Chip Wars by Chris Miller — Part II beating Russia with silicon

Just about everything in this post is from Chip Wars by Chris Miller, some are direct quotes, others are paraphrases.   A few things are my own, but they’re pretty obvious.

Even though I was vitally interested in computers as a neurologist starting in the early 70s and got one as soon as I could afford it (an Alpha Micro which I really couldn’t), Chip Wars covering the early history of silicon based computers taught me a lot.  It starts with Shockley’s invention of the transistor in 1948 and goes from there.   I won’t try to summarize that, but if you want an extremely well written early history of the period, read this book.  It isn’t dry and the personalities of the main characters are well fleshed out.

What I didn’t realize was just how much of the early development of silicon computers was driven by the military.  In particular Defense Advanced Research Projects Agency (DARPA) funded a lot of academics and their research into computation. Today every chip company uses tools from one of 3 chip design companies founded and built by graduates of DARPA programs

Guided munitions during the early Vietnam war used vacuum tubes that were hand soldered (Sparrow III) they broke down 2/3 of the time only 10% hit their target.  Bombs fell 420 feet from target.    Some 800 bombs had tried to take out a bridge in Vietnam and failed.   A set of wings was added to direct the bomb’s flight along with a laser guidance system which worked as follows.  A small silicon wafer was divided into 4 quadrants and placed behind a lens.  The laser reflecting off the target would shine through the lens onto the silicon,.  If the bomb veered off course one quadrant would receive more of the laser’s energy than the others and circuits would move the wings to reorient the bomb’s trajectory so the laser was shining straight through the lens.    A simple laser sensor and a few transistors made the bomb accurate.  This was 1972.  Spoiler alert — we still lost.

Photolithography is a crucial technology for drawing circuits on silicon, and it is mentioned many times throughout the book.   It’s basically a simple idea— turning a microscope lens upside down to make something big look smaller.  It was initiated by  Lathrop at Texas Instruments in 1958  It took thousands of experiments and a lot of interaction with suppliers etc to make it work.   There will be much more about photolithography in the next posts.

When owning a copy machine was a crime and no one without security clearance could use a computer, Russia had no way to educate a truly huge number of computer programmers.  So they basically used espionage to copy our technology.

This didn’t work.   The Soviet copy it strategy was flawed, because they couldn’t scale up the manufacturing process reliably, something Grove at Intel and Chang at Texas Instruments fixated on and spent countless hours improving.  Moreover the US had access to technology in optics, chemistry purified materials.  They knew the temperature at which chemicals needed to be heated or how long photoresist should be exposed to light.  Every step of the process of making chips involved specialized knowledge that was rarely shared outside a specific company — often not written down.    So it couldn’t be stolen.

As Silicon valley crammed more transistors onto silicon chips, building them became steadily harder.    Russia stole the equipment to make them, but they had no way to get spare parts.  The Russian military didn’t trust the chips produced in country, so they minimized the use of electronics and computers in military systems.    The math they put into guidance computers was simpler to minimize the strain on the onboard computer.

The copy it strategy left the Russians 5 years behind.

Russia’s defense chief of staff Ogarkov knew this and said in ’83 to Leslie Gelb  “The Cold War is over and you have won”.  Interestingly, Star Wars begun the same year is nowhere mentioned in the book.  It was derided as implausible, but the Russians knew they couldn’t match it.

The Soviets only customer for computers was the military, while the US had a large civilian market which created companies with a wide variety of expertise in everything needed for them (pure silicon wafers, advance optic for lithography).  The Russians also had no international supply chain. The Gulf War ’91 on Saddam Hussein — US smart weapons (Paveway) using more advanced electrons decimated the best equipment of Russia.

As I said in part I (https://luysii.wordpress.com/2024/03/10/chip-wars-by-chris-miller-part-i/) Chip Wars is really about manufacturing, not the abstract computational problems and programming I was interested in.

