Mini WorldLaplace Life & Medicine / Plate 67 — Salivary amylase

Enzymes, Heat and pH

Salivary amylase cuts starch in a cleft of exactly the right shape. Warmth speeds it up until the fold comes apart; a shift in pH changes the charges in the cleft.

Scroll

01

A Cleft That Cuts Starch

Starch is a long string of glucose, each joined to the next by an α-⁠1,4 bond. The α-⁠amylase in saliva cuts the string from the inside, into maltose and dextrins of various lengths.

On the plinth is human salivary α-⁠amylase, built from the atoms of Protein Data Bank entry 1MFV and enlarged sixty million times. The blue is a sugar chain in its cleft: in the crystal the enzyme rebuilt a drug molecule, acarbose, into a chain of six rings that fills the cleft’s six places. Where the chain runs on out of the cleft, it is drawn.

The first enzyme ever isolated was an amylase: in 1833 Payen and Persoz drew it from germinating barley and called it diastase. The “-ase” at the end of most enzymes’ names comes from it.

Human salivary α-⁠amylase (PDB 1MFV) · 496 amino acids · the sugar chain fills 6 places in the cleft

02

Between Sugars −1 and +1

The six places for sugars in the cleft are numbered −4 to +2, and the enzyme cuts between −1 and +1. Three acidic side chains face that spot: Asp197 attacks the carbon of sugar −1, Glu233 hands a proton to the oxygen that comes free, and Asp300 is needed too. A water molecule finishes the job and the bond is broken.

Having cut one bond, the enzyme is unchanged: it lets the pieces go and takes the next. It is a catalyst, not used up by the reaction.

Without an enzyme, bonds between sugars in water at 25 °C would take about 11 million years for half of them to break. Human α-⁠amylase cuts about 175 bonds a second at 37 °C (on a test substrate). That is roughly 10¹⁶–10¹⁷ times faster.

Asp197 · Glu233 · Asp300 · a water molecule (schematic) · about 10¹⁶–10¹⁷ times faster (computed)

03

A Loop Closes In

In 1894 Emil Fischer compared an enzyme and its substrate to a lock and key: the shapes fit from the start. In 1958 Daniel Koshland proposed induced fit: when the substrate binds, the enzyme changes shape and brings its catalytic groups into place.

Salivary amylase has such a part: residues 304 to 310, a glycine-⁠rich loop. With no sugar it hangs loose; when a sugar chain lies in the cleft, it becomes ordered and moves in towards it.

Laid over each other, the crystal without sugar (PDB 1SMD) and the one with it (1MFV) put residue 308 about 4.6 Å apart (computed). The two crystals were grown differently, so the number is only a size.

Loop 304⁠–⁠310 · with sugar (1MFV) and without (1SMD) · residue 308 moves about 4.6 Å (computed)

04

The Shape Is a Fold

The cleft is not carved. The enzyme is one chain of 496 amino acids, one bead to each; only when the chain folds does the cleft appear.

It folds into three domains: A, a (β/α)₈ barrel with the active site at its top; B, with no definite shape, a calcium ion held between it and A to steady the structure; and C, a Greek-⁠key barrel.

A chloride ion sits under the cleft, bound to Arg195, Asn298 and Arg337, and helps the catalysis. What holds the fold together is mostly many weak, non-⁠covalent bonds: hydrogen bonds, attractions between charges, oily side chains packed inside.

One chain, 496 amino acids · domains A, B and C · one calcium ion, one chloride ion

05

Warmer Is Faster

Molecules jostle with heat. A little warmer, and the sugar chain and the enzyme meet more often, and more of the meetings carry enough energy to get over the reaction’s barrier: the reaction speeds up.

By how much depends on the activation energy. For starch cut by pig pancreatic α-⁠amylase, studies give about 49⁠–⁠50 kJ/mol; at that value, going from 27 °C to 37 °C makes the reaction about 1.9 times faster (computed). Reported activation energies run from about 20 to 129 kJ/mol, or 1.3 to 5.3 times. “Twice as fast for every 10 °C” is only a rough rule.

At about 50 kJ/mol: 27 → 37 °C is about 1.9 times faster (computed)

06

Too Hot, the Fold Comes Apart

Warmer still, and the jostling starts to break the weak bonds that hold the fold. The chain loosens, the cleft loses its shape, and the sugar chain no longer fits. Enzymes like salivary amylase do not fold back by themselves.

A measured “optimum temperature” is not one number: the longer the test, the more enzyme is lost at high temperature, and the lower the fastest temperature. In this page’s model the fastest is about 49 °C for a 2-⁠minute test and about 41 °C for a 30-⁠minute one; at 60 °C the fold comes apart in seconds.

