!CTA 5NIVERSITATIS !

GRICULTURAE 3UECIAE

Acta Universitatis Agriculturae Sueciae

Doctoral Thesis No. 2021:65 • Novel potato starch • Xue Zhao

Doctoral Thesis No. 2021:65

A simplified method was developed for determining the internal molecular structure of Doctoral Thesis No. 2021:65
whole starch without prior amylopectin isolation. Applying the method, varied internal Faculty of Natural Resources and Agricultural Sciences
structural parameters of potato starch were obtained from different genetic backgrounds.
Various internal structure parameters were found to affect the thermal properties of potato
starch. A dense structure of building blocks led to higher gelatinisation temperatures and Novel potato starch
enthalpy. Retrogradation was found to be favoured by more large building blocks and
New structure and beneficial qualities
many short internal chains.

Xue Zhao, the author of this thesis, conducted her PhD studies in Food Science at the Xue Zhao
Department of Molecular Sciences, SLU, Uppsala. She received her MSc and BSc degrees
in Animal Science from SLU, Sweden and Northwest A & F University, China, respectively.

Acta Universitatis Agriculturae Sueciae presents doctoral theses from the Swedish
University of Agricultural Sciences (SLU).

SLU generates knowledge for the sustainable use of biological natural resources. Research,
education, extension, as well as environmental monitoring and assessment are used to
achieve this goal.

Online publication of thesis summary: http://pub.epsilon.slu.se/

ISSN 1652-6880
ISBN (print version) 978-91-7760-807-3
ISBN (electronic version) 978-91-7760-808-0
 Novel potato starch

New structure and beneficial qualities

Xue Zhao
Faculty of Natural Resources and Agricultural Sciences
Department of Molecular Sciences
Uppsala

DOCTORAL THESIS
Uppsala 2021
Acta Universitatis agriculturae Sueciae
2021:65

Cover: Microscopic images of potato starch granules

(photo: X. Zhao)

ISSN 1652-6880
ISBN (print version) 978-91-7760-807-3
ISBN (electronic version) 978-91-7760-808-0
© 2021 Xue Zhao, Swedish University of Agricultural Sciences
Uppsala
Print: SLU Service/Repro, Uppsala 2021
Novel potato starch

Abstract
This thesis presents a simplified method for determining the internal molecular
structure of whole starch without prior amylopectin isolation. The structure of potato
and barley whole starches, the thermal properties of starch from potato lines with
different genetic backgrounds and the relationship between molecular structure and
functional properties of starch were examined in the thesis.
The internal B-chain distribution and building block composition of amylopectin
were characterised effectively by degrading starch into β-limit dextrins (β-LDs), α-
limit dextrins (α-LDs) and building blocks. Great variations in internal structure
were observed for starches from different plant sources and genetic backgrounds.
The general composition of intermediate and large building blocks and the
proportion of fingerprint B-chains (Bfp-chains), in size order, were determined for
starches with decreasing amylose content.
Thermal properties (gelatinisation and retrogradation) of potato starches were
investigated using differential scanning calorimetry. Amylopectin lines with a high
degree of mutations in multiple genes showed a broader gelatinisation temperature
range and lower enthalpy of gelatinisation and retrogradation. Various internal
structure parameters were found to affect the thermal properties of potato starch. A
dense structure of building blocks led to higher gelatinisation temperatures and
enthalpy, while retrogradation was found to be favoured by more large building
blocks and many short internal chains.
The high-amylose potato line T-2012 was shown to have higher levels of
resistant starch and dietary fibre than the parental variety after cooking. The level of
resistant starch increased further after one extra day of cold storage. T-2012 had a
very large fraction of long outer amylopectin chains and intermediate-sized inner
amylopectin chains, and more intermediate and large building blocks, than the
parental potato. The unique amylopectin structure of T-2012 starch favoured
formation of recrystallised amylopectin that did not split as easily as ordinary potato
starch and was resistant to enzyme digestion.

Keywords: potato, starch, amylopectin internal structure, building blocks, resistant

starch, dietary fibre, gelatinisation, retrogradation

Author’s address: Xue Zhao, Swedish University of Agricultural Sciences,

Department of Molecular Sciences, P.O. Box 7015, 750 07 Uppsala, Sweden
Ny potatisstärkelse

Sammanfattning
Avhandlingen beskriver en förenklad metod för att bestämma den interna
molekylära strukturen hos stärkelse utan tidigare isolering av amylopektin.
Strukturen hos potatis- och kornstärkelse, de termiska egenskaperna hos potatislinjer
med olika genetisk bakgrund och sambandet mellan molekylär struktur och
stärkelsens funktionella egenskaper studerades i avhandlingen.
Den interna fördelningen av B-kedjornas längd samt förgreningarnas densitet hos
amylopektinet karakteriserades effektivt genom att bryta ned stärkelse till β-LD och
α-LD, dvs den interna strukturens byggstenar. Stora variationer i stärkelsens interna
struktur erhölls från olika växtkällor och olika genetiska bakgrunder. Det visade sig
att andelen halvstora och stora byggstenar samt andelen sk fingeravtryck-kedjor
minskade för stärkelserna med minskande amyloshalt.
Termiska egenskaper (gelatinisering och retrogradering) av potatisstärkelser
undersöktes med hjälp av differentiell scanningskalorimetri. Amylopektinlinjerna
med en hög grad av mutationer i flera gener gelatiniserade i ett bredare
temperaturintervall och hade en lägre entalpi för såväl gelatinisering som
retrogradering. Olika detaljer i den interna strukturen av potatisstärkelse visade sig
påverka stärkelsens termiska egenskaper. Tät struktur hos byggstenarna ledde till
högre gelatiniseringstemperaturer och entalpi. Retrogradering visade sig gynnas av
fler stora byggstenar och många korta interna kedjor.
I avhandlingen visade sig att högamylospotatislinjen T-2012 hade högre nivåer
av resistent stärkelse (RS) och kostfiber efter tillagning, jämfört med
ursprungssorten. RS-nivån ökade ytterligare efter ett extra dygn i kylförvaring. T-
2012 hade en mycket stor fraktion långa yttre kedjor och medelstora inre kedjor i
amylopektinet, och mellanstora såväl som stora byggstenar jämfört med
ursprungssorten. Denna unika amylopektinstruktur gynnade bildningen av
retrograderat, dvs kristalliserat, amylopektin efter tillagning. Den kristalliserade
stärkelsen bryts inte ned lika lätt som vanlig potatisstärkelse och blir delvis resistent
mot enzymatisk hydrolys.

Nyckelord: potatis, stärkelse, inre struktur av amylopektin, byggstenar, resistent

stärkelse, kostfiber, gelatinisering, retrogradering

Författarens adress: Xue Zhao, Sveriges lantbruksuniversitet, Institutionen för

molekylära vetenskaper, P.O. Box 7015, 750 07 Uppsala,
 Dedication

To my beloved wonderful family

“The more I learn, the more I realise how much I don’t know.”

Albert Einstein
Contents

List of publications........................................................................... 9

Abbreviations ................................................................................ 11

1. Introduction .......................................................................... 13
1.1 Starch composition ..................................................................... 14
1.2 Starch granule............................................................................. 14
1.3 Starch.......................................................................................... 16
1.3.1 Amylose .......................................................................... 16
1.3.2 Amylopectin .................................................................... 16
1.3.3 External molecular structure of amylopectin................... 17
1.3.4 Internal molecular structure of amylopectin .................... 17
1.3.5 Starch structure characterisation .................................... 19
1.4 Genetically modified potato starches .......................................... 19
1.5 Nutritional properties................................................................... 20
1.5.1 Dietary fibre .................................................................... 21
1.5.2 Resistant starch.............................................................. 21
1.6 Thermal properties...................................................................... 21
1.7 Starch of interest......................................................................... 23

2. Aims..................................................................................... 25

3. Materials and methods......................................................... 27

3.1 Potato varieties and developed lines .......................................... 27
3.2 Potato tubers and starch preparation.......................................... 29
3.3 Nutritional properties of potato tubers......................................... 29
3.4 Amylose content of starch samples ............................................ 30
3.5 Characterisation of starch granules ............................................ 31
3.6 Production of β-limit dextrin, α-limit dextrin and building blocks . 31
3.7 Chain and building block distribution of starches........................ 33
3.8 Thermal properties of potato starches ........................................ 35
 3.9 Statistical methods...................................................................... 35

4. Results and discussion ........................................................ 37

4.1 Amylose content ......................................................................... 37
4.2 Starch granular structure ............................................................ 38
4.3 Starch molecular structure .......................................................... 41
4.3.1 Starch chain distribution (Paper III) ................................ 41
4.3.2 Amylopectin chain distribution (Paper III) ....................... 42
4.3.3 Starch internal chain distribution .................................... 44
4.3.4 Amylopectin internal chain distribution ........................... 47
4.3.5 Building block structure .................................................. 48
4.4 Nutritional properties of potato line T-2012................................. 52
4.5 Influence of starch structure on thermal properties..................... 54

5. Conclusions ......................................................................... 59

6. Future research ................................................................... 61

References.................................................................................... 63

Popular science summary ............................................................. 71

Populärvetenskaplig sammanfattning ............................................ 73

Acknowledgements ....................................................................... 75
List of publications
This thesis is based on the work contained in the following papers, referred
to by Roman numerals in the text:

I. Zhao, X.*, Andersson, M. & Andersson, R. (2018). Resistant

starch and other dietary fibre components in tubers from a high-
amylose potato. Food Chemistry 251, 58-63.

II. Zhao, X.*, Andersson, M. & Andersson, R. (2021). A simplified

method of determining the internal structure of amylopectin from
barley starch without amylopectin isolation. Carbohydrate
Polymers 255, 117503.

III. Zhao, X.#, Jayarathna, S.#, Turesson, H., Fält, A-S., Nestor, G.,
González, M.N., Olsson, N., Beganovic, M., Hofvander, P.,
Andersson, R. & Andersson, M.* (2021). Amylose starch with no
detectable branching developed through DNA-free CRISPR-Cas9
mediated mutagenesis of two starch branching enzymes in potato.
Scientific Reports 11, 4311.

IV. Zhao, X., Hofvander, P., Andersson, M., & Andersson, R.

Amylopectin internal structure and thermal properties of potato
starches with a wide variation in amylose content (manuscript).

Papers I-III are reproduced with the permission of the publishers.

*Corresponding author.
#
First authorship shared.

9
The contribution of Xue Zhao to the papers included in this thesis was as
follows:

I. Designed the experiments together with the supervisors,

conducted the experimental work, collected and analysed the
data, and wrote and revised the manuscript.

II. Planned the study together with the supervisor and conducted all
the experiments. Was responsible for data collection and
evaluation. Wrote and revised the manuscript.

III. Participated in designing and conducting the characterisation of

starch structure and evaluating the results. Had responsibility for
writing and revising the manuscript.

IV. Designed the experiments, carried out the experimental work and
performed the statistical analysis. Was responsible for writing the
manuscript.

10
10
Abbreviations

ANOVA Analysis of variance

DF Dietary fibre
DM Dry matter
DMSO Dimethylsulphoxide
DP Degree of polymerisation
DSC Differential scanning calorimetry
GBSS Granule-bound starch synthase
GLM General linear model
HPAEC High performance anion exchange chromatography
HPSEC High performance size exclusion chromatography
LD Limit dextrin
MALLS Multiple-angle laser light scattering
PAD Pulsed amperometric detection
PLS Partial least squares (regression analysis)
RS Resistant starch
SBE Starch branching enzyme
SS Starch synthase
UDMSO Urea-dimethylsulphoxide

11
1. Introduction
Potato (Solanum tuberosum) is one of the most important, nutritious, high-
yielding and starch-rich staple crops worldwide (Birch et al., 2012). Potatoes
comprise a major part of the human diet in the Nordic countries and potato
consumption has also increased greatly elsewhere in the world in recent
decades (Eriksson et al., 2016).
Potato is also grown as a crop for starch production. Starch is an
inexpensive raw material and is widely used for food and non-food
applications. Starch is the most abundant food energy source for the human
population and animals globally. However, native starch has some
drawbacks when it is used in manufactured foods. In order to overcome these
disadvantages, much effort has been devoted to physical and/or chemical
modification of native starch in food processing. However, the chemical and
physical modifications devised to date are money-, time- and energy-
consuming, as well as labour- and chemicals-intensive.
Work is underway to make starch production and downstream processing
more sustainable, and one part of the solution could be to use modern
breeding technologies to develop new starch qualities in a crop. Therefore,
increased understanding of the effect of genetic modification on molecular
structure and functional properties of starch is vitally important. The ability
to tailor starch at the genetic level with desired functional properties for
multiple applications, without the need for further chemical or physical
modification, would provide an environmentally and economically friendly
and sustainable approach for developing novel desirable starches in the near
future.

13
1.1 Starch composition
Starch is primarily extracted from plant tubers, cereal grains and legume
seeds. In potato, starch comprises about 15-20% of the tuber by fresh weight
and over 80% of the dry matter content (Bertoft & Blennow, 2016).
Therefore, starch is considered a major factor affecting the functional
properties of potato products developed in the food industry.
Starch consists of two main components, amylose and amylopectin, which
are built up of a number of glucose monomers. The molecular weight of
amylose and amylopectin is in the order of 105-106 Da and 107-109 Da,
respectively. In native potato starch, the ratio of amylose to amylopectin is
approximately 1:4 (Zeeman et al., 2010). However, certain lines known as
waxy potato contain very little or no amylose (Vamadevan & Bertoft, 2015)
and some high-amylose potato lines have an amylose content of up to 80%
(Menzel et al., 2015).

1.2 Starch granule

Starch molecules are organised into highly ordered granules (Figure 1), the
morphology and size of which differ between plant species. The diameter of
potato starch granules is within the range 10-100 μm, and potato starches
generally have larger and smoother granules than cereal starches (Jane et al.,
1994). The starch granules from native potato have alternating amorphous
and semi-crystalline rings, and characteristic Maltese crosses can be seen in
polarised light under the microscope (Jenkins et al., 1993).
The organisation of the alternating rings in potato starch granules is still
not well known. However, it is generally believed that amylose is primarily
located in the amorphous parts of the granules, since the majority of amylose
is in the amorphous state in granules. Amylopectin is mainly responsible for
the architecture of the semi-crystalline rings, although components of both
amylose and amylopectin can be found in the amorphous and semi-
crystalline rings inside the granules (Vamadevan & Bertoft, 2015). The semi-
crystalline rings of starch granules comprise stacks of alternating amorphous
and crystalline lamellae with a general repeat distance of ~9 nm (Jenkins et
al., 1993).

