Kevin James' Paper Review

This web page was created for an undergraduate course at Davidson College

Paper to be reviewed:

PEX11 promotes peroxisome division independently of peroxisome metabolism.

Xiaoling Li and Stephen J. Gould

Department of Biological Chemistry, The Johns Hopkins University School of Medicine, Baltimore MD 21205

The Journal of Cell Biology, Volume 156, Number 4, February 18, 2002 643-651.

Summary of Previous Knowledge:

This paper deals with peroxisomal membrane proteins; therefore it is important to understand a few facts about peroxisomes. Peroxisomes are the organelles in the cell responsible for lipid metabolism. Of importance to this paper is thier role in B-oxidation of fatty acids (Wanders and Tager, 1998; Tabak et al., 1999). Peroxisomes import their proteins and lipids from the cytosol. It is hypothesized that they grow by taking up more lipid and protein, and then divide following this growth phase (Lazarox and Fujiki, 1985). We thus note that peroxisomes can arise via this growth and division system (Purdue and Lazarow, 2001). The purpose of this paper will be to gain insite into peroxisome division.

It is previously known that cells with defects in peroxisomal fatty acid B-oxidation have fewer peroxisomes. This leads to the hypothesis that lipid metabolism play a role in determining the number of peroxisomes in the cell (Chang et al., 1999). It is also know that loss of PEX11 is directly correlated with peroxisome abundance (increased PEX11 leads to increased peroxisome abundance, decreased PEX11 leads to decreased peroxisome abundance in cells). This finding leads to an alternative hypothesis, that PEX11 plays a direct role in peroxisome abundance (Gould and Valle, 2000). There is one additional piece of information that makes determing which hypothesis is correct a little more confusing. It was found that removing PEX11 from yeast affects the oxidation of medium chain fatty acids (MCFAs), and that affecting this oxidation lowers peroxisome abundance. This seems to lend evidence to the hypothesis that it is metabolism, not PEX11, that directly affect peroxisome division. This paper will deal with these two hypotheses, and in the end support a direct role of PEX11 and a secondary role of lipid metabolism in peroxisome division.

Summary of Results:

The authors previously showed that overexpression of PEX11 caused an increase in peroxisomes in human cells (Schrade et al., 1998). Their first experiment attempts to show that this increase involves three distinct steps. Their design for this test utilizes immunofluorescence microscopy in normal human fibroblasts (GM5756). For this paper they use the B form of PEX11. Into the human fibroblasts they microinjected pcDNA3-PEX11Bmyc. This added DNA will lead to expression of PEX11Bmyc fusion protiens, where myc is the added tag which will be used to identify the presence of the protein later. Following microinjection they take pictures of the cells using immunoflurorescence microscopy (data shown in Figure 1). Figure 1 can be divided into 2 groups. A,C, and E show cells probed with anti-myc Ab's. B,D, and F show these same cells probed with anti-PEX14 Ab's. The authors include the anti-PEX14 photos as a control that shows where in the cells the peroxisomes are. The wt protein PEX14 should only show up in peroxisomes since it is a peroxisome protein. The authors can then conclude that what is tagged in A,C, and E must be peroxisomes because they are in the same place as the tagged PEX14 proteins. Panel A shows the cells after 1.5 hours and shows the presence of peroxisomes. Panel C shows the cells after 4.5 hours. The peroxisomes appear more elongated than in Panel A. The authors include a 'zoomed in' box which aids in seeing this. Finally, in panel E (48 hours later) there has been a large increase in the number of peroxisomes. It should be noted that the intensity of fluorescence appears to be the same for panels A/B and C/D. Panel E appears slightly darker than panel F; however, the increase in peroxisome abundance is still noted and the difference seems of little importance. We can also tell by viewing the anti-PEX14 cells that not all cells took up the microinjected DNA.


Figure 1 adequatly shows that overexpression of PEX11B greatly increases peroxisomes. What is not as clear is the idea of 3 distinct steps. While the zoom in box in C seems to show some peroxisomes longer than those in A, it would be helpful if the same cells were shown in the same location at all 3 time periods. This would allow us to see if the peroxisomes move during elongation and division, and to see individual peroxisomes change shape.. It would also be nice to show a cell not microinjected probed with anti-PEX14 to get a baseline visual on the abundance of peroxisomes.

