Never underestimate the ability of scientific inquiry...

Today I'm going to blog about something that I actually have some expertise in, for once. This has been a blog a long time in coming, an idea that came from an especially bad argument someone defending the existence of the supernatural: Science can't measure everything therefore it is unreasonable to dismiss the existence of the supernatural. This is backwards logic. Since by definition one can have no knowledge of the supernatural, discussing the topic as if any knowledge about it can be had is irrelevant. One example this person gave me refuting the ability of science to measure everything was in the form of a question: "How can I know that I love my wife?"

There are several things wrong about this example, which underscore the problem with the argument itself. Indeed, Science can't measure everything - at the moment. The implication of the above statement is that if something like emotion can't be quantified by any current method available, then it will never be measureable. A bold statement indeed, especially in light of the second premise behind the above challenge which is itself incorrect. I can indeed demonstrate that I love my wife. In fact, I have access to the equipment needed for the experiment.
A magnetic resonance imaging (MRI) system is an expensive piece of lab equipment, especially the new generation of high-field systems, but man can they do some amazing stuff. I sat in on a General Electric luncheon in Toronto earlier this spring and was floored not just by the quality, but the applications of MRI. So, how does one go about using MRI to investigate what we call love? This is a special application of MRI called functional MRI, or fMRI for short. The ability to detect brain activation comes from a peculiar decoupling of neuron function and blood flow. What I mean by this is that when the inputs to a neuron from other neurons (delivered by connections between neurons called synapses) cause it to activate and transmit a current down the axon and be the input to the next neuron. This activation of a neuron results in membrane depolarization (ion gradients across membranes, which are kept high by ion transport proteins in the cell membrane, equalize, resulting in electrical transmission down an axon) an increase in metabolism, as it tries to recover the electrical potential across cell membranes. That is, glucose and oxygen usage rise sharply upwards. Local blood flow is sensitive to neuronal activation and increases to meet the increased demand for oxygen, but actually overshoots the requirements of tissue.

Oxygen is carried in red blood cells by hemoglobin. When the oxygen molecules are delivered, hemoglobin becomes deoxyhemoglobin. From a MRI perspective, this is an important change. Deoxyhemoglobin has unpaired electrons (supplied by the iron it contains) that, when oxygen is bound, are not normally there. These unpaired electrons cause very strong local fluctuating magnetic fields, which cause the signal from protons in water molecules (which is how we do MRI) to dephase. In essence, these local fluctuating fields cause the signal to dephase, resulting in a decreased signal that is acquired to produce the image.

How does this show us where brain activation is happening? Well, when tissue is running at a higher metabolic rate, more oxygen gets taken up from blood, increasing the local concentration of deoxyhemoglobin. This deoxyhemoglobin would cause a signal drop to to the dephasing phenomenon, but blood flow local to this increase in oxygen uptake increases to meet the added burden. However, since blood flow overshoots what is necessary, it actually clears out more deoxyhemoglobin relative to tissue that is not running at a higher metabolic rate. Thus, what is seen in the image is an increased signal intensity in those regions where the brain is activated. This is known as the blood-oxygen level dependent (BOLD) phenomenon. It's somewhat more complicated than this, but these are the main events which lead to our ability to do fMRI.

When I made my reply to the question "How can I know that I love my wife?", I wasn't up on the literature in the area. But the experiment was quite obvious to me, so I wasn't all that surprised when I was listening to a CBC radio program called "Between You and Me" where the host was discussing with Helen Fisher of Rutgers University some of the fMRI experiments she has done to study the emotion we call love. So I went into the literature earlier and dug out a couple of papers.

One of the trickiest things in science is designing experiments. The answer from an experiment may not be the answer to the question you were interested in asking, but answering another question altogether. For instance, in Blind Faith Richard Sloan describes the results of a study which seemed to show a significant health benefit related to the amount one attends church. The more often subjects went to church seemed to be healthier than those that did not. What the author of the study failed to account for was what is known as a confounder, which is essentially a word used to describe a monkey wrench thrown into the works. The simple fact of the matter is that healthier people are able to attend church more often than those that are not as healthy. Studies of the type Fisher is engaged in are no different and take very careful designing to remove counfounding variables.

