Deep dive: Cultivated meat end products
Learn about the science behind cultivated meat end-product design and characterization, and the opportunities for innovation and open-access research they present.
Cultivated meat end products
Introduction
The goal for researchers and companies investigating cultivated meat is to develop methods of producing end products to meet or exceed customer expectations for conventional meat while incurring fewer externalized costs. For some whole-cut products grown on a scaffold, the end product may simply be harvested from the bioreactor and packaged without further processing. For other products, cultivated meat producers may feed mature cells into the same processing methods used for plant-based or conventional meat. Still other products may be structured using 3D printing or other methods. Because cultivated meat is composed of many or all of the same cell types in a similar arrangement as conventional meat, the relationship between the properties of the cells and tissues and those of the final meat product will be essentially the same irrespective of the production process. Therefore, much of what has already been learned by meat science researchers will apply to cultivated meat. This page provides an overview of some of the considerations relevant to cultivated meat end products.
Basics of muscle structure as relevant to meat
Most meat comes from the skeletal muscle of terrestrial or aquatic animals, the structure of which has implications for the properties of meat. Organ meats and other products will have different properties depending on the structure of the originating tissue. This section will focus on skeletal muscle, as most cultivated meat companies are starting with these products.
Skeletal muscle is typically comprised of around 90% muscle fibers, with the remaining 10% consisting primarily of fat and connective tissue (Listrat et al. 2016). Mature muscle fibers are long, multinucleated cells and typically show strong alignment parallel to one another within the muscle tissue. Muscle fibers are surrounded by a sheath of extracellular matrix (ECM) tissue called the endomysium. Groups of muscle fibers are further organized into bundles—called fascicles—by a more rigid ECM layer called the perimysium. A third ECM layer, the epimysium, surrounds the entire muscle.
The overall structure of a terrestrial vertebrate muscle consists of muscle fibers aligned along the long axis of the muscle, which are further subdivided into fascicles by the perimysium. Individual fibers within a terrestrial vertebrate muscle are often shorter than the muscle itself, with the surrounding connective tissue holding the fibers together. In contrast, fish muscles are organized into thin sheets called myomeres, with the fibers aligned perpendicular to the long axis of the muscle and spanning from one side of the muscle sheet to the other. Myomeres are separated from one another by layers of fat and connective tissue called myosepta. Crustacean and bivalve muscles show an organization roughly similar to that of terrestrial vertebrates, with muscle fibers running in a single direction and no myosepta (Yang et al. 2021; Sun et al. 2019). However, there is evidence that crustaceans may differ in having few or no intramuscular adipocytes and—for some taxonomic groups—in the fine-scale organization of the ECM (discussed later in this section). Cephalopod muscles show a substantially different organization, with fiber bundles running in three approximately orthogonal directions due to their reliance on jet propulsion for swimming (Gosline and DeMont 1985; Kier 2016).
Muscle fibers
Mature muscle fibers (myofibers) express a characteristic set of proteins required for muscle contraction, most notably actin and myosin. The properties of these proteins influence the muscle’s texture when consumed as meat and its response to cooking. Muscle fibers also express variable amounts of the oxygen carrier protein myoglobin, the content of which affects the color, nutritional properties, and flavor of the meat. Myoglobin-rich oxidative or “red” muscle is typically involved in sustained, low-intensity activities such as walking or standing and relies on aerobic respiration. Glycolytic or “white” muscle is used mainly for high-intensity bursts of activity such as running or flying and is dependent on anaerobic respiration. Many studies have found that white muscle tends to contain more intramuscular fat than red, though this is not a strict rule (Listrat et al. 2016).
Whereas oxidative and glycolytic fibers in terrestrial vertebrate muscles are typically interspersed with one another—in different ratios depending on the muscle and, in some cases, the part of the muscle—fish typically show a robust spatial separation between red and white muscle. In most fish species, the red muscle is confined to a thin band on either side of the animal, just below the skin, while most of the muscle is purely glycolytic. A band of “pink” muscle with intermediate properties separates the red muscle from the white.
Adipocytes & lipid profiles
Adipocytes are the primary fat storage cells in most animal tissues, including muscle. Mature white adipocytes (the main type found in muscle) contain a single, large lipid droplet within the cytoplasm. Triglycerides, the primary storage lipids, are mostly (~80%) stored within the adipocytes, while 5–20% are found as lipid droplets inside the myofibers (Listrat et al. 2016). Chemically, triglycerides consist of a single glycerol molecule covalently linked to three fatty acids. Phospholipids, the primary constituent of cell membranes—regardless of cell type—also contribute to the fat content of meat. Phospholipids are chemically similar to triglycerides, except that one of the three fatty acid chains is replaced by a negatively charged phosphate group. This negative charge results in an amphipathic molecule capable of interacting with both the hydrophilic environment of either the cytoplasm or the extracellular space as well as the hydrophobic environment within the cell membrane, making phospholipids the ideal structural components of membranes.
Within these broad classes, the structure of the fatty acid chains determines the characteristics of specific fats. Fatty acids are carboxylic acids with a hydrocarbon chain. In the case of triglycerides and phospholipids, the carboxylic acid is linked to the glycerol moiety. Fatty acids differ in two primary ways: chain length and degree of unsaturation. Saturated fatty acids contain no double bonds and therefore have the maximum number of hydrogen atoms possible given the length of the carbon chain (hence the name). Unsaturated fatty acids contain one or more double bonds, which can lower their melting temperature relative to saturated fatty acids of similar length. This explains why food products high in saturated fat (e.g., butter) remain solid at room temperature, while those higher in unsaturated fat (e.g., olive oil) are liquid. Triglycerides and phospholipids can “mix and match” their fatty acid chains, leading to a wide variety of molecular diversity within the broader category of lipids.
In meat, the amount of fat and the lipid profile are key considerations relevant to nutrition, sensitivity to oxidation, texture, and flavor, including distinctive species-specific flavors. Fat in mammalian muscle can be either intermuscular (located around the muscle) or intramuscular (interspersed within the muscle). The intramuscular fat is the primary determinant of meat sensory quality, as most intermuscular fat is usually removed (Listrat et al. 2016). The extent of intramuscular fat deposition determines the amount of marbling in the final meat product, with a high degree of marbling typically considered a positive attribute (Li et al. 2020). Fat in fish may be subcutaneous, within the perimysium, or within the myosepta. In the context of meat quality, fat within the myosepta is considered the most important (Listrat et al. 2016). The structure of aquatic invertebrate muscles has been less extensively studied than those of other common food animals. However, the low content of triglycerides relative to phospholipids in crustacean muscle (Chapelle 1977; Zhao et al. 2015; Lu et al. 2020; Shu-Chien et al. 2017) suggests that intramuscular adipocytes may be less prevalent in these animals, with muscle cells themselves being the primary determinants of the lipid profile.
The extracellular matrix
The skeletal muscle extracellular matrix (ECM) surrounds individual muscle fibers, fiber bundles (fascicles), and entire muscles. It provides mechanical strength as well as essential cues for the cells contained therein. Collagen content varies within and between species but generally ranges between 0.75–15% of muscle dry weight within cows, pigs, chickens, and fish (Listrat et al. 2016). The ECM’s composition and structure, especially that of collagen, is an essential determinant of meat texture and cooking properties (discussed later). The organization of the ECM also differs in key ways between species, which may be relevant for developing cultivated meat scaffolds and end products.
The ECM in fish is not quite as well understood as that of terrestrial vertebrates, though it is thought to follow the same overall organization as in other vertebrates (Listrat et al. 2016). However, Sleboda et al. (2020) did not find clear evidence of a distinct perimysium in carp muscle. It is unclear whether this is a feature of all fish or whether meaningful differences exist among fish species in the presence or organization of perimysial tissue. In the latter case, it might be interesting to investigate whether these differences correlate with differences in the texture of individual myomeres. At least in terrestrial meat, perimysium is a critical determinant of meat grain (Listrat et al. 2016). Aquatic invertebrates also share many features of this organization, with some interesting interspecies differences. Mizuta et al. (1994) observed distinct epimysium, perimysium, and endomysium by light microscopy of prawn, shrimp, lobster, and crayfish muscle. However, the perimysium and endomysium were difficult to distinguish in prawn, shrimp, and lobster. In contrast, crab muscle showed wider muscle fibers with a thick endomysium but no distinct perimysium and fine collagen fibers that extended into the muscle fiber area.
Myosepta in fish, which surround and separate individual muscle segments, have been described as structurally analogous to the epimysium of other vertebrate muscles (Listrat et al. 2016). However, Sleboda et al. (2020) found ultrastructural evidence that they should be considered tendons. Consistent with this, Bricard et al. (2014) found tenocytes within the myosepta. Whether they are considered the anatomical equivalent of the terrestrial vertebrate epimysium or the tendons, myosepta make a vital contribution to the sensory qualities of fish muscle as the primary source of fat and because of their breakdown upon heating, leading to the characteristic flaky texture of fish.