The authors next wish to explore the specificity and extent of PEX11B driven peroxisome growth and division. To do this they perform two seperate microinjections into normal human fibroblasts. They transfected pcDNA3-PEX11Bmyc into one set of cells and pcDNA3-PMP34myc to another set. PMP34 is yeast peroxisomal protein with function in B-oxidation pathways. The control for this experiment was a set of unstransfected cells. The authors used confocal flurorescence microscopy to view the cells. Peroxisome abundance was determined by adding up the peroxisomes in the widest part of the cell (Peroxisome Per Section/Cell). This data is shown in Figure 2A. Untransfected cells showed a peroxisome abundance of 94 +/- 36. Cells transfected with the yeast PMP showed 101 +/- 37 (statistically the same as the untransfected cells as shown by error bars in figure). The cells transfected with PEX11Bmyc showed a peroxisome abundance of 964 +/- 341. This data shows a 1000% increase in the abundance of peroxisomes of cells transfected with PEX11Bmyc. They show the microscopy data in Figures 2B-2E. B and C were both transfected with PMP34myc and B was probed with anti-myc, while C was probed with anti-PEX14. These two pictures show a low abundance of peroxisomes by both probes. D and E were both transfected with PEX11Bmyc. D was probed with anti-myc, and E was probed with anti-PEX14. D and E both show an incrased abundance of peroxisomes compared to B and C (which did not recieve PEX11). Panel D glows brighter than E, and this seems to confirm that the anti-myc Ab shows stronger binding than the anti-PEX14 (as this is what we noticed in Figure 1).

The authors include that they also performed this test using human PEX3 in place of PMP34. They say that overexpression of PEX3 caused no increase in peroxisome abundance just as overexpression of PMP34 did not lead increased peroxisome abundance. Data for this test was unfortunatly not shown.

Conclusion Drawn: Increase in peroxisome abundance is induced by overexpression of PEX11B specifically, and thus it is not just a result of overexpressing any PMP.


A primarily difficulty for the reader in this section is that it is impossible for us to check their data. One cannot count the peroxisomes in the sections to back up the values given by the authors. This would be more problematic if the difference was not so dramaticly (and statistically) different. Another problem is that they only show data for the PMP34myc. They mention the PEX3 in this section and allude to testing 10 human PMP's in the discussion, but his data is not shown. It could thus be argued that other PMP's not tested, or whose data isn't shown, could lead to similar direct relationships with peroxisome abundance. While the effect of changing PEX11 is dramatic, it does not rule out the fact that other proteins could lead to similar results. They should therefore tone down the comment regarding general PMP overexpression. I feel comfortable with these data showing the implied results because the difference is so large (1000%) and they used controls that worked (transfected with PMP34 myc and untransfected).

The next task of the authors is to show that the increase in peroxisome abundance seen above was caused directly by PEX11B and not instead as a result of altered metabolism. The previously held hypothesis (van Roermund et al., 2000) is that in order for PEX11B to increase the number of peroxisomes, there must be a functioanl B-oxidation pathway present. They authors can test this by using cells that do not have such a functinal pathway. They explain that in human cells there are two sets of B-oxidation enzymes (Wanders and Tager, 1998; WAnders et al., 2001a). For this reason they decide to use cells that lack all peroxisomal metabolic function, yet still have peroxisomes. The cells they choose are PBD005 which lack all peroxisomal metabolic activity. They once again transfect cells with pcDNA3-PEX11Bmyc or pcDNA2-PMP34. They let the cells incubat for two days and then viewed them using indirect immunofluorescence. The control was once again untransfected cells. They calculated peroxisome abundance as in Figure 2. Figure 3 shows the data for this experiment. Figure 3A shows that untransfected cells had 32 +/- 16 peroxisomes per section/cell, 35 +/- 15 for cells transfected with PMP34myc, and 979 +/- 388 for cells transfected with PEX11Bmyc. It is apparent that the cells transfected with PEX11Bmyc once again showed a large increase in peroxisome abundance compared to cells transfected with PMP34myc, which showed no increase compared to untransfected cells. The authors do note the discrepency between untransfected GM5756 cells and unstransfected PBD005 cells. They say that this was predicted based on knowledge of the lowering of peroxisomal abundance that occurs in cells with disfunctional peroxisomal B-oxidation. Plates 3B-3E provide visual evidence for the increase in peroxisomal abundance in cells overexpressing PEX11Bmyc. Plates B and C were transfected with PMP34myc, and Plates D and E were transfected with PEX11Bmyc. B and D were probed with anti-myc, and C and E were probed with anti-PEX14. Comparing these plates shows that it is in fact peroxisomes that are being tagged (because the PEX14 probed cells glow in the same place the myc probed cells do). The difference in peroxisome abundance between the two different transfections is quite remarkable.

