The viruses in our brains

PNMA2 (ParaNeoplastic antigen MA2) is a protein initially found as the target of the immune response (autoantibodies) producing a nasty dementing neurologic disease (Paraneoplastic encephalitis).  The PNMA2 protein is exclusively expressed in neurons which implies that neurons are using it for something.   This is teleological thinking, usually looked down on, but always needed in molecular biology and cellular physiology.

What PNMA2 does is amazing.  It forms icosahedral viral capsids which are released from cells (in culture) as nonEnveloped capsids.  It isn’t clear if this normally happens in our  brains.    Probably it doesn’t, and when the capsid somehow gets out of the producing cell or neuron immunological hell breaks loose and autoimmune encephalitis is the result.

PNMA2 is derived from one of the long terminal repeat retrotransposons (LTR retrotransposons), viral remnants that make up 8% of the human genome (https://en.wikipedia.org/wiki/LTR_retrotransposon). This explains why it makes particles that look like viruses.  Such particles can contain RNA, so big pharma is interested in them as a way of delivering mRNA drugs.

Totally off topic but yesterday I read a paper about E. Coli DNA gyrase, an amazing enzyme which untangles DNA ( Science vol. 384 pp. 227 – 232 ’24 ).

Here is what it does.   If you’ve got some venetian blinds in your home twist it 20 or so times (keeping the ends fixed, and you have the DNA double helix, with two strands winding around each other.  Now to read or copy a single strand, you must grab both strands where you want this to happen  and pull them apart keeping the ends of the venetian blind fixed.  This immediately increases the coiling elsewhere. Since there are only 10 nucleotides/turn of the double helix, copying a gene for a 100 amino acid protein means you are removing 33 twists from the separated strands (and producing new ones elsewhere).   The cords of the venetian blind quickly become a tangled mess when this happens.  This is where DNA gyrase comes in.  It cuts both strands of the DNA double helix, holding on to the cut ends, and slides an intact double helix of the twisted DNA through the cut.   Sounds fantastic doesn’t it?  Hard to see how evolution could come up with something like this but it did.

The paper contains the following passage toward the end

“A second model based on a sign-inversion reaction wassuggested to describe introduction of (–)SC by this enzyme (28). This model proposed that the enzyme binds to a positive crossover followedby a DNA strand passage through a DNA double-strand break that results in a sign inversion.”

(28) is 28. P. O. Brown, N. R. Cozzarelli,Science206, 1081–1083 (1979).

The paper is 45 years old and has now been shown to be correct.  N. R.  Cozzarelli is my late good friend and Princeton classmate Nick, and it is very nice to see him honored here.

A few words about Nick.  Although Princeton was full of rich kids, they still had the brains to take in someone like Nick whose father was an immigrant shoemaker in Jersey City.  Nick worked his way through Princeton waiting on tables in commons (where all Freshmen ate).  I can still see the time that some rich preppie jerk gave him a hard time about the service.

Nick got his PhD at Harvard and later became a professor at Berkeley where he did his great work.  Nick later edited the Proceedings of the National Academy of Sciences (USA) for 10 years before his very untimely death over 20 years ago from Burkitt’s lymphoma.  R. I. P. Nick.

Axiomatize This !

“Analyze This”, is a very funny 1999 sendup of the Mafia and psychiatry with Robert DeNiro and Billy Crystal.  For some reason the diagram on p. 7 of Barrett O’Neill’s book “Elementary Differential Geometry” revised 2nd edition 2006 made me think of it.

O’Neill’s  book was highly recommended by the wonderful “Visual Differential Geometry and Forms” by Tristan Needham — as “the single most clear-eyed, elegant and (ironically) modern treatment of the subject available — present company excpted !”

So O’Neill starts by defining a point  as an ordered triple of real numbers.  Then he defines R^3 as a set of such points along with the ability to add them and multiply them by another real number.

O’Neill then defines tangent vector (written v_p) as two points (p and v) in R^3 where p is the point of application (aka the tail of the tangent vector) and v as its vector part (the tip of the tangent vector).

All terribly abstract but at least clear and unambiguous until he says — “We shall always picture v_p as the arrow from point p t0 the point p + v”.

The picture is a huge leap and impossible to axiomatize (e.g. “Axiomatize This”).   Actually the (mental) picture came first and gave rise to all these definitions and axioms.

