Category Archives: Molecular Biology

Don’t get your hopes up — but

Amyotrophic lateral sclerosis (ALS) is a God-awful disease, where patients progressively weaken and die because they aren’t strong enough to breathe, remaining mentally intact the entire time. A recent paper [ Science vol. 348 pp. 239 – 242 ‘ 15 ] showed that a drug already released by the FDA for treating hypertension — Wytensin (Guanabenz) was of benefit in a mouse model of the disease. So the drug is out there. If I were still in practice, I’d certainly give it a shot in my patients — off-label use be damned. Even better, enterprising organic chemists synthesized an analogue of Wytensin (Sephin1) which doesn’t lower blood pressure, but which still works in the mouse model.

Here’s why you shouldn’t get your hopes up too high. [ Nature vol. 4564 pp. 682 – 685 ’08 ] The work using SOD1 mutant mice (the mouse model of ALS mentioned above) is quite sloppy and nearly 12 drugs with benefit in mouse models have had no benefit in clinical trials. Minocycline which was effective in 4 studies in mice actually made things worse in a clinical trial of over 400 patients .

Now for a bit of background. Most cases of ALS aren’t familial, but a few are. One protein Superoxide Dismutase 1 (SOD1) was found to mutated in about 20% of familial ALS. It’s been studied out the gazoo, and some 140 different mutations have been found in its 153 amino acids in familial cases.

It’s hard to conceive of them all acting the same way, and literally thousands of papers have been written on the subject. It does seem clear that aggregated proteins occur in the dying neurons of ALS patients, but whether they are made mostly of SOD1 remains controversial (although it is present in the inclusions to some extent). Mature SOD1 is a 32 kiloDalton homodimeric metalloenzyme, in which each monomer contains Cu and Zn and one intrasubunit disulfide bond. It is one of the most abundant cellular proteins. It has a tendency to aggregate when overexposed.

The mouse results are impressive, as it improved established disease. In vivo, Sephin1 prevented the motor morphological and molecular defects of two unrelated protein misfolding diseases in mice (Charcot Marie Tooth 1B and ALS ! ! !). The mice had a mutant SOD1 (G93A). SOD1 mutants bind to Derlin1 on the the cytosolic side of the endoplasmic reticulum (ER) membrane blocking degradation of ER proteins causing ER stress. Very impressive ! ! ! !

None dare call it junk

There has been a huge amount of controversy about whether all the DNA we carry about has some purpose to carry out — or not. Could some of it be ‘junk’?.

At most 2% of our DNA actually codes for the amino acids comprising our proteins. Some (particularly the ENCODE consortium) have used the criterion of transcription of the DNA into RNA (a process which takes energy) as a sign that well over 50% of our genome is NOT junk. Others regard this transcription as the unused turnings from a lathe.

All agree however, that bacteria use a good deal of their small genomes to code for protein. The following paper quotes a figure of 84 – 89%.

Consider the humble leprosy organism.It’s a mycobacterium (like the organism causing TB), but because it essentially is confined to man, and lives inside humans for most of its existence, it has jettisoned large parts of its genome, first by throwing about 1/3 of it out (the genome is 1/3 smaller than TB from which it is thought to have diverged 66 million years ago), and second by mutation of many of its genes so protein can no longer be made from them. Why throw out all that DNA? The short answer is that it is metabolically expensive to produce and maintain DNA that you’re not using

If you want a few numbers here they are:
Genome of M. TB 4,441,529 nucleotides
Genome of M. Leprae 3,268,203 nucleotides
1,604 genes coding for protein
1,116 pseudoGenes (e.g. genes that look like they could code for proteins, but no longer can because of premature termination codons.

This brings us to the organism described in the paper — Trichodesmium erythraeum — a photosynthetic bacterium living in the ocean. When conditions are right it multiplies rapidly causing a red algal bloom (even though it isn’t an algae which are cellular). It’s probably how the Red Sea got its name.

