Learning from fish who can regenerate their eyes

Written by: Kaitlyn Sadtler

Original Article: Mitra et al. Journal of Cell Biology 2019
The Gist of It:
Some animals, such as fish and salamanders, have amazing regenerative capabilities. However, when it comes to mammals like us, these traits are largely lost. You might notice this by something as simple as the scar you got from falling off your bike as a kid – the scar tissue is dense, unorganized collagen whose goal was just to shut off the injury from the outside world, but there are no hair follicles, no sweat glands, sometimes even no pigment in those scars. Your skin didn’t regenerate; it healed, but it scarred. Researchers looking at different organisms that are able to regenerate are trying to understand how these animals can re-grow fully functional body parts after injury. Recent research from the Indian Institute of Science Education and Research focused on the ability of zebrafish to regenerate their retina, a part of the eye. Zebrafish have a special find of cell called Muller glia-derived progenitor cells (MGPCs) that share characteristics with stem cells, the individual cells that have the potential to develop into lots of different kinds of cells. These MGPCs are required for zebrafish to be able to regenerate their retina. Mitra and colleagues were interested in how exactly these cells were being regulated to promote regeneration. They focused in on a certain gene – Myc. Myc can directly affect the activation of other genes. They found that Myc was at the center of what is called a “gene regulatory network”. This is a collection of genes in which each gene affects the activity of others in that network. In the MGPCs, Myc promotes the activity of a gene called lin28a (which is important in regeneration), and in neighboring cells that sit next to these MGPCs, Myc can actually inhibit this lin28a gene. These activities together help balance and regulate regeneration in the zebrafish retina. These studies to understand how zebrafish regulate their regenerative capacities could lead to new discoveries and therapeutics to promote regeneration in higher mammals (maybe even us!).
Screen Shot 2019-10-29 at 11.19.44 AM.png

Myc turns lin28a on in Muller glia-derived progenitor cells (MGPCs) but turns lin28a off in neighboring cells to balance regeneration in the zebrafish retina

The Nitty Gritty:
Mitra et al. used a zebrafish model of retinal damage and regeneration. Using tubala:gfp transgenic fish that express GFP in Muller glia-derived progenitor cells (MGPCs, previously described to be necessary for retinal regeneration), they evaluated the participation of Myca and Mycb isoforms in regeneration. myca and mycb are expressed throughout the retina at 12 hours post-injury, and by 2 days post-injury their expression is restricted to the injury site. If these genes were knocked down with morpholino-modified antisense oligonucleotides (MOs), there was a defect in regeneration. Furthermore, knockdown of both myca and mycb resulted in a combinatorial defect, suggesting the potential for multiple pathways being affected by myca and mycb signaling. Max, a partner of Myc, was also shown to function with Myc and stimulate proliferation through inhibition of Max-Myc interactions with the 10058-F4 inhibitor. In addition to max, they analyzed a previously reported gene, ascl1a, and its interaction with Myc. Inhibition of mycb either with MO or 10058-F4 downregulated down ascl1a and mycb, while upregulating myca and max. ChIP (chromatin immunoprecipitation) revealed Myc binding to the ascl1a promoter. To further analyze this transcriptional network and the seemingly confounding finding that ascl1a knockdown increases mycb early after modulation, they evaluated the contribution of insm1a (previously reported to be induced by the activity of Ascl1a). Knockdown of insm1a with MOs led to increases in myca and mycb expression. Based on these interactions of important gene regulatory proteins, they evaluated expression and activity of lin28a, previously described as a critical mediator of retinal regeneration. Knockdown of Myc significantly increased lin28a selectively in retinal non-MGPCs. Myc bound to the promoter of lin28a and recruited Hdac1 (histone deacetylase 1). Furthermore, delta-notch signaling was shown in repress lin28a expression. These data suggest that Myc acts in a dual role as an activator of lin28a in MGPCs and a repressor in neighboring cells (in concert with Hdac1 activity).
Original Research Article: Mitra, S., et al. “Dual regulation of lin28a by Myc is necessary during zebrafish retina regeneration.” J Cell Biol 218.2 (2019): 489–507.