Conclusion Drawn: PEX11B increases peroxisome abundance completely independent of all peroxisomal metabolic activities.

This conclusion is drawn because the human cells used showed no peroxisomal metabolic activity, yet PEX11B was still able to greatly increase their abundance of peroxisomes.


What is of concern in this section is the fact that they used cells devoid of all peroxisomal metabolic pathways rather than just the B-oxidation pathways predicted to affect peroxisomal abundance. They draw the conclusion that since there were no peroxisomal metabolic pathways functioning in the cells, then PEX11 is independent of metabolic activities. It could also be imagined that in a wt cell there is an interaction between PEX11 and another metabolic pathway that leads to peroxisomal abundance variation. While the data in Figure 3 shows very convincingly that PEX11B increases peroxisomal abundance without metabolism occuring, it seems that this could be an inflated result that occurs when the normal metabolic pathways are removed that may place a counter role in PEX11 peroxisomal increase. There is a slight difference in the peroxisomal increase between Figure 2A and 3A. Perhaps this means that some metabolic pathway slightly reduced PEX11B's ability to increase peroxisomes.

In both 2B-2E and 3B-3E it would have been nice to see an untransfected cell probed with the a PEX14 Ab to get a feel for the effect of introducing a plasmid on the cell.

The authors next switch from testing PEX11B in human cells to testing PEX11p yeast cells in order to determine if the yeast equivolent can lead to peroxisome abundance increase directly, and thus independent of B-oxidation levels. The first task was to get a baseline range of peroxisomes in yeast (BY4733) grown on glucose and oleic acid media. In order to detect the peroxisomes they introduced GFP/PTS1 via a consitutivly expressed plasmid (pPGK1-GFP/PTS1). This protein is labled with GFP protein and will be taken into all peroxisomes allowing them to glow green (Kalish et al., 1996). They also introduced a plasmid with the high copy GAL1 promoter not hooked up to any gene as a control for later experiments. The authors simply count the number of peroxisomes in each cell (120 cells counted for each media type). Figure 4A and 4B show this data. The cells grown on glucose had, on average, a lower number of peroxisomes per cell than did the cells grown in oleic acid. The authors predicted this based on previous research; however, they did not expect there to be such a range in the number of peroxisomes in each cell. The graphs in Figure 4 compare #peroxisomes vs cell frequency. For glucose the range was from 1-12 peroxisomes, and for oleic acid the range was 4-21. The constant level expected by the authors was based on a different strain of yeast (BJ1991). It is important to keep this range in the controls in mind when viewing plates 4C-4I.

The next set of experiments will be conducted on media without fatty acids to prevent peroxisomal metabolic activity. The authors created two new strains of yeast with which to test the effect of PEX11 on peroxisome abundance. They deleted the chromosomal copy of PEX11 and introduced the plasmid pPGK1-GFP/PTS1, and called this strain XLY1. They then addeda high-copy plasmid (with GALI promoter) that expressed different proteins: Ypr128Cp, Pex13p, and Pex11p. Ypr128Cp carries adenine in the peroxisome membrane (Palmieri et al., 2001; van Roermund et al., 2001). Pex13p is a PMP that works to import proteins to the peroxisome. They grew the XLY1 cells with the varying plasmids for 17 hours on either galactose or glucose. They then counted the number of peroxisomes by looking for the glowing GFP in each peroxisome.

When the XLY1 cells, with GAL1 promoter attached to no gene, were switched from glucose (4C) to galactose (4D), the authors say that the peroxisome abundance increased. It is mentioned that glucose-repression causes the cells to have fewer peroxisomes. When the XLY1 cells were given a plasmid with the GAL1 promoter attached to PEX 13 (4E), on galactose from D-I, they showed the same peroxisome abundance as the cells in 4D. The cells given plasmid with GALI promoter attached to Ypr128c (4F) also did not increase peroxisome abundance above the control 4D. The XLY1 cells given the plasmid with the GAL1 promoter attached to Pex11 did show an increased peroxisome abundance compared to 4D. From this set of experiments of XLY1 the authors conclude that only Pex11 was able to increase peroxisome abundance when overexpressed. Furthermore, since there were no fatty acids in the growth media, they conclude that the effects of metabolism are not directly involved in the increase in peroxisome abudnance.

The authors conducted one more experiment to test this hypothesis. They develope a strain XLY2 which lacks a functional POX1 gene. This means that it cannot successfully perform B-oxidation. In these cells, as in the above experiemnt with XLY1, the Pex11 plasmid successfully led to increased peroxisome abundance (4H), and the Pex13 plasmid did not (4I).

