The Plant Journal (2018) 96, 163–175 doi: 10.1111/tpj.

14026

A role for the carbohydrate-binding module (CBM) in

regulatory SnRK1 subunits: the effect of maltose on SnRK1
activity
Ana Ruiz-Gayosso, Rogelio Rodr
ıguez-Sotres, Eleazar Mart
ınez-Barajas and Patricia Coello*
Departamento de Bioquımica, Facultad de Quımica, Universidad Nacional Auto noma de Me xico, 04510, Ciudad de Me xico,
Me xico

Received 5 April 2018; revised 19 June 2018; accepted 26 June 2018; published online 12 July 2018.
*For correspondence (e-mail pcoello@unam.mx).

SUMMARY

SnRK1 is a protein kinase complex that is involved in several aspects of plant growth and development.
There are published data indicative of a participation of SnRK1 in the regulation of the synthesis and degra-
dation of starch, although the molecular mechanism is not known. In this work, we performed electron
microscopy to explore the in vivo localization of the regulatory and catalytic subunits that constitute the
SnRK1 complex. The results indicated that all the subunits are present in the chloroplast and, in particular,
the SnRK1 bc and SnRK1 b3 subunits are associated with starch. Furthermore, the regulatory subunits bind
maltose, a relevant product of starch degradation. The kinase activity of immunoprecipitated complexes
containing the bc regulatory subunit was positively regulated by maltose only in the complexes obtained
from Arabidopsis leaves collected at dusk. Recombinant complexes with the SnRK1a1 catalytic subunit,
SnRK1bc and three different b subunits showed that maltose only had an effect on a complex formed with
the b3 subunit. Truncation of the CBM domain form SnRK1 bc abolished the maltose activation of the com-
plex and the activity was significantly reduced, indicating that the CBM is a positive regulator of SnRK1. A
model of the SnRK1a1/bc/b3 complex suggests the presence of two putative maltose-binding sites, both
involving ligand interactions with the bc subunit and the a subunit.

Keywords: SnRK1, starch, maltose, chloroplast, CBM.

INTRODUCTION
Plants can sense and integrate a wide variety of external
and internal stimuli that compromise their growth and
development via signaling pathways that regulate gene
expression and metabolism. SnRK1 (sucrose non-ferment-
ing related kinase 1) is a protein kinase involved in low-
energy and carbon-status signaling (Baena-Gonza
lez et al.,
2007; Fragoso et al., 2009; Coello et al., 2011). SnRK1 also
participates in fundamental processes such as cell prolifer-
ation, embryo and pollen development, germination,
cotyledon growth and senescence (Lumbreras et al., 2001;
Zhang et al., 2001; Lovas et al., 2003; Wingler et al., 2009;
Radchuk et al., 2010; Gue
rinier et al., 2013) as well as regu-
lation of leaf starch metabolism. In photosynthetic tissue,
mutants exhibiting reduced SnRK1 activity accumulate leaf
starch during periods of darkness (Baena-Gonza
lez et al.,
2007). Furthermore, snrk1 mutants fail to mobilize transi-
tory leaf starch under conditions of phosphate starvation
(Fragoso et al., 2009). In addition, overexpression of potato
SnRK1 in tobacco leaves was seen to increase the starch
and soluble sugar content and stimulate the expression of
starch biosynthesis genes such as sucrose synthase, ADP-
glucose pyrophosphorylase and soluble starch synthase
(Wang et al., 2017). These results indicate a SnRK1 involve-
ment in the control of synthesis and/or starch degradation,
if so, a fine-tuned regulation of SnRK1 would be required
to control in concert these antagonistic processes.
The SnRK1 protein kinase is the plant ortholog of AMPK
(AMP-activated protein kinase) in mammals and SNF1 (su-
crose non-fermenting 1) in yeast. It shares with these pro-
teins a heterotrimeric structure consisting of a catalytic a
subunit and two regulatory b and c subunits. The Ara-
bidopsis genome encodes three catalytic subunit isoforms
(SnRK1a1, SnRK1a2, and SnRK1a3), three isoforms of the
regulatory b subunit (SnRK1b1, SnRK1b2, and SnRK1b3),
and one c subunit (SnRK1c1). Additionally, there is a plant-
exclusive SnRK1bc subunit that contains a carbohydrate-
binding module (CBM) at the N termini (present usually in
some of the b subunits) as well as the classic cystathionine

© 2018 The Authors 163

The Plant Journal © 2018 John Wiley & Sons Ltd
164 Ana Ruiz-Gayosso et al.
b-synthase domain (CBS), commonly found at the C ter-
mini of the c subunits (Lumbreras et al., 2001; Gissot et al.,
2006; Polge and Thomas, 2007). In particular, the SnRK1b1,
SnRK1b2 and SnRK1b3 subunits act as scaffolds, which
enables complex formation via association with the C-
terminal SNF1 complex (ASC) domain (Gissot et al., 2006;
Figure S1). In mammals, the CBM known as the glycogen-
binding domain (GBD) has been proposed to be a regula-
tory domain that inhibits AMPK activity because this
domain binds to glycogen in vitro and acts as a sensor of
glycogen as stored carbon source (Polekhina et al., 2003;
McBride et al., 2009; Koay et al., 2010). Additionally, the
presence of the GBD could recruit the AMPKb subunits to
the vicinity of potential targets involved in glycogen meta-
bolism such as glycogen synthase and glycogen phospho-
rylase (Hudson et al., 2003).
The role of the CBMs of SnRK1b1, SnRK1b2, and
SnRK1bc has been formerly evaluated, but evidence
regarding the ability of these CBMs to bind starch is con-
troversial. In a previous report, SnRK1b2 and SnRK1bc
monomers were found to bind starch from Arabidopsis
leaf in vitro via the CBM and in the presence of the
polysaccharide, SnRK1 activity decreases significantly


(Avila-Casta ~ eda et al., 2014); however, in a later study the
n vided.
isolated subunits failed to bind starch from maize grains or
related carbohydrates (Emanuelle et al., 2015).
The CBMs of SNF1/AMPK/SnRK1 belong to the CBM48
family (CAZy database), which is related in sequence to the
CBM20 family (Christiansen et al., 2009; Janecek et al.,
2011). It is well known that the CBM20 family is present in
a wide variety of both amylolytic and non-amylolytic
enzymes that are capable of binding starch and one or
more starch-related carbohydrates such as maltose,
maltotriose, maltoheptaose, maltodecaose, and a-, b- and
c-cyclodextrin (Christiansen et al., 2009). Due to the evolu-
tionary and structural relationship between both CBM
families, it is possible that members of the CBM48 family
could bind other starch-related carbohydrates.
Starch is the main storage polysaccharide in plant cells.
During the day, transitory starch is synthesized and stored
as compact granules inside chloroplasts. At night, leaf
starch is almost completely degraded to supply sugars to
sink tissues. Starch degradation involves a group of
enzymes driving granule solubilization by phosphoryla-
tion/dephosphorylation reactions (Yu et al., 2001; Hejazi
et al., 2009; Nashilevitz et al., 2009) and releases a variety
of soluble sugars such as malto-oligosaccharides (MOS) of
variable length and maltose (Crichley et al., 2001; Chia
et al., 2004; Streb and Zeeman, 2012). Maltose, being the
sugar with the largest export flux from the chloroplast to
cytoplasm, is also the main product of this nocturnal starch
degradation (Niittyla€ et al., 2004; Weise et al., 2004). The
maltose produced by BAM (b-amylase) activity is trans-
ported to the cytoplasm, where it is metabolized to release
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
one unit of glucose while the second unit is transferred to
a soluble heteroglycan (Chia et al., 2004; Fettke et al., 2004,
2005; Lu and Sharkey, 2004).
Previously, our group identified the subcellular localiza-
tion of SnRK1b2 and SnRK1bc by immunohistochemical
techniques and found that the proteins were distributed


