How do we come to recognize expertise in another person and integrate new information with our prior assessments of that person's ability? The brain mechanisms underlying these sorts of evaluations—which are relevant to how we make decisions ranging from whom to hire, whom to marry, and whom to elect to Congress—are the subject of a new study by a team of neuroscientists at the California Institute of Technology (Caltech).

In the study, published in the journal Neuron, Antonio Rangel, Bing Professor of Neuroscience, Behavioral Biology, and Economics, and his associates used functional magnetic resonance imaging (fMRI) to monitor the brain activity of volunteers as they moved through a particular task. Specifically, the subjects were asked to observe the shifting value of a hypothetical financial asset and make predictions about whether it would go up or down. Simultaneously, the subjects interacted with an "expert" who was also making predictions.

Half the time, subjects were shown a photo of a person on their computer screen and told that they were observing that person's predictions. The other half of the time, the subjects were told they were observing predictions from a computer algorithm, and instead of a face, an abstract logo appeared on their screen. However, in every case, the subjects were interacting with a computer algorithm—one programmed to make correct predictions 30, 40, 60, or 70 percent of the time.

Subjects' trust in the expertise of agents, whether "human" or not, was measured by the frequency with which the subjects made bets for the agents' predictions, as well as by the changes in those bets over time as the subjects observed more of the agents' predictions and their consequent accuracy.

This trust, the researchers found, turned out to be strongly linked to the accuracy of the subjects' own predictions of the ups and downs of the asset's value.

"We often speculate on what we would do in a similar situation when we are observing others—what would I do if I were in their shoes?" explains Erie D. Boorman, formerly a postdoctoral fellow at Caltech and now a Sir Henry Wellcome Research Fellow at the Centre for FMRI of the Brain at the University of Oxford, and lead author on the study. "A growing literature suggests that we do this automatically, perhaps even unconsciously."

Indeed, the researchers found that subjects increasingly sided with both "human" agents and computer algorithms when the agents' predictions matched their own. Yet this effect was stronger for "human" agents than for algorithms.

This asymmetry—between the value placed by the subjects on (presumably) human agents and on computer algorithms—was present both when the agents were right and when they were wrong, but it depended on whether or not the agents' predictions matched the subjects'. When the agents were correct, subjects were more inclined to trust the human than algorithm in the future when their predictions matched the subjects' predictions. When they were wrong, human experts were easily and often "forgiven" for their blunders when the subject made the same error. But this "benefit of the doubt" vote, as Boorman calls it, did not extend to computer algorithms. In fact, when computer algorithms made inaccurate predictions, the subjects appeared to dismiss the value of the algorithm's future predictions, regardless of whether or not the subject agreed with its predictions.

Since the sequence of predictions offered by "human" and algorithm agents was perfectly matched across different test subjects, this finding shows that the mere suggestion that we are observing a human or a computer leads to key differences in how and what we learn about them.

A major motivation for this study was to tease out the difference between two types of learning: what Rangel calls "reward learning" and "attribute learning." "Computationally," says Boorman, "these kinds of learning can be described in a very similar way: We have a prediction, and when we observe an outcome, we can update that prediction."

Reward learning, in which test subjects are given money or other valued goods in response to their own successful predictions, has been studied extensively. Social learning—specifically about the attributes of others (or so-called attribute learning)—is a newer topic of interest for neuroscientists. In reward learning, the subject learns how much reward they can obtain, whereas in attribute learning, the subject learns about some characteristic of other people.

This self/other distinction shows up in the subjects' brain activity, as measured by fMRI during the task. Reward learning, says Boorman, "has been closely correlated with the firing rate of neurons that release dopamine"—a neurotransmitter involved in reward-motivated behavior—and brain regions to which they project, such as the striatum and ventromedial prefrontal cortex. Boorman and colleagues replicated previous studies in showing that this reward system made and updated predictions about subjects' own financial reward. Yet during attribute learning, another network in the brain—consisting of the medial prefrontal cortex, anterior cingulate gyrus, and temporal parietal junction, which are thought to be a critical part of the mentalizing network that allows us to understand the state of mind of others—also made and updated predictions, but about the expertise of people and algorithms rather than their own profit.

The differences in fMRIs between assessments of human and nonhuman agents were subtler. "The same brain regions were involved in assessing both human and nonhuman agents," says Boorman, "but they were used differently."

"Specifically, two brain regions in the prefrontal cortex—the lateral orbitofrontal cortex and medial prefrontal cortex—were used to update subjects' beliefs about the expertise of both humans and algorithms," Boorman explains. "These regions show what we call a 'belief update signal.'" This update signal was stronger when subjects agreed with the "human" agents than with the algorithm agents and they were correct. It was also stronger when they disagreed with the computer algorithms than when they disagreed with the "human" agents and they were incorrect. This finding shows that these brain regions are active when assigning credit or blame to others.

"The kind of learning strategies people use to judge others based on their performance has important implications when it comes to electing leaders, assessing students, choosing role models, judging defendents, and so on," Boorman notes. Knowing how this process happens in the brain, says Rangel, "may help us understand to what extent individual differences in our ability to assess the competency of others can be traced back to the functioning of specific brain regions."

The study, "The Behavioral and Neural Mechanisms Underlying the Tracking of Expertise," was also coauthored by John P. O'Doherty, professor of psychology and director of the Caltech Brain Imaging Center, and Ralph Adolphs, Bren Professor of Psychology and Neuroscience and professor of biology. The research was supported by the National Science Foundation, the National Institutes of Health, the Betty and Gordon Moore Foundation, the Lipper Foundation, and the Wellcome Trust.