Can we have personalized therapies for familial Parkinson’s disease?

Written By: Sravya Kotaru

Original article: Yi et al. Human Molecular Genetics 2019
The Gist of It:
Familial or early-onset Parkinson’s disease occurs when individuals inherit harmful changes or mutations in certain genes. A common gene mutated in familial Parkinson’s disease contains the information to produce a protein called Parkin. Parkin gets rid of damaged mitochondria (the cellular compartments that generate energy for a cell to live). If Parkin fails to function properly, a buildup of damaged mitochondria can eventually lead to cell death. When Parkin failure leads to loss of nerve cells controlling skeletal muscles, one loses proper control over body movements and develops Parkinson’s disease, with common symptoms including tremors and stiffness.
Imagine Parkin (or, in fact, any protein) as a tiny machine with different parts that work together to do a job. Mutations can occur in different parts (or domains) of Parkin, each affecting a different coordinated step in Parkin’s function. Some mutations have a minor effect and can be tolerated without complete loss of function – these are called benign (i.e. harmless). Others can be debilitating, causing total failure of Parkin function – these are called pathogenic (i.e. harmful and disease-causing). The best way to ‘fix’ mutant Parkin and treat disease is to identify which pathogenic mutation a patient has and tailor treatment to counteract that specific mutation.
The first step to achieving this kind of customized treatment for Parkin mutations is knowing which are pathogenic and what step of function they interfere with. In this study, researchers classified 215 different Parkin mutations into five categories – pathogenic, likely pathogenic, likely benign, benign, and uncertain significance. This was based on whether each individual mutation causes disease in patients or is tolerated in healthy individuals. They used computer modeling to study the effect of each mutation on Parkin’s structure to determine which step of Parkin’s function was altered, if any. For example, some mutations kept parts (domains) of Parkin from doing their job, some kept Parkin from interacting with partner proteins, whereas some others didn’t have any obvious effect. But the most interesting mutations were those that, when present in combination with a pathogenic mutation, could actually counteract the pathogenic effects of the first mutation and recover normal function. This work opens up the possibility of using drugs that mimic such rescue mutations as a treatment for patients that have the corresponding pathogenic mutation, a.k.a. personalized therapy.

Classifying observed Parkin mutations as pathogenic or benign identifies whether patients with Parkinson’s disease have a pathogenic mutation that requires customized treatment. Some mutations that function better than normal can rescue pathogenic mutations and clue us in to design drugs to mimic their effect.

The Nitty Gritty:
Mutations in the E3 ubiquitin ligase Parkin are a leading cause of early-onset or familial Parkinson’s disease, as they interfere with Parkin’s ability to orchestrate the clearance of damaged mitochondria through mitophagy. Most common Parkin mutations are missense and affect function without altering expression, serving as ideal candidates for genotype (mutation)-specific treatment. A prerequisite for such a customized treatment is differentiating pathogenic mutations from non-pathogenic ones. In this study, the authors classified 215 observed Parkin missense mutations into five clinical categories using the Sherloc framework – pathogenic, likely pathogenic, likely benign, benign, unknown significance. This framework uses data from various Parkinson’s disease databases and other exome databases, accounting for segregation within affected families as well as allele frequencies and homozygosity in healthy and diseased individuals. In addition, they also categorized the mutations based on protein stability and mitophagy activity of mutants compared with wild-type Parkin in cell-based assays. They found that clinically pathogenic or likely pathogenic mutations had severely reduced mitophagy and either lower or normal protein levels. Alternatively, clinically benign and likely benign mutations had moderate to no effect on mitophagy with normal protein levels. Moreover, modeling these mutations onto known Parkin structures in different stages of activation indicated that pathogenic mutations often introduce severe steric clashes in various binding and active sites, unlike the benign mutations that only cause minor steric hindrance. An additional group of missense mutations was identified that increased mitophagy compared to wild-type and could rescue some pathogenic Parkin mutations when introduced in cis. Structural modeling showed that these hyperactive mutations relieved the impact of a steric clash from the pathogenic mutations and allowed the necessary protein interactions and structural changes for Parkin activation and function. For example, R234Q and R256C destabilize the RING0:REP interface, whereas M458L destabilizes the RING0:RING2 interface; all of them favoring the Parkin active state.
In conclusion, using a combination of cell-based assays and structural studies predicted the pathogenicity of a mutation. Developing drugs that mimic distinct hyperactive mutations in cis to pathogenic mutations can lead to customized genotype-specific therapies for patients with Parkinson’s disease.
Original Research Article: Yi, W., et.al. “The landscape of Parkin variants reveals pathogenic mechanisms and therapeutic targets in Parkinson’s disease.” Hum Mol Genet 28.17 (2019): 2811–2825.

