We populate the world of cells, like the world of people, with heroes and villains. There are cells that are our friends and saviours, and cells that want to kill us. There are “good” and “bad” bacteria. There are killer cells, rogue cells, zombie cells. The metaphor of cell as organism is thriving and filled with personalities.
Pathogenic bacteria – “germs” – were easily assimilated into this menagerie: they are invaders from outside, the invisible enemy with which we are always at war. But cancer cells are harder to conceptualize, for the same reason that they are harder to combat. They are a part of us: our dark side.
Cancer is not a disease like any other. It’s not ultimately caused by a pathogen like a virus or bacterium, although some viral infections can trigger it. Indeed, part of the puzzle, as well as part of the reason why “curing” cancer is such a challenge for medical science, is that many causes all result in the same symptom: the uncontrolled proliferation of cells as tumours, which wreak physiological havoc that can be catastrophic.
Cancer is not, as is sometimes implied, a modern affliction. It has always been with us, and it afflicts most other animal species. While it’s understandable that we should frame such a potentially lethal condition in terms of dysfunction, breakdown and invasion – a condition on which we must “wage war” – it’s not clear that this helps us to comprehend or come to terms with cancer. It makes more sense to see it as something cells naturally do: as an inevitable consequence of being multi-celled beings.
One increasingly popular and fruitful way of thinking about cancer is in evolutionary terms. For it highlights a puzzle about our situation as cell communities, which is that the normal evolutionary process has been largely eliminated from it. As bacteria multiply and a colony spreads, random mutations occur during cell division, and this gives natural selection variation on which to work. That’s how antibiotic resistance arises: in the presence of antimicrobial agents, bacterial cells that acquire by chance some resistance gain a strong advantage and soon come to dominate the colony. The more you ply bacteria with antibiotics, the stronger becomes the selective pressure that can generate resistance.
You would expect variation also to arise from copying errors in DNA replication as our own cells divide and proliferate during growth and tissue replenishment. And it does: as we saw earlier, the human body is full of small differences in the precise sequence of our genome. Mostly these mutations have no effect, but it’s reasonable to imagine that some will convey a reproductive advantage, for example by letting some cells divide faster.
That sort of “selfish” cell proliferation is precisely what gives rise to cancer, so it’s important for the reproductive success of the whole organism that it is kept in check. The human body contains several mechanisms to actively control “in vivo evolution” that might otherwise get out of hand. For a start, fidelity of DNA replication in cell division is scrupulously checked by “proofreading” enzymes. There are genes that regulate the cell cycle, acting as brakes to keep cell division under control. What’s more, our immune system is constantly on the lookout for pathologically dividing cells. That’s why cancer may also result from immune suppression.
One theory of cancer even goes so far as to say that it is not a failure of the body’s mechanisms for keeping cell proliferation in check, but rather a natural response to environmental factors that subject cells to stress, which is encoded in our genomes from some very ancient evolutionary epoch but normally kept on a leash. “It may be triggered by mutations, but its root cause is the self-activation of a very old and deeply embedded toolkit of emergency survival procedures,” suggests physicist Paul Davies. That’s a controversial and minority view, but it illustrates how cancer may need new narratives in comparison to most stories of disease. From an evolutionary perspective, cancer is just the kind of thing you’d expect cells to do. Just as civilization and socialization require us to suppress ruthless, selfish atavistic urges, so cells in an animal body must learn restraint and cooperation, and be policed to ensure compliance. It’s just as Thomas Hobbes said in Leviathan: the body, like the body politic of the state, relies on a suppression of the State of Nature. With cancer, as with anarchy, the alternative is a life that is nasty, brutish and short. (Let me just remind you to keep your Metaphor Alert switched on here.)
* * *
Military imagery – a “war on cancer” – adds to the impression that we are fighting off some invading pathogen rather than trying to keep control of our own cells and flesh. That confusion is invited by the fact that things entering our body from outside can indeed trigger cancer. They do so by interfering with cell regulatory mechanisms that normally suppress tumours.
One of the earliest associations of cancer with environmental factors was the discovery in 1775 by English physician Percivall Pott that there was an unusually high incidence of cancer among men who had worked in their youth as chimney sweeps – soot being, we now know, horribly abundant in carcinogenic chemical compounds. The problem for understanding the causal factors was that these associations seemed to be so varied.