The picture is figure 1.1 on p. 7 — it’s a stick figure of a box shaped like an orange crate sitting in a drawing of R^3 with 3 orthogonal axes (none of which is or can be axiomatized).  p sits at one vertex of the box, and p + v at another.  An arrow is drawn from p to p + v (with a barb at p + v) which is then labeled v_p.  Notice also, that point v appears nowhere in the diagram.

What the definitions and axioms are trying to capture is our intuition of what a (tangent) vector really is.

So on p. 7 what are we actually doing?  We’re looking at a plane in visual R^3 with a bunch of ‘straight’ lines on it.  Photons from that plane go to our (nearly) spherical eye which clearly is no longer a plane.  My late good friend Peter Dodwell, psychology professor at Queen’s University in Ontario, told me that the retinal image actually preserves angles of the image (e.g. it’s conformal). 1,000,000 nerve fibers from each eye go back to our brain (don’t try to axiomatize them).   The information each fiber carries is far more processed than that of a single pixel (retinal photoreceptor) but that’s another story, and perhaps one that could be axiomatized with a lot of work.

100 years ago Wilder Penfield noted that blood flowing through a part of the brain which was active looked red rather than blue (because it contained more oxygen).  That’s the way the brain appears to work.  Any part of the brain doing something gets more blood flow than it needs, so it can’t possibly suck out all the oxygen the blood carries.  Decades of work and zillions researchers have studied the mechanisms by which this happens.  We know a lot more, but still not enough.

Today we don’t have to open the skull as Penfield did, but just do a special type of Magnetic Resonance Imaging (MRI) called functional MRI (fMRI) to watch changes in vessel oxygenation (or lack of it) as conscious people perform various tasks.

When we look at that simple stick figure on p. 7, roughly half of our brain lights up on fMRI, to give us the perception that that stick figure really is something in 3 dimensional space (even though it isn’t).  Axiomatizing that would require us to know what consciousness is (which we don’t) and trace it down to the activity of billions of neurons and trillions of synapses between them.

So what O’Neill is trying to do, is tie down the magnificent Gulliver which is our perception of space with Lilliputian strands of logic.

You’ve got to admire mathematicians for trying.

Biden looked good tonight

As a retired clinical neurologist, I’m often asked about President Biden’s mental competence.  I thought he did well in the debates with Trump 11/20 — https://luysii.wordpress.com/2020/09/29/first-debate-what-did-the-neurologist-think/

Tonight President Biden gave a long (1 hour and 7 minutes) State of the Union speech.  He showed no sign of dementia.  While he garbled a few words here and there, he was coherent throughout  — although he had teleprompters on either side of him.  In particular he remained vigorous throughout and not old and feeble.

I have been concerned that he might have occult hydrocephalus due to his ruptured intracranial aneurysm — see https://luysii.wordpress.com/wp-admin/post.php?post=6414&action=edit

One of the earliest signs of this disorder is a gait disturbance and I was watching for it.  Unfortunately, given the crush of people he waded through entering and leaving, he was unable to take more than a single step at a time, so no statement can be made about his gait.

In the early stages of dementia, people have good and bad days, but his performance tonight was reassuring that he can function as the President.

What if our most common assumption about Alzheimer’s disease is wrong?

Although the “Abeta protein aggregates cause Alzheimer’s disease” has had quite a run, it is not our most common assumption about Alzheimer’s disease.

 

Any guesses?

The assumption is hidden in the deep in the semantics of Alzheimer’s disease.  By simply naming it we are tacitly assuming that Alzheimer’s disease is just one thing.   The history of medicine is the history of splitting diagnostic categories with the passage of time due the accumulation of  causal knowledge.

Infections were characterized by the type of fever they produced before Pasteur.  Breast cancer was characterized by pathology before it was molecularly split depending which hormone receptors were present.  No one would dream of treating it all the same way today.

Yet here we have massive clinical trials of single therapies in Alzheimer’s disease because we don’t know any better.

Because people vary, in all clinical trials the responses to a given drug vary patient to patient. It is worth studying those responding best and those responding worst to a therapy in terms of the data taken on entry (MMSE, age, sex, education, pre-existing disease, smoking drinking, etc. etc.).  Such analysis might tell us something about the underlying causes in addition to predicting who will and who won’t respond in the future.

In particular, Cassava Sciences’ recent release of two years of open label Simufilam administration should be studied this way.  The second year is problematic as some dropped out, some continued to receive the Simufilam, some did not, but all patients still in the study at one year had been on the drug for a year, and their clinical data (ADAS-Cog etc. etc.) is in Cassava’s possession.

