Flavor-switchable scaffold for cultured meat with enhanced aromatic properties

Cultured meat research has largely emphasized structural scaffolds and achieving cell mass and differentiation to mimic the properties of slaughtered meat, yet key organoleptic properties such as flavor remain challenging to replicate due to differences in amino acid profiles and the roles of blood, fat, and connective tissues. Flavor of cooked meat arises from Maillard reaction products (aldehydes, alcohols, sulfur-containing compounds) formed when peptides react with reducing sugars at high temperatures (above 150 °C). The disparity in amino acid profiles between in vitro tissues and traditional meat complicates generating comparable Maillard flavors in cultured meat. Furthermore, volatile synthetic flavor compounds can be lost during long cell culture periods due to burst release and evaporation, leaving little residual flavor by the time of consumption. To address this, the study proposes a materials-based strategy that introduces a switchable flavor compound (SFC) into a gelatin methacryloyl scaffold to stably retain flavor precursors during culture and release meaty flavor molecules specifically upon cooking at Maillard-relevant temperatures, thereby enhancing the sensorial similarity of cultured meat to conventional beef.

Prior work focused on mimicking meat structure: Jeong et al. fabricated steak-type cultured meat using 3D printing, and Liu et al. produced meatball-type constructs via cellularized microtissues. Sensory aspects have begun to be explored: Joo et al. compared cultured chicken muscle to traditional chicken and found taste remained difficult to mimic without additional treatments because of amino acid differences; Liu et al. reported cultured fat exhibited fruity and creamy notes due to lipid oxidation. The authors’ previous work showed enhanced proliferation and differentiation of murine and bovine myoblasts could induce more meat-like flavors. In conventional food science, synthetic Maillard reaction products such as sulfur-containing furfuryl mercaptan are used to impart roasted meat aromas, but their volatility leads to rapid loss, a problem exacerbated in cultured meat by long culture times (e.g., 15–28+ days). These gaps motivate strategies that stably entrap and thermally release desired volatiles to emulate cooking-driven flavor formation.

