A key for every lock in the universe – the Blind Locksmith ?
This essay was written by Benedict Seddon and was first published in the 2011/12 Mill Hill Essays.
In the Matrix trilogy of movies, machines controlled the fate of humankind. Every living person was plugged into a computer simulation of Sydney, Australia, in the 1990s. In this virtual reality world people walked side by side with computer software constructs that were projected as people. In the course of saving humankind from this virtual imprisonment, our lead protagonists, Neo,Trinity and Morpheus encounter various ‘software’ individuals who help them in their quest. One such character was the ‘Keymaker’. He had the unerring ability to make any key, for any lock, on any door in the entire virtual world. I am not talking about that kiosk in the shopping centre that also reheels your shoes. Being a sci-fi blockbuster these are naturally not ordinary doors, but portals to special and important locations. The Keymaker has remarkable powers of prescience – “I’ve been waiting for you”. He files the final burrs off of a newly cut key moments before it is needed. He has near supernatural powers of both anticipation and foreknowledge of the keys that he requires and the doors that they unlock. Watching the film as an immunologist, many of the Keymaker ‘s remarkable powers reminded me of our body’s immune system. It too is always poised, ready to react against an invading organism. It seems to know what is coming round the next corner. It can recognise and react against virtually any bacteria or virus in our environment and then remember it in the future, almost like it is intelligent and has a mind of its own. Many of the features that make the immune system so remarkable, are mediated by a single family of proteins: antibodies. In this article, I will explore what antibodies are, how they work and how they are made by the immune system. I will also look at the equally intriguing question of why we need this remarkable protective system at all.
Our immune system has an impressive arsenal at its disposal designed to keep the various bacteria, viruses and other pathogens in our environment at bay. Because there are fundamental differences between our chemistry and that of bacteria and viruses, we have successfully evolved immune mechanisms that recognise and exploit these differences to identify and eliminate such foreign invaders. These immune mechanisms are intrinsic, or ‘innate’, to every individual; they are hard-wired into our genetic code. Everyone has the same set of ‘innate’ immunological tools that work in an identical way. They are an extremely effective tool. Without our innate immune system we would very soon be overwhelmed by the various microbes that live alongside us. However, because this system is hardwired it is somewhat inflexible. Any bug which learns how to evade the system could potentially run rampant through our bodies. Fortunately we have also evolved a more sophisticated system of sensors to complement the activity of our innate immune system. These sensors are highly adaptive and reactive to the microbes in the environment.When exposed to a particular microbe, this adaptive system has the capacity to learn to recognise it. These powers of recognition are accompanied by equally impressive powers of memory. Once it has learnt, the adaptive immune system takes a long time to forget. You may remember having chicken pox as a child. So does your adaptive immune system, and that is exactly why you do not catch it again when exposed to that spotty younger sibling. It is this property of the adaptive immune system that vaccines exploit: they teach the immune system to recognise a disease-causing microbe by exposing it to a harmless version. Perhaps the most impressive feature of this system is just how adaptive it really is; the scope of recognition by the sensors is mind-boggling. Every one of us has the capacity to generate sensors that recognise any conceivable microbe in the entire universe. This sounds like a bold claim and one difficult to test without going boldly where no man has gone before, to seek out all these microbes. But as you read on, you will see that not only is the theory behind this claim solid but that the reality is equally likely to be true.
Like many of the nuts and bolts that make the body work, the ‘sensors’ I refer to, are composed of protein molecules that have a specific shape or ‘structure’ that is suited to their purpose. These sensor proteins are ‘antibodies’. Scientists have had an appreciation of the properties of antibodies as early as the late 19th century, even without understanding precisely what they were. Antibodies are particularly abundant in blood, and scientists studying blood serum from immune individuals found the serum to have remarkable properties: it could neutralise the harmful activity of bacterial toxins. When added directly to bacteria, it was noticed that the bacteria clumped together and in some cases were even directly destroyed by something in the serum. These early studies also demonstrated the highly specific nature of the adaptive immune system. Serum from someone who was immune to a particular infection only exhibited these properties when confronted with that particular bug or toxin, and showed no immunity towards other unrelated bacteria. This was an important observation, because it revealed what we term ‘specificity’ within the adaptive immune system: the ability to specifically recognise a particular microbe or toxin. This specificity is a central property of the antibody molecule. A single antibody will recognise and bind to a very particular target, and only that target. Successful recognition of a microbe by an antibody is bad news for that microbe. Antibodies can block the infectivity of microbes, preventing them from getting into the body, and they can neutralise the adverse effects of toxins. For invaders that do successfully get into the body, antibodies encourage other immune cells to engulf and destroy their target and activate potent immune proteins in the serum to also destroy the target.
