Category Archives: Medicine in general

Could Gemfibrozil (Lopid) be used to slow down (or even treat) Alzheimer’s disease?

Is a treatment of Alzheimer’s disease at hand with a drug in clinical use for nearly 40 years? A paper in this week’s PNAS implies that it might (vol. 112 pp. 8445 – 8450 ’15 7 July ’15). First a lot more background than I usually provide, because some family members of the afflicted read everything they can get their hands on, and few of them have medical or biochemical training. The cognoscenti can skip past this to the text marked ***

One of the two pathologic hallmarks of Alzheimer’s disease is the senile plaque (the other is the neurofibrillary tangle). The major component of the plaque is a fragment of a protein called APP (Amyloid Precursor Protein). Normally it sits in the cellular membrane of nerve cells (neurons) with part sticking outside the cell and another part sticking inside. The protein as made by the cell contains anywhere from 563 to 770 amino acids linked together in a long chain. The fragment destined to make up the senile plaque (called the Abeta peptide) is much smaller (39 to 42 amino acids) and is found in the parts of APP embedded in the membrane and sticking outside the cell.

No protein lives forever in the cell, and APP is no exception. There are a variety of ways to chop it up, so its amino acids can be used for other things. One such chopper is called ADAM10 (aka Kuzbanian). ADAM10breaks down APP in such a way that Abeta isn’t formed. The paper essentially found that Gemfibrozil (commercial name Lopid) increases the amount of ADAM10 around. If you take a mouse genetically modified so that it will get senile plaques and decrease ADAM10 you get a lot more plaques.

The authors didn’t artificially increase the amount of ADAM10 to see if the animals got fewer plaques (that’s probably their next paper).

So there you have it. Should your loved one get Gemfibrozil? It’s a very long shot and the drug has significant side effects. For just how long a shot and the chain of inferences why this is so look at the text marked @@@@


How does Gemfibrozil increase the amount of ADAM10 around? It binds to a protein called PPARalpha which is a type of nuclear hormone receptor. PPARalpha binds to another protein called RXR, and together they turn on the transcription of a variety of genes, most of which are related to lipid metabolism. One of the genes turned on is ADAM10, which really has never been mentioned in the context of lipid metabolism. In any event Gemfibrozil binds to PPARalpha which binds more effectively to RAR which binds more effectively to the promoter of the ADAM10 gene which makes more ADAM10 which chops of APP in such fashion that Abeta isn’t made.

How in the world the authors got to PPARalpha from ADAM10 is unknown — but I’ve written the following to the lead author just before writing this post.

Dr. Pahan;

Great paper. People have been focused on ADAM10 for years. It isn’t clear to me how you were led to PPARgamma from reading your paper. I’m not sure how many people are still on Gemfibrozil. Probably most of them have some form of vascular disease, which increases the risk of dementia of all sorts (including Alzheimer’s). Nonetheless large HMOs have prescription data which can be mined to see if the incidence of Alzheimer’s is less on Gemfibrozil than those taking other lipid lowering agents, or the population at large. One such example (involving another class of drugs) is JAMA Intern Med. 2015;175(3):401-407, where the prescriptions of 3,434 individuals 65 years or older in Group Health, an integrated health care delivery system in Seattle, Washington. I thought the conclusions were totally unwarranted, but it shows what can be done with data already out there. Did you look at other fibrates (such as Atromid)?

Update: 22 July ’15

I received the following back from the author

Dear Dr.

Wonderful suggestion. However, here, we have focused on the basic science part because the NIH supports basic science discovery. It is very difficult to compete for NIH R01 grants using data mining approach.

It is PPARα, but not PPARγ, that is involved in the regulation of ADAM10. We searched ADAM10 gene promoter and found a site where PPAR can bind. Then using knockout cells and ChIP assay, we confirmed the participation of PPARα, the protein that controls fatty acid metabolism in the liver, suggesting that plaque formation is controlled by a lipid-lowering protein. Therefore, many colleagues are sending kudos for this publication.

Thank you.

Kalipada Pahan, Ph.D.

