Thursday, 10 September 2026

Petfood Palatability: The Physical Signals That Drive Preference

THE PHYSICS OF PETFOOD PALATABILITY

The same recipe, in a different physical state, is a different food to the animal. How much of palatability was never about the formula at all?

Take a single cut of beef, the same animal, the same muscle, and put it into four different kitchens.

In the first it is thrown onto a grill and left to char, and it comes off blackened at the edges, smoky, its surface a crust of browned reactions, eaten hot with the hands. In the second it is cut small and braised for hours until it collapses, and it arrives soft and yielding in a dark paprika gravy, the meat and the sauce by now almost one thing. In the third it is simmered gently and served in a clear rich liquor, tender pieces in a broth that carries its own separate warmth and smell. And in the fourth it is ground, salted, seeded with bacteria, stuffed and hung for months in a cool cellar until it is dense and dry and sharp, and then it is sliced thin and eaten cold.

A barbecue. A gulasch. Meat in gravy. A salami. The same animal stands behind all four. Yet no one who has eaten would call them the same food. No one wants the salami hot from a grill or the barbecue cold from a fridge in thin wet slices. We do not merely accept that these are different dinners. We feel it, in the mouth, without being taught.

So here is a question worth carrying into the rest of this essay, because the petfood industry answers it every day without quite noticing. If the same beef can become four foods that eat nothing alike, how much of what we call palatability was ever really about the recipe at all, and how much was about the state the food arrived in?

Movement I

THE OBJECT IN THE BOWL

Now carry that cow into the pet food aisle, because the same thing has happened there, and we have been much slower to admit it.

The industry does not sell one wet food. It sells a chunk in gravy, and a chunk in jelly, and a pâté, and a mousse, and a loaf, and alongside them a dry kibble that shares more with a biscuit than with any of the wet formats. These are not simply five recipes. They may share the same nutritional intention, the same protein source, the same flavour direction, and yet become radically different physical objects that an animal meets in radically different ways. The chunk must be picked up and broken. The pâté yields at once. The gravy carries smell and moisture separately from the solid it surrounds, so that a cat can, and often does, drink the sauce and leave the meat, or the reverse. The kibble is hard and dry and can deliver a disproportionate part of its immediate palatability signal from a surface rather than from its body. To the formulator these may be one product line in several presentations. To the animal they are several foods.

And this is the quiet assumption the whole category rests on, that if we get the formula right the food will follow. But the animal never encounters the formula. It encounters an object, with a temperature, a texture, a moisture, a shape in the mouth, and a smell that the physical state either locks away or sets free. The recipe is what we wrote down. The object is what the animal actually met. We have spent a great deal of this series arguing about the first, and comparatively little about the second, and the animal has been judging the second all along.

None of this is a failure of formulation. It is simply a part of palatability that sits in the physics rather than the chemistry, and that has never had the attention the chemistry receives. So the question is not whether recipe matters. Of course it does. The question is the one we rarely put to ourselves plainly. How much of the animal's verdict was shaped, from the first encounter, by the plain physical facts of how the food arrived?

The recipe is what we wrote down. The object is what the animal met.

Movement II

THE TEMPERATURE OF THE MEAL

Start with the lever that costs almost nothing and is discussed almost never. Temperature. Not the heat of cooking, which the animal never sees, but the plain warmth or cold of the food in the moment it is set down.

Anyone who has kept a cat has watched the small drama. A pouch or a can straight from the fridge is opened and spooned out, and the cat approaches, lowers its head, and declines, and the food sits congealing while the animal looks at its owner with what feels like reproach. The same food, left to reach room warmth, or briefly warmed, is eaten. Nothing in the recipe changed between the refusal and the meal. Only the temperature did. And the question that ought to follow is the one the industry rarely asks out loud. If the same food can be refused cold and eaten warm, what exactly was the cold food failing to do?

One important part of the answer, as far as the science can currently show it, is that the colder food was giving the nose less to work with. When aging cats were offered the same chunks-in-gravy product at three temperatures, chilled to 6 degrees, at room temperature near 21 degrees, and warmed to 37 degrees, they preferred the warmer food in every comparison, and the warmest most of all. What makes the study worth dwelling on is what it ruled out. The gravy's thickness barely changed across the three temperatures, so the difference was not that the cold food was harder to eat. What changed was the air above the bowl. Warming released markedly more of the volatile compounds that carry smell, including the sulfur notes associated with cooked meat and an acid tied to palatability, while some of the plant-like volatiles fell away. The cold food was not a different recipe. It was the same recipe with its aroma still locked inside it.

For a cat this may matter more than it would for us, because olfaction appears to be especially important in how a cat first investigates and accepts a food. Much of whether it approaches, and often whether it eats at all, seems to be settled by what reaches it through the air before a whisker touches the food. Warm the food and you set more of that aroma free; chill it and you hold much of it back, muffling the very signal the animal leans on for its first decision. The food that a cat meets cold is, in a real sense, a quieter food, and a quiet food asking a scent-led animal to commit is asking a great deal.

It is tempting to see an evolutionary fit here, since freshly killed prey is warm and not refrigerated, and a warmed food sits nearer that temperature than a chilled one. Whether a cat actually reads the warmth itself as part of a fresh-prey signal, rather than simply enjoying the fuller aroma that warmth releases, is a harder question than the tidy story admits, and one the evidence has not settled. What is clear is the direction, not yet the full reason.

And there is a deeper principle underneath the single study, one that belongs to physics rather than to cats. The molecules that carry smell are volatile, which is only to say that they leave the food and enter the air more readily as the food warms. Aroma release is temperature-dependent by its nature, and in humans a good deal of what we loosely call taste is really smell arriving by the back of the throat as we chew. How large that retronasal contribution is for a cat is not something we should assume from our own mouths, but the underlying physics, that warmth lifts volatiles from food, does not care whose nose is downwind. This is not a quirk of one product or one panel of elderly cats. It is true of a stew on a human stove and true of a chunk in gravy in a bowl, and it means the temperature of a food is never neutral. It is always either releasing the signal or holding it back.

Which raises something the industry knows and rarely says plainly. The animal does not choose the temperature. We do, or the room does. A pet has no fridge and no microwave and no sense of when the pouch was opened. It meets the food at whatever warmth we happen to have left it, and that warmth is doing quiet work on the aroma either way. Nor is warmth a free lever, because the same heat that frees the smell also hurries the spoiling. A wet food left out to reach an appetising warmth is also a wet food warming toward the temperatures at which it turns, and the window in which a bowl is both fragrant and safe is narrower than we like to admit. Temperature, in other words, is not a small convenience at the edge of palatability. It is a variable we control on the animal's behalf, it governs both the smell and the safety of the meal at once, and the animal lives entirely at the mercy of the choice we make for it.

We did not change the recipe. We changed how loudly it could speak.

There is a fairness in noting that the effect has been studied particularly carefully in older cats, a population in which age-related changes to smell and taste may make food presentation especially consequential. But the finding points somewhere uncomfortable for how we test and how we feed. A palatability trial run at one temperature has measured the food at that temperature and nowhere else. A diet judged wanting in the bowl may have been judged wanting cold. How much of what we have recorded as a preference for a food was, without our noticing, a preference for a temperature?

Movement III

HOW MUCH WATER

Temperature governs how loudly a food can speak. Water changes the language in which it speaks, and it does so along a spectrum most owners never think about even as they choose a point on it in the aisle.

At one end sits the dry food, a kibble holding less than about a tenth of its weight as water, closer in physical character to a biscuit than to anything an animal would meet in the wild. At the other end sits the wet food, most of its weight water, often three-quarters or more, soft and yielding and packaged in a can or a pouch because it cannot be stored any other way. And between them, less discussed and quietly interesting, sits the semi-moist food, somewhere around a fifth to a third water, soft enough to yield like meat yet stable enough to live on a shelf. Three foods, one nutritional intention, and three completely different physical propositions for the animal that has to eat them.

