). As predicted from the increased affinity of nitroTyr–CaM-138 for eNOS at low calcium concentrations, the eNOS activity was stimulated to a significantly larger extent in lysate supplemented with nitroTyr–CaM-138 over WT–CaM (Fig. 4), particularly at intermediate calcium levels (225–750 nm [Ca²⁺]_free). Specifically, we see a 60–70% increase in NO production by nitroTyr–CaM-138 over WT–CaM at these intermediate calcium levels.

Figure 4.

NitroTyr-CaM stimulates production of NO by eNOS in HEK293–eNOS cell lysate. HEK293–eNOS cell lysate was supplemented with WT– or nitroTyr–CaM and eNOS co-factors. The production of NO^• was measured via NO selective electrode at different Ca²⁺ concentrations. This confirms that nitration of CaM at tyrosine 138 stimulates eNOS activity at low concentrations of Ca²⁺. The values shown here represent the averages of three measurements ± S.D. NS, not significant.

Nitration of CaM does not impair calcium binding

It has been shown that CaM PTMs regulate target affinity and activity by altering interactions with specific targets, but another potential mechanism would be for the PTM to alter CaMs affinity for Ca²⁺. Dansylated CaM (dansyl-CaM) fluorescence has also been used for monitoring conformational changes in CaM as a result of interactions with Ca²⁺ (19). To analyze the Ca²⁺-induced structural changes of the different dansyl-CaM species, we performed Ca²⁺ fluorescence titration experiments (Fig. 5). Fitting the data (Equation 2 in “Experimental procedures”) gives the EC₅₀(Ca²⁺), the free calcium concentration at which the dansyl-CaM is half-saturated with Ca²⁺, and n, the Hill cooperativity constant (Table 4). The EC₅₀(Ca²⁺) values for all CaM forms were ∼0.3 μm [Ca²⁺]_free. This unaltered Ca²⁺ affinity for the WT– and nitroTyr–CaMs is in good agreement with previously published dansyl-CaM Ca²⁺ affinity data (27). The Ca²⁺ fluorescence titration did show a difference for the Hill cooperativity coefficient for the WT–CaM as compared with the nitroTyr–CaMs (Table 4). This indicates that there is slightly lower cooperativity for the nitroTyr–CaMs as compared with the WT–CaM. The fluorescence titration data confirm that tyrosine nitration on CaM does not independently change its affinity for Ca²⁺, which indicates the increased nitroTyr–CaM–eNOS affinity at low Ca²⁺ concentrations is due to cooperative binding of both target eNOS and Ca²⁺ binding to nitrated CaM.

Figure 5.

CaM–Ca²⁺ affinity assessed by dansyl-CaM fluorescence following titration with Ca²⁺. Fluorescence changes of dansyl-CaM (WT, nitroTyr-99, and nitroTyr-138) were monitored at different Ca²⁺ concentrations to confirm that CaM–Ca²⁺ affinity is not changed in response to CaM tyrosine nitration. The normalized fluorescence is shown for CaM under assay conditions described under “Experimental procedures.” The values shown here represent the averages of three measurements ± S.D.

Discussion

Previous work has shown that eNOS is almost exclusively CaM-free under resting conditions but is half-bound at roughly a quarter of maximum intracellular [Ca²⁺]_free (50% saturation of the eNOS peptide at a [Ca²⁺]_free of 228 nm) (27). Although unmodified CaM is inactive under resting [Ca²⁺]_free, tyrosine phosphorylation of CaM increases binding affinity for NOS compared with unphosphorylated CaM (4). The sites of phosphotyrosine regulation on CaM are also shared by tyrosine nitration, which makes it important to understand how this ox-PTM alters CaM signaling. CaM is maintained at low basal levels of tyrosine phosphorylation and nitration.

In response to insulin signaling or infection by Rous sarcoma virus, the amount of phosphotyrosine increases up to 50%, and nitrotyrosine levels increase up to 30% because of oxidative stress and immune responses (5, 20). The balance of CaM phosphorylation and nitration further broadens the pool of regulatory components that control eNOS function (5, 21–23).