14
14
Figure 1. Schematic diagram of starch granule architecture. Source: adapted from
Carbohydrate Polymers 57 (2004), 211-224.

The external part of amylopectin is the part of chain segments from the
non-reducing end to their outermost branch points (Figure 2). These external
amylopectin chains form double helices that make up the crystalline lamellae
(Perez & Bertoft, 2010). The internal part of amylopectin is the part of chain
segments from the outmost branches to the reducing end (Figure 2). This part
contains most of the branching points and is involved in the conformation of
the amorphous lamellae of starch granules (Perez & Bertoft, 2010).
According to the building block backbone model (Bertoft, 2004), the
backbone of amylopectin is located here. The stacks of the lamellae are built
up of layers of amylopectin molecules (Bertoft, 2013).
Depending on the organisation of the double helices in starch granules,
three types of X-ray diffraction patterns are displayed and are used to classify
starch into A- B- or C-type (Buléon et al., 1998; Imberty et al., 1991). Tuber
and root starches generally show a B-type pattern, while the A-type pattern
is found in cereal starch (Buléon et al., 1998). The C-type pattern, which is
a mixture of B- and A-type crystallites, is found in legume starch (Buléon et
al., 1998). The crystals in A-type starch are more densely packed than those
in B-type starch, and less water is contained in A-type starch (Imberty et al.,
1991). Complexes of crystalline amylose-lipid show a V-type pattern
(Buléon et al., 1998).

15
1.3 Starch
Starch is one of the most complex materials in nature, with characteristic
structures that differ between granular populations and within single
granules. Knowledge of starch molecular structure is of vital importance in
order to further biosynthesis of starch and understand its structure-property
relationships.

1.3.1 Amylose
Amylose is essentially a very long linear molecule consisting of α-(1,4)-
linked D-glucosyl chains with degree of polymerisation (DP) in the range of
2000-5000 residues (Hoover, 2001). A minor proportion of branches can also
be found in the amylose molecule (Zhu et al., 2013).

1.3.2 Amylopectin
Amylopectin is a large and highly branched molecule with thousands of α-
(1,4)-linked D-glucosyl chains connected to each other through α-D-(1,6)-
branches. These chains are divided into three types, i.e. A-, B- and C-chains
(Peat et al., 1952). One amylopectin molecule contains only one C-chain
with a free reducing end and the C-chain carries B-chains and A-chains. B-
chains can carry both other B-chains and A-chains, while A-chains do not
carry other chains. The chain-length distribution of starch is primarily
represented by the chain-length distribution of amylopectin after
debranching by the enzymes pullulanase and isoamylase. The chain-length
distribution is an important characteristic of starch molecular structure.
Amylopectin in a native potato starch has an average amylopectin chain
length of DP35, with a peak at DP13 (Menzel et al., 2015). Similar results
for native potato amylopectin have been reported in earlier studies (Jane et
al., 1999; Koch et al., 1998).
The backbone model assumes a variable and flexible arrangement of the
amylopectin chains (Bertoft, 2004). In the model, clusters are oriented
perpendicular to a backbone and external segments of clustered chains form
double helices (Figure 2). This model, which indicates the crystalline and
amorphous lamellae inside starch granules, offers an alternative solution to
explore how amylopectin structure influences starch properties and to better
understand the biosynthesis of starch. Within the model, the long chains of
amylopectin form a backbone where building blocks are bound (Bertoft,

16
16
2013). In cereal starches extensive branches connect to the backbone,
whereas in tuber starches, like potato, the backbone is probably also bound
with a few, shorter branches (Bertoft et al., 2012a, 2012b).

Figure 2. The building block backbone model of amylopectin structure. External chains
are shown in red and internal chains are shown in black. The black part is the φ, β-limit
dextrin (φ, β-LD). Source: structure adapted from Carbohydrate Polymers 57 (2004),
211-224.

1.3.3 External molecular structure of amylopectin

The external part of amylopectin is the part of chain segments from the non-
reducing end to their outermost branch points (Figure 2). These external
amylopectin chains include all A-chains and the external parts of the B-
chains. They form double helices that constitute the crystalline lamellae of
starch granules (Perez & Bertoft, 2010).

1.3.4 Internal molecular structure of amylopectin

The internal part of amylopectin is the part from outmost branches to the
reducing end. This part is involved in the conformation of the amorphous
lamellae of starch granules (Bertoft, 2017). In recent years, multiple studies
have revealed the importance of amylopectin internal structure in
determining physicochemical properties of starch (Zhu, 2018).

Internal chain-length distribution

The internal chain-length distribution of starch is mainly the internal B-chain
distribution of amylopectin and it can be determined by debranching β-limit
dextrins (β-LDs) or φ, β-LDs with pullulanase and isoamylase. However, the

17
internal chain-length distributions in the amylose fraction may also make a
slightly different contribution, because of the increased amylose content
when de-branched β-LDs are studied in whole starch samples.
In order to study the internal B chains, starch scientists have divided them
into different categories with different naming systems and varying ranges
of DP, depending on different botanical sources, starch genotypes and
properties (Bertoft, 2004; Källman et al., 2015; Zhu, 2018). Therefore, it is
very important to carefully define the categories of internal B-chains using
additional information, such as DP range or elution interval, so as to avoid
confusion over the actual molecular structure referred to in different studies.
Based on differences in the composition of internal B-chains, the
amylopectins can be summarised into four groups (Bertoft et al., 2008).
Amylopectins in group one have the highest amount of short internal B-
chains and the lowest amount of long internal B-chains, while those in group
four have the lowest amount of short internal B-chains but the highest
amount of long internal B-chains. The second and third groups of
amylopectins are structurally intermediate between the first and fourth
groups. Starches containing group one amylopectins have an A-type
polymorph, whereas those containing group four amylopectins have a B-type
polymorph (Bertoft et al., 2008).

Building blocks
The amylopectin branches are concentrated in the form of building blocks in
the internal part of amylopectin. Building blocks are conventionally obtained
by partial α-amylolysis and successive extensive hydrolysis with the
enzymes β-amylase and phosphorylase (Bertoft et al., 2011a, 2011b).
Building blocks, as one of the basic internal structural units, provide
information on aspects of starch structure and have a major impact on the
physicochemical properties of starch (Källman et al., 2015). Building blocks
can be categorised into five groups based on differences in their size, from
the most common smallest building blocks carrying two chains per block to
the largest carrying an average of about 10-12 chains per block (Bertoft et
al., 2012a, 2012b). Building blocks have tight branching patterns and the
distance between branching points is no longer than DP 3 (Bertoft et al.,
2012b).
It is estimated that only 1-2% of total branches in whole starch may
originate from the amylose fraction (Zhu et al., 2013). Thus, the branched

18
18
amylose fractions in normal starches might not interfere with the results
when studying the composition of building blocks of whole starch.

1.3.5 Starch structure characterisation

Most of the conventional methods used to date for characterising
amylopectin structure involve many time-consuming and labour-intensive
steps, including amylopectin isolation and partial α-amylolysis of
amylopectin into domains, clusters and building blocks at the end (Bertoft et
al., 2011a, 2011b, 2012a, 2012b; Källman et al., 2013). These steps in total
require a relatively large sample amount to obtain enough material for final
structure characterisation at the level of building blocks (Lemos et al., 2019;
Zhu et al., 2013). Furthermore, it has been shown that D-amylolysis of E-LD
or amylopectin (i.e. the order of enzyme addition of E-amylase and D-
amylase) gives practically the same result (Bertoft, 1989). Zhu et al. (2013)
made an attempt to reduce the workload in analysing amylopectin internal
structure in maize starch without separating amylopectin. However, the
whole starch sample was still hydrolysed in two steps by partial D-
amylolysis, with clusters produced in the first step and then hydrolysed again
in the second step with D-amylase and E-amylase to produce building blocks.
Therefore, it is very important to have a good method for effective
characterisation of starch internal structure in order to better explore the
relationships between the complex structure and properties of starch
components. The internal structure of whole starch can be studied after
extensive removal of the external chains using an exo-acting enzyme β-
amylase only, or in combination with phosphorylase. The resulting β-LDs or
φ, β-LDs contain all intact branch points, while all A-chains and the external
parts of B-chains are degraded into maltotriosyl or maltosyl stubs (Bertoft,
2004).
Examining the structural features of whole starch is important, since it
can provide an alternative viewpoint and a broader picture for better
understanding the relationships of structure-synthesis and structure-function.

1.4 Genetically modified potato starches

Starch biosynthesis involves many enzymes, including granule-bound starch
synthase (GBSS), several types of starch synthases (SS1, SS2, SS3, SS4, SS5
and SS6) and starch branching enzymes (SBE1 and SBE2) (Sarka &

19
Dvoracek, 2017). GBSS is the only starch synthase isoform in amylose
synthesis, while SSs and SBEs act in amylopectin biosynthesis. SSs are
involved in glucan chain extensions by transferring ADP-glucose residues to
the growing glucan chains at the non-reducing end. SS1 catalyses the
formation of chains of amylopectin with DP 8-12 from chains with DP 6-7.
SS2 catalyses the synthesis of amylopectin chains with DP of about 6-10 to
form chains with DP 12-25. SS3 catalyses the synthesis of chains with DP
25-35 or longer. SS4 is suggested to regulate the initiation of starch granules
and SS5 is a novel starch synthase isoform found in cereals (Sarka &
Dvoracek, 2017). SS5 and SS6 are also found in potato, but they have not
yet been thoroughly characterised (Van Harsselaar et al., 2017). The enzyme
SBE1 is involved in the formation of branches leading to B-chains, while
SBE2 is involved in the biosynthesis of A-chains. Both SBE1 and SBE2
synthesise α-D-(1,6)- linkages that form the amylopectin branches (Sarka &
Dvoracek, 2017).
There are generally no major differences in amylose content between
potato cultivars and the ratio of amylose to amylopectin cannot be changed
through conventional cross-breeding, but only through mutational breeding
or genetic tools (Karlsson et al., 2007). Potatoes with an increased amylose
content and/or altered starch molecular structure have been developed
through targeting two SBEs using conventional gene silencing technologies
(Andersson et al., 2006; Schwall et al., 2000) and most recently CRISPR-
Cas9 (Tuncel et al., 2019; Zhao et al., 2021). Likewise, potatoes that
synthesise pure amylopectin starch have been obtained by eliminating GBSS
activity using conventional mutagenesis, antisense, RNA-interference
(RNAi) and genome editing (Andersson et al., 2003; Andersson et al., 2017;
Jacobsen et al., 1991; Visser et al., 1991).

1.5 Nutritional properties

Potato is rich in starch and starch is the main component of the energy in
most human and animal diets. Unfortunately, potato starch is digested
quickly and results in a high blood sugar level when eaten. A potato with
more dietary fibre (DF) or a potato with starch that is partly resistant to
digestion (resistant starch, RS) or digested more slowly, would be beneficial
for human health and would help control glycaemic index and body weight.

20
20
1.5.1 Dietary fibre
The health benefits of DF are well-known, e.g. reducing calories in foods,
decreasing insulin and glucose responses, boosting faecal output and
transition, and promoting the production of short-chain fatty acids and the
growth of beneficial gut bacteria (Brown et al., 2001). Dietary fibre is the
fraction of plant material that is retained in the small intestine after enzymatic
digestion. It contains RS, cellulose, hemicellulose, pectin, gums, mucilages
and lignin. These components can be chemically determined as RS, non-
starch polysaccharide residues and Klason lignin (Theander et al., 1995).

1.5.2 Resistant starch

Resistant starch forms part of DF. It is determined as the sum of starch and
degradation products which cannot be absorbed by healthy people in their
small intestine (Englyst & Cummings, 1984; Englyst et al., 1982). In the
large intestine, resistant starch is fully or partly fermented. Resistant starch
(RS) is classified as five types: I) inaccessible starch, II) resistant granules,
III) retrograded amylose, IV) chemically modified starch and V) the complex
of amylose and lipid (Hasjim et al., 2013). Type II RS consists of resistant
granules which are found in raw potatoes. Cooking and then cooling starches
can transform ordinary starch to RS through the processes of gelatinisation
and retrogradation. When starch is heated in excess water, the granules
absorb water and amylose leaches out from the granules (Figure 3), so the
granules start to swell and lose their molecular order (Srichuwong et al.,
2005; Vamadevan & Bertoft, 2015). When gelatinised starch is cooled,
amylose and amylopectin recrystallise and become ordered again (Figure 3).
The recrystallised amylose is resistant to human α-amylase (type III RS).
The content of RS in food products is reported to be affected mainly by
the amylose content, amylopectin structure and some other factors like starch
granule phenotypes and different temperature treatments (Karlsson et al.,
2007; Leeman et al., 2006; Lehmann & Robin, 2007; Tian et al., 2016).

1.6 Thermal properties

Gelatinisation and retrogradation, two commonly investigated thermal
properties analysed using differential scanning calorimetry (DSC), are very
important for various industrial applications of starch (Srichuwong & Jane,
2007). These thermal properties vary considerably between different plant

21
samples and between varieties/lines within species, depending on the starch
composition and the molecular structure of the starch (Vamadevan & Bertoft,
2015).
When starch is heated in excess water, the granules undergo an order-
disorder phase transition called gelatinisation (Srichuwong et al., 2005;
Vamadevan & Bertoft, 2015). The gelatinisation process involves water
uptake by the amorphous region of starch granules, loss of birefringence with
swelling of the starch granules due to water hydration and loss of crystalline
order as a result of dissociation of double helices and amylose leaching by
taking up heat (Srichuwong et al., 2005; Vamadevan & Bertoft, 2015). When
gelatinised starch is cooled, disorganised amylose and amylopectin
recrystallise and become ordered (Figure 3), a process known as
retrogradation (Srichuwong et al., 2005).

