In this study, five distinct hemp seed protein concentrates , produced at industrial scale using dry or wet protein enrichment technologies , were comparatively investigated in terms of their functionality, structuring potential during wet extrusion, and the molecular properties of their most abundant component, protein. In addition, their macro-nutrient composition , mineral profile, and phenolic fractions were comparatively investigated to elaborate on the underlying mechanisms for protein molecular properties and structuring behaviour of hemp protein concentrates during processing. Finally, selected hemp protein concentrates were subjected to wet extrusion and the resultant High Moisture Meat Analogues were investigted for anisotropy, viscoelasticity, and water mobility as a proxy for mouthfeel. Field pea protein, one of the most common plant-based protein alternatives, was also included in this study for comparative purposes. Pea protein is also made up of albumins and globulins, it is abundant and cheap, and pea plants are grow in moderate climates . Dietary pea and hemp proteins are less connected to GMO questions and are not listed as allergenic. Nevertheless, compared to other seeds, the content of phytic acid in the hemp seed cotyledon is reported to be higher , with values as high as 22.5 mg/g in hemp meal . The presence of phytic acid could further reduce the bioavailability of multivalent cations and its separation from the plant tissue is challenging due to its low solubility in water. Although phytic acid can be vastly reduced during pilot scale AE-IP , this would not be the case for dry fractionated hemp protein concentrates, which could limit their nutritional interest. Therefore, in this work, phytic acid content was measured in all flour samples and HMMA prototypes. Protein powders were analysed for moisture, ash, protein, and fat,drying rack cannabis and then the total carbohydrate content was obtained by weight difference with the other components. Moisture was analysed using an automated moisture analyser following AACC method 44–15.02 .
Ash content was analysed according to AOAC 923.03 method using a muffle furnace at 550 ◦C for 4.5 h. Protein content was determined using AACC method 46–30.01 with an automated Dumatherm N Pro protein analysis system . The released nitrogen was converted into protein by multiplying the value of nitrogen by a factor of 6.25. Lipid content was determined by Low-Field Proton Nuclear Magnetic Resonance following the AOAC method 2008.06 . All analyses were run in triplicate. Since hemp seeds are very low in starch , total dietary fibre was analysed using the AOAC official method 991.43 . The extraction and analysis of extractable polyphenols and non-extractable polyphenols , the latter consisting of non-extractable hydrolysable polyphenols and non-extractable proanthocyanidins , were performed as described by Pico et al. , with modifications.Specifically,the quantification of EPP and HPP in the solutions was performed using the highly phenolic selective Fast Blue BB reaction, developed by,to account for the lack of specificity of the Folin–Ciocalteu assay due to numerous interferences . NEPA content in the solution obtained by depolymerisation by butanolysis was measured at 555 and 450 nm to detect anthocyanins and xanthylium compounds, respectively . EPP and HPP fractions were expressed as mg of gallic acid equivalents /100 g dry matter, while the anthocyanins-NEPA fraction was expressed as mg of delphinidin equivalents /100 g dry matter. All analyses were performed at least in triplicate. The main processing steps used to obtain the hemp seed protein concentrates are detailed in Table 1. All hemp protein materials were first subjected to cold pressing-expelling for lipid extraction, although they differed in their protein enrichment steps . Hemp 1 and Hemp 3 were subjected to dry fractionation steps, with Hemp 3 being the only sample coming from dehulled seeds. Meanwhile, Hemp 1, Hemp 4 and Hemp 5 were fractionated through AE-IP. The proximate composition of the resulting hemp protein concentrates as well as of pea protein is shown in Table 2, which was very similar to the data provided by the manufacturers.
The oil , carbohydrate and protein content of the hemp samples were distinctly different from each other. Compared to the pea sample used in this study, hemp protein concentrates were significantly richer in oil, carbohydrates and ash, and they exhibited a lower protein content. In terms of macro-nutrient abundance, Hemp 1, Hemp 2, and Hemp 5 were characterized by a lower lipid content than Hemp 3 and Hemp 4 . The lower lipid content of Hemp 1 compared to Hemp 3 , and of Hemp 2 versus Hemp 4 , confirms a very efficient oil extraction performed by the manufacturer of Hemp 1 and Hemp 2. Hemp 1 contained the highest amount of carbohydrates , which can be attributed to the choice of using whole seeds and dry fractionation. Conversely, Hemp 5 presented the lowest carbohydrate and the highest protein content, which is explained by the double AE-IP. The ash content was rather similar for the Hemp 1–4 samples, ranging from 7.3 to 8.8% d.b. Hemp 5 exhibited a significantly lower ash content than the rest , which could suggest that minerals were removed during the isoelectric precipitation step , which was performed twice for Hemp 5. The mineral profile revealed that all hemp samples were generally rich in the macro-elements potassium , phosphorous and magnesium , which are needed in the amount of >50 mg/day in the human diet. Generally, hemp samples were also richer in the macro-elements calcium and the micro-elements manganese, copper, and zinc than the pea counterpart. Hemp samples had a significantly lower sodium content than pea . Interestingly, lower sodium content for dry fractionated hemps, Hemp 1 and 3, was observed compared to wet-fractionated hemps and pea, suggesting that the sodium in is coming from the alkaline extraction step, normally done with sodium hydroxide. The high amount of potassium along with a relatively low sodium content leads to a high K/Na ratio, which is believed to be related to cardio protective effects as it promotes a high K intake that is considered to be inversely related to blood platelet aggregation and stroke incidence . Nonetheless, compared to other seeds, the content of phytic acid in hemp seeds is reported to be higher , which could further reduce the bioavailability of multivalent cations, including Zn2+, Fe2+/3+, Ca2+, Mg2+, Mn2+, and Cu2+. Hemp 3 was significantly richer in magnesium and phosphorus than the other hemp samples, which could be due to differences in genotype, environmental and soil conditions as well as higher ash content as a consequence of the protein enrichment processing .