The industry has long observed that animals often take more readily to wet food than to dry, with semi-moist somewhere between, though the ranking is a tendency rather than a law and turns on the species, the individual, the product and the test. What is more certain, and more interesting, is that a good part of whatever difference exists is physics rather than flavour. Water is the medium in which taste-active substances dissolve and reach the tongue, so a wet food presents much of its flavour already in solution, ready to be perceived the moment it is in the mouth. Aroma is a subtler matter, because how much scent a food releases into the air depends not only on its water but on its fat, its proteins, its matrix and the particular volatile in question, and more water does not simply mean more smell. What water changes most is not whether a food has aroma but how the food organises and releases its sensory chemistry. A dry kibble is not silent; it carries scent in its headspace, which is part of why the outside of a kibble is where so much of its immediate palatability is deliberately placed. But a dry kibble carries relatively little flavour in solution until saliva begins to rehydrate it in the mouth, so a large part of its first, decisive impression has to be loaded onto its surface rather than built through its body.

That distinction, between sensory chemistry distributed through a food and sensory chemistry concentrated on its surface, is one of the most consequential facts in the category, and it follows directly from moisture. In a wet food, sensory chemistry can be distributed through the matrix. In a dry food, much of the immediate palatability signal is deliberately concentrated at the surface, so the animal meets that layer first, in the first second, dry against the tongue. The two formats are not louder and quieter versions of one food. They are two different architectures of flavour, and moisture is a large part of what decides which architecture you are building.

The flavour of a wet food is largely in the food. Much of the immediate flavour of a dry food is deliberately put on it.

The semi-moist food is the most ingenious point on the spectrum, and the reason repays a moment. Its appeal is softness and higher moisture, closer to the mouthfeel of meat than a hard kibble can manage, and yet it survives on a shelf without a can. The key is that the important number is not simply how much water a food contains, but how much of that water is free to participate in chemistry and support microbial growth, which is measured as water activity. A food can hold a good deal of water and still keep well if enough of that water is bound rather than free. Water activity is not the whole of the trick, because shelf stability also leans on solutes, acidity, preservation and packaging together. Salt is one of those solutes, and it is worth naming here because it is quietly doing physical work long before anyone asks whether it can be tasted. Dissolved in the food, salt lowers the activity of the water present, making less of it available to the microorganisms that would otherwise exploit it, which is part of why salting has preserved meat for as long as people have wanted meat to last. In a pet food, salt is one of the levers that helps hold a soft, moist texture at a moisture that would otherwise spoil. Its first job, in other words, is not flavour at all. It is physics. Water activity, then, sits at the heart of the compromise, and it lets a semi-moist food strike a clever balance, wet enough in the mouth to please the animal, stable enough in the microbial sense to keep. It shows how far the physical form alone, quite apart from the recipe, can be engineered to change what the animal experiences.

So the question that closes this lever is not which format is best, because that depends on the animal, the owner, the budget and the shelf. The question is subtler and more uncomfortable. When an animal prefers the wet food to the dry, as animals so often do, how much of that was a verdict on the flavour we formulated, and how much was a verdict on water itself, changing the way that sensory chemistry was organised, released and encountered?

Movement IV

THE SHAPE OF THE MOUTHFUL

There is a soft, semi-moist dog food, strongly aromatic and easy to mould, that anglers have quietly used for years as fishing bait. It is not hard to see why. It holds together on a hook, it releases its scent into the water, and it can be pressed into whatever shape the moment requires. Not one of those virtues has anything to do with the dog it was formulated for. They are pure physics, the properties of the object rather than the recipe, and they are compelling enough to work on a fish. A food designed for one species, doing some of its most persuasive work underwater, on another that it was never meant to feed. If the physical form of a food can do that, we should take seriously what it is doing inside the mouth it was actually built for.

Because the animal does not swallow a formula. It takes hold of an object, of a particular size and hardness and shape, and its mouth has to do something with it, and what the mouth can comfortably do differs between our two species. Dogs tend to eat quickly and bolt much of what they are given. Cats tend to take smaller bites and spend longer over a meal, and their jaws do not move sideways to grind; their teeth are built to shear and tear. That anatomy is part of why texture, size and shape appear to weigh more heavily for cats than for dogs, and why a cat is quicker to reject a piece that does not sit right in the mouth. For the cat, texture is often not a finishing touch on palatability. It can be a gate the food has to pass through before taste is ever consulted.

The consequences are surprisingly specific. How a kibble behaves in the mouth is a question of food mechanics, of its hardness and brittleness and the force needed to fracture it, and those depend on many things at once, on density and porosity and moisture and the way it was made, as well as on its size and shape. A cat that must break every piece with shearing teeth is sensitive to the result. Shape matters too, and not decoratively; a kibble with sharp edges can be awkward to grasp and uncomfortable to bite, and there are indications that cats favour rounded, easily handled forms, some so specifically that flat-faced breeds are offered kibbles shaped to be caught more easily by a short muzzle. Lower-density pieces that fracture readily tend to be appreciated. None of these are matters of flavour. They are matters of how the object meets the mouth, and where the fit is wrong the flavour may never get its hearing.

For the cat, texture is not the finish on palatability. It is the gate it has to pass first.

Salt belongs in this movement too, and not for the reason a human cook would expect. On our side of the bowl, salt earns its place partly by changing the food as a physical material, drawing out and dissolving proteins, altering how a mixture binds and sets and holds together, shaping the very texture we have been talking about. But cross to the animal's side and the familiar human logic loosens. Experimental work suggests that cats do not show the straightforward attraction to added salt that humans do, and that sufficiently high concentrations become aversive to them, so that even a sodium-depleted kitten will not reliably choose a salted diet over an unsalted one. One plausible explanation lies in their carnivorous history. An animal whose ancestral diet supplied its sodium through animal tissue may have faced a very different selection pressure from omnivores repeatedly meeting sodium scarcity, though whether that fully explains the feline response is harder to establish. What matters here is the practical point. Salt can be technologically important to the food, doing real work on its water and its texture, without functioning as the simple flavour reward a human cook would assume, which is one more sign of how far the animal's palate has diverged from the one we keep, without noticing, projecting onto it.

We salt our food for pleasure. For the cat, salt plays by different rules.

Wet food is where physical form becomes almost a language of its own, because here the same nutritional intention can be presented as a smooth pâté that yields at once, or a firm loaf that must be broken, or discrete chunks suspended in a gravy, or the same chunks set in a jelly, or a whipped mousse. These are not decorative distinctions. A pâté asks nothing of the teeth and gives its flavour immediately. A chunk asks to be seized and sheared and holds its flavour in a body the animal has to work into. And the chunk formats do something stranger still, because the chunk and its surrounding gravy or jelly are, in effect, two foods in one bowl, a solid and a sauce with their own separate textures and their own separate loads of smell and flavour. Watch a cat lick every trace of gravy from around the chunks and walk away, or clean up the chunks and leave the jelly congealing at the edges, and you are watching something precise. On the label there was one product. In the formulation system there was one recipe. In the bowl, the animal ran its own separation and chose. The manufacturer made one food. The cat found two.

The manufacturer made one food. The cat found two.

So the uncomfortable question that this lever leaves is the same one the others keep raising, in a new costume. When a cat turns from a food, we are inclined to record that it did not like the taste. But how often had the taste even been reached? How often was the refusal a verdict passed at the lips, on a piece too large or too hard or too sharp or too oddly shaped to be worth the trouble, delivered long before a single flavour molecule was ever given the chance to make its case?

Movement V

THE ACID EDGE

The last of the physical levers is the quietest, and the one that sits closest to the border between physics and chemistry, which is why it is worth handling carefully. The acidity of a food, its pH, is not a flavour in the way that a savoury peptide is a flavour. It is a condition of the whole food, a property of its water and its salts, and it shapes how the food is perceived without being an ingredient the animal could point to.

Cats, the palatability literature suggests, do care about it. Given a choice they tend to prefer a mildly acidic food over a neutral or an alkaline one, and acidulants are used in cat foods partly for that reason. One reading of why is evolutionary, that a carnivore eating fresh meat and organs is eating mildly acidic material, so a food in that range reads as right. That is a plausible interpretation rather than a settled fact, but the preference itself is real enough to matter to anyone formulating for cats. So far this looks like one more thing we could simply add and be done with.

But the lever has a hinge in it, and the hinge is the interesting part. The same cat that is drawn to a mild acidity will refuse a strong sourness. A sharply sour taste is aversive to cats, as it is to many animals, one of the signals a cautious eater treats warily. It is tempting to tie this neatly to spoilage, to say that sourness means rot, but the chemistry of decay is not that tidy. Spoiling food can turn more acidic when certain bacteria produce acids, and it can turn more alkaline when others release amines and ammonia from breaking-down protein, and it throws off sulfur compounds and other volatiles besides. Rising sourness is one possible sign among several, not a clean dial from fresh to foul. What we can say more safely is narrower and still useful. Cats like a little acid and reject a lot of it, so pH is not a setting you can push in one direction for more reward. It is a window, and both of its edges matter.

