Structural Reconstruction and Enhanced Digestive Resistance in High-Amylose Maize Starch–Fatty Acid Complexes via Debranching and Heat–Moisture Treatment

TL;DR

This study shows how to create highly resistant starch by modifying maize starch with enzymes and heat treatments, improving its stability for food applications.

Contribution

A novel sequential strategy combining debranching and heat-moisture treatment to enhance RS5 formation and digestive resistance.

Findings

Extended debranching increased complexing index and resistant starch content significantly.

PHT outperformed ANN, achieving 69.2% RS content in DH24-LOA complexes.

Resistance arises from amorphous-to-crystalline chain transitions and macroscopic densification.

Abstract

The development of thermally stable Type 5 resistant starch (RS5) is critical for functional food applications to modulate glycemic responses. This study investigated the structural assembly and enzymatic resistance of RS5 complexes prepared from high-amylose maize starch (HAMS) via a sequential strategy coupling pullulanase debranching with heat–moisture treatment (HMT). HAMS was debranched for varying durations (0–24 h) to generate short, linear glucan chains, subsequently complexed with myristic acid (MA) or linoleic acid (LOA), and further modified by pressure–heat treatment (PHT) or annealing (ANN). Extended debranching (24 h) significantly enhanced the complexing index and resistant starch (RS) content. While saturated MA promoted higher crystallinity of the hexagonal Bravais lattice, unsaturated LOA effectively enhanced resistance through steric hindrance despite lower crystallinity. Notably, PHT generally outperformed ANN, with the highest RS content (69.2%) achieved in DH24-LOA complexes treated at 120 °C with 10% moisture. Multi-scale structural analyses revealed that resistance originated from the transition of amorphous chains into highly ordered, thermally stable nanocrystals possessing a hexagonal Bravais lattice and the densification of the macroscopic architecture. These findings demonstrate that coupling pullulanase debranching with optimized PHT is a potent strategy to engineer high-performance RS5 ingredients with superior digestive resistance.

Figures

Full text

1. Introduction

As a primary dietary carbohydrate, starch plays an irreplaceable role in the food industry [1]. However, the excessive consumption of rapidly digestible starch (RDS) in modern diets is increasingly associated with the prevalence of metabolic disorders, such as type 2 diabetes and obesity. During thermal processing, starch undergoes gelatinization and structural disintegration, which enhances its susceptibility to amylolytic enzymes and consequently induces rapid postprandial glycemic spikes [2,3]. Therefore, the development of resistant starch (RS), which resists digestion in the small intestine and ferments in the colon, has emerged as a critical strategy for modulating glycemic response and promoting gut health [4]. Among the RS subtypes, Type 5 resistant starch (RS5) is characterized by amylose–lipid inclusion complexes that possess a unique hexagonal Bravais lattice and superior thermal stability [5,6].

Unlike native starch granules, where crystallinity is primarily governed by the double helices of amylopectin (typically possessing orthorhombic or hexagonal crystal structures) [7], RS5 complexes are formed by the co-crystallization of amylose and lipids. Crystallographically, these inclusion complexes typically pack into nanocrystals possessing a hexagonal Bravais lattice (specifically the V_6h_ allomorph) [8,9], which is structurally distinct from the native granular architecture.

The multiscale structure of starch–lipid complexes determines their nutritional function and quality, which can be effectively modulated through processing treatments. Starches with a high proportion of amylose are considered ideal precursors for RS5 because their linear chains exhibit strong interactions with fatty acids (FAs), favoring the formation of highly ordered inclusion complexes that limit enzymatic hydrolysis [10,11,12]. High-amylose maize starch (HAMS), typically containing 50–90% amylose, is widely used as a raw material for preparing RS5 [13]. However, native starch granules often contain a fraction of branched amylopectin that exerts steric hindrance [14]. To overcome this, enzymatic modification using pullulanase efficiently hydrolyzesα-1,6 glycosidic bonds, promoting the release of short-chain amylose [15,16]. Crucially, this debranching process fundamentally alters the crystallization mechanism: rather than the retrogradation of branched amylopectin, the liberated short, linear glucan chains are free to co-crystallize with lipids. This favors the formation of thermally stable inclusion complexes that are distinct from the native granular structure. Previous studies on debranched starch (DH) have demonstrated that increasing the amylose content or degree of debranching generally reduces RDS while increasing slowly digestible starch (SDS) and RS fractions, making debranched high-amylose starches highly feasible substrates for RS production [17,18]. Regarding the lipid component, the formation and order of the complexes are significantly influenced by the chain length and unsaturation of the FA. Saturated FAs like myristic acid (MA C14:0) typically form highly ordered structures with strong enzymatic resistance, whereas unsaturated FAs like linoleic acid (LOA C18:2) may offer different complexation behaviors due to steric effects [19].

To further reinforce the structural integrity of these complexes, physical modifications such as Heat–Moisture Treatment (HMT) have been employed. HMT, which encompasses both pressure heating (PHT) and annealing (ANN), facilitates the rearrangement of molecular chains within amorphous and crystalline regions, thereby enhancing crystalline perfection and rigidity without destroying the granular structure [20]. PHT modifies starch by constraining chain mobility under elevated temperatures and restricted moisture, promoting internal structural rearrangements that lead to higher RS content and reduced solubility [21,22,23,24]. Similarly, ANN treatment applied to starches from various botanical sources has been shown to improve crystallite perfection and increase resistant starch fractions by optimizing the hydration and thermal energy input [25,26,27]. These hydrothermal strategies offer an environmentally friendly approach to tailoring the digestibility of starch–lipid complexes [28].

