Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK α2β2γ1 by the Glucose Importagog SC4

Article

Structural Determinants for Small-Molecule

Activation of Skeletal Muscle AMPK a2b2g1 by the

Glucose Importagog SC4

Graphical Abstract

Authors

Kevin R.W. Ngoei,

Christopher G. Langendorf,

Naomi X.Y. Ling, ..., Bruce E. Kemp,

Jonathan B. Baell, Jonathan S. Oakhill

Correspondence

clangendorf@svi.edu.au (C.G.L.),

joakhill@svi.edu.au (J.S.O.)

In Brief

Therapeutic activation of the metabolic

regulator AMPK in skeletal muscle is a

validated strategy to combat type 2

diabetes. Ngoei et al. have solved the

crystal structure of the activator SC4

complexed to a skeletal muscle AMPK

isoform, identifying important binding

determinants that will advance

development of AMPK-targeting

therapeutics.

Highlights

d

SC4 is a potent allosteric activator of AMPK complexes

containing the b2 isoform

SC4 stimulates b2-AMPK in cells, and glucose uptake by

isolated skeletal muscle

Binding to b2-AMPK is mediated by 4⁰-nitrogen in the SC4

core and b2 residue Asp111

We identiﬁed an interaction between b2 N terminus and

AMPK autoinhibitory domain

Ngoei et al., 2018, Cell Chemical Biology 25, 1–10

June 21, 2018 ª 2018 Elsevier Ltd.

https://doi.org/10.1016/j.chembiol.2018.03.008

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

Cell Chemical Biology

Article

Structural Determinants for Small-Molecule

Activation of Skeletal Muscle AMPK a2b2g1

by the Glucose Importagog SC4

2

3

Kevin R.W. Ngoei,^1,12Christopher G. Langendorf,^1,12,Naomi X.Y. Ling, Ashfaqul Hoque, Swapna Johnson,

*

Michelle C. Camerino,^3,4Scott R. Walker,^3,4Ylva E. Bozikis,^3,4Toby A. Dite,²Ashley J. Ovens,^2,5William J. Smiles,²

Roxane Jacobs,⁶He Huang,⁷Michael W. Parker,^8,9John W. Scott,^1,5Mark H. Rider,⁶Richard C. Foitzik,^4,10

Bruce E. Kemp,^1,5Jonathan B. Baell,^3,11and Jonathan S. Oakhill^2,5,13,

*

¹Protein Chemistry & Metabolism, St. Vincent’s Institute of Medical Research, University of Melbourne, Fitzroy, VIC 3065, Australia

²Metabolic Signalling Laboratory, St. Vincent’s Institute of Medical Research, University of Melbourne, Fitzroy, VIC 3065, Australia

³Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, VIC 3052, Australia

⁴Cancer Therapeutics CRC (CTx) Pty Ltd, Level 3, 343 Royal Parade, Parkville, VIC 3052, Australia

⁵Mary MacKillop Institute for Health Research, Australian Catholic University, Melbourne, VIC 3000, Australia

6

´

Universite catholique de Louvain (UCL), de Duve Institute, Avenue Hippocrate 75 bte 74.02, 1200 Brussels, Belgium

⁷Jiangsu National Synergistic Innovation Center for Advanced Materials (SICAM), School of Pharmaceutical Sciences, Nanjing Tech

University, No. 30 South Puzhu Road, Nanjing 211816, People’s Republic of China

⁸ACRF Rational Drug Discovery Centre, St. Vincent’s Institute of Medical Research, University of Melbourne, Fitzroy, VIC 3065,

Australia

⁹Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, Parkville,

VIC 3052, Australia

¹⁰MecRX Pty Ltd, Level 9, 31 Queen Street, Melbourne, VIC 3000, Australia

¹¹School of Pharmaceutical Sciences, Nanjing Tech University, No. 30 South Puzhu Road, Nanjing 211816, People’s Republic of China

¹²These authors contributed equally

¹³Lead Contact

*Correspondence: clangendorf@svi.edu.au (C.G.L.), joakhill@svi.edu.au (J.S.O.)

https://doi.org/10.1016/j.chembiol.2018.03.008

SUMMARY

INTRODUCTION

The AMP-activated protein kinase (AMPK) abg het- Type 2 diabetes (T2D) is a progressive disease, characterized in

its early stages by varying degrees of pancreatic b cell dysfunc-

ꢀ

tion and insulin resistance. Insulin secretagogs (Greek agogos;

erotrimer regulates cellular energy homeostasis

with tissue-speciﬁc isoform distribution. Small-

molecule activation of skeletal muscle a2b2 AMPK

complexes may prove a valuable treatment strategy

for type 2 diabetes and insulin resistance. Herein,

we report the small-molecule SC4 is a potent, direct

AMPK activator that preferentially activates a2 com-

plexes and stimulates skeletal muscle glucose

uptake. In parallel with the term secretagog, we

propose ‘‘importagog’’ to deﬁne a substance that

leading), such as the sulfonylureas (SUs) and glinides, which

stimulate insulin secretion from remaining functional b cells,

are widely used to overcome initial decline in production

(Hanefeld, 2007; Carpio and Fonseca, 2014; Thrasher, 2017).

However, chronic SU use is associated with b cell apoptosis,

exhaustion, and desensitization, and retention of glycemic con-

trol over time usually requires progression to combinatorial

therapy. Other complications, including weight gain, cardiotox-

icity, and hypoglycemia, warrant the continued development of

induces or augments cellular uptake of another sub- therapies with alternative modes of action to SUs. Peripheral in-

sulin sensitization by the thiazolidinediones (TZDs), for example,

is one such successful strategy. Another emerging strategy for

stance. Three-dimensional structures of the glucose

importagog SC4 bound to activated a2b2g1 and

a2b1g1 complexes reveal binding determinants, in

particular a key interaction between the SC4 imida-

glycemic control, underpinned by its implicated roles in exer-

cise-induced glucose clearance, is targeting of AMP-activated

protein kinase (AMPK) to promote glucose disposal and storage

zopyridine 4⁰-nitrogen and b2-Asp111, which pro-

in skeletal muscle (O’Neill, 2013; Cokorinos et al., 2017; Myers

vide a design paradigm for b2-AMPK therapeutics.

et al., 2017).

The a2b2g1/SC4 structure reveals an interaction

between a b2 N-terminal a helix and the a2 autoin-

AMPK is a myristoylated abg heterotrimer that plays an impor-

tant role in regulating metabolism in eukaryotes by balancing

hibitory domain. Our results provide a structure-

function guide to accelerate development of

nutrient supply with energy demand (reviewed in Hardie, 2015;

Oakhill et al., 2012). Elevations in cellular AMP/ATP and ADP/

potent, but importantly tissue-speciﬁc, b2-AMPK ATP ratios, induced by energy stress, are detected by AMPK

and promote phosphorylation of the AMPK catalytic a-subunit

therapeutics.

Cell Chemical Biology 25, 1–10, June 21, 2018 ª 2018 Elsevier Ltd.

1

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

activation loop residue T172 (a-pT172) by LKB1 and CaMKK2, efforts. PF-739 was co-crystallized with the a1b1g1 complex,

triggering the AMPK signaling cascade. Current models of whereas the only previous structures of the b2-AMPK hetero-

allosteric regulation depict exchange of ATP for AMP at binding trimer captured the analogous ADaM site in the apo form, and

sites in the g subunit, leading to sequestration of a tri-helical consequently was not informative on the binding determinants

autoinhibitory domain (AID) located immediately C-terminal important for drug action (Li et al., 2015). In this study, we

to the catalytic kinase domain (Chen et al., 2013). AMPK acts describe the ﬁrst crystal structures of phosphorylated a1b2g1

to restore cellular energy balance through phosphorylation of and a2b2g1 with an occupied ADaM site. Crystallization was

numerous key metabolic enzymes, resulting in inhibition of aided by inclusion of the potent, small-molecule b2-AMPK

ATP-consuming anabolic pathways and promotion of ATP- activator SC4, and comparative structural analysis reveals the

generating catabolic processes.

b2-speciﬁc residue D111 as a key mediator of drug activation.

The potential of AMPK drugs has been pre-clinically validated We consider that the results provide a structure-function guide

by recent and extensive in vivo characterization of high-potency to accelerate development of more potent, but importantly tis-

pan-AMPK activators PF-739 (Cokorinos et al., 2017) and sue-speciﬁc, b2-AMPK therapeutics.

MK-8722 (Myers et al., 2017). Uniquely, these drugs induced

robust, skeletal muscle AMPK-dependent glucose disposal, RESULTS

and improved glucose homeostasis, in diabetic mice and non-

human primates, both important attributes for T2D therapeutics. SC4 Is a Potent Activator of a2b2-AMPK Complexes

With MK-8722, however, chronic dosing was associated with To expand the repertoire of fully characterized b2-AMPK

cardiac related complications (glycogen accumulation and agonists, we selected and synthesized the small-molecule

hypertrophy) reminiscent of chronic AMPK activation (Hardie, SC4 from the patent literature (Tonogaki et al., 2013) (Fig-

2014), potentially compromising the therapeutic use of these ure 1A). Structurally, SC4 shares the ring-3 (2-methylbenzoic

pan-AMPK activators in a clinical setting.

acid/o-toluic acid) and core ring-2 (imidazopyridine) groups

Mammalian AMPK consists of a catalytic a subunit and with 991, but differs in the presence of a nitrogen atom at

regulatory b and g subunits. The b subunit is myristoylated at the 4⁰position. When screened against all 12 AMPK heterotri-

position Gly2, a modiﬁcation important for spatiotemporal regu- meric combinations (Table S1), SC4 activated complexes

lation of AMPK signaling (Oakhill et al., 2010; Zhang et al., containing either b isoform, but at 0.5 mM displayed a selec-

2017). Each subunit has multiple isoforms (a1, a2, b1, b2, g1, tivity preference for AMPK a2 (all six a2 complexes were

g2, g3) that are differentially expressed and regulated signiﬁcantly activated, compared with two [a1b1g1 and

throughout the human body. For example, a1- and a2-AMPK a1b1g3] out of six a1 complexes) (Figure 1B and Table S2).

complexes contribute similarly to total AMPK activity in the hu- Importantly, SC4 signiﬁcantly activated a2b2g1 (EC₅₀17.2

man heart (Kim et al., 2012), whereas a2 and b2 are highly ex- 1.6 nM) and a2b2g3 (EC₅₀82.1 1.3 nM), the predominant

pressed, and almost exclusively associated, in human skeletal AMPK isoform complexes in human skeletal muscle (Wojtas-

muscle (Thornton et al., 1998; Cheung et al., 2000; Wojtaszew- zewski et al., 2005), but %5 mM failed to signiﬁcantly stimulate

ski et al., 2005). Since skeletal muscle is the primary organ for activities of the a1-equivalent complexes, a1b2g1 and a1b2g3

glucose disposal and storage, and b2-AMPK has an estab- (Figure 1C).

lished role as a driver of insulin-independent skeletal muscle

We performed further biochemical analyses to determine the

glucose clearance (Scott et al., 2008), a2b2 complexes are re- mechanism of b2 activation by SC4. Deletion of the b-subunit

garded as important drug targets to treat metabolic diseases, CBM in both AMPK a2b1g1 and a2b2g1 abolished SC4 activa-

in particular T2D. However, ﬁrst-generation direct-acting tion, whereas AMP activation was largely preserved (Fig-

AMPK drugs (e.g., A769662, salicylate, PF-06409577, and, to ure S1A). This strongly indicates that SC4 occupies a pocket

a lesser extent, 991), show strong b1-isoform selectivity (Cool in a2b2g1 analogous to the ADaM site in b1-AMPK complexes.

et al., 2006; Scott et al., 2008; Hawley et al., 2012; Cameron Exchange of the phosphorylation site b-S108 for Ala increased

et al., 2016; Xiao et al., 2013). All these AMPK drugs likely the SC4 activation EC₅₀by ꢀ4-fold for both a2b2g1 and

occupy a pocket formed between the b-subunit carbohydrate a2b1g1, without signiﬁcantly altering maximal activation (Fig-

binding module (CBM, b1 residues 76–156) and the N-terminal ure S1B and Table S2). This is in contrast to A-769662, which

lobe of the a-subunit kinase domain (Xiao et al., 2013; Calabr- has an absolute requirement for S108 phosphorylation to

ese et al., 2014) (termed the ADaM [allosteric drug and activate b1-AMPK, but broadly in line with similar experiments

metabolite] site [Langendorf and Kemp, 2015]). This pocket is conducted with 991 (Willows et al., 2017a). As previously

stabilized in the b1 isoform by phosphorylation of the CBM shown with A-769662 (Scott et al., 2014), SC4 synergistically

residue S108, which sensitizes b1-AMPK to ADaM-site ago- activated dephosphorylated a1b1g1 in the presence of

nists and may play a role in pro-survival pathways (Dite et al., AMP (>500-fold activation); however, AMP/drug synergistic

2017). Furthermore, b1-pS108 renders phosphorylation of activation did not translate to the a2b2g1 isoform complex

a-T172 dispensable for activation by ADaM-site drugs, (Figure S1C).

whereas unphosphorylated AMPK remains sensitive to syner-

gistic activation in the presence of both AMP and ADaM-site SC4 Triggers AMPK Cellular Signaling

drugs (Scott et al., 2014; Dite et al., 2017).

