Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: Following 15 days of NSC differentiation and maturation, immunocytochemistry confirmed the presence of mature neural networks in the microfluidic chips. A ) Tiled image of a fluorescently labelled cortical neural network structured within a microfluidic chip, overlaid by a schematic of the design. B ) Brightfield image of the developing neural network in a microfluidic chip, showing the area containing the axon tunnels, synaptic compartment and active zones, while C ) Brightfield image from the cell chamber area. D ) F luorescently labelled LRRK2 neural network with markers for neurons (MAP2, green), neuron specific microtubules (beta-III tubulin, red), and kainic acid receptors (GRIK5, magenta) together with the counterstain Hoechst (blue), with E) showing equivalent markers in a control neural network (10um scale bar). The remaining images show the neural networks fluorescently labelled with markers for F ) neurons (MAP2, red) expressing (CaMKII, green), G ) with presynaptic vesicles (Piccolo, green), postsynaptic densities (PSD95, red) and F-actin (Phalloidin, blue) expressed in the axon tunnels and synaptic area, H ) neuronal specific microtubules (beta-III tubulin, red) together with CaMKII (green), and I ) neurofilament heavy (green) together with AMPA receptors (red) and Hoechst counterstain. 50μm scale bars.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques: Immunocytochemistry, Expressing

Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: A ) Microfluidic chip design, with (1) indicating the top cell chamber/node, (2) the inlet/outlet connected via the synaptic compartment, and (3) the axonal/dendritic tunnels. Directionality of the inter-nodal connectivity is imposed on the network through different tunnel lengths, where mainly the axons from the top and bottom cell chambers connect to the dendrites and axons from the middle cell chamber in the two synaptic compartments. B ) Outline of microfluidic chip interfaced with the custom made multielectrode array for electrophysiological investigation. C ) Electrode layout in the area connecting the bottom and middle cell chamber through axonal/dendritic tunnels and the synaptic compartment. D ) Timeline of the experiment, indicating the time points and assays used to assess the LRRK2 and control neural networks.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques:

Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: A ) Microfluidic chip design, with a red circle indicating the top cell chamber used for confined KA stimulation (10μM). B ) Line-graph where the reactive oxygen species (ROS) production of a structured neural network following a targeted KA stimulation has been analysed, with a clear difference between the ROS production in the top cell chamber compared to the two other chambers. Images (10X) were taken from each of the cell chambers every 5 minutes over the course of one hour following a KA stimulation. C, D and E ) Representative images of fluorescently labelled ROS from each of the cell chambers, 45 minutes after the targeted KA stimulation. F ) Tiled image of the middle cell chamber of a structured neural network fluorescently labelled with Calcein-Am (green) and Ethidium homodimer-1 (red) 24 hours post KA. G ) Close-up of the area selectively chosen for analysis. H ) Bar-graph showing the percentage of viable cells counted in each chamber, for each condition with standard deviation bars. A statistically significant difference was found between the control and the LRRK2 groups in the PBS condition (p=0.0004, N1=N2=12, DF=18.17) by post hoc Tukey’s multiple comparisons test, where N= the number of images counted. No significant difference was found between the KA and control condition within the same group, nor between the different chambers, demonstrating that the KA stimulation was sublethal as intended.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques: Standard Deviation

Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: The structured neural networks were labelled with TMRM, and the active mitochondria contained within 4 axonal tunnels were imaged live at two time points (baseline and after KA or PBS) for 3 neural networks from both groups, i.e. the control and LRRK2 neural networks. Additionally, image z-stacks were taken from the synaptic compartment at baseline for both groups to capture the height span of the active mitochondria. A and B ) Volumetric view of the area containing fluorescently labelled mitochondria in the synaptic compartment from an A ) LRRK2 neural network (height = 44μm) and a B ) control neural network (height = 9μm). One of the z-slices from each of the stacks making up the volumetric figure in A and B are shown in C and D , respectively. Some autofluorescence in the PDMS walls of the microfluidic chips outline the structure of the synaptic compartment and axon tunnels (30μm scale bars). E ) shows a bar-graph with scatter plots of the mean height in the synaptic compartment measured to contain fluorescently labelled mitochondria in each group, with standard deviation bars. An independent samples t-test showed that the LRRK2 neural networks contained active mitochondria within 3 times the height span of the control neural networks (p<0.0001, N=6) where N equals the number of networks investigated in each group. G ) shows fluorescently labelled mitochondria contained within a single axonal tunnel in a control neural network, at baseline (20μm scale bar). F ) Bar-graph with the median number of mitochondria counted within single axonal tunnels at baseline for both control and LRRK2 neural networks, with range bars and scatter plots. The LRRK2 neural networks displayed significantly more TMRM labelled mitochondria contained within the axonal tunnels compared to the control neural networks at baseline (p=0.0114, U=659.5, N 1 =48, N 2 =40) by Mann-Whitney U test. H and I ) Bar-graphs with the median number of mitochondria measured at each timepoint, with range bars and scatter plots, for both the PBS and KA stimulated condition, within both groups. H ) The control neural networks were found by Wilcoxon matched-pairs signed rank test to have significantly fewer active mitochondria after KA stimulation compared to the baseline (p=0.0432, pairs=24), while the LRRK2 neural networks I ) showed the same trend without it being statistically significant. A statistically significant difference was found between the two timepoints in the PBS condition of the LRRK2 neural networks however (p=0.029, pairs=16), with more active mitochondria being measured after PBS addition.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques: Standard Deviation, MANN-WHITNEY

Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: Images were taken from the synaptic compartment area and show cortical neural networks fluorescently immunolabelled with Piccolo (green) and PSD95 (red). A and B Representative images from the two conditions (KA and PBS) in the control neural networks, where A ) shows immunolabelling in the synaptic compartment of a KA stimulated neural network and B ) from the synaptic compartment of a PBS neural network. Both images are enlarged (with representative full-view images in the top right corner) (10μm scale bars). E ) Similarly, representative image from the KA stimulated condition in an LRRK2 neural network (20μm scale bar), illustrating the lower level of morphological detail available due to the density of neurites contained within the synaptic compartment. C ) Bar-graph with the median ratio of neuritic boutons contained within the synaptic compartment for each condition in the control neural networks, with range bars. A statistically significant reduction of neuritic boutons was found in the KA stimulated condition (Mann-Whitney U=73, n 1 =22, n 2 =18, p=0.0004) compared to the PBS condition, where n equals the number of images analysed. D and F ) Bar-graphs with the median size of the synaptic contacts (Piccolo/PSD95 co-occurrence) with range bars, measured within the synaptic compartment for both the control and LRRK2 neural networks, respectively. For the control neural networks, a statistically significant difference was found in the synaptic size measurements between the conditions (Mann-Whitney U=153.5, n 1 =26, n 2 =24, p=0.0017), with much larger areas of cooccurrence between Piccolo and PSD95 found at the neurites of the KA stimulated condition compared to the PBS condition. For the LRRK2 neural networks, no significant difference was found between the conditions in synaptic contact size. Furthermore, to illustrate the difference in neuritic density G ) shows 10μm thick z-stack volume projections of the LRRK2 neural network in E , with Piccolo and PSD95 merged (top), followed by Piccolo (green) and PSD95 (red) alone.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques: MANN-WHITNEY

Journal: bioRxiv

Article Title: Structural and functional alterations associated with the LRRK2 G2019S mutation revealed in structured human neural networks

doi: 10.1101/2020.05.02.073726

Figure Lengend Snippet: A ) Bar graph of mean MFR with standard deviations measured for each group at the three different experimental timepoints (baseline, stimulation, 24hours post), with each group consisting of measurements from 3 different networks in each condition (PBS and KA) from both the control and LRRK2 group. The average MFR of the LRRK2 neural networks are consistently higher at all timepoints compared to the control neural networks, particularly at baseline. B ) The same data as in A ) but plotted as a line graph displaying the variation of the measures from the baseline condition. 24 hours post KA stimulation both the LRRK2 and control neural networks display a drop in MFR, however, the difference from the baseline is much larger for the control networks. C ) Network overexcitation induced by the KA stimulation. The graph on the left shows the activity measured at a single electrode located in the top cell chamber of an LRRK2 neural network during the KA stimulation, with the red line indicating the threshold set for spike detection. The firing rate (Hz) profile of a single neuron recorded by the same electrode is plotted on the right (see Suppl.fig.4 for spike sorting details), where a drastic increase in firing can be observed from 350-500 seconds into the recording, followed by an abrupt activity drop. D ) Line graph of the relative total network correlation change measured for each group at the three experimental timepoints in relation to the baseline values. At the 24 hours post KA or PBS addition timepoint, a similar increase in total network correlation change can be observed for the PBS condition of both the LRRK2 and control neural networks (+44-48%) relative to the baseline, while the KA stimulated networks display a similar decrease in total network correlation (−25-27%). E-G ) Schemaball correlation maps of a representative LRRK2 neural network during the baseline, KA stimulation, and 24 hours post, respectively. Colour intensity of the lines interconnecting each channel indicates the correlation between their activity (from r 0-1), with 0 being black and 1 being bright yellow.

Article Snippet: Control human induced pluripotent stem cell (iPSC)-derived H9N neural stem cells (NSCs) (ax0019) and iPSC derived H9N NSCs homozygously carrying the LRRK2 G2019S (GGC>AGC) mutation (ax0310) (Axol Bioscience, Cambridge, United Kingdom) were cultured and expanded on 0.01 % poly-L-ornithine (PLO) (Sigma) and L-15 laminin (L15 medium containing 1:60 natural mouse laminin and 1:41 sodium bicarbonate) coated culture vessels in neural expansion medium (ax0030) supplemented with human FGF2 and EGF (ax0047 and ax0047X), and kept in a standard humidified air incubator (5% CO 2 , 20%O 2 , 37°C) (full cell culture protocol, as well as further information on each cell line available in the supplementary data).

Techniques: Activity Assay