Structural effects of microalgae additives on the starch gelatinisation process

1 STRUCTURAL EFFECTS OF MICROALGAE ADDITIVES ON THE STARCH 1 GELATINISATION PROCESS 2 3 Marta Martínez-Sanz*, María José Fabra, Laura G. Gómez-Mascaraque and Amparo López-Rubio 4 5 6 Food Safety and Preservation Department, IATA-CSIC, Avda. Agustin Escardino 7, 46980 7 Paterna, Valencia (Spain). 8 9 *Corresponding author: Tel.: +34 963200022; fax: +34 963636301 10 E-mail address: [email protected] 11 12 2 Abstract 13 This work presents a detailed structural characterisation of the starch gelatinisation process 14 and the effect of the addition of three microalgae species, Nannochloropsis gaditana sp., 15 Scenedesmus almeriensis and Spirulina, by means of an advanced approach consisting of 16 temperature-resolved simultaneous SAXS/WAXS experiments, combined with DSC. 17 Furthermore, regular and high amylose corn starch were utilised to evaluate the impact of 18 the amylose content. 19 The presence of microalgae has been seen to limit water accessibility towards the interior of 20 starch granules, reducing granule swelling and, thus, hindering the arrangement of 21 amylopectin helices into highly ordered crystalline structures. As a result, more 22 heterogeneous lamellar structures, with reduced apparent crystallinity, are attained. Despite 23 the existence of lipidic compounds in the three microalgae species, the tough cell walls in 24 Nannochloropsis and Scenedesmus impede their release towards the aqueous medium. In 25 contrast, the weak cell walls in Spirulina are disrupted by stirring, allowing cell components 26 to be released. The diffused lipids form helical inclusion complexes with the amylose chains 27 and promote the crystallisation of V-type structures. The presence of amylose-lipid 28 complexes counteracts the limited water swelling effect and results in the formation of more 29 crystalline and homogeneous lamellar structures. This result is relevant for the food industry 30 due to the potential of Spirulina to affect the processability and nutritional characteristics of 31 starch-based products. 32 33 Keywords: starch; microalgae; gelatinisation; SAXS; WAXS; DSC 34 35 3 1. Introduction 36 Starch is one of the most widely known dietary polysaccharides since it is the main 37 component in raw and processed foods. Native starch is found in vegetal resources forming 38 water-insoluble semi-crystalline granules which present a complex hierarchical architecture 39 characterised by at least four structural levels: (i) granules, (ii) semi-crystalline amorphous 40 growth rings, (iii) lamellae (i.e. stacks of alternating amorphous and crystalline structures) 41 and (iv) linear amylose and branched amylopectin chains. The crystalline regions are thought 42 to consist mainly of amylopectin side chains which are organized in double helices (Imberty, 43 Buléon, Tran, & Pérez, 1991), whereas the amorphous regions are mainly composed of linear 44 amylose chains and less ordered amylopectin branch points (Pérez & Bertoft, 2010). 45 Depending on several factors such as the amylose/amylopectin ratio and the amylopectin 46 chain length, these helices can be arranged differently into A-type (mainly found in cereal 47 starches and related to short double helices) or B-type (long double helices found in tubers 48 and high amylose starches) crystalline unit cells (Cheetham & Tao, 1998; Hoover, 2001; 49 Paul J Jenkins, Cameron, & Donald, 1993; Salman, et al., 2009). Water has a great influence 50 in the long range order of starches, constituting up to 4-7% and 25-27% of the crystalline 51 unit cell in the A and B polymorphs, respectively (Imberty, Chanzy, Pérez, Bulèon, & Tran, 52 1988; Imberty & Perez, 1988; Sarko & Wu, 1978). 53 54 Many food processing methods involve the use of combined heat and humidity, leading to 55 starch gelatinisation which involves the break-up of the starch structure. Essentially, the 56 gelatinisation process is initiated by the access of water towards the interior of starch 57 granules, which swell and absorb water. Subsequently, structural changes occur at the 58 different architectural levels, leading to crystallinity losses, disruption of the crystalline and 59 lamellar structure of starch and amylose molecules leaking out of the starch granules forming 60 4 a continuous gel. This complex gelatinisation phenomenon is governed by a number of 61 factors such as starch botanical origin, plant growth conditions, extraction methods, water 62 content, presence of additives, heating rate and thermal history (Imberty, et al., 1991; Liu, 63 Yu, Xie, & Chen, 2006; Tester & Morrison, 1990; Varavinit, Shobsngob, Varanyanond, 64 Chinachoti, & Naivikul, 2003; Waigh, Gidley, Komanshek, & Donald, 2000). A detailed 65 understanding of the structural changes taking place during starch gelatinisation, at the 66 different length-scales, as well as the effect of different parameters, is highly relevant to 67 several scientific and industrial fields. For instance, in many food processing methods such 68 as baking of bread, extrusion of cereal-based products, thickening, and gelling of sauces, 69 starch gelatinisation is a key factor to produce a desirable texture or consistency of the end 70 product (Biliaderis, Maurice, & Vose, 1980). From the human nutrition perspective, the 71 degree of starch gelatinisation in food products is known to affect its digestion rate (Joergen 72 Holm, Lundquist, Björck, Eliasson, & Asp, 1988). Furthermore, the gelatinisation process is 73 the basis for the processing methods applied to produce starch-based biopolymeric materials 74 (Li, et al., 2011; Liu, Xie, Yu, Chen, & Li, 2009). 