Take X-rays: discovered in 1895, they were at first thought to be an amusing and even beneficial form of radiation for several decades, until they were found to be associated with skin cancers and leukaemia. The same with radioactivity: miners of uranium ore (pitchblende) proved to be particularly prone to lung cancer, and both Marie Curie and her daughter Irène Joliot-Curie died prematurely from cancer surely caused by their experiments on radioactive materials. It wasn’t until the link between smoking and lung cancer was discovered in the 1950s that these diseases became associated with lifestyle more generally: what you eat, where you live and work. Many chemical agents, some of them common in foodstuffs and additives, are now known to be carcinogenic: liable to create a raised cancer risk. To complicate matters even more, some cancers are linked to viral infections. The sexually transmitted but largely innocuous human papillomavirus, for example, can increase the risk of cancer, especially of the cervix.
It took a long time to figure all this out: to appreciate that these external agents may in some way or another introduce mutations to genes vital for keeping cell division under control. Some of these genes, called proto-oncogenes, perform important roles in the cell cycle, and a mutation – sometimes of just a single base in the DNA sequence – can convert them to a faulty form called an oncogene that triggers uncontrolled division. A gene called Myc is one of these proto-oncogenes, having roles that are many and varied.1 Other cancer-causing mutations appear in genes that actively inhibit the formation of cancer cells, for example by slowing down cell division, repairing damaged DNA or telling cells when to die. These are called tumour suppressor genes, and if mutations disrupt their function, cells may again run amok. One such gene, called p53, is a component of the cell cycle’s braking machinery: when it is switched on it can either halt the cycle and initiate DNA repair or induce cell death if the DNA is beyond mending. Both the activation of p53 and its knock-on effects are complicated, but about half of all cancers are thought to involve p53 in some fashion.
These cancer-causing gene mutations can be produced in the body by the action of some carcinogenic chemical or radiation (ultraviolet light or X-rays, say), or they can arise spontaneously during cell division, or be inherited. Recent years have seen a debate about whether genetic mutations appear primarily at random or because of lifestyle – an incendiary question because it carries implications about whether cancer is to some extent just “bad luck” or primarily “self-inflicted”.
As well as having tumour suppressor genes, our healthy cells have another important mechanism that stops them becoming cancerous. If their genetic systems go awry, they are likely to undergo a kind of self-inflicted death: suicide, if you will. Death is in fact the fate of any somatic cell lineage: cells will automatically die (a process called apoptosis) after they have divided a certain number of times, called the Hayflick limit (see here), which for humans is around 50. This makes sense, because each time a cell divides there will inevitably be some mistakes in replicating its DNA: no enzyme machinery is good enough to ensure perfect fidelity in copying two sets of three billion base pairs. So cells that replicate indefinitely will acquire ever more faulty and aberrant genomes. Much better, then, to set a limit and let those degenerating cells die off, while replenishing the tissue by production of fresh lineages.
Apoptosis also commonly happens for cells that lose their proper role in the growth and maintenance of the body. It takes place, for example, in cells grown in a petri dish that become separated from the main mass – a reason to suggest that cells are inherently “social” entities, acquiring their raison d’être only in relation to others and depending on signals from their neighbours to sustain them. “Apparently the only thing our cells do on their own,” says biologist Martin Raff, “is kill themselves, and the only reason they normally remain alive is that other cells are constantly stimulating them to live.”
Cell death is also involved in the sculpting of the body from growing tissue: it is what removes flesh from between the separate digits of the growing hands and feet, which begin as paddle-like masses in the embryo. If this fails to happen, some fingers or toes may remain connected by an intervening web, as they are (for good adaptive reason) on the feet of ducks.
But cancer cells can evade apoptosis and go right on dividing, so that a tumour grows and grows. They do so by undermining the built-in “division counter” that cells possess. This counter takes an ingenious form. At the tips of our chromosomes are stretches of DNA called telomeres, which don’t encode proteins but are merely end caps. Each time the chromosomes are replicated, the telomeres are incompletely copied and get progressively shorter. When the telomeres are completely worn away, things go wrong with the unprotected tips of the chromosomes – they may fuse with one another, for instance – and the resulting chaos prevents the cell from growing and primes it for death.
Cancer cells avoid this by manufacturing an enzyme called telomerase, which repairs telomeres. Cells in the early embryo need to conduct such repair too, which is why our cells have telomerase genes in the first place – but usually they are mostly silenced thereafter.2 In cancer cells, these genes are reawakened.
Thus there are many requirements that cells must satisfy before they develop into a malignant tumour – it’s not simply the case that any old bit of DNA damage will set that fateful process in train. The brakes of the cell cycle must be removed to enable unchecked proliferation. The cell must evade tumour suppressors, and they must fire up their telomerase genes to avoid apoptosis, and stay hidden from the immune system (on which, more later). And they must actively induce the body to grow new blood vessels so as to keep cells deep within a tumour supplied with nutrients. They must eventually be able to spread throughout the body in the lethal process called metastasis.
