This suggests that low-affinity starch-binding domains are a recurring feature in plastidial starch metabo-lism, and supports the hypothesis that reversible binding, effectuated through l
Trang 1domains from glucan, water dikinase and a-amylase
involved in plastidial starch metabolism
Mikkel A Glaring1,2, Martin J Baumann1, Maher Abou Hachem1, Hiroyuki Nakai1, Natsuko Nakai1, Diana Santelia3, Bent W Sigurskjold4, Samuel C Zeeman3, Andreas Blennow2and Birte Svensson1
1 Enzyme and Protein Chemistry, Department of Systems Biology, Technical University of Denmark, Kongens Lyngby, Denmark
2 VKR Research Centre Pro-Active Plants, Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of
Copenhagen, Frederiksberg, Denmark
3 Department of Biology, ETH Zu¨rich, Switzerland
4 Department of Biology, University of Copenhagen, Denmark
Introduction
Starch is deposited as water-insoluble granules in the
amyloplasts of tubers, roots and seeds for long-term
storage, and in the chloroplasts of green tissues during
the day as short-term storage for the following night
(transitory starch) The granules are composed of two different polymers of glucose linked via a-(1,4)-glyco-sidic bonds: amylose, which is essentially linear, and amylopectin, which also contains a-(1,6)-glycosidic
Keywords
carbohydrate-binding module; starch
metabolism; starch-binding domain;
a-amylase; a-glucan, water dikinase
Correspondence
B Svensson, Enzyme and Protein
Chemistry, Department of Systems Biology,
Technical University of Denmark, Søltofts
Plads, Building 224, DK-2800 Kongens
Lyngby, Denmark
Fax: +45 45886307
Tel: +45 45252740
E-mail: bis@bio.dtu.dk
(Received 17 November 2010, revised 5
January 2011, accepted 31 January 2011)
doi:10.1111/j.1742-4658.2011.08043.x
Starch-binding domains are noncatalytic carbohydrate-binding modules that mediate binding to granular starch The starch-binding domains from the carbohydrate-binding module family 45 (CBM45, http://www.cazy.org) are found as N-terminal tandem repeats in a small number of enzymes, primarily from photosynthesizing organisms Isolated domains from repre-sentatives of each of the two classes of enzyme carrying CBM45-type domains, the Solanum tuberosum a-glucan, water dikinase and the Arabid-opsis thaliana plastidial a-amylase 3, were expressed as recombinant pro-teins and characterized Differential scanning calorimetry was used to verify the conformational integrity of an isolated CBM45 domain, reveal-ing a surprisreveal-ingly high thermal stability (Tmof 84.8C) The functionality
of CBM45 was demonstrated in planta by yellow⁄ green fluorescent protein fusions and transient expression in tobacco leaves Affinities for starch and soluble cyclodextrin starch mimics were measured by adsorption assays, surface plasmon resonance and isothermal titration calorimetry analyses The data indicate that CBM45 binds with an affinity of about two orders
of magnitude lower than the classical starch-binding domains from extra-cellular microbial amylolytic enzymes This suggests that low-affinity starch-binding domains are a recurring feature in plastidial starch metabo-lism, and supports the hypothesis that reversible binding, effectuated through low-affinity interaction with starch granules, facilitates dynamic regulation of enzyme activities and, hence, of starch metabolism
Abbreviations
AMY3, a-amylase 3; AtAMY3, Arabidopsis thaliana a-amylase 3; CBM, carbohydrate-binding module; DSC, differential scanning calorimetry; GWD, a-glucan, water dikinase; IPTG, isopropyl thio-b- D -galactoside; ITC, isothermal titration calorimetry; PWD, phosphoglucan, water dikinase; SBD, starch-binding domain; SPR, surface plasmon resonance; StGWD, Solanum tuberosum a-glucan, water dikinase;
TEV, tobacco etch virus; YFP⁄ GFP, yellow ⁄ green fluorescent protein.
Trang 2branch points Together, these polymers are laid down
as alternating semicrystalline and amorphous layers to
form the supramolecular granule structure The
semi-crystalline layers are made up primarily of packed
double helices formed by adjacent glucose chains in
amylopectin The amorphous layers are comprised
mainly of amylopectin branch points and interspersed
amylose [1] This semicrystalline structure offers a
sig-nificant challenge for degrading enzymes and, for the
efficient amylolysis of raw starch, requires either
sur-face substrate-binding sites on the catalytic module [2]
or, more commonly, a specialized
carbohydrate-bind-ing module (CBM) known as a starch-bindcarbohydrate-bind-ing domain
(SBD) [3]
CBMs are noncatalytic structural domains that
mediate the binding of proteins to polysaccharides,
thus bringing the appended catalytic modules into
close contact with the substrate, enabling efficient
hydrolysis of insoluble polysaccharides, such as starch
and cellulose [4] Based on their amino acid sequences,
CBMs are currently grouped into 61 families SBDs
are found in CBM families 20, 21, 25, 26, 34, 41, 45,
48, 53 and 58 (http://www.cazy.org) [5,6] The majority
of characterized SBDs are found in extracellular
microbial amylolytic enzymes, where they enhance
binding to starch and related a-glucans A significant
and so far largely uncharacterized number of SBDs
occur in nonamylolytic enzymes from all domains of
life [3,6] Among these is a recently established small
family of SBDs, named CBM45 [7] CBM45s are
found primarily in photosynthesizing organisms in
only two classes of intracellular enzyme: the a-glucan,
water dikinases (GWDs, EC 2.7.9.4), which
phosphor-ylate starch, and the plastidial a-amylases (AMYs,
EC 3.2.1.1) Two types of CBM45-containing GWD
have been identified in plants, one of which is
plastid-ial and essential for normal starch metabolism
(GWD1⁄ R1 ⁄ SEX1) [8,9] The second GWD, called
GWD2 in Arabidopsis, is extraplastidial and has no
apparent role in starch degradation [10] The plastidial
a-amylase AMY3 is not required for normal transitory
starch metabolism in Arabidopsis [11], but a functional
role in planta has been inferred from knock-out studies
in phosphoglucan phosphatase (SEX4) and quadruple
debranching enzyme (ISA1⁄ ISA2 ⁄ ISA3 ⁄ LDA) mutant
backgrounds [12,13] In both enzyme classes, the
CBM45s are present as N-terminal tandem domains,
separated by a linker domain of varying length No
