Fig. 6: Effect of combined RA and SAG modulation on hTLS.
From: Modelling co-development between the somites and neural tube in human trunk-like structures
a, LaminB1::mTagRFP-T 120 h hTLS structures following RA inhibition with BMS493, showing depletion of PAX6 expression from the neural tube (n = 3–5 per condition). b, Quantification of HCR expression by neural tube volume (%) with exposure to SAG and BMS493. The dashed line indicates the smoothed trend with 0.95 confidence interval (grey ribbon). The box plots represent mean (centre line) and interquartile range (bounding box), with whiskers representing the largest and smallest value no further than 1.5× the interquartile range and outliers represented as black dots. n.s., non-significant. *P < 0.05; **P < 0.01 obtained by two-sided Wilcoxon test compared with all samples. c,d, Effect of variable SAG exposure without (c) (n = 5–6 per condition) and with (d) RA inhibition by BMS493 on 144 h LaminB1::mTagRFP-T hTLS (n = 5–7). e, Schematic diagram of proposed reciprocal signalling occurring between somites and the neural tube of hTLSs. Non-canonical Wnt signalling from the neural tube probably biases ALDH1A2 expression in the somites towards the neural tube (far left), followed by subsequent RA signalling from the somites to the neural tube (middle left). This leads to PAX6 expression being biased towards the somitic side (middle), or in the context of SAG exposure, increased OLIG2 expression adjacent to the somites (middle right). This patterning is dependent on endogenous RA signalling because inhibition leads to FOXA2 expression in the absence of other mediolateral patterning (far right). Asterisks indicate somites; dashed lines outline the hTLS. Scale bars, 100 μm. RA, retinoic acid; IM, intermediate mesoderm; LPM, lateral plate mesoderm.