All this can easily sound like a systematic and purposeful plan. It’s no wonder, then, that cancer cells are so often cast as rogues, renegades and killers. Some researchers express this in remarkably teleological terms, talking of pre-cancerous tumour cells employing all manner of dastardly “tricks” in a selfish effort to ensure their unchecked proliferation, as if these cells are intent on “outwitting the system”. Meanwhile, the natural defences of the cell are depicted as labouring valiantly to prevent this disaster: the p53 gene, for example, is in cancer expert Robert Weinberg’s words “the arbiter between life and death, an ever-watchful guardian that monitors the cell’s well-being and sounds the death alarm if the machinery of the cell is damaged.” Quite a responsibility.
There’s a pedagogical value in such language that chimes with our own sense of cancer as a terrifying and deadly enemy. But it’s worth keeping the metaphor always in view as no more or less than that. It may be that on other occasions, for other purposes, it is better to accept cancerous growth as something innate to cells rather than as a pathological deviancy. It surely results, after all, from the same evolutionary “imperative” that gave rise to us in the first place. It reminds us that there is a contingency in our emergence and continued existence as a cell community. As individuals and as generic human forms, we are just one of the possible outcomes from the potential that exists within our formative cells as they multiply and communicate.
* * *
I said that one of the things cancer cells must do is avoid the immune system. Within this narrative, our immune cells are the body’s police, patrolling the body on the lookout not just for invaders like viruses and bacteria but “antisocial” cells of our own. If cancer cells are seen as the villains, immune cells are the heroes.
But according to today’s understanding, the immune system does much more than this. Sure, it includes cells called lymphocytes that will identify, catch and digest invaders. But it also plays an important role in many diseases that are more to do with malfunctions of the tissues than with pathogens: brain and heart diseases, obesity, arthritis, diabetes. In part, that’s because the immune system controls what is almost the default response to disease or malfunction in the body: inflammation and repair. According to immunologist Lydia Lynch, “we can’t ignore the immune system in any disease.”
Few areas of biology have advanced more in the past several decades than immunology. At the same time, few are more forbidding to outsiders, more bristling with impenetrable acronyms and hedged with exceptions and complications. The traditional picture was at least fairly straightforward in broad outline: the immune system was considered to fend off pathogens such as bacteria and viruses by generating a huge variety of specialized white blood cells called B and T cells, with “receptor” proteins on their surfaces among which one of them might have the right shape to latch on to the intruder. When that happens (the story went), the machinery of the immune system kicks into gear to eliminate the foreign particle, for example by recruiting cytotoxic or so-called killer T cells to kill and ingest the infected cells. Meanwhile, T cells are “trained” to recognize the body’s own cells from the molecules called HLA proteins (never mind the abbreviation – it’s another misnomer) on their surfaces. No one’s HLA proteins exactly match anyone else’s, but they fall into general classes that determine the compatibility of tissues for transplants.
This is all more or less right, but is far from the whole story. For one thing, the generation of B and T cells with diverse receptors constitutes only one part of the immune response, called the adaptive immune system. In the late 1980s, immunologist Charles Janeway proposed that there is also an innate immune system that deploys immune cells with standardized receptors for common pathogens. Other researchers verified Janeway’s bold idea, for which a Nobel prize was awarded in 2011 – sadly, too late for Janeway, who died in 2003. The innate immune system has its own suite of “killer cells”, while being able also to awaken the adaptive immune system. It is the first line of defence, and is an evolutionary older strategy, still dominant in plants, fungi and other primitive multi-celled organisms. It has the advantage of being much faster to respond, albeit at the cost of lower flexibility for recognizing new types of pathogen. The adaptive immune system, meanwhile, maintains a memory of past encounters and thereby confers the immunity on which vaccination depends.
A major aspect of the immune response is tissue inflammation, a signal of the body’s defences going into action. However, it’s all too common for the immune cells to respond to false alarms, leading for example to allergic responses, with all the accompanying discomfort of mucus generation and skin irritation as the body tries to destroy or expel foreign entities that actually pose no threat. Immunity can also fail in the essential task of discriminating “friend” from “foe”, so that it attacks the body’s own cells in disorders of “auto-immunity” that include rheumatoid arthritis, type 1 diabetes and muscular dystrophy.
