Linköping Studies in Health Science No. 105
Studies of recombinant forms of
Aleuria aurantia lectin
Johan Olausson
Department of Clinical and Experimental Medicine, Faculty of Health Sciences
Linköpings Universitet, SE-581 85 Linköping. Sweden
Linköping 2009
© Johan Olausson, 2009
Paper I has been reprinted with permission from the copyright holder.
ISBN 978-91-7393-558-6
ISSN 1100-6013
Table of contents
List of papers ...................................................................................................................... 6 Populärvetenskaplig sammanfattning ................................................................................ 7 Abstract .............................................................................................................................. 8 Abbreviations ..................................................................................................................... 9 Background ...................................................................................................................... 10 Protein‐carbohydrate interactions ............................................................................................. 10 Lectins ....................................................................................................................................... 10 Classification of lectins .............................................................................................................. 11 Lectins as research tools ............................................................................................................ 12 Aleuria aurantia lectin ............................................................................................................... 12 Structure of AAL ........................................................................................................................ 12 Binding characteristics of AAL.................................................................................................... 13 Fucose ....................................................................................................................................... 13 Fucosylation as a disease marker............................................................................................... 14 Aims of the thesis.............................................................................................................. 15 Paper I....................................................................................................................................... 15 Paper II...................................................................................................................................... 15 Methods ........................................................................................................................... 16 Construction of plasmids ........................................................................................................... 16 His-tagged rAAL – Paper I .......................................................................................................................... 16 His-tagged monomeric AAL – Paper II ....................................................................................................... 16 His-tagged monovalent AAL – Paper II....................................................................................................... 16 Circular dichroism – Paper I and II.............................................................................................. 17 Enzyme Linked Lectin Assay – Paper I and II ............................................................................... 19 Surface Plasmon Resonance – Paper I and II. ............................................................................. 19 Tryptophan fluorescence – Paper I............................................................................................. 22 Hemagglutination – Paper I and II ............................................................................................. 22 Analytical Ultracentrifugation – Paper II.................................................................................... 23 Results and Discussion ...................................................................................................... 25 Finding a high affinity binding site in recombinant AAL – Paper I ............................................... 25 Construction of a monomeric and a monovalent fucose binder – Paper II................................... 28 4
Tack!................................................................................................................................. 30 References ........................................................................................................................ 31 5
List of papers
This thesis is based on the following papers;
I.
Olausson, J., Tibell, L., Jonsson, B-H. and Påhlsson, P. (2008) Detection of a
high affinity binding site in recombinant Aleuria aurantia lectin. Glycoconj J.
25, 753-62
II.
Olausson, J., Tibell, L., Jonsson, B-H. and Påhlsson, P. (2009) Production and
characterization of a monomeric and a monovalent form of Aleuria aurantia lectin.
In manuscript
6
Populärvetenskaplig sammanfattning
Förändringar av de kolhydrater som sitter fast på de flesta av våra proteiner i kroppen har
kunnat knytas till flera olika typer av sjukdomar. Man säger att sjukdomar ger ett förändrat
kolhydratmönster. Till exempel uppvisar alla typer av tumörceller ett förändrat
kolhydratmönster jämfört med normalvävnad. Vår forskargrupp har tidigare visat att proteiner
som finns i blodbanan (plasmaproteiner) ofta uppvisar kolhydratmönster som går att relatera
till en viss sjukdom. Ett exempel är plasmaproteinet orosomukoid vars förändringar i
kolhydratmönstret har visat sig kunna användas för diagnostik av till exempellevercirrhos.
För att mäta förändringar av kolhydratmönstret som kan uppstå vid sjukdom används
proteiner som känner igen och binder till kolhydrater. Dessa naturligt förekommande
kolhydratbindande proteiner kallas lektiner. Vi har nyligen visat att ett lektin vid namn
Aleuria aurantia lektin (AAL) kan användas för att hitta de förändringar i kolhydratmönstret
som uppkommer vid leversjukdom. AAL binder till en kolhydrat som kallas fukos och det är
just mängden av fukos som förändras vid leversjukdom. I detta arbete har vi detaljstuderat
AAL för att få bättre kunskap om hur AAL binder till fukos. Detta för att det ska vara möjligt
att förbättra AALs förmåga att upptäcka olika sjukdomstillstånd.
Vi har producerat AAL i bakterier och visar att bakterieproducerat AAL kan binda mycket
starkt till fukos. Vi har också kunnat göra två mindre former av bakterieproducerat AAL som
båda kan binda fukos. Den minsta av dessa former binder något annorlunda till fukos än de
övriga två AAL formerna. Denna AAL form tror vi kan ha nya användningsområden inom
diagnostik. Studierna har dels gett oss större förståelse för hur AAL binder till fukos men de
visar också att DNA-teknik kan användas för att påverka ursprungsegenskaperna hos ett
lektin. De DNA-tekniska metoderna kan vara viktiga redskap för att skapa nya tillförlitliga
kolhydrat bindande reagens för sjukdomsdiagnostik.
7
Abstract
The presented work describes construction and analysis of recombinantly produced forms of
Aleuria aurantia lectin (AAL). The binding properties of the produced AAL forms were
studied using techniques such as tryptophan fluorescence, hemagglutination analysis, ELISA
and surface plasmon resonance analysis.
Lectins are proteins that are ubiquitous in nature with the ability to bind specifically to
different types of carbohydrates. The physiological function of different lectins is not always
known, but they are involved in many recognition events at molecular and cellular levels. In
research, lectins are widely used for structural and functional studies of complex
carbohydrates, and they are also used to detect changes in the carbohydrate pattern on
glycoproteins in different diseases.
With the use of recombinant technology it is now possible to refine properties of lectins such
as decreasing the valency and alter specificity and affinity. This may be a way of constructing
more suitable reagents for use in diagnostic glycosylation analysis assays.
AAL has been extensively used in different types of research for its ability to bind the
monosaccharide fucose and to fucose-containing oligosaccharides. It is composed of two
identical subunits where each subunit contains five binding sites for fucose. AAL was
expressed recombinantly (rAAL) and its properties was investigated. These studies reveled
that one of the binding sites in rAAL had unusually high affinities towards fucose and fucosecontaining oligosaccharides with Kd-values in the nanomolar range. This binding site is not
detected in AAL that have been exposed to fucose during its purification, and therefore we
proposed that this site may be blocked with free fucose in commercial preparations of AAL.
Normally lectin-oligosaccharide interactions are considered to be of weak affinity, so the
finding of a high affinity site was interesting for the future study of recombinant forms of
AAL. The next step was to produce recombinant AAL forms with decreased valency. This
was done using site-directed mutagenesis. First a monomeric form of AAL (mAAL) was
constructed and then a monovalent form of AAL, containing only one fucose-binding site
(S2-AAL) was constructed. Both of these forms had retained ability to bind fucose. The
binding characteristics of mAAL were similar to that of rAAL, but mAAL showed decreased
hemagglutinating activity. S2-AAL showed a lower binding affinity to fucosylated
oligosaccharides and did not bind to sialylated fuco-oligosaccharides such as sialyl-LewisX.
This study shows that molecular engineering techniques could be important tools for
development of reliable and specific diagnostic and biological assays for carbohydrate
analysis.
8
Abbreviations
AAA
AAL
AGP
BSA
CD
ELLA
Fuc
GAG
Gal
GalNAc
GlcNAc
HA
HAI
ITC
Kd
koff
kon
mAAL
Man
MIC
nAAL
Neu5Ac
PA – IIL
PBS
RA
rAAL
RSL
S2-AAL
SE
SPR
SV
Anguilla anguilla agglutinin
Aleuria aurantia lectin
α1-acid glycoprotein
Bovine serum albumin
Circular dichroism
Enzyme linked lectin assay
Fucose
Glycosaminoglycan
Galactose
N-acetylgalactosamine
N-acetylglucosamine
Hemagglutination
Hemagglutination inhibition
Isothermal titration calorimetry
Dissociation constants
Dissociation rate constant
Association rate constant
His-tagged recombinant monomeric Aleuria aurantia lectin (AAL-Y6R)
Mannose
Minimum inhibiting concentration
Native Aleuria aurantia lectin
Sialic acid
Pseudomonas aeruginosa lectin
Phosphate buffered saline
Rheumatoid arthritis
His-tagged recombinant Aleuria aurantia lectin
Ralstonia solanacearum lectin
His-tagged recombinant site 2 of Aleuria aurantia lectin
Sedimentation equilibrium
Surface plasmon resonance
Sedimentation velocity
9
Background
Protein-carbohydrate interactions
Protein-carbohydrate interactions are normally considered as weak affinity interactions.
