Dr Clare Sansom, Senior Associate Lecturer in Birkbeck’s Department of Biological Sciences reports on the 2019 Rosalind Franklin Memorial Lecture, delivered by Professor Emily Rayfield on how computational tools are reshaping our understanding of form and function in fossil animals.

Since 2016, the School of Science at Birkbeck has held an annual lecture named for one of its most distinguished alumni, Rosalind Franklin. This lecture, which is always given by a notable woman scientist, forms part of the school’s Athena SWAN programme. Each Rosalind Franklin lecturer’s research field is closely aligned to one of the three departments that make up the School, Biological Sciences, Earth and Planetary Sciences, and Psychological Sciences. This was the year of the earth sciences, and the lecturer, Emily Rayfield, was a professor of paleobiology at the University of Bristol. She gave an engaging talk about how computer modelling is helping us understand the biology and behaviour of fossil animals, beginning with the dinosaurs.

Emily drew a contrast between the techniques used to study living animals and prehistoric ones. With living animals, there is a chance, at least, that we will be able to observe their behaviour, but with prehistoric ones all we have to go on is the fossils they leave behind. So how can we approach a question such as, what – and how – did dinosaurs eat? We can begin to understand this problem by relating the skulls, and particularly the jawbones, of living animals to their diets. Plotting the measured bite force of reptiles’ jaws, including those of the closest living relatives of dinosaurs, the crocodiles and alligators, against the size of those jaws, and then scaling up to the size of dinosaur jaw bones has suggested that the largest would have had a bite force of over 10 tonnes.

It is possible to get some idea of what fossil animals ate by, literally, looking at their dung. Fossilised faeces, or coprolites, are frequently found, and geologists can estimate the size of the creatures that produced them, as well as finding clues to their diets. The largest that have been seen are likely to have come from Tyrannosaurus rex. This monster, one of the largest land-dwelling carnivores that ever lived, measured over 40’ from nose to tail and stood about 12’ tall at the hips. Skeletal fragments found in dinosaur coprolites include those from some of the first birds. Fossilised bite marks and teeth, which differ in shape and size between herbivorous and carnivorous animals, can also fill gaps in the picture of dinosaur diets.

Feeding is only one aspect of animal behaviour, although an important one. Emily opened out her question to ask what the shape of an extinct animal’s bones can tell scientists about its behaviour more generally. Bones all respond to externally applied forces of stress (load per unit area) and strain (stretch per unit length). Wolff’s Law, which dates from 1892, states that any change in the function of bones, and therefore in the stresses and strains that they are exposed to, is directly followed by changes in their shape. In human terms, if an individual overloads his or her bones by, for example, taking up a strenuous sport, the bones will gain mass, whereas disuse will cause bone loss in astronauts exposed to low gravity as well as the chronically sick.  Mechanical loads experienced in utero affect the shape of the developing embryo and, going back to the example of feeding, animals that experience different diets develop different-shaped jaws. This can be observed in individuals of the same species, with mice raised on only soft food developing less efficient jaws than those raised on hard pellets. It is also reflected in species differences in both living and extinct animals: animals and their environments may have changed dramatically since ‘deep time’, but the laws of physics – and the basic structure of the cells and tissues they operate on – have not.

It is, of course, impossible to measure the stresses and strains that a dinosaur bone will have been subjected to, but it is not impossible to deduce them. This is where computers come in, via a mathematical technique known as finite element analysis.  In this, a complex structure is broken down into a number of simple shapes. A force is applied to each element and a computer program is used to estimate how it moves and changes shape.

To apply this technique to a fossil, you need to start with a digital model of that fossil, and this can now be done quite easily using a CAT scanner similar to those used in medicine. The model is then completed by adding any bones missing from that specimen. The model is combined with information from living relatives to estimate the stresses and strains on the bones.

Professor Emily Rayfield

Armed with all this data on the forms of, and loads experienced by, dinosaur skulls, it is possible to ask complex questions about their mechanics and evolution. It is now quite well known that modern birds evolved from a group of dinosaurs, and this begs the question of how they evolved their characteristic, but extremely diverse, beaks. Some herbivorous dinosaurs in the group known as the theropods (three-toed) had beaks and comparing models of similar sized dinosaur skulls with and without beaks has suggested that a beak reduced stress and strain during feeding. Large theropods were found to have experienced proportionally lower stress during feeding than smaller ones, with the exception of Spinosaurus, which had much higher stress than expected for its size.

At the end of the lecture, Emily moved on from the largest land-based fossils to look at some of the smallest: a group of primitive shrew-like mammals known as the ‘Jurassic fissure mammals’ that lived in crevices between rocks some 200 million years ago. Working with Pamela Gill, an expert on the anatomy of these creatures, Emily examined the fossilised jaws and teeth of two species and predicted differences in the speed and strength of their bites. Comparing patterns of wear on the teeth of these mammals with modern bats suggested a similar range of insect diets. This implies that, even at the very beginning of the mammalian radiation, species that occupied similar niches were beginning to diversify their diets; and it provides another example of how studies of the mechanics of fossil bones can lead to insights into the lives of animals from hundreds of millions of years ago.