Figure 3. Changes in starch granules during gelatinisation and retrogradation.

It is generally known that retrogradation occurs more easily in starches

containing more amylose or longer glucan chains (long amylopectin external
chains or intermediate material) (Sasaki et al., 2000). The molecular
structure of starch also influences its thermal properties. Amylopectin with
longer internal chains tends to form a more ordered packing structure of
double helices in the granules, which gives higher thermal stability. Longer
internal amylopectin chains have also been found to favour the formation of
recrystallised amylopectin with higher thermal stability, due to the more
ordered structure (Zhu, 2018). Studies on the retrogradation properties of
starch have also found that the long chains (DP >~37) and intermediate-sized
chains (DP 17-20) in amylopectin promote retrogradation, as indicated by
high ΔH measured using DSC (Silverio et al., 2000). Very short chains (DP
<12 or 10) have the opposite effect (Silverio et al., 2000). Amylopectin with
large building blocks has a dense branching structure that apparently

22
22
facilitates recrystallisation (Källman et al., 2015). Previous studies have also
shown that the composition of internal amylopectin chains determines the
polymorph type of starch (Bertoft et al., 2008). Polymorph type is known to
greatly influence the physicochemical properties of starch (Zhu, 2018). Thus,
starch internal structure indirectly influences the properties of starch in terms
of granular structure (Zhu, 2018).

1.7 Starch of interest

Barley starch and potato starch were the main types studied in this thesis.
Barley has many genetic variations with varying molecular structure
available for research, and therefore barley starch is ideal for use in
development of methods for structure characterisation.
Potato is an important crop in Sweden and potato starch was selected for
study in the thesis due to its economic value and research interest. Potato
starch is widely used in industry. In particular, it is preferred in the food
industry because of its neutral flavour and clear pastes due to its low content
of lipids and protein compared with cereal and legume starches. Moreover,
potato starch is known to contain a high level of phosphate. The high
phosphate content of starch granules promotes larger granule size and
increased susceptibility to gelatinisation, and is important in potato starch
rheology (Bergthaller, 2004). Potato starch is also preferred in the paper
industry, owing to good suspensibility of its amylopectin for coating printing
paper and the high molecular weight of its amylose for barrier film in
packaging paper.

23
2. Aims
Much is known about the effect of the molecular structure of starch on its
properties, but there is still a significant gap in current knowledge about
genetically tailored starches, knowledge that is essential when seeking to
alter the genome to reach desirable functionality. The overall aims of this
thesis were to develop and optimise a method for effectively characterising
starch internal structure and to apply the optimised method to a wide range
of potato starches, developed using conventional genetic modification or
targeted mutagenesis, in order to further explore the relationships between
molecular structure and functional properties of starch.

Specific objectives of the work were to:

¾ Investigate nutritional properties of tubers from a high-amylose

potato line T-2012 (Paper I).

¾ Develop a simplified method without amylopectin separation for

effectively determining the internal structure (internal B-chain
distribution and building block composition) of starch (Paper II).

¾ Characterise the molecular structure (chain-length distribution) of

starch and the crystalline type and molecular organisation of starch
granules from “amylose potato lines” (Paper III).

¾ Apply the optimised method to “amylose and amylopectin potato

starches” in analysis of their internal structure (IV).

25
 ¾ Study the effect of genetic modification on starch structure and the
relationship between molecular structure and functional properties
of starch (Papers I, III and IV).

26
26
3. Materials and methods
The structure and properties of potato starch were investigated in 22 different
potato lines and three parental varieties, provided either as tubers or isolated
starch. Methods of great importance for the results presented in the thesis are
explained in detail below. Other methods are presented with only brief
descriptions, but are described in detail in Papers I-IV.

3.1 Potato varieties and developed lines

All potato lines studied in the thesis were developed at the Department of
Plant Breeding, Swedish University of Agricultural Sciences (SLU) (Alnarp,
Sweden). The wild-type potato varieties Dinamo (Paper I), Desiree (Papers
III and IV) and Kuras (Paper IV) were studied as reference.

T-2012 line
In Paper I, the nutritional properties of potato tubers from a high-amylose
line (T-2012) were investigated and the results were compared with those of
the parental potato cultivar Dinamo. The potato line T-2012 was developed
through genetic modification, by down-regulating two starch branching
enzymes genes (SBE1 and SBE2) using RNA-interference (RNAi). This
potato line was also included in the analyses in Papers III and IV.

“Amylose lines”
Thirteen potato lines (82007, 82050, 82079, 104011, 104032, 104001,
104005, 104006, 104016, 104018, 104034, 104010 and 104023) with an
increased amylose content were developed by DNA-free genome editing in
Paper III, by inducing mutations in SBE1 alone or in combination with SBE2,
using the CRISPR-Cas9 technique. Some representative “amylose lines”
were also included in Paper IV (Table 1).

27
“Amylopectin lines”
In Paper IV, eight potato lines (L1, L2, L3, L4, L5, L6, L7 and L8) with a
decreased amylose content were developed through DNA-free genome
editing, by inducing mutations in GBSS alone or in combination with one or
two starch synthase enzymes genes (SS2 and SS3), using the CRISPR-Cas9
technique.
Based on the allelic dosage of mutations in the different genes, the potato
lines analysed in this thesis (except for T-2012) were divided into eight
groups. Detailed information on these groups is provided in Table 1.
Table 1. Origin and mutations of potato lines investigated in this thesis
Line Parent Group Mutations (CRISPR-Cas9) Paper
“Amylose lines” SBE1 SBE2
82007 Desiree Group 1 × ˗ III
82050 Desiree ˵ × ˗ III, IV
82079 Desiree ˵ × ˗ III, IV
104011 Desiree ˵ × ˗ III
104032 Desiree ˵ × ˗ III
104001 Desiree Group 2 × partial III
104005 Desiree ˵ × partial III
104006 Desiree ˵ × partial III, IV
104016 Desiree ˵ × partial III
104018 Desiree ˵ × partial III, IV
104034 Desiree ˵ × partial III
104010 Desiree Group 3 × × III, IV
104023 Desiree ˵ × × III, IV
“Amylopectin lines” GBSS SS2 SS3
L1 Kuras Group 4 partial ˗ ˗ IV
L2 Kuras Group 5 × ˗ × IV
L3 Kuras Group 6 × partial partial IV
L4 L3 Group 7 × × partial IV
L5 L3 ˵ × × partial IV
L6 L1 Group 8 × × × IV
L7 L1 ˵ × × × IV
L8 L1 ˵ × × × IV
“×” = full mutation, “-” = no mutation, “partial” = partial mutation.

28
28
3.2 Potato tubers and starch preparation
Tuber cube preparation
To analyse nutritional properties of a high-amylose potato (Paper I), potato
tubers from line T-2012 and its parental variety Dinamo were peeled and cut
into about 1-cm3 cubes. The cubes were fully cooked using an autoclave for
5 min at 120 °C. After cooking, the potato cubes were stored cold at 4 °C for
0, 1 or 2 days before analysis. Uncooked potato cubes were prepared
similarly, but without cold storage.

Starch isolation
When studying the effect of genetic modification on the molecular structure
of whole starch, it is important to achieve good isolation of starch samples.
In this thesis, starch of the RNAi line T-2012 (Papers I, III and IV), all
“amylose potato lines” (Papers III and IV), and the parental variety Desiree
(Papers III and IV) was isolated from mature tubers (Larsson et al., 1996),
with minor modification of the duration of each buffer washing step and
sedimentation after extraction overnight to ensure that small granules were
retained. Starch of “amylopectin lines” (Paper IV) and the parental varieties
Dinamo (Paper I) and Kuras (Paper IV) was kindly provided by a Swedish
starch producer (former Lyckeby Starch AB, Kristianstad, Sweden).
In Paper II, the starch from 10 barley cultivars and breeding lines was
used for optimisation of a method for effectively characterising starch
internal structure from whole starch samples without prior amylopectin and
amylose separation. The barley starches were the same samples studied using
conventional methods in a previous doctoral thesis produced within the
research group (Källman, 2013).

3.3 Nutritional properties of potato tubers

Resistant starch and total starch
Nutritional properties of potato tubers from line T-2012 were studied in
Paper I. The content of RS was analysed according to AOAC Method
2002.02 (McCleary & Monaghan, 2002), using a K-RSTAR 09/14 kit from
Megazyme (Bray, Ireland). This method is aimed at determining
“physiologically RS”, since the analytical conditions correspond to those
in the human small intestine. Physiologically resistant starch is

29
determined as the part of starch not digested after 16 h incubation with
amyloglucosidase and pancreatic α-amylase at 37 °C, while non-resistant
starch is determined as the part of the starch that has been hydrolysed. The
sum of RS and non-RS is the content of total starch.

Dietary fibre
Dietary fibre constituents were analysed based on AOAC Method 994.13
(Theander et al., 1995), after removal of non-resistant starch by
amyloglucosidase and thermostable α-amylase. The content of DF
constituents was then chemically determined as the sum of sugar residues,
uronic acid residues and Klason lignin. The sugar residues included
rhamnose, arabinose, xylose, mannose, galactose and glucose.

Type III resistant starch

Based on the assumption that the content of type III RS in uncooked tubers
was zero, any increase in glucose residues quantified in DF analysis after
cooking was taken as the type III RS content in cooked potato tubers.

3.4 Amylose content of starch samples

In Paper I, the amylose content in potato starch from the high-amylose line
T-2012 and the parental cultivar Dinamo was measured using a colorimetric
assay, by determining the absorbance of starch-iodine complex according to
a standard curve with defined amylose contents (Morrison & Laignelet,
1983). The amylose content in potato starch from the “amylose lines”, the
parental cultivar Desiree and the RNAi line T-2012 (Paper III) was also
analysed colorimetrically, but based on stabilised starch-iodine complex
formation with trichloroacetic acid (Chrastil, 1987). In brief, the isolated
starch was dissolved in urea-dimethylsulphoxide (UDMSO), and then 0.01
N KI-I2 solution, alone or in combination with 0.5% trichloroacetic acid, was
added for incubation before the absorbance analysis. The amylose content in
Paper III was double checked using an enzymatic assay with an
amylose/amylopectin kit (Megazyme, Bray, Co, Ireland) (Paper III).
In Paper IV, the amylose content of potato starch from the “amylopectin
lines” was carefully measured using high performance size exclusion
chromatography (HPSEC) (Sveriges stärkelseproducenter, Kristianstad,
Sweden).

30
30
3.5 Characterisation of starch granules
Crystalline type
The X-ray pattern of potato starch from line T-2012 (Papers I and III) and
Dinamo (Paper I), all the “amylose lines” and Desiree (Paper III) was
determined using a PANalytical X’Pert alpha1 powder X-ray diffractometer.
After spreading the starch sample onto 1.5 x 2 cm2 silicon wafers, the starch
was scanned between diffraction angle (2θ) 5° and 40°.

Microstructural analysis
The phenotype of starch granules from line T-2012 (Papers I and III),
Dinamo (Paper I) and the representative “amylose lines” and Desiree (Paper
III) was studied using light microscopy after staining the purified starch with
iodine solution (Paper I) and Lugol’s solution (Paper III). A Nikon Eclipse
Ni-U light microscope (Nikon, Tokyo, Japan) was used in Paper I and a Leica
DMLB light microscope (Leica Microsystems, Wetzlar, Germany) in Paper
III. The images of starch granules were documented with a Nikon DS-Fi2-
U3 camera in Paper I and an assembled Leica DFC450C camera in Paper III.
The birefringence of starch granules from line T-2012 and Dinamo (Paper
I), line 104016 (Group 2), line 104010 (Group 3) and Desiree (Paper III) was
studied. A starch/water dispersion (50 mg/mL) was freshly prepared and then
the starch granules were visualised with a 20× objective under polarised
light. In Paper I, the Nikon Eclipse Ni-U light microscope referred to above
was applied. In Paper III, a light microscope (Leica DMLB, Wetzlar,
Germany) equipped with an infinity X-32 digital camera (DeltaPix,
Samourn, Denmark) was used.

3.6 Production of β-limit dextrin, α-limit dextrin and

building blocks
In Paper II, a method for effectively characterising the internal molecular
structure of barley whole starch was developed and optimised. In Paper IV,
this optimised method was applied in determining the internal molecular
structure of potato starch with a wide range of amylose contents, originating
from different gene (SBE1, SBE2, GBSS, SS2, SS3) mutations.
Without prior amylopectin fractionation, the whole starch samples from
barley and potato were solubilised in 90% aqueous dimethylsulphoxide
(DMSO) by heating and stirring. The dissolved starch was then treated with

31
barley β-amylase for 1 h (Figure 4). After β-amylolysis, the whole starch was
degraded into β-LDs and maltose, and the amylose without branching was
removed (Figure 5). Finally, the β-LDs were precipitated by ethanol through
cold storage, followed by centrifugation (Figure 4).

Figure 4. Flowchart of experimental production of β-limit dextrins, α-limit dextrins and

building blocks from whole starch.

The β-LDs obtained were re-dissolved in water with the help of heating,
stirring and scraping. Extensive treatment with α-amylase was then applied
overnight to produce α-limit dextrins (α-LDs) (Figure 4).
Finally, in order to remove all interconnecting chains, the α-LDs were
further hydrolysed into building blocks from the amylopectin fraction by
extensive β-amylolysis overnight (Figure 4). By the end of the process, the
amylose fraction had been degraded into mainly maltotriose, maltose and
glucose, together with a minor proportion of very small building blocks
originating from the branched amylose (Figure 5).

32
32
Figure 5. Schematic illustration of degradation of starch into β-limit dextrin (β-LD), α-
limit dextrin (α-LD) and building blocks by β-amylase and α-amylase (Paper II).

3.7 Chain and building block distribution of starches

The chain-length distribution of whole starch from the potato “amylose
lines” was studied in Paper III. The whole starch samples were dissolved in
UDMSO (0.6 M urea in 90% aqueous DMSO) and then directly debranched
by isoamylase and pullulanase before analysis of the chain-length
distribution, using HPSEC and high performance anion exchange
chromatography (HPAEC).

