cd/ wun (lw2) 15, 427-434 QLmgmrnC3roy,UKWd13!32
Ca*+/calmodulin-regulated nitric oxide synthases H.H.H.W. SCHMIDT’*, J.S. POLLOCK*, M. NAKANE*, U. FC)RSTERMANN’** and F. MURAD’12 ‘Northwestern UniversityMedical School, Chkago, USA *Abbott Laboratories,Abbott Park, USA
Abstract- NO synthase (NOS) catalyzes the oxidation of Larginine to L-citrullineand nitric oxide (NO) or a NO-releasingcompound. At least three fsoformsof NOS exist (types i-iii). The activities of the type I isoform purified from brain and the type iii isoform purified from endotheliai tils are regulated by the intracellular free calcium concentration @a*+$) and the Ca*‘-binding protein caimoduiin. At resting [Ca2yi, both isozymes are inactive; they become fully active at [Ca*+li 2 500 nM Ca*+. Longer lastly increases in [Ca*+j may downregulate NO formation, for in vitro phosphoryiationby Ca Mmoduiin protein kinase Ii decreases the VIII, of NOS. Besides the conversion of L-arginine, type I NOS, Ca2t/caimodulindependently,generates He and reduces cytochromec/Pa. Other redox activities, i.e. the reduction of nitrobluetetrazoiiumto diformazan (NADPHTrase) or of quinoiddihydrobiopterin to tetrahydrobiopterin, by NOS appear to be Ca %aimodullnindependent
Nitric oxide or a nitric oxide-releasing compound (NO) is an endogenous paracrine substance and intracellular messenger [l, 21. In the majority of cases, NO acts by activating the guanosine cyclic S’S’-monophosphate (cGMP)-forming enzyme, soluble guanylyl cyclase (GTP pyrophosphate-lyase (cyclizing), EC 4.6.1.2) 13, 41. The second messenger formed, cGh4P. regulates various protein kinases, nucleoside 3’,5’-monophosphate phosphodiesterases, and ion channels [5].
Ca2+-dependent NO synthesis In mammalian cells and tissues, NO is synthesized from &uginiue [6-91 by NO synthases (NOS. EC *Pltacnt address, Universitat Wclnburg. Gemany
1.14.23) [lo]. NOS activity can be assayed by de&mining either the formation of NO or the co-product L-citrullhre [lo]. To detetmine NO, a bioassay based on the activation of soluble guanylyl cyclase and subsequent increases in cGMP levels in RPL-6 detector cells [l 11,or a chemical assay to determine the oxidation products of NO, nitrite and nitrate [12-151, have been employed. All NOSs tequire L-arginhre, 02, NADPH, FAD. PMN and tetrahydrobiopterin W-19, 23, 241 and all are inhibited by @-n&o- and NO-methyl-L-arginhre r20.211. Three NOSs have been purified [lo, 22261, characterized [lo, 22-301 and molecularly cloned [31-331. The type I NOS. which is localized predominantly in the cerebellum 1341,pancreas 1351, and specialized epithelial cells [361in kidney, lung, stomach and uterus, and the type III isoform, which
427
C.C.E
428
m
100 A
s g 8 3 se
50 0
10
100 (trifluopaszine), IlM
4.0
> 10
~50 (G?‘), IM EQo (Mg+) nodep.-nodqx&nca
135
0.35
0.3
uo dep.
no dep.
The dialyzed pool from the 2’,5’-ADP Sepharose caabeappliedtoa calmod& agarose af3aity colnmninthepreseaceof2mMCa&aadNOStype I caa then be eluted with 5 mM BGTA aad 10% glycerol in a single sharp proteia peak (Fig. 3) coataiaiag a single protein staiaiag band by SDS/PAGE analysis (Fig. 4). For the binding of NOS to calmodulin agarose, the optimal ratio between calmoduliu agarose aad non&belled Sepharose 4B is 1:8 (v/v, equivalent to 150 p.g calmodulia/ml packed gel; both Sigma) resulting ia a fiaal concentration of about 8.6 am01 calmodulia/ml packed gel. This allows the preparation of a high afMty column aad specific elutioa conditions (5 mM EGTA, 10% glycerol) 1231. Lower concentrations of EGTA, omitting glycerol or choosing a higher calmodulin/ ml gel conceatratioa leads to no, or only partial, elutioa of NOS, substantial peak-broadening, or required chaotropic elutioa conditions to recover NOS. After SDS/PAGE, semi+ transfer to a&ocellulose membrane, incubation with biotinylated cahnodulia and aa avidia alkaline phosph&ase coajugate 1431,NOS Ca2’-depeadently bound biotinylated calmodulia. These results identified NOS as a calmodulia-biadiag protein [23]. The induced type II NOS. which in its ~+.n%ied form is not regulated by calmoduliu or Ca , coataias consensus sites for calmodulia binding [31, 331. It is, therefore, possible that type II NOS may post-translationally bind calmodulin. This pmcess may be irreversible sad lead to aa irreversible activation of the enzyme. Type II NOS squires immunological activation of cells in order to be expressed. IatenHiagly, this induction process can be partially blocked by the intracellular Ca2’ inhibitor TMB-8 [ 141.
