Hyperactive LRRK2 kinase impairs the trafficking of axonal autophagosomes

ABSTRACT

Parkinson disease (PD)-causing mutations in the LRRK2 (leucine rich repeat kinase 2) gene hyperactivate LRRK2 kinase activity. Here, we discuss our recent work linking LRRK2 hyperactivation to defective axonal autophagosome transport in neurons. In three different models, we observed that expression of the most common causative mutation for PD, LRRK2^G2019S, disrupts processive autophagosome transport in a kinase-dependent manner. Mechanistically, we found that hyperactive LRRK2 recruits SPAG9/JIP4, a motor adaptor known to bind to LRRK2-phosphorylated RAB proteins, to the autophagosomal membrane. Increased SPAG9/JIP4 levels induce abnormal recruitment and activation of kinesin-1, which we propose results in an unproductive tug-of-war between anterograde and retrograde motors bound to autophagosomes. Disruption of autophagosome transport correlates with defective autophagosome maturation, suggesting that hyperactive LRRK2 may impair efficient degradation of autophagosomal cargo. Our work demonstrates that LRRK2 hyperactivation is sufficient to induce defects in autophagosome transport and maturation, further establishing a role of defective autophagy in the pathogenesis of PD.

Mutations in the LRRK2 (leucine rich repeat kinase 2) gene are the most common genetic cause of Parkinson disease (PD). Pathogenic LRRK2 mutations hyperactivate LRRK2 kinase activity, but it is unclear how hyperactive LRRK2 causes neurodegeneration. Previous studies have established a subset of RAB proteins, key regulators of intracellular transport, as robust substrates of LRRK2 kinase. This raises the possibility that defects in intracellular trafficking contribute to LRRK2-mediated neurodegeneration. In our recently published work [1], we investigated the effects of LRRK2 hyperactivation on neuronal autophagy, a degradative pathway that is essential for neuronal health and highly dependent on efficient axonal transport of autophagic vesicles (AVs).

In neurons both in vitro and in vivo, AVs constitutively form in the distal axon, containing cargos such as aggregated proteins and dysfunctional organelles. Once formed, AVs are transported to the soma by the molecular motor cytoplasmic dynein. The opposing motor kinesin-1 is also bound to axonal autophagosomes but remains inhibited, allowing for the smooth, processive retrograde movement of these organelles. During this transport, AVs fuse with lysosomes, maturing into degradative organelles.

We investigated how axonal AV transport is affected by the most frequent pathogenic LRRK2 mutation, LRRK2^G2019S, using three different model systems: rat hippocampal neurons overexpressing LRRK2, cortical neurons from a G2019S knockin (KI) mouse, and human iPSC-derived neurons gene-edited to express the G2019S mutation. In all three models, we observed that LRRK2^G2019S significantly disrupts the processivity of axonal AV transport by increasing both the frequency of pausing and the number of reversals during retrograde organelle motility. Impairment of AV transport is dependent on hyperactive LRRK2 kinase activity, as either expression of a kinase-inactive double mutant or application of the LRRK2 kinase inhibitor MLi-2 rescues the transport defect. In contrast to these effects on AV transport, expression of LRRK2^G2019S does not affect the transport of LAMP1-positive late endosomes/lysosomes and does not alter axonal microtubule dynamics, suggesting that hyperactive LRRK2 induces a specific impairment of AV transport.

