Sometimes serine is serene – other times it’s sabotaging other proteins by providing the blade to “protein scissors” called serine proteases. That’s really not fair of me to characterize them in a bad light – you can’t blame these proteins for just doing their job. In the body they have lots of crucial jobs – serine proteases are involved in digestion, hormone activation, blood clotting, immune system activation, and much more – and serine’s not acting alone – in one of the most biochemically beautiful examples of atomic cooperation, it serves as part of a catalytic triad, getting help from histidine and aspartate. But it’s all about right place, right time!
It’s Day 20!!!!! of #20DaysOfAminoAcids – the bumbling biochemist’s version of an advent calendar. Amino acids are the building blocks of proteins. There are 20 (common) ones, each with a generic backbone to allow for linking up through peptide bonds to form chains (polypeptides) that fold up into functional proteins, as well as unique side chains (aka “R groups” that stick off like charms from a charm bracelet). Each day I’m going to bring you the story of one of these “charms” – what we know about it and how we know about it, where it comes from, where it goes, and outstanding questions nobody knows.
We’ve talked a lot about how our bodies can recycle amino acids and use them (or parts of them) to build new molecules – other amino acids and even other things like sugars & fats. But in order to recycle individual amino acids you have to “individualize” them – you need to cut up the big ole proteins into single letters and process them individually.
This involves breaking up those super-sturdy peptide bonds linking up the amino acid letters. The key is to make them less stable and more irresistible to attackers. The carbonyl C (carbon double-bonded to oxygen) is the most vulnerable because, although it might not look like it, it’s partly positive because the O is hogging the electrons they’re sharing. So the C is happy to add some negativity (it’s electrophilic) if a positivity-seeker (nucleophile) comes along. And, since each atom can only form a limited number of bonds, you end up with some electron rearranging leading to the peptide bond being broken.
So you can break a peptide bond through nucleophilic attack – so now we just need a strong nucleophile. If you deprotonate (remove a proton, H⁺ from) an alcohol (-OH) group you get -O⁻. This O is negative (fully this time because it has more electrons than protons, they’re not just unevenly distributed) and really nucleophilic. That alcohol group can be part of water, but it doesn’t have to be – it can, for example, be sticking off of a protein in perfect position to attack…
So where in proteins can we find a hydroxyl group? Given that you know that today’s post is about serine you probably won’t be surprised if I tell you serine has one! Serine (Ser, S) is the alcohol version of Cysteine (Cys, C) – they only differ by 1 atom (Ser has a hydroxyl (-OH) instead of a thiol (-SH)) after their methylene (CH₂) linker. Unlike the glutamate (Glu, E) or the aspartate (Asp, D) we looked at, serine is *NOT* usually deprotonated. So it’s normally neutral, but it does have the potential to lose a proton to give you an alkoxide anion (-CH₂-O⁻) (note – we call proton-donors acids).
The reason it’s *not* usually deprotonated but those other guys are is that Ser has a much harder time handling the negative charge that comes with losing a proton (and thus throwing off the balance of protons (positively-charged) and neutrons (negatively-charged)). Glu & Asp are able to handle the negative charge well, even embrace it, because they can share the extra electron through resonance, aka electron delocalization – and they can’t get that stabilizing effect if they’re protonated. Plus, the -OH in Glu & Asp is part of a carboxylic acid group, meaning it’s next to a carbonyl (C=O), which draws e⁻ density away from the hydroxyl O, making the O more positive and thus happy to get rid of a positive proton.
But Ser is next to a methylene group, which doesn’t draw away electrons – it actually does the opposite – groups like this are e⁻ *donating*, so it makes the O more negative, so it wants to keep H⁺. Plus, it doesn’t have that resonance bonus if it deprotonates, so it holds on tighter to its proton and doesn’t normally act as an acid unless conditions are very basic (there aren’t many free H⁺ around)
But basicity is in the eye of the beholder! We normally describe and compare acid strength (willingness to give up a proton) with a value called pKa which which tells you the pH at which half of the thing will be protonated – above the pKa, there are fewer protons around, so the thing is more likely to deprotonate whereas below the pKa there are more protons around, so it’s likely to be protonated. Protonation is reversible so the deprotonated form can “change its mind” – act as a base and take a proton – thus we can call the deprotonated form of an acid its “conjugate base” and, by comparing the pKa to the pH you can predict which form will dominate.
You can look up charts of pKas for the various amino acids that have protonatable/deprotanatable forms. And if you do this you’ll see that Asp has a pKa of ~ 3.7 & Glu has a pKa ~ 4.3 (Glu). What about serine? It’s hydroxyl pKa is so high (~13) that isn’t usually even listed! So how the heck is this thing ever going to be deprotonated physiological (normal bodily) pH, which is around 7.4?!
