PKU: PhenylKetonUria biochemistry & history of its discovery, detection, & treatment

[@thebumblingbiochemist]

Phe’s aromatic part is based off of a benzene ring, which is the simplest aromatic hydrocarbon, consisting of 6C & 6H (thus having the formula C₆H₆). This benzene ring is hard to break - but if you add an -OH (hydroxyl) group to it makes breakdown easier, especially with the help of dedicated enzymes. Adding an -OH does more than just aid in breakdown - when that -OH is attached in the “para” position (directly opposite the attachment point) it gives you a whole new amino acid - tyrosine (Tyr, Y). In addition to being used to make proteins, Tyr is a precursor to catecholamine hormones (think dopamine, adrenaline (aka epinephrine), and noradrenaline (aka norepinephrine)) & thyroid hormones. It’s also a precursor to the skin pigment melanin - and since people with PKU have defects in the enzyme responsible for the -OH adding (hydroxylation), they’re often fair-skinned. That enzyme, PAH (when functional) is powerful. But even superheroes need help, and PAH gets help from a cofactor (non-protein helper molecule that binds to it) called tetrahydrobiopterin (THB or BH₄). BH₄ gets modified in the process and has to get recycled, so in order to break down Phe, you need working PAH and working BH₄ recycling, and a problem with either of those can lead to elevated levels of phenylalanine (hyperphenylalaninemia or HPA). If HPA is due to defects with PAH, it’s called PKU and if it’s due to other causes, like mutations in genes involved with BH₄ recycling problems, it’s referred to as “non-PKU HPAs” Even within “true” PKU, there’s a lot of variability because there are hundreds of different (known) mutations that have been found in PAH. Some just “weaken” the enzyme, some shut down production altogether, etc. And the disease is autosomal recessive meaning it’s on a non-sex-chromosome (autosomal) and you need 2 defective copies to show symptoms - so patients are usually “compound heterozygous” - they have different mutations in each copy. Some patients with mild PKU associated with low levels of PAH activity can benefit from BH₄ supplementation - this is the rationale behind the PKU drug Kuvan, which is a stabler version of BH₄. It’s not fully known why PKU mainly affects the brain - especially given that most of it is in the liver. There are lots of theories and it probably involves a combination of them. One thing is that all the big bulky amino acids have to compete for the same brain receptor to get into the brain (you might remember this from my post on the turkey makes you sleepy myth bit.ly/2DYj647 ). So having too much Phe might outcompete the other amino acids that need to get in, so your brain doesn’t get enough of them. And, once in the brain, what’s there for Phe to do? Normally about half of the Phe you eat is turned into Tyr. But if you can’t turn it into Tyr it’ll just build up. And your cells don’t want that, so some of it goes through an alternative pathway in which it donates its amino group through a process called transamination. This amino group pass off occurs via an intermediary - Phe passes its amino group to 2-oxoglutarate (aka α-keto-glutarate) to make glutamate which can then pass it off to other things and/or send it through the urea cycle for disposal as urea. α-keto-glutarate has a lot of jobs though. It’s a key intermediate of the energy-producing citric acid cycle (aka Krebs cycle aka TCA). So dealing with the Phe overload might deplete the α-keto-glutarate stock, shutting down energy production When Phe gives up its amine group it doesn’t just let it go, it swaps it out for a double-bonded oxygen. So now you have a carbonyl (-C=O) sandwiched between carbons on either side. We call such an arrangement a ketone. So phenylalanine has become a phenylketone. This ketone’s called phenylpyruvate. That ketone can be reduced - have that =O turn into an -OH- to give you phenyllactate. Or it can be decarboxylated (lose CO₂) to give you phenylacetate. These phenylalanine “metabolites” were the key to discovering PKU.