33
 The internal chain-length distribution of whole starch from the barley
samples (Paper II) and the potato “amylose lines” and “amylopectin lines”
(Paper IV) was studied based on β-LDs. The whole starch samples were first
degraded into β-LDs and then debranched with isoamylase and pullulanase.
The internal chain-length distribution was analysed by HPSEC and HPAEC.
The composition of building blocks from whole barley and potato
starches was investigated in Papers II and IV, respectively, using HPSEC
and HPAEC, after transforming the whole starches into β-LDs, α-LDs and
final building blocks by β-amylase, α-amylase and then β-amylase again (see
Figure 5).
The HPSEC analysis was carried out according to a previous study
(Andersson et al., 2009), with minor modifications. The samples were eluted
at 35 °C using 0.1 M NaNO3 with 0.02% NaN3 at a flow rate of 0.5 mL/min,
through the serially connected guard column and two OHpak SB-802.5 HQ
columns (Shodex, Showa Denko KK, Miniato, Japan). Multiple-angle laser
light scattering (MALLS) and refractive index detectors (Wyatt Technology
Corp., Santa Barbara, CA) were used. The molecular weight was determined
using ASTRA software (Wyatt Technology Corp., Santa Barbara, CA), with
a dn/dc ratio of 0.147 (Andersson et al., 2009).
An HPAEC Series 4500i (Dionex Corp., Sunnyvale, CA, USA) was used
for the analyses. A pulsed amperometric detector (PAD) and a BioLC
gradient pump were incorporated into the HPAEC system. A CarboPac PA-
100 (4×250 mm) analytical column (Dionex, Sunnyvale USA) was used for
the separation and a guard column was also coupled with. A 25 μL portion
of sample was injected at 25 °C and the eluents 0.15 M NaOH (A) and 0.50
M NaOAc+0.15 M NaOH (B) were eluted at a flow rate of 1 mL/min. The
de-branched whole starches (Paper III) and β-LDs (Papers II and IV) were
separated with the following gradient: 0-15 min, 15-28% eluent B; 15-45
min, 28-55% B; 45-75 min, 55-70% B; and 75-80 min 70-15% B. The PAD
response of de-branched whole starches and β-LDs was converted to molar
percentage (M%) and then normalised by total molar weight (Koch et al.,
1998). The building blocks were separated with a shorter gradient: 0-20 min,
15-28% eluent B; 20-35 min, 28-50% B; 35-45 min, 50% B; and 45-50 min
50-15% B. The PAD response of building blocks was directly reported as
relative peak area, without conversion.

34
34
3.8 Thermal properties of potato starches
Gelatinisation and retrogradation properties of potato starches were
investigated by differential scanning calorimetry (DSC) with a modulated
DSC250 (TA Instruments, New Castle, DE, USA). A starch:water ratio of
1:3 was used during gelatinisation and the scanning temperature range was
20-120 °C, with a heating rate of 4 °C/min. Retrogradation of potato starch
was studied in gelatinised starch at a starch:water ratio of 1:2, after cold
storage at 5 °C for three days. The scanning temperature range was -5-120
°C, with a heating rate of 10 °C/min.

3.9 Statistical methods

In Paper I, the content of total starch, resistant starch, other dietary fibres and
its constituents was compared between the RNAi line T-2012 and the parent
variety Dinamo. Statistical analyses were performed using the general linear
model (GLM) procedure in Statistical Analysis System (Version 9.4; SAS
Institute, Cary, NC). One-way analysis of variance (ANOVA) was applied
for uncooked potato tubers and two-way ANOVA was applied for cooked
potato tubers, since storage time was included as another factor. Tukey’s
multiple comparison test was used to determine the statistical significance of
any differences observed.
In Papers II and IV, differences in starch structure based on HPAEC
results between the groups of barley lines and between the groups of potato
lines, and differences in gelatinisation and retrogradation parameters of
potato starch between the groups of potato lines, were analysed by one-way
ANOVA, using Minitab 18 (State College, PA, USA). Tukey comparisons
(Paper II) and Fisher comparisons (Paper IV) were applied in the statistical
analysis.
The relationship between molecular structure and thermal properties of
potato starch (Paper IV) was evaluated by partial least squares (PLS)
regression analysis, using SIMCA 16.0.1 software (Sartorius Stedim Data
Analytics AB, Umeå, Sweden).

35
4. Results and discussion
The structure and properties of potato starch were investigated for 22
different potato lines and three wild-type varieties, provided either as tubers
or isolated starch, in Papers I, III and IV. The effects of genetic modification
on starch structure and on functional properties were evaluated and assessed
in detail in those papers.
A simplified analytical method was developed for effective
characterisation of the internal molecular structure of whole starch without
amylopectin separation (Paper II). By applying this method, structural
information on internal chain-length distribution and building block
composition of whole starch from 14 potato samples was obtained in Paper
IV. The internal starch structure was also correlated to thermal properties in
Paper IV.
The starch granular structure, starch and amylopectin chain distribution
of potato lines with increased amylose content were studied and compared
with those of the parental variety in Paper III.
The nutritional properties of potato tubers of the genetically modified
RNAi line T-2012 and the effect of cooking and storage treatment on these
nutritional properties were studied in Paper I.

4.1 Amylose content

In Paper I, the apparent amylose content of field-grown T-2012 starch and
the parental variety Dinamo was found to be 57% and 41%, respectively,
based on a spectrophotometric iodine-based method. The high-amylose
potato line T-2012 was produced through transgenesis, by down-regulating
two starch branching enzymes involved in building amylopectin. When the

37
branching of amylopectin was down-regulated, the proportion of amylose
was successfully increased.
In Paper III, the amylose content in starch from lines in Group 3 (see
Table 1) and from the RNAi line T-2012 was found to be 98% and 40%,
respectively, based on an enzymatic method. However, lines in Groups 1 and
2 and the parental variety all had an amylose content of around 25%, with no
significant differences according to the enzymatic method. The amylose
content was also measured using a colorimetric method in Paper III. The
results indicated an amylose content of 159-168% in lines in Group 3, 40-
48% in lines in Group 2, 31-35% in lines in Group 1 and 38% in the parental
variety. The RNAi line was found to have an amylose content of 87%, which
is similar to the first published value of 89% for T-2012, obtained using the
same colorimetric method (Menzel et al., 2015). The colorimetric method
was found to overestimate the amylose content, while it was also influenced
by variations in the chain-length distribution of the amylopectin molecule.
In Paper III, none of the Group 1 and 2 lines were found to have as high
an amylose content as that found in the RNAi line T-2012 included in the
analysis. Some previous studies on potato have shown that antisense
suppression of SBE1 alone does not influence the amylose content (Safford
et al., 1998). Lines with completely mutated SBE1 alleles together with two
to three mutated SBE2 alleles (Group 2) only displayed a slightly increased
amylose ratio based on the colorimetric method. Thus, mutations in all alleles
of both SBE1 and SBE2 (Group 3), using the genome-editing method, were
necessary in order to develop a starch with a significant increase in amylose
content detectable using the applied methods.
The amylose content of potato starch in almost all “amylopectin lines”
was below the detection limit of 0.5%, according to the HPSEC method.
However, a line with a partial mutation in GBSS was found to have an
amylose content (14.5%) close to that of the parental variety Kuras (19.2%)
(Paper IV). Hence, full knockout of GBSS was needed to develop a waxy
potato, while to develop a short-chain amylopectin starch SS2 and SS3 also
needed to be mutated.

4.2 Starch granular structure

Most of the potato “amylose lines” displayed the B-type X-ray diffraction
pattern (Figure 6a), with diffraction peaks at 15° (broad), 17° (strong) and a

38
38
doublet at 22-24° 2θ, which is the common pattern for tuber starches
(Hizukuri et al., 2006). However, the diffraction intensity of starch from the
Group 3 lines decreased for most of the peaks compared with the other lines.
Moreover, the Maltese crosses of starch granules from Group 3 were not
visible under polarised light, but were evident in starch from the parent
variety Desiree and line 104016 from Group 2 (Figure 6b).

Figure 6. a) X-ray diffraction patterns of potato lines from Groups 1, 2 and 3. Average
diffraction intensity is shown for each group, with the parental variety Desiree and the
high-amylose line T-2012 included for comparison. b) Images of selected potato starches
using polarised light microscopy. Scale bar = 40 μm.

Compared with the parent potato Dinamo, starch granules from RNAi line
T-2012 displayed more triangle- or rod-shaped and smaller granules with
deep fissures and irregular surfaces (Figure 7A, 7B). Similar results have
been reported previously (Menzel et al., 2015). Under polarised light, the
Maltese crosses of T-2012 starch granules overlapped and some granules

39
showed less birefringence than those of the parental starch Dinamo (Figure
7C, 7D). The altered appearances of the Maltese cross in the RNAi T-2012
starch may indicate some changes in molecule order within the starch
granule.

Figure 7. Microscopic images of potato starch granules (Paper I). A) T-2012 stained with
iodine, B) Dinamo stained with iodine, C) T-2012 under polarised light, D) Dinamo
under polarised light).

X-ray diffraction of starch is correlated with crystalline organisation of

double helices within the starch granules, while the amylopectin fraction
mostly contributes to the crystallinity of granular starches (Hizukuri et al.,
2006). Thus, the considerably reduced X-ray diffraction pattern of the Group
3 starches was probably attributable to the major loss of the amylopectin
fraction in starches from that group. Moreover, the amylose content inversely
affects the level of crystallinity (Cheetham & Tao, 1998). Hence, the high
amylose content of the Group 3 starches might also be the reason for the
lower level of crystallinity in those potato lines. However, the RNAi line T-
2012, with high amylose content, showed a similar X-ray diffraction pattern

40
40
to the parental variety. This is probably because there was no real increase in
amylose molecules, just an increase in long-chain amylopectin.
The Maltese crosses of starch granules generally indicate a highly ordered
granular structure, with molecules arranged in a radial pattern (Bertoft,
2017). It is also known that granular thickness, the degree of crystallinity and
the orientation of the crystallites all affect the apparent intensity of
birefringence (French, 1984). Therefore, the complete disappearance of
Maltese crosses in the Group 3 starches was probably due to a deviation in
crystallite orientation from their radial arrangement. The polarised light
microscopy and X-ray results both revealed that the Group 3 starches have
the ability to organise into granules with a poorly ordered arrangement of
crystalline amylose, even with the essential loss of the amylopectin fraction.
These results are in line with findings reported for amylose-only barley
starch (Carciofi et al., 2012).

4.3 Starch molecular structure

4.3.1 Starch chain distribution (Paper III)

Compared with the wild-type variety Desiree, changes in the chain
distribution pattern were observed in both the amylopectin and amylose
fractions based on the HPSEC results (Paper III), in the order: “amylose
lines” of Group 1, Group 2, T-2012 and Group 3 (Figure 8). Following this
order, the amylopectin fraction decreased gradually to almost absence. No
amylopectin fraction was detected in the Group 3 starches when analysed by
HPAEC. However, the total amount of the amylose fraction increased
constantly, with a shift from a more short-chain amylose fraction to a more
long-chain amylose fraction. Moreover, the T-2012 RNAi line showed a
unique fraction around elution volume 13 mL with amylose-like long glucan
chains (Figure 8).
The starch chain distribution suggested that mutation of SBE1 alone or
combined with SBE2 had different effects on starch molecular structure, as
has been suggested in previous studies (Andersson et al., 2002; Andersson
et al., 2002; Blennow et al., 2005; Brummell et al., 2015; Rydberg et al.,
2001; Tuncel et al., 2019). The reason is believed to be that the potato SBE
isoforms play different roles (Rydberg et al., 2001) during amylopectin
construction (Kossmann & Lloyd, 2000). It has been shown that suppression

41
of both SBEs in potato induces drastic effects on starch molecular structure
(Hofvander et al., 2004; Schwall et al., 2000; Tuncel et al., 2019). This is
consistent with the HPSEC results obtained in this thesis, which showed that
the Group 3 starches experienced a dramatic increase in amylose fraction and
lost a fraction of amylopectin chains.

Figure 8. Chain distribution of debranched starches in potato lines from Groups 1, 2 and
3 after normalisation for the peak area, analysed with HPSEC. The parental variety
Desiree and the high-amylose line T-2012 are included for comparison. The elution
volume of 13 mL is a breakpoint, before which the primary amylose fraction eluted and
after which the main amylopectin fraction eluted. The arrows from left to right indicate
the three populations of amylose chains, i.e. the long-chain, intermediate and short-chain
amylose fraction, respectively.

4.3.2 Amylopectin chain distribution (Paper III)

Besides the overall structural features of potato whole starch, the molar
proportion distribution of amylopectin chains of debranched starches was
analysed in detail by HPAEC (Figure 9).

42
42
Figure 9. Chain distribution of debranched starches on a relative molar basis (M%) with
degree of polymerisation (DP) 6-50, based on HPAEC analysis with averages of potato
lines from Groups 1 and 2. No peak was detected for the starches from Group 3. The
chain-length distribution from the HPAEC analysis primarily represented chains
originating from the amylopectin fraction. The parental variety Desiree and the high-
amylose line T-2012 are included for comparison.

Starch from the wild-type potato Desiree had a predominant peak at DP13
and a shoulder peak at DP11. Compared with the wild-type potato starch, the
starch from T-2012 showed a sharp and high peak at DP11 and a large
increase in chains of DP19-42, but a large decrease in chains of DP7-10 and
DP12-18 (Figure 9). A similar distribution of amylopectin chains in T-2012
potato starch has been reported previously (Menzel et al., 2015), with the
peak shifted to DP22 and the proportion of chains of DP6-18 decreasing in
particular. In general, the chain distribution of starch from Group 1 lines was
similar to that of starch from the parental variety Desiree, with only slightly
increased proportions of DP6 and DP12-21 chains and decreased proportions
of DP7-11 and DP22-33 chains (Figure 9). In contrast, the starch from Group
2 lines analysed in Paper III had a reduced proportion of short chains
(DP≤13), but an increased proportion of DP14-33 chains compared with the

43
parental variety (Figure 9). Short chains with DP6-8 are suggested to be the
amylopectin outermost chains and these chains probably have characteristic
profiles depending on the crop and different genotypes (Bertoft, 2004).
Many studies on cereals and some previous studies on potato have
reported that suppression of a single SBE1 does not affect, or only slightly
affects, the molecular structure of starch (Safford et al., 1998). This was
supported by the HPSEC results obtained in this thesis, which indicated that
the chain distribution of amylopectin fraction in Group 1 starches was similar
to that in the parental variety. However, an altered amylopectin chain
distribution was found in the Group 1 starches in analyses with higher
resolution using HPAEC. This indicates that complete knockout of SBE1
alone affected the starch structure somewhat.

