Báo cáo Y học: Substrate selectivity and sensitivity to inhibition by FK506 and cyclosporin A of calcineurin heterodimers composed of the a or b catalytic subunit potx

Kinetic assays of CaNa and CaNb phosphatase activity Kinetic analyses of the in vitro phosphatase activities of CaNa and CaNb were carried out to determine their Km and Vmaxvalues toward

Substrate selectivity and sensitivity to inhibition by FK506

and cyclosporin A of calcineurin heterodimers composed

of the a or b catalytic subunit

Brian A Perrino1, Andrew J Wilson2, Patricia Ellison3and Lucie H Clapp2

1

Department of Physiology & Cell Biology, University of Nevada School of Medicine, Reno, NV, USA;2Center for Clinical Pharmacology, University College London, UK; 3 Department of Biochemistry, University of Nevada School of Medicine,

Reno, NV, USA

The calcineurin (CaN) a and b catalytic subunit isoforms are

coexpressed within almost all cell types The enzymatic

properties of CaN heterodimers comprised of the regulatory

Bsubunit (CnB) with either the a or b catalytic subunit were

compared using in vitro phosphatase assays CaN containing

the a isoform (CnAa) has lower Kmand higher Vmaxvalues

than CaN containing the b isoform (CnAb) toward the PO4

-RII, PO4-DARPP-32(20–38) peptides, and

p-nitrophenyl-phosphate (pNPP) CaN heterodimers containing the a or b

catalytic subunit isoform displayed identical calmodulin

dissociation rates Similar inhibition curves for each CaN

heterodimer were obtained with the CaN autoinhibitory

peptide (CaP) and cyclophilin A/cyclosporin A (CyPA/CsA)

using each peptide substrate at Kmconcentrations, except for

a five- to ninefold higher IC50 value measured for CaN

containing the b isoform with p-nitrophenylphosphate as

substrate No difference in stimulation of phosphatase

activity toward p-nitrophenylphosphate by FKBP12/FK506 was observed At low concentrations of FKBP12/FK506, CaN containing the a isoform is more sensitive to inhibition than CaN containing the b isoform using the phosphopep-tide substrates Higher concentrations of FKBP12/FK506 are required for maximal inhibition of b CaN using PO4 -DARPP-32(20–38) as substrate The functional differences conferred upon CaN by the a or b catalytic subunit isoforms suggest that the a:b and CaN:substrate ratios may determine the levels of CaN phosphatase activity toward specific sub-strates within tissues and specific cell types These findings also indicate that the a and b catalytic subunit isoforms give rise to substrate-dependent differences in sensitivity toward FKBP12/FK506

Keywords: calcineurin; calmodulin; dephosphorylation; Ser/ Thr protein phosphatase

Calcineurin (CaN) is a ubiquitously expressed Ca2+

/CaM-dependent protein phosphatase that is a critical component

of several Ca2+-dependent signaling pathways CaN

regulates a number of transcription factors and ion

channels and is involved in the regulation of T-cell

activation, long-term depression of postsynaptic potential,

synaptic vesicle recycling, and cardiac and skeletal muscle

hypertrophy [1,2] CaN is a heterodimer of a catalytic A

subunit (CnA) (58–61 kDa) and a Ca2+-binding Bsubunit

(19 kDa) Three CnA isoforms (a, b, c) have been

described The expression of CnAc is restricted to testis,

while the CnAa and CnAb isoforms are present in all

tissues examined [3] The physiological significance of the

expression of multiple CaN catalytic subunits within

the same cell or tissue is unknown Overall, the

amino-acid sequences of CnAa and CnAb are 81% identical [4]

However, the amino-acid sequence identity is 90% within

the core catalytic region, the CnB-binding helix, the CaM-binding domain, and the autoinhibitory domain [4] Mammalian CnAa or CnAb subunits exhibit extensive sequence homologies, with only one or five amino-acid changes between human and rat CnAa and CnAb, respectively [4] The most striking differences between the CnAa and CnAb catalytic subunit isoforms are the 12 Pro residues within the first 24 amino-acid residues of CnAb and multiple amino-acid differences C-terminal of the autoinhibitory domain [4]

It has been proposed that the two isoforms may exhibit substrate preferences and may also be selectively targeted to distinct subcellular locations [2] Variations in the amount and ratio of CnAa:CnAb have been noted within and between tissues [3] For example, although CnAa is more abundant than CnAb in mammalian brain, the CnAa/CnAb ratio in the striatum is 4 : 1, while in the cerebellum the ratio

is 2.5 : 1 [5] Similarly, CnAa is more abundant in kidney, but its expression is restricted to the tubules, while CnAb expression was observed only in the glomerular region [6]

In contrast, CnAb is more abundant in T and Bcells [6] In addition, in hepatocytes and some neurons, CnAa is found

in the cytoplasm and nucleus, while CnAb is found only in the cytoplasm [7,8] Together these findings raise the possibility of substrate-dependent functional differences between the a and b CaN catalytic subunit isoforms To determine whether the CnA a and b isoforms impart functional differences to CaN phosphatase activity, we have

Correspondence to B A Perrino, Department of Physiology & Cell

Biology, Anderson Medical Bldg MS352, University of Nevada

School of Medicine, Reno, Nevada, 89557,

Tel.: + 1 775 784 6396, Fax: + 1 775 784 6903,

E-mail: perrino@physio.unr.edu

Abbreviations: CaN, calcineurin; CnB, calcineurin regulatory

Bsubunit.

