All human cell programming-derived neural cell types were generated, cultured and/or expanded using previously described methods: iNSCs20, smNPCs21, ltNES22, RGL-NPCs23, as well as forward programmed NGN2- and ASCL1/DLX2-neurons24. IPSdMiG were generated via a proprietary protocol from LIFE & BRAIN GmbH (patent application number EP20162230). Details are found in the Supplementary Methods.
AAV production and transduction
AAV vectors were produced in HEK293T cells using a standard triple-transfection protocol and an AAV helper plasmid encoding AAV rep and cap genes, an AAV vector plasmid carrying the transgene as well as an adenoviral helper plasmid.51 Wild-type cap genes were modified through insertion of DNA oligonucleotides encoding 7mer or 9mer peptides whose sequences and properties were reported19 and are provided in Supplementary Table S1.
For parallel AAV capsid screening in 96-well plate format, crude cell lysates from triple-transfected 6-well plates containing non-purified AAV particles were used. Therefore, 10 µl AAV were added per well with a multichannel pipette directly into the medium (200 µl) of the pre-plated cells (1:20 dilution). For the characterization of ltNES-derived neuroglial differentiation cultures and AAV-mediated gene knockdown, cells were cultured in 3.5 cm dishes and transduced with crude lysates at a 1:20 dilution.
For purification by iodixanol gradient density centrifugation, cells were collected by scraping 72 hours after transfection, followed by centrifugation at 500 × g for 15 minutes at room temperature. The cell pellet was resuspended in 20 ml Benzonase buffer and lysed by five freeze-thaw cycles in liquid nitrogen and a 37°C water bath. Cell debris was sonicated for 1 minute and subjected to two centrifugations steps, each at 4,000 × g for 15 minutes at 4°C. The virus-containing lysate was purified via iodixanol density gradient ultracentrifugation (290,000 × g, 120 minutes, 4°C) and collection of the 40% phase. This AAV-containing fraction was rebuffered in 1xDPBS using amicon spin columns (Merck, Darmstadt, Germany) and AAVs were stored in aliquots at -80°C. Viral genome titers were determined by qPCR using the AAV quantification kit from Norgen Biotek Corporation (Ontario, Canada). Concentrated AAVs were used for NGN2 overexpression at a dose of 6x104 vg/cell.
To generate an AAV vector expressing an shRNA against SLC25A1, two complementary oligonucleotides were designed, each comprising the sense (5′-GCCATCCGCTTCTTCGTCATGA-3′) and antisense strand (5′-TCATGACGAAGAAGCGGATGGC-3′), as well as the central loop sequence TCAAGAG. The 5’ ends were flanked by the sequence CACC on the forward strand and by AAAA on the reverse strand. Following annealing, these flanking sequences produced overhangs that were used for direct ligation using a Golden Gate protocol into the BbsI-digested AAV vector plasmid TRISPR52 (details of this plasmid will be reported elsewhere).
To clone the pAAV-EF1a-NGN2-3xFLAG expression vector, plasmids pAAV-EF1a-NGN2-RFP-NLS (details of this plasmid will be reported elsewhere) and NR6AI-C-term-3xFLAG (kindly provided by Nils Braun) were used as templates. The NGN2 and 3xFLAG sequences in these plasmids were PCR-amplified with appropriate overhangs for cloning using primers 5′-tgtcgtgaggaatttcgacggatccgccaccatgttcg-3′ plus 5′-ctttataatctttatcatcatcatctttataatcgatacaatccctggctatgg-3′ (for NGN2) or 5′-ccatagccagggattgtatcgattataaagatgatgatgataaagattataaag-3′ plus 5′-ttcaacatgcctccgcccttatcatttatcatcatcatctttataatc-3′ (for 3xFLAG). Gibson assembly was performed using the 2x Gibson assembly master mix (New England Biolabs, Frankfurt am Main, Germany). Thermo Fisher Scientific’s (Waltham, USA) plasmid DNA Miniprep and Maxiprep kits were used for DNA isolation.
