Developmental effect of RASopathy mutations on neuronal network activity on a chip

RASopathies are a group of genetic disorders caused by mutations in genes encoding components and regulators of the RAS/MAPK signaling pathway, resulting in overactivation of signaling. RASopathy patients exhibit distinctive facial features, cardiopathies, growth and skeletal abnormalities, and varying degrees of neurocognitive impairments including neurodevelopmental delay, intellectual disabilities, or attention deficits. At present, it is unclear how RASopathy mutations cause neurocognitive impairment and what their neuron-specific cellular and network phenotypes are. Here, we investigated the effect of RASopathy mutations on the establishment and functional maturation of neuronal networks. We isolated cortical neurons from RASopathy mouse models, cultured them on multielectrode arrays and performed longitudinal recordings of spontaneous activity in developing networks as well as recordings of evoked responses in mature neurons. To facilitate the analysis of large and complex data sets resulting from long-term multielectrode recordings, we developed MATLAB-based tools for data processing, analysis, and statistical evaluation. Longitudinal analysis of spontaneous network activity revealed a convergent developmental phenotype in neurons carrying the gain-of-function Noonan syndrome-related mutations Ptpn11D61Y and KrasV14l. The phenotype was more pronounced at the earlier time points and faded out over time, suggesting the emergence of compensatory mechanisms during network maturation. Nevertheless, persistent differences in excitatory/inhibitory balance and network excitability were observed in mature networks. This study improves the understanding of the complex relationship between genetic mutations and clinical manifestations in RASopathies by adding insights into functional network processes as an additional piece of the puzzle.

Overview of the datasets for the recordings of spontaneous activity of both genotypes (A: Ptpn11, B: Kras).The data show in detail the group size per experiment, and include information on the total number of recordings (# recordings), recordings that were excluded due to the exclusion criteria 3 (Table S 6) (#exclusions) and the total number of evaluated recordings (#evaluated).1.

A: Sample sizes for spontaneous activity recordings of RASopathy model
Selection of valid arrays through the MATLAB routine: A network on an array is categorized as active and further evaluated only when a minimum number of active electrodes (if not otherwise stated: 10) at DIV 21-24 is reached.An electrode is considered active with a minimum mean firing rate (MFR) of 0.1 Hz.

2.
Exclusion of whole MEA plate: In the experimental series conducted with KrasV model, on one plate a significant decline in activity (manifested as a decrease in MFR and the drop of number of active electrodes) was observed in a majority of arrays on DIV27/28 .Given the potential for systematic errors introduced by confounding factors, the plate was consequently excluded from statistical analysis starting from that time point.

3.
Exclusion of arrays due to loss of vitality: During the cultivation period, networks may suffer damage and experience a decline in vitality.This is reflected in a marked reduction in wMFR as well as a loss of active electrodes.In the arrays utilized, the activity at an electrode mirrors the neural activity of a group of neurons.A loss of active electrodes implies a substantial portion of the network has disintegrated.Networks are excluded only when both exclusion criteria are met: wMFR < 70%, and a drop of electrodes by at least 2.

4.
For the analysis of individual parameters assessing population-wide neuronal activity, the mean value across wells originating from a single animal was calculated.Consequently, animals with only one valid well were excluded from this animal-wise analysis.However, this rule was not applied to the well-wise analysis.

Fig. S 2 Fig. S 4
Fig. S 2 Experimental schema for testing the effect of disinhibition on neuronal activity in Ptpn11 D61Y .Individual steps of experiments are depicted on a timeline.Created with BioRender.com

Fig. S 5
Fig. S 5 General schemes of data analysis procedures.aAutomatized analysis of spontaneous activity recordings in time series: raw data is converted to spike list files (.csv) in Axion Integrated Studio (AxIS) 2.4.2.Further processing is performed by custom-written MATLAB software package.Data set is converted and rearranged to a .matcell array containing the time points of spikes and corresponding electrode name sorted in columns.Parameters describing spiking and bursting were calculated and output on well-level.For network bursting and connectivity parameters, time stamp matrix is calculated beforehand.Then, the parameters are either directly rearranged according to genotype, animal or treatment and averaged according to time units or principle component analysis (PCA) is switched in between.b The functional network features spiking, bursting and network bursting resulting from PCA dimension reduction projected on principle component (PC) 1.

Fig. S 6 Fig. S 7
Fig. S 6 General scheme of automatized analysis of evoked activity by electrical stimulation.Input data derives from Axion Integrated Studio (AxIS) 2.4.2.Input consists of .csvspike list files, containing information about electrode name and time point of each spike chronologically.Spike list files are entered from spontaneous activity recordings and evoked activity recordings, as well.Information about electrical stimulations is given in AxIS Spike .spkfiles.Data processing is performed by custom-written MATLAB functions.A session comprises triggering of six electrodes in a row with test pulses at 0.2 Hz for 5 min each.Selection of valid wells performs on a spontaneous activity (SA) recording prior to test stimulus (STIM) 1 defined by z (number of the related file in the folder containing all recordings) inserted by the user.Groups of wells are defined in the .xlsxgrouping file.MFR from spontaneous activity recordings is used to normalize the evoked activity on each electrode.

Fig. S 8 Fig. S 9
Fig. S 8 Homoscedasticity plots for recorded individual features describing spontaneous activity in longitudinal recordings in Ptpn11 D61Y .Absolute values of residual versus predicted values of several parameter describing spontaneous network activity.A data point represents the averaged value across the wells belonging to one preparation/animal on one plate (preparation level).Calculation is based on the data recorded on DIV21.
Fig. S 10 Homoscedasticity plots for recorded PC1 projected data sets for Ptpn11 D61Y and Kras V14I .Absolute value of residual versus predicted values of the projection of feature vectors describing the fields spiking, bursting and network bursting onto principle component (PC) 1.A data point represents one well (well-level).

Fig. S 11
Fig. S 11 Effect of disinhibition on neuronal activity in Ptpn11 D61Y with time.Line graphs with point representing mean ± SEM of several parameters describing network activity as a function of time prior and after application of bicuculline (bic) on DIV 33.Baseline values (averaged values over a time interval of 20 min prior bic treatment) were set as 100 %, values upon bic treatment were related to baseline.

Fig. S 12
Fig. S 12 Assessment of evoked activity in Ptpn11 D61Y RASopathy model.Scatter dot plot demonstrating.evMFRnormed upon stimulation pre TET (STIM1), black lines indicate median values, one data point refers to one well.Significance was tested by Mann-Whitney (MW) to check differences in mean values and Kolmogorov-Smirnov (KS) to check differences in data distributions.

Fig. S 13
Fig. S 13Relative frequency for evMFRnormed upon STIM2 and STIM 3 for control and Ptpn11 D61Y

Table S 2
Features that were used in the principle component analysis and a brief description about their calculationThe mean number of bursts in individual arrays per time weighted by the relation of currently active electrodes to number of active electrodes to reference point (e.g.time of maturity in time series), (Hz)Burst durationMean duration of bursts calculated as the time from the first spikes to the last spike in individual arrays.

Table S 3
Normality tests performed to test normality in the data sets of MEA-derived parameters; Four tests were conducted to lower the risk to fail in identifying the proper distribution.Here: 'no' indicates non-normal

Table S 4
Identification of pathologically affected networks by test stimulus (STIM).