4.3.3 Starch internal chain distribution

The pattern of internal chain distribution of potato whole starch was not
similar to that of barley whole starch. Potato whole starch had a remarkably
high proportion of very long internal chains which eluted before elution
volume 13 mL (Figure 10). Moreover, the proportion of long internal B-
chains was greater in the potato amylopectin fraction compared with in
barley amylopectins, while potato whole starch lacked the dominant
proportion of fingerprint B-chains (Bfp-chains) which elute between elution
volume 16.5 mL and 18 mL in barley whole starch (Figures 10 and 11). It
has been reported previously that tuber starches with a B-type polymorph
contain the lowest amount of short internal B-chains and the highest amount
of long internal B-chains, while cereal amylopectins with an A-type
polymorph contain the highest amount of short internal B-chains and lowest
amount of long internal B-chains (Bertoft et al., 2008).

Internal chain distribution of potato starch (Paper IV)

In the HPSEC analysis, the debranched β-LDs of potato “amylose lines”
from Group 1, Group 2, T-2012 and Group 3 were found to have more (or
even an almost full proportion) of the very long internal B-chains from the
amylose fraction (eluted before elution volume 13 mL) than the wild-type
potato Desiree (Figure 10). Correspondingly, there was a lower proportion
of internal B-chains originating from the amylopectin fraction in starch from
potato lines from Group 1, Group 2, T-2012 and Group 3 than in starch from
the variety Desiree (Figure 10). Compared with the parental variety Kuras,

44
44
the debranched β-LDs of potato starch from “amylopectin lines” showed a
decreased proportion of very long internal B-chains originating from the
amylose fraction, while the degree of mutation increased gradually (Figure
10). Consequently, more internal B-chains that eluted around elution volume
14 mL and between elution volume 15 and 16.5 mL were present in starch
from “amylopectin lines” with an increased degree of mutation.

Figure 10. Mean chain distribution of de-branched E-limit dextrins (β-LDs) in starches
from potato lines with the SBE1, SBE2, GBSS, SS2, and SS3 mutation, analysed with
HPSEC after normalisation of the peak area from elution volume 11 mL to 18 mL
(nothing eluted before 11 mL). Wild-type potatoes (Desiree and Kuras) and T-2012 are
included for comparison. For group information on the mutation lines, see Table 1.

45
 On studying the internal chain distribution of whole potato starch samples
by HPSEC analysis, major variations were found for the amylose fraction
between and within the “amylose lines” and “amylopectin lines” compared
with the wild-type potato starch (Figure 10). The proportion of very long
internal chains was correlated with the amylose content of the starch
samples. Moreover, some variations in long internal chains in T-2012 starch
may be explained by the content of debranched intermediate material, which
can vary from 9% to 4% in high-amylose and native starches (Tang et al.,
2001).

Internal chain distribution of barley starch (Paper II)

Figure 11. Mean distribution of de-branched E-limit dextrins (β-LDs) in starch from
barley varieties with the wax mutation, starch from barley varieties with the amo1
mutation and normal starch, analysed with HPSEC after normalisation of the total area
from elution volume 12 mL to 18 mL (nothing eluted before 12 mL).

Despite differences in patterns, the HPSEC results for the debranched β-LDs
of barley starch revealed a similar structural feature in starch internal chain
distribution to that of potato starch. Starch from barley varieties with the
amo1 mutation had a higher proportion of very long internal B-chains which

46
46
eluted before 13 mL and a lower proportion of intermediate internal B-chains
which eluted between elution volumes 13.7 and 16.5 mL than starch from
normal barley and waxy barley varieties (Figure 11).

4.3.4 Amylopectin internal chain distribution

Significant variations in amylopectin internal chains of whole starch samples
from potato were found between different genotype groups, based on the
distribution analysis of different chain categories using HPAEC. The results
are shown in Table 2.

Internal chain distribution of potato amylopectin (Paper IV)

The molar percentages of Bfp-chains from potato whole starch samples in the
form of β-LDs varied from 16.9% (“amylopectin line”) to 29.8% (“amylose
line”), whereas that of long internal B-chains (DP 42-51) ranged from 1.1%
(high-amylose line T-2012) to 2.6% (“amylopectin line”) (Table 2).
However, starch from T-2012 showed a very different pattern in chain-length
distribution, with a high proportion of internal B-chains with DP 21-24 and
a considerably lower amount of internal B-chains with DP ≥25 compared
with all other starch samples (Table 2).
Table 2. Mean composition of chain categories of β-limit dextrins (β-LDs) with degree
of polymerisation (DP) 4-51, based on relative molar composition analysed by HPAEC,
in potato starch of lines from Group 1 (n=2), Group 2 (n=2), Group 7 (n=2), Group 8
(n=3) and wild type (n=2). Results for T-2012 are included for comparison
Sample DP 4-7 DP 8-12 DP 13-20 DP 21-24 DP 25-41 DP 42-51

Group 8 16.9c 27.3b 28.6a 7.1a 17.5ab 2.6a

Group 7 19.9c 27.4ab 23.8b 5.9b 20.5a 2.4a

Wild type 25.1b 28.7ab 20.6c 5.7b 18.1ab 1.7b

Group 1 29.8a 30.4a 19.4c 5.6b 13.5b 1.3b

Group 2 28.7a 30.1ab 19.8c 6.0b 13.6b 1.7b

T-2012 26.1 27.7 22.6 9.7 12.8 1.1

Different superscript letters indicate statistically significant differences (p<0.05).

Bfp-chains are chains of DP4-7.
The starch of potato lines from Group 3 has almost no amylopectin, and no internal amylopectin chains
were detected.
Values for Groups 4-6 are not included in the statistics because there was only one sample per group.

47
 The potato starch of “amylopectin lines” from Group 8, with the highest
degree of mutations in GBSS, SS2 and SS3, contained the lowest proportion
of Bfp-chains (p≤0.05) and the highest proportion of chains with DP 13-24
(p≤0.001) of all the potato starches (Table 2). In contrast, the potato starch
of “amylose lines” from Group 1 and Group 2 showed the highest proportion
of Bfp-chains (p<0.05) of all the potato starches. The potato starch of
“amylopectin lines” from Group 7 had the second lowest proportion of Bfp-
chains (p≤0.05). It also had the second highest proportion of chains with DP
13-20 (p<0.01) (Table 2) among all the potato starches analysed except that
of line from Group 5 (Paper IV). The potato starch of “amylopectin lines”
from Group 7 showed a higher proportion of internal B-chains with DP 25-
41 than the starch from the “amylose line” potatoes (p<0.05). For the long
internal B-chains with DP 42-51, the potato starch of “amylopectin lines”
from Group 7 and Group 8 had a significantly higher proportion than the
starch from the wild-type and the “amylose line” potatoes (p<0.05). For the
“amylose line” starches, no significant difference was found between the
lines from Group 1 and Group 2 or with the starch of wild-type potatoes in
the chain categories except for Bfp-chains (Table 2).
In general, different combinations of mutated GBSS, SS2 and SS3 resulted
in different effects on amylopectin internal chain distribution. Clear
significant variations were found in the molar proportion of Bfp-chains that
correlated with the phenotype and genetic background of the potato lines
(Table 2). In descending order, the ranking was “amylose lines” from Group
1 and Group 2, wild type, and “amylopectin lines” from Group 7 and Group
8.

4.3.5 Building block structure

Building block composition of potato starch (Paper IV)

The differences in building block groups revealed by HPAEC and the
differences in composition of building blocks from potato whole starch
samples were related to the genetic background of potato lines (Table 3). In
general, the proportion of intermediate (G4) and large (G5 and G6) building
blocks increased gradually in the starch of potato lines in the order: Group 8,
Group 7, wild type, Group 1 and Group 2. The proportion of small (G2 and
G3) building blocks declined gradually following the same order. The starch

48
48
of “amylose lines” from Group 1 and Group 2 had significantly more
(p<0.01) intermediate (G4) and large (G5 and G6) building blocks, and fewer
(p<0.05) small (G2 and G3) building blocks, than the starch of “amylopectin
lines” from Group 7 and Group 8 (Table 3). For building blocks G3, G4 and
G5, the starch of lines from Group 1 and Group 2 also showed significant
differences (p<0.05).
However, a very different distribution of building blocks was found in the
starch of T-2012, where building blocks G2 (42.2%) and G4 (34.6%) were
the major components, whereas the starch from all the other samples mostly
contained G2 and G3 building blocks (Table 3). As a consequence, the starch
of T-2012 had the highest proportion of intermediate (G4) and large (G5 and
G6) building blocks, and the lowest proportion of small (G2 and G3) building
blocks, of all the starch samples analysed in this thesis (Table 3).

Table 3. Mean composition of building blocks in five different size groups (G2-G6), based
on relative peak area analysed by HPAEC, in potato starch samples of lines from Group
1 (n=2), Group 2 (n=2), Group 7 (n=2) and Group 8 (n=3) and in the wild type (n=2).
Results for T-2012 are included for comparison
Sample G2 G3 G4 G5 G6

Group 8 63.4a 28.0a 6.5d 1.7d 0.4b

Group 7 62.2a 26.8b 8.5cd 2.0cd 0.5b

Wild type 61.3ab 23.9c 11.3bc 2.7bc 0.7b

Group 1 57.2bc 22.8d 14.9b 3.7b 1.5a

Group 2 54.9c 19.0e 19.4a 4.9a 1.8a

T-2012 42.2 12.6 34.6 7.9 2.7

Different superscript letters indicate statistically significant differences (p<0.05).

The starch of potato lines from Group 3 has almost no amylopectin and no building blocks were detected.
Values for Group 4-6 are not included in the statistics because there was only one sample per group.

49
Figure 12. Mean distribution of building blocks with two and more chains in starch from
potato lines with the SBE1, SBE2, GBSS, SS2 and SS3 mutation, determined by HPSEC
analysis after normalisation of the peak area from elution volume 13 mL to 18 mL
(nothing eluted before 13 mL). Numbers 2-6 indicate building block size groups G2-G6.
The wild-type potato starches and T-2012 are included for comparison. For full
information on the mutation lines, see Table 1.

The HPSEC results of building block composition for the potato starch
samples (Figure 12) agreed very well with the distribution of building block
analysed by HPAEC (Table 3). The proportion of intermediate (G4) and

50
50
large (G5 and G6) building blocks increased gradually in the starch of potato
lines with the order Group 8, Group 7, Group 6, Group 5, Group 4, wild type,
Group 1, Group 2 and T-2012. The proportion of small (G2 and G3) building
blocks declined gradually following the same order. The pattern of building
block composition of potato whole starch and barley whole starch was
generally similar, with the slight difference that potato whole starch had a
comparatively larger proportion of intermediate (G4) building blocks than
barley whole starch (Figures 12 and 13).

Building block composition of barley starch (Paper II)

Figure 13. Mean distribution of building blocks with two and more chains in starch from
barley varieties with the wax mutation, starch from barley varieties with the amo1
mutation and normal starch, determined by HPSEC analysis after normalisation of the
total area from elution volume 13 mL to 18 mL (nothing eluted before 13 mL). Numbers
2-6 indicate building block size groups G2-G6, which were assigned according to the old
division of building blocks (Bertoft et al., 2011a).

51
4.4 Nutritional properties of potato line T-2012
In Paper I, the nutritional properties of potato tubers from the high-amylose
line T-2012 were investigated and compared with those of the parental potato
cultivar Dinamo. T-2012 had a slightly lower total starch content (65-68%
of dry matter (DM)) than the parental potato Dinamo (72-76% of DM). This
supports findings in a previous study that high-amylose potato lines
generally have a reduced starch content, but an increased content of free
sugars (such as fructose and glucose) (Hofvander et al., 2004). However, the
content of DF excluding RS in T-2012 (10-14% of DM) was significantly
(p<0.001) higher than that in Dinamo (5-7% of DM), irrespective of whether
it was cooked or not.
The RS in uncooked tubers is type II RS, where starch granules resist
enzyme digestion (Brown et al., 1995). T-2012 contained less RS (30% of
DM) than the parent potato Dinamo (56% of DM), which could be due to the
altered starch granular structure in T-2012 (Figure 7). It is known that besides
amylose content, granular structure also has a strong impact on enzyme
resistance (Themeier et al., 2005). Native potato starch granules are less
accessible to enzymes, due to the homogeneous, compact and large granules,
and the smooth surfaces of these granules (Lehmann & Robin, 2007). The
starch granules in T-2012 were much smaller, with a more irregular shape
and surface and deep fissures (see Figure 7). This probably renders the
granules more accessible to enzymes, and therefore the type II RS content
decreased in raw tubers of this high-amylose potato line.

52
52
Figure 14. Content of resistant starch (RS) in boiled potato tubers of the parent potato
Dinamo and the high-amylose line T-2012 (Paper I).

T-2012 had much higher level of physiological RS (13-20% of DM) than

the parent potato (4-7% of DM) after cooking (Figure 14). This was most
probably due to the higher amylose content in T-2012, since a high level of
amylose usually gives a high level of RS in cooked food (Leeman et al.,
2006; Lehmann & Robin, 2007). After cold storage, the level of RS in T-
2012 increased further, but the level of type III RS (i.e. recrystallised
amylose) did not show significant differences between T-2012 and the
parent, or between different storage durations. When the level of
recrystallised amylose did not differ, part of the increased level of RS after
cold storage might be recrystallised amylopectin, which needs some time to
form. Recrystallised amylopectin is generally not resistant to enzyme
degradation, since it has a much lower melting point (35-80 qC) than
recrystallised amylose (120-170 qC) (Silverio et al., 2000; Sievert &
Pomeranz, 1990). However, genetic modification was found to have a major
effect on amylopectin structure, with the RNAi line T-2012 having a very
large fraction of long outer starch chains and intermediate-sized inner

53
amylopectin chains, and more intermediate and large building blocks,
compared with the wild-type potato. Longer internal amylopectin chains
have been found to promote formation of recrystallised amylopectin with
higher thermal stability, due to a more ordered structure (Zhu, 2018). This
unique amylopectin has properties that are similar to amylose. After cooking
and cooling, the modified amylopectin recrystallises and thereafter is not as
easily split as ordinary potato starch. Therefore in total, T-2012 had almost
three-fold higher content of RS than the parental potato after cooking and
cold storage.
The cell wall composition of T-2012 was also altered indirectly by the
genetic modification, giving a greater amount of glucose residues (cellulose)
and a smaller amount of other polysaccharides than in the parent (Figure 15).

