Team 8
Mackenzie Lynch, Fabiana Delgadillo, Polina Klishina, Sofia Orrantia
Introduction to ALS
Amytrophic Lateral Sclerosis (ALS) is a neurodegenerative disease that impacts nerve cells responsible for muscle movement. Symptoms include challenges with walking, talking, and breathing (in its latest stages). For about 10% of people, a genetic cause can be named, but other risk factors including aging and gender. At its diagnosis, the average survival time for ALS is two to five years, with symptoms significantly worsening during the last three to five months. Current resarch into ALS aims to understand how changes in genetic makeup can contribute to the symptoms and progression of the disease. Because few genes have been tested, our goal is to discover new genes related to ALS and their impact on muscular decline.
To carry out our project, we have chosen a gene of interest call DLL. Using the Drosophilia melanogaster model of ALS, we manipulated expression of DLL in glial (supportive) cells. To mimic ALS symptoms, we used an overexpression of TDP43. Using either RNA interferance (to reduce function) or UAS lines (to increase expression), we crossed and analyzed flies using a a petri dish assay. In this behavior test, flies were placed in a petri dish and left to move on their own. Recording over a period of three-four weeks, this unbiased assay allowed us to observe mobility changes in our gene of interest flies.
Gene of Interest: DLL
DLL, formerly known as Distel-less, is a Drosophilia gene primarily responsible for DNA binding to a transcription factor, allowing for limb development in the flies. Using the Catanese data set, we found DLL to have a Log2fC (changes of expression in ALS) of +4.6, meaning is 4.6 times more expression in ALS cases! In this data set, the signifance of the log2fC (adjusted p-value) was 0.003. This means the results were unlikely to have happened by chance. With these numbers, we are confident in using this gene for our experiment.
Because flies share up to 70% of their genome with humans, DLL has two primary human gene matches, called DLX1 and DLX6. In humans, the DLX group of genes are thought also to be a transcription factor involved in the morphogenesis of developing limbs and sensory organs.
Selecting Progeny
Above is the genotypes of our parent flies. We were able to use 100% of progeny. Did not sort out unwanted phenotypes such as curly wings or stubble. As controls, we used alrmGAL4 (astrocytes) and TDP43.
Methods
After collecting the virgins, we made crosses with them to obtain offspring for testing. Female flies had an AlrmGAL4 genotype, while male flies had three different genotypes: TDP-43, wild type, and DLL (our gene of interest). Distinguishing between males and females also requires some knowledge. The easiest way to determine the sex of the flies is to examine their genitals. Female genitals are pale, while male genitals are dark. Additionally, one can use dark bristles on the front pair of legs to identify females, although we did not utilize this method.
After making a cross, we waited for pupae to appear. Once they hatched, we flipped the vials to prevent mating between offspring and parents. Flipping the vials is a straightforward process. We tapped the vials with flies to prevent them from flying away, covered these vials with an empty vial, and then flipped them. This way, parents would appear in a new vial while the pupae remained in the initial vial. It was also important to relabel the vials and create a "frame" for the date of birth.
We waited at least one week for offspring to hatch, and once they hatched, we prepared them for a behavior test. Our group determined that analyzing three flies per video was optimal, so we separated all the offspring into several vials with a specific number of flies in each. Although we were testing three flies, we placed five flies per vial, as testing flies can encounter various unexpected circumstances: some flies may be damaged, and some may simply fly away.
Our group chose a petri dish test, and here are the steps for it:
Flip your flies into an empty vial.
Leave them on ice for 2 minutes.
Then flip them into a petri dish. (We also used a white background to improve video quality.)
Let them wake up for 5 minutes. This step is crucial as flies need time to recover their motor functions after being on ice.
Now you are ready to record your flies.
Install the light on the stand so that it shines on the sides of the petri dish.
Record for the first 30 seconds to ensure everything is set up correctly.
Show the label with genotype, date of birth, date of recording, and team number at the beginning of the video.
Record for an additional 2 minutes.
Analyze your videos!
AnimalTA
Published by Violette Chiara, AnimalTA is a program designed to process and track a given video. Through AnimalTA, the user is able to correct tracking errors, analyze trajectories, assess spatial components, and interactions between targeted subjects.
In our lab we used AnimalTA to analyze the average speed and total distance traveled of the flies.
Hard to detect, image above shows the red line to adjust video size.
Before and after concealing a fly using modify background
After creating 'New project' and clicking 'add video(s)' to upload your desired video, your screen should look like this.
Using the crop video feature, the user can crop the video length by using 'select video start' and 'select video end' to shorten video or remove sudden movement that could result in tracking errors.
The user can also adjust the video size both vertically and horizontally by moving the red line on the edges to desired location.
When finished, click validate.
'Stabilize video' is not needed unless there is light movement or vibration seen throughout the video
Click on 'Create background' and make sure to check if there is no flies in the back ground by clicking on 'modify background'. If there is no flies, move on to the next step.