Weak biological interactions are determined by a dissociation constant of about 0.10-0.01
mM [1]. Protein-carbohydrate binding is mainly formed by two types of interactions:
hydrogen bonds formed from the hydroxyl groups of the carbohydrate, and van der Waals
interactions formed by stacking of the hydrophobic face of carbohydrates to the side chains of
aromatic amino acids [2]. The protein-carbohydrate interactions in biological processes is
often multivalent interaction between protein and carbohydrate, that creates an avidity effect
that compensate the low affinity generally observed for each individual carbohydrate binding
site. In the human body, several biological processes are mediated by weak interactions
between proteins and carbohydrates, one example being rolling of leukocytes along the
endothelial surface during an inflammatory process [3]. During rolling, the speed of the
leukocytes decreases so that they can extravasate through the endothelium to the site of
inflammation. This rolling event is mediated by the selectin family of cell adhesion molecules
including L-, P-, and E-selectin [3, 4]. L-selectin is expressed on the surface of leukocytes
whereas P- and E-selectins are expressed on the surface of endothelial cells. L-selectin
interacts with carbohydrates on endothelial cells, while P-and E-selectins interact with
carbohydrates on the leukocytes [3, 4].
Lectins
Proteins that interact with naturally occurring carbohydrates are generally referred to as
“lectins”. When first discovered lectins were defined as phytoagglutinins because they were
commonly detected by their ability to induce erythrocyte agglutination and were found almost
exclusively in plants. This definition was later revised to define lectins as a class of proteins
of non-immune origin that binds carbohydrates without modifying them. This definition is
still used, but today a broader definition of lectins is starting to be accepted. Now lectins are
defined as proteins which specifically bind (or crosslink) carbohydrates. One exception to the
former definition is actually the oldest known lectin called ricin [5]. This lectin has been
revealed to act as the enzyme RNA-N-glycosidase [6]. Today it is known that lectins are
ubiquitous in nature and found in all kinds of organisms such as animals, fungi, bacteria and
even viruses [7].
Lectins are structurally different, and their binding properties are dependent on the exact
three-dimensional structure and amino acid sequence of the lectin [7, 8]. A specific lectin
usually binds a certain carbohydrate epitope [9].
The biological function for most lectins is still unclear. The function differs depending on the
type of lectin and the species that expresses the lectin but in general they are important in
many recognition events at molecular or cellular level. It is known that bacteria uses lectins to
bind host carbohydrates at epithelial cell surfaces and use these attachment points for further
10
colonization [10]. In fungi, lectins can be of importance for host association (symbiosis or
parasitism) [11]. In humans lectins are involved in cell-cell recognition, leukocyte rolling,
cancer metastasis, fertilization and microbial adherence to host tissues [7, 12, 13]. Plant
lectins are proposed to be involved in the defense against pathogens since lectins bind to
glycoproteins on the surface of parasites [14]
Classification of lectins
Lectins are not easily classified but some attempts have been made. One way of classifying
lectins is according to the small carbohydrate haptens they recognize, e.g. mannose - (Man),
galactose/N-acetylgalactosamine- (Gal/GalNAc), N-acetylglucosamine- (GlcNAc), fucose(Fuc) and sialic acid- (Neu5Ac) binding lectins [15]. This conventional method is practically
useful but it is not necessarily relevant for refined specificity because lectins in the same
category, for example galactose-specific lectins, can show different binding specificity [16].
Since lectins are proteins it is also possible to classify lectins through their evolutionary
kinship. Animal lectins are generally classified according to homologies in the carbohydrate
recognition domain [12, 17, 18]. Some examples of animal lectin families are P-type, C-type,
Galectins (former S-type), I-type and Glycosaminoglycan-binding proteins. The special
features of these families are:
• P-type lectins are lectins that recognize Man-6-P [17].
• C-type lectins are a large family with diverse structure and function that requires calcium
for recognition of carbohydrates (not all calcium-requiring lectins are C-type lectins) [17].
Selectins are a subfamily of C-type lectins distinguished by their specific function in
leukocyte adhesion to endothelial cells through sialyl-LewisX recognition [3, 19].
• Galectins all share galactose specificity and requires free thiols for stability [16].
• I-type lectins are members of the immunoglobulin superfamily. Siglecs are a subfamily of
these molecules that specifically recognize sialic acids [18].
• Glycosaminoglycan-binding proteins (GAG-binders) binds with clusters of positively
charged amino acid that line up against internal regions of the anionic GAG chains. This is
unlike other lectins that bind their carbohydrate ligand with a relatively narrow binding
pocket [20].
Plant lectins have also been classified in families and the largest and most studied is the
legume lectin family [21]. These lectins have a shared consensus sequence and most of them
require Ca2+ and Mn2+ for carbohydrate binding. But since there is such a tremendous
diversity of carbohydrate-binding specificities among plant lectins they are often classified
according to the carbohydrate they bind.
The three dimensional structure has been solved for more than 550 lectins and they all show
proteins that are oligomeric, usually made up of two or four subunits (homo- or hetromers)
11
with typically one single carbohydrate binding site per monomer. Based on their tertiary
structure lectins have been divided into different folding families. Most of these families
contain several β-strands in the core structure [21, 22]. The tertiary structures are similar
within many lectin families [22]. This similarity in the tertiary structure has been extensively
studied for plant lectins and these studies suggest that lectins are conserved throughout
evolution and thus may play important physiological functions [23-25].
Lectins as research tools
Since lectins recognize specific carbohydrate structures they are used to detect and separate
cells, bacteria, and viruses with respect to their carbohydrate content. Lectins are also useful
tools for investigating the structure and function of different carbohydrate chains on
glycoproteins and glycolipids [26, 27]. In clinical research, lectins can be used to detect
changes in the carbohydrate pattern on glycoproteins in different diseases [28-35].
Today many researchers consider lectins as "deciphers of the glycocode". The use of lectins
may be great tools to analyze glycosylation patterns and to facilitate the development of novel
treatments for many diseases in which carbohydrate recognition plays a key role. To achieve
this goal, it is important to learn more about lectin structure, function and binding properties.
Aleuria aurantia lectin
The Aleuria aurantia lectin is obtained from the mushroom Aleuria aurantia and was initially
characterized by its binding to fucosylated blood group antigens [36]. When AAL was
characterized by Kochibe et al. they showed that AAL had the ability to precipitate blood
group substances in an agar gel that contained 0.5 % of SDS (sodium deoxycholate), Triton
X-100 or Tween 80 [36]. The fact that AAL tolerated these detergents made it a promising
tool for purifying and analyzing fucosylated oligosaccharides and glycoproteins. AAL has
during the years been widely used to estimate the extent of fucosylation on glycoproteins and
to fraction glycoproteins [28, 37-41]. AAL has shown to be a non-glycosylated dimer with
two identical 312-amino acid subunits and both of these subunits specifically recognize
fucose [42, 43]. Each subunit has a calculated molecular weight of about 33 400 Dalton [44].
Structure of AAL
Crystallisation studies of AAL show that each AAL monomer consists of a six-bladed βpropeller fold and of a small antiparallel two-stranded β-sheet [42, 43]. The antiparallel twostranded β-sheets have an important role in the dimerization of AAL [42]. Between each of
the six β-blades, a cleft is formed by the loops that connect the blades. Fucose can bind to five
of these six clefts (site 1-5) in each monomer [42, 43]. AAL differs from other fucose-specific
lectins, such as Anguilla anguilla agglutinin (AAA, isolated from the serum of eel) and
Pseudomonas aeruginosa lectin (PA – IIL, isolated from the bacteria Pseudomonas
aeruginosa) that also forms dimers but only have one fucose-binding site per monomer [45,
46]. The recently characterized fucose-binding lectin from the bacterium Ralstonia
solanacearum (RSL) forms trimers with two binding sites per monomer [47, 48].
12
Before the determination of the crystal structure of AAL, no lectins had been found in the
group of six-folded β-propeller proteins. However when RSL was crystallized it also was
included in the group of six-folded β-propeller proteins [48]. RSL forms a homotrimer that
reproduces a six-bladed β-propeller fold [48]. Besides these two lectins a couple of other
proteins are included in this family, for example glucose dehydrogenase and the low-density
lipoprotein (LDL) receptor [22]. In contrast to other six-bladed proteins almost all residues in
AAL and RSL are used to form the propeller structure, and therefore AAL and RSL form the
most compact structures among all known six-bladed proteins [42, 43, 22]. AAL also differs
from other β-propeller proteins, which normally are disc like, in the way that AAL monomers
are shaped like short cylinders with a hydrophobic central cavity [42, 43].
The dimerization of AAL creates a “back to back” association of the two cylinder-shaped
monomers with the fucose-binding sites exposed to the solvent on each side of the dimer [42].
The AAL-monomers that forms the dimer are almost equivalent with the exception of a
marginal contact between site 4 in one monomer and site 5 in the other monomer [42].
Binding characteristics of AAL
The dissociation constant (Kd) for interaction between free fucose and AAL has been
determined to 16 µM using equilibrium dialysis [36]. However, since AAL has multiple
binding sites for fucose this value represents an average Kd-value for all binding sites.
The crystal structure of AAL was first determined by Wimmerova et al. [42] who showed that
each AAL monomer could bind five fucose residues. In a similar study performed by
Fujihashi et al. it was shown that three fucose molecules per monomer at sites 1, 2 and 4 were
present after dialysis against a fucose-free buffer. This indicated that the binding affinity for
fucose was higher at sites 1, 2 and 4 than at sites 3 and 5.