Figure 15. Composition of the cell walls of the parent potato Dinamo and the high-
amylose line T-2012, as revealed by dietary fibre analysis of uncooked potato tubers.

4.5 Influence of starch structure on thermal properties

In Paper IV, the thermal properties of starch in the 15 potato lines and two
wild types included in the thesis were investigated. Gelatinisation was
examined using DSC with a starch:water ratio of 1:3. The starch of the high-
amylose potato line T-2012 and “amylopectin lines” from Group 8 had a

54
54
broader gelatinisation temperature range. T-2012 had the highest peak (75.8
°C) and endset (87.4 °C) temperature, whereas the “amylopectin lines” from
Group 8 had the lowest mean onset (p≤0.001) and peak temperature (p<0.01)
(Figure 16). The onset temperature for gelatinisation was ~50-54 °C and the
peak temperature was ~59-61 °C in the potato lines from Group 8, compared
with ~60-65 °C and ~64-76 °C, respectively, in the other potato lines. These
results are in line with previous findings that starch with more heterogeneous
crystals leads to a broader temperature range of gelatinisation (Källman et
al., 2015).
The relationship between gelatinisation and structural features was
studied by PLS regression analysis. Structural parameters were able to
explain most of the variance with the first two components in onset, peak and
endset temperatures and gelatinisation enthalpy (Paper IV). The onset
temperature of gelatinisation and the gelatinisation enthalpy were positively
correlated with short internal B-chains (DP 4-7 and DP 8-12), and large and
intermediate building blocks (G6, G5 and G4). The peak and endset
temperatures of gelatinisation were positively affected by intermediate and
large building blocks (G4, G5 and G6), and short internal B-chains of DP 4-
7 and intermediate B-chains of DP 21-24.
The mean gelatinisation enthalpy of potato starch from the different lines
gradually increased from “amylopectin starch” (15 J/g in Group 8), to wild
type starch (18.7 J/g), and finally to “amylose starch” (19.2 J/g in Group 2)
(Paper IV). The PLS regression coefficients showed that the thermal
properties were influenced by the size of the building blocks in amylopectin.
Presence of more large building blocks and more short internal B-chains
resulted in higher gelatinisation enthalpy (p<0.05). A previous study
concluded that more perfect crystals are formed within the granule,
contributing to a higher peak temperature of gelatinisation, if the
amylopectin building blocks have short chains and a dense structure
(Källman et al., 2015).

55
Figure 16. Gelatinisation onset, peak and endset temperatures of starch from the different
groups of potato lines.

In Paper IV, the retrogradation of potato starch was also investigated after
gelatinisation at a starch:water ratio of 1:2 and storage at 5 °C for three days.
There were no clear differences in the onset and peak melting temperatures
of retrograded starch between the different potato lines. The onset and peak
of melting temperature was ~39-40 °C and ~64-66 °C, respectively, for all
potato starches (Figure 17). However, T-2012 had the lowest endset (70.6
°C) temperature, whereas the “amylose lines” from Group 2 had the highest

56
56
mean endset (84.2 °C) temperature (p<0.05). As a result, T-2012 had the
narrowest melting temperature range, while the amylose lines from Group 2
had the widest range (Figure 17).

Figure 17. Retrogradation onset, peak and endset temperatures of starch from the
different groups of potato lines.

Although no significant effect of structural parameters was seen for the

onset, peak and endset melting temperatures of retrograded potato starch, the
starch of potato lines from Group 8 (3.1 J/g; p<0.001) and Group 7 (6.0 J/g;
p≤0.001) had lower retrogradation enthalpy than the starch from other lines
(~8 J/g), indicating that fewer crystals were formed. When the retrogradation
enthalpy of the potato starch was modelled by PLS analysis, the regression
coefficients showed that the retrogradation enthalpy was influenced by the
different chain categories, size of the building blocks in amylopectin and the

57
amylose content. More short internal B-chains (DP 4-7 and 8-12) and more
large and intermediate building blocks (G6, G5 and G4) resulted in higher
retrogradation enthalpy. This indicates that molecular network re-formation
in gelatinised starch is favoured by a high density of branching and low levels
of long internal chains.
Amylose content also contributed to the amylopectin gelatinisation and
retrogradation properties of the different starches. The enthalpy of potato
starch gelatinisation and the retrogradation and gelatinisation temperatures
were positively affected by the amylose content. It is also possible that the
correlation between starch thermal properties and amylose content is
attributable to a correlation between starch structure and amylose level. The
starch of potato lines from Group 8, with a lack of amylose and the most
affected amylopectin, also had the highest degree of mutations, with the
lowest levels of short internal B-chains and large building blocks found in
this type of starch.
However, the starch of T-2012 line, with high amylose content, resulted
in lower retrogradation enthalpy but the highest peak temperature of
gelatinisation. This might be because oven cooking at 105 °C was not
sufficient to achieve full gelatinisation of the starch in T-2012. A previous
study indicated that the starch of T-2012 needs higher gelatinisation
temperature, since the starch showed no granule swelling and viscosity under
the conditions applied in rapid viscosity analysis with the standard method
(up to 95 °C) (Menzel et al., 2015). When the gelatinised starch of T-2012
was oven-cooked again at a higher temperature (around 120 °C), its
retrogradation enthalpy increased to 8.0 J/g. Moreover, the peak and endset
temperatures of retrogradation and the temperature range increased
dramatically (Paper IV). The very large fraction of long outer starch chains
and large building blocks of T-2012 could favour the formation of more
perfect crystals compared with the wild-type potato starch.

58
58
5. Conclusions
Starch is a glucose polysaccharide consisting of amylose (essentially
long linear molecule) and amylopectin (highly branched molecule).
Amylopectin is constructed from A-, B-, and C-chains (carrying the
reducing end) connected through branching points. Knowledge of starch
internal structure (the structure of chain segments from the outmost
branches to the reducing end) is important in understanding the
biosynthesis of starch and the starch structure-property relationship.
However, there are challenges in measuring the levels of different chain-
length and building block categories, especially when newly developed
starches do not fall into any of the categories defined previously.
In this thesis, a method for determining the internal structure of starch
without prior amylopectin isolation was developed and optimised.
Whole starches of potato and barley were degraded into β-LDs, α-LDs
and building blocks with this simplified method. The internal B-chain
distribution and building block composition were characterised
effectively, providing a deeper understanding of the relationship
between molecular structure and functional properties.
Potato whole starches had a markedly higher proportion of long
internal chains from both amylose and amylopectin fractions compared
with barley whole starches, while potato whole starches were lacking the
dominant proportion of Bfp-chains that is found in barley whole starches.
The pattern of building block composition of potato whole starch and
barley whole starch was generally similar, with slightly different
proportions of the intermediate (G4) building blocks. The general
descending order Group 2>Group 1>wild type>Group 7>Group 8) was
determined for the composition of intermediate and large building blocks
and proportion of Bfp-chains. These structure parameters appeared to be

59
related to genetic background and could be used to predict some
physicochemical properties of starches from different plant sources and
genetic backgrounds.
The high-amylose potato line T-2012 was found to have higher levels
of resistant starch and dietary fibre than the parental variety after
cooking, and may therefore have beneficial effects on human health. The
resistant starch content was influenced both by the amylose content and
by the amylopectin structure. T-2012 starch had a very large fraction of
long outer chains and intermediate-sized inner amylopectin chains, and
more intermediate and large building blocks, compared with the wild-
type potato. The unique amylopectin structure of T-2012 starch
promoted the formation of a unique recrystallised amylopectin with
properties similar to those of amylose. After recrystallisation, this
amylopectin did not split as easily as ordinary potato starch and was
resistant to enzyme digestion. The information obtained in this thesis on
the unique amylopectin structure in T-2012 can be useful in designing
functional starch and healthier food in future.
Links between structural features and thermal properties
(gelatinisation and retrogradation) were also uncovered in this thesis.
Various parameters of the internal structure of potato starch were found
to be significantly correlated with the gelatinisation temperature,
enthalpy of gelatinisation and retrogradation of potato starch. Dense
structure of the building blocks led to higher gelatinisation temperatures
and enthalpy. Retrogradation was found to be favoured by presence of
more large building blocks with many short internal chains.

60
60
6. Future research
Following on from this thesis, further work needs to be performed in order
to customise starch with predicted properties, through:

¾ Further studies on the effect of genetic modification on starch

structure and the relationship between molecular structure and
functional properties of starch

¾ Investigations on the nutritional properties of the “amylose lines” of

potato developed via DNA-free CRISPR-Cas9-mediated
mutagenesis.

¾ Further studies on amylose starch with no detectable branching.

¾ Development of a method to zoom in on amylose structure using

HPSEC.

¾ Comparative investigations of amylose structure from potato lines

with different amylose contents.

¾ Wider application of the simplified analytical method to diverse

starch samples from different plant sources, to characterise the
internal molecular structure of whole starch.

¾ Studies on the correlation between amylose structure and functional

properties of starch.

61
References
Andersson, L., Andersson, R., Andersson, R. E., Rydberg, U., Larsson, H., &
Åman, P. (2002). Characterisation of the in vitro products of potato starch
branching enzymes I and II. Carbohydrate Polymers, 50(3), 249-257.
https://doi.org/10.1016/S0144-8617(02)00059-0
Andersson, L., Rydberg, U., Larsson, H., Andersson, R., & Åman, P. (2002).
Preparation and characterisation of linear dextrins and their use as
substrates in in vitro studies of starch branching enzymes. Carbohydrate
Polymers, 47(1), 53-58. https://doi.org/10.1016/S0144-8617(01)00162-X
Andersson, M., Melander, M., Pojmark, P., Larsson, H., Bulow, L., & Hofvander,
P. (2006). Targeted gene suppression by RNA interference: An efficient
method for production of high-amylose potato lines. Journal of
Biotechnology, 123(2), 137-148.
https://doi.org/10.1016/j.jbiotec.2005.11.001
Andersson, M., Trifonova, A., Andersson, A. B., Johansson, M., Bulow, L., &
Hofvander, P. (2003). A novel selection system for potato transformation
using a mutated AHAS gene. Plant Cell Reports, 22(4), 261-267.
https://doi.org/10.1007/s00299-003-0684-8
Andersson, M., Turesson, H., Nicolia, A., Falt, A. S., Samuelsson, M., &
Hofvander, P. (2017). Efficient targeted multiallelic mutagenesis in
tetraploid potato (Solanum tuberosum) by transient CRISPR-Cas9
expression in protoplasts. Plant Cell Reports, 36(1), 117-128.
https://doi.org/10.1007/s00299-016-2062-3
Andersson, R., Fransson, G., Tietjen, M., & Åman, P. (2009). Content and
molecular-weight distribution of dietary fiber components in whole-grain
rye flour and bread. Journal of Agricultural and Food Chemistry, 57(5),
2004-2008. https://doi.org/10.1021/jf801280f
Bergthaller, W. (2004). Developments in potato starches. In Ann-Charlotte
Eliasson (Ed.), Starch in food. Structure, function and applications (pp.
241-257): Woodhead Publishing.
Bertoft, E. (1989). Investigation of the fine-structure of amylopectin using alpha-
amylase and beta-amylase. Carbohydrate Research, 189, 195-207.
https://doi.org/10.1016/0008-6215(89)84097-2
Bertoft, E. (2004). On the nature of categories of chains in amylopectin and their
connection to the super helix model. Carbohydr Polym, 57(2), 211-224.
https://doi.org/10.1016/j.carbpol.2004.04.015
Bertoft, E. (2013). On the building block and backbone concepts of amylopectin
structure. Cereal Chemistry, 90(4), 294-311.
https://doi.org/10.1094/CCHEM-01-13-0004-FI

63
Bertoft, E. (2017). Understanding starch structure: Recent progress. Agronomy-
Basel, 7(3). https://doi.org/10.3390/agronomy7030056
Bertoft, E., & Blennow, A. (2016). Structure of potato starch. In Jaspreet Singh &
Lovedeep Kaur (Eds.), Advances in potato chemistry and technology. San
Diego, California, United States: Nikki Levy.
Bertoft, E., Källman, A., Koch, K., Andersson, R., & Åman, P. (2011a). The
building block structure of barley amylopectin. Int J Biol Macromol,
49(5), 900-909. https://doi.org/10.1016/j.ijbiomac.2011.08.001
Bertoft, E., Källman, A., Koch, K., Andersson, R., & Åman, P. (2011b). The
cluster structure of barley amylopectins of different genetic backgrounds.
Int J Biol Macromol, 49(4), 441-453.
https://doi.org/10.1016/j.ijbiomac.2011.04.014
Bertoft, E., Koch, K., & Åman, P. (2012a). Building block organisation of clusters
in amylopectin from different structural types. Int J Biol Macromol, 50(5),
1212-1223. https://doi.org/10.1016/j.ijbiomac.2012.03.004
Bertoft, E., Koch, K., & Åman, P. (2012b). Structure of building blocks in
amylopectins. Carbohydrate Research, 361, 105-113.
https://doi.org/10.1016/j.carres.2012.08.012
Bertoft, E., Piyachomkwan, K., Chatakanonda, P., & Sriroth, K. (2008). Internal
unit chain composition in amylopectins. Carbohydr Polym, 74(3), 527-
543. https://doi.org/10.1016/j.carbpol.2008.04.011
Birch, P. R. J., Bryan, G., Fenton, B., Gilroy, E. M., Hein, I., Jones, J. T., . . . Toth,
I. K. (2012). Crops that feed the world 8: Potato: are the trends of
increased global production sustainable? Food Security, 4(4), 477-508.
https://doi.org/10.1007/s12571-012-0220-1
Blennow, A., Wischmann, B., Houborg, K., Ahmt, T., Jorgensen, K., Engelsen, S.
B., . . . Poulsen, P. (2005). Structure function relationships of transgenic
starches with engineered phosphate substitution and starch branching.
International Journal of Biological Macromolecules, 36(3), 159-168.
https://doi.org/10.1016/j.ijbiomac.2005.05.006
Brown, I. L., McNaught, K. J., Andrews, D., & Morita, T. (2001). Resistant starch:
Plant breeding, applications development and commercial use. In B. V.
McCleary & L. Prosky (Eds.), Advanced dietary fibre technology (pp.
399-412). Oxford; Malden, MA: Blackwell Science Ltd.
Brown, I. L., McNaught, K. J., & Moloney, E. (1995). Hi-maizeTM: new directions
in starch technology and nutrition. Food Australia, 47(6), 272-275.
Brummell, D. A., Watson, L. M., Zhou, J., McKenzie, M. J., Hallett, I. C.,
Simmons, L., . . . Timmerman-Vaughan, G. M. (2015). Overexpression of
starch branching enzyme II increases short-chain branching of
amylopectin and alters the physicochemical properties of starch from
potato tuber. Bmc Biotechnology, 15. https://doi.org/10.1186/s12896-015-
0143-y