This study aims to engineer high-performance RS5 complexes by coupling pullulanase debranching with optimized HMT. Unlike traditional ANN, which requires prolonged treatment times (24–72 h), PHT was employed as a time-efficient route to engineer thermally stable (Type II) complexes capable of withstanding industrial food processing conditions. We systematically investigate the effects of debranching duration and hydrothermal conditions (temperature and moisture) on the multi-scale structure and in vitro digestibility of HAMS complexed with MA and LOA, which were selected to represent medium-chain saturated and long-chain polyunsaturated FAs, respectively. We propose that combining debranching pretreatment with PHT or ANN will act in a complementary manner on the multiscale structure of the complexes and thus further enhance their resistance to enzymatic hydrolysis. The findings will elucidate the structure-function relationships governing RS5 formation, providing a theoretical basis for developing functional food ingredients with enhanced enzymatic resistance and thermal stability.

2. Materials and Methods

2.1. Materials

High-amylose maize starch (HAMS), containing 74.6% amylose, was sourced from Quanyin Xiangyu Biotechnology Co., Ltd. (Beijing, China). Myristic acid (MA C14:0) and linoleic acid (LOA C18:2) were procured from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). The amylose assay kit and D-glucose assay kit were purchased from Megazyme International Ltd. (Bray, Ireland). Microbial pullulanase (E2412, EC 3.2.1.41, ≥1000 NPUN/g), porcine pancreatic α-amylase (from porcine pancreas, 8 U/mg), and amyloglucosidase (300 U/mL) were acquired from Sigma-Aldrich (Shanghai) Trading Co., Ltd. (Shanghai, China). All other chemicals used in this study were of analytical grade.

2.2. Preparation of Type 5 Resistant Starch (RS5)

2.2.1. Preparation of Debranched High-Amylose Maize Starches (DHs)

HAMS was dispersed in a phosphate buffer (0.05 M, pH = 4.4) to achieve a concentration of 10.0% (w/w, dsb). To ensure complete granular disruption, the suspension was heated in a boiling oil bath at 130 °C with vigorous stirring for 1 h. This high-temperature gelatinization step was essential to fully disentangle the resistant HAMS granules, consistent with protocols for high-amylose starches [29]. After cooling to 50 °C, pullulanase (9.75 U/g starch) was added. The mixture was incubated at 50 °C for varying durations (0, 4, 8, 12, and 24 h) to obtain products with different amylose contents. The enzymatic reaction was terminated by heating the suspension in a boiling water bath (100 °C) for 30 min. The suspension was centrifuged at 4500 rpm for 10 min. The resulting precipitates were dried at 45 °C for 12 h, thoroughly ground, and passed through a 125-μm sieve. The final samples were designated as DH4, DH8, DH12, and DH24, corresponding to their respective debranching times.

2.2.2. Preparation of DH-FA Complexes

The DH suspensions were maintained at 100 °C. MA and LOA were prepared in 40% (v/v) ethanol and separately added to the suspensions to achieve a lipid concentration of 20% (w/w, dsb). The mixtures were stirred at 100 °C for 1 h to facilitate complexation. The precipitates were collected by centrifugation (4500 rpm, 10 min), washed with absolute ethanol to remove free lipids, and dried at 45 °C. To further reinforce the structure, the complexes were subjected to PHT to induce the transition into thermally stable Type II polymorphs [30], or ANN to enhance crystalline perfection. Specifically, samples were adjusted to different moisture contents (10–30% for PHT; 60–80% for ANN) and treated at targeted temperatures (110–130 °C for PHT; 70–90 °C for ANN) in sealed stainless-steel vessels.

2.3. Heat–Moisture Treatment (HMT) of RS5 to Enhance Resistance

To further enhance the resistance and optimize the crystalline structure of the starch–lipid complexes, the resulting Type 5 Resistant Starch (RS5) was subjected to two distinct heat–moisture treatments (HMTs): pressure heating (PHT) and annealing (ANN). The substrate for all treatments was the DH24-FA complex, which demonstrated optimal RS content.

2.3.1. Pressure Heating (PHT)

A known quantity of dry DH24-FA complexes was weighed and placed into a glass container. The samples were adjusted to moisture contents of 10% and 30% (w/w) by adding the required amount of distilled water. The mixture was sealed and equilibrated at room temperature overnight to ensure homogeneous moisture distribution. The samples were then subjected to PHT using an autoclave at temperatures of 110 °C, 120 °C, and 130 °C for 1 h at each condition. After cooling to room temperature, the samples were dried at 45 °C overnight in a convection oven, and subsequently ground thoroughly to pass through a 125-μm sieve. The resulting PH-modified complexes were designated as PHT-T-M, where T is the temperature and M is the moisture content (e.g., PHT-110 °C-10%) [31].

2.3.2. Annealing (ANN)

Annealing treatment was applied to the DH24-MA and DH24-LA complexes following a method adapted from [31]. The DH24-FA complexes (4 g, dsb) were adjusted to high moisture contents of 60% and 80% (w/w) by mixing with the calculated amount of distilled water. The high moisture samples were sealed and equilibrated at room temperature overnight. The samples were then heated in an oven at temperatures of 70 °C, 80 °C, and 90 °C for an extended duration of 24 h at each condition. After the treatment, the samples were cooled to room temperature, dried at 45 °C overnight in an oven, and fully ground to pass through a 125-μm sieve. The resulting ANN-modified complexes were designated as ANN-T-M, where T is the temperature and M is the moisture content (e.g., ANN-70 °C-60%).

2.4. Determination of Amylose Content

The amylose content was determined using a commercial assay kit (K-AMYL; Megazyme, Bray, Ireland) based on the Concanavalin A (Con A) precipitation method, following the protocol described by [24]. Briefly, starch samples were completely dispersed in DMSO. The amylopectin fraction was specifically precipitated by Con A, and the amylose in the supernatant was enzymatically hydrolyzed to D-glucose. Concurrently, a separate aliquot of total starch was hydrolyzed to determine the total glucose concentration. The absorbance of the resulting glucose was measured at 510 nm using the GOPOD reagent. The amylose content was calculated using Equation (1):

$$[eqn]$$

where Asupernatant and Atotal starch are the absorbance values of the Con A supernatant and the total starch sample, respectively. The constants 6.15 and 9.2 represent the dilution factors for the respective extracts [24]. All measurements were performed in triplicate, with standardized corn starch (supplied in the kit) used as a control, and values were corrected against a reagent blank.

