Báo cáo khoa học: Regulation of the muscle-specific AMP-activated protein kinase a2b2c3 complexes by AMP and implications of the mutations in the c3-subunit for the AMP dependence of the enzyme docx

kinase a2b2c3 complexes by AMP and implications of the mutations in the c3-subunit for the AMP dependence of the enzyme Kerstin Lindgren*, Mattias Ormestad*, Ma˚rten Persson, Sofia Marti

kinase a2b2c3 complexes by AMP and implications of the mutations in the c3-subunit for the AMP dependence of the enzyme

Kerstin Lindgren*, Mattias Ormestad*, Ma˚rten Persson, Sofia Martinsson, L Thomas Svensson and Margit Mahlapuu

Discovery Research, Biovitrum AB, Go¨teborg, Sweden

The AMP-activated protein kinase (AMPK) is an

evo-lutionarily conserved enzyme that is important for

metabolic sensing both within individual cells and at a

whole body level [1–3] AMPK activation has been

shown to increase glucose uptake and fatty acid

oxida-tion in skeletal muscle, suppress glucose output from

the liver, and diminish adiposity [4–7], and as a result AMPK is considered a promising novel drug target for the treatment of diabetes and related metabolic disorders

AMPK is a heterotrimeric complex consisting of a catalytic a-subunit and regulatory b- and c-subunits,

Keywords

allosteric regulation by AMP; AMP-activated

protein kinase; c3 isoform

Correspondence

M Mahlapuu, Discovery Research,

Biovitrum AB, Biotech Center, Arvid

Wallgrens Backe 20, SE-413 46 Go¨teborg,

Sweden

Fax: +46 31 749 1101

Tel: +46 31 749 1126

E-mail: margit.mahlapuu@biovitrum.com

*These authors contributed equally to this

study

(Received 9 February 2007, revised 20

March 2007, accepted 4 April 2007)

doi:10.1111/j.1742-4658.2007.05821.x

The AMP-activated protein kinase is an evolutionarily conserved hetero-trimer that is important for metabolic sensing in all eukaryotes The muscle-specific isoform of the regulatory c-subunit of the kinase, AMP-activated protein kinase c3, has a key role in glucose and fat metabolism in skeletal muscle, as suggested by metabolic characterization of humans, pigs and mice harboring substitutions in the AMP-binding Bateman domains of c3

We demonstrate that AMP-activated protein kinase a2b2c3 trimers are allosterically activated approximately three-fold by AMP with a half-maximal stimulation (A0.5) at 1.9 ± 0.5 or 2.6 ± 0.3 lm, as measured for complexes expressed in Escherichia coli or mammalian cells, respectively

We show that mutations in the N-terminal Bateman domain of c3 (R225Q, H306R and R307G) increased the A0.5 values for AMP, whereas the fold activation of the enzyme by 200 lm AMP remained unchanged in compari-son to the wild-type complex The mutations in the C-terminal Bateman domain of c3 (H453R and R454G), on the other hand, substantially reduced the fold stimulation of the complex by 200 lm AMP, and resulted

in AMP dependence curves similar to those of the double mutant, R225Q⁄ R454G A V224I mutation in c3, known to result in a reduced gly-cogen content in pigs, did not affect the fold activation or the A0.5values for AMP Importantly, we did not detect any increase in phosphorylation

of Thr172 of a2 by the upstream kinases in the presence of increasing con-centrations of AMP Taken together, the data show that different muta-tions in c3 exert different effects on the allosteric regulation of the a2b2c3 complex by AMP, whereas we find no evidence for their role in regulating the level of phosphorylation of a2 by upstream kinases

Abbreviations

AICAR, 5-aminoimidazole-4-carboxamide-1-b- D -ribonucleoside; AMPK, AMP-activated protein kinase; AMPKK, AMP-activated protein kinase kinase; CAMKKb, Ca2+⁄ calmodulin-dependent protein kinase b; CBS, cystathionine-b-synthase; TAK1, transforming growth factor b-activating kinase 1; ZMP, 5-aminoimidazole-4-carboxamide riboside monophosphate.

all of which are required for its activity [8,9] In the

mammalian genome, two isoforms of the a-subunit (a1

and a2) and the b-subunit (b1 and b2), and three

iso-forms of the c-subunit (c1, c2 and c3), are present,

with 12 possible heterotrimeric combinations The

cat-alytic a-subunit contains a conventional serine⁄

threo-nine protein kinase domain [10], and phosphorylation

of Thr172 within the activation loop of the a-subunit

is essential for the activity of the enzyme Several

upstream kinases have been shown to phosphorylate

AMPK at Thr172, including tumor suppressor LKB1

[11–13], Ca2+⁄ calmodulin-dependent protein kinase b

(CAMKKb) [14,15] and transforming growth

fac-tor b-activating kinase (TAK1) [16] The three AMPK

c-subunit isoforms differ in their N-terminal sequences

but share the four tandem repeats of a structural

module known as the cystathionine-b-synthase (CBS)

domain, located in the C-terminal region [17] First

recognized by Bateman [18], CBS motifs contain about

60 residues and are found in a range of diverse

pro-teins The basic functional unit is believed to contain

two CBS motifs, which associate closely together [19],

binding ligands containing adenosine [20], and the

term ‘Bateman domain’ has been suggested to refer to

the structure formed by two tandem repeats [21] In

mammalian AMPK, the CBS domains have been

shown to bind the allosteric activator of the kinase,

AMP, whereas the two pairs of CBS domains both

bind one molecule of AMP [20] In terms of the

struc-ture of AMPK, the b-subunit has been reported to act

as a scaffold for binding of the a-subunit and

c-sub-unit [22]