[@thebumblingbiochemist]

And here’s the story: It starts in Norway in the 1930s - a mom named Borgny Egeland had 2 kids - a 4-year-old boy named Dag and his 6 and a half year-old sister Liv. They had severe developmental delays. And an unusual symptom - musty-smelling pee. Borgny was desperate to find answers and help for her children and, after going to a lot of doctors who didn’t have much biochem background, she struck gold - around 1934 she went to see a doctor at the University of Oslo named Asbjörn Fölling. She went to him because she knew of his biochemistry background - and this background would prove crucial. I want to emphasize this because I know a lot of my followers are pre-med - so I hope this inspires you to see the value in biochem you’re “forced to take” Anyways, her husband Harry had taken a course from Dr. Fölling in dental college and knew he was interested in metabolic disorders. Borgny & Harry talked it over and Borgny decided to try to “use her connections” to get her children help. Her sister knew Dr. Fölling, so Borgny asked her to ask him if he thought there could be a link between the smell and the symptoms. Dr. Fölling said he didn’t know but, being a fine gentleman, offered to examine their pee, so she brought some of Liv’s pee to him. He ran all the routine tests on it, but all those normal tests came back normal - there wasn’t any pus or blood in it, sugar levels were A-okay… So he put on his thinking (and smelling cap). He knew that there’s an odor associated with untreated diabetes. That was a different odor, a “fruitier” one, but a smell none the less, and it’s caused by a build up of ketones (things with that -C-(C=O)-C- sandwich). And there was a test for that. Ferric chloride (an iron cation (Fe²⁺ balanced out by chlorine anions (Cl ⁻) is brownish but if you add it to a substance containing ketones, it’ll turn purply and this could indicate diabetes. The kids had no symptoms of diabetes and their sugar levels were normal, but they did have a smell so he thought, “oh, what the hell!” He went ahead and added some ferric chloride to the girl’s pee - and what did he see? Not purple, but GREEN! What could that mean?! Could the son be the same? More urine was brought and yes - green again! He had never seen this before - had never even heard of it, so he wanted to make sure it was legit. Maybe he’d mis-mixed one of his chemicals or something? Or maybe one of herbal remedies or the aspirin the children were occasionally taking was causing it. So he asked the mom to bring more after stopping any treatment. And he got the same result. In fact, every time the mom brought pee (he requested fresh batches be provided every other day) it turned green - a deep green that faded in a few minutes. So now he wanted to figure out why. But he didn’t know what he was looking for, so his number 1 priority was “urine - get me more!” He ended up taking ~20L of it in order to isolate out the green-maker from all the other chemicals in the pee. But once he found it, it wasn’t like it just told him what it was - instead, what he’d done was use a bunch of organic chemistry to fractionate out various things in the urine such as by “extracting” them into different solvents (basically given a choice of liquids to hang out in, different molecules will choose different ones, so you can separate molecules based on their preferences for different solvents). So he could extract stuff, test them to with ferric chloride to see where the green-in-ator was, then further fractionate that part, etc, and keep following it.

[@thebumblingbiochemist]

