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.

Carly’s cancer

Carly Florina had breast cancer surgery (bilateral mastectomy) 2 March 2009 at Stanford University Hospital followed by chemotherapy and radiation. She was given an excellent prognosis for full recovery —

So far so good and it’s coming up on 7 years. But it is reasonable to ask just what her prognosis really is, particularly as she may be our next president. I asked an old friend and colleague who has been involved in research on breast cancer and in many of the clinical trials of therapy over the past 35 years.

So I wrote the following — I’m writing you for some idea what the chances of someone with breast cancer being free for 6+ years (Carly’s surgery was 2/09) will be free for the next 5+? I know that there are all sorts of statistics on survival in breast cancer (because the cohort is so large). If anyone would know them, it would be you.

and got this back

Impossible to answer your question. Too many variables and NO DATA or info. Many people, docs and patients alike call ductal carcinoma in situ,” cancer” but cure rate is 99%. If she was one of those then, of course, she’s likely to be cured . Stage 1 ,luminal A tumors (even though real cancers) have excellent prognoses—probably > 90% cured. For other real cancers Lots depend on stage, hormone receptors ad infinitum. On thin ice lumping anyone into a broad statement without lots more info

just what you’d expect from an circumspect intelligent expert

So I dug a bit more and sent him this

I tried to find out just what type of breast cancer Carly had. No luck, but various newspaper articles show that she did receive postop chemo causing her hair to fall out as well as radiation. Would ductal carcinoma in situ (Dcis) be treated this way? Would stage 1 luminal A tumors be treated this way?

He replied

Dcis definitely no. luminal a probably shouldn’t be. Sounds like a significant cancer. Next issue is did she get antihormonal therapy. Estrogen receptor tumors are the ones that tend to relapse after 5 years. ER neg. tumors while more aggressive overall seldom recur >5 yrs after dx. The radiation part doesn’t mean much unless she had a mastectomy since all lumpectomy patients get radiation. – If she had mastectomy and chemo and radiation it was probably a poorer risk tumor. Even chemo might not be so bad—–we give chemo to node neg tumors which could end up with very good long term prognosis.AMONG RELAPSES in ER pos pts. 15% recur before year 5 and 17% recur after year 5. However overall likelihood of relapse depends on whether or not she had positive or negative nodes and was ER + or Neg. Sorry to be so wordy but prognosis has been improving steadily. I would guestimate that we’re curing about 70% of women with newly diagnosed breast cancer—excluding dcis who are virtually all cured.

I realized that I’d neglected to tell him that she’d had a bilateral mastectomy as well and got the following back after I did.

If she indeed had radiation after a mastectomy as well as chemo it speaks for a more aggressive presentation. Rule of thumb—-post mastectomy xrt reserved for patients with > 4 positive nodes or tumors >5 cm in size. Today, many are giving post mastectomy xrt to 1-3 positive nodes although that was very controversial for years . newer data impies benefit. So, just guessing, but she probably had positive nodes—a poorer prognostic sign for long term—but only if she was estrogen receptor pos. as noted in prior email.

So there you have it — she’s fortunately well presently, but the tumor and prognosis doesn’t sound that good. Still unknown are histologic type of the tumor, presence or absence of spread to lymph nodes (and if so how many), estrogen receptor positivity, which would certainly give us a better idea of her ultimate prognosis (and the country’s should she become president).

I take no pleasure in any of these posts. See Both Carly and Hillary are brilliant women it would be an honor to know and I wish them both the best. FYI Hillary was valedictorian of her class at Wellesley.

So why write about their potential health problems? Look at the sad saga of Hugo Chavez who claimed he was cured in July elected in the fall with death before he could take office in March of the following year — see Also consider the last months in office of Woodrow Wilson and Franklin Roosevelt and the results of the League of Nations and the Yalta conference when they were both impaired.

My wife asked my why I pick on female candidates, and I’ll address Christie’s massive obesity should he rise in the polls.

The amount of BS you can get published about the brain

Hype about brain therapies isn’t confined to second rank journals The following is a direct quote from Neuron vol. 87 pp. 940 – 941 ’15.

” When a journal of the standing of Nature carries a headline ‘‘Shocks to the brain improve mathematical abilities’’ (Callaway) concerning two studies, we have to ask how this will be perceived. It is doubtful that the non-scientific reader will note that the studies have not been independently replicated, that only one of them tested mathematics, that the gains are as small as being milliseconds faster at some simple sums, and that only six people were tested in follow-up.