If there is flies, right click on the mouse next to the fly to copy color. After copying the color, click on top of the fly to conceal the fly.
After both:
Clicking on 'define arena' and defining the arena in accordance to the shape of the area where your flies are and
Clicking on 'define scale' and inputting the length of the arena in 'real length' which in our case was 10cm
Click on 'prepare tracking' and this screen should appear
Using 'Threshold' , slide the curser till only the flies are visible in the black background.
Using 'Distance Threshold, slide the curser till around 1.30
Using 'Number of individuals per arena' slide the cursor to reflect the number of flies in the video.
Click on 'Validate' when finished.
Click on 'Begin tracking' and on the pop-up window click on desired video and then click on validate and then wait till video is done processing
When finished click on 'Prepare analyzes' and then 'Analyses of individual trajectories' to see data collected.
Download the data by clicking on 'Run analyze'.
Results: Temporal Changes in Locomotion and Model Fit Analysis
Wildtype analysis based on aging
Our graph results demonstrate an increase in average speed as flies age, from week 1 to week 3 from Wildtype. However, it is expected that their locomotor function would decrease with aging. Thus, several factors might have influenced these results, such as some individual flies not moving or being out of range during the recording of behavior tests, among others.
Consider the coefficient of determination, 𝑅^2, which is a statistical measure that provides information about the goodness of fit of a model.
Values closer to 1 suggest a better fit of the model to the data.
TDP43: Average Speed
Our graph portrays the expected decreasing relation in motor function as flies age.
Our R^2 is 0.973, the regression line (exponential line) approximates the actual data.
Motor function while aging: DLL
Our graph shows the decrease of motor function recorded using the petri dish test on DLL flies.
Our R^2 is far away from 1 or even 0.95, which explains the unreliability of our model to fit our data.
Analyzing the graph
Notice how the Error Bars of the standard deviation overlap!
They behave this way, when there's a chance that the difference is not statistically significant.
However, we need to check the P-test and compare it to the alpha value to make sure that our prediction is right.
Different Samples per Gene: ANOVA Analysis
In this analysis, we examine three samples corresponding to distinct genes at week 1:
Sample 1: WildType (AlarmGAL4)
Sample 2: TDP43 (with AlarmGAL4)
Sample 3: DLL (with AlarmGAL4)
As indicated in the table, the P-value is 0.669, while the significance level (alpha) is set at 0.05. Consequently, the difference in average speed between the genes at week 1 is statistically insignificant.
Furthermore, although our group anticipated an increase in motor function loss as the subjects aged, we did not expect to observe significant changes at the initial week 1 comparison.
Conclusion
Both TDP43 and DLL have a negative effect on average speed from early life. DLL had the most immediate effect on motor skills with the average speed being the lowest out of all of the levels (Wildtype, DLL, and TDP43).
Through our ANOVA result, we can conclude that there was no significant difference in average speed the presence of DLL or TDP43 but it must be noted that further testing is needed since we only had one trial of 3-week data of DLL to base our conclusion with.
Next Steps & Challenges
Advice/ Thoughts For Future Semesters:
With the introduction of the petri dish assay, we encountered several challenges that could be improved upon next semester. First, it was especially difficult to transfer the flies from the vial to the petri dish, but we found putting the flies on ice to be helpful. Additionally, using CO2 after the flies are completed with their behavior makes the transfer back to the vial seamless. With that, however, managing to insert a CO2 probe into the petri dish without losing flies is an issue that could be sorted out next semester.
AnimalTA was by far one of the most challenging aspects of the course, as it was a new learning experience for everyone, including Dr. B. Our amazing group member, Fabi, noted that watching Youtube videos on the software itself made it easier to learn about how to analyze our fly videos. Still, we run into challenges with the system recognizing flies and recording their movement for the entire video. For us, practice made progress when it came to AnimalTA, and that is the advice we could give to future semesters.
We would like to have done experiments using the 12-well petri dish, as it would show us how a lone fly would move in a smaller environment. This may make analysis easier.
More control data established in the beginning of the semester and more experimental data done later in the semester. This way, we can have something to compare our gene of interest data to before we begin.
Limitations of this Study:
For our group specifically, we did not have a returning student to guide us with organization and basic concepts in the course. With a returning student, we could have used methods from previous semesters and spent less time on figuring out how to organize crosses and communicate as a team. (We did do a good job in the end!)
Running into issues with both the tracking software and the petri dish assay made it harder to move fully into analyzing results.
We don't know if the mobility changes in flies are specifically related to our purposeful manipulation of genes, age, or environmental factors. It is hard to control a flies environment in order to narrow it down to one variable.
Future Crosses:
Besides further work on the DLL gene, we would like to work on a gene that down regulated (weaker) in ALS. Our gene is 4.6x more expressed in ALS cases, so it would be worth it to compare to a gene in the opposite direction. For example, we could used RDL (GABRR1 in humans), which is -1.5 times less expressed in ALS pateints (Ziff). A cross for this gene can be found below:
