Carbohydrate Polymers 59 (2005) 153–163

www.elsevier.com/locate/carbpol

Structural and physicochemical characteristics of winter squash

(Cucurbita maxima D.) fruit starches at harvest
David G. Stevensona,1, Sang-Ho Yooa,b, Paul L. Hurstc, Jay-lin Janea,*
a
Department of Food Science and Human Nutrition, Iowa State University, 2312 Food Sciences Building, Ames, IA 50011, USA
b
Department of Food Science and Technology, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143-747, South Korea
c
New Zealand Institute for Crop and Food Research Limited, Private Bag 11 600, Palmerston North, New Zealand
Received 26 April 2004; revised 19 August 2004; accepted 22 August 2004
Available online 11 November 2004

Abstract

Structures and physicochemical properties of starch isolated from fruit (pepo) of seven winter squash cultivars (Cucurbita maxima D.)
were studied. Squash starches exhibited the B-type X-ray diffraction pattern and had most granules with diameters of 1.5–2.5, 6–8, and
11–13 mm. Squash amylopectin had more long branch chains (25.4–29.3% DPR37), and squash starch had relatively small absolute amylose
contents (12.9–18.2%). Squash starch amylopectins had weight-average molecular weights ranging from 2.03 to 3.22!108 g/mol, gyration
radii of 294–337 nm, and unusually low polydispersity (1.21–1.81). Isoamylase-debranched amylopectins showed average chain-length
varied from DP 26.5 to 28.1. Squash starches had onset gelatinization temperature ranging from 60.6 to 63.5 8C and enthalpy changes (DH)
from 15.9 to 17.4 J/g. Retrogradation rate for squash starches ranged from 41 to 55% after 7 days at 4 8C. Squash starch pastes, measured by
using a Rapid Visco-Analyzer, had peak viscosity of 174–233 RVU, final viscosity of 193–244 RVU, and setback of 79–100 RVU, with
pasting temperature of 65.6–68.8 8C.
q 2004 Elsevier Ltd. All rights reserved.

Keywords: Starch; Starch structure; Starch function; Amylose; Amylopectin; Winter squash; Pumpkin; Cucurbit

1. Introduction et al., 2001; Corrigan, Irving, & Potter, 2000; Culpepper &
Moon, 1945; Cumarasamy, Corrigan, Hurst, & Bendall,
Starch is the major carbohydrate in plant storage organs. 2002; Daniel, Brecht, Sims, & Maynard, 1995; Harvey,
Starch structures and physicochemical properties have Grant, & Lammerink, 1997; Hurst et al., 1995; Merrow &
been characterized for storage organs of many plant Hopp, 1961; Schales & Isenberg, 1963; Smittle, Hayes, &
families. However, Cucurbitaceae, which includes squash Williamson, 1980).
and pumpkins, have had little published results on fruit Sugimoto, Yamashita, Oomori, et al. (1998), investigated
(pepo) starch characteristics, despite the fruit dry matter fruit starch properties of one cultivar each of Cucurbita
starch content ranging from !3% (Corrigan, Hurst, & maxima D. and C. moschata D. during development. Squash
Potter 2001; Schales & Isenberg, 1963) to O60% (Hurst, starch was shown to exhibit a B-type X-ray pattern, granule
Corrigan, Hannan, & Lill, 1995; Merrow & Hopp, 1961). diameter of 2.9–8.8 mm, amylose content ranging from 14 to
Studies have shown that texture of squash fruit is related to 23%, and onset gelatinization temperature of 62–66 8C.
its starch content and possibly starch properties (Corrigan Starch granules were also reported to have a larger diameter
range (3–35 mm) for the same C. maxima D. cultivar
(Yoshida, 1989). Starch content and granule size variation
* Corresponding author. Tel.: C1 515 294 9892; fax: C1 515 294 8181. between external and internal cellular layers of the fruit
E-mail address: jjane@iastate.edu (J.-l. Jane). were also reported (Yoshida, 1989).
1
Present address: Cereal Products and Food Science Unit, National
Center for Agricultural Utilization Research, Agricultural Research
There are few cultivars of squash with limited starch
Service, United States Department of Agriculture, 1815 N. University properties characterized. To our knowledge, there has been
Street, Peoria, IL 61604, USA. no report on starch retrogradation, phosphorus content
0144-8617/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbpol.2004.08.030
154 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

of starch, absolute amylose content, and amylopectin with ethanol, and then recovered by filtration using What-
structures such as molecular size, gyration radii, man No. 4 filter paper. Purified starch cake was dried in a
and amylopectin branch chain-length distribution in any convection oven at 35 8C for 24 h.
Cucurbita L. sp. starch. In this study, we characterized
structures and physicochemical properties of starch isolated 2.3. Starch granule morphology
from fruit of seven cultivars of C. maxima D., grown in the
same year, at the same location and using identical Starch granules, spread on silver tape and mounted on a
procedures. Structures and physicochemical properties of brass disk, were coated with gold/palladium (60/40) for all
starch will be related to their textural and eating qualities five replicates of each cultivar. Sample images were
reported elsewhere. observed at 1500! magnification under a scanning electron
microscope (JOEL model 1850, Tokyo, Japan), following
the method of Jane, Kasemsuwan, Leas, Zobel, and Robyt
2. Materials and methods (1994).