He needed to get it super pure in order to be able to characterize it as accurately as possible, so once it was “pretty pure” he turned to crystallization - when you evaporate out the solvent the molecules will come together to form crystals - this involves them organizing into orderly lattices, kicking out other molecules in the process. so you can keep dissolving and crystalizing to get better and better purity. He was able to get pure crystals of it after six recrystallizations. But, as you know if you’ve gotten distracted in the middle of mixing or weighing salt crystals and then wonder if you put the right one in the right thing, crystals of different things often look pretty darn similar… And there was nothing really special-looking about these crystals, so he turned to some chemical tests. He combusted it & quantified the different elements - this gave him the empirical formula C₉H₈O₃ (note: an empirical formula is the simplest “ratio” of different elements - so the true thing could be C₉H₈O₃ or C₁₈H₁₆O₆, etc.). He found that it was acidic (able to donate a proton (H⁺)) and by “titrating” it with a base (proton-taker) to see how much base was needed to neutralize it, he determined that there was 1 acidic group per molecule. The measurements were made more difficult by the fact that it oxidized easily. Oxidation is the loss of electrons, often (but not always) associated with addition of oxygen and/or loss of hydrogen and its counterpart is reduction (gain in electrons, often (but not always) associated with loss of oxygen and/or gain of hydrogen). More on such “redox reactions” here: bit.ly/2R3vIPf In order to prevent oxidation from messing with his results, he expelled oxygen gas from all his solutions by replacing it with nitrogen gas and performed his experiments under an atmosphere of nitrogen. That oxidation was problematic in that sense, but helpful in another sense - the sense of smell! Mild oxidization made it smell like benzaldehyde and stronger oxidation produced oxalic acid (he knew this because it could be precipitated by calcium & benzoic acid & detected by distillation). Sorry for getting jargonny - I’m not going to explain it all just wanted to put that out there in case folks are interested - but you don’t really need to understand how he did it, just know that he started to suspect that he was looking at phenylpyruvic acid. and if you want more details, here’s a good article describing it (in English!) www.pediatrics.org/cgi/content/full/105/1/89 His suspicion was strengthened when he measured the melting point and found it to be 155°C, consistent with the melting point for phenylpyruvic acid, but he wanted to be sure. So he made some pure phenylpyruvic acid, & mixed it with his crystals and measured the melting point again (sometimes if you mix 2 different things with the same melting point the melting point drops). The melting point stayed the same, so he concluded that his green-in-ator and the phenylpyruvic acid were one and the same! Apparently phenylpyruvic acid forms a complex with Fe²⁺ that looks green (unlike the diabetic ketone complex that looks red). The smell came from phenylacetic acid/phenylacetate - metabolic by-products “metabolites” of phenylpyruvic acid. This was the first time a metabolic disorder had been linked to a neurological condition, and Fölling wondered if there were other patients out there. So he collected urine samples from 430 patients with developmental disabilities - and 8 tested positive. That 8 included 2 more pairs of siblings - it was definitely looking to be genetic. And further “pedigree” studies, which test affected and non-affected family members to get at the inheritance pattern, showed it to be recessive - 2 faulty copies needed, so (barring spontaneous mutations) each parent has to be a carrier and each carrier/carrier couple has a 1/4 chance of having the disease (and a 1/2 chance of being a carrier). Those pedigree studies came a little later - he wanted to get his initial findings out ASAP - they were published in German in 1934 in a paper titled the German version of “On the excretion of phenylpyruvic acid in the urine as an anomaly of metabolism in connection with mental retardation.” He suggested the name phenylpyruvic oligophrenia. It’s called Fölling’s Disease in Norway & it got the “phenylketonuria” name from another Dr. - Lionel Penrose - because of the phenylpyruvic acid detected in patients’ urine. Speaking of that phenylpyruvic acid…

[@thebumblingbiochemist]

First, a nomenclature thing. It has one of those carboxyl groups, which can be in a protonated, neutral, form or a deprotonated, negatively-charged, form. In its protonated form it’s “phenylpyruvic acid” and when it’s deprotonated it’s called “phenylpyruvate.” Which state it’s in depends on the pH (measure of how many free protons are around to take). Thankfully, both can be abbreviated PPA! Also, thankfully, PPA can be detected using the ferric chloride test, which was how Fölling was able to find it in the patients’ urine. Once Fölling published his results, other doctors started testing their patients’ urine. And many more cases were discovered. But that knowledge wouldn’t really be helpful unless doctors could do something to intervene once they knew. So they needed to know, “where was the PPA coming from?” Fölling suspected it was being made as an alternative breakdown product of phenylalanine, but to show that he needed to figure out how much phenylalanine was in the blood. If there was a problem breaking it down, there should be a buildup of it, but he didn’t have a way to measure it. So he turned to his bacteriologist friend for help - the friend found a strain of bacteria called Proteus vulgaris that was able to convert phenylalanine to phenylpyruvic acid, which then they could measure by color using the ferric chloride test. This confirmed that there was indeed “extra” phenylalanine present. (These days, Phe is measured in the blood using mass spectrometry (mass spec), a technique which breaks molecules into charged fragments and measures their charge to mass ratio to get a “fingerprint” they can interpret and compare to known “fingerprints”) So there was some problem with phenylalanine metabolism in these patients - some holdup in a “normal” pathway causing an alternative pathway to be used. But what was that problem? Enter George Jervis. In 1947, he reported that the holdup is due to inability to hydroxylate phenylaline (add that -OH onto it to get Tyr). He figured this out by feeding healthy adults & PKU patients phenylalanine, tyrosine, or phenylpyruvic acid, then drawing blood at various timepoints and performing a “Millon reaction” - basically you add a dye and it interacts with tyrosine (or other “Millon-reacting substances” to form a red product. In healthy people all 3 (Phe, Tyr, & PAA) produced a rise in Millon-reacting substance (tyrosine), peaking an hour or so after ingestion for the Phe & PPA, followed by a return to baseline (phenylpyruvic acid can converted into phenylalanine in the body so it can then be turned into tyrosine). PKU patients peaked with tyrosine but remained flatlined (and low) with either of the other 2 suggesting that they couldn’t convert them to tyrosine www.jbc.org/content/169/3/651.full.pdf So now they knew the source of the problem, but how to intervene? Could they make food without phenylalanine? Yes, but with extreme difficulty…