(Received 12 March 2002, revised 16 May 2002,

accepted 10 June 2002)

initiated in vitro studies of the enzymatic characteristics of

CaN heterodimers composed of either CnAa or CnAb

CaNa or CaNb heterodimers were obtained by

coexpress-ing each CnA isoform with CnBin the

Sf21/baculovirus-expression system CaM binding to each CnA isoform

within the CaN heterodimer was measured using

stopped-flow techniques We compared Km and Vmax values

obtained from assays of the phosphatase activities of both

CaN heterodimers towards two peptide substrates, and

pNPP We also compared the inhibition of CaN

phospha-tase activity toward the two peptide substrates by the CaN

autoinhibitory peptide, and the FKBP12/FK506 complex

The activation of CaN phosphatase activity towards pNPP

by the FKBP12/FK506 complex was also measured Our

results are the first indication that CaN heterodimers

composed of either CnAa or CnAb exhibit differences in

substrate selectivity and sensitivity to

immunophilin/immu-nosuppressant inhibition

E X P E R I M E N T A L P R O C E D U R E S

Materials

Rabbit anti-CnAa IgG, rabbit anti-CnAb IgG, and

horseradish peroxidase-conjugated goat anti-(rabbit IgG)

Ig were purchased from Chemicon T-4 DNA ligase,

restriction enzymes, Grace’s supplemented insect cell

medium, antibiotic/antimycotic solution, Pluronic F-68,

and bacterial culture media, were obtained from Gibco/

BRL Fetal bovine serum was purchased from Atlanta

Biologicals CaM–Sepharose was purchased from

Amer-sham Pharmacia Antibiotics, FKBP12, and pNPP were

obtained from Sigma Cyclophilin A, cyclosporine A, and

FK506 were obtained from Calbiochem Phospho-RII

peptide, CaP, and BioMol Green reagent were purchased

from BioMol Phospho-DARPP-32(20–38)

[LDPRQVE-MIRRRRPT(PO4)PAML] was purchased from American

Peptide Company Human CnAb cDNA was generously

provided by M M Lai and S Snyder (The Johns Hopkins

University School of Medicine [9]) All other materials and

reagents were of the highest quality available commercially

Recombinant CaNa and CaNb expression and purification

The 1.6 kb Sal1-Not1 CnAb fragment was ligated into

SalI-NotI cut pSE420 (InVitrogen) The pSE420/CnAb

con-struct was restriction digested with EcoRI–NotI and the

EcoRI–NotI CnAb cDNA ligated into EcoRI–NotI cut

pVL1393 (InVitrogen) Sf21 cells were transfected with the

pVL1393/CnAb construct using the Bac-N-Blue

Transfec-tion kit from InVitrogen Recombinant CnAb

baculovirus-es were screened by plaque assay and Wbaculovirus-estern blotting using

anti-CnAb Ig, and amplified and titered by plaque assay as

described [10] The coinfection, expression and purification

of baculovirus-expressed CaN containing the rat brain

CnAa subunit and rat brain CnB, or the human CnAb

subunit and rat brain CnBwas carried out as described,

except that monolayer cultures of Sf21 cells were used for

CaN expression [11] The phosphatase activities of the

purified CaN heterodimers were not further stimulated by

the addition of purified CnB, indicating that the CaN

heterodimers are composed of a 1 : 1 molar ratio of CnA/

CnB(data not shown) [10]

Phosphatase assays Dephosphorylation of PO4-RII peptide, PO4 -DARPP-32(20–38) peptide, and pNPP by CaN was carried out at

30°C in 50 lL reaction volumes in duplicate The assays were carried out in CaN assay buffer (40 mM Tris/HCl,

pH 7.5, 6 mMMg(C2H3O2)2, 8 mMascorbic acid, 100 mM NaCl, 0.1 mMCaCl2, 0.5 mMMnCl2, 0.5 mM dithiothrei-tol, 0.1 mgÆmL)1 bovine serum albumin) The reactions were initiated by addition of substrate, and the peptide dephosphorylation assays terminated by the addition of

100 lL of BioMol Green reagent, while the pNPP dephosphorylation assays were terminated by the addition

of 2 lL of 65% K2HPO4[12] The assay times are indicated

in the Figure legends The concentrations of CaN, CaM, CaP, FKBP12, and substrates are indicated in the Figure legends The Kmand Vmaxvalues were determined by linear regression analysis (PRISMsoftware) of inverse plots of the data from phosphatase assays in which the concentrations

of substrates were varied The FK506 or CsA stocks (1 mM

in dimethylsulfoxide) were diluted 80-fold in H20 in a glass tube before being added to the samples The final dimethyl-sulfoxide concentration of 0.05% in the assays had no effect

on CaN phosphatase activity FKBP12 and FK506 or CyPA and CsA were preincubated together on ice for

10 min, followed by incubation with CaN in assay buffer for 10 min prior to the start of the phosphatase assays The amount of phosphate released from the peptide substrates was determined by comparing the A620 values obtained from the experimental samples to the values generated from the K2HPO4 standard curve according to the manufac-turer’s (BioMol) instructions Dephosphorylation of pNPP was monitored by measuring the A410values [13] The data were best fit to a second-order polynomial equation by nonlinear regression analysis

Rate constant measurements The Lys75 to Cys CaM mutant (CaM C75) was labeled at Cys75 with the fluorescent probe acrylodan (Molecular Probes) essentially as described by Waxham et al [14,15] Dissociation rates of CaM from CaN isoforms were determined using a temperature-controlled stopped-flow fluorimeter (Hi-Tech SF 61-DX-2) equipped with a 150-Watt Hg-Xe lamp The excitation was at 365 nm and emission was monitored using a 399-nm cut-off filter Acrylodan-labeled CaM(C75) [CaM(C75)ACR] (0.1 lM) and either (0.3 lM) CaNAa or CaNAb in 25 mMMops,

pH 7.0, 150 mM KCl, 0.5 mM CaCl2 were rapidly mixed with native CaM (10 lM) in the same buffer at 20°C Rate constants were derived by fitting the experimental data using the Kinetasyst software supplied with the Hi-Tech stopped-flow fluorimeter In both cases, the best fit was obtained to a double-exponential model, where each rate accounted for approximately 50% of the observed ampli-tude change

R E S U L T S

Expression and purification of CaNa and CaNb CaN heterodimers composed of the Ca2+-binding B subunit and either the a or b catalytic subunit isoform were

generated by coinfecting Sf21 cells with recombinant CnB

baculovirus and either recombinant CnAa or CnAb

bacu-loviruses The CaN heterodimers were obtained by CaM–

Sepharose chromatography as described in Experimental

Procedures, and analyzed by SDS/PAGE and Western

blotting The purified CaNa and CaNb heterodimers are

90–95% pure as indicated by the Coomassie stained

SDS-polyacrylamide gel shown in Fig 1A The CnAb subunit

has a slightly slower mobility in SDS/PAGE, consistent with its higher molecular mass (59 kDa) compared to CnAa (57.6 kDa) Immunoblotting the purified CaN heterodimers with isoform-specific antibodies confirm that CnAa and CnAb proteins are expressed by the appropriate recombin-ant CnAa and CnAb baculoviruses (Fig 1B,C)