When it is claimed that tDCS (transcranial Direct Current Stimulation) can improve problem-solving abilities (Chi and Snyder, Brain Res. vol. 1353 pp. 168 – 175 ’10), the casual reader will not notice that only one-third of subjects improved, that there is no evidence that effects are sustained beyond 3 min, that there was no active control stimulation, and that there was no control for order effects.”

Caveat emptor. The quote is from an excellent review about brain stimulation with either magnetic fields or electrical current (direct, alternating or random) — Neuron vol. 87 pp. 932 – 944 ’15 — e.g. Non-invasive human brain stimulation, by 3 of the adults in the room.

It gets worse — here’s part of an older post — with a link to the whole thing —

How badly are thy researchers, O default mode network

If you Google “default mode network” you get 32 million hits in under a second. This is what the brain is doing when we’re sitting quietly not carrying out some task. If you don’t know how we measure it using functional mMRI skip to the **** and then come back. I’m not a fan of functional MRI (fMRI), the pictures it produces are beautiful and seductive, and unfortunately not terribly repeatable.

If [ Neuron vol. 82 pp. 695 – 705 ’14 ] is true than all the work on the default network should be repeated.


Because they found that less than half of 71 subjects studied were stably awake after 5 minutes in the scanner. E.g. they were actually asleep part of the time.

How can they say this?

They used Polysomnography — which simultaneously measures tons of things — eye movements, oxygen saturation, EEG, muscle tone, respiration pulse; the gold standard for sleep studies on the patients while in the MRI scanner.

You don’t have to be a neuroscientist to know that cognition is rather different in wake and sleep.


Time to get busy

Well I asked for it (the answer sheets to my classmate’s book on general relativity). It came today all 347 pages of it + a small appendix “Light Orbits in the Schwarzschild Geometry”. It’s one of the few times the old school tie has actually been of some use. The real advantages of going to an elite school are (1) the education you can get if you want (2) the people you meet back then or subsequently. WRT #1 — the late 50s was the era of the “Gentleman’s C”.

It should be fun. The book is the exact opposite of the one I’d been working on which put the math front and center. This one puts the physics first and the math later on. I’m glad I’m reading it second because as an undergraduate and graduate student I became adept at mouthing mathematical incantations without really understanding what was going on. I think most of my math now is reasonably solid. I did make a lot of detours I probably didn’t need to make — manifold theory,some serious topology — but that was fun as well.

When you’re out there away from University studying on your own, you assume everything you don’t understand is due to your stupidity. This isn’t always the case (although it usually is), and I’ve found errors in just about every book I’ve studied hard, and my name features on errata web pages of most of them. For one example see

Reproducibility and its discontents

A widely cited (but unreproducible) therapeutic study caused the most difficult moral choice I ever faced in decades of neurologic practice. Lack of reproducibility is currently in the news big time thanks to the work of 270 or so volunteer psychologists [ Science vol. 349 pp. 910 – 911, 943, aa47161 to 8 ’15 (28 August) ]. They tried to reproduce 100 significant studies appearing in 3 high level psychology journals in 2008. They had the cooperation of the original authors. Sadly, they were only replicate slightly over 1/3, and the effect sizes of the original 100 were half that on replication, and in a few cases, the effects were opposite.

In what follows you will encounter reasoning of almost talmudic intricacy. Don’t give up. The real world effects were far from trivial, and undoubtedly some died of therapeutically induced intracerebral hemorrhage.

The study was large and done under NIH auspices [ New England J. Med. vol. 333 pp. 1581 – 1587 ’95 ] and had an extremely logical basis — if some strokes are due to blood clots blocking vessels, administer an agent to break up the clot and restore circulation, before the part of the brain lacking circulation irreversibly dies. The agent was Tissue Plasminogen Activator (tpa).

624 patients were studied (so it had significant statistical power to detect an effect) . Treatment had to be initiated within 3 hours (about this more later) As compared with patients receiving placebo, patients with tPA were 30% more likely to have minimal or no disability at 3 months (the absolute increase in good outcome was 11 – 13%) — e.g. without tpa roughly 30% of the placebo group had good outcomes at 3 months, with tpa the figure was 41 – 43%. So the therapy was no guarantee of success.

There was a huge amount of interest in the paper and letters to the NEJM soon followed. One from Mass. General noted that rapid neurologic improvement expected after recanalization was rare in the treated group (there were few differences at 24 hours between the two groups ). This surprised my cardiologist friends, long used to giving clot busting (thrombolytic) agents to patients with acute heart attacks (myocardial infarctions) due to coronary occlusion and watching EKG changes vanish before their eyes along with improved cardiac function.