Each binding sites in AAL includes arginine, glutamic acid and tryptophan as characteristic
amino acids. All of these amino acids form hydrogen bonds with fucose [42]. In site 2 seven
hydrogen bonds are formed between fucose and AAL, in site 4 six hydrogen bonds are
formed whereas 5 hydrogen bonds are formed between AAL and fucose in site 1, 3 and 5
[42]. This probably gives site 2 the highest affinity for fucose followed by site 4.
AAL has been shown to bind to fucosylated oligosaccharides irrespective of linkage (Fucα12, Fucα1-3, Fucα1-4 or Fucα1-6). However, affinity for oligosaccharides with α1-6 linked
fucose is somewhat higher than for the other linkages. Therefore AAL has mostly been used
for detection and fractionation of oligosaccharides containing α1-6 linked fucose [38, 44, 49].
Studies on whether there is specificity differences between the individual binding sites in
AAL have not been performed.
Fucose
Fucose is a 6-deoxyhexose (Fig. 1) and a common component of many N- and O-linked
glycans and glycolipids produced by mammalian cells. In contrast to other hexoses in
mammals fucose has an L-configuration [50]. Fucosyltransferases (FucTs) are responsible for
13
the catalysis of fucose transfer from donor guanosine-diphosphate fucose to various acceptor
molecules including oligosaccharides, glycoproteins, and glycolipids [51]. Fucose normally
exists as a terminal modification but it can also serve as substrate for the addition of other
monosaccharides [52]. Fucose is commonly linked α-1,2 to Gal or α-1,3, α-1,4 or α-1,6 to
GlcNAc. Fucose linked α-1,6 is commonly found to the reducing terminal GlcNAc on
complex type N-linked glycans [51]. Fucose linked α-1,2 to Gal forms the H antigen which is
the precursor structure for formation of the A and B blood group antigens. Fucosylated
structures are involved in a variety of important biological and pathological processes in
eukaryotic organisms including tissue development, angiogenesis, fertilization, cell adhesion,
inflammation, and tumour metastasis [51]. In prokaryotic organisms fucosylation appears to
be less common. Still fucosylated structures are present in prokaryotes and these structures
has been suggested to be involved in molecular mimicry, adhesion, colonization, and
modulating the host immune response [51].
Figure 1. Structure of the monosaccharide α-L-fucose.
Fucosylation as a disease marker
Changes in fucosylation patterns are commonly associated with many disease states such as
inflammation, liver disease and cancer [28, 39-41, 53]. Using AAL, Rydén et al. have shown
that the fucosylation of the plasma protein α1-acid glycoprotein (AGP) also known as
orosomucoid, is significantly higher in patients with rheumatoid arthritis and liver cirrhosis,
and that the degree of fucosylation correlated with the severity of the disease [39]. Therefore
they proposed that the fucosylation of AGP could be useful in diagnosis of inflammatory
disease and in early detection of liver cirrhosis.
Changes in glycosylation are also common in all types of tumor cells [28, 53]. On these cells
N-linked oligosaccharides often become more branched and contain more fucose compared
with non-malignant cells [53, 54]. The changed glycosylation pattern of tumor cells can give
rise to expression of specific tumor-associated oligosaccharide structures [53, 54]. These
structures could potentially be used as cancer markers, and some of them may have a
prognostic value.
14
Aims of the thesis
Glycosylated plasma proteins and/or cell surface proteins have in various studies shown
disease-specific or disease-state specific changes in glycosylation pattern. Thus there is an
obvious need for glycosylation analysis in routine diagnostics. Therefore it will be important
to obtain reliable carbohydrate binders with high specificity. In our lab Aleuria aurantia lectin
(AAL) has been used in glycodiagnostic assays to determine fucosylation of AGP during
different pathological conditions. This lectin has been shown to bind to fucose and
fucosylated oligosaccharides with relatively high affinity and broad specificity compared to
other lectins. The overall aim of this study was to use molecular biology techniques to
produce recombinant AAL and recombinant AAL forms and to study their binding
characteristics in detail. More specific aims were:
Paper I
The aim of this study was to produce a recombinant form of AAL lectin (rAAL). This rAAL
would then be purified in the absence of free fucose. The aim was further to characterize the
purified rAAL and determine its affinity to fucose and different fucose-containing
oligosaccharides.
Paper II
A problem of using multivalent lectins for diagnostic purposes is that differences in affinity
between different binding sites can reduce the accuracy and performance of the assay.
Another problem is the fact that binding specificity can differ between different binding sites
in the lectin giving unwanted cross reactivity. To solve some of these problems, it is of
interest to produce monovalent lectins with more specific carbohydrate binding properties.
The aim of this study was to create a functional monomeric form of AAL and to create a
monovalent form of AAL by expressing only one binding site of AAL and to investigate their
binding characteristics.
15
Methods
Construction of plasmids
His-tagged rAAL – Paper I
Initially we obtained the plasmid pGEM-AAL as a generous gift from Dr. Yoshiho Nagata at
the Chiba University, Japan. Instead of using this plasmid to produce AAL we incorporated
the AAL-gene in the pET28b vector, which will produce rAAL with a Histidine tag (His-tag).
The His-tag enabled purification of rAAL without adding free fucose during purification. The
plasmid pGEM-AAL was digested with the two restriction enzymes NdeI and XhoI. The gene
product (AAL = 939 bp) was purified from an agarose gel and then added to pET28b that had
previously been digested with NdeI and XhoI, purified and dephosphorylated. This mixture
was ligated with T4 DNA-ligase and then transformed into E.coli. Transformed E.coli was
cultured on agar plates and bacterial colonies were screened for pET28b containing AAL
insert.
His-tagged monomeric AAL – Paper II
To create monomeric AAL a single amino acid was changed in AAL. Site directed
mutagenesis of pET-28b-AAL was used to change the codon TAC to CGC in the AAL coding
region giving a switch in amino acid from Tyr6 to Arg6.
His-tagged monovalent AAL – Paper II
Initially we tried to use modified primers and PCR to amplify a part of the AAL gene
comprising only binding site 2 (S2-AAL). The primers were modified to create an NdeI
cleaving site in the 5’-end of S2-AAL and an XhoI cleaving site in the 3’-end. This PCR
product was then cleaved with NdeI and XhoI and added to cleaved (NdeI and XhoI), purified
and dephosphorylated pET28b. The mixture was ligated with T4 DNA-ligase and then
transformed into E.coli. However this approach was not successful most probably due to
failure of the restriction enzymes to cleave the PCR products prior to ligation. Our second
approach was to use site directed mutagenesis of pET-28b-AAL (Fig. 2 A). An NdeI cleavage
site was inserted at nucleotide position 150 in the AAL coding sequence. This was followed by
the insertion of a stop codon site at nucleotide position 480 in the AAL coding sequence (Fig.
2 B). The resulting plasmid was cleaved with NdeI to remove the 5’-segment of AAL
corresponding to nucleotides 1-150 and then ligated (Fig. 2 C). The coding region of the
resulting plasmid corresponds to nucleotides 150-480 (encoding S50 – G160) in the original
AAL sequence. When expressed, we observed that approximately 50 % of S2-AAL was found
in the supernatant and 50 % in inclusion bodies. We succeeded to purify S2-AAL from
inclusion bodies using denaturation with guanidinium chloride and renaturation on the Ni2+column. However, due to ease of purification only S2-AAL purified from supernatant was
used in Paper II.
16
Figure 2. Modification of pET28b-AAL to express S2-AAL. a Shows were AAL is located in pET28b-AAL
and restriction enzyme cleavage sites. b Indicates insertion of new NdeI and stop codon sites. c The final
pET28b-S2-AAL plasmid.
Circular dichroism – Paper I and II
Circular dichroism (CD) spectroscopy is a form of light absorption spectroscopy that is
sensitive to the secondary structure of proteins. Changes in a CD signal reflect changes in the
protein structure. The definition of CD is that it measures the difference in absorbance of left
circularly polarized light (LCPL) and right circularly polarized light (RCPL) for a specific
substance: CD = Abs (LCPL) – Abs (RCPL) [55-58]. The fact that CD measures circularly
polarized light is the major difference compared to other absorption spectroscopic methods
that measures the absorption of isotropic light. A molecule that has the ability to absorb RCPL
and LCPL must be structurally asymmetric and exhibit absorbance [56, 57]. Such a molecule
is said to be a CD active molecule. For instance proteins are CD active molecules because all
amino acids except glycine contain a chiral carbon, and are thus asymmetric [56].
CD is reported in units of absorbance or ellipticity and both can be normalized for the molar
concentration of the sample [55, 56]. The CD spectra in Paper I and Paper II have the unit of
molar ellipticity ([θ]) which is ellipticity that has been corrected for the concentration of the
sample. The unit of molar ellipticity is (deg cm2/dmol). To calculate molar ellipticity, the
sample concentration (g/L), molecular weight (g/mol) and the cell path length (cm) must be
known [55, 56]. If the sample is a protein, the mean residual weight (average molecular
weight of the amino acids it contains) is used instead of the molecular weight.
