Another Partner for LRRK2: Will This Relationship Endure?

A decade after the discovery that mutations in leucine-rich repeat kinase 2 (LRRK2) cause Parkinson’s disease, researchers are still trying to get a handle on how the giant, 51-exon, multi-catalytic protein exerts its pathogenic power. Now, researchers led by Valina and Ted Dawson at Johns Hopkins University in Baltimore report that LRRK2 phosphorylates a ribosomal protein, resulting in a boost in protein translation. Mutations that enhance the kinase activity of LRRK2 trigger a toxic surge in protein production that results in neuronal injury and death, according to their April 10 Cell paper. Other researchers in the field were cautious about the findings, but hoped they will be confirmed by others. 

Valina Dawson proposes that LRRK2’s effects on the ribosome lead to a gradual disruption of proteome homeostasis, with some proteins being more affected than others. “Initially the cells can compensate,” Dawson said, “but over time you end up in that slow slide toward misregulation and disease.”

“Understanding the basic biology of how LRRK2 exerts its effects, and identifying its immediate downstream targets, is a terribly important goal in the PD research field,” Dario Alessi of the University of Dundee in Scotland told Alzforum. However, Alessi, who was not involved in the study, questioned the rigor of the biochemical findings. To his mind, they were not compelling enough to support the mechanism proposed by the authors. “It will be key for other labs to reproduce this result,” he said. “In that case, this will be seen as a major advance.”

Lost in Translation.

Overexpression of a LRRK2 mutant with elevated kinase activity triggered a shortening of neurites in human dopaminergic neurons (top). Co-expression of a ribosomal protein proposed to be a LRRK2 substrate rescued this defect when rendered phosphor-deficient (bottom panel).

Dominantly inherited mutations in LRRK2 are the most common cause of familial PD, and a majority of known LRRK2 mutations occur in the kinase domain. The most common LRRK2 mutation, G2019S, ups the activity of the kinase. Intense research has tried to identify the proteins phosphorylated by LRRK2, but with limited success (see Tsika et al., 2012). Alessi estimated that a dozen potential substrates have surfaced, mostly from in-vitro studies, but most have failed to survive as bona fide substrates when mammalian cells or tissues were examined (see Kumar et al., 2010; Trancikova et al., 2012).  “The question is whether the data here are any different than the previous 12,” he said.

Valina Dawson said that an enduring issue with the identification of LRRK2 substrates has been that researchers have difficulty linking them to LRRK2 pathology. “We wanted to find the pathophysiologic targets of the LRRK2-G2019S mutant,” she said.

To pluck out a genuine pathological LRRK2 substrate from the thousands of proteins within the cell, first author Ian Martin and colleagues started by pulling out a tagged version of LRRK2 from cells and looking for phosphoproteins that came with it. The researchers next separated the phosphoproteins and analyzed their peptide makeup by mass spectrometry. Using those sequences, they then cloned and purified a majority of the 161 interacting proteins they identified, and tested which of them LRRK2 phosphorylated. They found 11 proteins that seemed to be LRRK2 substrates, and 10 were ribosomal proteins.

The investigators expanded their search to other components of the ribosome, and discovered the s15 ribosomal subunit was also phosphorylated by LRRK2 at threonine-136. In vitro, the LRRK2-G2019S mutant boosted the phosphorylation of s15. However, the LRRK2 protein appeared to phosphorylate itself far more readily than s15, which led some researchers to wonder if the ribosomal subunit represents a bona fide substrate. Dawson said they did not quantitate the number of phosphorylation events on LRRK2 versus s15. However, autophosphorylation is part of the normal biology of LRRK2, she said.

To determine whether s15 mediated the toxicity of LRRK2-GS, the researchers expressed both proteins in rat cortical neurons and in human cortical and dopaminergic neurons derived from a human embryonic stem cell line. When expressed together, LRRK2-GS and wild-type s15 drove up markers of cell death as well as the shortening of neurites. However, the rise in these toxic symptoms subsided when the researchers swapped s15 threonine-136 with an alanine residue (s15-TA), blocking phosphorylation. The s15-TA variant also prevented the loss of dopaminergic neurons and rescued deficits in climbing seen in fruit flies expressing LRRK2-GS. These results suggested that LRRK2 could mediate its neurodegenerative effects through phosphorylation of s15. 

