Performing Search, as a Transcription Regulator

Billions of years have created some weird tricks in DNA search.

Search is all around us, as we increasingly rely on search engines to find everything we need on the internet, want to watch, or want to buy. Search looks into databases, which hold the sought-after information. All our accounts, all the domain names, all the products... everything is held in databases of one kind or another, and those databases are indexed in clever ways to provide virtually instant pointers from the question we ask to the answer held online. AI merely puts a linguistic gloss on this, and most people are still encountering AI first as a feature of search, such as the top of current Google search results.

Well, our genomes are databases as well- rich and ancient storehouses of jewels that encode the body and its doings. How does search work there, and what is search even for? At any moment, each cell of the body has certain needs, stresses, and goals, as expressed in its DNA programming. The tools available are proteins and RNAs, which carry out the cell's functions. The needs may arise from signals coming from previously expressed receptors, say, for insulin, which may trigger and tell the cell to take up glucose from the blood. The receptor turns on a kinase, which may turn on another kinase, which turns on a transcription regulator, which goes into the nucleus and ... does a search. This regulator is searching for places (specific sequences) in the genomic DNA where it can bind, after which it helps to turn on (or off) the nearby gene, executing the desired function / tool. 

General introduction to transcription regulators (or "factors") and their role in gene activation and the whole process of gene expression.

Obviously a very different kind of search than what Google does on our behalf across documents, but there are similarities. Internet search depends on patterns, matching the user's input with the vast corpus of the internet also held as text symbols. Transcription regulators match patterns, in this case patterns of DNA that they like to bind, which may occur only once in the genome, or occur tens of thousands of times. The pattern here is a complementary physical/electrochemical shape, rather than an abstract same-symbol match. The genome is, to a protein, truly vast. Our three billion-base genome is forty million times larger than an average regulatory protein of, say, fifty kilodaltons (kDa). Search is also, here, a difficult problem, which researchers have been wondering about for decades. Several recent papers discuss different aspects of the problem and shed some modern light on it.

We have roughly 1600 transcription regulators in our genomes, so there is something going on all the time. DNA is always being queried. And what it replies with is RNA- a transcript issued/copied from a gene, which either goes off to instruct creation of a protein, or is itself functional in some way. So, how do proteins bind to DNA, executing their search? It was transformative when the first atomic structures of such proteins were solved. They were clearly complementary with their DNA targets, with nicely positioned positive charges to mate with the backbone of the DNA and amino acid fingers reaching into the helix to feel the shapes of the nucleotides they wanted to bind. All very neat, and paradigmatic for bacteria whose genomes are quite small. But there is more to the story. Binding sites in human genes tend to be quite short- five to seven bases. That really isn't enough to be very specific, across a vast genome. Eukaryotes have developed several weird tricks, as it were, to encourage efficient search over much larger genomes and at the same time increase precision while maintaining evolvability and flexibility.

Eukaryotes have nucleosomes, chromatin, and packaging. The DNA is not just splayed out randomly, but wound up on protein spools. One would think that this would impair search by regulators. But paradoxically, there is a fine balance between hunting around on a given piece of DNA for a preferred site (one-dimensional search, 1D), and jumping off, letting go, and trying somewhere else (by diffusion; three-dimensional search, 3D). The compaction of genomic DNA into nucleosomes that wind up most of the DNA while leaving linking DNA in between free appears to provide a nice balance of landing spots that allow searching regulators to jump very long distances (in linear DNA terms) while not going very far in absolute terms. Regulators vary in how aggressively they can plow through nucleosomes to try out their internal DNA sites, but many (called pioneer factors) can do so.

Secondly, transcription regulators cooperate with other proteins to create longer, more complex DNA sites for precise gene identification and higher binding affinity. As biologists have characterized the enhancers and promoters of important developmental genes, they have found that DNA binding sites occur in bunches, and have much weaker effects when broken down and separated. Sometimes there is direct side-to-side cooperation between two regulators that bind the DNA. At other times, they combine with other non-DNA-binding proteins to create complexes at such sites. The DNA recognition sequences of these combinatorial sites can be changed significantly, even beyond (our) recognition, by the addition of cooperative proteins. This is something that makes prediction of where a given regulator binds particularly perilous. 

Thirdly, many regulators contain not only DNA binding domains, but also extra disordered domains that facilitate DNA search and binding. This has been a recent realization that accounts for some of the speed and flexibility of regulator search and DNA interaction in eukaryotes. The stable crystal structures of paradigmatic bacterial regulators are not the whole story, and indeed are insufficient to explain what is happening in the much larger setting of our own cells. The authors note that eighty percent of human gene regulators have large disordered domains, (called IDRs, for intrinsically disordered region), upwards of 500 amino acids long. These never showed up in crystal structures, naturally. Being disordered, they are also poorly conserved. So, they have been difficult to study. 

Comparison of binding by one regulator, MSN2, which has a large IDR, to its genomic sites. At top is its native binding pattern, across a whole genome. At bottom are mapped its core motif occurrences on that DNA. Second from top is the MSN2 protein mutated to contain only its core motif-binding domain, and third from top is the MSN2 protein mutated to remove that domain and retain everything else. Note how different the patterns of binding by each of these proteins are, though how each approximates to some degree the wild-type pattern.

In related work, researchers have divided up such proteins into the core binding site part and the IDR part. They find that both parts work partially, directing binding to some of the native sites around the genome. In fact, the IDR part does a more statistically accurate job than the core DNA binding motif. This is fascinating, showing that in eukaryotes, a new search mechanism arose, supplementing discrete and precise binding with a floppy / fuzzy code in the IDR and its binding sites. It turns out that regions of hundreds of bases around core target sites (which in one case amount to only the motif AGGGG) are preferentially bound by the respective IDR protein domain, with multiple weak interactions that remain structurally uncharacterized. In fact, neither the protein structures responsible, nor the DNA sites they bind are known yet, though deletion studies through IDR domains show that binding is distributed throughout.

Relationship between IDR binding site size, and the ratio of 1D vs 3D search time, by simulation. The bottom axis is size of the IDR binding region, the Y axis is time taken for search. Time spent in total (yellow) goes down to minimum at an optimum between 1D search that is slowed by longer IDR-binding regions, while 3D search is strongly accelerated by longer IDR-binding regions.

