Food & alt-protein
Enzymatic approaches to gluten
Why the immunogenic gluten peptides survive digestion, the role of transglutaminase-2 deamidation and HLA-DQ2/DQ8 presentation, how prolyl endopeptidases work — and why the clinical evidence does not support them as a treatment.
Coeliac disease is an immune response to specific gluten peptides in genetically susceptible people. Understanding why enzymes are an attractive idea, and why they have so far failed as a therapy, requires following the peptide.
Why gluten survives digestion
Gluten proteins are unusually rich in proline and glutamine. Proline is a structural oddity among amino acids: its side chain loops back to the backbone nitrogen, creating a rigid kink. Human gastric and pancreatic proteases — pepsin, trypsin, chymotrypsin — and the brush-border peptidases all cleave poorly next to proline, and humans have no prolyl endopeptidase in the digestive tract.
The result is that gluten digestion stalls, leaving long peptides intact where other dietary proteins would have been reduced to amino acids and di- or tripeptides. The best-characterised is a 33-residue fragment of α-gliadin, which contains several overlapping copies of the immunodominant epitopes and is essentially resistant to further breakdown.
What happens next
These peptides reach the lamina propria. There, tissue transglutaminase 2 deamidates specific glutamine residues, converting them to negatively charged glutamate. That single chemical change matters enormously: the HLA-DQ2 and DQ8 molecules that confer susceptibility have binding pockets that strongly prefer negatively charged residues, so deamidation dramatically increases the affinity with which the peptide is presented to T cells.
The resulting T-cell response drives the mucosal damage. This chain — proline-rich peptide survives digestion, transglutaminase deamidates it, DQ2/DQ8 presents it — explains both why only some people are affected and why the offending molecule is so specific.
The enzymatic idea, and its limits
If the peptide’s survival is the problem, degrading it before it reaches the intestine should prevent everything downstream. Prolyl endopeptidases, from Aspergillus niger and other sources, cleave after proline and can degrade gluten peptides in vitro. Some are acid-stable enough to work in the stomach; combinations pairing a prolyl endopeptidase with a glutamine-specific protease degrade gluten faster than either alone.
The clinical evidence does not support these as a treatment for coeliac disease, and this should be stated without hedging. Trials in which participants consumed meaningful gluten while taking such enzymes have generally failed to show protection of the intestinal mucosa. The reasons are mechanistic: gastric emptying moves food onward faster than the enzyme can process gram quantities of a protein embedded in a food matrix, dose-for-dose capacity is far below the gluten content of ordinary meals, and incomplete degradation still leaves immunogenic fragments.
The defensible claim is much narrower — potential assistance with inadvertent trace exposure alongside a strict gluten-free diet, not permission to eat gluten. Products marketed as digestive aids in this category are not therapies, and coeliac disease is currently managed by gluten avoidance.
Other approaches target later steps instead: transglutaminase-2 inhibitors to block deamidation, and agents acting on the immune response itself. These are investigational, and results to date are early.