2.5. Molecular Weight of Amylose

The molecular weight of amylose in the samples was determined by gel permeation chromatography (GPC), the method referred to [32].The GPC analysis was performed using an Agilent 1260 Infinity II system (Agilent Technologies, Santa Clara, CA, USA). Chromatography-grade dimethyl sulfoxide (DMSO) was used as the mobile phase. The column temperature was maintained at 45 °C, and the solvent flow rate was set to 1 mL/min. The injection volume for each sample was 20 μL. The GPC system was equipped with a tandem column setup, consisting of a PLPolaegel-M column (7.5 × 50 mm, 8 μm) and a PLPolargel-M column (7.5 × 300 mm, 8 μm). Detection was carried out using a refractive index detector.

2.6. In Vitro Digestibility of DH-FA Complexes Modified by HMT

The digestion profiles of the RS5 complexes, following debranching and HMT modifications, were evaluated using a modified version of the method described by [33]. Briefly, starch samples (300 mg) and guar gum (25 mg) were dispersed in distilled water (7.5 mL) and gelatinized in a boiling water bath for 10 min. After cooling to room temperature, 2.5 mL of sodium acetate buffer (0.4 M, pH 5.2, containing 0.18% CaCl_2_) was added. The mixture was equilibrated at 37 °C with continuous shaking (160 rpm) for 15 min. Hydrolysis was initiated by the addition of an enzyme cocktail (2.5 mL) prepared by mixing porcine pancreatic α-amylase (Sigma-Aldrich, 8 U/mg) and amyloglucosidase (Sigma-Aldrich, 300 U/mL). Aliquots (250 μL) were withdrawn at specific intervals (20 min and 120 min) and immediately transferred into 10 mL of 66% ethanol (v/v) to terminate enzymatic activity. The suspension was centrifuged at 2500× g (4500 rpm) for 10 min. The content of released D-glucose in the supernatant was quantified employing a Megazyme assay kit (GOPOD format). Consequently, the fractions of rapidly digestible starch (RDS, hydrolyzed within 20 min), slowly digestible starch (SDS, digested between 20 and 120 min), and resistant starch (RS, remaining undigested after 120 min) were calculated based on the glucose release kinetics [33].

2.7. Complexation Index (CI) of DH-FA Complexes Modified by HMT

The CI was determined according to a slightly modified procedure described by [12]. Briefly, samples (0.3 g, dsb) were dispersed in centrifuge tubes to form a 6% (w/w) suspension. The slurry was fully gelatinized by heating in a boiling water bath for 20 min and subsequently cooled to room temperature. After adding deionized water (25 mL), the mixture was vortexed for 2 min and centrifuged at 4000 rpm for 15 min to separate the insoluble fraction. An aliquot of the supernatant (0.5 mL) was diluted with deionized water (15 mL) and mixed with 2 mL of iodine solution (2.0% KI and 1.3% I_2_, w/v). Native HAMS without FA was used as the blank control. The absorbance was recorded at 690 nm using a spectrophotometer (Unico UV-2800, Shanghai, China). All measurements were performed in triplicate. The CI value was computed using the following equation:

$$[eqn]$$

where Acontrol denotes the absorbance of the reference starch without lipids, and Acomplex corresponds to the absorbance of the starch–lipid complexes (sample).

2.8. Fourier-Transform Infrared (FTIR) Spectroscopy

The molecular framework of the specimens was examined using Fourier-Transform Infrared (FTIR) spectroscopy. This analysis was carried out at ambient temperature utilizing a Bruker INVENIO^®^ instrument (Bruker, Karlsruhe, Germany). For sample preparation, each material was combined with potassium bromide (KBr) at a 1:100 mass proportion and subsequently pulverized into a fine powder with a mortar. This homogeneous mixture was then compacted to form an IR-transparent KBr disk for transmission analysis. Spectral data were acquired across the 4000 to 400 cm^−1^ wavenumber range. The parameters were set to a resolution of 4 cm^−1^ with 32 scans accumulated for each specimen. Subsequent data processing, including baseline correction and spectral smoothing, was executed using OMNIC software (version 8.2.0).

2.9. X-Ray Diffraction (XRD)

Structural characterization was performed using X-ray diffraction (XRD) analysis. Data were collected with a Bruker D8 Advance diffractometer (Bruker AXS, Karlsruhe, Germany), which utilized a copper Kα radiation source (λ = 0.154 nm) set to 40 kV and 40 mA. The resulting diffraction patterns spanned a 2θ range from 4° up to 40°. Measurements were taken at a scanning speed of 2° per minute, applying a step size of 0.02°. The relative crystallinity (RC) percentage was quantified using Jade software (version 6.0).

2.10. Morphological Structure

To facilitate subsequent morphological visualization, the specimens were initially coated with a thin layer of gold via sputtering. The resulting surface topography was subsequently analyzed using a field-emission scanning electron microscope (FE-SEM). The Hitachi instrument (model SU8020, Hitachi High-Tech Corporation, Tokyo, Japan) was operated at an accelerating potential of 2.0 kV.

2.11. Thermal Properties of DH-FA Complexes Modified by HMT

The thermal behavior of the modified complexes was evaluated using a PerkinElmer DSC 4000 calorimeter (Waltham, MA, USA). This analytical procedure was slightly adjusted from the protocol detailed by [34]. In brief, starch specimens (10 mg) were loaded into stainless-steel high-pressure pans and mixed with distilled water to achieve a 1:3 (w/w) dry matter-to-water ratio. After sealing, the vessels were left to hydrate at ambient temperature for 2 h. Thermal scanning was performed against an empty reference crucible from 20 to 150 °C with a linear ramp of 10 °C/min. The onset temperature (T_O_), peak temperature (T_P_), conclusion temperature (T_C_), and enthalpy change (ΔH) were obtained from Pyris Software (Version 13.3). Each measurement was replicated three times.

