Genotypic diversity and genome-wide association study of gelatinization and retrogradation properties of potato starch

Abstract

Potato starch gelatinization and retrogradation properties underpin functional performance in food and industrial applications, yet their genetic architecture in potato remains incompletely resolved. We profiled gelatinization and retrogradation traits by differential scanning calorimetry (DSC) in a diverse potato panel across two environments. Wide genotypic variation in all the traits was observed, with highly significant genotype, environment, and genotype by environment (G × E) effects. Gelatinization temperatures were strongly correlated with each other and positively associated with the retrogradation percentage. A genome-wide association study (GWAS) identified loci near ADP-glucose pyrophosphorylase large subunit (AGPL) on chromosome 1 for onset gelatinization temperature, glucan water dikinase (GWD) on chromosome 5) for conclusion gelatinization temperature, and ADP-glucose pyrophosphorylase small subunit (AGPS) and starch branching enzyme (SBE) on chromosome 7 for the retrogradation peak temperature. These findings implicate that carbon flux, starch phosphorylation, and branching are key factors determining the starch functional properties. Together, the results provide a foundation for targeted breeding strategies aimed at improving starch quality for both industrial and food applications.

Highlights

• •Wide variation in potato starch gelatinization and retrogradation traits observed.
• •GWAS identified key loci in AGPL, GWD, AGPS, and SBE genes.
• •Loci near AGPL1 and GWD govern starch gelatinization properties.
• •AGPS1 and SBE1 influence retrogradation and structural reassembly.
• •It implies amylopectin branching and phosphorylation affect crystalline organization.

1. Introduction

Potato (Solanum tuberosum L.) ranks as the fourth most important food crop worldwide, playing a crucial role in global food security. As the largest potato producer, China contributed over 95.57 million metric tons in 2023, underscoring its pivotal position in the worldwide potato industry (Ritchie et al., 2023). Starch is the primary component of potato tubers, making up approximately 75% of their dry matter, and it has a wide range of economic applications, including food processing, pharmaceuticals, and the production of biodegradable materials (Ahmed, Zhou, Pang, Jin, & Bao, 2018; Lu et al., 2016). Potato starch is highly valued in the food industry for its neutral flavour profile, free from the typical taste of cereal-based starches, and its functionality as a thickener, emulsifier, filler, adhesive, and auxiliary raw material (Altemimi, 2018; Li et al., 2017; Saeedyzadeh et al., 2017).

Gelatinization, a key functional property of starch, plays a crucial role in its behaviour in food systems, influencing starch digestion rates and glycemic response (Ek et al., 2014). When heated in excess water, starch undergoes irreversible swelling, losing its crystalline structure. Amylose then solubilizes and leaches out, increasing system viscosity (Thomas & Atwell, 1999). The onset temperatures (To) range from 52.2 °C to 67.6 °C of potato starch significantly lower than those of other starches like arrowroot, cassava, and corn, which vary from 50.6 °C to 77.7 °C (Ahmed, Zhou, Pang, Xu, et al., 2018; Wang et al., 2018). The conclusion temperatures (Tc) for potato starch span from 67.7 °C to 79.7 °C (Chung et al., 2014). Moreover, the enthalpy of gelatinization (ΔHg) of potato starch shows a range from 6.5 J/g to 22.2 J/g, typically broader than the 4.6–13.2 J/g reported for other starches (Wang et al., 2018; Zhu & Liu, 2020). Genetics strongly determines starch structure, which in turn influences properties such as gelatinization temperature and the energy absorbed (melting enthalpy) during starch gelatinization (Lai et al., 2016). The amylopectin fine structure also determines the thermal and retrogradation properties. For instance, potato starches with more short chains and fewer long chains of amylopectin displayed lower gelatinization temperature but higher retrogradation (Ahmed, Ru, Han, et al., 2019).

Retrogradation is the process by which gelatinized starch molecules, particularly amylose, reassociate into a lower-energy crystalline state upon cooling, forming a firm gel (Zaidul et al., 2007). This phenomenon occurs in two steps: first, the retrograded amylose forms crystalline regions, and second, an ordered structure develops within amylopectin. The rate of retrogradation depends on factors such as the starch concentration, amylose content, chain length, temperature, and other substances in the system. High-amylose starches retrograde more quickly due to their greater propensity for crystallization.

Genotypic variation within potato populations plays a crucial role in determining starch's gelatinization and retrogradation properties. Recent research has identified specific single-nucleotide polymorphisms (SNPs) that are closely associated with differences in starch composition and functionality (Ahmed, Ru, Cheng, et al., 2019). These genetic markers facilitate the selection of parent lines capable of producing progeny with enhanced starch traits, thereby presenting new opportunities for improving the quality of potato starch. For instance, quantitative trait locus (QTL) linked to amylose content has been successfully incorporated into breeding programs to optimize starch quality (Ahmed, Zhou, Pang, Xu, et al., 2018).