There is, separately, a real fact about acid and keeping. Lowering a food's pH is one of the oldest ways of preserving it, because many of the microbes that spoil food grow poorly as acidity rises, which is why acidified and fermented foods keep where neutral ones do not. So on the food's side, acidity is partly a preservation tool. Whether the animal's own taste for mild acidity evolved as a freshness cue, or for some other reason, or as a mixture of causes, is less certain than the neat story would have it. It is enough to notice that acid does real work on two fronts at once, on how long a food lasts and on whether a cat will accept it, and that a formulator has to satisfy both.

Step back and the physical levers we have walked through share a quality worth naming. Each of them is a case of the same thing: physics deciding how the chemistry of a food reaches the animal, or whether it reaches the animal at all. Temperature governs how much of the aroma leaves the food for the air. Water, and how much of it is free, governs how flavour dissolves and travels and how the food keeps. Texture and geometry govern whether the mouth can comfortably process the piece at all. Phase, the chunk and its separate sauce, governs whether the animal can take one part and leave another. And pH shifts taste, protein state, preservation and the food's chemical equilibria all together. None of these is the flavour. All of them decide what becomes of the flavour on its way to being judged. The recipe lists what is in the food. Physics settles what the animal actually gets.

We have brushed against this before in these Conversations, and only now does it show its shape. An earlier one followed fat down the road of oxidation and found that the same molecule which signals a fresh kill signals rancidity a few steps later, so that the animal is not reading whether the fat has oxidised but how far along the reaction has run. That was a story about how a chemical signal is presented to the animal in time, and this is the same story told in space and state, about how a whole food is presented to the animal in the physical form we give it. In both, the thing we call palatability turns out to live not purely in the composition, but in the meeting between a food in a particular state and an animal equipped to read it.

And here the physical lever quietly becomes a clinical one, which is a pattern this series keeps meeting. A cat's urinary health depends on the acidity of what it eats, because a diet that pushes the urine too far in one direction invites one kind of stone and too far in the other invites another. So the pH of a cat food is never chosen for palatability alone. It is chosen at the meeting point of what the cat will find attractive, what will keep it from harm, and what the food's own ingredients and preservation will allow. The window that palatability wants and the window that health requires are not guaranteed to be the same window, and the formulator lives in the overlap.

There is a last thing pH reveals, and it reframes the whole essay. The cat works with a relatively sparse gustatory apparatus, only a few hundred taste buds where a dog has thousands and a human more still, and it lacks a working sweet receptor entirely. Yet it remains one of the most discriminating animals we formulate for, the quickest to refuse, the hardest to please. That is not a contradiction. It is a warning against equating palatability with taste. The tongue is only one member of the committee, and for the cat, evidently, not the loudest. Smell, temperature, texture, geometry, the acidity the animal will accept only within a narrow band, the physical state of the food in every respect we have walked through: all of them get a vote. And all of them are, in one way or another, physics.

So the question this final lever leaves is the largest one. We have spent a career, as an industry, refining the flavour of foods for an animal that judges as much by other means as by the tongue. How much of the palatability we have been chasing was ever really in the flavour at all, and how much was waiting, all along, in the physical state of the food, in the plain facts of how warm it was, how wet, how hard, how shaped, and how sour, before the tongue was ever asked for its weak and final opinion?

Movement VI

THE SAME COW

Return, at the end, to the four kitchens.

The barbecue and the gulasch and the salami were the same cow, and no one who ate them was fooled into thinking they were the same dinner. We knew, in the mouth, that the fire and the braise and the ferment had made three different foods out of one animal, and we never once mistook the recipe for the meal. We grant ourselves that distinction without a thought. We have been slower to grant it to the animal in the bowl.

Because the pet food industry has spent most of its ingenuity, and most of these Conversations, on the recipe: the protein, the fat, the palatant, the chemistry of the thing. All of that is real and none of it is wasted. But the animal has never eaten the recipe. It has eaten an object, warm or cold, wet or dry, soft or hard, sharp or mild, and it has judged that object with its whole body, its nose and its mouth and its long inheritance of caution, well before the tongue was ever consulted. The formula was our question. The object was the animal's answer.

None of this asks us to think less of flavour. It asks us to notice how much of what we credited to flavour was being carried, quietly, by the plain physics of the food. And it suggests a way of holding the two together that does not make them rivals. Chemistry supplies the signals a food can send, the savoury peptide, the fresh-kill volatile, the amino acid the tongue is tuned to. Physics decides when, where and in what state those signals reach the animal, or whether they reach it at all. Physical form, in this light, is the delivery architecture of palatability. Chemistry writes the message. Physics controls its delivery.

Chemistry writes the message. Physics controls its delivery.

And it is worth pausing on the strangest thing of all, which is how the animal reads that delivery. The cat carries no pH electrode and no thermometer, no hygrometer to weigh the moisture and no instrument to measure the hardness of a piece or the water activity of a paste. It does not measure these properties one at a time. It encounters their consequences together, in the few seconds between approaching the bowl and eating or turning away. Our instruments separate the food into variables, temperature here, moisture there, acidity, hardness, geometry, each on its own dial. The animal integrates them into behaviour. And that behaviour is the only answer we ultimately receive. The food arrives whole. The animal answers whole. We are the ones who divide both into measurements.

We have spent these Conversations taking the food apart. We have looked at the volatile that announces a fresh kill, the fat that carries its own clock of freshness, the plasma that holds a chunk together, and the enzyme that cuts a protein until savour and bitterness begin to trade places. All of it matters, and none of it reaches the animal in the abstract. It arrives warm or cold, wet or dry, dissolved or held fast, on a surface or inside a body, in a chunk that fractures, a gravy that can be licked away, a jelly that keeps its water still, or a pâté in which there is nowhere for one part to hide from another. The same cow can be four dinners. The same formula can be five foods, because a formula has no temperature, no bite and no mouthfeel until we turn it into something an animal can actually eat.

The recipe tells us what we made. The animal only ever meets what it became.

References

1. Eyre, R., Trehiou, M., Marshall, E., Carvell-Miller, L., Goyon, A. & McGrane, S. (2022). Aging cats prefer warm food. Journal of Veterinary Behavior 47:86-92. doi:10.1016/j.jveb.2021.09.006

2. Bradshaw, J.W.S., Goodwin, D., Legrand-Défretin, V. & Nott, H.M.R. (1996). Food selection by the domestic cat, an obligate carnivore. Comparative Biochemistry and Physiology Part A 114(3):205-209. doi:10.1016/0300-9629(95)02133-7

3. Cook, N.E., Rogers, Q.R. & Morris, J.G. (1996). Acid-base balance affects dietary choice in cats. Appetite 26(2):175-192. doi:10.1006/appe.1996.0014

4. Yu, S., Rogers, Q.R. & Morris, J.G. (1997). Absence of a salt (NaCl) preference or appetite in sodium-replete or depleted kittens. Appetite 29(1):1-10. doi:10.1006/appe.1996.0088

5. Fettman, M.J., Coble, J.M., Hamar, D.W., Norrdin, R.W., Seim, H.B., Kealy, R.D., Rogers, Q.R., McCrea, K. & Moffat, K. (1992). Effect of dietary phosphoric acid supplementation on acid-base balance and mineral and bone metabolism in adult cats. American Journal of Veterinary Research 53(11):2125-2135.

6. Hagen-Plantinga, E.A., Orlanes, D.F., Bosch, G., Hendriks, W.H. & van der Poel, A.F.B. (2017). Retorting conditions affect palatability and physical characteristics of canned cat food. Journal of Nutritional Science 6:e23. doi:10.1017/jns.2017.17

7. Calderón, L.A. et al. (2024). Measuring palatability of pet food products: sensory components, evaluations, challenges, and opportunities. Journal of Food Science 89(9):5359-5382. doi:10.1111/1750-3841.17511

8. Shin, T., Nahm, I., Maeyama, T., Miyazaki, J., Matsuo, H. & Yu, Y. (1995). Morphological study of the laryngeal taste buds in the cat. The Laryngoscope 105(12):1315-1321. doi:10.1288/00005537-199512000-00010

About the Author

Dr. rer. nat. habil. Dr. Seronei Chelulei Cheison is the founder and Chief Executive Officer of Sinonin Biotech GmbH, where he works with petfood companies on palatant sourcing strategies, the innovation of palatability enhancers, and the design of palatability assessment. A trained food enzymologist, he led alternative protein and palatant development at Mars Petcare before founding Sinonin, and his research on protein chemistry spans two decades, a doctorate from Jiangnan University, and a habilitation from the Technical University of Munich.

