This web page was produced as an assignment for an undergraduate course at Davidson College.
Mitochondrial pathology and apoptotic muscle degeneration in Drosophila parkin mutants
Jessica C. Greene
Alexander J. Whitworth
Laurie A. Andrews
Mel B. Feany
Leo J. Pallanck
PNAS 100(7): 4078-4083
This paper focused on the prospect that mitochondrial dysfuncation, stemming
from loss of function mutations in the parkin gene, is a major causitive factor
in Parkinson’s Disease (PD). Previous research indicated that loss of
function mutations in the parkin gene, which encodes a ubiquitin-protein ligase,
resulted in a familial form of PD known as autosomal recessive juvinile parkinsonism
(AR-JP). In order to investigate their hypothesis, they created a Drosophila
model of this particular disorder. Using the Berkeley Drosophila Genome Project
Database, the investigators identified a Drosophila parkin ortholog, which had
a 42% amino acid identity with the human parkin gene. The human parkin gene
is characterized by an N-terminal ubiquitin-like domain, two carboxy-terminal
ring finger motifs, and an inbetween ring-finger domain. The ortholog identified
in the Drosophila gemone is the only gene in the Drosophila genome that has
all three of these characteristics. The investigators proceeded to generate
loss of function mutations in the Drosophila parkin gene with the use of a P
element that maps close to Parkin. A large collection of deletion alleles of
the parkin gene were generated including alleles that removed all of the parkin
Through behavioral assays and histological analysis, the investigators found that flies bearing null alleles in trans to the Df(3L)Pc-MK deletion chromosome, which removes the parkin gene, are viable throught the adult stage of development but exhibit a slight developmental delay, reduced longevity, defects in flight and climbing ability, and while females are fertile, male flies expressing a parkin null allele are sterile. The authors claim that loss of parkin function results in a spermatid individualization defect in the male germ line, and the locomoter defects result from apoptic muscle degeneration, as indicated by mitochondrial pathology.
Parkinson’s disease in humans in characterized by loss of dopaminergic neurons in the substantia nigra, however, upon analyzing the development of the major brain centers in the Drosophila parkin mutants, no neuronal loss was observed in any of the cell groups. One of the major differences between Drosophila parkin mutants and AR-JP are that different tissues are affected by loss of parkin function. This paper suggests, however, that the underlying molecular mechanisms that are responsible for the pathology may be highly conserved between the two species.
A: This figure displays the amino acid sequences of Drosophila Parkin (D-Park) and human Parkin (H-Park). The domains discussed earlier, the N-terminal ubiquitin-like domain, ring-finger domains, and in-between domains are all highlighted. There are a few things that the figure does not clearly state that are important for clear understanding of this figure. In between the amino acid sequences of H-Park and D-Park is either a + sign, a letter which codes for a particular amino acid, or a blank space. A plus sign indicates that although H-Park and D-Park do not have identical amino acids at that position, a conservative substitution has been made. If a letter appears, that indicates that both H-Park and D-Park have the same amino acid at that position. A blank space simply means that while the amino acids are not identical at the position, neither a conservative nor nonconservative substitution has been made. There are also dashed lines present in both the H-Park and D-Park sequences. This was done simply to maximize the alignment between the two sequences.
B: Figure 1B shows a Northern Blot which uses a parkin-specific probe. In the first lane, labeled E, containing poly (A)+ RNA from wild type embryos, a single band is seen which is approximately 1.6kb in length. It is interesting to note that in the next lane, labeled L, containing third instar larvae, the single band that is observed appears to be slightly larger in molecular weight, and is approximately 1.65kb in length. Finally in the lane labeled A, containing poly(A)+ RNA from adult flies, there are two bands visible, meaning that there are two forms of the parkin transcript that are present in adults. One band is approximately 1.7kb in length and appears to be fainter than a band which is approximately 1.6kb in length. The 1.7kb band was referred to as the minor transcript and the 1.6kb band was referred to as the major transcript. The fact that there are two bands visible when a parkin-specific probe is used indicates that some sort of transcript modification of the parkin gene takes place. This is further supported by the fact that the transcripts in lanes E and L are of slightly different molecular weight. Thus, throughout development, the role and/or expression of the parkin gene may change, as indicated by modification of the transcript. Also, it would have been good to see a standard that appeared in all lanes so that intensity of the bands could have been analyzed accordingly. There is no way to be sure that the same amount of RNA was loaded into all lanes unless a standard is present.
