Jeremy Taylor is author of "Body by Darwin: How Evolution Shapes Our Health and Transforms Medicine" (University of Chicago Press, 2015), a former career science television documentary maker with the BBC, and author of "Not A Chimp: The Hunt For The Genes That Make Us Human" (OUP, 2009). He contributed this article to Live Science's Expert Voices: Op-Ed & Insights.
Our ailments have evolved with us, from aggressive pathogens to the aches and pains associated with bipedalism. In this chapter from "Body by Darwin: How Evolution Shapes Our Health and Transforms Medicine," author Jeremy Taylor explores how our bodies, immune systems, many microbes, and possibly even parasitic worms have co-evolved to protect human health — and how a number of studies reveal evidence that our modern, pristine society may be erasing some of the benefits from those complex biological relationships.
To learn more, read Taylor's related Op-Ed for LiveScience's Expert Voices "Friends for Life: How Good Bugs Keep You Healthy (Op-Ed) ", more about the book from the University of Chicago Press, and the excerpt on the hygiene hypothesis below.
Reprinted with permission from "Body by Darwin: How Evolution Shapes Our Health and Transforms Medicine" by Jeremy Taylor, published by the University of Chicago Press. © 2015 by Jeremy Taylor. All rights reserved.
Absent Friends: How the Hygiene Hypothesis Explains Allergies and Autoimmune Diseases
Throughout the 1990s the Johnson family was being ripped apart by the increasingly violent, self-abusive, uncontrollable behavior of their son Lawrence. He was a very disturbed child and would become highly agitated, smash himself in the face, bang his head against walls, try to gouge out his eyes, and bite his arms until he bled. At age two and a half, he was diagnosed with autism, and as he got older, things got worse. If traffic lights failed to change according to his inner timetable as he passed them while walking down the street, he would explode with rage. He could not deal with crowded places like restaurants or movie theaters, and frequently had to be physically restrained from hurting himself. His doctors tried anti-depressive medication, anti-seizure drugs, anti-psychotics, and lithium, among others, to no avail.
His parents were at their wits' end. But because Lawrence's father, Stewart, is an active, coping, problem-solving sort of man, he tried to think his way through his son's illness and became a self-taught scholar on autism. A clue soon emerged. "We noticed that when Lawrence had a fever, all that would go away. And it was true 100 percent of the time. If his temperature went up and he had a fever from a cold, flu, or sinus infection, he would stop hitting himself, he would be calm, he was like a different child. We talked to other parents with autistic kids, and they all said the same thing."
Was it simply the lethargy of feeling poorly that damped down Lawrence's bad behavior? Some scientists have suggested that fever changes neural transmission in the brain, while others have invoked changes in the immune system. No one knew for sure. But everybody who either had to live with Lawrence or care for him said the same thing: "We're very happy when he gets sick, because life is wonderful!" However, whenever his fever subsided, the frightening behavior returned. By 2005, when Lawrence was fifteen, Stewart and his wife decided they could no longer take care of Lawrence on their own. While Lawrence was away at a specialized summer school, they reluctantly applied to have him taken into care for the rest of his life. "Lawrence was going to go and be somebody else's problem because he was killing our whole family."
It was at that precise moment, with Lawrence's dismal future decided, that they received a phone call from summer camp. They feared the worst. "But they said, ‘We don't know what's going on, but Lawrence is behaving completely normally. He's fine, he's not freaking out, he's not hitting himself, he's not throwing his food, he's participating in all the activities, he's interacting. . . .'" Stewart drove up to the camp to find his son perfectly calm, engaged with the other children, and pleased to see him. They got into the car for the two-hour drive home, and Lawrence announced that he wanted to go out for dinner. He hadn't been to a restaurant in two or three years. "But now he wants to go to this place that's extremely noisy and crowded, which is a place we'd normally cross the street to avoid. He would normally never wait in line, but we waited forty-five minutes, we got served, he ate, we had a wonderful dinner!"
Stewart drove the family home, his mind whirling. Later that night he was helping Lawrence get undressed for bed when he noticed that he was completely covered in chigger bites from his thighs to his ankles. Chiggers, biting mite larvae, are very common in the summertime where the family lived, and Lawrence had been bitten by the chiggers in the long grass of summer camp. What could be the link between chigger bites and the total remission of Lawrence's autistic symptoms? It drove Stewart back into the literature to discover that chiggers elicit a very powerful immune response as they drill into the skin and release saliva to digest host tissue. They then drop off, leaving a patch of hardened scar tissue that itches for days until the immune reaction subsides. For ten days, while Lawrence's immune system fought the chigger infection instead of him, they had a wonderful time, but as the bites faded and the itching stopped, he returned to his violent and self-destructive behavior. "I immediately said, ‘That's it. I've seen enough. There's an answer here. At least some part of my son's symptoms are an aberrant immune response.'"
Stewart knew that his son's physician, autism expert Dr. Eric Hollander of the Albert Einstein College of Medicine in New York, had done research showing that there was nine times the incidence of autoimmune disease in the first-order relatives of autistic children, compared to those of normal children. Lawrence has a peanut allergy; Stewart suffers from myasthenia gravis, an autoimmune disorder that causes fatigue and weakened muscles; and his wife is asthmatic. His family medical history was consistent with research that links autism to autoimmunity and allergy. Back in 1971, researchers at Johns Hopkins University reported on a family study where the youngest son in the family had a multiple diagnosis of autism, Addison's disease (an autoimmune condition affecting the adrenal glands), and candidiasis (infection with the opportunistic yeast, Candida albicans). The next older brother had hypoparathyroidism, which can have autoimmune origins, Addison's disease, candidiasis, and type 1 diabetes. The next older brother had hypothyroidism, Addison's disease, candidiasis, and alopecia totalis — an autoimmune condition that causes the loss of all head hair. The oldest son, and firstborn, was symptom-free, like his parents.
In 2003 Thayne Sweeten, from Indiana University School of Medicine, reported that the rate of autoimmune disorders in the families of children with autism was even higher than in the relatives of children with autoimmune diseases. Disorders included hypothyroidism, Hashimoto's thyroiditis (where the thyroid gland is attacked by autoantibodies and immune cells), and rheumatic fever. Sweeten said that the finding of increased autoimmune illness in grandmothers, uncles, mothers, and brothers of autistic children "suggests a possible mother-to-son transmission of susceptibility to autoimmune diseases in the families of autistics," and speculated that autoimmunity or chronic immune system activation could account for some of the biochemical anomalies seen in autistic individuals, including high uric acid levels and iron-deficiency anemia, which are also seen in autoimmune disorders.
Research conducted on Danish children between 1993 and 2004 by Dr. Hjördis Atladóttir agreed with Sweeten by finding higher rates of autism in children born to mothers who suffered from celiac disease (where you cannot tolerate gluten). The study also linked autism to a family history of type 1 diabetes and to children whose mothers suffered from rheumatoid arthritis.
Chiggers, remission of autism, and autoimmunity all began to come together in Stewart Johnson's forensic mind. If immune disorder — a hyperactive immune system — was causing his son's autism, he needed something to damp it down. His research led him to the work of Joel Weinstock, David Elliott, and colleagues, then at the University of Iowa. Weinstock's group had reported success in medical trials where they had treated a small number of patients with Crohn's disease, an autoimmune inflammatory bowel disorder, with the eggs of an intestinal parasite — the pig whipworm. They had treated one group of twenty-nine patients with 2,500 live whipworm (Trichuris suis) eggs, delivered via a tube into their stomach, every three weeks for twenty-four weeks. By the end of the treatment period, 79 percent of the patients had responded dramatically; the whipworm eggs had driven their Crohn's disease into remission. "I was impressed," says Stewart. "These were real scientists doing real work and getting results; this was not fringe stuff. It certainly looked good for Crohn's, so I thought, ‘Maybe I'm right with this.' So I wrote all this up and did a sort of mini research paper with references, and I gave it to Eric Hollander."
Hollander was intrigued. "Stewart is a very smart guy and he'd done a lot of intense research and he just pulled out the literature and we talked it over. It seemed like a plausible hypothesis and a reasonable thing to try." Hollander obtained the necessary clearance to administer the treatment and helped Stewart with the task of importing the whipworm eggs from Germany. They began with a low dose, for fear of side effects. Stewart took the eggs as well — he wasn't about to foist a seemingly bizarre treatment on his son that he was not prepared to share. The initial results were extremely disappointing. They only saw the "good" Lawrence for four noncontiguous days during the whole twenty-four weeks of treatment. Stewart rang the manufacturers and they told him that the results were actually predictive of someone who ultimately does respond and will respond better at a higher dose. So they went to the same level that Weinstock had set for his Crohn's patients — 2,500 eggs per treatment. Within eight days Lawrence's symptoms completely evaporated, and they have stayed away ever since. The old Lawrence only briefly returned four times when they experimented by taking him off the eggs for a few days. So far, as long as he keeps taking the eggs, his autism symptoms appear to be kept at bay.
Stewart Johnson had discovered the hygiene hypothesis, which links the bacteria, fungi, and helminths (parasitic worms) in our guts, on our skin, and in our airways and vaginas, and a host of autoimmune and allergic diseases. There is mounting evidence that the composition of all these organisms, living on us and inside us — collectively known as our microbiota — can offer protection against a formidable list of autoimmune diseases, including the inflammatory bowel diseases Crohn's disease and ulcerative colitis, type 1 diabetes, rheumatoid arthritis, multiple sclerosis, and, as we have seen, mental health. Some studies suggest that they can also protect against a comparable list of common atopic or allergic illnesses like eczema; food, pollen, and pet allergies; hay fever, rhinitis, and asthma.