Sinonin Biotech GmbH is a partner in the ZEST project (Grant Agreement No. 101157382) and the PROSCALE project (Grant Agreement No. 101288362), both funded by the Circular Bio-based Europe Joint Undertaking (CBE JU) under the European Union's Horizon Europe research and innovation programme. Views and opinions expressed are those of the author only and do not necessarily reflect those of the European Union or CBE JU. Neither the European Union nor the granting authority can be held responsible for them.

Tuesday, 1 September 2026

Protein Hydrolysate Palatability in Pet Food: Where to Stop the Cut

Where to Stop, The Friday Conversation No. 7. From papaya leaf to reactor, the protein hydrolysis continuum from intact protein to free amino acids.

On protein hydrolysates in petfood: when breaking protein makes food more palatable, when it does not, and why the animal decides where to stop.

In parts of this world, when the meat is tough, you wrap it in the leaf of the papaya tree and leave it a while before it meets the fire. In others you bury it, or hang it in the cold until the flesh gives up its stiffness on its own. My grandmother's generation did not call any of this chemistry. They called it good sense, and they were right, and they were also, without a word of the vocabulary, running a controlled enzymatic reaction on a kitchen table.

The papaya leaf carries papain. The pineapple carries bromelain, the fig its ficin, and the meat, left to itself in the cold, carries its own quiet proteases that go on working long after the animal has stopped. All of them do the same thing. They cut the long protein chains of muscle into shorter pieces, and the shorter pieces eat more tenderly, release more savour, and give the tongue more to find. Tenderisation is not softening in the way a soak in water softens. It is scission. It is the protein being taken apart, a little, on purpose, for the pleasure of the eater. And humans have been doing it, deliberately and with real skill, for far longer than they have had a name for the enzyme that does the work.

There is an older cousin to this, and the line between them is thinner than it looks. Long before refrigeration, communities learned to let meat and fish sit under the work of time and microbes, and to prize what came out: the deep, resonant flavour of a cured ham, a fermented fish, a sauce drawn from anchovies left to their own slow dissolution. We call that fermentation, and it is more than enzymes, microbes and time and salt all playing their parts. But enzymes are unmistakably in it, because much of what fermentation does to flavour, it does by proteolysis, by cutting protein into the fragments and free amino acids the tongue reads as savoury. Fermentation for flavour and tenderisation for tenderness are, at the level of the protein, close kin. Both are the controlled disassembly of protein for a sensory reward.

So the question this essay begins with is not a modern one at all. It is the one the grandmother answered with a papaya leaf, asked again with instruments: what happens, exactly, when you take a protein apart, and why has every food culture that ever lived gone to such trouble to do it?

Movement I

FROM THE LEAF TO THE REACTOR

The difference between the papaya leaf and the modern reactor is not the chemistry. It is the control.

When you wrap a tough cut in a papaya leaf, you get whatever the leaf happens to give you, for as long as you happen to leave it, at whatever temperature the kitchen happens to be. The enzyme works where it lands and stops when the heat of the fire finally destroys it, and the result is a better piece of meat and a great deal you never measured. It is real skill, but it is skill of the hand and the eye, passed down and adjusted by taste. What the last century added was not a new reaction. It was the dial. Choose the enzyme, and you choose which bonds along the chain get cut. Choose the temperature and the acidity and the time, and you choose how far the cutting goes. Stop it when you decide to stop it, and you fix the product exactly where you want it. The grandmother cut her protein by feel. The industry learned to cut it to a number.

The grandmother cut her protein by feel. The industry learned to cut it to a number.

That number has a name, the degree of hydrolysis, and it is simply the proportion of the protein's bonds that have been cut, from a whisper of cleavage to extensive fragmentation into small peptides and free amino acids. Everything that matters in this essay hangs on where along that range you choose to stop, because the protein is a different material at every point, and it behaves differently in the bowl and in the gut depending on where you left it.

This is not a marginal craft. Enzymatic hydrolysis is now a major industrial route to these ingredients, and the trade has grown into a substantial one precisely because controlled cutting turns out to be useful in ways the papaya leaf only hinted at. The proteins that go into it are drawn from wherever good protein can be had: whey and casein from milk, collagen from hide and bone, muscle and organ from slaughter, poultry, and a large and growing share from fish. And increasingly from plants, too, soy above all, along with pea and rice and wheat gluten, which matter more to this story than their share of the market suggests. Plant proteins are where hydrolysis has long been most notorious for the bitterness it can bring, and soy hydrolysate is the substrate on which a great deal of what we know about bitter peptides was first learned. In pet food the pull is strongest at the premium and therapeutic end, where hydrolysates are prized for two properties above all, and it is worth being precise about what those two properties are, because the rest of this essay is about the fact that they do not always come together.

The first is digestibility. A protein already cut into peptides has undergone part of the proteolysis the animal would otherwise perform for itself, and in some formulations and physiological circumstances that can alter the rate and pattern with which its nitrogen becomes available, which matters most for the young, the old, the recovering, and the compromised gut. The second is that extensive hydrolysis can reduce antigenicity, breaking the protein into fragments less able to be recognised by the immune system as the shape it once reacted to, which is why extensively hydrolysed proteins are reached for in the elimination and management diets of the allergic animal. These are real reasons for hydrolysing a protein, and they are why a formulator reaches for the tool. But neither of them tells us whether the resulting food will be eaten. Both are reasons of the gut, not the mouth, and a diet the animal refuses has a nutritional value of zero no matter how digestible or how hypoallergenic it is on paper. So the question this essay is really about is not what hydrolysis does to a protein. It is what hydrolysis does to a protein's chances at the bowl. And there the story stops being a list of benefits and becomes something far less obedient.

Movement II

THE SAME CUT

Here is the thing the papaya leaf never had to reckon with, because it never cut very far. The same scission that delivers the digestibility and the hypoallergenicity also does two other things, in the same stroke, that no one asked it to do. It is not that hydrolysis has a benefit and, separately, a side effect. It is that the cut is single and its consequences are plural. You cannot take the protein apart for one reason and leave the other reasons uninvited.

Consider what the cut does to taste, which turns out to be two opposing things at once. Cleaving protein releases free amino acids and small peptides that profoundly change flavour, and to a human tongue some of them carry a savoury character, the deep meaty resonance that has drawn every long-cooked stew and every aged cheese and every drop of fish sauce toward the same chemistry. But here the caution this whole series keeps returning to becomes almost the point. The cat does not read that chemistry as we do. Its savoury sense, the work of the palatability scientists tells us, is not built around glutamate the way ours is; it appears to lean more on nucleotides, with amino acids playing a supporting part, and the glutamate and aspartate that anchor the human idea of umami do not map cleanly onto how a cat responds at all. Kokumi, the mouth-filling roundness that is less a taste than an enrichment of the others, appears to be functional in the cat too, one more channel through which a broken-down protein might reach the animal.

So when hydrolysis unlocks a savour that sat mute inside the intact chain, we should be careful whose savour we mean, and honest that we do not fully know which of these channels a given cut is feeding, and which it is starving. We can release the molecules confident that they are appetising and be describing our own tongue rather than the animal's. But whatever the cat makes of that half, the very same cutting, carried further or aimed differently, can also do the opposite. It exposes the hydrophobic amino acids that had been folded safely inside the protein, and hydrophobicity is strongly associated with bitterness, though it is not the whole story. Hydrophobicity is a useful first clue to where the bitterness may reside, never a complete explanation of it: the more hydrophobic peptides tend to be both the more bitter ones and the ones that behave differently when you try to separate a hydrolysate by that property.

Some years ago I worked on exactly this, removing the bitterness and the salt from a whey protein hydrolysate at the same time by letting the bitter, hydrophobic peptides interact selectively with an adsorbent and lifting them out on that interaction. It works because bitterness is not scattered randomly through a hydrolysate; it rides on a physical characteristic you can act on. But knowing that bitterness and hydrophobicity travel together is a long way from knowing, for a given protein and a given purpose, where the balance tips. And the relationship between how far you cut and how bitter the result is turns out to be more interesting than a straight line. Bitterness depends on the substrate, the sequence, the size of the peptides and the specificity of the enzyme, and as the cutting proceeds it can rise as bitter hydrophobic peptides are liberated, peak, and even fall again as further cutting breaks those same peptides down. Every cut changes the peptide population, and the bitterness rides that changing population up and down rather than simply climbing. There is no monotonic dial you can simply turn less of. There is a moving target.