What Fisher and others have shown is that people in a state of what is called romantic love (other states being attachment and lust) show brain activation when the name of their beloved partner was mentioned. The paradigm also included giving the subjects names of neutral friends or describing hobbies they were passionate about. Regions of the brain which are recruited are parts of dopaminergic systems (that is, the primary neurotransmitter in these regions is dopamine). Dopaminergic systems are typically involved in reward/motivation and include such regions as the right ventral tegmental area and right caudate nucleus. This suggests that dopaminergic reward pathways are important in the general arousal component of romantic love.

So, it's quite clear that the poser of the question we started with is in error not only in what we can measure now, but also more fundamentally in the process leading up to the erroneous conclusion. Claiming that Science can not know everything is no reason to believe in the supernatural (which by definition lies outside our experience), nor is there any reason to suggest that we can't at some point in the future measure all things within our experience and then some.

Is this reductionist? Absolutely. But those that use the word in derision are simply ignorant. No one is attempting to demean the emotion of love as a whole by trying to understand from whence it arises. And no understanding of the highest level, our experience of love, can come without understanding the next lower level. Indeed, we gain a lot of insight into other behavior. The same dopaminergic pathways are also involved in gambling addiction. The mind as a whole can not be understood without understanding how neurons work, but not a single neuroscientist will say that such emergent phenomena as the mind can be understood by simply looking at its most basic components. By analogy, we can not understand how a clock works without knowing how gears work. By the same token it is difficult to make a clock by simply looking at a gear. We need to understand each level of organization.

I want to say one more thing here about a new phenomonen popping up, a pseudoscience known as neurotheology. I have one word to describe it: nonsense. What those which push neurotheology are trying to suggest comes from a study on nuns in what they subjectively called "a state of union with God". There are those out there that make outrageous claims that this is some kind of "picture of God". Absolute rot. There is nothing in these images which can not be accounted for by a self-induced change in brain function. As Richard Sloan pointed out in a speech to the Freedom From Religion Foundation on the topic, you will see the brain "light up" while eating a piece of cheese. Does that mean that if you acquired images under such tasty circumstances that you are viewing a picture of Gouda? Hardly. What this is is confusing what we feel with what we believe is the source. However, we've known for a long time that the conscious mind can have a profound effect on brain function and the existence of god is utterly unnecessary in explaining what the fMRI data shows with these nuns.

References:

Fisher H, Aron A, Brown LL. Romantic love: An fMRI study of a neural mechanism for mate choice. J Comp Neurol 493:58-62 (2005)

Beauregard M, Paquette V. Neural correlates of a mystical experience in Carmelite nuns, Neurosci Lett 405:186-90 (2006)

Fisher HE, Aron A, Brown LL. Romantic love: a mammalian brain system for mate choice. Phil Trans R Soc B 361:2173-86 (2006)

Ortigue S, Bianchi-Demicheli F, de C Hamilton AF, Grafton ST. The neural basis of love as a subliminal prime: An event-related functional magnetic resonance imaging study. J Cogn Neurosci 19:1218-30 (2007)

In Search of the Protocell - The Work of Jack Szostak

The origin of life problem is perhaps the most important question to ever have been the focus of scientific scrutiny. The only other question that I think rates of similar importance is the origin of the universe. Both questions have special obstacles to overcome before any answers are within sight. 

The Miller-Urey experiment of the 50s was a lightningrod for research into this question, but the euphoria caused by the viewpoint that the answers were near quickly receded once the scope of the problem was realized and it was decades before the excitement was rekindled in the scientific world. It's no surprise that the search is a difficult one. Remember, researchers are trying to compress the millions of years that undoubtedly were required for nature to give life a kick start into the lifespan of humans. Couple this with only a limited knowledge of what the conditions were at the time life started except in the grossest terms with the possibility that trace elements may be essential for the synthesis of life greatly compounds the issue. Anyone thinking that if life arose through naturalistic processes means it should be both easy, and that a mere 50 years of research should have resulted in the creation of protolife really doesn't have a good grasp of the scope of the problem. 