three-dimensional structure is available for CBM45,
but a recent bioinformatic analysis produced a rough
model and identified two tryptophan residues as
puta-tive binding sites in the N-terminal CBM45 from
Ara-bidopsis GWD1 [14] These two tryptophans have
previously been experimentally confirmed as pivotal for the starch-binding ability of potato GWD [7] The metabolism of plastidial starch in leaves of plants is a tightly regulated process The available pho-tosynthate has to be balanced with the energy and car-bohydrate needs of the plant during the subsequent dark period Perturbations in this process lead to severe phenotypes and retardation of growth [15,16] Plastidial starch metabolism has been well character-ized in the model plant Arabidopsis thaliana, and numerous enzymes are involved in the process of building and degrading the insoluble granule (for recent reviews, see refs [17–19]) Many of these enzymes contain SBDs representing several different CBM families, and starch and⁄ or a-glucan binding has been demonstrated in a number of cases [20–26] Starch binding in the plastidial system is influenced by potential regulatory mechanisms that can reversibly affect the interaction with the granule [24,27] Binding
of potato GWD to starch in planta has been shown to
be diurnally regulated [23] and potentially influenced
by the redox status of the enzyme [23,27] The redox status also influences the binding of the CBM48-con-taining glucan phosphatase SEX4 to starch in vitro [24] Phosphoglucan, water dikinase (PWD⁄ GWD3,
EC 2.7.9.5), a second type of plastidial GWD carrying
an N-terminal CBM20 SBD, shows a relatively low affinity towards the starch mimics, cyclodextrins [20], and it has been proposed that this could facilitate more dynamic interactions with starch, allowing the modulation of affinity necessary for metabolic regula-tion in the plastid [6,20,28]
In this article, we report the characterization of SBDs from representatives of the two classes of enzyme carrying CBM45-type domains, the Sola-num tuberosum a-glucan, water dikinase (StGWD) and the A thaliana plastidial a-amylase 3 (AtAMY3), sug-gesting that the evolution of low-affinity domains is a recurring and functionally important theme in plastid-ial starch metabolism
Results and Discussion
Identification and bioinformatic analysis of CBM45s from GWD and AMY3
Family CBM45 sequences were obtained from the car-bohydrate-active enzymes database (CAZY, http:// www.cazy.org) Furthermore, a search of the translated nucleotide database at the National Center for Bio-technology Information (http://www.ncbi.nlm.nih.gov) uncovered several additional sequences with homology
to StGWD and AtAMY3, which were included in the
Trang 3analysis Alignment of these sequences showed not
only extensive conservation in the catalytic domains,
suggesting a preservation of function, but also the
exis-tence of the N-terminally appended CBM45 domains
(data not shown) The sequences used in the
subse-quent analysis included CBM45s from 45 proteins
from primarily photosynthetic eukaryotes (plants,
green algae and red algae), as well as from
apicom-plexan parasites (Doc S1) A common characteristic
of these organisms is the presence of starch or
starch-like crystalline polysaccharides The starch synthesis
ability of the apicomplexan parasites is believed to be
derived from an endosymbiosis of a red alga, with
following loss of photosynthetic capacity [29]
The CBM45s are present as tandem domains in the
N-terminal part of StGWD and AtAMY3 and
sepa-rated by a linker of approximately 200 and 50 amino
acids, respectively (Fig 1A) The alignment of all
iden-tified CBM45s revealed that each contains five
aro-matic amino acids that are widely conserved across all
species (Fig S1) These residues are also present in
StGWD and AtAMY3 (Fig 1B) The ability to bind
to starch has been associated with certain consensus
aromatic residues in other CBM families [6] and,
although there is evidence that two of the aromatic
residues (W139 and W194) are necessary for the
starch-binding ability of StGWD [7], the lack of
struc-tural information on CBM45 precludes the assignment
of residues to specific binding sites as has been
per-formed for CBM20 from Arabidopsis PWD⁄ GWD3
(AtPWD) [20]
It was clear from the collected sequences that the
tandem organization of two domains is a common
characteristic of most CBM45-containing enzymes,
suggesting that this is essential for the functionality of
the appended catalytic modules Isoforms of GWD
and AMY3 from the green algae Chlamydomonas reinhardtii (CreGWD) and Micromonas (MiGWDb, MpGWDb and MpAMY3; Doc S1) contained only one identifiable CBM45 domain Whether this repre-sents a simple misannotation or a distinct function of single-CBM45 SBDs is unknown A previous analysis
of a recombinant truncated StGWD lacking CBM45-1 showed an altered specificity on soluble substrates with
a preference for the phosphorylation of shorter glucan chains [7] A phylogenetic tree based on the complete amino acid alignment of all CBM45s showed obvious groupings of the individual domains from plant sequences (Fig S2) Most of the nonplant sequences formed a separate, mixed group, reflecting the evolu-tionary distance and low homology between these sequences Overall, it appears as though CBM45s and the tandem structure of these domains arose early in evolution, perhaps in an ancestor of the current photo-synthetic eukaryotes It has been proposed that GWD sequences were a prerequisite for the appearance of semicrystalline starch-like polymers [29] If this is indeed the case, the appended CBM45 domains could represent a truly ancient SBD and, perhaps, be one of the first CBMs dedicated to starch binding
Expression and purification of CBM45s from potato GWD and Arabidopsis AMY3
In order to characterize the CBM45s from StGWD, several expression constructs were produced and tested Because there is no structural information on CBM45, putative domain borders were assigned on the basis of the predicted secondary structure and homol-ogy to other CBMs Thirteen constructs with an N-ter-minal tobacco etch virus (TEV) protease-cleavable Histidine (His)-tag, containing both the single and
A
B
Fig 1 Overview of the CBM45 domains.