The combination of core binding and loosely unstructured binding in one regulatory / search protein provides powerful benefits. In dimensionality terms, if the effective landing site is expanded from five to five hundred bases, then the time required for 3D search through the space of the nucleus is dramatically shortened. Secondly, loose binding by the IDR then promotes an "octopus"-like 1D search along the local DNA, resulting in efficient settling on the core binding site to get ultimately precise positioning. The ultimate affinity of the regulator with the local DNA is also enhanced compared to what it could manage over a five base pair site. The researchers conclude that with these domains, the search problem is, in net terms, reduced by one dimension, from 3D to 2D. The surrounding areas of DNA that have marginal affinity for the IDR domain are called "antenna" regions, and the author's simulations show how they alter search behavior.

Schematic explanation of the current work, describing how IDR domains help to speed up the transition from 3D search through space, to 1D search across the DNA. And then also to facilitate 1D search by preventing full detachment from the DNA while the core binding motif continues to search by diffusion for its binding site (yellow).

For computers and databases, search is a huge problem that has led to technical innovation, as well as large drains on resources. Every search engine combs the internet, gobbling up all available information, creating indexes, and updating them constantly in order to give us the instant access we want. This infrastructure has been raised to a new level by AI, which transforms search into a new form, combining it with language translation and prediction methods that allow a search for corkscrew to bring back results for wine. Whether it understands anything is unlikely, but the desire to upgrade search from a simply determinative process to one that is more fuzzy and richly interpretive, and thus more useful, is not a new phenomenon.

• Rights? What rights?
• Goodbye to independent agencies.
• What are those crows saying?
• We have our sociopath, they have theirs.

Strides in Cancer Treatment

A new paper shows that CART therapies can be unleashed against solid tumors.

We are finally in the payoff period in the decades-long war on cancer. Slowly, painfully, precision approaches are being developed to treat specific molecular lesions in ways that are superior to the old blunderbuss kill-everything approaches. At first, these treatments had only marginal effects, at astounding costs. But increasingly, the effects are lengthening and cures are in sight in some forms of cancer. One unexpected area of revolutionary progress has been immunotherapies, which in various ways help our immune systems attack cancers. It turns out that many cancers have tricks to hide from the immune system, and once those tricky dampening molecules are circumvented, dramatic reductions are possible. One paper recently described an anticancer vaccine made up of a witch's brew of targeting molecules, cancer antigens, and adjuvants, that achieves strong anti-melanoma action.

Another one of these immunotherapies is CART, or chimeric antigen receptor T-cell therapy. T cells have a receptor repertoire, just as B-cells do, which target things to be attacked- foreign pathogens, diseased states, etc. at molecules called antigens. One problem in cancer is that the cells are, originally at least, our own, so they mostly evade immune detection by having few "foreign" antigens. But there are nevertheless some antigens, comprised of normal molecules that are out of place (such as DNA found outside the cell) and "neoantigens" that are proteins expressed from the mutations in cancer cells. Additionally, as mentioned above, cancer cells express additional molecules (PD-L1) that can dampen even the immune response that does get generated by these few cancer antigens. So, the chimeric part of CART is taking the patient's own T-cells and engineering some of them to express new anti-antigen receptors that are relevant to the patient's cancer. Perhaps there is a mutant fusion protein that the cancer depends on. Perhaps the cancer displays an unusual surface molecule. Perhaps the tables need to be turned and PD-L1 targeted. There are many possible targets. 

CART therapies have, to date, been mostly directed at blood tumors. Solid tumors have extra protection in their micro-environments, and have not been good targets, though they necessarily have blood supplies and thus exposure to systemic T-cells. A recent paper blows open this field by revealing a magic molecule that plays a very significant role in the structure of solid tumors- the urokinase receptor. The urokinase plasminogen activator receptor (uPAR) is heavily expressed on senescent cells and many solid tumors, but rarely expressed elsewhere. Indeed, its expression correlates with tumor aggressiveness. Plasminogen is a protease that is sort of a cleanup crew for the circulatory system and body generally. It breaks up blood clots, and digests follicle tissues allowing ovulation. It encourages wound healing and discourages fibrosis- the buildup of scar tissue. However, in the cancer setting, the same activity seems to encourage fibrosis in a sort of constant wound healing state. Reviews in this field are rather confused about the direction of action. But one thing is clear- uPAR has myriad signaling activities relevant to tissue repair and immune activation that are not all dependent on the uPA (plasminogen activator) and plasmin activation system. Indeed, it is expressed not just in cancers, but in many other fibrotic settings.

A wide array of proteins are assessed here for their expression in a cancer tissue sample. uPAR is in red at the upper left. The matrix on the right shows the correlation of expression in a wide variety of cell types and tissues, like cancer-associated fibroblasts (CAFs), monocytes/macrophages (Mo/Mac), and with the protein fibroblast activation protein alpha (FAP).

The authors sought to target CART cells against uPAR, principally as a targeting device, since this marks many solid tumors and correlates with metastasis and rapid cancer progression, in addition to inflammation and fibrosis. While only tested in mice, the results were remarkable. 

"CAR T cells targeting the D2-D3 domain of uPAR display broad antitumor activity in xenograft, syngeneic, and patient-derived models, including in adjuvant and combination settings, supporting the concept that targeting conserved malignant cell states can enable therapeutic strategies that transcend tumor type. ... our prior work shows that uPAR CAR T cells targeting senescent cells remodel fibrotic tissues, and, as shown herein, this remodeling is associated with CAR T cell infiltration and cytotoxic activity. Similarly, parallel work demonstrates that uPAR CAR T cells exhibit potent efficacy in glioblastoma models and can co-target supportive stromal cells."

This is to say that these CART cells target not only tumor cells, but the surrounding solid tissues (stromal cells) that they rely on. That is the key to defeating solid tumors. It also indicates that other autoimmune and fibrotic conditions may be addressable with this therapy as well. 