The cat makes this harder still, and in a way that ought to keep us humble. An obligate carnivore might be expected to have surrendered much of the bitter-detection apparatus that other animals use to avoid the toxins of plants. It did not; the domestic cat carries a substantial repertoire of bitter receptors that respond, in the laboratory, to bitter compounds. But a receptor firing in a dish is not the same as an animal tasting bitterness, still less disliking it, and the older work on how cats actually respond to bitter stimuli is genuinely murky, some of it pointing to responses that lack the clean specificity we would want before saying with any confidence what the cat perceives. So the honest position is a question rather than a claim. We know the cat kept the machinery, and we know a little of what it does with some of it: cats will reject quinine, so the apparatus is not idle. But quinine is a plant alkaloid, and the bitterness a hydrolysate carries is a different chemistry, a matter of peptides and exposed residues, and how the cat meets that is far less clear. We call these bitter receptors, but they earn the name only by resemblance to ours; the label is borrowed from the one species that can say what it tastes. When the cat's version fires, we do not actually know that the animal experiences bitterness as we would recognise it. The researchers who first characterised these receptors said as much, that the cat may detect a narrower, or simply a different, range of bitter things than we do, and that its bitter world has scarcely been studied. It might register something we have no word for, and the cat cannot tell us, because the one instrument that could settle the matter, its own report, is the one a cat does not have.

So the questions stack. Does the bitterness a cut liberates register at all in the cat; if it does, does it register as bitterness or as some other thing we cannot name; and if as bitterness, does a carnivore that meets such compounds in prey rather than in poison read them as an aversion to be masked, or make very little of them? I do not think anyone can answer that from a receptor, and until someone answers it some other way, a great deal of what we say about bitter hydrolysates and cats is really being said about our own tongues. The savour and the bitterness are not two ingredients you can order separately. They are two faces of the one act of cutting, and where you stop decides which face is showing.

The same cut that unlocks the savour exposes the bitter. You do not get to invite one and refuse the other.

Now consider what the cut does to structure, and here the essay meets its two predecessors. An intact protein can build things. It can be heat-set into a gel, woven into a network that holds water and fat and gives a chunk its bite, which is the entire subject of the two conversations that came before this one. Cut that protein into peptides and you take that ability away. A hydrolysed protein has been relieved of the very architecture that let it bind, because the long chains that entangled and cross-linked and held are now short pieces that cannot. So the tool that improves digestion dismantles structure in the same motion. The gain in digestibility is paid for in structure, and the loss is not incidental. It is the direct arithmetic of cutting a long thing into short things.

This is why there is no single best form of a protein, only a best form for a purpose, and it is worth laying the spectrum out honestly. At one end sits the intact protein: full structure, its savour locked up, and asking the animal's gut to do the work of breaking it down. At the far end sits the pool of free amino acids: no peptide bonds left for a protease to cleave, no peptide structure left to build a network with, and a sensory character that can be powerful without necessarily being attractive. And in the enormous middle sits the hydrolysate, wherever you chose to stop it, carrying some of each virtue and some of each cost. But it would be too tidy to call this a single spectrum with the good things at one end and the bad at the other. Every movement of the dial changes several properties at once, and they do not move in step, and they do not reach their best points together. Digestibility, structure, savour, bitterness: each follows its own curve as the cutting proceeds, rising and falling on its own schedule, and the peaks and troughs do not line up. The dial does not have a setting that is best at everything, and not merely because the properties compete along a line, but because they are several different response curves laid over one another, and no single point sits at the top of all of them.

Movement III

WHEN A HYDROLYSATE IS NOT A PALATANT

The word hydrolysate describes what we did to the protein. The word palatant describes what the animal thought of the result. They are not the same word, and they do not always describe the same thing, and the entire difficulty of this trade lives in the gap between them.

A hydrolysate can be, by every measure a laboratory can put to it, an excellent thing. Its nitrogen readily available, its antigenicity low, its solubility complete, its amino acid profile answering the animal's needs point for point. And it can still be met at the bowl with refusal, because none of those virtues is the same as being wanted. The cut that made it so digestible may have carried the bitterness up onto its peak. The peptides that make it dissolve so cleanly may be exactly the ones the cat's kept receptors object to. A protein can be hydrolysed perfectly for nutrition and badly for palatability, and the two failures are invisible to each other. The nutritionist's instruments will report a triumph while the animal walks away.

Hydrolysate describes what we did to the protein. Palatant describes what the animal thought of the result.

Collagen offers a particularly visible demonstration of what progressive dismantling can do, because it has been walked down that road in front of us for a century. In its native state it is highly ordered and structurally formidable, with real technological and nutritional consequences of its own. Disrupt that ordered triple helix through denaturation and partial hydrolysis into gelatine, and its behaviour changes dramatically: it becomes dispersible and can set a thermoreversible gel, a reminder in passing that the cut can be made by chemistry as readily as by the papaya leaf's biology. Hydrolyse it further into smaller collagen peptides and that gel-forming capacity largely disappears, leaving a soluble, available fragment that builds nothing. One parent protein, progressively dismantled into materials with profoundly different functions, and not one point along that road that is best at everything. Where you would stop, for collagen or for any protein, depends entirely on what you were trying to make, and whether the animal agrees is a separate question the road cannot answer.

None of this is waste rescued from a bin. The tougher streams a slaughter leaves, the connective tissue and the harder material, are already valorised into the rendered meals and fats that are real ingredients in their own right, and hydrolysis is simply a further rung on that ladder, a way of taking something already useful and giving it a different functionality, perhaps more soluble, perhaps more digestible, and potentially more useful within a palatability system. But none of those transformations guarantees that the animal will want it. You can climb every rung of nutrition and functionality and still arrive at a material the animal declines, because willingness is not a rung on that ladder at all. It is the judgement waiting at the top, and it belongs to the eater.

Which leaves a question worth sitting with, if you make or buy these ingredients. Before the animal has told you, how would you know whether the excellent hydrolysate in front of you is a good palatant? What, exactly, would you measure? We have instruments for every rung of the ladder and none for the judgement at the top, and I am not sure the industry has been honest with itself about how much of what it calls palatability prediction is really just the hope that the two will coincide.

Movement IV

WHERE TO STOP

So the whole art, in the end, comes down to a decision the papaya leaf never had to make consciously: where to stop the cut.

Stop too early and the protein may remain little changed, much of its original structure intact, its immunoreactive epitopes potentially preserved, and whatever sensory potential the cutting might have released still largely locked away. Cut too far and you have a pool of free amino acids that the gut absorbs in an instant, the structure gone entirely and the savour at risk of tipping over into a rawness the animal may or may not accept. Somewhere between those, for any given protein and any given purpose, there is a place to stop that gives the animal enough of what it needs and spares it enough of what it will not tolerate. That place is different for a hypoallergenic diet than for a digestibility aid, different for a fish protein than for a collagen, different for a cat than for a dog. It is not one setting. It is a judgement made freshly each time, against the specific material and the specific animal and the specific job the ingredient has to do.

I am not going to tell you, in this essay, how that judgement is made. Partly because it is the working life of people like me and not a thing to be given away in a Friday paragraph, but also because the answer is not a universal number. It belongs to the substrate, the process, the purpose, and, ultimately, the species doing the eating. It is the accumulated feel for a tool that the industry has spent decades learning, the modern inheritor of exactly the skill the grandmother had with her leaf, only now aimed with instruments and held to a number. What can be said plainly is what the decision is between, and this whole conversation has been an attempt to say it: between digestion and structure, between savour and bitterness, between what the gut will take and what the mouth will accept, all of them moving at once, all of them tied to the single fact of the cut.

There is no setting that is best at everything. There is only the setting that is best for this.