(A) Domain structure of StGWD and
AtAMY3 showing the chloroplast transit
peptide (TP, black), tandem CBM45s (grey)
and catalytic domain (light grey) The size
of the proteins is given in amino acids (aa).
(B) Sequence alignment of CBM45 domains
1 and 2 from StGWD and AtAMY3 created
using C LUSTAL W2 (http://www.ebi.ac.uk/
tools/clustalw2) Identical (*), conserved (:)
and semiconserved (.) residues are indicated
below the alignment The arrows indicate
the five conserved aromatic amino acid
residues.
Trang 4double modules with or without the intervening
sequence, were produced and tested for soluble
expres-sion (Fig S3) Two of each of the recombinant single
modules, as well as the double module (CBM45-1B,
amino acids 77–217; CBM45-1E, amino acids 109–217;
CBM45-2A, amino acids 405–545; CBM45-2B, amino
acids 405–551; CBM45-1,2, amino acids 77–551;
Fig S3), were further purified Although the
CBM45-1s and double module could be expressed with
reason-able yields, they precipitated rapidly after His-tag
puri-fication and were not suitable for detailed analyses
The CBM45-2s were stable at pH 8.0 and could be
subjected to TEV protease cleavage and dialysis
with-out significant loss of protein The more stable of the
two resulting proteins (CBM45-2A) was chosen for
detailed characterization This protein precipitated
slowly when stored at 4C The general problem of
aggregation observed with the recombinant CBM45s
suggests that the isolated SBD is destabilized as a
result of the exposure of hydrophobic surface, which
would otherwise be packed on other domains in the
native full-length GWD
Initial differential scanning calorimetry (DSC)
screening of the His-tagged StGWD CBM45-2A
indi-cated a high unfolding temperature This protein gave
rise to a broad asymmetric thermogram with
Tm= 78.1C (data not shown) Proteolytic removal
of the N-terminal His-tag yielded a symmetric single
peak thermogram with Tm= 84.8C at pH 8.0
Inter-estingly, the unfolding was partially reversible, as
dem-onstrated by the 83% and 74% area recovery of the
second and third scans, respectively (Fig 2) Fitting a
two-state model to the reference- and baseline-corrected
calorimetric trace resulted in an excellent fit, yielding
DH = 414.1 ± 0.7 kJÆmol)1, attesting to the high
ther-mal stability and conformational integrity of
CBM45-2A from StGWD This CBM displayed similarly high
conformational stability at pH 7.0 (Tm= 87.1C), but
the reversibility was decreased significantly and the
pro-tein was prone to aggregation (data not shown) The
reason for this extraordinarily high stability is
unknown, but it strongly suggests that CBM45-2A is
correctly folded and justifies the use of the isolated
domain to investigate the binding properties of
CBM45 Thermostability is often associated with lower
structural flexibility, which may influence ligand
inter-actions, but how this affects the binding properties of
the domain is unclear In the case of the CBM20
domain from Aspergillus niger glucoamylase, binding
site 2, which is characterized by high conformational
flexibility and large rearrangements on binding,
dis-plays higher affinity towards b-cyclodextrin when
com-pared with the structurally rigid site 1 [30,31]
AtAMY3 was expressed as either the full-length pro-tein, excluding the chloroplast transit peptide (amino acids 68–887), or as the tandem CBM45s (amino acids 68–388) Two versions of the double module carrying
a His-tag at either end gave rise to some soluble pro-tein but, as these propro-teins were prone to aggregation and rapid degradation, they could not be characterized any further In contrast, full-length AtAMY3 carrying
a C-terminal His-tag was soluble and was produced in satisfactory yields (2–3 mgÆL)1) in a fermentor This recombinant AtAMY3 was capable of releasing reduc-ing sugars from amylopectin, glycogen and b-limit dex-trin at both pH 6.2 and pH 8.0 and 30–37C (data not shown), indicating that the recombinant protein was correctly folded
Affinity measurements using surface plasmon resonance (SPR) and isothermal titration calorim-etry (ITC)
It has been shown previously that the isolated CBM20 domain from AtPWD has relatively low affinity towards cyclodextrins [20] SPR was employed to mea-sure the affinity of StGWD CBM45-2A for selected soluble oligosaccharides The domain was biotinylated, immobilized on a streptavidin-coated chip and probed for carbohydrate-binding ability at pH 8.0 (Fig 3) The resulting dissociation constants (Kd) towards both a- and b-cyclodextrin, as well as
6-O-a-maltosyl-b-Fig 2 Differential scanning calorimetry (DSC) analysis of StGWD CBM45-2A Reference subtracted thermograms of 50 l M
StGWD CBM45-2A in 25 m M Hepes, pH 8.0 The full black line, grey line and broken black line are the thermograms of the first, second and third scans, respectively, at a rate of 1 CÆmin)1.