2.12. Statistical Analysis

The collected data were initially arranged and processed utilizing Microsoft Excel (version 2019, Microsoft Corporation, Redmond, WA, USA). Subsequently, all graphical representations were generated using Origin software (version 8.5, OriginLab Corporation, Northampton, MA, USA). Statistical assessment involved a one-way analysis of variance (ANOVA). To identify specific differences among the means, this was followed by Tukey’s honestly significant difference (HSD) post hoc test. The criterion for statistical significance was established at the p < 0.05 level. All computations for the statistical analysis were executed using IBM SPSS Statistics (version 24, IBM Corp., Armonk, NY, USA).

3. Results

3.1. Amylose Content and Molecular Characteristics of HAMS After Pullulanase Debranching

The molecular characteristics of the starch substrate are crucial for the efficiency of starch–lipid complexation and the formation of RS5 inclusion complexes [35,36]. As summarized in Table 1, pullulanase treatment substantially increased the apparent amylose content of HAMS. The amylose content of native HAMS was 74.6%. After 4, 8, 12, and 24 h of debranching, the values rose to 83.0%, 85.4%, 87.1%, and 87.7% for DH4, DH8, DH12, and DH24, respectively (p < 0.05). The increase plateaued between 12 and 24 h, indicating that the generation of apparent amylose reached saturation within the first 12 h of enzymatic treatment.

This time-dependent pattern is consistent with the catalytic behavior of pullulanase, which preferentially hydrolyzes the α-1,6-glycosidic linkages at branch points [37,38]. The progressive removal of these linkages decreases branch density and shifts more chains into the linear, amylose-like soluble fraction quantified by the ConA assay [36,37]. The lack of a further significant increase in amylose content between DH12 and DH24 suggests that accessible branch points were largely cleaved by 12 h and that extended treatment primarily refined the chain-length distribution without altering the proportion of linear chains [37,38].

Gel permeation chromatography confirmed extensive molecular fragmentation. The weight-average molecular weight (Mw) declined markedly from 445,489 g/mol in native HAMS to 138,610 g/mol in DH24, indicating the breakdown of large, highly branched macromolecules into smaller linear species [39]. In addition to Mw, the number-average molecular weight (Mn) decreased from 124,114 g/mol to 30,513 g/mol (Table 1), confirming the effective cleavage of glycosidic bonds and the generation of shorter glucan chains essential for inclusion complex formation. While the PDI generally trended downward to 3.42 at DH24, a slight fluctuation was observed at DH12 (PDI = 4.57). This transient increase is likely attributed to the accumulation of intermediate-sized fragments at this specific stage of hydrolysis. Despite this fluctuation, the final reduction in PDI at DH24 reflects a transition towards a narrower and more uniform chain-length distribution. This homogeneity is functionally critical, as it facilitates the orderly alignment of amylose helices around fatty acid ligands, thereby promoting the formation of highly crystalline and thermally stable RS5 structures.

In this study, DH24 combined the highest apparent amylose content, the lowest molecular weight, and the smallest polydispersity among all debranched samples. These features indicate a predominance of mobile, linear glucan chains capable of stabilizing amylose helices. Consequently, DH24 was selected as the optimal substrate for subsequent fatty acid complexation and hydrothermal treatment.

3.2. In Vitro Digestibility Profile of RS5 Complexes Modified by Debranching and HMT

The in vitro hydrolysis of starch is mediated by the synergistic action of α-amylase and amyloglucosidase, while the enzymatic resistance of RS5 fractions is fundamentally governed by the hierarchical architecture of the starch–lipid complexes [40]. As shown in Figure 1A,B, pullulanase-induced debranching profoundly reshaped the digestibility profile. The RS fraction increased substantially from 38.0% in HAMS-MA complexes to 51.0% in DH24-MA, with a parallel trend observed in LOA systems where RS rose from 39.1% to 53.6%. Notably, the RS content exhibited a rapid rise between DH0 and DH12 followed by a plateau from DH12 to DH24. This trend mirrors the saturation of apparent amylose content described in Section 3.1, suggesting that once a critical threshold of short linear chains is generated, further debranching primarily refines chain length distribution rather than driving major shifts in digestibility. Mechanistically, this enhanced resistance is attributed to two synergistic factors facilitated by the liberated short linear chains. First, the increased abundance of mobile, uniform linear chains provides more potential nucleation sites for single helices, thereby promoting the formation of inclusion complexes with FAs [41,42,43,44,45,46,47]. Second, the hydration of these hydrophilic chains fosters a compact, erosion-resistant gel network that physically restricts enzymatic access to the substrate, favoring the accumulation of SDS and RS [43].

Hydrothermal processing, comprising PHT and ANN, further refined the semi-crystalline structure of the DH24-FA complexes, driving distinct digestibility shifts via moisture-mediated reorganization. In general, PHT exerted a more potent effect on RS enhancement than ANN for both MA and LOA complexes (Figure 1C,D). For example, PHT at 120 °C with 10% moisture yielded the highest RS content of 69.2% for DH24-LOA, surpassing the maximum of 64.8% achieved by ANN at 80 °C with 60% moisture. The superior effect of PHT suggests that the combination of high thermal energy and restricted moisture creates an optimal environment for the mobility of glucan chains within amorphous regions. This facilitates the perfection of existing crystallites and the nucleation of robust new microstructures that are highly resistant to hydrolysis [15,48,49]. However, thermodynamic stability limits were evident; when thermal input was excessive (e.g., 130 °C at 30% moisture), the RS content declined significantly. This reduction indicates the melting or unravelling of hexagonal Bravais lattice inclusion complexes back into a disordered, digestible state, potentially exacerbated by excessive swelling of the starch matrix which disrupts the protective gel network.