The molecules of life in the prebiotic world were all over the place, and not just on this planet. We know this not only from the Miller-Urey experiment itself, though we now think were the conditions at the time were somewhat different (which does not change the conclusions drawn from that experiment, or from similar ones which simulated what we now think the conditions at the time were), but meteorites have been found with complex organic molecules which could have seeded a barren Earth with the raw materials for the synthesis of life. Amphiphilic molecules (molecules possessing both water-loving (hydrophilic) and water-hating (hydrophobic) regions) have been generated by a variety of means simulating conditions found naturally: ultraviolet radiation of ice particles in the vacuum of space and at hydrothermal vents. Such molecules would provide the first cell membranes. 

A plan for synthesizing life was put forward by Szostak in 2001. It is based on a heterotophic model (cell structure first) rather than on an autotrophic one (metabolism first). First, create a spontaneously-replicating membrane through which small molecules can diffuse but bar larger molecules synthesized from these precursors from escaping. Next, create a replicase - a molecule mediating polymerization of a second molecule - a template containing protogenetic information to be copied.  The template could be RNA complimentary in sequence to the replicase or an unfolded replicase. RNA molecules can be encapsulated in vesicles and the whole cell self-assemble. This compartmentation inevitably results in the replicase component being subject to variation and natural selection.

Under the right conditions, amphiphilic molecules in solution can form micelles, or vesicles. This is similar to what soap, another amphiphilic class of molecules, does. Soap molecules (in the correct range of concentrations) cling together to form balls with the water-loving heads facing outward. In the case of vesicles, the molecules stand tail-to-tail with their hydrophilic heads facing outward from both the inner and outer surfaces of the ball. 

These vesicles would provide microenvironments for retaining and protecting primitive oligonucleotides (short sequences of RNA or DNA, typically of less than 20 bases). It is unlikely that early cell membranes would be made up of the same types of molecules which make up those in modern cells: phospholipids. Membranes made up of phospholipids are far too efficient at keeping out negatively charged ribonucleotides. Modern cells have evolved specific transport proteins to take in nutrients, but the earliest cells would have had no such mechanism available to them. 

Rather, the earliest cells would have used less efficient amphiphilic molecules, such as fatty acids, through which small molecules like ribonucleotides (such as uridine monophosphate, which make up RNA) could pass accross by simple diffusion. One hypothesis for both vesicle formation and RNA synthesis is respectively the interaction of fatty acids and ribonucleotides with clays. There is a growing body of evidence that this is a viable mechanism by which both of these process could happen. The clay montmorillonite has long been known to be able to catalyze RNA from activated ribonucleotides, but it can also greatly increase the rate of formation of fatty acid vesicles. The clay has a positively charged surface which attracts and concentrates the negatively charged fatty acids and thus facilitates their formation. Fatty acid membranes are also permeable to magnesium, a divalent cation necessary in many biochemical reactions and itself increases membrane permeablilty to negatively charged ribonucleotides. 

The surprise is that vesicles created in the presence of montmorillonite will also incorporate clay particles! It was immediately obvious to Szostak that this provides not only a mechanism for vesicle formation, but a method of synthesizing RNA oligonucleotides from ribonucleotides which diffuse through the membrane. Oligonucleotides formed within the vesicle are unable to escape the interior and are trapped. (As an aside, it also provides an explanation as to why L- rather than D-amino acids are utilized in protein synthesis. D- and L-amino acids are non-superimposable versions of each other, rather like the mirror image of your hand is not superimposable on your physical hand. Amino acids synthesized in an isotropic medium would be an equal (racemic) mixture of both optical isomers. These optical isomers have exactly the same physical properties bar one - each rotates the plane of polarized light in opposite directions. However, catalysis by a surface breaks the symmetry and one optical isomer would be selected over the other. It just so happens that L-amino acids were the ones selected. For sugars like glucose, it is the D-optical isomer that is used in biochemical reactions.)