Treatment effects from the CART therapy in mice, against several tumors. The red graphs are controls, and the blue graphs are treatments. Top is the tumor volume over time, while at bottom is survival of the mice over time.  Lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), high-grade serous ovarian carcinoma (HGSOC), and pancreatic ductal adenocarcinoma (PDAC).

The results of treatment of xenografted human ovarian tumors into susceptible mice, at 3 weeks, bottom. On the left is the control, while the other two sets were treated with CART cells against uPAR.

The authors note that relapses were seen occasionally, but that in these cases, the uPAR target was still highly expressed. That suggests firstly that it is difficult for tumors of these targeted types to do without uPAR, and secondly that something else went wrong with the tailored CART therapy, other than that its target went away. Perhaps future work can enhance its penetration or activity. The researchers also strained to make their model systems as human-relevant as possible, using cancer tissue transplanted (xenografted) from human cell lines, human CART cells, and mice with transplanted immune systems from humans. This work is thus not only a scientific breakthrough of the highest order, but is a technical tour de force as well. It also ends up with a variety of patent declarations and commercial ties, indicating that this breakthrough is being fully milked by its inventors and commercialized at breakneck speed.

One major problem with this mode of therapy is that CART cells require a great deal of engineering. First, antibodies against uPAR were developed in mice or other species. Then the genes from those immune systems were recovered from those mice, to get the precisely recombined gene that expressed the antibody with highest binding activity against uPAR. Then that gene, hooked up to new transmembrane and intracellular domains, (specially selected to activate the T cell they will be put into), was introduced into a transformation vector and put into the T cells collected from the diseased mice. In humans, this treatment routinely runs a half a million dollars. It is incredibly ornate, and one expects that gene therapy will someday allow the patient's T cells to be directly modified in the body, without all the collection and laboratory work, (which takes months), given a high-quality gene encoding the antibody fragment that is generally applicable- not tailored to a specific patient.

• The example of Spain.
• AI does insurance... as you would expect!
• AI is not what it is cracked up to be. And way more expensive than it has to be.
• Japan is surprisingly willing to keep importing fossil fuels, despite exchange rate degradation.
• Renewables and batteries have stabilized California's grid and made electricity cheaper.
• Map of where electricity has gotten more expensive in the last year.
• How other animals deal with inequality.
• What economic warfare looks like.
• Can we survive the internet?
• Electric cars are good for everyone.

The History and Future of a Single Mutation

The CCR5delta32 confers resistance to HIV. Where did it come from?

We are edging into an age of precision medicine, where the causes of our maladies will be known in molecular detail, allowing treatments that address them at the root. Given the parlous state of medicine today, in the midst of financial breakdown and a continued mediocre level of basic diagnosis, it is hard to believe this is a corner we can turn. But vaccines have long been in this category, of addressing the precise pathogenic causes of disease, and oncology is fitfully getting there, given advances in DNA sequencing and in treatments based on specific mutations.

HIV is also a beneficiary of this approach, since the discovery of its pathogen led directly to a variety of effective (if not yet permanent) treatments. A researcher in China created gene-edited humans with a specific mutation that will render them resistant to HIV. The mutation he chose for this work is called CCR5delta32, and it does not naturally exist in Chinese populations. 

But it does exist in European populations, at a roughly 10% rate in single copy. When present in two copies, it provides complete immunity to HIV, while if present in one copy, it slows infection substantially. A recent paper rooted through the available ancient and present genomes to figure out where this mutation came from. 

CCR5 is a cell surface receptor for cytokines 3, 4, and 5. These are all pro-inflammatory cytokines, and they interact with multiple receptors. Here, as in so many other respects, the immune system is riven with redundancy, so that it can grapple with as many contingencies as possible. Cytokines are signaling molecules for the immune system, which is an unusual organ, being dispersed all over the body with numerous cell types all patrolling around, and communicating with each other by long- and short-range chemical messages. It turns out that the major form of HIV uses the CCR5 protein to get into our immune cells, explaining why CCR5delta32, which is totally non-functional, has such a dramatic effect on HIV susceptibility. 

While people carrying CCR5delta32 are generally fine, this defect does confer a variety of subtle changes to their susceptibility to other infectious diseases and cancers. That explains why this mutation has settled at its low level in the European populations, probably balancing the occasional benefit against a specifically CCR5-seeking pathogen against its natural functions that form the basis of its existence in the first place as a part of immune system that is conserved in all mammals. The Chinese gene-editing researcher came under withering criticism not only for breaching the generally agreed moratorium on human germline gene editing, but also because the net effect of this mutation is, on the whole, negative, raising risks of numerous diseases, despite its beneficial effect on HIV. 

The authors run several models and populations in an attempt to time the origin of the CCR5delta32 mutation, and portray its positive selection over the ensuing millenia.  CHG- Caucasus hunter-gatherer; EHG- Eastern hunter-gatherer; WHG- Western hunter-gatherer; ANA- Anatolian Neolithic ancestry. The bottom axis is time, and the Y axis is the frequency of the mutation in these populations. "Modern DAF" refers to the inclusion of the data set of current (not ancient) population frequencies, (top), which the authors claim leads to continued rates of selection (last 2,000 years) that are artifactual.

So where did it come from? The new authors gather up a large variety of population samples from around the world, and from ancient humans, back to about ten thousand years ago. They find the first instance of the mutation in one sample at 5.8 thousand years ago. After that, its frequency rises dramatically, up to about two thousand years ago, when it levels off. They conclude that this mutation originated about seven to nine thousand years ago, in the steppes of Eastern Europe / Western Asia, and was under strong positive selection at first, spreading to the current frequency of about 10% of the population / alleles. All occurrences on other continents can be accounted by the spread from this source.

Does this mean that HIV was prevalent long, long before the current pandemic? Hardly. The authors can not say anything about it, but one theory would be that some other disease had a similar profile. It certainly was not the Black Death, as the authors show that this mutation had no change in frequency over that gruesome pandemic. Another hypothesis is that general reduction in inflammatory response might be beneficial in some settings, as has been found for Covid-19, though here again, this mutation does not have any known positive or negative net effect on Covid-19 susceptibility or course. 

It is amazing that we have enough sequences of ancient DNA to be able to reconstruct this kind of thing- to be able to trace where and when some influential mutation occurred, and how it traveled and spread. It is a tour-de-force of bio-archeological reconstruction.

