A Heart Attack After Lunch (Thankfully in a Dish)

The session kicked off with a heart attack (I hope the food did not have too much cholesterol!). However in this case,  Adam Engler at UCSD, has developed assays to cause heart attack in a dish in his laboratory. They designed a material, which can increase in stiffness in a step wise fashion. This helps them to assay how dynamically changing matrix affects heart development and may lead to a heart attack in older hearts. They then looked at the 9p21, a large non-coding RNA. Single nucleotide polymorphisms in this RNA increases the risk of familial heart attack. They show that indeed this ncRNA can regulate connexin expression levels. Alteration in these levels leads to an increased risk of heart attack. This is such a fantastic model to study heart attacks in vitro. Heart being a mechanical tissue, these materials now provide amazing ways to look at those aspects in a dynamically changing mechanical landscape.

Next Krystyn Van Vliet from MIT attempted to address an important issue — why  mesenchymal stem cells (MSCs) are not used for any FDA approved therapy.  She postulates that this is because of heterogeneity in the stem cell population. To make the cells segregate reliably into distinct populations based on cell diameter, cell stiffness and nuclear fluctuations, they build a microfluidics based device. They show that based on these three features, they can separate the cells that are still stem-like and have not exited the cell cycle. Moreover, this device is now in clinical trials in Singapore and provides a promising way to use MSCs for reliable therapeutic purposes with the major ill effects of the same.

This was followed by a talk by Yunn Hwen Gan from NUS. Her lab studies Burkholderia pseudomallei, a pathogen that causes melioidosis. This bacterium infects mammalian cells and, using a typeIII/IV secretion system, it leads to cell fusion and form large multinucleate cells. This induces interferon 6 and cytosolic DNA, which helps mount an immune response. The interesting part of their discovery was that, in this bacterium, unlike other bacteria, it uses the secretion system once it has infected the host and is inside the host cell. This provides interesting ways to think about therapy.

The next talk of the session was by Samuel Safran from Weizman Institute. He educated me about some very important theories on how holes close. This becomes particularly relevant when cells are challenged to squeeze through narrow pores and this may create holes in the nuclear membrane. The hole closing would depend on line tension (leading to shrinkage) and lateral stress (growth). Outflow of liquid from the hole in a 3D system would also contribute to this. The flow of fluid – chromatin in the case of nucleus, that is very viscous- would change the dynamics of hole closure. This is an important problem, as inability to close the holes, would be detrimental to the cell.

The Final Morning at the Mechanobiology Meeting

After the banquet, there was one last morning of scientific sessions.  It started with thanking those that made the meeting happen.  One of the organizers, Linda Kenney noted that over 30% of the speakers at the meeting were women, and encouraged others to strive to do the same when they plan meetings.   This was followed by an action packed morning session.

Xavier Trepat from IBEC, showed some very interesting studies to understand mechanisms of how cancer cells migrate out of tumors. Using tumor cells of epithelial origin (A341 cells) and from clinical samples of lung cancers that these cells could indeed bind using cell-cell adhesions mediated by E cadherin and N cadherin. To demonstrate that these two cadherins can form heterotypic adhesions they presented different cadherin (E,N,P) on magnetic beads to the cells. A magnetic field was applied for a minute and assayed how the percentage of beads bound to the cell. Here they saw that N-cad- Ecad adhesions were formed. These adhesions between cancer cells and CAFs(cancer activated fibroblasts) actually assist migration where the CAFs act as the leader cell and pull the cancer cells along their path of migration. Using traction force microscopy, they also observed that the leader cell actually pulls onto the CAFs. In some interesting preliminary experiments, the Trepat group also shows that when cancer cells associate with CAF, they lead to the polarization and subsequent migration of CAF by a yet unknown mechanism.

This was followed by a somewhat related talk by Kevin Chalut from the University of Cambridge. In his lab, they have developed a different method to make hydrogels of different rigidities in a more reliable manner. Changing only the stiffness of hydrogels they show that OPC proliferation relies on soft matrices. There was a nice study on comparing pre- and post- implantation tissue with stem cells grown on soft matrices. On comparison of gene expression profile, they found that pre-implantation embryos were comparable to soft gels whereas post-implantation embryos were comparable to stem cells grown on stiffer gels. Further they went on to show that in the brain, regeneration occurs better when the matrix is softer. Using decellularized tissues, they showed that ageing matrices are more stiff and support proliferation and regeneration to a far lesser extent compared to matrices from neonatal tissues, which are softer. Wow, so this ties in with the talk suggesting that massaging prevents ageing. So get a massage today to relax all that stiffness, and perhaps maintain a softer matrix!

