Category Archives: Molecular Biology

The next big drug target

So many of the molecular machines used in the cell are composed of many different proteins held together by nonCovalent interactions. The Mediator complex contains 25 – 30 proteins with a mass of 1.6 megaDaltons, RNA polymerase contains 12 subunits, the general transcription factors contain 25 proteins, our ribosome with a mass of 4.3 megaDaltons contains 47 in the large subunit and 33 in the small. The list goes on and on — proteasome,nucleosome, post-synaptic density.

The typical protein/protein interface has an area of 1,000 – 2000 square Angstroms — or circles of diameter between 34 and 50 Angstroms. [ Proc. Natl. Acad. Sci. vol. 101 pp. 16437 – 16441 ’04 ]. Think of the largest classical organic molecule you’ve ever made (not any polymer like a protein, polynucleotide, or polysaccharide). It isn’t anywhere close to this.

Yet I’m convinced that drugs targeting these complexes, will be useful. Classical organic chemistry will be useless in designing them. We’ll have to forget our beloved SN1, SN2, nonclassical carbonium ions etc. etc. We need some new sort of physical organic chemistry, one not concerned with reaction mechanism, but with van der Waals interactions, electrostatic interactions. At least stereochemistry will still be important.

The problem is much harder than designing enzyme inhibitors, or their allosteric modifiers, because the target is so large.

What follows are some notes on the protein protein interface I’ve taken over the years to get you started thinking. Good luck. Don’t expect any neat answers. There is a lot of contention concerning the nature of the binding occurring at the interface.

Many of the references aren’t particularly new.  In my reading, I don’t try for the latest reference, but the newest idea that I’m unfamiliar with.  I think they pretty much cover the territory as it stands now.

[ Proc. Natl. Acad. Sci. vol. 108 pp. 603 – 608 ’11 ] A very interesting article argues that worms and humans have about the same number of proteins (20,000) because if they had more, nonspecific protein protein interactions would cause disease. The achievable energy gap favoring specific over nonspecific binding decreases with protein number in a power law fashion (in their model). The optimization of binding interfaces favors networks in which a few proteins have many partners and most proteins have just a few — this is consistent with a scale free network topology.

[ Proc. Natl. Acad. Sci. vol. 101 pp. 16437 – 16441 ’04 ] The hot spot theory of protein protein interactions says that the binding energy between two proteins is governed in large part by just a few critical residues at the binding interface. In a typical interface of 1000 – 2000 square Angstroms, only 5% of the residues from each protein contribute more than 2 kiloCalories/mole to the binding interaction. (This is controversial — see later)

[ Proc. Natl. Acad. Sci. vol. 99 pp. 14116 – 14121 ’02 ] Specific replacement of amino acids in the interface by alanine (alanine scanning or alanine mutagenesis) and measuring the effect on the interaction has led to the idea that only a small set of ‘hot spot’ residues at the inferface contribute to the binding free energy. A hot spot has been defined as a residue that when mutated to alanine leads to a significant drop in the binding constant (typically 10 fold or higher — should know how many kiloCalories this is — I think 2 or 3 ). This was well worked out for human growth hormone (HGH) and its receptor. Subsequently ‘many’ other studies have suggested that the presence of a few hot spots may be a general characteristic of most protein/protein interfaces.

However there is extreme variation in the size, shape, amino acid character and solvent content of the protein/protein interface. It is not obvious from looking at structural contacts which residues are important for binding. Usually they are found at the center of the interface but sometimes the key residues can lie on the periphery. Peripheral residues serve as an O-ring to exclude solvent from the center. A lowered effective dielectric constant in a ‘dryer’ environment strengthens electrostatic and hydrogen bonding interactions. An interaction deleted by alanine mutagenesis in the periphery can be replaced by a water molecule in the periphery and hence cause less loss in stability (this calls the whole concept of alanine scanning into question).

Interestingly, there is no general correlation between ‘surface accessibility’ and the contribution of a residue to the binding energy.

Polar residues (Arg, Gln, His, Asp, and Asn) are conserved in interfaces. This implies that they are hot spots — implies ? don’t they know? haven’t they tested? However, many interaction hot spots involve hydrophobic or large aromatic residues (also hydrophobic). It is unclear whether buried polar interactions are energetically net stabilizing or merely facilitating specificity (how would you tell?).

Some residues without significant contacts in the interface apparently contribute substantially to the free energy of binding when assayed by alanine scanning mutagenesis, because of destabilization of the unbound protein.

This a report of a free energy function (using packing interactions, hydrogen bonds and an implicit solvation model) which predicts 79% of all interface hot spots. They think that a description of polar interactions with Coulomb electrostatics with a linear distance dependent dielectric constant. ??? The latter ignores the orientation dependence of the hydrogen bond. Also the assumption that acidic or basic residues largely buried in the interface are charged may be wrong. The enthalpic gains of ionization are offset by the cost of desolvating polar groups, and the loss in side chain conformational entropy.

[ Proc. Natl. Acad. Sci. vol. 101 pp. 16437 – 16441 ’04 ] It is of interest to find out if hot spot theory applies to transient protein protein interactions (such as those involved in enzyme catalysis). This work looked for them in the process of protein substrate recognition for the Cdc25 phosphatase (which dephosphorylates the cyclin dependent kinases). Crystal structures of the catalytic domains of Cdc25A and Cdc25B have shown a shallow active site with no obvious features for mediating substrate recognition. This suggests a broad protein interface rather than lock and key interaction. This is confirmed by the activity of the Cdc25 phosphatases toward Cdk/cyclin protein substrates which is 6 orders of magnitude greater than that of peptidic substrates containing the same primary sequence — this suggests a broad protein interface rather than a lock and key interaction. The shallow active sites also correlates with the lack of potent speicific inhibitors of the Cdc25 phosphatases, despite extensive search. This work finds hot spot residues in the catalytic domain (not the catalytic site) of Cdc25B located 20 – 30 Angstroms away from the active site. They are involved in recognition of substrate. The residues are conserved across eukaryotes.

[ Proc. Natl. Acad. Sci. vol. 101 pp. 11287 – 11292 ’04 ] One can study the effects of mutating a single amino acid on two separate rates (the on rate and the off rate) the ratio of which is the equilibrium constant. Mutations changing the on rate, concern the specificity of protein protein interaction. Mutations only changing the off rate do not affect the transition state of protein binding (don’t see why not). Mutations in bovine pancreatic trypsin inhibitor (BPTI) have been found at positions #15 and #17 which differentially affect on and off rates. K15A decreases by 200 fold in the on rate and by a 1000 fold increase in the off rate. But R17A doesn’t change the on rate but also increases the off rate by 1000 fold.