The denouemont came in 1991 with the Gulf War. US smart weapons (Paveway) using more advanced electronics decimated the best equipment of Russia.

Things haven’t changed much in Russia. p335

“The fact that Russia faced shortages of guided cruise missiles within several weeks of attacking Ukraine is partly due to the sorry state of its semiconductor industry.”

Next up: The coming competition with China

It passeth all understanding

While we were away in Taiwan we were mercifully away from the internet as well. So playing catchup in midFebruary, I was pleased to see that Cassava Sciences had released an open label trial of two years on Simufilam showing no cognitive decline in 47 patients with mild Alzheimer’s (defined as MMSE of 21 – 26) when a yearly decline of 3 on the ADAS-Cog would have been expected.  I thought that the stock would have exploded, but nothing happened.  It passeth all understanding.  Remember that two monoclonal antibodies have been approved by the FDA which decreased the rate of decline by 25 – 33%.  Perhaps the  patients in the antibody studies were sicker, as simufilam for 2 years didn’t help 32 patients with moderate Alzheimer’s (defined as an MMSE of 16 – 20) who declined by the expected 11 points on ADAS-Cog in 2 years.

Stabilizing early Alzheimer’s disease for 2 years is definitely a therapy worth having.  It will be fascinating to see what this group does in the next two years.

Perhaps the onslaught of negative articles in the press and Science have taken their toll and nothing Cassava Sciences says is credible.   Patient observation in the double blind placebo controlled study of Simufilam will be complete in the second half of this year, and analysis of the data (done by an outside group having nothing to do with Cassava Sciences) in a study of this sort usually takes 1 – 3 months, so we will have a definitive answer around New Year’s.  Note also that the FDA has accepted the way the study is to be performed, so there will be no requests for additional work at its conclusion.

A brand new theory of traumatic cerebral edema

I saw and tried to treat traumatic cerebral edema (brain swelling) in the 3 and half years I worked with 2 very busy neurosurgeons.  A paper this month in Nature (volume 623 pp. 992 – 1000 ’23) argues that we’ve had  its cause  all wrong.

Neurologists are well aware of brain swelling around a stroke because of damage to blood vessels by the stroke.  Such swelling is localized, but can get worse and worse (e. g. larger and larger) and kill the patient by compressing the brain stem.  Similarly brain swelling around tumors is (initially) local to them.  Both are thought to be due to leaky abnormal blood vessels

The Nature paper argues that acute post-traumatic brain edema is to due to suppression of glymphatic and lymphatic flow.  Since the glymphatic system was only described 10 years it’s time for some background.

First — a bit of history. The tissue of the brain is so tightly packed that it is impossible to see the cells that make it up with the usual stains used by light microscopists. People saw nuclei all right but they thought the brain was a mass of tissue with nuclei embedded in it (like a slime mold). Muscle is like that — long fibers with hundreds of nuclei here and there. It wasn’t until that late 1800′s that Camillo Golgi developed a stain which would now and then outline a neuron with all its processes. Another anatomist (Ramon Santiago y Cajal) used Golgi’s technique and argued with Golgi that yes the brain was made of cells. Fascinating that Golgi, the man responsible for showing nerve cells, didn’t buy it. This was a very hot issue at the time, and the two received a joint Nobel prize in 1906 (only 5 years after the prizes began).

How tightly packed are the cells in the brain? The shortest wavelength of visible light is 4000 Angstroms. Cells in the brain are packed far more tightly. To see the space between the brain cell external membranes you need an electron microscope (EM). Just preparing a sample for EM really fries the tissue. Neurons are packed together with less than 1000 Angstroms between them. So how much of this is artifact of preparation for electron microscopy has never been clear to me. One study injected a series of quantum dots of known diameter into the cerebral spinal fluid (CSF) to see the smallest sized dot that could insinuate itself between neurons [ Proc. Natl. Acad. Sci. vol. 103 pp. 5567 – 5572 ’06 ]. The upper limit was around 350 Angstroms. No wonder the issue was contentious when all they had was light microscopy.