75 76 Microalgae constitute one the most promising sustainable feedstocks for the manufacture 77 of plant-derived products. Due to their ease of cultivation, high growing rates and 78 productivity and possibility of adapting the harvesting conditions to modify their 79 composition, microalgae have gained a great deal of interest for their use in a wide range 80 of applications such as the production of biofuels (Mata, Martins, & Caetano, 2010) and 81 food commodities (Draaisma, et al., 2013; Spolaore, Joannis-Cassan, Duran, & Isambert, 82 2006). Their interest from the human nutrition perspective lays in their interesting chemical 83 composition. In particular, the high protein content of several species and their interesting 84 lipidic profile, with relatively high amounts of ω3 and ω6 fatty acids (Spolaore, et al., 2006), 85 5 make them attractive for the production of nutritional supplements or food additives. The 86 incorporation of microalgae into pasta and bakery products to enhance their nutritional 87 profile is currently being investigated and has been reported in several recent works (De 88 Marco, Steffolani, Martínez, & León, 2014; Monica Fradique, et al., 2010; Mónica 89 Fradique, et al., 2013; Kadam & Prabhasankar, 2010). The components present in the 90 microalgae may interfere with the gelatinisation process in starchy foods and, thus, the 91 processing conditions may also be affected. Therefore, investigating the effect of different 92 microalgae on the gelatinisation of starch may be of interest to determine their impact on 93 the processability and digestibility of starch-microalgae blends. In this work, we have 94 investigated the structural changes undergone by two starches, with different amylose 95 contents, as well as the effect of three different microalgae species, during the gelatinisation 96 process. To this end, an advanced approach of temperature-resolved simultaneous small 97 angle and wide angle X-ray scattering (SAXS/WAXS) experiments, combined with DSC 98 characterisation, has been utilised. The use of scattering techniques is particularly 99 advantageous to study the starch gelatinisation process, since it enables the characterisation 100 of different structural levels (crystalline and lamellar structures) in starch as the temperature 101 is raised and at high relative humidity conditions. 102 103 2. Materials and methods 104 2.1 Materials 105 Corn starch (27-28% amylose) and high amylose starch (70% amylose) powders were 106 supplied by Roquette (Roquette Laisa España, Benifaió, Spain). The three different 107 microalgae species, i.e. Nannocloropsis gaditana sp., Spirulina and Scenedesmus 108 almeriensis, in the form of dry powders, were kindly donated by Dr. Acién from the 109 University of Almeria (Spain). 110 6 111 2.2 TEM characterisation of microalgae cell walls 112 The raw microalgae were dispersed in water by vortex stirring at a concentration of 4 g/L. 113 One drop (8 μl) of the prepared suspensions was allowed to dry on a carbon-coated grid (200 114 mesh). The microalgae cell walls were stained with a 2 % (w/w) solution of uranyl acetate. 115 TEM was performed using a JEOL 1010 equipped with a digital Bioscan (Gatan) image 116 acquisition system at 80 kV. 117 118 2.3 Preparation of starch and starch/microalgae dispersions 119 Starch dispersions (samples designated as “corn” and “amylo” for the corn starch and high 120 amylose starch, respectively) were prepared by adding 0.5 g of starch into 1 mL of water 121 and subjecting the samples to vortex stirring for 2 min. For the starch/microalgae samples, 122 0.004 g of the microalgae powder were also added prior to stirring. The dispersions were 123 immediately used for the SAXS/WAXS and the DSC experiments. 124 125 2.4 Temperature resolved SAXS/WAXS experiments 126 Combined small and wide angle X-ray scattering (SAXS and WAXS, respectively) 127 experiments were carried out in the Non Crystalline Diffraction beamline, BL-11, at ALBA 128 synchrotron light source (www.albasynchrotron.es). The starch and starch/microalgae 129 dispersions were placed in sealed 2 mm quartz capillaries (Hilgenburg Gmbh, Germany). 130 The energy of the incident photons was 12.4 KeV or equivalently a wavelength, λ, of 1 Å. 131 The SAXS diffraction patterns were collected by means of a 9 CCD detector Quantum 132 ADSC 315r with an active area of 315 x 315 mm2, an effective pixel size of 102 x 102 µm2 133 and a dynamic range of 16 bits. The sample-to-detector distance was set to 6488 mm, 134 resulting in a q range with a maximum value of q = 0.25 Å-1. Additionally, the WAXS 135 7 diffraction patterns were collected by means of a 3 CCD detector Rayonix LX255-HS with 136 an active area of 85 x 255 mm2, an effective pixel size of 44 x 44 µm2 and a dynamic range 137 of 16 bits. In this case, the sample-to-detector distance was set to 144.9 mm, corresponding 138 to a maximum q value of 7.87 Å-1. This detector was tilted with a pitch of 26.4 degrees. 