Paper Review
"This web page was produced as an assignment for an undergraduate corse at Davidson College."
PEX11 Promotes Peroxisome Division Independently of Peroxisome Metabolism
By Xiaoling Li and Stephen Gould
From the Journal of Cell Biology, Volume 156, Number 4, February 18, 2002; 643-651
Department of Biological Chemistry, The Johns Hopkins University School of Medicine, Baltimore, MD 21205
Summary and Critique
By Sarah Henry
Abstract
It has previously been proposed that PEX11 proteins have a direct role in peroxisomal fatty acid oxidation and affect peroxisome abundance indirectly. Li and Gould attempt to prove in this paper that PEX11 proteins act directly in peroxisome division and are indirectly involved in metabolism.
Introduction
The paper begins with an introduction to the role of peroxisomes in the cell. This is helpful since many people reading the paper might not know much about peroxisomes and understanding their role in metabolism and division is vital to understanding this paper. Peroxisomes metabolize lipids and toxins using H2O2. Although many people don't realize it, peroxisomes are very important in protecting the cell from damage by peroxides and O2 radicals and defects in them can lead to lethal diseases.
Due to the existence of the cytoplasm-to-peroxisome protein import pathway, it has been proposed that peroxisomes grow by the uptake of proteins, which is followed by division. The mechanism of this division is poorly understood. It has previously been shown that defects in peroxisomal fatty acid oxidation reduce peroxisomal abundance, indicating a possible role of metabolic control. The influence of PEX11 proteins on abundance has also been investigated in yeast, humans, rodents, and protozoan and the conclusion that Li and Gould propose is that PEX11 has a direct role in peroxisome division. There have been various studies regarding the role of PEX11 proteins in metabolism and therefore, it is unknown whether the role of PEX11 is a secondary, indirect consequence of peroxisome division or whether it is the primary cause of peroxisome division.
This paper displays the results of experiments performed to test the relationship between PEX11 proteins and peroxisome division and to also understand the role of peroxisome metabolism. The authors attempt to demonstrate that PEX11 proteins cause peroxisome division in the absence of metabolism and that reducing the amount of PEX11b protein causes decreased peroxisomal abundance in the absence of peroxisomal metabolism. Li and Gould propose that PEX11 proteins play a direct role in peroxisome division and that the loss of PEX11 proteins indirectly inhibits peroxisome metabolism.
The introduction is well written and provides appropriate introductory information regarding peroxisomes. More information about oxidation, especially the difference between long chain fatty acid oxidation and medium chain fatty acids (MCFAs) might have been provided, but the information was not absolutely necessary.
RESULTS
The Peroxisome-Proliferating Activity of Human PEX11ß
There are two forms of PEX11 (alpha and beta), which are both integral peroxisomal membrane proteins (PMPs).
Figure 1-PEX11ß-mediated Peroxisomal Proliferation is a multistep process In the first experiment, wild type human fibroblasts were injected with PEX11bmyc on a plasmid. Then one set of cells was labeled with anti-myc antibodies and another set were labeled with antibodies against PEX14, an endogenously expressed peroxisomal membrane protein (PMP). Fluorographs of the cells at 1.5-2hr., 4-8hr., and 24-48hr. were displayed in order to show the three kinetic steps of the process. It is easy to see from the fluorographs that as time progressed, the peroxisome number increased in the cells treated with the anti-myc antibody. The control cells didn't divide as rapidly as the cells injected with PEX11b and labeled with the myc antibody. The cells look "normal" at the initial time point, they are elongated at the second time point, and their number has greatly increased by the third time point. Figure 1 shows the general increase in peroxisome division due to the addition of PEX11b protein, but the data is only qualitative.