430
CELL CALCIUM
A
F 25 F E 2 3 I!
P
212170116-
0.4 94;+
0.6
53-
0.8
43-
Dye Front -
1
2
4.6 4.8 5.0 5.2 5.4 log (molecular mass)
Fig. 4 SDS/PAGE of rat brain NO synthase befxv and after afhity clunatogmphy on cabdulin aganxe. Lane 1, column load (NADPH-eluate of a 2’,5’-ADP Sephawe): lane 2, EGTA-eluate of the calmodulin agamse column. Pane.1B, the appamnt mokulat mass of the single protein band in the purified soluble NO synthase preparationis designated by the dotted line and is 155 f 3 kD. With pemxissionfrom [23]
NOS is a Ca2t/calmodalin-dependent NADPH oxidase
Vma that is 10-100 times higher than that for oxygen:L-arginineturnover[29]. The mechanism probablydoes not involve superoxideradicalformIndependent of the presence of r_+#ine, NOS type ationbecausecytochro~ c reductionis not preventI oxidizes NADPH which is accompaniedby a ed by superoxidedismutase. It is probablydue to a interaction betcahnociulin-dependent uptake of molecular oxygen direct Ca2+kahnodnlin-dependent [28] (Fig. 5). Underthese conditions,formationof ween NOS and cytochron~c. NOS binds with high hydrogen peroxide is observed. This Hz02 prod- affinity to a cytochromec column and can only be uction is reduced in the presence of L-arginine and eluted under chaotropic conditions [291. When abolished in the presence of tetrahydrobiopterin reconstitutedwith cytocbromeP450.NOS induces a [30]. Thus, NOS type I can reduce molecular moderate Ca2’-independent hydroxylation of oxygen to hydrogen peroxide in a similar way to the N-ethyhnorphine t29]. These results suggested that NOS participates in cellular electron transfer proNADPH oxidase of neutrophils. cesses 1291. Moreover, it was suggested that NOS is also a tetrahydmbiopterin reductase and NADPH diaphdrNOS is a Cazt/calmodulin-dependentcytochrome ase and that these activities correspond to reductase sites of NOS which are independent of Ca2’/ l%dUctase cahnodulin and different from the cytochrome c/ NOS also reduces cytuchrome c (Fig. 5) with a P450reductase site 1271.
CaZ’/CAIMODULIN-REGULATED NITRIC OXIDE SYNTHASES
431
2.0
k
1.5
B 6
1.0
!i t
0.6
0.0
0.t
,
0
2
a
1
0
2
8
1
n&w6
Fig. 5 Cahnodulin-dependent NADPH- and molecular oxygen-consumption, He formation and qtochmc c mduuion by NO “. synthaae. Purified brain NO synthasc was incubated in the presence of NADPH and ca” and in the abacnce (filled cimlcs) or pn#ace (open circles) of cahwduk (A, B. D). For details, see [22,29,30]. The diffemnt psnels s&v: (A) NADPH co~~osuraption as monitotwl by the dcueasc in abwrbawe et 340 am; (B) oxygen consumption as detecmined with an oxygen-sensitive ekctmdq (C) EI$)z formation as de&mhd by fenic thiocyanate formation (absorbance at 492 mn); and (D) cytochrome c (0.2 mM) &u&n as determiaed by the incnwc in absorbance at 550 nm. With pemkion from [22,29,30]
Caz~/caImod~iinkinase II and NO synthase
pboryltion
Recently, Bredt et al. obtained cDNA clones of NOS type I fromratbrainaad foundpossiblephos-
In vitro a?+/msequeace WI. dependent protein kinase II dhosphoryla&sNOS Phosphorylation is Ca*- and 0%. 6) WI.