Other cell types present in muscle
A small percentage of the volume of vertebrate muscle is made up of other cell types, including fibroblasts, neurons, endothelial cells, and smooth muscle cells. While these cells are unlikely to have substantial direct effects on the nutritional or organoleptic properties of meat, they have indirect effects that may be important to consider in the context of cultivated meat (Bomkamp et al. 2022). Fibroblasts impact the mechanical properties of muscle—and thus the texture of meat—by secreting ECM components. Unless the scaffold convincingly replicates the mechanical properties of the ECM, ECM secretion by fibroblasts may be needed to achieve the correct texture in cultivated meat, which may explain why fibroblasts were the second most commonly used cell type cited by cultivated meat manufacturers in a 2020 survey.
Skeletal muscle in vivo is innervated by motor neurons that carry signals from the brain to trigger muscle contraction. There is some evidence that co-culture with motor neurons or direct electrical stimulation may promote the growth of muscle cells (Das et al. 2020; Furuhashi et al. 2021) and thus may be helpful for cultivated meat. Endothelial and smooth muscle cells make up the walls of blood vessels and play an essential role in vivo in transporting oxygen, nutrients, and waste products to and from muscles. Although tissue-engineered constructs containing true vascular networks have been demonstrated at small scales (Nishiguchi et al. 2011), achieving cell-based vasculature at the scale needed for a food product may be difficult. Therefore, alternative strategies such as simple channels without a lining of vascular cells (Miller et al. 2012) may prove to be more cost-effective. From an end product standpoint, fibroblasts, neurons, endothelial cells, and smooth muscle cells can likely be considered dispensable. However, depending on the chosen strategy, they may be important for some upstream parts of the meat cultivation process. They would then be present in the final product, as in conventional meat.
Differing needs for structural complexity across product categories
Meat products—whether cultivated or conventional—differ in how much they contain and showcase the 3D structure of muscle tissue. Meat products are often described as “structured” or “unstructured.” Ong et al. (2020) further distinguished between processed meats such as surimi, containing single cells or ruptured cells and no tissue structure to speak of; minced meats such as hamburgers, containing only small pieces of muscle and fat tissue that contribute to the structure of the product; and meat cuts such as steaks, consisting of large pieces of muscle tissue in or near to their natural state.
Within the broader category of meat cuts, some applications showcase the product’s structure in ways that require a high level of structural complexity. Others may have somewhat more lenient requirements. Meat products such as stews, soups, fajitas, steak sandwiches, poke, and pulled pork are distinct from processed or minced meat products in that they retain a good deal of the structure of the original muscle. However, such products are also distinct from whole meat cuts in that they are served in smaller pieces and often in dishes that heavily feature condiments or other non-meat ingredients that contribute additional flavors and textures. We can think of the meat cuts category as consisting of two rough subcategories, which we will refer to here as “semi-structured” and “fully-structured.” Semi-structured cuts may provide a useful stepping stone to more complex fully-structured cuts that need to stand on their own as “center of the plate” items. The earliest cultivated meat products on the market are likely to fall mainly on the unstructured end of the spectrum. As the technology improves, it will become possible gradually to replicate more structured and complex products.
If processed meats are the “low-hanging fruit” of the cultivated meat industry and fully-structured meat cuts are the “high-hanging fruit,” we might think of both minced meats and semi-structured meat cuts as in between points on that spectrum. Furthermore, we might expect that companies will be able to produce semi-structured cultivated meat well before they become entirely successful in producing fully-structured products. For example, a cultivated steak that is unsuccessful as a “center of the plate” item might be very successful when sliced and served in a sandwich, taco, or fajita.
Prototypes demonstrated by cultivated meat startups already cover much of the spectrum between unstructured and structured. The borders between these categories are blurry, especially as the distinction between fully-structured and semi-structured cuts depends in part on the intended cooking application. Examples of semi-structured prototypes might include Aleph Farms’ thin-cut steak or BlueNalu’s yellowtail. Several prototypes have also been presented in academic papers, though these are consistently at or below the centimeter scale. Some of these academic prototypes include some level of complex 3D structure. However, how that structure translates to consumer-relevant texture and flavor will be more easily evaluated with larger-scale products and more sophisticated analyses of food-relevant properties.
Access our alt protein photo library
Looking for more examples of cultivated meat product prototypes? Our photo library contains Creative Commons-licensed product images available for use from cultivated meat companies.
Implications of the carcass balancing problem
The mix of products produced by the conventional meat industry is fundamentally constrained by the muscle structure of the relevant animal. This phenomenon is often referred to as the carcass balancing problem. For example, only a small portion of an expensive cut such as filet mignon, derived from one end of the psoas major muscle, can be produced from a single cow. For obvious economic reasons, meat producers might prefer to yield 100% filet mignon from a given animal, but this is at odds with the anatomy of a cow. According to a white paper by Infor, “the problem lies in the design of the raw material itself.” In early 2020, the carcass balancing problem manifested in the form of increased demand for cheap cuts of meat and decreased demand for high-end cuts as people cooked at home instead of visiting restaurants during the early days of the Covid-19 pandemic.
Conventional meat producers are essentially forced into processing at least 38% of their beef (185 pounds of lean trim per 490 pounds of boneless, trimmed beef; exact percentages will vary by species and breed) into unstructured, lower-value products due to the carcass balancing problem. In contrast, a cultivated meat producer could hypothetically produce nearly 100% of the highest-value structured products if that was what their customers demanded. They would also have the flexibility to shift their product ratios in preparation for peaks in demand for particular product categories, such as for specific holidays, or in response to seasonal fluctuations in demand. Furthermore, if the differentiation and maturation phase for unstructured products is cheaper and less resource-intensive than for structured products, producers could increase their ratio of unstructured products directly and avoid some of the effort of producing fully structured cuts. This is in contrast to the current approach for conventional meat producers, whereby a fixed ratio of fully-structured cuts is produced only to grind them up during times of high demand for unstructured products. This would ultimately reduce the cost to both the customer and the environment.
As in the conventional meat industry, taking advantage of co-products from the cultivated meat production process may allow for more efficient production and additional revenue generation. For example, Shiok Meats has announced plans to create shrimp flavor paste and powder as co-products from their cultivated shrimp production process. Wild Earth has demonstrated a chicken broth made from cultivated chicken cells, which contains cell culture supernatant as an ingredient. In place of the carcass balancing problem, cultivated meat may have what we could call the “bioproduct balancing consideration,” in which the challenge is to design one or more bioprocesses to produce products efficiently in the ratios customers want. One possible solution would be to develop a unique bioprocess for each desired product and optimize each process to minimize waste production. However, it may be more efficient for a company to design one or a few bioprocesses with multiple co-products, the ratios of which can be altered as desired within a specific—and likely fairly wide—window.
Figure 10. The cuts of meat that can be produced from a single cow (A) are fundamentally constrained by the animal’s anatomy. Cultivated meat could be produced in the same ratios (B) by taking the appropriate number of cells following the proliferation phase and maturing them on appropriate scaffolds under the correct conditions to replicate the various cuts. These ratios could also be easily altered (C) to produce more of some cuts and fewer of others. This diagram is meant for illustrative purposes only, and is not meant to imply that cells from a single proliferation bioreactor would be fed into multiple downstream processes as shown here. Instead, it is more likely that, in practice, product ratios would be primarily controlled by altering the number of proliferation bioreactors feeding into each downstream process at a given time.
Realizing the benefits of cultivated meat will require solutions to several challenges, including developing scaffolds or alternative technologies to replicate fully-structured meat cuts. Once these challenges have been addressed, the long-term success scenario for cultivated meat is one in which product ratios are governed by demand rather than by the properties of the animal in question. We can also consider the implications of the carcass balancing problem on shorter-term scenarios. For instance, what if cultivated ground beef could be produced economically at scale, but semi- or fully-structured cultivated meat cuts were of low quality or disproportionately expensive? If only some conventional products (e.g., ground beef) were replaced, one might expect that the carcass balancing problem would lead to the waste of lower-value conventional products or at least major shifts in the relative prices and abundance of these products. Similar logic would apply if conventional meat were partially replaced by unstructured plant-based, fermentation-derived, or hybrid products.