And here is where the title of this conversation finally asks its real question. Where to stop the cut is not, in the end, a question the laboratory can close. You can measure the degree of hydrolysis. You can measure the size of the peptides and the quantity of free amino acids, the solubility, the digestibility, and with enough patience even the bitterness. There are instruments now, electronic tongues and noses and the models built on top of them, that read these properties and predict, sometimes rather well, how a food is likely to be received. But prediction is not preference, and correlation is not consent. Every one of those readings is a number, and not one of those numbers is the animal's yes. The best of them forecast the verdict. None of them is the verdict. Which ought to unsettle us more than it does, because we specify these ingredients to numbers all the same. When we write a hydrolysate to a target, are we describing what the animal will accept, or only what the quality sheet can check? The two are not the same, and the gap between them is where a great many refused bowls have quietly lived.

The laboratory measures hydrolysis. The animal measures palatability.

The grandmother knew when the meat was ready. She knew it the way you know these things, by the eating. We have better tools than she had, and a great many more numbers, but at the very end we are still waiting on the same verdict she was, and it is not a verdict any instrument can return. We can say, with all the precision in the world, exactly how far we cut the protein. Only the animal can tell us whether we stopped in the right place.

Which means the question was never really only where to stop the cut. It was how to learn to ask the animal sooner, before the batch is made, before the diet is formulated, and before the refusal comes back from the bowl too late to do anything about it. I do not think we have answered that question yet. I am not sure we have been asking it in the right language. And that, more than any degree of hydrolysis, is the conversation I would like us to be having.

The bowl, as always, decides.

References

1. Cheison, S.C., Wang, Z. & Xu, S.-Y. (2007). Use of macroporous adsorption resin for simultaneous desalting and debittering of whey protein hydrolysates. International Journal of Food Science & Technology 42(10):1228-1239. doi:10.1111/j.1365-2621.2006.01461.x

2. Cheison, S.C. & Kulozik, U. (2017). Impact of the environmental conditions and substrate pre-treatment on whey protein hydrolysis: A review. Critical Reviews in Food Science and Nutrition 57(2):418-453. doi:10.1080/10408398.2014.959115

3. Liu, B., Li, N., Chen, F., Zhang, J., Sun, X., Xu, L. & Fang, F. (2022). Review on the release mechanism and debittering technology of bitter peptides from protein hydrolysates. Comprehensive Reviews in Food Science and Food Safety 21(6):5153-5170. doi:10.1111/1541-4337.13050

4. Fu, Y., Chen, J., Bak, K.H. & Lametsch, R. (2019). Valorisation of protein hydrolysates from animal by-products: perspectives on bitter taste and debittering methods. International Journal of Food Science & Technology 54(4):978-986. doi:10.1111/ijfs.14037

5. McGrane, S.J., Gibbs, M., Hernangomez de Alvaro, C., Dunlop, N., Winnig, M., Klebansky, B. & Waller, D. (2023). Umami taste perception and preferences of the domestic cat (Felis catus), an obligate carnivore. Chemical Senses 48:bjad026. doi:10.1093/chemse/bjad026

6. Laffitte, A., Gibbs, M., Hernangomez de Alvaro, C., Addison, J., Lonsdale, Z.N., Giribaldi, M.G., Rossignoli, A., Vennegeerts, T., Winnig, M., Klebansky, B., Skiles, J., Logan, D.W. & McGrane, S.J. (2021). Kokumi taste perception is functional in a model carnivore, the domestic cat (Felis catus). Scientific Reports 11:10527. doi:10.1038/s41598-021-89558-w

7. Sandau, M.M., Goodman, J.R., Thomas, A., Rucker, J.B. & Rawson, N.E. (2015). A functional comparison of the domestic cat bitter receptors Tas2r38 and Tas2r43 with their human orthologs. BMC Neuroscience 16:33. doi:10.1186/s12868-015-0170-6

8. Lei, W., Ravoninjohary, A., Li, X., Margolskee, R.F., Reed, D.R., Beauchamp, G.K. & Jiang, P. (2015). Functional analyses of bitter taste receptors in domestic cats (Felis catus). PLoS ONE 10(10):e0139670. doi:10.1371/journal.pone.0139670

9. Cho, M.J., Unklesbay, N., Hsieh, F.-H. & Clarke, A.D. (2004). Hydrophobicity of bitter peptides from soy protein hydrolysates. Journal of Agricultural and Food Chemistry 52(19):5895-5901. doi:10.1021/jf0495035

10. Schrieber, R. & Gareis, H. (2007). Gelatine Handbook: Theory and Industrial Practice. Wiley-VCH, Weinheim.

About the Author

Dr. rer. nat. habil. Dr. Seronei Chelulei Cheison is the founder and Chief Executive Officer of Sinonin Biotech GmbH, where he works with petfood companies on palatant sourcing strategies and the innovation of palatability enhancers. A trained food enzymologist, he led alternative protein and palatant development at Mars Petcare before founding Sinonin, and his research on enzymatic protein hydrolysis spans two decades, a doctorate from Jiangnan University, and a habilitation from the Technical University of Munich.

Thursday, 27 August 2026

Animal-Free Pet Food: Can It Stay Palatable Without Plasma?

THE FRIDAY CONVERSATION · No. 6

Can We Build a (Still) Palatable Chunk Without Animal Blood Plasma?

Palatability is not only texture. It is also taste and aroma. But removing animal blood plasma from a wet food is easy; removing it without the animal noticing is the hard part, and that is where the whole replacement problem is really decided.

Dr. rer. nat. habil. Dr. Seronei Chelulei Cheison

Illustration of a plant-protein network with one strand missing at its centre, holding water and fat weakly

The plant kingdom supplies almost every strand. Whether one protein locks the net, or the ensemble does, is the open question.

A soybean contains a protein called albumin. It contains proteins called globulins. Open a seed catalogue of the plant kingdom and you will find, listed in plain type, the same words we use for the proteins in blood: albumin, globulin, and, if you go looking, even a red pigment that carries iron in a ring of haem, the leghaemoglobin that gives a soy nodule its bloody blush and gives a certain famous plant burger its bleed. On paper, the plant has everything the plasma has. The names line up.

So the obvious question, the one a formulator asks in the first meeting, is why we cannot simply reach for the plant version and be done with it. The plant makes albumin. Plasma is largely albumin. Why is this hard?

It is hard because a name is not a function. The word albumin, in a seed, means only that the protein dissolves in water and sets when you heat it. It was borrowed, a century ago, precisely because those plant proteins reminded chemists of the egg. The word tells you how the protein behaves in a test tube, not what job it does in a chunk. Plant albumin and blood albumin are cousins in name and strangers in work. Leghaemoglobin made the colour of blood, and it made it convincingly enough to sell. It did not make blood.

The plant has albumin. The blood has albumin. Cousins in name, strangers in work.

And that is the whole essay in a sentence, so let me say it plainly before we go further. We are not, in this series, trying to replace a protein. We are trying to replace what a protein does, in a chunk, to an animal that will either finish the bowl or walk away from it. The plasma question was never really about plasma. It was about the bowl.

Movement I

THE ANIMAL IN THE CHUNK

Plasma is animal blood, and there is a great deal of animal in it: a tonne of the powder is the recovered plasma of something like three thousand pigs, or nine hundred cattle, though not one of them died for it, since they were killed for meat and the blood is only what ran out. That last point matters, because it tells you what the animal-free project is and is not about. It is not about saving those animals; the meat industry will kill them with or without a use for their blood. It is about something simpler and, in its way, harder to argue with. A growing number of people want to feed their animals a diet with nothing in it that was ever an animal, for reasons of their own that the industry does not get to overrule, and plasma is the quiet animal ingredient sitting in the one product they cannot yet have.

That product is the wet chunk. Structurally, plant-based kibble is largely a solved engineering problem. The wet chunk in gravy is another matter: holding together, holding its water, keeping its shape through the retort and the shelf and the fork, it is the wall the vegan project keeps hitting, and plasma is mortared into it.

No animal dies for its plasma. They die for meat, and the blood is what runs out.

So the ambition is easy to state and hard to meet. We want the chunk without the animal. We want it to bind, to hold, to bite, and above all to be eaten, with nothing in it that was ever an animal. What we are really asking for, though we rarely put it this baldly, is an animal-free plasma: something that is not blood and has never been near an animal, but that does, in the chunk, the work that blood did. The question is whether that thing can exist. And that is where the honest trouble begins.

Movement II

THE THINGS THAT ALMOST WORK

The first instinct is to reach for a gum. Our pantry is full of them, and several will hold a chunk together, so let me be fair to each before I say why fair is not enough.

