News this Week

Science  15 Aug 2014:
Vol. 345, Issue 6198, pp. 716
  1. This week's section

    Comet chaser arrives at quarry

    Rosetta's view of comet 67P/Churyumov-Gerasimenko on 3 August.

    PHOTO: ESA/ROSETTA/MPS FOR OSIRIS TEAM MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA

    Rosetta has finally met its cometary match. Last week, after a 10-year journey, the European Space Agency's spacecraft arrived at comet 67P/Churyumov-Gerasimenko, a lopsided, bilobed lump of dust and ice sometimes likened to a 4-kilometer-long rubber duckie. The duo will now take a lap around the sun, and Rosetta will watch as the comet heats up and releases ice in a halo of gas and dust (Science, 1 August, p. 502). The spacecraft is now flying triangles in front of the comet, staying 100 kilometers away; in several weeks, Rosetta will settle into a 30-kilometer orbit around the comet. Mission scientists are gearing up to map the comet's surface in hopes of finding a smooth area on which to drop Philae, a lander the size of a washing machine, this November. To help it stick the landing in the comet's feeble gravity field, Philae will use two harpoons. Then, it will drill into the comet to gather surface material for analysis.

    “The international community was comfortable in allowing two relief agencies to provide all of the clinical care for the Ebola victims in three countries.”

    Ken Isaacs of Samaritan's Purse, one of those agencies, criticizing the response to the Ebola threat at a House of Representatives subcommittee hearing on 7 August.

    Mummification's origins pushed back

    Resin on fibers from Predynastic Egypt, viewed in polarized light.

    PHOTO: © RON OLDFIELD

    In prehistoric Egypt, between 4500 and 3100 B.C.E., the hot, dry desert did the work of mummification by desiccating bodies naturally. It wasn't until much later, around 2200 B.C.E., that artificial embalming using resins began—or so scientists thought. New evidence pushes back the origin of artificial mummification in ancient Egypt by 1500 years, researchers at the University of York in the United Kingdom and Macquarie University in Sydney, Australia, reported online this week in PLOS ONE. The researchers examined funeral wrappings dating to between 4300 and 2900 B.C.E. and excavated pit graves in cemeteries dating to between 4500 and 3350 B.C.E. in the Badari region in Upper Egypt. The team identified embalming agents on the linen wrappings made from ingredients such as pine resin, gum, aromatic plant extract, and natural petroleum. Some 3000 years later, the researchers note, Egyptians at the height of the Pharaonic period used recipes with the same ingredients in similar proportions to produce more famous mummies.

    By the numbers

    5 — Hours of sleep per night that astronauts got aboard the International Space Station and the space shuttle, according to a study of 85 crew members in The Lancet Neurology. An official flight schedule mandates 8.5 hours.

    60% — Fraction of plants, in a study of 300 species across the United States, that moved downhill over the last 40 years—likely seeking wetter regions despite warmer temperatures at lower elevations, according to a study in Global Change Biology.

    Around the world

    Brisbane, Australia

    Patent or perish?

    A suggestion by Australia Minister for Industry Ian Macfarlane that the government may consider tying science funding for universities to the number of patents they generate has drawn sharp criticism. “We need to ensure that commercialisation of [intellectual property] … is part of the process of giving taxpayers' money to researchers,” Macfarlane said in a 6 August speech, according to a transcript by the Australian Broadcasting Corporation. Political opponents and researchers decried the idea, saying it could shift funding from basic research, particularly in areas with no obvious potential for commercialization. A focus on patents could create “perverse incentives,” Aidan Byrne, chief of the Australian Research Council, told The Australian. But several commentators note that Macfarlane's ministry may have little impact on university research funding policy, which is primarily the responsibility of the education ministry. http://scim.ag/Ausminpatents

    Bangalore, India

    India bans dissections

    India's Western Ghats bullfrog.

    PHOTO: AJITH U/FLICKR

    A campaign to bar dissections in India's university classes has scored a major victory. India's University Grants Commission, which sets India's standards for university education, has banned the dissection of animals in zoology and life science university courses. That follows a decision in March by the Medical Council of India to prohibit animal dissection in undergraduate medical courses; it may extend the ban to postgraduate courses. The group People for the Ethical Treatment of Animals has argued that computer models and simulations can replace dissection and that several frog species are now endangered due to large numbers of zoology students collecting them for experiments. But some educators contend that university classrooms aren't prepared to offer alternatives. http://scim.ag/Indiadissect

    Washington, D.C.

    Wolverine protection withdrawn

    Citing scientific uncertainty in climate change models, the U.S. Fish and Wildlife Service (USFWS) on 12 August decided not to list the wolverine as threatened. The decision is controversial because the agency's scientists and two independent review panels had previously endorsed the listing proposal, only to be overruled in May by a regional director. Once hunted, trapped, and poisoned nearly to extinction, the wolverine has bounced back in the mountain west. But global warming threatens their snowy dens. In May, a regional official decided that climate change models could not predict a lack of dens precisely enough to justify the listing. USFWS Director Dan Ashe defended the decision: “Reasonable professionals can look at the same information and come to a good but different conclusion,” he said. http://scim.ag/_wolver

    Newsmakers

    Three Q's

    PHOTO: R. STONE/SCIENCE

    A draconian antigay law enacted in Uganda in February (and struck down on a technicality on 1 August) generated worldwide censure—and the scientific community's response has been a trial by fire for the Uganda National Academy of Sciences' president, Nelson Sewankambo, 62. Sewankambo spoke with Science about the controversy. http://scim.ag/Sewankambo

    Q:What were the origins of the antigay law?

    A:The international community tends to think that intolerance for gays started after Christian activists visited Uganda in 2009. [But] intolerance for gays has been going on for years. [The Christian] group may have accelerated antigay feelings within the government.

    Q:After Parliament passed the antigay legislation last December, when was your academy asked to weigh in?

    A:We were brought into the picture very late, a week or so before the report was required. I said, this academy … has to use evidence and produce a credible report. I also knew that this report would define the reputation of this academy. So we declined.