Li and Gould do a good job of using this basic first experiment to begin the argument that PEX11b proteins are sufficient for peroxisome division. The insets are a helpful addition to the fluorographs since the cells are difficult to see. The only problem that I have with this experiment is that they never showed any data or stated that they tested their experimental design. How do we know that their plasmid was functional and that it was uptaken by all the cells? And therefore, can the authors truly claim that PEX11b is causing the increased proliferation in the cells? It would have been nice to have seen some information on the way they screened their cells to ensure that they had cells that had the plasmid. What was the selectable marker on their plasmid? Did they perform a functional test to ensure that the myc tag wasn't disrupting the conformation of the PEX11b protein?
Figure 2-Overexpression of PEX11b Increases Peroxisome Abundance in Wild-type Human Skin Fibroblasts
Normal fibroblasts cells were transfected with either PEX11bmyc expression vectors or another PMP protein involved in oxidation of fatty acids, PMP34myc, and were treated with either anti-myc antibody or PEX14 antibody. The amount of peroxisome division was determined by counting the number of peroxisomes in the widest region of the cell. Untransfected cells served as a control in this experiment. The level of peroxisome abundance was similar in untransfected cells and in cells treated with PMP34myc. However, the cells treated with PEX11b had a 1,000% increase in the number of peroxisome cells over the controls. Example confocal fluorescence images were shown for both the PEX11myc injected cells and the PMP34myc cells. A graph representing data from 30 cells/sample was presented to quantify the data. The authors also state that they performed experiments where they over expressed other PMP proteins and that these experiments resulted in no increase in peroxisome abundance, but it would have been nice to have seen some of this data. Therefore, Li and Gould conclude that PEX11b proteins have a specific ability to increase peroxisome division that is not found when other PMP proteins are over expressed.
From the data provided, this seems to be a reasonable conclusion to make and this experiment was a good second step following the first experiment. This was an effective experiment and the results support the claim that the increase in peroxisome division is not just due to PMP properties, but rather, is specific to PEX11b.
Figure 3-The Peroxisome-Proliferating Activity of Human PEX11b is Independent of Peroxisome Metabolism
It was previously proposed that PEX11 proteins are directly involved in fatty acid oxidation and thus indirectly effect peroxisome division. This hypothesis was tested in this paper by using cells lacking all peroxisomal metabolic functions, but containing peroxisomes. The homozygous human mutant cell line, PBD005, that is defective in many metabolic activities, was transfected with PEX11bmyc or PMP34myc plasmids and treated with the either anti-myc antibody or anti-PEX14 antibody respectively. Figures also show example confocal microscopy results similar to the last figure and has a graph providing quantitative data. Even when the mutant cells lacked metabolic activity, the untransfected and PMP34myc cells had similar low levels of abundance while the cells injected with PEX11myc had thirty times as much abundance as the control cells. This experiment suggests that PEX11b stimulates the division of peroxisomes independently of metabolic function.
Again, this is a reasonable conclusion to draw from the data provided. It is difficult to see specific cells in their fluorographs, but I think they were showing them for consistency with the other figures. I was impressed that the authors explained why the control levels were lower here than in figure 2. They explained that the cells that they used in this experiment had defects in peroxisomal oxidation and therefore, this decreased the baseline amount of peroxisome abundance.
Figure 4-Fatty Acid b-oxidation is Not Required for PEX11-mediated Peroxisome Proliferation in Yeast
The next experiment was designed to test whether or not yeast PEX11p could cause peroxisome division without metabolites flowing through the peroxisomal b-oxidation pathway. First, the number of peroxisomes in the laboratory yeast strain was determined by importing green fluorescent protein (GFP) into the peroxisome using transformation of a plasmid. It has previously been determined that peroxisomes are the only site of fatty acid b-oxidation in S. cerevisiae as well as that cells grown on oleic acid (OLA) have a higher number of peroxisomes than those grown on glucose. The results of this experiment are depicted on graphs plotting the number of peroxisomes/cell versus the cell frequency. One hundred and twenty cells were observed under fluorescence microscopy for each sample. The first two figures (A&B) show the difference in the number of peroxisomes present between cells grown in glucose versus those grown in oleic acid and the results coincide with past experiments. Again, an adequate explanation is given for the unexpected variability in peroxisome abundance. The cell line used is competent for degradation while other previously studied cells were not.