Nevertheless, it cannot be stressed strongly enough that autism is a complex, multi-factorial illness and that the therapeutic application of all the science associated with the hygiene hypothesis, for a variety of autoimmune and allergic diseases, is still very much in its infancy and largely unproven — Lawrence Johnson's treatment, for instance, is a one-off experiment, not the result of tried-and-tested medicine. But much of the research is compelling, and if it can translate into medical therapy, it could represent, within a few years, nothing short of a revolution in medicine.
Dramatic improvements in hygiene, sanitation, and water quality from late Victorian times to the present day, allied to extensive use of antibiotics and population-wide vaccination, have raised the quality of life and life expectancy throughout the developed world. But, perplexingly, post-industrial society — while largely eradicating epidemics of polio, whooping cough, dysentery, measles, and many other potentially lethal or debilitating infections — has fallen prey to new, major growing epidemics of autoimmune and allergic disease.
Take bowel disease for example. According to Weinstock's research, prior to the twentieth century, inflammatory bowel disease (IBD) was largely unknown. Between 1884 and 1909, hospitals in London were averaging two cases of ulcerative colitis per year at most, and Crohn's disease only became recognized in 1932. But during the second half of the twentieth century, IBD gained in scope and prevalence. Currently, IBD in the United States affects between 1 and 1.7 million people. The current estimate is that 2.2 million people in Western Europe and the United Kingdom have IBD. Although once thought to have stabilized, the incidence of Crohn's disease continues to gradually rise in England, France, and Sweden. IBD is less prevalent in Eastern Europe, Asia, Africa, and South America. However, as countries in these regions develop socioeconomically, IBD increases. Moreover, when people move from a country with a low prevalence to a country with high prevalence of IBD, their children acquire a higher risk of developing IBD.
Although type 1 diabetes has been around for centuries, it also is on the rise — and too fast for genetic change to be implicated. This guilt-by-association link between affluent, hygienic, Westernized countries and the rise of autoimmunity also extends to multiple sclerosis (MS), which has a low incidence in tropical regions that increases as you move north of the equator. In the United States, prevalence is twice as high north of parallel 37 than below. Infectious agents, genetics, and vitamin D levels have all been implicated, but, intriguingly, adult immigrants leaving Europe for South Africa have a threefold higher risk of developing multiple sclerosis than those migrating at age fifteen or less, suggesting a protective environmental effect in the adoptive country, operating only on the young.
The opposite trend is seen among the children of immigrants to the United Kingdom from India, Africa, and the Caribbean (all regions of low prevalence), where the risk of developing MS is higher than among their parents but similar to children born in the UK. Jorge Correale, a neurologist from Buenos Aires, points out that MS is rising steadily in all developed nations. In Germany, the incidence of MS doubled between 1969 and 1986, he says, while MS has increased twenty-nine-fold in Mexico since 1970, in line with steadily improving living standards. Correale also reports a fascinating inverse relationship between MS and the distribution of a very common intestinal parasitic whipworm,Trichuris trichiura, which used to be extremely common in the southern United States and still is common throughout the developing world. MS prevalence, he explains, falls steeply once a critical threshold of 10 percent of the population infected is reached. In a similar way, common atopic diseases like eczema and asthma are relatively rare in the developing world, whereas levels of helminth infection are relatively high. Weinstock recalls how the "penny dropped" for him while ruminating in his seat on an aircraft waiting interminably for takeoff from Chicago's O'Hare Airport. He was thinking of cause and effect — I do something and then something happens. He suddenly realized that the answer to the conundrum of rising levels of bowel disorder and other autoimmune diseases lay in "what doesn't happen but used to." In other words, it wasn't about what new aspects of our environment could be contributing to autoimmunity but what had been taken away from our modern environment that might leave us open to it. "In the historical environment we had filthy streets — horse dung was a major feature — and many people walked barefoot or were poorly shod. But now we have built roads and sidewalks and we wear proper shoes, and so our ability to transmit these organisms from one another went down lower and lower. Then we cleaned up the food supply, the water . . . everything got clean. As a result worms disappeared. And when you look at the incidence of deworming from population to population and the rise of immune-mediated diseases, you find that they are inversely related. That's a negative correlation, it doesn't prove that worms are effective — but it is a smoking gun."
Modern sanitation has proved disastrous to most helminths, says Weinstock. Indoor plumbing and modern sewage treatment spirit away their eggs before they can spread infection, as do frequent baths and laundered clothing. Cleaning fluids disinfect utensils and domestic surfaces, blocking their transmission. Sidewalks and shoes obstruct the common hookworms Necator americanus, Ancylostoma duodenale, andStrongyloides stercoralis, while modern food processing kills Diphyllobothrium, Taenia, and Trichinella larva. These changes have all but eradicated helminths from industrialized countries. Until the 1960s, trichinosis was endemic in the northeastern and western United States through the eating of contaminated pork. There are now less than twenty-five cases a year. Removal of these parasites has undoubtedly reduced a huge amount of morbidity in the population, but the baby has been sluiced away with the bathwater — the protection afforded by these organisms has gone with it.
A classic example of this two-edged sword effect comes today from East Africa. Educational learning specialists looking at achievement in Kenyan schools had assumed that lack of textbooks and flip charts were the major factors holding pupils back but discovered, to their surprise, that helminth gut parasites were much more important. Huge deworming programs have since almost eradicated bilharzia and hookworm, and exam results have consequently risen dramatically. However, the unwanted by-product of deworming has been a dramatic rise in eczema and other allergies among Kenyan and Ugandan children. In tropical Africa the irritation of these skin conditions often goes untreated, and constant rubbing and scratching of the sores leaves children open to infection and septicemia.
Correale treats multiple sclerosis patients in Argentina. Out of a group of twenty-four patients in whom he was monitoring the progression of the disease, he identified twelve patients who were carrying mild loads of intestinal parasites. He followed them for a little over four years, regularly checking their immunological function and performing MRI scans to identify lesions in their brains and spinal cords. The infected individuals had significantly fewer relapses, fewer lesions, and better measurements on all disability scores. He extended his observation to seven years, but after five years, four of the patients had to be given anti-helminthic treatment — their worm infections were causing gut pain and diarrhea. As soon as the helminths were swept out of their systems, all the MS indicators rose and they were soon undistinguishable from uninfected patients.
Erika von Mutius, an expert on allergies from Munich University, had a unique opportunity in the unification of East and West Germany to test her theory that high levels of air pollution and crowded, poor living conditions would lead to higher levels of asthma, hay fever, and other atopic illnesses. She expected that children from wealthier West Germany — with its cleaner environment, high levels of sanitation, and lower levels of polluting heavy industry — would show less atopic illness than children in East Germany. To her surprise, the opposite was true. East German children — who characteristically lived in cramped domestic conditions surrounded by lots of siblings, pets, and other animals and who spent longer hours in day care — showed far less allergy and asthma than their West German counterparts. Her revised conclusions are that early exposure to a variety of childhood microbial infections from brothers and sisters, other children, and animals somehow conditions the immune system such that it becomes more tolerant of potential allergens later in life.
She has since gone on to compare urban and rural communities all over Europe to show that children who grow up on traditional farms — where they come into contact from birth onward with livestock and their fodder, and they drink unpasteurized cow's milk — are protected from asthma, hay fever, and other allergic sensitizations. Von Mutius explains that throughout Switzerland, Austria, and Germany, where traditionally farming has been the main source of subsistence, most farms are involved in dairy production but may also keep other animals such as horses, pigs, and poultry. In addition, some farmers raise sheep and goats. Most farmers also grow grass, corn, and grain for fodder, and many farmhouses place fodder, people, and animals close together under one roof. Furthermore, women work in stables and barns before, during, and after pregnancy, and children a few days old are taken into stables, where mothers can look after them while working. Von Mutius stresses that several things seem of paramount importance for inducing tolerance to allergens. Exposure to microbes very early in life, even in utero, is vital, as is the diversity of animal species with which children come into contact and directly relates to the richness, in number and species diversity, of microbes.
Of all the autoimmune diseases, type 1 (or early-onset) diabetes is rapidly becoming the main scourge of life in the modern, hygienic Western world. Rates are set to double over the next decade among European children under the age of five. But the hot spot is Finland, with the highest rates of type 1 diabetes in the world. In an effort to find out why, Mikael Knip and colleagues, from the University of Helsinki, have been conducting an extraordinary population survey designed to disentangle genetics and environment in the causation of this life-threatening disease, where the body attacks the insulin-producing beta cells of the pancreas, causing high blood sugar. Although insulin treatment has saved lives and can stabilize the condition, many people suffer in the long-term from blindness and kidney damage.
Karelia, the old home of the Karelian people, is a large northern European landmass that used to belong to Finland but was partly ceded to Russia during World War II. As a result, the country has been partitioned into Russian and Finnish Karelia ever since. Although Russian and Finnish Karelians have the same genetic makeup, including the same susceptibilities to diabetes, the differences in their socioeconomic status and health could not be more stark. According to Knip, one of the steepest standard-of-living gradients in the world exists at the border between Russian and Finnish Karelia, with the latter having eight times the gross national product of the former. That is even greater than the difference between Mexico and the United States. Yet the incidence of type 1 diabetes, and a host of other autoimmune diseases, is far higher on the Finnish side. Finnish Karelians have six times the incidence of diabetes, five times the frequency of celiac disease, six times more thyroid autoimmunity, and much higher allergy levels than Russian Karelians.
