Soybean [Glycine max (L.) Merr.] is the primary cost-effective source of protein used worldwide. Soybean processing generates a range of by-products, including flour, oil, and lecithin, among others (Alarape et al., 2024; Oviedo-Rondón et al., 2024). Soybean meal, a by-product obtained from oil extraction, is a high-quality protein source containing approximately 48% crude protein. It is characterized by high digestibility and a favorable amino acid profile, providing essential amino acids such as lysine, tryptophan, threonine, isoleucine, and valine, which are typically limited in cereal-based diets (Lagos and Stein, 2017; Sotak-Peper et al., 2017). In addition, soybean meal provides metabolizable energy due to its residual lipid content and supports digestive health through the presence of insoluble fiber. These properties make soybean meal a key ingredient in feed formulations for poultry, swine, and ruminants. However, the nutritional quality of soybean grains can be influenced by theirs protein composition and the occurrence of antinutritional factor that may impair nutrient digestibility and absorption (Huang et al., 2023; Toomer et al., 2023; Xiao et al., 2025).

Antinutritional factor (ANFs) are detrimental in animal feeding, particularly for monogastric species such as poultry and swine (Nahashon et al., 2013). In soybean, the main ANFs of nutritional relevance include lectins, saponins, enzyme inhibitors, phytates and raffinose family oligosaccharides, all of which substantially reduce the nutritional value of soybean meal (Di et al., 2024). These compounds interfere with the activity of digestive enzymes, cause damage to the intestinal epithelium, decrease nutrient digestibility, trigger inflammatory responses, and consequently impair feed conversion efficiency in animals (Liener, 1994; Samtiya et al., 2020).

In the industry, soybean meal and its derivatives intended for animal consumption undergo heat treatment to eliminate antinutritional factors (Thakur et al., 2022). However, excessive processing can affect protein solubility, thereby reducing amino acid bioavailability and ultimately compromising the quality of soybean-derived proteins (Parsons et al., 1992; van den Berg et al., 2022). Thus, heat treatment increases production costs and can negatively impact the nutritional quality of soybean meal.

Protein content and amino acid composition are critical factors in formulating animal feed derived from soybean grains. Protein concentration can vary depending on genotype, soil profile, and nutrient availability (Guo et al., 2022). Although soybean protein is rich in some essential amino acids, such as lysine, it is not considered a complete protein (Patil et al., 2017; Singer et al., 2019, 2023). Soybean meal is particularly deficient in sulfur-containing amino acids, such as methionine and cysteine, thus requiring supplementation when used as the sole protein source (Panthee et al., 2006). Therefore, the development of soybean cultivars with low levels of antinutritional factors and high protein content, combined with a balanced composition of essential amino acids, is essential to meet industrial demands and to minimize adverse effects on animal health.

New Breeding Technologies (NBTs) have been employed to enhance soybean traits for animal nutrition. Over time, approaches such as mutagenesis, marker-assisted selection, genomic selection, transgenesis, genetic modifications using CRISPR-Cas and RNA interference have significantly contributed to the development of cultivars with optimized protein profiles and reduced concentrations of antinutritional compounds. Within this framework, the present review aims to summarize the major advances achieved in improving soybean protein quality and reducing antinutritional factors through the application of NBT-based strategies.

The popularity of soybeans is due to their high nutritional value, making them a fundamental food source for both human and animal consumption (Sudarić, 2020; Deng and Kim, 2024). Their composition is well balanced, containing significant amounts of proteins, fats, and carbohydrates. With approximately 35–40% protein in their dry weight, soybeans are an excellent source of protein, providing all essential amino acids (Sudarić, 2020; Modgil et al., 2021). This makes them comparable to animal-derived proteins and highly suitable for vegetarian and vegan diets. In addition, soy protein is widely used in the processed food industry, such as in tofu, soy milk, and protein isolates (Sudarić, 2020; Modgil et al., 2021; Deng and Kim, 2024).

Regarding fat content, soybeans contain about 18–20% lipids, which are predominantly unsaturated fats, including omega-3 and omega-6 fatty acids that play an important role in cardiovascular health (Hymowitz, 1970). Soybeans also contain lecithin, a phospholipid extensively used in the food and cosmetics industries (List, 2015). Carbohydrates account for roughly 30% of soybean composition, with a notable proportion of dietary fiber, which promotes digestive and intestinal health (Medic et al., 2014). Polysaccharides such as cellulose, and oligosaccharides such as raffinose and stachyose, are also present in significant amounts, although they can cause gastrointestinal discomfort in some cases (Dierking and Bilyeu, 2008; Zeng et al., 2021; Elango et al., 2022).

