A collection of familial PD knock-in models, related to Supplemental figure 1.

(a) Scheme of the knock-in strategy where the first common exon in all Drosophila transcripts (left) was replaced by an attP-flanked mini-white gene using CRISPR/Cas9 mediated homologous recombination, creating a null mutant, and then replaced by wild-type or pathogenic mutant human or fly CDS using PhiC31 mediated integration (right). The chromosomal positions of the genes are indicated. All knock-ins are in the same isogenic genetic background (cantonized w1118). (b) Phenotypic analysis of PD mutants (1) Mitochondrial membrane potential measured by ratiometric TMRE fluorescence at neuromuscular junction (NMJ) boutons in 3rd instar larvae. n≥4 animals per genotype and 10 boutons from ≥3 NMJs per animal. (2) Depolarisation amplitude quantified from electroretinograms (ERGs), recorded after 7 days of light exposure. n≥20 animals per genotype. (3) Quantification of startle-induced-negative geotaxis (SING) at 5±1 d after eclosion (young) and 25±1 d after eclosion (old). 95% of pink1 mutants died < 25 d and were tested at 15 d. Values are normalized to control (see methods). Variance of control measurements are in grey. Bars are mean ± s.e.m; *, p<0.05 ANOVA/Dunnett’s test.

Single-cell RNA sequencing reveals that olfactory projection neurons are impaired across young PD fly models, related to Supplemental figure 2.

(a) tSNE of 226k cells characterized by single-cell RNA sequencing of entire young (5 day old) fly brains (controls and PD knock-in mutants). Colors indicate the cell types; 186 of which were identified (See methods). Key cell types are encircled, including dopaminergic neurons (DAN, blue), Mushroom body Kenyon cells (orange) and olfactory projection neurons (OPN, green). (b-f) tSNE of the cells from 5 selective PD knock-in mutants (5 day old). Black cells are those with a significant transcriptomic change. Key cell types labeled in (a) are indicated. Note that OPN are consistently affected across mutants, while DAN are not at this early stage.

Common DEGs in cholinergic neuron-rich brain regions of human PD patients and PD fly models, related to Supplemental figure 3.

(a-a”) Schematic of the sunburst plot indicating GO terms for each sector (a) and the mapping of the DEGs in nucleus basalis of Meynert (NMB), nucleus accumbens and putamen brain samples idiopathic PD patients (with LRRK2 risk mutations) and controls (a’) and mapping of the DEGs found commonly in fly and human samples (a”). Inner rings represent the different GO categories (indicated in a), with their subcategories in the outer rings, rings (in a’-a”) are color-coded according to enrichment Q-value. (b) Gene Ontology analysis of DEG in cholinergic neurons of young PD fly models and NBM neurons of PD patients. Redundant terms were removed. Color: adjusted p-value. (c) Schematic of the DEGs found commonly in fly PD models (blue) and human PD samples (black) manually sorted according to their previously described synaptic functions. Supplemental File 7 contains the summarized results of the DEG analysis of fly brains and postmortem human brain samples and the SynGO analysis.

Synaptic defects in OPN of PD mutants.

(a) Schematic of a head-fixed awake fly for live Ca2+-imaging through a window in the head capsule. (b) Confocal image of GCaMP3-fluorescence expressed in OPN using GH146-Gal4, indicating the locations of the cell bodies, antennal lobe and the calyx (Scale bar: 50 µm). (c-d) Confocal image of the stimulus-induced change of fluorescence (peak amplitude) in the synaptic region of the calyx of a control and a hLRRK2G2019S knock-in animal (c) and quantification of fluorescence change (± SEM) over time (d); Arrowhead: time of stimulus application (10 mM Nicotine, See methods). (e, f) Images of GFP fluorescence marking the synaptic area of OPN in the calyx of control (e) and hLRRK2G2019S knock-ins (f); Scale bar is 20 µm. (g) Quantification of GCaMP3 peak amplitude at OPN synapses in the calyx following stimulation (10 mM Nicotine) in controls, wild-type CDS knock-ins, and in the PD knock-in mutants where the wild-type CDS is not (-) or is (+) expressed in OPN using GH146-Gal4 (OPN>wt CDS). Note that the hPINK1 control could not be determined as the combination of nSyb-Gal4>UAS-hPINK1 (expression in all neurons) in the Pink1 knock-out background interferes with the OPN-specific expression of Gal4 to drive UAS-GCaMP3 expression (Fig. 4g (left bar; “nd”). In contrast, this issue is not present in hPINK1P399L mutant knock-in flies (nor the other flies used in the study) that could be rescued by OPN-specific expression of hPINK1 (Fig. 4g (right bar); “OPN>wt CDS +”). (h) Quantification of the GFP fluorescence area of OPN synapses in the calyx (based on GCaMP3 signal) in controls, wild-type CDS knock-ins, and in the PD knock-in mutants where the wild-type CDS is not (-) or is (+) expressed in OPN (OPN>wt CDS). For (g, h): n≥5 animals per genotype. *, p<0.05 in ANOVA/Dunnett.

OPN dysfunction causes hyposmia in PD mutants.

(a, b) Olfactory performance of PD knock-in flies and controls when given the choice between a blend of motor oil and banana odors (a) or a blend of beer and wine odors (b). (c, d) Olfactory performance of PD knock-in flies and LRRKKO when the relevant wildtype CDS is not (–) or is expressed selectively in OPN (“OPN>wt CDS +”, green label) or T1 cholinergic interneurons (“T1>hRab39B +”, red label), as a negative control. For (a-d): Bars are mean ± s.e.m. n≥5 assays, ≥200 flies per genotype. *, p<0.05 in ANOVA/Dunnett.

Cholinergic neuron activity rescues dopaminergic defects that occur at later stages in the life of PD mutants, related to Supplemental figure 4.

(a, left) Synaptic area of DAN innervating the mushroom body in aged PD models (AuxR927G, synjR258Q or LRRK2G2019S and with or without GH146-Gal4 driven expression of the wildtype PD gene (aux or synj) or EndoAS75D, respectively. Bars: mean and SEM. n≥5, *p<0.05 in ANOVA, Dunnett’s test. (a, right) SING of the PD models with or without GH146-Gal4 driven expression of wildtype gene or endoAS75D. Mean and SEM. n≥5, * p<0.05 in 2-way-ANOVA. Grey zone: variance of controls. (b, left) Odor choice performance, stimulus-induced changes in synaptic Ca2+ signal and OPN synapse area of young controls and hLRRK2G2019S flies with or without chronic nicotine feeding (up to one day before testing). Bars: mean and SEM. n≥5 assays, * p<0.05 in ANOVA, Dunnett’s test. (c) SING, stimulus-induced changes in synaptic Ca2+ and DAN synapse area of aged controls and hLRRKG2019S flies with or without chronic application of nicotine. Bars: mean and SEM. n≥5 assays, * p<0.05 in ANOVA, Dunnett’s test. (d) Confocal images of differentiated (60 days) wildtype and LRRK2G2019S ventral midbrain DAN labelled with the ventral midbrain marker FOXA2, dopaminergic marker TH and neuronal marker MAP2. Scale bar: 20 µm. (e) Scheme of the treatment protocol and spontaneous Ca2+ activity (e’) and amplitude (e’’) of human induced DAN, 2 days after a 20-day nicotine or nicotine + mecamylamine treatment. Bars: mean and SEM. n≥60 DAN from 3 independent differentiations, * p<0.05 in ANOVA, Dunnett’s test.