139 Based on previous experiments, an exposure time, of 2 seconds was selected for both 140 detectors. 141 142 Samples were heated from 30 ºC to 110 ºC at a heating rate of 2 ºC/min. Data were collected 143 in frames of 30 seconds, followed by a period of 30 seconds in which the samples were 144 protected from the beam by a local shutter. Each data frame thus corresponds to a 145 temperature range of 1 ºC, with one data frame every 2 ºC. The data reduction was treated 146 by pyFAI python code (ESRF) (Kieffer & Wright, 2013), modified by ALBA beamline 147 staff, to do on-line azimuthal integrations from a previously calibrated file. The calibration 148 files were created from well-known standards, i.e. Silver behenate (AgBh) and Cr2O3 for 149 SAXS and WAXS respectively. The intensity profiles were then represented as a function 150 of q (SAXS) and 2θ (WAXS) using the IRENA macro suite (Ilavsky & Jemian, 2009) 151 within the Igor software package (Wavemetrics, Lake Oswego, Oregon). 152 153 2.5 SAXS/WAXS data fitting 154 SAXS data were fitted using the Igor NIST analysis macro suite (Kline, 2006) and applying 155 a mathematical function consisting of a power-law term plus one Gaussian-Lorentzian 156 peak, similar to that previously reported for several starch samples (A. Lopez-Rubio, et al., 157 2007; Salman, et al., 2009): 158 8 𝐼𝐼(𝑞𝑞)=𝐴𝐴∙𝑞𝑞−𝑚𝑚+�𝑅𝑅·�𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚·�1 + �2·(𝑞𝑞−𝑞𝑞𝑚𝑚𝑚𝑚𝑚𝑚) Δ𝑞𝑞 �2�−1��+�(1−𝑅𝑅)·�𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚·𝑒𝑒𝑒𝑒𝑒𝑒�−1 2·159 �𝑞𝑞−𝑞𝑞𝑚𝑚𝑚𝑚𝑚𝑚 Δ𝑞𝑞 �2���+ 𝑏𝑏𝑏𝑏𝑏𝑏 (1) 160 The first term in equation (1) corresponds to the power-law function (where 𝐴𝐴 is a prefactor 161 and 𝑚𝑚 is the power-law exponent) to account for the underlying diffuse scattering, the 162 second and third terms correspond to the Lorentzian and Gaussian functions used to 163 describe the starch lamellar peak (where 𝑞𝑞𝑚𝑚𝑚𝑚𝑚𝑚 is the peak position, 𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚 is the intensity of 164 the peak, Δ𝑞𝑞 is the full width at half maximum and 𝑅𝑅 is the Lorentzian to Gaussian ratio of 165 the peak shape) and the fourth term accounts for the incoherent background. 166 167 WAXS peak fitting was performed in Igor, following the procedure described in a previous 168 work (Amparo Lopez-Rubio, Flanagan, Gilbert, & Gidley, 2008). The obtained values 169 from the fitting coefficients are those that minimize the value of Chi-squared, which is 170 defined as: 171 𝜒𝜒2=∑�𝑦𝑦−𝑦𝑦𝑖𝑖 𝜎𝜎𝑖𝑖�2 (2) 172 where y is a fitted value for a given point, 𝑦𝑦𝑖𝑖 is the measured data value for the point and 173 𝜎𝜎𝑖𝑖 is an estimate of the standard deviation for 𝑦𝑦𝑖𝑖. The curve fitting operation is carried out 174 iteratively and for each iteration, the fitting coefficients are refined to minimize 𝜒𝜒2. The 175 crystallinity index XC was determined from the obtained fitting results by applying the 176 following equation: 177 100(%) ×= ∑ Total Crystal C A A X (3) 178 where ATotal is the sum of the areas under all the diffraction peaks and ΣACrystal is the sum of 179 the areas corresponding to the crystalline peaks. 180 9 181 2.6 Differential Scanning Calorimetry (DSC) 182 DSC measurements of starch and starch/microalgae dispersions were performed on a Perkin-183 Elmer DSC 8000 thermal analysis system using N2 as the purging gas. Approximately 10 184 mg of samples were weighted and added into hermetically sealed aluminium sample pans. 185 The sample treatment consisted of heating step from 0 ºC to 180 ºC at a heating rate of 2 186 ºC/min. Before evaluation, similar runs of an empty pan were subtracted from the 187 thermograms. The DSC equipment was calibrated using indium as a standard. Measurements 188 were done, at least, in triplicate. 189 190 3. Results and Discussion 191 3.1 Thermodynamic characterisation of the starch gelatinisation process 192 The gelatinisation process of starch and starch-microalgae suspensions in excess water was 193 studied by means of DSC characterisation by heating the samples at a rate of 2ºC/min. 194 Typical DSC profiles for the corn starch and high amylose based samples are shown in 195 Figures 1A and 1B, respectively, and the extracted gelatinisation parameters are summarised 196 in Table 1. All the samples show a broad melting peak that corresponds to the gelatinisation 197 endotherm. In the case of corn starch the endotherm extends over a temperature range of ca. 198 60-70ºC with the maximum at 64 ºC, whereas the process takes place within the range of ca. 199 70-90ºC, with the maximum at 78ºC, for the high amylose starch. Similar gelatinisation 200 temperatures have been previously reported for corn starch (Liu, et al., 2006; Yoshimura, 201 Takaya, & Nishinari, 1996) and high amylose starch (Liu, et al., 2006) samples. The higher 202 gelatinisation temperature for the high amylose starch is partly a result of the amylose 203 fraction being less prone to swelling than the amylopectin (Hermansson & Svegmark, 1996; 204 Yuryev, Kalistratova, van Soest, & Niemann, 1998). Moreover, despite the higher packing 205 16 The estimated crystallinity index values for the samples at the initial temperature of 30ºC, 302 listed in Table 2, evidence a structural effect of the microalgae in decreasing the apparent 303 