Surprisingly, the PNAS paper comes up with an estimate that brain extracellular space comprises 20% of brain volume. I find this hard to accept given the above. So how does the brain get rid of waste products? It turns out that there is a circulation of cerebrospinal fluid (CSF) of sorts. Inject a tracer that you can follow into the CSF. After a period of time the tracer enters the brain along arteries (not veins) and after still more time it leaves the brain along the  horribly named glymphatic system which drains into the cervical lymphatic vessels.

Rather than leaky vessels the Nature paper above says that traumatic brain swelling is due to a suppression of glymphatic flow which gets out of the brain through lymphatic vessels.  This is held to be a response to excessive systemic release of norepinephrine, a well known response to trauma of any sort.  That’s pretty far out, except the the authors were able to back up their idea by attenuating (not cure) traumatic cerebral edema (in experimental animals) by blocking the sympathetic nervous system.  Norepinephrine binds to a variety of receptors in the body (called adrenergic receptors), so the authors used 3 different drugs

propranolol (Inderal) which blocks beta adrenergic receptors

prazosin (Minipress) which blocks alpha1 adrenergic receptors

atipamezole which blocks alpha2 adrenergic receptors.

I’m sure some plucky neurosurgeon will try it out, as the therapy we used (high dose corticosteroids) back then  didn’t always work for traumatic brain edema, but it usually worked beautifully for the edema around tumors.

Opiate receptors 51 years on

It seems like only yesterday that Candace Pert found the morphine receptor  as a graduate student at Johns Hopkins 2 years after graduating BrynMawr at 24 and getting (academically) screwed by Solomon Snyder, who may have a department named after him at Hopkins but who will never get the Nobel prize.

Things seemed so simple back then, and the route to the nonaddicting opiate seemed clear.  Well its 51 years and counting since 1972 and things have become incredibly complicated as we learn more and more about our opiate receptors (there are 4).

A marvelous review is available (if you have a subscription) replete with multiple cryoEM structures of multiple receptors with multiple ligands at  resolution approaching amino acid size (better than 3.5 Angstroms).  Cell vol. 186 pp. 5203 – 5219 ’23.

Just in terms of combinatorial size, the possibilities are quite large.

There are 4 types of G Protein Coupled Receptors (GPCRs) binding opiate peptides — mu, delta, kappa and nociceptive.  Although we have all sorts of small molecule drugs binding to them (morphine, heroin, fentanyl and worse), their natural ligands in our brains are protein fragments (of which there are 20) derived from 4 precursor proteins) — so that’s 80 possibilities there.

To get anything done inside the brain, each of the 4 GPCRs binds to G proteins of the G(i/o) class of which there are 7.

But wait, there’s more.  All good things come to an end, and to stop signaling the intracellular part of the GPCR is phosphorylated by the awfully named GRKs (G receptor kinases) of which there are 7.

Once the G proteins are phosphorylated, they leave the intracellular part of the GPCR, and another protein (arrestin) binds to the same area of the GPCR.  There are 4 known arrestins.

The possibilities are multiplicative since they’re independent — so its

20 x 4 x 7 x 7 x 4 = 15,680 different possible interactions.

For a long time it was thought that arrestins terminated opiate peptide signaling, dragging the receptor inside the cell and schlepping it to either the lysosome or the proteasome where it was then destroyed by proteolysis.

Not so.  After arrestin binding, the opiate receptors can be found in endosomes where they can continue to signal, so we’re not really sure just what the effects of the arrestins are on opiate signaling.

Now let’s hear it for the biochemical ingenuity of plants.  Even the smallest opiate peptide (met-enkephalin) has 5 amino acids, far too large to insert itself in the 7 alpha helices of the GPCR crossing the cell membrane.  So they bind to the extracellular surface of the GPCR, particularly extracellular loop 2 (ICL2) which contains 21 amino acids.  Yet plants have figured out how to make small molecules like morphine which bind to the parts of the GPCR in the cell membrane.  It’s hard for me to see an evolutionary push (selection pressure) for them to do this.

