In one of my very first lab meetings, my PI, Dr. Alan Beggs, showed us a photograph of a group of the lab’s patients. They were living with SELENON-related and myotubular myopathy, just some of the rare muscle diseases our lab studied. Some were infants who could not sit upright without support. Others used wheelchairs or depended on ventilators. Some were standing, but had rigid spines. The photograph had babies, school-age children, college students, and parents. What stayed with me the most were the patients not in the photo, whose conditions made them too sick to travel. The work we did wasn’t just about understanding rare disease, but giving families answers where few exist.
What I spent the rest of the summer building was a very small piece of that.
About SELENON-myopathy
This summer, I was a lab intern in the Beggs Lab and the Manton Center for Orphan Disease Research at Boston Children’s Hospital. My project studied SELENON-congenital myopathy (SELENON-CM), a rare congenital neuromuscular disease affecting roughly 1 in 200,000 births. Children with the condition present early with spinal rigidity, muscle weakness and scoliosis. While they are able to walk for most of their lives, they succumb to respiratory sufficiency over time. There is currently no cure for SELENON-CM.
The SELENON gene encodes for production of selenoprotein-N (SelN), which is on the membrane of the endoplasmic reticulum. Muscle cells depend on calcium to contract and relax, and SelN regulates the SERCA channel that pumps calcium back in between contractions. Additionally, SelN is a reductase that helps the cell manage its internal chemical balance between oxidation and reduction. When SelN is not produced, calcium levels are not regulated and oxidative stress increases, damaging muscle tissue.
Under my mentor, Dr. Pamela Barraza-Flores, we aim to understand what the loss of SELENON does, using mice, zebrafish, and immortalized cell models. In the past, the lab has studied SELENON rescue, phenotyping, and histology. Through this work, the lab has generated a large list, hundreds of compounds long, of candidates to test, but no efficient way to screen them.
Flow Cytometry Assay
I worked on developing a mid-throughput screen for testing possible candidates, by creating a reproducible flow-cytometry based assay that measures oxidative stress at the single cell level to discriminate between SELENON-deficient and wild-type cells, in both immortalized human and mice models. Flow cytometry is a quantitative way to measure properties of the cells, including size and fluorescence. Low oxidative stress should correspond with SELENON-healthy cells (also known as wild-type cells) and high oxidative stress should correspond to the SELENON knockouts. A treatment that is effective should reduce the knockout levels of stress back towards a healthy baseline, where an assay could shortlist candidate compounds based on how much they lower oxidative stress. We hypothesized that oxidative stress would be elevated in SELENON-knockout cells compared to the wild-type lines, with a clear separation to test small molecules against.
To develop the assay, I worked with mouse immortalized C2C12 myoblasts wild-type and SELENON-knockout lines as well as human immortalized myoblasts, using wild-type and two SELENON-knockout cell lines generated with CRISPR. This was to ensure the results were reproducible across model types.
Since these cells were immortalized, they were grown continuously in culture, receiving a media change every 2 days. We measured the growth of the cells by assessing the confluency, which was how closely packed the cells were. Flow cytometry needs enough cells to produce a reliable read, but cells that are too crowded will experience stress from overcrowding. To ensure that we had our target confluency, I would periodically split the cells by counting how many were in the current well and resuspending in a new plate with only 50,000 cells. I aimed to have 50-60% confluency, or concentration of cells on the day I ran the experiment.
For the staining protocol, CellROX was a deep red reagent used to measure oxidative stress, while the Sytox Blue Dead cell stain distinguished living cells, which was important for gating. We also included a positive and negative control to establish baselines of what stressed cells should express. The positive control was treated with TBHP, a stable form of hydrogen peroxide that increases oxidative species, while the negative control was treated with NAC, an antioxidant that balances the oxidative species.
After cells were stained with both dyes and their respective treatments in both the wild-type and knockout lines, they were harvested with trypsin to pull them off the plates, and put into FACS tubes to be read in the flow cytometer.
In the flow core, I worked with Sakshi Bubna and Dr. Ron Matthieu to process my samples. Flow cytometry measures the physical and chemical properties of cells as they pass through light. Once a tube was inserted, the cells passed one by one in front of a laser beam that measured forward scatter, which corresponds to the size of the cells, and side scatter, which corresponds to the fluorescence and granularity of the cells. These measurements are then plotted against each other. In step 3 of the figure below, for example, forward scatter and side scatter were plotted on the two axes.
From there, different combinations of measurements were used to gate for cells of interest, and filter for debris and dead cells. Gating means drawing a boundary around the population of interest for analysis. I then plotted APC-A, the intensity of CellROX expressed, on the x-axis, and the count of cells that shared that intensity on the y-axis. The peak of that histogram tells you the level of oxidative stress carried by the largest share of cells in the sample. Below is an example of what gating looked like performed in an app called FlowJo, with a picture of the final APC-A plot of interest, also known as a reactive oxidative stress histogram, on the right of the figure below.