crystallinity of the starch granules, being this effect more obvious for the Nannochloropsis 304 species. Hydration has been shown to be a relevant factor affecting starch crystalline 305 structure, with a positive linear relationship between moisture content and crystallinity 306 index (Cheetham, et al., 1998; Qiao, et al., 2017). This is due to the fact that water favors 307 molecular motion and chain organization, thus promoting the formation of more ordered 308 structures (Qiao, et al., 2017). Therefore, the decreased apparent crystallinity observed for 309 the starch-microalgae suspensions is most likely a consequence of a limited moisture 310 accessibility towards the interior of starch granules, i.e. a certain amount of the added water 311 is taken up by the microalgae. The WAXS apparent crystallinity drop is in contrast with the 312 DSC results, which show that the gelatinisation enthalpy of corn starch is unaffected by the 313 presence of microalgae. The discrepancy between DSC and WAXS arises from (i) the lower 314 sensitivity of DSC to determine crystalline transitions as compared to WAXS and (ii) the 315 susceptibility of these techniques to identify crystalline or molecular order. With regards to 316 (i), it has been shown that X-ray scattering is indeed able to detect residual crystallinity at 317 temperatures higher than those at which the DSC gelatinisation endotherm is completed 318 (Wang, Zhang, Wang, & Copeland, 2016). Furthermore, in relation to (ii), whereas the 319 thermal transition determined by DSC accounts for the dissociation and unwinding of 320 helical structures (amylopectin double helices, amylose-lipid complexes and V-type single 321 helices), WAXS is only susceptible to those helical structures which are packed forming 322 crystallites. Thus, the presence of microalgae does not perturb the molecular order (i.e. 323 amount of helical structures) but restrains water access towards the interior of the granules, 324 hindering the organisation of these helices into crystallites and hence, reducing the apparent 325 crystallinity estimated by WAXS. The temperature range at which the crystallinity starts 326 17 decreasing for all the samples, i.e. 60-70ºC (cf. Figure 4A), is well correlated to the 327 gelatinisation temperature range determined from the DSC characterization (cf. Table 1). 328 329 330 Figure 2. Temperature-resolved WAXS patterns of corn starch samples. (A) pure corn 331 starch, (B) corn starch-Spirulina, (C) corn starch-Nannochloropsis and (D) corn starch-332 Scenedesmus. 333 334 As shown in Figure 3, the WAXS patterns from the high amylose corn starch samples differ 335 significantly from the corn starch patterns, with peak positions characteristic of the B-type 336 18 crystalline allomorph. A transition from A-type to B-type crystallinity has been previously 337 reported for corn starches when increasing the amylose content across a range of 0-84% 338 (Cheetham, et al., 1998). This is a consequence of longer amylopectin helical chains found 339 in high amylose starches, which are preferentially crystallised into the B-type allomorph 340 (Gernat, Radosta, Anger, & Damaschun, 1993). As observed for the corn starch, the peak 341 indicative of V-type crystallinity is detected in the sample containing Spirulina and sharp 342 peaks are visible in the scattering patterns of high amylose starch with Spirulina (peaks 343 located at ca. 15.0º, 20.4º, 26.6º and 29.2º) and Scenedesmus (peaks located at ca. 23.0 º 344 and 29.7º). The formation of V-type crystallites as a result of amylose-lipid complexes may 345 have important implications not only for the industrial processing of starch-based products, 346 but also from the nutritional perspective. For instance, it has been reported that when 347 amylose is complexed with lipids it is hydrolyzed and absorbed in the gastrointestinal tract 348 to the same extent as free amylose but at a slower rate (J. Holm, et al., 1983). Additionally, 349 the V-type amylose-lipid complexes have been shown to produce a decrease in the amylose 350 solubility, hence increasing the gelatinisation temperature (Eliasson, Carlson, & Larsson, 351 1981) and postpone retrogradation, thus enabling longer storage times (Krog, 1971). It 352 should be mentioned that the fact that a strong impact on the gelatinisation temperature is 353 not clearly observed in this work is most likely related to the low microalgae loadings added 354 into the starch aqueous suspensions (only 0.8 wt.-% with regards to the starch weight). The 355 calculated crystallinity values at 30ºC are 21.5%, 26.8%, 13.9% and 17.9% for the high 356 amylose, high amylose-Spirulina, high amylose-Nannochloropsis and high amylose-357 Scenedesmus samples, respectively. It is also worth noting that, as seen in Figure 4B, the 358 crystallinity of all the samples decreases significantly within the temperature range of ca. 359 70-90ºC, which is in agreement with the temperature range determined from the DSC 360 gelatinisation endotherms. However, it should be noted that, in agreement with previous 361 19 studies (Wang, et al., 2016), the samples still present residual structural order at 362 temperatures higher than the DSC gelatinisation endotherm end temperature. It is also 363 worth noting that even though at the final temperature of 110ºC the starch crystalline 364 structure has been almost completely destroyed, the peaks characteristic of the crystallised 365 lipids present in the Spirulina and Scenedesmus microalgae are still detected, indicating the 366 high thermal stability of these compounds. 