This web page was produced as an assigment for an undergraduate course at Davidson College.

Mitochondrial Pathology and Apoptotic Muscle Degeneration in Drosophila parkin Mutants

Jessica C. Greene, Alexander J. Whitworth, Isabella Kuo, Laurie A. Andrews, Mel B. Feany, Leo J. Pallanck


This website is a summary and critique of the paper named above. All information found in this website is from that paper unless otherwise indicated. For a link to the original paper, click here.

Background of Parkinson’s Disease Research:
The mechanism through which Parkinson’s disease (PD) affects those afflicted with it is unclear. This paper is one part of an ongoing investigation into what causes PD and the mechanism through which it acts.

In order to understand the history of Parkinson’s research, here is a brief list of breakthroughs and a description of their relevance.

One of the first breakthroughs in this field of research was the discovery that a Parkinson’s-inducing compound (1-methyl-4-phenylpyridinium) acted by inhibiting mitochondrial complex I of select cells, causing cell death to dopaminergic neurons. This led to the association of mitochondrial dysfunction and PD. Research then began to attempt to elucidate a mechanism through which mitochondrial dysfunction occurs.

Attempts to find this mechanism led researchers to monogenic forms of PD, such as autosomal-recessive juvenile parkinsonism (AR-JP). AR-JP has some differences from PD - unlike PD, Lewy bodies do not form in AR-JP, although dopaminergic neurons are still selectively killed. Researchers located the parkin gene that leads to AR-JP when it has a loss-of-function mutation.

Researchers then began to identify the function of parkin, and decided that parkin encodes an ubiquitin protein ligase. The function of the Parkin protein was determined from Parkin’s tertiary structure. The function of ubiquitin protein ligase is to join a specific intracellular substrate with the ubiquitin polypeptide, which targets the substrate for degradation (Doherty, 1996). Therefore, researchers reasoned that AR-JP occurred because of a failure to label specific substrates for degradation. Research in this area then began to focus on which specific cellular substrates Parkin targets. In this paper, the authors claim to create a Drosophila model of AR-JP by using P-element insertion to create a null parkin mutant to investigate the role of Parkin and to further elucidate the relationship between AR-JP and PD. They also make observations of Drosophila and the effect that mutation of the parkin gene has on the Drosophila phenotype.

Summary of Experimental Results:
Figure 1a: In order to create the fly model for AR-JP, the authors scanned a Genome Database for Drosophila DNA sequences with similarity to the human Parkin gene. Figure 1a shows the Drosophila Parkin protein (D-Park) in conjunction with the human Parkin (H-Park). Specific domains, such as the N-terminal ubiqiutin-like domain, or the RING finger domains are identified and boxed. The Drosophila Parkin protein has 59% similarity to the human Parkin, but only 42% of Drosophila Parkin protein amino acids match the human amino acids. (The measurement of similarity between proteins must also account for charge similarity of amino acids.) Also, the authors note that the Drosophila parkin gene codes for a protein that structurally matches the human Parkin protein by also containing the following structures: a ubiquitin-like domain, ring-finger domains, and an in-between ring finger domain. The authors state that this protein is the only Drosophila protein that contains all of those domains. The authors create a reasonably good case for considering Parkin as the Drosophila ortholog of the human Parkin protein based on primary and tertiary structure. Ideally the primary structures would exhibit more similarity so that the author's claim would be more definitive, and the authors would prove that the parkin protein in Drosophila matches human Parkin better than other human ubiquitin ligases.

Figure 1b: This is a Northern blot of poly(A)+ to determine if Parkin is present at different developmental stages in Drosophila. The authors probed this blot using a parkin specific probe. They claim to detect Parkin in wild-type embryos, larvae, and adults. This figure is missing several key elements – both positive and negative controls, as well as a loading control. A positive control would be helpful to ensure that the probe was unquestionably binding to parkin. Perhaps the positive control could be a sample of purified parkin mRNA. A negative control formed from the parkin null mutants for each of the three developmental stages would also be helpful, so that you could ensure that the two different adult parkins that appear are actually parkin and not artifacts. A loading control is necessary since they use this figure to make assumptions about the relative quantities of parkin in the different developmental stages. Without a loading control, it is impossible to make those assumptions. There is some evidence to show that the parkin is present in the three different developmental states, but without a positive control to show that the probe is actually binding parkin, there is no definitive proof that parkin is present. Also, there is not enough evidence to support the author’s claim that adult Drosophila produce more parkin than embryonic or larval Drosophila.