between the chloroplast and the cytoplasm (Avila-Cas-
~
taneda et al., 2014). To obtain additional information with
regard to the function of the CBM, here we report trans-
mission electron microscopy evidence of the in vivo local-
ization of the SnRK1b and SnRK1bc subunits. In addition,
we investigated the ability of the SnRK1b and SnRK1bc
subunits to bind other types of sugars, such as trehalose-
6-phosphate (T6P), or glucose-6-phosphate (G6P), which
inhibit SnRK1 activity (Nunes et al., 2013) and maltose,
which is the main product of starch degradation. From
intrinsic protein fluorescence (IPF) and differential scan-
ning fluorimetry (DSF) data, maltose was the only ligand
of the CBM-containing subunits, at physiological concen-
trations. When tested on different SnRK1 recombinant
complexes, maltose affected the activity of only SnRK1a1/
SnRK1bc/b3, and evidence of a positive regulation of the
catalytic activity of this complex by its CBM is also pro-
RESULTS
Localization of the SnRK1bc and SnRK1b subunits
The role of the CBMs in the SnRK1b and SnRK1bc subunits
of the SnRK1 complex is not very well understood. Previ-
ous results from our laboratory revealed an interaction of
SnRK1bc and SnRK1b2 with starch purified from leaves in


an in vitro assay (Avila-Casta ~ eda et al., 2014). To obtain
n
more resolution of the cellular localization of the SnRK1
complex, we performed immunogold detection of the b
subunits using antibodies against SnRK1b1, SnRK1b2,
SnRK1b3, and SnRK1bc subunits to identify their in vivo
subcellular distribution. Although the label was found both
in cytoplasm and chloroplast (Figure 1, left column), the
analysis revealed significant label inside the chloroplast for
SnRK1bc and SnRK1b3 and notably, immunogold markers
were observed in starch, suggesting an association of both
proteins to the polysaccharide (Figure 1). For SnRK1b1 and
SnRK1b2, some signal was detected inside the chloroplast,
but only weak immunogold labeling of the starch was
observed, especially for the former, indicating the occur-
rence of low association with the polysaccharide (Fig-
ure 1). Pre-immune serum used as control did not produce
detectable gold signal (Figure S2). The specificity of the
SnRK1bc and SnRK1b3 antibodies was evaluated using an
antibody blocked by the soluble respective antigen and
this resulted in loss of signal; additionally, all the antibod-
ies were tested in a western blot showing that a single
band at the correct molecular weight was recognized
© 2018 The Authors
 Maltose and SnRK1 activity 165

(Figure S3). Interaction of SnRK1b3 with the starch is diffi-
cult to explain due to the absence of a CBM in the protein;
however, dimers between SnRK1bc and SnRK1b3 are
formed at least during co-expression of both recombinant
proteins (Maya-Bernal et al., 2017), so it might be possible
that binding of SnRK1b3 to starch depends on the CBM
domain present in SnRK1bc. In addition to localization in
the chloroplast, all the subunits were also localized in the
cytoplasm. We also evaluated whether the catalytic sub-
units SnRK1a1 and SnRK1a2 were present in this orga-
nelle. The results indicated that both catalytic subunits
were localized inside the chloroplast, and a fraction of
these subunits were associated with the starch (Figure S4).
The co-localization of the SnRK1 subunits inside the
chloroplasts could drive the formation of different SnRK1
complexes.
Binding of maltose to the SnRK1bc and SnRK1b subunits
The carbohydrate-binding specificity of the CBMs in
SnRK1bc and SnRK1b subunits has not been explored in
enough detail. The AMPK complex can bind a range of
glycogen mimetic oligosaccharides, including maltose. As
maltose is a disaccharide produced by starch degradation
and can accumulate inside and outside chloroplasts, we
analyzed the capacity of the subunits to bind this carbohy-
drate. Initially, intrinsic fluorescence spectroscopy was uti-
lized to measure binding. The emission spectra of both
SnRK1b1 and SnRK1b2 showed fluorescence quenching
after the addition of maltose (Figure S5a and b). SnRK1bc
was expressed as a soluble SnRK1bc/b3 dimer because the
SnRK1bc protein accumulates in inclusion bodies when
expressed alone (Maya-Bernal et al., 2017). The dimeric
protein SnRK1 bc/b3 also showed a clear fluorescence
quenching (Figure S5c). The E. coli maltose-binding pro-
tein (MBP) was expressed and used as a positive control
showing a similar response (Figure S5d) and the dimer
SnRK1bc (DCBM)/b3, which lacks the CBM of the SnRK1bc
subunit, was used as a negative control (Figure S5e). Satu-
ration kinetics of the fluorescent quenching resulted in
classical binding curves (Figure 2a). SnRK1b1, SnRK1b2,

Figure 1. Subcellular localization of the regulatory SnRK1b and SnRK1bc subunits in Arabidopsis leaves. Immunogold assays were used to detect SnRK1b1,
SnRK1b2, SnRK1b3 and SnRK1bc using specific antibodies. The first column represents a general view of the field, and magnified views of specific areas are
included. The organelles indicated are cytoplasm (C), chloroplast (Ch), mitochondria (M), vacuole (V), and starch (S). Scale bars are shown.

© 2018 The Authors

The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
166 Ana Ruiz-Gayosso et al.

Figure 2. Monomers of SnRK1b and SnRK1bc as well as SnRK1bc-b dimers are capable of binding maltose. (a) Intrinsic protein fluorescence (IPF) measurements
were performed to evaluate the capacity of one CBM to bind maltose. Relative fluorescence values were used to calculate the Kd of each protein. Maltose-bind-
ing protein (MBP) was used as a positive control, whereas SnRK1bc (DCBM) was used as a negative control. Differential scanning fluorimetry (DSF) was per-
formed to measure the maltose-binding activity of SnRK1b1 (b) and SnRK1b2 (c) monomers or dimers of SnRK1bc or SnRK1bc (DCBM). The same method was
used to determine the Kd values of the SnRK1bc/b3 and SnRK1bc (DCBM)/b3 dimers (d). The data represent the values of three biological replicates.