Knip managed to get cooperation from the Russian medical authorities and collected medical data, stool samples, blood samples, and swabs from skin and nasal passages from thousands of children on both sides of the border. They found that Russian Karelians experienced much higher loads of microbial infection by the time they were twelve years old, and they tended to have more diverse colonies of microbes in the gut, where bacterial species known to play an active role in protecting and maintaining the gut lining were more common. They also noted biochemical evidence of much better regulated immune systems. Vitamin D deficiency has often been cited as a causative agent for type 1 diabetes, yet the researchers found that vitamin D levels were generally lower on the Russian and Estonian sides of the border than in Finland, with its much higher rates of the disease. To put it bluntly, Russian Karelians are poorer and dirtier than their counterparts on the Finnish side, but in terms of immune-related diseases, they are much healthier.
Does early exposure to a variety of bacteria, fungi, and helminths (which, traditionally, we would have inhaled, been infected by, or ingested from birth) work the same way as a childhood vaccination — such as the triple vaccine against measles, mumps, and scarlet fever — in stimulating immunity?
The original form of the hygiene hypothesis suggested that it did. The hypothesis began in the nineteenth century in the context of allergies. In 1873 Charles Harrison Blackley had noted that hay fever was associated with exposure to pollen but that farmers rarely experienced the condition. Later, in the 1980s, David Strachan, from St. George's Hospital in London, observed that having many older siblings correlated with a diminished risk of hay fever. The assumption was that this "grubby brother syndrome" of postnatal infections, rife in large families, protected against allergies. Strachan's theory, therefore, suggested that the immune system is primed by early exposure to develop acquired immunity to these diseases, just as occurs with childhood vaccination, and that our almost pathological obsession with hygiene in the West has removed these important stimuli. However, over the last ten years, a number of clues have arisen that suggest there is something much more profound going on.
The first clue lies in the huge length of evolutionary time that we humans have been exposed to certain bacteria, fungi, and helminths, compared to more modern pathogens like cholera and measles. George Armelagos, of Emory University, says that between the Paleolithic period, some 2.5 million years ago, up to about 10,000 years ago, our human ancestors would have been frequently in contact with saprophytic mycobacteria that abound in soil and decomposing vegetation. Paleolithic diets would likely have contained a billion times more of these non-pathogenic bacteria, like lactobacilli, than do diets today because of the unprocessed food our ancestors ate and because they stored food in the ground. They would also have been exposed to chronic infection by a variety of helminth worms. Molecular analysis of tapeworms, explains Armelagos, shows that they were ubiquitous parasites in human guts 160,000 years ago, before the out-of-Africa human exodus. A severe hookworm infection would cause poor health in any human, but it would be unlikely to kill its host. Once established, most helminthic infections would have been almost impossible to dislodge before the advent of modern drug treatment, and a hyperaggressive immune response against them would eventually do far more damage to the human host than to the worms — they had to be tolerated.
Only with the establishment of the first cities over six thousand years ago did humans transition to more crowded settlements, and a new raft of serious epidemic diseases — cholera, typhus, measles, mumps, smallpox, and many others — arose. These more modern diseases have not been around long enough to cause as much evolutionary change in us as more ancient infections. Helminths, fungi, mycobacteria, and commensal bacterial species have been part of our "disease-scape," as Armelagos coins it, for eons. We, and they, have had time to evolve together — we have coevolved. Graham Rook, from University College London and a doyen of the field, not surprisingly calls these organisms "old friends" and has renamed the hygiene hypothesis, having given it this deep-time coevolutionary dimension, the "Old Friends Hypothesis."
The second clue to this extraordinary coevolution lies in the fact that we need early exposure to these "old friend" microorganisms not merely to activate our immune systems but to establish, build, and mature our immune systems in the first place. There is no finer example of how we humans have coevolved with the bugs inside us than our interaction with bacteria during childbirth and the first crucial months of life.
During pregnancy there are important changes in the spectrum of bacterial species that populate the vagina. Both numbers of species and total numbers of bacteria decrease, but, within that, several species enrich their presence. Most of them are Lactobacillus — the genus of bacteria that commonly make up our probiotic yogurts, specifically L. crispatus, L. jensenii, L. iners, and L. johnsonii. They keep the mucosal lining of the vagina slightly acid, which protects it against pathogens, but they are also important members of a stable gut microbiota; when they are swept into a baby's mouth and involuntarily ingested as it pushes its way into the world, they will rapidly populate the infant gut, protecting it against pathogenic bacteria like Enterococcus. As the baby exits the vagina, it also becomes "infected" by bacteria present in traces of feces from its mother's anus. Particularly common species are a number of facultative anaerobes (bacteria that can operate in the presence of oxygen but also switch to fermentation in anaerobic conditions). They are able to establish themselves in the infant gut very early on because it contains oxygen, but they may also rapidly drive the gut toward a state where obligate anaerobes, including one of the principle "friendly" bacteria, Bifidobacterium, can thrive and take over.
The baby is born with a gut that is almost completely sterile and must be populated immediately with bacteria. If it is breast-fed, it starts to receive one of the most extraordinary products in the natural world. Human breast milk contains a complex array of fats and sugars — fast food — but it also contains immunoglobulin A, an antibody that protects the lining of the human gut and prevents pathogens from attacking and perforating it. It has also been estimated that a breast-fed infant receives over 100 million immune cells every day, including macrophages, neutrophils, and lymphocytes, together with a host of cytokines, chemokines, and colony-stimulating factors — molecules that signal between cells of the immune system and promote their growth. Over seven hundred species of bacteria have been found in human breast milk, many of them—like Lactococcus, Leuconostoc, and Lactobacillus — are capable of digesting milk sugars. Bifidobacteria, members of one of the most potent probiotic genera, are also predominant.
Some of the main solid components of breast milk are very complex, long-chain sugars called oligosaccharides. They weigh in at approximately ten grams in every liter, and human breast milk contains between ten and a hundred times more oligosaccharides than any other mammal milk tested. Yet the baby is totally incapable of digesting these molecules — it simply doesn't have the enzymes to do the job. Why human breast milk contains such high amounts of indigestible material has perplexed scientists for years, but it is now known that they are never destined for the baby at all. They are specifically produced to nourish the bifidobacteria that accompany them in breast milk. B. longum, for instance, has seven hundred unique genes that code for enzymes involved in breaking down oligosaccharides. These friendly bugs are parachuted into the baby's gut along with their own exclusive packed lunch. It gives bifidobacteria a head start in the rough-and-tumble of bacterial competition in the infant gut. Babies' guts rapidly digest and absorb all the simpler sugars so that practically the only sugar molecules left undigested by the time food reaches the large intestine will be the oligosaccharides. This means that all the bacterial species there are forced to compete for this sole source of carbon. It is the exclusive enzymatic tool kit of bifidobacteria that gives them the competitive edge. Breast-fed babies, in Darwinian terms, are ecological niches for bifidobacteria. Babies, in turn, have evolved to take full advantage of these friendly bugs. One Finnish study, for instance, has shown that bifidobacteria represent 90 percent of the gut microbiota in three-to five-day-old infants.
Probiotic bacteria like this have crucial functions. The immature, sterile gut of a newborn baby is at the mercy of aggressive pathogens, and its naive immune system is not yet developed and programmed to repel invaders. Probiotic species can act as decoy receptors to bamboozle pathogens and can prevent microbial pathogens from sticking to the gut wall. They seem to protect against necrotizing enterocolitis, for instance. They also help the establishment of a rich biofilm of probiotic-laden mucus that protects the gut, and they direct the development of a well-regulated immune system. The gut has its own immune system, distributed throughout the gut wall, known as the gut-associated lymphoid tissue, and probiotic species have been shown to be vital for its proper development. A number of experiments illustrate how potent this effect is. When oligosaccharides are experimentally added to infant-food preparations in controlled trials, they lower circulating levels of immunoglobulin E (IgE), which is an important marker for allergies. They also reduce atopic dermatitis, diarrhea, and upper-respiratory-tract infections. This baby/oligosaccharide-laden breast-milk/probiotic microbes system is one of the most elegant examples of coevolution that science has ever uncovered and has been essential for the survival of babies across evolutionary swathes of time.
At the risk of fashioning yet another stick to beat modern mothers over the head with, a number of studies point to the downside of both Cesarean section and bottle-feeding. Babies born via elective Cesarean section are more likely to be initially colonized by bacteria we commonly associate with the skin, and they experience lower initial colonization with "friendly" bacteria like Bifidobacterium. They take over five months to establish a stable, healthy microbiota. Bottle-fed babies show higher counts of clostridia, enterobacteria, Enterococcus, andBacteroides, which can all be opportunistic pathogens. Dr. Christine Cole Johnson, of the Henry Ford Hospital in Detroit, followed over one thousand babies out to two years old. C-section babies were five times more likely to develop allergies than babies who had been delivered normally. Research elsewhere adds celiac disease, obesity, type I diabetes, and even autism to the risk list. One obvious solution that might counter these early disadvantages would be to deliberately "infect" C-section babies with probiotics at birth. This is exactly what Maria Dominguez-Bello has done, together with colleagues from the United States and Puerto Rico. They have incubated gauze swabs for one hour in the vaginas of women electing for C-section and then painted the babies — first in the mouth, then on the face, and finally all over the rest of the body — with the swabs as soon as they are surgically removed from the womb. They have shown that the babies acquire their mothers' vaginal bacterial populations and show high bacterial species diversity in the gut after birth, which decreases somewhat once breastfeeding begins until it resembles the pattern of microbes present in mother's milk.