    Q:Is the law shelved for now?

    A:The government is free to appeal. They are still thinking through it. Members of Parliament are free to reintroduce it.

  2. Parenting

    Unsettled questions trail IVF's success

    1. Kelly Servick

    Researchers are still struggling to understand the potential long-term effects of assisted reproduction.

    Back in 2001, developmental biologist Richard Schultz stood before a meeting of reproductive clinicians in New York City to present some uncomfortable results. Adult mice that had developed in vitro as early embryos, he reported, showed subtle differences from controls, including poorer spatial memory. He thought that his audience of experts on human in vitro fertilization (IVF)—extracting and combining sperm and egg in a culture medium, then transferring the embryo back into the mother's uterus—would be interested. “I was naive,” he says.

    Several attendees pounced on the idea that IVF could have long-term effects, and one asked how he dared to conduct such an experiment. Clinicians seemed to fear that his finding would threaten the fertility industry, or even provoke litigation from patients, Schultz says. “I have never been attacked ad hominem in a public summit such as this.”

    In the audience that day was Paolo Rinaudo, then a medical fellow looking for his first research project. He, for one, took Schultz's results to heart. “This was really a key moment for me,” Rinaudo says. “I was completely flabbergasted” by the suggestion that IVF could have consequences in adulthood. He now studies possible metabolic effects of IVF in mice and also performs the procedures for couples at the University of California, San Francisco, Medical Center.

    Based on existing research, he's able to tell patients that the known risks of assisted reproductive technologies (ART), which include variations on in vitro fertilization, are relatively small: a slight increase in pregnancy complications, preterm birth, and low birth weight, and possibly an increased, though still very low, risk of birth defects.

    But as Rinaudo and others learn more about the sensitivity of early embryos, they have begun to wonder if the manipulations of IVF may have additional subtle effects that appear as children grow. So far the evidence is equivocal. IVF boosters can point to more than 5 million children conceived this way, most apparently healthy. But the track record is short: The oldest IVF baby, Louise Brown, is just 36. “What will happen to these kids when they are middle-aged?” Rinaudo wonders. Animal studies hint that these people may be at higher risk of agerelated disorders, such as heart disease, diabetes, or hypertension, though these effects haven't been confirmed in people. A few other studies suggest subtle cognitive differences in ART children, but sample sizes are small.

    No researchers are suggesting that ART shouldn't be done. But some argue that to inform parents accurately about risks, they need more and larger studies. Rinaudo and clinician-researchers like him straddle two worlds: a multibillion-dollar industry full of eager patients, and a contradictory scientific literature full of vague concerns. Most fertility researchers are trying to improve success as measured by a single, clear standard—the birth of an apparently healthy baby. Only a few are trying to discern whether IVF leaves a subtle legacy in children.

    TWO EVENINGS A MONTH, the conference room in Aberdeen Maternity Hospital in the United Kingdom fills with couples planning to start fertility treatments. At these gettogethers, they realize that “there are plenty of other people in the same boat,” says Abha Maheshwari, a gynecologist at the hospital who also studies the possible consequences of ART. Worldwide, more than 45 million couples face the same painful situation.

    Driven by the urgent desires of these couples, scientists have turned ART into one of medicine's dramatic success stories. In the United States, the odds of a live birth with the method have risen from the low teens in the 1980s to roughly 40% today for women under 35. That's largely due to lab conditions that more closely mimic a mother's body, and to new technology—notably, the use of intracytoplasmic sperm injection (ICSI), in which a sperm is inserted directly into the egg to make up for defects in sperm.

    An empty petri dish awaits an egg for in vitro fertilization.

    PHOTO: SCIENCE PHOTO LIBRARY

    But as ART took off in the clinic, some basic scientists remained wary. “It didn't rest upon a solid foundation of experimental work,” says Schultz, citing the lack of animal research documenting physiology, behavior, longevity, and susceptibility to disease later in life. “There really weren't any solid data in terms of ‘Were these children at risk?’”

    When asked why they suspect ART might have long-term effects, most researchers start by citing one name: David Barker. In landmark studies in the 1980s, the physician and epidemiologist linked chronic diseases like diabetes and heart disease to malnourishment in the womb, as measured by low birth weight.

    As Barker's ideas gained popularity in the mid-1990s, some wondered whether the formative effects of fetal development reached further “upstream,” to the period before a drifting clump of embryonic cells nestles itself into the uterine wall to become a fetus. In 2000, developmental biologist Tom Fleming and his colleagues at the University of Southampton in the United Kingdom showed that pregnant rats that were malnourished during this preimplantation stage gave birth to offspring with high blood pressure and abnormally sized organs, even if the mothers resumed a normal diet for the rest of pregnancy. Apparently, the early embryo's struggle for resources in nutrient-poor surroundings led to irreversible changes in development. “The embryo needs to fight for itself,” Fleming says.

    That raised a new question: Are embryos that develop in the foreign environment of a petri dish and undergo physical manipulation in the lab compelled to “fight for themselves” in particular ways? The question spurred Schultz's animal studies, and a few labs have begun to explore these issues with human embryos—the surplus from IVF procedures, donated by parents.

    In vitro embryos are not malnourished—quite the opposite. Compared with the relatively weak, watery solution in a mother's fallopian tube, commercially available cell culture medium, which is sometimes enriched with growth factors, is “a little bit like rocket fuel,” says reproductive biologist Helen Picton of the University of Leeds in the United Kingdom, who is exploring the effects of the medium. “You really set [the embryos] off at a great speed.”

    Picton adds that in human embryo research, with only a few donated embryos of a handful of cells each, “the hard work is characterizing what's normal.” Adding growth factors to the medium appears to alter the expression of certain genes known in mice to be associated with cell fate—which cells survive, and which will develop into part of the placenta or part of the fetus. But researchers are far from linking these gene expression patterns to any health risks.

    A microneedle punctures an egg during intracytoplasmic sperm injection (ICSI).