Table 1 Kd values for maltose binding and SnRK1bc/b3 bound maltose in a saturable manner with
Kd values in the micromolar range (Figure 2a, Table 1). In
IPF DSF
Protein Kd (lM) Kd (lM)
addition, DSF assays were performed to determine the
thermal stability of the protein in the presence of increas-
MBP 5.099
0.467 4.277
1.696 ing amounts of maltose. We estimated the Kd using the Tm
SnRK1b1 10.800
0.992 13.345
0.706 at different maltose concentrations (Figure 2b, c and d;
SnRK1bc/b1 ND 3.002
0.206
SnRK1bc(DCBM)/b1 ND 12.584
0.871
Table 2). These data confirmed the observation derived
SnRK1b2 13.900
1.226 12.803
0.429 from the fluorescence quenching as all the subunits could
SnRK1bc/b2 ND 3.013
0.144 bind maltose, and the Kd values obtained for the dimers
SnRK1bc(DCBM)/b2 ND 8.986
0.796 SnRK1bc/b1 and SnRK1bc/b2 were in agreement with the
SnRK1bc/b3 7.451
0.808 3.572
0.568 previous estimates, except perhaps for the SnRK1bc/b3-
SnRK1bc(DCBM)/b3 NA NA
maltose Kd, which was about half the fluorescence

© 2018 The Authors

The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
 Maltose and SnRK1 activity 167

0.060
0.001

0.068
0.001
0.083
0.001
estimate, but still in the same order of magnitude. The data
also reveal a four-fold reduction in the apparent Kd for mal-

ND
tose of the SnRK1bc/b1 and the SnRK1bc/b2, with respect

sec1
Kcat
to the Kd of SnRK1b1 and SnRK1b2. Notably, all three
dimers, SnRK1bc/b1, SnRK1bc/b2, and SnRK1bc/b3, bound

9.00
0.89

7.99
0.59
9.78
1.62
to maltose with a similar Kd, but the removal of the CMB in
the bc subunit (SnRK1bc-ΔCBM) form SnRK1bc/b1 and

ND
Km

lM
SnRK1bc/b2 dimers did not abolish maltose binding. No
signal was observed with the CBM-free SnRK1bc/(ΔCBM)/
1
mg

b3. In order to test the binding specificity, we performed

SnRK1a1/bc/b3

61.82
0.98

69.25
1.48
84.71
0.60
1

DSF assays using other carbohydrates, but sucrose, T6P,

nmol min

and G6P did not bind to any of the dimeric complexes (Fig-
ND
Vmax

ure S6).
Effect of maltose on SnRK1 activity
0.001
0.001
0.002
0.001

To detect complex formation in vivo, we performed

1

0.051
0.052
0.055
0.054

immunoprecipitation (IP) of SnRK1 complexes from leaf

Kcat

sec

extracts obtained at two time points during the day (dawn

and dusk) using antibodies against the SnRK1bc subunit
0.20
1.34
1.37
0.14

(Figure S7). The IP experiment showed high activity asso-

11.91
11.01
10.57
7.76

ciated with the SnRK1bc subunit, and only complexes

Km

lM

formed at dusk showed a significant increase in activity

SnRK1a1/bc/b2

when maltose was added to the assay (Figure 3). To

1.09
1.10
1.52
1.12
min1 mg1

determine whether a specific complex was stimulated by

maltose, we expressed recombinant SnRK1 subunits, and

53.00
54.00
56.82
55.88
nmol
Vmax

the kinase activity was assayed in the presence of maltose

(Maya-Bernal et al., 2017). Neither was the SnRKa1 activity
0.003
0.005
0.001
0.005

stimulated by maltose nor was the activity of the com-

plexes formed by SnRK1a1/bc/b1 and SnRKa1/bc/b2

Table 2 Kinetic parameters of SnRK1 complexes at increasing concentrations of maltose

1

0.054
0.051
0.057
0.053
Kcat

sec

1.20
0.56
2.09
0.76








13.39
11.24
12.12
10.81
Km

lM
SnRK1a1/bc/b1

0.49
0.53
0.67
0.64
min1 mg1

54.91
54.19
58.37
54.13
nmol
Vmax

0.032
0.001

0.036
0.004
ND
ND
1
Kcat

sec

6.89
0.54

8.61
1.48
ND
ND
Km

lM

Figure 3. SnRK1 kinase activity in immunoprecipitates from leaves har-

32.71
0.68

36.97
0.54
min1 mg1

ND, not determined.

vested at dawn and dusk. Crude extract (CE) proteins extracted from
SnRK1a1

rosettes harvested at dawn or dusk were incubated with primary rabbit anti-
ND
ND
nmol
Vmax

SnRK1bc antibodies. Kinase activity was measured in the immunoprecipi-

tates and in the CE at each time point in the presence (gray) or absence
(black) of maltose. Values are the mean
SEM of five replicates. Specific
[Maltose]

activity for CE at dawn and dusk were 6.095
0.137 nmol Pi min1 mg1
and 1.86
0.056 nmol Pi min1 m1, respectively. The letters indicate sta-
1000
200

tistically significant differences (P < 0.001).

lM

10
0

© 2018 The Authors

The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
168 Ana Ruiz-Gayosso et al.
Figure 4. Effect of maltose on SnRK1 kinase activity in heterotrimeric com-
plexes assembled in vitro. SnRK1a1/bc/b heterotrimeric complexes, includ-
ing SnRK1bc (DCBM), were assembled in vitro to test the effect of maltose
binding on SnRK1 activity. As a control, SnRK1a1 recombinant subunit
activity in the absence of carbohydrate was measured and considered as 1
on the relative activity scale. Values are the mean
SEM of three repli-
cates. The letters indicate statistically significant differences (P < 0.001).
(Figure 4). Interestingly, the activity of the complex
formed by SnRK1a1/bc/b3 increased more than 30% when
1 mM maltose was added. Unexpectedly, the complex
formed by the SnRK1bc subunit that lacked the CBM had
less activity than the catalytic subunit alone, indicating
that the CBM acts as a positive regulator of the catalytic
subunit. The negative effect was not observed when
SnRK1a1/bc (DCBM)/b1 or SnRK1a1/bc (DCBM)/b2 were uti-
lized, indicating that the presence of at least one CBM is
important for kinase activity. No effect was observed with
the SnRK1a1/bc (DCBM)/b3 complex when maltose was
added, suggesting that the presence of the bc–CBM is
necessary for maltose to have an effect (Figure 4) and
while the CBM in b1 or b2 can render the complex active,
these CBMs seem not responsive to maltose. When the
effect of different carbohydrates on SnRK1a1/bc/b3 activity
was evaluated, we observed activation by maltose and no
effect was observed with the remaining carbohydrates
tested (Figure 5).
To determine if maltose was a K- or V-type activator of
the different complexes, we tested the saturation kinetics
of the complex at variable concentrations of the AMARA
peptide and fixed concentrations of Mg2+ and ATP, in the
absence or presence of maltose. As shown in Figure 6, the
kinetics for all the complexes followed a Michaelis–Menten
response, but an addition of up to 1 mM maltose only
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Figure 5. In vitro-assembled heterotrimeric SnRK1 activity in the presence
of soluble sugars. Heterotrimeric complexes were assembled, and the
SnRK1 activity was measure in the presence of 1 mM sucrose, trehalose-6-
phosphate (T6P), glucose-6-phosphate (G6P) and maltose. The SnRK1a1
subunit activity in the absence of carbohydrate was measured and consid-
ered as 1. Values are the mean
SEM of three replicates. The letters indi-
cate statistically significant differences (P < 0.01).
affected the SnRK1a1/bc/b3 complex. From the kinetic
parameters (Table 2) maltose was a V-type activator of the
SnRK1a1/bc/b3 complex. From the Vmax versus maltose
replot, the apparent KA for maltose was ~0.6 mM, which is
in the physiological range of concentrations for this starch
degradation product (Weise et al., 2005).
Structural model for the heterotrimeric SnRKa1/bc/b3
complex and maltose binding
Molecular docking studies of maltose onto the modeled
three-dimensional SnRK1a1/bc/b3 complex were performed
in a similar fashion to the studies performed to predict the
AMP-binding modes on the same protein in a previous
report by Maya-Bernal et al. (2017) (Figure 7a, b). Of the
eight putative maltose-SnRK1a1/bc/b3 complexes identi-
fied, only poses b and c (Table S1) were predicted to have
0
negative binding energies with theoretical ΔG0 values low
enough to be considered potentially meaningful. An over-
view of the three-dimensional models of these two puta-
tive complexes is shown in Figure 7. Both docking poses
were located at an interface between the SnRK1bc CBM
domain and the SnRK1a1 N-terminal domain. The pose-b-
binding site was formed by greater contribution of resi-
dues from the SnRK1bc CBM domain, and the ligand
formed contacts with the SnRK1a1 N-terminal domain at
four residues of a loop (Figure 7c). In contrast, the pose-c-
binding site was dominated by residues from the SnRK1a1
N-terminal domain and four residues of a loop from the
SnRK1bc CBM domain (Figure 7d). Surprisingly, the abso-
0
lute value for the theoretical ΔG0 of pose c was twice as
© 2018 The Authors