Well that’s a rather broad overview of what’s in the paper, but there is much much more.  Reading it is like eating intellectual fruitcake — it is far too dense to be ingested and digested at one sitting.

Two further tidbits to whet your interest.  The paper contains a detailed discussion (with pictures of the structures) of why fentanyl is so much more potent than morphine.  But of course such things require as knowledge of organic chemistry.  The benzene ring of Fentanyl engages in direct pi pi electron interactions with the toggle switch amino acids tryptophans #295 and #328 (W295 and W328).  Benzene doesn’t contain an isolated benzene moiety.  Also the phenylethyl group of Fentanyl interacts hydrophobically with a minor binding pocket of morphine found between transmembrane 2 (TM2) and TM3.     Meat and drink for an old organic chemist such as yours truly.

Tidbit #2.  On activation of the opiate GPCR, there is inward movement of TM5 and outward movement of TM6 along with clockwise rotations of TM6.  This is initiated by ligand reconstruction of the polar network in the binding pocket, the collapse of the sodium pocket and rearrangement of the proline, isoleucine and phenylalanine triad (all 3 found on 3 different transmembrane segments (TMs).

Put the multiple structures shown in the paper are thousands of times more instructive than the previous two paragraphs, confirming yet again the old adage.

Neurologic Velcro

Neurons must in some way respond to mechanical stimuli or you wouldn’t have a sense of touch.  A remarkable ion channel called piezo2 in the axon membrane opens in response to stretch of the axon causing a nerve impulse to fire.  They are huge proteins with anywhere from 2,100 to 4,700 amino acids.  Three come together to form an ion channel.  Each monomer contains 35 or so transmembrane segments, arranged in an arc.  (If I could ever figure out how to get an image from a paper or captured on the web into a WordPress document, I’ll update this post, and I’m going to work on it this coming weekend. ) In the meantime content yourself with figure 1 from this https://www.nature.com/articles/s41586-019-1505-8.  The 3 arcs are called blades and are made mostly of alpha helices.   The complex is huge, fitting into a circle of diameter 270 Angstroms.   The structure distorts the membrane, indenting it so the ion is channel formed by the junction of the 3 monomers is closed.  Stretch the membrane and the indentation disappears opening the channel, ions flow inside the axon and a nerve impulse is fired.

Enter Neuron vol. 111 pp. 3211 – 3229 ’23 which has fantastic pictures of 3 types of mechanically sensitive sense organs — the Pacinian corpuscle, the Meissner corpuscle and the lanceolate complex found around hair follicles.  Each picks up a different type of mechanical stimulus, yet they all use the same ion channel — piezo2.

The answer was totally unexpected — each axon has hundreds to thousands of small (1 micron) projections (containing piezo2) looking exactly like velcro (see figure 8 p. 3225) and these are tacked to the nonneuronal cells of the sense organ by adherens junctions, so that when any part of the sense organ is moved piezo2 is stretched and the axon fires.  Again apologies — I hope to get these figures in here over the weekend.

So the specificity of the sense organ for the type of mechanical stimulus doesn’t lie in the neuron’s axon which is the same in all 3, but in the non-neuronal cells it is attached to.

In defense of Cassava Sciences — Part II — clinical data

Cassava Sciences has been under attack  since 8/21 reaching a crescendo recently — https://www.science.org/content/article/co-developer-cassava-s-potential-alzheimer-s-drug-cited-egregious-misconduct and https://www.science.org/content/blog-post/saga-cassava.  Both contain demands that their current double blind placebo controlled studies on Simufilam, their Alzheimer drug, be stopped.

Full disclosure.  I do not own any Cassava stock and have not received anything from them for writing about them (aside from a free meal from Lindsay Burns 11/21 at the CTAD (Clinical Trials in Alzheimer’s Disease) in Boston.   There was and is no quid pro quo about anything I’ve written.