South-central Idaho and the surface of Mars have an interesting geological feature in common: amphitheater-headed canyons. These U-shaped canyons with tall vertical headwalls are found near the Snake River in Idaho as well as on the surface of Mars, according to photographs taken by satellites. Various explanations for how these canyons formed have been offered—some for Mars, some for Idaho, some for both—but in a paper published the week of December 16 in the online issue of Proceedings of the National Academy of Sciences, Caltech professor of geology Michael P. Lamb, Benjamin Mackey, formerly a postdoctoral fellow at Caltech, and W. M. Keck Foundation Professor of Geochemistry Kenneth A. Farley offer a plausible account that all these canyons were created by enormous floods.

Canyons in Malad Gorge State Park, Idaho, are carved into a relatively flat plain composed of a type of volcanic rock known as basalt. The basalt originated from a hotspot, located in what is now Yellowstone Park, which has been active for the last few million years. Two canyons in Malad Gorge, Woody's Cove and Stubby Canyon, are characterized by tall vertical headwalls, roughly 150 feet high, that curve around to form an amphitheater. Other amphitheater-headed canyons can be found nearby, outside the Gorge—Box Canyon, Blue Lakes Canyon, and Devil's Corral—and also elsewhere on Earth, such as in Iceland.

To figure out how they formed, Lamb and Mackey conducted field surveys and collected rock samples from Woody's Cove, Stubby Canyon, and a third canyon in Malad Gorge, known as Pointed Canyon. As its name indicates, Pointed Canyon ends not in an amphitheater but in a point, as it progressively narrows in the upstream direction toward the plateau at an average 7 percent grade. Through Pointed Canyon flows the Wood River, a tributary of the larger Snake River, which in turn empties into the Columbia River on its way to the Pacific Ocean.

Geologists have a good understanding of how the rocks in Woody's Cove and Stubby Canyon achieved their characteristic appearance. The lava flows that hardened into basalt were initially laid down in layers, some more than six feet thick. As the lava cooled, it contracted and cracked, just as mud does when it dries. This produced vertical cracks across the entire layer of lava-turned-basalt. As each additional sheet of lava covered the same land, it too cooled and cracked vertically, leaving a wall that, when exposed, looks like stacks of tall blocks, slightly offset from one another with each additional layer. This type of structure is called columnar basalt.

While the formation of columnar basalt is well understood, it is not clear how, at Woody's Cove and Stubby Canyon, the vertical walls became exposed or how they took on their curved shapes. The conventional explanation is that the canyons were formed via a process called "groundwater sapping," in which springs at the bottom of the canyon gradually carve tunnels at the base of the rock wall until this undercutting destabilizes the structure so much that blocks or columns of basalt fall off from above, creating the amphitheater below.

This explanation has not been corroborated by the Caltech team's observations, for two reasons. First, there is no evidence of undercutting, even though there are existing springs at the base of Woody's Cove and Stubby Canyon. Second, undercutting should leave large boulders in place at the foot of the canyon, at least until they are dissolved or carried away by groundwater. "These blocks are too big to move by spring flow, and there's not enough time for the groundwater to have dissolved them away," Lamb explains, "which means that large floods are needed to move them out. To make a canyon, you have to erode the canyon headwall, and you also have to evacuate the material that collapses in."

That leaves waterfall erosion during a large flood event as the only remaining candidate for the canyon formation that occurred in Malad Gorge, the Caltech team concludes.

No water flows over the top of Woody's Cove and Stubby Canyon today. But even a single incident of overland water flow occurring during an unusually large flood event could pluck away and topple boulders from the columnar basalt, taking advantage of the vertical fracturing already present in the volcanic rock. A flood of this magnitude could also carry boulders downstream, leaving behind the amphitheater canyons we see today without massive boulder piles at their bottoms and with no existing watercourses.

Additional evidence that at some point in the past water flowed over the plateaus near Woody's Cove and Stubby Canyon are the presence of scour marks on surface rocks on the plateau above the canyons. These scour marks are evidence of the type of abrasion that occurs when a water discharge containing sediment moves overland.

Taken together, the evidence from Malad Gorge, Lamb says, suggests that "amphitheater shapes might be diagnostic of very large-scale floods, which would imply much larger water discharges and much shorter flow durations than predicted by the previous groundwater theory." Lamb points out that although groundwater sapping "is often assumed to explain the origin of amphitheater-headed canyons, there is no place on Earth where it has been demonstrated to work in columnar basalt."

Closing the case on the canyons at Malad Gorge required one further bit of information: the ages of the rock samples. This was accomplished at Caltech's Noble Gas Lab, run by Kenneth A. Farley, W. M. Keck Foundation Professor of Geochemistry and chair of the Division of Geological and Planetary Sciences.

The key to dating surface rocks on Earth is cosmic rays—very high-energy particles from space that regularly strike Earth. "Cosmic rays interact with the atmosphere and eventually with rocks at the surface, producing alternate versions of noble gas elements, or isotopes, called cosmogenic nuclides," Lamb explains. "If we know the cosmic-ray flux, and we measure the accumulation of nuclides in a certain mineral, then we can calculate the time that rock has been sitting at Earth's surface."

At the Noble Gas Lab, Farley and Mackey determined that rock samples from the heads of Woody's Cove and Stubby Canyon had been exposed for the same length of time, approximately 46,000 years. If Lamb and his colleagues are correct, this is when the flood event occurred that plucked the boulders off the canyon walls, leaving the amphitheaters behind.

