We are grateful for the interest in our paper describing EndoA as a LRRK2 target in synaptic recycling, and would like to respond to comments raised on Alzforum regarding our work.

The model we propose, where both too much and too little LRRK2-dependent EndoA1 phosphorylation would result in reduced endocytosis, explains our data. It is also in line with results other groups obtained specifically with rat neurons. Shin et al., 2008 report that the rate of endocytosis is reduced both when LRRK2 is knocked down using shRNA or by expressing the kinase active LRRK2G2019S. We would like to note that some confusion exists about the effects of different mutations in LRRK2 causing PD. The G2019S mutation is clearly a gain of function, enhancing phosphorylation. The others have unclear effects on LRRK2 kinase function. Our model, in which fine-tuning of activity is important, provides at least an interesting new perspective on the discussion of whether gain or loss of function of the kinase activity is critically involved in the...  Read more

We are grateful for the interest in our paper describing EndoA as a LRRK2 target in synaptic recycling, and would like to respond to comments raised on Alzforum regarding our work.

The model we propose, where both too much and too little LRRK2-dependent EndoA1 phosphorylation would result in reduced endocytosis, explains our data. It is also in line with results other groups obtained specifically with rat neurons. Shin et al., 2008 report that the rate of endocytosis is reduced both when LRRK2 is knocked down using shRNA or by expressing the kinase active LRRK2G2019S. We would like to note that some confusion exists about the effects of different mutations in LRRK2 causing PD. The G2019S mutation is clearly a gain of function, enhancing phosphorylation. The others have unclear effects on LRRK2 kinase function. Our model, in which fine-tuning of activity is important, provides at least an interesting new perspective on the discussion of whether gain or loss of function of the kinase activity is critically involved in the disease.

With regard to kidney phenotypes and potential other pathways in which LRRK2 might be involved, we agree that LRRK2 is likely to phosphorylate other substrates in addition to EndoA and that this even might be cell-type dependent. As such we expect that LRRK2 may impinge on other pathways; however, we anticipate that in these pathways, as well, LRRK2 will have a “tuning” effect. We are quite confident that our studies in neurons are important to unravel the role of LRRK2 in the central nervous system, and that the disturbed neurotransmission phenotype we study is relevant for the understanding of the neurobiological role of LRRK2 in general. Such studies might well turn out to be relevant for Parkinson’s disease. We recommend the News and Views article by Heutinck and Verhage. It clearly explains our strategy, which focuses first on the synaptic role of LRRK2, especially since little is known about the function of this protein.

It is correct that LRRK2 is broadly expressed in neurons; however, this feature per se does not make a role of LRRK2 in endocytosis necessarily indirect. Again, we think that LRRK2 acts in other pathways. Given the mild phenotypes in the deficient flies, mice or other species, it likely acts in a regulating and fine tuning way as we describe here for EndoA. Our work indicates beyond doubt that fruit fly LRRK or mammalian LRRK2 in vivo and in vitro are both necessary and sufficient for EndoA(1) phosphorylation.

EndoA(1) has been firmly implicated in the actual formation and uncoating of synaptic vesicles (Verstreken et al., 2003 ; Milosevic et al., 2011). The most parsimonious explanation follows that LRRK2 acts via EndoA in endocytosis. The remarkable observation that EndoA1-3 mouse triple knockouts result in neurodegeneration, yet other mouse knockouts in endocytic genes do not (P. De Camilli personal communication) requires further investigation, also in relation to other clinical LRRK2 mutants than the G2019S that we used in our studies.

We take issue with the statement that the fly LRRK data ought to be interpreted with caution because mammalian LRRK2 would not be present at presynaptic terminals. In Mandemakers et al., 2012, we do not report that LRRK2 is absent from synapses. Although it is not enriched, LRRK2 is present at synapses and is detected in the synaptic vesicle fraction after subcellular fractionation. Synaptic localization of LRRK2 has previously been shown by many others (i.e. Piccoli et al., 2011, Stafa et al., 2012). Mammalian LRRK2 rescues all the synaptic endocytosis defects that we report in Drosophila, indicating the mammalian protein is able to act in the pathways affected by loss of fly LRRK. Noteworthy, our loss of function studies in LRRK null mutant flies are performed in the absence of potential compensation of LRRK1 that may confound interpretation of synaptic phenotypes in mammalian systems. Finally, other researchers have found synaptic transmission defects in LRRK2 deficient mammalian neurons that are in line with the data we reported (Piccoli et al., 2011; Shin et al., 2008).

Given the extensive conservation of endocytic proteins and mechanisms across species (Lloyd et al., 2000; Littleton et al., 2000), our observation that human LRRK2 can phosphorylate fly and human EndoA1 indicates conservation of this regulatory mechanism, as well. We would like to remind Alzforum’s readers of the power of Drosophila genetics to unravel fundamental, conserved biological pathways. The field of neurodegeneration is replete with examples where salient novel features were originally discovered in small genetic models such as the fruit fly and the worm.

In conclusion we hope that future work by our lab and others will rigorously test our hypothesis on the role of LRRK2 in synaptic vesicle endocytosis and how this pathway contributes to the development of PD.

View all comments by Bart De Strooper
View all comments by Wim Mandemakers
View all comments by Samer Matta
View all comments by Patrick Verstreken