It also seems that vertical transmission from mother to baby via breast milk can pass on either good or bad traits directly from one generation to another. There is, for instance, a link between maternal obesity and microbial species diversity in breast milk. Obese mothers produce species-impoverished milk compared to leaner mothers. Their milk is poorer in beneficial probiotic species, and there are higher counts of potentially pathogenic bacteria like Staphylococcus and Streptococcus. There is evidence that if you receive the gut microbiota that is associated with obesity, you are at higher risk of becoming obese yourself and of developing insulin resistance.
Allergic mothers can pass their aberrant immune settings to their babies. The breast milk of allergic mothers contains lower counts of probiotic bifidobacteria. Although several months down the line, breast-fed and bottle-fed babies will stabilize around a similar mix of microbial species in their gut microbiota, it is the pattern of early colonization, in the first few days, that seems vital for the proper direction of the immune system.
How do "friendly" bacteria get into breast milk in the first place? According to Christophe Chassard, from Switzerland, there is a spectrum of bacterial species that is common to the mother's gut, her breast milk, and the baby's gut. It appears that bacteria are actively translocated from gut to breast by crossing the gut wall and entering mesenteric lymph nodes, from which they are transported through the lymphatic system to the mammary glands. This completes the cycle of vertical transmission from mother to baby. However, it does suggest that the baby is largely dependent upon the gut health of its mother. If she has a healthy gut microbiota, the baby will quickly benefit; but if she has a species-poor depleted microbiota, her baby will be similarly compromised.
It took a restaurant owner in London's Covent Garden to realize that human breast milk could be a big seller. He began marketing a brand of ice cream called "Baby Gaga," churned with generous donations from a prodigious nursing mother, Victoria Hiley, together with Madagascan vanilla pods and lemon zest. You could have it served, reported BBC News, with the optional extra of a rusk and a shot of Calpol (an infant pain reliever) or Bonjela teething gel. "Some people will hear about it and go ‘yuck,' but actually it's pure organic, free-range, and totally natural," said Hiley, while the restaurateur, Matt O'Conner, added: "No one's done anything interesting with ice cream in the last hundred years!" Of course it couldn't last. Westminster Council's food inspectors descended and demanded the refrigerators be cleared, as they couldn't be sure it was fit for human consumption!
Within a week or so after birth, the infant gut, originally sterile, has become colonized by up to 90 trillion microbes. The number of microbes in our guts eventually exceeds the total number of cells in our bodies by a factor of ten; our gut microbiota weighs considerably more than either our brain or liver; and the total number of microbial genes exceeds the number of genes in the human genome by a factor of a hundred. These microbes are not transit tourists, passing through, but long-term residents. Although it has long been recognized that much of this microbiota is benign, it was conventionally assumed that we, and the bugs inside us, simply dine at the same table. We passively allow the bugs to take a proportion of the nutrients flowing through our gut and give them somewhere warm and oxygen-free to live, while they feed us scraps from their digestion, like vitamins B, H, and K, which we cannot manufacture ourselves, and break down sugars like butyrate, which help our metabolism. But it has become clear that our relationship with our "old friends" goes far beyond such symbiosis. We have evolved such a close mutual interdependence with our microbiota that it no longer makes sense to distinguish between the two genomes. Scientists now refer to the existence of a meta-genome to represent the combined genomes of human and microbiota, a superorganism in which we humans are the junior partner and without which we could no longer exist. Scientists are asking two linked and fundamental questions: First, how can our bodies tell the difference between "old friends" (commensal bacteria, fungi, and intestinal worms) and dangerous pathogens, so that they tolerate the former while attacking the latter? Second, what happens to human health when these "old friends" are absent or depleted? This is what is allowing them to close in on the full story of the evolution of this mutual codependency, the development of the human immune system, and a pharmacology of the near future geared toward removing the vast epidemics of allergic and autoimmune disease currently ravaging Westernized societies.
To understand how the "old friends" manipulate our immune systems in order to masquerade as "self," we need a few basic facts about what the immune system is and how it is organized. We humans have two immune systems: the innate immune system, common to the whole animal kingdom — both invertebrates and vertebrates — and an adaptive immune system that only exists in vertebrates.
The innate immune system reacts to pathogens in a totally nonspecific way — it cannot offer long-lasting, protective immunity because it has no memory of past insults. It leaps into action whenever a pathogen is sensed, by producing an inflammatory reaction at the site of injury or infection. This serves to cordon off the infected area, dilates the surrounding blood vessels, and recruits a number of immune cells to the site to fight the infection. The inflammation is caused by cytokines — molecules that pass signals between immune cells — together with histamines and prostaglandins. The most important "pro-inflammatory" cytokines are tumor necrosis factor-alpha (TNF-α), interferon-gamma (interferon-γ), and interleukins 1, 6, 7, and 17. The innate immune system also includes the complement system of blood plasma proteins that helps or complements other immune factors by attacking and disrupting pathogens, labeling them so that they can be targeted by other cells, and recruiting more inflammatory factors to the battlefield.
The main cells of the innate immune system are collectively known as leukocytes (white blood cells), but there are many different types. Mast cells are common in all mucosal surfaces, like the lining of the gut and lungs, and release histamine, cytokines, and chemokines — a type of cytokine that acts as a signpost to other immune cells and directs them to the site of action. Most important are phagocytes, a group of active scavengers of pathogens that includes macrophages (literally "big eaters"); neutrophils, which come laden with killer chemicals like hydrogen peroxide, free radicals, and hypochlorite — nature's bleach; and dendritic cells, particularly common in the gut wall, whose main job is to engulf the foreign proteins that make up the coats of pathogenic bacteria and viruses and then "re-present" them on their cell surface in a form that can be recognized by cells of the adaptive immune system. They are one of the bridges between the innate and the adaptive immune systems.
Two key members of the adaptive immune system are types of white blood cells called lymphocytes. The first is the B cell, which is born in the bone marrow and travels while still immature to various lymphoid tissues such as the spleen, the lymph nodes, and the immunological tissues of the gut wall. As it matures, it will become capable of producing receptor molecules on its surface in response to the detected presence of antigens on invading microorganisms. These receptors are immunoglobulin molecules with a tip that is hypervariable. The genes responsible for this hypervariable region can be rapidly mutated to produce an almost infinite number of combinations, so an exact fit can be made to bind to any specific antigen protein on a pathogen. Thus, naive B cells can rapidly produce the correct lock for any antigen key, bind to it, and neutralize it. At this point the B cell either transmogrifies into a plasma cell and becomes a factory for that particular antibody, churning out millions of copies that free-float in the blood ready to bind to more antigen, or it becomes a memory B cell, which can survive for long periods in the body, remembering the antigen that activated it, and ready to immediately go into action whenever that antigen is sensed again. The second key member of the adaptive immune system is the T cell or T lymphocyte. Its precursor travels from the bone marrow to the thymus gland (hence the T) to complete several further stages of maturation.
One group of T cells, the effector T cells, present receptors on their cell membranes that are also hypervariable and can be manufactured to recognize any antigen complex detected on the coat of invading viruses and bacteria. They do not release antibodies into the blood but directly attack the invaders and destroy them. They can rapidly expand their population number to destroy all invaders bearing the same antigen signature, and a number of them linger on in the blood and lymph for as long as twenty years, a semipermanent memory. It is this adaptive immune system memory that gives us natural immunity in which subsequent infections with the same pathogen become milder because the memory B cells and T cells are like a gun already cocked. It is also the basis for vaccination, which introduces dead, inactivated, or attenuated pathogens, or the antigens stripped from the outer coat of one specific pathogenic organism, to establish permanent clones of memory lymphocytes, which can leap into action should the real pathogen arrive on the scene. So, adaptive immunity recognizes specific antigens, produces specific receptors and antibodies to fight them, remembers any specific pathogen, and rapidly deploys should it reappear. Both the B cells and the T cells go through an instruction process either in the bone marrow or thymus that weeds out those cells that react too strongly to "self" antigens, the protein markers for the host body's own cells. Thus, in the main, they are "taught" to tell the difference between "self" and "non-self" and act accordingly.
Crucial to an understanding of allergy and autoimmunity are two other types of T lymphocytes. First are the T helper cells (Th cells, for short). They are so called because they assist or help other white blood cells to mount a challenge to pathogens. Once activated by antigens "presented" by other cells, they produce cytokine-signaling molecules that can either damp down or inflame the immune response. The three types of Th cells we are concerned with are the Th1 cells, which are implicated in autoimmune diseases; the Th2 cells, which are implicated in the response to intestinal worms and common allergens (and are therefore implicated in allergy and all the atopic illnesses); and the Th17 cells, which are very potent defenders against a range of microbial invaders and whose cytokine products can be highly inflammatory — as such they are also frequently associated with a range of autoimmune diseases. Finally, we have the regulatory T cells (Tregs). Their job is to prevent the Th1 or Th2 cells, and other lethal T cells, from getting out of control by suppressing or regulating their activity. This ensures that the inflamed, cytotoxic immune response can be reined in once any particular pathogen has been dealt with. In the early days of the hygiene hypothesis, it was assumed that Th1 cells and Th2 cells were antagonistic, operating rather like a seesaw such that when Th1 cell populations were high, they inhibited the production of Th2 cells, thus preventing allergies. Conversely, it was thought that if the seesaw tipped the other way, so that Th2 populations were high, Th1 populations would be depressed and so autoimmunity would be prevented. However, it soon became obvious that in some autoimmune conditions, patients were also atopic (like the Johnson family) and that allergies and autoimmune disorders were simultaneously rising in all Western countries. This is what has led to a reinterpretation of immune system dynamics, and it is now believed that it is the regulatory T cells that operate as a master switch to turn down all effector T cells — both Th1 and Th2 — in the immune system.
