Biochemist Dorothee Kern and her colleagues reconstructed extinct forms of two cell division proteins, and then watched in the lab as one boosted the activity of the other. What Kern’s team observed represents the earliest known instance of two proteins interacting in such a way, the researchers reported February 21 in the journal Science.

That interaction, called allosteric regulation, occurs between many modern proteins, but “how such a feature evolved is unexplored territory,” Kern says. Over the eons, allosteric regulation has become an indispensable tool for cells. As cells developed more elaborate networks of protein activity, Kern explains, they needed more nuanced controls to tweak them.

“Allosteric regulation is arguably the biggest evolutionary step in the development of higher organisms,” she says.

Kern, a Howard Hughes Medical Institute (HHMI) Investigator at Brandeis University, wants to unravel how proteins and other molecules work together — at atomic resolution and in real time. Pinpointing how proteins behave as they send cellular signals, bind to other molecules, or kickstart chemical reactions could give scientists a new window into drug discovery.

Kern chose to study a protein called Aurora A because it’s vitally important in cell division — abnormalities can drive cancer growth — and because it’s turned on by a protein partner called TPX2.

During cell division, Aurora A helps to distribute chromosomes equally between daughter cells. TPX2 binds to an “allosteric site” of Aurora A — a spot apart from the active site of the protein. Then, TPX2 brings Aurora A to the site of the action and switches the protein into high gear. Kern wanted to know how the two proteins’ partnership evolved.

Organisms across the evolutionary tree make some version of the Aurora A and TPX2 proteins. First, Kern, study coauthor Adelajda Hadzipasic, and their colleagues at Brandeis compared versions of Aurora A and TPX2 from a diverse set of species — including bacteria, plants, and humans. Then, they did a bioinformatics analysis, traveling back in time to deduce the ancient amino acid sequences of the two proteins’ oldest common ancestors. Next, they used those sequences to recreate the ancient proteins.

“Evolution used to be speculative subject, but it is now experimental. That’s what’s really exciting about this technique,” says Christopher Miller, an HHMI Investigator alum and recently retired biochemist who used to work next door to Kern at Brandeis.

Experiments in the lab let Kern’s team map the proteins’ entwined histories. Initially, around 1.5 billion years ago, Aurora A had to self-activate — no TPX2 required. After TPX2 showed up, about a half billion years later, the protein increased Aurora A’s productivity by pulling it to the job site. “That created an evolutionary increase in fitness,” she says.

Because teaming up with TPX2 boosted Aurora A’s efficiency, the partnership endured. That gave the two proteins time to dial in a more sophisticated relationship, Kern says. Over time, TPX2 took on an additional, new role — as Aurora A’s activator. Today, the modern Aurora A/TPX2 pair can work 10 times faster than Aurora A alone.

And when Kern’s team combined ancient Aurora A with modern TPX2, they found that the two proteins bound together surprisingly well — despite almost a billion years of evolution between them.

That wasn’t expected, Kern says. Conventional wisdom held that proteins that work together also evolve together — so a change in one partner is matched by a complementary change in the other. Modern TPX2, the researchers thought, would likely have amassed too many changes to connect with such an ancient partner.

Kern solved this mystery. The sections of Aurora A and TPX2 that touch stayed essentially the same over 1 billion years, her team discovered, even as other parts of the proteins morphed and shifted over time.

Investigating the details of such interactions could provide a springboard for creating new types of drugs, says Neel Shah, a chemist at Columbia University. Though there’s a long road between identifying proteins’ allosteric sites and actually designing drugs that can take advantage of them, he says, “these findings could be a starting point.”

Fishermen from California to Washington caught almost all the available legal-size male Dungeness crab each year in the last few decades. However, the crab population has either remained stable or continued to increase, according to the first thorough population estimate of the West Coast Dungeness stocks.

“The catches and abundance in Central California especially are increasing, which is pretty remarkable to see year after year,” said Kate Richerson, a research scientist at NOAA Fisheries’ Northwest Fisheries Science Center in Seattle. Richerson is the lead author of the new study published in the journal Fisheries Research. “There’s reason to be optimistic that this fishery will continue to be one of the most productive and on the West Coast.”

Other recent research has suggested that the West Coast’s signature shellfish could suffer in the future from ocean acidification and other effects related to climate change. That remains a concern, Richerson said, but the study did not detect obvious signs of population-level impacts yet.

Fishing Regulation Success

The secret to the success of the Dungeness crab fishery may be the way fishing regulations protect the crab populations’ reproductive potential. Male Dungeness crabs mature and begin reproducing one to two years before they can be caught, so crabs can reproduce even with heavy fishing pressure. Female Dungeness crab can store sperm for more than a year, allowing them to reproduce even in the absence of numerous males. Fishermen must also return females to the water, further protecting the reproductive capacity of the population.

“The management system that is used for Dungeness crab seems to be a perfect fit for their life history because it allows the population to reproduce and grow even with the intensive harvest,” Richerson said.