367 368 The fact that the corn and high amylose starch present similar crystallinity values contrasts 369 with previous works stating that the crystallinity index is negatively correlated with the 370 amylose content (Gernat, et al., 1993). This may be true for low to intermediate amylose 371 contents, where the A-type allomorph is preserved; however, it could be hypothesised that 372 the longer helical structures present in starches with greater amylose contents and which 373 are crystallised into the B-type conformation, do not significantly affect the overall 374 crystallinity index. In fact, it has been shown that increasing the amylose content in B-type 375 starch does not have a strong impact on the crystallinity as it does in A-type starch 376 (Cheetham, et al., 1998). The complexation of amylose with the lipids from the added 377 microalgae seems to have a positive impact on the starch crystallinity. Thus, while the 378 general trend when incorporating microalgae into the suspensions is a reduction in the 379 apparent crystallinity (especially in the case of the Nannochloropsis) due to a reduced 380 swelling of the starch granules, this effect is reverted in the case of the high amylose starch 381 loaded with Spirulina, as in this case a greater amount of lipid-amylose complexes are 382 formed. 383 384 20 385 Figure 3. Temperature-resolved WAXS patterns of high amylose starch samples. (A) pure 386 high amylose starch, (B) high amylose starch-Spirulina, (C) high amylose starch-387 Nannochloropsis and (D) high amylose starch-Scenedesmus. 388 389 Table 2. Lamellar peak parameters and crystallinity (XC) for the native samples (T=30ºC). 390 The lamellar peak parameters were obtained by fitting the experimental SAXS data to the 391 sum function of a power-law plus a Gaussian-Lorentzian peak and crystallinity was obtained 392 from the peak fitting of the WAXS patterns. Standard deviation values on the last digit are 393 shown in parentheses. 394 21 Lamellar peak position (Å-1) Intensity of the lamellar peak (a.u.) Width of the lamellar peak (Å-1) Lamellar repeat distance (Å) XC (%) Corn 0.06358 (1) 110.4 (1) 0.0258 (1) 9.88 21.3 (3) Corn-Spirulina 0.06335 (1) 72.6 (1) 0.0233 (1) 9.92 16.7 (4) CornNannochloropsis 0.06418 (2) 53.9 (1) 0.0269 (1) 9.79 13.9 (2) CornScenedesmus 0.06390 (1) 62.8 (1) 0.0257 (1) 9.83 18.0 (3) High Amylose 0.06508 (1) 21.2 (1) 0.0239 (2) 9.65 21.5 (2) High AmyloseSpirulina 0.06575 (2) 16.2 (1) 0.0197 (1) 9.56 26.8 (3) High AmyloseNannochloropsis 0.05239 (2) 24.6 (1) 0.0322 (3) 11.99 13.9 (2) High AmyloseScenedesmus 0.06456 (2) 14.4 (1) 0.0250 (2) 9.73 17.9 (2) 395 396 22 397 Figure 4. Evolution of the crystallinity index for the corn starch (A) and high amylose starch 398 (B) samples. 399 400 3.2.2. Lamellar structure 401 The crystalline regions constituted by the tightly packed helical structures are alternated with 402 amorphous regions, forming the next structural level known as lamellae. Due to their 403 different physical density, the crystalline and amorphous domains present distinct X-ray 404 scattering length density and, thus, the presence of these lamellar structures gives rise to the 405 appearance of scattering features in the SAXS patterns of starch samples. To investigate the 406 structural changes taking place in the lamellar arrangement during the gelatinisation process, 407 as well as the effect of the microalgae addition, the SAXS patterns were collected 408 simultaneously to the WAXS patterns when heating up the samples. As shown in Figure 5, 409 at the initial temperature of 30ºC all the samples present a well-defined peak located at ca. 410 0.06 Å-1, which corresponds to the lamellar peak characteristic of hydrated native starch 411 (Blazek & Gilbert, 2011; Chanvrier, et al., 2007; Salman, et al., 2009). The position of this 412 peak is related to the thickness of crystalline and amorphous lamellae. Additionally, the peak 413 23 area depends on the scattering length density difference between the crystalline and 414 amorphous domains. To extract detailed structural information from the obtained SAXS 415 patterns, the experimental data within the q range of 0.045-0.2 Å-1 were fitted using a 416 mathematical function based on the sum of a power-law plus a Lorentzian-Gaussian peak 417 (cf. section 2.5) and the main parameters extracted from the fits are gathered in Table 2. The 418 first observation is that while the lamellar distance is not strongly affected by the starch 419 amylose content, remaining within the range of 9-10Å typically reported for native starches 420 (P. J. Jenkins & Donald, 1995), the intensity of the lamellar peak is greatly reduced for the 421 high amylose starch. This is probably originated by a lower scattering length density contrast 422 in this sample. In fact, it has been reported that while the lamellar distance remains constant, 423 the size of the crystalline regions increases and the electron density difference between the 424 crystalline and amorphous regions