Our results show nitroTyr–CaM-138 remains bound even at 20 nm [Ca²⁺]_free, indicating that eNOS in the presence of nitroTyr–CaM-138 is active under resting conditions, analogous to what was seen for phosphoCaM-99. Both of the nitroTyr–CaM species bind the same as the WT–CaM with identical subnanomolar affinities under saturating [Ca²⁺]_free; however, they exhibit a slight decrease in affinity for the eNOS peptide at the resting [Ca²⁺]_free of 50 nm. Previously, phosphorylation of eNOS at Ser-1179 by Akt or PKA has shown a gain of function by allowing CaM to bind at lower [Ca²⁺]_free levels, effectively lowering the calcium signal required for eNOS activation (24). In the same vein, nitration of CaM at Tyr-138 provides a similar but newly uncovered mechanism for enhancing CaM interactions and thereby activating eNOS independent of an increase in the [Ca²⁺]_free level (Fig. 6). Tyr-138 serves as a structural coupler between the N-terminal domain and the central linker of CaM through hydrogen bonding with Glu-82 (Fig. 6) (25), potentially explaining the alteration in eNOS association seen following nitration at this site. Although phosphorylation of CaM has been shown to alter the affinity of CaM for eNOS, it is also clear that phosphorylation of CaM or eNOS also influences the activity of eNOS in a manner independent from CaM–eNOS affinity. Based on this, we expect that tyrosine nitration of CaM will impact eNOS function in addition to the changes to affinity.

Figure 6.

Potential effects of CaM nitration on eNOS function. The two sites of physiological tyrosine nitration in calmodulin are shown, with Tyr-99 in yellow, Tyr-138 in orange, and Ca²⁺ shown as green spheres corresponding to the van der Waals radius. Hydrogen bonding between Glu-82 in purple and Tyr-138 is key to structural coupling between the N- and C-terminal lobes of CaM. Nitration of Tyr-138 lowers the pK_a of the residue resulting in a weakened hydrogen bond, likely lowering interlobe coupling. The carbonyl oxygen of Tyr-99 (yellow) involved in chelating Ca²⁺ in EF-hand III is also shown. A gain of function may occur if holoCaM (Protein Data Bank code 3CLN) binds to a target protein, like eNOS, with greater affinity (eNOS–CaM complex, Protein Data Bank code 1NIW). Further, a gain of function could occur if nitroTyr–CaM binding induces increased target protein activity. Nitration of these sites leads to different effects on eNOS, and the gain-of-function change caused by nitroTyr–CaM-99 leads to decreased activity of eNOS, whereas the gain of function caused by nitroTyr–CaM-138 leads to increased activity of eNOS.

To study the influence of CaM nitration on the catalytic activity of full-length WT-eNOS, we monitored NO production, NADPH oxidation, and cytochrome c reduction. Previous work with the phosphomimetic CaM mutation Y99E showed a 40% decrease in NO production, whereas the control mutation Y99Q showed only a 20% decrease in NO production (23). At 2 mm [Ca²⁺]_free, similar to the Y99E mutation, nitroTyr–CaM-99 resulted in a 40% decrease in NO production and was less efficient with a 10% decrease in NADPH to NO ratio as compared with WT–CaM. Strikingly nitroTyr–CaM-138 was able to support more NO production and more efficiently than WT–CaM, with a 33% increase in NO output and a 62% efficiency increase. This increase in output and efficiency is comparable with what was seen for eNOS regulation by the phosphomimetic mutations S1179D or S617D (18, 26).

Because we do not expect the affinity of nitroTyr–CaM for eNOS to be altered at saturating [Ca²⁺]_free as compared with WT–CaM, we may be able to ascribe the change in eNOS function to a shift in eNOS domain dynamics (16). Under intermediate calcium conditions (225 nm [Ca²⁺]_free), both nitroTyr–CaMs retained 10% more eNOS activity than the WT–CaM, indicating that nitration will regulate CaM to increase NO at such intermediate calcium levels in the cell.

Because CaM regulates many proteins, the nitroTyr–CaM pool may impact target proteins other than eNOS. Regulation of eNOS is also multifaceted and is also dependent on interactions with proteins other than CaM and subcellular localization. To approach this problem, we first assessed eNOS-nitroTyr–CaM association in HEK293–eNOS and EA.hy926 human endothelial cell lysate. We saw that both nitroTyr–CaM species when supplemented to cell lysate at 20% of the total CaM pool interacted with eNOS via immunoprecipitation of eNOS via the V5 tag on CaM. Assessing eNOS function, we saw that supplementation with nitroTyr–CaM, particularly at site 138, led to significantly more eNOS activity at intermediate calcium levels representing a gain-of-function even in the presence of native CaM, other CaM target proteins, and other eNOS regulatory interactions. Under this condition, the nitroTyr–CaM-138 does stimulate 60–70% more eNOS activity in cell lysate than does WT–CaM, consistent with a gain of function.