Genetic basis of starch's gelatinization and retrogradation behaviors has been conducted in potatoes. QTLs related to gelatinization temperature were identified on chromosomes 2, 3, and 5 (Werij et al., 2012). Notably, two major QTLs on chromosome 2, SGT02–2 and SGT03–2, were consistently significant across two years. Additional minor QTLs, STG02–5 and SGT03–3, were detected but were only significant in one year. Interestingly, co-localization of QTLs was observed, with the starch synthase gene SS2 mapped to the same position as SGT02–2 and SGT03–2 on chromosome 2 (Werij et al., 2012). Furthermore, two starch phosphorylase genes, plastidic phosphorylase A (StPhola) and cytosolic phosphorylase (StPho2), also coincided with the QTLs for gelatinization temperature on chromosome 2, indicating overlapping genetic regulation of starch-related traits (Werij et al., 2012).

We previously examined the genotypic diversity and genetic architecture of the phosphorus content (PC), apparent amylose content (AAC), pasting and textural properties of potato starch using a diverse panel of 102 potato accessions grown in two environments (Ahmad et al., 2025). In that study, we identified 74 significant loci distributed across all 12 chromosomes, with several loci nearby the genes involved in starch biosynthesis and metabolism, such as starch branching enzyme (SBE), glucan water dikinase (GWD), and phosphoglucomutase (PGM). Notably, many loci exhibited pleiotropic effects on multiple starch traits, providing critical insights into the genetic regulation of starch functionality. These findings established a robust foundation for exploring additional starch properties and their genetic determinants. However, the genetic factors controlling the gelatinization and retrogradation properties of potato starches have not been previously investigated through genome-wide association study (GWAS).

The objective of this study is to identify significant SNPs and candidate genes associated with starch gelatinization and retrogradation properties in potato germplasm using GWAS. By integrating these findings with those from our previous research, we seek to provide a comprehensive understanding of the genetic basis of potato starch functionality, offering valuable insights for breeding programs aimed at improving starch quality.

2. Materials and methods

2.1. Plant materials

A diverse panel of 102 potato varieties was cultivated at the Huajiachi Campus field of Zhejiang University, Hangzhou, China, in 2019 and 2020, following a randomized complete block design. The plant materials, cultivation conditions, and experimental design are described in detail in Ahmad et al. (2025) and Rasheed et al. (2024). At maturity, the potato accessions were harvested, and the tubers were carefully washed, manually peeled, and cut into smaller pieces before starch extraction. A total of 97 genotypes from environment E1 and 98 genotypes from environment E2 produced sufficient tubers for starch extraction. To assess the genotype by environment interaction (G × E) effects, 93 potato accessions common to both environments were analyzed, ensuring consistency and reliability in evaluating the genetic and environmental influences on starch characteristics.

2.2. Starch isolation

Potato starch was isolated as described by Hu et al. (2025) and Ahmad et al. (2025). The final starch was dried at 37 °C, ground, and sieved for further analysis.

2.3. Gelatinization properties

The gelatinization behaviour of starch was evaluated using a differential scanning calorimetry Q20 (TA Instruments, New Castle, DE), following the procedure outlined by Ahmed, Zhou, Pang, Xu, et al. (2018). For this analysis, 2.0 mg of starch (on a dry weight basis) was combined with 6 μL of distilled water in an aluminum pan. The pan was hermetically sealed and allowed to equilibrate at room temperature for 2 h to ensure proper hydration. The sample was then heated from 30 °C to 130 °C at a steady rate of 10 °C/min, with an empty sealed pan used as a reference. Nitrogen gas was employed as the purge gas with a 50 mL/min flow rate to maintain the DSC atmosphere. The gelatinization onset temperature (To), peak temperature (Tp), conclusion temperature (Tc), and enthalpy (ΔHg) were determined using the Universal Analysis software (Version 4.4 A).

2.4. Retrogradation properties

Retrogradation properties were determined using the same DSC pans analyzed for gelatinization. After the initial thermal scan, the gelatinized starch samples were cooled to room temperature and subsequently stored at 4 °C for 7 days to facilitate retrogradation. Following the storage period, the samples were rescanned using the DSC, with the temperature range set from 30 °C to 110 °C and a heating rate of 10 °C/min. The onset temperature (RTo), peak temperature (RTp), conclusion temperature (RTc), and enthalpy (ΔHr) of the retrograded starch were measured using the Universal Analysis software (Version 4.4 A). The retrogradation percentage (R%) was calculated using the formula below.