Figure 6. SnRK1 kinetics in the presence of maltose. Activities of the het-
erotrimeric complexes SnRK1a1/bc/b1 (a), SnRK1a1/bc/b2 (b) and SnRK1a1/
bc/b3 (c, dashed line) were measured with increasing concentrations of sub-
strate (AMARA) in the presence of different concentrations of maltose to
determine the kinetic parameters. To determine the effect of maltose on the
catalytic subunit, the kinetic parameters were measured with recombinant
SnRK1a1 (c, continuous line). Values are the mean
SEM of three repli-
cates.
© 2018 The Authors
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Maltose and SnRK1 activity 169
large as that of pose b, suggesting that a more stable com-
plex is formed by this conformation and indicating the pos-
sible formation of stable SnRK1bc–maltose and SnRK1a1–
maltose complexes by the dissociated subunits. However,
if this last complex exists, the activity of the catalytic sub-
units is not modified by maltose (Figure 5) in the absence
of SnRK1bc and SnRK1b3 subunits.
DISCUSSION
SnRK1 is a protein kinase involved in different aspects of
plant metabolism and development. Most of the functions
of SnRK1 are performed via phosphorylation of enzymes
and transcription factors regulating metabolic processes
and the expression of multiple genes. The identification of
cytoplasmic and nuclear SnRK1 targets suggests an intra-
cellular location that has been experimentally corroborated
(Tsai and Gazzarrini, 2012; Williams et al., 2014). For the
regulatory subunits, different approaches have indicated
the presence of the b subunits in different organs, suggest-
ing a ubiquitous location of the proteins along the whole


plant. Fragoso et al. (2009) and Avila-Casta ~ eda et al.
n
(2014), reported that both catalytic and regulatory SnRK1b2
and SnRK1bc subunits are localized in the cytoplasm and
in the chloroplast. Even though no other work has shown
the SnRK1 subunits to be localized in chloroplasts, func-
tional studies have shown a reduction in the phosphoryla-
tion of chloroplastic proteins in snrk1a1/a2 mutants,
suggesting a direct effect of the kinase in the regulation of
the phosphorylation of chloroplastic proteins (Nukarinen
et al., 2016). To contribute additional data in support of the
chloroplastic localization of SnRK1 subunits, we performed
immunogold labeling and used electronic microscopy to
identify the SnRK1 subunits using specific antibodies. In
addition to identifying all the SnRK1 subunits inside the
chloroplast, SnRK1bc was also seen to be attached to the
starch granules, confirming the results of in vitro experi-
ments previously performed using recombinant proteins


(Avila-Casta ~ eda et al., 2014). Interestingly, SnRK1b3,
n
which does not contain a CBM, was also localized in the
starch. We propose that SnRK1bc forms a dimer with
SnRK1b3 and carries it to the chloroplast, as evidence of
dimer formation was established during the expression of
recombinant complexes (Maya-Bernal et al., 2017).
SnRK1b1 and b2 were not associated with the starch
despite being localized inside the chloroplast (Figure 1).
These data indicate that other factors determine the inter-
action of the CBM with polysaccharides, and it is possible
that formation of the heterotrimeric complex plays an
important role in this interaction. Both catalytic subunits
were also localized in the chloroplast and specifically inter-
acted with the starch, suggesting potential complex forma-
tion by either of the catalytic subunits, bc or b3. Additional
experiments should be performed to demonstrate that an
active kinase is present and that a specific complex is
170 Ana Ruiz-Gayosso et al.

Figure 7. Predicted maltose-binding sites in the SnRK1a1/b3/bc complex from Arabidopsis thaliana. (a) Superposition of the two docking poses predicted to
have low binding energies after a 5-ns MD simulation in explicit solvent with ions. (b) Close-up of the superposition of B with the A subunit as transparent car-
toons. In (a) and (b), the b pose complex has a reddish overtone, and the c pose complex has a cyan overtone. In (c) and (d), close-ups of the binding sites are
depicted, showing the pockets for the b pose (c) as translucent cyan surfaces and for the c pose (d) as red translucent surfaces; the neighboring-chain bc resi-
dues are labeled in gray boxes, and those from chain b3 are labeled in blue boxes. The subunits in (c) and (d) are shown as translucent gray (bc) or blue (b3) car-
toons; yellow text was used for contrast only. [Colour figure can be viewed at wileyonlinelibrary.com]

formed inside this organelle. When comparing our
immunolocalization results with published proteomic data
using the SUBA4 database, we found that only SnRK1b2
has been identified as a chloroplastic protein. Because,
however, only a fraction of the Arabidopsis chloroplast
proteome has been covered through proteomics directly
(Jensen and Leister, 2014), the absence of SnRK1 subunits
among the identified proteins does not demonstrate their
absence in the organelle, but rather additional experiments
need to be performed.
We also asked whether the CBM present in the regula-
tory subunits could bind other sugars. Interestingly, mal-
tose, a sugar produced during starch degradation,
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
exhibited binding to all the regulatory subunits (Figure 2).
Kd values obtained for the monomers were slightly higher
than the values for the dimers, and the values obtained for
the dimers were similar to those that were obtained when
SnRK1bc/b3 was used. A possible interpretation of these
results is that the CBM of SnRK1bc is the only one that is
functional when a CBM of another regulatory subunit is
also present. In a previous report by Emanuelle et al.,
2015, no evidence was found for the binding of maltose to
the b regulatory subunit with a fixed 10 mM maltose con-
centration using intrinsic tryptophan fluorescence spec-
troscopy. In contrast, in this work, using fluorescence and
thermal shift measurements, we obtained evidence for the
© 2018 The Authors