Here’s were that context thing I’ve been stressing comes into play – the values you see in the chart are usually for free amino acids – but when amino acids are around others, all bets are off – and the protons might be off as well… at least temporarily! In the active site of serine proteases, Ser’s hydroxyl H gets yanked off – but don’t worry – there’s a substrate carbonyl in place to comfort it and eventually it gets its proton back! so it can do it all over again… We’ll look in more detail later on in the post.
A lot of the amino acids were discovered via acid hydrolysis – scientists used acid to help break up proteins to try to figure out what was in them – and they did that in beakers and test tubes. Your stomach also uses acid (hydrochloric acid) to help break down proteins, but only enough to help denature (unfold) them to make them easier for the real cutters to go to work – these cutters are enzymes called proteases. Enzymes are usually proteins (sometimes protein/RNA complexes (like the case with ribosomes) or just RNA (we often call such enzymes ribozymes) and they mediate and speed up reactions by doing things holding the molecules together in the right positions for whatever they need to do and providing an optimal environment for the reaction to take place.
Quick terminology note – “protease” and “peptidase” are often used interchangeably but technically a peptidase specializes in cutting short chains of amino acids, whereas a protease specializes in cutting long chains of amino acids (which in their folded form are better known as proteins). “Endo”protease/peptidase refers to ones that cut in the middle of peptide chains and exoproteases/peptidases chew off the ends.
Pepsin (which is a protease but not a serine one) does some preliminary protein snipping in the stomach and then, in the intestines, the bulk of the protein cutting takes place, thanks to the enzyme trypsin (which is excreted by the pancreas and *is* a serine protease) – trypsin chops proteins into a few letters long, then more enzymes in & on the surface of intestinal cells finish the job, freeing up individual amino acids.
Trypsin is just one of many serine proteases, all with the same core MO. Basically they all have this conserved catalytic triad of a serine (which directly interacts with the peptide) helped out by a histidine (His, H) & an aspartate (Asp, D), which, through a “charge relay” help get the serine to give up a proton and become more reactive (we’ll get more into this below). Around this conserved core, evolution (as allowed for by gene duplication giving natural selection multiple copies to work with and random mutation providing variation for it to act on) has “played around” with the surrounding parts, allowing different serine proteases to specialize in cleaving at different spots on proteins, and adapting to work in different environments.
I thought it would be especially fitting today, as the last *official* day of #20DaysOfAminoAcids to discuss serine proteases because the catalytic triad is one of the most beautiful examples of amino acids working together to get a job done! Here’s an overview of what we’re gonna see – when serine proteases cut, half of the thing they cut gets “stuck on them” so you need a second cutter to cut it off. So they use a “ping-pong” mechanism. First, the serine gets activated so it cuts the peptide bond. Now half of the peptide’s stuck to the protease & 1/2 is released. But you can’t use the serine to cut off the stuck part because the serine’s what it’s stuck on! So it brings in a water molecule. And it activates it. Then the activated water frees the serine to do it again.
In order to do all that activating (of serine & water), you need the help of the triad’s other members, His & Asp. You don’t want to activate Serine until go time, so all’s calm until substrate binds to a specificity-determining pocket. This gets the protein to shift a little and this makes Asp’s O- move closer to His, drawing away His’s H, so it’s more willing to steal Ser’s H, giving you an “alkoxide ion”
And, like we discussed before, Ser really doesn’t want negative charge, so it goes on the nucleophilic attack, attacking the carbonyl carbon of the peptide bond in the substrate. And this might make Ser happy, but it kinda throws everything else out of whack – carbon’s left with a really unstable tetrahedral (bound to 4 things) intermediate – it’s gonna break – but the key is to break it the “right way” – and this is helped out by a “side-pocket” of the enzyme called the oxyanion hole – it helps stabilize this unstable intermediate so that it splits the right way – breaking the peptide bond between the carbon that was the carbonyl carbon and the nitrogen, with the nitrogen snatching off the proton that His stole from Ser on its way out.
Congrats – you’ve just cut the peptide & released the first part. But, before you start celebrating… you still have the other part stuck on Ser… So now we have to get that off. And this is the “slow” part – because we have to get water in and activated. The His is going to do the activating again, just like it did with Ser (remember that the peptide took the proton from His that His took to Ser, so there’s now room for another proton there!). So the nitrogen of His snatches a proton from water, giving you a hydroxyl ion, which for reasons akin to the reasons the alkoxide ion attacked the carbonyl C, will nucleophilically attack the carbonyl carbon. So you get another tetrahedral intermediate that gets broken by Ser’s O taking a proton from His.