The organism only uses 64% of its genome to code for its protein. The most interesting point is that 86% of the nonCoding (for protein anyway) DNA is transcribed into RNA.

The authors wrestle with the question of what the nonCoding DNA is doing.

“Because it is thought that many bacteria are deletion-biased (47, 77), stable maintenance of these elements from laboratory isolates to the natural samples suggest that they may be required in some fashion for growth both in culture and in situ.”

Translation: The nonCoding DNA probably isn’t junk.

They give it another shot.

“Others have hypothesized that the conserved repeat structures observed in some bacteria could function as recombination-dependent “promoter banks” for adaptation to new conditions, thereby allowing relatively quick “rewiring” of metabolism in subpopulations”

Plausible, but why waste the energy transcribing the DNA into RNA if it isn’t doing anything for the organism doing the transcribing?

Never assume that what you can’t measure or don’t understand is unimportant.

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.

One reason our brain is 3 times that of a chimpanzee

Just based on the capacity of the skull, our brain is 3 – 4 times larger than that of our closest primate relative, the chimp. Most of the increase in size occurs in the cerebral cortex (the gray matter) just under the skull. Our cortex is thrown into folds because there is so much of it. Compare the picture of the mouse brain (smooth) and ours, wrinkled like a walnut

We now may have part of the explanation. A fascinating paper studied genetic differences between the progenitor cells from which the cortex arises (radial glia) in man and mouse. They found 56 protein coding genes expressed in our radial glia not present in the mouse (out of 20,000 or so).

One in particular called by the awful name ARHGAP11B is particularly fascinating. Why? Because it’s the product of a gene duplication of ARHGAP11A. When did this happen — after the human line split off from the chimp 6 million years ago. Chimps have no such duplication, just the original

Put ARHGAP11B into a developing mouse and its cortex expands so much it forms folds.

There has been all sorts of work on the genetic difference between man and chimp. There almost too many — [ Nature vol. 486 pp. 481 – 482 ’12 ] — some 20,000,000. Finding the relevant ones is the problem. ARHGAP11A is by far the best we’ve found to date.

Another fascinating story is the ‘language gene’ discovered in a family suffering from a speech and language disorder. It’s called FOXP2. Since the last common ancestor of humans and mice (70 megaYears ago) there have been only 3 changes in the 715 amino acids comprising the protein. 2 of them have occurred in the human lineage since it split with the chips 6 megaYears ago. So far no one has put the human FOXP2 gene into a chimp and got it to talk. For more details see

There is all sorts of fascinating molecular biology about what these two genes actually do in the cell, but that would make this post too long,. This is, in part, a chemistry blog and just what FOXP2 and ARHGAP11A actually do involves some beautiful and elegant chemistry — look up RhoGAP and Winged Helix transcription factors. Ferrari’s are beautiful cars, and become even more beautiful when you understand what’s going on under the hood. Chemistry gives you that for molecular, cellular and organismal biology.

Of what use is an inactive enzyme?

Why should a cell take the trouble make an enzyme protein with no enzymatic activity? It takes metabolic energy to store the information for a protein in DNA, transcribe the DNA into RNA and then translate the RNA into protein. Is this junk protein a la junk DNA? Not at all — and therein lies a tale.

All sorts of nasty bugs inveigle their way into cells, among them viruses (such as influenza) whose genome is made of RNA, rather than DNA. Not only that, but in many virus their genome is not single stranded (like mRNA) but double stranded with two RNA strands base paired to each other (just like DNA, except for an extra oxygen on the ribose sugars in the backbone).

Nucleated cells don’t contain much double stranded RNA (dsRNA) outside the nucleus, so it almost always means trouble. An extremely elegant mechanism exists to find and respond to such RNA. Recall that double helix molecules can reach enormous lengths.The 3.2 billion base pairs of our genome, if stretched out, would be more than a yard.