Then the authors created a strain called XLY1 by deleting the chromosomal copy of the PEX11 gene and introducing GFP/PTS1, a plasmid with a GAL1 promoter that expresses Ypr128Cp, PEX13, and PEX11. The strains were grown in medium lacking fatty acids with either glucose or galactose. The number of peroxisomes increased when the cells were moved from glucose to galactose (figures C&D). As observed in a previous experiment, the number of peroxisomes did not increase when PMP proteins other than PEX11 were over expressed (figure E). PEX13 was used here and is a PMP protein involved in peroxisome matrix protein import. Ypr128Cp was also used in this experiment and is the "adenine nucleotide carrier of the peroxisome membrane" (figure F). In this experiment, when PEX11 was expressed, the number of peroxisomes increased even in the absence of metabolic substrates (figure G). This indicates that PEX11 stimulates the division of peroxisomes through a mechanism totally independently of any metabolic process.
To further test the previous findings, the authors made another strain called XLY2, which is a pox1 derivative of XLY1. Pox1 encodes an enzyme that is essential for fatty acid oxidation. As seen in the previous experiment with the XLY1 cells, no increase in peroxisome abundance was observed when PEX13 (figure H), another PMP protein was expressed, but the peroxisome number did increase when PEX11 was expressed (figure I).
This is a very well controlled experiment, but they might have explained the data a little better. They state that PEX11 causes increased peroxisomes in XLY2 the same as in XLY1, but the graphs look somewhat different, although the patterns are similar. I'm also not quite sure why they switched to yeast in this experiment after doing the first three experiments with human cells. I think that they switched because mammalian peroxisomes don't oxidize MCFAs and because they wanted to test their hypothesis in two organisms with different systems, but they could have made the reason for the switch more clear.
Figure 5-Loss of Mouse PEX11b has an Indirect Effect on Peroxisome Metabolism
The data gathered previously in this paper suggests two possible hypotheses for the function of PEX11 proteins. Either, PEX11 proteins have two independent functions; one in peroxisome division and one in fatty acid oxidation, or PEX11 proteins have an indirect effect on metabolism in addition to their primary effect on division. The authors created mice lacking the PEX11b gene and performed an experiment observing the effect of this deletion on peroxisome abundance. They hypothesized that if PEX11b functions mostly in fatty acid oxidation, then peroxisome abundance should be the same in cells with and without the PEX11b gene when grown in medium lacking the substrates needed for the oxidation pathway. But, if PEX11b functions primarily in division, then peroxisome abundance should be decreased in the cells lacking the PEX11 gene.
Immunoflourescence was performed similarly as before, treating the cells with either anti-PEX14 or anti-catalase antibody and counting the number of labeled cells. Catalase is an enzyme used in metabolic processes to convert H2O2 to water and oxygen. Li and Gould state that the data shows that the cells lacking the PEX11 gene had about half the number of peroxisomes that the wild-type cells had and state that this affirms the hypothesis that the PEX11b affects peroxisome abundance independently of metabolism.
I agree with this conclusion based on the previous experiments, but the data from this experiment isn't as impressive. The experimental design is good, but the error bars on the data are huge and cause the data to almost overlap while they state that the PEX11b-/- cells show half as much division as the PEX11b+/+ cells. I'm not sure what the purpose of the two fluorographs was since I can't see any difference in the two sets. I'm also not sure why they switched to the anti-catalase enzyme. The researchers again switched organisms in this experiment, using mice, but I think they chose this organism because most gene deletions are done in mice.