• When you escape reality, and morality.
• Some environmental benefits are flowing from the current war.
• We may be at peak oil, courtesy of the US.

Pumping Calcium

An ornate ion pump manages rapid outflow of calcium.

In the beginning, the egg cell experienced a wave of calcium release, triggered by union with a sperm cell. This blocked other sperm from entering, and prepared the egg to become a zygote and embark on embryogenesis. It is but one example of the pervasive role of calcium signaling among animals. Another is the muscle activation cycle, which relies on calcium release from the specialized sarcoplasmic reticulum (in response to a nerve activation) to get the cell as a whole contracting. Generally, calcium is kept very low in the cytoplasm, and high in the endoplasmic reticulum and outside the cell. Thus, channels gated by electrical activation or other signals can cause rapid cytoplasmic calcium spikes and signal widely within a cell. 

On the flip side, there have to be pumps that keep the cytoplasmic concentration low, and a recent paper elucidates the structure of one such pump that is remarkably fast, while also closely regulated. It is an impressive machine. PMCA2 is an ATP-using calcium pump that sits in the plasma membrane and carries out what is called the Post-Albers cycle. This is a flip-switch mechanism for pumping ions, where ATP drives conformational switches alternately exposing ion binding sites to each side of the membrane. When the pore is open to the cytoplasm, there is no competition from higher concentrations outside, so the active site can bind one internal calcium, given a high-affinity site. Then, after the conformational switch, the pore is exposed to the outside, and at the same time the site is reconfigured to be lower-affinity, releasing the calcium ion into a high concentration environment. Neurons especially use calcium signaling extensively to operate synapses and regulate growth and development. Their rapid and frequent signaling requires a pump that has especially high capacity. PMCA2 operates at a maximal rate of several thousands of Ca2+ ions pumped per second.

Cartoon of the Post-Albers cycle, which is shared by a large family of active ATP-using pumps that transfer ions against their chemical concentration gradient. M is the main transmembrane domain of the pump, where the ions traverse the membrane. The N, P, and A domains are regulatory, especially binding and cleaving ATP  at an interface between the N, P, and A domain. The cycle links power steps (1,2) with conformational changes that carefully gate the pumping process.

And that is not all. Since calcium has a charge of 2+ and this pump does not intend to alter charge across the membrane, the pump simultaneously has binding sites for counter-ions (generally two OH-) that are transferred in the opposite direction from the calcium. Not only that, but every pump of this kind requires regulation of various kinds. PMAC2 is activated by phosphatidyl inositol 4,5 bisphosphate (PIP2), which is another important signaling molecule generated by specific PI kinases in response to activation of G-protein coupled receptors or protein kinase C, which may respond to external signals. In very general terms, these tend to be pro-growth or stress-induced pathways. These regulatory processes can tune the overall rate of recovery from rapid Ca2+ signaling events, by adjusting the level and activity of pumps like PMAC2. 

ATP binds at the N/P/A domain interface, and its hydrolysis (and loss of ADP) generates extensive shape changes, including into the transmembrane M domain. At the very bottom, the calcium ion is shown in green, bound inside the M domain pumping channel. The motions here are subtle, but enough to dramatically reshape the calcium channel.

The authors, using various substrate variants and other tricks, were able to develop structures of PMAC2 in several steps of the pumping cycle, using cryo-electron microscopy. The ATPase site in the N domain (red) is far from the channel that conducts the calcium ion (brown, far bottom). They show extensive shape changes from binding or losing the ATP molecule, though they mostly concern the intracellular domains (red, blue, yellow). The effects on the transmembrane pore domain are rather subtle, shown on right. The authors claim that, compared to other pumps of this large family, the structural changes are significantly less, suggesting that evolution for speed has caused the mechanism to become more efficient, with less wasted motion per conduction event, at least in the channel region itself.

Relation of the PIP2 binding domain (orange/red stick figures) to the calcium core binding site. PIP2 appears to be essential for rapid pump operation. At bottom is shown some schematics of the gating provided by PIP2 in bound and unbound states, especially via the D873 side chain (negatively charged aspartic acid).

They also find that the activating molecule PIP2 is neatly parked right next to the main calcium binding and conduction region, and is more or less essential for enzyme activity. In the graph above (e), they show that several single mutations made in the calcium binding high affinity site, for example switching the negatively charged D873 for the positively charged K (lysine), kills ion pumping activity. Mutation of the PIP2 binding pocket (KKQ->TLL, around position 347) likewise kills enzyme activity.

Relation of the counter-ion channel (red dots) with the calcium channel. Both are essential parts of the mechanism. Closeups with the coordinating protein side chains shown on the right.

The whole mechanism is alluded to in the last figure, where the central calcium binding site is shown, with the general direction of calcium pumping. The counter-ion transport area is shown nearby as a flurry of red dots (standing for water molecules, which at this scale are interchangeable with OH ions). Specific single mutations in either area, either changing negatively charged E412 to positively charged lysine at the calcium binding pocket, or changing polar S877 in the water/hydroxy binding area to the bulky and hydrophobic F (phenylalanine), each kill pumping activity (graph). 

While it would be ideal to have a more dynamic representation of what is going on, the new structures give tremendous detail, including the associated ATP, PIP2, calcium, and water molecules. The mutations also nail down several functional points. Obviously a rather intricate and well-oiled machine that keeps its bit of cellular calcium homeostasis on an even keel. It is hard to believe that the sum of thousands of machines like this one is life, but the deeper we look the more true that appears to be.

• Gullibility-maxxing.
• There is a serious bifurcation in Catholic thought.
• Reality has a well-known liberal bias.
• Well, that didn't work out so great.
• Is the internet going to collapse?

Not Every Transcript is Golden

 Reflections on junk DNA, and junk transcripts.