How does hyperactive LRRK2 interfere with processive AV transport? We found that AVs isolated from brains of G2019S KI mice have increased levels of the motor adaptor protein SPAG9/JIP4. Increased levels of SPAG9/JIP4 correlate with higher levels of the anterograde motor kinesin-1 on G2019S KI AVs. A microtubule pelleting assay showed increased levels of activated kinesin in lysates from G2019S KI MEFs. Further, overexpression of SPAG9/JIP4 in mouse cortical neurons is sufficient to disrupt axonal AV transport in live-imaging experiments. Why is SPAG9/JIP4 erroneously recruited to the AV membrane following LRRK2 hyperactivation? We found that isolated AVs from G2019S KI mice have increased levels of LRRK2-phosphorylated RAB proteins, which bind to SPAG9/JIP4. Therefore, we propose a model, in which hyperactive LRRK2 recruits SPAG9/JIP4 to the AV membrane through its binding to LRRK2-phosphorylated RAB proteins. SPAG9/JIP4 then causes an abnormal activation of the anterograde motor kinesin, leading to an unregulated tug-of-war with the retrograde motor dynein and disrupting the normal pattern of processive retrograde AV transport (Figure 1).

Figure 1.

Model depicting the effect of hyperactive LRRK2 on AV transport and maturation. Normal LRRK2 activity allows for processive retrograde AV transport, facilitating en route fusion with lysosomal vesicles. Different motor adaptors (not shown) inhibit kinesin and promote dynein activity, resulting in processive retrograde motility. Hyperactive LRRK2 enhances recruitment of SPAG9/JIP4 to the autophagosomal membrane via binding to LRRK2-phosphorylated RABs, leading to an abnormal activation of kinesin and causing an unproductive tug-of-war between anterograde and retrograde motors. Disruption of processive AV transport is accompanied by inefficient autophagosome-lysosome fusion and impaired AV acidification

Further studies will be required to determine which specific LRRK2-phosphorylated RAB protein, or combination of RAB proteins, mediates SPAG9/JIP4 recruitment to the AV membrane. It will also be interesting to investigate if/how binding of SPAG9/JIP4 to phospho-RABs affects the coordinated regulation of the dynein and kinesin motors bound to AVs. These mechanistic studies will shed more light on the integrated regulation of AV trafficking in neurons.

Another intriguing question is how the observed impairment of AV transport may contribute to neurodegeneration in PD. Previous reports have found that efficient autophagosome-lysosome fusion in the axon is dependent on retrograde AV transport. Using an mCherry-EGFP-LC3 tandem construct to assess AV acidification, we observed a lower percentage of acidified AVs in the proximal axon of G2019S KI neurons, suggesting that disruption of AV transport by LRRK2^G2019S impairs maturation of axonal autophagosomes. We speculate that deficient transport and inefficient acidification cause a delay in the degradation of cargos engulfed by axonal AVs. This delay may, over time, lead to the accumulation of protein aggregates and/or dysfunctional organelles in the axon. It is interesting to note that SNCA/alpha-synuclein, a key protein in PD pathology, is an established substrate of autophagy and predominantly found in the distal axon, the primary location of AV biogenesis. Multiple groups have reported exacerbated formation of SNCA aggregates in axons of LRRK2^G2019S neurons after treatment with pre-formed fibrils of SNCA. One possible explanation is that the observed defect in AV transport impairs efficient clearance of SNCA from the axon, making LRRK2^G2019S neurons more vulnerable to aggregate formation when seeded by pre-formed fibrils.

In summary, our study links pathogenic hyperactivation of LRRK2 to defects in neuronal AV transport. The impairment of AV transport induced by LRRK2^G2019S is remarkably consistent across three different model systems. Downstream of hyperactive LRRK2, our work implicates recruitment of the motor adaptor SPAG9/JIP4 in the disruption of AV transport, inducing abnormal activation of kinesin and thereby causing an unproductive tug-of-war between anterograde and retrograde motors bound to AVs. Our findings show that hyperactivation of LRRK2 causes defects in transport and maturation of neuronal autophagosomes, further establishing autophagy as a key pathway in PD pathogenesis.

Funding Statement

This work was supported by the German Research Foundation (BO 5434/1-1 to C.A.B.), the National Institutes of Health (R37 NS060698 to E.L.F.H.), and the Michael J. Fox Foundation (grant 15100 to E.L.F.H.).

Disclosure statement

No potential conflict of interest was reported by the author(s).