The Floyd A. Davis, M.D., Endowed Chair of Neurology


Departments of Neurological Sciences, Biochemistry and Pharmacology

So there you have it.  An idea worth pursuing according to Dr. Pahan, but one which he can’t (or won’t).  So, dear reader, take it upon yourself (if you can) to mine the data on people given Gemfibrozil to see if their risk of Alzheimer’s is lower.  I won’t stand in your way or compete with you as I’m a retired clinical neurologist with no academic affiliation. The data is certainly out there, just as it was for the JAMA Intern Med. 2015;175(3):401-407 study.  Bon voyage.


There are side effects, one of which is a severe muscle disease, and as a neurologist I saw someone so severely weakened by drugs of this class that they were on a respirator being too weak to breathe (they recovered). The use of Gemfibrozil rests on the assumption that the senile plaque and Abeta peptide are causative of Alzheimer’s. A huge amount of money has been spent and lost on drugs (antibodies mostly) trying to get rid of the plaques. None have helped clinically. It is possible that the plaque is the last gasp of a neuron dying of something else (e.g. a tombstone rather than a smoking gun). It is also possible that the plaque is actually a way the neuron was defending itself against what was trying to kill it (e.g. the plaque as a pile of spent bullets).

Why drug discovery is so hard (particularly in the brain): Reason #28: The brain processes its introns very differently

Useful drug discovery for neurologic and psychiatric disease is nearly at a standstill. It isn’t for want of trying by basic researchers and big and small pharma. A recent excellent review [ Neuron vol. 87 pp. 14 – 27 ’15 ] helps explain why. In short, the brain processes its protein coding genes rather differently.

This post assumes you know what introns, exons and alternate splicing are. For pretty much all the needed background see the following.


When splicing first came out I started making a list of proteins which were alternatively spliced. It is now safe to assume that any gene containing introns (95% of all protein coding genes [ Proc. Natl. Acad. Sci. vol. 112 pp. 17985 – 17990 ’08 ]) results in several protein products due to alternative splicing. The products produced vary from tissue to tissue, probably because most tissues express different splicing regulators.

Here are a few. A2BP1 (aka Rbfox1, aka FOX1) is a brain specific RNA splicing factor found only in postmitotic terminally differentiated neurons. It is deleted in 10% of glioblastomas. Another is nSR100 (neural Specific Related protein of 100 kiloDaltons) — see later.

To show how crucial alternative splicing is for the every existence of the brain, consider this. The neuronal splicing regulator PTBP2 is barely expressed in most tissues. It is upregulated in neurons. Both PTBP1 and PTBP2 are repressors of neural alternative splicing (but some genes are actually enhanced). In a given region of the brain either PTPB1 or PTBP2 is expressed (but not both). PTBP1 promotes skiping of a neural specific exon (exon #10) in PTBP2 transcripts. This exposes a premature termination codon in PTBP2 leading to nonsense mediated decay (NMD). PTPB1 is expressed in most nonNeural tissues and neural precursor cells, but is silenced in developing neurons by the microRNA miR-124. The mRNA for PTBP2 contains an alternative exon which triggers nonsense mediated decay (NMD) when skipped. Inclusion of the exon requires positive transacting factors such as nSR100 in neurons. Repression is mediated by PTBP1 in undifferentiation. microRNAs (which ones?) downregulate PTBP1 during neuronal differentiation, relieving the negative regulation of PTBP2. Depletion of PTBP1 in fibroblasts is enough for PTBP2 induction and neuronal transdifferentiation.

It gets more complicated still. PTBP1 inhibits splicing of introns at the 3′ end of some genes involved in presynaptic function. This results in nuclear retention and turnover via components of the nuclear RNA surveillance machinery. As PTBP1 is downregulated during neuronal differentiation, the target introns are spliced out and the mature mRNAs are found.

Now we get to microExons, something unknown until 2014. For more details see —
Briefly, microexons are defined as exons containing 50 nucleotides or less (the paper says 3 – 27 nucleotides). They have been overlooked, partially because their short length makes them computationally difficult to find. Also few bothered to look for them as they were thought to be unfavorable for splicing because they were too short to contain exonic splicing enhancers. They are so short that it was thought that the splicing machinery (which is huge) couldn’t physically assemble at both the 3′ and 5′ splice sites. So much for theory, they’re out there.

