Farnesyltransferase Haplodeficiency Reduces Neuropathology and Rescues Cognitive Function in a Mouse Model of Alzheimer Disease*

F1.largeThe paper that I chose to focus on is called “Farnesyltransferase Haplodeficiency Reduces Neuropathology and Rescues Cognitive Function in a Mouse Model of Alzheimer Disease.” This paper focuses on the extent that isoprenylation plays in the formation of amyloid-beta plaques in Alzheimer Disease (AD) and furthers knowledge from a paper that we discussed in class earlier. The authors were aware from previous studies that the use of statins reduced protein farnesylation and geranylgeranylation and therefore amyloid precursor protein (APP) metabolism through inhibition of HMG-CoA Reductase, however there was not much information about the relative importance of two proteins that perform the protein prenylation. These enzymes, called farnesyltransferase (FT) and geranylgeranyltransferase-1 (GGT), are involved in anchoring proteins to the membrane. This is important in AD because it effects the GTPase superfamily of Ras, Rab, and Rho proteins that are involved in APP metabolism. These GTPases are found in different parts of the brain and are involved in the signaling pathway to activate the processing of APP to make amyloid-beta (AB). The alternative splicing of APP to AB-42, a 42 amino acid protein, instead of AB-40, a 40 amino acid protein, has been associated with the physiological effects of AD. The accumulation of AB-42 on the surface of neurons eventually form plaques and increase phosphorylated Tau protein concentrations, which inhibits communication between neurons and causes neurodegeneration such as microgliosis.

To uncover more about the role of protein prenylation in AD, the researchers bred AD model mice to be haplodeficient for the gene encoding for either FT or GGT. Western blot analysis confirmed that the level of FT or GGT in the haplodeficient mice was lowered by ~50% as compared to WT AD mice. The authors also found that levels of AB-42 and AB-40 levels were decreased in the cerebrum through the analysis of brain lysates by ELISA (Figure 2). The same trend was seen in the hippocampal and cortical areas of the haplodeficient mice (Figure 3).  To take these findings further, the authors went on to show that APP processing was trending towards the production of a non-amyloidogenic state in FT-haplodeficient mice. In other words, amyloid-alpha was being produced instead of amyloid beta. This is completed by an enzyme called a secretase. The version of the amyloid protein that is produced depends on the cleavage that is carried out by either alpha or beta-secretase. This produces small carboxyl (CTF) and amino terminal fragments (sAPP) that the authors could detect by immunoblot analysis. The data shows that in FT haplodeficient mice there is an increase of alpha-CTFs and sAPPs, which suggests that AB production is lowered since fewer beta fragments were seen after secretase splicing. There was no difference in the clearance of AB between WT and haplodeficient mice in data collected in a similar experiment targeting the apoE and LRP1 proteins. Neuroinflammation was also being influenced by less protein prenylation. Haplodeficient mice had lower levels of IBA-1 (a marker for microgliosis) and GFAP (a marker for astrocyte activiation), which are involved in neruoinflammation associated with AD. Next, what the authors referred to as a “battery of behavioral tests” were used to assess the implications of their work on a physiological level. Interestingly, the FT haplodeficient mice consistently performed as well as the WT mice, but the GGT deficient mice had the same performance as the AD mice with no gene knockouts. The authors go on to explain that since FT and GGT perform different roles in the progression of AD it is not surprising to see confounding results about the effect statins can have on AD. Targeting FT could prove to be an effective strategy for curing or treating patients with AD.


17 Replies to “Farnesyltransferase Haplodeficiency Reduces Neuropathology and Rescues Cognitive Function in a Mouse Model of Alzheimer Disease*”

  1. Matt – Cool paper! I was really excited by the results, so I Google Scholar-ed this paper to see if anyone has cited this paper in trying to inhibit FT as an Alzheimer’s therapy. Those studies must be in the works as I did not see anything (or, ya know, those experiments just failed miserably…) Anyway, my question for you is an impact question by way of a methods question. In showing the differential effects of knocking out FT or GTT, the authors were making headway in flipping paradigms about statin treatments and the progression of AD in general. Further, as the authors seemed to tell us at every paragraph, results of this nature have not been shown in vivo to this extent yet. These points of distinction seem to be setting us up for a high impact type of discovery. But yet here we are in the JBC. One reason for this, as they report, is that the mechanism for the rescue of cognitive function is completely unknown. But I am theorizing that the mouse model has more to do with it. The authors simply report that they bought the mice, and in a more recent paper out of that lab (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0108845), the authors simply cite themselves in reporting this model of AD-mouse behavior as if they’ve already established the soundness of the method. So without further ado (and I do apologize for all the ado up to this point), here are my questions (certainly not all of them need to be answered): What is it about these mice that make them an appropriate model for AD progression? What are the limitations? How are they “manufactured”? Is there anything out there assessing the strength of the model? To what extent does the mouse model affect the paper’s impact?

