| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular Medicine |
From the Johns Hopkins University School of Medicine (L.S., H. Kyo Lim, H. Kyoung Lim, V.M., T.B., M.P., S.B., S.R., M.A., D.N., R.R., A.S., D.E.B.), Baltimore Md; Department of Anesthesiology and Pain Medicine (H. Kyo Lim), Yonsei University, Wonju, Korea; University of Pittsburgh Medical Center (K.I., F.K.), Pa; University of Miami Miller School of Medicine (J.M.H.), Fla; and Department of Chemistry (L.D.C., D.W.C.), University of Pennsylvania, Philadelphia.
Correspondence to Dan E. Berkowitz, Johns Hopkins University School of Medicine, 720 Rutland Ave, Traylor 621, Baltimore MD 21209. E-mail dberkowi{at}jhmi.edu
| Abstract |
|---|
|
|
|---|
Key Words: arginase NO synthase S-nitrosylation aging
| Introduction |
|---|
|
|
|---|
Given the interaction between NOS and arginase signaling, we hypothesized that S-nitrosylation of arginase1 might be an important posttranslational modification mechanism that regulates its activity. The protein sequence for human arginase1 contains 3 cysteines, C45, C168, and C303, with C303 being very close to R308 and the C-terminal S-shaped tail of arginase. The latter mediates 54% of the intersubunit interactions to form the active arginase homotrimer.21–23 Although the monomeric forms of arginase are active, they display only 15% to 25% activity compared with the trimer.21 We therefore studied S-nitrosylation of each cysteine residue of human arginase1, as well as the effect of S-nitrosylation on enzyme activity and the stability of the arginase1 trimer in vitro and ex vivo. Finally, we examined whether altered arginase nitrosylation could contribute to the pathobiology of vascular aging by limiting NO bioavailability, thereby contributing to impaired endothelial function. Because aging is accompanied by an increase in oxidative/nitrosative stress and inducible NOS (iNOS) expression,24,25 we determined whether iNOS plays a role in this mechanism. We demonstrate that arginase1 is activated by nitrosylation of C303, that this activation results from increased stabilization of the arginase trimer, that nitrosylation is iNOS-dependent, and that nitrosylation of arginase contributes to age-related endothelial dysfunction and impaired NO signaling.
| Materials and Methods |
|---|
|
|
|---|
Animals
Twelve old (22 to 24 months) and young (10 to 14 weeks) male Wistar rats were used in this study and were purchased from Harlan Laboratories (Haslett, Mich).
S-Nitrosylation
S-Nitrosylation of arginase was followed using the NitroGlo Assay kit (Perkin Elmer, Wellesley, Mass), which is based on the biotin-switch method.26
Arginase Activity Assay
The activity of arginase was measured colorimetrically as previously described.5
Cloning and Mutagenesis
Human arginase1 cDNA was obtained from Invitrogen and cloned into the mammalian expression vector pcDNA 3.1C-myc-His. Mutations were performed using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, Calif; Table I in the online data supplement).
Cell-Based Assays
PC 12 cells and RAW 264.7 were grown according to the protocol of the American Type Culture Collection. Human aortic endothelial cells (HAECs) were maintained according to the protocol of the supplier (Intracel Co, Frederick, Md).
Arginase Assay in Blood Vessels
Dissected rat thoracic and tail artery were treated as indicated. Rings were homogenized and arginase activity and biotin-switch assays performed.
NO Production in Isolated Rat Aorta
NO production in aortic rings was measured using DAF-FM fluorescence as previously described.27
Vascular Reactivity and Endothelial Function in Rat Arteries
Tail arteries were dissected and the external diameter measured using video microscopy as previously described.28
| Results |
|---|
|
|
|---|
|
Arginase1 Is Modulated by S-Nitrosylation
To determine whether S-nitrosylation altered arginase1 function, we determined the kinetics of purified bovine arginase1 (Figure 1C through 1E) and purified recombinant human arginase1-myc-his (rhArginase1) (supplemental Figure IA) with increasing concentrations of GSNO or DEANONOate (data not shown). S-Nitrosylation activated arginase1 by effecting a significant reduction in the Km value (2-fold decrease for bovine at 1000 nmol/L GSNO [Figure 1C through 1E] accompanied by a modest (10%) change in the value of Vmax normalized to mg enzyme used in assay [Figure 1D, inset]). There was approximately a 6-fold decrease in Km for wild-type (WT) rhArginase1 at 500 nmol/L GSNO (supplemental Figure IA and Figure 4A at pH 9). Furthermore, treatment of bovine arginase1 with ascorbate to reduce SNO bonds led to a decrease in its activity (Figure 1F), which was matched by a loss of SNO content (Figure 1G) measured by Western blotting.