formation of maltose-b subunits alone or in complex with
bc. The difference may be due to the quenching of the fluo-
rescent signal at high maltose concentrations caused by
the binding of maltose to low-affinity sites, possibly expos-
ing aromatic residues to solvent or to other groups in the
protein. This binding could obscure the signal observed at
low maltose concentrations, which occurs due to the bind-
ing of maltose to high-affinity sites (likely to be physiologi-
cally relevant). In addition, the absence of binding of
trehalose 6P strongly supports the evidence that this car-
bohydrate binds to SnRK1 via an unidentified factor. This
intermediary is present only in young, growing leaves,
where inhibition of SnRK1 activity by trehalose 6P is
observed (Zhang et al., 2009). A study of the sequences
and structures of animal proteins containing GBDs in b
subunits indicated that the amino acids involved in sugar
binding are not conserved in plant subunits (Emanuelle
et al., 2016). However, the presence of a binding site sug-
gests the possibility that a different site is generated and
participates in maltose interaction. In fact, molecular dock-
ing studies proposed the presence of two maltose com-
plexes formed between SnRK1bc-CBM and the N-terminal
domain of SnRK1a1. This site is different from the site
identified for cyclodextrin binding (Polekhina et al., 2003).
Clearly, additional experiments are required to identify the
maltose-binding site and to determine whether this site is
the same one that binds starch.
To evaluate the effect that maltose had on kinase activ-
ity, we immunoprecipitated complexes formed with bc
subunits from leaf protein extracts collected at two differ-
ent times during the day. SnRK1 activity was significantly
higher at dawn than at dusk, indicating that kinase activity
is differentially regulated over the course of the day. This
behavior may reflect a physiological scenario in which,
after a period of darkness, the plant has consumed all of
its reserves, and SnRK1 is activated. In contrast, during the
day, carbohydrates and starch are synthesized, and the sig-
nal that indicates abundance is perceived, leading to a
decrease in SnRK1 activity. Immunoprecipitated complexes
using anti-bc antibodies indicated that the activities of only
the complexes obtained at dusk increased when maltose
was added. The differences in response could be due to
complex composition. Analysis of maltose regulation in
recombinant expressed complexes indicated that only the
complex formed with a1/bc/b3 was stimulated by maltose.
These results suggested that at dusk, most complexes are
formed by a1/bc/b3, whereas at dawn, mainly a1/bc/b1 or
b2 participated in complex formation. Alternatively, the
presence of a protein and/or a post-translational modifica-
tion could affect maltose sensitivity. Verification of these
possibilities requires further experimentation.
The function of the GBD in AMPK has been evaluated,
and it was postulated that the binding of AMPK to the a1–6
branched carbohydrates present in glycogen inhibit AMPK
© 2018 The Authors
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Maltose and SnRK1 activity 171
activity, so AMPK can sense not only energy availability in
the form of AMP and ATP but also long-term energy
resources in the form of glycogen, as proposed by McBride
et al. (2009). Furthermore, high levels of muscle glycogen
have been shown to repress AMPK activation. In heterotri-
meric complexes, deletions of b1 subunits lacking the GBD
exhibited activities that were reduced by approximately
30–40% compared with that activities exhibited by the full-
length proteins (Hudson et al., 2003). In yeast, the involve-
ment of glycogen in Snf1 activity is uncertain as mutants
that do not accumulate glycogen do not exhibit differential
activity. Moreover, deletion of GBD from Gal83, the most
abundant b subunit, results in constitutive kinase activity,
indicating that the GBD is important for sensing glucose
abundance, as described by Mangat et al. (2010). In SnRK1,
deletion of the CBM in the bc subunit exhibited a reduction
in kinase activity when a complex with SnRK1a1 and
SnRK1b3 was evaluated (Figure 4). However, when the
deleted bc was combined with complete b1 or b2 subunits,
no reduction in activity was observed, indicating that a
CBM from any of the b subunits could substitute the
deleted domain.
Evidence from different laboratories has suggested that
SnRK1 is involved in the synthesis and degradation of
starch. It has been proposed that SnRK1 increases the
carbon flux toward starch biosynthesis by upregulating
the expression of SUS (sucrose synthase) and AGPase
genes and downregulating the expression of SPS (Wang
et al., 2017). In addition, the data presented in this work
indicated that particular subunits of the heterotrimeric
complex were associated with the starch. Previous results
indicated that the association of SnRK1 with starch inhib-


ited SnRK1 activity (Avila-Casta ~ eda et al., 2014). Based
n
on those results, we hypothesize that during the day,
association of SnRK1 to the polysaccharide regulates the
activity of the kinase and inhibits the phosphorylation of
specific targets. In addition, all the SnRK1 subunits were
also present in the stroma of the chloroplast and in the
cytoplasm; therefore, we hypothesize that CBM-contain-
ing subunits play additional roles. The data obtained
from this work indicated that maltose could bind all the
b and bc subunits, but the kinase activity was upregu-
lated only when the recombinant complex SnRK1a1/bc/b3
was expressed. Based on these results, we proposed the
model depicted in Figure 8. In this model, we hypothe-
size that during the night, when maltose starts to accu-
mulate, the activity of SnRK1 increases, promoting
maltose metabolism via an as yet undefined mechanism.
In this model, the absence of the catalytic a subunits
(Baena-Gonza
lez et al., 2007; Fragoso et al., 2009) can
promote maltose accumulation, inhibiting further starch
degradation. This condition would explain the starch-
accumulated phenotype of the mutants of the catalytic
subunits; we are currently testing this model.
172 Ana Ruiz-Gayosso et al.

Figure 8. Proposed model to explain the effect of maltose on SnRK1 heterotrimeric complex. During the day, transitory starch is synthesized and stored in the
chloroplast as granules. At night, the leaf starch is almost completely degraded due a group of enzymes that drive the starch granule degradation to produce
maltose, which is transported to cytoplasm. The SnRK1 complex composed by SnRK1a1/bc/b3 can bind maltose (blue dashed arrow) produced during starch
degradation. This binding activates the kinase (red arrow), which can promote the carbon flux from starch to degradation products. Fru-6-P (fructose-6-phos-
phate), Glc-6-P (glucose-6-phosphate), Glc-1-P (glucose-1-phosphate), ADP-Glc (ADP-glucose), MOS (malto-oligosaccharides), Glc (glucose), SHG (soluble
heteroglycan), SHG-Glc (soluble heteroglycan with the additional glucose from maltose), ATP (adenosine triphosphate), PPi (pyrophosphate), PGI (phosphoglu-
cose isomerase), PGM (phosphoglucomutase), AGPase (ADP-glucose pyrophosphorylase), SS (starch synthase), BE (glucanotransferase 1,4-a-glucan branching
enzyme), DBE (a-1, 6-glucosidase starch debranching enzyme), GWD (a-glucan water dikinase), PWD (phosphoglucan water dikinase), SEX4 (starch excess 4),
BAM (b-amylase), AMY (a-amylase), ISA3 (isoamylase 3), DPE1 (disproportionating enzyme 1), MEX1 (maltose transporter 1), DPE2 (disproportionating enzyme
2), PHS2 (a-glucan phosphorylase). [Colour figure can be viewed at wileyonlinelibrary.com]