Conclusion Drawn: In yeast, as in humans, Pex11 has the ability to directly control peroxisome abundance independent of metabolism.


The previous sections were very convincing in the presentation of statistically distinct data. Figure 4 is not as clear due to the range of peroxisome numbers in the cells. It is unclear why so many cells were so unaffected by changes in PEX11 and some were drastically changed (in regards to their peroxisomal abundance). However, they set up good controls with which to compare the effects of PEX11 vs effects of other PMP's and promoter only plasmid. While there is some overlap in 4G and 4D, it is clear that the peak of each is independent of the other. For this reason I am willing to believe that PEX11 is shown to increase peroxisome abundance in yeast as well as humans. I am not willing to conclude from the XLY2 data that other PEX13 does not also affect the abundance of peroxisomes. In the previous Figures it was clear that they had no effect; in this figure however, the difference between 4H and 4I is minimal. Adding this data did not help their case and could have been ommited. They have once again shown that with no fatty acids present, PEX11 is able to alter the abundance of peroxisomes. This strengthens the claim made earlier; however, just because PEX11 can increase peroxisomes without metabolism doesn't mean that in wt cells it does not work along with metabolic pathways to do so.

The authors now propose two possible mechanisms for PEX11's affect on peroxisome division and medium chain fatty acid oxidation. Either PEX11 proteins directly affect peroxisome division and fatty acid oxidation, or PEX11 proteins affect on fatty acid metabolism is indirect. To test this they created mouse strains. One strain is -/- for PEX11B, and the other is +/+. The idea is that if PEX11B has a direct role in oxidation, then it is increased when increased fatty acids are around. This would lead to an increase in peroxisomes because we're shown that increasing PEX11B increases peroxisomes. Thus, when cells are grown in a media without fatty acids, the level of PEX11B would not increase, nor would peroxisomes, based on this theory. The strain -/- for PEX11B should show the same number of peroxisomes as the +/+ strain if this is the case when no fatty acids are present because by this thoery there are no acids to encourage an increase in PEX11B production. We see in Figure 5A that there is a difference between the +/+ and the -/- mouse cell culture's peroxisome abundance. The authors deduce from this that PEX11B increases peroxisome abundance directly and is not mediated by the presence of fatty acid metabolism. A control for this experiment was growing the cells in normal media in addition to the fatty acid free media. It is apparent that each genotype showed the same peroxisome abundance in both media it was grown in. Viewing Figure 5B-5E shows the different numbers of peroxisomes in +/+ vs -/- mouse cells. It was by counting the glowing peroxisomes that they came up with the numbers in 5A.


This experiment is well predicted and carried out. The data does support their explanation of PEX11B having a direct role in peroxisome division because when niether cell is performing peroxisomal metabolism, there is still a difference in the abundance of peroxisomes. It is curious however that the level of peroxisomes in the -/- is so close to that of the +/+. In previous experiements the authors had been using overexpression of PEX11 to show its affect vs other PMP's. With the peroxisome levels being so close in these mouse cells it seems that we must consider whether or not overexpression of PEX11 is somehow overshadowing a role of another protein in peroxisomal abundance.

Overall this paper convinced me that increased PEX11 can greatly increase peroxisomal abundance with or without peroxisomal metabolic activity present. They then showed that a lack of PEX11 can decrease peroxisomal abundance in cells, which further improved the argument.

Future Research:

A key to understanding the role played by PEX11 in peroxisomal division would be to learn what other protiens PEX11 interacts with. There are a few ways we could go about figuring this out. We could perform an immunoprecipitation with an Ab against PEX11 and use a mild detergent to get it out of the membrane. This would pull out PEX11 with any associated proteins still bound. We could then seperate the proteins and run them on a gel. This would give us a starting place from which to explore new proteins that may play a role in increasing peroxisome abundance. It is assumed that there will be at least one protein interacting with the integral membrane PEX11.

The authors mention in the discussion that VPS1 and MYO2 have been shown to be important for peroxisome division (Hoepfner et al., 2001). These proteins are involved in the movement and pulling apart of the peroxisome. Since these proteins are also able to increase peroxisome abundance it seems that we should look for a connection to PEX11. It could be the case that VPS1 would be found binding to PEX11 in the immunoprecipitation, and this could easily be tested by looking at the MW. If this is not the case, then it would still be useful to view all three of these proteins simultaneously in the cell. This could be done by attaching different colored marker proteins (GFP, red fluorescent, blue fluorescent) and looking to see if any of the three proteins are ever co-localized in the peroxisome. You could also explore whether or not they are all turned on at the same time, in sequence, or randomly compared to one another.