Followup of the 624 was impressive (609), and results one year out showed that the 11 – 13% absolute increase in good outcome persisted [ New England J. Med. vol. 340 pp. 1781 -1787 ’99 ]. Of interest was the risk of another stroke during this time — was 6% in both treated and untreated. Hopefully that figure is lower now (I don’t know as I’m no longer in practice and reading the literature).

There are a variety of scales to measure stroke severity and stroke outcome. The tpa and placebo groups had the same average National Institutes of Health Stroke Severity scores (NIHSS scores) on entry to the study — the higher the number the more severe the stroke. The authors used something called the modified Rankin score to measure outcome (here lower is better).

There was a huge push to get everyone to use tpa for stroke hitting the ER in the first few hours. The evidence seemed pretty flimsy and the effect pretty minimal to me (and to most clinical neurologists that I talked to). Hospitals used it as an advertising tool. The legal profession geared up to sue docs who didn’t use it. Fortunately the criteria for use were quite strict (within 3 hours of stroke, no hypertension, etc. etc.). The academic industrial complex weighed in, with the author of one of the subsequent studies threatening to testify in malpractice cases against docs who didn’t use it.

There was a lot of resistance on the part of clinicians and ER docs to using the therapy, but we had no blogosphere then.

Results when the therapy was adopted in the community were not happy ones. Here’s just one example [ Stroke vol. 30 pp. 34 – 39 ’99 ] A study of 138 patients treated in Houston and Calgary from 12/95 to 7/98 showed a much higher rate of symptomatic intracerebral hemorrhage (9 % == 13/138) and the any hemorrhage rate was also high (30% — 42/138). The only predictor was blood glucose. 7/13 symptomatic hemorrhage patients died and the rest were severely disabled.

This was not a therapy to be undertaken lightly, particularly if you thought the evidence was flimsy.

A technique was available (TransCranial Doppler) to actually look at blood flow inside the head in someone with a stroke, and this was later applied to patients receiving tpa. It involved continuous ultrasound monitoring, which one plucky neurologist actually thought was breaking up the clot (rather than the tpa) Robinson, L. Clototripsy? Stroke 2000; 31 2024. Clototripsy is a play on lithotripsy a therapy in widespread use to focus sound energy on kidney stones to break them up.

It was suggested that a 3 pronged study be done on patients with acute stroke due to arterial occlusion (1) tpa alone (2) TransCranial Doppler alone (3) Tpa + TransCranial doppler. This was never done.

A later study [4] Alexandrov AV, Molina CA, Grotta JC, Garami Z, Ford, SR, Alvarez-Sabin J, Montaner J, Saqqur M, Demchuk AM, Moye LA, Hill MD, Wojner AW. Ultrasound-Enhanced Systemic Thrombolysis for Acute Ischemic StrokeN. Engl. J. Med. 2004; 351: 2170 – 2178
was devastating (although the authors didn’t realize it). Why?

29% of patients receiving tPA alone had a modified Rankin score of 0 – 1 at 3 months (e.g. a good result). But 29% is practically the same as 26% — the percentage of good outcome in patients in the original 1995 study who received no tPA whatsoever. A mere difference of 3 points on the initial NIHSS score between the two groups is not explanatory.  The 41 – 43% good outcome of the original study was nowhere in sight.

So tpa was actually no better than placebo when retested 9 years later. The authors had failed to replicate the original work.

I tried to get the following letter in the New England Journal back then.

To the editor:

The initial results of continuous 2 MHz transcranial Doppler (TCD) monitoring during tPA infusion for acute ischemic stroke were exciting and unprecedented. [1] Since no study up to then had replicated the NINDS results [2], I argued that the dramatic results with continuous TCD were probably real, and quite likely independent of tPA [3]. I urged that a three-pronged study (tPA alone, TCD alone, and tPA plus TCD) should be carried out.

The recent two-pronged study [4] also fails to replicate the NINDS study. Only 29% of patients receiving tPA alone had a modified Rankin score of 0 – 1 at 3 months. Even worse, 26% of patients in the NINDS study receiving no tPA whatsoever had this outcome. A mere difference of 3 points on the initial NIHSS score between the two groups is not explanatory. After 9 years our diagnostic techniques have improved and we are presumably smarter about whom to treat. It’s time to do the study I suggested and to stop regarding tPA for stroke as ‘standard of care’.

[1] Alexandrov AV, Demchuk AM, Felberg RA, Christou I, Barber PA, Burgin WS, Malkoff M, Wojner AW, Grotta JC. High rate of complete recanalization and dramatic clinical recovery during tPA infusion when continuously monitored with 2-MHz transcranial Doppler monitoring. Stroke. 2000; 31: 610 – 614.