Final talk of the session was delivered by Shivashankar. He suggested that matrix constraints could determine the chromatin packing in the cells. This could be critical to bring together co-regulated genes on different chromosomes into gene clusters. Within these boundary conditions there could be an erasure of the genomic memory of the cells and help transform them on the genetic landscape to cells of another type. This was demonstrated by the growth of fibroblasts on fibronectin presenting rectangular substrates and for a few days lead to the crowding of cells and they began to express E-Cadherin. Finally, they propose to use nuclear features as a biomarker for several diseases in the tissue.

After this session, poster prizes were awarded. Really exciting posters, ranging from broad topics of integrins to cadherins and engineering substrates to genomic data analyses in cancer cells and to unconventional (previously annotated with very different functions!) genes expressed in cancer tissue were recognized.

What Makes Neurons Contract to Generate Tension?

BPJ_111_7.c1.inddWhen preparing for the cover image for the October 4 issue of Biophysical Journal, we started with an image that resembled an art painting, probably something between Monet and Pollock (not claiming that it’s up to their standards). It was pretty, at least to us, but we thought that it didn’t tell a story. So we thought to ourselves, “What is the message that we want to tell if we have the chance?”

This is how the current version of our cover image came about. On the left, it has a bunch of curved green and red lines, while on the right there is the same set of lines but straightened. These are actually real experimental images achieved by genetically encoding the neuronal membrane to fluoresce in green—thanks to the technology enabled by the late Roger Tsien, while staining the cytoskeleton, in this case microtubules, in red. The implication is that buckled (curved) neurons would always contract and become straight again in less than 5 minutes with the contraction vs. time profile following an exponential decay. We used genetic and pharmaceutical tools to study this phenomenon and found that the acto-myosin machinery was the active driver, while microtubules contributed in a resistive/dissipative role.

The neurons contracted because the machineries were trying to build up a mechanical tension, which was shown to be critical for vesicle clustering, a process central to signal propagation across neurons. Yet, we do not know how tension leads to clustering. It’s like we know that we can drive from New York to Los Angeles, but we do not know the path. Unfortunately, there isn’t an app for that in our case. Knowing the players (tension generators) involved is the first step to answer this question.

There are quite a few implications, mostly to our understanding of mechanics in neuroscience. A better understanding always leads to new ideas and approaches to solve problems, which, in the context of nervous system health, are usually costly and sometimes a matter of life and death.

– Alireza Tofangchi, Anthony Fan, Taher Saif

Before the Banquet

After a meal of local delights, everyone heads back to more exciting science.

Yusuke Toyama started the session by telling us about how forces in the system when apoptotic cells extrude out from the epithelial monolayer. Drosophila epithelium was used as the model system. This is a complex process where the cells cant leave a hole in the monolayer. To prevent formation of this hole, there is a complex gymnastic of forces and remodeling of cell-cell junctions. He showed that at early time points when perhaps apoptotic cells decide to leave, the tension at the junctions is relaxed, this is followed by an increase in tension to restore it to normal levels at the junctions once the cells have been extruded. What a beautiful intricate mechanism to ensure that no holes are left and the monolayer is not weak at any point in space. These force changes were also observed in monolayers on a dish using traction force microscopy.

Switching gears, Marvin Whitely told us about building houses for bacteria. This is an amazing technology where they can use a few bacteria to colonize a pico liter home made of PDMS, and this is used to study the social lives of bacteria. One of the most important questions is the simultaneous existence of Pseudomonas and Staphylococcus in chronic infections or cystic fibrosis in humans. These are drug resistant and are a simply a real pain! They show that these bugs coexist in separate clusters and don’t intermix. These clusters are 4-13um apart, suggesting that they are not randomly positioned. Guess what causes proper positioning? One bacteria produces food (lactate) for the other but (perhaps to protect its space) it also produces a toxin (H2O2) for the other. They further show that in bacterium 2, the gene for dispersal is under the promoter that responds to peroxidase levels. At high peroxidase levels these bacteria disperse. Bacterium 1 is a great neighbor to provide food at the same time ensuring that the neighbors don’t get too close. If only humans could have such beautiful designs in society, imagine not crowded cities but everyone had food to eat.