    PHOTO: MAURO FERMARIELLO/SCIENCE PHOTO LIBRARY/CORBIS

    In the years since Schultz presented his results, Rinaudo says most evidence from animal models has been reassuring—the procedure doesn't seem to affect longevity, for example. But ART-conceived animals do appear more likely to have problems metabolizing glucose, and this effect may be transmitted to subsequent generations, probably through epigenetic changes—alterations in inherited chemical markers that attach to DNA and affect its expression. Two rare disorders related to epigenetic marks, Beckwith-Wiedemann and Angelman syndromes, also appear to be more common in ART children, although the absolute risk is still extremely low, with fewer than 100 cases of either disorder reported in these children.

    Perhaps because of all the uncertainty, Rinaudo says that most of his clinical colleagues remain focused on improving the numbers of live births with ART, rather than exploring its possible consequences. “We haven't been able yet to metabolize this information,” he says, “because there are not really studies in humans that show long-term problems.”

    HUNTING FOR SUCH PROBLEMS among the millions of ART children is far from straightforward. Easiest to spot are negative outcomes early in life, and studies have repeatedly shown small increases in low birth weight, preterm birth, cerebral palsy, and other birth defects in ART children. “They have higher risk of all the anomalies, all the complications in pregnancy you can name,” Maheshwari says. The best estimates put their overall risk of a major birth defect at about 30% higher than the general population. That's a modest increase in a small risk: 4% to 5% versus 3% to 4%.

    But Maheshwari adds that with any increased risk, researchers have to ask: “Is it because of IVF processes, or is it because we are treating a different population?” Couples who seek ART tend to be older, and for some, the difficulty conceiving may be related to abnormalities in sperm or eggs that also affect embryos. The procedures also increase the odds of multiple births as much as 20-fold in women under 35, according to a paper Rinaudo and colleagues published in Molecular Human Reproduction in 2013. And however they are conceived, twins and triplets face increased risks of all kinds, from pregnancy complications and low birth weight to cognitive and behavioral problems.

    A control group of fertile couples willing to undergo ART for no reason could resolve such questions, but that's not practical. But there's a natural control group: children born to couples diagnosed as infertile who then waited long enough for ART that they conceived spontaneously. In 2006, researchers in the Netherlands used a national database to find 6343 Dutch women diagnosed with infertility who conceived without IVF or the ovulation-stimulating drugs that usually precede it. Compared with these controls, babies born after both fertility drugs and IVF were more likely to have very low birth weight and to be born very preterm—before 32 weeks. Another group used the same database to conclude that the ART children, aged 8 to 18 at the time of testing, had slightly higher blood pressure and blood glucose levels than their spontaneously conceived counterparts—potential indicators of risk for heart disease or diabetes later in life.

    Those findings hint that ART carries an extra overlay of risk. Still, most studies to date are relatively small and lack statistical power, says reproductive epidemiologist Michael Davies of the University of Adelaide in Australia. “People are dragging their feet somewhat,” he says. “They've yet to find the utterly compelling body of data.”

    Early reports of mild cognitive differences in very young ART children, for example, have been criticized for small sample size and other biases. Davies recently submitted a grant for a study that would link 11,000 IVF births to school records and census data measuring sight, hearing, and disability, in part to follow up on controversial suggestions that IVF children have more contact with disability services.

    His group also did a large analysis of data from South Australian clinics and birth defect registries, published in 2012 in The New England Journal of Medicine, showing that the direct sperm-into-egg injection method of ICSI appeared to increase the risk of birth defects, but only when embryos were transferred to the uterus while fresh, not preserved by freezing and rethawed for later use.

    Such efforts to pick apart the risks of different ART techniques are critical, Maheshwari says. ICSI, which accounts for more than half of ART procedures in both the United Kingdom and the United States, can increase the chance of a live birth, but the Aberdeen clinic recommends against it unless defective sperm make it absolutely necessary. The dilemma highlights the tension between offering parents the best odds of conceiving and the least risk to the embryo, Maheshwari says. National registries that list live birthrates at different IVF centers heighten the pressure to demonstrate efficiency, she says. “This is what drives the market, and that's what drives your practices.”

    A more dramatic and controversial example is the practice, still relatively common in the United States but increasingly discouraged in most developed countries, of returning multiple embryos to a woman's uterus simultaneously. This reduces the chance a couple will have to go through (and pay for) another round of IVF to get pregnant, but raises the chance of multiple pregnancies and their associated risks.

    A newer technique in which an embryo is allowed to develop in culture for 5 days rather than three presents another tradeoff. The practice appears to boost the odds of pregnancy, but Maheshwari thinks it may also increase the risk of preterm delivery. Rinaudo also worries about lengthening the exposure of early embryos to an unusual environment, but he says, “It's very difficult to make a recommendation when the evidence is not there.”

    He's still not sure how many vague concerns about long-term health to share with hopeful parents-to-be. For now, he tells couples that some animal studies suggest subtle adverse effects later in life, but it's too early to draw firm conclusions. He keeps animal studies, like Schultz's mouse work, on hand, should patients want to see them.

    Usually, he says, they don't ask.

  3. Parenting

    Nature's first functional food

    1. Trisha Gura*

    Breast milk feeds helpful microbes, fights harmful ones, provides immunity, and jump-starts a newborn's life.

    Blood, urine, saliva, and spinal fluid. Those are the human bodily fluids most explored by scientists over the decades. Yet any woman who has ever nursed a newborn will cite a major omission: breast milk. Researchers long ago unraveled the basics of this maternal liquid. But until recently, few have given it serious attention with sophisticated analytical techniques. Breast milk was “ignored as not modern,” for the last half-century, says epidemiologist Ardythe Morrow of Cincinnati Children's Hospital Medical Center in Ohio. But now, it's “an exploding area of science.”

    The resurgence has its origins in a long-standing conundrum: Breast milk abounds with complex carbohydrates called oligosaccharides that humans can't digest but beneficial bacteria can thrive on. Fifty years ago, when the oligosaccharides were discovered, investigators lacked the technology to deduce their structure and determine their effect on what is now called the infant gut microbiome (the myriad bacteria that naturally reside in human intestines, beginning at birth). Unable to progress significantly, scientists lost interest in milk-microbe connections.