2.1. Plant material

2.4. Crystalline structure
Seven squash cultivars were planted in summer of 1998
Crystallinity of starch granules was studied using X-ray
at an Iowa State University farm site 1.7 miles south of
diffractometry. X-ray diffraction patterns were obtained
Ames, Iowa (geographical location 41858 0 57.5 00 N,
with copper, Ka radiation using a Siemens D-500
93838 0 22.9 00 W), in a completely randomized block (3.05!
diffractometer (Siemens, Madison, WI). Analysis was
3.05 m blocks) with 18 replicates (four plants/replicate).
conducted following the procedure of Song and Jane
Normal crop husbandry was followed as required. Five
(2000). Degree of crystallinity was calculated based on
replicates of each cultivar were randomly selected for
the method of Hayakawa, Tanaka, Nakamura, Endo, and
analysis of starch characteristics. Squash cultivars studied,
Hoshino (1997). The following equation was used to
with range of fruit weight observed in parentheses, were
determine percent crystallinity:
three buttercups [Delica (0.79–2.68 kg), Kurijiman
(0.75–2.13 kg) and Sweet Mama (1.61–3.25 kg)], Crystallinity ð%Þ Z Ac =ðAc C Aa Þ100
one Halloween-type [Prizewinner (13.7–70.2 kg)], one
where Ac is the crystalline area on the X-ray diffractogram
Hubbard-type [Warren Scarlet (2.70–8.12 kg)], one
and Aa is the amorphous area on the X-ray diffractogram.
Crown-type [Whangaparoa Crown (1.82–6.03 kg)] and
one Native American Indian squash [Lakota (1.24–
3.78 kg)] (Coyne, Reiser, Sutton, & Graham, 1995). 2.5. Molecular weight distribution and gyration
Pictures of winter squash fruits for all the cultivars have radius of amylopectin
been reported previously (Stevenson, 2003). Seeds of
Kurijiman, Warren Scarlet and Whangaparoa Crown were Weight-average molecular weight and z-average gyra-
purchased from Webling and Stewart Ltd, Petone, New tion radius of amylopectin were determined using high-
Zealand; Delica seeds from Yates New Zealand Ltd, performance size-exclusion chromatography equipped with
Onehunga, New Zealand; Prizewinner seeds from King’s multi-angle laser-light scattering and refractive index
Seeds, Auckland, New Zealand; Sweet Mama seeds from detectors (HPSEC-MALLS-RI). Starch samples, duplicate
Henry Field Seed and Nursery Co., Shenandoah, IA, and measurements of each replicate, were prepared as described
Lakota seeds from W. Atlee Burpee and Co., Warminster, by Yoo and Jane (2002a). The HPSEC system consisted of a
PA. Squash fruit maturity was adjudged when stalks became HP 1050 series isocratic pump (Hewlett Packard, Valley
woody (Hawthorne, 1990), and this stage had been Forge, PA), a multi-angle laser-light scattering detector
previously shown to have the highest starch content (Irving, (Dawn DSP-F, Wyatt Tech. Co., Santa Barbara, CA) and a
Hurst, & Ragg, 1997). HP 1047A refractive index detector (Hewlett Packard,
Valley Forge, PA). To separate amylopectin from amylose,
2.2. Starch isolation Shodex OH pak KB-G guard column and KB-806 and
KB-804 analytical columns (Showa Denko K.K., Tokyo,
Starch was isolated from squash fruit by using the Japan) were used. Operating conditions and data analysis
method reported by Badenhuizen (1964) with slight are described by Yoo and Jane (2002b).
modifications (Kasemsuwan, Jane, Schnable, Stinar, &
Robertson, 1995). Two fruits per replicate, peeled and 2.6. Phosphorus content
deseeded, were used for starch isolation. Squash pulp was
ground in 0.01 M HgCl2 and then filtered through 106 mm The phosphorus content of the starch was determined
mesh. The starch residue was washed with 10% toluene in following the method of Smith and Caruso (1964) except
0.1 M NaCl, washed three times with distilled water, twice that 5 g of starch was used and all glassware was soaked
 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163 155

24 h in 0.625% sodium molybdate (w/v) in 1.75 N and sealed. The sample was equilibrated for 2 h and scanned
H2SO4:10% ascorbic acid (w/v) mixture (4:1) to remove at a rate of 10 8C/min over a temperature range of 10–
residual phosphorus. Duplicate analysis of each replicate 100 8C. An empty pan was used as the reference. The rate of
was conducted. starch retrogradation was determined using the same
gelatinized samples, stored at 4 8C for 7 d, and analyzed
2.7. Amylose contents using DSC as described previously (White, Abbas, &
Johnson 1989). Analysis of starch thermal properties was
Apparent amylose was determined following the carried out in triplicate for each replicate of each cultivar.
method of Takeda and Hizukuri (1987), and the iodine
affinity was determined by using a potentiometric 2.10. Starch pasting properties
autotitrator (702 SM Titrino, Brinkmann Instrument,
Westbury, NY). The absolute amylose content of starch Starch pasting properties were analyzed using a
was calculated by subtracting iodine affinity of amylo- Rapid Visco-Analyser (RVA-4, Newport Scientific, Syd-
pectin fraction from that of the defatted whole starch, ney, Australia) (Jane et al., 1999). Starch suspension (8%,
following the procedure of Lu, Jane, Keeling, and w/w, dsb, 28 g total weight) was equilibrated at 30 8C for
Singletary (1996). Amylose content of a 90% dimethyl 1 min, heated at a rate of 6.0 8C/min to 95 8C, maintained
sulfoxide (DMSO)-defatted starch sample was also at that temperature for 5.5 min, and then cooled to 50 8C
determined by treating the starch sample with concana- at a rate of 6.0 8C/min. Constant paddle rotating speed
valin A (Yun & Matheson, 1990) and measuring glucose (160 rpm) was used throughout entire analysis, except for
content, by the glucose oxidase method, of the supernatant rapid stirring at 960 rpm for first 10 s to disperse sample.
using amylose/amylopectin assay kit (Megazyme Inter- Analysis was in duplicate for each replicate of each
national Ireland Ltd, Wicklow, Ireland). Determination of cultivar.
amylose content, by both methods, was duplicated for
each squash cultivar replicate. 2.11. Data analysis