    1. Hi Zach! This is a really relevant question, as models for human diseases in mice are not always perfect. These mice have been engineered to have two common mutations that are found within certain groups of individuals with early onset AD. Basically, it is as if these mice have been extremely unlucky genetically, which causes the physiological effects generally associated with AD. There has been some debate previously about the extent to which the mouse model of AD fully represents the human form of AD. The mouse model, while not perfect, does produce the formation of amyloid-beta plaques, as well as the appearance of phosphorylated-Tau protein. While not entirely perfect, this model is generally accepted because the physiological effects match up well between the model and the human form of AD.

  2. Hey Matt, I enjoyed this paper. It’s material tied nicely into what we’d covered in class and I didn’t feel completely lost when reading it.

    My question is in regard to their final statement, that ‘specific inhibition of protein farnesylation could be a potential strategy for effectively treating AD.’ The purpose of this study was clearly to dissect the implications of FT and GGT deletions… And we see:
    FT delete= decrease neuroinflam., and rescued spatial learning and memory func.
    GGT delete= decrease neuroinflammation. NO learning and memory impact.

    Back to my question, why would we ONLY want to inhibition of protein farnesylation? Surely the neuroinflammation associated with geranylgeranylation can’t be good, so why would we allow that to persist? Obviously the study wanted to differentiate between the two, but if we could inhibit farnesylation AND geranylgeranylation with the use of statins, why wouldn’t we do that? I can’t see the point of specificying to inhibit only farnesylation… Does geranylgeranylation have a positive effect on AD that I’m missing?

    Thanks again for the great write-up!

    1. Hi Ryan! I had the same kind of thought while I was reading this myself, and I definitely agree with you. My best guess as to why the authors chose to focus more on farnesylation is that they deemed the recovery of cognitive function more important than neuroinflammation. The effects of geranylgeranylation should be more closely reviewed in future research to deduce geranylgeranylation’s role in AD.

  3. Nice job, Matt! I really appreciated how closely this paper related to topics that we discussed in class. It was very easy to follow. I find it very interesting how farnesyl isoprenylation and geranyl isoprenylation result in different downstream effects following cell anchoring. I’m curious as to why this could be because farnesyl and geranyl only differ by a few carbons units. The paper mentions that the mechanistic details about this are still to be determined, but do you think that the difference of a few carbon units could have some effect on the way that these two groups interact with other proteins? Maybe the structural difference could also affect downstream activity inside the cell? No need for additional literature searching to answer my question, etc. – just curious to see if you had any thoughts. Thanks!

    1. The most reasonable explanation is that protein farnesylation and geranylgeranylation are attached to proteins by recognition of different sequence motifs. Most likely proteins are selectively farnesylated or gernaylated based on their different functions, which is exemplified by the different physiological effects the authors saw when selectively inhibiting FT or GGT.

  4. Hi Matt! Great job selecting an article that links so well with our class content! My first question is why they kept the mice such a short period of time when Alzheimer’s Disease is a degenerative disorder that typically occurs over time? I would have been interested to see if their results changed over time. Additionally, was concerned about the potential for extraneous variables in this experiment.
    Also, to the best of my knowledge, this pathway is not the only physiological pathway to involve isoprenylated G proteins. How can we be sure that these results are really due the effects of FT and GGT on amyloid beta only? If FT did turn out to be a therapy for AD, how can we be sure that this wouldn’t be problematic for other cholesterol metabolites? We saw that the use of statins had a large impact on downstream cholesterol metabolites, is it possible that the use of FT could throw off this pathway as well? Thanks!

    1. The justification for looking at these mice in such a short period of time is that these mice are genetically engineered to have a severe case of early onset (familial) AD. The specific mutations were identified in populations of people with unusually severe cases of familial AD, specifically from Sweden and London. The general thinking here is that the severity of AD in these mice is high enough that the form of AD present in mice resembles that of a human patient by the time symptoms have shown themselves.
      There is no way to be 100% that the results the authors are seeing are from the inhibition of FT or GGT alone, but the data seems to suggest that conclusion. the important thing to note here is that these mice were only haplodeficient in FT or GGT, meaning that there was still some level of protein prenylation occurring in the cell. If FT does become a potential therapy for AD, test would need to be run on other important prenylated proteins to ensure that FT inhibition does not effect any aspects of metabolism as a whole.

  5. Hi Matt,
    Great job with your review! As I was reading through the discussion section of the paper, I noticed that the authors spent quite a bit of time arguing that the use of statins to inhibit isoprenylation is highly problematic; they do not simply state their objections, but instead spend about one-third of a page detailing the problems of this older method, while spending little time supporting the methods they actually used in their study. While I definitely think the authors needed to highlight these shortcomings, why do you think such a long-winded explanation was necessary? Their in vivo genetic approach may be new in the study of isoprenylation, but obtaining knockout mice to test the effects of a particular protein is hardly a novel idea, or one that would need much justification in 2015. While I know that JBC isn’t as strict with page limits as some of the more high impact journals, do you think a few sentences would have been sufficient to make their argument, or was an entire paragraph actually necessary? Finally, in terms of overall organization, does this argument belong in the discussion section, or might it be more effective in the introduction, where they are setting up the background information for their study, and identifying how this study fills a gap in the literature?

