The researchers at Shihezi University in China first screened seven probiotic strains for their effects on a feed substrate containing cottonseed meal, corn flour and wheat bran. Cottonseed meal served as the main protein source, corn flour supplied fermentable carbohydrate and wheat bran provided a porous fibrous structure. The mixture was fermented for 48 hours at 40% moisture with individual strains or without inoculation, after which soluble protein, crude protein, microbial crude protein, ammonia nitrogen and pH were assessed.

All seven strains increased soluble-protein content relative to the uninoculated control, but C. tropicalis , Saccharomyces cerevisiae and E. faecalis produced the strongest overall results when the indicators were ranked together. The ranking used soluble protein, crude protein and microbial crude protein as positive indicators, ammonia nitrogen as a negative indicator and pH as a range-based indicator. The three best-performing strains were then tested in combinations using an orthogonal experimental design.

Four combinations were evaluated, each with or without each organism. The combination containing C. tropicalis and E. faecalis , without S. cerevisiae , had the highest measured soluble protein, at 34.28% of crude protein, the highest crude protein, at 19.4% of dry matter, and the highest microbial crude protein, at 43.29% of dry matter. It also had the lowest ammonia nitrogen, at 1.26 g/kg total nitrogen, and the lowest pH, at 4.19, among the four combinations tested.

These findings selected the two-organism culture for the subsequent rumen incubation experiment; they did not establish a molecular mechanism for its performance or demonstrate benefits in animals. The selected culture was used to ferment three substrates with different protein structures. S1 had a soluble-protein to insoluble-protein ratio of 1:2, with soluble protein representing 33.5% of crude protein; S2 used a ratio of 1:2.5 and 28.7% soluble protein; and S3 used 1:3 and 25.0% soluble protein.

Before fermentation, the formulations contained 18.2%, 17.1% and 24.3% crude protein on a dry-matter basis, respectively. The experiment therefore compared the targeted protein ratios together with the ingredient adjustments used to create them, rather than changing only the ratio in an otherwise chemically identical feed. S2 delivered the highest total volatile fatty acid concentration, 144.08 mmol/L, compared with 116.15 mmol/L for S3 and 103.08 mmol/L for S1.

The individual volatile fatty acids followed the same overall pattern. S2 contained 62.85 mmol/L acetate, 25.78 mmol/L propionate and 11.57 mmol/L valerate, all significantly higher than the corresponding values in S1 and S3 at the study’s statistical threshold. S3 produced intermediate values for total volatile fatty acids, acetate, propionate and valerate, while S1 had the lowest concentrations of these measures.

Butyrate and isobutyrate did not differ significantly among the three groups, and the acetate-to-propionate molar ratio was also similar, ranging from 2.40 to 2.49. Substrate degradation showed a somewhat different pattern. Crude-protein disappearance was significantly higher in S1 and S2 than in S3, although S1 and S2 did not differ significantly from one another.

Dry-matter disappearance increased over the 48-hour incubation and was numerically highest in S2 at the endpoint, followed by S3 and S1. Sampling time and the interaction between time and treatment were significant for dry-matter disappearance, but the treatment main effect was not. Culture pH declined gradually, with values fluctuating within 6.47–7.64, and treatment did not significantly affect pH.

Ammonia nitrogen differed with both time and treatment, while the time-by-treatment interaction was not significant. S1 and S3 maintained higher ammonia-nitrogen concentrations during part of the incubation, showing that nitrogen accumulation and the strongest overall endpoint response cannot be equated in a simple way. For the microbial analysis, the researchers amplified and sequenced the V3–V4 region of the 16S rRNA gene from the incubated communities, using six biological replicates per treatment.

Richness estimates, including the Chao, Ace and Simpson indices, did not differ significantly among S1, S2 and S3, although the Shannon index differed between S1 and S3. Principal-coordinate analysis showed distinct community structures among the three treatments, with a significant difference in beta diversity. At genus level, Streptococcus and Ruminobacter were relatively more abundant in S2, while Rikenellaceae_RC9_gut_group was less abundant in S2 and S3 than in S1.