Previously, we have shown that AMPK c3 is

selec-tively expressed in glycolytic (white, fast-twitch type II)

skeletal muscle [23], and is thus the only AMPK

sub-unit isoform exhibiting strictly restricted tissue-specific

expression Furthermore, in both human and rodent

skeletal muscle, c3 primarily forms complexes with the

a2 and b2 isoforms [23,24] Two naturally occurring

missense mutations have been identified in the first

CBS domain of the pig c3 gene, resulting in increased

(R225Q) or decreased (V224I) skeletal muscle glycogen

content [25–27] R225Q carriers are also characterized

by a higher oxidative capacity in white skeletal muscle

fibers [28,29] Transgenic mice with skeletal

muscle-specific expression of the R225Q substituted form of

mouse c3 replicate the phenotype observed in pigs,

with elevated glycogen levels and increased fat

oxida-tion in muscle tissue [30,31] Recently, R225W and

R307C substitutions in the human c3 gene have been

reported to result in increased muscle glycogen

tent, indicating that the function of AMPK c3 is

con-served across the mammalian species [32] The R225Q

mutation in c3 is equivalent to the R302Q substitution

in c2, the dominant isoform of the c-subunit in the heart The R302Q mutation, together with several other substitutions identified in CBS domains of the c2 gene, cause Wolff–Parkinson–White syndrome in humans, which is characterized by electrophysiologic abnormalities and hypertrophy of the heart, and also

by an increase in cardiac glycogen content [33–36] Recent studies using transgenic mice overexpressing c2 harboring these mutations showed a disease phenotype that closely mimicked the human condition [37–40] Direct binding, modeling and mutagenesis studies on AMPK c1 and c2 suggest that the mutations in CBS domains directly interfere with AMP binding, as the mutations are predicted to lie around the mouth of the AMP-binding cleft [20,41] However, the mechanism

by which the mutations lead to the observed pheno-types remains controversial

In this study, we have utilized AMPK a2b2c3 hetero-trimers expressed in Escherichia coli to characterize the enzyme kinetic parameters and regulation mechanisms for these physiologically relevant complexes We have also evaluated the impact of CBS domain mutations on a-subunit phosphorylation and AMP dependence of a2b2c3, to improve our understanding of the functional consequences of the substitutions

Results

In this work, we used AMPK a2b2c3 heterotrimers expressed in E coli to characterize the regulation of these complexes by AMP as well as to evaluate the effect of substitutions within the AMPK c3 gene on the activity of the enzyme Neumann et al [42] des-cribed gaining milligram amounts of the AMPK a1b1c1 in E coli using a tricistronic vector with T7 RNA polymerase controlling transcription of a single a1b1c1 messenger, with each subunit carrying its indi-vidual ribosome-binding site In our hands, the expres-sion of AMPK a2b2c3 in bacteria by a similar approach gave a lower yield of the trimer, due to solu-bility problems However, AMPK a2b2c3 complex could be purified from the soluble protein fraction by using a polyhistidine tag fused to a2, and the forma-tion of a stable trimer was demonstrated by SDS⁄ PAGE and western blot analysis using subunit-specific antibodies (Fig 1A,B) As E coli is not capable of phosphorylating Thr172 of the a-subunit, the bacteri-ally expressed trimers were initibacteri-ally inactive, but could

be activated through phosphorylation by the upstream kinase LKB1 (Fig 1B) After incubation with LKB1, the enzymatic activity of the bacterially expressed AMPK a2b2c3 trimers was similar to that of the

a2b2c3 complexes purified from transiently transfected

mammalian cells, with respect to both the total activity

and the AMP dependence (Fig 1C) We measured the

activity of the recombinant AMPK a2b2c3 complexes

over a range of AMP concentrations, and calculated

the concentration of AMP giving half-maximal

stimu-lation (A0.5) The bacterially expressed phosphorylated

AMPK a2b2c3 trimers had an A0.5 value of 1.9 ±

0.5 lm (Table 1), which is very similar to the A0.5for

a2b2c3 complexes expressed in COS7 cells

(2.6 ± 0.3 lm)

In order to study the effect of the mutations within AMPK c3 on the activity of the enzyme, we intro-duced several missense mutations into the c3 gene (Fig 2), and investigated the effect of these substitu-tions on the allosteric activation of the kinase by AMP Our main focus was to study the effects of mutations in positions V224, R225 and R307 in the first Bateman domain, given that substitutions in these positions occur in vivo in the pig and⁄ or human AMPK c3 genes [25,26,32] Also, the physiologic con-sequence of the R225Q mutation in the AMPK c3

A

C

D

in E coli (A) SDS ⁄ PAGE of bacterially expressed AMPK a2b2c3 complexes after purification by nickel–ion and gel filtration chroma-tography, stained with silver stain (B) Representative western blot

of AMPK a2b2c3 complexes expressed in E coli and COS7 cells using subunit-specific antibodies raised to either a2, b2, c3 or phosphorylated Thr172 in the a-subunit (apThr172) Bacterially expressed complexes were phosphorylated by incubation with an upstream kinase, LKB1 (C) The kinase activity of the AMPK a2b2c3 trimers expressed as pmoles of phosphate transferred to the SAMS peptide in the absence (open bars) or presence (black bars) of AMP (160 l M ) Data are presented as mean ± SEM of

n ¼ 4 In the activity assay, the equivalent amounts of hetero-trimers expressed in bacteria or mammalian cells were used, as estimated by western blot (B) (D) Western blot analysis of AMPK a2b2c3 complexes expressed in E coli using antibodies to a2 or apThr172 The trimers carry either a wild-type (WT) or R225Q ⁄ R454G version of c3 AMPK complexes were phosphorylated by LKB1 or CAMKKb in the absence or presence of AMP (0, 50, 200

or 500 l M AMP).

Table 1 Effect of the mutations in the AMPK c3 gene on the AMP dependence of the enzyme Fold stimulation reflects the activation

of the corresponding AMPK complexes by 200 l M AMP relative to the basal activity in the absence of added AMP (calculated from the data presented in Fig 3B) Significant difference in the AMP-sti-mulated versus basal activity for each complex was determined by two-sided Student’s t-test A0.5(concentration of AMP giving half-maximal stimulation) was calculated from the curves presented in Fig 3C (curve fitting by KALEIDAGRAPH 4.03 with the Michaelis–Men-ten equation) Values are means ± SEM NS, not significant; ND, not determined.