Another way to test PEX11 for its interactions with other proteins would be the yeast two-hybrid system. The fact that PEX11 is an integral membrane protein would make this more difficult because each portion of the protein that sticks out of the membrane would have to be tested individually to see if it could bind another protein. Depending on the Kyte-Doolittle model of PEX11 this would involve a few or a lot of domains to test (I haven't seen this data to know). One could either do random selection of the proteins to attach to the activation protein, or select only proteins suspected to play a role in binding PEX11. If a protein (attached to activation protein) was found that did bind to PEX11, then we would see transcription of a reporter gene because PEX11 is fused to the GAL4 promoter that is attached to this reporter gene. If this method worked we could pinpoint which portion of the PEX11 protien binds the other protein. We would also know what other protein(s) bind to PEX11. We could then design experiments to see if just having that portion is sufficient to lead to increased peroxisomal abundance, or if the entire PEX11 is needed.

The authors bring up questions about PEX11's role in fatty acid oxidation. They do not believe that PEX11 is directly involved in bringing fatty acids because it is not similar to other proteins that do the same thing. They predict that PEX11 alters the peroxisome membrane in a way that leads to a change in import of fatty acids. It would be a good test to quantify how a variation of PEX11 in the peroxisome membrane affects fatty acid concentration inside the peroxisome. This could be performed by using plasmids with promoter hooked up to PEX11 in cells whose PEX11 genes were deleted. The promoters could be of different 'strengths' and would thus lead to different levels of PEX11 expression. These different cells would have their peroxisomes tested for fatty acid concentration and the relationship could thus be observed.

Another way to see how PEX11 affects the peroxisome membrane would be to show localization and movement of PEX11 within this membrane. The FRAP method could be used to show whether or not PEX11 moves freely within the membrane or is anchored. FRAP would involve attaching glowing proteins to PEX11 proteins. You would then bleach out a portion of the peroxisome membrane and see if over time glowing returned to that area. If this occured then you would know that PEX11 was able to move throughout the membrane because it had passed through the previously bleached spot. This could also give diffusion rate of PEX11. It would be of great utility to understand where PEX11 is localized in the membrane at different stages of peroxisome growth and division. By fusing GFP to PEX11 you could follow the PEX11 proteins to see if they moved in predictable ways at different stages of this process. It seems like a good prediction to say that PEX11 might group together if it is directly involed with the proteins used to expand and pull apart the peroxisome. If PEX11 is only a necessary middle-man some pathway (a limiting reagent in this path) that it could be more spaced out in the membrane.



Wanders and Tager, 1998. Lipid metabolims in peroxisomes in relation to human disease. Mol. Aspects Med. 19:69-154.

Tabak et al., 1999. Peroxisomes: simple function but complex maintenence. Trends Cell. Biol. 9:447-453.

Kalish et al., 1996. Characterization of a novel component of the peroxisomal protein import apparatus using fluorescent peroxisomal proteins. EMBO J. 15: 3275-3285.

Palmieri et al., 2001. Identification and functional reconstitution of the yeast peroxisomal adenine nucleotide transporter. EMBO J. 20: 5049-5059.

van Roermund et al., 2001. Identification of a peroxisomal ATP carrier required for medium-chain fatty acid beta-oxidation and normal peroxisome proliferation in S. cerevisia. Mol. Cell. Biol. 21: 4321-4329.

Wanders et al., 2001. Single peroxisomal enzyme deficienceis. In The Metabolic and Molecular Bases of INherited Diseases. Vol.2. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle, editors. McGraw-Hill, New York. 3219-3256.

Lazarox and Fujiki, 1985. Biogenesis of peroxisomes. Annu. Rev.Cell Biol. 1:489-530.

Purdue and Lazarow, 2001. Peroxisome biogenesis. Annu. Rev. Cell Dev. Biol. 17: 701-752

Gould and Valle, 2000. The genetics and cell biology of the peroxisome biogenesis disorders. Trends Genet. 16:340-344.

Chang et al., 1999. Metabolic control of peroxisome abundance. J. Cell Sci. 112: 1579-1590.

Hoepfner et al., 2001. A role for Vps1p, actin, and the myo2p motor in peroxisome abundance and inheritance in S. cerevisia. J. Cell Biol. 155: 979-990.


Dr. Campbell's Homepage

Molecular Biology Homepage

Email Kevin James with comments or questions