Well we have at least 4 genes which bind dsRNA and then signal trouble. They all make a molecule called 2′ – 5′ oligoadenyic acid (2-5A) from ATP, so they are called OligoAdenylate Syntheses (OASs). The 2-5A, once made wanders about the cell until it finds another enzyme called RNAase L. 2-5A binds to RNAase L causing it to dimerize and become active. RNAase L then destroys all the RNA in the cell, killing it along with the invading virus. Pretty harsh, but it’s one way to stop the virus from spreading and killing more cells.

A recent paper concerns OAS3, which has 3 catalytic modules rather than just one like most enzymes. Even worse, 2 of the 3 catalytic modules can’t make 2-5A (but they still can bind dsRNA). OAS3 is a large protein (over 1,000 amino acids), so it has some length to it. The 3 catalytic modules are spread out along OAS3 with the active catalytic module at one end and one of the inactive modules at the other.

The modules at both ends bind dsRNA, but only the active module makes 2-5A when it does. Interestingly, the inactive module binds dsRNA much more strongly than the active one.

OK, you’ve got the picture — what possible use is this rather Byzantine set up?

See if you can figure it out.

It’s incredibly clever and elegant, and shows the danger to regarding anything within the cell as functionless (or junk). Teleology rides supreme in molecular and cellular biology.

Give up?

OAS3 essentially acts as a molecular ruler making 2-5A only when long dsRNA (e.g. over 50 nucleotides long) binds to it. The inactive module gloms onto longish dsRNA, holding it tightly until till Brownian motion brings it to the other end of OAS3 activating the catalytic module to make 2-5A. This is good as the cell normally contains all sorts of shorter RNA duplexes (the binding of microRNAs to the 3′ end of mRNAs come to mind — but they are much shorter (22 nucleotides at most).

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.

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.

When the active form of a protein is intrinsically disordered

Back in the day, biochemists talked about the shape of a protein, influenced by the spectacular pictures produced by Xray crystallography. Now, of course, we know that a protein has multiple conformations in the cell. I still find it miraculous that the proteins making us up have only relatively few. For details see —

Presently, we also know that many proteins contain segments which are intrinsically disordered (e.g. no single shape).The pendulum has swung the other way — “estimations that contiguous regions longer than 50 amino acids ‘may be present” in ‘up to’ 50% of proteins coded in eukaryotic genomes [ Proc. Natl. Acad. Sci. vol. 102 pp. 17002 – 17007 ’05 ]

[ Science vol. 325 pp. 1635 – 1636 ’09 ] Compared to ordered regions, disordered regions of proteins have evolved rapidly, contain many short linear motifs that mediate protein/protein interactions, and have numerous phosphorylation sites compared to ordered regions. Disordered regions are enriched in serine and threonine residues, while ordered sequences are enriched in tyrosines — this highlights functional differences in the types of phosphorylation. Interestingly tyrosines have been lost during evolution.

What are unstructured protein segments good for? One theory is that the disordered segment can adopt different conformations to bind to different partners — this is the moonlighting effect. Then there is the fly casting mechanism — by being disordered (hence extended rather than compact) such proteins can flail about and find partners more easily.

Given what we know about enzyme function (and by inference protein function), it is logical to assume that the structured form of a protein which can be unstructured is the functional form.

Not so according to this recent example [ Nature vol. 519 pp. 106 – 109 ’15 ]. 4EBP2 is a protein involved in the control of protein synthesis. It binds to another protein also involved in synthesis (eIF4E) to suppress a form of translation of mRNA into protein (cap dependent translation if you must know). 4EBP2 is intrinsically disordered. When it binds to its target it undergoes a disorder to ordered transition. However eIF4E binding only occurs from the intrinsically disordered form.

Control of 4EBP2 activity is due, in part, to phosphorylation on multiple sites. This induces folding of amino acids #18 – #62 into a 4 stranded beta domain which sequesters the canonical YXXXLphi motif with which 4EBP2 binds eIF4E (Y stands for tyrosine, X for any amino acid, L for leucine and phi for any bulky hydrophobic amino acid). So here we have an inactive (e.g. nonbonding) form of a protein being the structured rather than the unstructured form. The unstructured form of 4EBP2 is therefore the physiologically active form of the protein.