Mutations in c3 Fold activation by AMP P-value A0.5(l M )

gene has been extensively studied using transgenic

mouse models [30,31] Additionally, we generated

H306R, H453R, R454G and R225Q⁄ R454G

substitu-tions in c3, which occur at the posisubstitu-tions corresponding

to R225Q and R307G, in the CBS2 and CBS4

domains (Fig 2) All complexes were expressed in

E coli and purified from the soluble fraction by

nickel–ion chromatography in quantities similar

to those of the trimers containing a wild-type c3

(Fig 3A), suggesting that there was no difference in

the assembly of the different mutants into a

hetero-trimeric complex Western blot analysis showed that,

in the conditions tested, the phosphorylation level of

Thr172 of a2 by the upstream kinase LKB1 did not

differ for any of the mutants (Fig 3A) The V224I,

R225Q, H306R and R307G substitutions in AMPK c3

did not affect the fold activation of the enzyme as

compared to the wild-type trimers, when measured in

the presence of 200 lm AMP (Fig 3B, Table 1)

How-ever, the H453R and R454G substitutions substantially

reduced the AMP dependence of the enzyme down to

approximately the same level as that of the double

mutant R225Q⁄ R454G (Fig 3B, Table 1) We

calcula-ted the A0.5 values for the mutant complexes, which

showed robust activation by AMP at a concentration

of 200 lm (V224I, R225Q, H306R and R307G;

Fig 3C) using the hyperbolic curve model (Hill

coeffi-cient¼ 1) All the substitutions tested, except V224I,

increased the A0.5value for AMP as compared to the

wild-type trimers (Fig 3C, Table 1; R2‡ 0.97)

Pharmacologic activation of AMPK can be

achie-ved using

5-aminoimidazole-4-carboxamide-1-b-d-ribo-nucleoside (AICAR) Once taken up by cells, AICAR is

phosphorylated to 5-aminoimidazole-4-carboxamide

riboside monophosphate (ZMP), which mimics the effects of AMP on AMPK In our hands, 160 lm ZMP caused a two-fold activation of the bacterially expressed wild-type AMPK a2b2c3 complexes, with half-maximal stimulation (A0.5) at 61 ± 19 lm ZMP ZMP failed to significantly increase the activity of the AMPK a2b2c3 complexes with an R225Q mutation in c3, when meas-ured at the concentration of 160 lm (Fig 3C)

The presence of AMP (0–500 lm) during the phosphorylation of the bacterially expressed AMPK a2b2c3 trimers with LKB1 or CAMKKb did not increase the phosphorylation level of Thr172 of a2

as shown by western blot analysis (Fig 1D) Simi-larly, the phosphorylation of AMPK trimers con-taining R225Q⁄ R454G mutations in c3 by LKB1 or CAMKKb did not differ at varying AMP concentra-tions (Fig 1D) In line with this, we did not detect any increase in the activity of the AMPK a2b2c3 com-plexes in the presence of AMP in the phosphorylation reaction by LKB1, as measured through their ability

to incorporate phosphate into the SAMS substrate, as compared to the complexes phosphorylated in the absence of AMP (data not shown)

Discussion

We have previously shown that AMPK c3 is the pre-dominant c isoform expressed in glycolytic skeletal muscle, where it primarily forms heterotrimers with the a2 and b2 isoforms [23] The present study des-cribes bacterial expression of AMPK a2b2c3 trimers, and provides the first characterization of the enzyme kinetic parameters and regulation mechanisms for these physiologically relevant complexes

Fig 2 An alignment of the amino acids in the four CBS domains in human AMPK c1, c2 and c3 Sequences were aligned using

CLUSTALX (1.83) A versatile coloring scheme was incorporated to highlight conserved fea-tures in the alignment, using a default color code of CLUSTALX Above the alignment is the conservation score generated by the program The enlargement shows details of the mutations, with red borders indicating the mutations that have been described in vivo in pigs (V224I and R225Q in c3) or humans (mutations in c2, R225W and R307C in c3), and red letters indicating the mutations analyzed in the present study.

The particular subunit isoforms present in the

AMPK complex determine the concentration of AMP

causing a half-maximal increase in the total activity as

well as the fold stimulation by AMP [42–46] Our

experiments show that AMPK a2b2c3 trimers are

acti-vated by AMP with a half-maximal stimulation (A0.5)

at 1.9 ± 0.5 or 2.6 ± 0.3 lm, as measured for

com-plexes expressed in E coli or mammalian cells,

respect-ively (Fig 3C, Table 1) Typically, 200 lm AMP

caused an activation of 2–4-fold of the AMPK a2b2c3

trimers expressed in bacteria or COS7 cells, relative to

the activity in the absence of AMP Previously,

allos-teric regulation of c3-containing AMPK complexes by

AMP has been questioned, on the basis of activity

measurements on rat brain protein immunoprecipitates

using antibodies to c3 [17] One plausible reason for

these discrepancies may be a poor specificity of the

antibodies applied, in combination with the fact that

protein lysate from brain was used, as we and others

have not been able to detect any c3 mRNA or protein

in this tissue [23,25] In addition, it is possible that the

antibody binding itself interferes with the

AMP-bind-ing properties of c3 Additionally, the low A0.5 value

characterizing the activation of a2b2c3 trimer by AMP

makes the system highly sensitive to the possible

pres-ence of any traces of AMP, either endogenous or as a

result of the presence of AMP-generating proteins as contaminants [47]

Several upstream kinases (AMPKKs) have been reported to activate mammalian AMPK through phos-phorylation of Thr172 in the a-subunit, i.e LKB1 [11–13], CAMKKb [15,48] and TAK1 [16] LKB1 is believed to be the major upstream kinase for AMPK

in skeletal muscle, as knocking out LKB1 almost com-pletely prevents both AICAR- and contraction-induced a2-AMPK signaling [49] Initially, AMP was thought

to increase phosphorylation of AMPK by AMPKK both by direct activation of the upstream kinase and

by making the AMPK a better substrate for AMPKK, through binding to the c-subunits [50] However, sev-eral recent studies have challenged this notion [15,48], and the role of AMP in phosphorylation of AMPK has remained controversial While this manuscript was

A

B

C

Fig 3 Effect of the substitutions in the AMPK c3 gene on the

activity of the enzyme (A) AMPK a2b2c3 complexes with either

wild-type c3 (WT) or mutant c3 harboring the indicated

substitu-tions were expressed in E coli, and equal amounts of the trimers

were phosphorylated in vitro by LKB1 Representative western

blots probed with antibodies to a2, b2, c3 or apThr172 are shown.