Kinetic assays of CaNa and CaNb phosphatase activity Kinetic analyses of the in vitro phosphatase activities of CaNa and CaNb were carried out to determine their Km and Vmaxvalues toward three different substrates; namely

PO4-RII peptide, PO4-DARPP-32(20–38), and pNPP The

PO4-RII peptide and pNPP have been extensively used to characterize the phosphatase activity of CaN [10,11,16,17] The PO4-DARPP-32(20–38) peptide contains amino-acid residues 20–38 of DARPP-32, which is a physiological substrate of CaN [18] PO4-Thr34 of the DARPP-32(20–38) peptide is dephosphorylated by CaN with Km and Vmax values similar to the values obtained with native DARPP-32 [18] In agreement with previous reports, the results in Fig 2 show that CaN phosphatase activity is characterized by different Kmand Vmaxvalues toward different substrates [19] However, the results also indicate that the phosphatase activities of CaN heterodimers containing the CnAa or CnAb catalytic subunit are characterized by different Km and Vmax values toward the same substrate For each substrate tested, CaN heterodimers containing the CnAa catalytic subunit are characterized by lower Kmand higher

Vmaxvalues compared to CaN heterodimers containing the CnAb catalytic subunit For both phosphopeptide sub-strates, the Kmvalues of CaN heterodimers containing the CnAb catalytic subunit are approximately threefold higher than the Kmvalues of CaN heterodimers containing the CnAa catalytic subunit (Fig 2A,B) With pNPP as sub-strate, the difference in Kmvalues is only twofold (Fig 2C)

Fig 1 SDS/PAGE and Western blot analysis of baculovirus expressed

CaN composed of CnAa or CnAb catalytic subunit isoforms CaN

heterodimers were expressed in Sf21 cells using recombinant

baculo-viruses, purified as described in Experimental procedures, and

ana-lyzed by SDS/PAGE (15%) and Western blotting Lane 1, CaN

heterodimer containing the CnAa catalytic subunit isoform (5 lg);

lane 2, CaN heterodimer containing the CnAb catalytic subunit

iso-form (3 lg) (A) Purified CaN heterodimers were separated by SDS/

PAGE and stained with Coomassie Brilliant Blue (B) Immunostaining

of purified CaN heterodimers using anti-CnAa Ig (C)

Immunostain-ing of purified CaN heterodimers usImmunostain-ing anti-CnAb Ig.

Fig 2 Kinetic analyses of CaN heterodimers composed of CnAa or CnAb catalytic subunit isoforms (A) Dephosphorylation of PO 4 -RII peptide CaNa or CaNb were each present at a final concentration of 5 n M CaM was present at a final concentration of 15 n M The reactions were allowed

to proceed for 7 min at 30 °C The concentrations of PO 4 -RII peptide used were 25 l M , 50 l M , 75 l M , 100 l M , and 150 l M (B) Dephospho-rylation of PO 4 -DARPP-32(20–38) CaNa or CaNb were each present at a final concentration of 50 n M CaM was present at a final concentration

of 150 n M The reactions were allowed to proceed for 10 min at 30 °C The concentrations of PO 4 -DARPP-32(20–38) used were 7 l M , 12 l M ,

17 l M , and 25 l M (C) Dephosphorylation of pNPP CaNa or CaNb were each present at a final concentration of 50 n M CaM was present at a final concentration of 150 n M The reactions were allowed to proceed for 20 min at 30 °C The concentrations of pNPP used were 10 m M , 15 m M ,

20 m M , and 30 m M , 50 m M , and 100 m M The results shown are representative of three assays performed in triplicate for each CaN heterodimer CaNa, d; CaNb, s.

CaN heterodimers containing the CnAa catalytic subunit

are characterized by approximately twofold higher Vmax

values toward the three substrates tested Together these

results indicate that CaN heterodimers used in these

experiments containing the CnAa catalytic subunit are

characterized by higher levels of phosphatase activity

toward these three substrates

Inhibition of CaNa and CaNb phosphatase activity by CaP

The CaN crystal structure shows that in the inactive state,

the CaN autoinhibitory domain lies over the catalytic site

[20] The amino-acid sequences of the CnAa and CnAb

catalytic domains are 90% identical, and the autoinhibitory

domain amino-acid sequences are 89% identical, suggesting

that CaN heterodimers containing the CnAa or CnAb

catalytic subunit would be equally inhibited by CaP, which

contains the autoinhibitory domain from CnAa [4,21]

However, because of the differences in Kmand Vmaxvalues

obtained with the CaN heterodimers used in these

experi-ments containing the CnAa or CnAb catalytic subunit

toward the same substrate, we examined the inhibition of

CaNa or CaNb phosphatase activity by CaP toward the

three substrates It has previously been reported that CaP

inhibits bovine brain CaN or baculovirus-expressed rat

brain CaNa with IC50values between 12 lMand 18 lM,

using32PO4-RII peptide as substrate [10,11] As shown in

Fig 3A, the phosphatase activities of CaN heterodimers

containing the CnAa or CnAb catalytic subunit toward

PO4-RII peptide are equally inhibited by CaP The IC50

values (10 lM-12 lM) and final extent of inhibition (90%

inhibition of phosphatase activity by 90 lMCaP) obtained

are similar to the previously reported values using32P-RII

peptide as substrate [10,11] Similar kinetics of inhibition

were also obtained for CaP with CaN heterodimers

containing the CnAa or CnAb catalytic subunit using

PO4-DARPP-32(20–38) as substrate (Fig 3B), giving IC50

values of 15 lMand 25 lM, respectively In addition, CaN

phosphatase activity is 85% inhibited by 90 lMCaP It has

been reported that bovine brain CaN phosphatase activity is 50% inhibited by 35 lMCaP using pNPP as substrate [21] Using CaN heterodimers containing the CnAa or CnAb catalytic subunit and pNPP as substrate, we measured IC50 values of 20 lMfor CaNa and 90 lMfor CaNb Further-more, in contrast to the results obtained using the phosphopeptide substrates, the phosphatase activities of CaNa and CaNb are only 70%, and 50% inhibited by