    Now, thanks to breakthroughs in analytical chemistry and a growing interest in the microbial fauna in the human body, as well as a movement touting the benefits of breast-feeding, those connections are being explored once again. Some researchers have focused on making better use of the microbiome fostered by milk, while others have documented how breast milk does more than feed a newborn and its “good” bacteria. Mother's milk also contains an evolving stockpile of compounds that thwart pathogens, foster a robust immune system, and perform other functions. Most recently, researchers have discovered that mom provides inactive enzymes in her milk that turn on in the infant gut and clip out bioactive molecules from other milk proteins. “Milk is really a genius fluid that was outrageously understudied,” says microbial ecologist David Mills of the University of California (UC), Davis. “If we can identify components of human breast milk that are important, then we can understand the wisdom of milk—and take advantage of them.”

    Breast milk supplies more than just nutrition for babies.

    PHOTO: KIDSTOCK/BLEND IMAGES/CORBIS

    AT FIRST BLUSH, BREAST MILK is a buffet of fats, proteins, and sugars, in a ratio of about 1-to-3-to-7. Until recently, scientists viewed the fluid mainly as food for a rapidly growing newborn. Yet, mother's milk also teems with immune cells, such as macrophages, stem cells for regeneration, and hundreds to thousands of bioactive molecules—some fatlike, others protein in nature, and still others that are indigestible oligosaccharides. These bioactive molecules can protect against infection, ward off inflammation, spur immune system and organ development—and, in the case of oligosaccharides, shape the infant microbiome, scientists have found. Milk “is not primarily about nutrition,” Morrow says. “Rather [it's about] immune protection.”

    Signs of that protection began to emerge almost 130 years ago with the discovery that breast-fed infants survived more often than bottle-fed ones. In searching for reasons why, pediatricians and microbiologists analyzed the bacterial composition of the feces of both types of infants, which reflects their gut microbiome, and found significant differences. Meanwhile, chemists at the turn of the 20th century noted that human breast milk contained “a different type of lactose” than cow's milk did. Then, in the 1950s, Nobel Prize winner Richard Kuhn and physician Paul György showed that this “different type” consisted of hundreds of oligosaccharides that promoted the growth of microbes called bifidobacteria. But the work paused, as chemists spent the next 50 years advancing techniques necessary to sort out the complexity of the oligosaccharides.

    In 2006, a team at UC Davis led by food chemist Bruce German and his graduate student Robert Ward helped reignite interest in the milk-microbiome connection. The group enlisted analytical chemist Carlito Lebrilla, who used mass spectrometry and a sophisticated chromatographic technique to characterize the oligosaccharide repertoire. Lebrilla identified nearly 200 unique human milk oligosaccharides (HMOs) and handed the mix to Mills, who, along with German, assumed a plethora of bacterial species would feast on the sugary cocktail.

    “But they didn't.” German says. “It was a jaw-smacking moment.”

    Only one type grew robustly: Bifidobacterium longum biovar infantis (B. longum bv. infantis). Scientists are not yet certain exactly how the bacteria first enters the infant's intestinal tract—possibly when a fetus begins swallowing amniotic fluid and later as the infant passes through the vaginal tract. Breast milk, which contains small numbers of the bacteria, is another possible route. Assisted by the HMOs in breast milk, the organism eventually multiplies to dominate the breast-fed baby's microbiome, making up as much as 90% of it. After weaning, it eventually drops to 3% of the adult microbiome. The UC Davis team wondered: Why did this organism so rule the budding infant microbiome?

    A year later, Mills came up with an answer. After sequencing the genome of B. longum bv. infantis, he learned that it carries the precise genes for all the enzymes needed to digest the milk's oligosaccharides—other bacteria, even closely related ones, don't share all of those genes. “Clearly this bacterium coevolved” with humans, German says.

    Why would a new mother expend energy producing food for an organism other than her offspring? German says it's pretty clear. “Mother is recruiting another life form to babysit her baby.”

    By babysitting, German refers to the growing list of protective functions in a newborn that B. longum bv. infantis appears to orchestrate. For example, the bacterium consumes the oligosaccharides whole before other nasty microbes can digest them, starving out the bad bugs. B. longum bv. infantis also produces molecules called short-chain fatty acids that serve as favorite food for other beneficial bacteria and help guide the cells that line the infant intestine in how to use energy and mount an immune defense.

    Apart from promoting good bacteria, human milk oligosaccharides themselves can ward off dangerous bacteria such as Salmonella, Listeria, and Campylobacter, causes of diarrhea and infant mortality. The HMOs closely mimic carbohydrate structures on the infant's gut cells. Because many pathogens typically attach to these cell-bound structures to infiltrate the gut, mother's milk, in essence, floods the infant's gastrointestinal tract with decoys that bind the pathogens and keep them off the intestinal wall.

    NEONATOLOGISTS ARE NOW HOPING to turn the growing understanding of B. longum bv. infantis to their advantage, most immediately by ameliorating, or preventing, an often deadly disease called necrotizing enterocolitis (NEC) that tends to develop in premature infants. Starting as a serious bacterial infection, NEC overwhelms the gut several weeks after birth. The intestinal tissue dies; surgeons try to remove it; and the baby often suffers severe and permanent disability, neurological damage, and—in 20% to 30% of cases—death.

    NEC can strike just when it seems a premature infant is out of the initial dangers of being born too early. “I can think of a baby,” says neonatologist Mark Underwood of UC Davis, “who one day was looking good and on the next day was pooping blood. By that night was dead. You see that in a kid who you thought was past the worst of it, and it is just so devastating.”

    Theories abound to explain the condition: Premature infants have a leaky gut that lets bacteria through, or cells lining their intestine exhibit an overreactive inflammatory response. Premature infants may also not get enough breast milk: Studies show that the risk of NEC plummets six- to 10-fold if preemies are given breast milk rather than formula.