Gellan is the one I reach for first, partly because it is already, in its way, an animal-free triumph. It is not dug or harvested; it is fermented, spun out of a bacterium in a tank, which is exactly the kind of process the back half of this essay will end up praising. It gels. It holds. But ask it how, and the trouble starts. High-acyl gellan sets into something soft and elastic that melts again when you warm it. Low-acyl gellan sets into something firmer that does not melt, but is stiff and brittle where the chunk wants to be yielding. And either way it will only gel if the water around it carries the right ions, calcium or potassium, in the right amount, which a wet recipe does not always politely supply. It is a gel with conditions.

Methylcellulose is stranger still, and it is the darling of the plant-burger, so it deserves a hearing. It does something almost no other food ingredient does: it sets when you heat it and melts when you cool it. That is not a typo. The burger holds on the grill because the methylcellulose is gelling in the heat, and it is at its most fragile on the plate as it cools toward the temperature at which it will actually be eaten. For a patty eaten hot off the flame, that can be made to work. For a chunk in gravy that sits in a bowl at room temperature while a cat decides whether to bother, a binder that is firmest when hot and weakest when cool is solving the wrong half of the problem.

And this is the pattern, once you line the candidates up. Gelatine sets as it cools and melts as it warms. Methylcellulose sets as it warms and melts as it cools. Gellan sets on cooling and either melts again or turns brittle, and only if the ions are right. Every one of them gels. Not one of them gels the way plasma does, which is to set when heated and then stay set, permanently and elastically, through the retort and the cooling and the shelf, holding water and fat in a network that does not care what the temperature does next. The plant proteins can be pressed into service too, and some of them, potato protein in particular, will form a true heat-set gel of their own. But pressed into the wet chunk, asked to survive the retort and the flood of gravy and still give a bite, they arrive short in one way or another, and the formulator ends up propping them with exactly the gums above.

So the roster is not empty. It is worse than empty. It is full of ingredients that each do part of the job and none of which does the whole of it, which is why the honest version of a plasma-free chunk today is not one clever replacement but a committee of them, a little gellan for set, an insoluble cellulose fibre working as a sponge to hold the water and steady the texture, a protein for body, a soluble cellulose gum to thicken the gravy and paper over the difference, each covering for what the others lack. And a committee is not a keystone. It is what you build when you do not have one.

Every one of them gels. Not one of them gels the way plasma does.

Which forces the real question, the one the gums cannot answer. What is it, exactly, that plasma has and none of these has? What is the single thing we keep failing to reproduce?

Movement III

THE PROTEIN I WAS TEMPTED TO NAME

There is a protein it is very tempting to name here, and I am going to name it, but I am going to be honest about how much weight the name can bear.

Plasma is largely albumin, and the plant has albumin, and we have already seen that the shared name is a coincidence of solubility rather than a shared job. But albumin is not, on its own, the obvious source of the one property the committee cannot reproduce: the setting into an elastic, water-holding solid that survives the retort and does not come apart in the gravy. And there is a candidate for that property with a very suggestive credential. Take fibrinogen out of plasma and what remains is called serum. The two fluids differ by essentially that one protein, and they behave differently in a gel. Fibrinogen is also a protein the plant kingdom has never made, and never had any reason to make, because plants do not bleed. So the temptation is obvious. Name fibrinogen the keystone, ferment that one protein, and the animal-free chunk falls into place.

Here is why I will not quite say that, and the reason matters more than the conclusion. In the body, fibrinogen does something very specific and very beautiful. When a vessel is breached, an enzyme called thrombin snips two small peptides from the molecule, and the sticky ends left behind reach out, clasp their neighbours, and assemble, without further instruction, into an insoluble elastic mesh. That is the clot, and it is genuinely a thing no plant protein does. But that is physiology, and a chunk in a can is not a wound. What happens in the powder is subtler than it first looks, and it took me a moment to get it right. The plasma is collected with an anticoagulant, commonly citrate or a phosphate, which does not attack the clotting proteins but locks away the calcium the coagulation cascade needs to proceed. So fibrinogen enters processing largely as fibrinogen, whole and uncleaved, rather than as a pre-formed fibrin clot, and thrombin is never generated from its precursor to do the cleaving. Spray-drying can preserve a good deal of protein functionality, though how much coagulation activity in particular survives depends on the process. What matters for the chunk is the step that comes next. The set that forms in the retort is principally a heat-induced protein gel, not a physiological fibrin clot: wet heat drives the whole population of plasma proteins, albumin and globulins and the intact fibrinogen together, into a collective aggregation, a gel formed by denaturation rather than by enzyme. And here is the twist that keeps fibrinogen in the frame even so, because it arrives whole rather than spent, it enters that thermal gelation as a full and unusually gelation-active protein. Fibrinogen clearly contributes materially to how the network behaves. But the published work on how plasma and its isolated fractions gel points just as strongly to interactions among the fractions, fibrinogen and albumin and the globulins influencing one another as they denature, as to any single protein doing the work alone, and only that reading is honest on what we actually know.

And there is a deeper reason to distrust the tidy answer, which is that I have made this mistake before, one conversation ago. In writing about meat I argued at length that reducing a food to its most quotable constituent misses the emergent system, that the value is in the ensemble and not the headline molecule. It would be a poor kind of consistency to spend that essay warning against reductionism and then open this one by declaring that plasma is really just fibrinogen in an ensemble. Plasma is a multi-protein system. Its gelation is collective. Some plant proteins do form strong, irreversible, heat-set networks of their own, so the point is not that plants cannot gel. It is more specific and, I think, stronger: no single plant ingredient appears to reproduce the full multifunctional behaviour of plasma under this process context, the setting and the water-holding and the fat-holding and the bite, all at once and all through the retort. Which single protein, if any, carries the irreplaceable part of that is a question I can pose sharply but cannot yet close, and I would rather pose it sharply than close it dishonestly.

The plant never made fibrinogen. Whether fibrinogen alone makes the chunk is a different question.

I can say all of this with more than a reader's confidence, because I have been down the near end of this road myself. Some years ago I worked on, and hold a patent for, a way of building the binding function of a wet pet food from a denatured whey protein rather than from blood, a micro-particulate protein engineered to set into the kind of matrix a chunk needs. It works, within its limits, and I am glad of it. But it taught me the lesson this essay is circling. You can engineer a protein to do a good deal of what plasma does. What you cannot easily do is get all of it, and get it from a plant, and get it animal-free, all at once. Whey is milk. Milk is an animal. Solving the plasma problem is not the same as solving the animal problem, and rebuilding one function is not the same as rebuilding the system. The binding can be reconstructed. The demand was never only for binding.

So the honest position, at the end of the roster and the patents and the gums, is this. The plant cupboard does not hold whatever it is that plasma has. A committee of hydrocolloids is a confession that we are working around an absence rather than filling it. And the absence has a named candidate, fibrinogen, with a real credential and a case that the evidence complicates rather than closes. Which leaves one door in the building we have not yet tried to open, and a more careful question to carry through it.

Movement IV

THE DOOR MARKED FERMENTATION

The careful question is this. If some part of what plasma does turns out to be genuinely irreplaceable, some protein or small set of proteins that no plant makes and no gum imitates, are we then simply stuck, holding a wet chunk we can build almost but not quite well enough. And the answer, built quietly over the last decade, is no. Not if we stop asking the plant to be the source at all.

We already make animal proteins without animals. Not by growing the animal, and not by finding a plant that fakes it, but by handing the gene to a microbe and letting the microbe do the work. The whey protein in a growing number of ice creams was never in a cow; it was fermented from a fungus carrying the instructions for it. Egg-white protein is being made the same way, and so, in laboratories and increasingly at scale, is serum albumin itself, the most abundant protein in the very plasma this essay has spent two instalments discussing. The proteins that were said to be too structurally clever for a plant to replace are not being replaced. They are being copied, faithfully, by organisms that can, increasingly, be taught to make proteins once thought inseparable from the animal.

Which turns the animal-free chunk from a search into a construction, and changes the questions we ask of it. Not what replaces plasma, but which of plasma's jobs actually matter to the animal, and what is the minimum animal-free system that can reproduce the functions that matter. Can plant protein carry the body while an insoluble fibre and a soluble gum manage the water? Can fat be placed and stabilised by design rather than by luck? And for whatever network function survives all of that as genuinely irreplaceable, the part the committee keeps failing to cover, can a protein grown in a tank stand in for one drained from an animal?