Boiling an egg is the same: the proteins in the white unfold and clump, and the clear white turns milky. Ovotransferrin goes first, at about 61⁠–⁠65 °C.

This page’s model: fastest about 49 °C over 2 min · about 41 °C over 30 min · the loose chain is schematic

07

pH Moves the Charges

The enzyme’s surface and its cleft carry many side chains that can gain or lose a proton. Asp and Glu lose one and turn negative; Lys and Arg gain one and turn positive; His sits between. Change the pH and these side chains gain or lose protons, and their charges change.

In folded proteins, the measured pK averages about 3.5 for Asp, 4.2 for Glu, 6.6 for His and 10.5 for Lys. At pH 7 nearly every Asp and Glu is negative; at pH 2 almost none is.

The catalysis needs two groups each in the right state: Asp197, which attacks, charged; Glu233, which hands over a proton, still holding it. Too acid and both hold their protons; too alkaline and both have let them go. So the rate against pH is a bell. Human salivary amylase is fastest at pH 6.9⁠–⁠7.1; the chloride ion raises the pKa of Glu233, helping it hold its proton.

Red: negative (Asp, Glu) · blue: positive (Lys, Arg, His) · schematic

08

Three pHs Down the Gut

Saliva runs at about pH 6.2⁠–⁠7.6, which suits salivary amylase. Swallowed, food reaches a stomach at about pH 1.0⁠–⁠2.5, where pepsin is fastest at about pH 1.5⁠–⁠2.5, and where neutral or alkaline solution would instead inactivate it.

Salivary amylase does not stop at once in the stomach: it is inactivated only once the stomach reaches pH 3.8⁠–⁠3.3, and starch shields it. After 60 minutes at pH 3 and 37 °C it keeps 56 % of its activity with starch, and 6 % without.

In the small intestine the pH is back to about 6.6⁠–⁠7.5, and trypsin from the pancreas is fastest at about pH 7.5⁠–⁠9. Each enzyme’s best pH matches the place it works.

Saliva 6.2⁠–⁠7.6 · stomach 1.0⁠–⁠2.5 · small intestine 6.6⁠–⁠7.5 · pepsin best at 1.5⁠–⁠2.5

09

Built for Hot Springs

Not every enzyme is at its best near 37 °C. In 1969 Thomas Brock and Hudson Freeze found a bacterium in Yellowstone’s hot springs, Thermus aquaticus. Its DNA polymerase, Taq, works fastest at about 75⁠–⁠80 °C, and at 97.5 °C it takes about 9 minutes for half of it to fail.

In 1988 PCR took up Taq: every round heats the DNA close to boiling to part its two strands, and a heat-⁠stable polymerase need not be added again each time.

Washing powders go the other way: proteases, amylases and lipases let a detergent clean at washes as cool as 20 °C.

Taq DNA polymerase: best at about 75⁠–⁠80 °C · half-⁠life about 9 min at 97.5 °C

10

Starch Turns Iodine Blue-⁠Black

The school test: starch solution, amylase and a buffer in a tube in a water bath; every 10 seconds one drop goes onto the iodine on a white spotting tile. While there is starch, the iodine turns blue-⁠black; once the starch is cut up, the iodine stays orange. The time until it stops turning blue-⁠black, turned upside down, is the rate.

The blue-⁠black comes from the amylose helix: iodine and iodide line up inside it. The shorter the chain, the paler: 40 to 130 glucose units give blue, 25 to 40 red, 10 to 25 orange, and fewer than 10 no colour at all.

The usual iodine is 0.01 mol per litre; even dilute iodine stains skin. Alkali takes the blue away, so in strongly alkaline tubes the iodine cannot be read.

A drop every 10 s · blue-⁠black: starch left · orange: the starch is gone

11

The Model Taken Apart

The model comes apart in five: domains A, B and C, the sugar chain from the cleft, and the calcium and chloride ions. A stays on its stand, B and C move off, and the sugar chain and the two ions come forward.

Each domain’s surface is computed again from its own atoms. Rest the pointer on a part and it is lifted out on its own and named.

Domain A · domain B · domain C · the sugar chain · the calcium and chloride ions

12

Timing the Tile

This page’s model: set the water bath’s temperature and the buffer’s pH; a drop goes onto the spotting tile every 10 seconds; watch for the iodine to stop turning blue-⁠black.

The rate climbs with temperature and drops when the fold comes apart; it is fastest near pH 7 and slower to either side. The curves’ shape is illustrative and its numbers are not measurements of human salivary amylase. Every move of a lever, or a press of “Start again”, starts the tubes afresh.