Conclusion

Here we sought to explore whether the nitroTyr PTM can alter the function of regulatory protein (27), in our case CaM. The central role CaM plays in regulating calcium signaling, abundant nitration of its tyrosine residues, and the potential presence of a cellular denitrase for nitrated CaM (5) make it an ideal candidate for assessing the impact of tyrosine nitration. Because nitroTyr can be incorporated site-specifically using GCE, it provides excellent opportunity to evaluate whether it can modulate protein function. Our results show that tyrosine nitration of CaM at sites 99 and 138 increase binding and activation of eNOS at lower Ca²⁺ concentrations. Most strikingly, CaM nitrated at site 138 binds eNOS even under resting physiological conditions, indicating that although only a subset of cellular CaM will be nitrated on tyrosine 138 at any given time, because of its gain of function, this subpopulation will activate eNOS at reduced [Ca²⁺].

Proper regulation of eNOS is important for healthy vascular function, and tyrosine nitration enhances CaM calcium sensitivity and activation of eNOS. CaM binding to NOS activates NOS by relieving the suppression of the electron transfer process. Under conditions of limited arginine substrate or improper coupling of N-terminal oxygenase domain and C-terminal reductase domain NOS has also been observed to produce superoxide through electron transfer to molecular oxygen (28). Because eNOS produces both superoxide and nitric oxide, which can react to form peroxynitrite (8, 29), the nitrating agent of CaM, tyrosine nitration may serve as a positive feedback mechanism to stimulate eNOS function. Based on the efficiency of nitric oxide production, we would predict from these results that nitration of CaM at Tyr-99 will lead to increased eNOS decoupling and therefore tyrosine nitration with lower NO bioavailability, whereas nitration at Tyr-138 will decrease uncoupling and the level of tyrosine nitration but increase the overall amount of bioavailable NO.

Experimental procedures

Antibodies used in this study

Nitrotyrosine antibody was obtained from Cayman Chemical (nitrotyrosine polyclonal antibody, item no. 10189540), V5 tag mAb was obtained from Invitrogen (V5 tag monoclonal antibody, mouse, catalog no. R960-25), NOS antibody was obtained from Cell Signaling Technology (NOS (pan) antibody no. 2977), and anti-rabbit and anti-mouse secondary antibodies were obtained from Li-Cor Biosciences (IRDye 800CW donkey anti-rabbit IgG secondary antibody P/N 926-32213; and IRDye 800CW donkey anti-mouse IgG secondary antibody P/N 926-32212).

Immunoblotting

Western blotting samples were separated on 15% SDS-PAGE gels, transferred to PVDF membrane, blocked with 5% nonfat milk in TBST, and probed with antinitrotyrosine (1:500) primary antibody, rocking for 16 h at room temperature. After rinsing three times with TBST, the membranes were than incubated with Li-Cor IRDye 800CW goat anti-rabbit IgG (1:10,000) secondary antibody, rocking for 1 h at room temperature, and washed three times for 5 min in TBST. The membrane was then scanned using a Li-Cor Odyssey 9120 imaging system.