Since phytic acid naturally stores phosphorous in plants , the high phosphorous content in dry-fractionated samples might be indicative of high levels of phytates in these particular samples. In fact, the phytic acid content of Hemp 1 and Hemp 3 was significantly higher than that of pea and wet-fractionated hemp samples , as shown in Table 2. These values are similar to those reported previously for hemp protein concentrates and pea . The higher values of Hemp 3 compared to Hemp 1, both being dry-fractionated samples, could be explained by the fact that only Hemp 3 seeds were hulled and that the majority of phytate is found in the cotyledon fraction . Therefore, pearling, or mechanical fractionation to remove coarse fractions, would get rid of those fractions low in phytic acid and, therefore, increase the weight percentage of phytic acid in protein concentrates. In contrast, a significant reduction of phytic acid during the precipitation step was already reported for wet-fractionated hemp, which was partially explained by a potential activation of plant phytases at the isoelectric pH of proteins or by the higher solubility of inositol phosphates during acidic precipitation of proteins. To the best of our knowledge, there are no regulations in the European Union regarding acceptable levels of phytic acid in hemp proteins. The total content of phenolics and the phenolic fractions of the pea and hemp protein flours were also studied, and the results are summarized in Table 2. Hemp protein concentrates possessed higher content of TP than pea protein concentrate . Our values are higher than those reported by Izzo et al. in hemp inflorescences using high resolution mass spectroscopy. However, it is noted that our study also focused on those phenolics not extracted during the aqueous-organic treatments commonly performed to analyse polyphenol content in foods. In fact, HPP was the main phenolic fraction found in all samples. Phenolic acids and monomeric flavanols are usually reported to remain mostly un-extracted , which could represent an important fraction of HPP since hemp inflorescences are rich in both . Several studies showed that hemp seeds are rich in various phenolic acids, lignanamides, phenolic amides and flavonoids . The comparison between samples revealed that Hemp 1 was significantly richer in EPP and HPP than the rest of samples, which suggest that many of these compounds remain attached to the abundant carbohydrate fraction in Hemp 1, which was not removed during processing.
Conversely, Hemp 3 exhibited the lowest EPP to HPP ratio, followed by Hemp 5, indicating that even more phenolics were present in bound form. Interestingly, these two hemp samples presented the highest protein content and relatively low presence of carbohydrates, which may have contributed to the loss of EPP during processing. However, further studies are needed to understand why free phenolics may have been lost during the pearling used for Hemp 3 dehulling or dual AE-IP processing. It should also be noted that differences found between phenolic fractions in the samples could also be attributed to other factors different to protein enrichment processing, commercial greenhouse supplies including but not limited to phenotype, seed maturity, growth and post harvest environmental conditions.The proanthocyanidins in hemp seed cake/meal have been identified to be essentially catechin polymers, i.e., procyanidins, with total values of 245–262 mg proanthocyanidins/100 g in non-treated hemp seed cake , which generally aligns with the reported results in this work. Nonetheless, some proanthocyanidins could be extracted as EPP fraction, as reported by P´erez-Jim´enez and Saura-Calixto with fruit peels. Hemp 1 and Hemp 2, followed by Hemp 5, displayed significantly lower NEPA than the rest of the hemp samples, suggesting that NEPA could have been co-extracted during the extraction of oil in Hemp 1 and Hemp 2 and during the more aggressive aqueous protein extraction of Hemp 5 . Hemp 4 followed by Hemp 3 possessed the highest NEPA content. Similar to flavan-3-ols, anthocyanins are highly susceptible to degradation during hydrothermal treatment . Hence, any protein enrichment technology that did not elevate the temperature of the samples could have partially protected some NEPAs from thermal degradation. The hemp protein fraction is comprised mainly by globular proteins in the form of globulins and albumins, whose relative abundance might impart different functionality. Reversed-Phase Chromatography has confirmed the quantitative dominance of the edestin fraction, accounting for up to 70% total hemp protein content . The edestin molecule is composed of six identical subunits, and each subunit consists of acidic subunits and basic subunits linked by one disulphide bond . Fig. 1 depicts the SDS-PAGE and Native-PAGE profiles of pea and hemp proteins. In reducing SDS-PAGE , edestin appeared to dissociate into its acidic and basic subunits, corresponding to bands at around 35 and 20 kDa, respectively. These results are consistent with previous studies by Shen et al. and Wang and Xiong which found roughly the same molecular weights for these fractions. Besides the bands of edestin acidic and basic subunits, three visible but less abundant bands appeared between 50 and 70 kDa. Tang, Ten, Wang, and Yang reported that the band at about 48.0 kDa was similar to the β-subunit of the trimeric β-conglycinin found in soybean, suggesting the presence of a 7S-vicilin-like protein. Nevertheless, it was a minor fraction compared to edestin and the albumins seen at <18 kDa. These findings are corroborated by the results from Mamone et al. using proteomic analyses, indicating that hemp protein isolate consisted of essentially three major storage proteins, 11S edestin, 7S vicilin-like protein, and albumin. In the absence of reducing agents, the disulfide bonds between AS and BS of edestin are not disrupted and an intense higher molecular weight band of ~55–60 kDa was instead visible .