In the case of ANN-treated complexes, digestibility followed a characteristic pattern where RS initially increased with temperature and moisture but declined as conditions approached the gelatinization onset (Figure 1E,F). Mechanistically, annealing enhances resistance by hydrating amorphous regions to increase molecular mobility, allowing double helices to realign into more uniform and compact crystallites. However, excessive moisture or heat can induce granular swelling that compromises structural integrity and diminishes resistance [20,26,30,50,51]. Collectively, these findings demonstrate that while debranching provides the essential structural building blocks, optimized hydrothermal treatment acts as a critical assembly step. This synergistic modification transforms the starch matrix into a highly ordered, dense architecture that maximizes enzymatic resistance [52].

3.3. CI of RS5 Complexes Modified by Debranching and HMT

The Complexing Index (CI) serves as an indirect measure of the degree of starch–lipid complexation, calculated based on the reduction in iodine binding capacity. It reflects the extent to which fatty acids (FAs) are entrapped within the hydrophobic cavity of amylose helices [46]. Consistent with the variation in digestibility profiles, the CI values were significantly modulated by the extent of enzymatic debranching, the molecular configuration of the FA, and the specific hydrothermal conditions [32,44,46,53].

As illustrated in Figure 2A, the initial enzymatic debranching step played a fundamental role in promoting inclusion complex formation. DH12 and DH24 complexes displayed markedly higher CI values than their HAMS and DH4 counterparts, implying a greater abundance of amylose-FA inclusion structures in the extensively debranched samples. This trend aligns with the increased generation of short, linear glucan chains (Table 1) [46] and can be attributed to the removal of α-1,6-glycosidic linkages. Debranching reduces the steric congestion of the branched amylopectin architecture, liberating mobile linear chains that can readily coil around hydrophobic FA guests [46,49,54,55,56]. Furthermore, the substantially higher amylose content in DH12 and DH24 provides additional helical segments capable of accommodating FA molecules, thereby supporting the formation of a denser network of inclusion complexes [46]. Since a higher content of such complexes is associated with improved resistance to enzymatic hydrolysis [57], these results explain why DH12 and DH24 systems exhibited greater RS contents than the corresponding HAMS and DH4 complexes (Figure 1). Consequently, DH24 samples demonstrated the greatest complexation potential among the debranched groups.

It should be noted that lipid type also plays an important role in modulating the CI of inclusion complexes. DH24-MA complexes demonstrated significantly higher CI values (65.3 ± 1.0^a^ %) compared to DH24-LOA complexes (54.1 ± 1.6^d^ %). This disparity suggests that the shorter, saturated hydrocarbon chain of MA (C14:0) facilitates more effective insertion and compact packing within the amylose cavity compared to LOA (C18:2). The linear geometry of MA allows for stronger interaction with the debranched amylose molecules, resulting in a higher yield of inclusion complexes [19,58,59,60]. In contrast, the cis-double bonds and kinked configuration of unsaturated LOA create pronounced steric hindrance, which limits the efficiency of inclusion and disrupts the continuity of the helical order [19]. However, interestingly, despite the lower CI, LOA complexes exhibited comparable or even slightly higher RS content than MA complexes (Section 3.2). This suggests that while LOA forms fewer perfect inclusion complexes, the steric bulk of the uncomplexed or loosely bound LOA chains may potentially create a steric barrier around the starch matrix. This barrier likely obstructs enzyme binding sites, sustaining high enzymatic resistance even with lower crystallite perfection. It is worth noting that while PHT typically pose a risk of oxidation for unsaturated fatty acids (e.g., LOA), the complexation process itself offers a protective mechanism [35]. Recent studies indicate that the amylose single helix serves as an effective physical barrier [61]. By encapsulating the lipid ligands within its hydrophobic cavity, the compact helical structure restricts the contact between oxygen and the sensitive double bonds of PUFAs, thereby retarding oxidative deterioration and enhancing the thermal stability of the bioactive lipids [6,62].

Subsequent HMT further optimized the complexation efficiency of the DH24-FA substrates, although the effects were highly dependent on the moisture–temperature interplay [15]. For PHT, the highest CI values were generally observed at lower moisture contents (Figure 2B). Specifically, PHT at 120 °C with 10% moisture yielded the peak CI for DH24-MA (73.0 ± 1.1^a^ %), representing a significant improvement over the untreated complex. PHT facilitates the movement of glucan chains and residual free FA, promoting their re-association into stable inclusion complexes [42]. However, increasing the moisture content consistently resulted in lower CI values compared to the lower-moisture counterparts (e.g., CI at 120 °C-30% moisture was lower than at 120 °C-10%). This decline suggests that excessive moisture under high-pressure conditions induces excessive swelling or partial gelatinization, which may destabilize the newly formed helices [30]. Similarly, ANN enhanced the CI by perfecting the crystalline order, with the optimal effect observed at 80 °C and 60% moisture (CI ≈ 72.6 ± 1.2^a^ % for DH24-MA) (Figure 2C). This aligns with the mechanism wherein annealing hydrates the amorphous regions, increasing chain mobility sufficiently to eliminate structural defects and incorporate loosely bound lipids into the helical structure [20,26,51]. However, when the temperature was raised to 90 °C or moisture increased to 80%, the CI values plateaued or decreased. This indicates that thermal energy exceeding the gelatinization onset or excessive hydration can disrupt the granular integrity and thermodynamic stability of the inclusion complexes [30].