Although ground beef is typically considered a meat industry by-product, the reality is somewhat more complex, especially in countries such as the U.S., where ground beef-based dishes have seen rising popularity over the last several decades (Close 2014). Due to price fluctuations, certain cuts of beef may be most profitable when sold in whole muscle form one week and as ground beef the next (Speer et al. 2015). In addition, the fat content of domestic beef trimmings is often higher than what the market demands in the ground beef market. Such by-products are often combined with imported lean beef to achieve the desired fat content in the final ground beef product. In both cases, beef suitable for sale as whole cuts—albeit mainly lower-value cuts—regularly ends up sold in ground form in substantial quantities.
As one hypothetical, let’s assume that alternative protein technologies can produce ground beef that competes successfully on taste and price with conventional products and that most consumers are willing to make the switch. Let’s also assume that whole-cut alternative protein products are not yet successful. Given those assumptions, what percent of the U.S. conventional beef market could be directly substituted by ground alternative beef while maintaining the current ratio between ground and whole-cut beef and without leading to waste of conventional ground beef? We can get a rough answer to our question by comparing the percentage of a cow carcass represented by trimmings versus the percentage of U.S. beef sales represented by ground beef. In 2014, Rabobank estimated that 62% of U.S. beef consumption was in the form of ground beef (Close 2014). In contrast, not counting fat trim and bone, a cow carcass yields around 38% lean trim beef (the fact that these percentages sum to 100% is purely coincidental).
For every 100 pounds of beef sold in the U.S., if we aim for 38% of conventional beef to be ground—the minimum allowable under the carcass balancing problem—and make up the difference using alternative proteins, we get:
Figure 11. Breakdown of various beef products according to a combination of consumer demand and the carcass balancing problem. Left: In the current scenario, 62% of beef sold in the U.S. is ground. The small inner ring in both panels indicates product ratios dictated by the carcass balancing problem. The surplus demand for ground beef relative to that dictated by the carcass balancing problem is made up by grinding lower-value whole-cut beef (shaded portion; it is assumed here that rib and loin steaks are all sold in whole-cut form). Right: In the hypothetical scenario where as much conventional ground beef as possible is replaced by alternative ground beef, it is possible to substantially reduce total conventional beef production while maintaining the overall balance between whole-cut and ground beef.
It would therefore be possible to maintain the current level of total U.S. beef sales and ratio of whole-cut to ground products while reducing the amount of conventional beef produced—and the associated negative externalities—by roughly 39%. This would lead to a corresponding 39% decrease in the availability of high-end whole cuts—e.g., rib and loin steaks—but would not require consumers to shift from whole cuts to unstructured products. The potential for displacement might be somewhat higher when in-home cooking is high relative to restaurant visits, as in the example mentioned above from early 2020, and will vary from country to country depending on the relative popularity of ground and whole-cut products. This is, of course, only a rough estimate based on broad categories, although meaningful subcategories exist within both the whole-cut and ground beef categories (Speer et al. 2015). In reality, the ratio between whole-cut and ground beef consumption would also be expected to change as the price differential changes, and prices of high-end whole cuts that are rarely or never ground would likely increase relative to other whole cuts.
This estimate should not be interpreted as a prediction. It is unlikely that unstructured alternative proteins will see this level of success before any meaningful success of structured products, especially considering that the availability of semi-structured cuts may substantially precede that of fully-structured cuts. Instead, this exercise is meant to place rough upper boundaries on the level of change that we might expect by substituting unstructured products alone. A more nuanced analysis of similar hypotheticals could yield additional insights, predictions, and indications of which alternative protein products have the potential to maximize effects on the near-term sustainability of our food system while minimizing disruptions to consumers. These analyses could include modeling more realistic scenarios for the relative success of ground and whole-cut alternative meats, predictions of price change impacts on demand for different cuts, consideration of semi-structured products, and other types of meat besides beef.
Despite its limitations, two key lessons we can take from this simple back-of-the-envelope estimate include:
- In markets where ground meat products are produced in higher quantities than those strictly required by the carcass balancing problem, such as in the U.S., unstructured alternative proteins alone can substantially impact total conventional meat consumption without asking consumers to switch from structured to unstructured products.
- For alternative proteins to command a larger market share in the absence of major consumer behavior change (i.e., consuming more processed and ground meat relative to whole cuts), innovations in semi-structured or fully-structured meat cuts will be required. As a rough estimate, around 39% may be the limit for displacement of conventional beef by ground alternative beef in the U.S., given the current demand for ground products.
Hybrids
Estimated production costs for cultivated meat have come down substantially in recent years, and future innovations have the potential to bring costs down further. Even so, producing cultivated meat at costs and scales capable of meaningfully competing with conventional animal meat is a daunting challenge. In addition, while cultivated meat produced using decarbonized energy sources is predicted to have lower impacts than conventional meat by most metrics, its environmental performance is still generally less favorable than that of plant-based meat. For these and other reasons, many companies are considering the strategy of hybrid meat products containing both cultivated animal cells and plant-based or fermentation-derived material. In some contexts, “hybrids” can include products composed of plant-based and fermentation-derived materials only or some combination of conventional meat and alternative proteins. In this deep dive, we will use the term “hybrids” to refer to products containing substantial amounts of both cultivated animal cells and plant-based or fermentation-derived material unless otherwise specified. For our purposes, hybrids do not include cultivated products with small, residual amounts of plant-based scaffolding or other material, but rather those where plant-based or fermentation-derived ingredients comprise an intentional and meaningful fraction of the final product.
Many of the first cultivated meat products on the market will likely be hybrids. In fact, as of early 2023, the only cultivated meat products on the market are cultivated/plant-based hybrid chicken products from GOOD Meat, available in Singapore. If costs decrease substantially over time, fully cultivated products may become more common, but hybrid products may still fill an important niche. Depending on how much cultivated meat producers can improve their environmental performance, hybrids may be preferred as the more sustainable option, at least for products where this does not negatively impact sensory performance. Some consumers might also prefer hybrids over purely cultivated products for sensory or nutritional reasons. In all cases, the goal is simultaneously to optimize for environmental sustainability, cost, sensory properties, and nutrition, with these factors being weighted differently depending on the product and the consumer. Cultivated, plant-based, and fermentation-derived materials can be considered ingredients that enable hybrid products to achieve this goal. For example, a small percentage of cultivated cells might play an outsized role in delivering a meaty flavor. In many cases, hybrids may offer the “best of both worlds” between cultivated and plant-based or fermentation-derived products.
A key question regarding hybrids is what percentage of the total product needs to be made up of cultivated animal cells—and what cell types are needed—to achieve the desired sensory properties. The lower this number turns out to be, the more potential there is for hybrid products to offer cost and environmental advantages over pure cultivated meat and to achieve price parity with conventional meat. The GOOD Meat chicken products mentioned above contain 70% chicken cells, and the 30% plant material is added to reduce costs and provide structure. Similarly, Bluu Seafood’s products are reported to contain fish cells as the main ingredient, with plant proteins added “to optimize cooking behavior and mouthfeel.”
However, lower percentages of cultivated cells may still yield a meaningful sensory benefit over plant-based products. SCiFi Foods claims that adding <20% cultivated beef cells to a primarily plant-based burger produces a major sensory enhancement. They plan to release their first product using a ratio in this range. Similarly, research by SuperMeat indicated that adding as little as 5% cultured hepatocytes to a plant-based meat product produced some sensory enhancement. For sausage and schnitzel, adding 13% hepatocytes was at least as good as adding 27%. In the case of ground beef, 27% hepatocytes—the highest percentage tested—produced the best sensory scores.
Companies are also targeting hybrids made from a combination of plant-based protein and cultivated fat, with the theory that a small percentage of fat may be more impactful in imparting flavor than an equivalent percentage of muscle or other cells. For example, Steakholder Foods (formerly MeaTech 3D), which recently acquired Peace of Meat, targets chicken fat for use in primarily plant-based products. Mission Barns follows a similar strategy for other species, starting with pork fat for bacon and sausages, as do Hoxton Farms and Cubiq Foods. As a rough upper limit on the percentage of such products that might be made up of animal cells, we can consider the fat content of some equivalent conventional products.
Hybrid approaches expand the “palette” of ingredients that can be used in creating food products using cultivated meat and other alternative protein technologies. The strategic use of hybrid strategies may help lower the barrier for cultivated meat products to enter the market, and in the long term may contribute to the development of products that are simultaneously tastier, healthier, and more sustainable.
Attributes & characterization methods
Flavor
Flavor is our means of assessing the chemical composition of a food product and is evolutionarily important as it helps us determine whether a food has valuable nutrients or is likely to be toxic. Flavor is also a crucial part of what makes eating an enjoyable experience. Although often perceived as a single sensation, flavor is made up of two distinct sensory modalities: taste and smell. Colloquially, when we ask, “How does the food taste?” we usually mean, “How is the flavor?”