Forty-eight hours after transduction with the AAV panels, cells were washed once with 1xDPBS before fixation with 4% PFA for 10 minutes at room temperature. Afterwards, cells were washed once with 1xDPBS before incubation with 2 µg/ml DAPI for 5 minutes at room temperature. DAPI was replaced by 200 µl 1xDPBS per well before YFP fluorescence was measured using an Infinite M Plex plate reader (Tecan Group, Männedorf, Switzerland; at 500 nm excitation and 540 nm emission) and/or immunofluorescence image acquisition and analysis (see below). For all other immunofluorescent stainings, blocking with 0.5% Triton-X and 10% FBS was performed after fixation, followed by overnight incubation with primary antibody solution at 4°C (1:1,000 rabbit GFAP from Merck Millipore, Burlington, USA, Cat. AB5804; 1:100 rabbit MAP2 from Abgent, San Diego, USA, Cat. AP2018E; 1:200 rabbit Nestin (NES) from Bio-techne, Minneapolis, USA, Cat. NB300-265; 1:1,000 mouse S100b from Sigma-Aldrich, St. Louis, USA, Cat. S2535; 1:100 mouse SOX2 from R&D systems, Minneapolis, USA Cat. MAB2018; 1:400 chicken TUBB3 from Merck Millipore, Cat. AB9354; 1:1,000 mouse TUBB3 from Eurogentec, Lüttich, Belgium, Cat. MMS-435P; 1:1,000 rabbit TUBB3 from biolegend, San Diego, USA, Cat. 802001; in 0.3% Triton-X and 5% FBS). The next day, three washing steps with 0.3% Triton-X were performed before secondary antibody solution was incubated for 2 hours at room temperature (1:500 donkey-anti-rabbit AF-647, Cat. A31573; 1:500 goat-anti-chicken AF-647, Cat. A21449; 1:1,000 goat-anti-mouse AF-555, Cat. A21424; 1:500 goat-anti-mouse AF-647, Cat. A21236; 1:1,000 goat-anti-rabbit AF-488, Cat. A11008; 1:1,000 goat-anti-rabbit AF-555, Cat. A21429; 1:500 goat-anti-rabbit AF-647, Cat. A21244; in 0.3% Triton-X and 5% FBS; all Thermo Fisher Scientific). After three additional washing steps, nuclei were counter-stained with DAPI. Wells were sealed with Mowiol plus DABCO (Carl Roth, Karlsruhe, Germany).
Image acquisition and analysis
Automated image acquisition of 96-well plates was performed on an epifluorescence Scan^R screening microscope equipped with the Scan^R acquisition software (Olympus Biosystems, Münster, Germany) or an INCell analyzer 2200 (GE Healthcare Bio-Sciences Corp, Piscataway, USA) with the respective INCell analyzer 2200 software. 10× and 20× objectives were used with appropriate excitation and emission filters to measure 9 and 16 positions per well, respectively. For AAV screening, Scan^R images were subsequently processed with a previously established pipeline for quantitative image analysis.53 This yielded the mean gray value of all cells in the analyzed images, total cell counts, and percentages of AAV-positive cells (determined via fluorescent reporter). Non-quantitative image processing of automatically acquired INCell images for qualitative display was performed with defined CellProfiler pipelines running the ‘ApplyThreshold’ (threshold strategy: manual) and ‘RescaleIntensity’ (rescaling strategy: divide each image by the same value) modules. Alternatively, images were manually captured with an AxioImager plus Apotome (Carl Zeiss, Oberkochen, Germany) and the AxioVision software. Quantification of manually acquired images was performed using a semi-automated, three-step pipeline specifically developed in the context of this project. 1) Cell segmentation: First, single cells and cell clusters were segmented using the DAPI channel. The segmentation was primary based on adaptive thresholding and Gaussian filtering. Objects were distinguished between invalid objects, cells and cell-clusters depending on size and shape. Cell clusters were broken up into cells by creating distance maps and their local maxima, which were used as the starting point of a watershed algorithm. 2) Background separation: Second, after an initial step of intensity normalization based on the maximum intensity, foreground and background were separated in all other (marker) channels. This could either be done by a thresholding based on intensity quantiles or by fitting Gaussian curves to the intensity distribution. Alternative to two Gaussians (foreground and background), we also found good results when using three Gaussians to distinguish between background, low- and high-expressing cells for some markers stained. Based on the locations of the Gaussians the separation threshold(s) were defined. 3) Classification: Third, the cells identified in step 1 were automatically denotated as marker-positive or -negative by comparison of their mean intensity in the marker channels with a manually set threshold or by probability assignment, if using the Gaussian-fitting approach from step 2. Further options applied to optimize the categorization process comprised measures of texture homogeneity, entropy or energy of objects. As a final step, automatic object categorization could be manually revised, if necessary. After quantification, images were post-processed with ImageJ for display. Image processing included adjustments of brightness and contrast, and was equally applied across single channel pictures.
In order to quantify the percentage of successfully transduced cells via flow cytometry, iNSCs were detached and singularized by Accutase (Thermo Fisher Scientific) treatment. GFP fluorescence was measured in FL1 on an Accuri C6 Plus flow cytometer (BD Biosciences, San Jose, USA). Events were gated for living single cells and data analyzed with the associated Accuri C6 Plus analysis software.