Further evidence supporting the team's theory can be found in Pointed Canyon. Rock samples collected along the walls of the first kilometer of the canyon show progressively more exposure in the downstream direction, suggesting that the canyon is still being carved by Wood River. Using the dates of exposure revealed in the rock samples, Lamb reconstructed the probable location of Pointed Canyon at the time of the formation of Woody's Cove and Stubby Canyon. At that location, where the rock has been exposed approximately 46,000 years, the surrounding canyon walls form the characteristic U-shape of an amphitheater-headed canyon and then abruptly narrow into the point that forms the remainder of Pointed Canyon. "The same megaflood event that created Woody's Cove and Stubby Canyon seems to have created Pointed Canyon," Lamb concludes. "The only difference is that the other canyons had no continuing river action, while Pointed Canyon was cut relatively slowly over the last 46,000 years by the Wood River, which is not powerful enough to topple and pluck basalt blocks from the surrounding plateau, resulting in a narrow channel rather than tall vertical headwalls."

Solving the puzzle of how amphitheater-headed canyons are created has implications reaching far beyond south-central Idaho because similar features—though some much larger—are also present on the surface of Mars. "A very popular interpretation for the amphitheater-headed canyons on Mars is that groundwater seeps out of cracks at the base of the canyon headwalls and that no water ever went over the top," Lamb says. Judging from the evidence in Idaho, however, it seems more likely that on Mars, as on Earth, amphitheater-headed canyons were created by enormous flood events, suggesting that Mars was once a very watery planet.

The paper presenting these results is entitled "Amphitheater-Headed Canyons Formed by Megaflooding at Malad Gorge, Idaho." The work was supported by grants from the National Science Foundation and NASA.

Introduction

Habitability, Taphonomy, and the Search for Organic Carbon on Mars: Introduction
Grotzinger, John P.

Research Articles

Volatile and Organic Compositions of Sedimentary Rocks in Yellowknife Bay, Gale crater, Mars
Ming, D. W. et al.

Mars' Surface Radiation Environment Measured with the Mars Science Laboratory's Curiosity Rover
Hassler, Donald M. et al.

Elemental Geochemistry of Sedimentary Rocks at Yellowknife Bay, Gale Crater, Mars
McLennan, S. M. et al.

In Situ Radiometric and Exposure Age Dating of the Martian Surface
Farley, K. A. et al.

A Habitable Fluvio-Lacustrine Environment at Yellowknife Bay, Gale Crater, Mars
Grotzinger, J. P. et al.

Mineralogy of a Mudstone at Yellowknife Bay, Gale Crater, Mars
Vaniman, D. T. et al.

Although researchers have determined the ages of rocks from other planetary bodies, the actual experiments—like analyzing meteorites and moon rocks—have always been done on Earth. Now, for the first time, researchers have successfully determined the age of a Martian rock—with experiments performed on Mars. The work, led by geochemist Ken Farley of the California Institute of Technology (Caltech), could not only help in understanding the geologic history of Mars but also aid in the search for evidence of ancient life on the planet.

Many of the experiments carried out by the Mars Science Laboratory (MSL) mission's Curiosity rover were painstakingly planned by NASA scientists more than a decade ago. However, shortly before the rover left Earth in 2011, NASA's participating scientist program asked researchers from all over the world to submit new ideas for experiments that could be performed with the MSL's already-designed instruments. Farley, W.M. Keck Foundation Professor of Geochemistry and one of the 29 selected participating scientists, submitted a proposal that outlined a set of techniques similar to those already used for dating rocks on Earth, to determine the age of rocks on Mars. Findings from the first such experiment on the Red Planet—published by Farley and coworkers this week in a collection of Curiosity papers in the journal Science Express—provide the first age determinations performed on another planet.

The paper is one of six appearing in the journal that reports results from the analysis of data and observations obtained during Curiosity's exploration at Yellowknife Bay—an expanse of bare bedrock in Gale Crater about 500 meters from the rover's landing site. The smooth floor of Yellowknife Bay is made up of a fine-grained sedimentary rock, or mudstone, that researchers think was deposited on the bed of an ancient Martian lake.

In March, Curiosity drilled holes into the mudstone and collected powdered rock samples from two locations about three meters apart. Once the rock samples were drilled, Curiosity's robotic arm delivered the rock powder to the Sample Analysis on Mars (SAM) instrument, where it was used for a variety of chemical analyses, including the geochronology—or rock dating—techniques.

One technique, potassium-argon dating, determines the age of a rock sample by measuring how much argon gas it contains. Over time, atoms of the radioactive form of potassium—an isotope called potassium-40—will decay within a rock to spontaneously form stable atoms of argon-40. This decay occurs at a known rate, so by determining the amount of argon-40 in a sample, researchers can calculate the sample's age.

Although the potassium-argon method has been used to date rocks on Earth for many decades, these types of measurements require sophisticated lab equipment that could not easily be transported and used on another planet. Farley had the idea of performing the experiment on Mars using the SAM instrument. There, the sample was heated to temperatures high enough that the gasses within the rock were released and could be analyzed by an onboard mass spectrometer.

Farley and his colleagues determined the age of the mudstone to be about 3.86 to 4.56 billion years old. "In one sense, this is an utterly unsurprising result—it's the number that everybody expected," Farley says.

Indeed, prior to Curiosity's geochronology experiment, researchers using the "crater counting" method had estimated the age of Gale Crater and its surroundings to be between 3.6 and 4.1 billion years old. Crater counting relies on the simple fact that planetary surfaces are repeatedly bombarded with objects that scar their surface with impact craters; a surface with many impact craters is presumed to be older than one with fewer craters. Although this method is simple, it has large uncertainties.

"What was surprising was that our result—from a technique that was implemented on Mars with little planning on Earth—got a number that is exactly what crater counting predicted," Farley says. "MSL instruments weren't designed for this purpose, and we weren't sure if the experiment was going to work, but the fact that our number is consistent with previous estimates suggests that the technique works, and it works quite well."