Antibodies are a remarkably potent and versatile defense system that has the capability to recognise just about any disease-causing agent. How can one molecule have the ability to recognise so many different targets? The answer is found both in the three-dimensional structure of the protein and most especially in the extraordinary nature of their genetic blueprints.
A complete antibody molecule is in fact made up of two pairs of proteins: two identical ‘heavy chains’ and two identical ‘light chains’. The ‘heavy’ chain is big and the ‘light’ chain is a smaller protein, half the size of the heavy chains.The bottom halves of each heavy protein stick together, leaving the top halves free to separate, thereby forming a ‘Y’ shape. Each free arm in the ‘Y’ is associated with a light chain, which is approximately the size of one of these free arms (see the cartoon in Figure 1). Critically, the section of the antibody structure that is responsible for recognising and binding is the same for all antibodies, regardless of what their target may be. At the tip of each arm of the ‘Y’, both the heavy and light chains have three protrusions, rather like stubby fingers. Together, these six stumps form a precisely shaped binding surface. When antibodies bind their targets on microbes or toxins, they are in fact targeting a very particular part or component of the microbe according to its shape. The shape of the target and the antibody binding surface are complementary to one another and fit together like 3D jigsaw puzzle pieces. The antibody will only bind effectively to its target on a microbe if the shape is a good match, much like a key opening a lock. These structural features explain how antibodies can be so particular about what they will bind.
If antibodies are so ‘specific’ about what they will recognise and bind, then it figures that we would need to have a very large selection of different antibodies in our immunological arsenal to cover all the different microbes we might encounter, or in other words, we need a potential ‘key’ for every conceivable infectious ‘lock’. It is this aspect of antibody biology that is most remarkable. How does the immune system know which key to cut to unlock a particular infection? The ‘Keymaker’ in the Matrix movies knew intuitively what key to cut and when. He seemed guided by supernatural powers of foresight that he did not even understand himself. Faced with a similar challenge, evolution has come up with a solution that is arguably even more remarkable than the Keymaker ‘s supernatural powers! Since the immune system cannot be clever about guessing which ‘keys’ to make nor can it see the future about which ‘doors’ need unlocking, it uses the only other alternative that will guarantee success – it makes every conceivable key in the universe! The immune system arms itself with an arsenal of different antibody types, the size of which is quite staggering. The number of different antibodies in any one individual is estimated at between one and ten billion. The immunological Keymaker is blind – it simply churns out as many different keys as possible, in the hope that a few will have the potential to protect us from any particular infections.
This seems like the ideal, if improbable, solution to the problem. If you have every conceivable key in your locker, then you will never be disappointed. But how can this be possible in practical terms? Keys for real door-locks work because it is practically impossible for a would-be burglar to guess what shape the key should be. To understand and appreciate the scale of the immune system’s achievement, we need to delve a little into proteins and genes so that we can understand how the immune system can synthesise the highly adaptive antibody molecule. A single antibody complex is made up of four proteins – two heavy and two light chain proteins. These individual proteins are themselves made from specific building blocks, ‘amino acids’, connected end-to-end in lines.There are 20 different types of amino acid, each with different chemical properties, sizes and shapes. When connected together the string of amino acids always twists into a particular shape and has particular chemical properties that depend on the precise order of the amino acids. If the order is the same, the 3D shape and chemical properties will always be the same. Cells in our bodies know how to make different proteins because the building instructions for each protein’s particular sequence of amino acids is encoded in our DNA molecules. The DNA encodes genetic blue-prints for every protein in our bodies, including antibodies, using a code of four different letters. The complete instruction set that encodes the amino acid composition for a single protein is a ‘gene’.