64
64
Buléon, A., Colonna, P., Planchot, V. & Ball, S. (1998). Starch granules: structure
and biosynthesis. International Journal of Biological Macromolecules
23:85-112. https://doi.org/10.1016/S0141-8130(98)00040-3
Carciofi, M., Blennow, A., Jensen, S. L., Shaik, S. S., Henriksen, A., Buleon, A., . .
. Hebelstrup, K. H. (2012). Concerted suppression of all starch branching
enzyme genes in barley produces amylose-only starch granules. Bmc
Plant Biology, 12. https://doi.org/10.1186/1471-2229-12-223
Cheetham, N. W. H., & Tao, L. P. (1998). Variation in crystalline type with
amylose content in maize starch granules: an X-ray powder diffraction
study. Carbohydrate Polymers, 36(4), 277-284.
https://doi.org/10.1016/S0144-8617(98)00007-1
Chrastil, J. (1987). Improved colorimetric determination of amylose in starches or
flours. Carbohydrate Research, 159(1), 154-158.
https://doi.org/10.1016/S0008-6215(00)90013-2
Englyst, H. N., & Cummings, J. H. (1984). Simplified method for the measurement
of total non-starch polysaccharides by gas - liquid-chromatography of
constituent sugars as alditol acetates. Analyst, 109(7), 937-942.
https://doi.org/10.1039/AN9840900937
Englyst, H., Wiggins, H. S., & Cummings, J. H. (1982). Determination of the non-
starch polysaccharides in plant foods by gas-liquid-chromatography of
constituent sugars as alditol acetates. Analyst, 107(1272), 307-318.
https://doi.org/10.1039/AN9820700307
Eriksson, D., Carlson-Nilsson, U., Ortiz, R., & Andreasson, E. (2016). Overview
and breeding strategies of table potato production in Sweden and the
Fennoscandian region. Potato Research, 59(3), 279-294.
https://doi.org/10.1007/s11540-016-9328-6
French, D. (1984). Chapter VII - Organization of starch granules. in starch:
Chemistry and technology (Second Edition) (eds. Whistler, R. L.,
Bemiller, J. N. & Paschall, E. F.) 183–247 (Academic Press).
Hasjim, J., Ai, Y., & Jane, J.-L. (2013). Novel applications of amylose-lipid
complex as resistant starch type 5. In Y.-C. Shi & Maningat C. C. (Eds.),
Resistant starch: Sources, applications and health benefits (pp. 79-94).
Chichester, United Kingdom: John Wiley and Sons Ltd.
Hizukuri, S., Abe, J-i., & Hanashiro, I. (Eds.). (2006). Starch: Analytical aspects.
Boca Raton, FL, USA: CRC Press.
Hofvander, P., Andersson, M., Larsson, C. T., & Larsson, H. (2004). Field
performance and starch characteristics of high-amylose potatoes obtained
by antisense gene targeting of two branching enzymes. Plant
Biotechnology Journal, 2(4), 311-320. https://doi.org/10.1111/j.1467-
7652.2004.00073.x
Hoover, R. (2001). Composition, molecular structure, and physicochemical
properties of tuber and root starches: a review. Carbohydr Polym, 45(3),
253-267. https://doi.org/10.1016/S0144-8617(00)00260-5

65
Imberty, A., Buléon, A., Tran, V. & Pérez, S. (1991). Recent advances in
knowledge of starch structure. Starch/Stärke 43(10):375-384.
https://doi.org/10.1002/star.19910431002
Jacobsen, E., Ramanna, M. S., Huigen, D. J., & Sawor, Z. (1991). Introduction of
an amylose-free (amf) mutant into breeding of cultivated potato, solanum-
tuberosum l. Euphytica, 53(3), 247-253.
https://doi.org/10.1007/BF00023276
Jane, J. L., Kasemsuwan, T., Leas, S., Zobel, H., & Robyt, J. F. (1994). Anthology
of starch granule morphology by scanning electron-microscopy. Starch-
Starke, 46(4), 121-129. https://doi.org/10.1002/star.19940460402
Jane, J., Chen, Y.Y., Lee, L.F., McPherson, A.E., Wong, K.S., Radosavljevic, M.,
et al. (1999). Effects of amylopectin branch chain length and amylose
content on the gelatinization and pasting properties of starch. Cereal
Chemistry, 76 (5), 629-637
https://doi.org/10.1094/CCHEM.1999.76.5.629
Jenkins, J. P. J., Cameron, R. E., & Donald, A. M. (1993). A universal feature in
the structure of starch granules from different botanical sources. Starch-
Starke, 45(12), 417-420. https://doi.org/10.1002/star.19930451202
Källman, A., Bertoft, E., Koch, K., Åman, P., & Andersson, R. (2013). On the
interconnection of clusters and building blocks in barley amylopectin. Int
J Biol Macromol, 55, 75-82.
https://doi.org/10.1016/j.ijbiomac.2012.12.032
Källman, A., Vamadevan, V., Bertoft, E., Koch, K., Seetharaman, K., Åman, P., &
Andersson, R. (2015). Thermal properties of barley starch and its relation
to starch characteristics. Int J Biol Macromol, 81, 692-700.
https://doi.org/10.1016/j.ijbiomac.2015.08.068
Källman, Anna. (2013). Barley starch molecular structure and properties.
(Doctoral thesis), Swedish University of Agricultural Sciences, Uppsala.
Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:slu:epsilon-e-1658
(2013:73)
Karlsson, M. E., Leeman, A. M., Bjorck, I. M. E., & Eliasson, A. C. (2007). Some
physical and nutritional characteristics of genetically modified potatoes
varying in amylose/amylopectin ratios. Food Chemistry, 100(1), 136-146.
https://doi.org/10.1016/j.foodchem.2005.09.032
Koch, K., Andersson, R., & Åman, P. (1998). Quantitative analysis of amylopectin
unit chains by means of high-performance anion-exchange
chromatography with pulsed amperometric detection. Journal of
Chromatography A, 800(2), 199-206. https://doi.org/10.1016/S0021-
9673(97)01151-5
Kossmann, J., & Lloyd, J. (2000). Understanding and influencing starch
biochemistry. Critical Reviews in Plant Sciences, 19(3), 171-226.
https://doi.org/10.1080/07352680091139204
Larsson, C. T., Hofvander, P., Khoshnoodi, J., Ek, B., Rask, L., & Larsson, H.
(1996). Three isoforms of starch synthase and two isoforms of branching

66
66
 enzyme are present in potato tuber starch. Plant Science, 117(1-2), 9-16.
https://doi.org/10.1016/0168-9452(96)04408-1
Leeman, A. M., Karlsson, M. E., Eliasson, A. C., & Bjorck, I. M. E. (2006).
Resistant starch formation in temperature treated potato starches varying
in amylose/amylopectin ratio. Carbohydr Polym, 65(3), 306-313.
https://doi.org/10.1016/j.carbpol.2006.01.019
Lehmann, U., & Robin, F. (2007). Slowly digestible starch - its structure and
health implications: a review. Trends in Food Science & Technology,
18(7), 346-355. https://doi.org/10.1016/j.tifs.2007.02.009
Lemos, P. V. F., Barbosa, L. S., Ramos, I. G., Coelho, R. E., & Druzian, J. I.
(2019). Characterization of amylose and amylopectin fractions separated
from potato, banana, corn, and cassava starches. Int J Biol Macromol,
132, 32-42. https://doi.org/10.1016/j.ijbiomac.2019.03.086
McCleary, B. V., & Monaghan, D. A. (2002). Measurement of resistant starch.
Journal of Aoac International, 85(3), 665-675.
https://doi.org/10.1093/jaoac/85.3.665
Menzel, C., Andersson, M., Andersson, R., Vazquez-Gutierrez, J. L., Daniel, G.,
Langton, M., . . . Koch, K. (2015). Improved material properties of
solution-cast starch films: Effect of varying amylopectin structure and
amylose content of starch from genetically modified potatoes. Carbohydr
Polym, 130, 388-397. https://doi.org/10.1016/j.carbpol.2015.05.024
Morrison, W. R., & Laignelet, B. (1983). An improved colorimetric procedure for
determining apparent and total amylose in cereal and other starches.
Journal of Cereal Science, 1(1), 9-20. https://doi.org/10.1016/S0733-
5210(83)80004-6
Peat, S., Whelan, W. J., & Thomas, G. J. (1952). Evidence of multiple branching in
waxy maize starch. Journal of the Chemical Society(Nov), 4546-4548.
Perez, S., & Bertoft, E. (2010). The molecular structures of starch components and
their contribution to the architecture of starch granules: A comprehensive
review. Starch-Starke, 62(8), 389-420.
https://doi.org/10.1002/star.201000013
Rydberg, U., Andersson, L., Andersson, R., Åman, P., & Larsson, H. (2001).
Comparison of starch branching enzyme I and II from potato. European
Journal of Biochemistry, 268(23), 6140-6145.
https://doi.org/10.1046/j.0014-2956.2001.02568.x
Safford, R., Jobling, S. A., Sidebottom, C. M., Westcott, R. J., Cooke, D., Tober,
K. J., . . . Gidley, M. J. (1998). Consequences of antisense RNA inhibition
of starch branching enzyme activity on properties of potato starch.
Carbohydrate Polymers, 35(3-4), 155-168. https://doi.org/10.1016/S0144-
8617(97)00249-X
Sarka, E., & Dvoracek, V. (2017). Biosynthesis of waxy starch - a review. Plant
Soil and Environment, 63(8), 335-341. https://doi.org/10.17221/324/2017-
PSE

67
Sasaki, T., Yasui, T., & Matsuki, J. (2000). Effect of amylose content on
gelatinization, retrogradation, and pasting properties of starches from
waxy and nonwaxy wheat and their F1 seeds. Cereal Chemistry, 77(1),
58-63. https://doi.org/10.1094/CCHEM.2000.77.1.58
Schwall, G. P., Safford, R., Westcott, R. J., Jeffcoat, R., Tayal, A., Shi, Y. C., . . .
Jobling, S. A. (2000). Production of very-high-amylose potato starch by
inhibition of SBE A and B. Nature Biotechnology, 18(5), 551-554.
https://doi.org/10.1038/75427
Sievert, D., & Pomeranz, Y. (1990). Enzyme-resistant starch. 2. Differential
scanning calorimetry studies on heat - treated starches and enzyme-
resistant starch residues. Cereal Chemistry, 67(3), 217-221.
Silverio, J., Fredriksson, H., Andersson, R., Eliasson, A. C., & Åman, P. (2000).
The effect of temperature cycling on the amylopectin retrogradation of
starches with different amylopectin unit-chain length distribution.
Carbohydr Polym, 42(2), 175-184. https://doi.org/10.1016/S0144-
8617(99)00140-X
Srichuwong, S., & Jane, J. L. (2007). Physicochemical properties of starch affected
by molecular composition and structures: A review. Food Science and
Biotechnology, 16(5), 663-674.
Srichuwong, S., Sunarti, T. C., Mishima, T., Isono, N., & Hisamatsu, M. (2005).
Starches from different botanical sources I: Contribution of amylopectin
fine structure to thermal properties and enzyme digestibility. Carbohydr
Polym, 60(4), 529-538. https://doi.org/10.1016/j.carbpol.2005.03.004
Tang, H., Ando, H., Watanabe, K., Takeda, Y., & Mitsunaga, T. (2001). Fine
structures of amylose and amylopectin from large, medium, and small
waxy barley starch granules. Cereal Chemistry, 78(2), 111-115.
https://doi.org/10.1094/CCHEM.2001.78.2.111
Theander, O., Aman, P., Westerlund, E., Andersson, R., & Petersson, D. (1995).
Total dietary fiber determined as neutral sugar residues, uronic acid
residues, and Klason Lignin (The Uppsala method): Collaborative study.
Journal of Aoac International, 78(4), 1030-1044.
Themeier, H., Hollmann, J., Neese, U., & Lindhauer, M. G. (2005). Structural and
morphological factors influencing the quantification of resistant starch II
in starches of different botanical origin. Carbohydrate Polymers, 61(1),
72-79. https://doi.org/10.1016/j.carbpol.2005.02.017
Tian, J. H., Chen, J. C., Ye, X. Q., & Chen, S. G. (2016). Health benefits of the
potato affected by domestic cooking: A review. Food Chemistry, 202,
165-175. https://doi.org/10.1016/j.foodchem.2016.01.120
Tuncel, A., Corbin, K. R., Ahn-Jarvis, J., Harris, S., Hawkins, E., Smedley, M. A.,
. . . Smith, A. M. (2019). Cas9-mediated mutagenesis of potato starch-
branching enzymes generates a range of tuber starch phenotypes. Plant
Biotechnology Journal, 17(12), 2259-2271.
https://doi.org/10.1111/pbi.13137