The inclusion in the final transcript of most identified neural microExons is regulated by a brain specific factor nSR100 (neural specific SR related protein of 100 kiloDaltons)/SRRM4 which binds to intronic enhancer UGC motifs close to the 3′ splice sites, resulting in their inclusion. They are ‘enhanced’ by tissue specific RBFox proteins. nSR100 is said to be reduced in Autism Spectrum Disorder (really? all? some?). nSR100 is strongly coexpressed in the developing human brain in a gene network module M2 which is enriched for rare de novo ASD assciated mutations.

MicroExons are enriched for lengths which are multiples of 3 nucleotides. Recall that every 3 nucleotides in mRNA codes for an amino acid. This implies strong selection pressure was used to preserve reading frames as 3n+1 and 3n+2 produce a frameshift. The microExons are enriched in charged amino acids. Most microExons show high inclusion at late stages of neuronal differentiation in genes associated with axon formation and synapse function. A neural specific microExon in Protrudin/Zfyve27 increases its interation with Vessicale Associated membrane protein associated Protein VAP) and to promote neurite outgrowth.

[ Proc. Natl. Acad. Sci. vol. 112 pp. 3445 – 3450 ’15 ] Deep mRNA sequencing of mouse cerebral cortex expanded the list of alternative splicing events TENfold and showed that 72% of multiexon genes express multiple splice variants. Among the newly discovered alternatively spliced exon are 1,104 exons involved in nonsense mediated decay (NMD). THey are enriched in RNA binding proteins including splicing factors. Another set of alternatively spliced NMD exons is found in genes coding for chromatin regulators. Conservation of NMD exons is found in lower vertebrates, but those involving chromatin regulators are found later into the mammalian lineage. So the transcriptome in the brain is even more complicated.

A bit more about the actual effects on protein structure of alternate splicing. The sites chosen for this aren’t random. Cell and tissue differentially regulated alternative splicing events are significantly UNDERrepresented in functionally defined folded domains in proteins, they are enriched in regions of protein disorder that typically are surface accessible and embed short linear interaction motifs (with other proteins and ligands). Among a set of analyzed neural specific exons enriched in disordered regions, 1/3 promoted or disrupted interactions with partner proteins. So regulated exon splicing might specify tissue and cell type specific protein interaction networks. They regard their inclusion/exclusion as protein surface microsurgery.

How much can a little microexon do to protein function? Here’s an example of a 6 nucleotide microexon (two amino acids). Insertion of the microExon in the nuclear adaptor protein Apbb1 enhances its interaction with Kat5/Tip60 a histone deacetylase. The microExon adds Arginine and Glutamic acid to a phosphotyrosine binding domain (PTB domain) which binds Kat4. This enhances binding.

Had enough? The complexity is staggering and I haven’t even talked about recursive splicing — that’s for another post, but here’s a reference if you can’t wait — [ Nature vol. 521 pp. 300 – 301, 371 – 375, 376 – 379 ’15 ]. Pity the drug chemist figuring out which alternatively spliced form of a brain protein to attack (particularly if it hasn’t been studied for microExons).

How little we know

Well it’s basic biochem 101, but enzymes only allow equilibrium to be reached faster (by lowering activation energy), they never change it. This came as a shock to the authors of [ Proc. Natl. Acad. Sci. vol. 112 pp. 6601 – 6606 ’15 ] when Cytosolic Nonspecific DiPeptidase 2 (CNDP2), a proteolytic enzyme, was found to tack the carboxyl group of lactic acid onto the amino group of a variety of amino acids, essentially running the proteolytic reaction in reverse. Why? Because intracellular levels of lactic acid and amino acids are in the high microMolar to milliMolar range. It’s Le Chatelier’s principle in action.

The compounds are called N-Lactoyl amino acids. No one had ever seen them before. They are part of the ‘metabolome’ — small molecules found in our bodies. God knows what they do. The paper was really shocking to me for another reason, because I had no idea how many members the metabolome has.

How large is the metabolome? Make a guess.

Well METLIN ( has 240,000, and Human Metabolome DataBase has 42,000. I doubt that we know what they are all doing. Undoubtedly some of them are binding to proteins producing physiologic effects. Drug chemists may be mimicking some of them unknowingly, producing untoward and unexpected side effects.

What’s even more shocking to me is the following statement from the paper. State of the art untargeted metabolomics studies still report ‘up to’ 40% unidentified, but potentially important metabolitcs which can be detected reproducibly. The unknown metabolites are only rarely characterized because of the extensive work required for de novo structure determination..