"Old friends" organisms prevent our immune systems from treating them with suspicion and attacking them because they stimulate the production of Treg cells, reducing the attack squadrons of effector T cells, and thus bringing about a state of immune tolerance. For instance, June Round and Sarkis Mazmanian have investigated the important probiotic bacterium Bacteroides fragilis, which is prominent in the gut of most mammals. They have shown that B. fragilis produces a specific symbiotic molecule called polysaccharide A (PSA), which signals directly onto one of the main receptors of regulatory T cells. If you remove PSA from the B. fragilis, it is immediately unmasked and Th17 cells move in and prevent it from colonizing the gut.
There is a general principle at work here. A wide range of microbes and fungi needed to be tolerated by human immune systems because they were ubiquitous, over millions of years, in food and water, and consequently infected us. Similarly, helminthic parasites needed to be tolerated because, although not always harmless, once they became established in us, they were difficult to dislodge and an immunological attack on them would cause disproportionate collateral damage. For instance, sustained attempts by the immune system to destroy the larvae of the nematode parasite Brugia malayi cause lymphatic blockage and elephantiasis. Over millennia, a state of interdependence seems to have evolved. These commensal organisms, sharing the dining table of our guts, needed to regulate our immune systems so that they could live inside us without being attacked, and immune regulation became a necessity for us to prevent our immune systems from overreacting to these long-term residents in such a way that we self-harm.
But this has meant that, in a sense, we have handed the control of our immune systems over to the microbiota inside us. The downside is that while immune regulation works perfectly well when we have a rich assortment of "friendly" bacteria, fungi, and worms in our guts, things start unraveling very fast when they are no longer there. Our powerful immune systems, evolved to work in the presence of relatively harmless endoparasites, lose their brakes and won't switch off, resulting in the chronic inflammation that leads to the modern plagues of allergic and autoimmune illness.
Matteo Fumagalli, now at the University of California, Berkeley, hypothesized that parasitic worms would have exerted very significant selection pressure on humans throughout their history. Even today, he says, over 2 billion humans are infected with parasitic worms that cause widespread childhood morbidity. They stunt growth, leave individuals open to other infections, and cause preterm births, low birth weights, and maternal mortality. It was ever thus. Fumagalli argues that the selection pressure exerted by worms would have been much greater than that exerted by bacteria and viruses or climate, and that it should be possible to see it reflected in our genomes, particularly among genes that are implicated in immune responses.
Using data from the Human Genome Diversity Project, he sampled 950 individuals from all over the world and correlated the occurrence of gene mutations with species diversity of helminths in the area of the world from which each individual came. Nearly one-third had at least one gene mutation significantly associated with helminth diversity, and in all they counted over eight hundred gene mutations. Many of these genes were involved in regulatory T cell function and in the activation of the macrophages of the innate immune system. Others were involved in the cytokines produced by the Th2 cells that are mobilized against helminth infection. Importantly, the gene variants they found give us a huge clue into the nature of the love-hate relationship between humans and helminths. While many of the genes that had evolved under pressure from helminths were associated with aggressive, pro-inflammatory responses geared to fighting helminth infections, other genes worked in the opposite direction by stimulating immune tolerance through regulatory T cells, anti-inflammatory cytokines, and other molecules that inhibit immune responses. Helminths are the masters of immune regulation. This is why they can persist in any host for a long time and persist as major human infections over millennia. Graham Rook sees the relationship between them and us as a game of chess, or as a dynamic tension between parasite and host immunity. He suggests that where helminth load was particularly high, there would have been selection for more pro-inflammatory gene variants either to counter the powerful immunoregulatory action of the helminths or to keep the immune system effective against other viral and bacterial insults in the face of helminthic regulation. This dynamic equilibrium is shattered as soon as the helminths are eradicated, leading to the inflammatory overshoot we see today in allergic and autoimmune disease.
Jim Turk is a biological security officer on the University of Wisconsin campus. It's his job to make sure all the laboratories handling pathogenic or recombinant organisms are safe and contained. Being in good shape had always been important to him — he used to run marathons. Then, in spring 2005, his speech suddenly became slurred. His wife feared he had had a stroke and made him go to the doctor, who ran some tests but could find nothing wrong. "You're overdoing it," the doctor said. "You're working full-time, you're going to university all the time, you've got a young family. You're tired and stressed." Jim noticed occasional balance problems and some numbness in his lower legs, but, because of his doctor's assurances, he brushed them aside. Then, in February 2008, he began serious indoor training for another marathon. After three or four minutes of hard running in the hot arena, he remembers, "I would lose control. I would be grabbing at railings. I'd be stumbling around. Of course being kind of dumb or in denial, I decided I was going to keep going back. ‘I must be in worse shape than I thought,' I told myself. And so I went back the next day and the next day, and it was always the same thing until one day I fell flat on my face." Still in denial, he busily set about coaching his son's baseball team only to notice that he was moving his legs wider and wider apart just to balance himself upright on the baseline. Bending down sent his head spinning. Back to the doctor he went. High blood pressure was quickly ruled out — he had the heart of a teenager — and then a neurology referral and an MRI scan made everything horribly clear. "My brain was peppered with these plaques. I had a lot of them — twenty or so." Another MRI scan showed that the plaques extended right down into his spine, and he was diagnosed with relapsing-remitting (early stage) multiple sclerosis in August 2008. Jim had joined over 2 million people worldwide who have MS. They mostly live in Westernized countries and tend to be young — mean age of onset is twenty-five. Although over fifty genes, lack of sunlight, vitamin D, viruses, smoking, and affluence have all been implicated, the strong inverse correlation between helminth infections and MS is striking, as is the fact that T cells are often dysregulated in people with MS. Each MS plaque is a small area of inflammation and tissue damage where the myelin sheath that coats and insulates the nerve fibers is destroyed. In relapsing-remitting MS, there are typically five to ten new plaques formed every year, though only one in ten, on average, will hit a critical area of the central nervous system like the optic nerve, cerebellum, or tracts of sensory nerve fibers.
A few nights after his diagnosis, Jim and his wife were watching a weekly feature on multiple sclerosis on their local television channel: "Dr. Fleming, from the University of Wisconsin, was on and he was talking about his new HINT (helminth-induced immunomodulatory therapy) trial using whipworm eggs. I already knew a little bit about the hygiene hypothesis, so I was immediately intrigued but, having a science background, realized it might be tough to get people to line up to swallow whipworm eggs just because of the way it sounds. I knew you wouldn't be able to see anything — the eggs are microscopic. But if you tell people they're drinking worm eggs . . ." Jim contacted Fleming's group the next day and became the first volunteer to be enrolled in the study. He took the worm eggs for three months and then had a sixty-day washout period.
John Fleming's interest in the potential of Trichuris suis had been spiked by the success of Joel Weinstock and his colleagues with inflammatory bowel diseases, where, as we discovered earlier, some 79 percent of patients who received the whipworm got better in a stage 1 trial. Although the pig whipworm eggs do not go through their full life-cycle in humans (they may not even hatch) and they last only a few weeks at most before being eliminated in feces, successive treatments seem to give the whipworm enough transient presence in the gut to modulate the immune system. Weinstock believes they may do this directly, by secreting immune-regulating molecules, or indirectly, by affecting the composition of the gut microbiota, which then does the job. Fleming says that the evidence for a failure in immune regulation being at the heart of MS is strong. Such regulation, he explains, is achieved by interaction between regulatory T cells, dendritic cells, regulatory B cells, cytokines, and other effector cell types. All can be disordered in MS sufferers, and over twenty mechanisms by which helminths influence immune regulation have so far been discovered.
Fleming had also taken notice of Jorge Correale's work in Argentina that reported much reduced severity in symptoms of MS in those patients with an intestinal worm infection. Correale had also shown that the worm effect was very specific to MS because the regulatory T cells produced in worm-infested patients were specific to myelin peptide, meaning they were protective of the myelin sheathing around nerves. Fleming obtained permission for the small initial phase 1 trial that involved only five newly diagnosed patients who had yet to receive any form of treatment for MS. Although the trial was deliberately short-lived and testing mainly for safety rather than efficacy, Fleming noticed a substantial reduction in new brain lesions in all five volunteers while they were taking the whipworm eggs, whereas the lesions increased in number once treatment had stopped.
Jim's MS is now being treated with conventional medication, and he tries to keep his illness at bay through a combination of careful diet and exercise. "In fact, if I didn't have MS, I'd probably be in the best shape of my life!" He has learned to contain his symptoms, especially his speech impediment, so well that most of his colleagues are totally unaware of his condition. But playing sports with the kids is a thing of the past. "As my kids grow up — they're both boys, thirteen and nine — I'd like to be able to play football in the backyard with them, shoot baskets, or go for long bike rides. I can still throw the ball a little — I just can't run with it. And I'd love to be able to go out for runs. I haven't been able to do that in six years. That's what I wish I could do the most."
