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

Mitochondrial pathology and apoptotic muscle degeneration in Drosophilia parkin mutants

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

Department of Genome Sicences, University of Washington, Seattle, WA 357730 and

Department of Pathology, Brigham and Women's Hospital and Harvard Medical School, Boston, MA 02115

Review by Holly Smith

Summary and Critique
The authors set out to determine the molecular mechanism responsible for the neurodegenerative disease known as Parkinson’s disease. Previous studies had shown that a form of Parkinson’s known as autosomal recessive juvenile parkisonism (AR-JP) develops when mutations occur in the parkin gene so that the protein that it encodes which is a ubiquitin- protein ligase, is no longer a product. The researchers use Drosophilia models into which they delete various sections of the parkin gene protein- encoding region in order to develop different alleles. By doing this, the researchers hope to determine the molecular mechanism by which the loss of Parkin leads to programmed cell death, or apoptosis. Because past research has shown that a specific toxin called MPP+ causes cell death within neurons, the authors propose that mitochodrial degeneration is the mechanism responsible for this cell death associated with loss of function of the parkin gene. Within the mutants that exhibit a null phenotype for the parkin gene function, a shorter lifespan, male sterility, and decrease in climbing and flying ability are observed.

Figure 1. A. This figure shows the protein sequence similarity between the human parkin protein and the Drosophila parkin protein. The Drosophila protein contains close to the same sequence of amino acids for not only the ubiquitin- like domain, but also the ring- finger and inbetween ring domains and is the only protein on the Drosophila genome to do so. When I look at this figure, I see a good deal of conserved amino acids, although within the N- terminal ubiquitin domain there is a bit of unsimilarity of amino acids that seems questionable. They say that the Drosophila protein and human protein contain 42% amino acid similarity, which is pretty good, but still less than have. However, they do say that overall similarity is 59% which is over half, so this seems sufficient. This figure is successful in proving that the Drosophila protein that they found is the homology of the human parkin gene.

Figure 1. B. The researchers used a Northern Blot to determine the presence of RNA within the wild- type embryos, third- instar larvae, and adults. This RNA was labeled with a parkin- specific probe. Both embryonic RNA and larval RNA seem to be expressed in lower amounts than adult RNA. Though the legend says that the major transcript expressed in embryonic, larval and adult RNA is at 1.7 kb, it seems that it may be a little closer to 1.6 kb. The adult RNA also contains a smear that is higher on the blot than the main binding site, seeming to suggest that some modifications exist within adult RNA that give these longer transcripts. The failure of this figure is that there is no loading control, without it, one cannot compare amounts of RNA expressed within the stages of the Drosophila lifespan.

Figure 1. C. The P element Transposon, park EP(3) LA1 was used to make many mutants of the parkin gene. This was done through the mobilization of park EP(3) LA1 with transposase. The first picture within this figure represents the wild- type parkin gene and shows where the transposon park EP(3) LA1 was inserted, 71 kb upstream from the start codon of the parkin gene. Three of the mutations generated, park 13, park 45 and park 25 are depicted below the diagram of the park wild- type gene. The deletions park13 and park 45 contain none of the protein encoding parkin gene. Park25 contains some of the protein- encoding portion of the parkin. Because both park13 and park45 do not contain any protein-encoding region, they are null phenotypes. Two additional mutants, are park7472, which is missing the first protein encoding sequence that encodes A46 through T and park74437 that is missing the portion of the gene that is encoding R431 through the stop codon. That the mutants park25 will be a null phenotype can be inferred because it is missing the start codon. Whether park Z472 and park Z4437 are null phenotypes will be null phenotypes, I believe is up to a functional test. Both the start and stop codons are present in the former, but the stop codon is missing in the later. This figure is essential in understanding how the mutant alleles were generated and is fairly easy to understand.

Figure 2. The researchers are comparing the phenotypes of wildtype parkin individuals with mutants with null parkin alleles. To compare the two phenotypes, they looked at the presence of the sperm within the seminial vesicle in wildtype and null individuals, the formation of the spermatids, and the formation of the axoneme and Nebenkern structures of the sperm. Within panels A and B that show the presence of the sperm within the seminal vesicle stained with DAPI, the researchers found that there were only mature sperm within the wild type seminal vesicle, and none within the mutant seminal vesicle. They reasoned this because they did not see any long thin nuclei (associated with mature spermatozoa) within the parkin mutants. Although it is clear that there are mature sperm present within the rupured testis in the wildtype, panel A, it seems unclear to me if the arrowhead actually points to mature sperm or not within the seminal vesicle. I would have liked for them to zoom in a little more on this top part next to the arrowhead, to me it seems a little fuzzy. In panel B, the parkin mutant picture seems pretty clear in that the cells pictured are roundish more than long, so I believe them when they say that these are cells of the testis sheath rather than sperm nuclei. Figure C seems pretty clear, they point to the developing spermatids. Although when one compares panels C and D it is obvious that there is a more regualar distribution of spermatid within the cyst within panel C compared to Panel D, I found it difficult to compare the sizes of the different Nebenkern structures within figure D. I think it would have been more helpful, once again, if the authors had zoomed in on this picture. I believe that Panels E and F are successful in demonstrating that within the mutant cells the Neberkern structures are irregular in size and electron- dense matrix is more spread out within the mutant cells compared to the control cells. Although I would have liked the authors to explain what the lines are around the spermatids and the Nebenkern in panels E and F and why they are more squiggly and around the Neberkern in Panel F, these two panels do well to convince me that there must be a mitochodrial dysfunction within the mutants since Nebenkern structures are derived from the mitochondria.