decreases with the amylose content (P. J. Jenkins, et al., 425 1995). This is a consequence of the larger amylopectin chain length observed for high 426 amylose starches, which results in the formation of the less densely packed B-type 427 crystallites. A-type structures are closely packed with water molecules between each double 428 helical structure, whereas B-type are more open and water molecules are located in the 429 central cavity formed by six double helices (Imberty, Buléon, Tran, & Pérez, 1991). Thus, 430 in low amylose starches (with short amylopectin chains which are crystallised into the A-431 type structure) the crystalline lamellae are smaller but more densely packed, whereas in high 432 amylose starches (with B-type crystallites formed by large amylopectin chains) crystalline 433 lamellae become larger but more loosely packed. 434 435 As clearly observed in Figure 5 and confirmed by the fitting parameters, the incorporation 436 of microalgae into the corn starch suspensions leads to a marked decrease in the relative 437 intensity of the lamellar peak. Conversely, the peak position and width are not significantly 438 24 affected. This indicates that the addition of microalgae reduces the scattering length density 439 contrast but does not affect the lamellae structural parameters. Together with the DSC and 440 WAXS characterisation, this suggests that the presence of microalgae does not affect the 441 molecular order but impairs the organisation of these helical structures into densely packed 442 and ordered crystalline domains by limiting to some extent the swelling of starch granules. 443 The reduced apparent crystallinity would then contribute to the observed decrease in the 444 scattering length density contrast. In the case of the high amylose starch, the presence of 445 microalgae does not have a strong effect on the lamellar peak intensity, but the peak position 446 and width are affected differently depending on the microalgae species. A significantly 447 broader lamellar peak, with a greater associated lamellar repeat distance of ca. 12Å is 448 observed upon incorporation of Nannochloropsis, being indicative of more heterogeneous 449 lamellae in the presence of this microalgae species. This effect can also be explained by a 450 decreased water accessibility towards the interior of starch granules, since it has been 451 reported that in the dry state only a small fraction of starch helices can be aligned and packed 452 to construct the semi-crystalline lamellar structure (Qiao, et al., 2017).Together with the 453 reduced crystallinity estimated from the WAXS patterns, this result indicates that the 454 presence of Nannochloropsis has a strong impact on the swelling mechanism of starch 455 granules. In contrast, the addition of Spirulina has the opposite effect, i.e. a greater degree 456 of crystalline order is attained (as indicated by the higher crystallinity index) and more 457 homogeneous lamellae are formed (as suggested by the reduced lamellar peak width). The 458 distinct effect of these microalgae on the starch structure and the gelatinisation process may 459 be explained based on their different composition and, mostly, on their distinct cell wall 460 structure. The total lipidic contents have been reported as ca. 6% for Spirulina (Matos, et al., 461 2016), 8-16% for Nannochloropsis (Matos, et al., 2016) and up to 30% for Scenedesmus 462 microalgae (Jiang, et al., 2017; Ma, et al., 2017). However, the Nannochloropsis microalgae 463 25 have been shown to possess a tough cell wall, formed by a cellulosic inner layer and an outer 464 layer rich in algaenans, which requires intense sonication treatments to be disrupted (Fabra, 465 et al., 2017; Scholz, et al., 2014). The existence of such resistant cell walls would then 466 impede the release of lipids from the cells towards the aqueous medium. The cell wall 467 structure of the Scenedesmus species also contains an inner cellulosic layer, which is then 468 covered by a thin middle layer and an outer pectic layer (Bisalputra & Weier, 1963). On the 469 other hand, the cell walls in Spirulina seem to be richer in proteins and they present porous 470 features which may facilitate the release of cell components (Van Eykelenburg, 1977). To 471 corroborate their different cell wall integrity, raw microalgae were characterised by TEM 472 and the crystalline components in their cell walls were stained with uranyl acetate. As 473 observed in Figure S2, while the Nannochloropsis microalgae presented intact cell walls, 474 intensely stained and with clearly defined edges, some of the cell walls in Spirulina 475 microalgae seemed to be disrupted as a result of the sample preparation process. Therefore, 476 despite the lower lipidic content of Spirulina, these components could more easily be 477 released through the weak cell walls, being then able to interact with the amylose fraction in 478 starch and forming complexes which induce a higher degree of crystallinity, as well as 479 promoting the formation of more homogeneous lamellae. 480 481 32 De Marco, E. R., Steffolani, M. E., Martínez, C. S., & León, A. E. (2014). Effects of 579 spirulina biomass on the technological and nutritional quality of bread wheat 580 pasta. LWT-Food Science and Technology, 58(1), 102-108. 