Discussion and Future Experiments
PEX11 proteins were identified in 1995 as yeast PMPs that promote peroxisome division when over expressed and cause a decrease in peroxisome abundance when PEX11 is absent, but the mechanism by which PEX11 functions is still unknown even after this paper in 2002. In 2000, van Roermund et. al. proposed that PEX11 proteins act primarily in medium chain fatty acid (MCFA) oxidation and affect peroxisome division indirectly by affecting metabolism. It is interesting to me that two lab groups can have almost opposite results, but I guess if certain experiments aren't performed, then relationships with other variables might not be noticed.
Li and Gould decided to reinvestigate the role of PEX11 in peroxisome division and metabolism due to some inconsistencies they noticed between previous data and hypotheses. Researchers had seen a decrease in peroxisome abundance with mutations in long chain fatty acid metabolism and oxidation, rather than in MCFA metabolism, so they wanted to know how these two processes were related; mammalian peroxisomes don't oxidize MCFA's, but they still have PEX11 proteins; and PEX11 appears to act as a transporter of MCFAs or ATP although it's not similar to transporters.
I think Li and Gould did a good job of proving that PEX11b was sufficient to promote peroxisome division in the absence of other stimuli to divide, but since they never addressed the role of PEX11a, they can't conclude that PEX11 is sufficient. Also, although they provided data for PMP34, PEX13, and Ypr128Cp they didn't provide any data for the other PMP's that they tested. I think that a few more examples of other PMPs over expressed, but not increasing peroxisome division would have been more convincing. I think their best data is that supporting the claim that the effect of PEX11b in increasing peroxisome division is independent of peroxisome metabolism. They counted many cells and the difference between the controls and the experiments was large. The test in which they deleted the PEX11 gene from the mouse was not as convincing because the error bars were so large that the data could be the same for both situations. This paper is well organized and presents the data in a logical sequence. Putting all the data together, it is easy to believe their claim that the PEX11b proteins promote peroxisome division independently of peroxisome metabolism.
Although Li and Gould proved to me that PEX11b was sufficient to promote peroxisome division, I think that they need to perform some more experiments to prove that PEX11b is necessary for peroxisome division. They did perform one experiment where PEX11b was deleted, but the results of this experiment weren't conclusive to prove that PEX11b is a necessary protein for peroxisome division. I think they should before similar experiments where they delete PEX11b from the cells in mice again, but also in human and yeast cells.
The authors also do a good job addressing possible interpretations of their results. Although it is virtually confirmed that PEX11 proteins have a direct effect on peroxisome division, what is their role in metabolism? The loss of PEX11 proteins causes a reduction in oxidation pathways in yeast and mammals, so there seems to be an indirect effect occurring by an unknown mechanism. One possible mechanism suggested by the authors is that since PEX11 proteins interact in maintaining physical properties of the peroxisome membrane, maybe alteration of these proteins leads to an impairment of the transport system across the membrane which could explain the different degrees of effect that partial inhibition of PEX11 proteins has on oxidation. This is consistent with the observation by van Roermund that "the defect in MCFA oxidation can be overcome by permeabilizing the peroxisome membrane." It has been hypothesized that the PEX11 proteins are transporters, but further tests are needed.
Some experiments that would be helpful in characterizing the PEX11b protein include a hydropathy plot, x-ray crystallography, sequencing of the gene, and gel electrophoresis. A hydropathy plot would provide information on the number of intermembrane domains that the PEX11b protein has and it also would provide information about the location of the cytoplasmic domain. Even more information about the structure of the PEX11b protein could be obtained from x-ray crystallography, but I'm not sure what is required to obtain one of these images. I figured that since there is an antibody available for the protein, that it would be possible purify the protein by way of some type of chromatography. Obviously, the gene for PEX11b is known since it was deleted in mice, so if this is the case then maybe obtaining the sequence of the gene would be helpful in determining the function of the protein. The secondary structure can be determined from the primary structure and maybe when the gene sequence was put into a sequence comparison program such as BLAST, genes with similar sequences in other species would be found that encode for proteins that have similar functions to PEX11b. Again, if the protein can be purified, a sample could be run on a gel to determine the molecular weight. This might be helpful in determining whether or not the protein has undergone post-translational modification. If I were continuing with this project I would also be very interested in learning more about the alpha form of this protein, but since I don't know anything about this protein, I can't design any experiments.