sites based on the predkkd amino acid
432
cELLcAu3IuM
60
5 +
102030405060 + + + + +
+
Time (min) Ca-CaM Kinase II
4 P CaM Kinase II 4a
Fig 6 Phoephmylation of type I NO qntbaae from rat brain by cabnodulin protein kinase II. Purifkd brain NO synthase was phospkylated by CaM kinase II for the indicated periods of time. As a contml, NO synthase @KS) was incubated for 60 min without CaCl&x&nodulin(with 1 mM EGl’A). With permissionfrom [44]
calmodulin-dependent and reaches a maximum at a final stoichiometry of about 9 moles of phosphate per mole of NOS monomef (Fig. 7) after 60 min. NOS is phosphorylated on both serine and thmonine, but not tyrosine. residues. Phosphorylation of NOS by Ca2t/calmodulin kinase II results in a marked decrease of activity to 33% of control, whereas the cofactor requirements do not change significantly. Both phosphorylated and nonphosphorylated NOS are totally dependent on NADPH, Ca2’ and calmodulin for activity [44]. ‘Ihe amino acid uences surrounding phosphor- ylation sites 5? /calmodulin kinase II have been determined for Ca to be Arg-X-X-SerR’hr. In the amino acid sequence of NOS predicted from cDNA [32], there are 8 possible phosphorylation sites for serine and 4 possible sites for threonine by Ca2’/ calmodulin kinase II. This is comparable to our stoichiometrical result of about 9 moles of phosphate per mole of NO synthase subunit. The number of the phosphorylated threonine sites (about 5 moles phosphate per mole of NO synthase subunit, Fig. 2) is more than the predicted value (4 moles per mole NO synthase subunit). This may suggest that Ca2t/cahnodulin kinase II also phosphorylates other sequences. It has been shown that Ca2t/calmodulin kinase II can phosphorylate the sequence of Lys-X-X-Ser/Thr, although the apparent Km was
Time
(min)
c
5
3
Fig. 7 PhosphD+tion of type I NO syntbase from rat brain by calmodulin protein kinaae II end effect on enxym activity. (A) The incorporation of ‘% was quantitkd by excising the NO synthase band from the gel and liquid scintillation spectrophotometry. (B) NO synthase activity was meaaued with and without priorphosp~latiou reaction with calmcduliu kinase II. With permissionfrom [44]
lo-fold higher than that for the sequence Arg-X-X-Ser/Thr. As mentioned above, brain NOS
Caz+/CALMODULIN-REGULATED NlTRIC OXIDE SYNTHASES
is highly dependent on free Ca2’ at physiologically relevant concentrations and shows an absolute requirement for calmodulin. Ca2+/calmodulinkinase lI is also Ca2t/calmodulin dependent and thought to be activated by calcium influx into the cells. Ca2+/calmodulinkinase II is present in sufficient amounts in brain tissue. Thus, an increase in intracellular Ca2+concentrations may not only stimulate NOS activity, but, at the same time, activate Ca2t/calmodulin kinase II. Phospholylation of NOS by this kinase could represent a negative feedback mechanism for regulation of NOS activity. The activity of Ca2t/calmodulin kinase II can become Ca2+-independent after autophosphorylation of the enzyme. Therefore, it is conceivable that activated NOS may be phosphorylated and inactivated by the autophosphorylated Ca2+-independent Ca2+/ calmodulin kinase II even after cytosolic free Ca2+ concentrations have returned to their basal levels. Future experiments will have to determine to what extent these mechanisms contribute to NOS synthesis in intact cells.
433
3. Arnold WP. Mittrl C!K- Katsuki S. Murad F. (1977) Pmt. Natl. AC&. Sci. USA, 74.32033207.
4. B&me E. Graf H.!khu11~ 0. (1978) Adv. Cyclic Nuckotide Rcs.. 9,131.143.
5. Walter U. (1989) Rev. Physiol. Biochem. Pharmacol.. 113. 41.88. 6. Iyengar R. Stuehr DJ. Marletta MA. (1987) Pmt. Natl. Acad. Sci. USA, 84.6369.6373. 7. Palmer RMJ. Ashton DS. Mcncada S. (1988) Natuse, 333, 664.666. 8. Schmidt HHHW. Nau H. Witffbbt W. et al. (1988) Eur. J. Pbarmacol., 154,213.216. 9. Schmidt HHHW. Klein MM. Nimomand F. B&me E. (1988) E!ur.J. Pbarmacol.. 148.293-295. 10. Bredt DS. Snyder SS. (1990) Pmt. Natl. Acad Sci. USA, 87,682.685. 11. Ishii K. Sheng H. Warner TD. F&atemuam U. Mmad F. (1991) Am. J. Physiol., 261, H598-H603. 12. Schmidt HHHW. Seifert R B&me E. (1989) FBBS Lat., 244.357-360. 13. Schmidt HHHW. Wilke P. Even B. B&hme E. (1989) Bicchem. Biophys. Res. Commun., 165.284-291. 14. Schmidt HHHW. Warner TD. N&me M. F&stermann U. Murad F. (1992) Mol. Pharmacol., 41, 615.624. 15. Stuehr DJ. Gmss SS. Sakuma I. Levi R Nathan CF. (1989) J. Exp. Med., 169,lOll.1020. 16. Marletta MA. Yoon PS. Iyengar R. Leaf’CD. Wisbnok JS. (1988) Biccbemistry, 27.8706-8711. 17. Kwon NS. Nathan CF. Stuehr DJ. (1989) J. Biol. Chem., 264,204%-20501.