The researchers do, however, acknowledge that there is some uncertainty in their measurement. One reason is that mudstone is a sedimentary rock—formed in layers over a span of millions of years from material that eroded off of the crater walls—and thus the age of the sample drilled by Curiosity really represents the combined age of those bits and pieces. So while the mudstone indicates the existence of an ancient lake—and a habitable environment some time in the planet's distant past—neither crater counting nor potassium-argon dating can directly determine exactly when this was.

To provide an answer for how the geology of Yellowknife Bay has changed over time, Farley and his colleagues also designed an experiment using a method called surface exposure dating. "The surface of Mars, the surface of Earth, and basically all surfaces in the solar system are being bombarded by cosmic rays," explains Farley, and when these rays—very high-energy protons—blast into an atom, the atom's nucleus shatters, creating isotopes of other elements. Cosmic rays can only penetrate about two to three meters below the surface, so the abundance of cosmic-ray-debris isotopes in rock indicates how long that rock has been on the surface.

Using the SAM mass spectrometer to measure the abundance of three isotopes that result from cosmic-ray bombardment—helium-3, neon-21, and argon-36—Farley and his colleagues calculated that the mudstone at Yellowknife Bay has been exposed at the surface for about 80 million years. "All three of the isotopes give exactly the same answer; they all have their independent sources of uncertainty and complications, but they all give exactly the same answer. That is probably the most remarkable thing I've ever seen as a scientist, given the difficulty of the analyses," Farley says.

This also helps researchers looking for evidence of past life on Mars. Cosmic rays are known to degrade the organic molecules that may be telltale fossils of ancient life. However, because the rock at Yellowknife Bay has only been exposed to cosmic rays for 80 million years—a relatively small sliver of geologic time—"the potential for organic preservation at the site where we drilled is better than many people had guessed," Farley says.

Furthermore, the "young" surface exposure offers insight into the erosion history of the site. "When we first came up with this number, the geologists said, 'Yes, now we get it, now we understand why this rock surface is so clean and there is no sand or rubble,'" Farley says. 

The exposure of rock in Yellowknife Bay has been caused by wind erosion. Over time, as wind blows sand against the small cliffs, or scarps, that bound the Yellowknife outcrop, the scarps erode back, revealing new rock that previously was not exposed to cosmic rays.

"Imagine that you are in this site a hundred million years ago; the area that we drilled in was covered by at least a few meters of rock. At 80 million years ago, wind would have caused this scarp to migrate across the surface and the rock below the scarp would have gone from being buried—and safe from cosmic rays—to exposed," Farley explains. Geologists have developed a relatively well-understood model, called the scarp retreat model, to explain how this type of environment evolves. "That gives us some idea about why the environment looks like it does and it also gives us an idea of where to look for rocks that are even less exposed to cosmic rays," and thus are more likely to have preserved organic molecules, Farley says.

Curiosity is now long gone from Yellowknife Bay, off to new drilling sites on the route to Mount Sharp where more dating can be done. "Had we known about this before we left Yellowknife Bay, we might have done an experiment to test the prediction that cosmic-ray irradiation should be reduced as you go in the downwind direction, closer to the scarp, indicating a newer, more recently exposed rock, and increased irradiation when you go in the upwind direction, indicating a rock exposed to the surface longer ago," Farley says. "We'll likely drill in January, and the team is definitely focused on finding another scarp to test this on."

This information could also be important for Curiosity chief scientist John Grotzinger, Caltech's Fletcher Jones Professor of Geology. In another paper in the same issue of Science Express, Grotzinger—who studies the history of Mars as a habitable environment—and colleagues examined the physical characteristics of the rock layers in and near Yellowknife Bay. They concluded that the environment was habitable less than 4 billion years ago, which is a relatively late point in the planet's history.

"This habitable environment existed later than many people thought possible," Grotzinger says. His findings suggest that the surface water on Mars at that time would have been sufficient enough to make clays. Previously, such clays—evidence of a habitable environment—were thought to have washed in from older deposits. Knowing that the clays could be produced later in locations with surface water can help researchers pin down the best areas at which to look for once habitable environments, he says.

Farley's work is published in a paper titled "In-situ radiometric and exposure age dating of the Martian surface." Other Caltech coauthors on the study include Grotzinger, graduate student Hayden B. Miller, and Edward Stolper.

Autism spectrum disorder (ASD) is diagnosed when individuals exhibit characteristic behaviors that include repetitive actions, decreased social interactions, and impaired communication. Curiously, many individuals with ASD also suffer from gastrointestinal (GI) issues, such as abdominal cramps and constipation.

Using the co-occurrence of brain and gut problems in ASD as their guide, researchers at the California Institute Technology (Caltech) are investigating a potentially transformative new therapy for autism and other neurodevelopmental disorders.

The gut microbiota—the community of bacteria that populate the human GI tract—previously has been shown to influence social and emotional behavior, but the Caltech research, published online in the December 5 issue of the journal Cell, is the first to demonstrate that changes in these gut bacteria can influence autism-like behaviors in a mouse model.

"Traditional research has studied autism as a genetic disorder and a disorder of the brain, but our work shows that gut bacteria may contribute to ASD-like symptoms in ways that were previously unappreciated," says Professor of Biology Sarkis K. Mazmanian. "Gut physiology appears to have effects on what are currently presumed to be brain functions."

To study this gut–microbiota–brain interaction, the researchers used a mouse model of autism previously developed at Caltech in the laboratory of Paul H. Patterson, the Anne P. and Benjamin F. Biaggini Professor of Biological Sciences. In humans, having a severe viral infection raises the risk that a pregnant woman will give birth to a child with autism. Patterson and his lab reproduced the effect in mice using a viral mimic that triggers an infection-like immune response in the mother and produces the core behavioral symptoms associated with autism in the offspring.