From a scientific perspective, taste refers strictly to the sensation of water-soluble compounds sensed by the taste buds. Five “basic tastes” have been well-characterized scientifically: umami, salty, sweet, sour, and bitter. Across the animal kingdom, receptors for these tastes can vary in their number, distribution, and location within the body, but the tongue is the primary site for taste reception. The tongue also captures other sensations important to our experience of eating food, some of which—including the taste of fatty acids, discussed below—may genuinely be additional tastes themselves.
A great deal of what we perceive as flavor arises from our sense of smell, or olfaction, sensed by receptors within the nasal cavity. The chemicals sensed by smell are volatile compounds released by the food spontaneously or during chewing. Orthonasal olfaction—which is often what we mean when we colloquially refer to “smell”—refers to taking air directly in through the nose by sniffing or breathing and sensing the volatile compounds in the inhaled air. Equally important is retronasal olfaction, in which volatiles released from food during chewing make their way from the back of the mouth into the nasal cavity and are sensed there. The combination of orthonasal olfaction before taking a bite of food, retronasal olfaction during chewing, and taste combine to give the perception of flavor.
The taste of meat can be described generally as strongly umami and fatty—though the status of “fatty” as a taste is debatable, described later—and a bit salty, with only minor notes of sweet, sour, and bitter. A wide variety of volatile compounds contribute to the aroma of cooked meat, most of which are produced by either lipid degradation or the Maillard reaction (Sohail et al. 2022). The flavor of meat ultimately arises from the combination of the two senses.
Certain amino acids—primarily glutamate and, to a lesser extent, aspartate—and nucleotides—including guanosine monophosphate and inosine monophosphate—evoke the taste sensation known as umami, which is often described as “meaty” or “savory.” While generally not perceived as pleasant in isolation, umami contributes substantially to the overall perception of “deliciousness” in various foods, including meats and cheeses, when experienced in the proper context. The relationship between umami amino acid and nucleotide concentration and perceived umami taste can be modeled by a simple mathematical relationship and depends on both the synergistic effects of umami amino acids and nucleotides as well as the concentration of umami amino acids alone (Yamaguchi et al. 1971; Coleman et al. 2022). Several specific glutamate-containing peptide sequences from 2–8 amino acids have been identified as especially effective in enhancing umami flavor in beef, pork, chicken, and fish (Dashdorj et al. 2015).
Although fat is not usually listed among the “primary” tastes, there is evidence that fatty acids—which may be present directly in food or produced during chewing from the breakdown of triglycerides by lipases in saliva—are sensed as a distinct taste by the fatty acid transporter CD36 and the G protein-coupled receptor GPR120 (Reed and Xia 2015; Keast and Costanzo 2015). Running et al. (2015) asked participants to sort various tastants, including fatty acids, according to perceptual quality. They found evidence for fatty acids as a distinct taste, with sensory differences between fatty acids with different chain lengths. Participants generally described the taste sensations associated with isolated fatty acids as unpleasant. Triglycerides are thought to provide mainly a textural signal, but it is unclear whether they are also taste-active (Reed and Xia 2015). Fatty acids and their breakdown products are also sensed through the olfactory system (discussed later). While the smell and the taste of isolated fatty acids have been reported as unpleasant, the presence of fat in food has a generally positive effect on the overall sensory experience (Reed and Xia 2015). At least in mice, there is evidence that taste is involved in the attraction to fatty foods, as animals lacking CD36 were less likely to prefer such foods (Laugerette et al. 2005).
What does this all mean for cultivated meat? The complexity of the mechanisms by which humans sense dietary fat, and the fact that the presence of fat can be either positive or negative, means that optimization of products’ fat content is unlikely to be straightforward. However, naturally occurring fatty acid profiles will provide a useful starting point. The impact of increased or decreased content of any particular form of fat may not be easily predictable and will ultimately need to be assessed in the context of a whole food product. Because breakdown products are involved, the effects of fat and fatty acid composition on the sensory properties of cultivated meat will also depend on aging conditions and preparation methods.
Sodium chloride, potassium chloride, and other inorganic salts are responsible for the slightly salty taste of meat (Dashdorj et al. 2015). Additional salt may also be added during aging or curing, or before, during, or after cooking, making additional contributions to the final dish’s overall taste and potentially affecting texture or juiciness.
Meat also contains smaller amounts of sweet, sour, and bitter compounds, which play supporting roles in modulating the overall flavor. Although the sugar content of meat is relatively low, the small amount of sweetness it contributes is thought to contribute positively to the overall flavor (Dashdorj et al. 2015). In addition, sugars contribute to the pool of precursors for reactions such as the Maillard reaction and caramelization that generate volatile compounds essential to the flavor of cooked meat (Dashdorj et al. 2015). Similarly, the acid content of meat is low relative to other foods. Various acids—including lactic, succinic, acetic, and citric—may contribute a subtle sour taste on their own and interact with other tastes, such as umami (Dashdorj et al. 2015). Perhaps more importantly, the concentration of acids in meat can modulate pH-sensitive reactions during aging and cooking (discussed later). Bitterness, while not a primary taste characteristic of meat, is sometimes present due to certain minerals, nucleotides, and amino acids and is generally thought of as a negative attribute, at least when present in excess (Dashdorj et al. 2015).
The aroma of meat is characterized by a wide variety of volatile compounds, most of which are produced during cooking by the Maillard reaction or lipid degradation (Sohail et al. 2022). The Maillard reaction is a form of non-enzymatic browning in which amino acids or peptides react with reducing sugars to produce a wide variety of volatile aroma compounds as well as pigments that impart a brown color. It is a key element in the development of meat flavors and is also important for many other foods, especially toasted or seared ones. The volatile products of the Maillard reaction include sulfur-containing compounds and nitrogen- and oxygen-containing heterocyclic compounds, all of which may be odor-active (Sohail et al. 2022). Oxidation and degradation of lipids produce a variety of compounds, including aliphatic aldehydes, ketones, alcohols, acids, lactones, esters, and furans (Sohail et al. 2022). These compounds may have positive, negative, or no effects on the aroma of meat. The mixture of products from lipid oxidation varies depending on the reaction conditions and rate. Lipid auto-oxidation during storage generally contributes to the development of rancidity (Domínguez et al. 2019), whereas thermal oxidation produced by cooking produces more positive flavors (Sohail et al. 2022).
Sohail et al. (2022) used partial least squares discriminant analysis to identify volatile compounds that distinguished beef, pork, chicken, and sheep meat based on previously published data. They identified many odor-active compounds that were reliably higher in some species than others, but few of these compounds were present in only one species. For example, 12-methyltridecanal and 2-octanone, both of which produce meaty odors in isolation and were identified as highly significant in differentiating beef odor from other meats, were also present in some pork samples. Additionally, 4-ethyloctanoic acid and 4-methyloctanoic acid are well-known contributors to the odor of sheep meat and sheep and goat milk (Kaffarnik et al. 2014) but were not among the main distinguishing compounds in this analysis, and the former was also identified in pork and beef. Odor-active compounds in meat are not limited to those likely to be described as “meaty” in isolation. Aroma descriptors of the compounds identified by Fan et al. (2018) as necessary constituents of chicken broth flavor include everything from “meaty,” “earthy, roasted,” and “meat broth” to “citrus, green,” “cooked potato,” and “cucumber.” The combination of these aroma compounds—in the right proportions—gives rise to the full complexity of meat aroma.
Flavor is undoubtedly one of the key attributes that will determine whether cultivated meat products are ultimately acceptable to consumers. One potential hypothesis is that simply producing the correct cell type from the correct species at the correct maturation stage will achieve this by default. Alternatively, further optimization may be required. Cultivated satellite cells from chicken and cow contained lower amounts of umami, bitter, and sour compounds than conventional meat (Joo et al. 2022), highlighting areas where media, bioprocess, and end product optimization will be required to produce cultivated meat products with familiar flavors. On the other hand, Dr. Sandhya Sriram of Shiok Meats described her team’s experience of being surprised by the extent to which cultured shrimp cells inherently produced much of the desired flavor in a 2019 panel discussion. She described the flavor as consisting of umami, “seafood-y,” and sweet attributes, though she also claimed that there was a need to tweak the flavor by adding plant extracts and salt. Open-access research on the flavor of cultivated meat is thus far limited, and future studies will need to clarify the relationship between various process parameters and the quality of the end product.