Recombinant Human Arginase1 Is S-Nitrosylated in Cells
Because purified human arginase1 was S-nitrosylated in vitro, we wished to determine whether S-nitrosylation occurs in a cell-based system. To this end, we studied PC-12 cells transfected with rhArginase1. This cell line was chosen because of its low level of endogenous arginase activity. Lysates from NO donor–treated transfected cells demonstrated enhanced arginase activity at both pH 9 (Figure 2A) and pH 7.4 (supplemental Figure IIB) in the context of constant arginase1 expression (Figure 2B and supplemental Figure IC).
|
Endogenously Produced NO Modulates Arginase Activity
To determine the effects of endogenous NO production on arginase activity, we stimulated RAW264.7 cells and HAECs with 100 U/mL interferon (IFN)
and 5 µg/mL lipopolysaccharide (LPS) to induce iNOS in the presence and absence of the iNOS inhibitor 1400W,18 and measured arginase activity (at pH 9) after 3 hours of stimulation. Induction of iNOS in RAW264.7 cells led to increased arginase activity (Figure 3A), whereas the expression of both arginase1 and arginase 2 remained constant (Figure 3B). This is important because the prolonged exposure of cells (>10 hours) to cytokines and LPS results in an increase in arginase expression.29 Furthermore, the iNOS inhibitor 1400W abrogated this increase, suggesting that increased arginase1 activity was iNOS dependent. Treatment of cell lysates with ascorbate markedly attenuated arginase activity in all cases. Finally, stimulated RAW264.7 cells showed increased activity at pH 7.4 in a time-dependent manner (Figure 3C), whereas arginase1 and -2 expression levels were unaffected (Figure 3D). This was matched by increased accumulation of nitrite in the cell culture medium (Figure 3E), confirming NO production. Similarly, HAECs treated with IFN
/LPS show increased arginase activity, which was reversed by the iNOS inhibitor 1400W and by ascorbate (Figure 3F), whereas expression of arginase1 and -2 remained constant (Figure 3G). The increased expression of NOS2 3 and 6 hours following stimulation was confirmed using immunoprecipitation and Western blotting.30 Accumulation of S-nitrosylated proteins in HAECs was confirmed using the Griess–Saville assay (Figure 3H). This further confirmed the modulation of arginase activity by endogenously produced NO.
|
Arginase1 Is S-Nitrosylated at C168 and C303
To identify the particular cysteine(s) that undergoes S-nitrosylation, we generated mutants of human arginase1 in which each of the 3 cysteines were modified to glycines, to obtain the 3 single mutants C45G, C168G, and C303G. These were overexpressed in PC-12 cells and purified. Their activity was the compared with WT arginase1 (WT). There was an increase in the specific activity (nmol of urea/[h · mg of enzyme · Km]) for both WT (6-fold) and C45G mutant (4-fold; Figure 4A and supplemental Figures IA and IIA through IIC). C168G exhibited a modest increase in activity, with increasing GSNO concentration; however, the activity levels were significantly lower than WT. C303G was relatively unaffected by GSNO treatment at pH 9 (Figure 4A). At pH 7.4, only C303G was unaffected by GSNO treatment (Figure 4B and supplemental Figure IIA through IIC). This suggested that both C168 and C303 are likely candidates for S-nitrosylation and that C303 is involved in the activation of arginase1. The role of S-nitrosylation of C168 is unclear. Additionally, we performed the biotin-switch assay on lysates of PC-12 cells transfected with arginase1 WT and the 3 single mutants and determined that C168G and C303G mutants had a reduction in total nitrosylation (Figure 4C). Furthermore, similar experiments with double mutants revealed lack of nitrosylated arginase1 in the C168/303G double mutant (Figure 4D), confirming that these 2 cysteines undergo S-nitrosylation.
|
S-Nitrosylation of Arginase1 at C303 Affects Trimer Stability
Arginase1 exists and acts in a homotrimeric form, and a loss of activity is observed when the trimer form is disrupted.6 Because C303 lies close to the S-shaped C-terminus tail of arginase, which, in large part, mediates the subunit–subunit interactions, we hypothesized that nitrosylation of this residue might influence trimer stability. We therefore tested the effect of S-nitrosylation on the trimerization of arginase1. Indeed, as seen in Figure 4E, GSNO treatment of intact PC-12 cells transfected with WT and the 3 single mutants led to increased trimer form, as determined by native PAGE in WT, C45G, and C168G, whereas in C303G, there was a decrease in the population of trimer when compared with untreated sample. This was observed in the setting of increased nitrosylation in all samples treated with GSNO, while arginase expression remained constant.