R% = (ΔHr / ΔHg) × 100.

2.5. Apparent amylose content, phosphorus content, pasting, and textural properties

The apparent amylose content (AAC), phosphorus content (PC), pasting, and textural properties were measured following the methods reported in Ahmad et al. (2025). All the data were reported previously (Ahmad et al., 2025) and used in this study for analysis of correlation.

2.6. GWAS

The genotypic data, consisting of 226,487 SNPs and 22,115 Indels, was generated using double-digested Restriction-site Associated DNA sequencing, as detailed by Rasheed et al. (2024). Further information on GWAS analysis, including methodology and LD decay, is provided in Ahmad et al. (2025).

2.7. Statistical analyses

All the measurements were technically conducted in duplicate. Descriptive statistics, Pearson correlations, ANOVA, and other data visualizations were performed using R, as described in Ahmad et al. (2025).

3. Results

3.1. Gelatinization properties

Apparent variation was observed in starch gelatinization traits among accessions and across environments as supported by ANOVA, violin plots, and histograms (Table 1; Fig. 1). The To, Tp, and Tc temperatures ranged from 58.8 to 68.5 °C, 63.1 to 73.8 °C, and 67.9 to 82.7 °C, respectively. The corresponding means across environments were 63.9 ± 1.5 °C (To), 67.5 ± 1.5 °C (Tp), and 72.7 ± 1.7 °C (Tc). The ΔHg ranged from 12.7 to 19.7 J/g, with a mean of 15.6 ± 1.1 J/g (Table 1). ANOVA revealed significant genotypic effects (p < 0.001) for all gelatinization traits (Table 1). The Tp and Tc was mainly determined by genotypic effects, whereas the To and ΔHg were mainly determined by the environmental effects. It was noted that the environmental effect was not significant for Tc. G × E effects were highly significant (p < 0.001) for all the traits, indicating that both genetic background and environmental factors jointly determined starch thermal behaviour.

Table 1.

Descriptive statistics and ANOVA of gelatinization and retrogradation properties.

		To (°C)	Tp (°C)	Tc (°C)	ΔHg (J/g)	RTo (°C)	RTp (°C)	RTc (°C)	ΔHr (J/g)	R%
Environment I
Mean + SD		63.5 ± 1.6	67.4 ± 1.8	72.7 ± 1.9	15.9 ± 1.3	49.2 ± 4.5	64.4 ± 2	74.2 ± 3.6	5.4 ± 2.2	35.3 ± 15.2
CV		2.5	2.6	2.7	7.9	9.2	3.1	4.9	40.4	43.2
Range		59.8–67.7	63.1–73.8	67.9–82.7	12.7–19.7	41.3–62.4	57.7–74.9	64.6–86.4	1.3–9.7	8.1–77.4
Environment II
Mean + SD		64.2 ± 1.3	67.6 ± 1.2	72.7 ± 1.5	15.3 ± 0.8	51.9 ± 1.7	64.5 ± 1.2	74.3 ± 2.3	5 ± 1.4	32.7 ± 9.2
CV		2	1.8	2.1	5.2	3.2	1.8	3.1	27.7	28.1
Range		58.8–68.5	64.4–72.1	69.3–77.7	12.9–17.8	45.7–56.6	61.4–66.8	67.6–81.1	2.4–8.4	16.2–54.6
Total
Mean + SD		63.9 ± 1.5	67.5 ± 1.5	72.7 ± 1.7	15.6 ± 1.1	50.5 ± 3.7	64.4 ± 1.6	74.2 ± 3	5.2 ± 1.9	34.0 ± 12.6
CV		2.3	2.2	2.4	7.0	7.2	2.5	4.1	35.2	37.1
Range		58.8–68.5	63.1–73.8	67.9–82.7	12.7–19.7	41.3–62.4	57.7–74.9	64.6–86.4	1.3–9.7	8.1–77.4
ANOVA
SOV	df	Mean Square
G	92	5.9***	6.3***	7.7***	2.57***	26.3***	5.95***	22.5***	8.5***	363.3***
E	1	42.3***	3.6*	0.08	26.4***	667.5***	0.91	0.95	15.7***	258.6**
G × E	92	1.8***	2.4***	3.8***	2.05***	19.4***	4.3**	13.4***	4.6***	197.5***
Error		0.54	0.73	1.67	1.12	2.9	2.785	7.39	0.70	34.7

Summary statistics represent the mean of two replications. SOV: source of variation; df: degree of freedom; CV: coefficient of variation; G: genotype; E: environment; G × E: genotype–environment interaction. To: gelatinization onset temperature; Tc: gelatinization conclusion temperature; Tp: gelatinization peak temperature; ΔHg: enthalpy of gelatinization; RTo: retrogradation onset temperature; RTp: retrogradation peak temperature; RTc: retrogradation conclusion temperature; ΔHr: enthalpy of retrogradation; R%: retrogradation percentage. Significance levels are denoted by asterisks *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 1.