Natural Variability

Crab numbers and reproduction rates do vary from year to year, mostly because of ocean conditions. That also may have contributed to the increasing numbers in Central California. They have risen over the last two decades and now average nearly five times abundance estimates from 1970 to 2000.

Central California crab numbers have increased enough that they are now closer to the size of populations in Northern California, coastal Washington, and Oregon. Those populations do not show the same growth trends as those in Central California, but remain stable overall.

However, a previous increase in the Central California landings from the 1930s to the late 1950s was followed by a dramatic crash about 1960. Catches remained low until the 1980s and then rebounded. Researchers believe those fluctuations likely reflected changing ocean conditions, and could happen again.

“If this is true, the recent increase in Central California crab abundance may be reversed when the system again shifts to a period of later spring transitions,” the scientists wrote. “This is likely to have a large impact on the fishery, as well as other interlinked fisheries in the area.”

Story Source:

Materials provided by NOAA Fisheries West Coast Region. Note: Content may be edited for style and length.

Making progress on that vision, MIT researchers have designed a system that lets these types of robots learn complicated tasks that would otherwise stymie them with too many confusing rules. One such task is setting a dinner table under certain conditions.

At its core, the researchers’ “Planning with Uncertain Specifications” (PUnS) system gives robots the humanlike planning ability to simultaneously weigh many ambiguous — and potentially contradictory — requirements to reach an end goal. In doing so, the system always chooses the most likely action to take, based on a “belief” about some probable specifications for the task it is supposed to perform.

In their work, the researchers compiled a dataset with information about how eight objects — a mug, glass, spoon, fork, knife, dinner plate, small plate, and bowl — could be placed on a table in various configurations. A robotic arm first observed randomly selected human demonstrations of setting the table with the objects. Then, the researchers tasked the arm with automatically setting a table in a specific configuration, in real-world experiments and in simulation, based on what it had seen.

To succeed, the robot had to weigh many possible placement orderings, even when items were purposely removed, stacked, or hidden. Normally, all that would confuse robots too much. But the researchers’ robot made no mistakes over several real-world experiments, and only a handful of mistakes over tens of thousands of simulated test runs.

“The vision is to put programming in the hands of domain experts, who can program robots through intuitive ways, rather than describing orders to an engineer to add to their code,” says first author Ankit Shah, a graduate student in the Department of Aeronautics and Astronautics (AeroAstro) and the Interactive Robotics Group, who emphasizes that their work is just one step in fulfilling that vision. “That way, robots won’t have to perform preprogrammed tasks anymore. Factory workers can teach a robot to do multiple complex assembly tasks. Domestic robots can learn how to stack cabinets, load the dishwasher, or set the table from people at home.”

Joining Shah on the paper are AeroAstro and Interactive Robotics Group graduate student Shen Li and Interactive Robotics Group leader Julie Shah, an associate professor in AeroAstro and the Computer Science and Artificial Intelligence Laboratory.

Bots hedging bets

Robots are fine planners in tasks with clear “specifications,” which help describe the task the robot needs to fulfill, considering its actions, environment, and end goal. Learning to set a table by observing demonstrations, is full of uncertain specifications. Items must be placed in certain spots, depending on the menu and where guests are seated, and in certain orders, depending on an item’s immediate availability or social conventions. Present approaches to planning are not capable of dealing with such uncertain specifications.

A popular approach to planning is “reinforcement learning,” a trial-and-error machine-learning technique that rewards and penalizes them for actions as they work to complete a task. But for tasks with uncertain specifications, it’s difficult to define clear rewards and penalties. In short, robots never fully learn right from wrong.

The researchers’ system, called PUnS (for Planning with Uncertain Specifications), enables a robot to hold a “belief” over a range of possible specifications. The belief itself can then be used to dish out rewards and penalties. “The robot is essentially hedging its bets in terms of what’s intended in a task, and takes actions that satisfy its belief, instead of us giving it a clear specification,” Ankit Shah says.

The system is built on “linear temporal logic” (LTL), an expressive language that enables robotic reasoning about current and future outcomes. The researchers defined templates in LTL that model various time-based conditions, such as what must happen now, must eventually happen, and must happen until something else occurs. The robot’s observations of 30 human demonstrations for setting the table yielded a probability distribution over 25 different LTL formulas. Each formula encoded a slightly different preference — or specification — for setting the table. That probability distribution becomes its belief.

“Each formula encodes something different, but when the robot considers various combinations of all the templates, and tries to satisfy everything together, it ends up doing the right thing eventually,” Ankit Shah says.

Following criteria

The researchers also developed several criteria that guide the robot toward satisfying the entire belief over those candidate formulas. One, for instance, satisfies the most likely formula, which discards everything else apart from the template with the highest probability. Others satisfy the largest number of unique formulas, without considering their overall probability, or they satisfy several formulas that represent highest total probability. Another simply minimizes error, so the system ignores formulas with high probability of failure.

