Here at Perlara we are excited to announce our probe into Pompe disease which, along with Cori disease, comprise our glycogen storage disease PerlQuest. For this project we will be collaborating with the Warren Center at the University of Notre Dame. Our goal is to develop an accurate Pompe disease fly model, and to test a select group of candidate drugs in hopes of finding a promising treatment for the disease.
Genetics of Pompe disease
Pompe disease is caused by different mutations in the GAA gene, which lead to the reduced function of an enzyme called acid alpha-glucosidase (also known as acid maltase). Acid alpha-glucosidase generally breaks down glycogen into glucose, so a protein that isn’t functioning properly will lead to a buildup of glycogen in muscles as well as in the liver. This buildup causes muscular symptoms and liver disease in patients.
Secondarily, defective GAA leads to a block in a process called cellular autophagy – one of the ways that cellular material like proteins and organelles get recycled. Glycogen is normally broken down in small vesicles called autophagosomes. When acid alpha-glucosidase isn’t functioning, these autophagosomes fill up with glycogen and can no longer bind to lysosomes, and autophagy is blocked.
Muscle biopsy showing large vacuoles in a case of Pompe disease (acid maltase deficiency, HE stain, frozen section). Source: Jensflorian, Wikipedia
Pompe disease in patients
In patients, Pompe disease, also known as glycogen storage disease type II, presents with myopathy (muscle weakness), hypotonia (poor muscle tone), an enlarged liver, and heart defects. There are 3 different onset patterns in Pompe disease. The first is “classic” infantile-onset, where symptoms begin to present at a few months old and can present as weak muscles and a failure to thrive. The next is known as “non-classic” infantile-onset, where symptoms present at around age 1. The least common is late-onset, where symptoms start to appear in late childhood, teen and even adulthood. Usually, with late-onset cases, symptoms are much milder (generally muscle weakness), and tend to progress slower than in the infantile-onset of the disease. There does tend to be a correlation between the onset and severity of the condition and the amount of enzyme activity left intact.
Developing a Pompe disease fly model
There is no known genetic equivalent (homolog) of GAA in flies. However, Jonathan Zirin, while working in Norbert Perrimon’s lab at Harvard, was able to successfully replicate the cellular process in Drosophila using a known side effect of the drug chloroquine. The drug is generally used to treat malaria, but 12% of patients on the drug can also experience myopathy as a side effect. Chloroquine (CQ) increases the pH of lysosomes, which disrupts their function, and prevents them from binding autophagosomes. This block prevents the process of autophagy, leading to a buildup of glycogen that can’t be broken down. Sound familiar? This process is very similar to what happens in the cells of Pompe patients, which is what makes this paper so promising!
Here at Perlara we will use third instar fly larvae fed on CQ-dosed food in order to observe and quantify myopathy caused by glycogen buildup. As a part of this process, the animals will undergo a starvation period, because nutrient deprivation causes a natural increase in cellular autophagy. When cells sense that there isn’t sufficient nutrient in the environment, they move glycogen into autophagosomes so that it can be broken down into a more usable form, glucose.
Zirin et al. used fluorescently tagged proteins in larval muscles to show how CQ prevents the degradation of autophagosomes by blocking autophagy from occurring (Figure 1). A surface protein of the lysosome is tagged red in these images, the autophagosomes are tagged green, and nuclei are tagged blue. In the first image you can see that there are no small pockets of green or red, meaning that there is no formation of autophagosomes or lysosomes. In image F the larvae have been starved (which should induce autophagy), and we see some red and green dots show up. The yellow arrow on image F is pointing to a small yellow dot that indicates that this particular pocket is positive for both the autophagosome tag and the lysosome tag – meaning that the two have fused. This fusion suggests that glycogen is being successfully broken down. In image G we see a slide of muscle from larvae that have been starved and treated with CQ. In contrast to the previous image, there is a clear buildup of both autophagosomes (green) and lysosomes (red) in image G.
Figure 1: Autophagasomes are labelled green, lysosomes are labelled red and nuclei are labelled blue. In the first image (image E) cells are taken from a fed larvae. In the second image (image F) larvae have been starved and there are some spots of green and red showing the induction of autophagy. The arrow in this image is pointing to a vesicle where the green (autophagasomes) and red (lysosomes) are colocalized, suggesting that glycogen is being broken down successfully. Image G shows starvation in the presence of CQ, and it is clear that there is a buildup of both unfused autophagosomes and lysosomes without colocalization, leading to a buildup of glycogen that is not being broken down. Source: Zirin, et al.
The next step in developing a Pompe disease fly model is to show if the appearance of autophagosomes indicates that glycogen is being taken up in order to be broken down into glucose. Otherwise, we would only know that these autophagosomes appear, not what is in them. In the next set of images (Figure 2) , glycogen is labelled red, the autophagosomes are labelled green, and nuclei are labelled blue. When larvae are treated with CQ and starved, you can see – with increased time of starvation – the red and green tags start to align with each other. This means that glycogen is being taken into the autophagosomes for degradation.
Figure 2: Glycogen is labelled red and autophagasomes are labelled green. These images are taken at 0, 3, 6, and 8 hours of starvation, and it is clear that as starvation continues, glycogen is colocalizing into the autophagasomes, showing that starvation induces autophagy. Source: Zirin, et al.
Now we know glycogen localizes in autophagosomes and that CQ treatment blocks autophagy which, together, prevent glycogen from being broken down. Next, we need a way to measure this phenotype. The Zirin et al. paper quantifies the CQ-induced myopathy by tracking some of the larval muscular defects (Figure 3A). We are currently working to try and reproduce those data. First, we will generate a dose curve of CQ, measuring its effects on larval mobility. We will take the larvae that have been both treated with CQ and starved, place them on agar and film their movement for 2 minutes. This video will be analyzed for path length and larval speed. In some of our early tests, we have seen a significant difference between larvae that have been treated with a high dose of CQ and larvae that have not been treated (Figure 3B).
| 3A |  |
| 3B |  |
Figures 3: These are two examples of the tracking done on non-CQ-treated larvae, and larvae treated with 50mg/ml of CQ. There is a stark difference between the speed of treated and untreated larvae (each point marks the position at every frame, these images are to scale and time between frames is equal). Source for image 3A: Zirin, et al.
Next steps
We will continue to develop this assay, optimize our CQ-treatment and starvation model, and perfect our larval tracking. After optimization of our Pompe disease fly model, we will start testing a small subset of drug compounds. Stay tuned to see what we learn from the results!