I do know Derek Lowe and have spent several pleasant afternoons discussing organic chemistry and drug development with him at my cousin’s New Year’s Day bash, before COVID put a stop to the affair.

I’ve known Lindsay since she was a teenager, when I was practicing neurology in Montana.  This was primarily through her parents who were sheep ranchers and good friends of my wife and me.

The basic position of the two articles above, is that some of the protein electrophoreses backing up Cassava’s model of Simufilam were either fudged or missing.

For details about Cassava’s model and the science behind it please see — https://luysii.wordpress.com/2023/04/23/the-science-behind-cassava-sciences-sava-the-latest-as-of-19-april-23/

Part I concerned models of disease and their usefulness.  This post concerns Cassava’s clinical data as interpreted by a clinical neurologist (me) with decades of clinical experience attempting to manage Alzheimer’s disease and other dementia (treat is too strong a word).  Derek is an organic chemist, Piller is a journalist.  Neither has any clinical experience with Alzheimer’s disease as far as I know.

I became excited about Cassava’s data in August 2021 when they released their data on the first 50 Alzheimer patients to complete 9 months on Simufilam (their drug) 100 milligrams twice daily.  Five of the 50 had close to 50% improvement in their ADAS-Cog score, which is used to measure changes in cognitive functioning in clinical trials.

Only a clinician dealing with demented patients would realize how unprecedented this is.  Demented patients don’t improve, particularly over 9 months although, like all of us, they fluctuate from day to day.

The following link contains a description of ADAS-Cog along with a link to Cassava’s data as reported 8/21 and much further discussion — https://luysii.wordpress.com/2021/08/25/cassava-sciences-9-month-data-is-probably-better-than-they-realize/

Over the 3+ decades as an active clinical neurologist I’d see at least one demented patient a week — that’s 1,560 patients.  And I never saw anything like this.

Not only that, but no therapy in current use including the monoclonal antibodies against aBeta produces improvement like this, even in a small subset — for details please see https://luysii.wordpress.com/2023/10/22/in-defense-of-cassava-sciences-part-i-models/

Well there are several possible explanations of this.

l. Cassava is lying.

2. The subjects didn’t really have Alzheimer’s disease and just entered the study because they were paid to do so.

3. Placebo effect.

The last is the easiest to deal with, based on my experience with Cognex (tacrine) released 30 years ago.  Initially Cognex was touted as helping Alzheimer’s disease (e.g. improving thinking and memory) although later it was held to slow the decline.   The local medical school advertised it aggressively (primarily as a marketing device).

So I put my Alzheimer patients on Cognex.  I wanted them to get better. They wanted to get better, and their families and caregivers certainly did.  Just about all of them thought it might have helped on followup visits in the first month.  I couldn’t see much difference.  By the second month, they weren’t sure, and later in the first year they didn’t think it helped, and most weren’t using the drug after 1 year.

Not only that, but I was in a call group with 4 other neurologists, and they saw exactly the same thing.  I was practicing in an area with a catchment area of over a million people, and every local neurologist I talked to had the same experience. People thought that “Cognex helped” for a month or two and then they didn’t

This a classic example of a placebo effect.  Moreover it occurred in a therapeutic trial for Alzheimer’s disease.  Crucially, the placebo effect was quite transient and  absent at 1 year.

Having said this, Cassava’s results in an open label trial of Simufilam in 200 patients are even more unprecedented. “47% of patients improved on ADAS-Cog over 1 year, and this group improved by 4.7 points”

Any open label study without controls is subject to the reasonable criticism that any benefits seen are due to the placebo effect.  Neurologic and Psychiatric disease trials can have large (33%) placebo effects (e.g. migraine, depression).  This is unlikely to be true of the open label trial based on my clinical experience with Cognex.

