A certain Nobel prize

Chemists will be green with envy to find out that a Nobel prize (possibly in Chemistry) is almost certain to be won by someone using using nothing fancier than formaldehyde, acrylamide and an ionic detergent (Sodium Dodecyl Sulfate — the SDS of electrophoresis fans everywhere)_. For details see Nature vol. 497 pp. 332 – 337 ’13 (16 May Issue). It’s by the same man (Karl Deisseroth) who already has a Nobel coming for the invention of optogenetics.

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 is 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 the light microscopy.

Your brain (and mine) is mostly fat. Light doesn’t get through fat very well at all. Deisseroth figured out a way to remove the fat leaving the other brain structures intact. The technique even works on brains fixed in formaldehyde for years. First they infused formaldehyde and acrylamide into brain tissue at 4 degrees Centigrade. The formaldehyde hardens the tissue, but it also links the acrylamide to the proteins making up the tissue. Then they raised the temperature to 37 Centigrade causing the acrylamide to polymerize. Then they infused sodium dodecyl sulfate into the tissue using electrophoresis. When the SDS was pulled out of the tissue (again by electrophoresis) pulling the fat (lipids) of the brain with it, this left what they call a hydrogel (which light could go through).

Using the technique it is possible to look through slabs of brain tissue 500 microns thick (5 million Angstroms thick) with a light microscope and see cell bodies and nerve fibers in their natural habitat (e.g. whole populations of neurons along with their projections). Even better you can stain the hydrogel with your antibody of choice and see what protein is where. Then you can wash this out and look at something else.

It is an incredible advance and certain to revolutionize our understanding of the brain. Look at the paper. The pictures are amazing and more are sure to follow from other workers. Definitely Nobel caliber work.

It is extremely amusing to me that this work could have been done 50 years ago. It just took someone smarter than you and I to think of it.

The DSM again

The Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-V) is in the news. The press has not been favorable, nor have two new books concerning it. Here are some links

l. A review of a book on it from today’s Nature (2 May ’13)–http://www.nature.com/nature/journal/v497/n7447/full/497036a.html
2. An article in the New York Times today concerning the Nature book and one other — neither favorable –http://www.nytimes.com/2013/05/02/books/greenbergs-book-of-woe-and-francess-saving-normal.html?ref=todayspaper&_r=0

Added 8 May ’13 The US National Institute of Mental Health (NIMH) will no longer use the Diagnostic and Statistical Manual of Mental Disorders (DSM) to guide psychiatric research, NIMH director Thomas Insel announced on 30 April. The manual has long been used as a gold standard for defining mental disorders. Insel described the DSM as ill-suited to scientific studies, and said the NIMH will now support studies that cut across DSM-defined disease categories.

But, as Ernst Mayr once said — nothing in biology makes sense except in the light of evolution. Keeping that thought in mind, what I wrote a few years ago is relevant today. Here’s the post. Although it starts off in Mathematics, it gives some history which helps explain why the DSM is the way it is.

Even so, psychiatric wisdom should be taken with a good deal of salt. A psychiatrist in my medical school class (1966) knew people who were thrown out of their psychiatric residencies because they were gay, and back then homosexuality was a psychiatric disease.

Here’s the post of 3 years ago

Reification in mathematics and medicine

Can you bring an object into existence just by naming and describing it? Well, no one has created a unicorn yet, but mathematicians and docs do it all the time. Let’s start with mathematicians, most of whom are Platonists. They don’t think they’re inventing anything, they’re just describing an external reality that is ‘out there’ but isn’t physical. So is any language an external reality, but when the last person who knows that language dies, so does the language. It will never reappear as people invent new languages, and invent them they do as the experience with deaf Nicaraguan children has shown [ Science vol. 293 pp. 1758 - 1759 '01 ]. Mathematics has been developed independently multiple times all over the world, and it’s always the same. The subject matter is out there, and not just a social construct as some say.

A fascinating book, “Naming Infinity” describes a Russian school of mathematicians who extended set theory beyond the work of the French and Germans. They literally believed that describing a mathematical object and its properties implied that the object existed (assuming the properties were consistent). The mathematicians involved were also very devout mystical Christians, who were called “Name Worshippers”. They thought that repeatedly invoking the name of Jesus would allow them to reach an ecstatic state. The rather contentious theory of the book is that their religious stance allowed them to imbue all names with powerful properties which could bring what they named into existence and this led to their extensions of set theory. Naturally the Communists hated them, and exterminated many (see p. 126). People possessed of all absolute truths dislike those possessed of a different set.

Docs bring diseases into existence all the time simply by naming them. This is why the new DSM-V (Diagnostic and Statistical Manual of Mental Disorders) of the American Psychiatric Association (APA) is so important. Is homosexuality a disease? Years ago the APA thought it was. If your teenager won’t do what you want, is this “Adolescent Defiant Disorder”? Is it a disease? It will be if the DSM-V says it is.