Such dysfunctions are normally avoided by a suite of immune mechanisms. Some types of T cell, for example, regulate the immune response, shutting it down once it has done its job. A part of this regulation involves a protein called CTLA-4, production of which is switched on after the T cells have been activated by detection of some foreign agent. CTLA-4 issues the instruction “ease off now”, acting as a brake or so-called “immune checkpoint”.
One of the most exciting and promising strategies for treating cancer in recent years enlists the immune system to identify and selectively destroy tumour cells. Cancer immunotherapy manipulates these checkpoint molecules, blocking their function so as to take the brakes off the immune response and unleash its full force on cancer cells. It uses protein drugs designed to latch on to specific target molecules on T cells and turn off their immune checkpoints.
Directing the immune system at cancer is a very old idea, but the checkpoint approach shows signs of really working, in some cases producing long-lived remission. It was pioneered in the 1990s by biologists James Allison of the University of California at Berkeley, who focused on manipulating CTLA-4, and Tasuku Honjo at Kyoto University, who worked on another checkpoint gene called PD-1. (Different brakes might be more effective for different types of cancer.) Allison and Honjo shared the 2018 Nobel prize in medicine for their work.
The challenge in using immune cells to fight cancer is that cancer cells are not, of course, your regular kind of invader. They are our own cells, which have switched to a state that endangers the organism as a whole. If the immune system has painstakingly acquired an ability to ignore the body’s native cells, how can it attack cancer cells? But in fact the immune system does possess some ability to distinguish cancerous from healthy cells. The changes that make a cell tumour-forming may already be enough to set off alarm bells for the patrolling T cells. At the same time, cancer cells acquire some ability to evade surveillance. It’s a finely poised game of hide-and-seek, in which efforts to intervene with drugs or agents that make cancer cells more visible to the immune system run the risk of triggering other, harmful autoimmune responses too. One approach under development involves genetically modifying a patient’s T cells so that they will attack tumour cells specifically, leaving others alone. Another is to tweak the microbes in a patient’s gut (see below) to boost the power of the treatment: certain gut bacteria make the immune system more responsive, but not everyone’s gut hosts them.
A checkpoint suppressor drug called Ipilimumab is now licensed by the United States’ Food and Drug Administration (FDA) for use against skin cancers. A course of treatment costs over $100,000 but, especially when used in conjunction with other immune-boosting drugs, the results have been encouraging. Cancer immunotherapy has in the past few years begun to make a real impact on treatments in the clinic, with some patients given these treatments making remarkable recoveries from conditions that would previously have been fatal. One woman recovered from a lung tumour the size of a grapefruit, and a six-year-old child was rescued from near-terminal leukaemia. Cancer immunotherapies seem to have particular promise for treating skin and blood cancers (melanomas and leukaemia). Researchers at the Fred Hutchinson Cancer Research Center in Washington state have reported initial clinical trials in which more than half of the patients treated went into complete remission, and for one particular kind of leukaemia, fully 94 per cent of the patients – most of whom would have been diagnosed as terminally ill – saw their symptoms vanish. According to Lydia Lynch, thanks to cancer immunotherapy, “we are now using the taboo word [in cancer research]: cure.”
* * *
Both cancer research and immunology show how fragile our sense of somatic integrity is. We might feel that we are unified beings making our way through an environment beset with external threats, but at the cellular level our health and existence depends on a constant and immensely complicated transaction between a diverse panoply of entities whose roles, for better or worse, may depend on their context as much as their content. And if it is not already disorienting enough to appreciate what our own cells are up to, we must factor into the equation the fact that around half3 of the cells in our body are not “ours” anyway.
They are the cells of organisms that exist symbiotically with us, living all over and within our flesh in an arrangement that is mostly of mutual benefit. The most familiar component of this so-called microbiome is the bacteria within our guts, which (among other things) assist digestion: these are the “good bacteria” that we are encouraged to supplement with probiotics and yoghurt. But there are also microorganisms on our skin, in our mouths, and in many parts of our bodies – not just bacteria but also fungi (such as yeasts) and the other type of prokaryotic single-celled organisms called archaea. In all, these microbial symbionts account for several pounds of our body mass.
Our own cells have adapted to live with these companions: not just to rub along, but to help one another. The gut bacteria are sufficiently important for wellbeing that a mother’s milk contains sugars produced especially to nourish them, which the infant itself can’t digest. The gut bacteria, in return, assist not just in digesting food but in building and repairing the body, for example by replenishing the gut lining and participating in the process by which the body stores fat. In some organisms, even fundamental biochemical processes may be outsourced to symbiotic microbes. Termites and some cockroaches eat wood but can’t digest it; that is done using the enzymes produced by their gut bacteria. The Planococcus mealybug makes some of its vital amino acids using enzymes from two types of symbiotic bacteria.