In the new Cell study, Mazmanian, Patterson, and their colleagues found that the "autistic" offspring of immune-activated pregnant mice also exhibited GI abnormalities. In particular, the GI tracts of autistic-like mice were "leaky," which means that they allow material to pass through the intestinal wall and into the bloodstream. This characteristic, known as intestinal permeability, has been reported in some autistic individuals. "To our knowledge, this is the first report of an animal model for autism with comorbid GI dysfunction," says Elaine Hsiao, a senior research fellow at Caltech and the first author on the study.

To see whether these GI symptoms actually influenced the autism-like behaviors, the researchers treated the mice with Bacteroides fragilis, a bacterium that has been used as an experimental probiotic therapy in animal models of GI disorders.

The result? The leaky gut was corrected.

In addition, observations of the treated mice showed that their behavior had changed. In particular, they were more likely to communicate with other mice, had reduced anxiety, and were less likely to engage in a repetitive digging behavior.

"The B. fragilis treatment alleviates GI problems in the mouse model and also improves some of the main behavioral symptoms," Hsiao says. "This suggests that GI problems could contribute to particular symptoms in neurodevelopmental disorders."

With the help of clinical collaborators, the researchers are now planning a trial to test the probiotic treatment on the behavioral symptoms of human autism. The trial should begin within the next year or two, says Patterson.

"This probiotic treatment is postnatal, which means that the mother has already experienced the immune challenge, and, as a result, the growing fetuses have already started down a different developmental path," Patterson says. "In this study, we can provide a treatment after the offspring have been born that can help improve certain behaviors. I think that's a powerful part of the story."

The researchers stress that much work is still needed to develop an effective and reliable probiotic therapy for human autism—in part because there are both genetic and environmental contributions to the disorder, and because the immune-challenged mother in the mouse model reproduces only the environmental component.

"Autism is such a heterogeneous disorder that the ratio between genetic and environmental contributions could be different in each individual," Mazmanian says. "Even if B. fragilis ameliorates some of the symptoms associated with autism, I would be surprised if it's a universal therapy—it probably won't work for every single case."

The Caltech team proposes that particular beneficial bugs are intimately involved in regulating the release of metabolic products (or metabolites) from the gut into the bloodstream. Indeed, the researchers found that in the leaky intestinal wall of the autistic-like mice, certain metabolites that were modulated by microbes could both easily enter the circulation and affect particular behaviors.

"I think our results may someday transform the way people view possible causes and potential treatments for autism," Mazmanian says.

Along with Patterson, Hsiao, and Mazmanian, additional Caltech coauthors on the paper, "Microbiota Modulate Behavioral and Physiological Abnormalities Associated with Neurodevelopmental Disorders," are Sara McBride, Sophia Hsien, Gil Sharon, Julian A. Codelli, Janet Chow, and Sarah E. Reisman. The work was supported by a Caltech Innovation Initiative grant, an Autism Speaks Weatherstone Fellowship, a National Institutes of Health/National Research Service Award Ruth L. Kirschstein Predoctoral Fellowship, a Human Frontiers Science Program Fellowship, a Department Of Defense Graduate Fellowship, a National Science Foundation Graduate Research Fellowship, an Autism Speaks Trailblazer Award, a Caltech Grubstake award, a Congressionally Directed Medical Research Award, a Weston Havens Foundation Award, several Callie McGrath Charitable Foundation awards, and the National Institute of Mental Health.

The Subaru Telescope, an 8.2-meter telescope operated by the National Astronomical Observatory of Japan, has been combing the night sky since 1999. Located at the Mauna Kea Observatories in Hawaii, the telescope has been systematically surveying each degree of space, whether it looks promising or not, in search of objects worthy of further investigation. One of the most fascinating objects to emerge from the Subaru Telescope's wide-field survey—Himiko—was discovered in 2009. Himiko, a "space blob" named after a legendary queen from ancient Japan, is a simply enormous galaxy, with a hot glowing gaseous halo extending over 55,000 light-years. Not only is Himiko very large, it is extraordinarily distant, seen at a time approximately 800 million years after the Big Bang, when the universe was only 6 percent of its present size and stars and galaxies were just beginning to form.

How could such an early galaxy have sufficient energy to power such a vast glowing gas cloud? In search of the answer to this question, Richard Ellis, the Steele Family Professor of Astronomy at Caltech, together with colleagues from the University of Tokyo and the Harvard-Smithsonian Center for Astrophysics, undertook an exploration of Himiko using the combined resources of the Hubble Space Telescope and the new Atacama Large Millimeter/submillimeter Array (ALMA) in Chile's Atacama Desert. The data collected through these observations answered the initial question about the source of energy powering Himiko, but revealed some puzzling data as well.

The Hubble images, receiving optical and ultraviolet light, reveal three stellar clumps covering a space of 20,000 light-years. Each clump is the size of a typical luminous galaxy dating to the epoch of Himiko. Together, the clumps achieve a prodigious rate of star formation, equivalent to about one hundred solar masses per year. This is more than sufficient to explain the existence of Himiko and its gaseous halo. The observation of the three stellar clumps is exciting in itself, as it means that Himiko is a "triple merger," which, according to Ellis, is "a remarkably rare event."

But a surprising anomaly emerged when Himiko was observed by ALMA. Although the giant gas cloud was bustling with energy at ultraviolet and optical frequencies, it was comparatively sleepy in the submillimeter and radio ranges that ALMA detects. Ordinarily, intense star formation creates dust clouds that are composed of elements such as carbon, oxygen, and silicon, which are heavy in comparison to the hydrogen and helium of the early universe. When these dust clouds are heated up by the ultraviolet light emitted by the developing stars, the dust reradiates the ultraviolet light out into the universe at radio wavelengths. But ALMA did not receive significant radio signals from Himiko, suggesting that heavier elements are not present. Also missing was the spectral signature associated with the emission of gaseous carbon, something also common in galaxies with intense star formation.