The cell and fly experiments relied on overexpression of either the substrate or the kinase or both. This drew caution from some researchers, including Bingwei Lu of Stanford University in Palo Alto, California. “The reliance on overexpression in cells throughout this study could be a cause of concern,” Lu said. He added that it would be interesting to test whether altered s15 phosphorylation is responsible for the disease-relevant phenotypes observed in neurons derived from patients carrying the LRRK2 G2019S mutation. Dawson said such experiments are underway in her lab, but that the researchers had initially wanted to establish the LRRK2-s15 link in cells with a common genetic background.

To further examine the effects of the phosphorylation of s15 by LRRK2 in vivo, the researchers measured protein translation rates in flies by feeding them 35S-methionine and measuring incorporation of the labeled amino acid into proteins in the brain. Flies expressing LRRK2-GS made slightly more protein than wild-type flies. Interestingly, expression of s15-TA reversed this boost. Anisomycin, a translation inhibitor, also reversed neuronal loss and restored climbing ability in flies expressing LRRK2-GS. 

The fly results hinted that the toxic effects mediated by LRRK2-GS played out in the form of elevated protein translation, triggered by phosphorylated s15. Dawson’s lab is now in the process of determining which proteins downstream of s15 could be causing pathogenic effects.

“The hypothesis that G2019S LRRK2 toxicity leads to bulk increase in protein synthesis is interesting, as it adds to the growing body of evidence linking LRRK2 and other PD-related proteins to changes in protein homeostasis in the cell,” Gaia Skibinski of the University of California, San Francisco, wrote to Alzforum. “More specifically it adds to the increasing lines of evidence linking neurodegeneration to translation misregulation.”

Many of the biochemical and neuronal effects observed in the paper were subtle, making it difficult to draw solid conclusions from the data, Alessi said. However, he noted that LRRK2 mutations do exert subtle biological effects that take decades to cause disease. “It’s possible that the effects you would observe would be incredibly small, and therefore hard to observe,” Alessi said. “This makes a study very difficult in this field.”—Jessica Shugart

Comments

  1. The paper highlights the ribosomal protein S15 as a novel pathogenic LRRK2 substrate. Their hypothesis is supported by both in vivo and in vitro models of LRRK2 pathogenesis.

    The hypothesis that G2019S LRRK2 toxicity leads to bulk increase in protein synthesis is interesting because it adds to the growing body of evidence linking LRRK2 and other Parkinson’s disease-related proteins to changes in protein homeostasis in the cell. More specifically, it adds to the increasing lines of evidence linking neurodegeneration to translation misregulation.

    S15’s role in G2019S LRRK2-induced degeneration doesn’t extend to all pathogenic LRRK2 mutations, as demonstrated by their work on R1441C. It will be interesting to see if S15’s role is limited to the G2019S mutation, or has broader implications for LRRK2-induced Parkinson’s disease.   

     

    View all comments by Gaia Skibinski

  2. The most important finding of the paper, in my view, is the biochemical and genetic demonstration that aberrant regulation of protein synthesis contributes to LRRK2-G2019S pathogenesis. This resonates with earlier studies in Drosophila Parkinson's disease (PD) models that link altered mRNA metabolism and translation to PD pathogenesis.

    The authors argue that the ribosomal protein s15 is a physiological substrate of LRRK2. The strongest evidence is that s15 phosphorylation is reduced by ~50 percent in the Drosophila LRRK2 null mutant. Clearly other unidentified kinase(s) can also phosphorylate s15. What is missing from the paper is whether s15 phosphorylation is reduced when LRRK2 is knocked out or knocked down in mammalian cells, and whether wild-type, phospho-deficient, and phospho-mimetic s15 exhibit differential activities in rescuing the cellular phenotypes observed in LRRK2 knockout animals, such as apoptosis of kidney cells. This is an important point, because the current controversy regarding whether 4E-BP is a genuine substrate of LRRK2 was caused by the inability  to confirm in mammalian cells the findings derived from fly models.