Not only will these vesicles form, they have been shown to be able to spontaneously grow and divide in a series of elegant experiments. It was found that if the high vesicle concentration decreased by slowly adding a dilute solution of fatty acids, the vesicles would actually grow rather than just form new micelles. Vesicle division can be accomplished by extruding them through a polycarbonate filter. This likely happens by elongating the micelles so that they are no longer spherical and resealing after being pinched-off. As confirmation of this, vesicles preloaded with fluorescent dye were run through a filter released the dye into the medium in amounts only slightly greater than what was predicted for this mechanism of division. Had complete membrane disruption and reformation of vesicles occurred, the entire contents of the micelles would have been dumped into the medium. Vesicle division thus strongly resembles cellular division via budding and their formation, growth and division require no complex machinery at all, only raw physical forces. This is consistent with our current hypotheses on how early cell membranes must have formed. It even supplies a means for the first genetic material to have been generated through ribonucleotide uptake and mineral-catalyzed oligonucleotide formation.

Now our good friend Darwin steps in. Vesicles under osmotic stress due to their encapsulated contents need to decrease osmotic pressure by increasing their volume (and hence their surface area) by capturing fatty acids. Either that, or explode, dumping their contents. They do this by stealing fatty acids from other vesicles. But this is not a random process. The encapsulated contents have something to say about how well a vesicle will relieve the stress. Thus, we have what may have been the first example of biological competition! In the paper which covers this research (Chen, 2004), however, the competition was purely for stealing fatty acids from isotonic micelles (that is, vesicles not under osmotic stress). In other words, they feed. Once a truly replicating protocell is synthesized, a goal not yet reached, natural selection will become paramount in importance. The replicase can easily mutate through random mutation (since there are no error correcting mechanisms yet) and those which replicate better than others, eat other vesicles more efficiently and, as a consequence, divide more often will become more prevalent. Sounds like evolution to me.

So, when vesicles divide, how can the genetic material split into two as well? This is a Holy Grail in abiogenesis research. Some RNA can act like enzymes (another tantalizing clue to the origin of bioactive molecules). Such RNA molecules are known as ribozymes. Hammerhead ribozymes, which can catalyze cleavage and ligation of RNA molecules, are thought to be important in an RNA world and allow a mechanism for self-replication in the presence of magnesium. Encapsulated hammerhead ribozymes perform this self-cleavage as well, a necessary first step in this line of study. Research continues in developing a truly self-replicating protocell, and the results to date are highly encouraging. Activated nucleotides permeating across amphiphilic membranes have been shown to non-enzymatically replicate - this is key - encapsulated DNA templates. It just remains to fill in the lines.

When all is said and done, is this going to show us how abiogenesis occurred? Maybe. Note the language that Szostak uses: "model protocell vesicles", "prebiotically plausible membrane", etc. It's very careful language. What these experiments and others give us is a possible pathway, not necessarily the pathway. Perhaps autotrophic and heterotrophic abiogenesis are not either/or propositions and both are possible but only one historically occurred. Unless someone invents a time machine that can take us back to that point in time (current theoretical designs can only take us - well, actually only particles, not us - back in time to the point at which the machine was turned on), it is unlikely that we will be at all confident in having found the pathway. But this is not the point. The point is to find a plausible mechanism whereby abiogenesis could have occurred naturally, and we are well on our way there.  

Sometimes the journey is more important than the destination.

References:

Szostak JW, Bartel DP, Luisi PL, Synthesizing Life, Nature 409387-390 (2001)

Hanczyc MM, Fujkiawa SM, Szostak JW, Experimental Models of Primitive Cellular Compartments: Encapsulation, Growth, and Division. Science 302:618-622 (2003)

Chen IA, Roberts RW, Szostak JW, The Emergence of Competition Between Model Protocells. Science 305:1474-1476 (2004)

Chen IA, Salehi-Ashtiani K, Szostak JW, RNA Catalysis in Model Protocell Vesicles, JACS 127:13213-13219 (2005)

Mansy SS, Schrum JP, Krishnamurthy M, Tobe S, Treco DA, Szostak JW, Template-directed Synthesis of a Genetic Polymer in a Model Protocell, Nature [Epub ahead of print] (2008)