Uncovering new insights into diabetes-associated pain

Written By: Abel B. Cortinas

Original Article: Cheng et al. Diabetes 2019
The Gist of It:
Diabetes is a debilitating disease that affects nearly 1 in 10 individuals in the United States (American Diabetes Association, 2015) and nearly 1 in 18 individuals globally (International Diabetes Federation, 2017). Unfortunately, amongst diabetics, one of the most common complications is known as diabetic peripheral neuropathy, sometimes abbreviated as “DPN.” DPN leads to nerve damage that can result in a number of symptoms including numbness, loss of feeling or sensation, and sometimes pain in your feet, legs, or hands. DPN is also associated with an increased susceptibility to foot and ankle fractures and ulcerations, as well as lower-limb amputations. There is currently no disease-modifying treatment or therapy for DPN, which means that the drugs that are available (for example, calcium ion channel inhibitors such as pregabalin, serotonin / norepinephrine reuptake inhibitors such as duloxetine, as well as opioids such as tapentadol) are meant only to help alleviate the many symptoms but don’t help cure the source of DPN. To learn more about the underlying causes of DPN, researchers at the Chinese Academy of Sciences and Shanghai Tech University studied the gene Sarm1, which is known to be responsible for axon degeneration, or put simply the breakdown of nerves that is a key feature of peripheral neuropathies. The researchers wanted to understand the role of Sarm1 in reducing nerve damage, specifically axonal degeneration, as well as the implications on alleviating DPN. Cheng and colleagues uncovered a number of interesting findings. Particularly, diabetic mice that had the Sarm1 gene selectively deleted had alleviated sensitivity to pain, less nerve fiber loss in their feet, and reduced rates of axon degeneration compared with diabetic mice that had the Sarm1 gene intact. Together, these findings support the potential role of the Sarm1 gene and the resulting protein in DPN and highlight a potentially novel strategy for the treatment of DPN either through inhibition of SARM1 or gene silencing / editing.
Screen Shot 2019-10-14 at 9.43.58 AM.png


The Nitty Gritty:
Using a streptozotocin (STZ)-induced diabetic mouse model in animals with a Sarm1-/- gene knockout and with disruptions of exons 3–6, Cheng et al. found that these mice exhibited normal pain sensitivity and intraepidermal nerve fiber density in the footpad comparable to that of wild type mice. Additionally, the Sarm1-deficient diabetic mice were rescued from the modulation of thermal and mechanical pain sensitivity found in STZ-induced diabetic mice with the Sarm1 gene intact when subjected to the hot plate, von Frey filament, as well as tail immersion behavioral tests. Building on this finding, Sarm1 gene deficiency was further studied to determine its protective effects with regard to diabetic peripheral neuropathy (also known as “DPN”). Sarm1 gene deficiency was found to successfully alleviate diabetes-induced intraepidermal nerve fiber loss in diabetic mice. Next, the Sarm1 gene and its protective effects on diabetic axon regeneration were investigated. In an experiment that studied the effects of Sarm1 gene ablation on diabetic neuronal degradation, Sarm1 knockout mice had significantly alleviated diabetes-induced axon degeneration and change of the g-ratio (which is a measurement of myelin sheath thickness and generally considered to be indicative of nerve conduction capacity) as well as a decrease in NAD+ in the sciatic nerve. Lastly, the deletion of Sarm1 also diminished the STZ-induced changes in the gene expression profile of the sciatic nerve. Taken together, Sarm1 gene deficiency provided strong evidence to support the targeting of axon degeneration as a potentially promising strategy to combat DPN.
Original Research Article: Cheng, Y., et al. “Sarm1 gene deficiency attenuates diabetic peripheral neuropathy in mice.” Diabetes (2019). doi: 10.2337/db18-1233

What is Tau protein and how can it stop your brain from signaling during dementia?

