Utilizing raw rapeseed press cake in foods: A case study on sensory quality and profile of selected bitter compounds in snack bars

This study demonstrates that incorporating raw rapeseed press cake into snack bars creates dose-dependent sensory attributes, particularly bitterness and astringency, which are primarily driven by specific metabolites like KSS, KSS-hexose, and goitrin rather than total protein content or dehulling treatment.

The Gist

Imagine you have a giant, golden factory that turns seeds into cooking oil. Once the oil is squeezed out, what's left behind is a dry, crumbly cake called Rapeseed Press Cake (RPC).

Think of this cake like the "grapefruit pulp" left after you squeeze juice. It's packed with protein and nutrients, making it a superfood in theory. But in practice, it's like a bitter, gritty brick that nobody wants to eat. It tastes terrible (bitter and astringent) and contains compounds that can be harmful in large amounts. Because of this, it usually ends up as animal feed, not human food.

The Big Question:
Can we take this "bitter brick," crush it up, and mix it into something delicious like a snack bar without ruining the taste? And if it does taste bitter, what exactly in the cake is causing that bitterness?

The Experiment: The "Bitter Bar" Test
The scientists in this study decided to play the role of culinary alchemists. They took raw, unprocessed rapeseed press cake and baked it into snack bars (think of them as energy bars made of dates, nuts, and coffee).

They made two types of bars:

1. Whole Seed Bars: Made from seeds with their tough outer shells (hulls) still on.
2. Dehulled Bars: Made from seeds where the shells were removed first.

They added the cake in increasing amounts: 0% (just the normal bar), 7%, 14%, and a heavy 21%. Then, they brought in a team of super-tasters (a trained sensory panel) to rate the bars on a scale of 1 to 15.

What They Found (The "Taste" Story)

• The Bitterness Ceiling: As they added more cake, the bitterness and astringency (that dry, puckering feeling like strong black tea) got worse. But here's the twist: once they hit 14%, the bitterness stopped getting much worse, even when they added 21%. It was like the taste buds hit a "ceiling" or got overwhelmed. The sweet, nutty flavors of the dates and almonds seemed to mask the extra bitterness.
• The "Astringency" Ramp: While bitterness hit a ceiling, the "dry mouth" feeling (astringency) kept getting worse and worse the more cake they added. It was a straight line.
• The Hull Surprise: You might think removing the tough outer shells (hulls) would make the cake taste better or have more protein. Surprisingly, it didn't. The "dehulled" bars actually tasted more intense in some ways (like smelling more like cabbage or popcorn) but weren't less bitter. The shells didn't seem to be the main source of the bad taste; the "meat" of the seed was.

The Detective Work: Finding the "Bad Guys"
The scientists didn't just rely on taste; they used high-tech microscopes (mass spectrometry) to hunt down the specific chemical molecules responsible for the bad flavors. They were looking for the "criminals" in the cake.

They found three main suspects:

1. Goitrin: A chemical that forms when a compound called progoitrin breaks down.
2. KSS & KSS-hexose: Complex flower-like molecules (flavonoids) that are naturally bitter.
3. Sinapic Acid: A phenolic compound known for bitterness.

The Big Discovery:

• The Transformation: When they mixed the cake into the snack bar, the chemistry changed. About 90% of the original "progoitrin" (a precursor) disappeared! It didn't just stay there; it broke down.
• The Real Culprit: The scientists calculated a "Dose-over-Threshold" score. This is like asking, "Is there enough of this chemical in the bar to actually be tasted?"
  • Progoitrin was present, but the amount was too low to be tasted.
  • Goitrin, however, was present in high enough amounts to be the main source of the bitterness. It was the "smoking gun."

The Takeaway for the Future
This study is like a roadmap for turning waste into wealth. It tells us:

1. We can use raw rapeseed cake in food, but we have to be careful about how much we add (around 14% seems to be the sweet spot before it gets too weird).
2. The hulls aren't the problem. Removing them doesn't fix the taste issue.
3. The real fix is breeding. Since we now know exactly which chemicals (like Goitrin and KSS) cause the bitterness, plant breeders can go back to the fields and grow new types of rapeseed that naturally have less of these "bad guys."

In a Nutshell:
Imagine trying to make a chocolate cake using a bitter, green vegetable. This paper says, "Yes, you can do it, but you need to know which part of the vegetable is making it taste bad." They found that the "bitterness" comes from specific chemicals that change when mixed with food. Now, instead of just throwing the vegetable away, we can either cook it carefully or breed new vegetables that are naturally sweeter, turning a waste product into a superfood snack.

Technical Summary

1. Problem Statement

Rapeseed press cake (RPC), a byproduct of oil extraction, is a protein-rich (30–40%) co-product with significant potential as a sustainable human food source. However, its utilization is currently limited by two major barriers:

• Regulatory and Safety: Raw RPC contains high levels of anti-nutritional compounds, specifically glucosinolates, preventing its approval as a food ingredient in the EU.
• Sensory Quality: Consumers perceive RPC as having an undesirable bitter taste and astringency, often associated with its use as animal feed.

While specific bitter compounds (e.g., sinapine, sinapic acid, kaempferol glycosides, and glucosinolates) have been identified in isolation, there is a lack of knowledge regarding how these compounds behave within a complex food matrix. It is unknown how the introduction of raw, unprocessed RPC affects sensory perception in a real food product and which specific metabolites drive the bitterness in that context.