I ran several flow cytometry experiments over my first month in the lab, trying out new stains, creating a rescue curve, and using the human model. The results of each experiment would inform the next, and I spent the majority of my time at the beginning troubleshooting, whether that was finding the right confluency to stain my cells at, the appropriate levels of an NAC rescue curve, and learning how to use FlowJo.
After gating, the reactive oxidative stress histograms for wild-type (blue) and knock-out lines were plotted against each other (red). In the figures below for 3 different experiments, there is a consistent separation in CellROX expression, or ROS, between WT and SELENON knockouts in both immortalized mice and human myoblasts. These were promising first results, but our next steps were to see if treatments, like the negative control of antioxidant, would help “rescue” the SELENON knockout to lower levels of stress.
In the next experiment, we plotted the positive controls for the WT and KO lines together, and repeated for the negative controls. There is clear separation between the WT and KO lines, with lower oxidative stress in the negative controls that were rescued with NAC.
Next, I investigated an NAC dose response curve, where the ideal level of NAC for recovery was 100 μM. This work is ongoing.
Conclusion
I found that flow cytometry can be used as a reliable tool to quantify changes in oxidative stress in myoblast cells. SELENON-knockout mouse and human myoblasts show higher CellROX fluorescence compared to wild-type, producing two distinct population peaks. Treatment with NAC successfully reduced CellROX fluorescence levels in human myoblast KO compared to no treatment, which means the assay can detect not just a difference, but a rescue in the knockouts.
Results from these experiments will advance early-stage therapeutic testing for SELENON congenital myopathy, but my future steps include further refining this assay. I will continue to repeat CellROX staining and NAC dose curve for reproducibility, test a new MitoSOX stain to separate mitochondrial from total cellular ROS, and use other similar antioxidants like NAC as additional positive controls. The purpose of testing MitoSOX is to further understand whether SELENON impacts the mitochondria more than the overall cell, and get specific about where its impact is.
Single-nuclei work
In parallel with the flow cytometry, I analyzed single-nuclei RNA from muscle biopsies of 16 healthy pediatric donors, using the Seurat package in R. The goal was to see what SELENON expression looked like in healthy donors and the role it played in the post. Two of my graphs of interest were a bar graph showing the percentage of SELENON-positive cells within each cell type and a comprehensive UMAP of the 16 samples with cell types labeled by SELENON expression level. The UMAP is a model that clusters cells by similarity, as shown by the groups in the figure.
As you can see, SELENON in purple is expressed at extremely low levels across nearly every cell type, and yet losing it is enough to cause a disease that shortens children's lives. In the UMAP, SELENON expression is most dense in the venous endothelial cells and muscle stem cells, but at low levels in comparison to the cell population. It gave me a different kind of context for my flow cytometry work, where I was measuring the downstream consequence of losing a protein that healthy muscle barely makes in the first place. I am continuing to analyze the R data, with the goal of filtering the SELENON-positive muscle cells to perform differential expression on other qualities they express. I also hope to add more patients to the analyzed cohort as the lab receives biopsies, to eventually have a repository that characterizes healthy SELENON.
Why this matters
Some of my colleagues had personally known or been part of families who lived with someone who had congenital myopathy. Working in a translational lab helped put into perspective how our research directly impacted patients.
One of the most impactful moments of my summer was hearing about the first participant that had been dosed in a clinical trial for X-linked myotubular myopathy, using an AAV vector our lab redesigned for MTM1 gene delivery. This patient went from not being able to sit up or roll over as a baby to kicking a soccer ball, playing baseball, running around, and being a normal kid just a couple years later. It was incredible how this patient was slowly able to gain back muscle strength, especially since they received the dosage so early.
Last week, I was having a conversation with a collaborator at Children's who studies the post-mortem tissues of patients who receive ASO treatments, to understand the biology of rare neurodegenerative diseases. ASOs are antisense oligonucleotide therapies that use short s strands of DNA to bind to mRNA to change how a gene is expressed. Rare diseases in her space are usually one gene and one specific mutation, which makes it easier to develop an ASO than complex diseases where hundreds of variants contribute a little. She finds patterns across post-mortem tissues that show up in the more complex diseases, arguing that rate disease research is a clearer path in understanding common disease and rare diseases across patients. In the rare disease space, you can develop a specialized treatment for one patient, diagnose what type of muscular dystrophy they have, and work with genetic counselors for the best treatment options. However, this also tends to be where money is the thinnest.
I came into this summer already interested in translational research, and how incredible findings make it off the lab bench to the patients. While my project helped develop a test to screen possible candidates, this type of infrastructure can be applied beyond rare muscle diseases to other diseases where oxidative stress is the symptom. Long term, I hope that my leadership-in-action project can help address accessibility for both labs, and patients, when it comes to screening therapies, receiving counseling, and developing personalized treatments.
This summer was truly an eye opening experience into what assay development in the rare disease space looks like. An extra special thank you to my mentor, Dr. Barraza-Flores and RA Won Lee for making this project possible and supporting my growth as a student, as well as Sakshi Bubna and Dr. Ron Mathieu in the Flow Core! Thank you to the Beggs Lab and Manton Center for their continuous support.