Both of these nondetections—of substantial radio waves and of gaseous carbon—are perplexing since carbon is ordinarily rapidly synthesized in young stars. Indeed, carbon emission has heretofore been recommended as a tracer of star formation in distant galaxies. But, as Ellis and his fellow astronomers found, Himiko does not contain the dust clouds of heavier elements that astronomers find in typical energetic galaxies. Instead its interstellar gas is composed of hydrogen and helium—primitive materials formed in the Big Bang itself.

Ellis and his fellow astronomers did not come to this conclusion quickly. They first carefully ruled out several other possible explanations for Himiko, including that the giant blob is being created by the magnification of a foreground object by a phenomenon known as gravitational lensing, or is being powered by a massive black hole at its center. Ultimately, the team concluded that Himiko is most likely a primordial galaxy caught in the moment of its formation between 400 million to 1 billion years after the Big Bang, a period astronomers term the cosmic dawn.

"Astronomers are usually excited when a signal from an object is detected," Ellis says, "but in this case it's the absence of a signal from heavy elements that is the most exciting result!"

The paper reporting the results of this research, titled "An Intensely Star-Forming Galaxy at Z ~ 7 with Low Dust and Metal Content Revealed by Deep ALMA and HST Observations," will be published in the December 1, 2013, issue of the Astrophysical Journal. The work was funded by NASA through a grant from the Space Telescope Science Institute, the World Premier International Research Center Initiative (WPI Initiative), and the Japan Society for the Promotion of Science (JSPS).

Difficulties in social interaction are considered to be one of the behavioral hallmarks of autism spectrum disorders (ASDs). Previous studies have shown these difficulties to be related to differences in how the brains of autistic individuals process sensory information about faces. Now, a group of researchers led by California Institute of Technology (Caltech) neuroscientist Ralph Adolphs has made the first recordings of the firings of single neurons in the brains of autistic individuals, and has found specific neurons in a region called the amygdala that show reduced processing of the eye region of faces. Furthermore, the study found that these same neurons responded more to mouths than did the neurons seen in the control-group individuals.

"We found that single brain cells in the amygdala of people with autism respond differently to faces in a way that explains many prior behavioral observations," says Adolphs, Bren Professor of Psychology and Neuroscience and professor of biology at Caltech and coauthor of a study in the November 20 issue of Neuron that outlines the team's findings. "We believe this shows that abnormal functioning in the amygdala is a reason that people with autism process faces abnormally."

The amygdala has long been known to be important for the processing of emotional reactions. To make recordings from this part of the brain, Adolphs and lead author Ueli Rutishauser, assistant professor in the departments of neurosurgery and neurology at Cedars-Sinai Medical Center and visiting associate in biology at Caltech, teamed up with Adam Mamelak, professor of neurosurgery and director of functional neurosurgery at Cedars-Sinai, and neurosurgeon Ian Ross at Huntington Memorial Hospital in Pasadena, California, to recruit patients with epilepsy who had electrodes implanted in their medial temporal lobes—the area of the brain where the amygdala is located—to help identify the origin of their seizures. Epileptic seizures are caused by a burst of abnormal electric activity in the brain, which the electrodes are designed to detect. It turns out that epilepsy and ASD sometimes go together, and so the researchers were able to identify two of the epilepsy patients who also had a diagnosis of ASD.

By using the implanted electrodes to record the firings of individual neurons, the researchers were able to observe activity as participants looked at images of different facial regions, and then correlate the neuronal responses with the pictures. In the control group of epilepsy patients without autism, the neurons responded most strongly to the eye region of the face, whereas in the two ASD patients, the neurons responded most strongly to the mouth region. Moreover, the effect was present in only a specific subset of the neurons. In contrast, a different set of neurons showed the same response in both groups when whole faces were shown.

"It was surprising to find such clear abnormalities at the level of single cells," explains Rutishauser. "We, like many others, had thought that the neurological abnormalities that contribute to autism were spread throughout the brain, and that it would be difficult to find highly specific correlates. Not only did we find highly specific abnormalities in single-cell responses, but only a certain subset of cells responded that way, while another set showed typical responses to faces. This specificity of these cell populations was surprising and is, in a way, very good news, because it suggests the existence of specific mechanisms for autism that we can potentially trace back to their genetic and environmental causes, and that one could imagine manipulating for targeted treatment."

"We can now ask how these cells change their responses with treatments, how they correspond to similar cell populations in animal models of autism, and what genes this particular population of cells expresses," adds Adolphs.

To validate their results, the researchers hope to identify and test additional subjects, which is a challenge because it is very hard to find people with autism who also have epilepsy and who have been implanted with electrodes in the amygdala for single-cell recordings, says Adolphs.

"At the same time, we should think about how to change the responses of these neurons, and see if those modifications correlate with behavioral changes," he says.

Funding for the research outlined in the Neuron paper, titled "Single-neuron correlates of abnormal face processing in autism," was provided by the Simons Foundation, the Gordon and Betty Moore Foundation, the Cedars-Sinai Medical Center, Autism Speaks, and the National Institute of Mental Health. Additional coauthors were Caltech postdoctoral scholar Oana Tudusciuc and graduate student Shuo Wang.