So we really don’t know everything that’s out there in our bodies, and even if we did, we don’t know what they are doing. Drug discovery is hard because we only dimly understand the system we are trying to manipulate. Until I read this paper, I had no idea just how dim this is.

Why making money in the stock market is as stressful than a missile attack (for me)

I made a ton of money in stock market in the past 7 weeks. I hated every minute of it. I found the stress very hard to take, particularly the anxiety and the inability to think of little else. As wives often do, my wife told me I’m nuts. “Don’t you remember how hard you worked for those neurosurgeons? This is the easiest money you’ve ever made”. I made probably more than half a year’s salary from them in absolute amount, and I was well paid.

But that was 25 years ago. For details see Yes, I was on every other night and probably was out between midnight and 8 AM every third night on call with a full day’s work to follow before sleep. It was physically demanding, but not particularly stressful mentally. “All you could lose was money, but in practice you could lose a life”. True, but I knew that I’d never make a mistake of omission, or commission or due to lack of knowledge (fairly arrogant but true, I really kept up with the literature in practice). Decisions didn’t always work out, initial diagnoses weren’t always correct, untoward reactions ensued (particularly to drugs), but I always knew that I given it my best shot. The stress came from doing everything right but still being unable to help, watching helplessly as patients deteriorated.

Thinking this over, after what my wife said, I realized that I was very much like a patient who really educated me.

Neurologists see headaches, lots of headaches. This young man came to see me about them, complete with his own (correct) diagnosis of their cause — a divorce in progress. Taking the history always starts things off, and a standard question is “When did the headaches start?” Oh, just after I got back from Riyadh. (This was during the Gulf War). Riyadh? Wasn’t there a missile attack that killed people while you were over there? Yes. Didn’t you have headaches then? No, but this divorce is killing me, doc. He was right.

So just like the shrinks say, it isn’t the situation itself, but how you perceive it.

Is natural selection disprovable?

One of the linchpins of evolutionary theory is that natural selection works by increased reproductive success of the ‘fittest’. Granted that this is Panglossian in its tautology — of course the fittest is what survives, so of course it has greater reproductive success.

So decreased reproductive success couldn’t be the result of natural selection could it? A recent paper says that is exactly what has happened, and in humans to boot, not in some obscure slime mold or the like.

The work comes from in vitro fertilization which the paper says is responsible for 2 -3 % of all children born in developed countries — seems high. Maternal genomes can be sequenced and the likelihood of successful conception correlated with a variety of variants. It was found that there is a strong association between change in just one nucleotide (e.g. a single nucleotide polymorphism or SNP) and reproductive success. The deleterious polymorphism (rs2305957) decreases reproductive success. This is based on 15,388 embryos from 2,998 mothers sampled at the day-5 blastocyst stage.

What is remarkable is that the polymorphism isn’t present in Neanderthals (from which modern humans diverged between 100,000 and 400,000 year ago). It is in an area of the genome which has the characteristics of a ‘selective sweep’. Here’s the skinny

A selective sweep is the reduction or elimination of variation among the nucleotides in neighbouring DNA of a mutation as the result of recent and strong positive natural selection.

A selective sweep can occur when a new mutation occurs that increases the fitness of the carrier relative to other members of the population. Natural selection will favour individuals that have a higher fitness and with time the newly mutated variant (allele) will increase in frequency relative to other alleles. As its prevalence increases, neutral and nearly neutral genetic variation linked to the new mutation will also become more prevalent. This phenomenon is called genetic hitchhiking. A strong selective sweep results in a region of the genome where the positively selected haplotype (the mutated allele and its neighbours) is essentially the only one that exists in the population, resulting in a large reduction of the total genetic variation in that chromosome region.

So here we have something that needs some serious explaining — something decreasing fecundity which is somehow ‘fitter’ (by the definition of fitness) because it spread in the human population. The authors gamely do their Panglossian best explaining “the mitotic-error phenotype (which causes decreased fecundity) may be maintained by conferring both a deleterious effect on maternal fecundity and a possible beneficial effect of obscured paternity via a reduction in the probability of successful pregnancy per intercourse. This hypothesis is based on the fact that humans possess a suite of traits (such as concealed ovulation and constant receptivity) that obscure paternity and may have evolved to increase paternal investment in offspring.