The CD spectrum is divided into a far-ultraviolet (far-UV) region (180-260 nm) and a nearultraviolet (near-UV) region (260-320 nm). In a far-UV CD spectrum the amide groups in the
polypeptide backbone are the major CD absorbents. Since these amide groups will have
different orientations depending on the secondary structure of the protein a CD spectra in the
far-UV region will give information on the presence of secondary structures such as α-helix,
β-sheet and random coil. These different structural elements give roughly the following
characteristic CD spectra: α-helical proteins have negative bands at 222 nm and 208 nm and a
positive band at 193 nm, proteins with well-defined antiparallel β-sheets have negative bands
at 218 nm and positive bands at 195 nm, whereas disordered proteins (random coil) have very
17
low ellipticity above 210 nm and negative bands near 195 nm (Fig. 3) [59]. But it is not only
the polypeptide backbone that affects the far-UV CD spectrum. It has been shown that
tryptophans with its aromatic side chains also affect the CD spectrum in almost the entire
wavelength region (180-310 nm) [60]. Since tryptophans contribute to the CD spectra also in
the far-UV region it will make predictions of the amount of various types of secondary
structure more difficult for proteins containing several tryptophans. One monomer of AAL
contains 16 tryptophans. Therefore, we did not predict the secondary structures of the studied
AAL forms on basis of the obtained CD spectra. Instead we used the data obtained from CDmeasurements to determine whether the different AAL forms were correctly folded and
whether there were differences in secondary structure between the different AAL forms.
Figure 3. Theoretical CD spectra for different structural elements in
the far-ultraviolet region.
In the near-UV region the side chains of aromatic amino acids (phenylalanine, tyrosine and
tryptophan) influence the CD spectrum the most. To give a CD signal these side chains have
to be situated in an asymmetric environment. Unfolded proteins do not give CD signals in the
near-UV region. Therefore a CD spectrum in this region gives information about the tertiary
structure around the aromatic side chains.
Requirements for obtaining a far-UV CD spectrum are a sample volume of 30-200 µL with a
protein concentration of 100 µg/mL to 1.5 mg/mL depending on type of cuvette used [55].
Signals in the near-UV region of the CD spectrum are generally more than an order of
magnitude weaker than in the far-UV region of the CD spectrum. Therefore these
measurements require more material with a sample volume of about 1 mL and a protein
concentration between 0.25-2 mg/mL [55]. The buffers in which the proteins are dissolved
during a CD experiment can not contain any molecules that have a high absorbance in this
region of the spectra (180-320 nm). Such molecules are for example dithiothreitol (DTT) and
imidazole [61]. In our experiments we never experienced any problems with lack of material
or high background absorbance from the buffer.
18
Both near- and far-UV CD spectra provide a fingerprint of the examined protein. CD
spectroscopy can therefore be used to study protein stability or folding intermediates
following different treatments such as heating or treatment with denaturating agents.
Enzyme Linked Lectin Assay – Paper I and II
One important application area for lectins is to detect specific carbohydrate structures on
glycoproteins. This can be done by different techniques such as lectin affinity
chromatography [62, 63], crossed affino-immunoelectrophoresis [64, 65], lectin
immunosensors [66], lectin blotting [65, 67, 68] and enzyme linked lectin assay (ELLA, lectin
ELISA) [69-72]. AAL has previously been used in different ELLA assays [69, 40]. The
advantages with the ELLA method are that it gives much information within a short time and
it is easy to perform and repeat. ELLA does not give any affinity values but it can be used for
specificity studies. We used ELLA to investigate if our lectin forms had retained ability to
bind a number of fucosylated oligosaccharide structures and if there was differences in
binding specificity. All ELLA experiments were performed in the same way. Initially a
microtiter plate was coated with synthesized BSA-oligosaccharide conjugates, containing on
average 15 oligosaccharide chains per conjugate. Then the wells were blocked with BSA
followed by addition of biotinylated lectin forms. The bound biotinylated lectin was
visualized by adding ExtrAvidin and phosphatase substrate and absorbance was measured at
405 nm. These experiments gave information about binding capacity and binding specificity
for the different lectin forms compared to the native lectin.
Surface Plasmon Resonance – Paper I and II
A sensor that uses biomolecules to detect a certain analyte is called a biosensor. One of the
most used types of biosensors is the affinity biosensor. An affinity biosensor can measure the
interaction between different biomolecules, for example between an antibody and an antigen
or between a lectin and a carbohydrate [73]. A common method to detect the interaction is
Surface Plasmon Resonance (SPR). The underlying physical principles of SPR are complex.
Simplified, the SPR method will give information on how the refractive index is changed near
the surface of a sensor chip. When an analyte binds to the sensor surface the refractive index
is changed and this will be measured and the interaction can be studied [74]. In 1990 a
commercial instrument using this technology was released by Biacore AB (Uppsala, Sweden)
and it was named BIACORE® [74]. There are several SPR-based systems but it is the
different BIACORE systems that are the most widely used [72]. An advantage with this
method is that data describing affinity as well as kinetics is obtained in a straightforward
manner [75].
Affinity biosensors based on SPR technique provide a powerful tool for the analysis of
protein carbohydrate interactions [76-83]. In addition to affinity measurements SPR can also
be used to determine the enthalpy, stoichiometry, kinetics and activation energy of an
interaction. A major advantage of SPR over other techniques such as isothermal titration
19
calorimetry (ITC) is that less amounts of protein are required.
We used BIACORE® 2000 and a CM5 sensor chip to study the affinity between the obtained
AAL forms and fucose or different fucosylated oligosaccharides. The sensor surface where
the interaction between ligand and analyte occurs is shaped as a small flow cell, 20-60 nL in
volume, through which an aqueous solution (running buffer) passes under continuous flow. In
order to detect an interaction, the ligand (in our case the lectin of interest) is immobilized onto
the sensor surface. It is important that the activity of the ligand is not disrupted during the
immobilization. After immobilization of the ligand, the analyte is injected (mono- and
oligosaccharides in our experiments) in aqueous solution (sample buffer) through the flow
cell under continuous flow. As the analyte binds to the ligand on the sensor surface the
refractive index will change due to the accumulation of molecules on the sensor surface. This
change in refractive index is measured in real time and the result is plotted as response or
resonance units (RUs) versus time in a sensorgram (Fig. 4). After the desired association time
a solution without the analyte, preferably the sample buffer, is injected over the sensor surface
resulting in dissociation of the analyte (Fig. 4). As the analyte dissociates from the ligand a
decrease in SPR signal is observed (Fig. 4). Finally the sensor surface is regenerated and a
new binding cycle can be performed (Fig. 4). From these data association rate constant (on
rate, kon) and dissociation rate constant (off rate, koff) as well as equilibrium constants can be
calculated. To obtain correct data in an SPR experiment it is important to measure the
background response. This is done to correct for unspecific binding of analyte or to correct for
any difference in the refractive index between the running- and sample buffers. The
background response must be subtracted from the sensorgram to obtain the actual binding
response. The background response is recorded by injecting the analyte through a different
reference flow cell on the same sensor chip, which has no ligand or an irrelevant ligand
immobilized to the sensor surface (blank). In Papers I and II we used a blank flow cell with
inactivated surface to record the background response. The advantage with this type of blank
reference flow cell is that it is easy to create and it gives a good estimation of the background
response. The disadvantage is that if there is any unspecific binding between the analyte and
the ligand this will not be corrected for. On the other hand there is no risk of overestimating
unspecific binding when this type of blank sensor surface is used. The ultimate reference
surface, if a lectin carbohydrate interaction is considered, would be a lectin that has the exact
same characteristics as the normal lectin but lacks the ability to bind the carbohydrate of
interest.
20
Figure 4. Sensorgram obtained from a typical SPR experiment. 1 A baseline is
recorded as running buffer flows over the immobilized ligand. 2 Injection of
analyte resulting in an increased response due to formation of analyte/ligand
complex. 3 Steady state equilibrium. 4 Injection of running buffer without
analyte. As a consequence the analyte disassociates from the ligand. 5 High flow
washing to remove remaining complexes.
The sensor surface of an ordinary BIACORE chip is coated with a 100-200 nm thick layer of
carboxymethylated dextran matrix. Commonly 2000 – 10 000 pg protein/mm2 (corresponding
to 2000 – 10 000 RU) is immobilized on the sensor surface when an experiment is performed.
It is important that the level of immobilized ligand is not too high. Firstly, with too high
ligand density the rate at which the surface binds the analyte may exceed the rate at which the
analyte can be delivered to the surface (mass-transport). In this situation the mass-transport
becomes the rate-limiting step and the kon will therefore be slower than the true kon. Secondly,
when the analyte dissociates from the ligand it can rebind to an unoccupied ligand before
diffusing out of the matrix and being washed from the flow cell. The consequence of this is
that the measured koff is slower than the true koff . These two kinetic artefacts (mass-transport
limitations and re-binding) are well known drawbacks for all surface-binding techniques.
Due to the fact that SPR measures the mass of material binding to the sensor surface, small
analytes as oligosaccharides will give low responses. Therefore we had to use quite high
surface concentrations of ligand in our experiments (about 10 000 RU) to receive a detectable
signal. Because of this high ligand concentration at the sensor surface there is a risk of masstransport limitations and/or re-binding events in our experiments. In our experiment this was
partly controlled for by increasing the flow rates of the injected analyte. These experiments
showed that kon was independent of flow rates, indicating that there were no mass-transport
effects. We used steady state equilibrium analysis to evaluate the affinities from our
experiments and not kinetic analysis which is more sensitive to mass-transport limitations and
re-binding.