The volatile compounds in a sample are often measured by separating the compounds using gas chromatography (GC), followed by either chemical measurement using mass spectrometry (GC-MS) or sensory detection using olfactometry (GC-O) (Starowicz 2021). GC-MS allows the precise concentration of the various volatiles in a sample to be quantified. In contrast, GC-O measures volatiles relative to their sensory thresholds and can include subjective descriptions of the odor evoked by each compound. Volatile profiles can also be measured using an electronic nose, consisting of an array of chemical sensors and a system for pattern recognition. Unlike GC-MS, an electronic nose does not produce a list of specific compounds and their concentrations. Still, it can be an efficient means of distinguishing samples on some metric of interest, e.g., spoilage. Electronic noses have been used successfully to measure a variety of meat and seafood samples, including cod (Olafsdottir et al. 2005), bigeye tuna (Sun et al. 2013), stinky mandarin fish (Li et al. 2013), and beef (Wijaya et al. 2017).
Once the composition of volatile compounds in a given sample is identified, the question remains of how these compounds contribute to the overall aroma. This can be answered by recombination experiments, in which compounds with sufficiently high odor activity values in the original sample—those present in concentrations greater than their odor threshold—are combined at the same concentrations as those measured in the sample. The recombined sample is then compared to the original using a triangle test or similar method to determine whether the odor has been successfully replicated. Following this, omission tests—in which one compound at a time is omitted, and the resulting mixture is compared to the control—can determine whether any compounds are dispensable despite their high odor activity values. These methods have been used to determine which volatile compounds are critical to the aroma of grilled mutton shashlik (Du et al. 2021) and chicken broth (Fan et al. 2018).
Instrumental analysis of taste-active compounds is somewhat more straightforward than that of aroma-active compounds, as a smaller number of chemicals is involved. Furthermore, these can be cleanly grouped into the five basic tastes, and within each category, the compounds evoking that taste are chemically somewhat similar. Electronic tongues can report the intensity of each of the five basic tastes and have widespread applications in food and other industries (Podrażka et al. 2017).
Few academic studies of cultivated meat have included instrumental analysis of flavor compounds, though this may be expected to change as the field moves beyond the initial proof of concept stage and into optimization of methods to improve organoleptic properties. An electronic tongue was recently used to assess the taste characteristics of a lab-scale cultivated meat prototype (Joo et al. 2022). Similarly, Lee et al. (2022) used both GC-MS and an electronic tongue to analyze the flavor characteristics of another prototype.
Texture
A food product’s texture describes how it responds to mechanical perturbations such as chewing. These responses are perceived through multiple senses and contribute heavily to our overall experience of the product. Texture is often quantified using instrumental tests, and the choice of test depends heavily on the product properties to be analyzed and the experimental goals. Standard tests used for meat products include Warner-Bratzler shear force (WBSF) and texture profile analysis (TPA), though a wide variety of other tests are also available (Schreuders et al. 2021).
Textural attributes such as tenderness depend on the product’s structure, including the ratios, properties, and spatial arrangement of muscle fibers, adipocytes, and connective tissue. In conventional meat, these structural attributes depend, in turn, on various pre- and postmortem variables (Listrat et al. 2016). While it is not yet clear exactly which factors will influence the texture of cultivated meat, we might expect that analogous pre- and post-harvest processes will exist. For example, while the texture of conventional meat depends in part on the animal’s diet and exercise level, which influence the amount and distribution of intramuscular fat, the texture of cultivated meat may depend on the ratio of muscle and fat cell precursors added to the bioreactor and the composition of the culture media. Post-harvest processes affecting texture may be more similar to those in conventional meat, though more research will be needed to understand the key similarities and differences.
Several studies have applied instrumental texture analysis methods to cultivated meat and other alternative proteins. For example, a comparison between lab-scale cultivated beef and rabbit prototypes, rabbit muscle, beef tenderloin, and ground beef revealed that the hardness of the cultivated meat prototypes was unchanged or increased after cooking. This response most closely resembled that of ground beef, which increased in hardness, as opposed to whole-muscle products, whose hardness decreased upon cooking (MacQueen et al. 2019). Interestingly, the same study revealed good alignment on a microscopic level. However, the cultivated meat prototypes lacked the densely packed tissue seen in the conventional meat/muscle controls. Improving the cell density of the final product could be an important area for optimization of the texture of whole-cut cultivated meat products. Other studies have compared cultivated meat products to their conventional counterparts using TPA (Ben-Arye et al. 2020; Paredes et al. 2022), single compression testing (Lee et al. 2022), rheology (Paredes et al. 2022), or by measuring the breaking force (Furuhashi et al. 2021). In addition, Zhang et al. (2021) used TPA to measure conventional Atlantic salmon’s hardness, cohesion, springiness, and chewiness before and after cooking to provide benchmarks for developing alternative seafood products.
Color
The appearance of food is often the first thing we experience about it and can substantially impact our willingness to try it and how much we ultimately enjoy it. The primary determinant of meat’s color is its myoglobin content (Listrat et al. 2016). As discussed previously, muscles vary in their content of slow-twitch and fast-twitch fibers. Because the highly oxidative slow-twitch fibers contain high levels of the oxygen-binding protein myoglobin—a distant relative of hemoglobin—they appear redder in color. The oxidation state of myoglobin in red meat alters its color substantially, with fresh meat appearing bright red due to oxygen-rich oxymyoglobin, whereas deoxymyoglobin appears purplish red. Denaturation of myoglobin during cooking leads to oxidation of its associated iron atom, which, together with Maillard reaction products, causes browning. Producing cultivated meat that is visually similar to its conventional counterpart will likely require either the addition of myoglobin—or similar compounds—or the identification of culture conditions that induce muscle cells to produce high levels of myoglobin, as they would in vivo. Indeed, small-scale cultivated meat prototypes grown with either hemoglobin or myoglobin added to the media produced colors similar to those found in cooked beef (Simsa et al. 2019). Additionally, the same study noted that the presence of myoglobin in the culture media improved bovine myosatellite cell proliferation.
Whereas most types of meat primarily derive their color from myoglobin, many seafood species are exceptions. As in terrestrial animals, the proportion of slow-twitch and fast-twitch fibers differs between fish species according to activity levels, with pelagic fish—such as tuna, which regularly cover long distances—having a higher proportion of slow-twitch fibers (Handbook of Seafood and Seafood Products Analysis). Unlike terrestrial animals, slow and fast muscle in fish show a strong spatial separation, with slow (red) muscle mainly confined to a small band just under the skin in most species and a small population of pink muscle fibers of intermediate location and properties (Listrat et al. 2016). Many species contain mostly white muscle with very little or no myoglobin content, but their meat is not always white in color. The distinctive “salmon” color in the white meat of salmon derives from carotenoids such as astaxanthin, which is derived from the fish’s diet and ultimately produced by microalgae. On the other hand, the small amount of red muscle appears gray or purplish. Astaxanthin also contributes to the color of other fish and shellfish, including shrimp. Incorporation of astaxanthin into the culture media for such species will likely be required to achieve the correct color and nutrition content.
While digital colorimeters accurately measure the color of a uniform surface and can be used with meat or seafood (Zhang et al. 2021), they are limited when it comes to capturing variation in color across an object. For this reason, some studies have opted to use images taken with a digital camera to measure the color of food and other objects (Yam and Papadakis 2004; León et al. 2006; Simsa et al. 2019). Instrumental measurements of color typically use the L*a*b* color space, in which L* represents an object’s lightness, a* represents its redness or greenness, and b* represents its blueness or yellowness. Relative to other color models, L*a*b* has the advantage of being consistent across measurement devices—although lighting and other environmental conditions still need to be standardized to make any comparisons meaningful—and close to perceptually uniform.
Subjective assessment of sensory properties
Various sensory analysis methods are regularly employed in the food industry and will be essential tools for cultivated meat researchers and companies to evaluate the success of prototypes and products. The choice of analysis method depends on the specific question a researcher aims to answer. Descriptive analysis attempts to quantify the perceptions of specific sensory attributes (for example, sweet, citrusy, or chewy) of a product without reference to whether the product as a whole is appealing or unappealing and is one of the most commonly used types of sensory testing (Djekic et al. 2021). Discrimination tests—a common example of which is the triangle test—ask whether samples are perceptually different from one another, again without reference to whether the samples are appealing or not. Affective, or hedonic, testing is concerned with whether the taster, who is usually untrained, likes the product, whereas descriptive analysis and discrimination tests are typically performed using trained sensory panelists (Djekic et al. 2021). Importantly, trained panelists are not very effective at predicting consumer preferences (Djekic et al. 2021), meaning that techniques like descriptive analysis, even if performed well, are not a substitute for hedonic tests with volunteers similar to the target consumer population.
In optimizing cultivated meat products, hedonic testing—whether people enjoy eating the products—is what ultimately matters. For products whose goal is to directly replace conventional meat, cultivated meat producers may also use discrimination tests to ensure that their products are not only delicious but also indistinguishable from conventional meat. Descriptive analysis can pinpoint specific differences between cultivated meat prototypes and conventional products. For example, suppose discrimination tests reveal that a prototype is easily distinguishable from conventional meat and hedonic tests reveal that most consumers prefer the conventional product. In that case, a cultivated meat producer might use descriptive analysis to understand the problem. If the descriptive analysis reveals that the cultivated product has lower levels of umami taste, optimization efforts could focus on amino acids, peptides, and nucleotides.
