Blessed are the cheesemakers

As metal arms twirl across a vat of liquid whose buttery yellow colour hints at its destiny, bacteria, enzymes and proteins are doing their own dance beneath the surface. As a well-washed and sanitised hand reaches in, small particles emerge from the pleasant warmth of this molecular waltz. ‘Our aim is to make world-class cheese,’ explains Mary Quicke, managing director of Quicke’s Traditional, a small dairy based in Newton St Cyres, a village in the south west of England. And although producing the tastiest cheeses is an art, success depends on mastering complex biochemical processes like those in the vat.

For Quicke, that starts with milk from cows on her farm, bred and fed specifically for cheesemaking, ensuring the best balance between fat and protein. Careful breeding also means her cows’ genes boost cheese output by coding for optimal structures in key casein milk proteins. In milk, casein proteins clump around calcium and phosphate ions, crucial for baby mammals’ bones and teeth but otherwise insoluble, forming a colloidal suspension of easy-to-swallow spherical micelle packages.

One particular casein protein stops the cheese party starting in milk, explains Paul McSweeney from University College Cork in Ireland. ‘The protein at the outside of the micelle, k-casein, protects the colloid from destabilisation, stopping other caseins precipitating in the presence of calcium,’ he says. ‘Casein micelles have net negative charge, so when two come together, like charges repel, and they part. The carboxy-terminal regions of k-casein strands also form a physical barrier that looks like protruding hairs. But they’re densely packed, so they scatter light, making milk look white.’

A casein point

In Quicke’s vat, this arrangement has broken down and become curds and whey, on its way to cheddar. If milk is left alone, bacteria quickly start converting its lactose sugar into lactic acid that can eventually start this curdling. This is probably how cheese was first made, but modern needs for safe storage and maturing demand a different approach. Quicke’s minimally pasteurises its milk, and like most modern cheesemakers adds a starter culture including lactic acid bacteria such as Streptococcus, Lactococcus and Lactobacillus. They work hand-in-hand to control which bacteria reach the final cheese, by outcompeting less welcome species and making the environment too acidic for them. And rather than these bacteria doing the curdling, acid conditions help an enzyme preparation known as rennet to do it. Their ongoing acidity development also controls the resulting solid curd’s texture.

Traditional rennet, which Quicke’s uses, comes from soaking a milk-fed calf’s stomach in brine. ‘Usually the enzyme is chymosin, and it cleaves the phenylalanine-105–methionine-106 bond in k-casein, about two-thirds of the way along the molecule,’ notes McSweeney. ‘It removes the carboxy-terminal region, shaving the hairs, and reduces the negative charge dramatically.’ In the freshly shaven micelles, casein’s strong self-assembly properties are liberated. The micelles form chains as they naturally bump into each other, at first small, and then larger and overlapping. ‘Eventually that creates a gel network with water and fat trapped in it, about 10 times as much fat as milk,’ McSweeney says. ‘And as cows transfer fat-soluble red-orange carotenoids from grass into their milk, that colour now becomes more apparent. Then during cheesemaking the gel is cut. It’s almost like a weak jelly, but when casein gel is cut, damaged or heated it undergoes syneresis, contracting and shrinking down on itself, squeezing out liquid. The liquid is the whey, the solid material is the curds.’

These names are linked to the nursery in more than just a rhyme, as chymosin’s biological role is stopping milk moving through baby mammals’ stomachs too fast. ‘If you’ve ever seen baby vomit, it’s curds and whey, effectively,’ McSweeney adds apologetically.

Worth its salt

Though salt promotes syneresis, creating high ionic strength that helps protein strands come together, governments are looking to set legal limits on its level in our food. Marianne Hammershøj from Aarhus University in Denmark is therefore examining alternatives.

‘The major challenges are cheese texture and structure,’ she says. ‘It becomes very soft and sticky if you remove or reduce the salt. Yellow cheeses break cleanly, but if you remove salt it becomes very elastic and doesn’t really break. The bad guys here are sodium ions, and you can easily get the structure with other ions, but that gives different flavours. The bitterness of potassium chloride is very dominant, so you can’t add very much. You can try to cover that flavour, but it adds more E-numbers, which isn’t usually what the dairy industry or consumers want.’

As the ‘A Pinch of Salt’ project leader Hammershøj is mapping how much salt concentrations can be reduced while still making acceptable cheese. ‘The aim is not to go all the way to zero but to reduce it so that the flavour and texture are OK, and the microbiology is under control,’ she says. ‘We are also trying out other starter cultures that look quite positive. Some of these may assist the casein network’s structure and cheese flavour. We’re also going to look into treating milk so the micelles make a much firmer network when you reduce the salt.’

At Quicke’s, a cheesemaker puts a glass pipette into his mouth, sucks out some whey and titrates to track lactic acid development. At the correct acidity, he drains the vat into a cooler on a lower level and removes the whey, taking most of the remaining lactose with it. The remaining curds are shaped and coloured like popcorn. As they’re scooped to the cooler’s sides, they continue coagulating into one giant lump that the cheesemakers then cut into strips and pile into a stack.

Breaking up the curd helps remove more whey, giving a firm but silky texture, Quicke explains. ‘This is called hand cheddaring. We’re squeezing out moisture, but not all of it, because then the lactic acid bacteria wouldn’t develop. By the time they’re done the strips will have the texture of a cooked chicken breast. Then, at the right acidity, we mill them into chip-sized pieces, and mix salt in, 1.8% by weight. That stops the lactic acid bacteria, removes some moisture, and adds to the flavour. We scoop them into a mill, which cuts them up, pump it down and back out again, and then consolidate into presses.’ Keeping more whey with the curds gives softer cheeses their moist and open texture. That allows mould growth, useful for blue-veined cheeses like stilton, but which can also lead to faster spoilage.