(B) Activity of the AMPK a2b2c3 complexes was measured by a

radioactive filter paper assay using SAMS substrate in the absence

(open bars) or presence (black bars) of AMP (200 l M ) The data

pre-sented are the mean ± SEM from three to four independent

phos-phorylation experiments (n ¼ 2 in the activity assay), expressed

relative to the activity of the wild-type AMPK complexes measured

in the absence of AMP (the activity of the wild-type complex

with-out AMP is set to 1) (C) The activity of the wild-type and mutated

AMPK a2b2c3 complexes was measured over a range of AMP

(from 0 to 160 l M ) or ZMP (from 0 to 160 l M ) concentrations The

data are expressed relative to the basal activity in the absence of

added AMP for each complex (n ¼ 3 in the activity assay; basal

activity is set to 0) The graph is plotted in KALEIDAGRAPH 4.03

(Syn-ergy Software), using curve fitting against the one-site Michaelis–

Menten equation; V ¼ V max · [S] ⁄ (A 0.5 + [S]), where V max is the

maximal activation, A0.5¼ [AMP] at 50% AMPK activation, and

V ¼ AMPK activity Given the very low level or lack of activation by

AMP, we were unable to measure A 0.5 values for H453R, R454G

and R225Q ⁄ R454G All the AMPK complexes were expressed in

E coli except for the wild-type trimers denoted by asterisks, which

were expressed in COS7 cells In (B) and (C), equivalent amounts

of AMPK were used, as estimated by western blot analysis (A).

in preparation, two studies reported that AMP failed

to promote phosphorylation of c1-containing AMPK

trimers, when recombinant upstream kinase and

AMPK preparations were used [44,47] In line with

these studies, we did not detect any increase in

phos-phorylation of a2 in a2b2c3 complexes by LKB1 or

CAMKKb in the presence of increasing concentrations

of AMP (Fig 1D)

Despite the considerable effort invested in this area,

the nature of the mutations in the CBS domains of the

AMPK c3 gene has remained controversial A previous

report provided evidence that the activity of AMPK

was reduced in skeletal muscle of c3 R225Q mutant

pigs [25] In resting muscle from c3 R225Q mutant

mice, AMPK activity was reported to be unaltered [30]

or reduced [31] However, in vivo activity measurements

are complicated by the potential inhibitory effects of

glycogen overload, which characterizes skeletal muscle

of mice and pigs carrying the c3 R225Q substitution,

on AMPK activation Therefore, the effect of the

mutation has been addressed in mammalian cells In

COS7 cells transfected with plasmids encoding a2b2c3

R225Q or V224I mutants, both substitutions resulted

in diminished AMP dependence of AMPK, as

com-pared to the wild-type trimers In addition, AMPK

total activity and phosphorylation of a2 were shown to

be markedly elevated in cells expressing a2b2c3

R225Q, as compared to the wild-type or

V224I-con-taining trimers [30] However, in our laboratory we

have failed to detect any increase in the basal activity

for R225Q complexes assayed from material purified

from transiently transfected COS7 cells (data not

shown) It has to be noted that the activity of the

AMPK is highly sensitive to the lysis protocol used

(hypoxia, glucose deprivation and mechanical stress

during the cell harvest activate AMPK), which

poten-tially complicates the comparisons of the results from

different research groups and may partly explain the

discrepancies observed In an effort to mimic the

R225Q mutation in c3, the equivalent position has

been mutated in AMPK c1 (R70Q) as well as in c2

(R302Q) [41,44,45,51] In the present study, we

evalu-ated the impact of CBS domain mutations (locations of

the substitutions are shown in Fig 2) on the AMP

dependence of the AMPK a2b2c3 using the enzyme

expressed in E coli, activated in vitro by an LKB1

pre-paration Importantly, we did not detect any difference

in phosphorylation of Thr172 of a2 by LKB1 when we

compared the different mutants to the wild-type trimers

(Fig 3A), although we have to acknowledge that the

semiquantitative nature of the western blot technique

makes the exact measurement difficult It is noteworthy

that for the mutations in the N-terminal Bateman

domain formed by CBS1–2 (V224I, R225Q, H306R and R307G), the fold activation of the complex by

200 lm AMP was similar to that of wild-type com-plexes (Fig 3B, Table 1) Mutations in the C-terminal Bateman domain formed by CBS3–4 (H453R and R454G), on the other hand, significantly reduced AMP stimulation and resulted in AMP dependence curves highly similar to that of the double mutant R225Q⁄ R454G (Fig 3B, Table 1) The R225Q, H306R and R307G mutations substantially increased the A0.5value for AMP as compared to the wild-type trimers (Fig 3C, Table 1) The V224I mutation, on the other hand, resulted in AMP dependence curves that were very similar to those of wild-type complexes (Fig 3C, Table 1) According to the current model of the allos-teric control of mammalian AMPK, the N-terminal and C-terminal pairs of CBS domains both bind one molecule of AMP [20,41] However, the different effects

on AMP activation observed in this study of substitu-tions in corresponding posisubstitu-tions of the two Bateman domains (H306R and R307G in CBS2 are equivalent

to H453R and R454G in CBS4, respectively, Fig 2) indicate that the two suggested AMP-binding sites in c3 are not functionally equivalent Notably, the AMP dependence curves of the AMPK a2b2c3 complexes appear hyperbolic rather than sigmoid (Fig 3C) While this manuscript was in preparation, a crystal struc-ture of the Schizosaccharomyces pombe AMPK was reported, demonstrating that one AMP molecule bound to a single site formed by the CBS domains of the c-subunit [52] Nevertheless, the absence of regula-tion of Saccharomyces cerevisiae AMPK activity by AMP [53,54] raises the possibility that nucleotide bind-ing in Sc pombe AMPK may differ from that in the human enzyme The present study does not explain the mechanisms by which the mutations in CBS domains

of AMPK c3 interfere with activation by AMP However, our data would be consistent with a single AMP-binding site being present in c3