[@thebumblingbiochemist]

You can’t just not give PKU patients any protein or their bodies will start breaking down their own proteins for parts. So they needed to find a way to remove phenylalanine from protein. Drs Horst Bickel, L. I. Woolf, and Evelyn Hickmans created a phenylalanine-free protein blend by taking the milk protein casein and hydrolyzing it - breaking it into the individual letters - then flowing that mix of letters through activated charcoal (carbon). Phe’s big ring hydrophobic nature makes it stick to the carbon, so it gets removed. Tyrosine stuck too, so they added some back in. The first patient was a 17-month-old girl named Sheila with PKU and a persistent mom. Bickel originally fed her only the Phe-Free food. The Phe levels in the blood went down but then they started to come back up as the child’s body started breaking down its own proteins to get Phe (turns out Phe is one of those “essential” amino acids that your body can’t make itself - and she still needed some). So he added a small amount of Phe and saw dramatic improvement - both in the blood Phe levels and in the patient’s neurological status. An improvement that was quickly reversed when he added higher levels of Phe back in to check that it really was excess Phe causing the problems. They end their classic 1953 Lancet article, titled “Influence of phenylalanine on phenylketonuria” bit.ly/2u1QkxY : “In this child at least the beneficial effects of a low phenylalanine intake seem unequivocal, although the degree of mental development finally obtained remains to be seen after further prolonged treatment. In view of the importance of phenylketonuria as a cause of mental deficiency, further controlled trials are being made, special attention being paid to very young children, who are likely to benefit most.” But finding those “very young children” and intervening before irreversible damage occurred, would require screening. First came the diaper test. Drop some ferric chloride on a urine-rich diaper and see if it turned green (they even made sticks coated with ferric chloride that could be used as “swabs”). Problem was, the test was only useful in babies at least a few weeks old because it’s detecting the PPA, which doesn’t build up in the urine until Phe levels get higher. So, if they wanted a way to test newborns before they even left the hospital they’d need to measure Phe directly - in the blood. Enter Robert Guthrie. Guthrie came up with a “bacterial inhibition assay” that could be performed with a small amount of blood (such as you can get from a “heelprick”) absorbed on a filter paper, then “punched out” and placed on an agar plate - those plates used for growing bacteria bit.ly/2UVJtBi But here’s the trick - the reason for the prick - in addition to the normal bacteria food filling up the agar mesh gel in the plates he added β-2-Thienylalanine. It mimics phenylalanine so confuses Bacillus subtilis bacteria & prevents it from growing. Unless there’s so much Phe in the patient’s blood that it “outcompetes” the faker and the bacteria can grow - which they do in a “halo” around the bloody paper punch. The more Phe in the blood, the bigger the halo (but worse the disease). A lot of advocacy and the test became widespread across the U.S. Nowadays, PKU is still tested through a heelprick and the test is still called a “Guthrie test,” but Phe levels are measured using mass spec - and the levels of other metabolites are also measured from it to detect other diseases.