3.4. Molecular Structure of RS5 Complexes Modified by Debranching and HMT

Fourier-transform infrared (FTIR) spectroscopy is a robust analytical tool for characterizing the overall chemical environment of carbohydrate systems. As shown in Figure 3 and in band assignments summarized in Table S1, the IR spectra of debranched starch and the complexes modified by PHT or ANN exhibited similar characteristic spectral bands. Consistent with the fact that the starch spectrum is a superposition of vibrational modes from its various components [63], we identified the major bands as follows: A broad absorption band near 3418 cm^−1^ is attributed to the stretching vibrations of O-H groups from amylose and amylopectin molecules, as well as tightly bound water. Distinct peaks at 2930 and 2853 cm^−1^ correspond to the asymmetric and symmetric stretching vibrations of methylene (-CH_2_-) groups, respectively [64]. The overall IR spectra of all samples showed similar patterns, with no new bands appearing and no disappearance of existing ones (Figure 3A–D). This indicates that the treatments (debranching, complexation, and subsequent PHT or ANN) did not induce covalent chemical modifications. Instead, they promoted a physical rearrangement of the glucan polymers mediated by hydrogen bonding [30,54]. While the carbonyl stretching (C=O) of free fatty acids is typically observed near 1700 cm^−1^, this band was not prominent in the complex spectra. This is likely due to the low concentration of lipids relative to the starch matrix and the potential interference from the broad water bending vibration (δ(H-O-H)) at ~1640 cm^−1^. Notably, the broad region between 3100 and 3500 cm^−1^ showed a clear red shift, which generally suggests strengthened hydrogen-bond networks within the modified complexes [65].

It should be noted that many empirical studies in carbohydrate research frequently utilize the deconvoluted transmittance intensity ratios of specific IR bands (e.g., 1047/1022 cm^−1^ and 995/1022 cm^−1^) to estimate the short-range molecular order [66,67]. However, from a rigorous physical chemistry and quantum mechanical perspective, the IR spectrum of starch is a complex superposition of over 100 independent vibrational modes originating from various components (amylose, amylopectin, lipids, water, proteins, and nanocrystals). Comparing relative transmittance intensities lacks physical meaning unless the true absorption coefficients are calculated using the Beer-Lambert law, which strictly requires precise knowledge of the sample thickness (e.g., the exact thickness of the KBr disk). Furthermore, assuming that different functional bands (such as 1047 cm^−1^ and 1022 cm^−1^) belong to the same dispersion function or vectorial basis is fundamentally flawed in such a complex multi-component system. Therefore, in this study, the empirical calculation of these ratios was deliberately discarded. The IR spectra were utilized solely to qualitatively confirm the basic chemical environments and the absence of new covalent bonds. The structural order, crystallinity, and the transformation of amorphous segments into highly ordered nanocrystals possessing a hexagonal Bravais lattice were exclusively evaluated using rigorous X-ray diffraction analysis (Section 3.5).

3.5. Crystalline Structure of RS5 Complexes Modified by Debranching and HMT

The X-ray diffraction patterns were analyzed to identify the crystalline structures formed during the sequential modification (Figure 4). Native HAMS exhibited reflections at 2θ ≈ 5.4° and 17.2°, characteristic of nanocrystals with a hexagonal crystal structure [29,68]. Upon debranching and complexation, the peaks corresponding to the hexagonal crystal structure disappeared, and a new set of diffraction peaks emerged. Precise determination of the peak positions revealed distinct reflections at 2θ = 7.63°, 13.20°, and 20.37°. According to the crystallographic indexing for starch–lipid inclusion complexes [8], these reflections correspond to the (100), (110), and (210) planes of the hexagonal Bravais lattice, respectively. The concurrent appearance of this complete set of reflections confirms the successful rearrangement of amylose and amylopectin molecules from the native hexagonal crystal structure into highly ordered nanocrystals possessing a hexagonal Bravais lattice.

In the case of DH-MA (myristic acid) complexes, additional diffraction peaks were observed at 2θ ≈ 21.50° and 24.05°. These reflections do not belong to the hexagonal Bravais lattice. Instead, they correspond to the crystalline domains of aggregated free fatty acids [19]. This behavior can be ascribed to the different crystallization tendencies of saturated and polyunsaturated chains. MA has a straight, fully saturated C14 chain that readily packs into a dense lattice; excess MA molecules that are not accommodated within amylose helices, or that are expelled during retrogradation, can therefore segregate and crystallize as pure-fatty-acid domains, giving rise to distinct XRD peaks. In contrast, LOA (C18:2) contains two cis double bonds that introduce kinks into the hydrocarbon chain, markedly reducing its ability to form well-ordered crystals. As a result, “free” LOA is more likely to remain molecularly dispersed or to form poorly ordered, amorphous regions that do not generate separate diffraction peaks in the 2θ range examined, so no additional FA-crystal peaks are resolved in the LOA-containing complexes [44].

The relative crystallinity (RC) of the DH-FA complexes increased with debranching time (Figure 4A,B), indicating that progressive removal of α-1,6 linkages facilitates the formation of more ordered crystalline domains [42]. Debranching treatment using pullulanase produces short, linear glucan chains, which increases molecular mobility and enhances the likelihood of chain alignment and subsequent aggregation [69]. Such ordered packing of glucan polymers in turn favors the development of well-organized crystalline domains within the granule [70]. This trend agrees with previous observations that moderate debranching promotes the formation of nanocrystals possessing a hexagonal Bravais lattice and RC in HAMS-FA systems, while overly extensive debranching can eventually reduce crystallinity by weakening amylose–lipid interactions or lowering the availability of complexable chains [71]. In the present study, DH-MA complexes consistently displayed slightly higher RC than the corresponding DH-LOA samples, suggesting that the straight C14:0 chain is incorporated more efficiently into amylose helices than the kinked C18:2 chain. Compared with DH24 alone, complexation with MA or LOA sharpened the characteristic peaks of the hexagonal Bravais lattice and elevated RC, confirming that inclusion complex formation contributes substantially to the long-range ordering of the systems.

HMT further reorganized the crystalline architecture of the DH24-FA complexes (Figure 4C–F). PHT under intermediate conditions (e.g., 110–120 °C at 10–30% moisture) produced the most pronounced increase in RC for both MA and LOA systems. This is attributed to the combination of high temperature and restricted water, which plasticizes the amorphous and semi-crystalline regions just enough to permit chain rearrangement without complete gelatinization. Under these conditions, imperfect nanocrystals and retrograded domains can reorganize into more regular, thermally stable structures, while free amylose molecules are driven to form additional inclusion complexes [20]. When the PHT temperature was further increased to 130 °C, RC decreased slightly, implying that excessive thermal input partially melted less stable crystallites and disrupted the ordered lamellae, in agreement with earlier observations in RS and CI results, and in previously reported high-amylose starch–lipid systems [30,72].