Violin plots and histograms showing the distribution of gelatinization and retrogradation properties of potato starch across two growing seasons.

3.2. Retrogradation properties

The retrogradation properties of potato starch showed considerable variation among accessions and across environments (Table 1; Fig. 1). The retrogradation onset (RTo), peak (RTp), and conclusion (RTc) temperatures ranged from 41.3 to 62.4 °C, 57.7 to 74.9 °C, and 64.6 to 86.4 °C, respectively, with mean values of 50.5 ± 3.7 °C, 64.4 ± 1.6 °C, and 74.2 ± 3.0 °C across environments (Table 1; Fig. 1). The enthalpy of retrogradation (ΔHr) ranged from 1.3 to 9.7 J/g, averaging 5.2 ± 1.9 J/g, indicating substantial variability in the energy required for melting the retrograded starch. The retrogradation percentage (R%) varied widely from 8.1 to 77.4%, with a population mean of 34.0 ± 12.6%, reflecting broad differences among genotypes.

Analysis of variance revealed highly significant genotypic effects (p < 0.001), emphasizing the contribution of genotypic variability to trait expression (Table 1). Environmental effects had pronounced effects for the RTo (p < 0.001) and ΔHr (p < 0.001) but had no effect on RTp and RTc. Additionally, G × E interactions were highly significant (p < 0.001) across all traits except for RTp. The RTp, RTc and R% were mainly controlled by genotypic effects, RTo and ΔHr were mainly controlled by environmental effects (Table 1).

3.3. Correlation of gelatinization and retrogradation properties

The correlation analysis of gelatinization properties revealed strong relationships across both environments (Table 2). The values along the diagonal represent the relationship of each trait with itself across both environments, while correlations above and below the diagonal correspond to E1 and E2, respectively (Table 2). A strong positive correlation was observed between To and both Tp (r = 0.83 in E1) and Tc (r = 0.59 in E1), with similar trends observed in E2. Similarly, Tp and Tc were highly correlated within both environments (r = 0.88 in E1, r = 0.87 in E2), indicating strong interdependence among these gelatinization properties. Interestingly, ΔHg in this study showed weak and no significant correlations with most gelatinization properties, contrasting with the strong relationships observed for gelatinization temperatures (Table 2).

Table 2.

Correlation of potato starch's gelatinization and retrogradation properties in two environments.

	To	Tp	Tc	ΔHg	RTo	RTp	RTc	ΔHr	R%
T0	0.52***	0.83***	0.59***	−0.11	−0.26*	0.24*	0.23*	0.25*	0.24*
Tp	0.86***	0.48***	0.88***	−0.12	−0.29**	0.23*	0.38***	0.34***	0.32**
Tc	0.63***	0.87***	0.37***	−0.05	−0.19	0.27**	0.44***	0.30**	0.27**
ΔHg	−0.16	−0.10	0.06	−0.03	0.08	0.10	0.09	−0.05	−0.21*
RTo	0.05	0.09	0.13	−0.01	0.23*	0.24*	−0.37***	−0.28**	−0.30**
RTp	0.42***	0.50***	0.43***	0.06	0.54***	0.19	0.43***	−0.05	−0.02
RTc	0.42***	0.52***	0.45***	0.22*	0.00	0.61***	0.28**	0.25*	0.34***
ΔHr	0.28**	0.34***	0.27*	0.04	−0.38***	0.03	0.45***	0.30**	0.91***
R%	0.31**	0.36***	0.26*	−0.13	−0.37***	0.02	0.40***	0.98***	0.25*

Significance levels are denoted by asterisks *p < 0.05, **p < 0.01, ***p < 0.001; The correlations above and below the diagonal correspond to E1 and E2, and those in diagonal in bold correspond to between E1 and E2.

Regarding retrogradation properties, the ΔHr and R% were strongly correlated in both environments (Table 2). Both ΔHr and R% also exhibited weak to moderate positive correlations with Tp (r = 0.34 for ΔHr, r = 0.32 for R% in E1) and Tc (r = 0.30 for ΔHr, r = 0.27 for R% in E1), indicating a connection between gelatinization and retrogradation traits (Table 2).