Trang 5cyclodextrin, were in the submillimolar range and
com-parable with the values obtained for the AtPWD
CBM20 (Table 1) [20] The related function of the
cat-alytic modules in AtPWD and StGWD is thus
matched by a similar range of affinity of their SBDs,
despite the fact that they have been assigned to two
different CBM families This property was confirmed
by SPR analysis of the binding of b-cyclodextrin to
AtAMY3 in a similar experimental set-up, using a
dif-ferent immobilization chemistry The Kd value
obtained was in the same range as for StGWD
CBM45-2A, suggesting that the weak binding of the
starch mimic b-cyclodextrin is a general feature of
CBM45s (Table 1) For both proteins, an approximate
two-fold variation was observed in the calculated Kd
value between identical experiments This was most
probably a consequence of the low affinity manifested
in low signal-to-noise ratios, particularly at high
cyclo-dextrin concentrations, resulting in elevated back-ground levels For this reason, cyclodextrin concentrations above 1 mm were excluded from the subsequent data analysis The data in Table 1 were obtained from a representative experiment giving the best fit to the binding curve (lowest v2) The StGWD CBM45-2A domain showed no detectable affinity towards maltoheptaose This is not surprising, as the affinity of SBDs for linear oligosaccharides is generally much lower than for cyclodextrins, because of the additional entropic penalty associated with the stabil-ization of the conformation of the linear ligand upon binding [32]
To corroborate the affinity range acquired in the SPR experiment, StGWD CBM45-2A was analysed by ITC with b-cyclodextrin at pH 7.0 and pH 8.0 Although binding was evident in both cases, the heat responses were small The data obtained at pH 7.0 were noisy, suggesting that the protein was more prone
to aggregation at this pH A one-site binding model was fitted to the integrated ITC data, giving a Kdvalue
of 0.68 ± 0.02 mm for the binding of b-cyclodextrin
to StGWD CBM45-2A at pH 8.0 (Fig 4), in good agreement with the value obtained using SPR (Table 1) The measured heat of dilution was negligible and was disregarded in the integrations The binding was driven by a favourable enthalpy change, which compensated for an unfavourable entropy change (DH =)42.1 ± 0.9 kJÆmol)1; TDS =)24.1 kJÆmol)1) The binding affinity at pH 7.0 (Kd= 0.44 mm) was similar to that at pH 8.0 This thermodynamic finger-print is consistent with the binding of b-cyclodextrin to other SBDs [33]
The observed binding affinity of CBM45 for cyclo-dextrins is considerably lower than that of other char-acterized SBDs from microbial amylolytic enzymes Analysis of CBM20 and CBM21 SBDs from two glu-coamylases gave Kdvalues of 14.4 and 5.1 lm, respec-tively, for the interaction with b-cyclodextrin [30,34]
0
5
10
15
20
25
30
35
-cyclodextrin (µM)
Fig 3 Surface plasmon resonance (SPR) analysis of b-cyclodextrin
binding to CBM45 The instrument response level (RU, response
units) is plotted (±SE) as a function of b-cyclodextrin concentration
for StGWD CBM45-2A (squares) and full-length AtAMY3 (triangles).
The full lines represent the fit to a one-site binding model The
experiments were carried out in triplicate at 25 C and pH 8.0.
Table 1 Dissociation constants for CBM45 determined using SPR Data are from representative experiments (±SE) giving the best fit (low-est v2) to the binding curves Each experiment was run in triplicate K d , dissociation constant; R max , maximum binding response; RU, response units; v 2 , chi-squared test value for the fitted curve.
a Previously published data [20].
Trang 6Similarly, a CBM41 SBD from Thermotoga maritima
pullulanase showed a Kdvalue of 42 lm for the
inter-action with b-cyclodextrin [35] An amylase from
Bacillus halodurans carrying both a CBM25 and a
CBM26 SBD gave Kd values in the range 0.01–1 mm
for the binding of various linear maltooligosaccharides
to the individual SBDs [36] As mentioned above, the
affinity for linear ligands is generally lower than for
cyclic ligands, and StGWD CBM45-2A showed no
binding to maltoheptaose Taken together, the data
presented here show that the binding affinity of the
CBM45 SBD is one to two orders of magnitude lower than the SBDs typically appended to microbial amylo-lytic enzymes This clearly distinguishes the CBM45s from these more thoroughly studied SBDs and, together with the previous report on the CBM20 domain from AtPWD [20], suggests that low-affinity interactions are a recurring characteristic of plastidial starch metabolism This would permit a more dynamic interaction with the starch granule, which may be nec-essary for the accurate control of the rate of degrada-tion [6,20,28] The glucan phosphoryladegrada-tion carried out
by GWD and PWD is an essential initial step in starch degradation in both tubers and leaves, and it has been suggested that the plant controls the release of energy from starch at this crucial step [18,19] It is possible that the low binding affinity of the single domain is masked by avidity effects of the tandem arrangement
of CBMs in the native enzyme This has recently been observed for the triple-CBM53-containing chloroplas-tic starch synthase III from Arabidopsis [25] The low binding affinity of native AtAMY3 towards b-cyclo-dextrin would suggest that this is not the case for the tandem CBM45s but, based on the current data, a small avidity effect cannot be entirely ruled out
AtAMY3 binding to starch in vitro The full-length AtAMY3 offered an advantage when examining the binding affinity of CBM45 SBDs to starch, as it displayed catalytic activity and, being a full-length enzyme, misinterpretation of binding data
as a result of instability or aggregation would most likely be minimal compared with the isolated CBM45s Hence, the starch-binding ability of purified recombi-nant AtAMY3 was demonstrated by incubation with starch isolated from leaves of tobacco plants Binding was carried out at 4C and the a-amylase activity of the unbound fraction was subsequently measured A one-site binding model was fitted to the binding iso-therm (Fig 5), resulting in a Kd value of 36 ± 6.8 mgÆmL)1 and maximum binding capacity (Bmax) of
93 ± 5.6% This affinity is up to two orders of magni-tude lower than that reported previously for the bind-ing of various CBM20 domains to starch [6] A similar experiment using maize starch resulted in comparable affinity, but substantially lower binding capacity (Kd= 21 ± 9.5 mgÆmL)1, Bmax= 42 ± 4.2%) A pre-vious binding analysis of a construct encompassing the isolated recombinant StGWD CBM45-1 to granular potato starch yielded a dissociation constant in the same range (Kd= 7.2 mgÆmL)1, Bmax= 53%) [7] This construct, however, contained approximately 70 amino acids of the C-terminal intervening sequence of
Fig 4 Isothermal titration calorimetry (ITC) analysis of StGWD
CBM45-2A interaction with b-cyclodextrin The top panel depicts
the raw heat response for each b-cyclodextrin injection, and the
bottom panel depicts the binding isotherm; open circles represent
the integrated binding heat of the data in the top panel, and the full
line is the fit of a one-site binding model to the integrated binding
data The experiment was carried out at 25 C by titrating 50 l M
protein in 25 m M Hepes, pH 8.0 with 4 m M b-cyclodextrin in the
same buffer.