Arginase1 Is S-Nitrosylated in Aging Rats
Because rat aortic endothelium is known to express arginase15,6,8,31 and arginase dysregulation appears to contribute, in part, to aging related endothelial dysfunction, we examined the effect of GSNO and ascorbate on arginase activity and nitrosylation in homogenates of rat aorta from old and young rats (n=4). Arginase1 derived from rat aortic tissue was modulated by exogenous NO (Figure 5A), and ascorbate reversed this effect. Furthermore, using the biotin-switch assay, we determined that endogenous arginase1 was significantly nitrosylated in old rat aorta compared with the young (Figure 5B).
|
Arginase Inhibition Enhances NO Production in Old Rat Aorta
We next confirmed the regulation of NOS by arginase in rat aorta by examining whether the inhibition of arginase enhances NO production. We measured NO production in rat aorta using the NO sensitive fluorescent dye DAF-FM (4-amino-5-methylamino-2',7'-difluorofluorescein) as previously described,27 with some modifications. There was a significant decrease in the baseline rate of the DAF fluorescence signal in aorta from old versus young rats (n=4; 2 -fold difference, P<0.05; Figure 6A). Incubation of the blood vessels from aged rats with 10 µmol/L S-(2-boronoethyl)-L-cysteine (BEC) enhanced NO production and restored it to the levels measured in young rats, such that they were statistically indistinguishable (Figure 6A). These changes were clearly measured as changes in the slope of DAF fluorescence (Figure 6A, inset, representative trace).
|
Arginase Inhibition Improves Endothelium-Dependent Vasodilation in Old Rats
It is well described that vascular aging results in impaired endothelial-dependent NO bioavailability and depressed vasorelaxant responses. We have also demonstrated previously that arginase1 reciprocally regulates endothelial NOS (eNOS) and is activated in endothelium of old rats.5 Furthermore, we have demonstrated that the knockdown of arginase1 restores endothelial-dependent vasodilator function in old vessels toward that of young. Because this physiologic phenomenon has been explored previously in the rat aorta,5 we tested the effect of arginase on endothelial function in another vascular bioassay: the rat tail artery. Because the vessels are pressurized and video-dimension analysis used to determine vascular tone/diameter, this is a good alternative physiologic bioassay. Firstly, arginase1 was expressed in tails of old and young rats (Figure 6C). Tail arteries from old rats were pressurized to 70 mm Hg, and preconstricted using phenylephrine. Responses to acetylcholine were tested either before or following incubation with BEC. Vasodilation was significantly impaired in old rat tail arteries (Emax=51.8±3.3; log(EC50)= –6.2±0.1) compared with young (Emax=88.3±3.0; log(EC50)= –6.4±0.1) (Figure 6B). Preincubation of vessels with BEC enhanced endothelial-dependent relaxation, such that they were toward that of the young blood vessels (Emax=68.2±3.3; log(EC50)=–6.6±0.1).
Endothelial iNOS Mediates Nitrosylation and Activation of Arginase1 in Aorta of Aging Rats
There is accumulating evidence that iNOS expression is enhanced in the endothelium of old rat aorta and contributes to age-related oxidative stress.24,25 We confirmed that iNOS is indeed expressed in the aorta of old but not young rats (Figure 7A). Endothelium-denuded rings had lower levels of NOS2, confirming its upregulation in the endothelium of old rat aorta.24,25 Furthermore, in our cell culture experiment, inhibition of iNOS led to reduced arginase1 activity. We thus tested the hypothesis that increased iNOS expression contributes to S-nitrosylation of endothelial arginase-1, thereby enhancing its activity. Aortic rings (endothelium intact and denuded) from old and young rats were incubated in the absence or presence of the iNOS-specific inhibitor 1400W for 16 hours, and arginase activity and nitrosylation were measured. As can be seen in Figure 7B, arginase activity was markedly reduced in endothelium-denuded compared with endothelium-intact rings in old rats. Interestingly, preincubation of rings from old rats with 1400W resulted in a significant decrease in arginase activity compared with untreated old controls. 1400W had no effect in the rings from young rats. The arginase activity levels in old rats with 1400W treatment were indistinguishable from the arginase activity in young rats. On the other hand, iNOS inhibition had no effect on arginase activity in the endothelium-denuded rings of either young or old rats. Arginase1 expression was greater in the intact rings when compared with the endothelium-denuded rings (Figure 7C). Furthermore, nitrosylated arginase1 was significantly higher in old rats compared with young (Figure 7D). Pretreatment of the rings with 1400W led to a significant decrease in nitrosylated arginase1 levels compared with controls (Figure 7D). At the same time, this led to decreased arginase trimer levels. This supports the idea that iNOS-dependent S-nitrosylation of arginase in the endothelium contributes to age-related increases in activity which, in turn, contributes to endothelial dysfunction as a result of impaired eNOS signaling.