3.4. Impact of PC, AAC, pasting, and textural attributes on gelatinization and retrogradation properties

Correlation analysis of PC, AAC, pasting properties, textural traits and gelatinization, reveals important relationships among these starch characteristics (Fig. 2). This provides insights into how PC, AAC, pasting and textural properties together influence gelatinization and retrogradation behaviors. PC showed moderate positive correlations with gelatinization temperatures, including To (r = 0.29), Tp (r = 0.33), and Tc (r = 0.32). Similarly, AAC exhibited moderate positive correlations with To (r = 0.26) and Tp (r = 0.22), and weak correlations with Tc (r = 0.12) (Fig. 2). Pasting properties such as breakdown (BD) showed weak negative correlations with To (r = −0.15), Tp (r = −0.11) and Tc (r = −0.07), while setback (SB) showed weak positive correlations with To (r = 0.18) and Tc (r = 0.11) (Fig. 2). Among textural properties, gel hardness (HD) showed moderate positive correlations with To (r = 0.31) and Tp (r = 0.33). Gel gumminess (GU) showed moderate positive correlations with To (r = 0.38) and Tp (r = 0.44).

Fig. 2.

Correlation analysis of phosphorus content (PC), apparent amylose content (AAC), pasting properties, and textural properties with thermal properties of potato starch. The heatmap depicts the strength and direction of correlations, with red indicating positive correlations and blue indicating negative correlations. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

For retrogradation properties, PC showed weak to moderate positive correlations with RTp (r = 0.28) and RTc (r = 0.34). AAC showed weak negative correlation with RTo (r = −0.26), no significant correlation with RTp (r = −0.02), and weak to moderate positive correlations with RTc (r = 0.13), R% (r = 0.16) and ΔHr (r = 0.21) (Fig. 2). In general, some pasting properties exhibited weak, while others showed moderate positive correlations with retrogradation properties in this study. PV showed weak positive correlations with RTo (r = 0.17), RTc (r = 0.19), but a significant positive correlation with RTp (r = 0.27). Textural properties such as HD showed weak positive correlations with RTc (r = 0.17) and ΔHr (r = 0.12) (Fig. 2). GU also showed weak positive correlations with RTp (r = 0.18) and R% (r = 0.15) (Fig. 2).

3.5. Genetic basis of gelatinization properties

A GWAS was conducted to elucidate the genetic determinants underlying DSC-derived gelatinization and retrogradation properties of potato starch across two environments. The chromosomal distributions of significant SNPs, their polymorphism patterns, and nearby candidate genes are summarized in Fig. 3, Fig. 4 and Table 3. QTLs were designated based on the corresponding trait, chromosome, and sequential order along each chromosome (Fig. 4, Table 3). Three genomic regions exhibited significant associations with gelatinization traits. A locus for the To on chromosome 1 (qTo-1) was located in proximity to AGPL1 (Sotub01g024100), which encodes the large subunit of ADP-glucose pyrophosphorylase (AGPase). Another major locus influencing the Tc was identified on chromosome 5, near GWD (Sotub05g014130). In addition, another QTL affecting Tc was detected on chromosome 9.

Fig. 3.

Manhattan plots represent a genome-wide association study regarding gelatinization and retrogradation properties. The x-axis represents the chromosome number and the position of each SNP, while the y-axis represents the negative logarithm (−log10) of the p-values for each SNP. The broken lines represent significant thresholds 4.02 × 10⁻⁶. The heat maps at the bottom of each chromosome represent SNP density. The QQ plot on the left shows the observed vs. expected −log10 values. The diagonal line shows the predicted distribution, whereas the other line indicates the observed p-values. The deviation from the diagonal line indicates a significant association.

Fig. 4.

Chromosomal mapping of significant SNPs (shown in black colors) and their associated traits in potato starch, along with key starch biosynthesis genes (shown in red colour). Each chromosome, labeled “Chrom” followed by its respective number, is shown with its total length in megabases. On the left side of each chromosome, the positions of identified SNPs are marked, while on the right, these SNPs are linked to traits they significantly influence. Green regions on the chromosome bars indicate the genomic regions where significant SNPs are associated with possible candidate genes located within a 2.36 Mb linkage disequilibrium. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 3.

Significant marker-trait associations for gelatinization and retrogradation properties.

Trait	QTL	Env	Chr	Position	Polymorphism	P-value	Location	Possible candidate gene
To
	qTo-1	E1	1	62,719,292	A/T	1.49E−07	Intergenic	AGPL1
Tc
	qTc-5	E2	5	10,515,105	C/CAATAATT	2.27E−07	Intronic	GWD
	qTc-9	E2	9	23,801,821	T/C	1.31E−06	Intergenic
RTo
	qRTo-4	E2	4	5,891,239	A/C	3.95E−06	Intergenic
RTp
	qRTp-2	E1	2	21,772,807	G/A	2.13E−06	Intergenic
	qRTp-5	E2	5	12,681,321	G/A	7.63E−07	Intergenic
	qRTp7–1	E2	7	51,892,773	CTA/C	2.45E−06		AGPS1
	qRTp7–2	E2	7	55,095,518	C/T	1.83E−06	Intronic	SBE1
	qRTp-11	E1	11	25,617,808	C/T	4.02E−06	Intergenic

Env; environment, Chr; chromosome.

