Category Archives: Medicine in general

The staggering implications of one axon synapsing on another

It isn’t often that a single paper can change the way we think the brain works.  But such is the case for the paper described in the previous post (full copy below *** ) if the implications I draw from it are correct.

Unfortunately this post requires a deep dive into neuroanatomy, neurophysiology, neuropharmacology and cellular molecular biology.  I hope to put in enough background to make some of it comprehensible, but it is really written for the cognoscenti in these fields.

I’m pretty sure these thoughts are both original and unique

Briefly, the paper provided excellent evidence for one axon causing another to fire an impulse (an action potential).   The fireror was from a neuron using acetyl choline as a neurotransmitter, and the fireree was a dopamine axon going to the striatum.

Dopamine axons are special.  They go all over the brain. The cell body of the parent neuron of the axon to be synapsed on uses dopamine as a neurotransmitter.  It sits in the pars compacta of the substantia nigra a fair piece away from the target they studied. “Individual neurons of the pars compact are calculated to give rise to 4.5 meters of axons once all the branches are summed”  — [ Neuron vol. 96 p. 651 ’17 ].”  These axons release dopamine all over the brain.  There aren’t many dopamine neurons to begin with just 80,000 which is 1 millionth of the current (probably unreliable) estimate of the number of neurons in the brain 80,000,000,000.

Now synapses between neurons are easy to spot using electron microscopy.  The presynaptic terminal contains a bunch of small vesicles and is closely apposed (300 Angstroms — way below anything the our eyes can see) to the post synaptic neuron which also looks different, usually having a density just under the membrane (called, logically enough, post-synaptic density).  Embedded in the postsynaptic membrane are proteins which conduct ions such as Na+, K+, Cl- into the postsynaptic neuron triggering an action potential.

But the dopamine axons going all over the brain have a lot of presynaptic specialization, but the post-synaptic neuron and its postsynaptic density is nowhere to be found.  This is called volume neurotransmission.

The story doesn’t end with dopamine.  There are 3 other similar systems of small numbers of neurons collected into nuclei, using different neurotransmitters, but whose axons branch and branch so they go all over the brain.

These are the locus coeruleus which uses norepinephrine as a neurotransmitter, the dorsal raphe nucleus which uses serotonin and the basal nucleus of Meynert which uses acetyl choline.

What is so remarkable about the paper, that it allows the receiving neurons to (partially) control what dopamine input it gets.

But dopamine doesn’t work at the synapse, and the 5 receptors for it (called G Protein Coupled Receptors — GPCRs) aren’t found there. None of the GPCRs conduct ions or trigger action potentials (immediately anyway).  Instead, they produce their effects much more slowly and change the metabolism of the interior of the cell.

Neither does norepinephrine all of whose receptors are GPCRs.  Serotonin does have one of its 16 or so receptors which conducts ions, but the rest are GPCRs.

Acetyl choline does have one class of receptors (nicotinic) which conducts ions, and which the paper shows is what is triggering the axon on axon synapse.  The other class (muscarinic) of acetyl choline receptor is a GPCR.

We do know that the norepinephrine and serotonin axons work by volume neurotransmission (not sure about those of the basal nucleus of Meynert).

Now the paper tested axon to axon firing in one of the four systems (dopamine) in one of the places its axons goes (the striatum).  There is no question that the axons of all 4 systems ramify widely.

Suppose axon to axon firing is general, so a given region can control in someway how much dopamine/serotonin/norepinephrine/acetyl choline it is getting.

Does this remind you of any system you are familiar with?  Maybe, because my wife went to architecture school, it reminds me of an old apartment building, with separate systems to distribute electricity, plumbing, steam heat and water to each apartment, which controls how much of each it gets.

Perhaps these four systems are basically neurological utilities, necessary for  the function of the brain, but possibly irrelevant to the computations it is carrying out, like a mother heating a bottle for her baby in water on a gas stove on a cold winter night.  The nature of steam heat, electricity, water and gas tell you very little about what is going on in her apartment.

The paper is so new (the Neuron issue of 21 September) that more implications are sure to present themselves.

Quibbles are sure to arise.  One is that fact that the gray matter of our brain doesn’t contain much in the way of neurons using acetyl choline as a neurotransmitter.  What it does have is lots of neurons using GABA which we know can act on axons, inhibiting axon potential generation.  This has been well worked out with synapses where the axon emerges from the neuron cell body (the initial segment).

The work was done in living animals, so no microscopy is available showing the synapse. Such work is sure to be done.  No classical presynaptic apparatus may be present, just two naked axons touching each other and interacting by ephaptic transmission.