Recombinant expression of homogenous site-specifically modified nitroCaM

The human CaM sequence was codon optimized for expression in Escherichia coli (GenScript) and cloned into a pBad vector to include a C-terminal His₆ affinity tag (Invitrogen). The protein was expressed via DH10B E. coli cells in autoinduction medium in the presence of 100 μg/ml ampicillin for 24 h at 37 °C with shaking at 250 rpm. The cells were pelleted at 5500 × g, resuspended in ∼10 ml binding/wash buffer (20 mm Tris, 500 mm NaCl, 20 mm imidazole, pH 7.4) at 4 °C, and lysed once with a Microfluidics M-110P microfluidizer set at 18,000 p.s.i. Cell debris was pelleted in Oakridge tubes at 20,900 × g for 25 min at 4 °C. Approximately 75 ml of supernatant was passed through an Acrodisc 32-mm syringe filter fitted with a 0.45-μm Supor membrane. The supernatant was loaded onto a 5-ml HisTrap nickel–nitrilotriacetic acid column at 1 ml/min, washed with 20 ml of wash buffer, and eluted with 0–100%, 30-ml linear gradient of elution buffer (20 mm Tris, 500 mm NaCl, 500 mm imidazole, pH 7.4) using an Amersham Biosciences AKTA explorer. Peak elutions were between 30 and 40% elution buffer, corresponding to 150–200 mm imidazole. The pure protein fractions then had CaCl₂ added to a concentration of 5 mm and loaded on a 5-ml HiTrap phenyl-Sepharose column, washed with 20 ml of wash buffer (50 mm Tris-HCl, 1 mm CaCl₂, 500 mm NaCl, pH 7.5), and eluted with 0–100%, 30-ml linear gradient elution buffer (50 mm Tris-HCl, 1 mm EDTA, pH 7.5). Pure peak fractions determined via SDS-PAGE analysis were pooled and dialyzed for 4 h into storage buffer (50 mm Tris-HCl, 100 mm NaCl, pH 7.5) twice, and fresh storage buffer was added for overnight dialysis. The purified CaM was concentrated via Centricon tubes (3-kDa molecular mass cutoff) to a concentration greater than 400 μm. All CaM used for co-immunoprecipitations, eNOS function assays, BLI measurements, and calcium-binding assays was prepared in this manner and used in this buffer unless noted elsewhere under “Experimental procedures.”

Expression and purification of CaM containing nitroTyr

E. coli DH10B was transformed with pBad–CaM–(99 or 138 TAG) and pDule–nitroTyr–5B (Addgene plasmid no. 85498) (30). The orthogonal aminoacyl-tRNA synthetase and cognate amber suppressing tRNA for incorporation of nitroTyr are expressed from the plasmid pDule containing the aforementioned synthetase/tRNA pair, a p15 origin, and a tetracycline resistance marker. The pBad–CaM plasmids contain CaM with an amber stop codon at the site of interest (99 or 138) to direct incorporation of nitroTyr. Similar to WT–CaM, expression of nitroTyr-containing CaM was in autoinduction medium (31) containing 100 μg/ml ampicillin, 25 μg/ml tetracycline, and 1 mm nitroTyr. Purification of nitroTyr-containing CaM was identical to purification of WT–CaM. Yields of ∼300 mg (liters of culture)⁻¹ were obtained for nitroTyr–CaM-99 and nitroTyr–CaM-138, compared with that of ∼375 mg (liters of culture)⁻¹ for WT–CaM expression (31).

Mass spectrometry

Purified CaM samples were diluted to a concentration of 10 μm, desalted on Millipore C₄ zip tips, and analyzed using an FT LTQ mass spectrometer at the Oregon State University Mass Spectrometry Facility. The samples included WT–CaM and nitroTyr-containing CaM. Spectra were collected using Tune-Plus (Thermo, version 2.2) page of Xcalibur (Thermo, version 2.0.5) using parameters described in Ref. 32. The spectra were averaged over the 3-min CaM eluted from the ZipTip using the Qual Browser (Thermo, version 2.0). The data were exported as text files containing two columns of m/z versus intensity to be analyzed by custom programs written in MatLab (32). The raw and deconvoluted MS data were deposited online.

Calmodulin dansylation and fluorescence measurements

Dansyl-CaM was prepared as previously described (23). CaM (1 mg/ml) was buffer-exchanged into 10 mm NaHCO₃ and 1 mm EDTA (pH 10.0) at 4 °C. 30 μl of 6 mm dansyl chloride (1.5 mol/mol of CaM) in acetone was added to 2 ml of CaM while it was being stirred. After incubation for 12 h at 4 °C, the mixture was buffer-exchanged into fluorescence buffer. Labeling yields were determined from absorbance spectra using an ϵ₃₂₀ of 3400 m⁻¹ cm⁻¹ and were compared with actual protein concentrations determined using the Bradford method with WT CaM used as the protein standard.