The concept of anchor residue arose in the study of peptide binding to class I MHC molecules (Major HistoCompatibility complex) In this system the carboxy terminal side chain of the peptide gets buried in pocket F of the MHC binding groove. Sometimes, one also finds a second anchor residue and even a third one buried at other positions.

The authors attempt to apply the anchor residue concept to protein protein interactions. They studied 39 different protein/protein complexes. They found them, and in some way conclude that these anchor residues are already in the ‘bound’ conformation in the free partner. The anchors interact with structurally constrained pockets matching the anchor residues. The presence of nativelike anchor side chains provides a readily attainable geometrical fit that jams the two interacting surfaces, allowing for the recognition and stabilization of a near-native intermediate. Subsequently an induced fit process occurs on the periphery of the binding pocket.

The analysis of ANY (really?) protein/protein complex at the atomic length scale shows that the interface, rather than being smooth and flat, includes side chains deeply protruding into well defined cavities on the other protein. In all complexes studied, the anchor is the side chain whose burial after complex formation yields the largest possible decrease in solvent accessible surface area (SASA). If SASA is over 100 square Angstroms, than only one anchoring interaction is present. For lesser SASA amino acids one anchor isn’t enough.

In all cases tested (39) latch side chains are found in conformations conducive to a relatively straightforward clamping of the anchored intermediate into a high affinity complex.

[ Proc. Natl. Acad. Sci. vol. 102 pp. 57 – 62 ’05 ] An analysis of the protein interface between a beta-lactamase and its inhibitor, shows that the interface can be divided into clusters (by means of cluster anlaysis) using multiple mutant analysis and xray crystallography. Replacing an entire module of 5 interface residues with alanine (in one cluster) created a large cavity in the interface with no effect on the detailed structure of the remaining interface. They obtained similar results when they did this with another of the 5 clusters.

Mutating a single amino acid at a time has been done in the past, but the results of single mutations aren’t additive (e.g. they aren’t linear — no surprise). The sum of the loss in free energy of all of the single mutations within a cluster exceeds by 4 fold the loss in free energy generated when all of the residues of the cluster are mutated simultaneously. The energetic effect of many single mutations is larger than their net contribution due to a penalty paid by leaving the rest of the cluster behind.

“Binding seems to be a result of higher organization of the binding sites, and not just of surface complementrity.”

[ Proc. Natl. Acad. Sci. vol. 103 pp. 311 – 316 ’06 ] Two different ‘interactomes’ both show the same power law distribution of node sizes. However, when the two major S. cerevisiae protein/protein interactions are experiments are compared with each other, only 150 of the THOUSANDS of interactions of each experiments are the same. A similar lack of agreement has been found for independent Y2H experiments in Drosophila.

This work says that desolvation of the interface is a major physical factor in protein/protein interactions. This model reproduces the scale free nature of the topology. The number of interactions made by a protein is correlated with the fraction of hydrophobic residues on its surface.

      [ Proc. Natl. Acad. Sci. vol. 108 pp. 13528 – 13533  ’11 ] The drugs they are looking for disrupt specific protein protein interactions (PPIs).   Tey use computational solvent mapping, which explores the protein surface using a variety of small probe molecules, along with a conformer generator to account to side chain flexibility.  They studied unliganded proteins known to participate PPI.  The surface cavities available at protein protein interfaces which can bind a smal molecule inhibitor are rather different than those seen in traditional drug targets.  The traditional targets have one or two disproportionately large pockets with an average volume of 260 cubic Angstroms — these account for the binding site for the endogenous ligand in over 90% of proteins.  The average volume of pockets at protein protein interfaces is only 54 cubic Angstroms, the same as for all protein surface pockets.  The interface ontains 6 such small pockets (on average). 
      The binding sites of proteins generall include smaller regions called hotspots which are major contributors to the binding free energy.  The results of experimental fragment screens confirm that the hot spotes of proteins are characterized by their ability to bind a variety of small molecules and that the number of different probe molecules observed to bind to a particular site predicts the importance of the site and predicts overall druggability.  
      This work shows that the druggable sites in PPIs have concave topology and both hydrophobic and polar functionality.  So the hotspots bind organic molecules having some polar groups decorating largely hydropobic scaffolds. Sos druggable sites have a ‘general tendency’ to bind organic compounds with a variety of structures.  Conformational flexibility at the binding site (by side chains?) allow the hotspots to expand to accomodate a ligand of druglike dimensions.  This involves low energy side chain motions within 6 Angstroms of a hot spot.
      So druggable sites at a PPI aren’t just sites complementary to particular organic functionality, but they have a general tendency to bind a variety of different organic structures.  
      The most important binding is that the druggable sites are detectable from the structure of the unliganded protein, even when substantial conformational adaptation is needed for optimal ligand binding.

[ Science vol. 347 pp. 673 – 677 ’15 ] Mapping the sequence space of 4 key amino acids in the E. Coli protein kinase PhoQ which drives the recognition of its substrate (PhoP). For histidine kinases mutating just 3 or 4 interfacial amino acids to match those in another kinase is enought to reprogram them. The key variants are Ala284, Val 285, Ser288, Thr289.

All 20^4 = 160,000 variants of PhoQ at these positions were made, of which 1,659 were functional (implying singificant degeneracy of the interface). There were 16 single mutants, 100 double, 544 triple and 998 quadruple mutants which were functional. There was an enrichment of hydrophobic and small polar residues at each position. Most bulky and charged residues appeared at low frequencies. Some substitutions were permissible individually, but not in combination. The combinations, ACLV, TISV, SILS, each involving aresidues found individually in functional mutants at high frequency, were quite impaired in competition against wildtype PhoQ — so the effects of individual substitutions are context dependent (epistatic). Of the 100 functional double mutants, only 23 represent cases where both single mutants are functional. THere are double mutants where neither single mutant is functional. 79/1,658 functional variants can’t be reached from the wild-type combination AVST) without passing through a nonfunctional intermediated. They talk about the Hamming distance between mutants.

Finally some blue sky stuff — implying that (as usual) Nature got there first

       [ Science vol. 341 pp. 1116 – 1120 ’13 ] Small Open Reading Frames (smORFs) code for peptides of under 100 amino acids.  This work has shown that peptides as short as 11 amino acids are translated and provide essential functions during insect development.  This work shows two peptides of 28 and 29 amino acids regulating calcium transport in the Drosophila heart.  The peptides are found in man.  
      They don’t think that smORFs can’t be dismissed as irrelevant, and function should be looked for. 
       [ Science vol. 1356 – 1358 ’15 ] The Drosophila polished-rice (Pri) sORF peptides (11 – 32 amino acids)trigger proteasome mediated processing converting the Shavenbaby transcription repressor into a shorter activator.
       They think that oORF/smORFs mimic protein binding interfaces and control protein interactions that way.