3.6. Genetic basis of retrogradation properties

GWAS for retrogradation traits revealed several significant loci associated with retrogradation RTo and RTp (Fig. 3, Fig. 4; Table 3. A significant QTL controlling RTo (qRTo-4) was detected on chromosome 4 at 5.891 Mb without an immediately apparent candidate gene. For RTp, multiple loci were detected across chromosomes 2, 5, 7, and 11. On chromosome 7, two major loci were distinguished: qRTp7–1 at 51.893 Mb, which colocated with AGPS1 (Sotub07g023520), encoding the small subunit of ADP-glucose pyrophosphorylase (AGPase), and qRTp7–2 at 55.096 Mb, positioned within an intron of SBE1 (Sotub07g029010). Additional significant associations for RTp were identified in intergenic regions on chromosome 2 (21.773 Mb), chromosome 5 (12.681 Mb), and chromosome 11 (25.618 Mb), although no clear candidate genes were evident at these loci (Table 3).

4. Discussion

4.1. Genotypic diversity and genetic basis of gelatinization properties

The observed variation in gelatinization properties among potato accessions indicates substantial genetic diversity within the panel. The thermal transition parameters recorded here are broadly consistent with earlier reports, though the ranges observed were generally wider (Table 1, Fig. 1). dos Santos et al. (2016) reported narrower intervals of To (64.6–66.7 °C), Tp (67.3–70.2 °C) and Tc (71.1–74.8 °C) for Brazilian cultivars, while gelatinization temperatures in Polish cultivars were reported to range from 62.6 to 64.0 °C for To, 68.3–69.9 °C for Tp, and 73.6–77.0 °C for Tc (Przetaczek-Rożnowska, 2017; Pycia et al., 2012).

In contrast, the present study revealed considerably broader gelatinization temperature ranges compared with commercial cultivars from Canada (59.8–62.8 °C To, 64.2–68.4 °C Tp, 73.8–79.7 °C Tc) (Chung et al., 2014). Similarly, in China, a study on 29 local cultivars revealed ranges of 62.2–67.6 °C for To, 66.1–71.1 °C for Tp, and 73.5–77.4 °C for Tc, which are comparable to the mid-to-upper ranges observed in this study (Ahmed, Zhou, Pang, Xu, et al., 2018). Collectively, these comparisons highlight the broad thermal diversity within the analyzed potato panel, possibly attributable to differences in amylopectin fine structure and genetic constitution among the accessions.

The ΔHg varied between 12.7 and 19.7 J/g, averaging 15.6 ± 1.1 J/g. This range overlaps closely with those reported by dos Santos et al. (2016) (15.5–17.8 J/g) and Chung et al. (2014) (14.1–16.5 J/g), while slightly exceeding values noted by Pycia et al. (2012) (11.1–15.3 J/g). It also approaches the lower end of the higher interval reported by Ahmed, Zhou, Pang, Xu, et al. (2018) in China (18.2–20.7 J/g). Higher ΔHg values typically correspond to more highly ordered crystalline regions and a greater proportion of stable double helices within amylopectin (Hoover, 2001).

The significant genotypic effects (p < 0.001) observed for all gelatinization traits, coupled with significant environmental effects for most gelatinization traits, emphasize the combined influence of genetic and environmental factors on starch thermal behaviour (Table 1). These findings align with reports that gelatinization properties are influenced by both intrinsic genetic factors and external environmental conditions, as demonstrated in studies from diverse regions (Ahmed, Zhou, Pang, Xu, et al., 2018; dos Santos et al., 2016). From an industrial perspective, the observed variation provides a valuable basis for aligning starch functionality with application requirements. Accessions with lower To and Tp are appropriate for processes favoring rapid or energy-efficient gelatinization, whereas higher Tc and ΔHg are advantageous where resistance to thermal treatment is required.

Three QTLs on chromosome 1, 5 and 9 were identified in this study. Werij et al. (2012) reported the QTLs identified on chromosomes 2, 3, and 5, the discrepancy may be derived from different materials and method used between two studies. The identified associations near AGPL and GWD are consistent with well-established structure–function relationships, in which amylopectin chain-length distribution and granule phosphorylation influence crystalline order, hydration, and the energy required to disrupt lamellae (Bertoft, 2017; Hoover, 2001; Imberty & Perez, 1988; Jane et al., 1999; Pérez & Bertoft, 2010; Reyniers et al., 2020; Vamadevan & Bertoft, 2015).