In addition to these macronutrients, soybeans are rich in bioactive compounds such as isoflavones (genistein, daidzein, and glycitein), which act as antioxidants and may contribute to bone and cardiovascular health, as well as provide benefits in hormonal regulation (Yamagata, 2019; Kanadys et al., 2021; Fatima et al., 2025). Another class of compounds found in soybeans are phytosterols, which help reduce LDL cholesterol (low-density lipoproteins) in the blood, benefiting heart health (Pelletier et al., 1995; Matvienko et al., 2002; Ilha et al., 2020).

Soybeans are also an excellent source of vitamins and minerals, being rich in B-complex vitamins, vitamin E, and minerals such as iron, magnesium, and phosphorus. These nutrients make soybeans a highly nutritious and functional food, also providing calcium, potassium, and phosphorus, which are essential for bone strengthening, muscle function, and fluid balance in the body (Kumar and Pandey, 2020).Additional compounds such as phytic acid, lectins, and saponins are considered antinutritional factors, as they can reduce the bioavailability of minerals, interfere with protein digestibility, and, in some cases, affect intestinal absorption (Viteri, 1997; Lott et al., 2000; Shi et al., 2004). However, these compounds have also been reported to possess certain biological activities, such as antioxidant and immunomodulatory effects, which are currently being investigated for potential health benefits when present in controlled amounts lipoxygenases (Liener, 1994; Samtiya et al., 2020; Salim et al., 2023; Di et al., 2024).

Therefore, soybeans are an extremely nutritious food with a diverse composition that offers numerous health benefits. Their content of protein, healthy fats, fiber, vitamins, and bioactive compounds makes them an excellent option for both human diets and animal nutrition (Samtiya et al., 2020; Toomer et al., 2023; Deng and Kim, 2024; Fatima et al., 2025).

Despite the benefits of soybean consumption, it contains several components that can cause harm when consumed. These are called antinutritional factors (ANFs), which are defined as “compounds that occur naturally or synthetically that interfere with the absorption, digestion, or utilization of nutrients by a body” (Liener, 1994; Woyengo et al., 2017). The main antinutritional factors present in soybean are trypsin inhibitors and lectins (Vasconcelos et al., 2006; John et al., 2021). Antinutritional factors can be classified as thermostable and heat-labile. In soybeans, the thermostable ANFs are saponins, tannins, oligosaccharides, allergens, and phytate, while the heat-labile ANFs are trypsin inhibitors, lectins, ureases, and lipoxygenases (Liener, 1994; Samtiya et al., 2020; Salim et al., 2023; Di et al., 2024). The antinutritional factors discussed in this review are summarized in Figure 1.

Lectins (hemagglutinin or agglutinin) are one of the main antinutritional factors in soybeans, containing 10 to 20 g.kg^-1 of lectin, considering that a lectin concentration greater than 7 g.kg^-1 is considered harmful to digestibility and unsuitable for direct consumption (George et al., 2008). In soybean, the seed lectin is controlled by the single Le1 gene (Glyma.02G012600). In le1 genotypes, a 3.5-kb insertion element disrupts the gene’s reading frame, preventing lectin from accumulating in the seeds (Goldberg et al., 1983; Vodkin et al., 1983; Vodkin and Raikhel, 1986; Okamuro and Goldberg, 1992; Shamimuzzaman and Vodkin, 2018). Lectin is a tetrameric glycoprotein of four identical subunits, each 30 kDa, totaling 120 kDa, which can reversibly bind to carbohydrates (Olsen et al., 1997; Sharon and Lis, 2001; Halder et al., 2016).Binding to carbohydrates occurs through Ala-86, Asp-88, Ala-105, Phe-128, Asn-130, Leu-214, Asp-215, and Ile-216, which can interact with intestinal cells, causing damage and changing their conformation. Furthermore, when consumed, the protein is not digested by proteases in the gastrointestinal tract and undergoes endocytosis in the intestine (Ma and Wang, 2010; Pan et al., 2013). When lectin enters the cell, it induces hyperplasia and hypertrophy of the small intestine. Several articles have demonstrated decreased performance in monogastric animals when fed raw soybean seeds. In pigs, soybean agglutinin (SBA) at a dose of 0.2% causes intestinal damage in piglets (Pan et al., 2013, 2018). In poultry and fish, the ingestion of raw soybean resulted in lower weight gains due to indigestion, a result of the morphological changes caused by the action of lectin in the intestine (Babot et al., 2021).