Dr Clare Sansom, Senior Associate Lecturer in Birkbeck’s Department of Biological Sciences reports on Dr Salvador Tomas’ talk, which explored the hypothesis of abiogenesis, which assumes life arose spontaneously from non-living matter, a few billion years ago. The talk focused on research directed to the development of protocells, which are produced in the laboratory as plausible ancestors of living cells, and can be used as models to study abiogenesis. In the future, scientists may be able to use this knowledge to create programmable, cell-like robots.

Tekno the Robotic Puppy, credit: Toyloverz

Birkbeck Science Week 2019 kicked off with a talk by Dr Salvador Tomas of the Department of Biological Sciences, intriguingly titled ‘Synthesising Life’. Introducing Salvador, the Executive Dean of the School of Science at Birkbeck, Nicholas Keep, explained that he had taken both his degrees at the Universitat de les Illes Balears in his native Balearic Islands. He moved to Birkbeck to set up his lab in 2006 following postdoctoral study in Sheffield, and he now holds the position of Senior Lecturer in Chemical Biology.

His lecture was every bit as engaging as its title suggests. He started by asking the question ‘what is life?’ and illustrated the answer by comparing a ‘cyberdog’ with the common-or-garden variety. At a basic level, both dog and cyberdog can be thought of as a network of transistors or cells that responds to input signals in different ways, but while the cyberdog is programmed to carry out whatever (presumably) menial tasks its owner demands, the dog is programmed for survival. This led to a formalised definition of ‘life’, as ‘a self-sustained chemical system capable of undergoing Darwinian evolution’.  Furthermore, if you zoom in hundreds of millions of times, the dog’s equivalent of the cyberdog’s uniform network of transistors is the bewildering complexity of ‘molecular machines’ inside every living cell.

The question of ‘how life came to be’ is perhaps almost as old as humanity itself. A few centuries ago speculations centred on the idea of ‘spontaneous generation’, suggesting that (for example) fish might have arisen directly from water or mice from hay. The development of pasteurisation in the mid-nineteenth century helped disprove this theory. We now understand that all living (and extinct) organisms evolved from a simple organism known as LUCA – short for the Last Universal Common Ancestor – but this begs the question: where did LUCA come from? To answer this question, you need to go back to the kind of conditions that scientists believe to have existed on an early Earth: a chemically rich ‘warm puddle’ of liquid in an oxygen-poor environment, much like those found in underwater volcanoes today.

LUCA would have been a single-celled organism containing a minimum set of biomolecules necessary for life, all coded for by a minimal segment of DNA, and, self-evidently, all its precursors must have been non-living. For decades, scientists have been trying to recreate the process of ‘abiogenesis’ by providing simple molecules with energy in a similar environment and investigating whether more complex molecules, the ‘building blocks’ for LUCA’s DNA and proteins, might be able to form. So far, it has proved possible to make the basic building blocks of proteins, the amino acids, and even, in some circumstances, to join several amino acids into a short chain, but not to connect hundreds of them to form a complete protein. Nucleic acids, the building blocks of DNA, are proving even more intractable.

Building blocks become biomolecules through a process in which each two units – amino acids or nucleotides – are joined together with the loss of a water molecule. This process requires energy, but the opposite one, in which the bond between the units is broken, can be spontaneous. Salvador used a set of simple blocks to illustrate how populations change over time, as combinations such as ‘AB’ are ‘born’ and ‘die’. If AB, for example, is made ‘sticky’ so it attracts more copies of A and B, it becomes ‘autocatalytic’ (that is, it helps form itself) and the AB population burgeons. Or at least, it does so until A or B is depleted, when an ‘extinction event’ occurs. The system becomes more complex with the addition of an energy supply and further building blocks, and it becomes possible to see how collections of units with specialist functions could evolve. Some types would specialise in storing information (the ancestors of DNA) and others in promoting bond formation (the ancestors of proteins).

This would be a resourceful molecular system, capable of building its own building blocks, but it would have one major disadvantage: its survival would depend on the proximity of different types of molecule. If it were in the ‘warm puddle’ of the early Earth, a single rainstorm could blow it away. Keeping the components together requires a third type of biomolecule. Lipids are molecules with a long ‘water-hating’ tail and a short ‘water-loving’ head, and in water they form double layers with the tails pointing towards each other. These lipid bilayers often form spherical vesicles, and any primitive biomolecules trapped inside such a vesicle will stay together come what may.

Vesicles containing both ‘DNA-like’ and ‘protein-like’ molecules can be thought of as ‘protocells’: or, if you like, putative ancestors of the ancestors of LUCA. Salvador explained that his own contribution to the evolving story of synthesising life was in exploring the chemistry inside such protocells. Something like a protocell is almost certain to have existed, and this will have evolved to be better programmed for survival through developing more efficient chemical ways of making use of resources, storing and using energy, and responding to stimuli. Reproducing this process by adding molecular machines and efficient, specialist switches to a blank vesicle or protocell can generate cell-like robots. Initially, these are likely to have a variety of useful but quite mundane functions in, for example, targeted drug delivery, but eventually they might do more: ‘life, but not as we know it’, perhaps?

Salvador ended by asking two questions: can we synthesise life, and if so, should we? Most of his audience agreed with him that the first was ‘not done yet, but seems likely in the near future’. Interestingly, however, a majority thought that it might be too risky to take very far.