C: Figure 1 C is a molecular map of the parkin transcript which shows the parkEP(3)LA1 insertion. This insertion, which appears 71 bp up from the parkin start codon, is a P element that is used to generate disruption of the parkin gene, and thus the creation of mutant parkin alleles. The insertion, mobilized using transposase, yielded 3 different deletion alleles of parkin, designated park13, park45, and park25. On the parkin transcript, the black boxes identify the parkin coding sequences and both park13 and park45 are deletion mutations that eliminate all of the parkin coding sequence, thus making them null alleles of parkin. park25 maintains a portion of the coding sequence. This work also yielded a precise excision of the P element, parkEP(3)LA1, which was used throughout the expreriment as a control, designated parkrva.
Male parkin mutants are sterile and this figure suggests that males undergo a spermatid individualization defect due to the presence of abnormal mitochondrial derivatives. Panels A, C and E are parkin+ while panels B, D and F and are parkin-. This figure was generated by electron microscopy. The arrow in panel A, which shows individual sperm released from a ruptured testis is clear, however, the arrowhead, which is intended to display the long thin nuclei that are located inside the seminal vesicle, is difficult to distinguish when compared to the round nuclei which correspond to cells in the testes sheath. Panel B clearly shows that mature sperm is absent from the seminal vesicle in parkin- mutants, however, the image does indicate that mature spermatids do form, but fail to individualize and remain as a syncytial bundle. Panels C,D, E, and F clearly show that parkin- mutants show well developed axonemes in developing spermatids, however, the Nebenkern are abnormally distributed and diffuse in the region that is highly electon dense in parkin+. This is particularly significant because the Nebenkern is a mitochondrial derivative, indicating that mitochondrial disfunction could be at the root of male sterility. It is also important to note the genotypes of parkin+ and parkin-. Parkin+ is parkrva/Df(3L)Pc-MK, and parkin- is park13/Df(3L)Pc-MK. Df(3L)Pc-MK is a chromosome which lacks the parkin gene, parkrva contains all of the parkin gene and lacks the P element parkEP(3)LA1, and park13 is a mutant parkin allele that lacks all of the parkin coding sequence. It would also be interesting to see data generated using park25/Df(3L)Pc-MK and park45/Df(3L)Pc-MK simply to ensure that all alleles containing loss of parkin function mutations yielded the same result.
Figure 3 illustrates that parkin function is required for normal wing posture, flight, and locomotion.
A and B: Panels A and A’ clearly show that control flies, which have a functional parkin gene exhibit normal wing posture. Panels B and B’, however, both show a fly with a downturned wing posture. The genotype of the fly in panels B and B’ is park25/Df(3L)Pc-MK. Park25, unlike park13, does not lack all of the parkin coding sequence, but rather just a portion of the coding sequence was deleted. This indicates that the portion of the parkin coding sequence that remains in park25 is not sufficient for expression of the wild-type wing phenotype.
C and D: In figures C and D it is clear from the data that parkin mutants exhibit impaired flight and climbing ability. It was unclear, however, how the last mutant allele in figure C was generated. The mutant allele reads parkz4437 and no where in the paper is this mutant allele discussed. It is also interesting to note that this mutant allele results in the most impaired flight ability of any muant allele. park13 and park45 mutants exhibit a similar flight index, which is consistent with the fact that both are null alleles of parkin. It would have been nice to see park25 mutants tested for flight index, since it was clear from panels B and B’ in this figure that they exhibited the downturned wing posture. It is unclear why all mutants were not tested in each assay. Also, upon analyzation of the data, it is reported by the authors that expression of the null alleles of parkin result in a downward wing posture phenotype which the mutant in the figure, park25, is not a null allele of parkin. Also, it is unclear why park13 was the only mutant tested in the climbing index assay. Showing all mutants in this graph would have given the claims more validity.
E and F: Figures E and F seek to demonstrate that the mutant phenotypes can be at least partically rescued through the use of the UAS/GAL4 system to express parkin in defined tissues, in this case, the mesoderm. There were two GAL4 lines that were found to rescue wing posture, flight index, and climbing index of parkin mutants to near wild type levels. There were three controls generated, however, only controls 1 and 2 appear in graph E while all three appear in graph F. The reason for this inconsistency is not clear. Also, not all parkin mutant alleles were tested in this assay. Only park13 was tested and the results would have been more conclusive if all parkin mutant alleles had been evaluated using the UAS-GAL4 rescue system in the mesoderm.