Every class in grad school seemed to begin with a discussion of units. Eventually, Don Voet got fed up and said he preferred the hand stone fortnight system and was going to stick to it. However, even though we all love quantum mechanics dearly for predicting chemical reactivity and spectra, it tells us almost nothing about the events going on in our cells. It’s a crowded environment with objects large and small bumping into one another frequently and at high speeds. At room temperature, a molecule of nitrogen is moving at 500+ meters a second or over 1100 miles an hour. The water in our cells is moving even faster (28/18 times faster to be exact). It’s way too slow for relativity however.

So it’s back classical mechanics to understand cellular events at a physical level, something that will be increasingly important in drug design (but that’s for another post).

The average thermal energy of a molecule at room temperature is kT.

What’s k? It’s the Boltzmann constant. What’s that? It’s the gas constant divided by Avogadro’s number.

I’m assuming that all good chemists know that Avogadro’s number is the number of molecules in a Mole = 6.02 x 10^23

What does the Gas constant have to do with energy?

It’s back to PChem 101 — The ideal gas law is PV = nRT

P = Pressure
V = Volume
n = number of moles
R = Gas constant
T = Temperature

Pressure is Force / Area

Force is Mass * Acceleration
Acceleration is Distance/ (Time * Time)
Area is Distance * Distance
Volume is Distance * Distance * Distance

So PV == [ Force/Area ] * Volume
== { [ Mass * (Distance / Time * Time) ] /( Distance * Distance ) } * ( Distance * Distance * Distance )
== Mass * (Distance/Time) * ( Distance/Time )
== Mass * Velocity * Velocity == mv^2

So PV has the dimensions of (kinetic) energy

The Gas Constant (R) is PV/nT ( == PV/T ) so it has the dimensions of energy/temperature

Now for some actual units (vs. dimensions, although things are much clearer when you think in terms of dimensions)

Force is measured in Newtons which is the force which will accelerate a 1 kiloGram object by 1 meter/second^2

Temperature is measured in Kelvin from absolute zero. A degree Kelvin is the same as 1 degree Celsius (1.8 degrees Fahrenheit)

Room temperature where most of us live is about 27 Centigrade or very close to 300 Kelvin.

So the Boltzmann constant (k) basically energy/temperature per single molecule, which is really what you want to think about when you think about physical processes in the cell.

At room temperature kT works out to 4.1 x 10^-21 Joules.

What’s a Joule? It’s the energy a force of one Newton produces when it moves an object one meter (or you can look at it as the kinetic energy one kilogram has after a force of one Newton has accelerated it over one Meter’s distance.

So a Joule is one Newton * meter

Well 10^-21 is 10^-12 times 10^-9. So what?

This means that at room temperature the average molecule has a thermal energy of 4.1 picoNewton – nanoMeters.

PicoNewtons just happens to be in the range of the force exerted by our molecular motors ( kinesin, dynein, DNA polymerases ) and nanoMeters the range of distances over which they exert forces (act).

Not a coincidence.

Since there are organisms which live at temperatures 20% higher, it would be interesting to know if their motors exert 20% more force. Does anyone out there know?

More interesting even than that are the organisms living at the mid-Ocean ridges where because the extremely high pressures, the water coming from the vents is a lot hotter. What about their motors?

It ain’t the bricks, it’s the plan

Nothing better shows the utility (and the futility) of chemistry in biology than using it to explain the difference between man and chimpanzee. You’ve all heard that our proteins are only 2% different than the chimp, so we are 98% chimpanzee. The facts are correct, the interpretation wrong. We are far more than the protein ‘bricks’ that make us up, and two current papers in Cell [ vol. 163 pp. 24 – 26, 66 – 83 ’15 ] essentially prove this.

This is like saying Monticello and Independence Hall are just the same because they’re both made out of bricks. One could chemically identify Monticello bricks as coming from the Virginia piedmont, and Independence Hall bricks coming from the red clay of New Jersey, but the real difference between the buildings is the plan.

It’s not the proteins, but where and when and how much of them are made. The control for this (plan if you will) lies outside the genes for the proteins themselves, in the rest of the genome (remember only 2% of the genome codes for the amino acids making up our 20,000 or so protein genes). The control elements have as much right to be called genes, as the parts of the genome coding for amino acids. Granted, it’s easier to study genes coding for proteins, because we’ve identified them and know so much about them. It’s like the drunk looking for his keys under the lamppost because that’s where the light is.

We are far more than the protein ‘bricks’ that make us up, and two current papers in Cell [ vol. 163 pp. 24 – 26, 66 – 83 ’15 ] essentially prove this.

All the molecular biology you need to understand what follows is in the following post —

Briefly an enhancer is a stretch of DNA distinct from the DNA coding for a given protein, to which a variety of other proteins called transcription factors bind. The enhancer DNA and associated transcription factors, then loops over to the protein coding gene and ‘turns it on’ — e.g. causes a huge (megaDalton) enzyme called pol II to make an RNA copy of the gene (called mRNA) which is then translated into protein by another huge megaDalton machine called the ribosome. Complicated no? Well, it’s happening inside you right now.

The faces of chimps and people are quite different (but not so much so that they look frighteningly human). The cell paper studied cells which in embryos go to make up the bones and soft tissues of the face called Cranial Neural Crest Cells (CNCCs). How did they get them? Not from Planned Parenthood, rather they made iPSCs (induced Pluripotent Stem Cells — differentiate into CNCCs. Not only that but they studied both human and chimp CNCCs. So you must at least wonder how close to biologic reality this system actually is.

It’s rather technical, but they had several ways of seeing if a given enhancer was active or not. By active I mean engaged in turning on a given protein coding gene so more of that protein is made. For the cognoscenti, these methods included (1) p300 binding (2) chromatin accessibility,(3) H3K4Me1/K3K4me3 ratio, (4) H3K27Ac.

The genome is big — some 3,200,000 positions (nucleotides) linearly arranged along our chromosomes. Enhancers range in size from 50 to 1,500 nucleotides, and the study found a mere 14,500 enhancers in the CNCCs. More interestingly 13% of them were activated differentially in man and chimp CNCCs. This is probably why we look different than chimps. So although the proteins are the same, the timing of their synthesis is different.

At long last, molecular biology is beginning to study the plan rather than the bricks.

