High-throughput micro-patterning platform reveals Nodal-dependent dissection of peri-gastrulation-associated versus pre-neurulation associated fate patterning

In vitro models of post-implantation human development are valuable to the fields of regenerative medicine and developmental biology. Here, we report characterization of a robust in vitro platform that enabled high-content screening of multiple human pluripotent stem cell (hPSC) lines for their ability to undergo peri-gastrulation-like fate patterning upon BMP4 treatment of geometrically-confined colonies and observed significant heterogeneity in their differentiation propensities along a gastrulation associable and neuralization associable axis. This cell line associated heterogeneity was found to be attributable to endogenous nodal expression, with upregulation of Nodal correlated with expression of a gastrulation-associated gene profile, and Nodal downregulation correlated with a neurulation-associated gene profile expression. We harness this knowledge to establish a platform of pre-neurulation-like fate patterning in geometrically confined hPSC colonies that arises due to a stepwise activation of reaction-diffusion and positional-information. Our work identifies a Nodal signalling dependent switch in peri-gastrulation versus pre-neurulation-associated fate patterning in hPSC cells, provides a technology to robustly assay hPSC differentiation outcomes, and suggests conserved mechanisms of self-organized fate specification in differentiating epiblast and ectodermal tissues.

We kindly thank Dr. Andras Nagy for providing the CA1 human embryonic stem cell line; Dr. Sean Palecek 33 for providing the 7TGP(H9) line; Dr. Gordon Keller for providing the GKH9(H9), HES2, and the PDX1-34 GFP(MEL1) lines; and Dr. Andrew Elefanty for providing the RUNX1-GFP(HES3), and the MIXL1-35 GFP(HES3) lines. We thank Dr. Emanuel Nazareth for providing us with the micro-contact printing stamps. 36 We gratefully acknowledge Dr. Alfonso Martinez Arias, Dr. James Briscoe Following implantation, embryos undergo a dramatic transformation mediated by tissue growth, cell 94 movements, morphogenesis, and fate specifications resulting in the self-organized formation of the future 95 body plan [1]. Post-implantation development to the neurula stage embryo is orchestrated by two vital 96 developmentally conserved events called gastrulation and neurulation. Gastrulation is the developmental 97 stage that segregates the pluripotent epiblast into the three multipotent germ layers, namelythe ectoderm, 98 the mesoderm, and the endoderm[2-4]. Closely following gastrulation, the ectoderm undergoes further fate 99 specification resulting in the patterned neural plate, neural plate border, and non-neural ectoderm regions 100 thereby setting the stage for the onset of neurulation[5-9]. As neurulation proceeds, morphogenetic 101 changes in these tissues result in the formation of the neural tube, the neural crest, and the epithelium 102 respectively [10]. Initiation of the morphogenetic restructuring of the epiblast and the ectoderm occurs due 103 to self-organized gradients of signalling molecules called morphogens, and morphogens belonging to the 104 transforming growth factor beta (TGFβ) superfamily, such as bone morphogenetic proteins (BMPs) and 105 Nodal, play vital roles in these developmental stages. 106 Two biochemical models, Reaction-Diffusion (RD) and Positional-Information (PI), have strongly influenced 107 our mechanistic understanding of self-organized fate specification during embryogenesis. The RD model 108 describes how a homogenously distributed morphogen can self-organize into a signalling gradient in a 109 developing tissue due to the presence of an interaction network between the morphogen and its inhibitor, 110 both of which are hypothesized to be diffusible molecules albeit with differential characteristic 111 diffusivities [11][12][13]. Recent interpretations of RD have proposed that higher order (>2 molecules) network 112 topologies can also underlie this self-organization [14]. The PI model describes how fate patterning can 113 occur in a developing tissue due to an asymmetric morphogen distribution. The classical version of this 114 paradigm hypothesized that the cells in the developing tissue sense the morphogen concentration in their 115 immediate vicinity and acquire fates according to a threshold model [15,16]. Recent studies have updated 116 this interpretation of the PI model and suggest that fates are acquired as a function of both the morphogen 117 concentration and time of induction [17,18]. Although both RD and PI have been incredibly valuable in 118 facilitating our comprehension of how developmental fates arise in a self-organized manner, these models 119 are typically studied in individual signalling pathways. How multiple signalling pathways may work in concert 120 to execute the rules specified by either RD or PI are not well understood. 121 Studying post-implantation developmental events, like gastrulation and neurulation, directly in human 122 embryos would unequivocally provide the most reliable interpretations of human development. While 123 valuable progress has been made of late in culturing human blastocysts in vitro [19,20], ethical concerns 124 preclude their maintenance beyond 14 daysprior to the onset of gastrulation. On the other hand, recent 125 studies on in vivo human development have provided some incredible insight into human development well 126 into the fetal stages [21]. However, these studies are performed on specimens acquired from terminations 127 or abortions that are typically accessible after the stages of gastrulation and neurulation have already 128 transpired. Consequently, investigation of the mechanisms underpinning early post-implantation human 129 embryonic development directly in human embryos is currently not possible. Nevertheless, the ability of 130 stem cells to self-organize into structures in vitro that mimic aspects of post-implantation human 131 development when provided appropriate biophysical and biochemical cues is well established [22][23][24][25][26][27]. We 132 and others have used human pluripotent stem cells (hPSCs) to demonstrate that BMP4 treatment of 133 geometrically confined hPSC colonies recapitulate some aspects of human peri-gastrulation-like self-134 organized fate patterning [24][25][26]. Although stem cell derived in vitro constructs are indisputably an 135 incomplete representation of embryos, they can serve to provide insights into some cell organizational 136 events that occur during the critically important post-implantation developmental stages. 