Designers can choose any one of the four criteria to preset before training and testing. Each has its own tradeoff between flexibility and risk aversion. The choice of criteria depends entirely on the task. In safety critical situations, for instance, a designer may choose to limit possibility of failure. But where consequences of failure are not as severe, designers can choose to give robots greater flexibility to try different approaches.

With the criteria in place, the researchers developed an algorithm to convert the robot’s belief — the probability distribution pointing to the desired formula — into an equivalent reinforcement learning problem. This model will ping the robot with a reward or penalty for an action it takes, based on the specification it’s decided to follow.

In simulations asking the robot to set the table in different configurations, it only made six mistakes out of 20,000 tries. In real-world demonstrations, it showed behavior similar to how a human would perform the task. If an item wasn’t initially visible, for instance, the robot would finish setting the rest of the table without the item. Then, when the fork was revealed, it would set the fork in the proper place. “That’s where flexibility is very important,” Shah says. “Otherwise it would get stuck when it expects to place a fork and not finish the rest of table setup.”

Next, the researchers hope to modify the system to help robots change their behavior based on verbal instructions, corrections, or a user’s assessment of the robot’s performance. “Say a person demonstrates to a robot how to set a table at only one spot. The person may say, ‘do the same thing for all other spots,’ or, ‘place the knife before the fork here instead,'” Shah says. “We want to develop methods for the system to naturally adapt to handle those verbal commands, without needing additional demonstrations.”

When 119 young adults consumed a meal while playing a simple computer game for 15 minutes, they ate significantly less than when they ate the same meal without distractions, said lead author Carli A. Liguori.

Liguori conducted the research while earning a master’s degree in food science and human nutrition at the University of Illinois at Urbana-Champaign. The findings were published recently in the Journal of Nutrition.

Participants’ food consumption was evaluated on two separate occasions — one day when they played the game while eating and on another day when they ate without distractions.

The game, called Rapid Visual Information Processing, tests users’ visual sustained attention and working memory and has been used extensively by researchers in evaluating people for problems such as Alzheimer’s disease and attention-deficit disorder.

The game randomly flashes series of digits on the computer screen at the rate of one per second. Participants in the study were instructed to hit the space bar on the keyboard whenever they saw three consecutive odd numbers appear.

“It’s fairly simple but distracting enough that you have to really be watching it to make sure that you don’t miss a number and are mentally keeping track,” Liguori said. “That was a big question for us going into this — how do you ensure that the participant is distracted? And the RVIP was a good solution for that.”

The participants, who had fasted for 10 hours before each visit, were told to consume as much as they wanted of 10 miniature quiches while they were either playing the game or eating quietly without distractions for 15 minutes.

The food was weighed and counted before and after it was given to each person.

After a 30-minute rest period, participants completed an exit survey that asked them to recall how many quiches they had been given and the number they had consumed. They also rated how much they enjoyed the meal as well as their feelings of hunger and fullness.

Liguori hypothesized that, in keeping with prior research, when people ate while using the computer game they would not only consume more food but would have poorer memory of what they ate and enjoy it less.

Instead, she found that participants ate less when they were distracted by the computer game. Moreover, participants’ meal memory — their ability to recall how much they had been served and eaten — was less accurate when they were distracted than when they ate quietly without the game.

However, participants’ consumption on their second visit was affected by which activity they had performed during their initial visit. The people who engaged in distracted eating on their first visit ate significantly less than their counterparts who did not experience the distracted eating condition until their second visit.

Moreover, when participants who engaged in the distracted eating on their first visit were served the quiches on their next visit, “they behaved as if they were encountering the food for the first time, as evidenced by a lower rate of consumption similar to that of those who began” with the non-distracted meal, according to the study.

“It really seemed to matter whether they were in that distracted eating group first,” said Liguori, who is a visiting faculty member in health and physical activity at the University of Pittsburgh. “Something about being distracted on their initial visit really seemed to change the amount they consumed during the nondistracted meal. There may be a potent carryover effect between the mechanism of distraction and the novelty of the food served.”

The results suggest that there may be a difference between distracted eating and mindless eating. Although the terms are often used interchangeably, Liguori hypothesized that they may be distinctly different behaviors with nuances that need to be explored.

Mindless eating may occur when we eat without intending to do so, Liguori hypothesized. For example, we grab a handful of candy from the jar at the office as we walk by or start snacking on chips because they happen to be in sitting front of us.

Conversely, distracted eating may occur when we engage in a secondary activity such as watching TV or answering emails while we are deliberately eating — for example, when we’re eating dinner, she said.

Although prior research indicated that people eat more when distracted, Liguori hypothesized that the differing results in her study may have been associated with examining within-person differences — comparing individuals’ consumption under the , rather than comparing individuals’ behavior to that of peers.

Or, she said, her findings could have been influenced by factors such as the type of distraction that was used, the type of food served or by using college students as the study population, limiting the diversity in participants’ age, race, food preferences and motivation to regulate their consumption.

The study was co-written by U. of I. food science and human nutrition professor Sharon M. Nickols-Richardson, who is also director of Illinois Extension and Outreach; and then-graduate student Cassandra J. Nikolaus.