Trang 7unknown function (amino acids 68–286) and the
struc-tural integrity of the protein was not verified
The data obtained not only support the affinity
range measured using SPR and ITC with the
cyclodex-trin starch mimics, but also verify the starch-binding
ability of AtAMY3 in vitro It cannot be precluded
that some binding may involve secondary binding sites
in the catalytic domain, but the demonstrated
starch-binding ability of different isolated CBM45s used in
both the present and other studies [7,10] suggests that
AtAMY3 does indeed interact with starch granules
through the tandem CBM45 domains
CBM45 interaction with starch granules in planta
It has been shown by immunoblotting that StGWD
binds to starch in planta in its active full-length form
[23] To investigate whether the isolated CBM45s from
StGWD function as SBDs in planta, they were
C-termi-nally fused to yellow fluorescent protein (YFP), either
singly or as the double module, and transiently
expressed in tobacco leaves The constructs were
tar-geted to the chloroplasts by an N-terminal fusion to the
transit peptide of Arabidopsis GWD1 In a similar
experimental set-up, fusions between green fluorescent
protein (GFP) and either full-length AtAMY3 or the
tandem CBM45s were analysed Investigation of
locali-zation using confocal laser scanning microscopy showed
clear targeting to the chloroplasts of mesophyll cells
and binding to disc-shaped transient starch granules
for StGWD CBM45-2, CBM45-1,2 and full-length
AtAMY3 (Fig 6) The StGWD CBM45-1 fusion, in
contrast, gave rise to numerous highly fluorescent inclu-sion body-like structures with no clear targeting (data not shown) The double module from AtAMY3 did not yield any visible signal The behaviour of these proteins
is likely to be affected by their instability and observed tendency to aggregate in isolated form (see above), sug-gesting that the CBM45s depend on packing contacts with other domains in the native enzyme Further support for the binding of isolated CBM45s to starch comes from a previous report showing binding to starch
in vitroof an StGWD construct encompassing
CBM45-1 [7] and an in planta localization analysis of GFP-tagged CBM45-1 from Arabidopsis GWD2 [10] In the present study, it has been demonstrated that both iso-lated single and double CBM45 domains from StGWD are capable of binding to starch, and that full-length AtAMY3 binds to starch both in vitro and in planta
Conclusion
In the present study, the carbohydrate-binding proper-ties of two representative plastidial enzymes containing CBM45-type SBDs were characterized The CBM45s
A
B
C
Fig 6 Transient expression of CBM45-YFP ⁄ GFP fusions in tobacco leaf mesophyll cells Single and double CBM45 domains from StGWD and the full-length AtAMY3 were fused to YFP or GFP, respectively, and transiently expressed in Nicotiana benthami-ana leaves by infiltration with Agrobacterium tumefaciens Expres-sion and localization were investigated by confocal laser scanning microscopy YFP ⁄ GFP fluorescence (green), chlorophyll autofluores-cence (red) and a merged image of the two channels are shown (A) StGWD CBM45-2 fused to YFP (B) StGWD CBM45-1,2 fused
to YFP (C) Full-length AtAMY3 fused to GFP Scale bar, 20 lm.
Starch (mg·mL –1 )
0
20
40
60
80
100
Fig 5 Binding of AtAMY3 to tobacco leaf starch in vitro
Recombi-nant AtAMY3 protein was incubated with starch isolated from
leaves of Nicotiana benthamiana for 45 min at 4 C Unbound
pro-tein was assayed for activity by measuring the release of
reducing-end sugars from amylopectin Each data point (±SE) is the average
of four independent experiments.
Trang 8were demonstrated to exhibit up to two orders of
mag-nitude lower affinity towards both cyclodextrins and
granular starch, compared with typical SBDs
encoun-tered in microbial amylolytic enzymes This behaviour
is analogous to that of the CBM20-type SBD from
AtPWD [20] and supports the hypothesis that
low-affinity SBDs are important for dynamic and reversible
interactions in starch metabolism [28] It remains
unclear how the functionality of these low-affinity
SBDs is integrated with other levels of regulation, such
as the observed diurnal effects of light and redox
conditions [23,27] Further studies will be required to
elucidate these details of starch metabolism and the
structural elements responsible for the lower affinity of
CBM45 The outcome of the present study
demon-strates the large functional diversity of SBDs that has
only started to be addressed, and the investigation of
SBDs occurring in nonhydrolytic, starch- and
glyco-gen-active enzymes will be essential to understand the
contribution of such atypical SBDs to this group of
important enzymes
Materials and methods
Cloning, expression and purification of CBM45
domains from potato GWD
DNA fragments of S tuberosum GWD (accession number
AY027522) were amplified as outlined in Fig S3 using the
primers given in Table S1 The PCR products were cloned
using Gateway technology (Invitrogen, Carlsbad, CA,
USA) via the entry vector pENTR⁄ TEV ⁄ D-TOPO, and
subsequently moved into the expression vector pDEST17
(Invitrogen) with an N-terminal TEV protease-cleavable
His-tag The expression vectors were transformed into
Escherichia coliBL21 cells Cultures were grown in 6· 1 L
scale in Tunair flasks (Sigma-Aldrich, St Louis, MO, USA)
at 37C, cooled to 16 C before induction with 0.5 mm
iso-propyl thio-b-d-galactoside (IPTG), and harvested 16 h
after induction For protein purification, cell pellets were
lysed in 15 mL Bugbuster (Novagen, Merck4Biosciences,
Nottingham, UK) with 5 lL of Benzonase Nuclease
(Sigma-Aldrich) Following centrifugation, the supernatant
was loaded onto a HisTrap HP, 5 mL column (GE
Health-care, Uppsala, Sweden) and eluted by a 40–400 mm
imidaz-ole gradient in 20 mm Tris⁄ HCl pH 8.0, 500 mm NaCl,
10% v⁄ v glycerol and 0.5 m betaine according to the
manu-facturer’s instructions
TEV protease cleavage of the His-tag was performed
with AcTEV protease according to the manufacturer’s
instructions (Invitrogen) For large-scale production,
incu-bation was performed overnight at room temperate using
25% of the recommended amount of protease The cleaved
untagged protein was purified by anion exchange on a
Mono Q 10⁄ 100 GL column (GE Healthcare) in 20 mm Tris⁄ HCl pH 8.0 and eluted with 20 column volumes of a 0–0.5 m NaCl gradient After dialysis to remove NaCl, the protein was stored at 4C
Cloning, expression and purification of
A thaliana AMY3
An AtAMY3 cDNA clone (At1g69830, accession number AY050398) was obtained from the RIKEN Arabidopsis Genome Encyclopedia (RARGE, http://rarge.psc.riken.jp) Full-length AtAMY3 excluding the chloroplast transit pep-tide and stop codon (amino acids 68–887) was amplified (primers AtCBM1-NcoI and AtpAMY-NotI, Table S1) and cloned into the NcoI and NotI sites of the expression vector pET-28a containing a C-terminal 6· His-tag The con-struct was transformed into E coli BL21 Rosetta (DE3) cells (Novagen) Protein expression was carried out in either
a 5 L bioreactor (Biostat B, B Braun Biotech International, Melsungen, Germany) on defined medium [37] by induction
at an absorbance at 600 nm (A600) of 5 with 0.1 mm IPTG
at 16C and harvesting after 22 h, or in shake-flasks by induction with 0.2 mm IPTG at 20C and harvesting after 16–18 h The cell pellet was resuspended in buffer A (20 mm Hepes pH 7.5, 500 mm NaCl, 40 mm imidazole, 40% v⁄ v glycerol, 0.1% v ⁄ v Triton X-100, 0.5 mm CaCl2) containing 2 mm dithithreitol, 0.1 lLÆmL)1 Benzonase Nuclease (Sigma-Aldrich) and one Complete Mini protease inhibitor tablet (Roche, Basle, Switzerland), and lysed using
a French press The lysate was incubated on ice for 30 min, clarified by centrifugation and filtered through a 0.22 lm filter The filtrate was applied to a HisTrap HP, 1 mL col-umn (GE Healthcare), washed with a 40–70 mm imidazole gradient for 10 column volumes, and eluted with 20 column volumes of a 70–400 mm imidazole gradient at 0.5 mLÆ min)1 Concentrated, partially pure AtAMY3 was applied
to a HiLoad Superdex 200 16⁄ 60 gel filtration column (GE Healthcare) and eluted in 20 mm Hepes pH 7.5, 150 mm NaCl, 25% v⁄ v glycerol and 0.5 mm CaCl2 The fractions containing AtAMY3 were pooled and concentrated to approximately 1 mgÆmL)1and stored at 4C
DSC analysis
DSC analysis was performed using a VP-DSC calorimeter (MicroCal, Northampton, MA, USA) with a cell volume of 0.52061 mL at a scan rate of 1CÆmin)1 Samples were dialysed in at least 500 volumes of 25 mm Hepes–NaOH,
pH 7.0 or pH 8.0, and degassed for 10 min at 20C Base-line scans collected with buffer in the reference and sample cells were subtracted from sample scans The reversibility of the thermal transitions was evaluated by checking the reproducibility of the scan on immediate cooling and rescanning The initial screening of the conformational sta-bility of purified StGWD constructs was performed using a