    Yet, even when mother's milk is provided in a feeding tube, NEC can still take hold. The problem may not be a lack of milk, but rather a deficient microbiome—one lacking enough B. longum bv. infantis, for example. Simply, premature infants carry premature intestinal fauna, Underwood and others say.

    In some studies, attempts to speed the maturation of the microbiome, just as physicians give substances to preemies to promote lung development, have shown little success. In one recent effort involving preterm infants, the UC Davis team tried out different combinations of formula and breast milk with and without the right oligosaccharides. The researchers discovered that none of the interventions changed the microbiota of the preemies. By contrast, some studies in full-term infants indicate that adding HMOs to formula or breast milk promotes beneficial microbes. So why are preemies not responding similarly?

    “It's a different environment that premature infants live in,” notes neonatologist Barbara Warner of Washington University School of Medicine in St. Louis. Catheters, breathing tubes, and other instrumentation breed infections by microbes not normally in an infant. Also, preemies get antibiotics for long periods of time, wiping out beneficial bacteria.

    DESPITE MIXED RESULTS, clinicians trying to help premature infants and other newborns now view mother's milk and its promotion of a good microbiome as so key that many are developing and testing supplements containing probiotics (the bacteria) and prebiotics—food that will stimulate the beneficial bacterial growth. Clinical researchers in Europe, Canada, and elsewhere have completed nearly 40 studies using pre- or probiotics, some containing B. longum bv. infantis, in premature infants, estimates Allan Walker of Boston's Massachusetts General Hospital and Harvard School of Public Health. The most recent, published online in January in The Journal of Pediatrics, showed that a probiotic product that contains B. longum bv. infantis, one readily available in North America, substantially reduced the frequency of NEC in a Montreal, Canada, neonatal intensive care unit.

    Illustration: G. Grullón/Science

    Few such studies have been done in the United States, however. Although for adults probiotics are considered supplements and go unregulated, introducing microbes into premature infants falls under the same U.S. Food and Drug Administration regulations as vaccines or new drugs. Anticipating increased development costs and time, U.S. researchers have either pursued other strategies or partnered with collaborators elsewhere.

    For example, Walker's group is working with neonatologists in Chile to test a specific sugar-containing molecule, or glycan, found in mother's milk. Studies in newborn mice, as well as in a tissue model of the human premature gut, show that the glycan damps down proinflammatory factors, while boosting anti-inflammatory ones. Chilean researchers will supplement mother's breast milk with the glycan and look for a reduction in the incidence of infection and death in premature infants.

    While also pursuing the idea of introducing B. longum bv. infantis or its food in preemies, the UC Davis researchers have gone back to basic studies of mother's milk for further inspiration. They began by analyzing the fluids taken from the gastric tubes of premature infants after feeding. To their surprise, the researchers learned that the infant stomach is not wildly acidic and stocked with aggressive enzymes, as previously thought. The group reported online in April in The Journal of Nutrition that the infant stomach seems geared to digest very few proteins, mainly β-casein, but also lactoferrin, lysozyme, and α-lactalbumin. All are found in breast milk. The enzymes that were present were not chopping up proteins randomly, but rather were clipping them at very specific sites in their sequence, generating 603 peptides. Some of these are known to inhibit the growth of pathogenic bacteria or regulate the immune system.

    What's more, the UC Davis team recently discovered that some of the digestive enzymes they found in infants' guts after feeding are actually produced by the mother and delivered, inactive, in her milk. Something in the infant's stomach later turns them on so they can free up the active peptides from the full milk proteins. The UC Davis team theorizes that supplementing milk with the latent enzymes will promote the formation of the beneficial peptides in premature infants.

    Industry is watching this and the oligosaccharide developments closely. “It is a rat race,” says nutritional scientist Lars Bode of UC San Diego. “Whatever formula company comes out with the first oligosaccharide to add to their product, that will make a huge shift in the market. Everyone is working on it.”

    For example, German, Mills, and the UC Davis team started a company called Evolve Biosystems Inc. to advance their own combination of prebiotics and probiotics. Because scientists cannot get enough donor human milk to isolate enough prebiotics, German's group is working instead with bovine milk, which contains small amounts of HMOs. Infants have been drinking dairy-based formulas for decades—without apparent harm—the researchers reason. Hence, the approach should get around some regulatory hurdles.

    But the sugars in cow's milk often differ substantially in their structure and abundance from those in human breast milk. Thus, investigators such as biochemist David Newburg of Boston College have set out to synthesize HMOs. In 2004, Newburg showed that an oligosaccharide found in breast milk called 2'-fucosyllactose is effective in warding off the pathogenic bacteria that cause cholera, a diarrheal disease. With his own money, he co-founded Glycosyn LLC to test the oligosaccharide and its derivatives. The company makes them by inserting the genes for various sugar-building enzymes into yeast and bacteria. Newburg predicts he is about a year away from marketing a nutritional supplement.

    This intense rush of research and commercial development only highlights the remarkable complexity and powers of the milk every mom naturally makes for her newborn. “Evolution has delivered the functional food,” Mills says. “We just need to understand it.”

    • * Trisha Gura is a science writer in Boston.

  4. Parenting

    The taste of things to come

    1. Emily Underwood

    Early postnatal, and even prenatal, experiences shape culinary tastes.

    Everything about the newborns was normal, except for their pungent aromas of fenugreek, curry, and cumin. At first, the physicians who delivered the four unrelated babies at a Tel Aviv hospital were worried—some strong smells at birth signal disease. But soon they discovered the apparent source of the odors. Hours before labor, the mothers said, they had eaten skhug and amba, spicy traditional Yemeni and Israeli foods.

    The obscure case study—described in a 1985 issue of the European Journal of Pediatrics—caught Julie Mennella's attention in the early 1990s. A biopsychologist at the Monell Chemical Senses Center in Philadelphia, Pennsylvania, Mennella studies why we like the foods that we do—what she describes as one of the “biggest mysteries of human behavior.” The Israeli study “was a great ‘aha!’ moment,” Mennella says, because it supported her hunch that infants can perceive—and perhaps even learn to love—the distinct flavors of their culture's foods while still in their mothers' wombs.

