Tyler O'Neal, Staff Editor GOVERNMENT

Gladstone researchers use machine-learning to discover new ways of controlling the spatial organization of induced pluripotent stem cells

Model organs grown from patients' own cells may one day revolutionize how diseases are treated. A person's cells, coaxed into heart, lung, liver, or kidney in the lab, could be used to better understand their disease or test whether drugs are likely to help them. But this future relies on scientists' ability to form complex tissues from stem cells, a challenging undertaking.

In their natural environment, stem cells form predictable patterns as they mature; over time, these patterns morph into the tissues of an adult organism. In the lab though, researchers have struggled to control the spatial organization of stem cells--an important step toward being able to create functional organs for research or therapeutic purposes. Some have turned to 3-D printing to lay out populations of stem cells in a desired shape. But the approach isn't always successful, with cells often migrating away from their printed locations. CAPTION Machine learning predicts conditions that will cause stem cell colonies to form desired patterns. (Left) Video showing simulated interactions between different stem cell populations. (Right) Stem cells grown in conditions dictated by the machine-learning program generate a colony that forms a bull's-eye pattern, as predicted. CREDIT Photo credit: Ashley Libby, David Joy, and Iman Haghighi, Gladstone Institutes{module INSIDE STORY}

Now, scientists at Gladstone Institutes, in collaboration with researchers at Boston University, have used a computational model to learn how to coax stem cells into forming new arrangements, including those that might eventually be useful in generating personalized organs.

"We've shown how we can leverage the intrinsic ability of stem cells to organize," said Gladstone Senior Investigator Todd McDevitt, PhD, a lead author of the study, which was published in the journal Cell Systems. "This gives us a new way of engineering tissues, rather than a printing approach where you try to physically force cells into a specific configuration."

"This works exemplifies the power of applying a computational approach to stem cell biology to make sense of the complexity in these cells," said Calin Belta, director of the Boston University Robotics Lab and co-corresponding author on the new paper.

Induced pluripotent stem (iPS) cells, similar to the stem cells found in an embryo, have the potential to become nearly every type of cell in the body. Researchers have found ways to direct iPS cells to become many of these cell types, including heart and brain. Some are already using these cells to model diseases in the lab or even transplant into patients. But clumps of cells in a Petri dish aren't the same thing as functioning three-dimensional organs.

"Despite the importance of organization for functioning tissues, we as scientists have had difficulty creating tissues in a dish with stem cells," said Ashley Libby, a graduate student in the UC San Francisco Developmental & Stem Cell Biology Program and co-first author of the new paper, who worked on the project with David Joy, a graduate student in the joint Graduate Program in Bioengineering from UC Berkeley and UC San Francisco (BioE). "Instead of an organized tissue, we often get a disorganized mix of different cell types."

McDevitt and his colleagues previously showed that blocking, or "knocking down," the expression of two different genes, ROCK1 and CDH1, affected the layout of iPS cells grown in a Petri dish. The scientists wondered whether they could predict the exact arrangement of cells that would result from altering ROCK1 and CDH1 by different degrees at different timepoints. But there were so many possible variables--the timing and degree of each gene knockdown, the duration of the experiment, the proportion of cells affected--that testing every possible combination would be too time-consuming. So McDevitt's group teamed up with the Belta Lab who could help them create a model to do the job.

The researchers used a CRISPR/Cas9 gene-editing system that could be induced to block expression of ROCK1 or CDH1 at any time during an experiment by adding a drug to the iPS cells. In addition, they engineered the system so that cells fluoresced in different colors when they lost expression of ROCK1 or CDH1, making it easier to study changes to the arrangement of the cells.

McDevitt's group carried out a handful of experiments using different doses and timing of the CRISPR/Cas9 system. Then, the computational researchers started inputting the results into a machine-learning program, designed to identify patterns within a dataset.

"Machine learning can predict what movie you might like based on your viewing history, but it can also generate new insights into biological systems by mimicking them." said Demarcus Briers, co-first author of the new paper who performed the work during his graduate studies at Boston University. "Our machine-learning model allows us to predict new ways that stem cells can organize themselves, and produces instructions for how to recreate these predictions in the lab."

The machine-learning program used results from the initial stem cell experiments to infer ways that ROCK1 and CDH1 affect iPS cell organization. With the model up and running, the researchers then began to probe whether it could compute how to make entirely new patterns, like a bull's-eye or an island of cells.