Trang 9protein concentration of 0.5 mgÆmL)1 in 25 mm Hepes–
NaOH, pH 8.0 The DSC analysis of the form with the
highest Tmvalue (StGWD CBM45-2A) was performed
fol-lowing cleavage of the His-tag with TEV protease (see
above), and subsequent repurification and dialysis of 50 lm
protein as mentioned above Origin v7.038 software with a
DSC add-on module was used for data analysis, Tm
(unfolding temperature, defined as the temperature of
maxi-mum apparent heat capacity) assignment and unfolding
enthalpy calculations
SPR analysis
Measurements of interactions with soluble ligands using
SPR were carried out on a Biacore T100 (GE Healthcare)
Domains of StGWD were biotinylated using EZ-Link
Sul-fo-NHS-LC-Biotin (Pierce, Thermo Scientific, Rockford,
IL, USA) in 10 mm Mes pH 6.8, 5 mm CaCl2 and 8 mm
b-cyclodextrin, and immobilized on a streptavidin-coated
chip (sensor chip SA, GE Healthcare) using a standard
pro-gram, aiming for a density of 1250 response units (RU)
AtAMY3 was immobilized on a carboxymethylated dextran
chip (sensor chip CM5, GE Healthcare) in 10 mm sodium
acetate pH 4.6, 20% v⁄ v glycerol, 1 mm CaCl2 and 2 mm
b-cyclodextrin, using a standard program, aiming for a
den-sity of 7500 RU Sensograms were collected at 25C in
25 mm Hepes pH 8.0, 150 mm NaCl, 0.5 mm CaCl2 and
0.005% v⁄ v P20 surfactant (GE Healthcare) at a flow rate
of 30 lLÆmin)1, contact time of 90–180 s and dissociation
time of 100–240 s Experiments were run in triplicate in the
range 0–2000 lm for each carbohydrate dissolved in the same
buffer All data evaluation was carried out using the Biacore
T100 evaluation software
ITC analysis
Experiments were performed using an MCS isothermal
titration calorimeter (MicroCal) Titrations were performed
by injecting 5 lL b-cyclodextrin in 25 mm Hepes–NaOH
pH 7.0 or 8.0 into a stirred (400 rpm) 1.3187 mL cell
con-taining 50 lm StGWD CBM45-2A in the same buffer For
each titration of enzyme, the dialysis buffer of the sample
was titrated as a control using the same b-cyclodextrin
stock to measure the heat of dilution The control titration
consisted of 10 injections of 1 lL in 2.5 s for the first
injec-tion and 5 lL for the rest, and with 180 s of equilibrainjec-tion
between injections Titrations of the protein were carried
out similarly, but were continued until no significant
response was observed on ligand injections Origin software
supplied with the instrument was used to analyse the data
Starch-binding assays
Tobacco leaf starch was isolated from 5-week-old
Nicoti-ana benthamiNicoti-ana The harvested leaves were homogenized
in 0.2% SDS in a polytron PT3000 blender (Kinematica
AG, Lucerne, Switzerland) and filtered sequentially through
2· 100 lm and 2 · 20 lm filtration cloth Following cen-trifugation, the starch pellet was washed twice in 0.2% SDS, three times in water, twice in 96% ethanol and air dried
Recombinant AtAMY3 (3 lg) was incubated with tobacco leaf starch granules in a 350 lL mixture containing
20 mm Hepes pH 7.5, 0.5 mm CaCl2, 0.05 mgÆmL)1 BSA and 0–200 mgÆmL)1 starch The suspension was incubated with gentle mixing at 4C for 45 min The supernatant (198 lL) containing unbound AtAMY3 was treated with
10 mm dithiothreitol for 20 min at 25C, and a-amylase activity was measured by adding 50 lL of a 25 mgÆmL)1 amylopectin solution (Fluka 10118, dissolved in 20 mm Hepes pH 7.5, 0.5 mm CaCl2) and incubating for 45 min at
37C Reactions were stopped by mixing with an equal volume of 0.5 m NaOH, and liberated reducing ends were determined by the 3-methyl-2-benzothiazolinone hydrazone method, as described previously [38] The activity (expressed as the percentage of bound AtAMY3 when com-pared with a no-starch control) was plotted against the starch concentration, and the data were fitted to a one-site binding model
Transient expression of CBM45s in tobacco
The CBM45s from StGWD were C-terminally fused to YFP, either singly (CBM45-1, amino acids 109–217; CBM45-2, amino acids 405–545) or as the entire double module (CBM45-1,2, amino acids 77–551) Fragments were PCR amplified using uracil-containing primers (Table S1) and cloned into the vector pPS48uYFP using an improved USER (uracil-specific excision reagent; New England Biolabs, Ipswich, MA, USA) cloning procedure [39] The chloroplast transit peptide of Arabidopsis GWD1 (amino acids 1–77) was fused to each construct by simultaneous cloning of both fragments as described previously [10] The full-length ORF of AtAMY3 (primers AMY3-F and AMY3-R, Table S1), as well as an N-terminal fragment (amino acids 1–391) covering both CBM45s (primers AMY3-F and AMY3SBD-R, Table S1), were fused to enhanced GFP in the binary vector pK7FWG2 [40] using GATEWAY cloning technology (Invitrogen) Constructs were transformed into Agrobacterium tumefaciens and tran-siently expressed by infiltration in Nicotiana benthamiana as described previously [41] Expression and localization were analysed by a confocal laser scanning microscope (TCS SP2, Leica Microsystems, Wetzlar, Germany) equipped with a 20· ⁄ 0.70 or 63 · ⁄ 1.20 PL APO water immersion objective A 488 nm laser line was used for excitation, and emission was detected between 520 and 550 nm for YFP fluorescence, between 510 and 535 nm for GFP fluores-cence, and between 600 and 750 nm for chlorophyll auto-fluorescence