EXPERIMENTAL PROCEDURES
Plant material
Wild-type Arabidopsis thaliana ecotype Columbia (Col-0) plants
were grown. For protoplast isolation, 3-week-old plants were
grown in a growth chamber at 22°C under short-day conditions
(8 h light/16 h dark). For immunogold labeling, young leaves from
3-week-old plants were fixed for ultrathin sectioning. For IP and
protein purification, Arabidopsis plants were grown under short-
day conditions. Rosette tissue was harvested and frozen in liquid
nitrogen at dawn (6 am) and dusk (6 pm) 3 weeks after germina-
tion.
with primary antibodies were washed twice with PBS for 5 min
and three times with freshly prepared blocking solution. For nega-
tive controls, the grids were incubated with pre-immune serum, or
the primary antibodies were omitted. For grids with blocked anti-
bodies, incubation with the specific peptide was carried out over-
night at 4°C. The sections were incubated with anti-rabbit
antibody (1:20 dilution) conjugated to 15-nm gold particles (Ted
Pella Inc., USA) in blocking solution for 1 h at room temperature
and were rinsed three times with deionized water. Finally, the
grids were examined by transmission electron microscopy using a
JEOL JEM-1200EXII, USA.
Maltose-binding assays

Determination of the subcellular localization of SnRK1b
and SnRK1bc subunits by immunogold labeling
Leaves from 3-week-old plants were fixed and treated for ultrathin
sectioning. Antibody labeling was performed as described by Wil-
son and Bacic (2012). Ultrathin leaf sections were blocked with
freshly prepared blocking solution containing 1% w/v BSA in 19
PBS (137 mM of NaCl, 10 mM of Na2HPO4, 2.7 mM of KCl, 0.18 mM
KH2PO4) pH 7.4 for 30 min. The grids were incubated with the pri-
mary rabbit-specific antibodies at 4°C overnight. The grids treated
Maltose binding was measured by two different techniques. For
the intrinsic fluorescence assays, the proteins were diluted to a
final concentration of 3 lM in a buffer containing 50 mM of NaCl
and 50 mM of Tris–HCl (pH 7.5). A freshly prepared maltose stock
was used to add increasing concentrations of ligand. The protein
solutions were excited at 280 nm, and the fluorescence spectra
were collected from 300 to 400 nm. As a positive control, MBP was
used. Dissociation constant (Kd) determination was performed by
considering the difference in the spectra after maltose treatment
and the initial spectrum, obtained in the absence of ligand.

© 2018 The Authors

The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175

The DSF was performed as described by Vivoli et al. (2014). The
proteins were diluted to a final concentration of 3 lM in a buffer
containing 100 mM NaCl, 50 mM HEPES–KOH (pH 7.0) and 19 Pro-
tein Thermal Shift dye (Applied Biosystems, USA). The thermal
denaturation protocol was performed using a StepOne real-time
PCR instrument (Applied Biosystems, USA). The data were
adjusted to the Hill equation.
Immunoprecipitation assays
Arabidopsis rosettes were harvested at dawn and dusk and frozen
in liquid nitrogen. The leaves were ground, and total soluble pro-
tein was extracted in homogenization buffer containing 100 mM of
Tricine–NaOH (pH 8.0), 0.5 mM of EGTA, 0.5 mM of EDTA, and
1 mM of benzamidine. Prior to homogenization, 1 mM of PMSF, 19
protease inhibitor cocktail (Sigma, Mexico), 19 phosphatase inhi-
bitors (50 mM of NaF, 25 mM of b-glycerolphosphate, 10 mM of
sodium pyrophosphate, and 2 mM of sodium orthovanadate) and
insoluble polyvinylpyrrolidone (2% w/v) were added. The homoge-
nate was centrifuged for 30 min at 13 0009 g, and insoluble mate-
rial was removed.
Total soluble protein (crude extract, CE) was clarified with 2.5%
w/v protein A-agarose (Sigma, USA) for 15 min at 4°C. The mix-
ture was centrifuged at 2000 g for 2 min, and the pellet was dis-
carded. Primary rabbit-specific antibodies were added to the
supernatant, and the mixture was incubated for 1 h at 4°C. Protein
A-agarose was added (2.5% w/v), and incubation was continued
for an additional 1 h. The mixture was centrifuged, and the pellet
was rinsed three times with wash buffer containing 50 mM Tris–
HCl (pH 7.5), 100 mM NaCl, 10% v/v glycerol and 0.05% v/v Triton
X-100. The pellet was resuspended and used in activity assays.
SnRK1 activity assays
The SnRK1 activity assays were performed in a final volume of
25 lL. The reaction mixture was prepared with 40 mM of HEPES
(pH 7.5), 5 mM of MgCl2, 200 lM of ATP containing 12.5 kBq (c-32P)
ATP (Perkin Elmer, USA), 4 mM DTT, 1X phosphatase inhibitor, 19
protease inhibitor cocktail (Sigma, Mexico), and 200 lM AMARA
peptide (AMARAASAAALARRR). According to the assay, AMARA
peptide concentration was modified, or maltose was added. The
assay was started by addition of the sample containing protein
kinases. After 6 min, 15 lL of the reaction mixture was transferred
to a square (2 9 2 cm) piece of phosphocellulose paper (Whatman
P81), which was immersed in 1% v/v phosphoric acid. The papers
were rinsed twice with phosphoric acid followed by acetone. The
membranes were air-dried, and the incorporation of 32P was quan-
tified by the addition of scintillation liquid (ACSII aqueous count-
ing scintillant, Amersham) and measurement in a Beckman
scintillation counter.
Purification of recombinant proteins and heterotrimer
assembly
The cDNAs corresponding to SnRK1a1 (At3g01090), SnRK1b1,
SnRK1b2, SnRK1b3, SnRK1bc, and SnAK2 (At3g45240) were cloned
into expression vectors as described by Maya-Bernal et al. (2017).
The resulting expression vectors were transformed into the E. coli
BL21 (DE3) strain to obtain the SnRK1b monomers, SnRK1bc/b
dimers or active SnRK1a1. Expression of the recombinant proteins
was induced, and the proteins were purified with a Ni-TED column
as according to the manufacturer’s instructions (Macherey-Nagel,
Germany). The purified proteins were used in the intrinsic fluores-
cence assays or for DSF. Assembly of the heterotrimers was per-
formed as described by Maya-Bernal et al. (2017) with 1.5 parts of
© 2018 The Authors
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Maltose and SnRK1 activity 173
SnRK1bc/b1 and SnRK1bc/b2, while the amount of SnRK1bc/b3
used was two times higher than that of SnR1a1.
Modeling of the SnRK1 complex SnRK1a1/bc-b3 and
docking studies
The three-dimensional structure of the SnRK1a1/b3/bc complex
was modeled based on the human AMPK complex (PDBID 4rer
and 4cfe) using homology modeling, molecular dynamics (MD)
simulations and PM7-LMO QM calculations as previously
described by Maya-Bernal et al. (2017). The search for maltose-
binding sites on the modeled plant SnRK1a1/b3/bc complex was
performed using repeated trials in AutoDock Vina as described by
Trott and Olson (2010). The docking events were clustered by
RMSD, and the eight clusters with the largest number of members
and lowest average energies were selected. These putative com-
plexes were subjected to 5-ns MD simulations in an explicit TIP3P
solvent periodic box with ions (0.15 M NaCl, neutral) using GRO-
MACS 4.6.7 MD software as described by Hess et al.(2008) with
the AMBER 99SB-ILDN forcefield as described by Lindorff-Larsen
et al. (2010), particle mesh Ewald electrostatics, LINCS bond
restriction, Berendsen V-rescaled thermostat, and Berendsen baro-
stat. The binding energy was estimated using the linear interac-
tion energy (LIE) method described by Chong Teoh et al. (2011)
using 50-ps intervals near the ends of the simulations. The ener-
gies calculated for the eight docking poses are shown in Table S1.
ACKNOWLEDGEMENTS