In developing nations, rural areas, and even one's own home, limited access to expensive equipment and trained medical professionals can impede the diagnosis and treatment of disease. Many qualitative tests that provide a simple "yes" or "no" answer (like an at-home pregnancy test) have been optimized for use in these resource-limited settings. But few quantitative tests—those able to measure the precise concentration of biomolecules, not just their presence or absence—can be done outside of a laboratory or clinical setting. By leveraging their discovery of the robustness of "digital," or single-molecule quantitative assays, researchers at the California Institute of Technology (Caltech) have demonstrated a method for using a lab-on-a-chip device and a cell phone to determine a concentration of molecules, such as HIV RNA molecules, in a sample. This digital approach can consistently provide accurate quantitative information despite changes in timing, temperature, and lighting conditions, a capability not previously possible using traditional measurements.

In a study published on November 7 in the journal Analytical Chemistry, researchers in the laboratory of Rustem Ismagilov, Ethel Wilson Bowles and Robert Bowles Professor of Chemistry and Chemical Engineering, used HIV as the context for testing the robustness of digital assays. In order to assess the progression of HIV and recommend appropriate therapies, doctors must know the concentration of HIV RNA viruses in a patient's bloodstream, called a viral load. The problem is that the viral load tests used in the United States, such as those that rely on amplification of RNA via polymerase chain reaction (PCR), require bulky and expensive equipment, trained personnel, and access to infrastructure such as electricity, all of which are often not available in resource-limited settings. Furthermore, because it is difficult to control the environment in these settings, viral load tests must be "robust," or resilient to changes such as temperature and humidity fluctuations.

Many traditional approaches for measuring viral load involve converting a small quantity of RNA into DNA, which is then multiplied through DNA amplification—allowing researchers to see how much DNA is present in real time after each round of amplification, by monitoring the varying intensity of a fluorescent dye marking the DNA. These experiments—known as "kinetic" assays—result in a readout reflecting changes in intensity over time, called an amplification curve. To find the original concentration of the beginning bulk RNA sample, the amplification curve is then compared with standard curves representing known concentrations of RNA. Since assays, such as those for HIV, require many rounds of DNA amplification to collect a sufficiently bright fluorescent signal, small errors introduced by changes in environmental conditions can compound exponentially—meaning that these kinetic measurements are not robust enough to withstand changing conditions.

In this new study, the researchers hypothesized that they could use a digital amplification approach to create a robust quantitative technique. In digital amplification, a sample is split into enough small volumes such that each well contains either a single target molecule or no molecule at all. Ismagilov and his colleagues used a microfluidic device they previously invented, called SlipChip, to compartmentalize single molecules from a sample containing HIV RNA. SlipChip is made up of two credit card-sized plates stacked atop one another; the sample is first added to the interconnected channels of the SlipChip, and with a single "slip" of the top chip, the channels turn into individual wells.

In lieu of PCR, the researchers used a different amplification chemistry on this chip called digital reverse transcription-loop-mediated amplification (dRT-LAMP), which produces a bright fluorescent signal in the presence of a target molecule during the amplification process. The dRT-LAMP technique eliminates the need for continuous tracking of the intensity of fluorescence; instead, just one end-point readout measurement is used. The resulting patchwork of "positive" or "negative" wells on the device, in combination with statistical analysis, enables single molecules to be counted.

"In each well, you are performing a qualitative experiment; the result is like a pregnancy test: either yes or no, positive or negative, for the presence of an HIV RNA molecule," says David Selck, a graduate student in Ismagilov's lab and a first author on the study. "But by doing a couple of thousand qualitative experiments, you end up getting a numerical, quantitative result: the concentration of HIV RNA molecules in the sample. By calculating the concentration from the number of wells that contain fluorescence—and therefore HIV—you're leveraging the robustness of many qualitative 'yes or no' experiments to fulfill the need for a quantitative, numerical result," he says.

When the researchers compared quantification results from dRT-LAMP to those obtained by the real-time, kinetic version of this chemistry, RT-LAMP, they found that the digital format provided accurate results despite changes in temperature and time, while the kinetic format could not. This finding adds to a body of research that the laboratory has been developing on the robustness of converting analog signals (i.e., a readout reflecting a changing concentration over time) into a series of positive or negative digital signals. Another recent paper, published in the Journal of the American Chemical Society, explored a variation on this analog-to-digital conversion.

Ismagilov's group also tested a way to take an image of the fluorescence pattern in the wells of the SlipChip and, from that image, determine the viral load—without the use of expensive microscopes or trained staff. They turned to a nearly ubiquitous 21^st-century technology: the smartphone.

The researchers placed the SlipChip in a makeshift darkroom (a shoebox with a hole in the top) and then photographed its wells using a smartphone outfitted with a special filter attachment—so that the smartphone flash would be able to "excite" the fluorescent DNA dye, and the smartphone camera could capture an image of the fluorescence. The resulting images were uploaded to Microsoft SkyDrive, a cloud-based server, where custom software—designed by the researchers—determined the viral load concentration and sent the results back in an email. These capabilities allow the digital approach to perform reliably with automated processing, regardless of how poor the imaging conditions may be. As an example of its simplicity, a 5-year-old child was able to use this cell phone imaging method to obtain quantitative results using strands of RNA extracted from a noninfectious virus (a video of this demonstration is available on the Ismagilov lab's YouTube channel).

"We were surprised that this cell phone method worked, because both cell phone imaging and automated processing are error prone," Ismagilov says. "Because digital assays involve simply distinguishing positives from negatives, we found that even these error-prone approaches can be used to count single molecules reliably."

The fact that this method is robust not only to changes in time and temperature but also is amenable to cell phone imaging and automated processing makes it a promising technology for limited-resource settings. "We believe that our findings of the robustness of digital amplification could signal a major paradigm shift in how quantitative measurements are obtained at home, in the field, and in developing countries," Ismagilov says.