Fluorescence emission spectra were recorded using a photon technology international (London, Canada) QuantaMaster spectrofluorimeter. Fluorescence measurements were taken on 50-μl samples consisting of dansyl-CaM (2 μm) in 30 mm MOPS, 100 mm KCl, and 10 mm EGTA (pH 7.2) with an increasing concentration of free Ca²⁺. The free Ca²⁺ concentration was controlled using the suggested protocol from the calcium calibration buffer kit from Invitrogen. The excitation wavelength for all of the dansyl-CaMs was set to 340 nm, and emission was monitored between 400 and 600 nm. Slit widths were set at 2 nm for excitation and 1 nm for emission. The relative fluorescence was calculated with the following equation,

$$\left(F-F_0\right)/\left(F_{\max }-F_0\right)$$ (Eq. 1)

where F is the measured intensity, F_max is the maximal intensity, and F₀ is the intensity without added Ca²⁺. The relative fluorescence was calculated using the fluorescence emission intensity at 475 nm. The Ca²⁺ sensitivities of the different dansyl-CaMs we determined as the EC₅₀(Ca²⁺) values, which were derived from fits of the relative fluorescence intensity increase upon addition of Ca²⁺ using the following equation,

$$\frac{{\left[{\text{Ca}}^{2+}\right]}_{\text{free}}^n}{{\left[{\text{Ca}}^{2+}\right]}_{\text{free}}^n+{\left[{\text{EC}}_{50}\left({\text{Ca}}^{2+}\right)\right]}^n}$$ (Eq. 2)

where relative fluorescence is obtained from Equation 1, and n is the Hill coefficient.

Octet Red96 biolayer interferometry measurements

All BLI measurements were made on a fortéBIO (Menlo Park, CA) Octet Red96 system using streptavidin sensors. The assays were performed in 96-well microplates at 37 °C. All sample volumes were 200 μl. The eNOS peptide was purchased from Genscript (Piscataway, NJ) homogeneously biotinylated at the N terminus. Tips were loaded with three different concentrations of biotin-eNOS peptide (33). After loading biotinylated eNOS peptide onto streptavidin sensors, a baseline was established in buffer composed of 30 mm MOPS, 100 mm KCl (pH 7.2), and varying free calcium concentration. Free calcium was controlled by mixing two buffers containing 10 mm EGTA and 10 mm Ca²⁺-EGTA in varying ratios. Free calcium concentration was calculated either from the Thermo Fisher scientific calcium calibration kit 1 instructions or the Maxchelator program (34). Association with the analyte CaM was then carried out in the same buffer for 90 s at CaM concentrations of 180, 60, and 20 nm. Dissociation was subsequently measured in buffer without CaM over 1200 s. CaM used in the BLI experiments was unaltered following the purification method outlined above.

Steady-state eNOS assays

NO synthesis and NADPH oxidation rates were determined using the oxyhemoglobin assay. The NO synthesis activity was determined by the conversion of oxyhemoglobin to methemoglobin using an extinction coefficient of 38 mm⁻¹ cm⁻¹ at 401 nm. The NADPH oxidation rates were determined following the absorbance at 340 nm, using an extinction coefficient of 6.2 mm⁻¹ cm⁻¹. Reaction mixtures (total volume 400 μl) contained 0.1–0.2 μm bovine eNOS, 0.3 mm dithiothreitol, 4 μm FAD, 4 μm FMN, 10 μm H₄B, 2 mm l-Arg, 0.1 mg/ml BSA, 2 mm CaCl₂, 0.2 mm EDTA, 2–5 μm CaM (WT or nitroTyr depending on the experiment), 100 units/ml catalase, 60 units/ml superoxide dismutase, 5 μm oxyhemoglobin, and 150 mm NaCl in 40 mm EPPS buffer, pH 7.6. The reaction was initiated by adding NADPH to a final concentration of 250 μm. In the assays without CaM, CaCl₂ was omitted, and EDTA concentration was increased to 0.45 mm. Cytochrome c reductase activity was determined by following the absorbance change for the reduction of cytochrome c by eNOS at 550 nm using an extinction coefficient of 21 mm⁻¹ cm⁻¹. Reaction mixtures (total volume 400 μl) contained ≤0.01 μm eNOS, 25 μm FAD, 25 μm FMN, 0.1 mg/ml BSA, 2 mm CaCl₂, 0.2 mm EDTA, 1.0–4.0 μm CaM (WT or nitroTyr depending on the experiment), 100 units/ml catalase, 40 units/ml superoxide dismutase, 65 μm cytochrome c, and 150 mm NaCl in 40 mm EPPS buffer, pH 7.6. The reaction was initiated by adding NADPH to a final concentration of 250 μm. In the assays without CaM, neither CaM nor CaCl₂ were added, and EDTA concentration was 0.45 mm. All steady-state assays were carried out at 25 °C using a Hitachi 3110 spectrophotometer. The reported values are the means ± S.D. of three or more determinations.