Design and synthesis of switchable flavor compound (SFC): SFC was designed with three moieties: two methacrylate-terminated binding groups (R1, R2) for covalent incorporation into gelatin methacryloyl (GelMA), and a thermoresponsive flavor group (R3) incorporating a disulfide linkage to a Maillard reaction product (furfuryl mercaptan; furan-2-ylmethanethiol). Disulfide exchange under heat enables release of the flavor moiety. Disulfide-linked Maillard products were formed by reacting 3-mercapto-1-propanol (11.6 mmol) with furfuryl mercaptan, 3-mercapto-2-pentanone, or 2-methyl-3-furanthiol (equimolar) in the presence of H2O2 (2.56 mmol) at 80 °C for 48 h. SFCs were prepared via urethane reaction with hexamethylene diisocyanate isocyanurate trimer (isocyanate trimer) and 2-hydroxyethyl methacrylate (molar ratio isocyanate trimer:HEMA:disulfide product = 1:2:1) in propylene carbonate. Reaction conditions: for furfuryl mercaptan SFC 30% w/v reagents, 0.5 mmol isocyanate trimer in 1 mL solvent, 80 °C for 4 days; for 3-mercapto-2-pentanone and 2-methyl-3-furanthiol SFCs 20% w/v, 0.3 mmol isocyanate in 1 mL solvent, 80 °C for 3 days. Structure confirmation by Raman spectroscopy (disulfide signals at 486, 515, 2570 cm−1) and 1H NMR (methacrylate, furan, and urethane signals). Hydrogel fabrication: GelMA was synthesized from fish gelatin by methacrylic anhydride functionalization, dialysis (5 days), and lyophilization. Scaffolds were prepared from 20% (w/v) GelMA in water with 0.1% (w/v) photoinitiator (2-hydroxy-4-(2-hydroxyethoxy)-2-methylpropiophenone, 12959). Groups: Gel-SFC (no SFC), Gel+SFC (0.5% w/v SFC with furfuryl mercaptan), Gel+FM (0.5% w/v pure furfuryl mercaptan physically mixed), and Gel+SFCV (mixture of SFCs each at 0.16% w/v: furfuryl mercaptan SFC, 3-mercapto-2-pentanone SFC, 2-methyl-3-furanthiol SFC). Precursor solutions were poured into 24-well plates and UV-polymerized for 3 h, then washed (four water washes, including swelling at 37 °C for 6 h) to remove unreacted molecules. Thermoresponsivity and stability testing: SFC UV-Vis spectra (200–600 nm) were recorded after heating sealed SFC solutions (1.86 µM) at 37, 80, or 150 °C up to 24 h. The 335 nm peak tracked furan mobility; onset of response at 80 °C, saturation ~12 h. Stability in open system at 37 °C was assessed by gravimetry over 14 days comparing SFC (0.37 mM) vs pure furfuryl mercaptan; 1H NMR verified SFC chemical stability. Hydrogel flavor stability: hydrogels were immersed in water for 15 days, then analyzed at room temperature or after 150 °C heating for 5 min. Volatiles were collected by HS-SPME (CAR/PDMS/DVB fiber) with autosampler adsorption (30 °C, 20 min for RT samples; 80 °C, 20 min for heated samples plus 40 min at RT) and analyzed by GC-MS (Agilent 8890/5977B; DB-WAX column; oven 40 °C 5 min, ramp 4 °C/min to 240 °C, hold 20 min; inlet 250 °C, split 20:1; He carrier). Cell culture on scaffolds: Hydrogels were lyophilized (−50 °C, 4 days) to aerogels, sterilized (70% ethanol, UV 2 h), and swollen in high-glucose DMEM. Primary bovine myoblasts (passage 3) were seeded at 2×10^6 cells per scaffold (16.6 mm^2×0.8 mm). Proliferation: 7 days in HG-DMEM with 10% FBS, 1% PS, 5 ng/mL bFGF, medium change every 2 days. Differentiation: 8 days in HG-DMEM with 5% horse serum and 1% PS. Groups post-culture: CM-SFC (Gel-SFC), CM+SFC (Gel+SFC), CM+SFCV (Gel+SFCV). Biological assessments included CCK-8 viability (days 1, 5, 7), immunostaining for F-actin (phalloidin) and nuclei (DAPI), myogenic differentiation via MHC staining (MF20) and quantification by bovine myosin-1 (MYH1) ELISA. Scaffold swelling and compressive modulus were measured (Supplementary) to interpret proliferation differences. Flavor analysis of cultured constructs: After differentiation, samples were heated at 150 °C for 5 min to induce volatilization. GC-MS quantified flavor notes for CM-SFC and CM+SFC using HS-SPME/DB-WAX as above. An electronic nose (HERACLES-II, Alpha MOS) assessed CM-SFC, CM+SFC, CM+SFCV, and beef brisket: 0.4 g per sample in 20 mL vials, heated 150 °C for 5 min; injection 125 µL/s, injector 200 °C, H2 carrier, dual columns (MXT-5 non-polar, MXT-1701 medium polarity) and dual FIDs at 260 °C. Principal component analysis evaluated similarity to beef. Flavor notes were assigned using the FEMA database; compounds classified as pleasant or off-flavor per study-defined categories.

- Thermoresponsive SFC behavior: UV-Vis revealed increased absorbance near 335 nm upon heating, indicating flavor group mobility; onset at 80 °C with saturation by ~12 h. At 37 °C, SFC showed negligible release, confirming selectivity for cooking temperatures. - Stability of flavor moiety: In an open system at 37 °C, pure furfuryl mercaptan rapidly evaporated (residual 60.9% at day 3; 6.76% at day 14), whereas SFC retained 93.8% mass at day 14; 1H NMR confirmed chemical stability over 14 days. - Hydrogel flavor retention and release: After 15 days aqueous incubation, Gel+SFC displayed enhanced retention and selective release of pleasant meaty volatiles upon heating to 150 °C, while Gel+FM (physical mix) showed poor retention. Before heating, Gel-SFC was dominated by off-flavor compounds; after heating, pleasant flavors increased. Flavor classification used FEMA notes and study-defined categories. - Cell compatibility and myogenesis: Myoblast morphology on CM-SFC and CM+SFC appeared similar. CCK-8 showed slightly lower day-7 viability on CM+SFC, attributed to reduced compressive modulus and higher swelling due to SFC crosslinking effects; however, there was no decline from day 1 to day 7, indicating slower proliferation rather than cytotoxicity. Differentiation was comparable: MHC-positive myotubes were observed in both groups, and ELISA showed similar MHC levels (CM+SFC normalized to CM-SFC ~1.0). - Flavor of cultured meat constructs: After heating at 150 °C, CM-SFC yielded primarily benzaldehyde (almond-like). CM+SFC produced meaty flavor notes with detection of furfuryl methyl disulfide and other sulfurous/meaty compounds derived from SFC volatilization. Electronic nose analysis showed higher ratios of pleasant/meaty notes in CM+SFC and CM+SFCV compared to CM-SFC. PCA (discrimination index=90) indicated CM+SFC was more similar to beef than CM-SFC, and CM+SFCV (with three SFC flavors: furfuryl mercaptan, 3-mercapto-2-pentanone, 2-methyl-3-furanthiol) was the closest to traditional beef among the cultured groups, reproducing a meat–floral/creamy–fruity dominance pattern akin to beef.

The study demonstrates that covalently tethered, thermoresponsive switchable flavor compounds embedded in GelMA scaffolds can bridge a key sensory gap in cultured meat by enabling on-demand release of meaty Maillard-like volatiles only at cooking temperatures. The SFC design leverages dynamic disulfide exchange to retain volatile flavor precursors during extended open-system cell culture and to trigger release during cooking, aligning with the thermal nature of culinary processing. The approach overcame volatility-driven losses typical of physically mixed flavors. Beyond the intended furfuryl mercaptan release, additional meaty and savory compounds (e.g., 2-methylthiophene, furfuryl methyl disulfide, 2,2-(dithiodimethylene)difuran) were detected, plausibly due to residual H2O2 used in synthesis generating oxidative and radical pathways at high temperature, thereby expanding the bouquet. Biologically, introducing SFC did not impair myogenic differentiation, though proliferation was modestly slower, potentially due to altered scaffold mechanics. Flavor analysis via GC-MS and e-nose confirmed that SFCs meaningfully shifted the volatile profile toward pleasant/meaty notes and that incorporating multiple SFC flavors (SFCV) enhanced similarity to beef. This materials strategy provides a controllable, cooking-triggered route to mimic aspects of the Maillard reaction in cultured meat, with potential extension to food-grade chemistries and plant-based scaffolds. PCA analyses support that multi-flavor SFCs can better approximate the complex flavor space of conventional meat.

This work introduces a flavor-switchable scaffold that covalently incorporates thermoresponsive switchable flavor compounds into GelMA to stably retain and then release meaty Maillard-like volatiles upon cooking. The system remains stable during cell culture, does not impede myogenic differentiation, and significantly enriches pleasant/meaty flavor notes in cultured meat, moving its aromatic profile closer to beef. Extending SFCs to include multiple Maillard-related thiol flavors (furfuryl mercaptan, 3-mercapto-2-pentanone, 2-methyl-3-furanthiol) further improved similarity to conventional beef per electronic nose and PCA. Future work should: (1) scale the diversity and quantity of SFC flavor agents to more comprehensively emulate the many Maillard products in traditional meat; (2) integrate strategies that increase cell content and leverage cell-derived proteins to synergize with SFC-driven flavors; (3) transition to fully food-grade chemistries (e.g., transglutaminase crosslinking, cysteine-based disulfides) and plant-derived protein scaffolds; and (4) adopt serum-free media to minimize confounding volatile interactions and better isolate SFC effects.

Key limitations include: (i) current scaffold chemistry (GelMA, methacrylic anhydride, photoinitiator) is biocompatible but not food-grade; no GelMA variant is FDA-approved for food use. (ii) Despite improvements, cultured meat with SFC still differs from traditional meat’s complex flavor; three flavor agents likely remain insufficient to replicate dozens of Maillard compounds. (iii) Slightly reduced proliferation on SFC-containing scaffolds may relate to altered mechanics; further optimization is needed. (iv) Serum-containing media may contribute volatiles that interact with SFC-derived compounds, potentially converting flavors to non-flavor or undesired products; serum-free conditions would clarify SFC contributions. (v) Use of animal-derived gelatin and serum conflicts with end-goal sustainability; plant-derived proteins and serum-free media should be validated.