Some time ago, a large project in molecular biology determined that most regions of the genome are transcribed. The authors and most observers took this to mean that most regions are functional, quite in contrast to the reigning theory up to that point, that our genomes host a smattering of genes floating in a sea of "junk" DNA. That theory was based on the now-ancient observations of reannealing curves for bulk DNA from humans and other species which found that most of our DNA re-anneals very quickly, due to the fact that it is repetitive. Most of our genomes (60%) are taken up with LINE repeats, SINE repeats, old retro-transposons, stray duplications, and other repetitive material that, at a first glance, seems like junk. There has been a battle ever since, between proponents of junk DNA and those who see function around every corner. As we learn more about the genome, many more functions have indeed come to light, like distant enhancers and regulatory RNAs of many flavors. But overall, there still seems to be a lot of junk. 

A recent paper took an oblique shot at this field, looking at the profusion of alternative gene transcripts, which can number into the hundreds for a single gene. (This was also reviewed.) These are generally called isoforms, and arise due to variable ways one gene's RNA products can be initiated, terminated, and spliced. So not only are most regions of the genome transcribed in some form, actively transcribed regions can be transcribed and processed in myriad ways to lead to different RNA products. Here again, there has been an analogous argument, about whether every such isoform has a function, or whether isoforms might arise from more or less sporadic processes, often as unintended and non-functional sparks coming out of the machinery. The importance of isoforms is very well documented in many cases, so the possibility of function, sometimes highly conserved, is not in question. Only the importance of every last variation in combinatorial collections of isoforms that can number into the hundreds.

Here is an image from the first page (of about six pages) of RNA transcripts coming off the notorious BRCA1 gene, which is intensely studied for its role in breast cancer. Each line is a distinct mRNA transcript. Each darker bar is an exon, which are separated by introns. The darker colored exons are in the protein coding region, while the lighter exons signify the untranslated upstream and downstream ends. I count about 315 transcripts described for this genetic locus. The idea that each of these has some evolutionarily constrained and important function is, on the face of it, absurd.

The authors took an interesting evolutionary approach, reasoning that species with larger population sizes experience more stringent purifying selection, and thus should (in theory) show tighter control over stray genomic products such as isoforms, if most transcript isoforms are neutral (or even deleterious) accidents, rather than intentional and functional forms. Thankfully, animals come in a wide range of population sizes, from insects to crocodiles and primates; very large to very small. While population size is hard to calculate, several convenient proxies are known, like lifespan, body size, etc. When they totted everything up, they saw clear correlations between these proxies and the number of alternative RNA products per gene- also termed transcript diversity. They sliced up the data by organ where the RNA was expressed, and by the source of the RNA variation- either different initiation, different termination, different splicing. In all cases the trend was the same. In species with larger population sizes, the diversity of transcripts was lower, agreeing with their hypothesis that when greater selecive force is available, the slop from the transcription and transcript processing machinery declines.

The authors draw correlations between alternative splicing (AI) diversity in an organism's cells and its population size. 

The authors additionally note that there is a similar relationship between alternative splice site usage and expression level of a gene. That is, the higher the gene expression, the less likely that minor splice sites are used, indicating that here again, higher selective pressure helps to clear out non-functional off-products of the transcription apparatus.

The correlations found here are only that- correlations. While significant, they are not terribly strong, let alone stark. So it is evident that our gene expression machinery has a lot of play in it, and this falls on a spectrum from deleterious to critically functional. It is, after all, machinery, not divine. It is also grist for evolution itself- it is useful to have some slop so that there is always some diversity in the gearing to accommodate new selective pressures. But the idea that just because a distinct transcript exists, it is biologically functional, or that, similarly, because a genomic region is transcribed, it is a "gene" rather than junk DNA.. that does not hold water. Every nucleotide in the genome has its own unique selective constraints, and for many of them, that constraint is zero.

• The world order, and our position in it, is crumbling.
• Whence Hungary?
• Another AI tax, as if gobbling up power wasn't bad enough.
• Mindless.

How 5S rRNA Gets Into the Ribosome

For a minor component, it gets a lot of molecular love.

As mentioned several times in this space, the ribosome, which synthesizes proteins according to mRNA instructions, is an extremely ancient and complicated machine. Its core, including the catalytic site, is RNA. This marks it as a hold-over from the RNA world, as the thing that made proteins, (probably tiny proteins at first), before proteins had become a thing. But boy has there been a lot of duct-taping since then. In humans, there are four ribosomal RNAs, eighty proteins pasted on the outside, and hundreds of other proteins or RNAs involved in assembling the ribosome, not to mention dozens of initiation factors and other regulators that help during translation.

A recent paper discussed the maturation of 5S ribosomal RNA, which is the smallest rRNA, and one whose function is more peripheral than the large central 16S and 23S rRNAs. It is present in all life forms, though ribosomes inside mitochondria do without it. Its processing is an interesting case study of the complexity that has accumulated over the eons. Exactly what the 5S rRNA does remains a bit unclear, though it clearly contributes to the dynamics of the large ribosomal subunit, and occupies the "central protuberance". One group ligated it into the large subunit 23S rRNA, showing that translation still worked quite well with the 5S portion stably tacked into the structure. But then they also found that these ribosomes fell with high frequency into an unproductive locked state, suggesting that the independent nature of the 5S rRNA plays an important role in the dynamics of the ribosome. 

At any rate, the assembly of 5S into the rest of the structure is a story in itself. There are multiple steps involved, some involving ATP-using helicases. As it comes off the gene, 5S rRNA is bound by two proteins- the TFIIIA regulatory factor that activates its transcription, and also La protein (aka La antigen), which is a storage protein, named after systemic lupus, for which it is one immune target. To be incorportated into ribosomes, the RNA is next bound by a complex of Rpl5 and Rpl11, which will remain with the 5S RNA and become part of the eventual ribosome. Next come Rpf2 and Rrs1, which are two assembly facilitators that bind as a complex. Then comes Rsa4, which is similarly an assembly protein that helps the whole mess bind to the proper place on the (immature) large ribosomal subunit. Lastly, Rea1 (called MDN1 in humans) is an ATP-driven RNA helicase that wrenches the whole 5S-containing protuberance into its final and quite different position. 

The authors provide a scheme for the stepwise processing and assembly of 5S rRNA into the ribosome, involving numerous assembly factors, ribosomal proteins, and a helicase. 