Acknowledgements We thank Dr B. Mayer, Dr E. B&me and their coworkers for helpful discussionand permissionto include some of their data in this review. T. Yamauchi (Tokyo MetropolitanInstitutefor Neurosciences, Tokyo, Japan) and T. Kuno (Kobe University School of Medicine, Kobe, Japan) for their kind gifts of protein kinases. We also thank S. Dorwin for her assistance with phosphoamino acid analysis. The skillful technical assistance of Z-J. Huang, KL. Kohlhaas and J. Kulc is gratefully appreciated. This work was supported by research grants DK 30787 and HL 28474 from the National Institutes of Health. HHHWS received a postdoctoral fellowship from the Deutsche Forschungsgemeinschaft (Germany); JSP a postdoctoral award (AR 08080) from the National Institutes of Health (USA); and UF a scholarship from the Heisenberg Foundation (Germany).
References 1. Palmer RMJ. Ferrige AG. Moncada S. (1987) Nature, 327, 524.526. 2. Ignam, LJ. Buga GM. Wood KS. Byms BB. ChaudhuriG. (1987) Proc. Natl. Acad Sci. USA, 84,9265X269.
18. Stuehr DJ. Kwon NS. Nathan CF. (1990) Biochem. Biophys. Res. Common,, 168.558-565. 19. Tayeh MA. Marietta MA. (1989) J. Biol. Cben~, 264, 19654.19658. 20. Hibbs Jr JB. Taintor RR. Vavrin Z. Racblin EM (1988) B&hem. Biophys. Res. Commun., 157.87-94. 21. l&ii K. chang B. Kenvin Jr JF. Huang ZJ. Mumd F. (1990) Eur. J. Pharmacol., 176,219.223. 22. Mayer B. John M. B&me E. (1990) FEBS L&t., 277, 215.219. 23. Schmidt HHHW.Pollock JS. Nakane M. Gorsky LD. F6rstermam1 U. Mumd F. (1991) Prw. Natl. Acad. Sci. USA, 88.365-369. 24. Pollock JS. F&sterma~ U. Mitchell JA. et al. (1991) Pmt. Natl. Acad. Sci. USA, 88.10480.10484. 25. Hevel JM. White KA. Marietta MA. (1991) J. Biol. Cbem.. 266,22789-22791. 26. Stuehr DJ. Cbo HJ. Kwon NS. Nathan CF. (1991) Prw. Natl. Acad. Sci. USA, 88,7773-7777. 27. Schmidt HHHW. Smith RM. Nakane M. Mumd F. (1992) Biochemistty, 31.3243-3249. 28. Mayer B. John M. Heinzel B. et al. (1991) FEBS Len., 288, 187.191. 29. Klatt P. He&e1 B. Jolm M. Kastner M. B6hme E. Mayer B. (1992) J. Biol. Cbem., 267,11374-11378. 30. Heinzel B. John M. Klatt P. B&me E. Mayer B. (1992) Biochem J., 281,627.630.
31. Lyons CR. Orloff GJ. Cunningham JM. (1992) J. Biol. Chem., 267.6370-6374.
32. Bredt DS. Hwang PM. Glatt CB. Lowenstein C. Reed RR.
434 Snyder SH. (1991) Nature, 351,716718. 33. Xie Q-W. C&o HJ. Calaycay J. et al. (1992) Science, 256, 225-228. 34. Bredt DS. Hwang PM. Snyder SK (1990) Nature.,347, 768-770. 35. Schmidt HHHW. Warner TD. Jshii IL Sheng H. Mumd F. (1992) science, 255.721-723. 36. Schmidt HHHW. Gap GD. Nakane M. Pollock JS. Miller MF. Mumd F. (195’2) J. Cytocbem. Histochem., Submitted. 37. F6rstermann U. Schmidt HHHW. Pollock JS. et al. (1991) B&hem. Pharmacol., 42,1849-1857. 38. F6mtermano U. Gorsky LD. Pollock JS. et al. (1990) Eur. J. Phamwol.. 183, 1625-1626. 39. Knowles RG. Palacios M. Palmer RMJ. Moncada S. (1989) FVoc. Natl. Acad. Sci. USA, 86.5159-5162.