So a lot of work should be done, the first of which should be replication. As the late Carl Sagan said “extraordinary claims require extraordinary evidence”.


As Mark Twain said ” There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.”


Understanding the Riemann curvature tensor is like doing a spinal tap

Back in the day when I was doing spinal taps, I spent far more time setting them up (positioning the patient so that the long axis of the spinal column was parallel to the floor and the vertical axis of the recumbent patient was perpendicular to the floor) than actually doing the tap.  Why? because then, all I had to do was have the needle parallel to the floor, with no guessing about how to angle it when the patient had rolled (usually forward into the less than firm mattress of the hospital bed).

So it is with finally seeing what the Riemann curvature tensor actually is, why it is the way it is, and why the notation describing it is such a mess.  Finally on p. 290 of Needham’s marvelous book “Visual Differential Geometry and Forms” the Riemann curvature tensor is revealed in all its glory.  Understanding it takes far less time than understanding the mathematical (and geometric) scaffolding required to describe it, a la spinal taps.

Over the years while studying relativity, I’ve seen it in one form or other (always algebraic) without understanding what the algebra was really describing.

Needham will get you there, but you have to understand a lot of other things first. Fortunately almost all of them are described visually, so you see what the algebra is trying to describe.  Along the way you will see what the Lie bracket of two vector fields is all about along with holonomy.  And you will really understand what curvature is.  And Needham will give you 3 ways to understand parallel transport (which underlies everything — thanks Ashutosh)

Needham starts off with Gauss’s definition of curvature of a surface — the angular excess of a triangle, divided by its area.

Here is why this definition is enough to show you why the surface of a sphere is curved.   Go to the equator.  Mark point one, then point two 1/4 of the way around the sphere.  Form longitudes (perpendiculars) there and extend them as great circles toward the North pole. You now have a triangle containing 3 right angles, (clearly an angular excess from Euclid who states that the sum the angles of a triangle is two right angles).  The reason, of course, is because the sphere is curved.

Ever since I met a classmate 12 years ago at a college reunion who was a relativist working with Hawking, I decided to try to learn relativity so I’d have something intelligent to say to him if we ever met again (COVID19 stopped all that although we’re still both alive).

Now that I understand what the math of relativity is trying to describe, I may be able to understand general relativity.

Be prepared for a lot of work, but do start with Needham’s book.  Here are some links to other things I’ve written about it.  It will take some getting used to as it is very different from any math book you’ve ever read (except Needham’s other book).

12 July 21 —

4 Dec 21 —

7 Mar 22 —

27 June 22 —

17 July 22 —

4 diseases explained at one blow said the protein chemist — part 1

A brilliant paper [ Science vol. 377 eabn5582 pp. 1 –> 20 ’22 ] explains how changing a single amino acid (proline) to another  can cause 4 different diseases, depending on the particular protein it is found in (and which proline of many is changed).

There is so much in this paper that it will take several posts to go over it all.  The chemistry in the paper is particularly fine.  So it’s back to Biochemistry 101 and the alpha helix and the beta sheet.

Have a look at this

If you can tell me how to get a picture like this into a WordPress post please make a comment.

The important point is that hydrogen bonds between the amide hydrogen of one amino acid and the carbonyl group of another hold the alpha helix and the beta pleated sheet together.

Enter proline : p//  Proline when not embedded in a protein has a hydrogen on the nitrogen atom in the ring.  When proline is joined to another amino acid by a peptide bond in a protein, the hydrogen on the nitrogen is no longer present.  So the hydrogen bond helping to hold the two structures (alpha helix and beta sheet) is no longer present at proline, and alpha helices and beta sheets containing proline are not has stable.  Prolines after the fourth amino acid of the alpha helix (e. g. after the first turn of the helix) produce a kink.  The proline can’t adopt the alpha helical configuration of the backbone and it can’t hydrogen bond.

But it’s even worse than that (and this observation may even be original).  Instead of a hydrogen bonding to the free electrons of the oxygen in the carbonyl group you have the two electrons on the nitrogen jammed up against them.  This costs energy and further destabilizes both structures.

Being a 5 membered ring which contains the alpha carbon of the amino acid, proline in proteins isn’t as flexible as other amino acids.

This is why proline is considered to be a helix breaker, and is used all the time in alpha helices spanning cellular membranes to cause kinks, giving them more flexibility.

There is much more to come — liquid liquid phase separation, prion like domains, low complexity sequences, frontotemporal dementia with ALS, TDP43, amyloid, Charcot Marie Tooth disease and Alzheimer’s disease.

So, for the present stare at the link to the diagram above.

Why Cassava’s 1 year results should allow compassionate use of Simufilam

Cassava reported results on 100 Alzheimer patients in an open label (e.g. no controls) trial of Simufilam for 1 year —  The average results were unimpressive (to the uninitiated) with only a minimal average overall improvement of an ADAS-Cog11 score of 1.5 points.  This is probably why the stock (SAVA) dropped a point yesterday after the news.  Since everything turns on ADAS-Cog11 here is a link to a complete description —  The test takes about 45 minutes placing it out of reach of a busy practicing clinical neurologist.

Why is even the 1.5 point improvement impressive to the initiated (me)?  Over 32 years in clinical neurology, I’d estimate that I saw at least 1 demented patient each week.  Now probably only 300 or so of the 1,664 were followed for a year.  Guess what?  None of them remained stable for a year, and all got worse.  Absolutely none of them  ever got better after a year.  So at least some stabilization of the disease is possible for a year.  The statistics say that Alzheimer patients lose 5 points a year on ADAS-Cog.

But that’s pretty small beer.  Who wants to keep a demented patient around but stable.  Here is the remarkable part of the Cassava results at a year.

63% of the 100 Patients Showed an Improvement in ADAS-Cog11 Scores, and This Group of Patients Improved an Average of 5.6 Points (S.D. ± 3.8). The statistics say that Alzheimer patients lose 5 points a year on ADAS-Cog.

This is unprecedented and is a strong argument for quick approval of Simufilam (or at least compassionate use).

The cynic will say that I’m just looking at the happy part of the Bell curve.  There must have been people who declined to average the improvement in the 63% down to a measly 1.5 points on the ADAS-Cog.

This is where clinical experience comes in.  No drug helps everyone with a given disease.  “Only 20% of cancer patients respond long term to a type of immune checkpoint blockade (of PD-1)” Science vol. 363p. 1377 ’19.  Nonetheless immune checkpoint blockade of several types was approved by the FDA, simply because there was nothing better available.

So if nearly 2/3 of Alzheimer patients will improve at one year on Simufilam, why not  let the FDA offer it to them now under compassionate use.



When to get your booster — a possibly useful post

I thought everyone knew that vaccines and boosters against the pandemic flu reach maximum protection against infection in a week or two, and then start declining.  Each of the zillions of papers on the subject say exactly this.

So if you’ve been isolating yourselves, as my wife and I have because of our age, the time to get that booster is a week or so before a time of likely exposure — in our case a visiting grandchild.

I thought everyone knew this, but maybe not.  My wife got an eMail today from her  85 year old cello teacher.   She summers on an island off the Maine coast near Mt. Desert Island (the home of Acadia National Park)


“We’re so far well, thanks to xxxx, I took his advice and got 2nd booster just before leaving middle of June. The whole island was shut down, the  only grocery store had signs SHUT TODAY, NO STAFF, went to Kneisel Hall in Blue Hill, people had not been able to rehearse together because they were actively sick, and so on. He and news are proving to be a blessing!

For this, we all thank you.”

Amyloid Structure at Last ! 3 The Alzheimer mutations

I am republishing this post from last October, because the excellent paper I’m going to write about has similar thinking.

Although the chemistry explaining why these mutations are associated with Alzheimer’s disease is exquisite and why they point to ‘the’ cause of Alzheimer’s disease — the amyloid fibril, billions have been spent in attempts to remove amyloid fibrils with no useful therapeutic result (and some harm)

Here’s the old post

The structure of the amyloid fibril formed by the aBeta42 peptide exactly shows why certain mutations are associated with hereditary Alzheimer’s disease.   Here is a picture

Scroll down to the picture above “Bonds that Tie”

If you need some refreshing on the general structure of amyloid, have a look at the first post in the series —

Recall that in amyloid fibrils the peptide backbone is flat as a flounder (well in a box 4.8 Angstroms high) with the amino acid side chains confined to this plane.  The backbone winds around in this plane like a snake.  The area in the leftmost loop is particularly crowded with bulky side chains of glutamic acid (single letter E) at position 22 and aspartic acid (single letter D) at position 23 crowding each other.  If that wasn’t enough, at the physiologic pH of 7 both acids are ionized, hence negatively charged.  Putting two negative charges next to each other costs energy and makes the sheet making up the fibril less stable.