Bacteroidetes and Bacillota together accounted for more than 83% of the sequences, and the groups contained 233, 159 and 217 unique operational taxonomic units, respectively. The microbial findings are consistent with the better fermentation measured in S2, but they do not demonstrate that Streptococcus or Ruminobacter caused the increase in volatile fatty acids or protein disappearance. The study did not measure functional gene expression, metagenomic functions or the activities of enzymes involved in nitrogen metabolism.

Its authors therefore describe the enrichment of these genera as a potentially important association. The selection of soluble and insoluble protein fractions followed the CNCPS protein-fractionation system: soluble protein comprised non-protein nitrogen plus soluble true protein, while insoluble protein was calculated as crude protein minus those two components. These analytical definitions helped quantify the protein structures that the probiotics fermented before the rumen incubation.

Several limitations constrain the conclusions. Fermentation can alter carbohydrate fractions as well as protein, but the researchers did not perform a comprehensive post-fermentation analysis of neutral-detergent fiber, acid-detergent fiber, starch or water-soluble carbohydrates. Consequently, differences in volatile fatty acid production, dry-matter disappearance, crude-protein disappearance and microbial composition cannot be attributed solely to the initial soluble-to-insoluble protein ratio.

The in vitro system also omits host immune responses, ruminal absorption and feed passage kinetics. Rumen fluid came from only three donor animals, only a limited set of strains was tested, and no animal growth, milk-production or feed-intake outcomes were measured. As a result, the findings support further in vivo testing but do not establish improved animal performance, reduced nitrogen excretion or a practical feeding recommendation.

Overall, the experiments show that probiotic fermentation can alter the protein structure of a cottonseed-meal-based substrate and that an intermediate soluble-protein proportion of 28.7% of crude protein corresponded to the most favorable measured rumen-fermentation outcomes. The C. tropicalis – E. faecalis combination also outperformed the screened single strains and the other tested combination in the feed-fermentation indicators. Yet the experiment demonstrates a laboratory association, not a causal mechanism or a proven benefit for cattle.

Validating whether the two organisms and the 1:2.5 ratio can improve nitrogen-use efficiency in vivo will require controlled animal studies, full compositional measurements of the fermented feeds and functional or omics-based analysis of the rumen microbiome.

Subject of Research: In vitro testing of probiotic-fermented feed protein structure and rumen fermentation in cattle

Article Title: In vitro investigation of probiotic-mediated enhancement of soluble protein utilization and rumen fermentation efficiency

Article References: Yang, X., Ma, X., Chen, N., Qi, Y., Li, S., Wu, Y., & Zhang, W. (2026). In vitro investigation of probiotic-mediated enhancement of soluble protein utilization and rumen fermentation efficiency. Frontiers in Veterinary Science, 13, Article 1899102. https://doi.org/10.3389/fvets.2026.1899102

Image Credits: AI Generated

DOI: 10.3389/fvets.2026.1899102

Keywords: probiotics, rumen fermentation, soluble protein, nitrogen utilization, cattle nutrition, cottonseed meal, volatile fatty acids, rumen microbiome, feed fermentation, in vitro study, vitro, investigation

Lipases, formally known as triacylglycerol ester hydrolases, are among the most versatile enzymes used in industry. They catalyze the hydrolysis of their natural triglyceride substrates as well as esterification and transesterification reactions, making them indispensable in the food, pharmaceutical, oleochemical, and biofuel sectors. Enzyme-catalyzed processes run under mild conditions with high selectivity and stability, allowing high-purity products to be made while reducing environmental impact compared with traditional chemical catalysts. The main obstacle remains economics: the cost of producing the enzyme strongly influences whether an enzyme-based bioprocess is commercially viable, which is precisely the problem the new study set out to address.

The research team, led by Tatiane de Souza Ribeiro and Gizele Cardoso Fontes Sant’Ana at the State University of Rio de Janeiro, exploited solid-state cultivation, or SSC, a technique in which microorganisms grow on a moist solid substrate in the absence or near absence of free water. The solid matrix simultaneously serves as a nutrient source and a physical support for fungal colonization. SSC is prized for its high yields and productivity, low contamination risk, simple bioreactor requirements, and low production costs, and it closely mimics the natural habitat of filamentous fungi, favoring hyphal growth and the secretion of extracellular enzymes. Using agro-industrial co-products in SSC further boosts sustainability, since these residues supply carbon, nitrogen, minerals, and moisture with little or no pretreatment.