40. Mayer B. B6hme E. (1989) FBBS Lctt., 256.211-214. 41. F6mtennano U. polloclr JS. Schmidt HHHW. Helkr M. Mumd F. (1991) Proc. Nat]. Acad. Sci. USA. 88.1788-1792. 42. Fatatem~annU. Schmidt HHHS. Pollcck JS. Heller M. Murad F. (1991) J. Cardiovasc. Pharmacol., 17 (Suppl3). s57s64. 43. Billing&y ML. I%xmypackerKR. Hoover CG. Brigati DJ. Kincaid RL. (1985) Proc. NaU. Acad. Sci. USA, 82, 7585-7589. 44. NalcaneM. Mitchell J. F&ste.nnannU. Mumd F. (1991) Biochem. Biophys. Res. Commun., 180,1396-1402. Please send reprint mquests to : Dr Hamld H.H.W. Schmidt, Klinis& Fom&qguppe, Univemit&t WtIrzburg, JosefSchneider& 2, W-8700 WUmburg, Getmany
Ca*+/calmodulin-regulated nitric oxide synthases H.H.H.W. SCHMIDT’*, J.S. POLLOCK*, M. NAKANE*, U. FC)RSTERMANN’** and F. MURAD’12 ‘Northwestern UniversityMedical School, Chkago, USA *Abbott Laboratories,Abbott Park, USA
Abstract- NO synthase (NOS) catalyzes the oxidation of Larginine to L-citrullineand nitric oxide (NO) or a NO-releasingcompound. At least three fsoformsof NOS exist (types i-iii). The activities of the type I isoform purified from brain and the type iii isoform purified from endotheliai tils are regulated by the intracellular free calcium concentration @a*+$) and the Ca*‘-binding protein caimoduiin. At resting [Ca2yi, both isozymes are inactive; they become fully active at [Ca*+li 2 500 nM Ca*+. Longer lastly increases in [Ca*+j may downregulate NO formation, for in vitro phosphoryiationby Ca Mmoduiin protein kinase Ii decreases the VIII, of NOS. Besides the conversion of L-arginine, type I NOS, Ca2t/caimodulindependently,generates He and reduces cytochromec/Pa. Other redox activities, i.e. the reduction of nitrobluetetrazoiiumto diformazan (NADPHTrase) or of quinoiddihydrobiopterin to tetrahydrobiopterin, by NOS appear to be Ca %aimodullnindependent
Nitric oxide or a nitric oxide-releasing compound (NO) is an endogenous paracrine substance and intracellular messenger [l, 21. In the majority of cases, NO acts by activating the guanosine cyclic S’S’-monophosphate (cGMP)-forming enzyme, soluble guanylyl cyclase (GTP pyrophosphate-lyase (cyclizing), EC 4.6.1.2) 13, 41. The second messenger formed, cGh4P. regulates various protein kinases, nucleoside 3’,5’-monophosphate phosphodiesterases, and ion channels [5].
Ca2+-dependent NO synthesis In mammalian cells and tissues, NO is synthesized from &uginiue [6-91 by NO synthases (NOS. EC *Pltacnt address, Universitat Wclnburg. Gemany
1.14.23) [lo]. NOS activity can be assayed by de&mining either the formation of NO or the co-product L-citrullhre [lo]. To detetmine NO, a bioassay based on the activation of soluble guanylyl cyclase and subsequent increases in cGMP levels in RPL-6 detector cells [l 11,or a chemical assay to determine the oxidation products of NO, nitrite and nitrate [12-151, have been employed. All NOSs tequire L-arginhre, 02, NADPH, FAD. PMN and tetrahydrobiopterin W-19, 23, 241 and all are inhibited by @-n&o- and NO-methyl-L-arginhre r20.211. Three NOSs have been purified [lo, 22261, characterized [lo, 22-301 and molecularly cloned [31-331. The type I NOS. which is localized predominantly in the cerebellum 1341,pancreas 1351, and specialized epithelial cells [361in kidney, lung, stomach and uterus, and the type III isoform, which
427
C.C.E
428
m
100 A
s g 8 3 se
50 0
10
100 (trifluopaszine), IlM
4.0
> 10
~50 (G?‘), IM EQo (Mg+) nodep.-nodqx&nca
135
0.35
0.3
uo dep.
no dep.
The dialyzed pool from the 2’,5’-ADP Sepharose caabeappliedtoa calmod& agarose af3aity colnmninthepreseaceof2mMCa&aadNOStype I caa then be eluted with 5 mM BGTA aad 10% glycerol in a single sharp proteia peak (Fig. 3) coataiaiag a single protein staiaiag band by SDS/PAGE analysis (Fig. 4). For the binding of NOS to calmodulin agarose, the optimal ratio between calmoduliu agarose aad non&belled Sepharose 4B is 1:8 (v/v, equivalent to 150 p.g calmodulia/ml packed gel; both Sigma) resulting ia a fiaal concentration of about 8.6 am01 calmodulia/ml packed gel. This allows the preparation of a high afMty column aad specific elutioa conditions (5 mM EGTA, 10% glycerol) 1231. Lower concentrations of EGTA, omitting glycerol or choosing a higher calmodulin/ ml gel conceatratioa leads to no, or only partial, elutioa of NOS, substantial peak-broadening, or required chaotropic elutioa conditions to recover NOS. After SDS/PAGE, semi+ transfer to a&ocellulose membrane, incubation with biotinylated cahnodulia and aa avidia alkaline phosph&ase coajugate 1431,NOS Ca2’-depeadently bound biotinylated calmodulia. These results identified NOS as a calmodulia-biadiag protein [23]. The induced type II NOS. which in its ~+.n%ied form is not regulated by calmoduliu or Ca , coataias consensus sites for calmodulia binding [31, 331. It is, therefore, possible that type II NOS may post-translationally bind calmodulin. This pmcess may be irreversible sad lead to aa irreversible activation of the enzyme. Type II NOS squires immunological activation of cells in order to be expressed. IatenHiagly, this induction process can be partially blocked by the intracellular Ca2’ inhibitor TMB-8 [ 141.