Sensory methods are also sometimes used to study a single aspect of the eating experience in isolation. For example, in one study, participants were asked to categorize liquid solutions of single taste-active molecules according to perceptual similarity to understand the mechanisms by which specific compounds are sensed (Running et al. 2015). A well-designed study avoids the confounds associated with other perceptual differences between samples—in this case, texture or color. However, such highly controlled experiments may be less relevant to perceptions of real food products, as the interplay between multiple sensory modalities is a typical feature of the eating experience. These cross-modal interactions may thus be considered a confound in mechanistic studies but an essential part of the subject under study in other contexts.
Some published academic studies where a cultivated meat prototype was produced included informal taste tests. For example, volunteers tasted a cultivated meat prototype produced by growing bovine cells on a textured soy protein scaffold and reported it to have “a pleasant meaty flavor and sensorial attributes, achieving a typical meat bite and texture” (Ben-Arye et al. 2020). A cultivated bovine fat prototype grown in an alginate hydrogel was reported to have “a creamy consistency typical of animal fat, with a discernible ‘beefy’ flavor” (Dohmen et al. 2022). While such observations reveal less about the characteristics and acceptability of these early prototypes compared to more in-depth sensory research, they are still valuable. For studies where the main goal is to create a proof of concept for a particular approach to producing cultivated meat—for example, a specific cell type or scaffolding material—general sensory evaluations may be sufficient. These provide a quick and easy way to determine to what extent one is “on the right track” in producing a prototype with general meat-like characteristics. As academic research into cultivated meat advances, studies aimed at fine-scale optimization of product properties will require more rigorous approaches to sensory evaluation of prototypes. Even for simple sensory evaluations, detailed descriptions of methods are essential to improve the reproducibility and interpretability of the resulting data. Reported methods should generally include panel demographics, details related to recruitment and training, and performance analysis (Djekic et al. 2021). A recent study of a hybrid chicken product containing cultivated fat produced by Believer Meats (Pasitka et al. 2022) included more formal sensory testing in the form of both hedonic and descriptive analyses performed by untrained volunteers.
Similarly, as cultivated meat companies move toward releasing consumer-facing products, detailed sensory evaluation will be an essential tool for ensuring that the taste, texture, and other properties match or exceed those of conventional meat. A blind taste test in early 2022 by SuperMeat produced encouraging results, with two of three food experts incorrectly identifying the cultivated chicken (served in minced form, without seasoning) as conventional and the third undecided.
Nutrition
In general, meat is a rich source of protein, though there is some variation in protein content between species and cuts. Fat content is substantially lower than protein and is somewhat more variable, but meat is still a significant source of fat and is often higher in saturated fat compared to other foods. Generally, marine-derived seafood contains more polyunsaturated fat compared to other meats.
| Product category | FoodData Central product description (ID) | Protein (g/100 g) | Total lipid (fat) (g/100 g) | Protein: Fat ratio | Fatty acids, total saturated (g/100 g) (% of total fatty acids) | Fatty acids, total monounsaturated (g/100 g) (% of total fatty acids) | Fatty acids, total polyunsaturated (g/100 g) (% of total fatty acids) |
|---|---|---|---|---|---|---|---|
| Beef | Beef, grass-fed, strip steaks, lean only, raw (169429) | 23.1 | 2.69 | 9:1 | 1.03 (48%) | 0.995 (47%) | 0.108 (5%) |
| Pork | Pork, fresh, loin, top loin (roasts), boneless, separable lean only, raw (168315) | 22.4 | 4.06 | 6:1 | 1.25 (38%) | 1.6 (49%) | 0.409 (13%) |
| Poultry | Chicken, broiler or fryers, breast, skinless, boneless, meat only, raw (171077) | 22.5 | 2.62 | 9:1 | 0.563 (34%) | 0.689 (41%) | 0.424 (25%) |
| Fish (lean) | Fish, tilapia, raw (175176) | 20.1 | 1.7 | 12:1 | 0.585 (40%) | 0.498 (34%) | 0.363 (25%) |
| Fish (fatty) | Fish, salmon, sockeye, raw (173691) | 22.2 | 4.69 | 5:1 | 0.814 (25%) | 1.37 (41%) | 1.12 (34%) |
| Crustacean | Crustaceans, shrimp, raw (175179) | 20.1 | 0.51 | 39:1 | 0.101 (30%) | 0.086 (25%) | 0.152 (45%) |
| Mollusc | Mollusks, octopus, common, raw (174218) | 14.9 | 1.04 | 14:1 | 0.227 (36%) | 0.162 (26%) | 0.239 (38%) |
It is reasonable to expect that cultivated animal cells will show roughly similar nutritional content to conventional meat from the same species. However, one study highlighted differences in amino acid composition between cultivated chicken and cow satellite cells and conventional meat, suggesting that media formulations and culture conditions should be optimized with nutrition in mind (Joo et al. 2022). While it is conceivable that simply inducing the cells to differentiate and mature will solve the issue, it will be necessary to verify this and other culture condition effects on nutrient profiles. Two separate studies of cultivated fat reported lipid profiles similar, but not identical, to fat from conventional meat (Dohmen et al. 2022; Yuen et al. 2022). Further optimization of media and culture conditions will allow for more precise control over amino acid and fat profiles, and may focus on either replicating the properties of conventional meat or optimizing directly for the desired taste and nutritional properties. Data submitted by UPSIDE Foods to the FDA showed a similar nutritional composition between the company’s cultivated chicken and conventional chicken (see pages 32, 78-82).
An important class of exception is that of compounds that are not produced directly by animal cell metabolism but are instead consumed through the animal’s diet. These include omega-3 fatty acids, essential amino acids, certain carotenoids, and vitamin B12. Because minerals such as iron and zinc are chemical elements, they cannot be synthesized by animals—or by any other living organism. Plants, animals, fungi, and other organisms ultimately take in minerals from the soil or their diet. Minerals may exist in different forms depending on their source—for example, much of the iron in animal tissues is complexed with heme, altering its absorption efficiency when consumed. Such minerals and nutrients will, therefore, need to be added to cultivated meat at some point in the production process if they are to be present in the final product. This can be accomplished by adding the relevant nutrients to the culture media (as is routinely done in cell culture), incorporating them into the scaffold, or adding them to the product in a final processing step. It is also possible to engineer cells to produce nutritionally important compounds that they usually would not. For example, bovine cells were engineered to produce carotenoids (Stout et al. 2020), and the same could be accomplished for other nutrients.
In addition to ensuring that the necessary nutrients are present in cultivated meat, it will be important to rigorously test their bioavailability, or how readily they are absorbed in usable form by the human body. For example, iron is more easily absorbed when complexed with heme. The presence of important vitamins, minerals, and nutrients is an important piece of the puzzle, but the other piece is the form those compounds take and any differences in their environment within the food product that might influence their absorption. Therefore, it will be necessary to rigorously test how well cultivated meat products are digested and their nutrients absorbed.
Post-harvest processes
Conventional meat undergoes a complex array of postmortem changes, including the partial breakdown of connective tissue, which leads to improved tenderness in aged meat (Listrat et al. 2016). Postmortem changes may be both positive and negative with respect to sensory quality and depend on factors including muscle pH, temperature, and glycogen stores. For example, the rate and extent of the decline in pH following slaughter is a critical determinant of the water-holding capacity of conventional meat, which in turn helps determine its perceived juiciness and overall acceptability (Listrat et al. 2016). Many of the same processes can be expected to occur in cultivated meat, and the added control over the relevant factors may lead to improved product quality and consistency and reduced food waste. Additional research is needed to understand how these post-harvest changes can be controlled through optimization of culture media formulations and pre- and post-harvest conditions. One post-harvest process with mainly adverse effects on the sensory quality of meat and seafood is the auto-oxidation of omega-3 fatty acids. Therefore, it will be helpful to develop improved methods for preventing the oxidation of omega-3 ingredients and final products.