Figure 3. A, A’, B, and B’ show that within the mutant phenotype fly with park 25 (which had a partial excision of coding sequences of the parkin gene), wing posture is more slanted to the ground than control flies that exhibit no genotypic deletion. This suggests that flight maybe impaired although not all of the parkin gene was deleted, thus the park25 deletion results in a null phenotype. C show that the mutants park13, park45, and park Z4437 result in reduced flight ability compaired to the control. Panel D shows that the climbing ability of mutants are always less than the control’s flying ability and that the mutants’ flying ability, as does the control fly’s ability decrease with age. Panel E does show that parkin function is necessary for normal flight because parkrvA or a fly with no deletion of the parkin gene coding region exhibits the highest flight index, while the rescued flies have the second highest flight indexes and the flies that would only have the deletion expressed- park13, Controls 1 and 2 have the lowest flight indexes. Panel F also successfully demonstrates that parkin function is necessary for normal climbing because the three genotypes that have the full parkin gene- parkrvA, rescue1 and rescue2 have the highest flight indexes and the four genotypes that do not have the full parkin gene- park13, control 1, 2, and 3 have the lower climbing indexes. It is curious why Panel E does not have 3 three controls when Panel F does. This third control is necessary in order to show that the UAS- park sequence has the same phenotype as a deletion when GAL4 promoter is not present.

Figure 4. I believe this figure to be one of the most convincing figures within the paper to establishe the link between mitochondrial dysfunction to indirect flight muscle degeneration. This figure shows that within the control flies that have the full parkin gene, the fly’s muscle, myofibril arrangement, and mitochondria are all normal. Within the mutant flies that contain some, but not all of the coding region of the parkin gene, muscle begins to degenerate, and myofibril arrangement becomes irregular. At this same time, mitochondria become swollen and deformed. When mutant cells are rescued with the by expressing the full parkin gene transgenically, normal muscle, myofibril arrangement, and less swollen mitochondria are seen again. To show that mitochondrial dysfunction is not just another symptom of parkinson’s, but also a cause of the disease, the last four panels convincingly show that within parkin mutants, degeneration of mitochondria occurs that does not normally occur within the control. As time goes by, degeneration of the mitochondria occurs before the disruption of the myofibril arrangement, which is shown to happen later on in mutants in panel E. Therefore, this figure convincingly shows that disruption of myofibril arrangement and degeneration of the mitochondria do not happen simultaneously or coincidentally, but the degeneration of the mitochondria happens before myofibril rearrangement.

Figure 5. This figure is a series of pictures taken at various time points within the life of the fly. Cells were stained using TUNEL staining to detect nuclei, and thus the occurrence of apoptosis. The first time that apoptosis occurs is within the 1 day adult. This figure is especially important when combined with the information that we gained from figure 4. It shows that the apoptosis of flight muscle that occurs in parkin mutants is a process that occurs later on within the mutant flies. Because mitochondrial degeneration within the flight muscle is a gradual process that begins within the pupal stages before apoptosis occurs, it can be gathered that mitochondrial degeneration is one of the factors that cause cell apoptosis later on in the fly’s life.

Figure 6. Figure 6 is not one of the reseachers strong figures. It shows that the alleles that are used to develop the mitochondrial deletions do not produce neuronal degeneration or dopaminergic neuron loss. In a perfect experiment, the deletions that have previously have shown to be the cause for a form or Parkinson’s disease, AR- JP, would cause not only muscle and mitochondrial degeneration within other tissues like the indirect flight muscle and within the germ line, but would also cause degeneration to the tissue in which degeneration usually characterizes the disease. The first two panels do not show dopaminergic neuron loss within the mutant, although dopaminergic nuron loss is what Parkinson’s is characterized by. However, there does seem to be some hope in relating the work that the researchers did back to the characterization of Parkinson’s disease because although there is no loss of the dopamiergic neurons, there is shrinkage of the cell body and decreased staining within the proximal dendrite. I would have like to see the same thing in this figure as what they showed in figure 4. I think it would have made the paper if they had shown two additional panels, one of a mitochondria within the dopaminergic neuron for a control fly and one of a mitochondria within the dopaminergic neuron for a mutant fly. If these figures had these two additional panels and the mutant mitochondria appears irregular and less electron dense than the control, this would strongly support the fact that the molecular mechanisms for the germ cells, indirect flight muscle and the neurons are all the same. The authors say in the discussion that “the underlying mechanisms responsible for pathology in these different tissues may be highly conserved” (Greene et al, 2003), but they do not take the extra step in checking to see by showing the mitochondria. They do not even state that they have tried to do this.