430
CELL CALCIUM
A
F 25 F E 2 3 I!
P
212170116-
0.4 94;+
0.6
53-
0.8
43-
Dye Front -
1
2
4.6 4.8 5.0 5.2 5.4 log (molecular mass)
Fig. 4 SDS/PAGE of rat brain NO synthase befxv and after afhity clunatogmphy on cabdulin aganxe. Lane 1, column load (NADPH-eluate of a 2’,5’-ADP Sephawe): lane 2, EGTA-eluate of the calmodulin agamse column. Pane.1B, the appamnt mokulat mass of the single protein band in the purified soluble NO synthase preparationis designated by the dotted line and is 155 f 3 kD. With pemxissionfrom [23]
NOS is a Ca2t/calmodalin-dependent NADPH oxidase
Vma that is 10-100 times higher than that for oxygen:L-arginineturnover[29]. The mechanism probablydoes not involve superoxideradicalformIndependent of the presence of r_+#ine, NOS type ationbecausecytochro~ c reductionis not preventI oxidizes NADPH which is accompaniedby a ed by superoxidedismutase. It is probablydue to a interaction betcahnociulin-dependent uptake of molecular oxygen direct Ca2+kahnodnlin-dependent [28] (Fig. 5). Underthese conditions,formationof ween NOS and cytochron~c. NOS binds with high hydrogen peroxide is observed. This Hz02 prod- affinity to a cytochromec column and can only be uction is reduced in the presence of L-arginine and eluted under chaotropic conditions [291. When abolished in the presence of tetrahydrobiopterin reconstitutedwith cytocbromeP450.NOS induces a [30]. Thus, NOS type I can reduce molecular moderate Ca2’-independent hydroxylation of oxygen to hydrogen peroxide in a similar way to the N-ethyhnorphine t29]. These results suggested that NOS participates in cellular electron transfer proNADPH oxidase of neutrophils. cesses 1291. Moreover, it was suggested that NOS is also a tetrahydmbiopterin reductase and NADPH diaphdrNOS is a Cazt/calmodulin-dependentcytochrome ase and that these activities correspond to reductase sites of NOS which are independent of Ca2’/ l%dUctase cahnodulin and different from the cytochrome c/ NOS also reduces cytuchrome c (Fig. 5) with a P450reductase site 1271.
CaZ’/CAIMODULIN-REGULATED NITRIC OXIDE SYNTHASES
431
2.0
k
1.5
B 6
1.0
!i t
0.6
0.0
0.t
,
0
2
a
1
0
2
8
1
n&w6
Fig. 5 Cahnodulin-dependent NADPH- and molecular oxygen-consumption, He formation and qtochmc c mduuion by NO “. synthaae. Purified brain NO synthasc was incubated in the presence of NADPH and ca” and in the abacnce (filled cimlcs) or pn#ace (open circles) of cahwduk (A, B. D). For details, see [22,29,30]. The diffemnt psnels s&v: (A) NADPH co~~osuraption as monitotwl by the dcueasc in abwrbawe et 340 am; (B) oxygen consumption as detecmined with an oxygen-sensitive ekctmdq (C) EI$)z formation as de&mhd by fenic thiocyanate formation (absorbance at 492 mn); and (D) cytochrome c (0.2 mM) &u&n as determiaed by the incnwc in absorbance at 550 nm. With pemkion from [22,29,30]
Caz~/caImod~iinkinase II and NO synthase
pboryltion
Recently, Bredt et al. obtained cDNA clones of NOS type I fromratbrainaad foundpossiblephos-
In vitro a?+/msequeace WI. dependent protein kinase II dhosphoryla&sNOS Phosphorylation is Ca*- and 0%. 6) WI.