581 Draaisma, R. B., Wijffels, R. H., Slegers, P. M., Brentner, L. B., Roy, A., & Barbosa, M. J. 582 (2013). Food commodities from microalgae. Current Opinion in Biotechnology, 583 24(2), 169-177. 584 Eliasson, A. C., Carlson, T. G., & Larsson, K. (1981). Some effects of starch lipids on the 585 thermal and rheological properties of wheat starch. Starch‐Stärke, 33(4), 130-586 134. 587 Fabra, M. J., Martínez-Sanz, M., Gomez-Mascaraque, L. G., Coll, J. M., Martínez, J. C., & 588 Lopez-Rubio, A. (2017). Development and characterization of hybrid corn 589 starch-microalgae films: effect of US pre-treatment on structural, barrier and 590 mechanical performance. Carbohydrate Polymers, under review. 591 Fradique, M., Batista, A. P., Nunes, M. C., Gouveia, L., Bandarra, N. M., & Raymundo, A. 592 (2010). Incorporation of Chlorella vulgaris and Spirulina maxima biomass in 593 pasta products. Part 1: Preparation and evaluation. Journal of the Science of 594 Food and Agriculture, 90(10), 1656-1664. 595 Fradique, M., Batista, A. P., Nunes, M. C., Gouveia, L., Bandarra, N. M., & Raymundo, A. 596 (2013). Isochrysis galbana and Diacronema vlkianum biomass incorporation in 597 pasta products as PUFA’s source. LWT-Food Science and Technology, 50(1), 598 312-319. 599 Gernat, C., Radosta, S., Anger, H., & Damaschun, G. (1993). Crystalline Parts of Three 600 Different Conformations Detected in Native and Enzymatically Degraded 601 Starches. Starch - Stärke, 45(9), 309-314. 602 Hermansson, A.-M., & Svegmark, K. (1996). Developments in the understanding of starch 603 functionality. Trends in Food Science & Technology, 7(11), 345-353. 604 Holm, J., Björck, I., Ostrowska, S., Eliasson, A. C., Asp, N. G., Larsson, K., & Lundquist, 605 I. (1983). Digestibility of Amylose-Lipid Complexes in-vitro and in-vivo. Starch 606 - Stärke, 35(9), 294-297. 607 Holm, J., Lundquist, I., Björck, I., Eliasson, A.-C., & Asp, N.-G. (1988). Degree of starch 608 gelatinization, digestion rate of starch in vitro, and metabolic response in rats. 609 The American journal of clinical nutrition, 47(6), 1010-1016. 610 Hoover, R. (2001). Composition, molecular structure, and physicochemical properties of 611 tuber and root starches: a review. Carbohydrate Polymers, 45(3), 253-267. 612 Ilavsky, J., & Jemian, P. R. (2009). Irena: Tool suite for modeling and analysis of small-613 angle scattering. Journal of Applied Crystallography, 42(2), 347-353. 614 Imberty, A., Buléon, A., Tran, V., & Pérez, S. (1991). Recent advances in knowledge of 615 starch structure. Starch‐Stärke, 43(10), 375-384. 616 Imberty, A., Chanzy, H., Pérez, S., Bulèon, A., & Tran, V. (1988). The double-helical 617 nature of the crystalline part of A-starch. Journal of Molecular Biology, 201(2), 618 365-378. 619 Imberty, A., & Perez, S. (1988). A revisit to the three-dimensional structure of B-type 620 starch. Biopolymers, 27(8), 1205-1221. 621 James, C. M., Al-Hinty, S., & Salman, A. E. (1989). Growth and ω3 fatty acid and amino 622 acid composition of microalgae under different temperature regimes. 623 Aquaculture, 77(4), 337-351. 624 Jenkins, P. J., Cameron, R. E., & Donald, A. M. (1993). A universal feature in the structure 625 of starch granules from different botanical sources. Starch‐Stärke, 45(12), 417-626 420. 627 33 Jenkins, P. J., & Donald, A. M. (1995). The influence of amylose on starch granule 628 structure. International Journal of Biological Macromolecules, 17(6), 315-321. 629 Jiang, L., Cheng, J., Pei, H., Zhang, S., Yang, Z., Nie, C., & Zhang, L. (2017). Coupled 630 microalgal cultivation with treatment of domestic sewage and high-level CO2. 631 Environmental Technology(just-accepted), 1-20. 632 Kadam, S., & Prabhasankar, P. (2010). Marine foods as functional ingredients in bakery 633 and pasta products. Food Research International, 43(8), 1975-1980. 634 Kieffer, J., & Wright, J. (2013). PyFAI: a Python library for high performance azimuthal 635 integration on GPU. Powder Diffraction, 28(S2), S339-S350. 636 Kline, S. R. (2006). Reduction and analysis of SANS and USANS data using IGOR Pro. 637 Journal of Applied Crystallography, 39(6), 895-900. 638 Krog, N. (1971). Amylose complexing effect of food grade emulsifiers. Starch‐Stärke, 639 23(6), 206-210. 640 Krueger, B., Knutson, C., Inglett, G., & Walker, C. (1987). A differential scanning 641 calorimetry study on the effect of annealing on gelatinization behavior of corn 642 starch. Journal of Food Science, 52(3), 715-718. 643 LeCorre, D., Bras, J., & Dufresne, A. (2011). Influence of botanic origin and amylose 644 content on the morphology of starch nanocrystals. Journal of Nanoparticle 645 Research, 13(12), 7193-7208. 646 Li, M., Liu, P., Zou, W., Yu, L., Xie, F., Pu, H., Liu, H., & Chen, L. (2011). Extrusion 647 processing and characterization of edible starch films with different amylose 648 contents. Journal of Food Engineering, 106(1), 95-101. 649 Liu, H., Xie, F., Yu, L., Chen, L., & Li, L. (2009). Thermal processing of starch-based 650 polymers. Progress in Polymer Science, 34(12), 1348-1368. 651 Liu, H., Yu, L., Xie, F., & Chen, L. (2006). Gelatinization of cornstarch with different 652 amylose/amylopectin content. Carbohydrate Polymers, 65(3), 357-363. 