There are a lot of things wrong with what the DSM has become (297 disorders in 886 pages in DSM-IV), but the original impetus for the major shift that occurred with DSM-III in the 70s was excellent. So it’s time for a bit of history. Prior to that time, it was quite possible for the same individual to go to 3 psychiatric teaching hospitals in New York City and get 3 different diagnoses. Why? Because diagnosis was based on the reconstruction of the psychodynamics of the case. Just as there is no single way to interpret “Stopping by Woods on a Snowy Evening” (see the previous post), there isn’t one for a case history. Freud’s case studies are great literature, but someone else would write up the case differently.

The authors of the DSM-III decided to be more like medical docs than shrinks. In our usual state of ignorance, we docs define diseases by how they act — the symptoms, the physical signs, the clinical course. So the DSM-III abandoned the literary approach of psychodynamics and started asking what psychiatric patients looked like — were they hallucinating, did they take no pleasure in things, was there sleep disturbance, were they delusional etc. etc. As you can imagine, there was a huge uproar from the psychoanalysts.

Now no individual fits any disease exactly. There are always parts missing, and there are always additional symptoms and signs present to confuse matters. The net result was that psychiatric diagnosis became like choosing from a menu in a Chinese restaurant, so many symptoms and findings from column A, so many from column B. (Update 2013 — Having been to China for 3 weeks this year, restaurant menus over there aren’t like that).

This led to a rather atheoretical approach, but psychiatric diagnoses became far more consistent. Docs have always been doing this sort of thing and still do (look at the multiple confusing initial manifestations of what turned out to AIDS back in the 80s). Different infections were classified by how they acted, long before Pasteur proved that they were caused by micro-organisms. Back when I was running a muscular dystrophy clinic, we saw something called limb girdle muscular dystrophy , in which the patients were weak primarily in muscles about the shoulders and hips. Now we know that there are at least 13 different genetic causes of the disorder. So there are many distinct causes of the same clinical picture. This is similar to the many different genetic causes of Parkinson’s disease I talked about 2 and 3 posts earlier. At least with limb girdle muscular dystrophy it is much easier to see how the genetic defects cause muscle weakness — all of the known genetic causes involve proteins found in muscle.

Where DSM-IV (and probably DSM-V — it’s coming out later this month) went off the rails, IMHO, is the multiplicity of diagnoses they have reified. Do you really think there are 297 psychiatric disorders? Not only that, many of them are treated the same way — with an SSRI (Selective Serotonin Reuptake Inhibitor). You don’t treat all infections with the same antibiotic. This makes me wonder just how ‘real’ these diagnoses are. However in defense of them, you do treat classic Parkinsonism pretty much the same way regardless of the genetic defect causing it (and at this point we know of genetic causes of less than 10% of cases).

There is a fascinating series of articles in Science starting 12 Feb ’10 about the new DSM-V. The first is on pp. 770 – 771. One of the most interesting points is that 40% of academic inpatients receive a diagnosis of NOS (Not Otherwise Specified — e.g. not in the DSM-IV — clearly even 297 diagnoses are missing quite a bit).

But insurance companies and the government treat this stuff as holy writ. Would you really like your frisky adolescent labeled with “prepsychotic risk syndrome” which is proposed for DSM-V. Also, casting doubt on the whole enterprise, are the radical changes the DSM has undergone since it’s inception nearly 60 years ago. We’ve learned a lot about all sorts of medical diseases since then, but strokes and heart attacks back then are still strokes and heart attacks today and TB is still TB. Do these guys really know what they’re talking about, and should we allow them to reify things?

That being said, cut psychiatry some slack. Regardless of theory, there are plenty of mentally ill people out there who need help. They aren’t going to go away (or get better) any time soon. Psychiatrists (like all docs) are doing the best they can with what they know.

That’s why it’s nice to be retired and reading stuff that it is at least possible to understand — like math, physics, organic chemistry and molecular biology. But never forget that it is trivial compared to human suffering. That’s why the carnage in the drug discovery industry is so sad — there goes our only hope making things better (written in 2010, but still true in 2013).

The weird way human memory works — Hopfield was right

Sometimes middle of the night thoughts are strange.  At 3AM today, I was trying to remember the name of the guy who wrote “Infinite Jest” and “The Broom of the System”.  The only thing that kept popping into my head was Richard Gordon Loomis, the name of my very excellent piano tuner, but a man with no known literary inclinations. I had no idea why this happened until I came up with the real name of the author this morning, 5 hours later — David Foster Wallace.

They don’t sound the same (except for the last syllable), but they have exactly the same rhythmic cadence of syllables when spoken.  Also he’s the only person I know who uses his middle name (the reason being that, amazingly enough, there is another piano tuner in the area named Richard Loomis).