90 lMCaP, respectively

Inhibition of CaNa and CaNb phosphatase activity

by FKBP12/FK506 or CypA/CsA The structurally unrelated immunophilin/immunosuppres-sant complexes of FKBP12/FK506 or CypA/CsA inhibit CaN noncompetitively by binding to the CnB-binding helix, CnB, and one side of the substrate-binding cleft of the catalytic site to alter the active-site geometry [16,17,20,22]

As the mechanism of inhibition of CaN by the immuno-philin/immunosuppressant complexes is different from that

of CaP, we examined the inhibition of CaN heterodimers containing the CnAa or CnAb catalytic subunit by FKBP12/FK506 or CyP/CsA As shown in the dose– response curves of Fig 4, using the two different phospho-peptide substrates, CaNa is more sensitive to inhibition by FKBP12/FK506 than CaNb With PO4-RII peptide as substrate, 50% inhibition of CaNa activity was achieved with 73 nMFKBP12 (in the presence of 500 nMFK506), compared to 50% inhibition of CaNb activity by 120 nM FKBP12 As expected, the high concentration of FKBP12 (200 nM) resulted in 90% inhibition of CaNa and CaNb with PO4-RII peptide as substrate (Fig 4A) Similarly, 50% inhibition of CaNa activity was achieved with 60 nM FKBP12, compared to 50% inhibition of CaNb activity

by 117 nM FKBP12 using PO4-DARPP-32(20–38) as substrate 200 nM FKBP12 resulted in 90% inhibition of CaNa using PO4-DARPP-32(20–38) as substrate However, 90% inhibition of CaNb phosphatase activity was achieved

by 1 lM FKBP12 using PO4-DARPP-32(20–38) as

Fig 3 Inhibition of CaN heterodimers composed of CnAa or CnAb catalytic subunit isoforms by CaP (A) The reactions proceeded for 10 min at

30 °C using 32 l M and 91 l M PO 4 -RII peptide for CaNa or CaNb, respectively CaNa or CaNb were each present at a final concentration of 5 n M CaM was present at a final concentration of 15 n M (B) The reactions proceeded for 10 min at 30 °C using 6 l M and 21 l M PO 4 -DARPP-32(20–38) for CaNa or CaNb, respectively CnAa or CnAb were each present at a final concentration of 50 n M CaM was present at a final concentration of

150 n M (C) The reactions proceeded for 20 min at 30 °C using 45 m M and 83 m M pNPP for CaNa or CaNb, respectively CaNa or CaNb were each present at a final concentration of 50 n M CaM was present at a final concentration of 150 n M The results shown are the averages ± SD from three assays in triplicate for each CaN heterodimer CaNa, d; CaNb, s.

substrate These findings indicate that the CaNb used in

these experiments is less sensitive than CaNa to inhibition

by FKBP12/FK506 when PO4-DARPP-32(20–38) is used

as substrate

With CyPA/CsA and PO4-RII peptide as substrate 50%

inhibition of CaNa activity was achieved with 342 nM

CyPA (in the presence of 2 lM CsA), compared to 50%

inhibition of CaNb activity by 456 nM CyPA A high

concentration of CyPA (1000 nM) resulted in 80%-90%

inhibition of CaNa and CaNb with PO4-RII peptide as

substrate (Fig 4B) Similar IC50 values were obtained for

CyPA/CsA inhibition of CaNa and CaNb using PO4

-DARPP-32(20–38) as substrate and 80–90% inhibition of

phosphatase activity was attained with 1000 nM CyPA

(Fig 4D) Using two different phosphopeptide substrates,

these findings indicate that CyPA/CsA results in similar

inhibition of both CaNa and CaNb These findings also

indicate that FKBP12/FK506 is a more potent inhibitor of

both CaNa and CaNb than CyPA/CsA

Activation of CaNa and CaNb phosphatase activity

by FKBP12/FK506 toward pNPP

In contrast to the inhibition of CaN phosphatase activity

toward phosphopeptide and phospho-protein substrates by

FKBP12/FK506, the phosphatase activity of CaN toward

the small organic compound pNPP is increased two- to

fourfold by FKBP12/FK506 [16,23] These observations are

consistent with the findings that the FKBP12/FK506

complex alters the conformation of the active site [20] To

examine the possibility that FKBP12/FK506 may have

different effects on the activities of CaNa and CaNb toward pNPP, we examined the activation of CaNa and CaNb phosphatase activities toward pNPP by FKBP12/FK506

As shown in Fig 5, a twofold increase in CaN phosphatase activity was observed with 200 nM FKBP12 and 500 nM FK506, in agreement with previous studies [16,23] How-ever, there was essentially no difference in the dose-dependent activation of CaNa or CaNb phosphatase activities toward pNPP by FKBP12/FK506

Determination of dissociation rates between CaM(C75)ACRand CaNa or CaNb

Because of the highly conserved amino-acid sequences of the CnAa and CnAb catalytic, CaM-binding, and CnB-binding domains, it was surprising to find differences in the phosphatase activities between CaNa and CaNb toward the same substrate The results of the kinetic analyses, and inhibition studies are summarized in Table 1 Although the amino-acid sequences of the CnAa and CnAb CaM-binding domains are identical, adjacent functional domains can influence the CaM-binding properties of CaM-binding

a helices [24] Thus, stop-flow analyses of Ca2+/CaM off rates were carried out in order to determine whether there are any differences in the Ca2+/CaM-binding properties of CaNa and Canb The dissociation rates of CaNa or CaNb from CaM(C75)ACR were determined by monitoring the rates of fluorescence decrease as CaM(C75)ACRbound to CaNa or CaNb is exchanged for excess unlabeled CaM (Fig 6) The data were best fit by a double-exponential model and yielded two essentially identical fast and slow

Fig 4 Inhibition of CaN heterodimers composed of CnAa or CnAb catalytic subunit isoforms by FKBP12/FK506 or CyPA/CsA FK506 or CsA were each present at a final concentration of 2 l M , and the FKB P12 or CyPA concentrations varied as indicated in the figure legends The reactions proceeded for

20 min at 30 °C using 32 l M (A) and 91 l M

(B) PO 4 -RII peptide for CaNa or CaNb, respectively CaNa or CaNb were each present

at a final concentration of 5 n M CaM was present at a final concentration of 15 n M The reactions proceeded for 20 min at 30 °C using

6 l M (C) and 21 l M (D) PO 4 -DARPP-32(20– 38) for CaNa or CaNb, respectively CaNa or CaNb were each present at a final concent-ration of 5 n M CaM was present at a final concentration of 15 n M The results shown are the averages ± SD from three assays in triplicate for each CaN heterodimer CaNa, d; CaNb, s.