Furthermore, re-evaluation of the XRD patterns using the second derivative criterion confirmed that the characteristic diffraction peaks at 2θ = 7.63° and 13.20° were completely absent in the ANN-treated samples (Figure 4E,F) and highly suppressed in the DH24-LOA complexes treated with PHT (Figure 4D). This indicates that under these specific processing conditions, the complexes exist predominantly in an amorphous state, lacking long-range crystallographic order.

3.6. Microscopic Structure of RS5 Complexes Modified by Debranching and HMT

Scanning electron microscopy (SEM) was used to characterize the morphological evolution of native and hydrothermally treated starch–lipid complexes (Figure 5). While native corn starch granules typically display a smooth, spherical or polygonal morphology [50], enzymatic debranching (DH) significantly altered this structure, leading to roughened surfaces and partial erosion. Subsequent complexation with lipids further intensified surface textural changes, likely due to the adsorption or adhesion of FAs onto the starch matrix [50].

Distinct morphological transitions were observed following different HMTs. PHT treated particles largely lost their original granular integrity, exhibiting irregular shapes, pores, and cracks. Notably, severe treatment conditions (high temperature and moisture) promoted the fusion of glucan polymers into large, irregular aggregates. While aggregation generally reduces the surface-to-volume ratio, the specific conditions optimal for RS formation (120 °C, 10% moisture) induced the formation of relatively larger, smoother, and more compact blocks. These dense structures create a steric barrier that physically restricts enzyme accessibility more effectively than loose aggregates, directly correlating with the highest enzymatic resistance observed (RS content peaking at 69.2%).

In contrast to the extensive aggregation seen in PHT, ANN treatment resulted in minimal changes to the overall particle size, suggesting that granular integrity was largely preserved. However, surface roughness increased markedly with rising temperature and moisture. Although the optimal ANN conditions (80 °C, 60% moisture) yielded a high RS content of 64.8%, this value remained lower than that of the PHT counterparts. The SEM images revealed that ANN samples possessed a rougher surface topology characterized by adhered debris of broken starch and lipids [50]. Unlike the dense, smooth encapsulation-like structures formed by PHT, this increased surface roughness and porosity in ANN samples likely provided more binding sites for amylolytic enzymes, thereby explaining the comparatively lower RS yield relative to PHT.

Furthermore, the fatty acid type significantly influenced the micromorphology. Under identical hydrothermal conditions, DH24-LOA complexes exhibited larger aggregate sizes and a smoother surface microstructure compared to their MA counterparts. This phenomenon can be attributed to the larger molecular volume and steric hindrance of the long-chain unsaturated LOA (C18:2), which likely facilitated “particle bridging” or the formation of a continuous lipid coating on the granule surface [71]. This denser structural barrier is consistent with the digestibility data, where DH24-LOA complexes achieved significantly higher RS levels (69.2% for PHT and 64.8% for ANN) compared to the DH24-MA complexes (63.2% and 58.8%, respectively) under equivalent processing conditions.

3.7. Thermal Properties of RS5 Complexes Modified by Debranching and HMT

Thermal stability is often correlated with starch digestibility resistance [50]. It is important to note that while the gelatinization of native starch is driven by the melting of amylopectin crystallites, the endothermic transitions observed in these modified complexes reflect the dissociation of the amylose–lipid inclusion complexes. All complex samples exhibited two distinct endothermic transitions (Figure 6). The endotherm observed at approximately 95–120 °C corresponds to the dissociation of less-ordered Type I complexes, while the higher-temperature endotherm around 120–140 °C is attributed to the melting of the more structurally organized Type II complexes [45,73]. An additional minor peak at lower temperatures was observed exclusively in DH24-MA samples. Since this temperature aligns with the melting point of pure myristic acid (Tm ≈ 53.5 °C), this transition is attributable to the melting of crystalline free fatty acids, which is consistent with the observation of distinct free fatty acid peaks (2θ ≈ 21.5° and 24.0°) in the XRD patterns (Section 3.5).

Comparing the DH24-MA and DH24-LOA complexes to the debranched starch control (DH24), a distinct numerical shift in transition temperatures was observed (Table S2), indicating enhanced thermal stability upon complexation. Specifically, DH24-MA exhibited consistently higher transition temperatures (TO/TP/TC) than DH24, increasing from 94.59/100.39/113.91 °C to 95.62/107.15/115.58 °C for Type I, and from 117.39/125.16/135.49 °C to 118.66/126.03/136.10 °C for Type II. A similar elevation in TP was observed for DH24-LOA (Type I: 101.59 °C; Type II: 127.02 °C) relative to DH24. This universal shift towards higher temperatures confirms that the nanocrystals possessing a hexagonal Bravais lattice formed by inclusion complexes possess superior thermodynamic stability compared to the retrograded amylose crystallites present in the debranched control.

The thermal properties of starch–lipid complexes are influenced by both temperature and moisture content during hydrothermal treatment [45]. HMT further reorganized these structures, with PHT and ANN exhibiting distinct effects under their optimal conditions for resistant starch formation. For PHT, the condition of 120 °C with 10% moisture resulted in the most dramatic enhancement of thermal stability. The peak temperature (TP) of the Type II complex in DH24-MA shifted significantly from 126.03 °C (untreated) to 133.98 °C (PHT). This substantial increase indicates that the high thermal energy of PHT effectively mobilized the molecular chains to form new, highly ordered crystalline domains that require significantly higher energy to disrupt [20]. In contrast, ANN at 80 °C with 60% moisture resulted in a more moderate increase in TP (to 128.26 °C). While ANN improved crystalline perfection through hydration-mediated rearrangement, it lacked the high-energy input required to induce the extensive formation of the ultra-stable Type II nanocrystals possessing a hexagonal Bravais lattice observed in PHT. Consequently, the superior thermal stability provided by PHT correlates directly with its higher efficacy in resisting enzymatic hydrolysis compared to ANN.