The researchers stress that there is still room for improvement, however. "While in this study we were examining robustness and used purified RNA, the next generation of devices will isolate HIV RNA molecules directly from patients' blood," says Bing Sun, a graduate student in Ismagilov's lab and a first author on the study. "We will also adapt the devices for other viruses, such as hepatitis C. By combining these improvements with the cell phone imaging method, we plan to create something that could actually be used in the real world," Sun adds.

The paper is titled "Increased Robustness of Single-Molecule Counting with Microfluidics, Digital Isothermal Amplification, and a Mobile Phone versus Real-Time Kinetic Measurements." In addition to Selck, Sun, and Ismagilov, the paper is coauthored by Mikhail A. Karymov, an associate scientist at Caltech. The work was funded by the Defense Advanced Research Projects Agency award number HR0011-11-2-0006, and by the National Institutes of Health award numbers R01EB012946 and 5DP1OD003584. Microfluid technologies developed by Ismagilov's group have been licensed to Emerald BioStructures, Randance Technologies, and SlipChip LLC.

Black holes—massive objects in space with gravitational forces so strong that not even light can escape them—come in a variety of sizes. On the smaller end of the scale are the stellar-mass black holes that are formed during the deaths of stars. At the larger end are supermassive black holes, which contain up to one billion times the mass of our sun. Over billions of years, small black holes can slowly grow into the supermassive variety by taking on mass from their surroundings and also by merging with other black holes. But this slow process can't explain the problem of supermassive black holes existing in the early universe—such black holes would have formed less than one billion years after the Big Bang.

Now new findings by researchers at the California Institute of Technology (Caltech) may help to test a model that solves this problem.

Certain models of supermassive black hole growth invoke the presence of "seed" black holes that result from the deaths of very early stars. These seed black holes gain mass and increase in size by picking up the materials around them—a process called accretion—or by merging with other black holes. "But in these previous models, there was simply not enough time for any black hole to reach a supermassive scale so soon after the birth of the universe," says Christian Reisswig, NASA Einstein Postdoctoral Fellow in Astrophysics at Caltech and the lead author of the study. "The growth of black holes to supermassive scales in the young universe seems only possible if the 'seed' mass of the collapsing object was already sufficiently large," he says.

To investigate the origins of young supermassive black holes, Reisswig, in collaboration with Christian Ott, assistant professor of theoretical astrophysics, and their colleagues turned to a model involving supermassive stars. These giant, rather exotic stars are hypothesized to have existed for just a brief time in the early universe. Unlike ordinary stars, supermassive stars are stabilized against gravity mostly by their own photon radiation. In a very massive star, photon radiation—the outward flux of photons that is generated due to the star's very high interior temperatures—pushes gas from the star outward in opposition to the gravitational force that pulls the gas back in. When the two forces are equal, this balance is called hydrostatic equilibrium.

During its life, a supermassive star slowly cools due to energy loss through the emission of photon radiation. As the star cools, it becomes more compact, and its central density slowly increases. This process lasts for a couple of million years until the star has reached sufficient compactness for gravitational instability to set in and for the star to start collapsing gravitationally, Reisswig says.

Previous studies predicted that when supermassive stars collapse, they maintain a spherical shape that possibly becomes flattened due to rapid rotation. This shape is called an axisymmetric configuration. Incorporating the fact that very rapidly spinning stars are prone to tiny perturbations, Reisswig and his colleagues predicted that these perturbations could cause the stars to deviate into non-axisymmetric shapes during the collapse. Such initially tiny perturbations would grow rapidly, ultimately causing the gas inside the collapsing star to clump and to form high-density fragments.

These fragments would orbit the center of the star and become increasingly dense as they picked up matter during the collapse; they would also increase in temperature. And then, Reisswig says, "an interesting effect kicks in." At sufficiently high temperatures, there would be enough energy available to match up electrons and their antiparticles, or positrons, into what are known as electron-positron pairs. The creation of electron-positron pairs would cause a loss of pressure, further accelerating the collapse; as a result, the two orbiting fragments would ultimately become so dense that a black hole could form at each clump. The pair of black holes might then spiral around one another before merging to become one large black hole. "This is a new finding," Reisswig says. "Nobody has ever predicted that a single collapsing star could produce a pair of black holes that then merge."

Reisswig and his colleagues used supercomputers to simulate a supermassive star that is on the verge of collapse. The simulation was visualized with a video made by combining millions of points representing numerical data about density, gravitational fields, and other properties of the gases that make up the collapsing stars.

Although the study involved computer simulations and is thus purely theoretical, in practice, the formation and merger of pairs of black holes can give rise to tremendously powerful gravitational radiation—ripples in the fabric of space and time, traveling at the speed of light—that is likely to be visible at the edge of our universe, Reisswig says. Ground-based observatories such as the Laser Interferometer Gravitational-Wave Observatory (LIGO), comanaged by Caltech, are searching for signs of this gravitational radiation, which was first predicted by Albert Einstein in his general theory of relativity; future space-borne gravitational-wave observatories, Reisswig says, will be necessary to detect the types of gravitational waves that would confirm these recent findings.

Ott says that these findings will have important implications for cosmology. "The emitted gravitational-wave signal and its potential detection will inform researchers about the formation process of the first supermassive black holes in the still very young universe, and may settle some—and raise new—important questions on the history of our universe," he says.

These findings were published in Physical Review Letters the week of October 11 in a paper titled "Formation and Coalescence of Cosmological Supermassive-Black-Hole Binaries in Supermassive-Star Collapse." Caltech coauthors authors on the study include Ernazar Abdikamalov, Roland Haas, Philipp Mösta. Another coauthor on the study, Erik Schnetter, is at the Perimeter Institute for Theoretical Physics in Canada. The work was funded by the National Science Foundation, NASA, the Alfred P. Sloan Foundation, and the Sherman Fairchild Foundation.