CaM dissociation constants for both CaNa and CaNb Fast

and slow rates of 4 s)1and 0.4 s)1, and 3.9 s)1and 0.4 s)1

were obtained for CaNa and CaNb, respectively These

results indicate that there are essentially no differences in

Ca2+/CaM-dissociation from the CnAa and CnAb

cata-lytic subunit isoforms The mechanistic basis for the two

rate constants is not clear, although Ca2+/CaM-binding to

the CaM-binding domain of CnA is modulated by Ca2+

-binding to CnB[25,26]

D I S C U S S I O N

Previous studies of the CaN a and b catalytic subunit

isoforms have mainly focused on their tissue and subcellular

distributions [5–8,27] These reports have provided

import-ant information demonstrating regional differences in

expression levels within tissues and differences in the

subcellular distribution of CnAa and CnAb Although it

has been generally assumed that CaNa and CaNb dephosphorylate the same set of substrates, the differences

in CnAa and CnAb localization and expression have been proposed to reflect differences in substrate selectivity between CaNa and CaNb [2,4] However, in the absence

of information concerning enzymatic differences between these two CaN catalytic subunit isoforms, the physiological significance of their differential distribution is unclear To address this question, we have examined the enzymatic properties of CaN heterodimers containing either the CnAa

or CnAb catalytic subunit In agreement with previous findings using purified mammalian brain CaN, we found that CaNa or CaNb heterodimers are characterized by different Kmand Vmaxvalues toward different substrates [19] However, our results also indicate that CaNa and CaNb heterodimers exhibit differences in phosphatase activity toward the same substrate, as indicated by the different Kmand Vmaxvalues obtained The results from the kinetic assays show that the CaN heterodimers containing the CnAa subunit have higher levels of phosphatase activity toward all three substrates tested, as indicated by lower Km and higher Vmaxvalues (Table 1) These findings indicate that the CaN catalytic subunits are characterized by different rates of phosphatase activity towards their sub-strates, and suggest that the CaNa:CaNb ratio within a cell

or tissue is an important determinant of CaN phosphatase activity toward specific substrates

The use of CnAa knockout mice has provided evidence that CaNa and CaNb exhibit selective phosphatase activity toward specific substrates within a cell or tissue Tau proteins are hyperphosphorylated in the brains of CnAa –/– mice [28] Similarly, hippocampal depotentiation is abol-ished while long-term depression and long-term potentia-tion are unaffected in CnAa –/– mice [29] Furthermore, CnAa –/– mice have impaired antigen-specific T-cell responses in vivo [30] Two possible interpretations of these findings are (a) CaNa selectively dephosphorylates tau proteins, and also specifically dephosphorylates substrates required for hippocampal depotentiation and antigen-induced T-cell responses, or (b) CaNa and CaNb have no substrate preferences, but the residual CaNb phosphatase activity is insufficient to dephosphorylate tau or participate

in hippocampal depotentiation and antigen-induced T-cell responses The findings that CnAb is the predominant isoform present in T-cells argues against impaired antigen-specific T-cell responses in CnAa –/– mice being due to insufficient CaN phosphatase activity [6] Our findings that the CnAa and CnAb catalytic subunits confer differences in substrate affinity and phosphatase activity (as shown by

Fig 5 Stimulation of CaN phosphatase activity toward pNPP by

FKBP12/FK506 FK506 was present at a final concentration of

500 n M , and the FKBP12 concentration varied as indicated in the

figure legend (A) The reactions proceeded for 45 min at 30 °C using

45 m M and 83 m M pNPP for CaNa or CaNb, respectively CaNa or

CaNb were each present at a final concentration of 5 n M CaM was

present at a final concentration of 15 n M The results shown are the

averages ± SD from three assays in triplicate for each CaN

het-erodimer CaNa, d; CaNb, s.

Table 1 Summary of kinetic parameters of CaNa and CaNb phosphatase activities toward three substrates The K m and V max values, CaP IC 50

values, FKBP12 IC 50 and CyPA IC 50 values were obtained from the inverse plots of V vs [S], the CaP dose–response experiments, and the FKBP12 and CyPA dose–response experiments, respectively.

PO 4 -RII peptide (l M ) PO 4 -DARPP-32 pNPP

K m 32 l M 91 l M 6.2 l M 21 l M 45 m M 83 m M

V max (lmolÆmin)1Æmg)1) 4.1 2.8 0.384 0.224 5.8 3.1

different Kmand Vmaxvalues) toward the same substrate

support the conclusion that CaN substrates are

differen-tially dephosphorylated by CaNa and CaNb in vivo

The differences in phosphatase activity and sensitivity to

FKBP12/FK506 or CyPA/CsA inhibition between CaNa

and CaNb that we observed may be due to subtle differences

in how the CnAa and CnAb catalytic subunits interact with

CnB The regulation of enzyme activity by two EF-hand

Ca2+-binding proteins is unique to CaN [4] Similar to

Ca2+/CaM-dependent kinases, Ca2+/CaM-binding to CnA

activates the enzyme by relieving inhibition due to an autoinhibitory domain, which is evidenced by an increase in

