This page was created as part of an undergraduate course at Davidson College.
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
PEX11 are peroxisomal membrane proteins known to affect peroxisomal division. Some researchers think that PEX11 affects division by directly influencing the peroxisomal metabolic pathways, but these authors show that PEX11 by itself promotes peroxisome division and affects metabolic pathways indirectly.
The peroxisome is a cellular organelle that must import all their protein and most of their lipids; they act in the metabolism of lipids, including the B oxidation of fatty acids. There are ideas on the biogenesis and development of peroxisomes, but little is known for sure. Peroxisomes are researched heavily because in mammals, including humans, peroxisomal dysfunction can result in several different lethal disorders.
Like development, little is known about peroxisome division, but the authors previously found that there may be metabolic control of peroxisome division(less B fatty acid oxidation results in a lower abundance of peroxisomes). It has also been found in yeast, humans, rodents, and protozoa that lower levels of PEX11 resulted in less peroxisomes and that overexpression of PEX11 led to an abundance of peroxisomes.
These findings are known from past research and led to the hypothesis by van Roermund et al. that the oxidation of medium chain fatty acids in the peroxisome releases a molecule that promotes peroxisome division, thus linking peroxisomal division directly to peroxisomal metabolism. The authors Li and Gould test this hypothesis in mammalian and yeast cells and propose their own idea on peroxisomal division: the authors believe that PEX11 acts directly to promote peroxisomal division and that PEX11 indirectly affects the metabolism through some other means, thus making peroxisomal abundance independent of metabolism.
These experiments examine the affect of one of the human PEX11 proteins (PEX11B) on peroxisomal division in human skin fibroblasts. The authors previously found that overexpression of PEX11B dramatically increased peroxisome abundance in cells. To verify this, the researchers took human skin fibroblasts and microinjected one group with PEX11Bmyc, a second group with an unrelated peroxisomal membrane protein (PMP34myc), and they left a third group untransfected. The fibroblasts were immunofluoresced with antibodies for the c-myc tag and for PEX14, and endogenously produced PMP. Peroxisomes were counted at the widest part of the cells under fluorescing microscope, and it was found that both the PMP34 and the untransfected cells had about 100 peroxisomes per section while the PEX11B cells had around 950 peroxisomes per section.
Figure 1 shows the immunofluoresced antibody-labeled fibroblasts transfected with the PEX11Bmyc DNA. ACE were labeled with anti c-myc and BDF were labeled with anti PEX14. Pictures of AB were taken 1.5 hours post injection, pictures CD at 4.5 hours post injection, and pictures of EF were taken 48 hours post injection. In both antibody groups, you can see the large increase in labeled tags on new peroxisomes: on the left you can see an abundance of labeled myc tags on PEX11 proteins, and on the right you can see the abundance of labeled PEX14 being produced inside new peroxisomes, both after 48 hours. Also, in the small inset box, you can see the peroxisomes (white specks) elongating at 4.5 hours, presumably dividing.
There's not much for me to critique here: they ran a straightforward experiment and got good, clean results. They had the control of the untransfected cells and the alternavtive experimental group with PMP34, both of which gave drastically different results from the PEX11 transfection. They also included the myc label on their genetic constructs, in case peroxisomal division was not obvious and easy to see under fluorescence they could check for the myc tag to make sure the cells had been properly injected. The only criticism I have is that in the figure they only show the data for the PEX11 cells, not for the PMP or the untransfected cells, which would normally make me have to doubt any conclusions they made. However, they do include the numerical data for the PMP and the untransfected cells, so I'll say I'd really like to see those two control groups included in the figure, but because they included the numerical data I'm willing to believe their conclusion-"the increase in peroxisome abundance induced by PEX11B reflects a specific activity of PEX11B and is not a general consequence of PMP overexpression."
First Experiments Continued
The authors now try to show that PEX11B causes an abundance of peroxisomes independent of peroxisome metabolism. The authors do this by seeing if PEX11B induced proliferation occurs in cells without a functional peroxisomal B-oxidation pathway. The researchers took a human skin fibroblast line known to have peroxisomes but lack all peroxisomal metabolic functions (cell line PBD005) including fatty acid B oxidation. They they performed the same experiment as before on the PBD005 cell line and the original GM5756 cell line used. In both cell lines, they found that untransfected cells and cells transfected with PMP34myc averaged around 35 pps, while the GM5756 and PBD005 cells transfected with PEX11B both averaged around 950 pps.