Nice try fellas, but this sort of thing is a body blow to the idea of natural selection as increased reproductive success.

There is a way out however, it is possible that what is being selected for is something controlled near to rs2305957 so useful, that it spread in our genome DESPITE decreased fecundity.

Don’t get me wrong, I’m not a creationist. The previous post described some of the best evidence we have in man for another pillar of evolutionary theory — descent with modification. Here duplication of a single gene since humans diverged from chimps causes a massive expansion of the gray matter of the brain (cerebral cortex).


Addendum 13 April

I thought the following comment was so interesting that it belongs in the main body of the text


Mutations dont need to confer fitness in order to spread through the population. These days natural selection is considered a fairly minor part of evolution. Most changes become fixed as the result of random drift, and fitness is usually irrelevant. “Nearly neutral theory” explains how deleterious mutations can spread through a population, even without piggybacking on a beneficial mutation; no need for panglossian adaptive hypotheses.

Here’s my reply

Well, the authors of the paper didn’t take this line, but came up with a rather contorted argument to show why decreased fecundity might be a selective advantage, rather than just saying it was random drift. They also note genomic evidence for a ‘selective sweep’ — decreased genomic heterogeneity around the SNP.

No wonder we get sick

“It is estimated that a human cell repairs 10,000 – 20,000 DNA lesions per day” This is the opening sentence of Proc. Natl. Acad. Sci. vol. 112 pp. 3997 – 4002 ’15, but no source for this estimate is given. The lesions range from single and double strand breaks in the sugar phosphate backbone of the DNA helix, to hydrolytic losses of a DNA base from the backbone, to chemical modification of the DNA bases themselves — oxidation etc. etc.

What needs explaining then, is why we stay as well as we do.

Should pregnant women smoke pot?

Well, maybe this is why college board scores have declined so much in recent decades that they’ve been normed upwards. Given sequential MRI studies on brain changes throughout adolescence (with more to come), we know that it is a time of synapse elimination. (this will be the subject of another post). We also know that endocannabinoids, the stuff in the brain that marihuana is mimicking, are retrograde messengers there, setting synaptic tone for information transmission between neurons.

But there’s something far scarier in a paper that just came out [ Proc. Natl. Acad. Sci. vol. 112 pp. 3415 – 3420 ’15 ]. Hedgehog is a protein so named because its absence in fruitflies (Drosophila) causes excessive bristles to form, making them look like hedgehogs. This gives you a clue that Hedgehog signaling is crucial in embryonic development. A huge amount is known about it with more being discovered all the time — for far more details than I can provide see

Unsurprisingly, embryonic development of the brain involves hedgehog, e,g, [ Neuron vol. 39 pp. 937 – 950 ’03 ] Hedgehog (Shh) signaling is essential for the establishment of the ventral pattern along the whole neuraxis (including the telencephalon). It plays a mitogenic role in the expansion of granule cell precursors during CNS development. This work shows that absence of Shh decreases the number of neural progenitors in the postnatal subventricular zone and hippocampus. Similarly conditional inactivation of smoothened results in the formation of fewer neurospheres from progenitors in the subventricular zone. Stimulation of the hedgehog pathway in the mature brain results in elevated proliferation in telencephalic progenitors. It’s a lot of unfamiliar jargon, but you get the idea.

Of interest is the fact that the protein is extensively covalently modified by lipids (cholesterol at the carboxy terminal end and palmitic acid at the amino terminal end. These allow hedgehog to bind to its receptor (smoothened). It stands to reason that other lipids might block this interaction. The PNAS work shows this is exactly the case (in Drosophila at least). One or more lipids present in Drosophila lipoprotein particles are needed in vivo to keep Hedgehog signaling turned off in wing discs (when hedgehog ligand isn’t around). The lipids destabilize Smoothtened. This work identifies endocannabinoids as the inhibitory lipids from extracts of human very low density lipoprotein (VLDL).

It certainly is a valid reason for women not to smoke pot while pregnant. The other problem with the endocannabinoids and exocannabinoids (e.g. delta 9 tetrahydrocannabinol), is that they are so lipid soluble they stick around for a long time — see

It is amusing to see regulatory agencies wrestling with ‘medical marihuana’ when it never would have gotten through the FDA given the few solid studies we have in man.