Vmax [11,31] Conversely, Ca2+-binding to CnBactivates CaN primarily by affecting the affinity of the catalytic site for substrate, as evidenced by the decrease in Km[11,31] Circular dichroism analysis has shown that CnBand CnA both undergo conformational changes upon Ca2+binding

to CnB[32] It appears that in the absence of Ca2+the catalytic core is in an inactive conformation and that Ca2+ -binding to CnBchanges the conformation of the catalytic core to allow substrate binding [25,33] The CaN crystal structure shows that the CnB-binding helix is immediately C-terminal of the b14 strand of one half of the b sandwich forming the catalytic core, providing a mechanism for transmission of Ca2+-induced conformational changes in CnBto the active site [22] In fact, the catalytic activity of CaN is sensitive to the amino-acid composition of the region linking the CnB-binding helix to the b14 strand of the catalytic core [34,35] Mutations S373P, H375L, and L379S decrease CaN activity, indicating the importance of this linker region to the activation of CaN by Ca2+-binding to CnB[34] Calcium binding to CnBalso influences the affinity of Ca2+/CaM for its binding domain on CnA [25,26] CnBhas two low affinity and two high affinity EF-hand Ca2+-binding loops [36] In the absence of Ca2+ -binding to the low affinity sites, the CaM binding domain interacts with the exposed side of the CnB-binding helix [26] Calcium binding to the low affinity sites on CnBdisrupts the interaction between the CaM-binding domain and the CnB-binding helix and increases the affinity of the CaM-CnB-binding domain for Ca2+/CaM [25,26] The regulation of the CaM-binding domain by CnBmay partly account for the slow and fast dissociation constants we measured using stopped-flow analysis

The FKBP12/FK506 complex inhibits CaN noncompeti-tively using PO4-RII and PO4-DARPP-32 as substrates [20] These findings are consistent with crystallographic data showing the active site and substrate-binding cleft are not directly blocked by the FKBP12/FK506 complex, support-ing the conclusion that alterations in the active site conformation affect the substrate-binding cleft and are responsible for the inhibition by FKBP12/FK506 [20,22] This mechanism of inhibition would account for the findings that CaN phosphatase activity toward the small organic molecule pNPP is increased by the FKBP12/FK506 complex (Fig 5) [16,23] However, the structural and enzymatic studies of the interactions between the FKBP12/FK506 complex and CaN have been carried out with CaN containing the CnAa catalytic subunit [16,20,22,23] The results presented here are the first direct studies of the inhibition of CaN containing the CnAb catalytic subunit by FKBP12/FK506 or CyPA/CsA For both immunophilin/immunosuppressant complexes, the

IC50 values are higher than previously reported [23,37] Differences in CaN preparations and methods of enzyme activity assays may account for the differences in IC50 values For these studies the CaN activity was assayed in the presence of ascorbic acid, which results in higher activity compared to activity in the absence of antioxidants [38,39] With both peptide substrates CaNa was more sensitive to inhibition by FKBP12/FK506 than CaNb (Fig 4) How-ever, CaNa and CaNb were both 90% inhibited by 200 nM FKBP12 using PO-RII at K concentrations Similarly,

Fig 6 Measurement of CaM dissociation rate constants from CaNa or

CaNb associated with CaM (C75) ACR using stopped-flow kinetics The

time course for CaM dissociation from CaNa (A) or CaNb (B) from

CaM(C75) ACR as determined using a stopped-flow fluorimeter.

CaM(C75) ACR (0.1 l M ) and either (0.3 l M ) CaNa (or CaNb) in

25 m M Mops, pH 7.0, 150 m M KCl, 0.5 m M CaCl 2 were rapidly mixed

with native CaM (10 l M ) in the same buffer at 20 °C The excitation

was at 365 nm and emission was monitored using a 399-nm cut-off

filter Each curve represents the average of four exchange reactions.

CaNa was 90% inhibited by 200 nMFKBP12 using PO4

-DARPP-32(20–38) at the Km concentration In contrast,

only 60% inhibition of CaNb by 200 nM FKBP12 was

measured using PO4-DARPP-32(20–38) as substrate, and

90% inhibition required 1000 nM FKBP12 Our results

using the two different phosphopeptide substrates show that

CsA is a less potent inhibitor of both CaNa and CaNb than

FK506 These results are consistent with the findings that

in vivo, CsA and FK506 are equally effective, but FK506 is

10-fold more potent in inhibiting CaN activity and IL-2

gene activation [37] Although FK506 and CyPA are

structurally unrelated compounds, biochemical and

muta-tional studies indicate that FKBP12/FK506 and CyPA/CsA

bind to a common site on the CaN heterodimer composed

of the CnB-binding helix, CnB and part of the

substrate-binding cleft of CaN [35,40] However, differences in the

interactions between these structurally dissimilar

immuno-philin/immunosupressant complexes and CaN will likely

contribute to differences in their inhibitory potency

As the substrate-binding cleft geometry is affected by

CnB, the interaction between CnA and CnB may affect how

the FKBP12/FK506 and CyPA/CsA complexes affect the

substrate-binding cleft Thus, both the catalytic subunit and

substrate may influence the degree of inhibition of CaN

phosphatase activity by FKBP12/FK506 and CyPA/CsA

The differences in phosphatase activity and sensitivity to

FKBP12/FK506 inhibition between CaNa and CaNb are

likely not due to differences in the linker region, as the

amino-acid sequences of the CnB-binding helix, and the

linker region of CnAa and CnAb are identical [41]