Written by: Abby Stahl

Original Article: Sohn et al. Neuron 2019
The Gist of It
Frontotemporal dementia (FTD), also known as Pick’s disease, occurs when cells in the front and sides of the brain die or become dysfunctional. The disease can cause changes in behavior and personality as well as difficulties with understanding or producing language. FTD is typically diagnosed in individuals between 40–60 years of age and is inherited in a third of all cases. One specific change, or mutation, in a protein called Tau has been identified as a major cause of FTD. Scientists wanted to know how this mutated Tau protein could affect the electrical signals that brain cells use to communicate. To answer this, they looked at areas called axon initial segments (AIS), a part of the exterior barrier of the cell that is essential in controlling the sending of these electrical signals. They compared the AIS in brain cells taken from healthy patients with those taken from patients with FTD. The researchers found that brain cells from patients with FTD have a shortened AIS and send electrical signals erratically. How does this happen? The team discovered that mutant Tau binds to another protein called EB3 more strongly than expected and brings it to the AIS; having the extra EB3 at the AIS causes the cell to become more rigid and less flexible, preventing the brain cell from sending electrical signals properly. When the amount of EB3 or mutant Tau was decreased in the cell, the AIS length was extended, and electrical signals returned to normal. This study nicely demonstrated how mutant Tau protein can stop the brain from signaling and identified potential new targets for FTD treatment.
Tau Neuron Paper.png

Brain cells with mutant Tau protein have shortened axon initial segments (AIS) and erratic electrical signaling. The mutant Tau protein binds EB3 at the AIS, over-stabilizing the connection between Ankyrin G (AnkG) and structural microtubule proteins and preventing normal AIS shortening and signaling

The Nitty Gritty
As signals enter neurons, they are integrated at the axon initial segment (AIS). If the strength of the signal surpasses a given threshold, the cell undergoes depolarization and emits an electrical signal down the axon. Interestingly, the AIS becomes physically shorter during depolarization, which can be measured by tracking the length of an associated protein called Ankyrin G (AnkG). In patients with FTD, mutant Tau is thought to impact the stability of AnkG at the AIS; however, the mechanism is unknown.
To study brain signaling ex vivo, researchers collected skin cells from patients with Tau mutation V337M and transformed them into induced pluripotent stem cells, then differentiated them into cortical neurons. Furthermore, the scientists made “isogenic control cells” from the same patient cells, in which they corrected the Tau mutation with CRISPR/Cas9 gene editing, so they could compare wild type and mutant Tau phenotypes within the same genetic background. Interestingly, the authors noticed that AnkG was shorter in cells with mutant Tau compared with healthy and isogenic control cells. When they added potassium chloride (KCl) to induce depolarization, AnkG could not get any shorter in the mutant cells. To test whether the contraction of the AIS would influence electrical signaling, they used multi-electrode arrays and found “firing rates” were similar, but cells with mutant Tau had more spikes in electrical signal within a given burst of activity and fewer network bursts per minute, meaning they were not functioning normally. To understand exactly how mutant Tau was causing these changes in AnkG length and electrical signaling, the authors studied EB3, which is known to be regulated by Tau and stabilize AnkG in the AIS. Using nuclear magnetic resonance (NMR) spectroscopy, the group found three potential binding sites where Tau could interact with EB3. They found that mutant Tau can bind to EB3 four times more strongly than the wild type Tau! Furthermore, EB3 was dispersed nicely throughout healthy cells but accumulated at the AIS in cells with mutant Tau, suggesting the interactions were playing a role in electrical signaling. Next, they treated cells with a small interfering RNA (siRNA) to block EB3 protein from being made. When EB3 was not available to interact with mutant Tau, AnkG was elongated, and normal electrical signaling was restored. Conversely, when EB3 was overexpressed within healthy cells, AnkG became shorter and did not decrease in size during depolarization, mimicking the diseased state. When siRNA was directed against Tau in mutant cells, EB3 levels decreased in the AIS, AnkG was elongated, and normal electrical signaling was restored. Finally, when Tau was suppressed but EB3 was overexpressed, there was no benefit in the disease cells, definitively showing that the improvements in electrical signaling were made through blocking the interactions of mutant Tau and EB3.
In summary, the authors showed that mutant Tau binds strongly to EB3, causing shortening of AnkG at the AIS. Shortened AnkG prevents normal electrical signals from being fired and likely contributes to the dysfunction of brain cells observed in FTD. Importantly, EB3 may be a new therapeutic target for treating FTD and restoring neuronal plasticity.
Original Research Article: Sohn, P.D., et al. “Pathogenic Tau impairs axon initial segment plasticity and excitability homeostasis.” Neuron (2019). doi: 10.1016/j.neuron.2019.08.008