Scary stuff

While you were in your mother’s womb, endogenous viruses were moving around the genome in your developing developing brain according to [ Neuron vol. 85 pp. 49 – 59 ’15 ].

The evidence is pretty good. For a while half our genome was called ‘junk’ by those who thought they had molecular biology pretty well figured out. For instance 17% of our 3.2 gigaBase DNA genome is made of LINE1 elements. These are ‘up to’ 6 kiloBases long. Most are defective in the sense that they stay where they are in the genome. However some are able to be transcribed into RNA, the RNA translated into proteins, among which is a reverse transcriptase (just like the AIDS virus) and an integrase. The reverse transcriptase makes a DNA copy of the RNA, and the integrates puts it back into the genome in a different place.

Most LINE1 DNA transcribed into RNA has a ‘tail’ of polyAdenine (polyA) tacked onto the 3′ end. The numbers of A’s tacked on isn’t coded in the genome, so it’s variable. This allows the active LINE1’s (under 1/1,000 of the total) to be recognized when they move to a new place in the genome.

It’s unbelievable how far we’ve come since the Human Genome Project which took over a decade and over a billion dollars to sequence a single human genome (still being completed by the way filling in gaps etc. etc [ Nature vol. 517 pp. 608 – 611 ’15 ] using a haploid human tumor called a hydatidiform mole ). The Neuron paper sequenced the DNA of 16 single neurons. They found LINE1 movement in 4

Once a LINE1 element has moved (something very improbable) it stays put, but all cells derived from it have the LINE1 element in the new position.

They found multiple lineages and sublineages of cells marked by different LINE1 retrotransposition events and subsequent mutation of polyA microsatellites within L1. One clone contained thousands of cells limited to the left middle frontal gyrus, while a second clone contained millions of cells distributed over the whole left hemisphere (did they do whole genome on millions of cells).

There is one fly in the ointment. All 16 neurons were from the same ‘neurologically normal’ individual.

Mosaicism is a term used to mean that different cells in a given individual have different genomes. This is certainly true in everyone’s immune system, but we’re talking brain here.

Is there other evidence for mosaicism in the brain? Yes. Here it is

[ Science vol. 345 pp. 1438 – 1439 ’14 ] 8/158 kids with brain malformations with no genetic cause (as found by previous techniques) had disease causing mutations in only a fraction of their cells (hopefully not brain cells produced by biopsy). Some mosaicism is obvious — the cafe au lait spots of McCune Albright syndrome for example. DNA sequencing takes the average of multiple reads (of the DNA from multiple cells?). Mutations foudn in only a few reads are interpreted as part of the machine’s inherent error rate. The trick was to use sequencing of candidate gene regions to a depth of 300 (rather than the usual 50 – 60).

It is possible that some genetically ‘normal’ parents who have abnormal kids are mosaics for the genetic abnormality.

[ Science vol. 342 pp. 564 – 565, 632 -637 ’13 ] Our genomes aren’t perfect. Each human genome contains 120 protein gene inactivating variants, with 20/120 being inactivated in both copies.

The blood of ‘many’ individuals becomes increasingly clonal with age, and the expanded clones often contain large deletions and duplications, a risk factor for cancer.

Some cases of hemimegalencephaly are due to somatic mutations in AKT3.

30% of skin fibroblasts ‘may’ have somatic copy number variations in their genomes.

The genomes of 110 individual neurons from the frontal cortex of 3 people were sequenced. 45/110 of the neurons had copy number variations (CNVs) — ranging in size from 3 megaBases to a whole chromosome. 15% of the neurons accounted for 73% of of the CNVs. However, 59% of neurons showed no CNVs, while 25% showed only 1 or 2.

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 75 other followers