Our resident gut microbiota — the mass of over two thousand bacterial species identified as frequent, long-term inhabitants inside us — is extremely complex. Our relationship with them is so close and intertwined that many of the metabolic signatures that can be identified in human blood, sweat, and urine actually come from our commensal bacteria, not us. When we look at any individual's response to a particular drug treatment, it may be that what we are looking at is the reaction of our gut microbe colonies, not our own bodies. Only vertebrates have such rich, enduring microbial colonies. All invertebrates have tiny gut colonies, sometimes only several species, and these are often transient. There is another interesting difference between invertebrates and vertebrates. Invertebrates do not have an adaptive immune system; they only have the more primitive innate immune system.
That observation has prompted Margaret McFall-Ngai, a medical microbiologist from the University of Wisconsin, to stand modern immunology on its head. She argues that the adaptive immune system evolved not only to protect us against environmental pathogens, but also to police the resident microbial community inside us. The pioneering microbiological research of the nineteenth century, for which we have to thank Koch and
Pasteur, was done in the context of human disease. It was set on its course only to view microbes as invaders of the human body, capable of causing infection. Pathogenic bacteria and viruses can mutate far faster than we can, and rapidly change the antigenic markers on their outside coat by which our immune system recognizes them as the enemy. We needed, states conventional immunology, an adaptive immune system with long-term memory and the ability to generate an almost infinite variety of matching antibodies to counter them.
While it is true that we could not effectively counter pathogenic infections without an adaptive immune system, it is also true, says McFall-Ngai, that the complexity of the pathogenic environment is dwarfed by the complexity of the microbial communities within us. To put this in perspective, it has been estimated that only twenty-five infectious diseases account for the vast majority of human death and disability, and ten of these could only have arisen after urbanization some six thousand years ago because they depend on human-to-human transmission. They could not have survived when humans existed in small, widely dispersed, roaming communities before the advent of agriculture and animal husbandry. Our gut microbiota can consist of thousands of species, and, although they may be friendly most of the time, bacteria are a perfidious bunch; they are opportunistic and can easily mutate to turn from being benign to being pathogenic when it suits them, for instance, when damage to the gut wall makes it leaky and they can no longer be safely corralled in the gut lumen.
So, not only has our microbiota been around far longer than most pathogens, but it is far richer in number and species diversity than the pathogenic environment. Without it, the "gut police" of our adaptive immune systems simply would not develop into a highly plastic system capable of discriminating between friendly microbes and the pathogens lurking among them, or detecting friendly bacteria who have turned rogue. According to Sarkis Mazmanian, our microbiota presents a challenge to the adaptive immune system because it contains an enormous foreign antigenic burden, which must be either ignored or tolerated to maintain health. In turn, it is in our microbiota's self-interest to maintain the health of its host. It is humbling to remind ourselves that we are simply an attractive niche or home for these microbes — one that they have fashioned to suit themselves. We, and our microbiota, have coevolved to work together to repel pathogens, because it is in our mutual interest.
For example, it has recently been shown that mice in the throes of a systemic bacterial infection switch on production of a sugar that specifically favors the population growth of friendly bacteria in their guts, which rally round to help repel ensuing infections. Gérard Eberl, of the Institut Pasteur, believes that in this superorganism of human and microbes, the immune system is never at rest; it is like a spring. The more the microbes colonize the niches within us or behave like pathogens, the stronger they pull the spring of immunity, he says, and that stronger spring of immunity pushes the microbes back. The immune system is therefore always under tension, the tension required to maintain homeostasis — the status quo. For instance, if we take away our gut microbiota with antibiotics, we can become susceptible to infection by Enterococcus. This is because it is the friendly bacteria that make our gut wall produce the toxic antibacterial peptides that normally fight this pathogen. Too weak an immune system, explains Eberl, leaves the superorganism vulnerable to old friends going bad and turning into opportunistic pathogens; too strong an immune system destabilizes our microbiota, and we progress into autoimmune disease.
McFall-Ngai says it is high time that we started viewing the microbes inside us as a whole organ, similar to but much more complex than the heart, liver, or kidney. In its complexity, it more closely resembles the brain. The brain is composed of over 80 billion neurons, massively interconnected; the microbiota is composed of over 80 trillion organisms massively communicating with one another via signaling molecules. Both have a memory, both can learn from experience, and both can anticipate future uncertainties. The gut has been called the "second brain" and has its own dedicated nervous system embedded throughout the gut wall. It is becoming increasingly clear that our gut microbes can communicate directly with our brains, and they are implicated in brain development, brain chemistry, behavior, and mental illness. They produce hundreds of neurochemicals, including most of the body's supply of serotonin, and there is two-way communication such that the mix of bacterial species in the gut can influence the brain and vice versa.
Much of the research that demonstrates this uses mice as the experimental model. For instance, Premysl Bercik has compared two strains of mice — one timid, one bold and adventurous. A group of animals from each strain were cultivated germ-free. The germ-free timid mice then had their guts inoculated with the gut contents of bold mice that had been reared normally. Germ-free mice from the bold strain were conversely inoculated with the gut contents of normally reared timid mice. Their behavior immediately switched. Timid mice became bold and bold mice, timid. John Bienenstock gave a timid strain of mice a broth heavily laced with a popular probiotic bacterium, Lactobacillus rhamnosus. After twenty-eight days, they were more willing to enter a maze than controls and less likely to give up and float on a forced swim test. The activity of stress hormones in the brain was reduced. Similarly, rearing germ-free mice in a confined area stresses them and raises the activity of the hypothalamus-pituitary-adrenal axis. This pushes up levels of corticosterone and adrenocorticotrophin, two stress hormones. But all this activity subsided after the mice were inoculated with another common probiotic, Bifidobacterium infantis. The reverse also happens. Michael Bailey has found that macaque infants from mothers who were stressed with loud noises during pregnancy had lower counts of friendly bacteria like Lactobacillus and Bifidobacterium in their guts, while another researcher found fewer lactobacilli in the stools of college students during exam week.
How can the bugs in our guts "talk" to the brain and vice versa? What is the link? In a very recent paper, Emeran Mayer and Kirsten Tillisch report in their fMRI brain-imaging study of the effects of probiotic bacteria on mood and brain activity in a group of healthy, normal female human volunteers. One group of women was untreated, while the other group was given fermented probiotic drinking yogurt twice a day for four weeks. They were imaged before and after treatment: while they were undertaking a task where they had to look at a range of emotional faces and while simply resting.
The researchers identified a pathway into the brain involving a tract of nerve fibers in the brain stem called the nucleus tractus solitarius. This receives inputs from the vagus nerve that innervates the gut. From this brain-stem nucleus, circuits were activated that ramified into higher brain centers, including the amygdala (the fear or emotional center of the brain), the insula, and the anterior cingulate cortex, all regions that are involved in processing emotional information. In those volunteers who had taken the yogurt, there was reduced activity in these circuits, suggesting decreased arousal or anxiety associated with the tasks. These women had calmer emotional reactions. Although the results should be interpreted with extreme caution, a reasonable working hypothesis is that the probiotic bacteria in the gut were able to signal to the brain using the vagus nerve — literally communicating gut feelings.
A recent paper by Joe Alcock, Carlo Maley, and Athena Aktipis has pulled together many sources of evidence suggesting that the bugs in our guts can influence what we eat, producing cravings for foods that will give them a competitive edge in the large intestine and inducing profound states of unease or dissatisfaction in us until we are sensually rewarded by particular foodstuffs, like chocolate, that will equally reward the bacteria with the food they hanker after. Our gut bacteria are using the vagus nerve to manipulate our behavior. This raises the intriguing possibility that by changing the species composition of our microbiota, we might be able to change eating habits and perhaps even ward off obesity.
What happens when the brain malfunctions? Inflammation and gut pathology are both involved in autism, for instance. Children with autism frequently have evidence of inflammation in their brains, and there is growing evidence that this inflammatory state is transmitted from their mothers during pregnancy. Alan Brown, professor of clinical psychiatry at Columbia University, has studied data from the Finnish Cohort Study, where 1.6 million blood samples were collected from over 800,000 women during their pregnancy. They matched the levels of C-reactive protein (CRP; a blood indicator of inflammation) with the risk of autism in their children. The risk was increased by 43 percent among mothers in the top twentieth percentile of CRP levels, and by 80 percent among women in the top tenth percentile. Since CRP is also elevated in the immune response to infections, this suggests that infections during pregnancy in some mothers, as well as inflammation caused by autoimmunity, can communicate an inflammatory state to their fetuses through the placenta. There is further evidence from Danish population-wide studies that autism risk is raised by 350 percent in babies born to mothers with celiac disease, and 80 percent in mothers who suffer from rheumatoid arthritis. Eric Hollander, the autism expert who treats Lawrence Johnson, suggests that something as simple as influenza in the mother-to-be can feed-forward an inflammatory state from mother to fetus via pro-inflammatory cytokines. These are similar responses to those triggered in fetuses when their mothers have lupus, an inflammatory autoimmune disorder that causes fever, swollen joints, and skin rashes. Autistic children may also inherit, in the genetic sense, a hyperactive immune system that leaves them vulnerable to developing autoimmune disease.
About 70 percent of children with autistic spectrum disorder have severe bowel irritation. They can suffer from diarrhea and painful abdominal distention, and this is related to their irritability, aggression, and self-harm. Endoscopic examination often reveals an inflammatory pathology very similar to Crohn's disease and ulcerative colitis. Stephen Walker, from the Wake Forest Institute, compared patterns of gene expression in biopsy material collected from the guts of autistic children with irritated bowels and adults with IBD. While there were significant differences in the pattern of genes affected, there was also significant overlap in a number of genes that were either turned up or tuned down in both conditions. This suggests autistic children with irritated guts, and non-autistic adults with bowel disease, both suffer from autoimmunity. Mouse models of autism show that helper T cells are made permanently hyperresponsive by maternal infection during pregnancy, and regulatory T cells are reduced in number.