It sounds like you could make a good kids’ book or song from it…. Ser steals a proton from His that His stole from water to replace the proton that the N-terminus stole from His that His stole from Ser! And with this, the C terminal is released and the enzyme is back to its starting state so it can do it all again. Yes, *now* you may celebrate! (although, who am I kidding – I say celebrate the small things!)
It’s a testament to the efficiency and biochemical beautifulness of this triad that so many enzymes throughout evolution in species far and wide have conserved it (and it’s even evolved separately a few times). But, if all these enzymes cut the same way, how do we bring in specificity? Kinda like how you can’t use a paper shredder to cut an encyclopedia because you stuff an encyclopedia into the opening slot of a paper shredder (and please please don’t try! encyclopedias don’t deserve that!), where a protease cuts has to do with how its specificity pocket is shaped and how welcoming the side chains sticking out into it are to different types of substrates (the thing that has to bind & get cut).
For example, trypsin has negatively-charged aspartate jutting out into it, attracting positively-charged amino acids, so it cuts the carboxy side next to lysine (Lys, K) & arginine (Arg, R). Chymotrypsin, on the other hand has a big, hydrophobic pocket making it great for cutting next to big bulky amino acids – tryptophan (Trp, W), phenylalanine (Phe, F), tyrosine (Tyr, Y), and leucine (Leu, L). Elastase has a really tiny pocket, so it can only cut chains with small amino acids – alanine (Ala, A), glycine (Gly, G), & serine (Ser, S)
Those guys are pretty generic, which is good for places like the intestines where you’re trying to break down a lot of different proteins (proteases themselves are super hardy in order to withstand the constant threat of scissoring). But if you were to release those generic-ish proteases into a cell, all would go to hell!
Of course, this raises the challenge of making the proteases inside cells and shipping them to where they need to go to work all without them chopping things up along the way – the solution? – don’t activate them until they arrive at their destination – digestive enzymes are made as inactive precursors called zymogens that have extra amino acids that have to be clipped off before the enzyme is active – clipping which is done by other proteases. For example, trypsin is made as the inactive trypsinogen, which travels safely to the intestines where it gets activated by another protease, enteropeptidase, which cuts off its tail, activating it. And then trypsin can turn the zymogen chymotrypsinogen into chymotrypsin and proelastase into elastase. A similar strategy is used with some hormones that are made as inactive prohormones and activated by cleavage.
Those cleavages are a lot more specific, which is possible because there are specialized proteases which have a lot pickier active sites. An example of this is thrombin – it’s tasked with cutting a protein called fibrinogen, helping blood to clot. Thrombin recognizes the sequence Leu-Val-Pro-Arg-Gly-Ser & cleaves between Arg & Gly.
These more specific proteases are often used in recombinant protein expression, where we stick a gene for a protein into cells (often bacteria or insect cells) and have them make the protein for us – if we add extra DNA letters onto the gene we can get the cells to add extra protein letters onto the protein, which we can use as a tag to help us purify them. And if we insert a protease cleavage site in between the tag and the protein we can use the matching protease to cut off the tag once we’re done using it so we can study the tagless protein.
Even the “generic-ish” ones like trypsin & chymotrypsin are useful in the lab. In mass spectrometry (mass spec) they’re used to cut proteins up into short peptides as part of the process of determining their identity. And they’re used in lower amounts for “limited proteolysis” where you add a little bit and track protein digestion products over time to try to figure out which regions are the most flexible, accessible, and thus easiest to cleave. This can be really helpful for identifying protein structural domains (parts of the protein that have distinct, stable, 3D shapes).
In addition to restricting where proteases cut (both where in the body and where in the peptide), our bodies have another cool way to prevent unwanted peptide cutting – SERPINS – they’re not snakes, they’re “serine scissor safety guards.” Their name comes from SErine PRotease INhibitors and they act as “fake substrates” that get stuck in the protease when it tries to cut them – it’s able to cut them but it can’t release them because, once cut, the serpin can (still bound to trypsin) shape-change, drag trypsin along with ti, and move the trypsin’s active site out of whack in the process.
An example of a serpin you might have heard mentioned in commercials for emphysema treatments is alpha-1-antitrypsin – in the bloodstream, it protects tissues from the serine protease elastase, which a type of immune cells called neutrophils secrete into sites of inflammation to break down connective tissue so that blood cells have an easier time getting in to go to work repairing damage. alpha 1-antitrypsin acts as a sort of “moat” that limits elastase’s area of action so it doesn’t just go chewing up connective tissue throughout your body. But smoking can modify this serpin so that it fails to do its job adequately, so the moat is breached and elastase is able to chew up the lungs. Some people have a genetic alpha-1-antitrypsin defficiency, so they’re more at risk for emphysema even in they don’t smoke. alpha 1-antitrypsin is just one example of more than 30 serpins have been discovered in humans. A couple others are antithrombin, which keeps thrombin in check during clot formation and antiplasmin, which inhibits plasmin so that clots can be dissassembled.