AGPL encodes a core enzyme of starch biosynthesis that regulates carbon flux into ADP-glucose, and perturbation at this locus has been associated with altered starch accumulation and amylopectin chain-length distribution in potato (Lloyd et al., 1999; Pfister & Zeeman, 2016; Stark et al., 1992). Structure–thermal studies indicate that increases in the proportion of longer amylopectin B-chains and more ordered crystallites elevate To and Tp (Hoover, 2001; Jane et al., 1999; Pérez & Bertoft, 2010; Vamadevan & Bertoft, 2015).

GWD installs C6 phosphate monoesters and, together with phosphoglucan water dikinase (PWD), modulates granule phosphorylation (Mikkelsen et al., 2004; Ritte et al., 2006; Zeeman et al., 2010). Genetic manipulation of GWD in potato reduces phosphate monoesters and alters DSC transitions and pasting behaviour (Kozlov et al., 2007; Ohnuma et al., 2023), and natural allelic variation at GWD associates with phosphate content in tetraploid germplasm (Uitdewilligen et al., 2022). Replicated mapping of phosphorylation-related traits in this region further supports a role for GWD-linked variation in determining Tc (Carpenter et al., 2015; Werij et al., 2012).

4.2. Genotypic diversity and genetic basis of retrogradation properties

The broad ranges observed for retrogradation temperatures and enthalpy underscore the substantial genetic and environmental diversity captured in the studied germplasm. Compared with reports from other regions, the present data span from lower to markedly higher temperature ranges, suggesting both genetic variation and environmental modulation of retrogradation properties. For example, narrower ranges reported in Brazilian accessions (dos Santos et al., 2016) and in Polish cultivars (Przetaczek-Rożnowska, 2017; Pycia et al., 2012) align with the lower segment of the current distribution, while higher values approaching those described in Chinese germplasm (Ahmed, Zhou, Pang, Xu, et al., 2018) reflect environments or genotypes that favor more extensive molecular reassociation.

The relatively low mean ΔHr values (5.2 J/g) observed in this study are comparable to those reported in Canada (Chung et al., 2014) and Poland (Przetaczek-Rożnowska, 2017) but lower than those from Brazil and China, implying reduced re-crystallization energy, potentially associated with differences in amylose content, molecular structure, or storage conditions (Ahmed, Zhou, Pang, Xu, et al., 2018; Chung et al., 2014; dos Santos et al., 2016; Przetaczek-Rożnowska, 2017).

The wide range of R% (8.1–77.4%) reinforces the strong genotypic component influencing the extent of retrogradation, consistent with prior findings in Polish and Canadian cultivars (Chung et al., 2014; Przetaczek-Rożnowska, 2017; Pycia et al., 2012) but contrasting with the narrower and generally lower R% reported in Chinese accessions (Ahmed, Zhou, Pang, Xu, et al., 2018). Together, the significant genotypic, environmental, and G × E effects confirm that starch retrogradation is governed by both heritable factors and external conditions, defining a complex physicochemical trait with potential implications for texture, shelf stability, and processing quality in potato-based products.

For RTo, a significant QTL (qRTo-4) was detected on chromosome 4; however, this SNP does not associate with any possible candidate gene (Table 3). For RTp, qRTp7–1 co-locates with AGPS1 (Sotub07g023520). AGPase controls carbon flux into starch and influences total starch accumulation (Geigenberger et al., 1999; Pfister & Zeeman, 2016). The qRTp7–2 locus associated with SBE highlights the contribution of branching to retrogradation kinetics. Lower branching frequency or a shift toward longer B-chains increases the tendency for double-helix formation during storage, typically elevating retrogradation rates and raising RTp or shifting the storage-time dependence of the peak (Brummell et al., 2015). For those loci where no clear candidate genes were found, there may have been variations in the regulatory elements or non-coding regions that regulate starch biosynthesis related genes.

This study provides novel insights into the genetic basis of potato starch gelatinization and retrogradation properties, highlighting the critical roles of key genes and their associated pathways. The identified QTLs and their associations with AGPL and GWD underscore the importance of carbon flux regulation and granule phosphorylation in modulating thermal transitions. Additionally, the discovery of loci linked to retrogradation properties, particularly those involving AGPS and SBE, reveals the influence of flux and branching on the reorganization of starch molecules during cooling and storage. Together, these findings advance our understanding of the genetic and molecular mechanisms shaping starch functional properties in potatoes. They also provide a foundation for targeted breeding strategies aimed at improving starch quality for both industrial and food applications. Future studies should explore the regulatory elements underlying these traits and investigate their interactions across diverse genetic backgrounds and environmental conditions.