What about depression? It cannot be said too clearly that depression is not an inflammatory disorder per se. Many patients presenting with depression do not have strong evidence of inflammation, and many individuals have high levels of inflammation markers in their blood without becoming depressed. Having said that, there is clearly a major subgroup of individuals who are more prone to reacting to a background inflamed state by becoming depressed. There is a fascinating example of the interplay between inflammation and depression in the effect of interferon-alpha (IFN-α) — which is a potent inflammatory cytokine — when it is used as a therapy for hepatitis C or cancer. At high doses, fully 50 percent of patients will develop major depression within three months of commencing therapy. Downstream, IFN-α induces a cascade of other inflammatory cytokines, like interleukin-6 and tumor necrosis factor-alpha (TNF-α), which also correlate with depression.
But is the converse true? If you remove the inflammatory cytokines, does the depression lift? One study examined patients who were being treated for Crohn's disease with a monoclonal antibody called infliximab. A proportion of them were also clinically depressed. The infliximab removed depressive symptoms but only in those individuals who had high levels of C-reactive protein in the bloodstream, indicating a state of high inflammation. This was because infliximab is a potent TNF-α agonist, so it is likely that while curing the Crohn's disease, it was also neutralizing the cytokine that stimulated the inflammation and therefore alleviated the depression.
In susceptible individuals, depression is only one of a number of disorders that seem to be caused by chronic low-grade inflammation. These include cardiovascular disease, stroke, diabetes, cancer, and dementia. Moderate increases in levels of chronic inflammation are enough to predict the future development of all these modern disease states in presently asymptomatic individuals. In the Whitehall study of UK civil servants, circulating
levels of C-reactive protein and interleukin-6 were inversely correlated with employment grade, implying that the lower the pecking order, the higher the background inflammation became. The psychologist Andrew Steptoe used this gradient to successfully predict occurrence of depression twelve years down the line. Further studies show that depressed individuals with histories of early life trauma or neglect release more interleukin-6 in response to stress tests. It could be that it is not simply the stress of modern Westernized lifestyles and workplaces that causes inflammation and thence depression, but a relative lack of immune regulation in Westernized societies that allows inflammatory cytokines to run amok in reaction to them. In which case, how, specifically, could the "old friends" hypothesis come into play in this scenario?
Tom McDade, of Northwestern University, has been comparing populations in developing countries with U.S. populations in an attempt to disentangle infection, inflammation, stress, depression, and morbidity. He notes that levels of C-reactive protein are transiently high in a tribe of Amazonian Ecuadorian Indians and correspond to frequent bouts of infection. But as soon as the infection subsides, so do the CRP levels. Their profile is a series of peaks and troughs, whereas in the United States, levels of CRP are stable and high even in the absence of high rates of infectious disease. This chronic, persistent inflammation indicates poor immunoregulation. McDade looked at a rural population in Cebu, in the Philippines. He measured the levels of microbial diversity in and around each village house by examining animal feces and measured the frequency of childhood diarrhea and the numbers of births during the dry season when infectious loads were highest. All these factors predicted low CRP levels in adulthood and reflected early high microbial exposure. McDade then looked at the effects on children of separation. All children, as you would expect, were distressed at losing their mothers, but that stress did not raise their CRP levels as long as they came from a home typified by a high level of microbial diversity. It did not even rise if they subjectively felt psychologically disturbed by the pain of separation. So, in rural Philippines, thanks to adequate childhood exposure to "old friends" microorganisms, temporary depression, social stress, and unhappiness never led to damaging inflammation. McDade also found that concentrations of the pro-inflammatory cytokine interleukin-6 were generally low in Philippine populations, while levels of the anti-inflammatory cytokine interleukin-10 were exceptionally high. Obese women in the United States generally have high levels of IL-6, but in the Philippines, women with similar waistlines don't. When they looked at men with high skin-fold thickness in the Philippines, it was not associated with high CRP levels, whereas this is exactly what you expect to find in the United States. There was, he concluded, protection against inflammation on all fronts in which the early microbial environment of upbringing seemed key.
We may be entering an era where microbiology and immunology, specifically the "old friends" hypothesis, begin to make a real impact on public health policy. For instance, Martin Blaser is deeply worried about the overuse of antibiotics. We are all aware of the increasing dangers of multiple antibiotic resistance because it is giving rise to a race of dangerous "super-bugs" that are becoming almost impossible to treat. But regular broad-spectrum antibiotic treatment is also killing off the friendly and useful commensal bacteria inside us with disastrous results. By the time they reach the age of eighteen, Blaser points out, American children, on average, will have received between ten and twenty courses of antibiotics, killing off friend and foe alike. In some cases, he says, our microbiota never recovers and we are fueling the dramatic increases we see in type 1 diabetes, obesity, inflammatory bowel disease, allergies, and asthma. Occurrence of IBD, for instance, rises with the number of antibiotic courses taken. Worse still is the industrial-scale administration of antibiotics to farm animals purely to assist them to put on weight. Antibiotics are routinely given to nearly one-half of all women in pregnancy in the United States, and since babies acquire their gut microbes from their mothers, each generation could be beginning life with a smaller endowment of friendly microbes than the last, and so on ad calamitas.
A scary scenario of what that calamity might be has recently arrived from Sven Pettersson, of the Karolinska Institutet in Stockholm. It is known, explains Pettersson, that there is an intestinal barrier to prevent the trillions of bugs in our guts from escaping into the body. That barrier is actually created and maintained by the friendly bacteria inside us. In experiments on mice, he has discovered that the gut microbiota exert a similar control on the impermeable blood-brain barrier that protects the brain from insult by a huge variety of molecules and microorganisms. Baby mice born from germ-free mothers had leakier blood-brain barriers that persisted throughout life. Although this research has yet to be transferred to humans, the implications are extremely worrying. If a depleted gut microbiota in the mother can lead to a defective blood-brain barrier in the baby, then the proper development of the brain, and later protection of the brain, might become heavily compromised. This may cause us to become much more wary of routine antibiotic treatment of pregnant mothers and Cesarean section for the delivery of their babies, because we already know that both interventions deplete the gut microbiota that babies inherit from their mothers.
A massive amount of research now amply demonstrates the degree to which a benign microbiota inside us can protect our health. But one researcher believes that the beneficial effect of microbes can extend to both the urban or rural environment around us. Ilkka Hanski, from the University of Helsinki, thinks that microbiology — specifically the "old friends" hypothesis — should be taken into account in town planning, particularly in considering the importance of green spaces. He has recently reported a significant relationship between skin allergies, vegetation, and land use in a heterogeneous group of 118 Finnish teenagers chosen to represent a range of living environments from towns to villages to farms. He took skin swabs to look at the diversity of skin bacteria, a skin allergy test to check levels of atopy, and measured land use and plant cover in the immediate vicinity of their houses and up to three kilometers away. He found a strong relationship between atopy and a group of bacteria called gammaproteobacteria, which were significantly less diverse in atopic individuals. He then went on to measure levels of the anti-inflammatory cytokine interleukin-10 in the blood and found that one gammaproteobacterium, Acinetobacter, was strongly linked to high levels of IL-10 in the blood of healthy individuals, but not in allergic individuals.
These "protective" bacteria are commonly found on plants and pollen, as well as in soil, which is why Hanski found such a strong association between diversity of skin bacteria, lack of atopy, and the richness of the surrounding vegetation, particularly in the less common species of flowering plants. The teenagers would have picked up the bacteria through contact with soil and vegetation, or through pollen or wind transfer. More and more of us worldwide are moving to cities where open green spaces may be few in number or completely absent. If we are dependent upon certain bacteria to encourage high levels of anti-inflammatory cytokines in our blood, thereby inducing immune tolerance, and if those bacteria, in turn, are dependent upon the richness of vegetation, then the importance of green spaces goes far beyond a feel-good factor and is at the root of allergic conditions and public health in general, where, as Hanski notes, they may have profound consequences.
In a similar vein, we have seen how Mikael Knip's Karelian study has identified the microbiota to be an important factor in type 1 diabetes. Russian Karelian children not only had a more diverse microbiota but higher levels of regulatory T cells in their blood. His continuing research hopes to more specifically identify the species of microbe most important for protection against autoimmune and allergic disease so that he can design a medical intervention in the near future for children at risk. Other allied metabolic disorders steeply on the rise, like type 2 diabetes and the obesity pandemic, have also been shown to relate to immune dysregulation and loss of microbial diversity in the gut.
There is a huge and growing literature on the effects of negative life events and loneliness or social isolation on our future health. Although many of these long-term epidemiological surveys show correlations between inflammation, stress, social isolation, and socioeconomic status, only one or two to date have looked at gut microbial diversity. Graham Rook, however, would bet the farm that, had they done so, they would have found that those individuals who showed poor resistance to the vicissitudes of Western life and had chronic high levels of inflammation, and mental or physical morbidity, would typically show reduced diversity in their gut microbiota and consequently compromised immunoregulation. The "old friends" hypothesis, thinks Rook, could be the "missing variable" in all these public health studies.