We thank Alejandra Avila for technical support. This work was
supported by PAPIIT IN216216, LANCAD-UNAM-DGTIC-215-2017
Supercomputing project. PAIP-FQ-UNAM 5000-9122 and 5000-
9126. CONACyT INFRA-2015-252001 and 252123. ARG was a recip-
ient of a postgraduate scholarship from CONACyT.
CONFLICTS OF INTEREST
The authors declare no conflicts of interest.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online ver-
sion of this article.
Figure S1. Structures of the three SnRK1 subunits.
Figure S2. Pre-immune controls for transmission electron micro-
scopy (TEM).
Figure S3. Blocked antibodies for SnRK1bc and SnRK1b3 immuno-
gold detection.
Figure S4. Subcellular localization of SnRK1a1 and SnRK1a2 in
Arabidopsis leaves.
Figure S5. Intrinsic fluorescence spectra of the SnRK1 regulatory
subunits.
Figure S6. Differential scanning fluorimetry (DSF) with other car-
bohydrates.
Figure S7. Proteins immunoprecipitated using the antibody
against SnRK1bc.
Table S1. Linear interaction energies for poses of maltose docked
into SnRK1a1/bc/b3 complex, estimated from molecular dynamics
simulations.
REFERENCES


Avila-Castan~ eda, A., Gutie
rrez-Granados, N., Ruiz-Gayosso, A., Sosa-Pei-
nado, A., Mart
ınez-Barajas, E. and Coello, P. (2014) Structural and
174 Ana Ruiz-Gayosso et al.
functional basis for starch binding in the SnRK1 subunits AKINb2 and
AKINbc. Front. Plant Sci. 5(5), 199.
Baena-Gonza
lez, E., Rolland, F., Thevelein, J.M. and Sheen, J. (2007) A cen-
tral integrator of transcription networks in plant stress and energy sig-
nalling. Nature, 448(7156), 938–942.
Chia, T., Thorneycroft, D., Chapple, A., Messerli, G., Chen, J., Zeeman, S.C.,
Smith, S.M. and Smith, A.M. (2004) A cytosolic glucosyltransferase is
required for conversion of starch to sucrose in Arabidopsis leaves at
night. Plant J. 37, 853–863.
Chong Teoh, T., Heidelberg, T. and Rizman-Idid, M. (2011) Systematic pro-
tein-protein docking and molecular dynamics studies of HIV-1 gp120 and
CD4: insights for new drug development. Daru, 19, 469–475 PMID:
23008694.
Christiansen, C., Hachem, M.A., Janecek, S., Vikso-Nielsen, A., Blennow, A.
and Svensson, B. (2009) The carbohydrate-binding module family 20–di-
versity, structure, and function. FEBS J. 276(18), 5006–5029.
Coello, P., Hirano, E., Hey, S.J., Muutcumaru, N., Mart
ınez-Barajas, E.,
Parry, M.A. and Halford, N.G. (2011) Evidence that abscisic acid pro-
motes degradation of SNF1-related protein kinase (SnRK)1 1in wheat
and activation of a putative calcium-dependent SnRK1. J. Exp. Bot. 63,
913–924.
Crichley, J.H., Zeeman, S., Takaha, T., Smith, A.M. and Smith, S. (2001) A
critical role for disproportionating enzyme in starch breakdown is
revealed by knock-out mutation in Arabidopsis. Plant J. 26(1), 89–100.
Emanuelle, S., Hossain, M.I., Moller, I.E. et al. (2015) SnRK1 from Arabidop-
sis thaliana is an atypical AMPK. Plant J. 82, 183–192.
Emanuelle, S., Doblin, M.S., Stapleton, D.I., Bacic, A. and Gooley, P.R.
(2016) Molecular insights into the enigmatic metabolic regulator, SnRK1.
Trends Plant Sci. 21, 341–353.
Fettke, J., Eckermann, N., Poeste, S., Pauly, M. and Steup, M. (2004) A gly-
can substrate of the cytosolic (Pho2) phosphorylase isoenzyme from
Pisum sativum L.: identification, linkage analysis and subcellular localiza-
tion. Plant J. 39(6), 933–946.
Fettke, J., Eckermann, N., Tiessen, A., Geigenberg, P. and Steup, M. (2005)
Identification, subcellular licalization and biochemical characterization of
water-soluble heteroglycan (SHG) in leaves from Arabidopsis thaliana L.:
distinct SHG reside in the cytosol and in the apoplast. Plant J. 43(4), 568–
585.
Fragoso, S., Esp
ındola, L., Pa
ez-Valencia, J., Gamboa, A., Camacho, Y.,
Mart
ınez-Barajas, E. and Coello, P. (2009) SnRK1 isoforms AKIN10 and
AKIN11 are differentially regulated in Arabidopsis plants under phos-
phate starvation. Plant Physiol. 149, 1906–1916.
Gissot, L., Polge, C., Jossier, M., Girin, T., Bouly, J.-P., Kreis, M. and
Thomas, M. (2006) AKINbetagamma contributes to SnRK1 heterotri-
meric complexes and interacts with two proteins implicated in plant
pathogen resustance through its KIS/GBD sequence. Plant Physiol. 142
(3), 931–944.
Gue
rinier, T., Millan, L., Crozet, P., Oury, C., Rey, F., Valot, B. and Glab, N.
(2013) Phosphorylation of p27(KIP1) homologs KRP6 and 7 by SNF1-
related protein kinase-1 links plant energy homeostasis and cell prolifera-
tion. Plant J. Cell Mol Biol. 75(3), 515–525.
Hejazi, M., Fettke, J., Paris, O. and Steup, M. (2009) The two plastidial
starch-related dikinases sequentially phosphorylate glucosyl residues
at the surface of both the A- and B-type allomorphs of crystallized
maltodextrins but the mode of action differs. Plant Physiol. 150(2),
962–976.
Hess, B., Kutzner, C., Van Der Spoel, D. and Lindahl, E. (2008) GROMACS 4:
algorithms for highly efficient, load-balanced, and scalable molecular
simulation. J. Chem. Theor. Comput. 4, 435–447. https://doi.org//10.1021/
ct700301q.
Hudson, E.R., Pan, D.A., James, J., Lucocq, J.M., Hawley, S.A., Green,
K.A., Baba, O., Terashima, T. and Grahame, H.D. (2003) A novel
domain in AMP-activated protein kinase causes glycogen storage bod-
ies similar to those seen in hereditary cardiac arrhythmias. Curr. Biol.
13(10), 861–866.
Janecek, S., Svensson, B. and MacGregor, E.A. (2011) Structural and evolu-
tionary aspects of two families of non-catalytic domains present in starch
and glycogen binding proteins from microbes, plants and animals.
Enzyme Microb. Technol. 49(5), 429–440.
Jensen, P.E. and Leister, D. (2014) Chloroplast evolution, structure and func-
tions. F1000Prime Rep. 6, 40.
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Koay, A., Woodcroft, B., Petrie, E.J. et al. (2010) AMPK beta subunits dis-
play isoform specific affinities for carbohydrates. FEBS Lett. 584, 3499–
3503.
Lindorff-Larsen, K., Piana, S., Palmo, K., Maragakis, P., Jl, Klepeis, Ro, Dror
and De, Shaw (2010) Improved side-chain torsion potentials for the
Amber ff99SB protein force field. Proteins, 78, 1950–1958. https://doi.
org//10.1002/prot.22711 PMID: 20408171.
Lovas, A., So
s-Hegedus, A., Bimbo
, A. and Ba
nfalvi, Z. (2003) Functional
diversity of potato SNF1-related kinases tested in Saccharomyces cere-
visiae. Gene, 321, 123–129.
Lu, Y. and Sharkey, T.D. (2004) The role of amylomaltase in maltose
metabolism in the cytosol of photosynthetic cells. Planta, 21(3), 466–
473.
Lumbreras, V., Alba, M.M., Kleinow, T., Koncz, C. and Pages, M. (2001)
Domain fusion between SNF1-related kinase subunits during plant evolu-
tion. EMBO Rep. 2(1), 55–60.
Mangat, S., Chandrashekarappa, D., McCartney, R.R., Elbing, K. and Sch-
midt, M.C. (2010) Differential roles of the glycogen-binding domains of
beta subunits in regulation of the Snf1 kinase complex. Eukaryot. Cell, 9,
173–183.