Figures 2 and 3
Data for the two tested cell lines (GM5756 and PBD005) are shown, both showing the same results in tables and immunofluorescence pictures. In both figures 2 and 3, panel B was transfected with PMP34myc and labeled with myc-both cell lines show very little peroxisomal activity in this panel. In panel C, cells were transfected with PMP34myc and labeled with PEX14-again showing little peroxisomal activity for both cell lines. In panel D, cells are transfected with PEX11Bmyc and labeled with myc-tons of peroxisomes producing PEX11 labeled with myc can be seen in both cell lines in this panel, and in panel E, cells are transfected with PEX11Bmyc and labeled with PEX14-again so much protein is fluorescing almost the entire cell is visible. In panels A, the average peroxisomes per section for each of the three differenct transections are shown, both for the GM5756 and the PBD005 cell lines. For both cell lines, the untransfected and PMP transfected cells have average pps values of under 50, while the PEX11B transfected cells for both cell lines have average values approaching 1000 pps.
Again, I feel like I have to search for things to critique here. While they give the numerical data for the untransfected cell lines, they don't show any of the visual immunofluorescence data, which would be nice to see. They also compare the two average pps of the two control groups to the PEX11 cells in the graphs in A, so while it would be nice to see the visual data, I'm willing to believe their conclusion. The BCDE panels are confusing because, unlike Figure 1, the cell in B is different than the cell in D, so you're not really comparing panels B and D or C and E directly against each other; instead you're mainly comparing panels B and C to D and E. I think these data and these figures firmly support their conclusion that "peroxisome-proliferating activity of human PEX11B is independent of all peroxisomal metabolic activities."
In the second group of experiments, the researchers were attempting to prove that PEX11B induced peroxisomal proliferation can occur without fatty acid B oxidation in S. cerevisiae cells, instead of human cells. To do this, they took different cell lines of the yeast and grew them in different metablic conditions, and also grew them gave them different plasmid inserts to produce different proteins, such as a PEX11 or a different PMP.
This figure shows 9 different diagrams, each with a yeast strain growing under different metabolic conditions. First, a laboratory strain (BY4733) was grown on fatty acid-free medium with glucose and low levels of peroxisomal activity were recorded by an inserted, constitutively expressed GFP which labels protein in the lumen of the peroxisome. After these cells from diagram A, were moved to a medium with fatty acid (oleic acid) they peroxisome activity increased in diagram B. However, the researchers attempt to prove that this effect-the growth on fatty acid media-is not directly related to the peroxisome activity. Next, they deleted the chromosomal PEX11 gene from BY4733 to get the strain XLY1, which they used for the rest of their experiments. They grew these cells on a glucose media with very low peroxisomal activity, and then grew this strain on galactose media with slightly higher peroxisomal activity. Next, they inserted the coding region for the PMP PEX13 with the GAL 1 plasmid construct, and grew these cells on galactose media with about the same results as the cells without PEX13. Next, they inserted the coding region for the protein Yprf128c, which acts as an adenine carrier in the peroxisome metabolic pathway; they grew these cells on galactose media and saw low peroxisome activity. Then they inserted the coding area for PEX11 and grew these cells on galactose media-this produced the highest peroxisomal activity of all the experiments. Finally, they tested a new strain of cells-the XL2 cells, which were a POX 1 derivative of XLY1. They grew this strain on galactose media with PEX13 inserted and saw fairly low peroxisomal activity, then grew XLY2 with PEX11 inserted on galactose media and saw peroxisomal activity comparable to XLY1 PEX11 or BY4733 on oleic acid.
The first thing that comes to mind is, why the switch from human to yeast? It seems though, that having more or less proved the their point in human cells, they now want to set precedence by showing the same effects in all the animal models previously studied. There are so many components to these experiments that they're pretty hard to follow closely; however, they did find strong support for their claim-that peroxisomal proliferation occurs independently of fatty acid B oxidation. . I might criticize them for not being able to make the same statement for yeast cells that they did for human cells-that the metabolic processes in general don't affect PEX11 induced proliferation, instead of specifically having to narrow it down to fatty acid B oxidation for yeast cells. I wonder if they couldn't have produced a yeast strain that possessed peroxisomes but no functoinal metabolic pathways, like the human PBD005 line. I would also have been interested to see a cell line with the PEX11 insert tested under more conditions-such as one culture growing on glucose media, which should repress division; one culture growing on galactose media, which should also repress, but not as much as the glucose; and finally one culture growing on oleic acid, which I would expect to have the highest abundance of all the groups. Through these experiments, I'd like to test the PEX11 overexpression culture in different metabolic conditions to check if any other proliferation was occurring.