A few examples will suffice to see how wide-ranging these implications could be because we are clearly looking at a cradle-to-grave phenomenon. Per Gustafsson, for instance, has examined the way that isolation and subjective feelings of unpopularity at school adversely affect health several decades down the line, correlating with psychiatric problems, cardiovascular problems, and diabetes. Gregory Miller and Steve Cole have gone further and specifically linked childhood stress and adversity to chronic low-grade inflammation. Using data from a large group of Vancouver adolescents, they show that depression and inflammation (as measured by high circulating levels of C-reactive protein and interleukin-6) co-occur, but only in those individuals who had suffered childhood adversity. CRP levels linger on after the depressive episode has ended, which may make such children vulnerable in the long term to persistent mood disorders, cardiovascular disease, diabetes, and autoimmunity, they say.
Bruce McEwan introduces the idea of "allostatic load" — the cumulative wear and tear on the organism as it tries to adapt to life's demands. They show how the stresses associated with gradients of socioeconomic status show up as chronic inflammation that upsets neuroendocrine function and can lead to heart problems, osteoporosis, metabolic disorders like diabetes, and cognitive decline. W. Thomas Boyce and Kathleen Ziol-Guest are even more explicit about the relationship between childhood poverty and adult disease. Although factors like diet and nutrition are also important, the chronic inflammatory processes set running by childhood exposure to negative life events are an additional common pathway to several chronic morbidities, they explain. Chronic inflammation is controlled by events in the brain, interplay between the hypothalamus, pituitary gland, and adrenal gland, and the cellular immune system where T lymphocytes are encouraged to differentiate into Th1 and Th2 cells, which can result in widespread tissue damage if inflammatory conditions persist. Low childhood socioeconomic status is associated with higher blood levels of C-reactive protein, cytokine IL-6, and the other pro-inflammatory cytokine, TNF-α, placing such children at greater risk of developing inflammatory diseases such as atherosclerosis, autoimmune disorders, and cancer. This is precisely where Tom McDade's research enters the picture by measuring these early environmental stressors, specifically the loss of the mother, in Filipino children and showing that the crucial protective factor against early insults that might normally lead to later metabolic and mental illness was microbial exposure in infancy.
We shouldn't really need the "old friends" hypothesis to tell us that we ought to provide better quality care for our aged population in residential homes. But a recent survey by Marcus Claesson and his colleagues at the University of Cork graphically shows how toxic the environment of care homes can be to its residents. Claesson identified 178 elderly people, with a mean age of 78, in southern Ireland and divided them into those who were still living in the community, those who were in a hospital for short-term rehabilitation, and those in long-term residential care. He took fecal microbial samples, dietary information, and measures of their immune status. He found that the gut microbiota in those individuals in residential care was much less diverse than in those living in a community setting, and this correlated with high and persistent inflammation and increased frailty. As we get old, our teeth don't work so well, we produce less saliva, our digestion suffers, and we get more constipated. All this damages our gut microbes. The combination of a poor, bland diet (and probably social isolation) in nursing homes simply makes the situation much worse, taking its toll on the microbiota, driving chronic inflammation, accelerating aging, and deteriorating health. Given a rapidly expanding aged population in Western countries, says Claesson, dietary interventions to prevent this accelerated morbidity and premature death should become a priority.
Across the world, researchers are trying to turn many of the microorganisms identified by the "old friends" hypothesis into mainstream medicine. You can tell when a new applied science is in its infancy because it is often populated with self-experimenting pioneers, people who try techniques out on themselves before subjecting patients to them, or people who venture into unproven therapies out of desperation. People like David Pritchard, from the University of Nottingham, who, while doing field research in Indonesia, became intrigued by the observation that individuals infected with hookworm seemed protected against allergic diseases. He later deliberately infected himself with hookworm larvae, via a scratch in the skin, in order to prove to his satisfaction that the negative effects of hookworm infection were tolerable compared with the hoped-for benefits in immunoregulation.
He is now connected with a major medical trial at Nottingham that aims to see if hookworms can moderate disease progression in multiple sclerosis. In 2004 a young man, who will remain anonymous, traveled to Thailand to deliberately infect himself with human whipworm eggs, procured from the feces of an infected girl, to see if they could cure an ulcerative colitis that was so resistant to cyclosporine treatment that he was in immediate danger of having his entire colon removed and replaced with a colostomy bag. He had previously approached Joel Weinstock for treatment with Trichuris suis, but Weinstock had to refuse him on ethical grounds. Within three months of self-treatment, bowels that had once produced over a dozen bloody movements a day were back to normal. It appeared that the worms were inducing high amounts of interleukin-22, which is important for healing the gut mucosa. He is now one of a number of volunteers helping P'ng Loke, at New York University, to investigate the effects of helminths on inflammatory bowel disease.
The road from bright biological insights to tried-and-tested pharmaceuticals is often a rocky one. In order for any new drug to reach the market, it has to be rigorously tested in large, randomized, double-blind trials that are capable of accounting for any placebo effect and dissociating it from any true efficacy of the drug being tested.
By this measure, the recent slew of medical trials for pig whipworm eggs has so far proved disappointing. A large trial of whipworm eggs among sufferers of Crohn's disease was recently halted because it could find no efficacy, and John Fleming's phase II trial of whipworm eggs for multiple sclerosis has also failed. The problem may be because the whipworm species chosen for both these trials is not well suited to human beings. Trichuris suis is a parasite of pigs — not humans. It may not even multiply in the human gut, and so the "infection" is flushed out in the feces within a couple of weeks. This is why Stewart Johnson's son has to continually take eggs to stand a chance of the treatment working. Because humans are not the natural home of Trichuris suis, the eggs may not be regulating human immune systems as effectively as would a human-specific helminth species.
We may, says Graham Rook, the pioneer of the "old friends" hypothesis, be using the wrong worm on the wrong people. The reason why Jorge Correale finds that tapeworm infection causes remission in symptoms among his multiple sclerosis sufferers in Argentina may be because, due to the endemic nature of tapeworm in parts of South America, they had asymptomatic tapeworm infections during early childhood when their immune systems were maturing. Their immune systems will have become calibrated by this previous early insult thanks to developmental or epigenetic effects by which the helminths regulate key immune system genes. This previous exposure would not be shared by Crohn's disease and multiple sclerosis sufferers in North America.
Meanwhile, Eric Hollander has achieved modest success with a small trial of Trichuris suis in adult patients with autism. There were reductions in scores on several psychological tests that measure autistic symptoms — reductions that hovered just below statistical significance. They were less likely to have temper tantrums or "act out," says Hollander, and were less compulsive and more tolerant of change. Hollander is now conducting a larger trial of Trichuris with children and younger adults with autism. Whatever the final results from Hollander, because of the acknowledged link between autism, poorly regulated immune systems, and gastrointestinal discomfort, Rosa Krajmalnik-Brown, at Arizona State University, examined the microbiota of autistic children and found them to be generally less diverse and lacking several important "friendly" bacterial species compared with the gut microbiota of normal children. She is currently running a trial with autistic children where she introduces fecal transplants taken from normal individuals to repopulate their guts with a diverse and protective microbiota. Fecal transplants could be a more effective way of rapidly repopulating the gut than orally taken probiotics, which may have a limited effect because when you take a few pills of probiotics or drink a glass of fermented probiotic yogurt, you are trying to improve the gut microbiota by introducing a few hundred million microorganisms by mouth into a lower gut population of scores of trillion. Also, common probiotic products abound withlactobacilli and bifidobacteria that, while vital in establishing infant immune systems, are only minority players in adult guts. We need new probiotics.
All these "old friends" researchers look forward to the near future when their efforts may lead to a "new pharma." Helminth researchers realize that we cannot go on forever feeding whipworm eggs to people or introducing hookworm larvae under the skin. Sarkis Mazmanian has shown the way through his work with the bacterium Bacteroides fragilis. He found that he could dispense with the whole intact bacterium because the polysaccharide A molecules extracted from them were capable of regulating the immune system. In a similar fashion, Joel Weinstock continues to research the mechanisms that helminths use to evade our immune systems. He hopes, in the future, to be able to isolate the molecules they use to do this and thus lay the foundation of a new evolution-inspired drug therapy for this range of illnesses.
This approach would greatly please Stewart Johnson, whose heroic, and ultimately successful, research saved his son Lawrence from the twilight world of a lifetime in institutionalized care. As Stewart says: "I did not see this coming. I did not expect this to work. I was just satisfying the scientist in me. To try and never stop. Until I die, I'm going to try. Even if it never works, I'm going to keep trying." Stewart has probably done more than anyone to raise the public profile of "old friends" therapy, and he hopes that someday a drug will be developed from helminths that will be easier for people to take. As he speculates: "What if these things down-regulate the immune response? You could have a world where, at no risk, maybe you wouldn't have autism — you wouldn't have any autoimmune disease. Maybe I'm just too close to it, but it seems every time you turn over a rock, it still fits, every time something new comes up, it just fits this model."
It is tempting to get carried away by the hubris surrounding "old friends" therapy and react to Johnson's optimism by rushing off to declare a toast "To absent friends!" be they parasitic worms, friendly bacteria, or a whole host of microorganisms from the natural environment. Then raise our glasses full of probiotic drinking yogurt or whipworm eggs, quaff them down, and quickly replenish our "old friends" as soon as possible. But it might be wise to exercise a little caution at this stage and keep the celebrations on ice. Despite the promise, there are likely to be many hard miles to be traveled before the "old friends" hypothesis fully materializes as a reliable, effective, tried-and-tested evolutionary medicine of the future.
Follow all of the Expert Voices issues and debates — and become part of the discussion — on Facebook, Twitter and Google+. The views expressed are those of the author and do not necessarily reflect the views of the publisher. This version of the article was originally published on Live Science.