However, proteolysis of the CnA N-terminus results in loss

of CaN activity, suggesting that the CnA N-terminus is

involved in enzyme activation [42] Indeed, as shown in the

crystal structure, the N-terminus of CnA interacts with the

C-terminal half of CnBas part of a CnB-binding cleft,

suggesting that the interaction of the CnA N-terminus with

CnBis involved in Ca2+CnB-dependent activation of the

enzyme [20,22] As noted previously, the N-terminus of

CnAb is different from CnAa, containing 12 Pro residues

within the first 24 amino acids [41] Information regarding

the interaction of the CnAb N-terminus with CnBis lacking

because only CaN containing CnAa has been crystalized

[20,22] However, the presence of 11 consecutive Pro

residues in the N-terminus of CnAb suggests that the CnAa

and CnAb N-termini interact with CnBdifferently, giving

rise to the different Kmand Vmaxvalues measured for CaNa

and CaNb using the same substrate

Polyproline motifs are involved in protein–protein

inter-actions [43] Molecular modeling indicates that an

11-residue type II polyproline helix exactly spans the length of

the central helix of CaM, and led to the proposal that the 11

consecutive Pro residues of CnAb may modulate the

interaction of CaM with the CaM-binding domain of CnAb

[41] However, as noted previously, the crystal structure

subsequently showed that the N-terminus of CnAa forms

extensive contacts with CnB Interestingly, the central helix

of CnBdiffers in length from the central helix of CaM by

only one amino-acid residue [44] These findings suggest

that the 11 Pro residues within the first 24 amino-acid

residues of CnAb may instead interact with the central helix

of CnBand affect its interaction with CnAb As Ca2+

-binding to CnBregulates the catalytic activity of CnA, this

proposal provides a potential mechanistic basis for the

differences in Kmand Vmaxvalues between CaNa and CaNb obtained using the same substrate

A C K N O W L E D G E M E N T S

This work was supported by National Institutes of Health Grants

NS-36318, DK-57168 (B.A.P), and The Medical Research Council, UK (G117/440) (L.H.C) L.H.C is a MRC Senior Fellow in Basic Science.

We thank M Neal Waxham (University of Texas Medical School at Houston) for the generous gift of CaM (C75), and Michael M Lai and Solomon H Snyder (The Johns Hopkins University School of Medicine) for providing the human CnAb cDNA.

R E F E R E N C E S

1 Olson, E.N & Williams, R.S (2000) Remodeling muscles with calcineurin Bioessays 22, 510–519.

2 Perrino, B.A & Soderling, T.R (1998) Biochemistry and phar-macology of calmodulin-regulated phosphatase calcineurin In Calmodulin and Signal Transduction (VanEldik, L & Watterson, D.M., eds), pp 169–236 Academic Press, New York.

3 Rusnak, F & Mertz, P (2000) Calcineurin: form and function Physiol Rev 80, 1483–1521.

4 Guerini, D & Klee, C.B (1991) Structural diversity of calcineurin,

a Ca2+ and calmodulin-stimulated protein phosphatase Adv Prot Phosphatases 6, 391–410.

5 Kuno, T., Mukai, H., Ito, A., Chang, C.D., Kishima, K., Saito, N.

& Tanaka, C (1992) Distinct cellular expression of calcineurin Aa and Ab in rat brain J Neurochem 58, 1643–1651.

6 Jiang, H., Xiong, F., Kong, S., Ogawa, T., Kobayashi, M & Liu, J.O (1997) Distinct tissue and cellular distribution of two major isoforms of calcineurin Mol Immunol 34, 663–669.

7 B osser, R., Aligue, R., Guerini, D., Agell, N., Carafoli, E & Bachs, O (1993) Calmodulin can modulate protein phospho-rylation in rat liver cells nuclei J Biol Chem 268, 15477–15483.

8 Usuda, N., Arai, H., Sasaki, H., Hanai, T., Nagata, T., Muramatsu, T., Kincaid, R.L & Higuchi, S (1996) Differential subcellular localization of neural isoforms of the catalytic subunit

of calmodulin-dependent protein phosphatase (calcineurin) in central nervous system neurons: immunohistochemistry on for-malin-fixed paraffin sections employing antigen retrieval by microwave irradiation J Histochem Cytochem 44, 13–18.

9 Lai, M.M., Burnett, P.E., Wolosker, H., Blackshear, S & Snyder, S.H (1998) Cain, a novel physiologic protein inhibitor of calci-neurin J Biol Chem 273, 18325–18331.

10 Perrino, B.A., Fong, Y.L., Brickey, D.A., Saitoh, Y., Ushio, Y., Fukunaga, K., Miyamoto, E & Soderling, T.R (1992) Char-acterization of the phosphatase activity of a baculovirus-expressed calcineurin A isoform J Biol Chem 267, 15965–15969.

11 Perrino, B.A., Ng, L.Y & Soderling, T.R (1995) Calcium reg-ulation of calcineurin phosphatase activity by its Bsubunit and calmodulin: role of the autoinhibitory domain J Biol Chem 270, 340–346.

12 Pallen, C.J & Wang, J.H (1983) Calmodulin-stimulated depho-sphorylation of p-nitrophenyl phosphate and free phosphotyro-sine by calcineurin J Biol Chem 258, 8550–8553.

13 Li, H.-C (1984) Activation of brain calcineurin phosphatase towards nonprotein phosphoesters by Ca2+, calmodulin, and

Mg2+ J Biol Chem 259, 8801–8807.

14 Putkey, J.A & Waxham, M.N (1996) A peptide model for cal-modulin trapping by calcium/calcal-modulin-dependent protein kin-ase II J Biol Chem 271, 29619–29623.

15 Waxham, M.N., Tsai, A & Putkey, J.A (1998) A mechanism for calmodulin (CaM) trapping by CaM-kinase II defined by a family

of CaM-binding peptides J Biol Chem 273, 17579–17584.

16 Liu, J., Farmer, J.D Jr, Lane, W.S., Friedman, J., Weissman, I & Schreiber, S.L (1991) Calcineurin is a common target of

cyclophilin-cyclosporin A and FKBP-FK506 complexes Cell 66,

807–815.

17 Liu, J., Albers, M.W., Wandless, T.J., Luan, S., Alberg, D.G.,

Belshaw, P.J., Cohen, P., MacKintosh, C., Klee, C.B & Schreiber,

S.L (1992) Inhibition of T-cell signaling by immunophilin-ligand

complexes correlates with loss of calcineurin phosphatase activity.

Biochemistry 31, 3896–3901.

18 Hemmings, H.C Jr, Nairn, A.C., Elliot, J.I & Greengard, P.

(1990) Synthetic peptide analogs of DARPP-32 (Mr 32,000

dopamine- and cAMP-regulated phosphoprotein), an inhibitor of

protein phosphatase-1 J Biol Chem 265, 20369–20376.

19 Klee, C.B., Draetta, G.F & Hubbard, M.J (1988) Calcineurin.

Adv Enzymol 61, 149–200.