Chemistry has a great role in this and will continue to do so. For instance, enhancers can be sequenced to see how different enhancer DNA is between man and chimp. The answer is not much (again 2 or so nucleotides per hundred nucleotides of enhancer). The authors did find one new enhancer motif, not seen previously called the coordinator motif. But it was present in man in chimp. Chemistry can and should explain why changing so few nucleotides changes the proteins binding to a given enhancer sequence, and it will be important in designing proteins to take advantage of these changes.

So why is chemistry futile? Because as soon as you ask what an enhancer or a protein is for, you’ve left physicality entirely and entered the realm of ideas. Asking what something is for is an entirely different question than how something actually does what it is for.  The latter question  is answerable by chemistry and physics. The first question is unanswerable by them.  The Cartesian dualism of flesh and spirit is alive and well.

It’s interesting to see how quickly questions in biology lead to teleology.

How ‘simple’ can a protein be and still have a significant biological effect

Words only have meaning in the context of the much larger collection of words we call language. So it is with proteins. Their only ‘meaning’ is the biologic effects they produce in the much larger collection of proteins, lipids, sugars, metabolites, cells and tissues of an organism.

So how ‘simple’ can a protein be and still produce a meaningful effect? As Bill Clinton would say, that depends on what you mean by simple. Well one way a protein can be simple is by only having a few amino acids. Met-enkephalin, an endogenous opiate, contains only 5 amino acids. Now many wouldn’t consider met-enkehalin a protein, calling it a polypeptide instead. But the boundary between polypeptide and protein is as fluid and ill-defined as a few grains of sand and a pile of it.

Another way to define simple, is by having most of the protein made up by just a few of the 20 amino acids. Collagen is a good example. Nearly half of it is glycine and proline (and a modified proline called hydroxyProline), leaving the other 18 amino acids to make up the rest. Collagen is big despite being simple — a single molecule has a mass of 285 kiloDaltons.

This brings us to [ Proc. Natl. Acad. Sci. vol 112 pp. E4717 – E4727 ’15 ] They constructed a protein/polypeptide of 26 amino acids of which 25 are either leucine or isoleucine. The 26th amino acid is methionine (which is found at the very amino terminal end of all proteins — remember methionine is always the initiator codon).

What does it do? It causes tumors. How so? It binds to the transmembrane domain of the beta variant for the receptor for Platelet Derived Growth factor (PDGFRbeta). The receptor when turned on causes cells to proliferate.

What is the smallest known oncoprotein? It is the E5 protein of Bovine PapillomaVirus (BPV), which is an essentially a free standing transmembrane domain (which also binds to PDGFRbeta). It has only 44 amino acids.

Well we have 26 letters + a space. I leave it to you to choose 3 of them, use one of them once, the other two 25 times, with as many spaces as you want and construct a meaningful sequence from them (in any language using the English alphabet).

Just back from an Adult Chamber Music Festival (aka Band Camp for Adults).  More about that in a future post

The elegance of metabolism control in the cell.

The current two pronged research effort on the possible use of Gemfibrozil (Lopid) to treat Alzheimer’s disease now has far wider implications than Alzheimer’s disease alone. As far as I’m aware, the combination of mechanisms described below to control a cellular pathway as never been reported before.

A previous post has the story up to 3 August — — you can read it for the details, but here’s some background and the rest of the story.

Background: One of the two pathologic hallmarks of Alzheimer’s disease is the senile plaque (the other is the neurofibrillary tangle). The major component of the plaque is a fragment of a protein called APP (Amyloid Precursor Protein). Normally it sits in the cellular membrane of nerve cells (neurons) with part sticking outside the cell and another part sticking inside. The protein as made by the cell contains anywhere from 563 to 770 amino acids linked together in a long chain. The fragment destined to make up the senile plaque (called the Abeta peptide) is much smaller (39 to 42 amino acids) and is found in the parts of APP embedded in the membrane and sticking outside the cell.

No protein lives forever in the cell, and APP is no exception. There are a variety of ways to chop it up, so its amino acids can be used for other things. One such chopper is called ADAM10 (aka Kuzbanian). ADAM10breaks down APP in such a way that Abeta isn’t formed. A paper in the 7 July PNAS (vol. 112 pp. 8445 – 8450 ’15 7 July ’15) essentially found that Gemfibrozil (commercial name Lopid) increases the amount of ADAM10 around. If you take a mouse genetically modified so that it will get senile plaques and decrease ADAM10 you get a lot more plaques.

I wrote the author (Dr. Pahan) to ask how they came up with Gemfibrozil (Lopid). He told me that a transcription factor (PPARalpha) helps transcribe the ADAM10 gene into mRNA, and that Gemfibrozil makes PPARalpha a better transcription factor.

I told him to datamine from HMOs to find out if people on Lopid had less Alzheimer’s, he said it would be hard to get such as grant to do this as a basic researcher.

A commenter on the first post gave me a name to contact to try out the idea, but I’ve been unable to reach her. So on 3 August, I wrote an Alzheimer’s researcher at Yale about it. He responded nearly immediately with a link to an ongoing clinical study in progress in Kentucky, actually using Gemfibrozil.

Both researchers (Dr. Jicha and Nelson) were extremely helpful and cooperative. What is so fascinating is that they got to Gemfibrozil by an entirely different route. There are degrees of Alzheimer’s disease, and there is a pathologic grading scheme for it. They studied postmortem brain of 4 classes of individuals — normal nondemented elderly with minimal plaque, non demented elderly with incipient plaque, mild cognitive impairment and full flown Alzheimer’s. They had studied the microRNA #107 (miR-107) in another context. Why this one of the thousand or so microRNAs in the human genome? Because it binds to the mRNA of BACE1 and prevents it from being made. Why is this good? Because BACE1 chops up APP at a different site so the Abeta peptide is formed.

How did Gemfibrozil get into the act? Just as Dr. Pahan did, they looked to see what transcription factors were involved in making miR-107, and found PPARalpha. So to make less BACE1 they give people Gemfibrozil which turns on PPARalpha which turns on miR-107, which causes the mRNA for BACE1 to be destroyed, hopefully making less Abeta. The study is in progress and will last a year, far too short with far too few people to see a meaningful cognitive effect, but not so short that they won’t see changes in the biologic markers  they are studying in the spinal fluids (yes 72 plucky individuals have agreed to take Gemfibrozil (or not) and have two spinal taps one year apart.

The elegance of all this is simply astounding. One transcription factor turns on a gene for a chopper which inhibits Abeta formation, and turns on a microRNA which stops an APP chopper producing Abeta from being made.