The A0.5 value for ZMP was increased for the R225Q AMPK complexes as compared to the wild-type heterotrimers, which was expected, as ZMP is thought to bind to the same site as AMP and activate AMPK in a similar manner This result may explain the impairment of AICAR-stimulated muscle glucose uptake described in R225Q transgenic mice, as com-pared to the wild-type littermates [30]

The present study does not answer the question of how the reduced AMP dependence of the R225Q vari-ants of c3 would lead to the increased glycogen content described in the skeletal muscle of the carriers of this mutation [25,30] Also, we did not detect any difference

in the regulation of AMPK complexes containing a

V224I form of c3 by AMP, whereas this mutation is

known to lead to reduced skeletal muscle glycogen

stores [26] Clearly, an important challenge is to

deter-mine the effect of the mutations in the CBS domains

on AMPK activity in the presence of physiologic AMP

and ATP concentrations, which is technically difficult

The in vitro AMPK activity assay is commonly

per-formed at 200 lm ATP, utilizing radiolabeled ATP,

which causes technical problems when the

concentra-tion is increased to close to physiologic levels [55,56]

The possibility of altered functionality of the AMPK

a2b2c3 trimers expressed in E coli, as compared to the

in vivocomplexes, also has to be acknowledged Native

AMPK is known to be post-translationally modified at

multiple sites, other than Thr172, in the a-subunit

[13,46,57], and the possibility that the absence of these

modifications has an impact on the activity of the

mutant complexes cannot be excluded Another

possi-bility would be that interactions other than with

AMP⁄ ATP, occurring in eukaryotic cells only, have an

impact on the activity of the enzyme Further studies

in relevant cell systems as well as in animal models are

required to investigate these issues

Experimental procedures

Expression and purification of recombinant

AMPK a2b2c3 in E coli

The tricistronic plasmids containing the three AMPK

sub-units a2 (mouse, accession number NP_835279) with

N-ter-minal polyhistidine tag, b2 (human, accession number

O43741) and c3 (human, accession number Q9UGI9) in

pET vector (Novagen, Madison, WI), were transformed

into competent host cells (E coli BL21 DE3 pLysS;

Nov-agen) Single colonies were used to inoculate 5 mL of LB

medium containing 100 lgÆmL)1 carbenicillin, 50 lgÆmL)1

chloramphenicol and 0.5% glucose Following incubation

in a shaker incubator (250 r.p.m.) overnight at 37C, the

starter cultures were used to inoculate 0.5 L of the above

medium Protein expression was induced with 1 mm

isopro-pyl thio-b-d-galactoside (final concentration) at an A600 of

0.5–0.7, and cultures were grown for an additional 3 h at

37C Cells were harvested (4000 r.p.m for 15 min at

4C; Allegra 6R), and the cell pellet was resuspended in

the lysis buffer containing 50 mm Hepes (pH 7.5), 50 mm

NaF, 5 mm sodium pyrophosphate, 1 mm EDTA, 1 mm

di-thiothreitol, 10% glycerol, 1% Triton X-100, lysozyme

(48 lgÆmL)1), and complete protease inhibitor cocktail

(Roche Diagnostics GmbH, Mannheim, Germany), and

placed on ice for 1 h before being sonicated (Dr Hielscher

GmbH, Teltow, Germany) for 3 min Insoluble material

was removed by centrifugation [20 400 g for 20 min at 4C

using an Avanti J-20 XP centrifuge (Beckman Coulter),

rotor type JA25.50] The supernatant was collected and diluted in buffer A containing 20 mm Tris⁄ HCl (pH 8.0),

150 mm NaCl, 15 mm imidazole, and 10% glycerol, and loaded onto a 5 mL His-Trap column (GE Healthcare, Uppsala, Sweden) Bound protein was eluted using a gradi-ent with buffer B (idgradi-entical to buffer A except for contain-ing 500 mm imidazole), and stored at ) 20 C until use In some cases, the AMPK complex was further purified by gel filtration chromatography with a Superdex 200 pg HiLoad

16⁄ 60 column (GE Healthcare) connected to A¨kta FPLC (GE Healthcare)

Expression and purification of recombinant AMPK a2b2c3 in COS7 cells

COS7 cells were cotransfected with cDNAs encoding mouse AMPK a2, b2 (human) and c3 [human, all cloned into pcDNA3.1(+); Invitrogen, Paisley, UK] using Lipofectin reagent, according to the manufacturer’s instructions (Invi-trogen) Immediately prior to the lysis, the cells were sub-mitted to hyperosmotic stress by incubating them for 30 min with sorbitol (final concentration 0.6 m) in the culture med-ium Cells were harvested 48 h post-transfection by rapid lysis: the medium was removed, and ice-cold lysis buffer (see above) was added Insoluble material was removed by centrifugation [15 700 g for 20 min at 4C using a 5415R centrifuge (Eppendorf), rotor type, 24 places fixed angle], and AMPK was partially purified from the soluble fraction using a DEAE–sepharose ion-exchange step (GE Health-care) The major part of the endogenous AMPK was deple-ted by immunoprecipitation of the relevant fractions with antibody to a1 prebound to protein A–sepharose as described previously [23]

Site-directed mutagenesis The mutations V224I, R225Q, H306R, R307G, H453R, R454G and R225Q⁄ R454G were introduced into AMPK c3 using the Quick Change II site-directed mutagenesis kit (Stratagene, La Jolla, CA) Mutagenesis primers were designed using Stratagene’s Tm calculator (primer seq-uences are available on request) Point mutations generated

in vitrowere confirmed by DNA sequencing

Western blotting Quantitative analysis of the expression of different AMPK subunits was performed as described previously [23]