4.3. Correlation of gelatinization, retrogradation and other starch properties

The strong correlations among To, Tp, and Tc showed interdependent nature of starch gelatinization, where changes in one temperature parameter proportionally influence others. These findings agree with those of Ahmed, Zhou, Pang, Xu, et al. (2018), who reported significant and positive correlations among gelatinization temperatures (To, Tp, and Tc) in potato starch. The positive correlation between ΔHr and R% aligns with the findings of Ahmed, Zhou, Pang, Xu, et al. (2018), which is evident because R% was calculated from ΔHr and ΔHg.

The correlation results demonstrate that PC, AAC, pasting and textural properties collectively influence the gelatinization and retrogradation properties of potato starch. Previous studies indicated that PC had no correlation with gelatinization temperatures (Karim et al., 2007; Hu et al., 2025). However, Hu et al. (2025) found that PC correlated negatively with To and Tp in high-PC potato starches but positively with To in low-PC potato starches. The positive correlation between PC and gelatinization temperature in this study (Table 2) may reflect the predominantly low PC content of the potato accessions used in this study. Karim et al. (2007) reported that high‑phosphorus starches typically exhibited lower ΔHg. Conversely, Hu et al. (2025) found that high-PC starches had higher ΔHg than low-PC starches, while Lu et al. (2012) observed a positive correlation between PC and both ΔHg and ΔHr. In contrast, no correlation between PC and either ΔHg or ΔHr was detected in this study, suggesting that the effect of PC on the melting energy of raw and retrograded starches is complex.

While AAC is a critical determinant of pasting properties, its relationship with gelatinization temperatures and retrogradation properties is complex. Moderate to weak positive correlation of ACC with gelatinization temperatureis consistent with findings by Zhu and Liu (2020). Negative or no significant correlation of RTo and RTp with ACC suggests that amylose may not uniformly impact all retrogradation properties, as also noted by Ahmed, Ru, Han, et al. (2019). Ahmed, Ru, Han, et al. (2019) and Zhu and Liu (2020) reported that longer-chain amylopectin, especially external long-chain amylopectin (B3 chains), contributes to higher gelatinization temperature, whereas they negatively correlate with ΔHr.

Regarding pasting properties, negative correlation between BD and gelatinization temperatures and positive correlation between SB and gelatinization temperatures are consistent with Ahmed, Zhou, Pang, Xu, et al. (2018) as high amylose levels and increased starch rigidity reduce swelling and viscosity. Interestingly, Ahmed, Zhou, Pang, Xu, et al. (2018) also observed positive correlations between SB and AAC, which supported with this study's findings that BD and SB exhibited weak associations with gelatinization and retrogradation properties. Zhu and Liu (2020) noted that pasting properties are influenced by the molecular structure of amylopectin, with shorter B1 chains negatively correlated with ΔHr and longer B3 chains positively correlated with RTp and RTc of retrograded starch.

Weak to moderate positive correlations of HD and GU with gelatinization temperature and retrogradation properties align with the finding of Zhu and Liu (2020), who reported that the higher double-helical arrangement of chains in amylose and amylopectin results in greater textural stability and higher gelatinization temperatures.

5. Conclusion

This work delivers the first genome-scale genetic underpinnings of DSC-derived gelatinization and retrogradation properties in potato starch across environments, bridging thermal behaviour to actionable genetics. Key loci near AGPL1 and GWD emerged as primary determinants of onset and conclusion gelatinization temperatures, respectively, implicating carbon flux into starch and granule phosphorylation in controlling crystalline order and lamellar mechanics. For retrogradation, signals at AGPS1 and SBE1 highlight the roles of flux and branching architecture in governing peak temperature of the retrogradated starch. The identified QTLs and candidate genes provide immediate entry points for marker-assisted selection, genomic prediction, and gene editing in potatoes, with the goal of tuning starch for faster cooking, enhanced thermal tolerance, or tailored texture.

CRediT authorship contribution statement

Daraz Ahmad: Writing – review & editing, Writing – original draft, Investigation, Formal analysis. Lin Zhang: Investigation, Formal analysis. Yan Gao: Investigation. Bowen Deng: Formal analysis, Data curation. Dianxing Wu: Investigation, Funding acquisition. Jinsong Bao: Writing – review & editing, Writing – original draft, Resources, Funding acquisition, Conceptualization.

Human and animal rights

This article does not contain any studies involving human or animal participants performed by any of the authors.

Funding

This work was supported by the Starry Night Science Fund of the Zhejiang University Shanghai Institute for Advanced Study (Grant Number SN-ZJU-SIAS-0012) and Nuclear Technology R&D Program (HJSYF2024-31).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.