As you might expect based on there being so many, they’re specialized to work on different proteases so your body can inhibit one without screwing up the activity of others. But in the lab, when we are purifying proteins and want to keep them from getting chewed up along the way, we want to prevent all the chewers, we don’t care about specificity. So we rely on “simpler,” much more generic, types of serine protease inhibitors.
The biggest place this comes into play is during the cell lysis part of the protein purification -lysis just refers to splitting up and in this case we’re splitting up the cells containing the proteins – (hopefully) *not* the proteins themselves! This exposes the cell insides to whatever’s outside and whatever’s inside the little membrane-bound compartments (some of which contain proteases – this membranous segregation is another way to protect unwanted cutting). Normally the proteins inside cells don’t get in contact with those unless they get an “escort” but when you break open the membranes, you lose that protective barrier. So we want to make sure that as soon as they’re released, they’re inactivated. So we add protease inhibitors to the solution (lysis buffer) that you resuspend the cells in before breaking them open.
We need something that will inactivate the proteases but not hurt our protein in the process. This can be especially tricky because the things we want to destroy are made up of the same stuff as the thing we want to save (amino acids). How can we stop it? We can add something to hide what the protease needs. And a serine protease needs serine. Hide that serine, and it’s like sticking a safety guard on the scissor’s blade. A safety guard we commonly turn to is PMSF (phenylmethylsulfonyl fluoride) which sticks its PMS onto the Ser’s O.
activesite-OH + PMSF -> active site-PMS + HF
PMSF doesn’t “sulfonylate” all the serines, just the ones in the active site, where there’s the perfect environment for it – the same conditions that make it easier to swap serine’s H for a protein part make it easier to swap it for a sulfonyl part. And, unlike removable safety guards, this guy gets glued on with a strong, covalent bond, making it “irreversible.” You can’t just use water to cut it off because the active sit isn’t built for that – it’s exquisitely made so it’ll only cut peptide bonds.
This doesn’t protect are proteins from all proteases – there are different kinds of proteases and they don’t all rely on serine. In addition to the serine proteases, there are cysteine proteases (similar except they use a thiol (SH) instead of an alcohol (OH)); aspartic proteases; and metalloproteases. The Ser & Cys ones use an amino acid for initial nucleophilic attack while the others use water which they activate through aspartate amino acid residues or metals. These different proteases have different “Achille’s heels,” so we often use “protease inhibitor cocktails” which are just a pre-mixed mix of inhibitors.
note: we make our own but sometimes the cocktails you buy pre-made have EDTA in them (but there are also EDTA-free ones). If you’re using IMAC (immobilized metal affinity chromatography), you don’t want EDTA! It’ll de-nickelize your nickel column! And you also don’t want to use EDTA if your protein needs metal.
In addition to having lots of different proteases, our bodies also have enzymes for breaking down sugars, fats, nucleic acids (DNA/RNA), etc. Amylases and friends help chop carbs, nucleases nudge apart nucleotides, lipase unlatch lipids, etc. This allows us to break down molecules for energy and/or parts, get rid of molecules that have extended their welcome or are damaged, and prevent molecules from entering our cells that were never welcome in the first place (e.g. invading microbes).
As is fitting for something so beautiful, serine gets its name from the Latin word for silk, sericum, because it was first found there (by Emil Gramer in 1865). Like threonine, which we looked at last week (as well as tyrosine), serine can get phosphorylated by kinases (have a negatively-charged phosphate group stuck on it) which can alter the shape and/or function of proteins, change who they hang out with, etc. In fact, serine is the most commonly phosphorylated amino acid, with the phosphorylated form called “phosphoserine”
Serine is nonessential in the dietary sense – our bodies can make it, one way is from the glycolysis (glucose-breakdown) intermediate 3-phosphoglycerate) via transamination (and a couple other steps). It can also be made from glycine with help from serine hydroxymethyltransferase (SHMT). Conversely, it can be used to make glycine – and cysteine – and it serves as a precursor for a lot of other important biomolecules including DNA & RNA letters.
how does it measure up?
systematic name: 2-Amino-3-hydroxypropanoic acid
coded for by: UCU, UCC, UCA, UCG, AGU, AGC
chemical formula: C3H7NO3
molar mass: 105.093 g·mol−1
link to PDB-101 page with structures: http://pdb101-beta.rcsb.org/motm/46
This post is part of my weekly “broadcasts from the bench” for The International Union of Biochemistry and Molecular Biology. Be sure to follow the IUBMB if you’re interested in biochemistry! They’re a really great international organization for biochemistry.