Why drug discovery is so hard: Reason #26 — We’re discovering new players all the time

Drug discovery is so very hard because we don’t understand the way cells and organisms work very well. We know some of the actors — DNA, proteins, lipids, enzymes but new ones are being discovered all the time (even among categories known for decades such as microRNAs).

Briefly microRNAs bind to messenger RNAs usually decreasing their stability so less protein is made from them (translated) by the ribosome. It’s more complicated than that (see later), but that’s not bad for a first pass.

Presently some 2,800 human microRNAs have been annotated. Many of them are promiscuous binding more than one type of mRNA. However the following paper more than doubled their number, finding some 3,707 new ones [ Proc. Natl. Acad. Sci. vol. 112 pp. E1106 – E1115 ’15 ]. How did they do it?

Simplicity itself. They just looked at samples of ‘short’ RNA sequences from 13 different tissue types. MicroRNAs are all under 30 nucleotides long (although their precursors are not). The reason that so few microRNAs have been found in the past 20 years is that cross-species conservation has been used as a criterion to discover them. The authors abandoned the criterion. How did they know that this stuff just wasn’t transcriptional chaff? Two enzymes (DROSHA, DICER) are involved in microRNA formation from larger precursors, and inhibiting them decreased the abundance of the ‘new’ RNAs, implying that they’d been processed by the enzymes rather than just being runoff from the transcriptional machinery. Further evidence is that of half were found associated with a protein called Argonaute which applies the microRNA to the mRBNA. 92% of the microRNAs were found in 10 or more samples. An incredible 23 billion sequenced reads were performed to find them.

If that isn’t complex enough for you, consider that we now know that microRNAs bind mRNAs everywhere, not just in the 3′ untranslated region (3′ UTR) — introns, exons. MicroRNAs also bind pseudogenes, SINEes, circular RNAs, nonCoding RNAs. So it’s a giant salad bowl of various RNAs binding each other affecting their stability and other functions. This may be echoes of prehistoric life before DNA arrived on the scene.

It’s early times, and the authors estimate that we have some 25,000 microRNAs in our genome — more than the number of protein genes.

As always, the Category “Molecular Biology Survival Guide” found on the left should fill in any gaps you may have.

One rather frightening thought; If, as Dawkins said, we are just large organisms designed to allow DNA to reproduce itself, is all our DNA, proteins, lipids etc, just a large chemical apparatus to allow our RNA to reproduce itself? Perhaps the primitive RNA world from which we are all supposed to have arisen, never left.

The dietary guidelines have been changed — what are the faithful to believe now ?

While we were in China dietary guidelines shifted. Cholesterol is no longer bad. Shades of Woody Allen and “Sleeper”. It’s life imitating art.

Sleeper is one of the great Woody Allen movies from the 70s. Woody plays Miles Monroe, the owner of (what else?) a health food store who through some medical mishap is frozen in nitrogen and is awakened 200 years later. He finds that scientific research has shown that cigarettes and fats are good for you. A McDonald’s restaurant is shown with a sign “Over 795 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 Served”

Seriously then, should you believe any dietary guidelines? In my opinion you shouldn’t. In particular I’d forget the guidelines for salt intake (unless you actually have high blood pressure in which case you should definitely limit your salt). People have been fighting over salt guidelines for decades, studies have been done and the results have been claimed to support both sides.

So what’s a body to do? Well here are 4 things which are pretty solid (which few docs would disagree with, myself included)

l. Don’t smoke
2. Don’t drink too much (over 2 drinks a day), or too little (no drinks). Study after study has shown that mortality is lowest with 1 – 2 drinks/day
3. Don’t get fat — by this I mean fat (Body Mass Index over 30) not overweight (Body Mass Index over 25). The mortality curve for BMI in this range is pretty flat. So eat whatever you want, it’s the quantities you must control.
4. Get some exercise — walking a few miles a week is incredibly much better than no exercise at all — it’s probably half as good as intense workouts — compared to doing nothing.

Not very sexy, but you’re very unlikely to find anyone telling you the opposite 50years from now.

It’s off topic, but I’d use the same degree of skepticism about the dire predictions of the Global Warming (AKA Climate change) people, particularly since there has been no change in global mean temperature this century.