Currently recruitment for two trials of Simufilam is nearly complete, and all that remains is to watch and wait for the results.  It is to be noted that the FDA has approved the trials as they stand and will accept the results without further trial modification.

This is why attempts to shut down the trials are so appalling. There is no drug for Alzheimer’s with benefits like this.  Interim analysis of the trials for safety has turned nothing up.  We will know in a year or two when the trials conclude.

This is a fairly sanitized version of what has gone on in the past two years.  The amount of invective Lindsay and Cassava have had to bear is incredible.   Fortunately Lindsay is keeping a diary, and the day to day events will make a great movie (or better yet an opera).

For the paranoid among you, it is obvious who would benefit from shutting down the current trials — those with lousy drugs for Alzheimer’s — think the monoclonals.

 

Chronic fatigue syndrome, long COVID and a metabolic cause

As a neurologist I saw a lot of people who were chronically tired and fatigued, because neurologists deal with muscle weakness and diseases like myasthenia gravis which are associated with fatigue.  Once I ruled out neuromuscular disease as a cause, I had nothing to offer then (nor did medicine).  Some of these patients were undoubtedly neurotic, but there was little question in my mind that many others had something wrong that medicine just hadn’t figured out yet — not that it hasn’t been trying.

Infections of almost any sort are associated with fatigue, most probably caused by components of the inflammatory response.  Anyone who’s gone through mononucleosis knows this.    The long search for an infectious cause of chronic fatigue syndrome (CFS) has had its ups and downs — particularly downs — see https://luysii.wordpress.com/2011/03/25/evil-scientists-create-virus-causing-chronic-fatigue-syndrome-in-lab/

At worst many people with these symptoms are written off as crazy; at best, diagnosed as depressed  and given antidepressants.  The fact that many of those given antidepressants feel better is far from conclusive, since most patients with chronic illnesses are somewhat depressed.

The above 3 paragraphs were written in 2017 when COVID19 was on nobody’s radar.  One definition of long COVID  (there are many) is symptoms persisting longer than 3 months after infection.  The symptoms of long COVID seem identical to chronic fatigue syndrome, and are just another piece of evidence that something is really wrong with these people.  I find this gratifying, and patients should too.

However, very little has turned up in a biochemical, metabolic, neurophysiologic sense to explain the fatigue.  r

One case of longstanding fatigue and exercise intolerance in a 38 year old women does have an explanation [ Proc. Natl. Acad. Sci. vol. 120 e2302738120 ’23 ].  Interestingly, the symptoms began after a case of mononucleosis at age 16.   There was over expression of Wiskott Aldrich Syndrome protein Family member 3 (WASF3 to you) in her muscles.   WASF3 disrupts mitochodrial respiratory supercomplex formation, so that the mitochondria make less ATP than the body needs.  Recall that you make and consume half your body weight of ATP each day.  ATP is where the metabolic rubber meets the road, and as much energy as food can give goes into making it.

Unfortunately the paper claims that immunoblots for WASF3 of 6 patients with chronic fatigue showed elevated levels compared to controls.  They show the blots and I’m unimpressed that there is much difference.

Time for other people to try to replicate this even though the cited work is from NIH

What makes us human (genetically at least) take 2

The 28 April 2023 issue of Science has a fabulous collection of papers about the genomes of 240 species of mammals (out of 6,500 total).  Now remember that only 2% of our 3,200,000,000 basepair genome is involved in coding for the amino acids making up our proteins.  So when Genome Wide Association Studies come up with a region linked to a trait or a disease, 80% of the time it’s in the 98%.  This is where comparison of our genome with the 240 comes in so handy.  11% of our genome is the same as these other species (rather than the expected 1%).  These fall into 100,000,000 different sites.

The ENCODE study defined Cis-Regulatory Elements (CREs) as small regions of the genome with evidence of regulatory.  Many of them are binding sites for transcription factors.  Far far from all of them are evolutionarily conserved (e.g. are the same in all mammalian species studied).  This implies that these elements emerged at the beginning of the mammalian lineage.  They predominate near genes functioning in metabolism and development.  The work found 439,000 deeply conserved candidate CREs (cCREs) accounting for 4% of the human genome.