sites based on the predkkd amino acid
432
cELLcAu3IuM
60
5 +
102030405060 + + + + +
+
Time (min) Ca-CaM Kinase II
4 P CaM Kinase II 4a
Fig 6 Phoephmylation of type I NO qntbaae from rat brain by cabnodulin protein kinase II. Purifkd brain NO synthase was phospkylated by CaM kinase II for the indicated periods of time. As a contml, NO synthase @KS) was incubated for 60 min without CaCl&x&nodulin(with 1 mM EGl’A). With permissionfrom [44]
calmodulin-dependent and reaches a maximum at a final stoichiometry of about 9 moles of phosphate per mole of NOS monomef (Fig. 7) after 60 min. NOS is phosphorylated on both serine and thmonine, but not tyrosine. residues. Phosphorylation of NOS by Ca2t/calmodulin kinase II results in a marked decrease of activity to 33% of control, whereas the cofactor requirements do not change significantly. Both phosphorylated and nonphosphorylated NOS are totally dependent on NADPH, Ca2’ and calmodulin for activity [44]. ‘Ihe amino acid uences surrounding phosphor- ylation sites 5? /calmodulin kinase II have been determined for Ca to be Arg-X-X-SerR’hr. In the amino acid sequence of NOS predicted from cDNA [32], there are 8 possible phosphorylation sites for serine and 4 possible sites for threonine by Ca2’/ calmodulin kinase II. This is comparable to our stoichiometrical result of about 9 moles of phosphate per mole of NO synthase subunit. The number of the phosphorylated threonine sites (about 5 moles phosphate per mole of NO synthase subunit, Fig. 2) is more than the predicted value (4 moles per mole NO synthase subunit). This may suggest that Ca2t/cahnodulin kinase II also phosphorylates other sequences. It has been shown that Ca2t/calmodulin kinase II can phosphorylate the sequence of Lys-X-X-Ser/Thr, although the apparent Km was
Time
(min)
c
5
3
Fig. 7 PhosphD+tion of type I NO syntbase from rat brain by calmodulin protein kinaae II end effect on enxym activity. (A) The incorporation of ‘% was quantitkd by excising the NO synthase band from the gel and liquid scintillation spectrophotometry. (B) NO synthase activity was meaaued with and without priorphosp~latiou reaction with calmcduliu kinase II. With permissionfrom [44]
lo-fold higher than that for the sequence Arg-X-X-Ser/Thr. As mentioned above, brain NOS
Caz+/CALMODULIN-REGULATED NlTRIC OXIDE SYNTHASES
is highly dependent on free Ca2’ at physiologically relevant concentrations and shows an absolute requirement for calmodulin. Ca2+/calmodulinkinase lI is also Ca2t/calmodulin dependent and thought to be activated by calcium influx into the cells. Ca2+/calmodulinkinase II is present in sufficient amounts in brain tissue. Thus, an increase in intracellular Ca2+concentrations may not only stimulate NOS activity, but, at the same time, activate Ca2t/calmodulin kinase II. Phospholylation of NOS by this kinase could represent a negative feedback mechanism for regulation of NOS activity. The activity of Ca2t/calmodulin kinase II can become Ca2+-independent after autophosphorylation of the enzyme. Therefore, it is conceivable that activated NOS may be phosphorylated and inactivated by the autophosphorylated Ca2+-independent Ca2+/ calmodulin kinase II even after cytosolic free Ca2+ concentrations have returned to their basal levels. Future experiments will have to determine to what extent these mechanisms contribute to NOS synthesis in intact cells.
433
3. Arnold WP. Mittrl C!K- Katsuki S. Murad F. (1977) Pmt. Natl. AC&. Sci. USA, 74.32033207.
4. B&me E. Graf H.!khu11~ 0. (1978) Adv. Cyclic Nuckotide Rcs.. 9,131.143.
5. Walter U. (1989) Rev. Physiol. Biochem. Pharmacol.. 113. 41.88. 6. Iyengar R. Stuehr DJ. Marletta MA. (1987) Pmt. Natl. Acad. Sci. USA, 84.6369.6373. 7. Palmer RMJ. Ashton DS. Mcncada S. (1988) Natuse, 333, 664.666. 8. Schmidt HHHW. Nau H. Witffbbt W. et al. (1988) Eur. J. Pbarmacol., 154,213.216. 9. Schmidt HHHW. Klein MM. Nimomand F. B&me E. (1988) E!ur.J. Pbarmacol.. 148.293-295. 10. Bredt DS. Snyder SS. (1990) Pmt. Natl. Acad Sci. USA, 87,682.685. 11. Ishii K. Sheng H. Warner TD. F&atemuam U. Mmad F. (1991) Am. J. Physiol., 261, H598-H603. 12. Schmidt HHHW. Seifert R B&me E. (1989) FBBS Lat., 244.357-360. 13. Schmidt HHHW. Wilke P. Even B. B&hme E. (1989) Bicchem. Biophys. Res. Commun., 165.284-291. 14. Schmidt HHHW. Warner TD. N&me M. F&stermann U. Murad F. (1992) Mol. Pharmacol., 41, 615.624. 15. Stuehr DJ. Gmss SS. Sakuma I. Levi R Nathan CF. (1989) J. Exp. Med., 169,lOll.1020. 16. Marletta MA. Yoon PS. Iyengar R. Leaf’CD. Wisbnok JS. (1988) Biccbemistry, 27.8706-8711. 17. Kwon NS. Nathan CF. Stuehr DJ. (1989) J. Biol. Chem., 264,204%-20501.