Response to cooking and other treatments
One of the critical processes in the transformation of raw meat to cooked meat is the denaturation of the myofibrillar proteins, actin and myosin. Denaturation of myosin is largely responsible for the textural differences between raw and cooked meat. At higher temperatures, actin denatures as well, leading to moisture loss and toughening. At this stage, the meat is likely to seem tough and overcooked. Concurrently with the denaturation of myofibrillar proteins, fat dissolves, influencing the meat’s mouthfeel. Whereas the denaturation of actin is associated with toughening, an opposite effect results from the breakdown and gelatinization of collagen. Gelatinization occurs over a relatively wide temperature range and requires more time at elevated temperatures compared to the denaturation of myofibrillar proteins (Lawrie’s Meat Science, 7th ed., Ch. 10). This explains why meats that are slowly cooked to high final internal temperatures—especially those from collagen-rich muscles—can end up very tender even though actin has been thoroughly denatured. Collagen from fish, especially cold-water fish, breaks down easily at lower temperatures, contributing to the delicate, flaky texture of cooked fish (Listrat et al. 2016). In cooked fish, most collagen has been converted to gelatin, whereas collagen is often only partially broken down in cooked terrestrial meat.
(* Based on Lawrie’s Meat Science, 7th ed., Ch. 10, Heat and Its Effects on Muscle Fibers in Meat, Baldwin (2012), Zhang et al. (2021), and Schoenbeck et al. (2000))
As discussed previously, meat browning during cooking is due to a combination of myoglobin denaturation and the Maillard reaction. The former process has been shown to depend somewhat on meat’s pH and oxygenation state (Schoenbeck et al. 2000), which points to the need for control over post-harvest conditions in cultivated meat. The Maillard reaction is most efficient at higher temperatures (140–165 °C) than those reached in the center of a meat product during a typical cooking process (62–74°C), which accounts primarily for the deep brown color of the crust of seared meat.
A challenge for cultivated meat and seafood producers is creating products that respond to cooking and other preparation methods in the same way as conventional meat. Cultivated meat products with highly specific preparation requirements may succeed with trained chefs or when sold as consumer goods with specific cooking instructions. However, more versatile products can more reliably serve as drop-in replacements in existing recipes. When Wildtype held its first cultivated salmon tasting in mid-2019, the product could be served raw but was reported to fall apart when cooked. Two years later, in mid-2021, they seemed to have overcome this hurdle with a product grown on a plant-based scaffold capable of standing up to raw, baked, and cold-smoked preparations. BlueNalu demonstrated raw, cooked, and acid-cured preparations of their cultivated yellowtail—produced using a form of extrusion bioprinting—in late 2019.
From these anecdotes, we can conclude that it is not certain that cultivated meat products will respond appropriately to various cooking methods. Therefore, this is a potential area that some producers will need to optimize. However, it also seems that this may be a relatively straightforward challenge. Cuts of conventional beef differ in their response to cooking based on the amount of collagen and level of collagen cross-linking within the meat (Lawrie’s Meat Science, 7th ed., Ch. 10). The same can be expected to be true for cultivated meat, so manipulations that alter the characteristics of collagen and other ECM proteins secreted by the cultivated cells can be expected to alter responses to cooking. If substantial amounts of scaffold material remain in the final product, these materials’ thermal properties may also influence responses to cooking.
White spaces in end product characterization
Fortunately for cultivated meat researchers aiming to produce delicious products, a substantial body of knowledge on the characteristics of meat already exists thanks to previous work by meat science researchers. Most of what needs to be done now is applying that knowledge to answer the question of how to recapitulate the desirable characteristics of meat using more sustainable methods. However, some gaps related to characterization of cultivated meat remain.
Better standardization of characterization methods within the cultivated meat field may help improve the reproducibility of data across labs. Because producing cultivated meat at scale remains challenging, researchers in academia and industry are limited in how rapidly they can iterate on methods for producing high-quality cultivated meat, given that creating large pieces is quite resource intensive. While this can be expected to get somewhat easier over time, improved methods for the characterization of small samples could help accelerate the process of optimizing production methods to improve product quality. Such methods would speed up the prototyping process by allowing researchers to test many combinations of cell types, media formulations, scaffold materials, and other parameters while collecting meaningful data useful for predicting flavor, texture, nutritional properties, and other product attributes.
There is also a need to collect more data to guide product development. This is especially true for seafood, given the lack of existing data and the diversity of food-relevant species. Open-access data on the properties of conventional meat and seafood in both raw and cooked states will provide a helpful baseline, allowing for optimization efforts by cultivated meat companies to proceed more rapidly and efficiently (Zhang et al. 2021). A project by researchers at EMBRAPA, funded by GFI, aims to collect data on the characteristics of several species of conventional seafood.
Regulation & safety
From a food safety perspective, cultivated meat may provide advantages over conventional meat in several areas. The use of antibiotics in conventional animal agriculture increases the chances of antimicrobial-resistant pathogens emerging. Cultivated meat, on the other hand, is likely to be produced without antibiotics and thus has the potential to reduce this risk. Documentation submitted by UPSIDE Foods to the FDA confirms (see page 40) that antibiotics and antifungals are used only during cell line development and not in the production stage. Cultivated meat may also reduce the risk of foodborne illness by avoiding the introduction of potentially pathogenic bacteria during and following slaughter. UPSIDE Foods observed (see Table 5.51, page 34) vastly reduced growth of bacteria and other microorganisms on their cultivated meat in comparison to samples of conventional ground chicken. This reduced microbial load could have implications for food safety and shelf life (though it may also come with risks, as discussed below). Finally, cultivated meat is likely to contain much lower levels of microplastics and other environmental contaminants that are ubiquitous in conventional meat, especially seafood. At this point, these anticipated advantages are based on data about food safety risks in conventional meat, knowledge of the general production process for cultivated meat, and the recently-published data submitted by UPSIDE Foods to the FDA as part of the regulatory process. Confirming these assumptions via open-access research and additional public disclosures of data by companies via regulatory bodies will be helpful for regulators and the public to assess the food safety implications of cultivated meat.
Regulators will also need to assess any food safety risks posed by the production process for cultivated meat. A recent educational video released by the FAO in partnership with Aleph Farms provides examples of some of the food safety controls companies are likely to employ in their production processes. Regulatory agencies in several countries are developing frameworks that will determine the regulatory path to market for cultivated meat. In December 2020, Singapore became the first country where cultivated meat could legally be sold, though as of early 2023, it remains the case that approval is required for each product. In the U.S., FDA and USDA have announced they will jointly regulate cultivated terrestrial meat and catfish (in case you’re curious, here’s the story about catfish). In contrast, FDA will have sole jurisdiction over all other cultivated seafood. In November 2022, FDA completed its first pre-market consultation with just one company, UPSIDE Foods. For UPSIDE’s products to be approved for sale, they will still need a grant of inspection from the USDA Food Safety and Inspection service (FSIS) for their facility. While this milestone provided valuable insights into the U.S. regulatory process, some details are yet to be determined.
A recent paper described likely sources of food safety hazards and proposed some general process diagrams for the cultivated meat industry based on consultation with cultivated meat companies and food safety experts (Ong et al. 2021). The authors concluded that most of the likely safety concerns in cultivated meat could be reasonably assessed using existing frameworks from other industries, though differences may exist. An ongoing project by New Harvest research fellow Sam Peabody aims to better understand the potential for bacterial growth in cultivated meat to ensure that proper food safety controls can be developed. Interestingly, it has been proposed that if harmful bacteria are present in cultivated meat, their growth might be aided by its low bacterial load (Lawton 2020). Therefore, it might be necessary to introduce a “microbiome” to cultivated meat as it is harvested, which would have the additional benefit of allowing for control of which benign microorganisms are introduced and at what level.
Consumer research
Successfully producing a cultivated meat product is worth very little if no one wants to buy and eat it. Therefore, cultivated meat producers must understand who they are producing products for and what their customers want. Consumer research has revealed a great deal about how various groups view the idea of cultivated meat, what factors predict acceptance or rejection, and what product attributes are most important.
Are people interested in cultivated meat?
Studies have shown substantial variation in the proportion of survey respondents who described themselves as willing or unwilling to try cultivated meat (Post et al. 2020). These results may reflect differences in the cultural background, food preferences, or prior knowledge about cultivated meat among the different populations surveyed. They may also reflect the design of the study and how the survey questions were worded. Another clear trend is that the proportion of people who indicate they are unsure whether or not they would try cultivated meat is often considerable. It is perhaps unsurprising that many people are undecided about a new category of food that is not yet on the market in most countries. To win over those currently undecided, it will be incumbent upon the companies in this emerging industry to demonstrate they are capable of making delicious products and are committed to doing so ethically, sustainably, and transparently.
* Based in part on Table 4 from Post et al. (2020); original data from YouGov 2013, Flycatcher 2017, Pew Research 2014, The Grocer 2017, Wilks and Phillips (2017), Surveygoo 2018, Bryant et al. (2019), Szejda et al. (2021a), Szejda et al. (2021b), Szejda et al. (2022).
Who is interested in cultivated meat?