We investigated SC4 cellular bioactivity in differentiated

Despite their demonstrated applicability, structural informa- C2C12 myotubes that express a2b2g3 (Bultot et al., 2016).

tion relating to how drugs bind to, and activate, b2-AMPK is lack- Incubation of these cells with SC4 induced dose-dependent in-

ing; this represents a major obstacle to drug development creases in phosphorylation of the AMPK protein substrate

2

Cell Chemical Biology 25, 1–10, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

Figure 1. Biochemical Analysis of SC4

(A) Structures of MK-8722, SC4, and 991. Arrow

indicates SC4 4⁰-nitrogen.

(B) Regulation of human AMPK isoforms by SC4.

Twelve AMPK heterotrimeric combinations, as

indicated, were puriﬁed from COS7 mammalian

cells and activities measured in the presence of

SC4 (0.5 mM) or AMP (100 mM). n = 4, data pre-

sented as mean AMPK activity (fold change relative

to untreated

SEM). Basal activities for each

isoform are displayed in Table S1. Statistical

analysis was performed using two-way ANOVA

with post hoc Dunnett’s multiple comparison test.

**p < 0.01 indicates signiﬁcant increase in activity

relative to untreated. hSM denotes isoforms

previously detected in human vastus lateralis

muscle (a2b2g1 > a2b2g3 = a1b2g1) (Wojtaszew-

ski et al., 2005).

(C) Dose-response curves for SC4 (0–5 mM) acti-

vation of AMPK isoforms a2b2g1, a2b2g3, a1b2g1,

and a1b2g3. n = 4, data presented as mean AMPK

activity (fold change relative to untreated SEM).

Isoform-speciﬁc EC₅₀values for SC4 and 991 (Xiao

et al., 2013) are displayed in the associated table.

See also Figure S3.

non-conserved CBM residue b2-D111 and the 4⁰-nitrogen in the the b2 subunit deleted. We were unable to obtain sufﬁcient

imidazopyridine ring-2 of the SC4 core (Figures 1A and 4C). expression of the truncated heterotrimer in mammalian cells;

b2-D111 makes another interaction with b2-R82, which forms therefore we used our Escherichia coli expression system and

a cationic p-stacking interaction with SC4 ring-2, similar to pre- activated WT and b2DNterm complexes during puriﬁcation by

vious b1-R83 drug-bound structures (Figures 4B and 4C). We CaMKK2-mediated phosphorylation of T172 (Figure 5G). SC4

speculated that the b2-D111-SC4 core interaction may account activated both AMPK complexes with similar potencies

for the increased b2 potency of SC4, relative to 991. Indeed, (p-a2b2g1 EC₅₀68.8 1.1 nM, p-a2b2DNtermg1 EC₅₀31.4

mutation of b2-D111 to Ala resulted in almost total loss of SC4 1.1 nM) and maximal fold activations (Figure 5H). Similarly,

activation, whereas mutation of the analogous b1 residue N111 100 mM AMP activated both AMPK complexes by a similar

to Ala had a less dramatic effect, increasing EC₅₀only ꢀ3-fold magnitude.

(Figure 4D and Table S2). Conversely, substitution of the SC4

4⁰-nitrogen for carbon (cSC4) resulted in ꢀ6-fold reduction in po- DISCUSSION

tency for a2b2g1 AMPK (EC₅₀: 16.3–99.2 nM) (Figure 4E).

The insulin-sensitizing and glucose-disposing properties of

AMPK activation in skeletal muscle in response to physical

exercise/contraction are well established and remain intact in

A b2-Subunit a Helix Interacts with the Autoinhibitory

Domain of a2

We detected unique electron density in the p-a2b2g1/SC4 com- individuals with T2D (Kjøbsted et al., 2016). Therefore, it is of little

plex that had not been seen in earlier AMPK structures. Despite surprise that small-molecule activation of AMPK in this tissue is

medium resolution, this enabled us to model main-chain density remarkably effective at lowering serum glucose levels and

for extra residues N-terminal to the CBM Pro78 in the only other improving glucose homeostasis in animal models of diabetes

available b2-AMPK structure (Li et al., 2015). The most striking (Cokorinos et al., 2017; Myers et al., 2017). A similar strategy

element of this extended structure is a short a helix, which we may be usefully employed to treat T2D in humans, given consis-

have assigned as comprising b2 residues Phe60 to Ser70, posi- tent AMPK isoform expression in skeletal muscle across

tioned within a shallow, open groove bordered by the three mammalian species. Importantly, these pan-AMPK activators

helices constituting the a2 AID (Figures 5A–5C). This groove is did not induce hypoglycemia, and exerted glycemic control

apparent, yet vacant, in other AMP-bound AMPK structures independently of increased insulin levels. This provides them

with a resolved AID (e.g., PDB: 4RER and 5IS0) (Figure 5C), with an important and unique advantage over current, incretin-

and in our p-a2b1g1/SC4 complex (Figure 5D), but is largely based therapeutics (e.g., SUs), chronic use of which has been

inaccessible in isolated human a2 AID (PDB: 2LTU) or inactive associated with pancreatic b cell exhaustion, desensitization,

kinase domain AID (PDB: 4RED) structures due to inward rota- and apoptosis (Hanefeld, 2007). Substances that cause or

tions of a helices 1 and 2 (Figure 5E). Consistent with the promote secretion of another substance, e.g., SUs (pancreatic

active-state structures, the AID in our p-a2b2g1/SC4 complex insulin secretion) and ghrelin (growth hormone release from the

is sequestered away from the kinase domain small lobe, and hypothalamus), are collectively termed secretagogs. A compar-

rotated by ꢀ180^ꢁsuch that the kinase domain-interacting sur- ative term for induced augmentation of cellular import does not

face now faces the g subunit CBS2. However, interaction with currently exist, thus we propose ‘‘importagog’’ to describe a

the b2 a helix in our structure results in an additional rotation, substance or compound that causes or promotes import or

centered around the C-terminal end of the AID a helix 3, culmi- uptake of another substance into a cell or tissue. Accordingly,

˚

nating in a further 11-A shift in AID a helices 1 and 2 toward insulin, TZDs, and other synthetic drugs with glucose-lowering

the g subunit CBS2 (Figure 5F).

To investigate the signiﬁcance of the b2/AID interaction for

actions can be classiﬁed as glucose importagogs.

In the current study we ﬁrst describe structural characteriza-

AMP or SC4 regulation, we expressed a truncated AMPK tion of the drug binding site in human skeletal muscle AMPK

complex (a2b2DNtermg1) with the N-terminal 68 residues of a2b2g1. Success in AMPK crystallization was provided by

Cell Chemical Biology 25, 1–10, June 21, 2018

5

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

addition of the nanomolar afﬁnity b2-AMPK activator SC4, the

fourth member of a class of drugs demonstrated to promote

insulin-independent skeletal muscle glucose uptake in vivo

(PF-739 and MK-8722) or ex vivo (991 and SC4). SC4 triggered

AMPK signaling in a range of cell types by a purely allosteric

mechanism, and in doing so acted independently of changes

to intracellular AMP/ATP ratio or pT172. This disconnect

between pT172 levels and AMPK activation by SC4 mirrors

previous observations using A-769662 (Scott et al., 2014) and

MK-8722 (Myers et al., 2017). Similarly, SC4-stimulated glucose

uptake in isolated mouse soleus, EDL, and epitrochlearis mus-

cles was not accompanied by detectable elevations in pT172.

Combined, these studies reinforce our view that pT172 in isola-

tion should not be regarded as a deﬁnitive measure of cellular

AMPK activity, particularly in response to drug treatment.

We recently showed that, when co-incubated with A-769662

and phenformin, signiﬁcant pT172-independent a2-AMPK

signaling could be detected in a1^ꢂ/ꢂ/a2^ꢂ/ꢂMEFs exclusively

expressing T172A-mutated a2 complexes (Dite et al., 2017);

this despite the generally smaller magnitude of AMP/A-

769662 dual-ligand co-activation of a2 complexes compared

with a1 (Langendorf et al., 2016). In contrast, treatment of

A549 cells, lacking LKB1 and genetically deleted for CaMKK2,

with 2-deoxyglucose (2DG, mimicking glucose starvation)

and 991 failed to stimulate T172 phosphorylation and only

modestly induced AMPK activity (Willows et al., 2017b). The

reason for this discrepancy is unclear, although we note that

991 poorly activated dephosphorylated, puriﬁed a1b1g1 in

the presence of AMP (117-fold) compared with A-769662

(>1,000-fold) (Scott et al., 2014), possibly accounting for the

weaker cellular response. Alternatively, AMP/drug (or AMP/

metabolite) activation of dephosphorylated AMPK may be

cell speciﬁc (governed by a1/a2 expression ratio) and highly

contextual in that individual energy stresses (e.g., glucose

starvation versus inhibition of mitochondrial function) give

rise to distinct effects. Willows et al. (2017b) raise the argu-

ment that, because activities achieved by dual-liganded, un-

phosphorylated AMPK are low (ꢀ10%) relative to that obtained

Figure 4. SC4 Occupies the b2-ADaM Site

following stoichiometric (100%) CaMKK2 phosphorylation,

(A) Cartoon representation of full-length p-a2b2g1 (a2, green; b1, cyan; g1,

physiological relevance of this mechanism as a means for

magenta) in complex with SC4 and two AMP molecules (g sites 3 and 4). Inset:

AMPK regulatory control remains doubtful. We note that in

surface representation centered on SC4 bound in the b2-ADaM site, formed

the same study the measured stoichiometry of cellular pT172

following 2DG treatment, which induced substantial

between the N lobe of the a-kinase domain and b2-CBM.

a

(B) Stereo view of the SC4-occupied ADaM site in p-a2b1g1 AMPK (a2, white;

b1, blue), superposed with compounds 991 (PDB: 4CFE, orange), PF-739

(PDB: 5UFU, purple), and PF-06409577 (PDB: 5KQ5, red), and their respective

interacting residues. Interacting residues (color matched with compound)

occupy a similar conformation, suggesting a common interaction proﬁle for all

compounds shown. b1-S108 is phosphorylated in 4CFE; phosphomimetic

b1-D108 is used in 5UFU and 5KQ5.

increase in ADP/ATP ratio, was at maximal 10%. This is

consistent with our own observations in COS7 cells, in which

pT172 levels reached ꢀ8% following treatment with ionomy-

cin, a Ca²⁺ionophore and CaMKK2 activator, and rapidly re-

turned to basal levels following removal of the stimulus (Scott

et al., 2014). Thus, pT172 is tightly constrained by strong

phosphatase(s) pressure and is unlikely to reach high stoichi-

ometry, even during periods of extreme cellular stress.