Special Issue: Heading toward the tenure-track — Hints, Tips, and Things to Remember

Written by: Kaitlyn Sadtler

(These are my own experiences with the tenure track and those from people I have had close contact with, make sure to talk to diverse people to learn a bit more about what the path might hold)
“Science” is a broad field with lots of different career pathways. Many people choose to go along the path that leads them to consider applying for faculty positions at universities or research institutions. The first thing you need to ask yourself headed into the faculty application season is – is this the right job for me, or am I applying because this is what I’ve been told you do (aka, get your PhD, do a postdoctoral fellowship [postdoc], become faculty)? It is very important to note the sheer number of career options for PhD-level scientists and to make sure you follow the path that is right for you. Today I am writing about faculty and tenure-track applications as that was my path.
First and foremost, the true process of applying to tenure-track positions starts well before the application season. It’s about what you’ve done during graduate school and your postdoc(s), between your publications and your work beyond the bench. There are certain things that are engrained into the faculty application process that will give you a leg up in most departments, like having a K99/R00 grant and a Science/Cell/Nature (SCN) paper. Knowing this jaded me from my first year in graduate school, so throughout my PhD I assumed there was no way that I could achieve this and focused on what was important – the science. Ultimately, I did land a first author Science paper from my PhD (and here is the part where my eyes would roll as someone with a 1st author SCN said “10 tips to get a faculty job”). At that point I was still unsure if I was going into academia but still wanted to give it a shot, so I did a postdoc. The good news for the majority of applicants is I never got a K99. I didn’t apply for one because I started the faculty application process about 1.5 years into my postdoc. If you start the K99 application right when you start your postdoc it will take several months to get the documents together (let’s call it 6 months), then it will go to the NIH for review (another 6 months), which takes you to just over a year after starting. Most grants don’t get funded on the first submission, so you do your edits and resubmit. All in all, it’s at least (normally) a 2-year process (if you’re applying right from the gate without preliminary data). Given the funding levels of grants, postdocs with more data who have had the time to get a publication from their postdoc lab are most likely to land the K99. Even if a K99 doesn’t sound like a possibility for you, there are other fellowships out there, including those from private foundations. Equally important to the research and data for grants, during your postdoc head out to conferences, especially in the year leading up to your faculty applications. If your PI is not able to fund you, there are frequently travel awards, and most fellowships come with some discretionary money (providing your university hasn’t taken it all away with health insurance bills – MIT I’m looking at you and that >$4000 chunk of money). This way, you can begin meeting investigators from different institutions and understanding what the culture is like across universities. And also – they can start meeting you!
Check for conferences in your field that have future faculty poster sessions – these are fantastic for meeting universities that are interested in hiring new faculty. You’ll also learn about what sorts of research different universities are looking for. Regardless of the quality of your application, if the department is looking for a certain set of skills to complement the investigators that are already there, your application might not be something they focus on. Furthermore, remember that wherever you are applying should be somewhere you want to go and a place that you want to live. Science is hard enough; you should be happy with your location and the university. If you want to buy a house and have a yard and be able to drive into work, maybe New York City isn’t for you. That being said, if you want to have an apartment right above restaurants and a solid public transit system and several grocery stores within walking distance, maybe avoid schools in more rural areas.
Once you’ve approached the application season, just like for all grants, it’s important to start writing your research statement early. Write, re-write, sit on it for a couple weeks, look at it again. Get friends and advisors to help read through it. Also, remember, you are applying to academia to help teach and train the next generation of students and scientists. If you aren’t excited about teaching or mentoring, there are other routes to lead a lab and do great science that have a lower teaching load. Do not expect a university to jump on your candidacy if you ask how long you have to teach.
Eventually, you will get all of the paperwork in and apply to all the listings; then there is the wait. Sort of. Often applications will be staggered; by the time I put in my last application in December, I had a notification for a phone interview that got scheduled in early January. It is important to take care of yourself during this application season, as (if things go well) you will be traveling a lot and, between trying to keep some semblance of research going back in the lab and the interviews and applications themselves, you’re going to get pretty worn down. So, take a breather, eat a cookie. Or go for a run. Whatever works for you.