The most constrained CREs  are involved in mRNA processing and embryonic development.

One of the cleverest papers in the bunch (and which I never would have thought of doing) looked for genome sequences which have been conserved in mammals but which have been lost in man (these are called hCONDELs), some are over 1 kiloBases long, but 96% are much smaller (under 20 nucleotides).   Only 5% of 10,032 hCONDELs are in protein coding regions.  Many involve transcription transcription factor binding sites — obviously this is rewiring of the control  structure of cellular function and metabolism. Transcription binding sites aren’t very long (usually under 50 nucleotides) which is why even a hCONDEL of under 20 nucleotides can have such a major functional effect on them.   Remember what makes us different from other species isn’t our proteins (bricks) but the plan — for more on this please see   https://luysii.wordpress.com/2015/09/27/it-aint-the-bricks-its-the-plan/.   Even more interesting, 80/800 hCONDELs changed function only in neural precursors (of the 6 cell types studied).

There is tons more in the issue.  Find a copy and read it.  It shows that to really understand humans we have to look outside ourselves (and our genomes).

Next up — primate genomes.  But first wish us a happy 59th wedding anniversary,.

The science behind Cassava Sciences (SAVA) — the latest as of 23 April ’23

It’s time for an update on the science  behind Cassava Sciences’ anti-Alzheimer drug, Simufilam.  It is  based on an older post of mine and a review of the published literature and my decades of experience as a clinical neurologist.

Disclaimer:  My wife and I have known Lindsay Burns, one of the Cassava Sciences principals since she was a teenager and we were friendly with her parents when I practiced neurology in Montana.

But as H. L. Mencken said, “A Professor must have a theory as a dog must have fleas”, and the reason I’m excited about Simufilam has nothing to do with the theory of the science behind it.  Simply put, the results of Cassava’s open label trial have never  been seen with Alzheimer’s patients.  10% improved by nearly 50% at 1 year, and over half did not deteriorate.  As a clinical neurologist with decades of experience seeing hundreds of demented people, I never saw anything like this, especially significant improvement after a year).  For more detail please see https://luysii.wordpress.com/2021/08/25/cassava-sciences-9-month-data-is-probably-better-than-they-realize/

Here is the science behind the drug.  We’ll start with the protein the drug is supposed to affect — filamin A, a very large protein (2,603 amino acids to be exact).  I’ve known about it for years because it crosslinks actin in muscle, and I read everything I could about it, starting back in the day when I ran a muscular dystrophy clinic in Montana.

Filamin binds actin by its amino terminal domain.  It forms a dimerization domain at its carboxy terminal end.  In between are 23 repeats of 96 amino acids which resemble immunoglobulin — forming a rod 800 Angstroms long.  The dimer forms a V with the actin binding domain at the two tips of the V, making it clear how it could link actin filaments together.

Immunoglobulins are good at binding things and 90 different proteins are known to which filamin A binds.  This is an enormous potential source of trouble.

As one might imagine, filamin A could have a lot of conformations in addition to the V, and the pictures shown in https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2099194/.

One such altered (from the V) conformation binds to the alpha7 nicotinic cholinergic receptor on the surface of neurons and Toll-Like Receptor 4 (TLR4) inside the cell.

Abeta42, the toxic peptide, has been known for years to bind tightly to the alpha7 nicotinic receptor — they say in the femtoMolar (10^-15 Molar) range, although I have my doubts as to whether such tiny concentration values are meaningful.  Let’s just say the binding is tight and that femtoMolar binding is tighter than picoMolar is tighter than nanoMolar is tighter than microMolar  binding etc., etc.

When aBeta42 binds to alpha7 on the outside of the neuronal plasma membrane  filamin A binds to alpha 7 on the inside making  aBeta42 binding even tighter.