We therefore argue that dual-ligand activation of unphos-

phorylated AMPK, which represents the major form of cellular

AMPK and is of course free from phosphatase regulation, rep-

resents a relevant magnitude of AMPK activation.

(C–E) Stereo view of p-a2b2g1 (a2, white; b2, light blue) and p-a2b1g1 (a2,

green; b1, dark blue) superposed to highlight the conformational changes in

b1/2 residue interactions with SC4 (yellow and gray, respectively). The side-

chain oxygen of the non-conserved residue b2-D111 is favorably positioned to

form an interaction with the SC4 4⁰-nitrogen and b2-R82. Some b2 side chains

have not been modeled due to poor electron density. Dose-response curves

for (D) SC4 (0–5 mM) activation of AMPK mutants a1b1(N111A)g1 and

a2b2(D111A)g1, and (E) cSC4 (0–5 mM) activation of a2b2g1. AMPK com-

plexes were puriﬁed from COS7 mammalian cells. n = 4, data presented as

mean AMPK activity (fold change relative to untreated SEM).

SC4, despite being a potent activator of p-a2b2g1, was un-

able to activate dephosphorylated a2b2g1 in the presence of

AMP (Figure S1C). Comparison of two crystal structures of

the a2b1g1/AMP/991 complex reveal that the activation loop

See Figure S4; Tables S2 and S3.

6

Cell Chemical Biology 25, 1–10, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

Figure 5. The b2 Subunit Interacts with

a2-AID

(A) Sequence alignments between (upper) a1/a2

AIDs, and b1/b2 residues N-terminal to the car-

bohydrate module (CBM). Secondary structural

elements are shown as cylinders for a1 (derived

from PDB: 2LTU; magenta) and a2 (derived from

our p-a2b1g1/SC4 structure; orange) and b2

(p-a2b2g1/SC4; cyan). Predicted secondary

structure elements for b1 and b2 are shown as

light-green cylinders (http://www.biogem.org/tool/

chou-fasman/; Chou and Fasman, 1974).

(B) Close-up view of the a2-AID groove (orange)

accommodating a b2 a helix N-terminal to the

CBM (cyan). Electron density in this region permits

reasonable main-chain, but not side-chain,

deﬁnition.

(C) Cartoon representation of the exposed a2-AID

groove, housing a short helix consisting of b2

residues 60–70, in the p-a2b2g1/SC4 structure

(a2, orange; b2, cyan). The a1-AID from 4RER

(green) is superposed for comparison, and

conserved residues forming the AID hydrophobic

core are highlighted.

(D) a2-AID from the p-a2b2g1/SC4 complex

(orange) superposed to a2-AID from the p-a2b1g1/

SC4 complex (gold).

(E) Cartoon representation of the isolated human

a2-AID NMR solution structure (PDB: 2LTU).

Hydrophobic core residues are shown as sticks.

(F) Orientations of AIDs in p-a2b2g1/SC4 (orange)

and a2b1g1/991 (green) complexes, structurally

aligned along the kinase large lobe

E helix

(p-a2b2g1/SC4 E helix in pink). The AID-interacting

b2 helix in our p-a2b2g1/SC4 structure has been

removed for clarity.

(G) E. coli expression and CaMKK2-activation of

a2b2g1 (WT) and a2b2(67–272)g1 (DNterm).

Puriﬁed preparations were immunoblotted as

shown.

(H) Dose-response curves for SC4 (0–10 mM)

activation of b2-WT and b2-DNterm AMPK. n = 4,

data presented as mean AMPK activity (fold

change relative to untreated

SEM). Basal

activities and maximal fold activation for SC4 and

AMP (100 mM) are displayed in the associ-

ated table.

adopts an almost identical active conformation regardless of genic studies in which disruption of the AID-kinase domain

the phosphorylation state of T172 (Xiao et al., 2013; Willows interaction led to loss of AMP allosteric regulation and reduced

et al., 2017b). This supports our model in which occupancy susceptibility to pT172 dephosphorylation (Chen et al., 2009),

of AMP/b1-ADaM sites stabilizes the unphosphorylated although the latter effect was not observed in a similar study

AMPK activation loop in an active conformation (Scott et al., (Li et al., 2015). It is tempting to speculate that the b2-AID

2014), but for reasons that are currently unclear this ligand- interaction might promote enzyme activity by stabilizing AMP-

induced structural restraint does not entirely translate to the induced sequestration, although our biochemical data using

b2 isoform.

non-myristoylated, bacterial-expressed AMPK does not support

An interesting feature of our p-a2b2g1/SC4 structure is this proposal (Figure 5H). However, the b-subunit N terminus

signiﬁcant movement of the a2-AID, most likely resulting from inherently contributes to cellular AMPK regulation since preven-

an interaction with a b2 a helix N-terminal to the CBM. The AID tion of N-myristoylation, or removal of b1 residues 1–63, resulted

is an important AMPK regulatory element that has been pro- in hyperactivate AMPK complexes when expressed in mamma-

posed to stabilize the kinase domain in an ‘‘open’’ inactive lian cells (Scott et al., 2008).

conformation through interactions with both small and large Is the interaction speciﬁc to a2b2g1 complexes? Intriguingly,

lobes. Autoinhibition is relieved by AMP-induced sequestration the interacting regions encompass two stretches of isoform

of the AID away from the kinase domain, supported by muta- sequence heterogeneity that, for the b-subunit, may also

Cell Chemical Biology 25, 1–10, June 21, 2018

7

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

manifest as divergence in secondary structure elements (Fig- skeletal muscle, where the immediate future for AMPK thera-

ure 5A). On the one hand, this interaction was not observed in peutics appears brightest.

our p-a2b1g1 structure presented here, or structures of AMP/

drug-bound, full-length a2b1g1 in which the AID was resolved

(Xiao et al., 2013; Willows et al., 2017b), even though the AID

groove is ‘‘accessible’’ in all these complexes. On the other

hand, deletion of the b N-terminal region and CBM in a2b1g1

perplexingly abolished AMP suppression of pT172 dephosphor-

ylation (Xiao et al., 2013), suggestive of intramolecular signaling

between g-nucleotide sites and the CBM that might be facilitated

by a direct interaction between the b1 N terminus and a2-AID.

Whether ADaM-site occupancy is required to stabilize the

interaction, and whether the interaction persists in the absence

of AMP, are also important questions to address in further

investigations.

SIGNIFICANCE

Small-molecule activation of b2-AMPK complexes has at-

tracted interest as a strategy to clinically improve glucose

homeostasis via skeletal muscle glucose uptake and

storage. Principal signiﬁcance of our results is ﬁrst provi-

sion of structural information of a b2-AMPK/drug complex.

2⁰-mannitol derivatives, such as MK-8722, are potent

pan-AMPKactivatorsthatinduceoff-targeteffects. Structural

comparisons described in this study reveal binding

determinants that can be exploited to increase both potency

and isoform selectivity, two criteria of AMPK drug develop-

ment that require equal consideration. We anticipate

success in both directions will be afforded by detailed

characterization of more b2-AMPK activators. Identiﬁcation

ofaninteraction between theb-subunitNterminusanda-sub-

unit autoinhibitory domain also opens the door to further

studies interrogating mechanisms of AMPK drug action.

Off-target tissue effects represent long-appreciated and sig-

niﬁcant obstacles to clinical use of systemic AMPK activators.

Accordingly, the pan-AMPK activator MK-8722 was reported

to induce reversible cardiac hypertrophy, although whether

this arose as a direct effect of activating cardiac AMPK or rep-

resents a secondary complication resulting, for example, from

systemic hypertension (Stanton and Dunn, 2017), is unclear.

Regardless, despite advances in tissue-speciﬁc delivery

systems, the challenges facing development of isoform/tissue-

speciﬁc AMPK activators remain signiﬁcant. Whereas pan-

AMPK activators may have been selected for in vivo analysis

based on high potency (Cokorinos et al., 2017; Myers et al.,

2017), we take the view that isoform selectivity requires at

least equal consideration when investigating AMPK-targeting

glucose importagogs. We anticipate that structural determi-

nants for SC4 binding presented here will aid development

of small-molecule agonists targeting skeletal muscle AMPK

a2b2 complexes. Tissue speciﬁcity may ultimately be realized

with activators of a2b2g3, found exclusively in skeletal mus-

cle, to avoid potential complications associated with stimu-

lating a2 signaling in the heart. Speciﬁcally, b2 potency is

conferred by interaction of the core ring-2 4⁰-nitrogen with

the non-conserved b2-D111 (SC4 versus 991). By mutational

analysis we showed that the analogous b1 residue N111 is

not a major binding determinant for SC4 binding, indicating

a therapeutically exploitable point of difference between the

two isoforms. Using puriﬁed enzyme, SC4 activated all six

a2-AMPK complexes, in comparison with only two a1 com-

plexes (Figure 1B). Comparisons of the binding modes of

SC4 to b1-AMPK with those of MK-8722 and PF-739 reveal

a possible explanation. A striking difference between these

compounds is the 2⁰-mannitol of MK-8722, common to PF-

739, in place of the SC4 2⁰-o-toluic acid group (Figure 1A).

This allows MK-8722 and PF-739 to make additional back-

bone contacts with conserved a2 residues F27 and N48 that

are absent in the b1/SC4 structure (Figure 4B). These contacts

likely strengthen the interaction between the a surface of the

ADaM site and PF-739, compared with SC4. Consequently,

while promotion of b2 afﬁnity conferred by the 2⁰-mannitol

substituent (PF-739 versus 991) may appear advantageous,

this group also promotes activation of all a1-AMPK complexes

(SC4 versus MK-8722 and PF-739) and disrupts the path to

isoform speciﬁcity. Overall, our ﬁndings lay the foundations

for rational reﬁnement of drug tissue speciﬁcity toward

STAR+METHODS

Detailed methods are provided in the online version of this paper

and include the following:

d KEY RESOURCES TABLE

d CONTACT FOR REAGENT AND RESOURCE SHARING

d EXPERIMENTAL MODEL DETAILS

B Animals

B Cell Lines

d METHOD DETAILS

B Synthesis of SC4

B 6-Chloro-5-iodo-2-(methylsulfonyl)-3-((2-(trimethyl-

silyl)ethoxy)methyl)-3H-imidazo[4,5-b]pyridine (1)

B Methyl 5-((6-chloro-5-iodo-3-((2-(trimethylsilyl)ethoxy)

methyl)-3H-imidazo[4,5-b]pyridin-2-yl)oxy)-2-methyl-

benzoate (2)

B Methyl 5-((6-chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-

3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo[4,5-b]

pyridin-2-yl)oxy)-2-methylbenzoate (3)

B 5-((6-Chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-3H-

imidazo[4,5-b]pyridin-2-yl)oxy)-2-methylbenzoic

acid (SC4)

B Synthesis of cSC4

B 6-Chloro-2’’-methoxy-4-nitro-[1,1’:4’,1’’-terphenyl]-3-

amine (4)

B 6-Chloro-2’’-methoxy-[1,1’:4’,1’’-terphenyl]-3,4-

diamine (5)

B 5-Chloro-6-(2’-methoxy-[1,1’-biphenyl]-4-yl)-1,3-dihy-

dro-2H-benzo[d]imidazole-2-thione (6)

B 6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(meth-

ylthio)-1H-benzo[d]imidazole (7)

B 6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(meth-

ylsulfonyl)-1H-benzo[d]imidazole (8)

B 6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(meth-

ylsulfonyl)-1-((2-(trimethylsilyl)ethoxy)

benzo[d]imidazole (9)

methyl)-1H-

8

Cell Chemical Biology 25, 1–10, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

B Methyl 5-((6-chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-

yl)-1-((2-(trimethylsilyl)ethoxy)methyl)-1H-benzo[d]imi-

dazol-2-yl)oxy)-2-methylbenzoate (10)

B Methyl 5-((6-chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-

1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoate (11)

B 5-((6-Chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-1H-

benzo[d]imidazol-2-yl)oxy)-2-methylbenzoic

acid (cSC4)

REFERENCES

Bricogne, G., Blanc, E., Brandl, M., Flensburg, C., Keller, P., Paciorek, P.,

Roversi, P., Sharff, A., Smart, O., Vonrhein, C., and Womack, T. (2011).