The choice of substrates was deliberate and data-driven. Cottonseed meal is rich in lipids—about 11.2 grams per 100 grams in this study—composed largely of triacylglycerols containing linoleic, oleic, and palmitic acids, all known inducers of lipase production. Wheat bran, meanwhile, has low lignin content, high nutritional value, favorable particle structure, and high porosity, all of which promote microbial colonization and oxygen transfer. The timing is opportune: Brazil surpassed the United States in cotton production in 2024, reaching more than 3.7 million tons, generating abundant cottonseed coproducts, and the country ranks among the world’s leading wheat and castor oil producers.

To optimize enzyme production, the researchers cultivated a mutant strain of Aspergillus niger, 11T53A14, on cottonseed meal alone or in a one-to-one blend with wheat bran, with and without 2 percent castor oil as an inducer, at initial moisture contents ranging from roughly 30 to 60 percent. The winning combination was pure cottonseed meal supplemented with castor oil at 54 percent initial moisture, which yielded a maximum lipase activity of 93.1 units per gram of dry mass after just 48 hours of cultivation. Activity rose 3.7-fold when moisture increased from 35 to 54 percent, and castor oil supplementation alone nearly doubled activity under the same moisture condition, underscoring the importance of both water availability and lipid induction in SSC.

Moisture control emerged as a central technical theme. Adequate moisture dissolves and transfers nutrients, promoting lipase activity, but excessively high moisture reduces substrate porosity and raises viscosity, inhibiting fungal growth, while insufficient moisture starves the fungus of diffusible nutrients. The cottonseed-wheat blend retained more water than cottonseed meal alone, likely because wheat bran’s high fiber content of about 41.6 percent creates larger pores and interstitial spaces. Cultivation pH drifted slightly downward during fermentation, a consequence of fatty acid release from triglyceride hydrolysis and acid production by microbial metabolism, though A. niger tolerates this range well during lipase production.

With enzyme extracts in hand, the team immobilized the lipases on a mesoporous poly(styrene-co-divinylbenzene) support synthesized by aqueous suspension polymerization. The hydrophobic copolymer, with a specific surface area of 259 square meters per gram and an average pore diameter of about 20 nanometers, was prepared with a toluene and n-heptane porogenic mixture that favors phase separation and large pore formation. Hydrophobic supports are especially effective for lipases because immobilization proceeds through interfacial activation, and the approach can double as a partial purification strategy—valuable here because the SSC-derived enzymes were used as crude extracts, skipping costly purification steps entirely.

Immobilization yields ranged from 67.3 to 98.2 percent, with the blend-derived enzymes approaching the 94.9 percent yield of a commercial A. niger lipase processed identically. Recovered activities, which measure how much immobilized enzyme remains catalytically active, were lower, spanning 5 to 30 percent, a common outcome attributed to conformational changes during immobilization, enzyme dimerization, and mass-transfer limitations as substrates and products diffuse through the support’s pores. Notably, the biocatalyst built from the crude SSC enzyme produced on cottonseed meal and castor oil achieved the highest recovered activity at 30 percent, outperforming the biocatalyst made from the commercial enzyme, suggesting the crude fungal extract was functionally comparable to its purified, market-ready counterpart.

The true test came in castor oil hydrolysis, the principal industrial route to ricinoleic acid, a high-value hydroxylated fatty acid used to synthesize sebacic, heptanoic, and undecylenic acids for polymers, lubricants, and cosmetics. Enzymatic hydrolysis proceeds under mild conditions, cutting energy consumption and avoiding degradation of heat-sensitive compounds. The SSC-derived biocatalysts achieved ester-to-free-fatty-acid conversions of 17.7 to 45 percent. The best preparation converted 22.3 percent of the oil, beating the commercial-enzyme biocatalyst on the same support at 9.7 percent, surpassing the commercial Lipozyme TL at 17.7 percent in this assay, and performing on par with Lipozyme 435 at 20.3 percent, though still trailing Lipozyme RM at 42.4 percent. Most strikingly, the leading biocatalyst maintained its conversion efficiency over five consecutive 24-hour hydrolysis cycles with no significant decline, a reusability profile that directly reduces process costs relative to free enzymes, which cannot be recovered. The authors conclude that agro-industrial by-products can serve as inexpensive substrates for lipase production and that SSC-derived immobilized biocatalysts represent sustainable, cost-effective alternatives for enzymatic hydrolysis at industrial scale.