68
68
Vamadevan, V., & Bertoft, E. (2015). Structure-function relationships of starch
components. Starch-Starke, 67(1-2), 55-68.
https://doi.org/10.1002/star.201400188
Van Harsselaar, J. K., Lorenz, J., Senning, M., Sonnewald, U., & Sonnewald, S.
(2017). Genome-wide analysis of starch metabolism genes in potato
(Solanum tuberosum L.). Bmc Genomics, 18.
https://doi.org/10.1186/s12864-016-3381-z
Visser, R. G. F., Somhorst, I., Kuipers, G. J., Ruys, N. J., Feenstra, W. J., &
Jacobsen, E. (1991). Inhibition of the expression of the gene for granule-
bound starch synthase in potato by antisense constructs. Molecular &
General Genetics, 225(2), 289-296. https://doi.org/10.1007/BF00269861
Zeeman, Samuel C., Kossmann, Jens, & Smith, Alison M. (2010). Starch: Its
metabolism, evolution, and biotechnological modification in plants.
Annual Review of Plant Biology, 61, 209-234.
https://doi.org/10.1146/annurev-arplant-042809-112301
Zhao, X., Jayarathna, S., Turesson, H., Falt, A. S., Nestor, G., Gonzalez, M. N., . . .
Andersson, M. (2021). Amylose starch with no detectable branching
developed through DNA-free CRISPR-Cas9 mediated mutagenesis of two
starch branching enzymes in potato. Scientific Reports, 11(1).
https://doi.org/10.1038/s41598-021-83462-z
Zhu, F. (2018). Relationships between amylopectin internal molecular structure
and physicochemical properties of starch. Trends in Food Science &
Technology, 78, 234-242. https://doi.org/10.1016/j.tifs.2018.05.024
Zhu, F., Bertoft, E., & Seetharaman, K. (2013). Characterization of internal
structure of maize starch without amylose and amylopectin separation.
Carbohydr Polym, 97(2), 475-481.
https://doi.org/10.1016/j.carbpol.2013.04.092

69
Popular science summary
A good understanding of the relationship between molecular structure and
functional properties of potato starch can help plant breeders produce
tailor-made starches with desired properties at the genetic level. These
starches can be used in various food applications, with beneficial effects on
human health and wellbeing.
Potato tubers are an important part of the human diet in the Nordic
countries and in many other countries worldwide. The starch contained in
potato tubers also has many valuable applications in food and non-food
industries, depending on the qualities of the starch. Starch can have different
properties depending on the ratio of amylose to amylopectin, granular size
distribution and the molecular structure of the individual starch components.
Native starch has some drawbacks when it is used in manufactured foods.
In order to overcome these drawbacks, many physical and/or chemical
modifications of native starch are applied in food processing. However,
modification of starch in this way requires much money, time and energy
and is labour- and chemicals-intensive. Efforts are underway to make starch
production and downstream processing more sustainable. One option could
be to use modern breeding technologies to develop new starch qualities in
crops such as potato and cereals.
This thesis evaluated a new potato line that contains a high content of
amylose. The results showed that cooked tubers of this high-amylose potato
line have a three-fold higher level of resistant starch than cooked tubers of
the parental potato. Resistant starch is part of dietary fibre, which has well-
known health benefits for humans such as contributing to low glycaemic
index, weight loss and probiotic effects. The resistant starch content in the
high-amylose potato tubers was found to be increased further by one extra

71
day of cold storage after cooking, due to the unique amylopectin structure of
the starch.
Native potato starch usually contains about 25% amylose and this
amylose content cannot be changed through conventional plant crossings. In
general, potato starch composition can only be adjusted through molecular
genetic techniques. The high-amylose potato line examined in this thesis was
developed by down-regulating two starch branching enzymes, using RNA
interference or the genome editing technique CRISPR-Cas9. Apart from
increasing the amylose content, the amylopectin structure of potato starch
was also altered by this molecular genetic modification.
To study the structure of amylopectin in the new potato line, an optimised
method for determining the internal structure of starch was developed.
Results obtained using this method showed that the high-amylose potato
starch had a very large fraction of intermediate-sized outer and inner
amylopectin chains, and more intermediate and large building blocks, than
the native potato starch. The amylopectin in the new potato line was found
to have properties similar to amylose. After cooking and cold storage, it did
not split as easily as ordinary potato starch and it was resistant to digestion.
Using this structural knowledge of the unique amylopectin in the new potato
line, functional starches and healthier foods can be designed in future.
The results also showed that various parameters relating to the internal
structure of potato starch were significantly correlated with starch thermal
properties (gelatinisation and retrogradation).
In summary, increased understanding of the effect of genetic modification
on molecular structure and functional properties of starch is of great
importance. So as to custom starch at the genetic level with desired
functional properties for multiple applications, without need for further
chemical or physical modifications of starch. This would be a green
alternative, economically friendly and sustainable approach to develop novel
desirable starches in the near future.

72
72
Populärvetenskaplig sammanfattning
Genom att ytterligare förstå förhållandet mellan molekylstruktur och
funktionella egenskaper hos potatisstärkelse kan vi hjälpa växtförädlare att
skräddarsy stärkelse med önskade egenskaper på genetisk nivå för olika
livsmedelsapplikationer som kan ha en positiv inverkan på människors hälsa
och välbefinnande.
Potatisknölar bidrar till en stor del av den mänskliga kosten i Norden och
i resten av världen. Stärkelsen som produceras av denna gröda har också
många värdefulla tillämpningar inom livsmedels- och livsmedelsindustrin.
Stärkelsens egenskaper är mycket viktiga i både livsmedels- och icke-
livsmedelsapplikationer, men stärkelsen kan ha olika egenskaper beroende
på förhållandet mellan amylos och amylopektin, granulstorleksfördelning
och molekylstrukturen hos de enskilda stärkelsekomponenterna.
Nativ stärkelse har dock vissa nackdelar när den används i en
livsmedelsapplikation. För att övervinna nackdelarna görs fysiska och/eller
kemiska modifieringar av nativ stärkelse. Emellertid är de kemiska och
fysiska modifieringarna av stärkelse både penga-, tid- och energikrävande,
liksom arbetskraft och kemikalieintensiva. Fokus läggs på att göra
stärkelseproduktion och nedströmsprocess mer hållbar, och en del av en
lösning kan vara att använda modern förädlingssteknik för att utveckla nya
stärkelsekvaliteter i en gröda.
I doktorandprojektet utvärderade vi en ny potatis som har högt innehåll
av amylos. Vi fann att denna högamylospotatis ger en tre gånger högre nivå
av resistent stärkelse i de kokta knölarna jämfört med modersorten. Resistent
stärkelse är en kostfiber, som har kända hälsofördelar för människokroppar,
som att bidra till ett lågt glykemiskt index och viktminskning samt har en
probiotisk effekt. Halten resistent stärkelse ökade ytterligare efter en extra

73
dags kylförvaring efter kokning, som ett resultat av den unika
amylopektinstrukturen.
Nativ potatisstärkelse innehåller vanligtvis cirka 25 % amylos och
amylosinnehållet kan inte ändras genom konventionella korsningar.
Potatisstärkelsekompositionen kan endast påverkas genom
molekylärgenetiska tekniker. De potatisar med hög amylos som används i
studien har utvecklats genom nedreglering av två
stärkelseförgreningsenzymer med RNA-interferens eller
genomredigeringsmetoded CRISPR-Cas9. Förutom att öka amylosinnehållet
har amylopektinstrukturen också förändrats.
För att studera amylopektinstrukturen har en optimerad metod för att
bestämma stärkelsens interna struktur utvecklats inom doktorandprojektet.
Genom att tillämpa metoden undersöktes potatisstärkelsen med hög amylos
vilket visade på en mycket stora fraktion av mellanstora yttre och inre
amylopektinkedjor och mellanstora såväl som stora byggstenar jämfört med
den ursprungliga potatisstärkelsen. Detta unika amylopektin har egenskaper
som liknar amylos. Efter tillagning och kylförvaring delas den inte lika lätt
som vanlig potatisstärkelse och blir motståndskraftig mot matsmältningen.
Kunskapen om den unika amylopektinstrukturen är viktig för att utforma
funktionell stärkelse och hälsosammare mat i framtiden.
Våra resultat visade också att olika parametrar för potatisstärkelsens inre
struktur signifikant korrelerade med stärkelseegenskaperna (gelatinisering
och retrogradering.
Sammanfattningsvis är ökad förståelse för effekten av genetisk
modifiering på stärkelsens molekylära struktur och funktionella egenskaper
av stor betydelse. Förståelsen behövs för att kunna designa stärkelse med
önskade funktionella egenskaper för flera applikationer, utan behov av
ytterligare kemiska eller fysiska modifieringar. Detta skulle vara ett grönt
alternativ; ett ekonomiskt- och hållbart tillvägagångssätt, för att utveckla nya
önskvärda stärkelser inom en snar framtid.

74
74
Acknowledgements
I wish to thank a number of people who gave me a great deal of support, help
and assistance throughout my PhD journey.

Professor Roger Andersson, my main supervisor, I greatly appreciated

your excellent supervision of my PhD studies, from starting the project to the
end of writing this thesis. Your expertise was always invaluable and it guided
me to formulate research questions. Thank you very much also for the
inspiration every time I encountered difficulties during method development.
I really appreciate the numerous meetings with you, and your insightful
feedback. These all helped me to sharpen my academic thinking, supported
me to progress in research and encouraged me to be more professional. You
were an inspiring and important mentor; I really want to adopt your strong
pedagogic strategy and great attitude to research.

Associate professor Mariette Andersson, my deputy supervisor, I wish to

thank you very much for your supervision, all the discussions and great
support during project planning, experiment progression and academic
writing. Special appreciation for sharing the amazing potatoes and starch.
Moreover, I am indebted to you for always guiding me so positively, which
made me feel confident in the project.

My industrial mentor, Magdalena Bergh, I sincerely thank you for the

time you took to complete my questionnaires and interviews. Big thanks for
the involvement in my PhD studies and for helping me to learn about the
Swedish food industry.

75
 I am grateful to all my co-authors, especially Shishanthi Jayarathna, Helle
Turesson, Gustav Nestor and Per Hofvander, for their important
contributions and invaluable inputs to the work included in this thesis.

To all former and present technical personnel at the Department of

Molecular Sciences, particularly Gunnel Fransson, Janicka Nilsson, Annica
Andersson and Henrik Hansson, I thank you all for skilful technical
assistance with analyses in the laboratory.

Special thanks to Rikard Landberg and Maud Langton for contacting me

and giving me the opportunity to shift my research interest into food science.
To present (Jaana, Eva-Marie, Erica, Anna) and former administrators, my
thanks for your help in administration. Thanks to Vadim for all your
signatures and the advice about X-ray analysis.

Warm thanks to my wonderful group members (Per, Glab, Santanu),

colleagues (Albina, Harald, Galia, Hanna, Xinmei, Su-Lin, Simon, Monika,
Daniel, Sumitha, Jolanta, Tomas, Åse, Nils, Alyona, Ali, Bettina, Gustav,
Jana, Sabine, Mats, Corine, Jerry, Maria and Yong) and lovely PhD fellows
(Laura, Louise, Sunera, Elin, Mathilde, Mikolaj, Topi, Anja, Mathias,
Thomas, Ludwig, Giselle, Yashaswini, Klara, Jonas, Laura, Ebba, Solja,
Hasi, Abhijeet, Fredric and Yan) and friends (Bing, Tong, Jing, Yanhong,
Miao, Li, Zhenxing, Hui, Chen, Shujing, Qing, Jun, Wenjia, Guo-Zhen, He,
Fengping, Huaxing, Lin, Ken, Shengjie, Yayuan, Xi, Ming, Ai, Yichun,
Liaozhen, Yanjun, Ke, Xinyi and Jie), who shared my PhD journey with
companionship, colourfulness and joyfulness.

I also want to thank the research schools Focus on Food and Biomaterials,
LiFT, Organism Biology and Sustainable Biomass Systems for all the
important courses, useful workshops and inspiring study trips across
Scandinavia and the globe.

Thanks to Lennart Hjelm’s Fund, Edvard Nonnen’s Fund and the

Swedish Foundation for Strategic Environmental Research for supporting
me to participate in national and international conferences.

76
76
 Finally, I would like to express my deepest appreciation to my beloved
family for all their love, understanding, patience and care over the years. I
thank my husband Hanqian Zhang who always encouraged me whenever I
appeared disappointed or depressed. To my little angel, my daughter Elsa,
your unconditional love makes me soft and kind. I also want to pay my
special respects to my parents, my younger sister and my parents-in-law, for
providing me with much invaluable practical help and care.

77
 !CTA 5NIVERSITATIS !GRICULTURAE 3UECIAE

Acta Universitatis Agriculturae Sueciae

Doctoral Thesis No. 2021:65 • Novel potato starch • Xue Zhao

Doctoral Thesis No. 2021:65

A simplified method was developed for determining the internal molecular structure of Doctoral Thesis No. 2021:65
whole starch without prior amylopectin isolation. Applying the method, varied internal Faculty of Natural Resources and Agricultural Sciences
structural parameters of potato starch were obtained from different genetic backgrounds.
Various internal structure parameters were found to affect the thermal properties of potato
starch. A dense structure of building blocks led to higher gelatinisation temperatures and Novel potato starch
enthalpy. Retrogradation was found to be favoured by more large building blocks and
New structure and beneficial qualities
many short internal chains.

Xue Zhao, the author of this thesis, conducted her PhD studies in Food Science at the Xue Zhao
Department of Molecular Sciences, SLU, Uppsala. She received her MSc and BSc degrees
in Animal Science from SLU, Sweden and Northwest A & F University, China, respectively.

Acta Universitatis Agriculturae Sueciae presents doctoral theses from the Swedish
University of Agricultural Sciences (SLU).

SLU generates knowledge for the sustainable use of biological natural resources. Research,
education, extension, as well as environmental monitoring and assessment are used to
achieve this goal.

Online publication of thesis summary: http://pub.epsilon.slu.se/

ISSN 1652-6880
ISBN (print version) 978-91-7760-807-3
ISBN (electronic version) 978-91-7760-808-0