Acknowledgements We thank Dr B. Mayer, Dr E. B&me and their coworkers for helpful discussionand permissionto include some of their data in this review. T. Yamauchi (Tokyo MetropolitanInstitutefor Neurosciences, Tokyo, Japan) and T. Kuno (Kobe University School of Medicine, Kobe, Japan) for their kind gifts of protein kinases. We also thank S. Dorwin for her assistance with phosphoamino acid analysis. The skillful technical assistance of Z-J. Huang, KL. Kohlhaas and J. Kulc is gratefully appreciated. This work was supported by research grants DK 30787 and HL 28474 from the National Institutes of Health. HHHWS received a postdoctoral fellowship from the Deutsche Forschungsgemeinschaft (Germany); JSP a postdoctoral award (AR 08080) from the National Institutes of Health (USA); and UF a scholarship from the Heisenberg Foundation (Germany).
References 1. Palmer RMJ. Ferrige AG. Moncada S. (1987) Nature, 327, 524.526. 2. Ignam, LJ. Buga GM. Wood KS. Byms BB. ChaudhuriG. (1987) Proc. Natl. Acad Sci. USA, 84,9265X269.
18. Stuehr DJ. Kwon NS. Nathan CF. (1990) Biochem. Biophys. Res. Common,, 168.558-565. 19. Tayeh MA. Marietta MA. (1989) J. Biol. Cben~, 264, 19654.19658. 20. Hibbs Jr JB. Taintor RR. Vavrin Z. Racblin EM (1988) B&hem. Biophys. Res. Commun., 157.87-94. 21. l&ii K. chang B. Kenvin Jr JF. Huang ZJ. Mumd F. (1990) Eur. J. Pharmacol., 176,219.223. 22. Mayer B. John M. B&me E. (1990) FEBS L&t., 277, 215.219. 23. Schmidt HHHW.Pollock JS. Nakane M. Gorsky LD. F6rstermam1 U. Mumd F. (1991) Prw. Natl. Acad. Sci. USA, 88.365-369. 24. Pollock JS. F&sterma~ U. Mitchell JA. et al. (1991) Pmt. Natl. Acad. Sci. USA, 88.10480.10484. 25. Hevel JM. White KA. Marietta MA. (1991) J. Biol. Cbem.. 266,22789-22791. 26. Stuehr DJ. Cbo HJ. Kwon NS. Nathan CF. (1991) Prw. Natl. Acad. Sci. USA, 88,7773-7777. 27. Schmidt HHHW. Smith RM. Nakane M. Mumd F. (1992) Biochemistty, 31.3243-3249. 28. Mayer B. John M. Heinzel B. et al. (1991) FEBS Len., 288, 187.191. 29. Klatt P. He&e1 B. Jolm M. Kastner M. B6hme E. Mayer B. (1992) J. Biol. Cbem., 267,11374-11378. 30. Heinzel B. John M. Klatt P. B&me E. Mayer B. (1992) Biochem J., 281,627.630.
31. Lyons CR. Orloff GJ. Cunningham JM. (1992) J. Biol. Chem., 267.6370-6374.
32. Bredt DS. Hwang PM. Glatt CB. Lowenstein C. Reed RR.
434 Snyder SH. (1991) Nature, 351,716718. 33. Xie Q-W. C&o HJ. Calaycay J. et al. (1992) Science, 256, 225-228. 34. Bredt DS. Hwang PM. Snyder SK (1990) Nature.,347, 768-770. 35. Schmidt HHHW. Warner TD. Jshii IL Sheng H. Mumd F. (1992) science, 255.721-723. 36. Schmidt HHHW. Gap GD. Nakane M. Pollock JS. Miller MF. Mumd F. (195’2) J. Cytocbem. Histochem., Submitted. 37. F6rstermann U. Schmidt HHHW. Pollock JS. et al. (1991) B&hem. Pharmacol., 42,1849-1857. 38. F6mtermano U. Gorsky LD. Pollock JS. et al. (1990) Eur. J. Phamwol.. 183, 1625-1626. 39. Knowles RG. Palacios M. Palmer RMJ. Moncada S. (1989) FVoc. Natl. Acad. Sci. USA, 86.5159-5162.
40. Mayer B. B6hme E. (1989) FBBS Lctt., 256.211-214. 41. F6mtennano U. polloclr JS. Schmidt HHHW. Helkr M. Mumd F. (1991) Proc. Nat]. Acad. Sci. USA. 88.1788-1792. 42. Fatatem~annU. Schmidt HHHS. Pollcck JS. Heller M. Murad F. (1991) J. Cardiovasc. Pharmacol., 17 (Suppl3). s57s64. 43. Billing&y ML. I%xmypackerKR. Hoover CG. Brigati DJ. Kincaid RL. (1985) Proc. NaU. Acad. Sci. USA, 82, 7585-7589. 44. NalcaneM. Mitchell J. F&ste.nnannU. Mumd F. (1991) Biochem. Biophys. Res. Commun., 180,1396-1402. Please send reprint mquests to : Dr Hamld H.H.W. Schmidt, Klinis& Fom&qguppe, Univemit&t WtIrzburg, JosefSchneider& 2, W-8700 WUmburg, Getmany