"The power of this model is that it can generate thousands of data points simulating things that it could take months for me to do in a lab," said Libby.

The simulations narrowed down a set of starting conditions that might lead to the desired arrangement of cells--informing researchers exactly when, where, and how to add drugs to their iPS cells to shut off ROCK1 and CHD1. Then, McDevitt and Libby could test those suggested conditions. The machine-learning system, it turned out, was correct--at least when it came to the bull's-eye and island patterns they were after. In the lab, for the first time, the researchers were able to reliably generate concentric circles of stem cell populations looped around each other.

"I was just blown away when I first saw the results," said Bruce Conklin, MD, a Gladstone senior investigator who also worked on the new study. "Modeling cell behavior is the Holy Grail of biology and this paper takes an important step forward in doing that."

The team would like to expand the model in the future--adding in the effects of other developmental genes to get an even wider variety of possible cell configurations. They also plan to work toward designing three-dimensional shapes in addition to the two-dimensional layouts they've already studied.

"We're now on the path to truly engineering multicellular organization, which is the precursor to engineering organs," said McDevitt, who is also the director of the BioE graduate program. "When we can create human organs in the lab, we can use them to study aspects of biology and disease that we wouldn't otherwise be able to."

Russian supercomputer model describes the dynamic instability of microtubules

Tyler O'Neal, Staff Editor GOVERNMENT

Researchers of Sechenov University together with their colleagues from several Russian institutes studied the dynamics of microtubules that form the basis of the cytoskeleton and take part in the transfer of particles within a cell and its division. The supercomputer model they developed describes the mechanical properties of protofilaments (longitudinal fibers that compose microtubules) and suggests how they assemble and disassemble. All the details of the study can be found in PLOS Computational Biology.

Microtubules are long hollow cylinders with walls consisting of tubulin molecules arranged helix-wise. Each cycle contains 13 pairs of α- and β-tubulin, so a microtubule is composed of 13 longitudinal fibers, protofilaments. Microtubules grow by the addition of tubulin from the cytoplasm. The protein binds more actively to one end (plus-end) of the microtubule and dissociates from the other one (minus-end) quicker. Both processes take place simultaneously, but their rate changes: when the concentration of tubulin is sufficient, microtubules grow faster than degrade, and when the concentration is low - vice versa. Between the phases of growth and shrinking, there is a period of stability, but it is very short. 

Though the mechanisms of microtubule's growth and shortening are well-studied, there are still quite a few questions about how the structure and properties of tubulin change. It is known that both molecules of tubulin (α- and β-tubulin) are connected to a molecule of guanosine triphosphate (GTP). GTP of β-tubulin can hydrolyze and turn into guanosine diphosphate (GDP) that causes the double tubulin molecule (dimer) to dissociate from a microtubule. The authors of the paper tried to understand how the properties of tubulin dimers and protofilaments depend on GTP hydrolysis and what provides the difference between plus- and minus-ends of microtubules. Above all, microtubules take part in cell division, and studies of these mechanisms will contribute to the search for yet unknown ways to suppress the replication of cancer cells. In particular, microtubules serve as molecular targets for an important anti-tumor drug, paclitaxel, that inhibits microtubule disassembly. {module INSIDE STORY}

Existing studies offer several models of possible changes in protein structure taking place upon GTP hydrolysis: a slight curving of tubulin dimers or weakening of longitudinal bonds between dimers without significant changes in their shape. Some researchers also suggest that hydrolysis may affect interactions between neighboring protofilaments. According to the authors, it was impossible to prove or refute any of these claims for a long time because of the lack of precise experimental data. In this research, they verified the first hypothesis and computed the 'behavior' of molecules using the latest available experimental structures obtained by cryo-electron tomography. They examined bonds between dimers as well as between α- and β-tubulin within them.

Scientists modeled the bending of the tubulin dimer and the whole protofilament, with GTP and GDP bound to them, throughout one millisecond, watching the angle and direction of the curvature and assessing the strength of bonds within and between dimers. The results showed that protofilaments with GTP and GDP-bound tubulin were bent almost to the same extent, so the first hypothesis was disproved. But it turned out that GTP influences the flexibility of the bonds between dimers: protofilaments made of tubulin connected with GTP were much more bendable compared with those containing GDP.