2.8. Amylopectin branch chain-length distribution All statistical significance tests were calculated by using
SAS (SAS Institute Inc., 1999) and applying Tukey
Amylopectin was fractionated using n-butanol as difference test (Ramsey & Schafer, 1996). Correlations
a complexing agent (Schoch, 1942). Amylopectin between squash starch structures and functional properties
(10 mg/ml) was debranched using isoamylase (EC 3.2.1.68 were conducted using SAS (1999) and the PROC CORR
from Pseudomonas amyloderamosa) (EN102, Hayashibara function specifying use of the Pearson correlation coeffi-
Biochemical Laboratories Inc., Okayama, Japan), following cient. A 5% level of significance was used to discriminate
the method of Jane and Chen (1992). Branch chain-length correlations of importance. Means of the squash cultivars
distribution of amylopectin was determined by using an were correlated, with nZ6 or 7 depending on whether
HPAEC system (Dionex-300, Sunnyvale, CA) equipped Prizewinner provided sufficient starch for the analysis.
with an on-line amyloglucosidase (EC 3.2.1.3, from
Rhizopus mold, A-7255, Sigma Chemical Co., St Louis,
MO) post-column reactor and a pulsed amperometric 3. Results and discussion
detector (HPAEC-ENZ-PAD) (Wong & Jane, 1997).
PA-100 anion exchange analytical column (250!4 mm, 3.1. Starch granule morphology
Dionex, Sunnyvale, CA) and a guard column were used for
separating debranched amylopectin samples. Gradient pro- Scanning electron micrographs show that squash starches
file of eluents and operating conditions were described of all cultivars have granules ranging mainly from 1.5–2.5,
previously (McPherson & Jane, 1999). Debranched amylo- 6–8 and 11–13 mm in diameter (Fig. 1). Squash starches
pectin was also analyzed by using a HPSEC equipped with a exhibited a mixture of spherical and polyhedral granules with
RI detector to determine the presence of extra-long branch some dome-shaped granules that had larger diameters on
chains (Yoo & Jane, 2002a). HPAEC-ENZ-PAD and HPSEC average. Yoshida (1989) reported that starch granules in the
analysis was duplicated for each replicate of each cultivar. endocarp region of squash fruit flesh had compound starch
granules, which was not found in the remainder of the fruit.
2.9. Thermal properties of starch The dome-shaped indented starch granules in our study were
more likely resulted from shrinkage of swollen granules than
Thermal properties of starch were determined by using a from compound starch. Squash starch granules displayed
differential scanning calorimeter (DSC-7, Perkin-Elmer, many surface indentations. Large granules of Whangaparoa
Norwalk, CT) (Jane et al., 1999). Approximately 2 mg of Crown (Fig. 1G), Lakota (Fig. 1C), Prizewinner (Fig. 1D)
starch (dry starch basis, dsb) was precisely weighed in an and Sweet Mama (Fig. 1E) had less incidence of indenta-
aluminum pan, mixed with 6 mg of deionized water tions, whereas Delica (Fig. 1A), Kurijiman (Fig. 1B)
156 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

Fig. 1. Scanning electron micrographs of Delica (A), Kurijiman (B), Lakota (C), Prizewinner (D), Sweet Mama (E), Warren Scarlet (F) and Whangaparoa
Crown (G) squash fruit starches (scale barZ10 mm).

and Warren Scarlet (Fig. 1F) frequently showed indentations. diameters up to 35 mm, but we did not observe granule
Whangaparoa Crown and Prizewinner had a high incidence diameter above 14 mm.
of indentations in medium-size starch granules. Granule
indentations could be due to non-uniform growth within 3.2. Crystalline structure
starch granules or collapse during drying. The granule size
distribution observed in our study was in agreement with that Squash starches all exhibited B-type X-ray diffraction
reported by Sugimoto, Yamashita, Suzuki, Morishita, and patterns (Fig. 2). Prizewinner, Warren Scarlet, Whanga-
Fuwa (1998b). Yoshida (1989) reported starch granule paroa Crown, Kurijiman, and Delica all had an additional
 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163 157

(35–55.7%) is greater than other native starches, such as

20–28% reported for a variety of A- and B-type starches
(Cooke & Gidley, 1992), 25–44% reported for wheat
starches (Fujita, Yamamoto, Sugimoto, Morita, & Yama-
mori, 1998), and 38% for normal maize (Keppel, 2001).

3.3. Amylose content

Iodine affinities for defatted whole starch and amylo-

pectin, and corresponding apparent and absolute amylose
contents are shown in Table 1. Both apparent and absolute
amylose contents were significantly different between
the starches of squash cultivars. Iodine affinities of the
squash amylopectin were greater than that of most other
native starches, indicating more long-branch chains of the
amylopectin, which resulted in substantially less absolute
amylose contents than the apparent amylose content
counterparts.
Concanavalin A method resulted in less amylose
contents than the absolute amylose content counterparts
obtained by the iodine-affinity method, but both sets of
results were in the same trend. The lower amylose content
obtained by using concanavalin A method could be due to
Fig. 2. X-ray diffraction patterns of Delica, Kurijiman, Lakota, Prizewinner, co-precipitation of amylose with amylopectin, resulting in
Sweet Mama, Warren Scarlet and Whangaparoa Crown squash fruit amylose lost when amylopectin was removed by
starches. centrifugation.
peak at 2qZ28.28 that was not observed in other cultivars.
The nature of the peak is not known. The percentage 3.4. Phosphorus content
crystallinity of Prizewinner, Warren Scarlet, Whangaparoa
Crown, Lakota, Kurijiman, Sweet Mama and Delica squash Phosphorus contents of squash starches ranged
starches, calculated based on X-ray diffractograms, was from 0.022 (Whangaparoa Crown) to 0.026% (Lakota)
48.0, 54.0, 42.9, 43.9, 55.7, 45.7 and 35.0, respectively. (w/w, dsb) and were not significantly different
The three buttercup cultivar squash starches, Delica, between cultivars (other data are not shown). Phosphorus
Kurijiman and Sweet Mama, showed different degrees of contents of all squash starches were higher than that
crystallinity. The percentage crystallinity of squash starch reported for all tuber and root starches, except potato

Table 1
Iodine affinities, apparent amylose, absolute amylose contents, and amylose content measured using concanavalin A (Con A) for squash fruit defatted starches