|
| Discussion |
|---|
|
|
|---|
The modulation of arginase1 through S-nitrosylation (Figures 1 through 4![]()
![]()
) is a novel finding and has implications for the regulation of NOS by arginase. The activity assays for arginase were performed at pH 9, which is recognized as the optimal pH for arginase activity in vitro and is widely used. Representative experiments were repeated at pH 7.4 to assess the validity of these observations under more physiologic conditions. Although, in general, we observed a lower activity of arginase1 at pH 7.4 (compared with pH 9), the trends (increase or decrease) in activity were consistent at both pH values.
In this study, low concentrations of NO donor enhanced arginase1 catalytic efficiency (kcat/Km ratio), predominantly because of a decrease in Km. Indeed, for WT rhArginase1, Km decreased from 10 mmol/L to 1.5 mmol/L (6-fold decrease; Figure 4A) at pH 9 and from 80 mmol/L to 40 mmol/L (2-fold decrease, Figure 4B) at pH 7.4. At higher concentrations of NO donor, there was a reversal from activation to deactivation. This was attributable to an increased Km, along with a modest decrease in specific activity (Vmax per milligram of enzyme; data not shown). However, this occurred at GSNO concentrations likely outside of the realm of physiologic or pathophysiologic significance. The kinetic data for the WT and 3 mutants (supplemental Figures IA and IIA through IIC and Figure 4A and 4B) suggested that both C168 and C303 are candidates for S-nitrosylation. Mutagenesis experiments confirmed this and showed that C303 undergoes nitrosylation and activates the enzyme at low GSNO concentrations, whereas C168 undergoes nitrosylation at high GSNO levels and leads to a deactivation of arginase1 (Figure 4A and supplemental Figures IA and IIA through IID). S-Nitrosylation of C168 and C303 may affect catalysis by causing slight structural changes that propagate
15 Å through the protein to the active site (Figure 4F). Of these, C303 is the closest (
13 Å) to the interface between subunits (Figure 4F), and S-nitrosylation of C303 may cause structural changes that result in the net stabilization of the arginase1 trimer (increased trimer in WT, decreased trimer in C303G with NO donor; Figure 4E). Because the high concentrations of GSNO required to nitrosylate C168 lie outside the physiologic/pathophysiologic concentrations of NO, we focused on the effects of C303 nitrosylation. Our data further suggested that the initial activation was important physiologically/pathophysiologically, as it might help regulate NOS activity by a negative-feedback mechanism (Figure 8). Although experimental evidence points to reciprocal regulation of NOS by arginase1,5,6,10 a direct comparison of Km values makes it seem unlikely. Some plausible explanations for this disconnect have included (1) the presence of endogenous NOS inhibitors (asymmetric dimethylarginines) in the cell and (2) proximity of NOS to L-arginine pools and cationic amino acid transporters.32,33 This study demonstrated that S-nitrosylation represents a further potential mechanism of nonessential activation of arginase1, reducing its Km to a concentration closer to NOS. This could explain, in part, the ability of arginase1 to regulate NOS function, particularly in the face of increased arginase1–SNO abundance.
|
We next determined whether arginase1 activity was affected by endogenous NO. We used RAW264.7 and HAECs stimulated with 100 U/mL IFN
and 5 µg/mL LPS for 3 hours. The induction of iNOS in stimulated HAECs was confirmed by immunoprecipitation and Western blotting (Figure 3H). In these systems, iNOS production induced by IFN
/LPS treatment led to increased arginase activity at constant arginase1 and arginase 2 expression (Figure 3).9 This was reversed by the use of the iNOS inhibitor 1400W and by treating the cell lysates with ascorbate, both of which led to reduced protein nitrosylation. We detected nitrosylated rhArginase1 even in cells transfected with rhArginase1 not treated with NO donors. This suggested that endogenous levels of NO are sufficient to nitrosylate arginase1 and modulate its activity.