Trang 10MAG was supported by a grant from the Danish
Research Council for Technology and Production
Sci-ences (grant no 274-06-0312) and MJB by a Hans
Christian Ørsted postdoctoral fellowship from the
Technical University of Denmark The financial
sup-port from the Carlsberg Foundation (to BS), the
Vil-lum Kann Rasmussen Foundation (to the VKR
Research Centre Pro-Active Plants) and ETH Zu¨rich
and the Swiss–South African Joint Research
Pro-gramme (grant no 08 IZ LS Z3122916, to SCZ and
DS) is gratefully acknowledged
References
1 Tester RF, Karkalas J & Qi X (2004) Starch –
composi-tion, fine structure and architecture J Cereal Sci 39,
151–165
2 Nielsen MM, Bozonnet S, Seo ES, Motyan JA,
Ander-sen JM, Dilokpimol A, Abou Hachem M, Gyemant G,
Naested H, Kandra L et al (2009) Two secondary
carbohydrate binding sites on the surface of barley
a-amylase 1 have distinct functions and display synergy
in hydrolysis of starch granules Biochemistry 48,
7686–7697
3 Machovic M & Janecek S (2006) Starch-binding
domains in the post-genome era Cell Mol Life Sci 63,
2710–2724
4 Guillen D, Sanchez S & Rodriguez-Sanoja R (2010)
Carbohydrate-binding domains: multiplicity of
biologi-cal roles Appl Microbiol Biotechnol 85, 1241–1249
5 Cantarel BL, Coutinho PM, Rancurel C, Bernard T,
Lombard V & Henrissat B (2009) The
Carbohydrate-Active EnZymes database (CAZy): an expert resource
for glycogenomics Nucleic Acids Res 37, D233–D238
6 Christiansen C, Abou Hachem M, Janecek S,
Vikso-Nielsen A, Blennow A & Svensson B (2009) The
carbo-hydrate-binding module family 20 – diversity, structure,
and function FEBS J 276, 5006–5029
7 Mikkelsen R, Suszkiewicz K & Blennow A (2006) A
novel type carbohydrate-binding module identified in
a-glucan, water dikinases is specific for regulated
plastidial starch metabolism Biochemistry 45, 4674–
4682
8 Lorberth R, Ritte G, Willmitzer L & Kossmann J
(1998) Inhibition of a starch-granule-bound protein
leads to modified starch and repression of cold
sweeten-ing Nat Biotechnol 16, 473–477
9 Yu TS, Kofler H, Hausler RE, Hille D, Flugge UI,
Zeeman SC, Smith AM, Kossmann J, Lloyd J, Ritte G
et al.(2001) The Arabidopsis sex1 mutant is defective
in the R1 protein, a general regulator of starch
degradation in plants, and not in the chloroplast hexose
transporter Plant Cell 13, 1907–1918
10 Glaring MA, Zygadlo A, Thorneycroft D, Schulz A, Smith SM, Blennow A & Baunsgaard L (2007) An extra-plastidial a-glucan, water dikinase from Arabidop-sisphosphorylates amylopectin in vitro and is not neces-sary for transient starch degradation J Exp Bot 58, 3949–3960
11 Yu TS, Zeeman SC, Thorneycroft D, Fulton DC, Dunstan H, Lue WL, Hegemann B, Tung SY, Umem-oto T, Chapple A et al (2005) a-amylase is not required for breakdown of transitory starch in Arabidopsis leaves J Biol Chem 280, 9773–9779
12 Kotting O, Santelia D, Edner C, Eicke S, Marthaler T, Gentry MS, Comparot-Moss S, Chen J, Smith AM, Steup M et al (2009) STARCH-EXCESS4 is a laforin-like phosphoglucan phosphatase required for starch degradation in Arabidopsis thaliana Plant Cell 21, 334–346
13 Streb S, Delatte T, Umhang M, Eicke S, Schorderet M, Reinhardt D & Zeeman SC (2008) Starch granule bio-synthesis in Arabidopsis is abolished by removal of all debranching enzymes but restored by the subsequent removal of an endoamylase Plant Cell 20, 3448–3466
14 Chou WY, Chou WI, Pai TW, Lin SC, Jiang TY, Tang
CY & Chang MDT (2010) Feature-incorporated align-ment based ligand-binding residue prediction for carbo-hydrate-binding modules Bioinformatics 26, 1022–1028
15 Smith AM & Stitt M (2007) Coordination of carbon supply and plant growth Plant Cell Environ 30, 1126–1149
16 Vriet C, Welham T, Brachmann A, Pike M, Pike J, Perry J, Parniske M, Sato S, Tabata S, Smith AM et al (2010) A suite of Lotus japonicus starch mutants reveals both conserved and novel features of starch metabo-lism Plant Phys 154, 643–655
17 Fettke J, Hejazi M, Smirnova J, Hochel E, Stage M & Steup M (2009) Eukaryotic starch degradation: integra-tion of plastidial and cytosolic pathways J Exp Bot 60, 2907–2922
18 Kotting O, Kossmann J, Zeeman SC & Lloyd JR (2010) Regulation of starch metabolism: the age of enlightenment? Curr Opin Plant Biol 13, 321–329
19 Zeeman SC, Kossmann J & Smith AM (2010) Starch: its metabolism, evolution, and biotechnological modifi-cation in plants Annu Rev Plant Biol 61, 209–234
20 Christiansen C, Abou Hachem M, Glaring MA, Vikso-Nielsen A, Sigurskjold BW, Svensson B & Blennow A (2009) A CBM20 low-affinity starch-binding domain from glucan, water dikinase FEBS Lett 583, 1159–1163
21 Delatte T, Umhang M, Trevisan M, Eicke S, Thorney-croft D, Smith SM & Zeeman SC (2006) Evidence for distinct mechanisms of starch granule breakdown in plants J Biol Chem 281, 12050–12059
22 Fettke J, Chia T, Eckermann N, Smith A & Steup M (2006) A transglucosidase necessary for starch