Already having made the points that PEX11 promotes peroxisomal proliferation independently of fatty acid oxidation, but that yeast lost some fatty acid oxidation when PEX11 was removed, the authors feel there are two possibilities: one, that PEX11 functions directly on both division and metabolism or two, that PEX11 functions directly on division but indirectly on metabolism. To test this, a knockout mouse was produced, lacking the PEX11 B gene. Fibroblasts from the knockout and a normal wildtype mouse were then cultured both in normal media and a serum free media, which lacked the substrates necessary for peroxisomal metabolism. The cells were then immunofluoresced with antibodies for peroxisomal enzyme catalase and PEX14. The levels of peroxisomal activity in were the same for in normal and serum free media for the wild-type and the knockout cells. The wildtype cells were recorded as having around 230 pps in both media, while the knockout had around 128 in both media. So, again, the authors show that peroxisomal dividing is independent of metabolic activity.
This figure is illustrated the same as Figures 2 and 3, with immunofluorescence data shown for the two different cell lines-the PEX11 normal mouse and the knockout mouse. The cells were labeled with antibodies for peroxisome enzyme catalase and PEX14, and like the previous data, the mouse with normal PEX11 produces peroxisomes in a much higher quantity than the knockout mouse. This data is shown both in the visual immunofluorescence images and is reflected in the graphs in A, which show the average peroxisome abundance for each mouse cell line. In both groups, the cells grown under normal conditions have about the same amount of peroxisomes as those grown under serum-free conditions that lack some metabolic activity. However, the normal mice cells have a much higher peroxisome abundance than the knockout mouse cells.
This set of experiments as well done. An initial concern is that they never validate the genotype of their knockout mouse, and the data for the production of their mouse is unpublished. However, their results are the expected results, and everything appears to be in order. I have to trust that the serum-free conditions really act as a metabolic detriment, and that these cells really were growing under very different conditions than the normally cultured cells.
Overall, I feel that the authors proved the ideas they were testing in each set of experiments. If I hesitate to agree, it would be because they don't show untransfected cell pictures in the first figures and they never validate the knockout mouse in the last experiment. Also, much of my willingness to trust their results is based on not knowing an extensive amount about the subject matter: I have to believe that cell line PDB005 has peroxisomes but no metabolic activity, I have to believe that all the genetic manipulation of the yeast cells worked, and I have to believe from the data that they actually produced a knockout mouse. Sometimes they make up for some shortcomings in other ways, such as giving the numerical data for the untransfected cell lines instead of the pictures, and in light of their results, this is enough to convince me, but not make me ecstatic about their conclusions. Overall, this was a well done group of experiments that provided strong evidence for the ideas behind them.
The researchers make some claims that I feel cannot be validated at this point. They say that PEX11 is "unique" in its ability to induce peroxisomal proliferation, but that's only relative to PMP34 or one of their other inserts. I would test other PEX proteins to see if they had any effect on the abundance of peroxisomes. It also seems that, because of differences in cells, the researchers make different conclusions for their experiments: in humans they can say metabolism does not affect peroxisomal division and abundance and in yeast and mice cells they can say that certain components of metabolism are not directly influencing peroxisomal division and abundance. I would like to run experiments with yeast and mice cells to attempt to prove the same as human cells-that metabolism in general does not affect peroxisomal division and abundance. I propose the idea for some extra experiments with yeast cells with different inserts grown under different conditions than the researchers used in the critique of Figure 4. Overall, I feel that these researchers made a large advance in proving that peroxisomal abundance and division occurs independently of some metabolic pathways, and that they showed this in many of the animal models previously used for experiments. I feel that this research does begin to set a precedence for human, yeast, and mice cells, that other research will have to follow. However, in future work, I would also like to see the same kinds of experiments run on any other animal models that have been used: rats, bacteria, etc., for comparison to Li and Gould's research.