Among the antinutritional factors, protease inhibitors stand out, which act by blocking digestive enzymes such as trypsin and chymotrypsin. These compounds reduce protein digestibility and cause a compensatory increase in pancreatic enzyme secretion, which can lead to hypertrophy and hyperplasia of the pancreas in monogastric animals (Liener, 1994). The main protease inhibitors present in soybean are the Kunitz-type trypsin inhibitor (KTI) and the Bowman-Birk-type trypsin inhibitor (BBI), which differ in structure, abundance, and thermal stability. The Kunitz-type trypsin inhibitors (KTI) are one of the main antinutritional factors, present in the proportion of 4–15 mg of trypsin inhibitor per gram of soybean flour (Park et al., 2023). KTI is a monomeric protein that weighs 21kDa, has two disulfide bonds and 181 amino acid residues. The trypsin inhibitor binds to the Asp189 region of trypsin through a positively charged amino acid, which can be a lysine or arginine, and prevents its activity (Sweet et al., 1974; De Meester et al., 1998; Song and Suh, 1998). Soybean harbors a multigene family encoding Kunitz-type trypsin inhibitors. The predominant seed-expressed inhibitor is encoded by KTI3 (Glyma.08G341500), located on chromosome 8, whereas related paralogs such as KTI1 (Glyma.01G095000) and KTI2 (e.g., Glyma.08G341400 in the reference genome) are located in distinct genomic regions, indicating that KTI genes contributing to trypsin inhibitory activity are not organized as a single tandem cluster (Jofuku and Goldberg, 1989; Zhang et al., 2021; Tian et al., 2025). The effects of Kunitz-type trypsin inhibitors on the digestibility of amino acids, especially lysine, methionine, and threonine, and on weight gain in livestock are well-documented, being consistently associated with reduced zootechnical performance (Herkelman et al., 1993; Palacios et al., 2014; Kuenz et al., 2022). This reduction results from the inhibition of pancreatic proteases, which stimulates the excessive secretion of digestive enzymes and increases endogenous amino acid losses, compromising the efficiency of dietary protein utilization (Herkelman et al., 1993; Perez-Maldonado et al., 2003; Kuenz et al., 2022).

The Bowman-Birk-type trypsin inhibitor is another type of protease inhibitor; however, it is found in smaller quantities in soybean, approximately 2 to 4 mg.g^-1 of dry soybean (Sessa and Wolf, 2001; Kim et al., 2024; Liu et al., 2025a). The Bowman-Birk trypsin inhibitor is a monomeric protein with a weight of 7–8 kDa that has two inhibitory loop sites that inhibit chymotrypsin (Phe, Leu, and Tyr) and trypsin (Lys or Arg) at independent sites; these inhibitory sites are stabilized by 7 to 8 cysteines (Gitlin-Domagalska et al., 2020). It has 7 disulfide bonds and 71 amino acid residues. There are at least 11 Bowman-Birk trypsin inhibitors in soybean; however, only three of them (Glyma.16G208900, Glyma.14G117700, and Glyma.09G158700) are more highly expressed in seeds (Kim et al., 2024). Recently, it was shown that this inhibitor, despite being in smaller quantity in the seed, is responsible for about 80% of chymotrypsin inhibition and 45% of total trypsin inhibition (Liu, 2024; Liu et al., 2025a). Furthermore, this protease inhibitor is not completely eliminated by heat treatment because its 7 disulfide bonds confer stability, maintaining the protein’s conformation upon heating (Sessa and Wolf, 2001; Qi et al., 2005; Sabotič et al., 2011). Similar to KTI, BBI inhibition also causes a reduction in animal performance and, as observed in the literature, its magnitude is even greater than that observed for the KTI inhibitor (Moreau et al., 2024; Kim et al., 2025).

Phytic acid (myo-inositol hexafosfato, IP₆) is the main storage form of phosphorus in soybean and other legume seeds, being essential for the plant’s initial development. However, this compound has a significant antinutritional effect because it forms insoluble complexes with minerals such as zinc (Zn), iron (Fe), and calcium (Ca), reducing their bioavailability for humans and animals (Graf and Eaton, 1984; Zhou et al., 1992; Punjabi et al., 2018). Phytic acid biosynthesis occurs in a pathway composed of four enzymatic reactions. First, glucose-6-phosphate is converted to myo-inositol-3-phosphate by the enzyme Myo-inositol-3-phosphate synthase (MIPS1 – Glyma.10G123400) (Redekar et al., 2017; DeMers et al., 2021). Next, Myo-inositol kinase (MIK – Glyma.05G098300) phosphorylates myo-inositol, forming inositol monophosphate (Hegeman et al., 2001; Jin et al., 2021). The subsequent phosphorylations, performed by inositol polyphosphate kinases (IPK1 – Glyma.12G149100; IPK2 – Glyma.12G149200), result in the production of IP₆ (Jin et al., 2021). Finally, Multidrug resistance-associated protein (MRP – Glyma.13G172300) transporters store the phytate in the seed vacuoles, ensuring phosphorus reserve and compound stability during development (Sashidhar et al., 2020). In addition to reducing mineral availability, phytate can also interact with proteins and digestive enzymes, such as trypsin and α-amylase, compromising the digestibility of proteins and starch (Bye et al., 2013; Theodoropoulos et al., 2018).

The raffinose family oligosaccharides (RFOs), such as raffinose and stachyose, are synthesized in soybean seeds by enzymes like galactinol synthase (GOLS), raffinose synthase (RS), and stachyose synthase (STS), whose genes are highly expressed during seed maturation (Dierking and Bilyeu, 2008; Valentine et al., 2017; de Koning et al., 2021; Jha et al., 2022). During germination and imbibition, the raffinose family oligosaccharides (RFOs) are degraded by α-galactosidases (α-GAL), which hydrolyze their α-1,6-galactosidic bonds, converting them into simple sugars like sucrose and galactose, providing energy to the developing embryo (de Koning et al., 2021). The total content of RFOs in soybean can exceed 40 mg.g^-1with stachyose being the most predominant sugar (Elango et al., 2022). When consumed by monogastric animals, these oligosaccharides are not digested due to the absence of endogenous α-galactosidases, passing intact through the intestinal tract to the large intestine, where they are fermented by bacteria, resulting in gas production, intestinal discomfort, and reduced weight gain (Dierking and Bilyeu, 2008; Zeng et al., 2021; Elango et al., 2022).

Saponins are triterpenoid or steroid glycosides widely distributed in plants. In soybean, they mainly derive from soyasapogenols A, B, and E, formed from the β-amyrin precursor through the sequential action of oxygenases (CYPs) and glycosyltransferases (UGTs) (Shibuya et al., 2006; Sayama et al., 2012; Yano et al., 2018; Sundaramoorthy et al., 2019). These compounds conjugate with sugars to generate different soyasaponins, which accumulate mainly in the seed coat and hypocotyl of the seeds. In plants, they exhibit antifungal and antibacterial activity, in addition to reducing herbivory due to low palatability (Timilsena et al., 2023). However, they are considered antinutritional factors because they form foam, impart a bitter taste, and can interact with membranes and minerals, reducing feed intake and weight gain in monogastric animals (Timilsena et al., 2023).

Antinutritional factors can cause alterations in metabolism when consumed without the proper treatment (Liener, 1994); thermal treatment is conventionally used for the elimination of ANFs in animal feed. However, this treatment increases the energy cost of grain processing and can, as a result of using high temperatures, decrease the nutritional value and solubility of the grains (Bonturi et al., 2022). Several modifications for limiting antinutritional factors are summarized in Table 1.

Soybean has always been traded quantitatively and not based on its quality. Therefore, historically, genetic improvement has always prioritized increasing plant productivity at the expense of grain quality (Lee et al., 2019), as there is a negative correlation between protein content and productivity (Lee et al., 2019; Zhang et al., 2021). Furthermore, the protein content is also negatively related to the oil content of the seed, which further hinders the manipulation of this attribute by genetic improvement (Hymowitz, 1970). This becomes more serious with the decrease in protein content over the years due to genetic improvement efforts for increased productivity. More recent soybean cultivars show a protein content of around 36%, while the ideal level considered for the industry is 42% (Medic et al., 2014; Patil et al., 2017; Assefa et al., 2018).

Soybean grain quality for animal feed is determined by the coordinated regulation of a limited number of core metabolic pathways, including carbon allocation, nitrogen assimilation, phosphorus storage, and the biosynthesis of antinutritional factors. Carbon allocation during seed filling controls the partitioning of sucrose between oil and protein biosynthesis, thereby directly influencing the oil–protein balance (Li et al., 2024). Nitrogen assimilation and transport regulate amino acid availability for storage protein synthesis, defining both protein content and quality (Kumar et al., 2025; Shi et al., 2025). Phosphorus storage pathways, mainly associated with phytic acid accumulation, affect mineral bioavailability and protein digestibility in animal diets. In parallel, secondary metabolic pathways lead to the accumulation of antinutritional compounds such as protease inhibitors, raffinose family oligosaccharides, and saponins, which negatively impact nutrient utilization, particularly in monogastric animals (Yao et al., 2023).

Soybean proteins have been classified by ultracentrifugation, which identified four protein fractions: 2S, 7S, 11S, and 15S, according to the Svedberg index (Jegadeesan and Yu, 2020). The 2S fraction, which represents 8% to 22% of soluble soybean protein, contains various enzymes, including trypsin inhibitors, such as the Bowman-Birk and Kunitz inhibitors. Other proteins, such as lectins and lipoxygenases, represent 1% to 3% of the total soybean protein (Vasconcelos et al., 2006). The 15S fraction, although poorly characterized, corresponds to about 5% of the total soybean protein and is known to contain polymers of other soybean proteins. Most of the soybean storage proteins, β-conglycinin and glycinin, are present in the 7S and 11S fractions, respectively, representing about 35% and 54% of the soluble soybean protein (Capriotti et al., 2014; Din et al., 2021; John et al., 2021). Glycinin is composed of six subunits that form a hexameric structure of 360 kDa. The six subunits are composed of basic units (B1a, B1b, B2, B3, and B4) and acidic units (A1a, A1b, A2, A3, A4, and A5) that are linked by disulfide bonds. This protein is encoded by several genes that have been separated into groups: group 1 composed of genes Gy1 (A1aB2), Gy2 (A2B1a); group 2 composed of Gy4 (A5A4B3) and Gy5 (A3B4); and group 3 composed of two pseudogenes (gy6 and gy8); and Gy7 (group of unclassified polypeptides). Glycinin has a higher concentration of sulfur-containing amino acids, such as methionine and cysteine, in which soybean is relatively deficient (Patil et al., 2017). The 7S fraction is composed of β-conglycinin, which is a trimer of 180 kDa, with about 15 genes (CG1–CG15) encoding the protein. Several strategies to increase soybean protein and protein quality are presented in Table 2.

Advances in molecular genetics and metabolic engineering have demonstrated that targeted manipulation of these pathways can shift metabolic fluxes toward higher protein content and lower antinutritional factor accumulation, without necessarily compromising seed viability (Yao et al., 2023). Consequently, genome editing and related biotechnological tools have emerged as effective approaches to directly modulate key enzymatic steps, transport processes, and regulatory nodes controlling soybean grain composition.

The advancement of biotechnology tools and computational biology has considerably expanded the possibilities for soybean improvement aimed at increasing protein content and enhancing the nutritional quality of the seeds. Emerging technologies, such as CRISPR/Cas guided by protein structural modeling, rational promoter engineering, and Prime Editing, represent a new frontier in the precise manipulation of genes related to seed storage metabolism. The application of the biotech tools to enhance digestibility and protein levels in soybean meal are presented in Figure 2.

A recent example is the work by Wang et al. (2025a), which combined the CRISPR/Cas9 system with structural predictions obtained through AlphaFold2 to edit the genes GmSWEET10a and GmSWEET10b, responsible for sucrose transport between the seed coat and the embryo. Alterations in the C-terminal structure of these proteins modified the shape of the transport tunnel, resulting in lines with different sugar translocation capacities. Variants with a more open tunnel showed higher oil accumulation and lower protein content, while mutations that restricted sucrose passage increased protein content. This integration of structural prediction and gene editing paves the way for protein engineering guided by molecular modeling, capable of generating genetic variability under rational control.

Another promising strategy is promoter engineering, which allows spatial and temporal modulation of genes related to nitrogen metabolism, storage protein synthesis, and carbon transport. The use of strong, seed-development-specific promoters can direct metabolic resources toward protein synthesis without compromising yield. This approach has already been tested in other legumes and could be applied to optimize the expression of key genes in soybean, minimizing undesirable pleiotropic effects.

Finally, Prime Editing, an evolution of the CRISPR system, emerges as a high-potential tool for precision breeding. This technique combines the action of a reverse transcriptase guided by RNA with a modified Cas endonuclease, allowing precise point mutations, insertions, and deletions without generating double-strand DNA breaks. Applying this technology in soybean could enable the controlled introduction of favorable variants in genes related to protein accumulation, structural stability of storage proteins, and the efficiency of sugar and amino acid transport, overcoming limitations observed in previous approaches based on random mutagenesis or RNAi.

The future of soybean genetic improvement, therefore, moves toward an integration of high-precision genome editing, AI-assisted structural modeling, and refined gene regulation, enabling the development of cultivars with superior nutritional profiles, adapted to the demands of industry and productive sustainability.