    1. I think that the authors were so long-winded about these shortcomings because the reader needed to understand the current state of research in this field in order to fully comprehend the paper in context with previous literature. The authors wanted to distinguish themselves from papers that use statins because these two methods have similar, but different effects on prenylation and on AD progression. In my opinion, this information belongs more in the introduction than in the discussion. The authors at least should have mentioned statins briefly in their introduction to put the paper’s overall goal in focus.

  6. Hey Matt,
    I really liked your analysis, it was pithy but covered all of the important parts and organized it very well. I have a question and I’m not entirely to sure to what extent you’ll be able to answer it but I’m curious of your thoughts. Normally, a protein is farnesylated, anchors to the membrane, and does its thing. Effectively, the farnesyl group is made and “used up”. Inhibiting FT, though, will allow for more intracellular, unused farnesyl-PP. All other things equal, these molecules won’t have anywhere else to go, so won’t it shift the equilibrium towards more lanosterol and eventually to higher cholesterol? If this is the case, do you think we should re-consider using statins rather than FT inhibitors?

    1. That’s a great point, and it very well might be true. It would be interesting to look at kinetic and thermodynamic data between the two competing reactions, making lanosterol or making geranyl-PP and isopentenyl-PP (the reverse reaction). Likely, because the forward reaction involves the loss of a pyrophosphate, the reverse reaction would be difficult to complete, shifting the equilibrium towards the production of lanosterol as you suggested. There should definitely be some research into the long-term effects on cholesterol levels when using inhibition of FT as a therapeutic treatment for AD.

  7. Hi Matt- I really enjoy that you picked a paper that relates to topics we’re learning about in class, and I really liked this one- I think it laid out its results well and did not overstate its significance. I am personally interested in what the authors also call “the most striking observation” of the study- the fact that farnesyltransferase haplodeficiency rescues learning and memory in mice, whereas geranylgeranyl transferase does not. The authors make a comment towards the end of their discussion that implies that statins may not be a good preventative/therapeutic agent for AD due to the fact that they would affect the prenylation pathways involving both FT and GGT because the positive results may cancel each other out. I was wondering- do you think that the authors are implying that statins may not only be unhelpful, but actually deleterious in people who have to take them? The authors go to great lengths to describe the importance of geranylgeranylprenylation’s physiological role. Given the fact that statins are considered fairly safe drugs, do you think that the authors could have overstated the importance of geranylgeranylprenylation in order to better explain their contradictory results?

    1. That does seem confusing, as the authors claim that there are positive effects of GGT inhibition, like neuroinflammation, despite these mice not regaining their cognitive function. They also cautioned the reader by saying that geranylgernaylation, and not farnesylation, was involved in negative side effects, such as erosive inflammatory arthritis in mice with deleted GGT genes. As for FT and GGT as potential drug targets, the authors make a point to compare their use against statins in order to highlight the differences between inhibition of HMG-CoA and and FT/GGT. Because different statins have different mechanisms of action and are prescribed in various dosages, it is more difficult to accurately depict the role statins play in AD.

  8. Hola amigo,

    Shocking performance here, really questionable effort ;).

    So I have this nice little quote from the paper here:
    “Also, limiting the availability of isoprenoids for prenylation
    protects neurons from A[Beta]-induced apoptosis via the activation of pro-survival signaling pathways”. I also did a little research, and discovered that Ras GTPases are more involved in regulating cell growth and division, whereas Rho GTPases are more involved in regulating cell scaffolding and organelle development. In their conclusion, the authors never really conjectured as to why haplodeficient mice for FT did not develop symptoms of Alzheimer disease where as haplodeficient GGT mice did. Do you think that overactive Ras signaling might have any part in the development of Alzheimer disease? How would you design (just a general scaffold, don’t murder yourself with details) an experiment that might help to further elucidate this conundrum?

    1. I am not entirely sure how Ras is involved with the development of AD. If Ras proteins are farnesylated and not geranylgeranylated, there would be an effect of Ras’ activity when FT was inhibited. Whether or not that plays a role in the actual formation of amyloid-beta plaques is not known. To me, it does not seem as though Ras or Rho proteins would be involved in the processing of amyloid-precursor protein by secreatses that creates the toxic AB-42. Ras and Rho are more likely involved in the progression of AD rather than the development of AD.

  9. Hey Matt,
    Great paper! As I read through it, I was left a little curious as to how the relatively broad scope of the statin-mediated response could be compared to the much more surgical genetic variant-based experimentation. Given that the methodologies of the statin approach are both broad-acting upon the farsenylgeneration and geranylgeranyation, is it possible for another mediated effect to play an equally important role as the genes identified within the study? Do you think it would be beneficial to combine both experimental options into one study so as to discern the interrelated nature of the two methods?

Comments are closed.