The tight binding causes signaling inside the cell  to hyperphosphorylate the tau protein forming the neurofibrillary tangle, which is more directly correlated with dementia in Alzheimer’s disease than the number of senile plaques.

In more detail, the high affinity aBeta42-alpha7 nicotinic cholinergic receptor binding activates the MAPK cascade (Mitogen Activated Protein Kinase cascade), ending in activation of the protein kinases ERK2, and JNK1.  Activated protein kinases catalyze the addition of phosphate to proteins forming an ester with the free hydroxyl groups of serine and/or threonine.  Activating ERK2 and JNK1 allows them to phosphorylate the tau protein leading to the neurofibrillary tangle of  Alzheimer’s disease (which is just a mess of hyperphosphorylated tau protein).

But there is still more about the mechanism which isn’t clear.  Recall that MAPK stands for Mitogen Activated Protein Kinase where a mitogen binds to a receptor on the cell surface, and a mitogen is nowhere in sight here, so there are still a few missing steps between aBeta42 binding to the alpha7 nicotinic cholinergic receptor and MAPK activation.  The references do show that MAPK signaling, ERK2 and JNK1 are activated when aBeta42 binds to the alpha7 nicotinic acetyl choline receptor.

Also the mechanism is radical in the extreme. The nicotinic acetyl choline receptor is a receptor all right but for acetyl choline. It is an ion channel and   looks nothing like the receptors that proteins and peptides bind to which are usually G Protein Coupled Receptors (GPCRs) or Receptors with Tyrosine Kinase activity (RTKs).  Also aBeta42 is not a mitogen.

So what does Sumifilam actually do — it changes the ‘altered’ conformation of filamin A getting it away from the alpha7 acetyl choline receptor and “indirectly reducing the high femtoMolar binding affinity of aBeta42 for alpha7” (and however this binding triggers tau hyperphosphorylation)  How do they know the conformation of filamin A has changed?  They haven’t done cryoEM or Xray crystallography on the protein.  The only evidence for a change in conformation, is a change in the electrophoretic mobility (which is pretty good evidence, but I’d like to know what conformation is changed to what).

So there you have it, after a fairly deep dive into protein chemistry, cellular physiology and biochemistry, the current thinking of how Simufilam works.

But even if the theory is completely wrong, the data in the link above must be regarded with respect.  Clinical blinded studies are ongoing, and the soon to be released Cognition Maintenance Study should  give us more information –https://luysii.wordpress.com/2023/03/02/the-cognition-maintenance-study-of-simufilam/

An incredibly clever experiment

Every now and then an experiment comes along that is so clever and definitive that all you can do is admire it and tell your friends.

The question to be answered is do bees have a cognitive map of their surroundings?  E.g. do they know where other objects are in relation to the hive and, more importantly to each other?

How in the world would you ever figure out a way to answer the question?

Anyone who has ever taken a college biology course has probably heard of the waggle dance.   The dance is performed by bees not strippers.  It tells other members of the hive where the dancer has found food (by the direction the bee moves during the dance) and how far away it is (by the length of the dance).

Well you don’t need a cognitive map of everything in your surroundings to do that.  Just a given direction and a given distance is enough to find the food.

The clever part (and a technical tour de force) was putting something on the bee allowing its flight pattern to be detected, and then (after the bee had performed its dance), taking it and moving it to a position away from the hive and watching what it did (incredibly clever to think of that).  The bee (most of the time) made a beeline for the original target, even though the flight was in a different distance and of a different length.  Clearly the bee had to know where the target was in relation to the bee’s new position.  The only reasonable explanation is that the bee possessed a map of where things were in relation to arbitrary locations in its environment (not just the hive).

Here’s an article about the experiment — Proc. Natl. Acad. Sci. vol. 120 e2303202120 ’23 — and a link to it — https://www.pnas.org/doi/10.1073/pnas.2303202120  — hopefully not behind a pay wall.