So there’s a whole research program for you. Take a given transcription factor, look at the protein genes it turns on. Then look at the microRNA genes it turns on and then see what protein mRNAs they turn off. Then see they affect the same biochemical pathway as do ADAM10 and BACE1.

The mechanism is so elegant (although hardly simple) that I’ll bet the cell uses it again, in completely different pathways.

One problem with PPARalpha is that it is said to affect HUNDREDS of genes (Mol. Metab vol. 3 pp. 354 371 ’14).  So Gemfibrozil is a nice story, but even if it works, we won’t really be sure it’s doing so by ADAM10 and microRNA-107.

Takes me right back to grad school

How many times in grad school did you or your friends come up with a good idea, only to see it appear in the literature a few months later by someone who’d been working on it for much longer. We’d console ourselves with the knowledge that at least we were thinking well and move on.

Exactly that happened to what I thought was an original idea in my last post — e.g. that Gemfibrozil (Lopid) might slow down (or even treat) Alzheimer’s disease. I considered the post the most significant one I’d ever written, and didn’t post anything else for a week or two, so anyone coming to the blog for any reason would see it first.

A commenter on the first post gave me a name to contact to try out the idea, but I’ve been unable to reach her. Derek Lowe was quite helpful in letting me link to the post, so presently the post has had over 200 hits. Today I wrote an Alzheimer’s researcher at Yale about it. He responded nearly immediately with a link to an ongoing clinical study in progress in Kentucky

On Aug 3, 2015, at 3:04 PM, Christopher van Dyck wrote:

Dear Dr. xxxxx

Thanks for your email. I agree that this is a promising mechanism.
My colleague Greg Jicha at U.Kentucky is already working on this:

Our current efforts at Yale are on other mechanisms:

We can’t all test every mechanism, but hopefully we can collectively test the important ones.

-best regards,
Christopher H. van Dyck, MD
Professor of Psychiatry, Neurology, and Neurobiology
Director, Alzheimers Disease Research Unit

Am I unhappy about losing fame and glory being the first to think of it?  Not in the slightest.  Alzheimer’s is a terrible disease and it’s great to see the idea being tested.

Even more interestingly, a look at the website for the study shows, that somehow they got to Gemfibrozil by a different mechanism — microRNAs rather than PPARalpha.

I plan to get in touch with Dr. Jicha to see how he found his way to Gemfibrozil. The study is only 1 year in duration, and hopefully is well enough powered to find an effect. These studies are incredibly expensive (and an excellent use of my taxes). I never been involved in anything like this, but data mining existing HMO data simply has to be cheaper. How much cheaper I don’t know.

Here’s the previous post —

Could Gemfibrozil (Lopid) be used to slow down (or even treat) Alzheimer’s disease?

Is a treatment of Alzheimer’s disease at hand with a drug in clinical use for nearly 40 years? A paper in this week’s PNAS implies that it might (vol. 112 pp. 8445 – 8450 ’15 7 July ’15). First a lot more background than I usually provide, because some family members of the afflicted read everything they can get their hands on, and few of them have medical or biochemical training. The cognoscenti can skip past this to the text marked ***

One of the two pathologic hallmarks of Alzheimer’s disease is the senile plaque (the other is the neurofibrillary tangle). The major component of the plaque is a fragment of a protein called APP (Amyloid Precursor Protein). Normally it sits in the cellular membrane of nerve cells (neurons) with part sticking outside the cell and another part sticking inside. The protein as made by the cell contains anywhere from 563 to 770 amino acids linked together in a long chain. The fragment destined to make up the senile plaque (called the Abeta peptide) is much smaller (39 to 42 amino acids) and is found in the parts of APP embedded in the membrane and sticking outside the cell.

No protein lives forever in the cell, and APP is no exception. There are a variety of ways to chop it up, so its amino acids can be used for other things. One such chopper is called ADAM10 (aka Kuzbanian). ADAM10breaks down APP in such a way that Abeta isn’t formed. The paper essentially found that Gemfibrozil (commercial name Lopid) increases the amount of ADAM10 around. If you take a mouse genetically modified so that it will get senile plaques and decrease ADAM10 you get a lot more plaques.

The authors didn’t artificially increase the amount of ADAM10 to see if the animals got fewer plaques (that’s probably their next paper).

So there you have it. Should your loved one get Gemfibrozil? It’s a very long shot and the drug has significant side effects. For just how long a shot and the chain of inferences why this is so look at the text marked @@@@


How does Gemfibrozil increase the amount of ADAM10 around? It binds to a protein called PPARalpha which is a type of nuclear hormone receptor. PPARalpha binds to another protein called RXR, and together they turn on the transcription of a variety of genes, most of which are related to lipid metabolism. One of the genes turned on is ADAM10, which really has never been mentioned in the context of lipid metabolism. In any event Gemfibrozil binds to PPARalpha which binds more effectively to RAR which binds more effectively to the promoter of the ADAM10 gene which makes more ADAM10 which chops of APP in such fashion that Abeta isn’t made.

How in the world the authors got to PPARalpha from ADAM10 is unknown — but I’ve written the following to the lead author just before writing this post.

Dr. Pahan;

Great paper. People have been focused on ADAM10 for years. It isn’t clear to me how you were led to PPARgamma from reading your paper. I’m not sure how many people are still on Gemfibrozil. Probably most of them have some form of vascular disease, which increases the risk of dementia of all sorts (including Alzheimer’s). Nonetheless large HMOs have prescription data which can be mined to see if the incidence of Alzheimer’s is less on Gemfibrozil than those taking other lipid lowering agents, or the population at large. One such example (involving another class of drugs) is JAMA Intern Med. 2015;175(3):401-407, where the prescriptions of 3,434 individuals 65 years or older in Group Health, an integrated health care delivery system in Seattle, Washington. I thought the conclusions were totally unwarranted, but it shows what can be done with data already out there. Did you look at other fibrates (such as Atromid)?

Update: 22 July ’15

I received the following back from the author

Dear Dr.

Wonderful suggestion. However, here, we have focused on the basic science part because the NIH supports basic science discovery. It is very difficult to compete for NIH R01 grants using data mining approach.

It is PPARα, but not PPARγ, that is involved in the regulation of ADAM10. We searched ADAM10 gene promoter and found a site where PPAR can bind. Then using knockout cells and ChIP assay, we confirmed the participation of PPARα, the protein that controls fatty acid metabolism in the liver, suggesting that plaque formation is controlled by a lipid-lowering protein. Therefore, many colleagues are sending kudos for this publication.

Thank you.

Kalipada Pahan, Ph.D.