    When Mennella began her career, relatively little was known about how and when humans form their preferences for different flavors. Despite a large literature on how maternal diet influences the overall nutrition and health of offspring, few researchers were looking at whether flavor preferences derive from what the mom consumes during pregnancy or soon after birth. Mennella recalls that doctors used to tell women still nursing their child to avoid spicy or pungent foods, based on “folklore” that infants inherently prefer bland foods.

    Over the past several decades, however, a convergence of findings from psychologists and neuroscientists, including Mennella, has shown that infants' tastes are malleable. Although our cravings for salt, sugar, and fat do have a genetic basis, new evidence shows that the majority of our food preferences are “not inborn,” Mennella says. Instead, the taste for a variety of flavorful foods, including bitter vegetables, is “part of culture, and something you have to learn.” Through creative experiments with newborns, Mennella and others have shown such learning can begin as early as in the womb.

    Babies form flavor preferences as early as in the womb.

    PHOTO: WENDELL/DAMBROSIA/WEESTOCK/CORBIS

    FOLLOWING YOUR NOSE is more than a figure of speech in Mennella's lab. An Italian-American for whom food is “very important,” she made pungent garlic the focus of her first attempts to pin down some of the origins of our flavor preferences. Working with Gary Beauchamp, director of the Monell research program, Mennella asked pregnant women to consume either plain-flavored or garlic capsules during the last weeks of their pregnancy. By that point, a fetus clearly has both the ability to taste and to smell—around the 24th week of gestation, for example, mucus plugs blocking the nasal passages dissolve, allowing amniotic fluid to bathe a fetus's newly formed olfactory receptors (see p. 751).

    Mennella then asked volunteers, recruited from around the university, to sniff unlabeled bottles of amniotic fluid and breast milk from both groups of mothers. If these amateur sommeliers could smell the garlic in a blind test, it was likely that the fetuses could, too, Mennella reasoned—and the volunteers easily sniffed out which came from the garlic consumers. Mennella next videotaped and documented how babies born to both maternal groups behaved when they were offered garlicky milk. Contrary to the old folk wisdom, the babies whose mothers had eaten garlic sucked their milk down with relish. In contrast, infants with garlic-free mothers grimaced and writhed to avoid it.

    Since that series of studies in the 1990s, Mennella and Beauchamp have obtained similar results with other flavors such as vanilla and carrot. Familiarity seems to breed affinity for substances such as alcohol, too, Mennella says. In one study, she and Beauchamp found that 1-year-olds with heavy-drinking or alcoholic parents prefer to mouth toys scented with ethanol over unscented or vanilla-scented toys.

    Research in mice hints at a potential neurobiological basis for these early associations, says Diego Restrepo, a developmental biologist at the University of Colorado School of Medicine in Aurora. In 2010, he fed cherry- or mint-flavored chow to pregnant mice, then studied which flavor their offspring preferred at 20 days old. Like the human infants, the young mice preferred to eat the chow flavor they'd been exposed to in utero.

    Neurons that relay information about smell from the nose to the brain respond individually to thousands of odorant molecules, allowing mice and humans to distinguish a vast array of different smells—possibly as many as 1 trillion odors in humans (Science, 21 March, p. 1370). On examining the brains of the two groups of rodents, Restrepo found that in mice exposed to the mint flavor in utero, the junctions between the olfactory sensory neurons that can detect mint and those that relay a mint signal to the amygdala, a brain region that processes emotion, were significantly enlarged. In mice exposed to cherry, those junctions between neurons relaying the cherry signal were bigger, suggesting that experience with the smell in utero had strengthened those connections, he says.

    A baby rejects a novel flavor (top), while another exposed to a flavor in utero accepts it (bottom).

    PHOTOS: CATHERINE FORESTELL AND JULIE MENNELLA/MONELL (2)
    PHOTO: NESTLE/PETIT FORMAT/SCIENCE SOURCE

    More evidence that maternal diet can influence the developing brain's flavor processing came from a study published last year in The FASEB Journal by researchers at the University of Adelaide in Australia. They fed pregnant rats either a monotonous diet or a smorgasbord of tempting treats, including hazelnut spread, peanut butter, chocolate biscuits, sweetened cereal, and a lard and chow mix. Then they observed how the offspring behaved around those foods. Mice whose mothers had a junk food diet developed abnormal neural reward circuitry resembling that seen in addiction.

    SOME RESEARCHERS DOUBT that prenatal or early postnatal experiences with flavor have much long-lasting effect. Adults are clearly capable of learning to love and hate new flavors, as anyone who's lost their taste for a favorite food after a bout of nausea can attest. And if you compare foods that college students enjoy with those they ate under their parents' watchful gazes, the correlation “is not zero, but it's very low,” says Paul Rozin, a psychologist at the University of Pennsylvania.

    But Mennella and others believe that there are sensitive periods for flavor learning, just as there are for the development of sight and other senses. At 3.5 months old, many babies will refuse to drink Nutramigen, a sour, bitter-tasting formula often prescribed to underweight infants who can't digest cow's milk. She and Beauchamp have shown that if the infants start the formula before that age, most will happily drink the nasty-tasting stuff, and even prefer it. Understanding why infants learn to accept different flavors could help guide treatment for premature babies, who frequently spend their first months of life being tube-fed and often gag when they first taste milk or formula, Mennella says.

    From an evolutionary point of view, infants need to be receptive to new flavors early in life in order to survive, says Regina Sullivan, a psychiatrist at New York University Langone Medical Center in New York City. Because baby mammals depend so much on their parents, it makes sense that a mechanism exists to ensure that newborns learn to eat what mothers can provide no matter how it tastes, she explains.

    Mennella hopes that understanding early-life influences on diet in children will help parents buffer children from an environment increasingly filled with sugar- and salt-laden foods. That is challenging because, as she and Beauchamp have found, infants are more sensitive than adults to both sweet flavors and bitter ones, making it harder for them to resist sugary foods and to eat vegetables.