How did the group responding best differ from those responding the least.  Lindsay is too busy dealing with the slings and arrows of outrageous fortune (courtesy of Science, the Wall Street Journal etc. etc.) to climb an academic totem pole to write it up.

I am particularly interested in the 5/50 patients Lindsay reported 8/21 who likely showed a 50% improvement at 9 months. As a clinical neurologist with decades of experience with demented patients, I never saw this degree of improvement at 9 months.  How did the 5 do at one year and, if continuing on Simufilam, how did they do at two years?   What was different about the 5 as a group vs. the 45 that didn’t do as well?

To have a look at the actual data Lindsay presented back then follow this link — https://luysii.wordpress.com/2021/08/25/cassava-sciences-9-month-data-is-probably-better-than-they-realize/

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.

Why me, O Lord, Why me?

It was very hard for my multiple sclerosis (MS) patients to understand why they were singled out for MS, given the publicity given to theories of viral causation, popular at least since I started getting seriously interested in neurology as a 3rd year medical student in 1964.  Herpes simplex (fever blisters) was a popular culprit, but we all know lots of people who’ve had them without coming down with MS.

The best explanation I could give them was of my med school classmate Marty, a Jewish kid from Pittsburgh.  Graduating in 1966 at the height of American involvement in Vietnam, all my classmates entered the service within a few years.  Marty was sent to Vietnam  as a GMO (General Medical Officer).  He was quickly sent back stateside as he developed a severe anemia.  Why?

Well malaria was endemic in Vietnam, and anyone going over received an antiMalarial as prophylaxis.  The malarial parasite does its damage by infecting red blood cells.  The antiMalarial drug he received inhibited a red cell enzyme Glucose 6 Phosphate Dehydrogenase (G6PD), essentially starving the parasites.  Marty had a partial deficiency of this enzyme.  Such deficiencies are relatively common in areas endemic for Malaria, as it is protective, just as the sickle cell trait is protective against Malaria in Africa.  A variety of other mutations in different red cell proteins arising in endemic areas are also protective (example Thalassemia in Greece, etc. etc.)

So if Marty had never been sent to Vietnam he would never have become anemic.  I’d tell my patients that they had some biochemical difference (totally unknown back then) that made them susceptible to complications of infection with a common organism.  Not very satisfying, but it was the best I could do.

In the case of another virus Epstein Barr Virus (EBV) which causes infectious mononucleosis, this explanation (50+ years later) turned out to be exactly correct.    Not only that it shows the extreme subtlety of what ‘causation’ in medicine actually means.

Not only must the unlucky people getting MS after EBV infection be different biochemically, they must be infected with a particular variant of EBV (not all EBV is the same, just as not all people or SARS-CoV-2 are the same).

That’s the view from 30,000 feet.  You can stop here but the full explanation is unsparingly technical.  It is to be found in Cell vol. 186 pp. 5675 – 5676, 5708 – 5718 ’23 ]

Here goes.

Long term control of Epstein Barr Virus is mediated by cytotoxic T lymphocytes which recognize parts of EBV proteins.  One such protein is EBNA1, and antibodies to amino acids #386 – #405 of EBNA1 cross react with amino acids #370 – #389 of a human protein called GlialCAM (for Glial Cell Adhesion Molecule) which is important in maintaining the myelin sheath around axons in the brain (MS is basically a destructive immune attack on myelin).

Such an antibody is called autoreactive, in the sense that it is reacting to a normal human protein.  Cells producing autoreactive antibodies (autoreactive cells) are normally eliminated by cytotoxic natural killer cells.  In the case of EBV specific T cells they are eliminated by natural killer cells expressing proteins NKG2C and NKG2D.  They target the autoreactive GlialCAM specific autoreactive B cells.  Some people have deletion of the gene coding for NKG2C rendering them more susceptible to MS after EBV infection.

But wait, there’s more. There are many EBV variants and some of them upregulate another human protein HLA-E by containing another protein (LMP1) which stabilizes HLA-E.  HLA-E blunts the natural killer cell attack on autoreactive GlialCAM cells.

So it’s a delicate dance of unfortunate events ‘causing’ MS.  A mild genetic defect in the human, and a genetic variant in the virus, both of which must occur for causation.

Who knew that medical causation could be so subtle.