This fits with John Hopfield’s theory of memory (he’s one of the inventors of the neural net) — having to do with chaos and attractors ( Neural networks and physical systems with emergent collective computational abilities. Proc. NatL Acad. Sci. USA Vol. 79, pp. 2554-2558, April 1982).   Just get anything near what you are trying to remember, and slowly (5 hours in this case), it converges to what you are trying to remember (an attractor in memory space).  It does show just how even peripheral parts of a concept (the cadence of what you are trying to remember when you speak it) are part of the concept itself.   Clearly, concepts are multidimensional.

Retinal physiology and the demise of the pure percept

Rooming with 2 philosophy majors warps the mind even if it was 50 years ago.  Conundrums raised back then still hang around.  It was the heyday of Bertrand Russell before he became a crank.  One idea being bandied about back then was the ‘pure percept’ — a sensation produced by the periphery  before the brain got to mucking about with it.   My memory about the concept was a bit foggy so who better to ask than two philosophers I knew.

The first was my nephew, a Rhodes in philosophy, now an attorney with a Yale degree.  I got this back when I asked –

I would be delighted to be able to tell you that my two bachelors’ degrees in philosophy — from the leading faculties on either side of the Atlantic — leave me more than prepared to answer your question. Unfortunately, it would appear I wasn’t that diligent. I focused on moral and political philosophy, and although the idea of a “pure precept” rings a bell, I can’t claim to have a concrete grasp on what that phrase means, much less a commanding one.

 Just shows what a Yale degree does to the mind.

So I asked a classmate, now an emeritus prof. of philosophy and got this back

This pp nonsense was concocted because Empiricists [Es]–inc. Russell, in his more empiricistic moods–believed that the existence of pp was a necessary condition for empirical knowledge. /Why? –>
1. From Plato to Descartes, philosophers often held that genuine Knowledge [K] requires beliefs that are “indubitable” [=beyond any possible doubt]; that is, a belief counts as K only if it [or at least its ultimate source] is beyond doubt. If there were no such indubitable source for belief, skepticism would win: no genuine K, because no beliefs are beyond doubt. “Pure percepts” were supposed to provide the indubitable source for empirical K.
2. Empirical K must originate in sensory data [=percepts] that can’t be wrong, because they simply copy external reality w/o any cognitive “shopping” [as in Photoshop]. In order to avoid any possible ‘error’, percepts must be pure in that they involve no interpretation [= error-prone cognitive manipulation].
{Those Es who contend  that all K derives from our senses tend to ignore mathematical and other allegedly a priori K, which does not “copy” the sensible world.} In sum, pp are sensory data prior to [=unmediated by] any cognitive processing.

So it seems as though the concept is no longer taken seriously.  To drive a stake through its heart it’s time to talk about the retina.

It lies in the back of our eyes, and is organized rather counter-intuitively.  The photoreceptors (the pixels of the camera if you wish) are the last retinal elements to be hit by light, which must pass through the many other layers of the retina to get to them.

We have a lot of them — at least 100,000,000 of one type (rods).  The nerve cells sending impulses back to the brain, are called ganglion cells, and there are about 1,000,000 in each eye.  Between the them are bipolar cells and amacrine cells which organize the information falling on the photoreceptors.

All this happens in something only .2 milliMeters thick.

The organization of information results in retinal ganglion cells responding to different types of stimuli.  How do we know?  Impale the ganglion cell with an electrode while still in the retina, and try out various visual stimuli to see what it responds to.

Various authorities put the number of retinal ganglion cell types in the mouse at 11, 12, 14, 19 and 22.  Each responds to a given type of stimulus. Here are a few examples:

The X-type ganglion cell responds linearly to brightness

Y cells respond to movement in a particular direction,

Blue-ON transmits the mean spectral luminance (color distribution) along the spectrum from blue to green.

From an evolutionary point of view, it would be very useful to detect motion.  Some retinal ganglion cells being responding before they should. How do we know this?  It’s easy (but tedious) to map the area of visual space a ganglion cell responds to — this is called its receptive field.  The responses of some anticipate the incursion of a moving stimulus — clearly this must be the way they are hooked to photoreceptors via the intermediate cells.

Just think about the way photoreceptors at the back of the spherical eye are excited by something moving in a straight line in visual space.  It certainly isn’t a straight line on the retinal surfaced.  Somehow the elements of the retina are performing this calculation and predicting where something moving in a straight line will be next.  Why  couldn’t the brain bedoing this?  Because it can be seen in isolated retinas with no brain attached.

Now for something even more amazing.  Each type of ganglion cell (and I’ve just discussed a few) tiles the retina. This means that every patch of the retina has a ganglion cell responding to each type of visual stimulus.  So everything hitting every area of the retina is being analyzed 11, 12, 14, 19 or 22 different ways simultaneously.

So much for the pure percept: it works for a digital camera, but not the retina.  There is an immense amount of computation of the visual input going right there, before anything gets back to the brain.