Another approach to deciding whether or not PEX11b is a transporter protein would be performing more functional experiments such as immunoprecipitation, affinity purification, fluorescence labeling, or monitoring the concentration of lipids inside the peroxisome. Since an antibody for PEX11 is available, one of the easiest experiments to perform would be an immunoprecipitation. In this experiment beads with antibody bound to them are introduced into the cell with detergent. The antibody binds and then the cells are centrifuged so the beads go to the bottom of the tube. Then after removing the initial liquid layer, the antibody is released from the beads and you have your protein and whatever it is bound to. Although there is no way to expect what other proteins will be bound to PEX11b, I think this would be an excellent start in characterizes the role of the protein. After reading this paper, I would expect to maybe see VSP1 or proteins involved in maintaining the physical structure of the membrane immunoprecipitated with PEX11b. Different mutants can be characterized by immunoprecipitation since protein-protein interactions can be studied. Each sample is run on a gel and the size is compared to wild type.
There is another way that the actions of PEX11b could be studied. A fusion protein could be made with GFP fused to PEX11b so that the movement of the PEX11b protein could be studied in the cell. Other proteins or molecules with which researchers think PEX11b might interact could be labeled various colors and then a confocal or fluorescence microscope could be used to observe the localization of the colored labeling and any co-localization. Organisms can be studied while alive when GFP is used, but cells could be stopped at different points in proliferation and studied as well. Based on the information presented in this paper, I would expect PEX11b to interact with proteins such as VPS1 and MYO2 that control peroxisome division, but not with proteins involved in metabolism.
A final way to determine whether or not PEX11b is involved in transporting lipids across the peroxisome membrane would be to over express and delete PEX11b from different groups of cells and somehow measure the concentration of lipids that are inside of the cell. If PEX11b is not a transporter of these molecules, then changing the amount of the protein shouldn't affect the amount of lipid inside the cell unless PEX11b regulates another protein that is involved in transportation.
Another question that easily follows after reading this paper is what is the mechanism by which PEX11 proteins cause peroxisome division? One factor to consider provided by the authors' states that peroxisome division depends on the concentration of PEX11 proteins while another states that post-translational modification may be the important regulator. Although not much is known about peroxisome division, it has recently been determined that the dynamin-related protein VPS1 is required for peroxisome division (Hoepfner et al., 2001). The authors state that "it will be interesting to determine whether PEX11 proteins participate in VPS1-mediated peroxisome division or whether they act in another process, such as coat-mediated peroxisome budding." There are a few experiments that they might be planning to perform to test this. I would start with immunoprecipitation as described above.
One of the questions that I thought would be interesting to address when I was reading the paper was the role of other proteins interacting with PEX11 and this is the next step that the author's state that they want to address. I would perform an immunoprecipitation experiment using the antibody for PEX11 first and then using antibodies for other known peroxisomal proteins as described previously. Another way to study protein-protein interaction in vitro is affinity purification, but both processes should give similar results.
Finally, is there a role for peroxisome metabolism in regulating peroxisome abundance? Studies previous to this paper showed that defects in enzymes in the metabolic process led to a reduction in peroxisome abundance. Some scientists have proposed that metabolite flux generates a signal, which regulates peroxisome division, but PEX11b levels are not altered in some human cells with metabolic problems. Therefore, Li and Gould propose that there are two mechanisms controlling peroxisome division. One in which PEX11 proteins are involved and another, independent of PEX11 proteins that is sensitive to metabolic flux through peroxisomal fatty acid oxidation.