At last a Science article about Alzheimer Rx that doesn’t trash Cassava Sciences

An article “Immunotherapies for Alzheimer’s Disease”  (Science vol. 382 pp. 1242 – 1244 ’23) avoids hype about these therapies and doesn’t trash Cassava Sciences or Simufilam (by ignoring it).   They note that “aducanumab, lecanemab, and donanemab are far from curative, but they all slow cognitive decline by ~25 to 30% over 18 months.   Continuing to avoid hype they say “reduction of brain Aβ in early symptomatic AD has a positive, but modest, clinical effect.”

They don’t ignore side effects. “ARIA-E (edema in the brain parenchyma or sulcal effusion—i.e., extravasated fluid in the leptomeninges along the sulcal spaces) and ARIA-H (hemosiderin deposits resulting from red blood cell breakdown products—i.e., microhemorrhages) (9). Although the incidence of ARIAs is increased by antibody treatment and is observed in roughly one-third of treated participants, in most individuals, it is asymptomatic.

This is similar to the rationale put out for the ‘asymptomatic hemorrhages’ seen after intravenous tissue plasminogen activator (TPA) for stroke.

Ask yourself, would you want any of them (ARIA-E, ARIA-H, asymptomatic hemorrhage) ?  I wouldn’t and the long term effects of such things are unknown.

They also deal with the cost of immunotherapy. “In the US, aducanumab, lecanemab, and donanemab each have an annual cost of $26,500, but associated imaging and monitoring means that costs for a year of treatment may exceed $75,000. ”

All in all, a very fair minded article.

But what really caught my eye was the following statement. “In cancer, many drugs are approved with modest effects on overall survival, but over time, data emerge showing exceptional responses in select individuals who are cured or are in very-long-term remission.”

This is exactly what I saw in my examination of Cassava’s data on the first 50 patients to complete 9 months of treatment on Simufilam.  Basically 5/50 had a greater than 50% improvement in their ADAS-Cog11.   This is why I’m so excited about the drug.  No antimicrobial cures all infections (because they are different).  Similarly, there is no compelling evidence that all Alzheimer’s disease is the same.

Now I am a retired clinical neurologist after 33 years of training and practice.  You never see results like this in Alzheimers disease.  I likely saw 1 demented patient a week during this time.  Although the 9 month results were not double blinded and open label, long clinical experience tells me that these results are spectacular and unprecedented (even if they only occur in 10% of patients getting Simufilam).

Here is a link to a post analyzing these results in far greater detail https://luysii.wordpress.com/2021/08/25/cassava-sciences-9-month-data-is-probably-better-than-they-realize/

What you can measure isn’t always what’s important

Back in the bad old days some residents would make sure that dying patients were in electrolyte (sodium, potassium, chloride CO2) balance, even though they were irrelevant to why the patient was dying (cancer, stroke, cardiac failure, etc. etc.).  Severe electrolyte imbalance can kill.  It was basically CYA.  While ‘lytes could easily be measured they were not what was important.

Similarly, the economy is great — inflation is no longer so bad and is decreasing, unemployment is low and the gross national product is increasing.  Tell that (Bidenomics) to the 60%+ of Americans living paycheck to paycheck according to three different polls conducted this year.  Inflation rate, unemployment rate, GNP are just irrelevant numbers to them.  They see diminished purchasing power every time they buy groceries (not included in inflation), buy gas, or try to eat out.

Similarly the controversy over protein electrophoretic patterns of Simufilam and whether they have been fudged is irrelevant to the far more important question of whether it helps people with Alzheimer’s disease think.  Here the important number is their scores on cognitive tests and their functioning on the activities of daily living.  I think it does.  For an elaboration please see — https://luysii.wordpress.com/2021/08/25/cassava-sciences-9-month-data-is-probably-better-than-they-realize/

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.

Ketamine again

Ketamine burst like a bombshell on depression research a few years ago.  I’ve written two other posts about it (reproduced after the ***).

The drugs we used for depression weren’t great.  They didn’t help at least a third of the patients, and they usually took several weeks to work for endogenous depression (e.g. depression not obviously triggered by external events).  They seemed to work faster in my MS patients who had a relapse and were quite naturally depressed by an exogenous event completely out of their control.

Because of the weeks of delay an incredible amount of work was done looking for the long term neurochemical and neurophysiologic changes produced by the antidepressants we had (tricyclic antidepressants, selective serotonin reuptake inhibitors — SSRIs)

Enter Ketamine which, when given IV, can transiently lift depression within a few hours.  You can find more details and references in an article in  Neuron ( vol. 101 pp. 774 – 778 ’19)  written by the guys at Yale who did some of the original work. However, here’s the gist of the article.