If you wish to read more about this — an excellent review is available, but it’s quite technical and not for someone coming to neuroanatomy and neurophysiology for the first time.  [ Neuron vol. 76 pp. 266 - 280 '12 ]

The New Clayden pp. 1029 – 1068

p. 1034 — “Small amounts of radicals are formed in many reactions in which the products are actually formed by simple ionic processes.”  Interesting — how ‘small’ is small?  

p. 1036 — A very improbable mechanism (but true) given in the last reaction involving breaking benzene aromaticity and forming a cyclopropene ring to boot.  

p. 1043 — Americans should note that gradient (as in Hammett’s rho constant) means slope (or derivative if the plot of substituents vs. sigma for a particular reaction isn’t a straight line).  However we are talking log vs. log plots, and you can fit an elephant onto a log log plot.  It’s worth remembering why logarithms are necessary iin the first place.  Much of interest to chemists (equilibrium constants, reaction rates) are exponential in free energy (of products vs. reactants in the first case, of transition state vs. reactions in the second).

p. 1044 — Optimally I shouldn’t have to remember that a positve rho (for reaction value) means electrons flow toward the aromatic ring in the rate determining step), but should gut it out from the electron withdrawing or pushing effects on the transition state, and how this affects sigma, by remembering what equilibrium constant is over what for sigma, and rho), but this implies a very high working memory capacity (which I don’t have unfortunately).  I think mathematicians do, which is why I’m so slow at it.  They have to keep all sorts of definitions in working memory at once to come up with proofs (and I do to follow them).  

If you don’t know what working memory is, here’s a link – http://en.wikipedia.org/wiki/Working_memory.  

Here are a few literature references 

        [ Proc. Natl. Acad. Sci. vol. 106 pp. 21017 - 21018 '09 ] This one is particularly interesting to me as it states that differences among people in working memory capacity are thought to reflect a core cognitive ability, because they strongly predict performance in fluid inteliigenece, reading, attentional control etc. etc.  This may explain why you have to have a certain sort of smarts to be a mathematician (the sort that helps you on IQ tests).  

       [ Science vol. 323 pp. 800 - 802 '09 ] Intensive training on working memory tasks can improve working memory capacity, and reduce cognitively related clinical symptoms.  The improvements have been associated with an increase in brain activity in parietal and frontal regions. 

I think there are some websites which will train working memory (and claim to improve it).  I may give them a shot. 

Unrelated to this chapter, but Science vol. 337 pp. 1648 – 1651 ’12, but worth bringing to the attention of the cognoscenti reading this –as there is some fascinating looking organometallic chemistry in it.  This is a totally new development since the early 60′s and I look forward to reading the next chapter on Organometallic chemistry.   Hopefully orbitals and stereochemistry will be involved there, as they are in this paper.  Fig 1 C has A uranium atom bound to 3 oxygens and 3 nitrogens, and also by dotted bonds to H and C.

p. 1050 — The unspoken assumption about the kinetic isotope effect is that the C-D and C-H bonds have the same strength (since the curve of potential energy vs. atomic separation is the same for both — this is probably true — but why?    Also, there is no explanation of why the maximum kinetic isotope effect is 7.1.  So I thought I’d look and see what the current Bible of physical organic chemistry had to say about it. 

Anslyn and Dougherty (p. 422 –> ) leave the calculation of the maximum isotope effect (at 298 Kelvin) as an exercise.  They also assume that the force constant is the same.  Exercise 1 (p. 482) says one equation used to calculate kinetic isotope effects is given below — you are asked to derive it 

kH/kD = exp [ hc (vbarH – vbarD)/2KT }, and then in problem #2 plug in a stretching frequency for C-H of 3000 cm^-1 to calculate the isotope effect at 298 Kelvin coming up with 6.5

Far from satisfying.  I doubt that the average organic chemist reading Anslyn and Dougherty could solve it.  Perhaps I could have  done it back in ’61 when I had the incredible experience of auditing E. B. Wilson’s course on Statistical Mechanics while waiting to go to med school (yes he’s the Wilson of Pauling and Wilson).   More about him when I start reading Molecular Driving Forces. 

On another level, it’s rather surprising that mass should have such an effect on reaction rates.  Bonds are about the distribution of charge, and the force between charged particles is 10^36 times stronger than that between particles of the same mass. 

p. 1052 — Entropy is a subtle concept (particularly in bulk thermodynamics), but easy enough to measure there.    Organic chemists have a very intuitive concept of it as shown here.

p. 1054 — Very slick explanation of the inverse isotope effect.  