This page’s model · a drop every 10 s · about 60 s at 37 °C and pH 7

Try thisAt pH 7, raise the temperature from 37 °C to 60 °C: how long until the iodine no longer turns blue-black? And back at 37 °C?

What the model leaves out

The protein’s surface is computed from the 3,⁠946 heavy atoms of PDB 1MFV and enlarged sixty million times; it treats each atom as a soft ball, which is close to, not the same as, a true molecular surface, and leaves out the hydrogens. The six sugar rings in the cleft come from the same coordinates: the enzyme built them from acarbose in the crystal, so they are not natural starch; the glucose units outside the cleft are drawn on with the step the chain takes inside it. The domains’ shades, the water, the chain of beads and the loosened chain are schematic, and the loosened chain is not a simulation. The beaker, tubes, thermometer, tile and pH paper are life-⁠size and not at the protein’s scale. The temperature and pH curves are this page’s model: the activation energy is pig pancreatic α-⁠amylase’s, about 50 kJ/mol; the unfolding rate, the two pK values of the pH curve and the 60-⁠second end point are chosen to illustrate, not measured on the human enzyme.

Easy to get wrong

An enzyme is not used up: after each bond it is as it was. Lock and key is only a start; many enzymes change shape as the substrate binds, and the 304⁠–⁠310 loop of salivary amylase is one case. Heat first speeds a reaction up and slows it only once the fold comes apart; an “optimum temperature” also depends on how long it is measured, and is not always 37 °C: Taq is fastest at about 75⁠–⁠80 °C. Salivary amylase cuts starch into maltose and dextrins, not straight into glucose. “Twice as fast for every 10 °C” is a rough rule. pH does not dissolve an enzyme: it changes the charges on its side chains; shielded by starch, salivary amylase keeps over half its activity after 60 minutes at pH 3. The “10¹⁶–10¹⁷ times faster” sets a turnover number on a test substrate at 37 °C against an uncatalysed rate at 25 °C, so it is only a size.

Common questions · 5
Is an enzyme used up in the reaction?

Having cut one bond, the enzyme is unchanged: it lets the pieces go and takes the next.

02Between Sugars −1 and +1
How much faster does salivary amylase make the cut?

Without an enzyme, bonds between sugars in water at 25 °C would take about 11 million years for half of them to break.

02Between Sugars −1 and +1
What is the optimum temperature of salivary amylase?

A measured “optimum temperature” is not one number: the longer the test, the more enzyme is lost at high temperature, and the lower the fastest temperature.

06Too Hot, the Fold Comes Apart
Why does pH change how fast an enzyme works?

The catalysis needs two groups each in the right state: Asp197, which attacks, charged; Glu233, which hands over a proton, still holding it.

07pH Moves the Charges
How is the starch–iodine test for amylase done?

The school test: starch solution, amylase and a buffer in a tube in a water bath; every 10 seconds one drop goes onto the iodine on a white spotting tile.