Maya-Bernal, J.L., Avila, A., Ruiz-Gayosso, A., Trejo-Fregoso, R., Pulido,
N., Sosa-Peinado, A., Z
un ~ iga-Sa
nchez, E., Mart
ınez-Barajas, E.,
Rodr
ıguez-Sotres, R. and Coello, P. (2017) Expression of recombinant
SnRK1 in E. coli. Characterization of adenine nucleotide binding to
the SnRK1.1/AKINbc-b3 complex. Plant Sci. 263(Supplement C), 116–
125.
McBride, A., Ghilagaber, S., Nikolaev, A. and Hardie, D.G. (2009) The glyco-
gen-binding domain on the AMPK beta subunit allows the kinase to act
as a glycogen sensor. Cell Metab. 9(1), 23–34.
Nashilevitz, S., Melamed-Bessudo, C., Aharoni, A., Kossmann, J., Wolf, S.
and Levy, A.A. (2009) The legwd mutant uncovers the role of starch
phosphorylation in pollen development and germination in tomato. Plant
J. 57(1), 1–13.
Niittyla€, T., Messerli, G., Trevisan, M., Chen, J., Smith, A.M. and Zeeman,
S.C. (2004) A previously unknown maltose transporter essential for
starch degradation in leaves. Science, 303, 87–89.
Nukarinen, E., Na€gele, T., Pedrotti, L. et al. (2016) Quantitative phospho-
proteomics reveals the role of the AMPK plant ortholog SnRK1 as a
metabolic master regulator under energy deprivation. Sci. Rep. 6,
31697.
Nunes, C., Primavesi, L.F., Patel, M.K., Mart
ınez-Barajas, E., Powers, S.J.,
Sagar, R. and Paul, M.J. (2013) Inhibition of SnRK1 by metabolites: tis-
sue-dependent effects and cooperative inhibition by glucose 1-phos-
phate in combination with trehalose 6-phosphate. Plant Physiol.
Biochem. 63, 89–98.
Polekhina, G., Gupta, A., Michell, B.J., van Denderen, B., Murthy, S., Feil,
S.C. and Stapleton, D. (2003) AMPK beta subunit targets metabolic stress
sensing to glycogen. Curr. Biol. 13(10), 867–871.
Polge, C. and Thomas, M. (2007) SNF1/AMPK/SnRK1 kinases, global
regulators at the heart of energy control? Trends Plant Sci. 12(1), 20–
28.
Radchuk, R., Emery, R.J.N., Weier, D., Vigeolas, H., Geigenberger, P., Lunn,
J.E. and Weber, H. (2010) Sucrose non-fermenting kinase 1 (SnRK1) coor-
dinates metabolic and hormonal signals during pea cotyledon growth
and differentiation. Plant J. 61(2), 324–338.
Streb, S. and Zeeman, S.C. (2012) Starch metabolism in Arabidopsis. Ara-
bidopsis Book, 10, e0160. American Society of Plant Biologists.
Trott, O. and Olson, A.J. (2010) AutoDock Vina: improving the speed and
accuracy of docking with a new scoring function, efficient optimization,
and multithreading. J. Comput. Chem. 31, 455–461. https://doi.org//10.
1002/jcc.21334 PMID: 19499576.
Tsai, A.Y. and Gazzarrini, S. (2012) AKIN10 and FUSCA3 interact to control
lateral organ development and phase transitions in Arabidopsis. Plant J.
69, 809–821.
Vivoli, M., Novak, H.R., Littlechild, J.A. and Harmer, N.J. (2014) Determina-
tion of protein-ligand interactions using differential scanning fluorimetry.
J. Vis. Exp. 91, e51809.
Wang, F., Ye, Y., Chen, X., Wang, J., Chen, Z. and Zhou, Q. (2017) A sucrose
non-fermenting-1-related protein kinase 1 gene from potato, StSnRK1,
regulates carbohydrate metabolism in transgenic tobacco. Physiol. Mol.
Biol. Plants, 23, 933–943.
© 2018 The Authors

Weise, S., Weber, A.P.M. and Sharkey, T. (2004) Maltose is the major form
of carbon exported from the chloroplast at night. Planta, 218(3), 474–482.
Weise, S., Kim, K., Stewart, R. and Sharkey, T. (2005) b-Maltose is the
metabolically active anomer of maltose during transitory starch degrada-
tion. Plant Physiol. 137, 756–761.
Williams, S.P., Rangarajan, P., Donahue, J.L., Hess, J.E. and Gillaspy, G.E.
(2014) Regulation of sucrose non-fermenting related kinase 1 genes in
Arabidopsis thaliana. Front. Plant Sci. 5, 324.
Wilson, S.M. and Bacic, A. (2012) Preparation of plant cells for transmission
electron microscopy to optimize immunogold labeling of carbohydrate
and protein epitopes. Nat. Protoc. 7, 1716–1727.
Wingler, A., Masclaux-Daubresse, C. and Fischer, A.M. (2009) Sugars, senes-
cence, and ageing in plants and heterotrophic organisms. J. Exp. Bot. 60
(4), 1063–1066.
© 2018 The Authors
The Plant Journal © 2018 John Wiley & Sons Ltd, The Plant Journal, (2018), 96, 163–175
Maltose and SnRK1 activity 175
Yu, T.S., Kofler, H., Ha€usler, R.E. et al. (2001) The Arabidopsis sex1 mutants
is defective in the R1 protein, a general regulator of starch degradation
in plants, and not in the chloroplast hexose transporter. Plant Cell, 13(8),
1907–1918.
Zhang, Y., Shewry, P.R., Jones, H., Barcelo, P., Lazzeri, P.A. and Halford,
N.G. (2001) Expression of antisense SnRK1 protein kinase sequence
causes abnormal pollen development and male sterility in transgenic
barley. Plant J. 28(4), 431–441.
Zhang, Y., Primavesi, L.F., Jhurreea, D., Andralojc, P.J., Mitchell, R.A.C.,
Powers, S.J., Schluepmann, H., Delatte, T., Wingler, A. and Paul, M.J.
(2009) Inhibition of Snf1-related protein kinase (SnRK1) activity and regu-
lation of metabolic pathways by trehalose 6-phosphate. Plant Physiol.
149, 1860–1871.