Despite the inconsistencies mentioned, these figures do provide evidence for the claims made by the authors: Drosophila parkin null mutants exhibit locomoter defects. It is also evident from panel D that Drosophila parkin null mutants begin adult life with a reduced climbing ability.
Figure 4 intends to display that the muscle degeneration present in parkin mutants as a result of a mitochondrial dysfucntion.
A,B and C: Panels A, B and C in this figure show longitudinal sections of indirect flight muscles (IFMs) from parkin+ flies (A, parkrva), parkin- flies (B, park25) and flies that have undergone transgenic rescue and are expressing parkin ectopically in the muscles.(C). Hematoxylin and eosin staining reveals that severe disruption of muscles integrity is present in Panel B, indicating that loss of parkin has a direct effect on the musculature. Panel C indicates that the musculature can be restored upon transgenic rescue. The results of these three panels are particularly convincing, however, it is also important to keep in mind that the tissue sectioning may have been inconsistent, and Panel B could have been simply a bad section rather than a defect in the musculature. Assuming that sectioning was done carefully, however, these panels indicate that the IFMs in parkin- show severe disruption of muscle integrity.
It is stated within the paper that the data indicate that not all muscles in the Drosophila are affected by loss of parkin function. While this data is not shown, it would have been nice to see a histological analysis of a muscle in a parkin- fly that was not affected for purposes of comparison.
D,G,E, H, F, I: When comparing parkin+ and parkin- adult sections through IFMs it is clear that the parkin- genotype results in an irregular myofibrillar arrangement and that the mitochondria are indeed swollen and malformed, even showing disintegration of the cristae. It also appears to be the case that transgenic rescue of mutant parkin alleles can restore the wild type phenotype in these areas.
J, K, L, M, N, O: It is evident that the mitochondria of 96-h pupae of parkin
mutants have less electron dense mitochondria than wild type 96-h pupae. By
120-h the parkin- pupae still maintain intact myofibril structure but the mitochondria
are grossly swollen.
It is also important to note, however, that the density of the myofibrils in parkin mutatns was variable, indicating that loss of parkin function cannot be directly correlated to a particular phenotype in myofibrils. However, mitochondrial disruption was characteristic of all IFM in parkin mutants, indicating that loss of parkin function does affect the mitochondria.
It should also be mentioned that park25 was the only mutant allele tested in this assay. As in the previous figure, the claims made by the authors would have been more substantiated if all mutant alleles that had been generated would have been assayed.
In this figure that authors attempt to identify the mechanism through which muscle degeneration proceeds. It was hypothesized that muscle degeneration proceeded through an apoptotic mechanism. In this assay, TUNEL staining was used, which is a common method for staining apoptotic nuclei. From the figure, it is clear that the only TUNEL positive panel was panel D, which depicted the IFM of a 1 day old parkin- fly. At 96 and 120 hours after puparium formation, no TUNEL staining was detected in parkin- flies, indicating that cell death does not occur immediately, but is the ultimate result. While this figure clearly shows that muscle degeneration proceeds through an apoptotic mechanism, it is not entirely clear that this is a result of the mitochondrial defect that was indicated earlier. Also, it is interesting to note that the parkin- genotype in this figure does not match the parkin- genotype in the other figures. In this case, the parkin- genotype is park25/park25, whereas in the previous figures, the parkin mutant allele had always been transheterozygous with the the Df(3L)Pc-MK deletion chromosome. It would have been more consistent if the authors had kept the genotypes the same, and tested all mutant alleles. Also, just as another control, the 96 and 120h APF parkin+ flies could have been shown in this figure.
Since it is known that loss of dopaminergic neurons in the substantia nigra are detected in human PD, it was important to analyze the effects of loss of parkin function on the brain in the Drosophila model. Brain sections were prepared from flies at 1, 10, and 30 days of age. Histologic analysis revealed appropriate development of the major brain centers andvreveaedl appropriate organization of the nervous system including well-formed ellipsoid body (eb), peduncle (p), subesophageal ganglia (sog), medulla (m), and lamina (l) (Panels A, parkin+ and panel B, parkin-). Upon immunostaining for tyrosine hydroxylase, no clear neuronal loss was observed in dopaminergic neurons of dorsomedial, dorsolateral, anteromedial clusters or medualla in any of the cell groups athough shrinkage of the cell body of the dorsomedical dopaminergic cell cluster (arrow) and decreased staining in the proximal dendrite (arrowhead) in panel D, parkin-, is evident when compared to Panel C, parkin+.