The marvelous paper (the source for much of this) Cell vol. 184 pp. 4857 – 4873 ’21 notes that there are 3 types of amyloid — pathological, artificial, and functional, and that the pathological amyloids are the most stable. The most stable amyloids are the pathological ones.  Why this should be so will be the subject of a future post, but accept it as fact for now

In 2007 there were 7 mutations associated with familial Alzheimer’s disease (10 years later there were 11). Here are 5 of them.

Glutamic Acid at 22 to Glycine (Arctic)

Glutamic Acid at 22 to Glutamine (Dutch)

Glutamic Acid at 22 to Lysine (Italian)

Aspartic Acid at 23 to Asparagine (Iowa)

Alanine at 21 to Glycine (Flemish)

All of them lower the energy of the amyloid fiber.

Here’s why

Glutamic Acid at 22 to Glycine (Arctic) — glycine is the smallest amino acid (side chain hydrogen) so this relieves crowding.  It also removes a negatively charged amino acid next to the aspartic acid.  Both lower the energy

Glutamic Acid at 22 to Glutamine (Dutch) — really no change in crowding, but it removes a negative charge next to the negatively charged Aspartic acid

Glutamic Acid at 22 to Lysine (Italian)– no change in crowding, but the lysine is positively charged at physiologic pH, so we have a positive charge next to the negatively charged Aspartic acid, lowering the energy

Aspartic Acid at 23 to Asparagine (Iowa) –really no change in crowding, but it removes a negative charge next to the negatively charged Glutamic acid next door

Alanine at 21 to Glycine (Flemish) — no change in charge, but a reduction in crowding as alanine has a methyl group and glycine a hydrogen.

As a chemist, I find this immensely satisfying.  The structure explains why the mutations in the 42 amino acid aBeta peptide are where they are, and the chemistry explains why the mutations are what they are.

, , , , , , , . No

What you breathe can get into your brain

It used to be ‘you are what you eat’, now it’s you are what you breathe according to a paper from the world capital of air pollution — China — Proc. Natl. Acad. Sci. vol. 118 e2117083119 ’22.par

They found small particles in the cerebrospinal fluid (CSF) of 8/25 people.  What is small?  Well, a hydrogen atom is 1 Angstrom in diameter, E. Coli is a cylinder 10,000 Angstroms in diameter and 20,000 Angstroms long and a red blood cell is 80,000 Angstroms. One particle they show is 500 Angstroms in diameter, a group of 3 was 2,500 Angstroms in diameter.

The paper reviews all sorts of theories of how particles might get into the brain.  One is through the nose, since nerve fibers go outside the skull to the nasal mucosa.  Others include the gut (your are what you eat after all).

So they experimented on mice having them breathe air containing particles 100 – 300 Angstroms made of titanium dioxide (TiO2) which are found in the air we breathe.  To make sure the nose wasn’t involved, they administered the particles into the trachea.  They were able to show the particles got  into the blood and then into the brain across the blood brain barrier, damaging it in the process.  Convincing electron micrographs of the brain are shown.    Whether such particles are harmful is another matter.  Just because they’re there doesn’t mean they’re causing trouble.  We have nearly 10 times as many bacteria on (and in) us as we have cells and we’re still here —

Finding the particles in human CSF is convincing evidence that they’re in the brain, but (at present) we have no electron micrographs of human brains showing that the particles are present in brain tissue itself.


The silence is deafening

3 weeks ago I published a post about a paper that I thought would be a real bombshell, in effect contradicting a paper in a prestigious journal, and strongly arguing from real data that the pandemic virus could have been made in a lab, quite possibly Wuhan.  .

Absolutely nothing has happened. No letters to PNAS (the source of the article) to Cell (the source of the criticized study).  With a question of this magnitude and importance  you’d think Nature or Science would weigh in about it.  The origin of the pandemic virus is certainly they’ve covered extensively.

So I’m going to send this to all concerned and see if I get any feedback.

Here is the original post.

Evidence that the pandemic virus was made in a lab


Everyone knows that the Chinese have been less than forthcoming about the origin of the pandemic virus (SARS-CoV-2).  An article in the current Proceedings of the National Academy of Sciences — arguesthat US data, which hasn’t been released, and some 290 pages of which has been redacted could shed a good deal of light on the subject (without any help from China).  One of the authors is an economist, but the other has serious biochemical chops —

Basically a variety of US institutions (see the paper — it’s freely available) have been working with the lab at Wuhan for years modifying the virus, long before the pandemic.  The paper names the names etc. etc. and is quite detailed, but I want to explain the evidence that the virus could have been produced (by human modification) at the Wuhan lab.  It has to do with a site in a viral protein which says ‘cut here’.