    However, I don't think that the authors have completely convinced me that LRRK2-G2019S phosphorylation of s15 triggers neuronal injury by ramping up translation. One key piece of data presented was that the phospho-mimetic s15, called TD, is toxic to cultured mammalian neurons when overexpressed. However, the authors show no data that TD s15 promotes bulk translation in these neurons, or that the toxicity of TD s15 can be blocked by partial inhibition of global translation. More importantly, they did not show whether TD s15 promotes translation and is toxic to dopaminergic neurons in vivo in transgenic animals.

    The reliance on overexpression in cells throughout this study could be a cause of concern. Human neurons derived from LRRK2-G2019S patient iPSCs are now available. It would be interesting to test whether altered s15 phosphorylation is responsible for the disease-relevant phenotypes observed in these neurons. Neurons derived from LRRK2-GS knock-in animals can also be used for this purpose. With regard to the levels of LRRK2 autophosphorylation being far higher than phosphorylation of s15, I don't think that observation alone can be used to argue against s15 being a bona fide LRRK2 substrate. Phosphorylation of 4E-BP by LRRK2 in vitro is also weaker than LRRK2 autophosphorylation. It could be that the in-vitro kinase assay condition favors LRRK2 autophosphorylation, which may not be the case in vivo, or that LRRK2 intrinsically has higher autophosphorylation activity compared to other kinases

    As we have discussed in our previous publications, increased translation could lead to neuronal injury through a number of mechanisms. First, given that protein synthesis is a highly energy-demanding process, stimulation of protein translation by pathogenic LRRK2 could perturb cellular energy and redox homoeostasis. This could be especially detrimental in aged cells or stressed post-mitotic cells such as dopaminergic neurons where energy reserve is already low. Second, increased protein synthesis could lead to the accumulation of misfolded or aberrant proteins, overwhelming the already-compromised ubiquitin proteasome and molecular chaperone systems in aged or stressed cells. Third, altered translation by pathogenic LRRK2 kinase may compromise synapse structure and function, which is known to involve regulated local protein synthesis. Deregulation of this process could lead to synaptic dysfunction and eventual neurodegeneration.

    The authors have tried to distinguish their study from our previous studies on LRRK2 and eukaryotic initiation factor 4E (eIF4E) binding protein (4E-BP), in which we made the first link between altered translation and LRRK2 pathogenesis (see Imai et al., 2008). Unfortunately, their data on LRRK2 and 4E-BP are all embedded in supplementary information and many details of their experiments are not available.

    First, the source of their p-4E-BP and 4E-BP antibodies is not clear. The p-4E-BP western blots look very different from those in published works, which generally show multiple bands of 4E-BP because 4E-BP is phosphorylated in many sites. The p-4E-BP blots in this paper show only a single band.

    Secondly, the result showing lack of 4E-BP phosphorylation by LRRK2-GS in flies is subjected to alternative explanation. The authors used a cell-type specific Ddc-Gal4 driver to express LRRK2-GS in dopaminergic neurons, but used whole fly head extracts to detect p-4E-BP. Assuming their antibody is specific to p-4E-BP, since LRRK2-GS is only expressed in ~200 dopaminergic neurons in an adult fly head composed of millions of neurons and non-neuron cells that all express endogenous LRRK2 and 4E-BP, it is unlikely that any stimulating effect of LRRK2-GS on 4E-BP phosphorylation can be observed. A ubiquitous Gal4 driver should be used to drive LRRK2-GS expression.

    Thirdly, the reporter experiment showing that the microRNA pathway does not mediate LRRK2-GS effects on mRNA translation is misleading. The effect of LRRK2 on the micro-RNA pathway is specific to particular microRNAs and mRNA targets (Gehrke et al., 2010). Based on the limited information described for their bicistronic reporter, it appears to have no miRNA binding sites. If so, one cannot expect manipulating the microRNA pathway to have any effect on reporter expression.

