Recipes

    Rennet

    ⭐ The flavour half of a traditional coagulant is now its own recipe — Cheese Lipase, Pregastric — which is what the powder sold to home cheesemakers as “lipase” actually is.

    I bought both vegetable and animal rennet when I started, because the books disagreed about which to use and I wanted to find out for myself. This page is what I’d have liked to read first.

    The short version: they are not interchangeable, the differences matter most in cheeses you age a long time, and the option with the worst public image is arguably the best one.

    What rennet actually is

    “Rennet” is a coagulant preparation, and the enzyme doing the work in the traditional version is chymosin — which, as coagulation covers, is remarkable for being almost absurdly specific. It cleaves one bond in κ-casein, Phe105–Met106, and largely leaves the rest of the milk’s protein alone.

    That specificity is the whole point, and it’s the axis every rennet on the market varies along. A coagulant that cuts only that bond gives you a clean set and a cheese that ages predictably. A coagulant that also chews on other proteins gives you a set — and then keeps working for the next eighteen months.

    Hold that thought; it explains the bitterness problem below.

    The four you can buy

    Animal rennet

    Extracted from the fourth stomach (abomasum) of a young ruminant, traditionally a calf. This is the original, and it’s what the pouch story is about.

    It is not pure chymosin. It’s a preparation containing chymosin plus pepsin and other enzymes, and the ratio varies — with the animal’s age and diet among other things. A young milk-fed calf yields a chymosin-rich extract; an older animal yields proportionally more pepsin. Pepsin is a less specific protease, so a pepsin-heavy rennet is a slightly blunter instrument.

    Practical implication: animal rennet is a variable natural product. Excellent, traditional, and not identical bottle to bottle.

    Fermentation-produced chymosin (FPC)

    The calf chymosin gene, expressed in a microorganism, fermented in a tank. Approved by the FDA in 1990 — the first bioengineered product in the US food supply, before the Flavr Savr tomato and before GM soy — and it is in most American cheese today, usually labelled as “enzymes.”

    On the merits, and I say this as somebody who bought the traditional stuff first:

    • it is pure chymosin, one enzyme, without the accompanying pepsin;
    • it is therefore identical every time — no animal-to-animal variation;
    • and because it is maximally specific, it does the least unwanted proteolysis during aging.

    It also requires no calf, which is what a great many vegetarians actually want — while being, in the technical sense, a product of genetic engineering, which many of the same people are wary of. I don’t have a tidy resolution to offer and I don’t think one exists. It’s just a genuinely awkward fact, and it’s covered further on today.

    Microbial rennet

    This is not the same thing as FPC, and the two get confused constantly — including on labels.

    Microbial rennet is a protease from a fungus (commonly Rhizomucor species). It is not chymosin. It’s a different enzyme that happens to clot milk, and it is genuinely vegetarian in the straightforward sense.

    Its weakness is specificity. It’s less precise than chymosin, so it does more general proteolysis — which is fine in a fresh cheese eaten next week and becomes a problem in a cheese you intend to keep.

    Vegetable / plant rennet

    The oldest alternative and the one with the best story: cardoon thistle, whose enzymes (cardosins) clot milk, and which Columella recommended in the first century AD alongside safflower seed and fig sap. Portugal’s Serra da Estrela and Spain’s Torta del Casar are still cardoon-set as a matter of PDO law, and they have a soft, spoonable, faintly bitter character you cannot get any other way.

    That last word is the warning as well as the attraction.

    🔴 Why vegetable rennet turns bitter in an aged cheese

    Here is the practical thing I wish somebody had told me plainly.

    Plant proteases are less specific than chymosin. They cut κ-casein — which is why they coagulate milk at all — but they also cut other caseins, and they carry on doing it inside the wheel for as long as the cheese exists.

    Proteolysis breaks proteins into peptides. Some peptides taste fine. Some short, hydrophobic peptides taste intensely bitter, and excessive or indiscriminate protein breakdown produces them. In a fresh cheese eaten within days, the enzyme hasn’t had time and it doesn’t matter. In a wheel you’re keeping for a year, it has all the time in the world.

    So the rule of thumb that actually holds:

    Plant and microbial coagulants for fresh and short-aged cheeses. Chymosin — animal or FPC — for anything you intend to age.

    The great cardoon cheeses are the exception that proves it: they are designed around that character, eaten at a few months, in a style built to carry it. That is not the same as it happening to your cheddar by accident at month nine.

    ⭐ Added later: cardoon is a sharper exception than that, and it corrected me elsewhere

    Writing Spain & Portugal turned up two things that make this section more precise rather than less.

    What’s actually in the flower. Cardoon’s clotting activity comes from cardosins A and B, two aspartic proteases — and cardosin A is similar in activity and specificity to chymosin, while cardosin B is similar to pepsin. So “plant rennet” is not one blunt thing: the thistle independently offers functional analogues of both animal-rennet enzymes, which is a large part of why cardoon works where a generic plant protease wouldn’t.

    🔴 And the bitterness relationship is not monotonic. Cheese made with cardoon has been measured as having a higher degree of proteolysis and a less bitter taste than the same cheese made with animal rennet.

    More proteolysis, less bitterness. Which means the rule stated above — read carefully — is right for the reason it gives and wrong if you compress it. The bitterness comes from which peptides, not from how much protein is broken down. A protease with different cleavage specificity can be more active overall and still generate fewer of the short hydrophobic peptides that taste bitter.

    I had compressed it. On the Italy page I used “broadly proteolytic therefore bitter” as a law to argue that Pecorino di Farindola’s pig rennet — pepsin, not chymosin — should produce a bitter cheese, against its makers’ claim that it never does. Cardoon is the counterexample, and the revision is published on both pages.

    The practical rule above still stands — plant and microbial coagulants for fresh and short-aged cheeses, chymosin for anything you’re keeping — because most plant proteases genuinely do misbehave over a year. What doesn’t stand is treating total proteolysis as a proxy for bitterness.

    IMCU — the number on the bottle

    Rennet is sold at different strengths, and “a quarter teaspoon” is a meaningless instruction without knowing which. The unit is IMCU — International Milk-Clotting Units.

    Get this wrong and the failure is specific and recognisable: too much rennet gives a fast, brittle set that shatters rather than cutting cleanly, expels whey too aggressively, and — because the excess enzyme stays in the curd — contributes bitterness later. Too little gives a weak set that never reaches a clean break.

    Check the bottle against the book. If the recipe doesn’t state a strength, work from the bottle’s own dose per gallon rather than the recipe’s volume. And measure it properly; this is not a place for eyeballing.

    Two more handling notes that cost me nothing to state and might save a batch:

    • Dilute it in cool, unchlorinated water immediately before use, and stir it in thoroughly but briefly. Chlorine deactivates it. Undiluted rennet hitting one part of the vat sets that part and not the rest.
    • Then stop stirring. Once the enzyme is distributed, agitation during the aggregation phase tears the network as it forms. Still milk, undisturbed, until the clean break.

    What I’d actually do

    Having bought both at the start: for anything going in the cave, chymosin. FPC if I want reproducibility and one less variable while I’m learning something else; traditional animal rennet when I want the older, slightly less predictable version and don’t mind that it varies.

    For fresh cheeses, it genuinely doesn’t matter much — and if you want to make something cardoon-set because the idea of using Columella’s coagulant appeals, that’s a reason. It appeals to me too.

    Next: cultures — the bacteria that do the rest of the work, and the meso-versus-thermo substitution that ruins more wheels than anything else. Being written now.

    Sources

    Comments (0)

    1. Loading…

    Comments are held for moderation — nothing appears until approved.