Though the frantic back-and-forth of casein rearrangement is largely over at this stage, the cheese is disappointingly bland. But the smell inside Quicke’s ‘cathedral of cheese’, a high-ceilinged warehouse where freshly-made cheddar ripens for up to two years, is intense. ‘Back in the dairy, tomorrow morning they’ll bump out today’s cheese and dunk it into hot brine to form the rind,’ she says. Then the cheesemakers will wrap it in muslin and lard and return it to the press, repeating the process for two more days before leaving it to mature. ‘That allows the cheese to breathe and lose moisture, which primarily drives maturing. Here the cheeses have lost some lard and gained populations of mould on the outside. In particular there’s a bug called Brevibacterium linens typically found on the rind of traditional cheddar.’

To brie or not to brie

Together, these secondary flora and the starter bacteria trigger the slow dance that breaks down cheese, developing flavours and textures – ones we treasure if done right. In cheeses like camembert and brie, white surface moulds, such as Penicillium camemberti, break down residual lactic acid. That deacidifies and helps to soften the cheese’s core by pulling calcium phosphate towards the surface. Harder cheeses age for months or even years, bacteria consuming the remaining lactose early in the process, making them suitable for even the lactose-intolerant.

Regardless of variety, cheeses typically contain the same flavour compounds, but in very different proportions. McSweeney estimates that in cheddar 300–500 compounds pass the threshold where they contribute flavour. The diversity comes partly because acidity and salt content make it hard for bacteria, both welcome and unwelcome, to live in cheese. This kills starter bacteria, bursting them open. That releases peptidase enzymes that can then cut casein proteins down to their constituent amino acids, which then break down further to produce flavour compounds. Methanethiol, dimethyl disulfide and dimethyl trisulfide, produced from methionine, are among the most important in cheddar.

Fat is also converted into flavour compounds. Cow’s milk fat is rich in short chain fatty acids like butyric acids, whose name comes from the Latin for butter, butyrum. Goat’s milk contains C6, C8 and C10 fatty acids, contributing to flavour differences between cheeses from the two animals. In stilton, fatty acids are converted to methyl ketones, mainly heptan-2-one and nonan-2-one, responsible for its distinct peppery ‘blue note’ flavours, says Christine Dodd from the University of Nottingham in the UK. And while P. roqueforti, the mould in the veins of blue cheese like stilton and roquefort, has long been known as important in this conversion, Dodd’s team found another organism also plays a crucial role.

That discovery came because secondary flora’s fickleness can be a threat as well as a boon to cheesemakers. ‘Stiltons are like babies: they’re all made the same way but they all come out different,’ Dodd jokes, quoting Liz Whitley, executive director of the International Dairy Federation. ‘Even cheeses from the same batch can come out with slightly different flavours and blue vein development. This is where the art of cheesemaking takes over from science. But it means you can occasionally get batch failures, as some can fail to blue completely, or give very poor blueing.’

Cheese blues

Seeking an explanation, Dodd and her teammates identified stilton’s entire microbial population using DNA profiling, looking at the rind, blue parts and white parts separately. The different areas hosted quite dissimilar groups of moulds, bacteria and yeasts, and when they tested their flavour profiles by gas chromatography–mass spectrometry, they found they differed significantly too. ‘We then made little model cheeses in the lab,’ she says. ‘We combined known organisms put into stilton, the lactic starter and P. roqueforti, with the individual components that we found. Certain yeasts could suppress mould growth completely, so we never got blueing. Some yeasts, particularly Yarrowia lipolytica, gave much higher ketone levels than just Penicillium and Lactococcus together.’

Dodd’s team thinks yeasts thrive off breakdown products spilled when other microorganisms die. ‘These organisms get unpredictably introduced into the cheese, and that’s a key thing about secondary flora – they’re not necessarily there every time,’ she emphasises. ‘If they’re not, you get a different product. Some cheesemakers see that as exciting and part of the art. But if you’re a large producer, knowing which organisms are responsible could help you to get better control.’

Another ripening problem can arise when bacteria contain decarboxylase enzymes, which turn histidine, tyrosine and tryptophan into histamine, tyramine and tryptamine. ‘They produce allergy-like responses; think antihistamine,’ says McSweeney. ‘A few years ago, a small-scale producer was selling a very strong extra-mature cheddar into a supermarket. People were getting a rash when they ate it, and that’s what caused the problem. Other amines, like putrescine and cadaverine, give you “off” flavours, as the names suggest. It’s not so much of a problem now because the industry has put a lot of effort into taking starter strains with high decarboxylase activity out of the system.’ With tyramine sometimes blamed, albeit without peer-reviewed evidence, for cheese giving people nightmares (see box below), perhaps these steps will dent that legend.

Eliminating decarboxylase is just one example of many that underline the high-tech power cheese producers today can wield to create their products more reliably. ‘For example, rennet means cheesemakers are the second largest user of industrial enzymes in the world after detergent makers,’ McSweeney says. ‘Much cheddar is produced by fermentation-derived chymosins, where they cloned the calf chymosin gene into carrier organisms, the same way they produce recombinant insulin. It’s an 8000 year old product, but if someone invented cheese now it would be hailed as a triumph of the biotechnology industry.’ 

Andy Extance is a science writer based in Exeter, UK