In our BIACORE experiments we used oligosaccharides as analytes and different AAL forms
as immobilized ligands. This approach has previously been successful for affinity
21
measurements between lectins and oligosaccharides [76-78]. The Biacore experiments could
also have been done with the different oligosaccharides immobilized on the sensor surfaces
and the AAL forms in the running buffer [79-80]. However, since some of the AAL forms are
multivalent this would result in an avidity effect that would complicate the interpretation of
the SPR data [81].
Tryptophan fluorescence – Paper I
The fluorescence of a folded protein is a mixture of the fluorescence from individual aromatic
residues. Most of the intrinsic fluorescence emissions of a folded protein are due to excitation
of tryptophan residues, with some contribution from tyrosine and phenylalanine.
Tryptophan fluorescence has shown to be a powerful tool to study the interaction between a
carbohydrate ligand and its protein binder, and has therefore been used to study the affinity
and specificity among different lectins [84-86]. Most proteins that bind carbohydrates contain
tryptophan residues [84]. The innate fluorescence of these tryptophans can be affected upon
binding due to changes in the microenvironment surrounding the tryptophans [87]. The
change in tryptophan fluorescence can be observed as increased or decreased intensity, with
or without change of the peak position of maximum intensity [87]. The tryptophans are
usually excitated at 295 nm and the emitted tryptophan fluorescence can vary in wave length
from 300-350 nm [84]. This variation depends on the environment surrounding the tryptophan
residues in the protein. Even if a change in tryptophan fluorescence is not a consequence of a
direct participation of tryptophans in the binding, the observed fluorescence change can still
be used to calculate binding constants [87]. Carbohydrates do not affect the tryptophan
fluorescence since they are transparent to ultraviolet light [87, 88].
We performed tryptophan fluorescence experiments by adding increased concentration of
fucosylated oligosaccharides to nAAL and rAAL in solution. The tryptophans were excited at
295 nm and a peak value of the emission spectra was obtained around 339 nm. For rAAL a
decreased fluorescence intensity was observed after addition of fucosylated oligosaccharides
(except NA2F) but no change was observed for nAAL. A non-fucosylated oligosaccharide
was used to correct for background fluorescence. The obtained data from the experiments
with rAAL was used to obtain binding curves and approximate Kd-values.
Hemagglutination – Paper I and II
Hemagglutination (HA) is a specific form of agglutination that involves erythrocytes. It is an
assay based on the ability of multivalent binders to cross link their antigen on erythrocytes
and thereby make them agglutinate. HA is commonly used to determine the ABO blood group
of blood donors and transfusion recipients. This is done with antibodies directed to the
different blood group antigens on erythrocytes. The wider term agglutination can involve
different types of cells and molecules and it can be either direct agglutination, caused by a
primary binder, or indirect agglutination caused by a secondary binder directed against the
first.
22
HA and hemagglutination inhibition (HAI) has been extensively used to study the specificity
of newly discovered lectins [89, 90]. AAL has previously been shown to agglutinate human
erythrocytes in a reaction that could be specifically inhibited by fucose and fucosylated
oligosaccharides [91, 36]. We used HA and hemagglutination inhibition (HAI) to investigate
if the produced AAL forms had retained or altered binding activity. Human type O
erythrocytes was used in our experiments and for the HA assay serial dilutions of the AAL
forms were added. A positive reaction was determined as the formation of visual aggregates.
If two lectins are compared and one gives agglutination at a lower concentration it can be
concluded that this lectin has binding sites with higher affinity available for agglutination than
the other lectin. For the HAI assay the AAL forms were preincubated with a serial dilution of
fucose or fucosylated oligosaccharides and then assayed for hemagglutinating activity. In this
way the minimum inhibition concentration (MIC) of oligosaccharide required to completely
inhibit the agglutination reaction could be determined.
Analytical Ultracentrifugation – Paper II
An accurate way to determine the molecular weight, binding constants and the hydrodynamic
and thermodynamic properties of macromolecules is through use of analytical
ultracentrifugation (AUC) [92, 93]. The technique was developed by Theodor Svedberg in
1923 [94]. An AUC application consists of a centrifugal force with the simultaneous real-time
observation of the resulting redistribution of the macromolecules during the centrifugation
[93]. The distribution of the macromolecules is monitored in real time through an optical
detection system, normally ultraviolet light absorption (280 nm). The acquired data gives the
spatial gradients that result from the application of the centrifugal field and their progression
with time. There are several advantages with this method. For example it usually does not
require any labeling or other chemical modification of the macromolecule/macromolecules to
be analyzed, and the material can often be reused after an experiment [93]. The main strength
with AUC is that the macromolecules of interest can be characterized in their native state in a
biologically relevant solution. Furthermore, the solution conditions can be varied to give more
information. Finally since the experiments are performed in free solution, there are few
complications due to interactions with matrices or surfaces [95]. In our studies we used AUC
to determine the molecular mass of the produced lectin forms and to determine whether they
were present in monomeric or multimeric forms. We used SPR-analysis (Biacore) for affinity
experiments due to the much higher throughput using this technique.
There are two main AUC methods used to study macromolecules namely sedimentation
equilibrium (SE) and sedimentation velocity (SV) [92, 93, 95]. In a SV experiment high
centrifugal force (typically ≥40 000 rpm) is used and the time course of the sedimentation
process is analyzed [95]. On the other hand in a SE experiment low centrifugal force is used
that permits the diffusion to balance the sedimentation such that a time-invariant equilibrium
gradient can be observed [95]. Both methods can be used to determine the molecular mass but
we used SE since this method is less sensitive to the shape of the macromolecules and
therefore reveals any heterogeneities of the sample [95]. After an equilibrium sedimentation
experiment is performed the buoyant molecular weight of the analyte is received. To convert
23
this value to an absolute molecular weight the partial specific volume for the analyte has to be
determined. This can be done experimentally, however we used a simpler technique that
computes the partial specific volume from a weight average of the partial specific volumes of
the amino acid residues in AAL. Calculated partial specific volumes of amino acids are in
agreement with the experimentally observed for most proteins [96].
24
Results and Discussion
Finding a high affinity binding site in recombinant AAL – Paper I
Usually monovalet interaction between lectins and carbohydrates are of weak affinity
(millimolar). But there have been reports of fungal and bacterial lectins that bind
carbohydrates with affinities in the nanomolar range [97]. Previous studies has shown that
AAL bind to different fucosylated oligosaccharides in the micromolar range [42]. We had
previously purified AAL from the Aleuria aurantia mushroom in our lab [69]. But in the
present study a recombinant form of AAL (rAAL) was produced to obtain larger quantities of
the lectin for further studies and characterization. We expressed AAL with a His-tag that
made purification fast, reliable and avoided the use of free fucose during purification steps.
Purified rAAL was analyzed by fucose-agarose affinity chromatography and was found to
have retained ability to bind fucose. Therefore the binding characteristics were studied in
more detail.
Secondary structure was examined using CD and it was obvious that the rAAL secondary
structure was very similar to nAAL. The near-UV spectra of rAAL gave information
indicating an ordered structure around the tryptophans in the protein. Since most of the
tryptophans are located in the binding sites these results indicated an ordered structure of the
binding sites.
Purified rAAL was biotinylated and used in ELLA experiments. The binding preference to
oligosaccharides with α1-2, α1-3 and α1-4 linked fucose was retained in rAAL. However,
interestingly, rAAL showed three to five times higher binding level compared to nAAL. To
check whether the His-tag was responsible for some of the gained ability to bind in the ELLA
experiment the His-tag was removed using thrombin digestion. Thrombin digested rAAL was
compared to untreated rAAL in an ELLA experiment. No difference in binding to different
fucosylated oligosaccharides was observed indicating that the His-tag did not contribute to the
binding between rAAL and fucosylated carbohydrates.
Biacore experiments was then performed to determine the affinity for rAAL towards different
fucosylated oligosaccharides. The Biacore sensorgrams obtained showed a remarkable
difference between rAAL and nAAL in affinity for all oligosaccharides except NA2F (for
oligosaccharide structures se Table I). When fucose, LNF I, LNF II, LNF III and LDFT was
injected over a sensor surface with immobilized rAAL distinct association and dissociation
phases was observed in the obtained sensorgrams. However when the same oligosaccharides
were injected over a sensor surface with immobilized nAAL pulse shaped sensorgrams with
extremely rapid association and dissociation phases were observed. This indicated that rAAL
has a higher affinity for these oligosaccharides compared to nAAL. When NA2F was injected
both rAAL and nAAL showed pulse shaped sensorgrams indicating similar affinity. When the
Kd-values were calculated it became clear that rAAL had two different Kd-values for fucose,
LNF I, LNF II, LNF III and LDFT. One in the micromolar range and one in the nanomolar
range. The Kd-values in the micromolar range was similar for both nAAL and rAAL whereas
the nanomolar Kd-value was only observed in rAAL.