[@thebumblingbiochemist]

Mandatory testing wasn’t the only governmental action - governments started requiring food-makers to put those warnings on food labels, “phenylketonurics: contains phenylalanine” when it’s found in places you’d might not expect it to be. Like WTF does F have to do with an artificial sugar? The sweetener aspartame is a dipeptide dipeptide of aspartic acid and phenylalanine. Breaking it down produces Phe (but on a molecule by molecule basis it’s way sweeter than sugar so, even though it still has some calories, you have to use such little amounts of it that they’re often considered “calorie-free”). So the warning gives people with PKU a heads-up. Another result of PKU scientists and advocates is that Lofenalac, a protein formula for PKU, became the 1st “medical food” in the US. That was in 1972. What’s happened since then? One thing is that companies dedicated to making yummier low/no-phenylalanine foods have formed. Medication-wise, there have been a couple advances. One is that BH₄ mimic, Kuvan - but that only works for patients with milder PKU who still have enough active PAH for that helper molecule (coenzyme) to have something to help! Another, more recent, advance (FDA-approved in May 2018) is a drug from the same company, BioMarin Pharmaceutical, called Palynziq (pegvaliase-pqpz). It’s an “enzyme substitution therapy” based on the idea of replacing the defective enzyme (PAH) so that even patients who don’t make any PAH can benefit. But PAH isn’t very stable and it requires cofactors, so they turned to a “substitute” - PAL (phenylalanine ammonia lyase). Instead of hydroxylating Phe, it deaminates it - but without using up α-ketoglutarate and without generating PPA - instead it makes trans-cinnaminic acid and ammonia, which the liver can metabolize and excrete. It was actually first tried in 1980 - but orally - and it didn’t survive digestion. Palynziq avoids that problem because it’s delivered via subcutaneous (under the skin) injection. To help increase stability & reduce immunogenicity (your body recognizing it as foreign and mounting an immune response against it) it’s “PEGylated” - it has the polymer polyethylene glycol attached to it. Even still, allergic reactions are common. ~1/2 of patients have side effects - rash, itching, swelling, even anaphylactic shock (in ~9% of trial participants). That’s not the only problem. It’s $$. The PAL gene they used comes from a cyanobacterium (Anabaena variabilis) made & it’s made & purified from E. coli bacteria. And costs close to $200,000 a year. But it may be worth it. Patients on it can eat normal food and feel less “brain foggy” (this suggests there may be other things missing in synthetic diets). here’s a great article on treatment: bit.ly/2P15lHI I also found an article about a PKU patient on the drug, whose mom Alison Reynolds is planning to cross-country ski across Norway to Fölling’s stomping grounds later this month to raise money and awareness for PKU. wapo.st/2HwJrbm And you now are all aware - more than you’d hoped probably, about PKU - so I hope this has helped informationitize you! But - disclaimer - I am NOT a doctor (not even the PhD kind - yet) so if you or your family, friend, etc. have PKU talk to a real medical doctor - hopefully with a biochemistry background! more on phenylalanine: blog form: bit.ly/phenylalaninearomatic ; KZbin: kzbin.info/www/bejne/amGsk2ZsjbV6kLs     recommended reading: Centerwall, S. A.; Centerwall, W. R. The Discovery of Phenylketonuria: The Story of a Young Couple, Two Retarded Children, and a Scientist. Pediatrics 2000, 105 (1), 89-103. doi.org/10.1542/peds.105.1.89. Jervis, G. A. STUDIES ON PHENYLPYRUVIC OLIGOPHRENIA. Journal of Biological Chemistry 1947, 169 (3), 651-656. doi.org/10.1016/S0021-9258(17)30882-7. posts on proteins and amino acids: thebumblingbiochemist.com/lets-talk-science/amino-acids/       KZbin channel on amino acids: kzbin.info/aero/PLUWsCDtjESrFQoCEsEmZX6NxnwlHzjHZ6   more on the glucose-alanine cycle: kzbin.info/www/bejne/h33Onp5jpJx8sM0   
                              more about all sorts of things: #365DaysOfScience All (with topics listed) 👉 bit.ly/2OllAB0 or search blog: thebumblingbiochemist.com