Cultivar Iodine affinity Apparent amylose Absolute amylose Amylose (%) content
content (%)a content (%)b measured by Con Ac
Whole starch Amylopectin fraction
Delica 5.51ab 1.89ab 27.8ab 18.2 12.4
Kurijiman 5.63a 2.35ab 28.3a 16.5 12.6
Lakota 4.45b 1.65b 22.3b 14.0 10.2
Prizewinner 5.31ab 2.73a 26.4ab 12.9 7.9
Sweet Mama 5.02ab 2.33ab 24.9ab 13.2 9.1
Warren Scarlet 5.41ab 1.84ab 27.1ab 17.9 11.4
Whangaparoa Crown 5.48ab 1.96ab 27.9ab 18.0 11.8
PZ0.03d PZ0.02 PZ0.03 PZ0.007e PZ0.07

Values with different letters denote differences at the 5% level of significance for each comparison between cultivars in the respective column.
a
Apparent amylose contents were averaged from two analyses for each of five replicates. Values were calculated from dividing iodine affinity by a factor of
0.199.
b
Absolute amylose contents were averaged from two analyses for each of five replicates. Values were calculated by subtracting iodine affinity for the
amylopectin fraction from the iodine affinity for the whole starch, divided by a factor of 0.199.
c
Amylose contents measured by concanavalin A were averaged from two analyses for each of five replicates.
d
P represents the probability of F-statistic exceeding expected for each comparison between cultivars in the respective column.
e
Conservative Tukey test was unable to detect differences between individual cultivars.
158 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

(Jane, Kasemsuwan, Chen, & Juliano, 1996; Kasemsuwan chain-length at DP 13–14. The chain-length distribution of
& Jane, 1996; Lim, Kasemsuwan, & Jane, 1994). amylopectin is summarized in Table 3. Average amylopectin
chain lengths varied from DP 26.5 to 28.1 for the squash
3.5. Amylopectin molecular weight and size cultivars and were not significantly different. However,
proportions of branch chains of different chain-lengths were
Average molecular weight, gyration radii, and dispersed significantly different (Table 3). The different chain-length
density of squash starches are shown in Table 2. Squash distributions correlated with properties of the starches as
starch weight-average molecular weight (Mw) ranged from discussed in Section 3.9. The greater contents of short branch
2.03 to 3.22!108 g/mol for the seven squash cultivars; chains (DP 6–12, 12.6–16.7%) of squash amylopectin than
there was no significant difference between cultivars. that of other B-type amylopectin (8.5–12.3%) (Jane et al.,
Squash amylopectin M w is comparable with other 1999) may contribute to the greater density of dispersed
starches of the B-type polymorphism, ranging from 1.7 to squash amylopectin. HPSEC chromatograms of debranched
3.4!108 g/mol (Yoo & Jane, 2002b). Squash amylopectins squash amylopectin showed only two distinct peaks of short
displayed substantially lower polydispersity (Mw/Mn) than and long branch chains; there were no extra-long branch
did amylopectin of other sources (Stevenson, 2003). The chains (DPw720), as in cereal amylopectin (Yoo & Jane,
polydispersity of Lakota, Prizewinner, Whangaparoa Crown 2002a) detected (Stevenson, 2003).
and Sweet Mama amylopectin ranged between 1.21 and
1.30 (Table 2), indicating highly uniform in size of the 3.7. Thermal properties
amylopectin molecules. Gyration radius of amylopectin was
not significantly different between the squash cultivars. Thermal properties of native squash starches are shown
Prizewinner and Whangaparoa Crown amylopectins, dis- in Table 4. Three buttercup squash cultivars, Delica,
persed in aqueous solutions, were somewhat denser than Kurijiman and Sweet Mama, showed similar onset gelati-
other squash cultivars, but differences were not significant nization temperatures (To), which were higher than others.
(Table 2). The density of the dispersed squash amylopectin The ranges of gelatinization temperature (TcKTo) (ROG) of
(10.1–12.2 g/mol/nm3) was larger than that of the B-type squash starches were the smallest comparing with other
non-waxy amylopectin (2.9–5.3 g/mol/nm 3) but was starches (Jane et al., 1999). Biliaderis, Maurice, and Vose
comparable to that of the ae-waxy maize amylopectin (1980) suggested that the greater the degree of amylopectin
(11.2 g/mol/nm3) (Yoo & Jane, 2002b). branching, the wider the melting temperature range. Our
results demonstrated a negative relationship between the
3.6. Amylopectin branch chain-length distribution ROG and the long branch chains of DPR37. The reduction
in ROG by increasing number of branch chains of DPR37
HPAEC-ENZ-PAD chromatogram of debranched amy- can be attributed to cooperative gelatinization induced by
lopectin of each squash cultivar is shown in Fig. 3. the long branch chains that extended through multiple
Chromatograms show that all squash starches have peak clusters. Enthalpy changes of gelatinization for all squash
starches were similar to that of other B-type starches (Jane
Table 2 et al., 1999), and there were no significant differences
Average amylopectin molecular weight, polydispersity, gyration radius and between cultivars.
density of squash fruit starches Squash starches, after being stored at 4 8C for 7 days,
Cultivara Mw!108 Polydisper- Rz r showed greater degrees of retrogradation than other starches
(g/mol)b sity (Mw) (nm)c (g/mol/nm3)d (Table 5) (Jane et al., 1999). Among the squash starch
Delica 2.41 1.44 310 8.1 cultivars, Lakota, Sweet Mama and Warren Scarlet
Kurijiman 2.03 1.81 294 8.0 displayed the greatest degrees of retrogradation. Prizewin-
Lakota 3.16 1.21 324 9.3 ner had the lowest percentage retrogradation, which
Prizewinner 3.22 1.23 317 10.1
agreed with its largest content of short branch chains
Sweet Mama 2.54 1.30 311 8.4
Warren Scarlet 2.91 1.35 337 7.6 (16.7% DP 6–12 and 1% DP 3–5).
Whangaparoa 2.70 1.29 304 9.6
Crown 3.8. Pasting properties
PZ0.12e PZ0.27 PZ0.11 PZ0.15