It is quite an amazing story of progressive assembly, all to attach an element of the ribosome that is hardly central, but is rather a relatively late accretion on the machinery. Nevertheless, it evidently deserves specialized attention for correct placement. 

A less schematic view of various steps heading toward ribosomal assembly. 5S rRNA is in teal, and the helicase Rea1 is in dark gold, mounted like a wrench at the top of the (late) structure.

• We are strangling Ukraine. Why?
• Building more housing reduces housing shortages.

We Have Rocks in Our Heads ... And Everywhere Else, Too

On the evolution and role of iron-sulfur complexes.

Some of the more persuasive ideas about the origin of life have it beginning in the rocks of hydrothermal vents. Here was a place with plenty of energy, interesting chemistry, and proto-cellular structures available to host it. Some kind of metabolism would by this theory have come first, followed by other critical elements like membranes and RNA coding/catalysis. This early earth lacked oxygen, so iron was easily available, not prone to oxidation as now. Thus life at this early time used many minerals in its metabolic processes, as they were available for free. Now, on today's earth, they are not so free, and we have complex processes to acquire and manage them. One of the major minerals we use is the iron-sulfur complex, (similar to pyrite), which comes in a variety of forms and is used by innumerable enzymes in our cells. 

The more common iron-sulfur complexes, with sulfur in yellow, iron in orange.

The principle virtue of the iron-sulfur complex is its redox flexibility. With the relatively electronically "soft" sulfur, iron forms semi-covalent-style bonds, while being able to absorb or give up an electron safely, without destroying nearby chemicals as iron alone typically does. Depending on the structure and liganding, the voltage potential of such complexes can be tuned all over the (reduction potential) map, from -600 to +400 mV. Many other cofactors and metals are used in redox reactions, but iron-sulfur is the most common by far.

Reduction potentials (ability to take up an electron, given an electrical push) of various iron-sulfur complexes.

Researchers had assumed that, given the abundance of these elements, iron-sulfur complexes were essentially freely acquired until the great oxidation event, about two to three billion years ago, when free oxygen started rising and free iron (and sulfur) disappeared, salted away into vast geological deposits. Life faced a dilemma- how to reliably construct minerals that were now getting scarce. The most common solution was a three enzyme system in mitochondria that 1) strips a sulfur from the amino acid cysteine, a convenient source inside cells, 2) scaffolds the construction of the iron-sulfur complex, with iron coming from carrier proteins such as frataxin, and 3) employs several carrier proteins to transfer the resulting complexes to enzymes that need them. 

But a recent paper described work that alters this story, finding archaeal microbes that live anaerobically and make do with only the second of these enzymes. A deep phylogenetic analysis shows that the (#2) assembly/scaffold enzymes are the core of this process, and have existed since the last common ancestor of all life. So they are incredibly ancient, and it turns out to that iron-sulfur complexes can not just be gobbled up from the environment, at least not by any reasonably advanced life form. Rather, these complexes need to be built and managed under the care of an enzyme.

The presented structures of the dimer of SmsB (orange) and SmsC (blue) that dimerize again to make up a full iron-sulfur scaffolding and production enzyme in the archaean Methanocaldococcus jannaschii. Note the reaction scheme where ATP comes in and evicts the iron-sulfur cluster. On right is shown how ATP fits into the structure, and how it nudges the iron-sulfur binding area (blue vs green tracing).

A recent paper from this group extended their analysis to the structure of the assembly/scaffold enzyme. They find that, though it is a symmetrical dimer of a complex of two proteins, it only deals with one iron-sulfur complex at at time. It also binds and cleaves ATP. But ATP seems to have more of an inhibitory role than one that stimulates assembly directly. The authors suggest that high levels of ATP signal that less iron-sulfur complex is needed to sustain the core electron transport chains of metabolism, making this ATP inhibition an allosteric feedback control mechanism in these archaeal cells. I might add, however, that ATP binding may well also have a role in extricating the assembled iron-sulfur cluster from the enzyme, as that complex is quite well coordinated, and could use a push to pop out into the waiting arms of target enzymes.

"These ancestral systems were kept in archaea whereas they went through stepwise complexification in bacteria to incorporate additional functions for higher Fe-S cluster synthesis efficiency leading to SUF, ISC and NIF." - That is, the three-component systems present in eukaryotes, which come in three types.

In the author's structure, the iron-sulfur complex, liganded by three cysteines within the SmsC protein. But note how, facing the viewer, the complex is quite exposed, ready to be taken up by some other enzyme that has a nice empty spot for it.

Additionally, these archaea, with this simple one-step iron cluster formation pathway, get their sulfur not from cysteine, but from ambient elemental sulfur. Which is possible, as they live only in anaerobic environments, such as deep sea hydrothermal vents. So they represent a primitive condition for the whole system as may have occurred in the last common ancestor of all life. This ancestor is located at the split between bacteria and archaea, so was a fully fledged and advanced cell, far beyond the earlier glimmers of abiogenesis, the iron sulfur world, and the RNA world.

• Triumph of stupidity and corruption.
• Oil fields are a mess.
• Homes are going nuts.
• But they should be getting closer.
• Gosh, there was a pedophilia ring after all.
• Who wants a concentration camp?

How do Anesthetics Work?

Ask a simple question, and get an answer that gets weirder the deeper you dig.

Anesthesia is wonderful. Quite simply, it makes bearable, even unnoticeable, what would be impossible or excruciating. It is also one of the most mysterious phenomena short of consciousness itself. All animals can be anesthetized, from bacteria on up. All sorts of chemicals can be used, from xenon to simple ethers, to complex fluoride-substituted forever chemicals, all with similar effects. Yet there are also complex sub-branches of anesthesia, like pain relief, muscular immobilization, shutdown of consciousness, and amnesia against remembering what happened, that chemicals affect differentially. It resembles sleep, and shares some circuitry with that process, but is of course is induced quite differently.

The first red herring was the Meyer–Overton rule, established back in 1899, that showed that anesthetic potency correlates closely with the lipophilicity of the chemical, from nitrogen (not very good) to xenon (pretty good) to chloroform (very good). All the forever (heavily fluorinated) chemicals used as modern anesthetics, like isoflurane and sevoflurane, have extremely high lipophilicity. This suggested that the mechanism of action was simply mixing into membranes somehow, altering their structure, and thus neuronal action.. something along that line. 

