such as glucose-stimulated insulin secretion, growth, or apoptosis,
remains controversial [5, 6]. Moreover, recent studies suggest that
chronic alteration ofβ-cell AMPK function might aggravate the
disease [5–8]. Several variables, including the type of cells studied
(β-cell lines versus isolated β-cells and/or islets), the context
(in vitro or in vivo), conditions of cell culture (glucose concentra-
tion, length of the experiments), or the use of different approaches
to AMPK activity modulation (e.g., pharmacological versus
genetic), have surely contributed to the mixed, even controversial,
results obtained by different research groups throughout the years.
AMPK consists of three subunits:α(catalytic),β(regulatory),
andγ(regulatory), each of them with 2 (α,β)or3(γ) isoforms. In
islets, AMPKα1 accounts for most of the catalytic activity [5]. The
kinase activity of the catalytic subunit is increased by phosphoryla-
tion of a threonine residue (in rat, Thr-172) by upstream kinases
such as LKB1 [9]. Binding of AMP allosterically activates AMPK,
promotes its phosphorylation by LKB1, and prevents its dephos-
phorylation by phosphatases (reviewed in [10]). Giving the central
role of AMP in increasing AMPK activity, it is not surprising that
adenosine analogues are often used by researchers and clinicians as
AMPK activators. One of the most widely used AMPK activators is
5-aminoimidazole-4-carboxamide riboside (AICAR), which is
intracellularly converted into the AMP analogue ZMP
[11, 12]. Nevertheless, being a potent analogue of AMP, ZMP
also affects the activity of multiple other enzymes important for
cell metabolism and has been recently demonstrated to act through
multiple AMPK-independent pathways [13–15]. Thus, our current
protocols for the study of AMPK function in human islets favor the
use of more potent and selective AMPK activators, such as com-
pound 13 (C-13, [16]) or compound 991 (C-991, [17]). On the
other hand, compound C has also been widely used as a rather
unspecific inhibitor of AMPK (seeChapter 12) [18].
Recently, we have generated mice with highlyβ-cell-restricted
deletion of both catalytic (α1 andα2) AMPK subunits [7]. These
animals represent a valuable model for study of AMPK function in
theβ-cell in vivo and also provide a convenient source of islets for
in vitro studies which are null for AMPK. Islets from AMPKdKO
animals (AMPKα1flox/flox, AMPKα2flox/flox, Ins1Cre/) and
littermate controls (AMPKα1flox/flox, AMPKα2flox/flox) were
isolated as described in Subheading3.1 and used for RNA sequenc-
ing [ 7, 19] and insulin secretion (Fig.1 [7]) as detailed in Sub-
headings3.4 and 3.5, respectively.
Here we aim to provide a detailed description of the protocols
that our group has applied to the study of AMPK function in mouse
and human pancreatic islets over the last 15 years.
The main methods described in this chapter are briefly listed
below.414 Aida Martinez-Sanchez et al.