25
The next step in the characterization was to use tryptophan fluorescence. These experiments
showed a dose-dependent quenching of the tryptophan fluorescence when LNF I, LNF II and
LNF III were added to a rAAL solution. However when NA2F was added no change in
tryptophan fluorescence was observed. When the same oligosaccharides were added to a
nAAL solution no change in its tryptophan fluorescence was observed regardless of
oligosaccharides. These results showed that rAAL has a high affinity binding site capable of
binding LNF I, LNF II and LNF III with affinities in the range 5-50 nM (Kd) that is blocked in
nAAL.
One interesting observation from the Biacore experiments was that if fucose was injected
prior to the injection of a fucosylated oligosaccharide no high affinity Kd-values in rAAL was
observed. But the fucose injection did not affect the low affinity Kd-values in rAAL or nAAL.
Thus, suggesting that there is a high affinity site in rAAL that is blocked in nAAL by free
fucose which is added to nAAL during its purification. This was supported by previous
observations [43]. We also performed a sugar analysis of both nAAL and rAAL. This
revealed that nAAL contained fucose, as expected, whereas rAAL did not which supports the
hypothesis of a blocked site in nAAL.
From the characterization we did not obtain direct evidence for which of the binding sites in
AAL that was the high affinity binding site. Based on the crystal structure of the different
binding sites in AAL we proposed that site 2 is the most probable high affinity site [42, 43].
However this remains to be confirmed.
26
27
Construction of a monomeric and a monovalent fucose binder – Paper II
To study the role of AAL dimerization for its binding properties a monomeric form of AAL
was produced. Previous crystallization studies had indicated amino acids that are important
for dimer formation of AAL [42, 43]. Site-directed mutagenesis was used to exchange these
amino acids. Two different constructs were made and one of the constructs resulted in a
monomeric form of AAL. This monomeric form showed to be about 32 kDa with a similar
secondary structure as rAAL. From the different binding experiments it could be concluded
that mAAL had the same binding preference as rAAL but with slightly decreased affinities.
The property that differed most between mAAL and rAAL was the ability to hemagglutinate
erythrocytes. A twenty fold higher concentration of mAAL was needed to agglutinate
erythrocytes compared to rAAL. This result was not surprising, since dimerization of AAL is
expected to enhance its possibility to create cross-linking of erythrocytes. The MIC-values for
fucose and LDFT were lower for mAAL compared to rAAL meaning that less concentration
of these ligands was needed to inhibit mAAL’s ability to produce hemagglutination compared
to rAAL. An explanation for this is that more sites involved in the agglutination are available
on rAAL compared to mAAL. Thus, the difference in MIC-values is not connected to
decreased affinities in individual sites of mAAL, but is rather connected with a decrease in
avidity.
Since a monomeric variant of AAL could be created that retained high affinity binding
towards fucosylated oligosaccharides, this encouraged us to make a monovalent fucose binder
composed of only binding site 2 (S2-AAL). Based on the results in Paper I and previous
crystallization studies of AAL [42, 43], site 2 was the strongest candidate to be the high
affinity binding site. This site was therefore the target for making a monovalent fucose binder.
A plasmid encoding a part of the AAL sequence comprising site 2 (S50 – G160) was
constructed and S2-AAL could be produced and purified. This protein had a greater tendency
to form inclusion bodies compared to rAAL and mAAL. We could purify S2-AAL from both
supernatant and inclusion bodies. S2-AAL showed the same binding characteristics in the
Biacore experiments independent of purification method. However, since enough material for
characterization studies was obtained from the bacterial supernatant this was used for further
characterization.
The secondary structure of S2-AAL was similar to that of mAAL, rAAL and nAAL. An AAL
monomer is built up of six β-blades and the S2-AAL was constructed to contain two of these
β-blades. The analysis of the secondary structure indicated that these β-blades were indeed
formed in S2-AAL.
ELLA experiments showed that S2-AAL had retained ability to bind fucose and fucosecontaining oligosaccharides. However one major difference between S2-AAL and the other
AAL forms was that S2-AAL did not bind to fucosylated and sialylated oligosaccharides such
as SLea and SLex.
28
Next S2-AAL was examined using Biacore and these experiments also showed that S2-AAL
had lost its ability to bind to sialylated-fuco oligosaccharides. But for the other fucosylated
oligosaccharides affinities could be calculated and S2-AAL showed as expected only one
binding site. These affinities were 2-5 times lower than the weak affinity binding in rAAL,
but the order of affinities to the different oligosaccharides were the same for S2-AAL and
rAAL with Kd-values in the order of LNF I >LNF II > LNF III > NA2F, LDFT.
From the obtained data it is impossible to conclude whether site 2 actually represents the
“high-affinity binding site” identified in Paper I. If site 2 is assumed to be the “high affinity
binding site” it means that S2-AAL has lost much of its affinity, but gained the ability to bind
NA2F. This is possible since (even though the secondary structure of S2-AAL is similar to
that of rAAL) there may be minor differences in the positions of different amino acids
affecting affinity and specificity. Another explanation is that site 2 is not the “high affinity
binding site”. Preliminary experiments to produce an AAL form comprising only site 4, which
is another potential candidate, have been performed.
Importantly, we have been able to create a recombinant monovalent carbohydrate binder with
affinity in the low micromolar range towards fucosylated oligosaccharides (NA2F and
LDFT).
29
Tack!
Förutom mig själv så har min excellente handledare Peter Påhlsson varit oumbärlig under
arbetet med denna licentiatavhandling. Med en så bra handledare blir forskarutbildningen
både lärorik och rolig. Jag kan bara ta av mig hatten och buga för dessa år, som varit de
lärorikaste i mitt liv hitintills. Du har hela tiden varit positiv och stöttat mig även då
omständigheterna förändrats under tiden. Jag hoppas verkligen vi kommer att ha fortsatt
kontinuerlig kontakt i framtiden.
Ett speciellt tack vill jag rikta till Ingvar Rydé för allt du gjort för mig trots att du egentligen
inte behövt. Med kontakter som dig går alla situation att lösa. Jag har verkligen haft tur att du
och Peter tagit hand om mig och hjälpt mig på alla sätt. En dag hoppas jag kunna återgälda i
alla fall något av all hjälp jag fått av er.
Andra personer som varit viktiga för min forskarutbildning är min biträdande handledare
Lena Tibell och Bengt-Harald ”Nalle” Jonsson. Tack för all hjälp med mitt arbete och för att
ni visat en sådan uppskattning för det jag gjort. Ni har verkligen gjort mig motiverad!
Som samtida forskarstudent i gruppen har Louise Levander varit mycket betydelsefull för
mig. Tack för all hjälp jag har fått och för att du stått ut men alla mina små egenheter. Jag
saknar all den skönsång som vi förgyllde dagarna med.
I övrigt vill jag tacka alla er som gjort min tid i Linköping väl värd att minnas. Jag tänker då
på alla mina underbara vänner och arbetskamraterna på cellbiologen plan 10.
Självklart vill jag tacka min familj för er outsinliga tro på mig och för att ni alltid stöttar mig.
Slutligen vill jag tacka min sambo Annica för att du gör mina dagar så underbara och för att
du stått ut med alla turer runt min licentiatavhandling. Du är helt enkelt bäst älskling och jag
hoppas vi får ett långt liv tillsammans.
30
References
1
Ohlson, S., Jungar, C., Strandh, M. and Mandenius, C. (2000) Continuous weakaffinity immunosensing. Trends Biotechnol. 18, 49-52
2
Sharon N, Lis H. (2002) How proteins bind carbohydrates: lessons from legume
lectins. J Agric Food Chem. 50, 6586-6591
3
McEver, R., Moore, K. and Cummings, R. (1995) Leukocyte trafficking
mediated by selectin-carbohydrate interactions. J Biol Chem. 270, 11025-11028
4
Nicholson, M., Barclay, N., Singer, M., Rosen, S. and van der Merve, A. (1998)
Affinity and kinetic analysis of L-selectin (CD62L) binding to glycosylationdependent cell-adhesion molecule-1. J Biol Chem. 273, 763-770
5
Stillmark, H. (1888) Uber Ricin, ein giftiges Ferment aus den Samen von
Ricinus communis L. und einigen anderen Euphorbiaceen. Thesis. Dorpat 1888.
6
Olsnes, S. and Kozlov, J. (2001) Ricin. Toxicon. 39, 1723-1728
7
Sharon, N. and Lis, H. (2004) History of lectins: from hemagglutinins to
biological recognition molecules. Glycobiology 14, 53-62
8
Sharon, N. and Lis, H. (1995) Lectins--proteins with a sweet tooth: functions in
cell recognition. Essays Biochem 30, 59-75
9
Nilsson, C. (2003) Lectins – proteins that interpret the sugar code. Anal Chem
15, 348-353
10
Hooper, L. and Gordon, J. (2001) Glycans as legislators of host-microbial
interactions: spanning the spectrum from symbiosis to pathogenicity.
Glycobiology. 11, 1-10
11
Gulliot, J. and Konska, G. (1997) Lectins in higher fungi. Biochem Syst Ecol.