RNA extraction and cDNA synthesis were performed with Qiagen’s (Hilden, Germany) RNeasy and Quanta biosciences’ (Beverly, USA) qScript kit, respectively, according to the manufacturers’ instructions. qPCR was performed on an Eppendorf (Hamburg, Germany) iCycler in 96-well format. Per reaction, 300 ng cDNA were diluted in a total volume of 19 µl qPCR master mix (ddH2O with 200 µM of each dATP, dGTP, dCTP and dTTP, 100 nM fluorescin calibration dye, 7.5x SyBr Green, 4% DMSO, 1x PCR Rxn buffer, 3 mM MgCl2, and 0.6 U Taq polymerase (Thermo Fisher Scientific)) and supplemented with 1 µl primer mix consisting of 5 µM forward and reverse primers (all IDT, Coralville, USA). Primer sequences were as follows: 18s F: 5’-TTCCTTGGACCGGCGCAAG-3’, R: 5’-GCCGCATCGCCGGTCGG-3’; ITGA6 F: 5’-AAAGCCTGCTCAATCCCTGA-3’, R: 5’-GTAAGTCAGCCACGCCAAAA-3’; ITGAE F: 5’-GGTTCTTTCTCCTTGGCACG-3’, R: 5’-CATCTGCCCCAAAACCTTCC-3’; ITGB1 F: 5’-AGCTTTAAAACCTGTGTGCCAT-3’, R: 5’-CAAAAGGTCCCCATTCAGCAA-3’; ITGB5 F: 5’-GGCTTGATCACAGCTCCCTA-3’, R: 5’-GGACAACAAGCTGGATGTGG-3’; ITGB8 F: 5’-AAGAGCAGTCACCTACCGAC-3’, R: 5’-CGTTCGAGTGTACAACCAGC-3’; GFAP F: 5’-ATCGAGAAGGTTCGCTTCCT-3’, R: 5’-CAGCCTCAGGTTGGTTTCAT-3’; MAP2 F: 5’-AAAGAAAAGGCCCAAGCTAAA-3’, R: 5’-GCTTCCAGACGAGGAGACA-3’; NEUN F: 5’-GGATGGATTTTATGGTGCTGA-3’, R: 5’-ACATGGTTCCAATGCTGTAGG-3’; NGN2 F: 5’-CTGGGTCTGGTACACGATTG-3’, R: 5’-GGTTGTTGGCCTTCAGTCTA-3’; PAX6 F: 5’-AATAACCTGCCTATGCAACCC-3’, R: 5’-AACTTGAACTGGAACTGACACAC-3’; SLC25A1 F: 5’-CCCCATGGAGACCATCAA-3’, R: 5’-CCTGGTACGTCCCCTTCAG-3’; SOX2 F: 5’-GTATCAGGAGTTGTCAAGGCAGAG-3’, R: 5’-TCCTAGTCTTAAAGAGGCAGCAAAC-3’; TUBB3 F: 5’-CCATCTTGCTGCCGACAC-3’, R: 5’-CAATAAGACAGAGACAGGAG-3’; VIM F: 5’-TCTCTGAGGCTGCCAACCG-3’, R: 5’-CGAAGGTGACGAGCCATTTCC-3’. qPCR reactions were run in triplicates and cycling conditions were as follows: 3 minutes at 95°C; 40 cycles of 15 seconds at 95°C, 20 seconds at 60°C and 30 seconds at 72°C; 1 minute at 95°C and 15 seconds at 55°C, followed by a temperature gradient to 95°C over the course of 20 minutes and finally 15 seconds at 95°C, before cooling down to 4°C. Ct values of the genes of interest were normalized to the house-keeping gene 18s and transformed to mean fold changes using the Eq. 2−∆∆Ct. qPCR products were size-validated by gel electrophoresis.
All statistical analyses were performed with R (version 3.5.1; the R foundation for statistical computing 2018). First, to determine whether variables met the assumptions of linear models, Kolmogorov-Smirnov’s, Shapiro Wilk’s and Levene’s test, and/or F-test for normal distribution and homogeneity of variance were performed. Since data were not normally distributed, two-sided Wilcoxon rank sum tests were calculated to compare two groups. Correlation was assessed using Kendall’s τ. The significance level was set to p < 0.05. Multiple testing on the same set of data was not performed.
Statement of Ethical Approvals
The collection of human somatic material (i.e., blood or fibroblasts) from healthy donors for iPSC reprogramming and iNSC conversion was in accordance with German law and approved by the ethics committee of the University of Bonn Medical Center (approval number: 275/08). All subjects gave written informed consent.