BUSTER Version 2.10.0 (Global Phasing Ltd).

Bultot, L., Jensen, T.E., Lai, Y.C., Madsen, A.L., Collodet, C., Kviklyte, S.,

Deak, M., Yavari, A., Foretz, M., Ghaffari, S., et al. (2016). Benzimidazole

derivative small-molecule 991 enhances AMPK activity and glucose uptake

induced by AICAR or contraction in skeletal muscle. Am. J. Physiol.

Endocrinol. Metab. 311 , E706–E719.

B Constructs for Crystallization

B Protein Expression and Puriﬁcation for Crystallization

B Crystallization

Calabrese, M.F., Rajamohan, F., Harris, M.S., Caspers, N.L., Magyar, R.,

Withka, J.M., Wang, H., Borzilleri, K.A., Sahasrabudhe, P.V., Hoth, L.R.,

et al. (2014). Structural basis for AMPK activation: natural and synthetic ligands

regulate kinase activity from opposite poles by different molecular mecha-

nisms. Structure 22 , 1161–1172.

B Protein Expression and Puriﬁcation for AMPK Assays

B AMPK Activity Assay

B Lentivirus Generation and Expression

B Cell Experiments

Carpio, G.R.A., and Fonseca, V.A. (2014). Update on safety issues related to

antihyperglycemic therapy. Diabetes Spectr. 27 , 92–100.

B Immunoblotting

Cameron, K.O., Kung, D.W., Kalgutkar, A.S., Kurumbail, R.G., Miller, R.,

Salatto, C.T., Ward, J., Withka, J.M., Bhattacharya, S.K., Boehm, M., et al.

(2016). Discovery and preclinical characterization of 6-chloro-5-[4-(1-hydrox-

B Adenine Nucleotide Measurements

B Mouse Skeletal Muscle Isolation and Glucose Up-

take Assay

ycyclobutyl)phenyl]-1H-indole-3-carboxylic acid (PF-06409577),

a direct

activator of adenosine monophosphate-activated protein kinase (AMPK),

for the potential treatment of diabetic nephropathy. J. Med. Chem. 59 ,

8068–8081.

B Incubation and Electrical Stimulation of Isolated Rat

Skeletal Muscles

d QUANTIFICATION AND STATISTICAL ANALYSIS

d DATA AND SOFTWARE AVAILABILITY

Chen, L., Jiao, Z.-H., Zheng, L.-S., Zhang, Y.-Y., Xie, S.-T., Wang, Z.-X., and

Wu, J.-W. (2009). Structural insight into the autoinhibition mechanism of

AMP-activated protein kinase. Nature 459 , 1146–1149.

SUPPLEMENTAL INFORMATION

Chen, L., Xin, F.J., Wang, J., Hu, J., Zhang, Y.Y., Wan, S., Cao, L.S., Lu, C., Li,

P., Yan, S.F., et al. (2013). Conserved regulatory elements in AMPK. Nature

498 , E8–E10.

Supplemental Information includes four ﬁgures and four tables and can be

found with this article online at https://doi.org/10.1016/j.chembiol.2018.

03.008.

Chen, V.B., Arendall, W.B., 3rd, Headd, J.J., Keedy, D.A., Immormino, R.M.,

Kapral, G.J., Murray, L.W., Richardson, J.S., and Richardson, D.C. (2010).

MolProbity: all-atom structure validation for macromolecular crystallography.

Acta Crystallogr. D Biol. Crystallogr. 66 , 12–21.

ACKNOWLEDGMENTS

Cheung, P.C., Salt, I.P., Davies, S.P., Hardie, D.G., and Carling, D. (2000).

Characterization of AMP-activated protein kinase gamma-subunit isoforms

and their role in AMP binding. Biochem. J. 346 , 659–669.

This research was undertaken on the MX2 beamline at the Australian Synchro-

tron, Victoria, Australia, part of ANSTO, and made use of the ACRF detector.

We thank the beamline staff for their assistance. We thank Sandra Galic and

Lisa Murray-Segal for assistance with preparation of primary hepatocytes

and isolation of mouse muscles. This work was supported by grants from

the Australian Research Council (ARC ID:130100988) and the National Health

and Medical Research Council (NHMRC ID: 1098459, 1145265). C.G.L. is an

EH Flack Fellow and is supported by the Jack Brockhoff foundation (JBF-

4206, 2016). M.W.P., B.E.K., and J.B.B. are NHMRC Research Fellows.

J.S.O. is an ARC Future Fellow. This study was supported in part by the Victo-

rian Government’s Operational Infrastructure Support Program.

Chou, P.Y., and Fasman, G.D. (1974). Prediction of protein conformation.

Biochemistry 13 , 222–245.

Cokorinos, E.C., Delmore, J., Reyes, A.R., Albuquerque, B., Kjøbsted, R.,

Jørgensen, N.O., Tran, J.L., Jatkar, A., Cialdea, K., Esquejo, R.M., et al.

(2017). Activation of skeletal muscle AMPK promotes glucose disposal

and glucose lowering in non-human primates and mice. Cell Metab. 25 ,

1147–1159.e10.

Cool, B., Zinker, B., Chiou, W., Kiﬂe, L., Cao, N., Perham, M., Dickinson, R.,

Adler, A., Gagne, G., Iyengar, R., et al. (2006). Identiﬁcation and characteriza-

tion of a small molecule AMPK activator that treats key components of type 2

diabetes and the metabolic syndrome. Cell Metab. 3 , 403–416.

AUTHOR CONTRIBUTIONS

Dite, T.A., Ling, N.X.Y., Scott, J.W., Hoque, A., Galic, S., Parker, B.L., Ngoei,

K.R.W., Langendorf, C.G., O’Brien, M.T., Kundu, M., et al. (2017). The auto-

phagy initiator ULK1 sensitizes AMPK to allosteric drugs. Nat. Commun.

8 , 571.

K.R.W.N. performed AMPK activity assays and glucose uptake assays. C.G.L.

puriﬁed and crystallized AMPK constructs. N.X.Y.L. designed, expressed and

puriﬁed AMPK for crystallization, and performed cellular assays and immuno-

blotting. A.H., T.A.D., A.J.O., W.J.S. and R.J. performed other experiments.

S.J., M.C.C., S.R.W., Y.E.B., and H.H. performed chemical syntheses of

SC4 and cSC4. M.W.P., J.W.S., M.H.R., B.E.K., R.C.F., J.B.B., and J.S.O. pro-

vided conceptual input. K.R.W.N., C.G.L., and J.S.O. designed the study and

wrote the manuscript with contributions from all authors.

Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010). Features and

development of coot. Acta Crystallogr. D Biol. Crystallogr. 66 , 486–501.

Evans, P.R., and Murshudov, G.N. (2013). How good are my data and what is

the resolution? Acta Crystallogr. D Biol. Crystallogr. 69 , 1204–1214.

Hanefeld, M. (2007). Pioglitazone and sulfonylureas: effectively treating type 2

diabetes. Int. J. Clin. Pract. Suppl. 20–27.

DECLARATION OF INTERESTS

Hardie, D.G. (2014). AMP-activated protein kinase: a key regulator of energy

The authors declare no competing interests.

balance with many roles in human disease. J. Intern. Med. 276 , 543–559.

Hardie, D.G. (2015). AMPK: positive and negative regulation, and its role in

Received: November 8, 2017

Revised: January 31, 2018

Accepted: March 19, 2018

Published: April 12, 2018

whole-body energy homeostasis. Curr. Opin. Cell Biol. 33 , 1–7.

Hawley, S.A., Fullerton, M.D., Ross, F.A., Schertzer, J.D., Chevtzoff, C.,

Walker, K.J., Peggie, M.W., Zibrova, D., Green, K.A., Mustard, K.J., et al.

Cell Chemical Biology 25, 1–10, June 21, 2018

9

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

(2012). The ancient drug salicylate directly activates AMP-activated protein ki-

AMP-activated protein kinase beta1-containing complexes. Chem. Biol. 15 ,

nase. Science 336 , 918–922.

1220–1230.

Scott, J.W., Ling, N., Issa, S.M., Dite, T.A., O’Brien, M.T., Chen, Z.P., Galic, S.,

Langendorf, C.G., Steinberg, G.R., Kemp, B.E., and Oakhill, J.S. (2014). Small

molecule drug A-769662 and AMP synergistically activate naive AMPK inde-

pendent of upstream kinase signaling. Chem. Biol. 21 , 619–627.

Kabsch, W. (2010). XDS. Acta Crystallogr. D Biol. Crystallogr. 66 , 125–132.

Kim, M., Shen, M., Ngoy, S., Karamanlidis, G., Liao, R., and Tian, R. (2012).

AMPK isoform expression in the normal and failing hearts. J. Mol. Cell.

Cardiol. 52 , 1066–1073.

Scott, J.W., Galic, S., Graham, K.L., Foitzik, R., Ling, N.X., Dite, T.A., Issa,

S.M., Langendorf, C.G., Weng, Q.P., Thomas, H.E., et al. (2015). Inhibition of

AMP-activated protein kinase at the allosteric drug-binding site promotes islet

insulin release. Chem. Biol. 22 , 705–711.

Kjøbsted, R., Pedersen, A.J., Hingst, J.R., Sabaratnam, R., Birk, J.B.,

Kristensen, J.M., Højlund, K., and Wojtaszewski, J.F. (2016). Intact regulation

of the AMPK signaling network in response to exercise and insulin in skeletal

muscle of male patients with Type 2 Diabetes: illumination of AMPK activation

in recovery from exercise. Diabetes 65 , 1219–1230.

Stanton, T., and Dunn, F.G. (2017). Hypertension, left ventricular hypertrophy,

and myocardial ischemia. Med. Clin. North Am. 101 , 29–41.

´

Lai, Y.C., Kviklyte, S., Vertommen, D., Lantier, L., Foretz, M., Viollet, B., Hallen,

Steinberg, G.R., O’Neill, H.M., Dzamko, N.L., Galic, S., Naim, T., Koopman, R.,

Jørgensen, S.B., Honeyman, J., Hewitt, K., Chen, Z.P., et al. (2010). Whole

body deletion of AMP-activated protein kinase b 2 reduces muscle AMPK ac-

tivity and exercise capacity. J. Biol. Chem. 285 , 37198–37209.

S., and Rider, M.H. (2014). A small-molecule benzimidazole derivative that

potently activates AMPK to increase glucose transport in skeletal muscle:

comparison with effects of contraction and other AMPK activators.

Biochem. J. 460 , 363–375.

Stephenne, X., Foretz, M., Taleux, N., van der Zon, G.C., Sokal, E., Hue, L.,

Viollet, B., and Guigas, B. (2011). Metformin activates AMP-activated protein

kinase in primary human hepatocytes by decreasing cellular energy status.

Diabetologia 54 , 3101–3110.

Langendorf, C.G., and Kemp, B.E. (2015). Choreography of AMPK activation.

Cell Res. 25 , 5–6.

Langendorf, C.G., Ngoei, K.R., Scott, J.W., Ling, N.X., Issa, S.M., Gorman,

M.A., Parker, M.W., Sakamoto, K., Oakhill, J.S., and Kemp, B.E. (2016).

Structural basis of allosteric and synergistic activation of AMPK by furan-2-

phosphonic derivative C2 binding. Nat. Commun. 7 , 10912.

Taylor, S.S., Knighton, D.R., Zheng, J., Sowadski, J.M., Gibbs, C.S., and

Zoller, M.J. (1993).

A template for the protein kinase family. Trends

Biochem. Sci. 18 , 84–89.

Li, X., Wang, L., Zhou, X.E., Ke, J., de Waal, P.W., Gu, X., Tan, M.H., Wang, D.,

Wu, D., Xu, H.E., et al. (2015). Structural basis of AMPK regulation by adenine

nucleotides and glycogen. Cell Res. 25 , 50–66.

Thornton, C., Snowden, M.A., and Carling, D. (1998). Identiﬁcation of a novel

AMP-activated protein kinase beta subunit isoform that is highly expressed

in skeletal muscle. J. Biol. Chem. 273 , 12443–12450.

McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C.,

and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Crystallogr.

40 , 658–674.

Thrasher, J. (2017). Pharmacologic management of type 2 diabetes mellitus:

available therapies. Am. J. Cardiol. 120 , S4–S16.

Tonogaki, K., Ino, A., Kojima, E., Katou, M., Iwatsu, M., Tanaka, N., and

Fujioka, M.. (2013). Hetero ring-Fused imidazole derivative having AMPK acti-

vating effect. US20130184240.

Myers, R.W., Guan, H.P., Ehrhart, J., Petrov, A., Prahalada, S., Tozzo, E.,

Yang, X., Kurtz, M.M., Trujillo, M., Gonzalez Trotter, D., et al. (2017).

Systemic pan-AMPK activator MK-8722 improves glucose homeostasis but

induces cardiac hypertrophy. Science 357 , 507–511.

Willows, R., Navaratnam, N., Lima, A., Read, J., and Carling, D. (2017a). Effect

of different g -subunit isoforms on the regulation of AMPK. Biochem. J. 474 ,

1741–1754.

O’Neill, H.M. (2013). AMPK and exercise: glucose uptake and insulin sensi-

tivity. Diabetes Metab. J. 37 , 1–21.

Willows, R., Sanders, M.J., Xiao, B., Patel, B.R., Martin, S.R., Read, J., Wilson,

J.R., Hubbard, J., Gamblin, S.J., and Carling, D. (2017b). Phosphorylation of

AMPK by upstream kinases is required for activity in mammalian cells.

Biochem. J. 474 , 3059–3073.

Oakhill, J.S., Chen, Z.P., Scott, J.W., Steel, R., Castelli, L.A., Ling, N.,

Macaulay, S.L., and Kemp, B.E. (2010). beta-Subunit myristoylation is the

gatekeeper for initiating metabolic stress sensing by AMP-activated protein ki-

nase (AMPK). Proc. Natl. Acad. Sci. USA 107 , 19237–19241.

Wojtaszewski, J.F., Birk, J.B., Frøsig, C., Holten, M., Pilegaard, H., and Dela, F.

(2005). 5⁰AMP activated protein kinase expression in human skeletal muscle:

effects of strength training and type 2 diabetes. J. Physiol. 564 , 563–573.

Oakhill, J.S., Scott, J.W., and Kemp, B.E. (2012). AMPK functions as an adeny-

late charge-regulated protein kinase. Trends Endocrinol. Metab. 23 , 125–132.

€

Schuttelkopf, A.W., and van Aalten, D.M. (2004). PRODRG: a tool for high-

Xiao, B., Sanders, M.J., Carmena, D., Bright, N.J., Haire, L.F., Underwood, E.,

Patel, B.R., Heath, R.B., Walker, P.A., Hallen, S., et al. (2013). Structural basis

of AMPK regulation by small molecule activators. Nat. Commun. 4 , 3017.

throughput crystallography of protein-ligand complexes. Acta Crystallogr. D

Biol. Crystallogr. 60 , 1355–1363.

Scott, J.W., van Denderen, B.J.W., Jorgensen, S.B., Honeyman, J.E.,

Steinberg, G.R., Oakhill, J.S., Iseli, T.J., Koay, A., Gooley, P.R., Stapleton,

D., and Kemp, B.E. (2008). Thienopyridone drugs are selective activators of

Zhang, C.S., Hawley, S.A., Zong, Y., Li, M., Wang, Z., Gray, A., Ma, T., Cui, J.,

Feng, J.W., Zhu, M., et al. (2017). Fructose-1,6-bisphosphate and aldolase

mediate glucose sensing by AMPK. Nature 548 , 112–116.

10 Cell Chemical Biology 25, 1–10, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

STAR+METHODS

KEY RESOURCES TABLE

REAGENT or RESOURCE

Antibodies

SOURCE

IDENTIFIER

AMPK a

Cell Signaling Technology

Abcam

Cat# 2793; RRID: AB_915794

AMPK b

Cat# 4150; RRID: AB_10828832

Cat# 2535; RRID: AB_331250

AMPK a-phosphoThr172

AMPK b-phosphoSer108

phosphoACC

Cat# 4181; RRID: AB_2169743

Cat# 3661; RRID: AB_330337

AMPK b1

Cat# ab58175; RRID: AB_940250

Cat# LCR-926-68071; RRID: AB_10956166

Cat# LCR-926-32210; RRID: AB_621842

Anti-rabbit IgG-IR680

Anti-mouse IgG-IR800

Chemicals, Peptides, and Recombinant Proteins

A-769662

LI-COR Biosciences

Abcam

Cat# ab120335

Cat# P7045

N/A

Phenformin

Sigma-Aldrich

SC-4

This study

cSC-4

This study

N/A

991

(Lai et al., 2014)

Sigma-Aldrich

N/A

Staurosporine

Cat# S4400

Cat# D6134

Cat# NET328250UC

Cat# NEC314050UC

Cat# E2311

N/A

2-deoxy-glucose

2-deoxy-[1,2-³H]-glucose

D-[1-¹⁴C]-mannitol

Fugene HD

Sigma-Aldrich

PerkinElmer Life Sciences

Promega

S108tide synthetic peptide

l-phosphatase

(Dite et al., 2017)

Cell Signaling Technology

(Oakhill et al., 2010)

Cat# P0753

N/A

Flag-CaMKK2

Deposited Data

a2b1g1/SC-4 structure

a2b2g1/SC-4 structure

This paper

PDB: 6B1U

PDB: 6B2E

Experimental Models: Cell Lines

COS7

ATCC

CRL-1651

CRL-1772

CRL-3216

N/A

C2C12

ATCC

HEK293T

ATCC

Immortalized MEFs WT

Immortalized MEFs AMPK b1^-/-/2^-/-

(Dite et al., 2017)

N/A

Experimental Models: Organisms/Strains

AMPK b2^-/-, on C57Bl/6 background

Oligonucleotides

(Steinberg et al., 2010)

N/A

AMPKa1-D141A ggtggtccatagagctttgaaacctgaaaatg

Sigma-Aldrich

N/A

AMPKb1-S108A

cccctcaccagagcccacaataactttgtagc

AMPKb1-N111A

Sigma-Aldrich

N/A

ccagaagccacaatgcctttgtagccatcc

AMPKb2-S108A

N/A

ccactgattaaggcccataatgactttgttgc

(Continued on next page)

Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018 e1

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

Continued

REAGENT or RESOURCE

SOURCE

IDENTIFIER

N/A

AMPKb2-D111A

Sigma-Aldrich

gattaagagccataatgcctttgttgccatcc

AMPKb2-DNterm

Sigma-Aldrich

N/A

ggtaccatggaggactccgtaaagc

Recombinant DNA

pBSSVD2005 SV40 large-T antigen expression

LeGO-iG2 AMPK b2-ﬂag

psPax2

A gift from David Ron

This study

Addgene #21826

N/A

A gift from Carl Walkley

(Scott et al., 2014)

pHCMV-EcoEnv

pET DUET AMPK his-a2/g1

pET RSF Duet AMPK b1

pET RSF Duet AMPK b2

pET RSF Duet AMPK b2(67-272)

pDEST27 GST-AMPK a1

pDEST27 GST-AMPK a2

pcDNA3 AMPK b1-ﬂag

pcDNA3 AMPK b2-ﬂag

pMT2-HA-AMPK g1

pMT2-HA-AMPK g2

pMT2-HA-AMPK g3

Other

Glutathione-Sepharose 4B

Streptavidin-Sepharose

GE Life Sciences

Cat#17075601

Cat#90100484

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed to and will be fulﬁlled by the Lead Contact: Jonathan

Oakhill (joakhill@svi.edu.au).

EXPERIMENTAL MODEL DETAILS

Animals

All animal procedures were approved by St. Vincent’s Hospital Animal Ethics Committee or the Ethics Committee of the Life Sciences

´

Sector, Universite catholique de Louvain (Male Wistar rats (170-180 g)). Whole body AMPK b2-KO C57BL/6 mice are described

previously (Steinberg et al., 2010).

Cell Lines

All cells were cultured in Dulbecco’s Modiﬁed Eagle’s medium (DMEM) containing 10% FBS and 1% penicillin/streptomycin

antibiotics, at 37^ꢁC with 5% CO₂. COS7 cells and C2C12 myoblasts were from the ATCC (Manassas, VA). C2C12

myoblasts were differentiated into myotubes by culturing in differentiation media (DMEM GlutaMax, supplemented with 2%

horse serum and 1% pencillin/streptomycin) for 7 days. MEFs were extracted from WT or homozygous AMPK b1b2 null embryos

(days 12–14 post-coitum), generated by crossing homozygous b1 and b2 null mice. WT and AMPK b1/2 double knockout

(b1/2-dKO) MEFs were immortalized by Fugene HD-mediated transfection with an SV40 large-T antigen expression construct.

(Dite et al., 2017).

e2 Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

METHOD DETAILS

Synthesis of SC4

Solvents were of analytical grade: acetonitrile (CH₃CN); ethyl acetate (EtOAc); dichloromethane (DCM); N,N-dimethylforma-

mide (DMF); methanol (MeOH); tetrahydrofuran (THF). Analytical TLC was performed on silica gel 60 F₂₅₄pre-coated

aluminium sheets (0.25 mm, Merck). Flash column chromatography was carried out with silica gel 60, 0.63–0.20 mm (70–230

mesh, Merck).

¹H and ¹³C NMR spectra were recorded at 400.13, and 100.62 MHz, respectively, on a Bruker Avance III Nanobay spectrom-

eter with BACS 60 sample changer, using solvents from Cambridge Isotope Laboratories. Chemical shifts (d, ppm) are reported

relative to the solvent peak (CDCl₃: 7.26 [¹H] or 77.16 [¹³C]; MeOH-d4: 3.31 [¹H] or 49 [¹³C]; DMSO-d6: 2.49 [¹H]). Proton res-

onances are annotated as: chemical shift (ppm), multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; dd, doublet

of doublets; td, triplet of doublets: v br, very broad), coupling constant (J, Hz), and number of protons. NH₂and OH protons are

in exchange.

Analytical HPLC was acquired on an Agilent 1260 Inﬁnity analytical HPLC coupled with a G1322A degasser, G1312B binary

pump, G1367E high performance autosampler, G4212B diode array detector. Conditions: Zorbax Eclipse Plus C18 Rapid reso-

lution column (4.6 3 100 mm) with UV detection at 254 nm and 214 nm, 30^ꢁC; sample was eluted using a gradient of 5–100%

of solvent B in solvent A where solvent A: 0.1% aq. TFA, and solvent B: 0.1% TFA in CH₃CN (5 to 100% B [9 min], 100% B

[1 min]; 0.5 mL/min).

Preparative HPLC was acquired on an Agilent 1260 Preparative HPLC coupled with a G1361A prep pump, G2260A

˚

autosampler, G1315D diode array detector. Conditions: Phenomenex Luna C8 (2) column 100 A Axia (250 3 21.2 mm; particle

size 5 mm) with UV detection at 254 nm and 210 nm, 30^ꢁC, ﬂow rate of 10 mL/min. Gradients are as speciﬁed in the individual

examples.

LCMS was performed on an Agilent 6100 Series Single Quad LCMS coupled with an Agilent 1200 Series HPLC, G1311A

quaternary pump, G1329A thermostatted autosampler, and G1314B variable wavelength detector (214 and 254 nm). LC

ꢁ

˚

conditions: Phenomenex Luna C8 (2) column (100 A, 5 mm, 50 3 4.6 mm), 30 C; sample (5 mL) was eluted using a binary

gradient (solvent A: 0.1% aq. HCO₂H; solvent B: 0.1% HCO₂H in CH₃CN; 5 to 100% B [10 min], 100% B [10 min]; 0.5 mL/min).

MS conditions: quadrupole ion source with multimode-ESI; drying gas temperature, 300^ꢁC; vaporizer temperature,

200^ꢁC; capillary voltage, 2000 V (positive mode) or 4000 V (negative mode); scan range, 100–1000 m/z; step size, 0.1 s

over 10 min.

Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018 e3

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

High resolution MS was performed on an Agilent 6224 TOF LCMS coupled to an Agilent 1290 Inﬁnity LC. All data were ac-

quired and reference mass corrected via a dual-spray electrospray ionisation (ESI) source. Each scan or data point on the total

ion chromatogram (TIC) is an average of 13,700 transients, producing a spectrum every second. Mass spectra were created by

averaging the scans across each peak and subtracting the background from ﬁrst 10 s of the TIC. Acquisition was performed

using the Agilent Mass Hunter Data Acquisition software ver. B.05.00 Build 5.0.5042.2 and analysis was performed using

Mass Hunter Qualitative Analysis ver. B.05.00 Build 5.0.519.13. Acquisition parameters: mode, ESI; drying gas ﬂow,

11 L/min; nebuliser pressure, 45 psi; drying gas temperature, 325^ꢁC; voltages: capillary, 4000 V; fragmentor, 160 V; skimmer,

65 V; octapole RF, 750 V; scan range, 100–1500 m/z; positive ion mode internal reference ions, m/z 121.050873 and

922.009798. LC conditions: Agilent Zorbax SB-C18 Rapid Resolution HT (2.1

3

50 mm, 1.8 mm column), 30^ꢁC;

sample (5 mL) was eluted using a binary gradient (solvent A: 0.1% aq. HCO₂H; solvent B: 0.1% HCO₂H in CH₃CN; 5 to

100% B [3.5 min], 0.5 ml/min).

6-Chloro-5-iodo-2-(methylsulfonyl)-3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo[4,5-b]pyridine (1)

Title compound was prepared according to literature procedure WO2012116145.

Methyl 5-((6-chloro-5-iodo-3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo[4,5-b]pyridin-2-yl)oxy)-2-

methylbenzoate (2)

To a solution of methyl 5-hydroxy-2-methylbenzoate (122 mg, 0.734 mmol) in DMF (6 mL) was added Cs₂CO₃(496 mg, 1.53 mmol)

and 6-chloro-5-iodo-2-(methylsulfonyl)-3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo[4,5-b]pyridine (300 mg, 0.612 mmol). The

reaction was stirred at room temperature for 2 h. The solvent was removed in vacuo, the residue acidiﬁed with a 10% a.q. citric

acid solution and extracted with EtOAc (33). The combined organic layers were washed with water, brine, dried (MgSO₄) and

concentrated in vacuo to give the title compound (337 mg, 96% yield). ¹H NMR (400 MHz, CDCl₃) d 7.93 (d, J = 2.68 Hz, 1H),

7.78 (s, 1H), 7.44 (dd, J = 2.70, 8.38 Hz, 1H), 7.35 (dd, J = 1.05, 8.38 Hz, 1H), 5.61 (s, 2H), 3.89 (s, 3H), 3.80 – 3.72 (m, 2H),

2.63 (s, 3H), 1.05 – 0.96 (m, 2H), -0.00 (s, 9H).

Methyl 5-((6-chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo[4,5-b]

pyridin-2-yl)oxy)-2-methylbenzoate (3)

A microwave vessel under nitrogen was charged with methyl 5-((6-chloro-5-iodo-3-((2-(trimethylsilyl)ethoxy)methyl)-3H-imidazo

[4,5-b]pyridin-2-yl)oxy)-2-methylbenzoate (337 mg, 0.587 mmol), Pd(PPh₃)₂Cl₂(21 mg, 0.0299 mmol), 4’-(4,4,5,5-tetra-

methyl-1,3,2-dioxaborolan-2-yl)-[1,1’-bipheny]-2-ol (190 mg, 0.642 mmol) and DMF (8 mL). The mixture was bubbled with

nitrogen for 5 min and a 1.0M solution of aq. K₂CO₃(0.875 mL, 0.875 mmol) added. The mixture was bubbled with

nitrogen for a further 5 min then heated at 120^ꢁC under microwave irradiation for 45 min. The reaction was cooled to room

temperature, concentrated, and then partitioned between water and EtOAc. The aq. layer was extracted with EtOAc (32)

and the combined organic phases washed with brine, dried over MgSO₄and the solvent removed in vacuo to give the

crude material. The product was puriﬁed by silica gel chromatography (0-30% EtOAc/petroleum benzine) to afford the title

compound (321 mg, 87% yield). ¹H NMR (400 MHz, CDCl₃) d 7.97 (d, J = 2.69 Hz, 1H), 7.93–7.86 (m, 3H), 7.61–

7.56 (m, 2H), 7.48 (dd, J = 2.71, 8.33 Hz, 1H), 7.37 (dd, J = 1.09, 8.34 Hz, 1H), 7.34–7.27 (m, 2H), 7.06–6.98 (m, 2H), 5.69

(s, 2H), 3.90 (s, 3H), 3.85–3.79 (m, 2H), 2.65 (s, 3H), 1.05–0.98 (m, 2H), -0.03 (s, 9H). LCMS: RT 6.052 min, m/z 484.1

[M-SEM+H]⁺.

5-((6-Chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-3H-imidazo[4,5-b]pyridin-2-yl)oxy)-2-methylbenzoic acid (SC4)

TBAF (1M in THF) (2.08 mL, 2.08 mmol) was added to a solution of methyl 5-[(6-ﬂuoro-5-(2’-hydroxybiphenyl-4-yl)-1-{(2-(trimethyl-

silyl)ethoxy]methyl}-lH-benzimidazol-2-yl)oxy ]-2-methylbenzoate (321 mg, 0.521 mmol) in THF (10 mL). The reaction was heated at

80^ꢁC for 6 h. The volatiles were removed in vacuo and the resultant residue dissolved in MeOH (6 mL) and a 2.5M aq. sol. of NaOH

(2 mL). The solution was heated at 70^ꢁC for 30 min after which the volatiles were removed in vacuo, the residue taken up in 10 mL

water and extracted with EtOAc. The combined organic layers were washed with brine, dried over MgSO₄and the solvent removed in

vacuo. The resultant crude material was puriﬁed by silica gel chromatography (0-100% EtOAc/petroleum benzine) to afford the title

compound (SC4) (70 mg, 28% yield, > 95% pure as judged by ¹H-NMR) as an off-white solid. Characterisation by ¹H NMR was in

agreement with the literature values. This compound has been synthesised previously in the literature: compound C-4,

WO2012033149 A1.

e4 Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

Synthesis of cSC4

6-Chloro-2’’-methoxy-4-nitro-[1,1’:4’,1’’-terphenyl]-3-amine (4)

4,5-Dichloro-2-nitroaniline (529 mg, 2.55 mmol) and 2’-methoxybiphenyl-4-boronic acid pinacol ester (0.71 g, 2.3 mmol) were

dissolved in 3 mL of 3:1 1,4-dioxane/water mixture. To this solution K₂CO₃(1.0 g, 7.7 mmol) and TBAB (82 mg, 0.26 mmol) were

added and the solution was degassed for 15 min. A catalytic amount of Pd(dppf)Cl₂(93 mg, 0.13 mmol) was then added and the

reaction mixture irradiated using a microwave reactor at 130^ꢁC for 3 h. After completion of the reaction as indicated by TLC analysis,

the reaction mixture was ﬁltered through celite and then evaporated to dryness. Puriﬁcation was performed by ﬂash chromatography

(5% EtOAc/petroleum spirits) to obtain the title compound as a yellow solid (519 mg, 63%). ¹H NMR (400 MHz, CDCl₃) d 8.28 (s, 1H),

7.67–7.59 (m, 2H), 7.51–7.45 (m, 2H), 7.45–7.40 (m, 1H), 7.40–7.33 (m, 2H), 7.29 (m, 1H), 7.06 (td, J = 7.5, 1.0 Hz, 1H), 7.04–7.00

(m, 1H), 6.85 (s, 1H), 3.85 (s, 3H); ¹³C NMR (101 MHz, CDCl₃) d 156.7, 148.1, 143.1, 139.3, 136.0, 131.0, 130.8, 129.5, 129.2,

128.8, 128.4, 127.0, 121.1, 120.8, 115.1, 111.5, 55.7; HPLC: RT 8.01 min > 99% purity at 254 nm; LCMS: m/z 354.8 [M+H]⁺;

HRMS: m/z 355.0844 [M+H]⁺, found m/z 355.0834

6-Chloro-2’’-methoxy-[1,1’:4’,1’’-terphenyl]-3,4-diamine (5)

6-Chloro-2’’-methoxy-4-nitro-[1,1’:4’,1’’-terphenyl]-3-amine (519 mg, 1.46 mmol) was dissolved in 1 mL ethanol and treated with

tin(II) chloride dihydrate (1.3 g, 5.9 mmol). The reaction mixture was then heated to 70^ꢁC for 24 h. After completion of reaction as

indicated by TLC analysis, KF (1 g, 17.2 mmol) in 2.5 mL of water was added and the solution stirred for 30 min. The reaction was

then quenched with water and washed with EtOAc (33). The combined organic layers were then dried with MgSO₄, ﬁltered

through a pad of silica and evaporated to dryness to yield the title compound as a yellow oil (426 mg, 89%). ¹H NMR (400 MHz,

CDCl₃) d 7.59–7.54 (m, 2H), 7.49–7.44 (m, 2H), 7.44–7.27 (m, 6H), 7.07–7.03 (m, 1H), 7.02 (m, 1H), 6.82 (s, 1H), 6.75 (s, 1H), 3.84

Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018 e5

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

(s, 3H); ¹³C NMR (101 MHz, CDCl₃) d 158.3, 156.5, 156.5, 148.1, 130.9, 130.9, 129.2, 129.1, 129.1, 128.8, 128.7, 128.6, 123.9, 121.0,

120.9, 111.3, 55.6; HPLC: RT 6.71 min > 85% purity at 254 nm; LCMS: m/z 324.9 [M+H]⁺.

5-Chloro-6-(2’-methoxy-[1,1’-biphenyl]-4-yl)-1,3-dihydro-2H-benzo[d]imidazole-2-thione (6)

To a mixture of 6-chloro-2’’-methoxy-[1,1’:4’,1’’-terphenyl]-3,4-diamine (100 mg, 0.3 mmol) dissolved in 0.5 mL ethanol, a solution of

KOH (21 mg, 0.37 mmol) dissolved in 0.1 mL water was added, followed by addition of CS₂(22 mL, 0.37 mmol). The solution was then

reﬂuxed for 5 h. After completion of the reaction as indicated by TLC analysis, the solution was carefully neutralised with 1M aq. HCl

and then washed with EtOAc (33). The combined organic layers were then washed with brine, dried with MgSO₄, ﬁltered and

evaporated to dryness. Puriﬁcation was performed by ﬂash chromatography (10% EtOAc/petroleum spirits) to obtain the title com-

pound as a white solid (50.1 mg, 45%). ¹H NMR (400 MHz, CDCl₃) d 7.62 (m, 2H), 7.50 (m, 2H), 7.43–7.31 (m, 3H), 7.11–6.92 (m, 3H),

3.85 (s, 3H); HPLC: RT 6.94 min > 98% purity at 254 nm; LCMS: m/z 364.8 [M-H]^-; HRMS: m/z 367.0666 [M+H]⁺, found m/z 367.0678.

6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(methylthio)-1H-benzo[d]imidazole (7)

To 5-chloro-6-(2’-methoxy-[1,1’-biphenyl]-4-yl)-1,3-dihydro-2H-benzo[d]imidazole-2-thione (0.28 g, 0.75 mmol) dissolved in 5 mL

acetone at 0^ꢁC, was added K₂CO₃(52 mg, 0.38 mmol) and iodomethane (24 mL, 0.38 mmol). The reaction was then stirred for 2 h

at room temperature. After completion of the reaction as indicated by TLC analysis, the reaction mixture was washed with EtOAc

(33). The combined organic layer was then washed with water, dried with MgSO₄, ﬁltered and evaporated to dryness to obtain

the crude product as a yellow oil (270 mg, 95%) which was directly used for the next step without further puriﬁcation. ¹H NMR

(400 MHz, CDCl₃) d 9.14 (s, 1H), 7.64–7.58 (m, 2H), 7.55–7.48 (m, 2H), 7.44–7.29 (m, 3H), 7.06 (m, 1H), 7.01 (m, 1H), 3.86 (s, 3H),

2.83–2.79 (m, 3H); ¹³C NMR (101 MHz, CDCl₃) d 156.5, 153.3, 138.5, 137.5, 134.8, 131.3, 131.0, 130.9, 130.2, 129.6, 129.5,

129.1, 129.1, 128.7, 126.7, 120.9, 111.3, 55.6, 14.7; HPLC: RT 6.8 min > 84% purity at 254 nm; LCMS: m/z 380.8 [M+H]⁺; HRMS:

m/z 381.0823 [M+H]⁺, found m/z 381.0821.

6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(methylsulfonyl)-1H-benzo[d]imidazole (8)

mCPBA (245 mg, 1.42 mmol) was added to a solution of 6-chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(methylthio)-1H-benzo[d]

imidazole (0.27 g, 0.71 mmol) in 5 mL DCM and the reaction stirred at room temperature for 10 min. The reaction mixture was

then washed with 10% aq. NaHCO₃solution, dried with MgSO₄, ﬁltered and evaporated to dryness. Puriﬁcation was performed

by ﬂash chromatography (25% EtOAc/petroleum spirits) to obtain the title compound as a white solid (110 mg, 37%). ¹H NMR

(400 MHz, CDCl₃) d 10.37 (d, J = 18.7 Hz, 1H), 7.97 (d, J = 46.5 Hz, 1H), 7.77–7.56 (m, 3H), 7.55–7.47 (m, 2H), 7.41 (dd, J = 7.5,

1.6 Hz, 1H), 7.36 (m, 1H), 7.10–6.99 (m, 2H), 3.87 (s, 3H), 3.44 (s, 3H); HPLC: RT 7.11 min > 99% purity at 254 nm; LCMS: m/z

410.9 [M-H]^-; HRMS: m/z 413.0721 [M+H]⁺, found m/z 413.0719.

6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(methylsulfonyl)-1-((2-(trimethylsilyl)ethoxy) methyl)-1H-benzo[d]

imidazole (9)

Et₃N (67 mL, 0.48 mmol) and SEMCl (56 mL, 0.31 mmol) were successively added to 6-chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-

(methylsulfonyl)-1H-benzo[d]imidazole (100 mg, 0.24 mmol) in 1 mL THF. The reaction was stirred for 1 h at room temperature

and then quenched with water. The aq. layer was extracted with EtOAc (33) and then the combined organic layers were washed

with 2M HCl, brine, dried with MgSO₄, ﬁltered and evaporated to dryness. Puriﬁcation was performed by ﬂash chromatography

(15-50% EtOAc/petroleum spirits) to obtain the title compound as a clear oil (67 mg, 52%). ¹H NMR (400 MHz, CDCl₃) d 7.93

(d, J = 46.5 Hz, 1H), 7.68 (m, 3H), 7.55–7.47 (m, 2H), 7.41 (dt, J = 7.5, 1.9 Hz, 1H), 7.39–7.32 (m, 1H), 7.07 (m, 1H), 7.04–6.98

(m, 1H), 5.92 (d, J = 6.2 Hz, 2H), 3.87 (d, J = 1.3 Hz, 3H), 3.74–3.63 (m, 2H), 3.55 (s, 3H), 0.95 (m, 2H), 0.05– -0.12 (m, 9H); HPLC:

RT 9.15 min > 93% purity at 254 nm; LCMS: m/z 542.8 [M+H]⁺; HRMS: m/z 543.1535 [M+H]⁺, found m/z 543.1529.

Methyl 5-((6-chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-1-((2-(trimethylsilyl)ethoxy)methyl)-1H-benzo[d]imidazol-2-

yl)oxy)-2-methylbenzoate (10)

6-Chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-2-(methylsulfonyl)-1-((2-(trimethylsilyl)ethoxy) methyl)-1H-benzo[d]imidazole (67 mg,

0.12 mmol) and methyl 5-hydroxy-2-methylbenzoate (31 mg, 0.19 mmol) were dissolved in 1 mL DMF to which K₂CO₃(66 mg,

0.48 mmol) was added and the reaction stirred at room temperature for 12 h. After completion of reaction as indicated by TLC anal-

ysis, the DMF was evaporated off and the residue was acidiﬁed with 1M aq. HCl. The aq. layer was washed with EtOAc (33),

combined organic layer washed with water and brine, dried with MgSO₄, ﬁltered and evaporated to dryness. Puriﬁcation was

performed by ﬂash chromatography (15% EtOAc/petroleum spirits) to obtain the title compound as a yellow oil (69 mg, 92%).

¹H NMR (400 MHz, CDCl₃) d 7.96–7.90 (m, 1H), 7.70–7.57 (m, 3H), 7.57–7.44 (m, 3H), 7.45–7.30 (m, 4H), 7.11–6.99 (m, 2H), 5.54

(d, J = 6.8 Hz, 2H), 3.88 (dd, J = 13.6, 3.5 Hz, 6H), 3.69 (dd, J = 16.5, 8.8 Hz, 2H), 2.64 (d, J = 3.4 Hz, 3H), 1.05–0.93 (m, 2H),

0.08–-0.06 (m, 9H); HPLC: RT 9.94 min > 99% purity at 254 nm; HRMS: m/z 629.2233 [M+H]⁺, found m/z 629.2236.

Methyl 5-((6-chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoate (11)

To a -78^ꢁC cooled solution of methyl 5-((6-chloro-5-(2’-methoxy-[1,1’-biphenyl]-4-yl)-1-((2-(trimethylsilyl)ethoxy)methyl)-1H-benzo

[d]imidazol-2-yl)oxy)-2-methylbenzoate (28 mg, 0.044 mmol) in 3 mL DCM was added 1M BBr₃in DCM (2 mL, 2 mmol) dropwise.

The reaction was stirred for 12 h allowing it to slowly warm up to room temperature. MeOH was carefully added to quench the

e6 Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

reaction and the reaction mixture evaporated to dryness. DCM was then added and the mixture ﬁltered to obtain the product as

an off-white solid (14 mg, 65%). ¹H NMR (400 MHz, MeOD) d 8.19–7.86 (m, 2H), 7.77–7.43 (m, 6H), 7.33 (m, 1H), 7.22–7.14

(m, 1H), 6.98–6.85 (m, 2H), 6.38 (d, J = 3.9 Hz, 1H), 3.97–3.90 (m, 3H), 2.71–2.61 (m, 3H); LCMS: m/z 484.7 [M+H]⁺.

5-((6-Chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoic acid (cSC4)

To a stirred solution of methyl 5-((6-chloro-5-(2’-hydroxy-[1,1’-biphenyl]-4-yl)-1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoate

(14 mg, 0.02 mmol) in 2 mL MeOH was added 2 mL 2M aq. NaOH solution and the reaction stirred at 45^ꢁC for 12 h. After completion

of the reaction as indicated by TLC analysis, the reaction mixture was neutralised to pH 7 using 6M aq. HCl solution and then

evaporated to remove MeOH. The residue was washed with EtOAc (33), dried with MgSO₄, ﬁltered and evaporated to dryness.

Puriﬁcation was performed by high performance liquid chromatography, eluting with 30-100% CH₃CN/water with 0.1% TFA, to

1

obtain the title compound as an off-white solid (2.5 mg, 26%). H NMR (400 MHz, MeOD) d 7.89 (m, 1H), 7.63 (m, 3H), 7.47–7.44

(m, 3H), 7.32 (m, 2H), 7.19–7.13 (m, 2H), 6.91 (m, 3H), 2.63 (s, 3H); HPLC: RT 6.59 min > 83% purity at 254 nm; LCMS: m/z 470.8

[M+H]⁺; HRMS: m/z 471.1106 [M+H]⁺, found m/z 471.1113.

Constructs for Crystallization

Heterotrimeric human AMPK His₆-a2b1g1 and His₆-a2b2g1 were expressed in E. coli strain Rosetta (DE3) using the pETDuet^TM-1

expression system (Novagen) (Scott et al., 2014). cDNAs for a2 and g1 were sequentially inserted into pETDuet^TM-1 multiple cloning

sites (MCS) 1 (BamH1/Not1) and 2 (MfeI/XhoI), respectively, resulting in incorporation of an N-terminal hexahistidine tag onto a2.

cDNAs for b1 or b2 was inserted into pRSFDuet^TM-1 MCS1 (NcoI/BamHI). All expression constructs were sequence veriﬁed.

Protein Expression and Puriﬁcation for Crystallization

Expression cultures were grown in Luria-Bertani broth and induced at 16^ꢁC with 0.25 mM isopropyl b-D-1-thiogalactopyranoside,

prior to overnight incubation. Cells were lysed using a precooled EmulsiFlex-C5 homogenizer (Avestin) and AMPK puriﬁed using

Nickel Sepharose and size exclusion chromatography as described previously (Langendorf et al., 2016). Phosphorylated AMPK

was generated by incubation with CaMKK2 (expressed in Sf21 cells as a C-terminal FLAG fusion, (Oakhill et al., 2010)), in the

presence of 2.5 mM MgCl₂, 0.5 mM ATP and 0.5 mM AMP (1 h, 22^ꢁC). Protein was re-puriﬁed using size exclusion chromatography,

concentrated to ꢀ10 mg/ml and ﬂash frozen in liquid nitrogen prior to storage at -80^ꢁC (storage buffer: 50 mM Tris.HCl, pH 8.0,

150 mM NaCl, 2 mM Tris(2-carboxyethyl)phosphine hydrochloride).

Crystallization

Full-length, phosphorylated a2b1g1 (4 mg/ml) was mixed with a three-fold molar excess of AMP and two-fold excess of S108tide

(NH₂-KLPLTRSHNNFVARRR-COOH) (Dite et al., 2017) and a 1:1 molar ratio of staurosporine and SC4. a2b1g1 crystals were grown

following the described method (Xiao et al., 2013; Langendorf et al., 2016). Protein/compound mixture was added 1:1 at 22^ꢁC with a

reservoir solution containing 8% PEG 3350, 0.1 M MgCl₂, 1.0% glucose, 0.001% cocamidopropyl betaine and 0.1 M imidazole

(pH 6.2), and incubated at 4^ꢁC for 1-2 weeks. Protein crystals were incubated with reservoir solution containing an additional 5%

glycerol, 5% PEG 400, 5% MPD, 5% sucrose, 5% sorbitol and 5% ethylene glycol for 1-2 min before ﬂash-cooling in liquid nitrogen.

Phosphorylated a2b2g1 protein was prepared as per a2b1g1. Crystals were grown by equally adding protein/compound mixture at

22^ꢁC with a reservoir solution containing 6-8% PEG 3350, 0.1 M MgCl₂, 0.001% cocamidopropyl betaine and 0.1 M imidazole

(pH 6.0), and incubating at 4^ꢁC for 1-2 weeks. Individual protein crystals were then transferred to a new crystallisation drop containing

identical reservoir solution, except the precipitant concentration was reduced to 4-5% PEG 3350, and incubated at 4^ꢁC for a further

1-2 weeks. Protein crystals were incubated with reservoir solution containing an additional 30% sorbitol for 2 min before ﬂash-cooling

in liquid nitrogen. Data were collected on the MX2 beamline at the Australian Synchrotron, part of ANSTO. Data were processed and

integrated using XDS (Kabsch, 2010) and scaled using AIMLESS from the CCP4 suite (Evans and Murshudov, 2013). The structure

was solved by molecular replacement using Phaser from the CCP4 suite (McCoy et al., 2007) and 4CFE (Xiao et al., 2013) as the

search model. Iterative rounds of model building and reﬁnement were performed using Coot (Emsley et al., 2010) and Buster

(https://www.globalphasing.com/buster/) (Bricogne et al., 2011) respectively. SC4 molecular coordinates and restraints were

€

generated using the PRODRG web-server (Schuttelkopf and van Aalten, 2004). Omit maps were calculated using Buster (Bricogne

et al., 2011). Structure validation and protein-SC4 interactions was performed using Molprobity (Chen et al., 2010), ﬁgures were

created using Pymol (https://www.pymol.org/).

Protein Expression and Puriﬁcation for AMPK Assays

Heterotrimeric AMPK was expressed in COS7 mammalian cells (Scott et al., 2014). For AMPK complex screening, 12 AMPK

heterotrimeric combinations (a1b1g1 through a2b2g3) were expressed in COS7 cells using transient transfection with FuGENE

HD, according to manufacturer’s protocol (Promega). Transfected AMPK heterotrimers consisted of N-terminal GST-tagged a1

or a2 (pDEST27 expression vector), C-terminal FLAG-tagged b1 or b2 (pcDNA3.1 expression vector) and N-terminal HA-tagged

g1, g2 or g3 (pMT2 expression vector). AMPK complexes from prepared COS7 lysates were puriﬁed using glutathione Sepharose.

Synergy experiments were conducted using l-phosphatase treated AMPK from COS7 cells. Brieﬂy, AMPK GST-a1/FLAG-b1/HA-g1

or GST-a2/FLAG-b2/HA-g1 complexes were immobilized on Glutathione Sepharose 4B, incubated on a rotating wheel with l-phos-

phatase for 1 h at 22^ꢁC (in Buffer A: 50 mM HEPES pH 7.4, 150 mM NaCl, 10% glycerol, 0.01% Tween-20, supplemented with 2 mM

Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018 e7

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

MnCl₂), followed by repeated washes in Buffer A and elution in Buffer A supplemented with 20 mM reduced glutathione. Lack of

AMPK phosphorylation on a-T172 and b-S108 was conﬁrmed by immunoblot, prior to assay.

Heterotrimeric a2b2DNtermg1 was generated by inserting cDNA for truncated b2 subunit (residues 67-272) into pRSFDuet^TM-1

MCS1 (NcoI/BamHI) and expressing in E.coli, as described above. Preparations were activated by CaMKK2 treatment during

puriﬁcation, as described above, prior to activity assay.

AMPK Activity Assay

AMPK activity assays using radiolabelled [g-³²P]-ATP were conducted as described previously (Scott et al., 2008). AMPK

heterotrimers puriﬁed from COS7 cells were immobilized on glutathione Sepharose as above and washed extensively with wash

buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 10% glycerol, 1 mM DTT and 0.1% Tween-20) prior to kinase reaction. Assays were

conducted in the presence of 100 mM SAMS synthetic peptide (sequence: NH₂-HMRSAMSGLHLVKRR-COOH), 5 mM MgCl₂,

200 mM [g-³²P] ATP for 10 min at 30^ꢁC SC4 or AMP (100 mM). For synergy experiments using dephosphorylated AMPK a1b1g1

and a2b2g1, additional ligand control A769662 (20 mM) was included. Phosphotransferase activity was quenched by spotting

onto P81 phosphocellulose paper (Whatman, GE Healthcare) followed by repeated washes in 1% phosphoric acid. ³²P transfer to

the SAMS peptide was quantiﬁed by liquid scintillation counting (Perkin Elmer).

Lentivirus Generation and Expression

Lentivirus expression constructs LeGO-iG2, second generation viral packaging vector psPax2 and ecotropic envelope vector

pHCMV-EcoEnv were gifts from Carl Walkley (St Vincent’s Institute of Medical Research). cDNA for human AMPK b2 was generated

with C-terminal FLAG-tag and cloned into LeGO-iG2 using BamHI/NotI restriction sites. Ecotropic lentivirus was generated by

transient transfection of HEK293T cells using calcium phosphate (Dite et al., 2017). 1 day prior to transfection, 2–2.5 3 10⁶

HEK293T cells were seeded per 10 cm culture dish. LeGO-iG2, psPax2 and pHCMV-EcoEnv plasmids (10 mg, 6.3 mg and 3.8 mg

per 10 cm culture dish, respectively) were mixed together with 2 M CaCl2 solution (244 mM ﬁnal concentration in 500 ml). The

DNA/CaCl2 solution was added drop-wise, with vortexing, into 500 ml of 2xHEPES-buffered saline, pH 7.06. After 20 min incubation

at RT, the mixture was transferred dropwise onto HEK293T cell culture and incubated. Within 16 h of transfection, cell were washed

with phosphate-buffered saline (PBS) and replaced with 6 ml fresh media. The lentivirus-containing supernatant was harvested after

48 and 72 h posttransfection and stored at ꢂ80^ꢁC.

iMEFs were transduced in 6-well plates by spinoculation. Brieﬂy, 1 day prior to transduction, 0.5 3 10⁵cells per well in a 6-well plate

were seeded in 2 ml media. 200 ml lentivirus supernatant in the presence of 8 mg/ml polybrene in 2 ml media was spun for 98 min at

25^ꢁC, 11003g (Heraeus Megafuge 2.0 R). Lentivirus containing media was replaced with an equal volume of fresh media after 24 h.

72 hours post-transduction, iMEFs were incubated with fresh media for 1 h and treated as indicated. Cells were harvested by washing

with ice-cold PBS, followed by rapid lysis in situ using 100 ml ice-cold lysis buffer (50mM Tris.HCl (pH 7.4), 150mM NaCl, 50mM NaF,

1mM sodium pyrophosphate, 1mM EDTA, 1mM EGTA, 1mM dithiothreitol, 1% (v/v) Triton X-100 and protease inhibitors), and cellular

debris was removed by centrifugation.

Cell Experiments

Differentiated C2C12 myotubes were pre-treated with fresh DMEM GlutaMax for 3 hours, prior to incubation with SC4 (0.03–3 mM) or

phenformin (2 mM) for 1 h. b1/2 dKO iMEFs were incubated with fresh DMEM for 3 h prior to incubation with SC4 (1–10 mM) or

phenformin (2 mM) for 1 h. All cells were washed with ice-cold phosphate buffered saline (PBS) and harvested by rapid lysis

using 150 ml ice-cold lysis buffer (50 mM Tris.HCl pH 7.4, 150 mM NaCl, 50 mM NaF, 1 mM sodium pyrophosphate, 1 mM EDTA,

1 mM EGTA, 1 mM DTT, 1% Triton X-100 and cOmplete protease inhibitor cocktail). Lysates were clariﬁed by centrifugation

(14,000 rpm; 3 min; 4^ꢁC) and ﬂash frozen in liquid N₂until processed.

Immunoblotting

Samples were separated by SDS-PAGE on a 12% gel (7% for ACC, previously enriched using streptavidin-Sepharose (Scott et al.,

2014)) and transferred to Immobilon-FL PVDF membrane (EMD Millipore). Membrane was blocked in PBS + 0.1% Tween-20 (PBST)

with 2% non-fat milk for 30 min at 22^ꢁC, then incubated for either 1 h or overnight with 1ꢁantibodies (dilutions in PBST). After washes

with PBST, membranes were incubated with anti-rabbit or anti-mouse IgG 2ꢁantibodies, ﬂuorescently-labelled with IR680 or IR800

dye, for 1 h. Immunoreactive bands were visualised on an Odysseyꢀ Infrared Imaging System with densitometry analyses

determined using ImageStudioLite software (LI-COR Biosciences).

Adenine Nucleotide Measurements

Adenine nucleotides were extracted from cells with perchloric acid (Scott et al., 2015). Relative concentrations were measured by

LC-MS on a QTRAPꢀ 5500 mass spectrometer (AB-SCIEX).

Mouse Skeletal Muscle Isolation and Glucose Uptake Assay

Soleus and extensor digitorum longus (EDL) were isolated from anaesthetized mice (mixture of xylazine (20 mg/kg) and ketamine

(100 mg/kg) in a volume of 10 ml/g body weight), transferred to glass vials containing 2 ml of basal buffer (Krebs-Henseleit bicarbonate

buffer pH 7.4 supplemented with 2 mM pyruvate, 8 mM mannitol and 0.1% BSA; gassed with 95% O₂+ 5% CO₂) and pre-incubated

e8 Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018

Please cite this article in press as: Ngoei et al., Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK a2b2g1 by the Glucose

Importagog SC4, Cell Chemical Biology (2018), https://doi.org/10.1016/j.chembiol.2018.03.008

in a shaking incubator at 30^ꢁC for 10 min for recovery (Steinberg et al., 2010). Muscles were incubated in basal buffer SC4 (10 mM))

for 30 minutes, in a shaking incubator at 30^ꢁC with a gas phase of 95% O₂+ 5% CO₂. 2-deoxy-D-glucose (2DG) uptake was

measured over 20 min by replacing basal buffer with glucose uptake buffer (basal buffer supplemented with 1.5 mCi/ml 2-deoxy-

[1,2-³H]-glucose and 0.2 mCi/ml D-[1-¹⁴C]-mannitol

SC4 (10 mM)). Muscles were washed in ice-cold PBS, homogenized in

300 ml ice-cold lysis buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 5 mM EDTA, 100 mM NaF, 1 mM Na₃VO₄, 10 mM sodium

pyrophosphate, 250 mM sucrose, 1% Triton X-100 and cOmplete protease inhibitor cocktail) using an electrical homogenizer,

and clariﬁed by centrifugation (14,000 rpm, 5 min, 4^ꢁC). The incorporated radioactivity in muscle extracts (100 ml) was measured

by liquid scintillation counting (Perkin Elmer), and calculated as the accumulation of intracellular 2-deoxy-[1,2-³H]-glucose.

Incubation and Electrical Stimulation of Isolated Rat Skeletal Muscles

Anesthetized rats were euthanized by cervical dislocation and epitrochlearis muscles were rapidly removed and mounted on an

incubation apparatus. Epitrochlearis muscles were incubated (Lai et al., 2014) in the presence of 991 (5 mM), SC4 (1 mM and 2 mM

) or vehicle (0.1% DMSO) for 60 min with or without contraction induced by electrical stimulation (200 ms trains delivered every

5 s at 100 Hz and 30 V) for 30 min.

QUANTIFICATION AND STATISTICAL ANALYSIS

Data analyses were performed using GraphPad Prism 7 software. Results from replicate experiments (n) were expressed as means

standard error (s.e.m.). All statistical tests were performed using one-way or two-way analysis of variance (ANOVA), or unpaired two-

tailed Student’s t test, where appropriate. Values for EC₅₀and maximum response were calculated from a non-linear regression ﬁt

using Prism 7 software.

DATA AND SOFTWARE AVAILABILITY

The co-ordinates for SC-4 complexed to a2b1g1 and a2b2g1 have been deposited in the PDB under ID codes 6B1U and 6B2E,

respectively.

Cell Chemical Biology 25, 1–10.e1–e9, June 21, 2018 e9

Article Doi

Structural Determinants for Small-Molecule Activation of Skeletal Muscle AMPK α2β2γ1 by the Glucose Importagog SC4

DOI: 10.1016/j.chembiol.2018.03.008

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.chembiol.2018.03.008

Full text of DOI:10.1016/j.chembiol.2018.03.008

Products guided by the article

R&D Labs maybe for 1364692-88-5

Relevant to this article

Hot Product