If that protein turns out to be fibrinogen, we ferment fibrinogen. If it turns out to be fibrinogen doing its work only in the company of others, we learn that, and we build the company too. The tool does not care which answer the evidence returns. It cares only that the answer, once known, can be made without the animal.

We are not looking for the plant that can be plasma. We are rebuilding what plasma did, job by job.

I will not pretend this is finished, or easy, or that I am a disinterested narrator of it. It is the problem I have given my working life to, and there is a name waiting on it for when the science is ready. But a name is a promissory note, and this essay is not the place to cash it. The honest thing to say is only that the door exists, that it is not locked, and that behind it is not a single clever substitution but a discipline: find the jobs that matter, identify the few that only an animal protein has ever done, and grow those and only those, while the plant kingdom does the rest of the work it was always able to do.

There is a reason this door has stayed shut, and it is not that no one noticed it. Plasma begins with an extraordinarily cheap raw material, blood from a slaughter that was going to happen anyway. The powder itself is not free; it has to be collected hygienically, separated, concentrated, stabilised and spray-dried, and those operations carry real cost. But even so, precision fermentation begins at a structural disadvantage, because it must deliberately manufacture, in a tank and on an energy bill, a protein whose biological synthesis the animal has already paid for. So the fermented protein does not only have to work. It has to earn its place against a mature co-product whose hardest cost, the making of the protein itself, was settled long ago inside a living body, and it has to do so for people who have decided, for reasons of their own, that a co-product of slaughter is still too much animal to accept. That is a real bar, and pretending it is low would be its own kind of dishonesty. But it is a bar that novel ingredients have cleared before, once the demand was certain enough to pull the cost of making them down.

And here the engineering has to hand the question back, because structure is not the same as palatability. A chunk can pass every instrument in the laboratory, match the old one for hardness and springiness and water held, and still be met at the bowl with the particular indifference a cat reserves for food that is almost right. Plasma may carry signals we have not fully named, chemical as much as physical, and reconstructing its texture job by job may still leave those signals behind. There is already a hint of this in the feeding data, where cats have preferred plasma-bound wet food over the same recipe bound with wheat gluten, and where some of plasma's biological effects appear to survive the can, which tells us the ingredient was never doing only mechanical work. The rheometer can tell us whether we rebuilt the chunk. It cannot tell us whether we rebuilt the food. Rebuilding what a food does is not quite the same as rebuilding what an animal recognises.

And there is one more reason to do this now rather than later, which is that the concession this essay opened with has an expiry date. I said at the start that no animal dies for its plasma, that the beast is killed for meat and the blood is only what runs out, and that is true today. But it is true only for as long as meat means slaughter. We are already producing cultivated chicken from animal cells, while companies work towards structured cuts of cultivated beef, learning, tissue by tissue, to make meat without killing the animal it came from. Plasma is just another tissue, a fluid one. If that promise holds, and it is still a promise, resting on the whole cell-culture logic proving genuinely clean and food-grade at scale, then the blood stops being a byproduct that was going to run out anyway. An animal-free steak in a gravy thickened with drained blood is not a destination. It is a halfway house.

But notice what that world does not do. It does not hand us a bowl of prime cultured muscle and call it dinner. Cultured meat is designer meat, grown almost without waste, and that is precisely why it will not, on its own, feed a cat. A prime cut is not a balanced diet; it is muscle, and a cat needs the whole formulated matrix around it, the organ nutrients and the taurine and the minerals and the fat and the moisture, composed and bound into something shelf-stable and palatable. That formulation does not disappear when the slaughterhouse does. If anything the cleaner technology frees land and feed and effort for exactly the novel ingredients this essay has been circling. The chunk still has to be built and held together, from material that increasingly will not have come from an animal at all. The binding problem is older than slaughter and outlasts it, because it belongs not to waste but to formulation itself.

And so we arrive back where the last essay left us, with the question turned finally right way round. We spent this whole conversation asking how to replace plasma, and the asking was half the error. We were never going to replace the ingredient, and we were never going to find its single secret and copy it in isolation. The ingredient is blood, an ensemble, and blood is not available to the animal-free chunk on any terms. What we can do, if we are careful about it, is reproduce not the ingredient but the jobs the ingredient did, drawing each from the source that does it most honestly, and let the plants do everything they were always waiting to help with.

Whether that chunk, when it is finally set and sliced and dropped into its gravy, is one an animal will finish, is not a question chemistry can answer. It is the one the instruments cannot reach, and it belongs, as it always has in this trade, to a single judge, sitting in front of a bowl, deciding in the only currency that has ever mattered whether to eat.

The bowl, as always, will decide.

References

1. Osborne, T.B. (1924). The Vegetable Proteins, 2nd ed. Longmans, Green & Co., London. (The albumin, globulin, prolamin and glutelin solubility classification of plant seed proteins.)

2. Shewry, P.R., Napier, J.A. & Tatham, A.S. (1995). Seed storage proteins: structures and biosynthesis. The Plant Cell 7(7):945-956. doi:10.1105/tpc.7.7.945

3. Ockerman, H.W. & Hansen, C.L. (2000). Animal By-Product Processing and Utilization. CRC Press. (Slaughter blood yields; plasma as a recovered co-product.)

4. Blázquez, E., Rodríguez, C., Ródenas, J., Segalés, J., Pujols, J. & Polo, J. (2020). Biosafety steps in the manufacturing process of spray-dried plasma: a review with emphasis on the use of ultraviolet irradiation as a redundant biosafety procedure. Porcine Health Management 6:24. doi:10.1186/s40813-020-00155-1

5. Dàvila, E., Parés, D., Cuvelier, G. & Relkin, P. (2007). Heat-induced gelation of porcine blood plasma proteins as affected by pH. Meat Science 76(2):216-225. doi:10.1016/j.meatsci.2006.11.002

6. Polo, J., Rodríguez, C., Saborido, N. & Ródenas, J. (2005). Functional properties of spray-dried animal plasma in canned petfood. Animal Feed Science and Technology 122(3-4):331-343. doi:10.1016/j.anifeedsci.2005.03.002

7. Weisel, J.W. & Litvinov, R.I. (2013). Mechanisms of fibrin polymerization and clinical implications. Blood 121(10):1712-1719. doi:10.1182/blood-2012-09-306639

8. Aro, N., Ercili-Cura, D., Andberg, M., Silventoinen, P., Lille, M., Hosia, W., Nordlund, E. & Landowski, C.P. (2023). Production of bovine beta-lactoglobulin and hen egg ovalbumin by Trichoderma reesei using precision fermentation technology and testing of their techno-functional properties. Food Research International 163:112131. doi:10.1016/j.foodres.2022.112131

9. Knychala, M.M., Boing, L.A., Ienczak, J.L., Trichez, D. & Stambuk, B.U. (2024). Precision Fermentation as an Alternative to Animal Protein, a Review. Fermentation 10(6):315. doi:10.3390/fermentation10060315

10. Rodríguez, C., Saborido, N., Ródenas, J. & Polo, J. (2016). Effects of spray-dried animal plasma on food intake and apparent nutrient digestibility by cats when added to a wet pet food recipe. Animal Feed Science and Technology 216:243-250. doi:10.1016/j.anifeedsci.2016.03.026

11. Cheison, S.C. & Murgueytio Riofrio, E.L. (2024). Pet Food Product. US Patent 12,096,781 B2 (filed 16 August 2018; granted 24 September 2024). Assignee: Mars, Incorporated.

Dr. rer. nat. habil. Dr. Seronei Chelulei Cheison is the founder and Chief Executive Officer of Sinonin Biotech GmbH, a biotechnology company focused on alternative protein and palatability enhancer innovation and application for petfood formulations. He formerly led alternative protein and palatant development projects at the Mars Petcare Global Innovation Centre in Verden, and spent close to two decades in academic research on enzymatic protein hydrolysis, holding a doctorate from Jiangnan University and a habilitation from the Technical University of Munich.

Sinonin Biotech GmbH is a partner in two consortia funded under the Circular Bio-based Europe Joint Undertaking: ZEST, on fungal fermentation of agricultural residues, and PROSCALE, on scalable microbial protein ingredients, which runs from September 2026 to August 2030.

Co-funded by the European Union under Grant Agreement No. 101157382 (ZEST) and Grant Agreement No. 101288362 (PROSCALE). Views and opinions expressed are however those of the author only and do not necessarily reflect those of the European Union or CBE JU. Neither the European Union nor the CBE JU can be held responsible for them.