Then we switched back to forces experienced by integrins, a topic close to my integrins. Alex Dunn from Stanford University, has developed (in his lab) and used (from other labs) beautiful FRET based sensors to generate force maps in the whole focal adhesions. These sensors are elastic springs that are tagged with a fluorophore at either end. They are attached to glass at one end and present the ligand at the other end. This was done at a single molecule precision using a few fluorescently labeled FRET probes whereas in a sea of unlabeled probes. Using this, they find that most integrins bear low force (if about 3pN) whereas a few integrins bear higher forces (in the range of 7pN). Alex gave a fantastic analogy of feet of a starfish that stick because of very many feet exerting very low forces. Now one can begin creating physical models of how focal adhesions are formed and maintained. This also helps the cells to provide a buffer wherein they can bear forces 10X(safety factor) higher than what they normally exert, and hence provides robustness in the system over a large dynamic range.

With that we had Pere Roca-Cusachs, tell us about integrins exerting forces on soft versus rigid substrates. In collaboration with Ada Cavalcanti, they use gold nano dots to present integrin ligands (RGD) to the cells. This was done on both soft and rigid substrates. Surprisingly, they find that cells can spread more easily on soft substrates with fewer ligands i.e if the ligands are more separated (~100nm) on the soft substrates, cells can still spread whereas on rigid substrates they spread only when the ligands are spaced not more than 60nm apart. This suggests a different mechanism of clutch engagement, wherein with low forces the adhesion can grow larger. They further test this by adding Blebbistatin (to inhibit myosin and lower forces) and observe that the adhesions grow in size.

Final talk of the session was by Priyamvada Chug from Ewa Paluch’s lab @ UCL. During her PhD, she studied the role of cortical thickness in exerting force, as the cells undergo cytokinesis during mitosis. They developed careful tools to measure cortical thickness and performed an RNAi screen to identify modulators of cortical thickness. The major players they isolated regulated the length of actin fibers. This was not regulated by Myosin II contractility only. Factors that shorten (e.g Capping proteins) or increase the length of actin fibers (Diaphanous) increase the cortical thickness. This was indeed a very interesting demonstration of non-monotonic dependence of cortical thickness on an optimal length of actin fibers

That led us into the last poster session. The posters were amazing ranging from engineering approaches to modulate the matrix to genomics approaches to understand regulation of various cancers. Beautiful imaging to answer such important questions was of course my favorite. I am definitely registering for the 2017 BPS annual meeting, where Eric Bretzig will deliver the national lecture.

This was followed by the conference banquet. An amazing venue – Singapore Yacht Club, overlooking the bay as the sunset. It was surprisingly not too humid and the cool breeze made the weather just perfect. I enjoyed discussing  lamin A/C, another mechanoregulated skeletal element I’m interested in, with fellow attendees. This was followed by interesting an interesting discussion about the U.S. Presidential election! I would safely say that this has been the most entertaining presidential race that I have witnessed and of course I look forward to more drama.

Some of us headed out to the city to enjoy more drinks and discuss science in the backdrop of the city center. The city does look colorful at night!

Fibrosis and Disease Diagnosis

Coffee break over, lets get back to some more interesting science.

It started with Shyni Varghese from UCSD telling us about fibrosis – a condition where the matrix stiffens. Using a genome wide screen, they identified several players that might be important for inflammation and fibrosis. They found fibulin5 to be important in progression of fibrosis. It is associated with elastin and is not essential for development and growth (as tested in knockout mice). So, it appears to be a great target for therapeutics.

This was followed by a talk from Cytovale, a three year old spin off company from Dino di Carlo’s lab. The company is based on the idea that nuclear mechanics can be used as a marker to identify disease states in the cells. One major disease they target in this first phase is sepsis, which leads to death in a large number of patients due to late detection. The major detection marker now is production of lactate. They plan to aid early diagnosis and hence better prognosis by using features in the nucleus as a biomarker. They have developed a PDMS device to stretch the tissues cells reliable and reproducibly. This device is called deformability cytometry cartridge. Using machine-learning algorithms including stochastic nearest neighbor, principle component analysis, DISNEY (developed in Columbia university) they identify a dozen features in the nucleus, which can separate healthy cells from unhealthy ones. The whole workflow takes about 10 minutes, hence is a fast method of diagnosis. With a customizable chip, they hope to use this for many other diseases and would like to hear and collaborate with new researchers to find out ways of enhancing their technology and developing it to diagnose other diseases and have other applications in research laboratory.

Next talk was by Natalie from Sydney where she studies the role of Calapin in membrane repair. They rupture the membrane by shooting it with silicon bullets of about 4um using a gene gun. Using this they identified a mutant of dysferlin, a calcium responsive gene that upon calpain mediated cleavage forms a synaptotagmin like mini dysferlin, which aids in vesicle fusion to the membrane in order to repair it.

The final talk of this session was by Victor Ma from Salaita Lab @ Emory. They use DNA based tension gauge to measure the forces of T cell binding to its receptor. Using this assay they are able to show that stronger binding of the TCR-ligand complex leads to a signal for the T cells to signal. This is a way to avoid signaling upon weak binding and increase specificity of T cell signaling.

Action packed session I must say! Lets get some lunch!

Super Resolution and Receptor Clustering!

Post lunch session – on super resolution microscopy and receptor clustering.  I am really looking forward to this.

First talk by Katharina Gaus from UNSW. She studies clustering of T cell receptors. The punch line of the talk is dense T cell receptor (TCR) clusters are signaling active. In an elaborate study where they use two color super resolution imaging they quantify both the TCRs and phosphorylation events. After looking at many parameters, they find that what matters to TCR signaling is the density of TCR clusters. Signaling clusters have a higher molecular density whereas non-signaling clusters had a lower density. To establish causality, they analyzed where would the information flow from – higher density clusters to signaling or signaling to increase in cluster density. Using correlation analyses they observe that information flows from cluster density to signaling. How is the density of clusters regulated, still needs to be established.

This was followed by a talk by Gregory Giannone from CNRS, Bordeaux. Using single particle tracking in PALM they have shown that integrins are present in immobile and confined diffusion states in the focal adhesions. Further, in collaboration with Val Weavers laboratory they showed that glycocalyx, mucin would create potential wells that would promote integrin clustering in locations where integrins bind ligand. In some recent studies they show that, Kindlin is required for the motility of integrins. They generate kindlin PH deletion mutants that can’t bind to PIP3 lipids. They propose a model where kindlin brings the integrins to the adhesions sites, which could be followed by replacement of kindlin by Talin.

More about Integrins and Cadherin Signaling

Day 1 continued with a talk by Bernard Wehrle-Haller. Beautiful work, addressing important questions about integrin biology. Noteworthy was the differences of the different integrin isoforms. The muscle isoform does not recruit Paxillin as efficiently as the isoform expressed in the fibroblasts. This suggests that it perhaps does not signal but is an efficient machine to provide attachment to the musculature. This demonstrates how the body makes small changes in its toolbox to adapt to vastly different needs. Another tool in the tool kit to regulate integrin function is via acetylation. They demonstrate that de-acetylated integrins are actually much more stable in an adhesion, wherein the recovery of these integrins upon FRAP was very slow compared to the integrins that are normally processed. This suggests many ways of regulating integrin function to adapt to the adhesions. I was amazed at how many unknowns are still there in integrin biology even after a massive body of understanding of focal adhesions on glass.

We switched from integrins to cadherins. Talk by Vania Braga from Imperial College London, was very interesting where she talked about their novel insights into how epithelial cell-cell adhesions form. She showed us the presence of two populations of actin – the junctional actin present at the junction and marked by overlap with the Ecadherin staining and the thin bundles present close to the junction. Unlike junctional actin, thin bundles exist even when cells do not form cell-cell junctions. They developed an automated platform to perform a genome wide RNAi screen to identify factors that regulate one or the other actin pools. One interesting factor they identified was EEF1A, which has a traditionally defined role in mRNA translation. This led to a 50% increase in the level of Ecadherin at the junctions without an increase in the overall levels of Ecadherin, demonstrating a probable existence of a moonlighting function of EEF1A. Wow, the traditional nomenclature does make us think that it would have only one function, in the omics era, a lot of this might change. When I chatted with Professor Braga later, she revealed that there is some old literature, which shows that EEF1A binds to actin, which needs to be revisited to understand junction formation. In the next part of her talk she showed that cortical tension also modulated junctions. They modulated cortical tension by presenting cells with substrates varying in geometry. What they found that lower tension leads to an increased number of cell-cell contacts whereas an increase in the same leads to more straight (a possible ready of tension bearing junctions) junctions.

Wow, lots of information of new tools and ways to think about cell-cell and cell-matrix adhesions.