20 Kissinger, C.R., Parge, H.E., Knighton, D.R., Lewis, C.T.,

Pelletier, L.A., Tempczyk, A., Kalish, V.J., Tucker, K.D.,

Show-alter, R.E., Moomaw, E.W., Gastinel, L.N., Habuka, N., Chen,

X., Maldonado, F., Barker, J.E., Bacquet, R & Villafranca, J.E.

(1995) Crystal structures of human calcineurin and the human

FKBP12-FK506 complex Nature 378, 641–644.

21 Hashimoto, Y., Perrino, B.A & Soderling, T.R (1990)

Identifi-cation of an autoinhibitory domain in calcineurin J Biol Chem.

265, 1924–1927.

22 Griffith, J.P., Kim, J.L., Kim, E.E., Sintchak, M.D., Thomson,

J.A., Fitzgibbon, M.J., Fleming, M.A., Caron, P.R., Hsiao, K &

Navia, M.A (1995) X-ray structure of calcineurin inhibited by the

immunophilin-immunosuppressant FKBP12-FK506 complex.

Cell 82, 507–522.

23 Swanson, S.K., Born, T., Zydowski, L.D., Cho, H., Chang, H.Y.,

Walsh, C.T & Rusnak, F (1992) Cyclosporin-mediated inhibition

of bovine calcineurin by cyclophilins A and B Proc Natl Acad.

Sci USA 89, 3741–3745.

24 Kwiatkowski, A.P & McGill, J.M (2000) Alternative splice

variant of c-calmodulin-dependent protein kinase II alters

acti-vation by calmodulin Arch Biochem Biophysics 378, 377–383.

25 Feng, B & Stemmer, P.M (2001) Ca 2+ binding site 2 in

cal-cineurin-Bmodulates calmodulin-dependent calcineurin

phos-phatase activity Biochemistry 40, 8808–8814.

26 Yang, S.A & Klee, C.B (2000) Low affinity Ca2+-binding sites of

calcineurin Bmediate conformational changes in calcineurin A.

Biochemistry 39, 16147–16154.

27 Strack, S., Wadzinski, B.E & Ebner, F.F (1996) Localization of

the calcium/calmodulin-dependent protein phosphatase,

cal-cineurin, in the hindbrain and spinal cord of the rat J Comp.

Neurol 375, 66–76.

28 Kayyali, U.S., Zhang, W., Yee, A.G., Seidman, J.G & Potter, H.

(1997) Cytoskeletal changes in the brains of mice lacking

calci-neurin A alpha J Neurochem 68, 1668–1678.

29 Zhuo, M., Zhang, W., Son, H., Mansuy, I., Sobel, R.A., Seidman, J.

& Kandel, E.R (1999) A selective role of calcineurin Aa in

synaptic depotentiation in hippocampus Proc Natl Acad Sci.

USA 96, 4650–4655.

30 Zhang, W., Zimmer, G., Chen, J., Ladd, D., Li, E., Alt, F.W., Wiederrecht, G.J., Cryan, J., O’Neill, E.A., Seidman, C.E., Abbas, A.F & Seidman, J.G (1996) T-cell responses in calcineurin Aa-deficient mice J Exp Med 183, 413–420.

31 Stemmer, P.M & Klee, C.B (1994) Dual calcium regulation of calcineurin by calmodulin and calcineurin B Biochemistry 33, 6859–6866.

32 Wolff, D.J & Sved, D.W (1985) The divalent cation dependence

of bovine brain calmodulin-dependent phosphatase J Biol Chem.

260, 4195–4202.

33 Wei, Q & Lee, E.Y.C (1997) Mutagenesis of the L7 loop con-necting b strands 12 and 13 of calcineurin: evidence for a structural role in activity changes Biochemistry 36, 7418–7424.

34 Jiang, B & Cyert, M.S (1999) Identification of a novel region critical for calcineurin function in vivo and in vitro J Biol Chem.

274, 18543–18551.

35 Kawamura, A & Su, M.S.S (1995) Interaction of FKBP12-FK506 with calcineurin A at the Bsubunit-binding domain.

J Biol Chem 270, 15463–15466.

36 Feng, B & Stemmer, P.M (1999) Interactions of calcineurin A, calcineurin B, and Ca2+ Biochemistry 38, 12481–12489.

37 Fruman, D.A., Klee, C.B., Bierer, B.E & Burakoff, S.J (1992) Calcineurin phosphatase activity in T lymphocytes is inhibited by

FK 506 and cyclosporin A Proc Natl Acad Sci USA 89, 3686– 3690.

38 Mitsuhashi, S., Shima, H., Kikuchi, K., Yazawa, M., Hatsuse, R., Maeda, K., Yazawa, M., Murayama, T., Okuma, Y & Nomura,

Y (2000) Development of an assay method for activities of serine/threonine protein phosphatase type 2B(calcineurin) in crude extracts Anal Biochem 278, 192–197.

39 Sommer, D., Fakata, K.L., Swanson, S.K & Stemmer, P.M (2000) Modulation of the phosphatase activity of calcineurin by oxidants and antioxidants in vitro Eur J Biochem 267, 2312– 2322.

40 Milan, D., Griffith, J., Su, M., Price, E.R & McKeon, F.D (1994) The latch region of calcineurin Bis involved in both immunosuppressant-immunophilin complex docking and phos-phatase activation Cell 79, 437–447.

41 Guerini, D & Klee, C.B (1989) Cloning of human calcineurin A: Evidence for two isozymes and identification of a polyproline structural domain Proc Natl Acad Sci USA 86, 9183–9187.

42 Hubbard, M.J & Klee, C.B (1989) Functional domain structure

of calcineurin A: mapping by limited proteolysis Biochemistry 28, 1868–1874.

43 Dalgarno, D.C., Botfield, M.C & Rickles, R.J (1997) SH3 domains and drug design: Ligands, structure, and biological function Biopolymers 43, 383–400.

44 Klee, C.B & Cohen, P (1988) The calmodulin-regulated protein phosphatase In Calmodulin (Klee, C.B & Cohen, P., eds), pp 225–248 Elsevier, Amsterdam.