653 Lopez-Rubio, A., Flanagan, B. M., Gilbert, E. P., & Gidley, M. J. (2008). A novel 654 approach for calculating starch crystallinity and its correlation with double helix 655 content: A combined XRD and NMR study. Biopolymers, 89(9), 761-768. 656 Lopez-Rubio, A., Flanagan, B. M., Shrestha, A. K., Gidley, M. J., & Gilbert, E. P. (2008). 657 Molecular Rearrangement Of Starch During In Vitro Digestion: Toward A 658 Better Understanding Of Enzyme Resistant Starch Formation In Processed 659 Starches. Biomacromolecules, 9(7), 1951-1958. 660 Lopez-Rubio, A., Htoon, A., & Gilbert, E. P. (2007). Influence of extrusion and digestion 661 on the nanostructure of high-amylose maize starch. Biomacromolecules, 8(5), 662 1564-1572. 663 Ma, C., Wen, H., Xing, D., Pei, X., Zhu, J., Ren, N., & Liu, B. (2017). Molasses 664 wastewater treatment and lipid production at low temperature conditions by a 665 microalgal mutant Scenedesmus sp. Z-4. Biotechnology for Biofuels, 10(1), 111. 666 Mata, T. M., Martins, A. A., & Caetano, N. S. (2010). Microalgae for biodiesel production 667 and other applications: A review. Renewable and Sustainable Energy Reviews, 668 14(1), 217-232. 669 Matos, Â. P., Feller, R., Moecke, E. H. S., de Oliveira, J. V., Junior, A. F., Derner, R. B., 670 & Sant’Anna, E. S. (2016). Chemical Characterization of Six Microalgae with 671 Potential Utility for Food Application. Journal of the American Oil Chemists' 672 Society, 93(7), 963-972. 673 Matveev, Y. I., van Soest, J. J. G., Nieman, C., Wasserman, L. A., Protserov, V. A., 674 Ezernitskaja, M., & Yuryev, V. P. (2001). The relationship between 675 34 thermodynamic and structural properties of low and high amylose maize 676 starches. Carbohydrate Polymers, 44(2), 151-160. 677 Pérez, S., & Bertoft, E. (2010). The molecular structures of starch components and their 678 contribution to the architecture of starch granules: A comprehensive review. 679 Starch‐Stärke, 62(8), 389-420. 680 Qiao, D., Zhang, B., Huang, J., Xie, F., Wang, D. K., Jiang, F., Zhao, S., & Zhu, J. (2017). 681 Hydration-induced crystalline transformation of starch polymer under ambient 682 conditions. International Journal of Biological Macromolecules. 683 Salman, H., Blazek, J., Lopez-Rubio, A., Gilbert, E. P., Hanley, T., & Copeland, L. (2009). 684 Structure–function relationships in A and B granules from wheat starches of 685 similar amylose content. Carbohydrate Polymers, 75(3), 420-427. 686 Sarko, A., & Wu, H. C. (1978). The Crystal Structures of A‐, B‐and C‐Polymorphs of 687 Amylose and Starch. Starch‐Stärke, 30(3), 73-78. 688 Scholz, M. J., Weiss, T. L., Jinkerson, R. E., Jing, J., Roth, R., Goodenough, U., Posewitz, 689 M. C., & Gerken, H. G. (2014). Ultrastructure and composition of the 690 Nannochloropsis gaditana cell wall. Eukaryotic cell, 13(11), 1450-1464. 691 Sevenou, O., Hill, S., Farhat, I., & Mitchell, J. (2002). Organisation of the external region 692 of the starch granule as determined by infrared spectroscopy. International 693 Journal of Biological Macromolecules, 31(1), 79-85. 694 Spolaore, P., Joannis-Cassan, C., Duran, E., & Isambert, A. (2006). Commercial 695 applications of microalgae. Journal of Bioscience and Bioengineering, 101(2), 696 87-96. 697 Tang, M. C., & Copeland, L. (2007). Analysis of complexes between lipids and wheat 698 starch. Carbohydrate Polymers, 67(1), 80-85. 699 Tester, R. F., & Morrison, W. R. (1990). Swelling and gelatinization of cereal starches. I. 700 Effects of amylopectin, amylose, and lipids. Cereal Chem, 67(6), 551-557. 701 Van Eykelenburg, C. (1977). On the morphology and ultrastructure of the cell wall of 702 Spirulina platensis. Antonie van Leeuwenhoek, 43(2), 89-99. 703 Varavinit, S., Shobsngob, S., Varanyanond, W., Chinachoti, P., & Naivikul, O. (2003). 704 Effect of amylose content on gelatinization, retrogradation and pasting 705 properties of flours from different cultivars of Thai rice. Starch‐Stärke, 55(9), 706 410-415. 707 Waigh, T. A., Gidley, M. J., Komanshek, B. U., & Donald, A. M. (2000). The phase 708 transformations in starch during gelatinisation: A liquid crystalline approach. 709 Carbohydrate Research, 328(2), 165-176. 710 Wang, S., Zhang, X., Wang, S., & Copeland, L. (2016). Changes of multi-scale structure 711 during mimicked DSC heating reveal the nature of starch gelatinization. 712 Scientific Reports, 6. 713 Yoshimura, M., Takaya, T., & Nishinari, K. (1996). Effects of Konjac-Glucomannan on 714 the Gelatinization and Retrogradation of Corn Starch As Determined by 715 Rheology and Differential Scanning Calorimetry. Journal of Agricultural and 716 Food Chemistry, 44(10), 2970-2976. 717 Yuryev, V. P., Kalistratova, E. N., van Soest, J. G., & Niemann, C. (1998). 718 Thermodynamic properties of barley starches with different amylose content. 719 Starch‐Stärke, 50(11‐12), 463-466. 720 Zobel, H. F. (1988). Starch Crystal Transformations and Their Industrial Importance. 721 Starch - Stärke, 40(1), 1-7. 722 723 35 Supplementary Material 724 725 726 Figure S1. X-ray diffraction patterns of the raw microalgae (freeze-dried). 727 728 729 36 730 731 Figure S2. TEM images of raw microalgae: (A) Spirulina, (B) Nannochloropsis and (C) 732 Scenedesmus. 733 734 37 735 736 Figure S3. Evolution of the lamellar distance extracted from the SAXS fitting parameters 737 for the corn starch (A) and high amylose starch (B) samples. 738 739