Structures of several general anesthetics. 8 is isoflurane.

But when researchers looked more closely there were some chemical differences that did not track with this hypothesis. Chiral enantiomers behaved differently in some cases, indicating that these chemicals do bind to something (that is, a protein) specifically. Also, variant genes started cropping up that conferred resistance to anesthesia or were found to bind particular anesthetics at working concentrations. Also, more complex, injectable anesthetics like fentanyl and propofol have slightly more defined targets and modes of action. So while anesthetics clearly partition to membranes, and the binding sites are often at protein-membrane interfaces, the modern theory of how they work is that they bind to ion channels and neurotransmitter receptors and affect their functions. Proteins generally have hydrophobic interiors, so the lipophilicity of these chemicals may track with binding / disrupting protein interiors as much as membrane interiors. And other proteins such as microtubules have been drawn into the discussion as well (leading indirectly to some very unfortunate theories about consciousness). 

But which key protein do they bind? Here again, mysteries abound, as they do not bind just one, but many. And not just that, they turn some of their targets on, others off. One target is the GABA receptor, which characterizes the major inhibitory neurons of the central nervous system. These are turned on. At high concentrations, anesthetics can even turn these receptors on without any GABA neurotransmitter present. Another is the NMDA receptor, which is the target of opioids, and of ketamine. These receptors are turned off. So, for some reason, still somewhat obscure, the net result of many specific bindings to an array of channels by an array of chemicals results in ... anesthesia.

A recent paper raised my interest in this area, as its authors demonstrated yet another target for inhaled anesthetics like isoflurane, and dove with exquisite detail into its mechanism. They were working on the ryanodine receptor, which isn't even a cell surface protein, but sits in the endoplasmic reticulum (or sarcoplasmic reticulum in muscles) and conducts calcium out of these organelles. This receptor is huge- the largest known- coding over five thousand amino acids (RYR1 of humans), due to numerous built-in regulatory structures. For example, it is sensitive to caffeine, but in a different location than where it is sensitive to isoflurane. Calcium is a very important signal within cells, key to muscle activity, and also to neuronal activation. The endoplasmic reticulum serves as a storehouse of calcium, from which signals can be sparked as needed by outside signals, including a spike in calcium itself (thus creating a positive feedback loop). These receptors (a family of three in human) are named for an obscure chemical (indeed a poison) that activates these channels, and all three are expressed in the brain. 

The authors were led to this receptor because mutations were known to cause malignant hyperthermia, a side effect of a few of the common anesthesia drugs where body temperature rises uncontrollably, driven from muscle tissue, where ryanodine receptors in the sarcoplasmic reticulum are particularly common and heavily used to regulate muscle activity and metabolism. That suggested that anesthetics such as isoflurane might bind to this receptor directly, turning it on. That was indeed the case. They started with cultured cells expressing each receptor family member in turn, and tested each receptor's response to isoflurane. Internal (cytoplasmic) calcium rose especially with the family member RYR1. That led to various control experiments and a hunt (by mutating and doctoring the RYR1 protein) for the particular region being bound by the anesthetic. After a lengthy search, they found residue 4000 was a critical one, as a mutation from methionine to phenylalanine reduced the isoflurane response about ten-fold. This is part of a binding pocket as shown below.

Structure of isoflurane, (B), bound to the RYR1 protein pocket. This is a pocket that happens to also bind another activator of this channel, 4-CMC. A layout of the whole active binding pocket is given on the right. At bottom are calcium channel responses of the wild-type and point mutant forms of RYR1, showing the dramatic effect these single site mutations have on isoflurane response.

Fine, but what about anesthesia? The next step was to test this mutation in whole mice, where, lo and behold, isoflurane anesthesia of otherwise normal mice was made slightly more difficult by this mutant form of RYR1. Additionally, these mice had no other observable problems- not in behavior, not in sleep. That is remarkable as a finding about anesthesia, but the effect was quite small- about 10% or so shift in the needed concentration of isoflurane. They go on to mention that this is similar in scale to knockouts or mutations in other known targets of anesthetic drugs:

• 10% shift in the curve from the M4000F mutation of RYR1
• 14% shift in the isoflurane curve from a mutation in GABA receptor, GABAAR.
• 5% shift in the isoflurane curve (though a 20% or more shift for halothane) for mutations in KCNK9, a potassium channel.

What this is telling us is that there are many targets for anesthetic drugs. They are spread over many neurotransmitter and physiological systems. They each contribute modestly (and variably, depending on the drug) to the net effect of any one drug. The various affected channels and membrane receptors curiously combine to achieve anesthesia across all animals and even microorganisms, which naturally also rely on channels and transmembrane receptors for their various sensing and motion activation needs. We are left with a blended hypothesis where yes, there are specific protein targets for each anesthetic that mediate their action. On the other hand, these targets are far from unique, spread across many proteins, yet are also highly conserved, looking almost like they are implicit in the nature of transmembrane proteins in general. 

One gets the distinct impression that there should be endogenous equivalents, as there are for opioids and cannabinoids- some internal molecule that provides sedation when needed, such as for deep illness or end-of-life crisis. That molecule has not yet been found, but the natural world abounds in sedatives, (alcohol is certainly one), so the logic of anesthesia becomes one of biological and evolutionary logic, as much as one of chemical mechanism.

• Jack Smith
• What is vice signaling?
• A do-nothing Friedman Fed, or a Trumpy Fed? It is hard to know which is worse. Or which we will get.

Jonathan Singer and the Cranky Book

An eminent scientist at the end of his career writes out his thoughts and preoccupations.

Jonathan Singer was a famous scientist at my graduate school. I did not interact with him, but he played a role in attracting me to the program, as I was interested in biological membranes at the time. Singer himself studied with Linus Pauling, and they were the first to identify a human mutation in a specific gene as a cause for a specific disease- sickle cell disease. After further notable work in electron microscopy, he reached a career triumph by developing, in 1972, the fluid mosaic model of biological membranes. This revolutionized and clarified the field, showing that cells are bounded by something incredibly simple- a bilayer of phospholipids that naturally order themselves into a remarkably stable sheet, (a bubble, one might say), all organized by their charged headgroups and hydrophobic fatty tails. This model also showed that proteins would be swimming around freely in this membrane, and could be integrated in various ways, ether lightly attached on one side, or spanning it completely, thereby enabling complex channel and transporter functions. The model implied the typical length of a protein alpha helix that, by virtue of its hydrophobic side chains, would naturally be able to do this spanning function- a prediction that was spot-on. He could have easily won a Nobel for this work.