25, 203-230
12
Gabius, H. (1997) Animal lectins. Eur J Biochem. 243, 543-576
13
Loris, R. (2002) Principles of structures of animal and plant lectins. Biochim
Biophys Acta 19, 198-208
14
De Hoff, P., Brill, L. and Hirsch, A. (2009) Plant lectins: the ties that bind in
root symbiosis and plant defense. Mol Genet Genomics. 282, 1-15
15
Lis, H. and Sharon, N. (1998) Lectins: Carbohydrate-Specific Proteins That
Mediate Cellular Recognition. Chem Rev. 98, 637-674
31
16
Barondes, S., Cooper, D., Gitt, M. and Leffler, H. (1994) Galectins. Structure
and function of a large family of animal lectins. J. Biol. Chem. 269, 2080720810
17
Drickamer, K. (1988) Two distinct classes of carbohydrate-recognition domains
in animal lectins. J. Biol. Chem. 263, 9557-9560
18
Powell, L. and Varki, A. (1995) I-type lectins. J. Biol. Chem. 270, 14243-14246
19
Magnani J. (2004) The discovery, biology, and drug development of sialyl Lea
and sialyl Lex. Arch Biochem Biophys. 426, 122-131
20
Herndon, M., Stipp, C. and Lander, A. (1999) Interactions of neural
glycosaminoglycans and proteoglycans with protein ligands: assessment of
selectivity, heterogeneity and the participation of core proteins in binding.
Glycobiology. 9, 143-155 (ev rew 2008)
21
Sharon, N. and Lis, H. (1990) Legume lectins. FASEB J. 4, 3198-3208
22
Fülöp, V. and Jones D. (1999) Beta propellers: structural rigidity and functional
diversity. Curr Opin Struct Biol. 9, 715-721
23
Streicher, H. and Sharon, N. (2003) Recombinant plant lectins and their
mutants. Methods Enzymol. 363, 47-77
24
Etzler, M., Allan. H. and Kisaillus, E. (1992) Plant lectins molecular biology,
synthesis and function. Glycoconjugates. 16, 521-539
25
Barondes, S. (1981) Lectins: their multiple endogenous cellular functions. Annu
Rev Biochem. 50, 207-231
26
Lis, H. and Sharon, N. (1986) Lectins as molecules and as tools. Annu. Rev.
Biochem. 55, 33-67
27
Wu, A., Lisowska, E., Duk, M. and Yang, Z. (2008) Lectins as tools in
glycoconjugate research. Glycoconj J. 27, DOI 10.1007/s10719-008-9119-7
28
Miyoshi, E. and Nakano, M. (2008) Fukosylated haptoglobin is a novel marker
for pancreatic cancer: Detailed analysis of oligosaccharide structures.
Proteomics. 8, 3257-3262
29
Hirabayashi, J. (2008) Concept, strategy and realization of lectin-based glycans
profiling. J Biochem. 144, 139-147
30
Gornik, O. and Lauc, G. (2007) Enzyme linked lectin assay (ELLA) for direct
analysis of transferrin sialylation in serum samples. Clin Biochem. 40, 718-723
32
31
Smetana, K. and André, S. (2008) Mammalian lectin as tool in glycochemistry
and histochemistry with relevance for diagnostic procedure. Methods Mol Biol.
418, 171-186
32
Zou, Y., Broughton, D., Bicker K., Thompson, P. and Lavigne, J. (2007) Peptide
borono lectins (PBLs): a new tool for glycomics and cancer diagnostics.
Chembiochem. 17, 2048-51.
33
Sharma, P., Gupta, B., Farhat Basir, S., Rani Das, H. and Nahar, P. (2008) Rapid
and sensitive detection of autoantibody in rheumatoid arthritis patients by heatmediated ELISA. Clin Biochem. 41, 97-102
34
Pusztai, A., Bardocz, S. and Ewen, S. (2008) Uses of plant lectins in bioscience
and biomedicine. Front Biosci. 13, 1130-1140
35
Liljeblad, M., Lundblad, A. and Pahlsson, P. (2002) Analysis of glycoproteins in
cell culture supernatants using a lectin immunosensor technique. Biosensors &
bioelectronics. 17, 883-891
36
Kochibe, N. and Furukawa, K. (1980) Purification and properties of a novel
fucose-specific hemagglutinin of Aleuria aurantia. Biochem. 19, 2841-2846
37
Amano, K., Takase, M., Ando, A. and Nagata, Y. (2003) Production of
functional lectin in pichia pastoris directed by cloned cDNA from Aleuria
aurantia. Biosci Biotechnol Biochem. 67, 2277-2279.
38
Yamashita, K., Kochibe, N., Ohkura, T., Ueda, I. and Kobata, A. (1985)
Fractionation of L-fucose-containing oligosaccharides on immobilized Aleuria
aurantia lectin. J Biol Chem. 260, 4688-4693
39
Ryden, I., Pahlsson, P. and Lindgren, S. (2002) Diagnostic accuracy of alpha(1)acid glycoprotein fucosylation for liver cirrhosis in patients undergoing hepatic
biopsy. Clinical chemistry. 48, 2195-2201
40
Ryden, I., Pahlsson, P., Lundblad, A. and Skogh, T. (2002) Fucosylation of
alpha1-acid glycoprotein (orosomucoid) compared with traditional biochemical
markers of inflammation in recent onset rheumatoid arthritis. Clinica chimica
acta; international journal of clinical chemistry. 317, 221-229
41
Hashimoto, S., Asao, T., Takahashi, J., Yagihashi, Y., Nishimura, T., Saniabadi,
A. R., Poland, D. C., van Dijk, W., Kuwano, H., Kochibe, N. and Yazawa, S.
(2004) alpha1-acid glycoprotein fucosylation as a marker of carcinoma
progression and prognosis. Cancer. 101, 2825-2836
42
Wimmerova, M., Mitchell, E., Sanchez, J., Gautier, C. and Imberty, A. (2003)
Crystal structure of fungal lectin: six-bladed beta-propeller fold and novel
fucose recognition mode for Aleuria aurantia lectin. J Biol Chem. 278, 2705927067
33
43
Fujihashi, M., Peapus, D., Kamiya, N., Nagata, Y. and Miki, K. (2003) Crystal
structure of fucose-specific Aleuria aurantia binding ligands at three of its five
recognition sites. Biochem. 42, 11093-11099
44
Fukomori, F., Takeuchi, N., Hagiwara, T., Ohbayashi, H., Endo, T., Kochibe,
N., Nagata, Y. and Kobata, A. (1990) Primary structure of a fucose-specific
lectin obtained from a mushroom, Aleuria aurantia. J Biochem. 107, 190-196
45
Wu, A., Wu, J., Singh, T., Liu, J. and Herp, A. (2004) Lectinochemical studies
on the affinity of Anguilla anguilla agglutinin for mammalian glycotopes. Life
Sci. Jul 16, 1085-1103
46
Mitchell, E., Sabin, C., Snajdrová, L., Pokorná, M., Perret, S., Gautier, C., Hofr,
C., Gilboa-Garber, N., Koca, J., Wimmerová, M. and Imberty, A. (2005) High
affinity fucose binding of Pseudomonas aeruginosa lectin PA-IIL: 1.0 A
resolution crystal structure of the complex combined with thermodynamics and
computational chemistry approaches. Proteins. 15, 735-746
47
Sudakevitz, D., Imberty, A. and Gilboa-Garber N. (2002) Production, properties
and specificity of a new bacterial L-fucose- and D-arabinose-binding lectin of
the plant aggressive pathogen Ralstonia solanacearum, and its comparison to
related plant and microbial lectins. J Biochem. 132, 353-358
48
Kostlánová, N., Mitchell, E., Lortat-Jacob, H., Oscarson, S., Lahmann, M.,
Gilboa-Garber, N., Chambat, G., Wimmerová, M. and Imberty, A. (2005) The
fucose-binding lectin from Ralstonia solanacearum. A new type of betapropeller architecture formed by oligomerization and interacting with fucoside,
fucosyllactose, and plant xyloglucan. J Biol Chem. 30, 27839-27849
49
Matsumura, K., Higashida, K., Hata, Y., Kominami, J., Nakamura-Tsuruta, S.
and Hirabayashi, J. (2009) Comparative analysis of oligosaccharide specificities
of fucose-specific lectins from Aspergillus oryzae and Aleuria aurantia using
frontal affinity chromatography. Anal Biochem. 386, 217-221
50
Becker, D. and Lowe, J. (2003) Fucose: biosynthesis and biological function in
mammals. Glycobiology. 13, 41-53
51
Ma, B., Simala-Grant, J. and Taylor, D. (2006) Fucosylation in prokaryotes and
eukaryotes. Glycobiology. 16, 158-184
52
Moloney, D. and Haltiwanger, R. (1999) The O-linked fucose glycosylation
pathway: identification and characterization of a uridine diphosphoglucose:
fucose-beta1,3-glucosyltransferase activity from Chinese hamster ovary cells.