Overall, I believe that the authors are successful in establishing that mitochondrial dysfunction is a major causative factor in Parkinson’s disease through their evidence of irregularity within the mitochondrial derived Nebenkern structures in the germ cells and the correlation of mitochondrial degradation with myofibril degradation. The fact that mitochondrial degradation occurs before myofibril degradation is convincing in supporting that mitochondrial degradation is a causative factor of Parkinson’s disease. However, where the authors fail is in correlating the mitochondrial dysfunction to neuronal degradation. A concern when evaluating this paper, however, like any paper is the issue of what evidence that the authors choose to present to us.Within each one of these figures 2, 3, 4, 5, and 6, only one of the mutants out of the 5 that where generated are shown as results with no mention as to what the results looked like for the other four. For example, in figure 2 there are irregular Nebenkern structures for the mutant generated by park 13. However, we do not know whether this was a rarity and that the other four mutants did not show irregularities in their Nebenkern structures. As always, the authors present only their best evidence.

When it comes to the question as to whether the mitochondrial dysfunction is necessary and sufficient as a cause for Parkinson’s, I do not believe that the authors were successful in proving either. Perhaps if their last figure had shown mitochondrial dysfunction in relation to neuronal degradation, it could have shown that it was necessary. However, in order to say than mitochondrial dysfuntion is sufficient, it is much more difficult. I believe that other factors must also be considered, like what other proteins that the parkin protein degrades and what that protein does if it is not degraded.
Future Research

As the researchers state, there needs to be further work done on the relevance of mitochondrial pathology to the spermatid individualization defect. Because spermatogenesis seems to occur normally within the mutants until the individualization stage, multiple cell germ line cysts should be isolated within wild type and mutant Drosophilia. Perhaps a test can be done for the presence of toxins that are like the mitochondrial toxin MPP+ that is found to induce cell death within the dopaminergic neurons by preventing the formation of the mitochondrial complex I. A test that could be done to determine the presence of a mitochodrial toxin that is related to MPP+ could be done through immunoflourescence. One could make an antibody through by injecting MPP+ into a rabbit and then labeling this antibody with GFP. This antibody can be added to the Drosophila germ line cells under which the individualization stage is taking place. If labeling can be seen within the mutant cells and not the wildtype, then it can be concluded that mitochondrial dysfunction is the mechanism by which the individualization defect is occurring.

Because little is known about how exactly the loss of parkin function triggers mitochondrial changes, investigations should be made into this. The parkin protein is a ubiquitin ligase protein, which means that parkin has the power to poly- ubiquitinate another protein (or mark it for degradation) and the parkin protein thus keeps the protein that it acts on in check. The pathway that leads to mitochondrial changes can be investigated by investigating what proteins the parkin protein, a ubiquitin- protein ligase acts on. This can be done in either two ways. Because the parkin protein most likely binds to the protein that it marks for degradation, the first way to isolate this protein would be through immunoprecipitation. You can generate an antibody for the parkin protein, since the researchers already know its amino acid sequence and bind this antibody to beads that can be used in immunoprecipition. Then you can electrophorece the isolated parkin protein on a gel and whatever comes out is the protein that the parkin protein degrades. However, this could lead to some confusion because the parkin protein is only active once it binds to the binds to the E2 ubiquitin-conjugating human enzyme 8 (UbcH8) at its C-terminal ring-finger. Therefore, steps must be made to avoid mistaking UbcH8 for the protein that parkin marks for degradation. Another way that the substrate for the parkin protein can be isolated is by maybe making an antibody to ubiquitin. Thus the ubiquitin marked protein can also be isolated through immunoprecipitation. Once the protein that parkin marks for degradation is known, what happens when there is a loss of function within the parkin gene will be understood much more and how this relates to the trigger of apoptosis and mitochondrial degradation will become more clear.

Although there is no animal model that exhibits dopaminergic neuron loss as a result of Parkinson’s yet, an additional test could be to test the relevance of the substrate of the parkin gene to the tissues that were affected within Drosophila by Parkinson’s, being the germ line and the indirect flight muscle. The presence of the substrate can also be identified within the affected cells by labeling the substrate with a radioactive antibody. An over abundance of this substrate within the mutants compared to the controls would suggest that this substrate has some effect on the cells that cause harm to them.


Greene JC, Whitworth AJ, Kuo I, Andrews LA, Feany LA, Pallanck LJ. 2003. Mitochondrial pathology and apoptotic muscle degeneration in Drosophila parkin mutants.PNAS 100(7):4078-4083.

Mosesson Y, Shtiegman K, Katz M, Zwang Y, Vereb G, Szollosi J, Yarden Y. 2003. Endocytsosis of receptor tyrosine kinases is driven by mono-, not poly-ubiquitylation. Journal of Biological Chemistry. [epub ahead of print]


Zhang Y, Gao J, Chung KK, Huang H, Dawson VL, Dawson TM. 2000 Nov 21. Parkin functions as an E2-dependent ubiquitin- protein ligase and promotes the degradation of the synaptic vesicle-associated protein, CDCrel-1.
Proc Natl Acad Sci 97(24):13354-9.


Molecular Biology Main Page

Course Materials

Biology Main Page


Questions? E-mail Holly Smith at