Data were obtained from two injections each of all five replicates. Pasting properties of squash starches are shown in Table 6.
a
Starch samples were dissolved in 90% DMSO solution and precipitated Peak viscosity of Lakota and Prizewinner starches were
with 5 vol. ethanol. Freshly prepared starch aqueous solution (100 ml; higher than that of others. Pasting temperature of squash
0.8 mg/ml) was injected to HPSEC system. starches was only 4–6 8C higher than respective To, which
b
Weight-average molecular weight.
c
z-Average radius of gyration.
could be attributed to the absence of phospholipids and the
d
Density is equal to Mw =R3z . low amylose content, facilitating the granule swelling after
e
P represents the probability of F-statistic exceeding expected for each gelatinization. The final viscosity of the squash starch was
comparison between cultivars in the respective column. similar to that of potato starch, but was higher than that of
 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163 159

Fig. 3. Relative peak area distributions of Delica, Kurijiman, Lakota, Prizewinner, Sweet Mama, Warren Scarlet and Whangaparoa Crown squash fruit
amylopectins analyzed by using a HPAEC-ENZ-PAD. Error bars represent standard error of the mean for each individual DP from two analyses of five
replicates. DP, degree of polymerization.

Table 3
Branch chain-length distributions of squash fruit amylopectins

Cultivar Peak DP Average Percent distribution Highest

CL detectable
I II DP 3–5 DP 6–9 DP 6–12 DP 13–24 DP 25–36 DPR37
DP
Delica 13 50 28.1 0.61 5.0b 14.9a 40.6b 14.7abc 29.3 72
Kurijiman 13 48 27.4 1.25 5.4ab 15.6a 41.0ab 14.2bc 28.4 76
Lakota 14 49 27.0 0.80 3.8c 12.6b 44.7a 16.5a 25.4 69
Prizewinner 13 49 26.5 1.01 6.1a 16.7a 41.9ab 14.1c 26.2 76
Sweet Mama 13 48 27.9 1.00 5.3ab 15.2a 40.1b 15.0abc 28.8 74
Warren Scarlet 14 49 26.9 1.43 5.4ab 15.4a 40.4b 16.0ab 27.1 68
Whangaparoa 13 49 27.2 0.74 5.5ab 15.9a 41.2ab 14.6bc 27.3 72
Crown
PZ0.26a PZ0.46 P!0.0001 P!0.0001 PZ0.01 PZ0.003 PZ0.13

Grouping of degree of polymerization (DP) numbers followed that of Hanashiro, Abe, and Hizukuri (1996). Values with different letters denote differences at
the 5% level of significance for each comparison between cultivars in the respective column.
a
P represents the probability of F-statistic exceeding expected for each comparison between cultivars in the respective column.
160 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

Table 4
Thermal properties of native squash fruit starches

Cultivara Native starch

To (8C) Tp (8C) Tc (8C) Range (8C)b DH (J/g)
Delica 63.4a 66.5a 69.7ab 6.3b 17.3
Kurijiman 63.5a 66.5a 69.8a 6.3b 16.9
Lakota 62.9ab 66.4a 70.4a 7.5ab 16.8
Prizewinner 60.9bc 65.0ab 69.7a 8.8a 15.9
Sweet Mama 63.5a 66.2a 69.8a 6.3b 17.4
Warren Scarlet 61.7abc 65.7ab 70.4a 8.7a 16.4
Whangaparoa Crown 60.6c 64.0b 67.7b 7.1ab 16.3
PZ0.0001c PZ0.001 PZ0.006 P!0.0001 PZ0.26

Values were calculated from three analyses for each of five replicates. Values with different letters denote differences at the 5% level of significance for each
comparison between cultivars in the respective column.
a
Starch samples (w2.0 mg, dsb) and deionized water (w6.0 mg) were used for the analysis; To, Tp, Tc and DH are onset, peak, conclusion temperature, and
enthalpy change, respectively.
b
Range of gelatinization is equal to TcKTo.
c
P represents the probability of F-statistic exceeding expected for each comparison between cultivars in the respective column.

most A-type starches (Jane et al., 1999). Setback viscosity of But it is positively correlated to onset gelatinization
all squash starches is higher than that of most other starches, temperature of retrograded starch (ToR) (rZ0.73, PZ0.05),
with exceptions of cattail millet, mungbean and green leaf which reflected the participant of amylose in the retro-
canna starches (Jane et al., 1999). High setback indicates that gradation and co-crystallized with amylopectin. Absolute
amylose forms network with amylopectin. amylose content was positively correlated to amylose
content measured by concanavalin A, but negatively
correlated to pasting peak viscosity, breakdown and final
3.9. Structure–property correlations
viscosity. The negative relationship between the apparent
amylose content and peak and final viscosity is in agreement
Correlation coefficients among selected squash starch
with previously reported (Blennow, Bay-Smidt, & Bauer,
structures and functional properties are shown in Table 7.
2001; Collado & Corke, 1997; Jane et al., 1999; Kuno,
Correlation coefficients are given in the text when not
Kainuma, & Takahashi, 2000; Wang, White, & Pollak,
included in Table 7. Apparent amylose content of squash
1993).
starches was negatively correlated to amylopectin chain
Squash amylopectin Mw was negatively correlated to
lengths of DP 13–24 and DP 25–36 and pasting peak
polydispersity, average amylopectin branch chain-length,
viscosity, breakdown, final viscosity and setback viscosity.
and amylopectin branch chains of DPR37, but positively
Table 5 Table 6
Thermal properties of retrograded squash fruit starches Pasting properties of squash fruit starches measured by Rapid Visco-
Cultivara Native starch Analyzer
To (8C) Tp (8C) Tc (8C) DH (J/g) % Retro- Cultivara Peak Break- Final Set- Pasting
gradation viscosityb downb viscosityb backb tempera-
Delica 36.3 54.0 64.5 8.1ab 46.6ab ture (8C)
Kurijiman 36.6 54.5 64.8 7.8ab 45.8ab Delica 174.8b 52.2 206.9 84.3 68.8
Lakota 35.2 53.3 65.4 8.9a 53.3a Kurijiman 179.2ab 63.1 195.3 79.2 67.8
Prizewinner 36.6 54.8 64.3 6.5b 40.8b Lakota 232.8a 88.7 243.8 99.7 67.2
Sweet 35.1 52.4 65.0 9.5a 54.7a Prizewinner 225.1a 76.7 244.3 95.8 66.5
Mama Sweet Mama 205.2ab 66.6 231.6 93.0 68.4
Warren 36.0 52.6 65.5 8.8ab 53.4a Warren 176.8b 52.6 217.5 93.2 68.0
Scarlet Scarlet
Whanga- 35.6 53.3 65.0 8.2ab 49.8ab Whangaparoa 173.7b 65.4 193.1 84.6 65.6
paroa Crown
Crown PZ0.01c PZ0.26 PZ0.49 PZ0.70 PZ0.16
PZ0.25b PZ0.31 PZ0.59 PZ0.008 PZ0.002
Values with different letters denote differences at the 5% level of
Values with different letters denote differences at the 5% level of significance for each comparison between cultivars in the respective
significance for each comparison between cultivars in the respective column.
column. a
8% (w/w) starch suspension.
a b
Same starch samples after gelatinization (see Table 4) were left for Viscosity measured in Rapid Visco-Analyzer units (RVU), 1 RVUZ12
7 days at 4 8C and rescan using DSC. centipoise.
b c
P represents the probability of F-statistic exceeding expected for each P represents the probability of F-statistic exceeding expected for each
comparison between cultivars in the respective column. comparison between cultivars in the respective column.
Table 7
Correlation coefficients (r!100) for selected squash starch structural and functional properties