Next, we examined whether nitrosylated arginase1 accumulates in aging rat aorta. Indeed, arginase activity and nitrosylation were enhanced in aorta from old rats (Figure 5A and 5B), and NO production was impaired (Figure 6A). GSNO treatment of homogenates of young rat aorta recapitulated the old phenotype. Conversely, ascorbate treatment of homogenates of old rat aorta significantly reduced arginase activity. BEC-mediated inhibition of arginase markedly enhanced NO production and restored it to levels of NO in young rats. Examining the rat tail artery using video dimensioning showed impaired vasodilatory response in old rats, which was restored by arginase inhibition (Figure 6B).
There is emerging evidence for the expression of iNOS in the endothelium of aging rat aorta (Figure 7A).22,23,34 Inhibition of iNOS with 1400W in aortic rings led to decreased arginase activity in endothelial intact rings from old rats (Figure 7B), accompanied by reduced arginase1–SNO (Figure 7D) and arginase1 trimer levels (Figure 7D). This was in contrast to the inhibition of eNOS, which results in increased arginase activity in endothelial cells.7 Furthermore, endothelium-denuded rings were insensitive to 1400W treatment (Figure 7B) and contained significantly less arginase1 (Figure 7C), suggesting that this is an endothelium-dependent phenomenon. In light of this, it is very likely that modulation of arginase1 activity via S-nitrosylation is a key mechanism in NOS–arginase reciprocal regulation and dysregulation with aging (Figure 8). The cytosolic distribution of both iNOS and arginase1 makes this spatially feasible. The role of iNOS as an important mediator of the pathobiology of protein S-nitrosylation is being recognized. Li et al35 recently demonstrated that radiation-induced activation of iNOS resulted in S-nitrosylation of the hypoxia inducible transcription factor-1
, leading to its enhanced stability. The nitrosylation of arginase1 represents another example in which increased protein S-nitrosylation may contribute to age-related vascular pathology. This is consistent with pathophysiologic processes described in age-related neurodegenerative diseases.36,37
In summary, we demonstrate that arginase1 is modulated by S-nitrosylation at physiologic/pathophysiologic levels of NO. At nanomolar concentrations of NO, activation of arginase1 is effected by stabilization of arginase1 trimer, leading to a decrease in the Km of arginase1 to a level that partly explains its ability to compete with NOS for the common substrate L-arginine, leading to a negative-feedback control of eNOS activity. Furthermore, iNOS-dependent NO production is responsible for arginase nitrosylation in the aorta of aging rats. This is an endothelium-dependent phenomenon: arginase1 is expressed largely in the endothelium and, hence, is mediated in this location. Our findings, therefore, further highlight the role of iNOS-dependent nitrosylation and posttranslational modification in the pathobiology of vascular aging.
| Acknowledgments |
|---|
Sources of Funding
This work was supported by NIH Grants R01 AG021523 (to D.E.B.), R01 HL70929 and P01 HL65608 (to K.I.); GM49758 (to D.W.C.); National Aeronautics and Space Administration grant NNJ05HF03G (to D.E.B.), National Space Biomedical Research Institute grant CA00405 through the National Aeronautics and Space Administration (to A.S.).
Disclosures
None.
| Footnotes |
|---|
Original received September 25, 2006; resubmission received June 13, 2007; revised resubmission received July 25, 2007; accepted August 3, 2007.
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
G. L. Pierce and T. J. LaRocca Reduced vascular tetrahydrobiopterin (BH4) and endothelial function with ageing: is it time for a chronic BH4 supplementation trial in middle-aged and older adults? J. Physiol., June 1, 2008; 586(11): 2673 - 2674. [Full Text] [PDF] |
||||
![]() |
P. M. Vanhoutte Arginine and Arginase: Endothelial NO Synthase Double Crossed? Circ. Res., April 25, 2008; 102(8): 866 - 868. [Full Text] [PDF] |
||||
![]() |
S. Ryoo, G. Gupta, A. Benjo, H. K. Lim, A. Camara, G. Sikka, H. K. Lim, J. Sohi, L. Santhanam, K. Soucy, et al. Endothelial Arginase II: A Novel Target for the Treatment of Atherosclerosis Circ. Res., April 25, 2008; 102(8): 923 - 932. [Abstract] [Full Text] [PDF] |
||||
![]() |
Z. S. Katusic Mechanisms of Endothelial Dysfunction Induced by Aging: Role of Arginase I Circ. Res., September 28, 2007; 101(7): 640 - 641. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |