AMPK Methods and Protocols

[2]. This binding site appears to be stabilized by an ionic interaction

between phosphorylated Ser108 onβ1 (which can be modified by

autophosphorylation) and the side chain of Lys31 onα2, while the

side chain of Lys29 directly interacts with the bound ligand. Con-

sistent with this, mutation of either Ser108 onβ1 or of Lys29 and

Lys31 onα2, to alanine (β1 S108A,α2 K29A/K31A), abolishes or

greatly reduces activation by A769662 or 991. We have shown that

AMPK containing theβ1-S108A mutation is not activated, or only

poorly activated, by the ADaM site ligands salicylate and A769662

but is still activated by quercetin in intact cells [6]. We have also

shown that AMPK containing the K40A/K42A mutations inα 1

(equivalent to K29A/K31A inα2) is not activated by A769662 in

intact cells, although it is still activated by an agent that binds in the

catalytic site, SU6656 [7]. Surprisingly, the K40A/K42A mutant,

although still allosterically activated by AMP in cell-free assays, is

not activated by increased Thr172 phosphorylation in intact cells

by agents that increase cellular AMP/ADP (phenformin, berberine,

troglitazone, or oligomycin) [7]. Thus, an intact ADaM site is

required for the effects of AMP-elevating agents to promote

Thr172 phosphorylation, although not for allosteric activation.

This may also be true of the S108A mutation inβ1, since quercetin,

the positive control used in the original study [6], may act in part by

an AMP-independent mechanism [3]. Thus, while the RG mutants

discriminate between agents acting at the AMP-binding and ADaM

sites, theβ1-S108A orα1-K40A/K42A mutants appear not to

distinguish between them in terms of Thr172 phosphorylation,

although doing so in terms of allosteric activation.

The levels of adenine nucleotides in intact cells can be deter-

mined either by capillary electrophoresis (CE), high-performance

liquid chromatography (HPLC), or liquid chromatography-mass

spectrometry (LC-MS). However, due to the low levels of AMP in

cells under basal conditions, CE or HPLC may not be sufficiently

sensitive to reliably detect and quantify AMP by ultraviolet absor-

bance alone. The ADP/ATP ratio, which can be used as a surrogate

for the AMP/ATP ratio [8], can be readily measured using CE or

HPLC, while the large increase in sensitivity of LC-MS also allows

AMP to be measured under all conditions. However, since the

recoveries of AMP, ADP, and ATP during MS analysis are not the

same, it is essential to run appropriate standards to correct for these

differences.

As an adjunct to measuring adenine nucleotides, the use of an

Extracellular Flux Analyzer to measure oxygen consumption rate

(OCR) can help to confirm whether a compound that increases

cellular ADP/ATP and/or AMP/ATP ratios is doing so by inhibit-

ing mitochondrial function.

Finally, a crucial point to note when harvesting AMPK for

kinase assays, as well as when monitoring the phosphorylation

status of AMPK or downstream targets by Western blotting, is

Intact Cell Assays to Determine Contributions of AMP-Binding and ADaM Sites 241