Since antibodies are proteins, it follows that there must be corresponding genes to provide instructions to make them. So does our genome contain all the genes for all the antibodies we need in the immune system? The human genome consists of 23 pairs of DNA strands that encode all our genes. It is big: each of the 46 strands is about one metre long and together they make up a code about 3,000,000,000 letters long. Amazingly, there is a complete and identical copy of this code in every cell in the human body, in the nucleus that lies at the centre of the cell. The human genome has somewhere in the region of 25,000 genes. If you have been keeping up with the number crunching, you may have spotted the problem. If each unique antibody requires its own gene, how can the immune system make more than a billion different antibodies when the entire genome is only a small fraction of this in size? An important clue came when scientists identified the source of antibody production. Antibodies are secreted by specialist immune cells called ‘B cells’. Crucially, an individual B cell can only make one type of antibody. Therefore, for every type of antibody in the body there is a unique B cell synthesizing and secreting it. When the location of the antibody gene was identified and the DNA sequence decoded, it became clear that individual B cells had different antibody genes to one another. However, I already told you that each cell in the body has a complete and identical copy of the genome, so how can B cells have different antibody genes to one another?
In fact, B cells all use the same set of genes to build their antibodies. The trick, however, is that evolution has combined two nifty genetic processes with a custom evolved genomic ‘word processor‘ that B cells can use to edit and make their own unique antibody gene. The two genetic processes are genetic mutation and the way that genes are structured in our genome. The process by which genes undergo spontaneous, random mutations is the raw material of evolution. If a change in the genetic code results in a new version of a protein that works better in the body then it has a good chance of being passed on to that individual’s offspring, preserving the new gene in the population.There are many kinds of mutation that can occur, but one of the more radical ones is when an entire gene is duplicated. This could happen to a particularly useful or adaptable protein. The two copies can evolve separately from one another and end up performing different but advantageous functions in the body. The second genetic process concerns the structure of genes themselves. The genetic code, or instructions, for a protein sequence is very rarely just a single continuous sequence of DNA. Rather, they are divided up into pages separated by stretches of DNA whose code has no meaning in terms of instructions for proteins. You can think of it like your local newspaper, in which the news articles are regularly interspersed with pages of adverts for local services and such like. To get just the news, you need to cut out and throw away all the adverts. This is similar to the way that a final and unbroken set of instructions for a gene is assembled for use.
The combination of the mutational phenomenon of gene duplication with the ‘page layout’ structure of genes holds the key to antibody gene diversity. When the gene structure for antibodies was decoded for the first time in the 1970s, what was revealed was something quite unique and extraordinary about the way the genetic instructions for antibody proteins were laid out in the genome. The genes for the heavy and light chains, that together assemble into fully functioning antibodies, each consist of a number of different pages of instructions. The DNA code for the tips of the antibody’s arms, that form the crucial binding site, are covered by three different pages of instructions, termed ‘Variable’, ‘Diversity’ and ‘Joining’, or V, D and J for short. Close inspection of the gene structure revealed that there is not just one copy of these critical instruction pages in the gene, but rather multiple copies of the same pages, all slightly different from each other (see Figure 3). The gene for the heavy chain contains ~6 different versions of the J page of instructions. For the D pages, there are ~27 different copies, while for the V pages, there are more than 120 different copies, all different from each other. During evolution, these critical pages had undergone multiple duplication and mutation events, thereby generating an enormous library of different instructions to choose from.
When B cells are made in the bone marrow, they do something that virtually no other cell in the body does. They directly edit their genome, in a process called recombination, choosing just one V, D and J page from the available library to make a fully functional heavy chain gene. This genetic ‘pick and mix’ allows them to generate their own custom antibody gene that is unique to that individual B cell. This radical process is only possible because the newly growing B cells in the bone marrow synthesise specialised ‘genome word processor’ proteins that can actively cut and edit the DNA. This gene editing occurs in a similar way for both the heavy and the light chain genes. Therefore, the number of possible different combinations between the two is astronomical.
There are approximately 20,000 different combinations of V, D and J for the heavy chain genes. The light chain genes do not in fact use the D page, and have slightly fewer choices of V and J instructions, so there are fewer combinations – approximately 500. But together there are a theoretical 10,000,000 different combinations of heavy and light chains possible. As if this were not enough, the genetic ‘word processor’ is deliberately sloppy in the way in which the cut pieces of DNA are re-stitched back together, and extra DNA codes can be inserted into the join at random. This results in even more variety in the final instructions. This extraordinary diversity in genetic code translates directly to the diversity of the 3D structure of the antibody molecule that is generated, resulting in an enormous variety in the specific shape of the antibody binding surface.