Again out of context — but more chemistry seems to be appearing in Nature and Science these days.   A carbon coordinated to 6 iron atoms ( yes six ! ! ! ) exists in an enzyme essential for life itself — the plant enzyme nitrogenase which reduces N2 to usable ammonia equivalents for formation of amino acids, nucleotides.   No mention seems to be made about just how unusual this is.  See Science vol. 337 pp. 1672 – 1675 ’12. 

p. 1061 — The trapping of the benzyne intermediate by a Diels Alder is clever and exactly what I was trying to do years ago in a failed PhD project — see https://luysii.wordpress.com/2012/10/04/carbenes-and-a-defense-of-pre-meds-and-docs/

p. 1064 — In the mechanism of attack on epichlorohydrin, the reason for the preference of attack on the epoxide isn’t given — it’s probably both steric and kinetic, steric because attack on the ring is less hindered — the H’s are splayed out, and kinetic, because anything opening up a strained ring should have a lower energy transition state. 

An incredible paper which may turn diabetes treatment on its head

First off, why should a neurologist be interested in diabetes?  That’s for endocrinologists isn’t it?  Not really.  One of the most serious aspects of diabetes is its acceleration of vascular disease (atherosclerosis), with an increased risk of heart attack and stroke (which is where the neurologist comes in).  So you have a diabetic who has survived a stroke. You want to prevent the next one, and you know that they’re in a much higher risk group for another stroke (1) because they just had one (2) because they’re more likely to have another because of their diabetes.  Also vascular disease makes any neurologic problem worse, dementia, neuropathy you name it.

So I’ve always tried to stay current on diabetes.  Which brings me to Proc. Natl. Acad. Sci. vol. 109 pp. 14972 – 14976 ’12 (11 September issue).  You’d think that nearly a century after Banting and Best’s discovery of Insulin that there were no new wrinkles left.  Not so.

Now for a bit of anatomy and physiology.  In a very simplistic way, you can regard diabetes as not enough insulin.  Insulin lowers blood sugar and is needed after you eat.   It is made in the pancreas, an organ which secretes digestive enzymes into your gastrointestinal tract.  It also secretes insulin, but into the blood rather than the gut.  Insulin is made in relatively small collections of cells (1,000 – 10,0000) called islets dispersed through out the pancreas.  They are .1 – .2 milliMeters in diameter, and even though we have over 1,000,000 of them, they constitute 2% of the mass of the pancreas. 85% of the islet cells make insulin (the beta cells), another cell type (the alpha cells) make another protein (glucagon), also secreted into the blood, making it a hormone by definition.  Glucagon raises blood sugar, by causing the liver to make glucose and then pump it out.

Streptozotocin (a glucose derivative) made by soil bacteria, has the nasty property of selectively killing pancreatic beta cells, leaving everything else alone.  Naturally experimentalists love it and have used it to produce severe diabetes (fatal if untreated) in lab animals, and then try new treatments to see if they can come up with something better.

Amazingly, here’s one they didn’t try until now.  Given our present tools, you can pretty much knock out any gene you want in a mouse embryonic stem cell, implant it in another mouse, and (after a lot of failure), make a strain of mice lacking that particular gene (these are what’s called knockout mice).  What they knocked out was not glucagon (which has effects all over the body, including the brain), but the receptor for glucagon.  So even though there’s plenty of glucagon around, lacking the receptor cells don’t respond to it.  It’s like hearing a language you don’t know — the sound is there all right but you can’t internalize it and react to what’s being said.

It had been known for a while that giving streptozotocin to mice lacking the glucagon receptor doesn’t produce diabetes (or much else).  The finding as been treated with a good deal of skepticism, being blamed on other factors.  The authors of the present paper, found a way to put the glucagon receptor back into the liver (using a virus).  When they did this to a knockout mouse living happily despite being treated with strpetozotocin, their glucose shot up.  The virus didn’t hang around forever, glucagon receptor levels in the liver dropped and blood sugar dropped along with it.   Remember that a normal mouse dies of diabetic complications fairly quickly after receiving streptozotocin.

You couldn’t ask for much better proof, and a new way to treat diabetes may have been found.  Amazing.

As in all of medicine there are caveats.  The glucagon receptor is found in other organs, heart and brain among them, so blocking the action of glucagon this way  may have many other effects.

The New Clayden pp. 877 – 908

p. 878 — “The transition state has 6 delocalized pi electrons and thus is aromatic in character”.  Numerically yes, but the transition state isn’t planar, and there is all sorts of work showing how important planarity is to aromaticity. 

p. 881 — It seems to me that the arrow is wrong in the equation at the bottom. Entropy should increase when a Diels Alder product is broken apart, and since deltaG = deltaH  - T * deltaS heating the product should break it apart not cause it to form.  I guess the heat shown is required to increase molecular velocity so that collisions result in reaction.   Enough kinetic energy will blow anything apart (see Higgs particle).

p. 890 — “It is not cheating to use the regioselectivity of chemical reactions to tell us about the coefficients of the orbitals involved.”    I do think that this sort of thing is  cheating when you use the regioselectivity of chemical reactions as an explanation.  They are adding nothing new.  A real explanation predicts new phenomena, the way the anomeric effect does, for example.  You should contemplate the point at which a description of something becomes an explanation (e.g. epistemology).   It’s not the case here, but it was the case for Newton’s laws of gravitation.  Famously he said Hypotheses non dingo (“I frame no hypotheses”).  It appears in the following

I have not as yet been able to discover the reason for these properties of gravity from phenomena, and I do not feign hypotheses.