A single dose of ketamine produced antidepressant effects that began within hours peaked in 24 – 72 hours and dissipated within 2 weeks (if ketamine wasn’t repeated).  This occurred in 50 – 75% of people with treatment resistant depression.  Remarkably one third of treated patients went into remission.

The incredibly rapid improvement in depression (hours) produced by ketamine is unprecedented and surely is telling us something vitally important about depression.  If only we could figure out what it is.

I think that one thing ketamine is telling us, is that depression is in some way an active process, which must be maintained somehow, and that ketamine is breaking up this process.

One problem with trying to figure out what ketamine is doing is that it is gone long before its therapeutic effects end.  For instance the elimination half life in man is 3 hours, but the antidepressant activity lasts 3 to 14 days.  How can it break up an active process if it’s not around.

So we turn to our friend the mouse.  They don’t talk, so how can you tell if they’re depressed.  We do have reasonable animal models of depression (tail suspension test, forced swim test).  We can at least get a handle on the anhedonia almost invariably found in depression using the sucrose preference test.

Throw ketamine at an animal and measure the biochemical or the neurophysiologic effect of your choice. There are zillions of them.  Throw just about anything at the brain, and all sorts of things change.  The problem is showing that the change is relevant.  Is the known blockade of NMDA receptors by ketamine how it helps depression.  Give enough ketamine to humans and you get out of body experiences and all sorts of craziness, not an antidepressant effect.

Enter Nature vol. 622 pp. 802 – 809 ’23.  Ketamine hangs around 13 minutes in mice, but its antidepressant effects (see above) last at least a day.    Here the antidepressant effect is measured by suppresion of burst firing in the lateral habenula (which is a tiny structure in man) something a very long way from clinical depression in humans.

So ketamine can’t interrupt an active process if it isn’t there.  But it is there according to the paper, which finds ketamine hanging around the NMDA receptor (actually a subtype of glutamic acid receptor) in the lateral habenula for a day, even though you can’t measure it anywhere else.

That’s interesting, but the paper notes that “Currently, there is an intense debate about whether the antidepressant effects of ketamine are mediated by NMDA receptors at all.”

Here are my two other posts on the subject

***

Published almost exactly 4 years ago

How does ketamine lift depression?

The incredibly rapid improvement in depression (hours) produced by ketamine is unprecedented and surely is telling us something vitally important about depression.  If only we could figure out what it is.  Clinicians were used to waiting weeks for antidepressants of all sorts to work.  As a neurologist, I’d see it work in a week or so in my MS patients depressed due a relapse.

Two recent papers show just how hard it is going to be [Neuron  vol. 104 pp. 182 – 182, 338 – 352 ’19 ]. First off you have to accept the idea that even though animals (usually mice) can’t tell us how they feel, we still have reasonable animal models of depression (tail suspension test, forced swim test).  We can at least get a handle on anhedonia using the sucrose preference test.

Throw ketamine at an animal and measure the biochemical or the neurophysiologic effect of your choice. There are zillions of them.  Throw just about anything at the brain, and all sorts of things change.  The problem is showing that the change is relevant.  Is the known blockade of NMDA receptors by ketamine how it helps depression.  Give enough and you get out of body experiences and all sorts of craziness, not an antidepressant effect.

Homer1a is a protein found at the synapse, and like all scaffold proteins, it interacts with a bunch of different proteins. It links another type of glutamic acid receptor (mGluR1 and mGluR5) to inositol 1, 4, 5 trisphosphate receptors (IP3Rs) on the endoplasmic reticulum.  It also links mGluR1 and mGluR5 to NMDARs and other ion channels.

So what?

Other work by the authors showed that knockdown of Homer1a (using small interfering RNA – siRNA) in the medial prefrontal cortex (mPFC) abolished the antidepressant effects (in animal models) to ketamine.  Well that’s good, but even better is that knockdown also abolished the antidepressant effects of a tricyclic antidepressant (imipramine).

The present work showed that increasing the expression of Homer1a (the protein comes in various isoforms) in the frontal cortex reduced depression in the various models.

Pretty good — all we have to do is increase Homer1a expression to have a treatment of depression.

Don’t get your hopes up, and this is why depression research is so — well depressing.

Increasing Homer1a expression in another brain region (the hippocampus) has exactly the opposite effects.

The four hour cure for depression: what is KetaminTe doing?

The drugs we use for depression aren’t great.  They don’t help at least a third of the patients, and they usually take several weeks to work for endogenous depression.  They seemed to work faster in my MS patients who had a relapse and were quite naturally depressed by an exogenous event completely out of their control.