Phosphorylation of bacterial trimer and AMPK activity assay

To activate the bacterial AMPK, recombinant trimers were incubated with LKB1 or CAMKKb in the presence of

0.2 mm ATP, 5 mm MgCl2and 1 mm dithiothreitol for 1 h

at 32C in a thermostated shaker Rat liver LKB1 was

purified as described previously [12] up to the Q-sepharose

step, except that the poly(ethylene glycol) precipitation of

the liver lysate was omitted Bacterially expressed

CAMKKb was a gift from D Carling [14,15] To determine

the relevant amount of upstream kinase versus AMPK to

be used in the phosphorylation reaction, the AMPK trimers

were incubated with increasing concentrations of AMPKK,

and the amount of upstream kinase corresponding to the

highest possible level of Thr172 phosphorylation was used

in the reaction AMPK activity was determined by in vitro

phosphorylation of the SAMS (HMRSAMSGLHLVKRR)

synthetic peptide substrate as previously described [58]

Briefly, kinase reactions were initiated by adding 4 lL of

the phosphorylated AMPK to 21 lL of assay buffer

con-taining 50 mm Hepes (pH 7.5), 80 mm NaCl, 8% glycerol,

5 mm MgCl2, 0.8 mm EDTA, 0.8 mm dithiothreitol,

0.2 mm ATP, 0.2 mm SAMS peptide, and [32P]ATP (final

concentration 0.016 lCiÆlL)1) in the presence or absence of

varying concentrations of AMP, as described in the figure

legends Reactions were incubated on a vibrating platform

for 3 h at 37C The reactions were terminated by adding

trichloroacetic acid to a final concentration of 13%,

fol-lowed by centrifugation Immediately thereafter, aliquots

were spotted onto Whatman P81 paper, and washed with

1% phosphoric acid, and the incorporation of 32P into

peptide substrate was measured in a liquid scintillation

counter

Acknowledgements

We gratefully acknowledge Professor David Carling

(Medical Research Council Clinical Sciences Center,

Imperial College, Hammersmith Hospital, London,

UK) for the recombinant CAMKKb preparation used

in this study and for discussions concerning the work

presented This work was supported by an Integrated

Project from the European Commission

(LSHG-CT-2005-518181)

References

1 Carling D (2005) AMP-activated protein kinase:

balan-cing the scales Biochimie 87, 87–91

2 Hardie DG, Hawley SA & Scott JW (2006)

AMP-acti-vated protein kinase-development of the energy sensor

concept J Physiol 574, 7–15

3 Kahn BB, Alquier T, Carling D & Hardie DG (2005)

AMP-activated protein kinase: ancient energy gauge

provides clues to modern understanding of metabolism

Cell Metab 1, 15–25

4 Hayashi T, Hirshman MF, Kurth EJ, Winder WW &

Goodyear LJ (1998) Evidence for 5¢-AMP-activated

protein kinase mediation of the effect of muscle contrac-tion on glucose transport Diabetes 47, 1369–1373

5 Merrill GF, Kurth EJ, Hardie DG & Winder WW (1997) AICA riboside increases AMP-activated protein kinase, fatty acid oxidation, and glucose uptake in rat muscle Am J Physiol 273, E1107–E1112

6 Minokoshi Y, Kim YB, Peroni OD, Fryer LG, Muller

C, Carling D & Kahn BB (2002) Leptin stimulates fatty-acid oxidation by activating AMP-activated pro-tein kinase Nature 415, 339–343

7 Ruderman NB, Saha AK & Kraegen EW (2003) Mini-review: malonyl CoA, AMP-activated protein kinase, and adiposity Endocrinology 144, 5166–5171

8 Kemp BE, Stapleton D, Campbell DJ, Chen ZP, Murthy S, Walter M, Gupta A, Adams JJ, Katsis F, van Denderen B et al (2003) AMP-activated protein kinase, super metabolic regulator Biochem Soc Trans

31, 162–168

9 Stapleton D, Woollatt E, Mitchelhill KI, Nicholl JK, Fernandez CS, Michell BJ, Witters LA, Power DA, Sutherland GR & Kemp BE (1997) AMP-activated pro-tein kinase isoenzyme family: subunit structure and chromosomal location FEBS Lett 409, 452–456

10 Kemp BE, Mitchelhill KI, Stapleton D, Michell BJ, Chen ZP & Witters LA (1999) Dealing with energy demand: the AMP-activated protein kinase Trends Bio-chem Sci 24, 22–25

11 Shaw RJ, Kosmatka M, Bardeesy N, Hurley RL, Witters LA, Depinho RA & Cantley LC (2004) The tumor suppressor LKB1 kinase directly activates AMP-activated kinase and regulates apoptosis in response to energy stress Proc Natl Acad Sci USA 101, 3329–3335

12 Hawley SA, Boudeau J, Reid JL, Mustard KJ, Udd L, Makela TP, Alessi DR & Hardie DG (2003) Complexes between the LKB1 tumor suppressor, STRAD alpha⁄ beta and MO25 alpha⁄ beta are upstream kinases in the AMP-activated protein kinase cascade J Biol 2, doi:10.1186/1475-4924-2-28

13 Woods A, Johnstone SR, Dickerson K, Leiper FC, Fryer LG, Neumann D, Schlattner U, Wallimann T, Carlson M & Carling D (2003) LKB1 is the upstream kinase in the AMP-activated protein kinase cascade Curr Biol 13, 2004–2008

14 Hurley RL, Anderson KA, Franzone JM, Kemp BE, Means AR & Witters LA (2005) The Ca2+⁄ calmodu-lin-dependent protein kinase kinases are AMP-activated protein kinase kinases J Biol Chem 280, 29060–29066

15 Woods A, Dickerson K, Heath R, Hong SP, Momcilo-vic M, Johnstone SR, Carlson M & Carling D (2005) Ca2+⁄ calmodulin-dependent protein kinase kinase-beta acts upstream of AMP-activated protein kinase in mam-malian cells Cell Metab 2, 21–33

16 Momcilovic M, Hong SP & Carlson M (2006) Mamma-lian TAK1 activates Snf1 protein kinase in yeast and

phosphorylates AMP-activated protein kinase in vitro.