    Fourthly, the genetic experiment showing that heterozygosity of eIF4E has no effect on LRRK2-GS toxicity by no means disproves the model that 4E-BP phosphorylation is relevant to LRRK2-GS toxicity. The hypomorphic eIF4E alleles only show partial reduction of eIF4E mRNA expression. No information is available as to whether this affects eIF4E protein expression or cap-dependent translation. A better way to  test the model would be to express WT and phospho-mutant forms of 4E-BP in LRRK2-GS background to see if there is genetic interaction, as we did  previously (Imai et al., 2008). 

    Finally, we carried out our previous studies on LRRK2 phosphorylation of 4E-BP using cell lines or transgenic flies expressing LRRK2-I2020T, whereas this study and the other cited mammalian studies all used LRRK2-GS. Although it is generally assumed that the T2020T and G2019S mutations all boost LRRK2 kinase activity, it may not be correct to assume that they act through the same pathogenic mechanism. Moreover, the phosphorylation sites in 4E-BP acted on by LRRK2 are also known targets for mTORC1, which are hyperactive in cultured mammalian cells. In order to observe 4E-BP phosphorylation by LRRK2 in mammalian cells, we used serum-starvation treatment to inhibit Ins/PI3K/mTORC1 signaling. Based on our reading of this study and the other published studies trying to duplicate our past work, it appears that this point was overlooked.

    Thus, until the right experiments using the appropriate reagents are performed in mammalian cells, it is premature to claim that the LRRK2-4E-BP connection is dead. In fact, the more scientific way to consider this study and previous work is that LRRK2 may act on more than one substrate (4E-BP, s15, etc) to regulate translation. Future studies exploring the relationship and relative contributions of 4E-BP and s15 phosphorylation to LRRK2 toxicity will test this hypothesis.

    References:

    Imai Y, Gehrke S, Wang HQ, Takahashi R, Hasegawa K, Oota E, Lu B. Phosphorylation of 4E-BP by LRRK2 affects the maintenance of dopaminergic neurons in Drosophila. EMBO J. 2008 Sep 17;27(18):2432-43. PubMed.

    Gehrke S, Imai Y, Sokol N, Lu B. Pathogenic LRRK2 negatively regulates microRNA-mediated translational repression. Nature. 2010 Jul 29;466(7306):637-41. PubMed.

    View all comments by Bingwei Lu

Make a Comment

To make a comment you must login or register.

References

Alzpedia Citations

  1. Leucine-rich repeat kinase 2 (LRRK2)

Paper Citations

  1. Tsika E, Moore DJ. Mechanisms of LRRK2-mediated neurodegeneration. Curr Neurol Neurosci Rep. 2012 Jun;12(3):251-60. PubMed.
  2. Kumar A, Greggio E, Beilina A, Kaganovich A, Chan D, Taymans JM, Wolozin B, Cookson MR. The Parkinson's disease associated LRRK2 exhibits weaker in vitro phosphorylation of 4E-BP compared to autophosphorylation. PLoS One. 2010 Jan 15;5(1):e8730. PubMed.
  3. Trancikova A, Mamais A, Webber PJ, Stafa K, Tsika E, Glauser L, West AB, Bandopadhyay R, Moore DJ. Phosphorylation of 4E-BP1 in the mammalian brain is not altered by LRRK2 expression or pathogenic mutations. PLoS One. 2012;7(10):e47784. Epub 2012 Oct 17 PubMed.

Further Reading

Papers

  1. Dauer WT, Guo M. Multiplying messages LRRK beneath Parkinson disease. Cell. 2014 Apr 10;157(2):291-3. PubMed.

Primary Papers

  1. Martin I, Kim JW, Lee BD, Kang HC, Xu JC, Jia H, Stankowski J, Kim MS, Zhong J, Kumar M, Andrabi SA, Xiong Y, Dickson DW, Wszolek ZK, Pandey A, Dawson TM, Dawson VL. Ribosomal protein s15 phosphorylation mediates LRRK2 neurodegeneration in Parkinson's disease. Cell. 2014 Apr 10;157(2):472-85. PubMed.

Annotate

To make an annotation you must Login or Register.