In evolutionary terms, antibodies are the latest and most sophisticated addition to the immune system. They are only found in vertebrates, which includes fish, amphibians, reptiles, birds and mammals. Interestingly, though, more primitive invertebrate organisms do not have any kind of adaptive immune system, so cannot make antibodies. Instead, they rely solely on their intrinsic, ‘innate’ immune mechanisms to protect themselves from infectious agents. Their innate immunity has proved more than adequate to protect them so they have not needed to evolve a more adaptive system like antibodies. It is an interesting question to wonder why vertebrates have developed an adaptive immune system. Fish were the first vertebrates to evolve, appearing in the seas about 450 million years ago. The appearance of antibodies in fish, together with the impressive scale of the gene duplication events involved in their development, suggests an extraordinarily rapid burst of evolution. There must have been some profound evolutionary pressure during the evolution of fish that resulted in the development of the antibody. What could it have been? One clue perhaps is the presence of extensive immune tissues surrounding the gut in fish. The gut is one of the most exposed surfaces in vertebrates. While we may consider our intestines as part of our inner workings, the gut is essentially a continuous tube that opens at its entrance and exit. As such, it is connected to the outside environment, albeit carefully internalised and controlled by our bodies. The development of the gut allowed vertebrates to take control of their food intake, creating a closed and controlled environment in which food could be broken down and absorbed with high efficiency. This efficiency boost no doubt gave the first fish a huge competitive advantage compared with invertebrates, whose feeding depended on far less efficient filtering of scraps directly from the water. While a large gut is an advantage for absorbing nutrients, it is also an open invitation to infectious agents that may inadvertently and unavoidably be taken in with the food. Therefore, it seems that the development of new immunological defense systems, in the form of antibodies, came about to combat this new risk. Since the risk was likely to be repeated from the same microbes again and again, it was useful for the system to be able to remember encounters and react even better in the future.
An interesting extension of this theory is the ‘jaw’ hypothesis. Most fish have hinged jaws that perhaps evolved to facilitate more aggressive feeding habits, allowing them to ingest larger animals than would be possible without a jaw. Such feeding habits, it is argued, were associated with a greater risk of damage to the gut and introduction of infection, as many invertebrates that were prey for the fish had hard exoskeletons and spines. A test of this hypothesis would be to examine the immune systems of fish without jaws. As luck would have it, there are still some surviving fish species that are so-called ‘jawless’ fish on account of their lack of a jaw. Hagfish and lampreys are both jawless fish species. Analysis of their immune systems could find no trace of either antibodies or B cells. As predicted by the hypothesis, these fish had not evolved the system of B cells and antibodies common to all other vertebrates. That is not the whole story, however. More recently it has come to light that their immune systems in fact do have a family of diverse proteins that are generated through a process of genomic rearrangement. They have a different protein structure to antibodies and depend on a completely different set of genes, but they serve much the same purpose. They are a diverse set of proteins with the apparent purpose of binding invading microbes. Although this dents the jaw hypothesis a little, it does support the idea that all fish were subject to some profound evolutionary pressure. For jawless fish to have developed a similar system of variable proteins similar in principle to antibodies, is an amazing example of convergent evolution. Two different animal groups evolved a similar solution to a common problem but by different means. Whatever this pressure was 450 million years ago, it seems that developing a system of diverse immune reactive proteins was the ideal solution.
The fight for life has resulted in many wonders of nature. It is easy to be impressed by the sheer diversity of organisms that inhabit our planet as it is accompanied by such visual diversity. The development of antibodies, however, is an equally remarkable wonder of nature. For our bodies to fight infection by directly editing our genetic code, a code so vital to the functioning of our bodies that it is found at the very centre of virtually every cell in our body, seems like an almost desperate and reckless solution. Who knows what could go wrong by such meddling? Probably there were many failures along the way, long since consigned to the scrap heap of evolution. Thanks, though, to this tireless process of refinement and selection, we are left with an immune system that truly unleashes the power of genetics, generating billions of different proteins from just a few genes. I will leave you with one final thought. How different would the Matrix movies have been if the ‘Keymaker’ had adopted the same solution to his key-making problems?