Yet his laws of gravity were used to predict all sorts of events never before seen, so they are explanatory in some sense.  

This sort of thing is just what a neurologist experiences learning functional neuroanatomy (e.g.  which part of the nervous system has which function).  Initially almost all of it was developed by studying neurologic deficits due to various localized lesions of the brain and spinal cord.  There’s a huge caveat involved — pulling the plug on a radio will stop the sound, but that isn’t how the sound is produced.  People with lesions of the occipital lobe lose the ability to see in certain directions (parts of their visual fields).  Understanding HOW the occipital lobe processes sensory input from the eyes has taken 50 years and is far from over.  

p. 892 — Unfortunately the rationale behind  the Woodward Hoffmann rules isn’t covered, so it appears incredibly convoluted and arbitrary.  Read the book — “The Control of Orbital Symmetry” which they wrote.  Also, unfortunately, the description of the rules uses the term ‘component’ in two ways.  At step two butadiene and the dienophile are each considered a component, as they are in steps 3, 4, and 5, then the two are mushed together into a single component in step 6. 

p. 894 — I haven’t been looking at the animations for a while, but those of the Diels Alder type reactions are incredible, and almost sexual.  You can rotate the two molecules in space and watch them come together and react.

p. 894 –”Remember, the numbers in brackets, [ 4 + 2 ] etc., refer to the numbers of atoms.  The numbers (4q +2)s and (4r)s in The Woodward Hoffman (should be Hoffmann) refer to the numbers of electrons.”  This is so very like math, where nearly identical characters are used to refer to quite different things.  Bold capital X might mean one thing, italic x another, script X still another.  They all sound the same when you mentally read them to yourself.  It makes life confusing. 

p. 894 — The Alder ene reactions — quite unusual.  The worst thing is that I remember nothing about them from years ago.  They must have been around as they were discovered by Alder himself (who died in 1958).  They produce some rather remarkable transformations, the synthesis of menthol from citronellal being one.  I wonder if they are presently used much in synthetic organic chemistry. 

p. 900 — How do you make OCN – SO2Cl, and why is it available commercially?

p. 904 — The synthesis of the sulfur containing 5 membered ring of biotin is a thing of beauty.  It’s extremely non-obvious beginning with a 7 membered ring with no sulfur at all. 

How fast is your biological clock ticking – II ?? Latest results.

The acceleration in genome sequencing capacity is just incredible.  In May 2010, I posted on the sequencing of the complete genomes of two parents and two of their kids.  Now in August of 2012 we have the results on the complete genomes of 78 parent children trios, along with 1,859 complete sequences from the same population, for comparison.

To get you started here’s the first post:

***

My family breeds about as fast as sequoias.  A cousin had a child at 46 who is presently burning up Columbia.  My brother had a child at 48, also doing OK.  But we do know that the older the parents, the more likely a kid is to have genetic problems (Sarah Palin & Trig).  So what are the odds and how do they change with age?

We don’t really know, but in the next five to ten years, we’ll have a good idea.  The first human genome project sequenced most of the 3,200,000,000 positions in our DNA. It cost billions and took years.  DNA sequencing technology marches on at an incredible pace.   A recent paper rated only 4 pages [ Science vol. 328 pp. 636 - 639 '10 ].  The complete genome sequence of 2 parents and 2 of their children was performed (to 1/100,000 accuracy yet).  There was known genetic disease in the family and the authors were looking for its cause.  The genome sequences of the parents and their children were compared position by position (using computers of course). If the base (adenine, thymine, guanosine, cytosine) at a given position differed from that of the parent supplying the surrounding DNA, a mutation had taken place between the generations.  Since they had looked at the entire genome, they could count the number of mutations they found.  The rate of mutation was 1.1 per 100 million positions, making about 30 new mutations between generations.

So this isn’t by guess and by gosh, but an actual mutation rate and a count.  While medical science marches on,  our biology has not.  We’ll soon be able to give the numbers of the mutations occurring between parents and progeny at a variety of parental ages, when enough of this sort of thing is done. And it definitely will be done as only 10 years separates the incredibly laborious first human genome project from this paper.

Interestingly, the authors didn’t mention anything about this application in their paper, so this may be (gasp) an original idea.