Social media is a great place to find support groups and advice on faculty applications. 

When you are at your interviews, make sure you know you are evaluating your fit in the department as much as they are interviewing you. If you didn’t have a great time at the interview, it’s probably not somewhere you should wind up. Chalk talks – the infamous part of faculty apps – are different depending upon the institution. Some places will let you get through your introductory spiel, while other places will have faculty start asking questions 2 minutes in. I had a couple where it was a sit-down interview where I was asked questions, a couple that were whiteboard presentations (no slides), and a couple that were PowerPoint presentations with questions. That being said, my very first chalk talk was one of the most miserable experiences of my life. And then I had several chalk talks that were great experiences – don’t get me wrong I was nervous as heck, but it felt like brainstorming about projects more than interviewing.
After the interviews come (hopefully) the offer(s). One thing I remember that I really wish I had more information on at this stage was how to budget for a startup package. After you get those offers, you submit your startup package request. Trying to remember all the little things you need to start up the lab is crazy. Small things you’re used to always being there but being quite innocuous – like a flammables cabinet – are suddenly up to you to remember to budget for. I tried to head this off by asking someone for an example spreadsheet, but instead of being sent theirs, they sent a chemist’s spreadsheet that was sent to them. Pro tip: get an example spreadsheet from someone in your field. Also, ask the university what will be included in your space; for example, do you need to budget for buying a biosafety cabinet? In general, for a startup position at an R1 university, in the biomedical world, > $1 million for a startup package is standard. This can change dependent upon the place; for example, everything is bigger in Texas, including the grants. Also, in different research institutes you can see more funding provided by the institute – at the NIH where I am, you negotiate an annual research budget plus your startup for equipment purchases. When you submit your budget, universities might come back with a counteroffer, which you are allowed to counter again. If they suggest that they can get you a used piece of equipment that works for your purposes, remember that that equipment will need servicing earlier than a new piece of equipment, and you might need to add potential service costs into your revised startup proposal.
Quite a few people apply for more than one application cycle. There are different opinions on how to approach this. Some people say to only apply once and not “poison the well” by applying before you’re “ready”. That being said, “ready” is a subjective term. I was told by some individuals to not apply when I did and that I should spend several more years in my postdoc, and I was told by others that I was ready and should definitely apply, and some even told me to not do a postdoc at all and apply right after graduate school. Ultimately, not all universities have openings each year, so if one year doesn’t work out for you, then check out what universities have openings the following year.
My overall advice: start early, make sure you have surrounded yourself with mentors who will advocate for you, take your time with writing, and remember – some interviews will be better than others, and if you come out of the application season with no final offers, it’s not the end of your venture into a tenure-track position. Ask for advice and for experiences from folks who have been on the tenure-track application trail recently (in addition to your mentors). Also, know that the academic landscape is evolving, and individuals from within the system are trying to drive change. And not to sound cliché, but, may the odds be ever in your favor. (Good luck!)