Using the revealed difference in bond rigidity between GTP and GDP-connected protofilaments, the authors concluded that more flexible bonds ease the straightening of protofilaments and thus facilitate the assembly of the microtubule.

'Based on simulations, we developed a simple model of dynamic instability of microtubules, i.e. their assembly and disassembly. A deeper understanding of this process on the molecular level would enable a targeted development of medicines able to affect the stability of microtubules and thus prevent the reproduction of tumour cells', said Philipp Orekhov, co-author of the paper and senior scientist at the Institute for Personalized Medicine, Sechenov University.

Curtin astrophysicist uses the Pawsey Supercomputing Centre, outback telescope to capture Milky Way center, discovers remnants of dead stars

Tyler O'Neal, Staff Editor GOVERNMENT

A radio telescope in the Western Australian outback has captured a spectacular new view of the center of the galaxy in which we live, the Milky Way.

The image from the Murchison Widefield Array (MWA) telescope shows what our galaxy would look like if human eyes could see radio waves.

Astrophysicist Dr. Natasha Hurley-Walker, from the Curtin University node of the International Centre for Radio Astronomy Research (ICRAR), created the images using the Pawsey Supercomputing Centre in Perth.

"This new view captures low-frequency radio emission from our galaxy, looking both in fine detail and at larger structures," she said.

"Our images are looking directly at the middle of the Milky Way, towards a region astronomers call the Galactic Centre."

The data for the research comes from the Galactic and Extragalactic All-sky MWA survey, or 'GLEAM' for short. This image shows a new view of the Milky Way from the Murchison Widefield Array, with the lowest frequencies in red, middle frequencies in green, and the highest frequencies in blue. Huge golden filaments indicate enormous magnetic fields, supernova remnants are visible as little spherical bubbles, and regions of massive star formation show up in blue. [The supermassive black hole at the centre of our galaxy is hidden in the bright white region in the centre.]{module INSIDE STORY}

The survey has a resolution of two arcminutes (about the same as the human eye) and maps the sky using radio waves at frequencies between 72 and 231 MHz (FM radio is near 100 MHz).

"It's the power of this wide frequency range that makes it possible for us to disentangle different overlapping objects as we look toward the complexity of the Galactic Centre," Dr. Hurley-Walker said.

"Essentially, different objects have different 'radio colors', so we can use them to work out what kind of physics is at play."

Using the images, Dr. Hurley-Walker and her colleagues discovered the remnants of 27 massive stars that exploded in supernovae at the end of their lives.

These stars would have been eight or more times more massive than our Sun before their dramatic destruction thousands of years ago.

Younger and closer supernova remnants, or those in very dense environments, are easy to spot, and 295 are already known.

Unlike other instruments, the MWA can find those which are older, further away, or in very empty environments.

Dr. Hurley-Walker said one of the newly-discovered supernova remnants lies in such an empty region of space, far out of the plane of our galaxy, and so despite being quite young, is also very faint.

"It's the remains of a star that died less than 9,000 years ago, meaning the explosion could have been visible to Indigenous people across Australia at that time," she said.

An expert in cultural astronomy, Associate Professor Duane Hamacher from the University of Melbourne, said some Aboriginal traditions do describe bright new stars appearing in the sky, but we don't know of any definitive traditions that describe this particular event.

"However, now that we know when and where this supernova appeared in the sky, we can collaborate with Indigenous elders to see if any of their traditions describe this cosmic event. If any exist, it would be extremely exciting," he said.

Dr. Hurley-Walker said two of the supernova remnants discovered are quite unusual "orphans", found in a region of the sky where there are no massive stars, which means future searches across other such regions might be more successful than astronomers expected.

Other supernova remnants discovered in the research are very old, she said.

"This is really exciting for us because it's hard to find supernova remnants in this phase of life--they allow us to look further back in time in the Milky Way."

The MWA telescope is a precursor to the world's largest radio telescope, the Square Kilometre Array, which is due to be built in Australia and South Africa from 2021.

"The MWA is perfect for finding these objects, but it is limited in its sensitivity and resolution," Dr Hurley-Walker said.

"The low-frequency part of the SKA, which will be built at the same site as the MWA, will be thousands of times more sensitive and have much better resolution, so should find the thousands of supernova remnants that formed in the last 100,000 years, even on the other side of the Milky Way."