Cheese Lab

The Cell Biology of Lactic Acid Bacteria

When you pitch a starter into a vat of warm milk, millions of bacterial cells come out of freeze-dried dormancy and start doing what they evolved to do. Understanding what's happening inside those cells — their metabolism, their defences, their by-products — turns starter selection from a recipe step into a design decision.

Abstract · TL;DR
  • LAB split into homofermentative (lactic acid only) and heterofermentative (acid + CO₂ + ethanol/acetate) metabolism.
  • Most LAB used in cheese are proteolytically weak, which is a feature — they acidify without over-breaking-down casein.
  • Bacteriocins like nisin are ancient biological warfare tools LAB use against competitors and mild pathogens.
Fig · LAB · Pitching starterReference plate
Freeze-dried lactic acid bacteria starter culture pellets being tipped from a foil sachet into a stainless cheesemaking vat, steam rising off warm milk

What LAB actually are

Lactic acid bacteria are a functionally defined group — Gram-positive, non-spore-forming, acid-tolerant, and united chiefly by the fact that their principal fermentation product is lactic acid. In cheese you mostly work with four genera: Lactococcus (classic mesophilic starter), Streptococcus (notably S. thermophilus, the thermophilic workhorse), Lactobacillus (starter and adjunct roles, now split taxonomically into many genera), and Leuconostoc (the 'aromatic' LAB that produces CO₂ and diacetyl). The taxonomy has been reshuffled aggressively in the last decade, but the working distinctions remain the same.

Homofermentative vs heterofermentative: the metabolic fork

Homofermentative LAB convert glucose (from lactose) almost entirely to lactic acid via glycolysis. This is what Lactococcus lactis and Streptococcus thermophilus do — they acidify efficiently and cleanly. Heterofermentative LAB run a different pathway (the phosphoketolase pathway) that produces lactic acid plus CO₂ plus ethanol or acetate. This is what Leuconostoc mesenteroides subsp. cremoris does — and it is why Gouda-style cheeses get their small eyes, and why DL-starters have a slightly buttery, complex flavour profile. Neither is 'better'; they do different jobs.

Insight

Citrate metabolism is the other production lever. Leuconostoc and the 'lactis' subspecies of Lactococcus can metabolise the small amount of citrate in milk into diacetyl — the buttery aroma compound that gives Dutch and buttermilk-style cheeses their character. If you want a neutral-flavoured Cheddar, you use a citrate-negative starter; if you want Gouda, you include a citrate-positive strain.

The growth curve in a vat

Pitched cells spend the first 20–40 minutes in a lag phase, repairing freeze-drying damage and adjusting to the warm milk. Then they enter logarithmic growth — doubling every 30 to 80 minutes depending on temperature and strain — producing lactic acid at increasing rates. As acidity climbs and nutrients deplete, they hit stationary phase; many cells eventually die and release their intracellular enzymes into the curd. Those released enzymes (peptidases, esterases) are a huge contributor to aged-cheese flavour. A starter's job isn't done when it stops acidifying — much of its contribution is posthumous.

Proteolysis: quiet but crucial

LAB have a well-studied proteolytic system — a cell-wall protease (PrtP in Lactococcus) cleaves casein into oligopeptides, specialised transporters haul those peptides into the cell, and a suite of intracellular peptidases break them down to amino acids the bacterium uses for growth. The system is tuned to grow the bacterium, not to flavour the cheese — but its by-products (free amino acids, which are flavour precursors) are precisely what aged-cheese flavour depends on. Different strains have measurably different peptidase portfolios, which is one of the reasons 'strain choice' is a real variable in commercial cheesemaking, not just a marketing line.

Bacteriocins: LAB fight back

Many LAB produce bacteriocins — small ribosomally-synthesised peptides that punch holes in the cell membranes of competitor bacteria. Nisin, produced by some Lactococcus lactis strains, is the most famous; it's active against a wide range of Gram-positive bacteria including Listeria and Clostridium and is approved as a food preservative (E234). Pediocin, produced by Pediococcus, is another. In a raw-milk cheese, native bacteriocin-producing strains contribute to the cheese's built-in pathogen suppression — one of the reasons well-made raw-milk cheeses can be safe despite not being pasteurised.

  • Lactococcus lactis: mesophilic, homofermentative, variably proteolytic, sometimes nisin-producing.
  • Streptococcus thermophilus: thermophilic, fast acidifier, limited proteolysis — pairs with Lactobacillus helveticus for Swiss and Italian styles.
  • Lactobacillus delbrueckii subsp. bulgaricus: thermophilic, acid-tolerant, strongly proteolytic — the flavour engine of yoghurt and some Italian cheeses.
  • Leuconostoc mesenteroides subsp. cremoris: mesophilic, heterofermentative, citrate-positive — the source of eyes and buttery notes in DL starters.

Why starter blends exist

A commercial starter is almost never one strain. It is a blend — often half a dozen strains — balanced for acidification rate, flavour profile, phage diversity, and salt tolerance. Blends perform more consistently than single strains and cover each other when conditions vary. The trade-off is that the exact balance is the starter company's trade secret, which is why two 'mesophilic aroma' blends from two suppliers can produce quite different cheeses from identical milk.

§ FAQ

Frequently asked

Why do some starters produce bitter cheese?+

Usually a peptidase imbalance. The cell-wall protease cuts casein into medium peptides that taste bitter; if the intracellular peptidases can't keep up with breaking those down to amino acids, bitterness accumulates. Strain choice, salt level, and aging temperature all affect the balance.

Are probiotic strains useful in cheese?+

Some are genuine probiotics (certain Lactobacillus and Bifidobacterium strains survive the cheese matrix and the gut reasonably well), and aged cheese is actually a decent delivery vehicle. But don't choose a cheese for probiotic content — the numbers are unreliable, and the flavour should come first.

§ Related
§ Related

Back to the library

Continue · Practice

Put this into practice

Design a cheese in the Lab →