Production of the stable HEK293–eNOS line

The bovine eNOS sequence in a pcDNA3 vector and empty pcDNA3 vector were transfected into HEK293 cells (ATCC) using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer's protocol. Stable HEK293 cell populations were selected for using G418 at 400 μg/ml. Expression of eNOS was confirmed by Western blotting using pan-NOS antibody (Cell Signaling Technology).

Immunoprecipitation of V5-tagged WT–CaM and NitroTyr-CaM from HEK293–eNOS cells

ENOS was co-immunoprecipitated from cells (either HEK–eNOS–expressing cells or EA.hy926 human endothelial cells) prepared in lysis buffer (50 mm Tris-HCl, pH 7.4, 150 mm NaCl, 1 mm EGTA, 1% Triton X-100, and 1× protease inhibitor mixture). Recombinant WT– or nitroTyr–CaM–V5 tag was supplemented into the lysate at 20% of the endogenous CaM concentration in the lysate along with 5 mm CaCl₂. The lysate with recombinant CaM was incubated at 4 °C overnight with 2 μl of mouse anti-V5 antibody (Invitrogen). The mixture was further incubated with 20 μl of protein A/G magnetic beads for 6 h at 4 °C. The magnetic beads were washed four times with lysis buffer containing 5 mm CaCl₂, resuspended in 40 μl of Laemmli buffer, and incubated at 55 °C for 10 min. Western blotting samples were separated on 4–22% gradient SDS-PAGE gels, transferred to PVDF membrane, blocked with 5% nonfat milk in TBST, and probed with anti-NOS (Pan) (1:1000) or anti-V5 (1:5000) primary antibodies rocking for 16 h at room temperature. After rinsing three times with TBST, the membranes were then incubated with Li-Cor IRDye 800CW goat anti-rabbit or anti-mouse IgG (1:10,000) secondary antibody, rocked for 1 h at room temperature, and washed three times for 5 min in TBST. The membrane was then scanned using a Li-Cor Odyssey 9120 imaging system.

eNOS-lysate reaction conditions

Reactions to determine eNOS activity in eNOS expressing HEK293 cellular lysate were carried out. Lysate extracted as above was supplemented with 2 mm l-arginine, 1 mm NADPH, 20 μm BH4 in solution containing 3 mm DTT, and 500 nm WT– or nitroTyr–CaM. For those reactions containing Ca²⁺ or EGTA, they were supplemented with 10 mm of either, whereas those reactions were supplemented with l-NAME to a concentration of 1 mm. The reaction was initiated by the addition of the NOS co-factors listed above and incubated for 30 min at 37 °C. The eNOS reactions were carried out at a 50-μl scale in triplicate and stopped via freezing in dry ice.

eNOS-activity assays

The amount of NO produced in lysate was measured amperometrically by an AmiNO700 NO selective electrode (Innovative Instruments, Inc.) as described previously (35).

Data Availability

The mass spectrometry data are publicly accessible online at Figshare (https://doi.org/10.6084/m9.figshare.11346536.v1).⁴

Author contributions

J. J. P. and R. A. M. conceptualization; J. J. P., H. S. J., and R. A. M. data curation; J. J. P., H. S. J., and R. A. M. formal analysis; J. J. P. validation; J. J. P., H. S. J., M. M. H., D. J. S., and R. A. M. investigation; J. J. P. visualization; J. J. P., M. M. H., and R. A. M. methodology; J. J. P., H. S. J., M. M. H., D. J. S., and R. A. M. writing-original draft; J. J. P., H. S. J., M. M. H., D. J. S., and R. A. M. writing-review and editing; D. J. S. and R. A. M. funding acquisition; R. A. M. supervision; R. A. M. project administration.