I was intrigued when I learned recently that Singer had written a book near the end of his career. It is just the kind of thing that a retired professor loves to do in the sunset of his career, sharing the wisdom and staving off the darkness by taking a stab at the book biz. And Singer's is a classic of the form- highly personal, a bit stilted, and ultimately meandering. I will review some of its high points, and then take a stab of my own at knitting together some of the interesting themes he grapples with.

For at base, Singer turns out to be a spiritual compadre of this blog. He claims to be a rationalist, in a world where, as he has it, no more than 9% of people are rational. Definition? It is the poll question of whether one believes that god created man, rather than the other way around. Singer recognizes that the world around him is crazy, and that the communities he has been a part of have been precious oases amid the general indifference and grasping of the world. But changing it? He is rather fatalistic about that, recognizing that reason is up against overwhelming forces.

His specific themes cover a great deal of biology, and then some more mystical reflections on balance and diversity in biology, and later, in capitalism and politics. He points out that the nature/nurture debate has been settled by twin studies. Nature, which is to say, genetics, is the dominant influence on human characteristics, including a wide variety of psychological traits, including intelligence. Environment and nurture is critical for reaching one's highest potential, and for using it in socially constructive ways, but the limits of that potential are largely set by one's genes. Singer does not, however, draw the inevitable conclusion from these observations, which is that some kind of long-term eugenic approach would be beneficial to our collective future, assuming machines do not replace us forthwith. Biologists know that very small selective coefficients can have big effects, so nothing drastic is needed. But what criteria to use- that is the sticky part. Just as success in the capitalist system hardly signals high moral or personal qualities, nor does incarceration by the justice system always show low ones. It is virtually an insoluble problem, so we muddle along, destined probably for continued cycles of Spenglerian civilizational collapse.

Turning to social affairs, Singer settles on "structural chaos" as his description of how the scientific enterprise works, and how capitalism at large works. With a great deal of waste, and misdirected effort, it nevertheless ends up providing good results- better than those that top-down direction can provide. He seems a sigh a little that "scientific" methods of social organization, such as those in Soviet Russia, were so ineffective, and that the best we can do is to muddle along with the spontaneous entrepreneurship and occasional flashes of innovation that push the process along. Not to mention the "monstrous vulgarity" of advertising, etc. Likewise, democracy is a mess, with most people totally incapable of making the reasoned decisions needed to maintain it. Again, the chaos of democracy is sadly the best we can do, and the duty of rational people, in Singer's view, is to keep alive the flame of intellectual freedom while outside pressures constantly threaten.

Art, and science.

What can we do with this? I think that the unifying thread that Singer was groping for was competition. One can frame competition as a universal principle that shapes the physical, biological, and social worlds. Put two children on a teeter-totter, and you can see how physical forces (e.g. gravitation) compete all the time, subtly producing equilibria that characterize the universe. Chemical equilibria are likewise a product of constant competition, even including the perpetual jostling of phospholipids to find their lowest energy configuration amidst the biological membrane bilayer, which has the side-effect of creating such a stable, yet highly flexible, structure. With Darwin, competition reaches its apotheosis- the endless proliferation, diversification, and selection of organisms. Singer marvels at the fragility of individual life, at the same time that life writ large is so incredibly durable and prolific. Well, the mechanism behind that is competition. And naturally, economics of any free kind, including capitalism and grant-making in science, are based on competition as well- the natural principle that selects which products are useful, which employees are productive, and which technologies are helpful. Waste is part of the process, as diversity amidst excess production is the essential ingredient for subsequent selection. 

And yet.. something is missing. The earth's biosphere would still be a mere bacterial soup if competition were the only principle at work. Bacteria (and their viruses) are the most streamlined competition machines- battlebots of the living world. It took cooperation between a bacterial cell and an archaeal cell to make a revolutionary new entity- the eukaryotic cell. It then took some more cooperation for eukaryotic cells to band together into bodies, making plants and animals. And among animals, cooperation in modest amounts provides for reproduction, family structure, flock structures, and even complex insect societies. It is with humans that cooperation and competition reach their most complex heights, for we are able to regulate ourselves, rationally. We make rules. 

Without rules, human society is anarchic mayhem- a trumpian, dystopian and corrupt nightmare. With them, it (ideally) balances competition with cooperation to harness the benefits of each. Our devotion to sports can be seen as a form of rule worship, and explicit management of the competitive landscape. Can there be too many rules? Absolutely, there are dangers on both sides. Take China as an example. In the last half-century, it revamped its system of rules to lower the instability of political competition, harness the power of economic competition, and completely transform its society. 

The most characteristic and powerful human institution may be the legislature, which is our ongoing effort to make rational rules regulating how the incredibly powerful motive force of competition shapes our lives. Our rules, in the US, were authored, at the outset, by the founders, who were- drumroll please- rationalists. To read the Federalist Papers is to see exquisite reasoning drawing on wide historical precedent, and particularly on the inspirations of the rationalist enlightenment, to formulate a new set of rules mediating between cooperation and competition. Not only were they more fair than the old rules, but they were designed for perpetual improvement and adjustment. The founding was, at base, a rationlist moment, when characters like Franklin, Hamilton, Madison, and Jefferson- deists at best and rationalists through and through, led the new country into a hopeful, constitutional future. At the current moment, two hundred and fifty years on, as our institutions are being wantonly destroyed and anything resembling reason, civility, and truth is under particularly vengeful attack, we should appreciate and own that heritage, which informs a true patriotism against the forces of darkness.

• Bill Mitchell on AI.
• How this drives politics.
• And can thought survive against an onslaught of slop?
• Paul Krugman on abasement.
• Is evil lack of religion, or too much religion?