    A handful of correlational studies suggest that children eat more fruits and vegetables later in life if their mothers did the same while pregnant or breast-feeding. Rozin notes, however, that it is difficult to determine how much those early experiences affect long-term outcomes because the foods we're exposed to at a very young age often continue to surround us up until we leave the nest.

    The most potent aspect of early flavor experiences may be their emotional power, Mennella says. Our first flavors are among “the oldest, most enduring memories that take us to our past,” she says. “Food is much more than a source of calories—it provides us with an identity.”

  5. Parenting

    An experiment in zero parenting

    1. Eliot Marshall

    A controversial study of Romanian orphans reveals long-term harm to the intellect.

    An unnerving sight greeted U.S. scientists a decade ago at an orphanage in Bucharest. The institution housed “many children rocking back and forth while sitting or on all fours, turning their head from side to side, or repeatedly bringing their hand to their face, often slapping themselves,” write neuroscientists Charles Nelson and Nathan Fox and psychiatrist Charles Zeanah in their book Romania's Abandoned Children, published in January. The children's empty motions, or “stereotypies,” are like the pacing of a tiger or elephant in the zoo. They are one of the abnormalities under scrutiny in a study of child neglect Nelson and his team launched in 2000.

    What the researchers saw was the legacy of a tragic scheme to boost Romania's population. Former dictator Nicolae Ceauşescu decided in the late 1960s that Romania needed to grow; the government taxed women who had fewer than five children, regardless of what they could afford. To care for thousands of unwanted children, the government filled orphanages. They weren't designed or funded to mimic family life, but they were busy.

    At their peak in the late 1980s, the orphanages housed nearly 170,000 children. Most grew up in a stunningly blank and unresponsive environment. Caregivers came and went in three shifts, and a single staffer might watch over 10 to 15 children. Infants spent time staring at the walls and ceiling, and a child might come in contact with 17 different caregivers in a single week. Nelson, a professor of pediatrics and neuroscience at Harvard Medical School in Boston, calls it an experiment in zero parenting.

    After Ceauşescu's fall and execution in 1989, people from around the world adopted thousands of the orphans. (Romania later put a brake on this.) Researchers have studied these early international adoptees—some with serious emotional problems—for 2 decades now, examining how child rearing can go wrong, how it alters behavior, and how foster parenting can rescue lives. The Nelson group, known as the Bucharest Early Intervention Project (BEIP), saw a chance to answer an additional, specific question: How does neglect in early childhood change neurological development?

    The worst injuries may be invisible. Children who lived in a Romanian orphanage beyond age 2 developed weaker brains than those who left earlier.

    PHOTO: WILLIAM SNYDER

    The researchers launched a radical and, at the time, controversial study. Using U.S. and private foundation funding, they teamed with Romanian officials to run a controlled study comparing a group left in orphanages with a group randomly selected for high-quality foster care in Romania. In a stream of reports since 2003, they have shown that children reared in institutions suffer an array of setbacks, some reflecting long-term changes in the brain (Science, 21 December 2007, p. 1937). The team is now analyzing data from the 12th year, measuring things like disparities in brain structure.

    Findings from the Bucharest project have buttressed those from smaller and less controlled studies, many based on U.S. children in troubled homes. “We are all reporting the same effects,” says Megan Gunnar, a clinical psychologist at the University of Minnesota (UMN), Twin Cities, who is not part of BEIP. “The brain needs stimulation to develop,” and when it doesn't get it, cognitive and emotional growth are stunted.

    Today, Nelson is convinced that early life without parenting can be “more disastrous for brain development” than living with an abusive caregiver. But the Romanian work also confirms that many children can bounce back to something like normality, if placed in a supportive environment. No country should be complacent, Nelson adds, noting that what he saw in Romania is “not all that different from the kind of neglect that many kids in the United States experience.”

    CHILD NEGLECT IS “A HUGE PROBLEM” and probably “the most common form of child maltreatment in the world,” says Seth Pollak, a clinical psychologist at the University of Wisconsin, Madison. Each year, the U.S. National Child Abuse and Neglect Data System tracks about 3 million case reports of maltreatment. These data suggest that neglect is nearly three times as prevalent as physical and sexual abuse combined. And the numbers may understate the problem because it's so difficult to see: “It's the absence of something happening between a parent and child when nobody else is around,” Pollak says: “How do you measure it?”

    Yet the consequences are haunting. The “creepiest thing” about the Romanian orphanages and other places where young children suffer chronic neglect is that “they're quiet,” Pollak says. In most U.S. elementary schools or child care centers, he says, “you hear talking and screaming and crying … it's just raucous.” But in “an environment where children are not being attended to, there is this kind of dead silence. … Children are learning: ‘Why should I cry, or gesture, or make eye contact if no one is responding?’”

    The problem is insidious because neglected children often don't have injuries and don't get the attention that victims of physical or sexual abuse receive. “They wind up coming to the emergency room when they're 5 years old and weigh 20 pounds,” Nelson says, long after the damage is critical.

    Neglect is difficult to study. Researchers normally can't control a child's environment, and they usually don't know what harm a child has experienced before he or she appears in the clinic. This is especially true for children living in poverty, says Gunnar, who sees many of them. “We can use fancy statistics to try to isolate the early effects from continuing and ongoing adversity,” but the results are uncertain.

    The opening of Romanian orphanages after Ceauşescu provided a rare chance to do well-controlled studies. A U.K. group then led by pediatric psychologist Michael Rutter at King's College London undertook the first, a large observational study. The English and Romanian Adoptee (ERA) project, as it was known, spent 2 decades comparing the development of children adopted by U.K. families with children who had not been institutionalized.

    The BEIP group chose a more direct approach: It partnered with Romanian officials to conduct research within the orphanage network. BEIP offered to finance foster care for a small group of children who would exit the institutions; in return, Romania opened the doors and helped with a testing program.