J Biol Chem 281, 25336–25343

17 Cheung PC, Salt IP, Davies SP, Hardie DG & Carling

D (2000) Characterization of AMP-activated protein

kinase gamma-subunit isoforms and their role in AMP

binding Biochem J 346 (Part 3), 659–669

18 Bateman A (1997) The structure of a domain common

to archaebacteria and the homocystinuria disease

pro-tein Trends Biochem Sci 22, 12–13

19 Zhang R, Evans G, Rotella FJ, Westbrook EM, Beno

D, Huberman E, Joachimiak A & Collart FR (1999)

Characteristics and crystal structure of bacterial

inosine-5¢-monophosphate dehydrogenase Biochemistry 38,

4691–4700

20 Scott JW, Hawley SA, Green KA, Anis M, Stewart G,

Scullion GA, Norman DG & Hardie DG (2004) CBS

domains form energy-sensing modules whose binding of

adenosine ligands is disrupted by disease mutations

J Clin Invest 113, 274–284

21 Kemp BE (2004) Bateman domains and adenosine

derivatives form a binding contract J Clin Invest 113,

182–184

22 Woods A, Cheung PC, Smith FC, Davison MD, Scott

J, Beri RK & Carling D (1996) Characterization of

AMP-activated protein kinase beta and gamma

subu-nits Assembly of the heterotrimeric complex in vitro

J Biol Chem 271, 10282–10290

23 Mahlapuu M, Johansson C, Lindgren K, Hjalm G,

Barnes BR, Krook A, Zierath JR, Andersson L &

Marklund S (2004) Expression profiling of the

gamma-subunit isoforms of AMP-activated protein kinase

suggests a major role for gamma3 in white skeletal

muscle Am J Physiol Endocrinol Metab 286, E194–E200

24 Birk JB & Wojtaszewski JF (2006) Predominant

{alpha}2⁄ {beta}2 ⁄ {gamma}3 AMPK activation during

exercise in human skeletal muscle J Physiol 577,

1021–1032

25 Milan D, Jeon JT, Looft C, Amarger V, Robic A,

Thelander M, Rogel-Gaillard C, Paul S, Iannuccelli N,

Rask L et al (2000) A mutation in PRKAG3 associated

with excess glycogen content in pig skeletal muscle

Sci-ence 288, 1248–1251

26 Ciobanu D, Bastiaansen J, Malek M, Helm J, Woollard

J, Plastow G & Rothschild M (2001) Evidence for new

alleles in the protein kinase adenosine

monophosphate-activated gamma(3)-subunit gene associated with low

glycogen content in pig skeletal muscle and improved

meat quality Genetics 159, 1151–1162

27 Estrade M, Vignon X, Rock E & Monin G (1993)

Glycogen hyperaccumulation in white muscle fibres

of RN- carrier pigs A biochemical and ultrastructural

study Comp Biochem Physiol B 104, 321–326

28 Lebret B, Le Roy P, Monin G, Lefaucheur L, Caritez

JC, Talmant A, Elsen JM & Sellier P (1999) Influence

of the three RN genotypes on chemical composition,

enzyme activities, and myofiber characteristics of por-cine skeletal muscle J Anim Sci 77, 1482–1489

29 Estrade M, Ayoub S, Talmant A & Monin G (1994) Enzyme activities of glycogen metabolism and mito-chondrial characteristics in muscles of RN- carrier pigs (Sus scrofa domesticus) Comp Biochem Physiol Biochem Mol Biol 108, 295–301

30 Barnes BR, Marklund S, Steiler TL, Walter M, Hjalm

G, Amarger V, Mahlapuu M, Leng Y, Johansson C, Galuska D et al (2004) The 5¢-AMP-activated protein kinase gamma3 isoform has a key role in carbohydrate and lipid metabolism in glycolytic skeletal muscle J Biol Chem 279, 38441–38447

31 Yu H, Hirshman MF, Fujii N, Pomerleau JM, Peter LE

& Goodyear LJ (2006) Muscle-specific overexpression of wild-type and R225Q mutant AMP-activated protein kinase gamma3-subunit differentially regulates glycogen accumulation Am J Physiol Endocrinol Metab 291, E557–E565

32 Costford SR, Chaudhry SN, Kavaslar N, Ahitus N, Pennacchio LA, Dent R, Mcpherson R & Harper M-E (2006) Novel mutations in the gamma-3 subunit of AMP-activated protein kinase affect protein function in humans In 13th Biochemistry of Exercise International Conference, Seoul, Korea, Abstract P010

33 Gollob MH, Seger JJ, Gollob TN, Tapscott T, Gonzales

O, Bachinski L & Roberts R (2001) Novel PRKAG2 mutation responsible for the genetic syndrome of ventri-cular preexcitation and conduction system disease with childhood onset and absence of cardiac hypertrophy Circulation 104, 3030–3033

34 Gollob MH, Green MS, Tang AS, Gollob T, Karibe A, Ali Hassan AS, Ahmad F, Lozado R, Shah G,

Fananapazir L et al (2001) Identification of a gene responsible for familial Wolff–Parkinson–White syn-drome N Engl J Med 344, 1823–1831

35 Arad M, Benson DW, Perez-Atayde AR, Mckenna WJ, Sparks EA, Kanter RJ, Mcgarry K, Seidman JG & Seidman CE (2002) Constitutively active AMP kinase mutations cause glycogen storage disease mimicking hypertrophic cardiomyopathy J Clin Invest 109, 357–362

36 Blair E, Redwood C, Ashrafian H, Oliveira M, Brox-holme J, Kerr B, Salmon A, Ostman-Smith I & Watkins

H (2001) Mutations in the gamma(2) subunit of AMP-activated protein kinase cause familial hypertrophic cardiomyopathy: evidence for the central role of energy compromise in disease pathogenesis Hum Mol Genet 10, 1215–1220