2. Methodology

The study employed a combined approach of food formulation, sensory science, and targeted metabolomics.

• Food Formulation:
  • Ingredients: Raw RPC was produced via cold-pressing from both whole and dehulled "canola quality" rapeseed seeds.
  • Prototype: Snack bars were formulated using a base of roasted almonds, coffee, flax, pumpkin, chia, dates, and raisins.
  • Experimental Design: The base was substituted with RPC at four levels: 0% (control), 7%, 14%, and 21%. Both whole-seed and dehulled-seed RPC were tested at each dosage.
• Sensory Analysis:
  • A trained panel of 10 experts evaluated the bars using descriptive analysis.
  • Attributes: 13 RPC-characteristic sensory attributes were defined (e.g., oxidized odor, cabbage flavor, bitter taste, astringent mouthfeel).
  • Protocol: Samples were presented monadically in a randomized block design over three sessions using anchored scales.
• Chemical Analysis (Metabolomics):
  • Targeted LC-MS/MS: Quantification of specific bitter compounds including:
    • Phenolics: Sinapic acid (sinapine quantification failed due to matrix effects).
    • Flavonoids: Kaempferol 3-O-(2‴-O-sinapoyl-β-sophoroside) (KSS) and KSS-hexose.
    • Glucosinolates: Progoitrin, gluconapin, glucobrassicin.
    • Hydrolysis products: Goitrin (derived from progoitrin).
  • Protein/Amino Acids: Crude protein content (via combustion) and free amino acid profiles were analyzed.
• Data Analysis:
  • Statistical analysis included 2-way and 3-way mixed-model ANOVA.
  • Principal Component Analysis (PCA): Correlated sensory attributes with metabolite levels.
  • Dose-over-Threshold (DoT): Calculated to determine if metabolite concentrations exceeded the sensory perception threshold for bitterness.

3. Key Contributions

• First Sensory Panel Assessment: This is the first study to evaluate the sensory profile of raw RPC (from both whole and dehulled seeds) incorporated into a whole-food matrix (snack bars).
• Metabolite-Sensory Correlation: It establishes a direct link between specific RPC-derived metabolites and perceived bitterness/astringency in a food product, moving beyond isolated compound studies.
• Discovery of Goitrin's Role: The study identifies goitrin as a primary driver of bitterness in RPC-containing foods, a finding previously unreported in this context.
• Dehulling Impact: It provides data on whether dehulling seeds prior to pressing improves the sensory profile or protein content of the final food product.

4. Key Results

Sensory Profile

• Dose-Dependence: Most sensory attributes (oxidized odor, bitter taste, cabbage flavor) increased with RPC dosage but plateaued at 14%. Adding 21% RPC did not significantly increase perceived intensity further, suggesting sensory adaptation or masking by the sweet, dense food matrix.
• Astringency: Unlike other attributes, astringent mouthfeel increased almost linearly across the full dose range (0–21%).
• Dehulling Effect: Bars made with dehulled RPC had higher intensities for specific attributes (overall odor, cabbage odor/flavor, popcorn flavor) compared to whole-seed RPC. However, dehulling did not increase bitterness or protein content in the final bar.

Metabolite Analysis

• KSS and KSS-Hexose: These compounds were present in raw RPC and increased in snack bars with dosage. Surprisingly, their levels in the snack bars were 5–9 times higher than expected based on raw RPC levels, suggesting hydrolysis of higher-order conjugates during food preparation or improved extractability in the matrix.
• Glucosinolates & Goitrin:
  • Raw RPC from whole seeds had lower total glucosinolates but higher goitrin levels than dehulled RPC, indicating pre-press hydrolysis in whole seeds.
  • In the final snack bars, only ~10% of the introduced progoitrin was recovered as goitrin, indicating extensive hydrolysis or further metabolism/loss during processing.
• Protein: Incorporating RPC did not significantly alter the crude protein content of the bars (remaining ~15–17%), nor did dehulling significantly boost protein levels in the final product.

Correlation and Thresholds

• PCA Findings: Bitterness and astringency were positively correlated with KSS, KSS-hexose, and goitrin. Sinapic acid showed a weaker correlation.
• DoT Analysis:
  • Progoitrin: DoT factors were <0.1 (below perception threshold).
  • Goitrin: DoT factors ranged from 0.3 to 1.4, indicating that goitrin concentrations in the bars were high enough to contribute to perceived bitterness.

5. Significance and Implications

• Food Product Development: The study demonstrates that raw RPC can be integrated into foods up to 14–21% without linearly increasing bitterness to unpalatable levels, likely due to matrix masking. However, astringency remains a linear challenge.
• Breeding Targets: The findings suggest that breeding rapeseed lines with reduced levels of KSS, KSS-hexose, and goitrin (rather than just total glucosinolates) would be more effective for improving the sensory quality of RPC-based foods.
• Processing Insights: Dehulling seeds before pressing offers no sensory or protein advantage for the final snack bar and may actually intensify certain off-flavors (cabbage/popcorn).
• Safety & Regulation: The identification of goitrin as a key bitter compound highlights its dual role as a sensory defect and a potential anti-thyroid agent, reinforcing the need for careful management of RPC in human diets.

In conclusion, the paper provides a foundational framework for utilizing raw RPC in human food, identifying specific chemical drivers of sensory defects and offering strategies for future product formulation and crop breeding.