Figure 1c: This figure is a molecular map of the parkin gene. The purpose of this figure is to show the different deletion alleles that the authors work with in this paper. The parkEP(3)LA1 insertion can be seen, as well as the coding portions of the parkin gene. This figure displays the extent of the parkin gene that is removed from the park13, park45, and park25 genes.

Figure 2: This figure shows the differences in gamete cells between wild-type Drosophila testes and parkin mutants (Drospohila that lack the parkin gene). Pictures A and B show testes, and it is obvious that sperm cells are present in the wt Drosophila but not in the parkin- mutants. Spermatogenesis for parkin- mutants supposedly differs from that of wild-type Drosophila at the 64-cell germ-line cyst stage. Pictures C and D show a late-stage 64-cell cyst. The parkin- mutants appear to have fewer developing spermatids than wild-type Drosophila, although a clear difference is difficult to observe. In Pictures E and F, a clear difference is readily observed. Picture E, (wt) shows regular normal sperm cells, with the mitochondrial derivatives displayed regularly. Picture F (parkin- mutants) shows a dramatic difference from Picture E, in that the Neberken is irregular and poorly distributed between individual sperm cells. This figure shows conclusive evidence that Drosophila parkin- mutants have abnormal spermatogenesis that would confer sterility when compared to wild-type Drosophila.

Figure3a and 3b: These pictures clearly display the differences in wing posture between wild-type Drosophila (3a) and parkin- mutants (3b). Although no data is shown for this, the posture of parkin- mutants apparently becomes more abnormal with age. Also, flies apparently exhibit abnormal posture more frequently at later ages, again, no data is given.

Figure 3c: This chart is a behavioral assay designed to test the ability of parkin mutants in flying. The parkrva mutant has a precise excision of the P insertion, called parkEP(3)LA1. These mutants have significantly greater ability to fly than the parkin deletion mutants – park13, park45. The parkZ4437 mutant was found among sterile males tested for parkin abnormalities. This mutant has a point mutation that gives it a premature stop codon. This appears to create a statistically significant less-functional fly than simply deleting the parkin gene. This is curious, but the authors do not offer an explanation as to why this may be the case. This figure effectively demonstrates that a parkin mutation reduces the flying ability of Drosophila.

Figure 3d: This chart demonstrates that flies with a parkin deletion have a decreased climbing ability and also shows that all Drosophila climbing ability decreases as age increases. It would be interesting to see the same data for the parkZ4437 mutant to see if the premature stop codon has the same affect as a deletion. This chart also demonstrates that as a fly gets older, its climbing ability decreases – both for wild-type Drosophila and mutants, so therefore, a mutant fly probably starts adulthood with less ability to fly than a wild type fly.

Figure 3e and d: In order to determine where the Drosophila mutant locomotion difficulty begins, the experimenters used the UAS/GAL4 system to express parkin in specific tissue, such as the mesoderm, which develops into muscles. The graph shows that rescued mutants have more climbing and flight abilities than mutants with no GAL4 rescue, and more than the controls. The controls possess either UAS or GAL, but not both.

Figure 4: This figure is designed to demonstrate the differences between muscle and mitochondrial development in wild-type, parkin-, and transgenic rescue Drosophila. In longitudinal sections (A-C) vacuoles and cellular debris can be clearly seen in parkin- mutants, and vacuoles are seen in rescued mutants, but the rescued sample appears much more similar to the wild-type than the parkin- sample. In samples D-F, you can see differences in myofibrils between the different samples. The important thing to note is that mitochondria are larger than normal and are malformed in the parkin- sample, but are normal in appearance in the rescued sample. This is also true in Figures H and I. In figures J-L, the authors demonstrate the temporal relationship between mitochondrial abnormalities and myofibril deformation. The mitochondria at 96 hours in the parkin- mutant are malformed with less electrons and cristae, but the myofibrils supposedly appear normal. It seems as though the myofibrils in the parkin- sample have striations that aren’t present in the parkin+ sample, but that may just be an artifact of the picture. At 120 hours, the parkin- sample has swollen mitochondria, but still has viable myofibrils. It would be beneficial if the authors had chosen to display the same sample of muscle at 96 and 120 hours, so that comparisons could be made more easily. However, it is clear that muscle deformation begins with mitochondrial deformation, then proceeds to myofibril deformation.