Beyond the headline results, the study sits within a broader industrial logic that makes it noteworthy. Aspergillus niger has long held GRAS status—Generally Recognized as Safe—which means enzymes derived from it face fewer regulatory hurdles in food and pharmaceutical applications than those from less-characterized microbes. The strain used here is also thermostable and 1,3-specific, with notable tolerance to glycerol, traits that matter in industrial biotransformations where reaction mixtures can become viscous and glycerol-rich as triglycerides are broken down.

The selection of castor oil as both an inducer during cultivation and the substrate for hydrolysis reflects a deliberate circularity. Brazil is the world’s second-largest castor oil producer, with roughly 87 percent of cultivation concentrated in the state of Bahia, and earlier work had shown that supplementing SSC media with 2 percent castor oil outperformed soybean, olive, corn, and palm oils as a lipase inducer. By using the same oil to induce enzyme production and to test the resulting catalyst, the researchers created a self-reinforcing value chain: a cheap regional commodity induces the enzyme, and the enzyme then upgrades that same commodity into ricinoleic acid, a platform molecule for sebacic, heptanoic, and undecylenic acids used in polymers, lubricants, and cosmetics.

The choice of a hydrophobic poly(styrene-co-divinylbenzene) support also deserves emphasis. Lipases possess a flexible lid over their active site that opens at oil-water interfaces, a phenomenon known as interfacial activation. Hydrophobic supports mimic this interface, locking the enzyme in its open, active conformation upon contact. This mechanism explains why immobilization on such materials can simultaneously purify and activate lipases, an advantage amplified here because the researchers deliberately avoided purifying their crude SSC extracts, eliminating one of the most expensive steps in industrial enzyme production.

The gap between immobilization yields and recovered activity, while seemingly disappointing, is typical of the field and instructive. High yields confirm that most enzyme protein attaches to the polymer, but conformational changes, enzyme dimerization, and diffusion limitations inside pores of roughly 20 nanometers can leave a fraction of the bound protein catalytically inaccessible. That the crude SSC-derived preparation achieved the highest recovered activity of any biocatalyst tested, including one built from a commercial purified enzyme, suggests that co-adsorbed components of the fungal extract may stabilize the lipase or that the crude enzyme’s intrinsic properties suit the support particularly well.

Reusability is where immobilization pays off economically. Free enzymes are discarded with the reaction mixture after a single use, whereas a heterogeneous biocatalyst can be filtered off and redeployed. Sustaining conversion across five 24-hour cycles without significant loss indicates strong physical retention and operational stability, though longer campaigns and different substrates would be needed to confirm industrial durability. Future work will likely need to address scale-up of SSC bioreactors, standardization of variable agro-industrial feedstocks, and enzyme loading optimization. Still, the demonstration that a crude, low-cost fungal extract can match commercial preparations on a synthetic support marks a meaningful step toward economically viable enzymatic hydrolysis in the oleochemical sector.

Subject of Research: Production of immobilized Aspergillus niger lipase biocatalysts via solid-state cultivation on agroindustrial residues for castor oil hydrolysis

Article Title: Heterogeneous biocatalysts based on porous polymer and lipase from Aspergillus niger obtained from SSC employing agroindustrial residues as raw material

Article References: de Souza Ribeiro, T., Torquato, E. C. C., Manoel, E. A., Cipolatti, E. P., da Cunha Costa, L., & Sant’Ana, G. C. F. (2026). Heterogeneous biocatalysts based on porous polymer and lipase from Aspergillus niger obtained from SSC employing agroindustrial residues as raw material. 3 Biotech, 16(10), Article 417. https://doi.org/10.1007/s13205-026-05042-0

Image Credits: AI Generated

DOI: 10.1007/s13205-026-05042-0

Keywords: Aspergillus niger, lipase, solid-state cultivation, enzyme immobilization, agroindustrial residues, cottonseed meal, wheat bran, castor oil, ricinoleic acid, poly(styrene-co-divinylbenzene), biocatalysis, green chemistry