ApA AbA Acon Mw PMw Rz DS CL DP13–24 DP25–36 DPR37 To ROG DH PV BK FV SB

D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

ApA 100
AbA 68 100
Acon 55 89** 100
Mw K58 K44 K68 100
PMw 59 39 65 K86** 100
Rz K44 K8 K33 73* K61 100
DS 11 K25 K41 38 K38 K29 100
CL 17 28 43 K72* 34 K44 K44 100
DP13–24 K72* K39 K24 57 K36 22 21 K46 100
DP25–36 K69 K1 0 46 K42 74* K53 K13 49 100
DPR37 57 39 46 K85** 56 K52 K36 87** K80* K46 100
To K21 K9 26 K59 49 K25 K72* 72* K6 14 53 100
ROG K14 K17 K47 81* K51 74* 27 K89** 25 25 K74* K72* 100
DH K15 5 31 K61 32 K28 K64 92** K23 13 69 91** K84* 100
PV K84* K90** K82* 68 K55 30 27 K44 72* 28 K67 K1 34 K17 100
BK K78* K75* K59 58 K44 6 41 47 87** 25 K75* K11 21 K25 89** 100
FV K80* K83* K87** 76* K66 57 9 K40 51 39 K58 K1 48 K14 92** 66 100
SB K83* K65 K76* 88** K81* 77* 5 K49 55 62 K71 K19 61 K25 81* 62 93** 100
PT K2 8 22 K45 35 11 K83* 66 K37 18 62 83* K40 78* K19 K47 3 K10

ApA, apparent amylose content; AbA, absolute amylose content; Acon, amylose content measured by concanavalin A; Mw, weight-average amylopectin molecular weight; PMw, polydispersity (Mw/Mn); Rz, gyration radius; DS,
density; CL, average amylopectin branch chain-length; DP13–24, proportion of amylopectin branch chain-lengths DP 13–24; DP25–36, proportion of amylopectin branch chain-lengths DP 25–36; DPR37, proportion of
amylopectin branch chain-lengths DPR37; DPR26, proportion of amylopectin branch chain-lengths DPR26 measured by HPSEC; To, onset gelatinization temperature; ROG, range of gelatinization temperature; DH,
enthalpy change of gelatinization; PV, peak viscosity; BK, breakdown; FV, final viscosity; SB, setback; PT, pasting temperature. *0.05 and **0.01 level of significance.

161
162 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163

correlated to Rz, ROG, final viscosity, and setback viscosity. References

The negative correlation observed between Mw and average
amylopectin branch chain-length and proportion of branch Badenhuizen, N. P. (1964). General method for starch isolation. In R. L.
chains of DPR37 agreed with findings reported by Yoo and Whistler, R. J. Smith, J. N. BeMiller, & M. L. Wolform (Eds.), Methods
in carbohydrate chemistry: Starch, 4 (pp. 14–15). London: Academic
Jane (2002b). The correlations indicated that squash
Press, 14–15.
starches with smaller amylopectin molecular weights and Biliaderis, C. G., Maurice, T. J., & Vose, J. R. (1980). Starch gelatinization
Rz possessed more long branch chains of DPR37 and, thus, phenomena studied by differential scanning calorimetry. Journal of
displayed narrower ROG. The negative correlation between Food Science, 45, 1669–1674.
the Mw and polydispersity indicated that reduction in the Blennow, A., Bay-Smidt, A. M., & Bauer, R. (2001). Amylopectin
molecular weight of amylopectin was resulted from aggregation as a function of starch phosphate content studied by size
exclusion chromatography and on-line refractive index and light
molecular breakdown.
scattering. International Journal of Biological Macromolecules, 28,
Average amylopectin branch chain-length was positively 409–420.
correlated to To (Jane et al., 1999; Li, Vasanthan, Rossnagel, Collado, L. S., & Corke, H. (1997). Properties of starch noodles as affected
& Hoover, 2001) and DH (Tang, Ando, Watanabe, Takeda, by sweetpotato genotype. Cereal Chemistry, 74, 182–187.
& Mitsunaga, 2001), but negatively correlated to ROG. The Cooke, D., & Gidley, M. J. (1992). Loss of crystalline and molecular order
percentage of branch chains of DP 13–24 was positively during starch gelatinisation: Origin of the enthalpic transition.
Carbohydrate Research, 227, 103–112.
correlated to peak viscosity and breakdown, and that of
Corrigan, V. K., Hurst, P. L., & Potter, J. F. (2001). Winter squash
branch chains of DP 25–36 was positively correlated to TcR (Cucurbita maxima) texture: Sensory, chemical, and physical measures.
(rZ0.86, PZ0.01), percent retrogradation (rZ0.77, PZ New Zealand Journal of Crop and Horticultural Science, 29, 111–124.
0.04) and Rz. Long amylopectin branch chains (DPR37) Corrigan, V. K., Irving, D. E., & Potter, J. F. (2000). Sugars and sweetness
were negatively correlated to peak viscosity, breakdown, in buttercup squash. Food Quality and Preference, 11, 313–322.
and setback viscosity, which agree with previous reports Coyne, D. P., Reiser, J. M., Sutton, L., & Graham, A. (1995). Lakota winter
squash, a cultivar derived from Native American Sources in Nebraska.
(Han & Hamaker, 2001; Li et al., 2001). However,
HortScience, 30, 1106–1107.
increasing peak viscosity with increasing amylopectin Culpepper, C. W., & Moon, H. H. (1945). Differences in the composition of
branch chain-length has also been reported (Sasaki & the fruits of Cucurbita varieties at different ages in relation to culinary
Matsuki, 1998; Shibanuma, Takeda, & Hizukuri, 1996). use. Journal of Agricultural Research, 71, 111–136.
These conflict results are attributed to that long amylopectin Cumarasamy, R., Corrigan, V., Hurst, P., & Bendall, M. (2002). Cultivar
chains can restrict swelling but also prevent breakdown of differences in New Zealand Kabocha (buttercup squash, Cucurbita
maxima). New Zealand Journal of Crop and Horticultural Science, 30,
swollen granules. To, Tp and DH were positively correlated
197–208.
to pasting temperature (rZ0.86, PZ0.01 for Tp) and ROG Daniel, A. L., Brecht, J. K., Sims, C. A., & Maynard, D. N. (1995). Sensory
was negatively correlated to DH. analysis of bush and vining types of tropical pumpkin. Proceedings of
the Florida State Horticulture Society, 108, 312–316.
Fujita, S., Yamamoto, H., Sugimoto, Y., Morita, N., & Yamamori, M.
(1998). Thermal and crystalline properties of waxy wheat (Triticum
aestivum L.) starch. Journal of Cereal Science, 27, 1–5.
4. Conclusion
Han, X. Z., & Hamaker, B. R. (2001). Amylopectin fine structure and rice
starch paste breakdown. Journal of Cereal Science, 34, 279–284.
Squash starches displayed B-type X-ray diffraction Hanashiro, I., Abe, J., & Hizukuri, S. (1996). A periodic distribution of the
patterns, consisted of low absolute amylose (12.9–18.2%) chain length of amylopectin revealed by high-performance anion-
and moderate phosphorus content (0.022–0.026%). exchange chromatography. Carbohydrate Research, 283, 151–159.
All squash starches had long amylopectin branch chain- Harvey, W. J., Grant, D. G., & Lammerink, J. P. (1997). Physical and
sensory changes during the development and storage of buttercup
lengths. Distinctive attributes of squash starches were low
squash. New Zealand Journal of Crop and Horticultural Science, 25,
polydispersity, narrow range of gelatinization, and high 341–351.
retrogradation rate. Squash starches had pasting temperature Hawthorne, B. T. (1990). Age of fruit at harvest influences incidence
4–6 8C higher than the onset gelatinization temperature. of fungal storage rots on fruit of Cucurbita maxima D. hybrid
Delica. New Zealand Journal of Crop and Horticultural Science,
18, 141–145.
Hayakawa, K., Tanaka, K., Nakamura, T., Endo, S., & Hoshino, T. (1997).
Quality characteristics of waxy hexaploid wheat (Triticum aestivum L.):
Acknowledgements Properties of starch gelatinzation and retrogradation. Cereal Chemistry,
74, 576–580.
The authors wish to thank Crop and Food Research, Hurst, P. L., Corrigan, V. K., Hannan, P. J., & Lill, R. E. (1995). Storage
Palmerston North, New Zealand, for financial support to rots, compositional analysis, and sensory quality of three cultivars of
DGS, Dr Wayne King for assistance with field site, Tracey buttercup squash. New Zealand Journal of Crop and Horticultural
Science, 23, 89–95.
Pepper for electron microscopy assistance, Scott Schlor- Irving, D. E., Hurst, P. L., & Ragg, J. S. (1997). Changes in carbohydrates and
holtz for X-ray diffraction assistance, David Knollhoff for carbohydrate metabolizing enzymes during the development, maturation,
assistance with harvesting and Dr Marian McKenzie for and ripening of buttercup squash (Cucurbita maxima D, Delica). Journal
helpful discussions. of the American Society of Horticultural Science, 122, 310–314.
 D.G. Stevenson et al. / Carbohydrate Polymers 59 (2005) 153–163 163

Jane, J., & Chen, J. F. (1992). Effect of amylose molecular size and Shibanuma, Y., Takeda, Y., & Hizukuri, S. (1996). Molecular and
amylopectin branch chain length on paste properties of starch. Cereal pasting properties of some wheat starches. Carbohydrate Polymers,
Chemistry, 69, 60–65. 29, 253–261.
Jane, J., Chen, Y. Y., Lee, L. F., McPherson, A. E., Wong, K. S., Smith, R. J., & Caruso, J.-L. (1964). Determination of phosphorus. In R. L.
Radosavljevic, M., & Kasemsuwan, T. (1999). Effects of amylopectin Whistler, R. J. Smith, J. N. BeMiller, & M. L. Wolform (Eds.), Methods
branch chain length and amylose content on the gelatinization and in carbohydrate chemistry: Starch, 4 (pp. 42–46). London: Academic
pasting properties of starch. Cereal Chemistry, 76, 629–637. Press, 42–46.
Jane, J., Kasemsuwan, T., Chen, J. F., & Juliano, B. O. (1996). Phosphorus Smittle, D. A., Hayes, M. J., & Williamson, R. E. (1980). Post-harvest
in rice and other starches. Cereal Foods World, 41, 827–832. quality changes in immature summer squash (Cucurbita pepo var.
Jane, J., Kasemsuwan, T., Leas, S., Zobel, H., & Robyt, J. F. (1994). Condensa). HortRes, 20, 1–8.
Anthology of starch granule morphology by scanning electron Song, Y., & Jane, J. (2000). Characterization of barley starches of
microscopy. Starch/Stäerke, 46, 121–129. waxy, normal, and high amylose varieties. Carbohydrate Polymers, 41,
Kasemsuwan, T., & Jane, J. (1996). Quantitative method for the survey of 365–377.
starch phosphate derivatives and starch phospholipids by 31P nuclear Stevenson, D. G. (2003). Role of starch structure in texture of winter squash
magnetic resonance spectroscopy. Cereal Chemistry, 73, 702–707. (Cucurbita maxima D.) fruit and starch functional properties (158, 163,
Kasemsuwan, T., Jane, J., Schnable, P., Stinar, P., & Robertson, D. (1995). 483–490 pp.). PhD Thesis, Iowa State University, Ames, IA.
Characterization of the dominant mutant amylose-extender (Ae1-5180) Sugimoto, Y., Yamashita, Y., Oomori, M., Suzuki, M., Wakita, H.,
maize starch. Cereal Chemistry, 71, 457–464. Morishita, M., & Fuwa, H. (1998). Developmental changes in the
Keppel, C. R. (2001). A study of glucose storage polymers: Teosinte starch, properties of squash starches. Oyo Toshitsu Kagaku: Journal of Applied
starch crystallinity and cyanobacterial glycogen. Masters of Science Glycoscience, 45, 11–19.
thesis, Iowa State University. Sugimoto, Y., Yamashita, Y., Suzuki, M., Morishita, M., & Fuwa, H.
Kuno, M., Kainuma, K., & Takahashi, S. (2000). Physicochemical (1998). The properties of starches in pumpkin (Cucurbita moschata
characteristics of low-amylose rice starches. Oyo Toshitsu Kagaku: DUCH.) during preservation. Oyo Toshitsu Kagaku: Journal of Applied
Journal of Applied Glycoscience, 47, 319–326. Glycoscience, 45, 33–39.
Li, J. H., Vasanthan, T., Rossnagel, B., & Hoover, R. (2001). Starch from Takeda, Y., & Hizukuri, S. (1987). Structures of branched molecules of
hull-less barley: II. Thermal, rheological and acid hydrolysis charac- amyloses of various origins, and molar fractions of branched and
teristics. Food Chemistry, 74, 407–415. unbranched molecules. Carbohydrate Research, 165, 139–145.
Lim, S., Kasemsuwan, T., & Jane, J. (1994). Characterization of Tang, H., Ando, H., Watanabe, K., Takeda, Y., & Mitsunaga, T. (2001).
phosphorus in starch by 31P-nuclear magnetic resonance spectroscopy. Fine structure of amylose and amylopectin from large, medium, and
Cereal Chemistry, 71, 488–493. small waxy barley starch granules. Cereal Chemistry, 78, 111–115.
Lu, T., Jane, J., Keeling, P. L., & Singletary, G. W. (1996). Maize starch Wang, Y., White, P., & Pollak, L. (1993). Physicochemical properties of
fine structures affected by ear developmental temperature. Carbo- starches from mutant genotypes of the Oh43 inbred line. Cereal
hydrate Research, 282, 157–170. Chemistry, 70, 199–203.
McPherson, A. E., & Jane, J. (1999). Comparison of waxy potato with other White, P. J., Abbas, I. R., & Johnson, L. A. (1989). Freeze–thaw stability
root and tuber starches. Carbohydrate Polymers, 40, 57–70. and refrigerated-storage retrogradation of starches. Starch, 41,
Merrow, S. B., & Hopp, R. J. (1961). Storage effects on winter squashes: 176–180.
Association between the sugar and starch content of and the degree of Wong, K. S., & Jane, J. (1997). Quantitative analysis of debranched
preference for winter squashes. Agricutural and Food Chemistry, 9, amylopectin by HPAEC-PAD with a post-column enzyme reactor.
321–326. Journal of Liquid Chromatography, 20, 297–310.
Ramsey, F. L., & Schafer, D. W. (1996). The statistical sleuth: A course in Yoo, S., & Jane, J. (2002a). Structural and physical characteristics of waxy
methods of data analysis. Belmont, CA: Duxbury Press (pp. 154). and other wheat starches. Carbohydrate Polymers, 49, 297–305.
SAS Institute Inc. (1999). The SASw system for Windowsw, version 8. Yoo, S., & Jane, J. (2002b). Molecular weights and gyration radii of
Cary, NC. amylopectins determined by high-performance size-exclusion chroma-
Sasaki, T., & Matsuki, J. (1998). Effect of wheat structure on swelling tography equipped with multi-angle laser-light scattering and refractive
power. Cereal Chemistry, 75, 385–391. index detectors. Carbohydrate Polymers, 49, 307–314.
Schales, F. D., & Isenberg, F. M. (1963). The effect of curing and storage on Yoshida, M. (1989). Some observations on the structure and starch granules
chemical composition and taste acceptability of winter squash. Journal of squash (var. Ebisu). Research Bulletin of the University Farm at
of the American Society of Horticultural Science, 83, 667–674. Hokkaido University, 26, 17–30.
Schoch, T. J. (1942). Fractionation of starch by selective precipitation Yun, S. H., & Matheson, N. K. (1990). Estimation of amylose content of
with butanol. Journal of the American Chemical Society, 64, starches after precipitation of amylopectin by concanavalin A.
2957–2961. Starch/Stäerke, 40, 302–305.