Figure 5: This figure demonstrates the way parkin affects flight muscles. After tagging indirect flight muscles with terminal deoxynucleotidyltransferase-mediated dUTP end labeling (TUNEL) staining, the cells were then observed for the presence of the stain, which would indicate apoptosis had occurred. (The stained nuclei can be observed if apoptosis occurs). In parkin- mutants, dye was observed after 1 day as an adult, and was not observed in parkin+ mutants or larvae. Therefore parkin- mutants undergo apoptosis in the indirect flight muscles.

Figure 6a and b: These figures show that parkin does not affect dopaminergic neurons. Because the authors selected 30-day-old flies to display, it can be assumed that the neurons are not be affected by loss of the parkin gene at any point during development or life because any changes would probably have occurred during the 30 days. This is significant because in humans, a mutation in the parkin gene causes cell death to dopaminergic neurons, so the lack of such an effect means that Drosophila parkin is not a perfect model for human parkin.

Figure 6c and d: These pictures show the dorsomedial cluster of cells in flies. While no cells are killed in this region, the cells are smaller in parkin- mutants than in wild-type Drosophila. Also, the dendrites in parkin- demonstrate less staining than in the wild-type sample. This demonstrates that parkin affects the neurons to some extent – just not to the same extent as in humans. Also, it selectively affects only a select few neurons, which is unusual. More research would have to go into discovering why the parkin mutation only affects certain neurons.

Critique and Future Experiments:
While the authors have possibly found an ortholog for parkin in Drosophila, and have created a Drosophila parkin- mutant, they have not presented convincing evidence that this model is a viable option for furthering research on AR-JP. Their claim that this gene is the homolog of the parkin gene is not completely proven. It would be interesting to examine the percentage of primary and tertiary structure similarity among all ubiquitin-ligase enzymes. It is possible that the Drosophila ubiquitin ligase that the authors have found is not the homolog for human Parkin, but rather another ubiquitin ligase. Until the authors prove that the Drosophila parkin is significantly more similar to human parkin than all other human ubiquitin ligases, their case is unproven. Also, while AR-JP is associated with a mutation in the parkin phenotype, it is defined by its phenotype, which includes selective cell death to dopaminergic neurons. The Drosophila model does not have this phenotype.

Assuming that Drosophila parkin is the homolog for human parkin, the Drosophila parkin mutants offer some possibility for illuminating some parts of the AR-JP molecular mechanism. Since mutations in parkin cause some mitochondrial abnormality in both humans and flies, using the Drosophila mutants to determine the mechanism through which parkin affects mitochondria would be an important next step in this field of research. The ideal next step would be to find the substrate that the Parkin protein tags with ubiquitin. Perhaps once the substrate was known, it could be understood how a failure to tag that particular substrate for degradation would affect mitochondrial activity. The authors state that several potential targets for the Parkin ubiquitin protein ligase have been identified. If these were tagged and then followed to see if they bind to the Parkin protein, the exact substrate for Parkin would be known, and research would be much closer to finding the molecular mechanism by which parkin mutations affect mitochondria. Perhaps this could be accomplished by selectively tagging each of the potential targets then running these targets individually on a column consisting of beads with ubiquitin protein ligase bound to the beads in oder to determine which bound to Parkin and which did not. If the exact target or targets were found, this would assist research in the field greatly.

Although the authors prove that mitochondrial dysfunction precedes muscle degradation in flies, this sequence of events can not be assumed to be true in humans, since there are significant differences in human and fly phenotypes associated with parkin mutations. Before making such assumptions, the authors would need to identify an animal that undergoes selective dopaminergic neuronal degradation when affected with a parkin mutation, and then prove that mitochondrial malformation precedes neuronal degradation in that animal model.

The Drosophila parkin- model will certainly have some application in attempting to solve some of the mysteries surrounding the molecular mechanisms of the human disease, AR-JP, and ultimately PD, but because it fails to have the same phenotype as human parkin mutations, it is not a perfect model. For example, Drosophila cannot be used to discover why parkin mutations selectively affect dopaminergic neurons.


Doherty, Fergus. 1996, August. 3.0 Enzymes of the Ubiquitin Pathway. Nottingham University Biochemistry Webpage <> Accessed: 5 May, 2003.

Green, JC; Whitworth, AJ; Kuo, I; Andrews, LA; Feany, MB; Pallanck, LJ. 2003, April 1. Mitochondrial pathology and apoptotic muscle degeneration in Drosophila parkin mutants. Proceedings of the National Academy of Science. <>





Molecular Biology

Send questions, comments, and suggestions to Jamie Causey.