10Starch Turns Iodine Blue-⁠Black

Sources

  1. RCSB PDB 1MFV: human salivary α-⁠amylase with the sugar chain made from acarbose (2.0 Å)
  2. N. Ramasubbu, C. Ragunath and P. J. Mishra (2003): J. Mol. Biol. 325, 1061 (the loop 304⁠–⁠310 becomes ordered on sugar binding and moves in)
  3. RCSB PDB 1SMD: human salivary α-⁠amylase without sugar (1.6 Å)
  4. N. Ramasubbu et al. (1996): Acta Cryst. D 52, 435 (496 amino acids, three domains, the calcium and chloride ions)
  5. C. Ragunath et al. (2008): J. Mol. Biol. 384, 1232 (the domains; 1MFV’s six sugars in subsites −4 to +2)
  6. S. Numao et al. (2002): Biochemistry 41, 215 (chloride raises the pKa of the acid/base catalyst E233)
  7. UniProt P0DUB6 (AMY1A): the function, products and cofactors of salivary amylase
  8. R. Wolfenden and Y. Yuan (2008): J. Am. Chem. Soc. 130, 7548 (a glucoside’s rate of hydrolysis in water at 25 °C)
  9. BRENDA EC 3.2.1.1: turnover numbers and pH optima of human α-⁠amylase
  10. G. M. Cooper (2000): The Cell, 2nd ed.: enzymes as catalysts (NCBI Bookshelf)
  11. D. E. Koshland (1958): PNAS 44, 98 (induced fit)
  12. J. Miłek (2021): Molecules 26, 4117 (activation energies and optimum temperatures of pig pancreatic α-⁠amylase)
  13. M. E. Peterson, R. M. Daniel, M. J. Danson and R. Eisenthal (2007): Biochem. J. 402, 331 (the measured optimum depends on the assay’s length)
  14. G. Feller, S. d’Amico and C. Gerday (1999): Biochemistry 38, 4613 (mesophilic α-⁠amylases unfold irreversibly)
  15. B. Alberts et al.: Molecular Biology of the Cell, 4th ed.: the shape of proteins (NCBI Bookshelf)
  16. A. Leau et al. (2023): Clin. Transl. Allergy 13, e12312 (heated egg white turns milky)
  17. I. Mannott, R. Bosse and M. Gibis (2026): Compr. Rev. Food Sci. Food Saf. 25, e70495 (ovotransferrin denatures at 61⁠–⁠65 °C)
  18. G. R. Grimsley, J. M. Scholtz and C. N. Pace (2009): Protein Sci. 18, 247 (the pK values of side chains in folded proteins)
  19. T. P. Silverstein (2021): Front. Mol. Biosci. 8, 764099 (the bell-⁠shaped pH curve)
  20. S. Baliga, S. Muglikar and R. Kale (2013): J. Indian Soc. Periodontol. 17, 461 (the pH of saliva)
  21. D. F. Evans et al. (1988): Gut 29, 1035 (the pH of the stomach and the small intestine)
  22. M. Fried, S. Abramson and J. H. Meyer (1987): Dig. Dis. Sci. 32, 1097 (salivary amylase is inactivated in the stomach)
  23. J. L. Rosenblum, C. L. Irwin and D. H. Alpers (1988): Am. J. Physiol. 254, G775 (starch protects salivary amylase in acid)
  24. X. Lin et al. (1993): Protein Sci. 2, 1383 (pepsin is inactivated in neutral and alkaline solution)
  25. T. D. Brock and H. Freeze (1969): J. Bacteriol. 98, 289 (Thermus aquaticus from Yellowstone’s hot springs)
  26. A. Chien, D. B. Edgar and J. M. Trela (1976): J. Bacteriol. 127, 1550 (Taq DNA polymerase works best at 80 °C)
  27. F. C. Lawyer et al. (1993): PCR Methods Appl. 2, 275 (75⁠–⁠80 °C; a half-⁠life of 9 minutes at 97.5 °C)
  28. R. K. Saiki et al. (1988): Science 239, 487 (PCR with a heat-⁠stable polymerase)
  29. P. Christakopoulos and E. Topakas (2012): Comput. Struct. Biotechnol. J. 2, e201209001 (diastase, 1833)
  30. J. Jensen and C. Puissant (2026): Eur. J. Appl. Physiol. 126, 629 (Payen and Persoz isolate diastase from germinating barley, 1833)
  31. Emerging Infectious Diseases 22 (2016): Etymologia: β-⁠Lactamase (the suffix “-ase” comes from diastase)
  32. Royal Society of Biology, Practical Biology: Investigating the effect of pH on amylase activity
  33. AQA: GCSE Biology (8461) required practicals handbook (the amylase test in a 35 °C bath)
  34. S. Pesek and R. Silaghi-⁠Dumitrescu (2024): Molecules 29, 641 (the iodine/iodide/starch complex; colour and chain length)
  35. Ricca Chemical: Tech Tips, Starch (alkali destroys the starch–iodine blue)
  36. CLEAPSS Student Safety Sheet 56: Iodine

Data

Human salivary α-⁠amylase with the sugar chain in its active site: RCSB Protein Data Bank entry 1MFV (N. Ramasubbu, C. Ragunath and P. J. Mishra, 2003); the same enzyme without sugar, for the loop: entry 1SMD (N. Ramasubbu, V. Paloth, Y. Luo, G. D. Brayer and M. J. Levine, 1996), laid onto 1MFV by its Cα atoms. PDB data are CC0. The surface is computed on the page from the atoms (docs/plates/enzyme/bake/bake.mjs, src/topics/enzyme/surface.js). Rates, pH optima and the school test from the sources listed below; the temperature and pH curves of the lab are this page’s model with illustrative parameters (docs/plates/enzyme/computed-⁠rates.py), not measurements of the human enzyme. The model, the chain beyond the cleft, the water, the loosened chain and the props are this page’s own drawing. Material presets and the model kit as on the site’s other model plates (public/looks/CREDITS.md).

The text, pictures and models on this page are licensed CC BY-NC 4.0: free to use and adapt for teaching and other non-commercial purposes, with credit to world.lapla.si. Third-party material listed above keeps its own licence.

Cite this plate

Laplace’s Mini World, Plate 67, “Enzymes, Heat and pH”, https://world.lapla.si/en/enzyme/, CC BY-NC 4.0, updated 2026-10-31.

Embed & API

Life & Medicine · all plates