***

This brings us to [ Nature vol. 488 pp. 439, 467 - 468 (editorial) and 471 - 475 (the actual paper) ].  Some 78 trios (momma poppa and baby) had their entire genomes sequenced, looking for changes found in the baby, not present in either parent.  The paper is from Iceland, which has a small population (317,000), and complete genealogical records going back 10 generations (thanks to National Censuses and Parish records).  In addition, to find out what a normal Icelandic genome was, they completely sequenced the genomes of 1,859 more.   The more times a single genome is sequenced, the more accurate it is, and 78 trios were sequenced to 30x coverage — making it quite accurate.  Knowing both the father’s genome and the mothers (which are of course different), any change in the baby from a carbon copy could be definitively linked to the parent providing the mutated sperm or egg.

The bottom line is that mothers contribute 15 mutations to the baby regardless of age, while fathers transmit more (25 to 65).  The most important point, is that the older the father, the greater the number of mutations transmitted.  Why do men transmit more than women — because they are continually producing new sperm, while the eggs each woman possesses were present at birth.  Anyone who has ever played the telephone game knows that messages get distorted when copied.  Nonetheless, the accuracy of the copying is incredible (around 1 error per 100 million bases copied).

Should you be worried about this?  Possibly, but look who they studied – of the 78 children studied 44 had autism spectrum disorder and 21 were schizophrenic.  They don’t say anything about the other 13.

I can’t fault the authors for looking for genetic information to help us all understand autism and schizophrenia, but remember these weren’t normal kids. This work may not generally apply.  The study definitely needs to be repeated with normal progeny before getting too excited.

As usual, even though there 4,933 mutations were discovered in this population, none were found more than once.  This has been discussed before.  For details see http://luysii.wordpress.com/2010/04/25/tolstoy-was-right-about-hereditary-diseases-imagine-that/ and

http://luysii.wordpress.com/2010/07/29/tolstoy-rides-again-autism-spectrum-disorder/

Would you be smarter if your mother received different prenatal advice?

Back in the day we worried a lot about the amount of weight a woman gained during pregnancy.  Too much weight gain increased the risk of pregnancy associated hypertension (and worse pre-eclampsia, and even worse eclampsia with fetal and even maternal death).  It also increases the likelihood of pregnancy associated diabetes, with its adverse effects on the fetus.

So when my wife was carrying our two boys, we watched her weight like a hawk (particularly since there was diabetes in her family).  Fellow med students and their wives did the same.

That may have not been so good according to a recent study of birth weight differences between identical twins [ Proc. Natl. Acad. Sci. vol. 109 pp. 11366 - 11371 '12 ].  To cut to the chase, they found that the bigger twin had a bigger brain and was smarter (by a few IQ points).   It was a big study (139 twin pairs or which 85 were identical — monozygotic if you want to impress your friends).  They only looked at mild levels of weight disparity — the difference in birth weight had to be less than 20% of the weight of the heavier twin.

Greater birth weight resulted in greater  brain volume as shown by magnetic resonance imaging (MRI), mostly due to more cerebral cortex.  Interestingly the cortex wasn’t any thicker in the heavier twin, there was just more of it (the brain was more wrinkled).

Interestingly, the fraternal twins were smarter than the monozygotics by having full scale IQs of 112 vs. 109.  What’s going on? The average IQ should be 100.  However, they excluded twins with IQs under 80, e.g. they only looked at 1.3 standard deviations (which is 15 points) below the mean, while keeping everything above it.   It’s an interesting mathematical problem, which I don’t have time to solve, to integrate the bell curve from 1.3 standard deviations below the mean to all the way above it, and see what the average would be — my guess is that this is why the average IQs are so high.

There is also something called the Flynn effect which you should know about, as long as we’re talking IQ. [ Science vol. 323 p. 989 '09 ] It was described  nearly 30 years about by  James Flynn of New Zealand.  He noted  that IQ scores rose steadily in the 20th century for children and adults in Western nations.  Using the late 20th century average IQ score of 100, the comparative score for the year 1900 was calculated to 60 — something clearly not true.  Actually the improvements were not in general knowledge or mathematics, but in abstract reasoning.

The conclusions of the PNAS paper seem solid, but behind your back I’ve made several inferential leaps, some of which can be checked fairly easily.  Do bigger babies have bigger brains, and if they do, does this persist throughout life.  Second do people with bigger brains have higher IQs — again the extremes must be cut off — down 2 -3 standard deviations from the average brain size, IQ is way down, and the smaller the brain down here the smaller the IQ.  Similarly with macrocephaly (large brain) — usually there is something wrong.  These are factual matters, whose results are probably in the literature already.

What’s great about the paper is that it controls for heredity, as the genomes of identical twins should be nearly identical — I’m not sure anyone has looked, and given the recent haphazard way our genomes vary between us, it should be.  For details please see https://luysii.wordpress.com/2012/07/31/how-badly-are-thy-genomes-oh-humanity/

The fourth inferential leap, is that bigger birth weight in a twin has to do with better nutrition in utero.  Clearly the bigger twin had more proliferation of cortical precursors resulting in more cerebral cortex.  So maybe your mother should have had a few more fudge cakes and ice creams when she was carrying you.