37 Arad M, Moskowitz IP, Patel VV, Ahmad F, Perez-Atayde AR, Sawyer DB, Walter M, Li GH, Burgon

PG, Maguire CT et al (2003) Transgenic mice over-expressing mutant PRKAG2 define the cause of Wolff–Parkinson–White syndrome in glycogen storage cardiomyopathy Circulation 107, 2850–2856

38 Sidhu JS, Rajawat YS, Rami TG, Gollob MH, Wang

Z, Yuan R, Marian AJ, Demayo FJ, Weilbacher D,

Taffet GE et al (2005) Transgenic mouse model of

ventricular preexcitation and atrioventricular reentrant

tachycardia induced by an AMP-activated protein

kinase loss-of-function mutation responsible for Wolff–

Parkinson–White syndrome Circulation 111, 21–29

39 Davies JK, Wells DJ, Liu K, Whitrow HR, Daniel TD,

Grignani R, Lygate CA, Schneider JE, Noel G, Watkins

H et al (2006) Characterization of the role of gamma2

R531G mutation in AMP-activated protein kinase in

cardiac hypertrophy and Wolff–Parkinson–White

syn-drome Am J Physiol Heart Circ Physiol 290, H1942–

H1951

40 Zou L, Shen M, Arad M, He H, Lofgren B, Ingwall JS,

Seidman CE, Seidman JG & Tian R (2005) N488I

mutation of the gamma2-subunit results in bidirectional

changes in AMP-activated protein kinase activity Circ

Res 97, 323–328

41 Adams J, Chen ZP, Van Denderen BJ, Morton CJ,

Parker MW, Witters LA, Stapleton D & Kemp BE

(2004) Intrasteric control of AMPK via the gamma1

subunit AMP allosteric regulatory site Protein Sci 13,

155–165

42 Neumann D, Woods A, Carling D, Wallimann T &

Schlattner U (2003) Mammalian AMP-activated protein

kinase: functional, heterotrimeric complexes by

co-expression of subunits in Escherichia coli Protein

Expr Purif 30, 230–237

43 Scott JW, Norman DG, Hawley SA, Kontogiannis L &

Hardie DG (2002) Protein kinase substrate recognition

studied using the recombinant catalytic domain of

AMP-activated protein kinase and a model substrate

J Mol Biol 317, 309–323

44 Sanders MJ, Grondin PO, Hegarty BD, Snowden MA

& Carling D (2006) Investigating the mechanism for

AMP activation of the AMP-activated protein kinase

cascade Biochem J 403, 139–148

45 Daniel T & Carling D (2002) Functional analysis of

mutations in the gamma 2 subunit of AMP-activated

protein kinase associated with cardiac hypertrophy and

Wolff–Parkinson–White syndrome J Biol Chem 277,

51017–51024

46 Stein SC, Woods A, Jones NA, Davison MD & Carling

D (2000) The regulation of AMP-activated protein

kinase by phosphorylation Biochem J 345 (Part 3),

437–443

47 Suter M, Riek U, Tuerk R, Schlattner U, Wallimann T

& Neumann D (2006) Dissecting the role of 5¢-AMP

for allosteric stimulation, activation, and deactivation

of AMP-activated protein kinase J Biol Chem 281,

32207–32216

48 Hawley SA, Pan DA, Mustard KJ, Ross L, Bain J, Edelman AM, Frenguelli BG & Hardie DG (2005) Calmodulin-dependent protein kinase kinase-beta is an alternative upstream kinase for AMP-activated protein kinase Cell Metab 2, 9–19

49 Sakamoto K, Mccarthy A, Smith D, Green KA, Grahame Hardie D, Ashworth A & Alessi DR (2005) Deficiency of LKB1 in skeletal muscle prevents AMPK activation and glucose uptake during contraction EMBO J 24, 1810–1820

50 Hawley SA, Selbert MA, Goldstein EG, Edelman AM, Carling D & Hardie DG (1995) 5¢-AMP activates the AMP-activated protein kinase cascade, and Ca2+⁄ calmodulin activates the calmodulin-dependent protein kinase I cascade, via three independent mechanisms

J Biol Chem 270, 27186–27191

51 Hamilton SR, Stapleton D, O’Donnell JB Jr, Kung JT, Dalal SR, Kemp BE & Witters LA (2001) An activating mutation in the gamma1 subunit of the AMP-activated protein kinase FEBS Lett 500, 163–168

52 Townley R & Shapiro L (2007) Crystal structures of the adenylate sensor from fission yeast AMP-activated pro-tein kinase Science 315, 1726–1729

53 Wilson WA, Hawley SA & Hardie DG (1996) Glucose repression⁄ derepression in budding yeast: SNF1 protein kinase is activated by phosphorylation under derepress-ing conditions, and this correlates with a high

AMP:ATP ratio Curr Biol 6, 1426–1434

54 Woods A, Munday MR, Scott J, Yang X, Carlson M & Carling D (1994) Yeast SNF1 is functionally related to mammalian AMP-activated protein kinase and regulates acetyl-CoA carboxylase in vivo J Biol Chem 269, 19509–19515

55 Bergeron R, Russell RR 3rd, Young LH, Ren JM, Marcucci M, Lee A & Shulman GI (1999) Effect of AMPK activation on muscle glucose metabolism in conscious rats Am J Physiol 276, E938–E944

56 Sabina RL, Kernstine KH, Boyd RL, Holmes EW & Swain JL (1982) Metabolism of 5-amino-4-imidazole-carboxamide riboside in cardiac and skeletal muscle Effects on purine nucleotide synthesis J Biol Chem 257, 10178–10183

57 Mitchelhill KI, Michell BJ, House CM, Stapleton D, Dyck J, Gamble J, Ullrich C, Witters LA & Kemp BE (1997) Posttranslational modifications of the 5¢-AMP-activated protein kinase beta1 subunit J Biol Chem 272, 24475–24479

58 Derave W, Ai H, Ihlemann J, Witters LA, Kristiansen

S, Richter EA & Ploug T (2000) Dissociation of AMP-activated protein kinase activation and glucose transport in contracting slow-twitch muscle Diabetes

49, 1281–1287