Glycobiology. 9, 679-687
34
53
Miyoshi, E., Moriwaki, K. and Nakagawa, T. (2008) Biological function of
fucosylation in cancer biology. J Biochem. 143, 725-729
54
Brockhausen, I., Schutzbach, J. and Kuhns, W. (1998) Glycoproteins and their
relationship to human disease. Acta Anta. 161, 36-78
55
Martin, S. and Schilstra, M. (2008) Circular Dichroism and Its Application tothe
Study of Biomolecules. Methods Cell Biol. 84, 263-293
56
Woody RW (1995) Circular dichroism Methods Enzymol. 246, 34-71
57
Kelly, S., Jess, T. and Price, N. (2005) How to study proteins by circular
dichroism. Biochim Biophys Acta. 1751, 119 – 139
58
Ignacio, T. (1995) Optical spectroscopy: General principles and overview.
Methods Enzymol. 246, 13-18
59
Greenfield, N. (2007) Using circular dichroism spectra to estimate protein
secondary structure. Nat Protoc. 6, 2876-2890
60
Freskgård PO, Mårtensson LG, Jonasson P, Jonsson BH, Carlsson U. (1994)
Assignment of the contribution of the tryptophan residues to the circular
dichroism spectrum of human carbonic anhydrase II. Biochemistry. 47,1428114288
61
Rodger, A., Marrington, R. and Roper, D. (2005) Circular dichroism
spectroscopy for the study of protein-ligand interactions. Methods Mol Biol.
305, 343-364
62
Kobata, A. and Endo, T. (1992) Immobilized lectin columns: useful tools for the
fractionation and structural analysis of oligosaccharides. J Chromatogr. 597,
111-122
63
Cummings, R. (1994) Use of lectins in analysis of glycoconjugates. Methods
Enzymol. 230, 66-86.
64
Rydén, I., Skude, G., Lundblad, A. and Påhlsson, P. (1997) Glycosylation of a1acid glycoprotein in inflammatory disease: Analysis by high-pH anion-exchange
chromatography and concanavalin A crossed affinity immunoelectrophoresis.
Glycoconjugate J. 14, 481–488
65
Faye, L. and Salier, J. (1989) Crossed affino-immunoelectrophoresis or affinoblotting with lectins: advantages and limitations for glycoprotein studies.
Electrophoresis. 10, 841-847
66
Liljeblad, M., Rydén, I., Ohlson, S., Lundblad, A. and Påhlsson, P. (2001) A
lectin immunosensor technique for determination of alpha(1)-acid glycoprotein
fucosylation. Anal Biochem. 288, 216-224
35
67
Gravel, P., Golaz, O., Walzer, C., Hochstrasser, D., Turler, H. and Balant, L.
(1994) Analysis of glycoproteins separated by two-dimensional gel
electrophoresis using lectin blotting revealed by chemiluminescence. Anal
Biochem. 221, 66-71
68
Iskratsch, T., Braun, A., Paschinger, K. and Wilson, I. (2009) Specificity
analysis of lectins and antibodies using remodeled glycoproteins. Anal Biochem.
386, 133-146
69
Rydén, I., Lundblad, A. and Påhlsson P. (1999) Lectin ELISA for analysis of
alpha(1)-acid glycoprotein fucosylation in the acute phase response. Clin Chem.
45, 2010-2012
70
Turner, G. (1992) N-glycosylation of serum proteins in disease and its
investigation using lectins. Clin Chim Acta. 208, 149-171
71
Pekelharing, J., Vissers, P., Peters, H. and Leijnse, B. (1987) Lectin-enzyme
immunoassay of transferrin sialovariants using immobilized antitransferrin and
enzyme-labeled galactose-binding lectin from Ricinus communis. Anal
Biochem. 165, 320-326
72
McCoy, J., Varani, J. and Goldstein, J. (1984) Enzyme-linked lectin assay
(ELLA): II. Detection of carbohydrate groups on the surface of unfixed cells.
Exp Cell Res. 151, 96-103
73
Karlsson, R. (2004) SPR for molecular interaction analysis: a review of
emerging application areas. J. Mol. Recognit. 17, 151–161
74
Rich, L. and Myszka, G. (2001) BIACORE J: a new platform for routine
biomolecular interaction analysis. J. Mol. Recognit. 14, 223–228
75
Johnsson, B., Löfås, S. and Lindquist, G. (1991) Immobilization of proteins to a
carboxymethyldextran-modified gold surface for biospecific interaction analysis
in surface plasmon resonance sensors. Anal Biochem. 198, 268-277
76
Kobayashi, Y., Nakamura, H., Sekiguchi, T., Takanami, R., Murata, T., Usui, T.
and Kawagishi, H. (2005) Analysis of the carbohydrate binding specificity of
the mushroom Pleurotus ostreatus lectin by surface plasmon resonance. Anal
Biochem. 336, 87-93
77
Duverger, E., Frison, N., Roche, A. and Monsigny, M. (2003) Carbohydratelectin interactions assessed by surface plasmon resonance. Biochimie. 85, 167179
36
78
Haseley, S., Talaga, P., Kamerling, J. and Vliegenthart, J. (1999)
Characterization of the carbohydrate binding specificity and kinetic parameters
of lectins by using surface plasmon resonance. Anal Biochem. 274, 203-210
79
Sota, H., Lee, R., Lee, Y. and Shinohara, Y. (2003) Quantitative lectincarbohydrate interaction analysis on solid-phase surfaces using biosensor based
on surface plasmon resonance. Methods Enzymol. 362, 330-340
80
Okazaki, I., Hasegawa, Y., Shinohara, Y., Kamasaki, T. and Bhikhabhai, R.
(1995) Determination of the interactions between lectins and glycoproteins by
surface plasmon resonance. J Mol Recognit. 8, 95-9
81
Shinohara, Y., Hasegawa, Y., Kaku, H. and Shibuya, N. (1997) Elucidation of
the mechanism enhancing the avidity of lectin with oligosaccharides on the solid
phase surface. Glycobiology. 7, 1201-1208
82
Karamanska, R., Clarke, J., Blixt, O., Macrae, J., Zhang, J., Crocker, P.,
Laurent, N., Wright, A., Flitsch, S., Russell, D. and Field, R. (2007) Surface
plasmon resonance imaging for real-time, label-free analysis of protein
interactions with carbohydrate microarrays. Glycoconj J. 25, 69-74
83
Strandh, M., Persson, B., Roos, H. and Ohlson, S. (1998) Studies of interactions
with weak affinities and low-molecular-weight compounds using surface
plasmon resonance technology. J Mol Recognit. 11, 188-190
84
Lee, Y. (1997) Fluorescence spectrometry in studies of carbohydrate-protein
interactions. J Biochem. 121, 818-825
85
Sallay, I., Hatakeyama, T. and Yamasaki, N. (1998) Studies on the carbohydrate
binding sites of the hemolytic lectin CEL-III isolated from the marine
invertebrate Cucumaria echinata. Biosci Biotechnol Biochem. 62, 1757-1761
86
Khan, F., Ahmad, A. and Khan, M. (2007) Interaction of Fusarium solani lectin
with monosaccharides and oligosaccharides: a fluorometric study. Photochem
Photobiol. 83, 966-970
87
Jolley, M. and Glaudemans, C. (1974) The determination of binding constants
for binding between carbohydrate ligands and certain proteins. Carbohydr Res.
33, 377-382
88
Glaudemans, C., Miller, C. and Nashed, E. (1997) The measurement of affinity
between ligand and antibody using ligand-induced antibody fluorescence
change. Carbohydr Res. 300, 169-170
37
89
Wang, H., Gao, J. and Ng, T. (2000) A new lectin with highly potent
antihepatoma and antisarcoma activities from the oyster mushroom Pleurotus
ostreatus. Biochem Biophys Res Commun. 275, 810-816
90
Sharma, A., Ng, T., Wong, J. and Lin, P. (2009) Purification and
characterization of a lectin from Phaseolus vulgaris cv. (Anasazi beans). J
Biomed Biotechnol. Article ID 929568, 9 pages, doi:10.1155/2009/929568
91
Debray, H. and Montreuil, J. (1989) Aleuria aurantia agglutinin. A new isolation
procedure and further study of its specificity towards various glycopeptides and
oligosaccharides. Carbohydr. Res. 185, 15–26
92
Lebowitz, J., Lewis, M. and Schuck, P. (2002) Modern analytical
ultracentrifugation in protein science: A tutorial review. Protein Sci. 9, 20672079
93
Schachman, H. (1992) Is there a future for the ultracentrifuge? In Analytical
ultracentrifugation in biochemistry and polymer science. Harding, S., Rowe, A.
and Horton, J. (eds), Royal Society of Chemistry, Cambridge, UK. 3–15
94
Svedberg, T., Pedersen, K. O. (1940) The Ultracentrifuge. Edited by R. H.
Fowler and P. Kapitza. Oxford University Press, London
95
Cole, J., Lary, J., Moody, T. and Laue, T. (2008) Analytical ultracentrifugation:
sedimentation velocity and sedimentation equilibrium. Methods Cell Biol. 84,
143-179
96
McMeekin, T. and Marshall, K. (1952) Specific volumes of proteins and the
relationship to their amino acid contents. Science. 116, 142-143
97
Imberty, A., Mitchell, E. and Wimmerová, M. (2005) Structural basis of highaffinity glycan recognition by bacterial and fungal lectins. Curr Opin Struct
Biol. 15, 525-534
38