Enter Ketamine which, when given IV, can transiently lift depression within a few hours.  You can find more details and references in an article in  Neuron ( vol. 101 pp. 774 – 778 ’19)  written by the guys at Yale who did some of the original work. However, here’s the gist of the article.  A single dose of ketamine produced antidepressant effects that began within hours peaked in 24 – 72 hours and dissipated within 2 weeks (if ketamine wasn’t repeated).  This occurred in 50 – 75% of people with treatment resistant depression.  Remarkably one third of treated patients went into remission.

This simply has to be telling us something very important about the neurochemistry of depression.

Naturally there has been a lot of work on the neurochemical changes produced by ketamine, none of which I’ve found convincing ( see https://luysii.wordpress.com/2019/10/27/how-does-ketamine-lift-depression/ ) untilthe following paper [ Neuron  vol. 106 pp. 715 – 726 ’20 ].

In what follows you have to have some basic knowledge of synaptic structure, but here’s a probably inadequate elevator pitch.  Synapses have two sides, pre- and post-.  On the presynaptic side neurotransmitters are enclosed in synaptic vesicles.  Their contents are released into the synaptic cleft when an action potential arrives from elsewhere in the neuron.  The neurotransmitters flow across the very narrow synapse to bind to receptors on the postsynaptic side, triggering (or not) a response of the postsynaptic neuron.  Presynaptic terminals vary in the number vesicles they contain.

Synapses are able to change their strength (how likely an action potential is to produce a postsynaptic response).  Otherwise our brains wouldn’t be able to change and learn anything.  This is called synaptic plasticity.

One way to change the strength of a synapse is to adjust the number of synaptic vesicles found on the presynaptic side.   Presynaptic neurons form synapses with many different neurons.  The average neuron in the cerebral cortex is post-synaptic to thousands of neurons.

We think that synaptic plasticity involves changes at particular synapses but not at all of them.

Not so with ketamine according to the paper.  It changes the number of presynaptic vesicles at all synapses of a given neuron by the same percentage — this is called synaptic scaling.  Given 3 synapses containing 60  50 and 40 vesicles, upward synaptic scaling by 20% would add 12 vesicles to the first 10 to the second and 8 to the third.   The paper states that this is exactly what ketamine does to neurons using glutamic acid (the major excitatory neurotransmitter found in brain).  Even more interesting, is the fact that lithium which treats mania has the opposite effects decreasing the number of vesicles in each synapse by the same percentage.

I found this rather depressing when I first read it, as I realized that there was no chemical process intrinsic to a neuron which could possibly work quickly enough to change all the synapses at once.  To do this you need a drug which goes everywhere at once.

But you don’t. There are certain brain nuclei which send their processes everywhere in the brain.  Not only that but their processes contain varicosities which release their neurotransmitter even where there is no post-synaptic apparatus.  One such nucleus (the pars compacta of the substantia nigra) has extensively ramified processes so much so that “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 ].  So when that single neuron fires, dopamine is likely to bathe every neuron in the brain.  We think that something similar occurs in the locus coeruleus of the lower brain which has only 15,000 neurons and releases norepinephrine, and also in the raphe nuclei of the brainstem which release serotonin.

It should be less than a surprise that drugs which alter neurotransmission by these neurotransmitters are used to treat various psychiatric diseases.  Some drugs of abuse alter them as well (Cocaine and speed release norepinephrine, LSD binds one of the serotonin receptors etc, etc.)

The substantia nigra contains only 450,000 neurons at birth, so you don’t need a big nucleus to affect our 80 billion neuron brains.

So the question before the house, is have we missed other nuclei in the brain which control volume neurotransmission by glutamic acid?   If they exist, could their malfunction be a cause of mania and/or depression?  There is plenty of room for 10,000 to 100,000 neurons to hide in an 80 billion neuron brain.

Time to think outside the box people. Here is an example:  Since ketamine blocks activation of one receptor for glutamic acid, could there be a system using volume neurotransmission which releases a receptor inhibitor?

Addendum 7 July — I sent a copy of the post to the authors and received this back from one of them. “Thank you very much for your kind words and interest in our work. Your explanation is quite accurate (my only suggestion would be to replace “vesicles” with “receptors”, as the changes we propose are postsynaptic). Reading your blog reassures us that our review article accomplished its main goal of reaching beyond the immediate neuroscience community to a wider audience like yourself.”