Since loss of function mutations in the parkin gene have little effect on the brain of parkin- Drosophila, there will need to be much more research to link the observations made in the Drosophila mutants to individuals with AR-JP.
There is sufficient evidence to suggest that loss of parkin function in Drosphila results in a mitochondrial defects in IFM and male germ line tissues. The obvious difference between Drosophila parkin mutants and AR-JP concerns the tissues that are affected by a loss of parkin function. Dopaminergic neurons are the primary tissues affected in AR-JP individuals, which is not consistent with what was observed in Drosophila. Nevertheless, as the authors indicate, the molecular mechanisms responsible for the pathology may be highly conserved.
One way to investigate this possibility would be to create a model in a species that is more closely related to the human, such as a mouse model. After locating the mouse parkin ortholog and generating mutant parkin alleles, mitochondrial pathology of tissue sections from the mouse brain could be analyzed to determine if a mitochondrial defect was indeed present.
There is also work needed to determine the mechanism by which loss of parkin function triggers the observed mitochondrial pathology and ultimately, cell death. It is known that parkin encodes a ubiquitin-protein ligase, which often function in concert with Ubiquitin-conjugating enzymes. Once the protein ubiquitin is activated, Ub-conjugating enzymes and Ub-ligases work together to ubiquitate a protein, which is then shuttled to a proteasome for degradation. (See ubiquitin-proteasome pathway) A mutation in a ubiquitin-protein ligase, such as parkin, may alter this pathway. Since mutations in parkin appear to result in mitochondrial defects, the role of the ubiquitin pathway in mitochondrial development should be investigated. Perhaps proteins that should be targeted for degradation are not in parkin mutants. It is also possible that a protein which should not be targeted for degradation is targeted as a result of a parkin mutation. In order to determine if parkin plays a role in mitochondrial defects, it would first be interesting to see if parkin is present in the mitochondria. One could attach the coding sequence for green fluorescent protein upstream of the parkin coding sequence and use a fluorograph to see where parkin is expressed. If parkin is present in the mitochondria, the next step would be to identify the protein in the mitochondria that parkin interacts with. One could clone the gene for a protein that interacts with parkin using the yeast two-hybrid system. Once this had been done several steps could be taken to see if this protein was indeed degraded as a result of a loss of parkin function mutation. For example, a probe could be generated for the protein that parkin interacts with and radioactively labeled. One could run a Western blot for both parkin+ and parkin- individuals and expose the blot to the probe which binds to the protein that interacts with parkin. If there was a molecular weight difference, one could speculate that this protein had been ubiquinated, or targeted for degradation. Running another western blot using a probe for ubiquitin could confirm this hypothesis. My guess is that the loss of function mutation in parkin results in the ubiquination of a protein that is vital for mitochondrial development, and thus, its subsequent degradation.
There are several papers which have previously made the link between ubiquination and sperm mitochondria which is particularly relevant for this paper, since a mitochondrial defect was observed in male germ line pathology.
Also, recent research has indicated that Parkin may promote the degradation of substrates localized in mitochondria and involved in the late mitochondrial phase of ceramide-mediated cell death. Loss of this function may underlie the degeneration of nigral dopaminergic neurons in patients with Parkin mutations.
Darios, F. et. al. 2003. Parkin prevents mitochondrial swelling and cytochrome c release in mitochondria-dependent cell death. Hum Mol Genet. 12(5):517-26.
Greene, Jessica C. et. al. 2003. Mitochondrial pathology and apoptotic muscle degeneration in Drosophila parkin mutants. PNAS. 100(7): 4078-4083.
Sutovsky, P. 2003. Ubiquitin-dependent proteolysis in mammalian spermatogenesis, fertilization, and sperm quality control: Killing three birds with one stone. Microsc Res Tech. 61(1): 88-102.
The Ubiqutin System. <http://homepages.bw.edu/~mbumbuli/cell/ublec/> Accessed 2003 May 1.
Davidson College Biology Department
Davidson College Molecular Biology
Questions or comments contact Erin Cobain