A relatively consistent finding is that prior familiarity with the concept of cultivated meat is often a predictor of positive attitudes toward it or willingness to try it. For example, Baumann and Bryant (2019) found that familiarity was a significant positive predictor of purchase intent among U.S. consumers. Similarly, Bryant et al. (2019) found that familiarity was positively associated with purchase intent in the U.S., China, and India. Children and adolescents in Germany who were more familiar with cultivated meat prior to the study showed a higher willingness to eat a cultivated meat burger and more positive attitudes toward both the cultivated burger and cultivated meat generally (Dupont and Fiebelkorn 2020). Among Italian consumers who were previously familiar with cultivated meat, Mancini et al. (2019) found more positive attitudes toward cultivated meat, a higher willingness to try cultivated meat, and a higher willingness to pay for cultivated meat. Similarly, Australian consumers familiar with cultivated meat showed a higher willingness to try it (de Oliveira Padilha et al. 2022).
A variety of demographic characteristics may predict attitudes toward cultivated meat. However, which characteristics are predictive—and in some cases, the direction of the relationship—has been found to differ between countries (Bryant et al. 2019). This suggests that it will be necessary for companies to consider cultural context when attempting to understand their primary target market in a given country. Notably, in the U.S., meat eaters reported more interest in purchasing cultivated meat than non-meat-eaters (Bryant et al. 2019), indicating the potential for cultivated meat to displace conventional meat purchases.
What do people want in a cultivated meat product?
Consumer research can also answer questions about what characteristics are the most critical determinants of cultivated meat products’ success. Both sensory properties (e.g., taste, texture) and price have been identified by consumer research studies as important factors influencing consumer acceptance of cultivated meat (Pakseresht et al. 2021; Grasso et al. 2019; Verbeke et al. 2015a; Gere et al. 2020; Ruzgys and Pickering 2020; Gómez-Luciano et al. 2019; Tucker 2014). Participants in some studies have expressed skepticism that cultivated meat would meet their sensory expectations (Pakseresht et al. 2021; Ruzgys and Pickering 2020; Wilks and Phillips 2017; Verbeke et al. 2015b; GFI & Kelton Global 2021). Such expectations may represent a barrier that would prevent some people from initially trying cultivated meat. Still, such consumers may update their expectations after trying cultivated meat or hearing positive reports from others who have tried it. Thus, if early cultivated meat products are successful from a sensory perspective, consumer skepticism in this area would represent a barrier to initial trials of cultivated meat but not repeat purchases.
A 2021 survey of potential alternative seafood consumers by Kelton Global and GFI revealed that good flavor and good texture were among the highest priorities when evaluating the appeal of cultivated seafood, with 78% and 72% of respondents, respectively, ranking these as somewhat or extremely important. Similarly, the top two reported barriers to choosing cultivated seafood were that consumers anticipated disliking the taste or texture. Consistent with this, a Total Unduplicated Reach and Frequency (TURF) analysis performed as part of the same study revealed that 78% of respondents found a hypothetical cultivated seafood product appealing if it had good flavor. Environmental, health, and additional organoleptic benefits had the potential to expand this reach.
It is worth noting that substantial proportions of the survey respondents indicated that environmental and health-related concerns were important to them. For example, 70% of respondents listed reducing plastic waste in the ocean as somewhat or extremely important. We can conclude from the TURF analysis that cultivated seafood products will need to be tasty to succeed in becoming more than a niche product. However, offering health and sustainability benefits—and messaging these benefits appropriately—will also be necessary for companies to expand their reach and differentiate their products from conventional products.
Much more consumer research into cultivated meat has been published in addition to that discussed here. Dive further into the literature using the list below.
Learn more about cultivated meat consumer research
Opportunities to improve consumer-facing properties relative to conventional meat
While cultivated meat has the potential to address the negative externalities associated with conventional meat production—in areas including climate, antimicrobial resistance, zoonotic disease risk, animal welfare, and global food security—it is worth delineating the benefits to the individual that a switch to cultivated meat might afford. Besides being less harmful to society and the planet, might cultivated meat also have the potential to be an inherently better product?
Food safety and shelf life
As mentioned above, cultivated meat may have some food safety advantages over its conventional counterpart. On the individual consumer level, the main advantages are likely to be a lower risk of foodborne illness, longer shelf life, and lower levels of contaminants such as mercury and microplastics. The results of GFI’s alternative seafood consumer research study support the hypothesis that these might be desirable attributes for many consumers. For example, 73% of respondents listed no chance of food poisoning and 57% listed longer shelf life as somewhat or extremely important factors that might lead them to select cultivated over conventional seafood.
Product ratios and species
The carcass balancing problem constrains the ratios of different products and co-products from a given animal species. Humans are a culinarily inventive species, and various cultures have found their own ways of using different parts of an animal based on their anatomical ratios. Even so, this represents a constraint on our food choices. There are also constraints related to specific species. For example, while bluefin tuna aquaculture has made some recent strides, it remains notoriously difficult to farm relative to other tuna species.
Our perception of certain foods can change dramatically based on their abundance or scarcity. For example, early European settlers in North America had access to a copious supply of lobsters and considered them a less desirable protein. Similarly, in the early nineteenth century, caviar (sturgeon eggs) was given out for free in North American saloons because the saltiness encouraged patrons to drink more. As populations of both lobster and sturgeon declined, both foods gained a new reputation as higher-end options.
Thus, the array of meat products available in the grocery store or on a restaurant menu—and our perception of these foods—is not dictated simply by what is inherently delicious. It also depends heavily on which species are easy to farm or catch and the relative sizes of their various muscles, organs, and tissues. Cultivated meat allows us to apply our culinary creativity in a world where product ratios are not constrained by the carcass balancing problem and species or breeds previously impractical as large-scale food sources may become practical. A few cultivated meat companies, including Vow and Orbillion, are already focusing on cultivated meat from exotic species and heritage breeds.
Optimizing nutrition and taste
Even while sticking to familiar product categories, there are numerous opportunities for cultivated meat producers to tweak their products’ properties to meet their customers’ wants and needs even better. For example, cultivated steaks could be engineered to contain more heat-stable actin or less heat-stable myosin isoforms to widen the “window” in which the meat is perfectly cooked. This could make it easier to achieve a perfectly cooked steak every time, even for home cooks with limited culinary experience. Further tuning the temperature sensitivity of fats, collagen, and myoglobin could optimize the cooking properties of meat products to cater to different preferences. Product composition could further be tuned with an eye toward taste and nutrition, such as by increasing the concentration of volatile compounds known to contribute to positive perceptions of flavor, reducing saturated fat content for those seeking to limit it for health-related reasons, or increasing the levels of compounds such as omega-3s and carotenoids thought to have health benefits.
Indeed, a substantial proportion of those surveyed in GFI’s alternative seafood consumer research study responded positively to the idea of certain product attributes that could be considered an improvement over conventional seafood. Respondents indicated that no bones (62%), less or no “fishy” smell (62%), and no skin (50%) would be somewhat or extremely important factors that might convince them to choose cultivated seafood over conventional. Therefore, products that successfully replicate the attributes of conventional seafood that people like while improving on what people dislike may find commercial success.
Interestingly, in FDA’s response memo to UPSIDE Foods in which it gave the green light to the company’s cultivated chicken, a distinction is made between components “used to support primary metabolism in cell culture” versus “inappropriate or indiscriminate food fortification.” The data submitted by UPSIDE indicated that several compounds, including iron and folate, were found at higher levels in their product than in conventional chicken. Because these compounds were added to the culture medium at levels intended to support the metabolism of the cells, FDA simply evaluated whether these elevated nutrient levels posed a safety concern at these levels observed in the final product. It will be important for cultivated meat manufacturers who intend to deliberately fortify their products to pay close attention to existing FDA policies and guidance on this topic, and to avoid “indiscriminate addition of nutrients to foods.”
Novel products
Cultivated meat also offers the opportunity to completely rethink what meat is and to produce products that prioritize deliciousness over familiarity. What would result if our conception of meat could be decoupled from how it occurs in an animal? Cultivated meat could be produced in novel forms or from a mixture of cells from multiple animal species. The Australian startup Vow is already pursuing this latter strategy. UPSIDE Foods is going to market with a more traditional strategy, but indicated in a recent Reddit post that they were also open to pursuing products that don’t exist yet. In response, commenters suggested a staggering variety of innovative ideas, though there was also enthusiasm for the more familiar options, and everything in between!
Once the technical challenges associated with cultivated meat have been overcome, cultivated meat will be limited only by our culinary creativity and appetites. If existing cultivated meat startups are any indication, companies will likely pursue strategies ranging from strictly focusing on familiar drop-in replacements to creating entirely new product forms. This diversity of strategies will allow the cultivated meat industry to meet consumers’ desires for delicious food while minimizing externalized costs.
View references featured in end products
Header image courtesy of Wildtype.