[2] The National Institutes of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N. Engl. J. Med. 1995; 333: 1581 – 1587

[3] Robinson, L. Clototripsy? Stroke 2000; 31 2024

[4] Alexandrov AV, Molina CA, Grotta JC, Garami Z, Ford, SR, Alvarez-Sabin J, Montaner J, Saqqur M, Demchuk AM, Moye LA, Hill MD, Wojner AW. Ultrasound-Enhanced Systemic Thrombolysis for Acute Ischemic Stroke
N. Engl. J. Med. 2004; 351: 2170 – 2178

It was not published.

Thank God for the blogosphere. Often irritating and irascible but extremely useful.

So what do you think I did faced with a patient satisfying the criteria for tpa, knowing that the therapy could (and had) killed, and not thinking that it was much good. Satisfy my conscience and my responsibility to the patient by not giving it and risk financial destruction, or do what I thought was wrong and  protect my family with the full weight of medical legal precident to back me up.

What would you do in this situation?

It’s important to realize, that all this took place a decade ago, and that new information and studies may have come out, showing that tpa actually is of use in acute stroke.  I wouldn’t rely on meta-analyses of the old studies however.

Nonetheless, the above reasoning is still valid.  Docs do the best they can with the information available to them at the time (which is NEVER enough) and this was the state of play when I was in practice.

Band Camp for Adults — 2015

This is not a scientific post, although it contains a lot of scientific types. Adult amateur chamber musicians are an interesting lot. The festival just concluded contained two people with books coming out, two organic chemists, two math profs, an english prof, multiple MDs, a retired foreign service officer and those were just the people I played with. Not everyone attending these things is so fancy and one of the best amateur cellists I ever played with was a moving man and probably the best violist ever was a 300 pound jail matron.

I’d not been to this one for 11 years or so, and it was amazing how people remembered the things I’d done back then. One classic neurotic back then was quite worried she was crazy. Her friend said she’d repeat over and over what I told her — I know crazy and you’re not crazy. Unfortunately she died an awful death of metastatic ovarian carcinoma, far too young in her 40s, as did another good friend, an RN.

Someone else brought up what I’d done for her at her first time at the festival. I’d totally forgotten about it. The first time my wife an I went there, I wasn’t assigned a group the first day and we didn’t know a soul. Everyone else appeared to know everyone else and had play dates arranged for the rest of the week. So the evening of the first day my wife and I were moping about in a local bar, when a gregarious participant came up to us, found out what was going on and set up the Dvorak piano quintet for me the next day.

The following year after the general initial assembly was over, I and the gregarious one got up and announced that we wanted 7 first timers for an immediate session — the Schumann piano quintet for me and a Mozart string quartet for her. The process has since been institutionalized.

I even met a reader of the blog, an excellent young violinist and organic chemist who I tried to steer into drug design. Probably not a good idea given the employment upheavals appearing nearly daily in Derek’s blog. One of the things we played was Vaughn Williams 6 studies on English Folk song. They have been scored for piano and violin, piano and viola, piano and clarinet and piano and cello. Few seem to know of them. They are each two piano pages long, extremely interesting musically and just not that hard to sight read.

Another great thing about the site, is that there is a whole piano and percussion building and many rooms have two pianos. This means two pianists can get together and play without squeezing onto the same bench. I strongly recommend trying two transcriptions of Bach concerti in C major and C minor. Both parts are quite interesting and well done musically, and you can switch so you’ll get to play each part. It’s  Peters edition #s 2200a, 2200b (BWV 1061, no BWV # given for 2200b). Start with the slow movements of each, the back and forth of the voices is great. I heard it today scored as a concerto for oboe and violin.

Unfortunately many of the people I played with 11 years ago had passed on, including Edwin Gould a violinist. Organic chemists of a certain age know him has the author of the ‘bible’ of physical organic chemistry back in the 60s.

Just by chance, two of the MDs were at places I’d trained — Colorado General Hospital and Children’s Hospital of Philadelphia, and it was fascinating to hear how they’d changed.

Also just by chance there were two graduates from Brown who would have been near classmates with my son, had he chosen to go there. Fascinating to hear about paths not taken.

One of the math profs has a book coming out and the other explained what a Toric variety is (David Cox, from whom I audited a course, wrote a 600 page book on the subject).

All in all an intellectually and musically stimulating week.

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.


Get every new post delivered to your Inbox.

Join 80 other followers