Developmental delay and movement defects in a worm model of Niemann-Pick Type A disease

In December of 2016, Perlara partnered with Wylder Nation to initiate an NPA PerlQuest, a research and drug discovery program for Niemann Pick Type A disease. If you’ve been following our blog, you’ve kept up with our fly team’s efforts in development of a fly model for NPA. Here I will describe the progress we’ve made with the nematode – C. elegans.

Why have 1 when you can have 3?

The gene causative for pathology in NPA disease is SMPD1. C.elegans have not one but three orthologs of SMPD1- asm-1, asm-2 and asm-3. While all three bear some similarity to human SMPD1, of the three, asm-3 is considered most similar to human SMPD1. An asm-3 null mutant (ok1744, strain name: RB1487) had been generated earlier and was available to us through the Caenorhabditis Genetics Resource Center, University of Minnesota.

Prior to our studies on the asm-3 mutant, there were four published papers which evaluated the role of the various asm orthologs in nematodes. Surprisingly, all three asm’s are thought to bear sphingomyelinase activity at acidic pH, though at different stages of life and development of the worm. Cumulatively, these papers also showed that asm-3 was important for regulating lifespan of animals. However, not much was known about a screenable phenotype in these worms or even whether the asm-3 null mutant animals could be a model for human NPA disease.

Uncovering new phenotypes

We embarked on characterizing the available asm-3 null, allele-ok1744. This strain bears a 1.5kb deletion in the asm-3 gene. Our thinking was if the reported phenotypes of lifespan extension were the only ones that exist, then we would generate new mutants that lacked all three asm gene orthologs. This would negatively impact the likelihood of us finding a viable worm (lacking all three gene orthologs) but it was worth a shot, especially if there were no screenotypes to be found in the asm-3 null mutant. With a back up plan ready, we tested the asm-3 null mutants on the following phenotypes:

1. ability to grow at rates similar to ‘normal’ or wildtype animals
2. ability to reproduce at rates similar to wildtype animals
3. ability to crawl on solid-agar media at speeds and shapes similar to wildtype animals
4. ability to swim (technically referred to as thrash) in liquid media at speeds and shapes to wildtype animals

We found that in growth, reproduction and in ability to swim, asm-3 nulls were much slower than their age-matched wildtype control animals. You can see this in Fig. 1 and observe the swimming behavior in this video—

 

 

We also quantified the levels of mRNA of asm-1, asm-2 and asm-3 in wildtype and asm-3 null mutants. We found that while asm-1 and asm-2 levels were not changed significantly from the wildtype animals, asm-3 transcript levels were zero compared to wildtype animals indicating that this model behaved like a true null mutant.

Figure 1.  Box and Whisker plot of Swimming behavior in Asm-3 nulls and Wildtype animals quantified as ‘head thrashes/second)

 

But what about the enzyme?

Because of the presence of asm-1 and asm-2 orthologs in worms, we sought to understand if the behavioral movement defects observed in the asm-3 null animal arose from reduction of asmase enzyme activity. After all, enzyme defect is part of the disease pathology observed in humans with NPA disease. We quantified the enzyme activity in wildtype control animals and asm-3 null animals of the same age. This biochemical test showed that asm-3 null animals have roughly 30% the enzyme activity of wildtype animals. The residual enzyme activity, we hypothesized, arises from enzyme variants associated with asm-1 and asm-2 genes (Fig.2)

Figure 2. Asmase enzyme activity in wildtypes and asm-3 null mutants. Asm-3 null mutants display 38% asmase activity compared to wildtype animals.

Because NPA disease is associated with increased sphingomyelin levels in humans, we also queried the asm-3 null animals for levels of sphingomyelin using a biochemical assay kit from Cell BioLabs Inc, San Diego, CA (Catalog No. STA 601). The kit relies on breaking down intact sphingomyelin derived from animal tissues using a series of enzyme steps as detailed below (Fig.3A). At the end of the reaction, a fluorescent substrate is measured. More fluorescence substrate equals more sphingomyelin to begin with. Through this assay, we determined that asm-3 null animals had more than twice the sphingomyelin present in wildtype worms at the same age!

Figure 3A. Steps of sphingomyelin breakdown reaction resulting in a fluorescent substrate that can be measured. Schematic obtained from user manual published with Sphingmyelin Assay kit from Cell Biolabs, Inc, (Cat No. STA 601)

Figure 3B. Normalized Sphingomyelin levels in WT and Asm-3 Null Mutants at age-matched adult worrms

Piecing together a few more parts of the puzzle

While all data thus far pointed towards the asm-3 null mutant showing phenotypes associated with NPA disease, we conducted a few more validation experiments to convince ourselves. One way to tell if a faulty gene is causative for defects seen in a disease is to re-introduce a normal copy of the gene and see if the defects disappear. In worms you can do this in two ways- a) you can mate the deficient animal to a wildtype (normal gene-bearing) animal and introduce a normal copy of the gene (also called as backcross) b) you can introduce a copy of the human gene and see if it replaces the function of the worm gene ortholog. We performed both a) and b) and assessed enzyme activity and swimming behavior on the backcrossed asm-3 worms and humanized SMPD1-bearing worms. Results from b) are still expected. Results from a) were exactly in line with what we expected! We found that adding a single copy of wildtype asm-3 gene back into the asm-3 null animal ‘normalized’ their swimming behavior though not quite to wildtype levels. In keeping with this ‘mid-level’ rescue of behavior, the asmase enzyme activity observed in the backcrossed animals is between that of the asm-3 null (lowest activity) and wildtype animals (highest activity) (Fig. 4)

With these data in hand, we were confident that the asm-3 null animals displayed many of the features of human NPA disease and is a suitable model an NPA drug discovery screen.

Are we there yet?

Last but not the least, while we had an NPA phenotype we needed to determine that we could conduct a screen on these animals with potential to find molecules that would rescue the disease. While we have spoken a lot about the swimming defect observed in the asm-3 null animals, this phenotype would make a primary screen a lot slower because we would need to sample videos of thousands of worms treated with several thousands of molecules. We instead decided to focus on the slowness of growth observed in the asm-3 null mutants. To conduct a small molecule screen, this phenotype would need to be captured with good signal to noise ratio in a high throughput format, i.e. the defect would have to be severe enough that separation between wildtypes and mutants can be reliably observed, yet not so severe that it cannot be rescued (e.g. if assay conditions cause death, a small molecule would likely have fewer chances of rescuing such a defect). As displayed in the figure below, under our protocols of liquid culture growth, we found that asm-3 null mutants do grow but at a rate significantly different from wildtype animals. This means we can look for molecules that rescue the growth defects in asm-3 null mutants and make them look more like wildtype animals.

References

1. Lin X, Hengartner M O, Kolesnick R, ‘Caenorhabditis elegans contains two distinct acid sphingomyelinases’, The Journal of Biological Chemistry, 1998, 273(23): 14374-14379.
2. Kim Y, Sun H, ‘Functional genomic approach to identify novel genes involves in the regulation of oxidative stress resistance and animal lifespan’, Aging Cell, 2007, 6 (4): 489-503
3. Kim Y, Sun H, ‘Asm-3 acid sphingomyelinase functions as a positive regulator of the Daf-2/Age-1 signaling pathway and serves as a novel anti-aging target’, PLOS One, 2012, 7(9): e45890.
4. Schmokel V, Mernar N, Wiekenberg A, Trotzmuller M, Schnabel R, Doring F, ‘Genetics of lipid-storage management in Caenorhabditis elegans embryos’, Genetics 2016, 202 (3) 1071-1083.