The Floyd A. Davis, M.D., Endowed Chair of Neurology


Departments of Neurological Sciences, Biochemistry and Pharmacology

So there you have it. An idea worth pursuing according to Dr. Pahan, but one which he can’t (or won’t). So, dear reader, take it upon yourself (if you can) to mine the data on people given Gemfibrozil to see if their risk of Alzheimer’s is lower. I won’t stand in your way or compete with you as I’m a retired clinical neurologist with no academic affiliation. The data is certainly out there, just as it was for the JAMA Intern Med. 2015;175(3):401-407 study. Bon voyage.


There are side effects, one of which is a severe muscle disease, and as a neurologist I saw someone so severely weakened by drugs of this class that they were on a respirator being too weak to breathe (they recovered). The use of Gemfibrozil rests on the assumption that the senile plaque and Abeta peptide are causative of Alzheimer’s. A huge amount of money has been spent and lost on drugs (antibodies mostly) trying to get rid of the plaques. None have helped clinically. It is possible that the plaque is the last gasp of a neuron dying of something else (e.g. a tombstone rather than a smoking gun). It is also possible that the plaque is actually a way the neuron was defending itself against what was trying to kill it (e.g. the plaque as a pile of spent bullets).

Why drug discovery is so hard (particularly in the brain): Reason #28: The brain processes its introns very differently

Useful drug discovery for neurologic and psychiatric disease is nearly at a standstill. It isn’t for want of trying by basic researchers and big and small pharma. A recent excellent review [ Neuron vol. 87 pp. 14 – 27 ’15 ] helps explain why. In short, the brain processes its protein coding genes rather differently.

This post assumes you know what introns, exons and alternate splicing are. For pretty much all the needed background see the following.


When splicing first came out I started making a list of proteins which were alternatively spliced. It is now safe to assume that any gene containing introns (95% of all protein coding genes [ Proc. Natl. Acad. Sci. vol. 112 pp. 17985 – 17990 ’08 ]) results in several protein products due to alternative splicing. The products produced vary from tissue to tissue, probably because most tissues express different splicing regulators.

Here are a few. A2BP1 (aka Rbfox1, aka FOX1) is a brain specific RNA splicing factor found only in postmitotic terminally differentiated neurons. It is deleted in 10% of glioblastomas. Another is nSR100 (neural Specific Related protein of 100 kiloDaltons) — see later.

To show how crucial alternative splicing is for the every existence of the brain, consider this. The neuronal splicing regulator PTBP2 is barely expressed in most tissues. It is upregulated in neurons. Both PTBP1 and PTBP2 are repressors of neural alternative splicing (but some genes are actually enhanced). In a given region of the brain either PTPB1 or PTBP2 is expressed (but not both). PTBP1 promotes skiping of a neural specific exon (exon #10) in PTBP2 transcripts. This exposes a premature termination codon in PTBP2 leading to nonsense mediated decay (NMD). PTPB1 is expressed in most nonNeural tissues and neural precursor cells, but is silenced in developing neurons by the microRNA miR-124. The mRNA for PTBP2 contains an alternative exon which triggers nonsense mediated decay (NMD) when skipped. Inclusion of the exon requires positive transacting factors such as nSR100 in neurons. Repression is mediated by PTBP1 in undifferentiation. microRNAs (which ones?) downregulate PTBP1 during neuronal differentiation, relieving the negative regulation of PTBP2. Depletion of PTBP1 in fibroblasts is enough for PTBP2 induction and neuronal transdifferentiation.

It gets more complicated still. PTBP1 inhibits splicing of introns at the 3′ end of some genes involved in presynaptic function. This results in nuclear retention and turnover via components of the nuclear RNA surveillance machinery. As PTBP1 is downregulated during neuronal differentiation, the target introns are spliced out and the mature mRNAs are found.

Now we get to microExons, something unknown until 2014. For more details see —
Briefly, microexons are defined as exons containing 50 nucleotides or less (the paper says 3 – 27 nucleotides). They have been overlooked, partially because their short length makes them computationally difficult to find. Also few bothered to look for them as they were thought to be unfavorable for splicing because they were too short to contain exonic splicing enhancers. They are so short that it was thought that the splicing machinery (which is huge) couldn’t physically assemble at both the 3′ and 5′ splice sites. So much for theory, they’re out there.

The inclusion in the final transcript of most identified neural microExons is regulated by a brain specific factor nSR100 (neural specific SR related protein of 100 kiloDaltons)/SRRM4 which binds to intronic enhancer UGC motifs close to the 3′ splice sites, resulting in their inclusion. They are ‘enhanced’ by tissue specific RBFox proteins. nSR100 is said to be reduced in Autism Spectrum Disorder (really? all? some?). nSR100 is strongly coexpressed in the developing human brain in a gene network module M2 which is enriched for rare de novo ASD assciated mutations.

MicroExons are enriched for lengths which are multiples of 3 nucleotides. Recall that every 3 nucleotides in mRNA codes for an amino acid. This implies strong selection pressure was used to preserve reading frames as 3n+1 and 3n+2 produce a frameshift. The microExons are enriched in charged amino acids. Most microExons show high inclusion at late stages of neuronal differentiation in genes associated with axon formation and synapse function. A neural specific microExon in Protrudin/Zfyve27 increases its interation with Vessicale Associated membrane protein associated Protein VAP) and to promote neurite outgrowth.

[ Proc. Natl. Acad. Sci. vol. 112 pp. 3445 – 3450 ’15 ] Deep mRNA sequencing of mouse cerebral cortex expanded the list of alternative splicing events TENfold and showed that 72% of multiexon genes express multiple splice variants. Among the newly discovered alternatively spliced exon are 1,104 exons involved in nonsense mediated decay (NMD). THey are enriched in RNA binding proteins including splicing factors. Another set of alternatively spliced NMD exons is found in genes coding for chromatin regulators. Conservation of NMD exons is found in lower vertebrates, but those involving chromatin regulators are found later into the mammalian lineage. So the transcriptome in the brain is even more complicated.

A bit more about the actual effects on protein structure of alternate splicing. The sites chosen for this aren’t random. Cell and tissue differentially regulated alternative splicing events are significantly UNDERrepresented in functionally defined folded domains in proteins, they are enriched in regions of protein disorder that typically are surface accessible and embed short linear interaction motifs (with other proteins and ligands). Among a set of analyzed neural specific exons enriched in disordered regions, 1/3 promoted or disrupted interactions with partner proteins. So regulated exon splicing might specify tissue and cell type specific protein interaction networks. They regard their inclusion/exclusion as protein surface microsurgery.

How much can a little microexon do to protein function? Here’s an example of a 6 nucleotide microexon (two amino acids). Insertion of the microExon in the nuclear adaptor protein Apbb1 enhances its interaction with Kat5/Tip60 a histone deacetylase. The microExon adds Arginine and Glutamic acid to a phosphotyrosine binding domain (PTB domain) which binds Kat4. This enhances binding.

Had enough? The complexity is staggering and I haven’t even talked about recursive splicing — that’s for another post, but here’s a reference if you can’t wait — [ Nature vol. 521 pp. 300 – 301, 371 – 375, 376 – 379 ’15 ]. Pity the drug chemist figuring out which alternatively spliced form of a brain protein to attack (particularly if it hasn’t been studied for microExons).

Kuru continues to inform

Neurologists of my generation were fascinated with Kuru, a disease of the (formerly) obscure Fore tribe of New Guinea. Who would have thought they would tell us a good deal about protein structure and dynamics?

It is a fascinating story including a Nobelist pedophile (Carleton Gajdusek) and another (future) Nobelist who I probably ate lunch with when we were both medical students in the same Medical Fraternity but don’t remember –

Kuru is a horrible neurodegeneration starting with incoordination, followed by dementia and death in a vegetative state in 4 months to 2 years. For the cognoscenti — the pathology is neuronal loss, astrocytosis, microglial proliferation, loss of myelinated fibers and the kuru plaque.

It is estimated that it killed 3,000 members of the 30,000 member tribe. The mode of transmission turned out to be ritual cannibalism (flesh of the dead was eaten by the living before burial). Once that stopped the disease disappeared.

It is a prion disease, e.g. a disease due to a protein (called PrP) we all have but in an abnormal conformation (called PrpSc). Like Vonnegut’s Ice-9 ( PrPSc causes normal PrP to assume its conformation, causing it to aggregate and form an insoluble mess. We still don’t know the structure of PrPSc (because it’s an insoluble mess). Even now, “the detailed structure of PrPSc remains unresolved” but ‘it seems to be’ very similar to amyloid [ Nature vol. 512 pp. 32 – 34 ’14]. Not only that, but we don’t know what PrP actually does, and mice with no PrP at all are normal [ Nature vol. 365 p. 386 ’93 ]. For much more on prions please see

Prusiner’s idea that prion diseases were due to a protein, with no DNA or RNA involved met with incredible resistance for several reasons. This was the era of DNA makes RNA makes protein, and Prisoner was asking us to believe that a protein could essentially reproduce without any DNA or RNA. This was also the era in which X-ray crystallography was showing us ‘the’ structure of proteins, and it was hard to accept that there could be more than one.

There are several other prion diseases of humans (all horrible) — mad cow disease, Jakob Creutzfeldt disease, Familial fatal insomnia, etc. etc. and others in animals. All involve the same protein PrP.

One can take brain homogenates for an infected animal, inoculate it into a normal animal and watch progressive formation of PrPSc insoluble aggregates and neurodegeneration. A huge research effort has gone into purifying these homogenates so the possibility of any DNA or RNA causing the problem is very low. There still is one hold out — Laura Manuelidis who would have been a classmate had I gone to Yale Med instead of Penn. n

Enter [ Nature vol. 522 pp. 423 – 424, 478 – 481 ’15 ] which continued to study the genetic makeup of the Fore tribe. In an excellent example of natural selection in action, a new variant of PrP appeared in the tribe. At amino acid #127, valine is substituted for glycine (G127V is how this sort of thing is notated). Don’t be confused if you’re somewhat conversant with the literature — we all have a polymorphism at amino acid #129 of the protein, which can be either methionine or valine. It is thought that people with one methionine and one valine on each gene at 129 were somewhat protected against prion disease (presumably it affects the binding between identical prion proteins required for conformational change to PrPSc.

What’s the big deal? Well, this work shows that mice with one copy of V127 are protected against kuru prions. The really impressive point is that the mice are also protected against variant Creutzfedlt disease prions. Mice with two copies of V127 are completely protected against all forms of human prion disease . So something about V/V at #127 prevents the conformation change to PrPSc. We don’t know what it is as the normal structure of the variant hasn’t been determined as yet.

This is quite exciting, and work is certain to go on to find short peptide sequences mimicking the conformation around #127 to see if they’ll also work against prion diseases.

This won’t be a huge advance for the population at large, as prion diseases, as classically known, are quite rare. Creutzfeldt disease hits 1 person out of a million each year.

There are far bigger fish to fry however. There is some evidence that the neurofibrillary tangles (tau protein) of Alzheimer’s disease and the Lewy bodies (alpha-Synuclein) of Parkinsonism, spread cell to cell by a ‘prionlike’ mechanism [ Nature vol.485 pp. 651 – 655 ’12, Neuron vol. 73 pp. 1204 – 1215 ’12 ]. Could this sort of thing be blocked by a small amino acid change in one of them (or better a small drug like peptide?).

Stay tuned.

Are you as smart as the (inanimate) blind watchmaker

Here’s a problem the cell has solved. Can you? Figure out a way to send a protein to two different membranes in the cell (the membrane encoding it { aka the plasma membrane }, and the endoplasmic reticulum) in the proportions you wish.

The proteins must have exactly the same sequence and content of amino acids, ruling out alternative splicing of exons in the mRNA (if this is Greek to you have a look at the following post — and the others collected under —

The following article tells you how the cell does it. Recall that not all of the messenger RNA (mRNA) is translated into protein. The ribosome latches on to the 5′ end of the mRNA,  subsequently moving toward the 3′ end until it finds the initiator codon (AUG which codes for methionine). This means that there is a 5′ untranslated region (5′ UTR). It then continues moving 3′ ward stitching amino acids together.  Similarly after the ribosome reaches the last codon (one of 3 stop codons) there is a 3′ untranslated region (3′ UTR) of the mRNA. The 3′ UTR isn’t left alone but is cleaved and a polyAdenine tail added to it. The 3′ UTR is where most microRNAs bind controlling mRNA stability (hence the amount of protein produced from a given mRNA).

The trick used by the cell is described in [ Nature vol. 522 pp. 363 – 367 ’15 ]. The 3’UTR is alternatively processed producing a variety of short and long 3’UTRs. One such protein where this happens is CD47 — which is found on the surface of most cells where it stops the cell from being eaten by scavenger cells such as macrophages. The long 3′ UTR of CD47 allows efficient cell surface expression, while the short 3′ UTR localizes it to the endoplasmic reticulum.

How could this possibly work? Once the protein is translated by the ribosome, it leaves the ribosome and the mRNA doesn’t it? Not quite.

They say that the long 3′ UTR of CD47 acts as a scaffold to recruit a protein complex which contains HuR (aka ELAVL1), an RNA binding protein and SET to the site of translation. The allows interaction of SET with the newly translated cytoplasmic domains of CD47, resulting in subsequent translocation of CD47 to the plasma membrane via activated RAC1.

The short 3′ UTR of CD47 doesn’t have the sequence binding HuR and SET, so CD47 doesn’t get to the plasma membrane, rather to the endoplasmic reticulum.

The mechanism may be quite general as HuR binds to thousands of mRNAs. The paper gives two more examples of proteins where this happens.

It’s also worth noting that all this exquisite control, does NOT involve covalent bond formation and breakage (e.g. not what we consider classic chemical reactions). Instead it’s the dance of one large molecular object binding to another in other ways. The classic chemist isn’t smiling. The physical chemist is.

The twists and turns of topoisomerase (pun intended)

It is very sad that my late friend Nick Cozzarelli isn’t around to enjoy the latest exploits of the enzyme class he did so much great work on — the topoisomerases. For a social note about him see the end of the post.

We tend to be quite glib about just what goes on inside a nucleus when DNA is opened up and transcribed into mRNA by RNA polymerase II (Pol II). We think of DNA has a linear sequence of 4 different elements (which it is) and stop there. But DNA is a double helix, and the two strands of the helix wind around each other every 10 elements (nucleotides), meaning that within the confines of our nuclei this happens 320,000,000 times.

I’ve written a series of six posts on what we would see if our nuclei were enlarged  by a factor of 100,000 (which is the amount of compaction our DNA must undergo to fit inside the 10 micron (10 millionths of a meter) in diameter nucleus (since if fully extended our DNA would be 1 meter long. So if you compacted the distance from New York to Seattle (2840 miles or 14,995,200 feet) down by this factor you’d get a sphere 150 feet in diameter or half the length of a football (US) field. Now imagine blowing up the diameter and length of the DNA helix by 100,000 and you’d get something looking like a 2,840 mil long strand of linguini which twists on itself  320,000,000 times. The two strands are 3/8th of an inch thick. They twist around each other every 9/16ths of an inch.

For the gory details start at and follow the links.

Well, we know that for DNA to be copied into mRNA it must be untwisted, the strands separated so RNA polymerase II (Pol II) can get to it.  Pol II is enormous — a mass of 500 kiloDaltons and 7 times thicker at 140 Angstroms than the DNA helix of 20 Angstrom thickness.

Consider the fos gene (which we’ll be talking about later). It contains 380 amino acids (meaning that the gene contains at least 1140 nucleotides ). The actual gene is longer because of introns (3,461 nucleotides), which means that the gene contains 346 complete turns of the double helix, all of which must be unwound to transcribe it into mRNA.

So it’s time for an experiment. Get about 3 feet of cord roughly 3/8 of an inch thick. Tie the ends together, loop one end around a hook in your closet, put a pencil in the other end and rotate it about 100 times (or until you get tired). Keeping everything the same, have a friend put another pencil between the two strands in the middle, separating them. Now pull on the strands to make the separation wider and move the middle pencil toward one end. In the direction of motion the stands will coil even tighter (supercoiling) and behind they’ll unwind.

This should make it harder for Pol II to do its work (or for enzymes which copy DNA to more DNA). This is where the various topoisomerase come in. They cut DNA allowing supercoils to unwind. They remain attached to the DNA they cut so that the DNA can be put back together. There are basically two classes of topoisomerase — Type I topoisomerase cuts one strand, leaving the other intact, type II cuts both.

Who would have thought that type II topoisomerase would be involved in the day to day function of our brain.

Neurons are extended things, with information flowing from dendrites on one side of the cell body to much longer axons on the other. The flow involves depolarization of the cell body as impulses travel toward the axon. We know that certain genes are turned on by this activity (e.g. the DNA coding for the protein is transcribed into mRNA which is translated into protein by the ribosome). They are called activity dependent genes.

This is where [ Cell vol. 1496 – 1498, 1592 – 1605 ’15 ] comes in. Prior to neuronal activity, when activity dependent genes are expressed at low levels, the genes still show the hallmarks of highly expressed genes (e.g. binding by transcription factors and RNA polymerase II, Histone H3 trimethylation of lysine #4 {H3K4Me3 } at promoters).

This work shows that such genes are highly negatively supercoiled (see above) preventing RNA polymerase II (Pol II) from extending into the gene body. On depolarization of the cell body in some way Topoisomerase IIB is activated, leading to double strand breaks (dsbs) within promoters allowing the DNA to unwind and Pol II to productively elongate through gene bodies.

There is evidence that neuronal stimulation leads to dsbs ( Nature NeuroScience vol. 16 pp. 613 – 621 ’13 ) throughout the transcription of immediate early genes (e.g. genes turned on by neural activity). The evidence is that there is phosphorylation of serine #139 on histone variant H2AX (gammaH2AX) which is a chromatin mark deposited on adjacent histones by the DNA damage response pathway immediately after DSBs are found.

Etoposide (a topoisomerase inhibitor) traps the enzyme in a state where it remains bound to the DNA of the dsb. On etoposide Rx, there is an increase in activity dependent genes (Fos, FosB, Npas4). Inhibition of topiosomerase IIB (the most prevalent topoisomerase in neurons) by RNA interference (RNAi) leads to blunted activity dependent induction of these genes. This implies that DNA cutting by topoisomerase IIB is required for gene activation in response to neuronal activity.  Other evidence is that knocking down topoisomerase  using RNA interference (RNAi) stops activity dependent gene transcription.

Further supporting this idea, the authors induced dsbs at promoters of activity dependent genes (Fos, fosB, Npas4) using the CRISPR system. A significant increase in transcription was found when the Fos promoter was targeted.

I frankly find this incredible. Double strand breaks are considered bad things for good reason and the cell mounts huge redundant machines to repair them, yet apparently neurons, the longest lived cells in our bodies are doing this day in and day out. The work is so fantastic that it needs to be replicated.

Social Note: Nick Cozzarelli is one of the reasons Princeton was such a great institution back in the 50s (and hopefully still is). Nick’s father was an immigrant shoemaker living in Jersey City, N. J. Princeton recognized his talent, took him in, allowing him to work his way through on scholarship, waiting tables in commons, etc. etc. He obtained a PhD in biochemistry from Harvard and later became a prof at Berkeley, where he edited the Proceedings of the National Academy of Sciences USA for 10 years. He passed away far too soon of Burkitt’s lymphoma in his late 60s. We were friends as undergraduates and in grad school.

I can only wonder what Nick would say about the latest twists of the topoisomerase story


Get every new post delivered to your Inbox.

Join 80 other followers