3.8. Mechanism of RS5 Complexes Modified by Debranching and HMT

The structural assembly of RS5 complexes in this study follows a hierarchical pathway involving enzymatic debranching, lipid complexation, and HMT as illustrated in Figure 7. The process initiates with the gelatinization of HAMS. This step disrupts the granular structure and exposes the internal amylose and amylopectin molecules. Subsequently, pullulanase selectively hydrolyzes the α-1,6-glycosidic linkages located at the branch points of amylopectin. This enzymatic cleavage removes the steric hindrance caused by the branched architecture and converts high-molecular-weight amylopectin into a pool of short, linear glucan chains. These liberated chains possess high conformational mobility, thermodynamically lowering the energy barrier required for the chains to wrap around hydrophobic ligands. Consequently, these chains serve as efficient hosts for FAs and facilitate the nucleation of inclusion complexes.

Following the initial complexation, HMT act as critical structural refinement stages. ANN operates via a hydration-mediated mechanism. Under conditions of high moisture and sub-gelatinization temperatures, water molecules plasticize the amorphous regions of the starch matrix. This hydration increases the mobility of the molecular chains and permits the lateral rearrangement of existing inclusion complexes into a more uniform register. However, the high moisture content during ANN can induce granular swelling. This swelling prevents the formation of a dense matrix and results in a structure with higher porosity compared to PHT samples.

In contrast, PHT functions as a high-energy densification process. Conducted at elevated temperatures with restricted moisture, PHT creates a unique thermodynamic environment. The high thermal energy mobilizes rigid molecular chains within both the amorphous and semi-crystalline regions. Simultaneously, the low moisture content elevates the melting temperature of the crystallites and prevents them from unraveling. This balance allows for the reorganization of imperfect inclusion complexes into highly ordered hexagonal Bravais lattice crystalline domains. Specifically, PHT at 120 °C with 10% moisture provides sufficient thermal input to organize sterically hindered lipid complexes into stable domains without causing hydrolytic degradation. The resulting structure is characterized by a multi-scale interlocking architecture. Hexagonal Bravais lattice nanocrystals are perfected at the molecular scale and packed into thermally stable domains at the crystalline scale. This hierarchy creates a formidable physical barrier that sterically hinders the binding and diffusion of amylolytic enzymes.

4. Discussion

In this study, the resistance of starch–lipid complexes to enzymatic digestion was modulated by the combined effects of molecular structure, fatty acid molecular geometry, and hydrothermal treatment conditions. Enzymatic debranching serves as the essential nucleation step. The removal of α-1,6-glycosidic linkages liberates short, linear glucan chains. These chains are thermodynamically preferred for wrapping around hydrophobic ligands compared to bulky amylopectin. The structural geometry of the fatty acid guest significantly modulates this assembly. Saturated MA facilitates tight initial packing and high crystallinity due to its linear conformation. Conversely, unsaturated LOA creates substantial steric hindrance due to its kinked structure. Although LOA complexes exhibit lower crystallite perfection compared to MA complexes, they demonstrate comparable or superior enzymatic resistance. This phenomenon suggests that the steric bulk of the uncomplexed or loosely bound LOA chains creates a significant physical barrier around the starch matrix. This barrier effectively obstructs enzyme binding sites and compensates for the lower structural order.

PHT demonstrated superior efficacy over ANN in enhancing resistance. This superiority is attributed to the ability of PHT to overcome higher energy barriers and induce structural densification. The high thermal energy provided by PHT mobilizes the rigid structures and hexagonal Bravais lattice nanocrystals within the amorphous regions without inducing complete gelatinization. This mobility allows for the repair of imperfect crystallites and the formation of new, dense aggregated structures. In contrast, ANN relies on a hydration-mediated glass transition mechanism. While ANN improves crystalline perfection, the high moisture content required for the process induces granular swelling. This swelling increases surface roughness and porosity, which prevents the formation of the highly dense, continuous matrix observed in PHT samples. Furthermore, thermodynamic stability limits were evident. Excessive thermal severity during PHT destabilized the equilibrium and led to the melting of hexagonal Bravais lattice nanocrystals. This reversion to an amorphous state highlights the necessity of precise thermal control.

Ultimately, the enhanced resistance is the result of a multi-scale structural evolution. At the molecular scale, HMT perfects the structural assembly of inclusion complexes. At the granular scale, PHT specifically promotes the formation of smooth, compact aggregates that physically exclude enzymes. This combination of helical perfection and macroscopic densification establishes a robust barrier against amylolysis.

5. Conclusions

This study developed a high-performance RS5 from HAMS by combining enzymatic debranching with hydrothermal modifications. The sequential process involving the pullulanase debranching of HAMS, complexation with FA, and subsequent hydrothermal treatment significantly reorganized the architecture of the material. Specifically, the resultant complexes exist as a semi-crystalline structure, where highly ordered nanocrystals possessing a hexagonal Bravais lattice are embedded within the amorphous starch matrix. Debranching acted as the foundational step by generating essential short linear chains to facilitate the initial nucleation of the inclusion complexes. Subsequent HMT served as a critical densification process. PHT at 120 °C with 10% moisture proved to be the most effective strategy. This condition drove the reorganization of imperfect amorphous structures into highly ordered and thermally stable nanocrystals possessing a hexagonal Bravais lattice that effectively hindered enzymatic hydrolysis. Structural characterization confirmed that the enhanced resistance was directly correlated with higher crystallinity and greater thermal stability. These findings provide a theoretical framework for the industrial production of high-performance RS5 ingredients. These modified complexes show significant potential for application in low-glycemic staple foods, such as noodles and bakery products, as well as functional dietary fiber supplements. Future work should focus on the in vivo physiological effects of these specific complexes to validate their prebiotic potential.