The butterfly effect in cancer

Fans of Chaos know all about the butterfly, where a tiny change in air current produced by a butterfly’s wings in Brazil leads to a typhoon in Java. Could such a thing happen in cell biology? [ Proc. Natl. Acad. Sci. vol. 112 pp. 1131 – 1136 ’15 ] comes close.

The Cancer Genome Project has spent a ton of money looking at all the mutations of all our protein coding genes which occur in various types of cancers. It was criticized as we already knew that cancer is effectively a hypermutable state, and that it would just prove the obvious. Well it did, but it also showed us just what a formidable problem cancer actually is.

For instance [ Nature vol. 489 pp. 519 – 525 ’12 ] is report from the Cancer Genome Atlas of 178 cases of squamous cell cancer of the lung. There are a mean of 360 exonic mutations, 165 genomic rearrangements, and 323 copy number alterations per tumor. The technical details in the rest of the paragraph can be safely ignored but the point is that there no consistent pattern of mutation was found (except for p53 which is mutated in over 50% of all types of cancer, which we knew long before the Cancer Genome Atlas). Recurrent mutations were found in 11 genes. p53 was mutated in nearly all. Previously unreported loss of function mutations were seen in the class I major histocompatibility (HLA-A). Several pathways were altered relatively consistently (NFE2L2, KEAP1 in 34%, squamous differentiation genes in 44%, PI3K genes in 47% and CDKN2A and RB1 in 72%). EGFR and kRAS mutations are rare in squamous cell cancer of the lung (but quite common in adenocarcinoma). Alterations in FGFR are quite common in squamous cell carcinomas.

This sort of thing (which has been found in all the many types of tumors studied by the Cancer Genome Atlas) lead to a degree of hopelessness in looking for the holy grail of a single ‘driver mutation’ which leads to cancer with its attendant genomic instability.

All is not lost however.

MCF-10A is an immortalized epithelial cell line derived from human breast tissue. It is capable of continuous growth, but is far from normal: (1) an abnormal complement of chromosomes ) (2) threefold amplification of the MYC oncogene, and (3) deletion of a known tumor suppressor . It does lack some mutations found in breast cancer. For instance, the Epidermal Growth Factor Receptor 2 (ERRBB2) is not amplified. The cell line doesen’t express the estrogen and progesterone receptors — making it similar to triple negative breast cancer.

A single amino acid mutation (Arginine for Histidine at amino acid #1047 ) in the catalytic subunit of a very important protein kinase (p110alpha of the PIK3CA gene) was put into the MCF-10A cell line (which they call MCF-1A-H1047R). The mutation was chosen because it is one of the most frequently encountered cancer specific mutations known. Exome sequencing of the entire genome showed that this was the only change — but the control sequences outside the exons weren’t studied, a classic case of the protein centric style of molecular biology.

In the (admittedly not completely normal) cell line, the mutation produced a cellular reorganization that far exceeds the known signaling activities of PI3K. The proetins expressed were stimilar to the protein and RNA signatures of basal breast cancer. The changes far exceeded the known effects of PIK3CA signaling. The phosphoproteins of MCF-1A-H1047R are extremely different. Inhibitors of the kinase induce only a partial reversion to the normal phenotype.

They plan to study the epigenome. This is signifcant as breast cancers are said in the paper to have tons of mutations changing amino acids in proteins (4,000 per tumor). In my opinion they should do whole genome sequencing of MCF-A1-H1047R as well.

The mutant becomes fully transformed whan a second mutation (of KRAS, an oncogene) is put in. This allows them to form tumors in nude mice. Recall that nude mice (another rodent beloved of experimental biologists — see the previous post on the Naked Mole Rat) has a very limited immune system, allowing grafts of human cells to take root and proliferate.

How close the initial cell line is to normal is another matter. Work on a similar cell line the (3T3 fibroblast) has been criticized because that cell is so close to neoplastic. At least the mutant MCF-1A-H1047R cells aren’t truly neoplastic as they won’t produce tumors in nude mice. However, mutating just one more gene (KRAS) turns MCF-1A-H1047R malignant when transplanted.

The paper is also useful for showing how little we really understand about cause and effect in the cell. PI3K has been intensively studied for years because it is one of the major players telling cells to grow in size rather than divide. And yet “the mutation produced a cellular reorganization that far exceeds the known signaling activities of PI3K”


Get every new post delivered to your Inbox.

Join 77 other followers