Here is more background than many readers will need, but the virus has affected us all and I want to make it accessible to as many as possible.

Proteins are linear strings of amino acids, just as this post is a linear sequence of letters, spaces and punctuation.

We have fewer amino acids (20 to be exact) than letters  and to save space each one has a one letter abbreviation (A for alanine V for valine, etc. etc.).  The spike protein (the SARS-CoV-2 protein binding to the receptor  for it on our cells) is quite long (1,273 amino acids all in a row).

Our genome codes for 588  proteins (called proteases) whose job it is to cut up other proteins. Obviously, it would be a disaster if they worked indiscriminately.  So each cuts at a particular sequence of amino acids. Think of the protease as a key and the sequence as a lock.  One protease called furin cuts in the middle of an 8 amino acid sequence RRAR’SVAS (R stands for aRginine and S for Serine).  This is called the furin cleavage site (FCS)

A paper (The origins of SARS-CoV-2: A critical review. Cell 184, 4848–4856 (2021) argued that the amino acid sequence of the FCS in SARS-CoV-2 is an unusual, nonstandard sequence for an FCS and that nobody in a laboratory would design such a novel FCS.  So, like many, I skimmed the paper and accepted its conclusions, as Cell is one of the premier molecular biology journals.

One final quote “The NIH has resisted the release of important evidence, such as the grant proposals and project reports of EHA, and has continued to redact materials released under FOIA, including a remarkable 290-page redaction in a recent FOIA release.”

Sounds like Watergate doesn’t it?


Watch this space

BMOR is a bad actor

RNA and proteins have long been known to interact, but classic molecular biology pretty much had proteins down as something that modified RNA function.   Not so for BMOR, a long nonCoding RNA (1,247 nucleotides) expressed in breast cancer cells metastatic to the brain.  BMOR binds to IRF3 (Interferon Regulatory factor 3) inhibiting its phosphorylation by TBK1 with subsequent movement to the nucleus where it stimulates interferon expression which then turns on hundreds of genes producing inflammation.  All this is described in Proc. Natl. Acad. Sci. vol. 119 e2200230119 ’22 —

May 26, 2022
119 (22) e2200230119
Not sure if it is behind a paywall.    Definitely worth a read because knocking down BMOR in breast cancer cells prevents them from spreading to the brain (probably  by using BMOR to turni off the brain’s immune response to them).  Even more interestingly, BMOR was found to be only substantially expressed in breast cancer metastasis to brain tissue versus breast cancer metastasis to nonbrain tissues.



Teleology as always raises its head.  What in the world is the normal function of BMOR?  It can’t be what it is doing in the animal model described here.  Why would a cell make something to help it kill the organism containing it?


Then of course, as is typical of all interesting research, larger questions are raised.  Are there other RNAs whose function is to modify protein function?  Remember that 75% of the genome is transcribed into RNA.  Most of this has been thought of as molecular chaff, like the turnings of a lathe.   Time pick up the chaff from the factory floor and give it a look.

If the right hand don’t get you, the left hand will

Do you know the source of the title?  I found it surprising.  Answer at the end.

Some cancer cells have elevated levels of an enzyme called PHosphoGlyceride DeHydrogenase (PHGDH, others have decreased levels.  Many cancers contain both types of cells.  Neither is good news.

Those cancers  with low levels of PHGDH  have slower growth.  That’s good news isn’t it?  No.  Such cells are more likely to metastasize.

Those with high levels of PHGDH are less likely to metastasize.  That’s good news isn’t it?  No. such cells grow faster.

So cancers with both types of cells are more aggressive.

Here’s how it works [ Nature vol. 605 pp. 617 – 617, 747 – 753 ’22 ].

PHGDH is on the pathway for synthesis of serine, an amino acid required for protein synthesis (like all of them).  So low levels of the enzyme result in less protein synthesis and less tumor growth.

So how is this bad?  PHGDH binds to another enzyme PFK (PhosphoFructoKinase) stabilizing it.  When PHGDH is low PFK enzyme levels are low, so the subsrate of PFK (fructose 6 phosphate) is diverted to making sialic acid, which modifies cell surface proteins making them more likely to migrate.

So blocking sialic acid synthesis reverses the effects of low PHGDH on cancer migration and metastasis — but it does potentiate cell proliferation.

You just can’t win

Things like this may explain other paradoxic and unexpected effects of enzyme blockade.

16 Tons by Tennessee Ernie Ford