I don’t know the answer to these questions, but as Mark Twain said :”There is something fascinating about science.  One gets such wholesale returns of conjecture out of such a trifling investment in fact”.

Why even great drugs have serious side effects in some patients

Finding good drugs is hard enough, but even great ones are often laid low by unexpected side effects.  This has to do with the tremendous genetic variation in people, about which, more later.  But first a true story from the past.

Neurologists treat epilepsy.  There was a period of 17 years when I was in practice when not a single new  drug against epilepsy (anticonvulsant) was introduced in the USA.  Each new drug would seem to be the answer for a small group of patients that nothing had helped before.

Felbamate (Felbatol) was one such anticonvulsant.  It helped people that nothing else touched. In the year after introduction some 150,000 people were taking it.   I had several very happy patients using Felbatol in the 90s.   1 year later the bomb dropped.  Ten cases of total bone marrow failure (aplastic anemia) had developed in patients taking the drug, a lethal complication.  Every neurologist (and probably every physician) got an urgent letter from the FDA.

Normally, unless there is an allergic reaction, anticonvulsants are never stopped suddenly.  They are tapered over a week or two.  Why?  Basically all anticonvulsants are sedating.  People adapt to this, and it’s like driving a car with one foot on the brake.  Remove the brake and the car shoots forward.  So neurologists all over the country brought patients into the hospital as the drug was immediately stopped.  We were quite worried that the previously uncontrolled seizures would flare.

I had one such patient.  Her family was quite worried about the possible side effects of suddenly stopping Felbamate.  I managed to control myself (hopefully) as I told them there was no side effect worse than death.  As risky as it is, there are still about 12,000 people taking the drug (after being carefully told about the risks) according to Wikipedia.  That’s how good a drug it is.

Why wasn’t this terrible complication picked up in the phase I, II, III studies of Felbamate — 10 cases in 150,000 people is 1/15,000, and no drug study for epilepsy was that large back then.  The incidence of epilepsy in adults is probably around 1%, meaning that some 1,500,000 people would have to be screened to find those 15,000.  So effectively there is no way to find such a rare complication before the drug was released.

A paper last month in Science (vol. 337 pp. 100 – 104 ’12) showed why this sort of thing is almost certain to happen again and again.

DNA sequencing is getting faster and cheaper all the time, so large numbers of people can have parts of their genomes sequenced.  A recent post https://luysii.wordpress.com/2012/07/31/how-badly-are-thy-genomes-oh-humanity/ discussed a paper that  sequenced roughly three quarters of the genes coding for proteins in some 2,439 people — e.g. 15,585 protein coding genes.

The Science paper was more circumspect.  They sequenced ‘only’ 202 genes coding for proteins in 14,002 people.  These genes were chosen quite carefully out of the 20,000 or so protein coding genes we have.  The 202 genes were known drug targets — say the neurotransmitter uptake proteins targeted by SSRIs and tricyclic antidepressants, the dopamine receptors targeted by antipsychotics.  So were the 14,002 people chosen to have their genes sequenced.  There were two ‘normal’ populations samples with 1,322 and 2,059 people each, and 12 populations chosen from people with particular diseases.  Most of these were European (12,514/14,002).

The findings essentially explain why we’ll always have rare side effects.  The total amount of DNA sequenced in each individual was 864,000 positions.  They found ‘rare’ variants (e.g. found in less than 1/200 people) quite commonly.  In fact in the group as a whole such rare variants occurred once every 21 positions in the Europeans.  The variants are the single nucleotide variants (SNVs).  Here’s a recap of just what a SNV is (for more detail see the link given above).  90% of the rare variants had never been seen before, even in these 202 proteins of great biologic and medical interest.

**** Recall that each nucleotide is one of four possibilities (A, T, G, C), and that each 3 nucleotides therefore has 4^3 = 64 possibilities.  61/64 combinations code for amino acids which, since we have only 20 gives a certain redundancy of the famed genetic code.   The other 3 combinations code for no amino acid (usually) and tell the machinery making proteins to stop.  Although crucial to our existence, these are called nonsense codons.

The genetic code is therefore 3fold degenerate (on average).  However, some amino acids are coded for by just 1 combination of 3 nucleotides while others are coded by as many as 6.  So some single nucleotide variants (SNVs) leave the amino acid coded for the same (these are the synonymous SNVs), while others change the amino acid (nonSynonymous SNVs), and possibly protein function.  *****

Certainly, not all of these variants will cause trouble, and our genomes are incredibly fault tolerant, as most of us carry very impaired genes for at least 35 of the proteins (e.g. they are truncated, so not a full protein is made).  Some almost certainly will cause unexpected reactions or side effects from a given drug.  There are so many SNVs out there.