It seems likely that the very growth of the embryo is partly orchestrated by the microbiome: these symbionts have been shown to activate certain genes needed during development in a variety of organisms including the zebrafish, fruit fly and mouse. Proper development of the mouse immune and digestive systems, for example, depends on chemical signals delivered by bacteria. Even fitness, mating and reproduction may be under the control of symbionts: bacteria in fruit flies make pheromones that influence the host’s mating preferences.
There are likely to be interactions of this kind involving our own microbiome too, further undermining the idea that our genome is a self-contained “instruction manual” for constructing a properly functioning human. When such higher organisms are grown in laboratory conditions that suppress their microbiome entirely, the resulting creatures typically have precarious health.
The microbiome’s role in our health is profound and perhaps even a little disturbing. The balance of our internal ecosystem is commonly altered by disease, and while it can be hard to distinguish cause and effect, there’s some reason to believe that tinkering with the microbiome could lead to cures or at least to alleviation of symptoms. The human gut is connected to the brain by a long nerve called the vagus nerve: a conduit that might give the microbiome some influence over our mental state. Some research suggests that probiotics – particular types of beneficial bacteria – could be used to treat conditions such as stress and depression, while the gut microbe Bacteroides fragilis has been implicated in some of the symptoms of autism. Such claims remain contentious, but it is by no means absurd to think that our microbiome has an influence well beyond a good digestion.
Our body’s microflora is highly personalized: no one else’s will be quite like it. What’s more, this ecosystem, like that of a jungle, varies over the territory it occupies – even the microbes on our left hand will be different to those on our right. And the function a microbe has may depend on where it is: a bacterium might be beneficial in the gut but pathogenic in the bloodstream. As with most stories in biology, context is everything.
Evidently, this collaborative community of non-human cells has to be tolerated by the immune system. In fact, the immune system might be engaged as much in managing the microbiome as it is in warding off hostile invaders. What’s more, our body’s microbes appear in turn to exert some influence over the mechanisms of immunity – commandeering it, you might say, to protect themselves against non-indigenous microbes.
Given this intimate and essential interaction between our human cells and those of the organisms we host, we need to respect and nurture the microbiome. We know all too well how rotten we can feel if it is knocked out of kilter after a course of antibiotics (although that’s by no means the only reason for the side-effects of those drugs). All the more reason then to be wary of a culture that sees fit to coat every surface with bactericidal agents – the vast majority of bacteria are, after all, harmless to humans. At the same time, the many contingencies of the microbiome should warn us once again against making any simplistic taxonomy of “good” and “bad” cells. As science writer Ed Yong has said:
There is no such thing as a “good microbe” or a “bad microbe”.
These terms belong in children’s stories. They are ill-suited for describing the messy, fractious, contextual relationships of the natural world.
The close connection between a host and its microbial symbionts, in which both contribute to the survival of what we might call the “joint organism” – which some dub the holobiont – complicates evolutionary theory. If a microbiome actively participates in the holobiont’s survival prospects, what then is the entity on which natural selection acts? But a microbiome is not inherited intact: yours will share some features with your mother’s but is not identical. Some of the microbes hosted in your birth mother will have been transmitted to you during passage down the birth canal, and also subsequently during physical contact and breast-feeding. But some will have been acquired from elsewhere, including other individuals around you during infancy. So it’s not clear how to incorporate the microbiome into evolutionary thinking: should it be considered a genetic pool entirely separate from that of the host, or at least partly connected to it?4 Evolutionary biologists are divided over the issue. Perhaps, suggests one of their veterans W. Ford Doolittle, the coherent evolving entity is neither host nor microbial symbiont but the process they enact together: the pattern of metabolism, say. Doolittle likens this to the way songs are perpetuated (while evolving!) by people singing them: it’s the song that survives and evolves along the way.
Some researchers even go so far as to suggest that we should think of organisms possessing a “hologenome” constituted by the genes of the host and its symbionts. Others dismiss or even deride that idea – although, as we’ll see later, it is far from obvious how biology can be parcelled up into individual organisms. At any rate, there’s nothing in this blurring of genomic selfhood that is incompatible with the standard Neodarwinian view of evolution. The question is whether or not it constitutes a useful way to tell the story – or rather, perhaps, of what kind of story you’re trying to tell.