    BEIP came under fire in Europe for including any children still in an institution. Such vulnerable kids were incapable of giving consent, critics argued. Zeanah, a pediatrician and psychiatrist at Tulane University in New Orleans, Louisiana, involved in designing BEIP protocols, claims that the furor was based on “a misunderstanding” about the project's link to international adoption efforts, and European Commission officials ultimately dismissed concerns. BEIP investigators still hear from critics who “just cannot accept randomization” of children to a group that may not benefit, Zeanah says, but it was essential to make the trial work.

    After screening to exclude children with genetic or neurological abnormalities, study leaders selected 136 children for the trial, ranging in age from 6 to 31 months. By random assignment, 68 went into a group that received BEIP-funded foster care and 68 were left to continue with “usual” care—that is, they lived in an institution. Foster caregivers were screened, trained, and paid a “living wage,” Nelson says. It was deluxe care even by U.S. standards. BEIP experts monitored and consulted with the caregivers, who also had 24-hour pediatric medical backup. As controls, BEIP included a group of 72 children from Romanian communities who had never lived in an institution. Researchers gathered test results at 30 months, 42 months, 54 months, 8 years, and 12 years. They have been funded to return at 15 years.

    Children in the institutional care group showed dramatic deficits on a variety of measures when compared with the community group. In an age-adjusted system that converts test results from young children into IQ-comparable scores, known as the developmental quotient (DQ) scale, the institutionalized kids at the outset earned an average DQ score of 74 compared with 103 for the community group. This was “two standard deviations below the mean” for children of that age, according to BEIP, suggesting “profound intellectual delay.”

    In time, the kids placed into foster care advanced toward the normal DQ and IQ ranges, though they continued to lag behind the community group. When BEIP tested children at 42 months, researchers found that the older a child was when placed into foster care, the lower the cognitive score was likely to be. Those who left the institution by 18 months scored above 90 on the DQ scale, whereas those who left at 24 to 30 months scored just over 80. Those who remained in institutions up to age 8 exhibited a “progressive decrease in IQ with age,” the BEIP authors wrote.

    Strangely, as they reached 8 years of age, the foster care children began to slow their climb into higher DQ and IQ scores and came to rest on a par with the children in the institutionalized group, with IQ scores hovering near 80. Both ranked “far below” children who had never lived in an institution, according to BEIP, perhaps because their recovery from the effects of early neglect had reached a ceiling.

    Scalp EEG readings from 8-year-olds reveal big differences in the alpha (high-frequency) range. The institutionalized group (far left) and those assigned to foster care after 24 months had statistically similar, weak results (cool colors). Those assigned to foster care before 24 months (third from left) or never in an institution had strong alpha results. The two leftmost groups appear to have significant delays in brain development.

    Meanwhile, the gradual dissolution of study categories and the start of regular schooling may have helped boost scores in the institutionalized group. By 8 years of age, only 14 of the 68 kids originally assigned to institutional care remained in place; most had settled with relatives or foster families. Yet, as planned, the analysis treated them as belonging to their initial group. As children bailed out of orphanages, their experience of family life may have boosted DQ and IQ scores.

    The good news is that IQ does seem to rebound in children who leave institutional care—the sooner they leave, the better, according to Nelson's group. BEIP found that children who left an institution by 24 months seemed to recover best. Other researchers agree about the need to get out early, but differ about timing. For example, the United Kingdom's ERA study found that children adopted into British families seemed to score just as well as community kids if they left the institution by 6 months of age. Both BEIP and ERA have found that certain behavioral problems, such as poor attention and hyperactivity, persist among kids who were in institutions for any period.

    BEIP also showed that the longer children lived in an institution before going into foster care, the more likely specific neurological changes would appear. The thesis underlying all of this work is that a baby's brain “expects” regular stimulation, the kind that comes from an attentive caregiver. If this doesn't occur, neurons don't grow properly. This leads to abnormalities that show up in neural activity and brain structure.

    BEIP researchers used small, sensor-laden caps to take snapshots of brain electrical activity, known as electroencephalography (EEG) readings, when children first joined the trial and at several intervals over the years. At the start, children living in institutions had lower EEG output than did children living in the community in the high-frequency alpha and beta EEG ranges, and larger output in the low-frequency theta range. This pattern, considered a sign of brain immaturity, persisted into the eighth year. Children placed in foster care before 24 months of age, however, developed EEG patterns with elevated alpha signals by 8 years of age that were “indistinguishable” from those of community children.

    Institutional care also seems to have left a mark on the children's brain structure. In a 2012 paper in the Proceedings of the National Academy of Sciences, members of the BEIP group reported on an imaging study of Romanian children 8 to 11 years old. Using magnetic resonance imaging, Margaret Sheridan of Boston Children's Hospital and others found that Romanian kids raised in an institution had significantly less brain tissue, especially gray matter (a mix of various cell bodies), than those who had never lived in an institution. But children who had been placed in foster care showed a smaller deficit. “These reflect profound effects on the brain,” Nelson says.

    Like other studies, the BEIP project suggests that brain development passes through critical periods. The first 24 months of life seem especially important for cognitive development, and Nelson says that the critical period for healthy attachment to a parent lasts through the first 20 to 22 months. For language, the window appears to be up to 16 months. But there are no sharp lines, says UMN's Gunnar. “The general story seems to be that the brain is remarkably plastic” and can find many ways around obstacles. But neglect makes things harder.

    The BEIP team is likely to retain its status as the only randomized controlled study of child neglect and its effects. Back when the trial began, some Romanian leaders were arguing that foster care would be worse than institutional care. But now that BEIP has demonstrated the opposite, the study most likely will never be repeated.

    Romania has changed its own policies, according to Zeanah. It began a foster care system when BEIP's results started coming out. Later, he says, it banned the placement of children younger than 2 in institutions. The number of children in Romanian institutions today has dropped to below 20,000; many, according to Zeanah, are severely handicapped children in group homes.

    Nelson says that his work and other Romanian studies have had global impact by, for instance, helping the United Nations Children's Fund in its efforts “to convince governments around the world to stop putting kids in institutions.” There's plenty of convincing to be done: Orphanages around the globe still house an estimated 8 million children.

Log in to view full text