137 In addition to providing insight into cell organizational aspects of early gastrulation-like behaviour, in vitro 138 models of human development are also of great value to the field of regenerative medicine. This is because 139 the ability of these models to specify early developmental cell fates highlights their suitability for 140 characterization of differentiation propensities of hPSC lines. It is well known that hPSC lineswhether 141 they are derived from embryos or from reprogrammed somatic cellshave an inherent bias to differentiate 142 toward specific lineages [28][29][30]. Consequently, assays that can characterize different hPSC lines to identify 143 these biases are of crucial importance to the field of regenerative medicine. Given the significance of 144 developing approaches for achieve this goal, multiple assays have been established to address this need. 145 The most prominent of these are assays like the teratoma assay[31], the scorecard assay[32], and the 146 pluritest [33]. Although each of these approaches have their benefits, they are either lengthy, tedious and 147 expensive (teratoma and scorecard), or do not directly measure differentiation of the hPSC lines (pluritest). 148 In contrast, the peri-gastrulation-like assay provides a quantitative measure of the generation of lineage-149 specific fates, is rapid and dramatically inexpensive in comparison to approaches like the teratoma assay. 150 In addition, it is readily amenable to high-content screening. However, to capitalize on the capabilities of 151 this assay, we require a robust micropatterning platform that readily enables high-content studies. Here, we report characterization of a high-content platform to produce microtiter plates that allow robust 164 geometric confinement of a variety of adherent cell types at single cell resolution. Employing this platform, 165 we tested the response of a panel of hPSC lines to a previously reported peri-gastrulation-like assay and 166 observed significant variability in the induction of the Brachyury (BRA) expressing region between the lines. 167 To probe the emergent differentiation trajectories of hPSC lines, we assessed their differentiation-168 associated gene expression profiles and found a switch-like response in the upregulation of gene profiles 169 associated with either gastrulation or neurulation. This switch in gene expression showed a strong 170 association with Nodal signalling; hPSC lines that exhibited higher levels of a gastrulation-associated gene 171 expression profile also upregulated Nodal signalling, and those that exhibited a higher neurulation-172 associated gene expression profile downregulated Nodal signalling. coverslips, with relative levels increasing with exposure times up to 12 minutes after which the relative 207 levels detected decreased (Fig. 1D). These findings indicate that, consistent with previous reports[40,45], 208 PLL-g-PEG incubation results in PEGylation of coverslips, and that the optimal exposure time to maximize 209 the presence of carboxyl functional groups on the PEGylated coverslips in our experimental setup was 12 210 minutes. To produce 96-well microtiter plates for patterned cell-culture surfaces, PEGylated large coverslips 211 (110mmx74mm) were photopatterned by DUV exposure through Quartz photo-masks for 12 minutes and 212 assembled to bottomless 96-well plates (Fig. 1E). Carboxyl groups were activated using carbodiimide 213 chemistry[43] (Fig. 1F) to enable covalent attachment to primary amines on ECM molecules. This "PEG 214 plates" platform enabled robust geometrical-confinement of a variety of cell types in colonies of a variety of 215 shapes and sizes (Fig. 1F, Sup. Fig. 1A-F). 216 Given the vital role that interactions between cells and the surrounding ECM play on cellular responses[46], 217 we next asked whether the approach of covalent attachment of ECM molecules interfered with fate 218 decisions of hPSCs micropatterned on the PEG plates. We opted to employ a recently reported two-day 219 assay using OCT4, and SOX2 expression as readouts to assess fate decisions in geometrically confined 220 hPSC colonies[30] (Fig. 1H), and directly compared fate acquisition of hPSCs on the PEG plates with μCP 221 plates, a micro-patterning technique that does not require any chemical immobilization of ECM molecules. 222 We observed a highly correlated (R 2 > 0.9) differentiation response between μCP and PEG plates ( Fig. 1I-223 J). Furthermore, the PEG plates responded in a more reproducible manner than the μCP plates both in 224 terms of the number of colonies achieved per well of a 96-well plate, and the number of cells attached per 225 colony (Sup. Fig. 1G-H . Although all hPSC lines tested expressed high levels of pluripotency markers at the start of 243 the differentiation culture ( Fig. S2), induction of BRA expression levels varied markedly between hPSC 244 lines at 48h after BMP4 treatment ( Fig. 2A,C). Notably, although the MEL1, and HES3-1 lines were unable 245 to induce the expression of BRA, they did differentiate as indicated by the reduction of SOX2 expression 246 relative to the starting population ( Fig. 2B-C, Fig. S2 Table S1)and cultured them 261 in conditions unsupportive of pluripotency for three days and analyzed differentiation marker gene 262 expression levels daily (henceforth -'EB assay') ( Fig. 3A). We observed strong variation in expression 263 profiles of differentiation associated genes between the test hPSC lines (Fig. 3Bi). To simplify data 264 interpretation, we used unsupervised K-means clustering to segregate the hPSC lines into 'Strong', 265 'Intermediate', and 'Weak' expressers for each gene tested. This analysis revealed distinct sets of 266 responses in the test lines where some lines upregulated expression of genes associated with gastrulation 267 while others upregulated expression of neurulation-associated genes (Fig. 3Bii) upregulation linked with the induction of gastrulation-associated genes and their absence linked with the 278 induction of neurulation-associated genes (Fig. 3Ci). Furthermore, clustering the hPSC lines with reference 279 to the expression profiles of Nodal and GDF3 by either unsupervised K-means clustering (Fig. 3Cii, S3), 280 or by hierarchical clustering based on Euclidean distance (Fig. S4) contained the identical hPSC line cohorts (Fig. 3Bii, S5A), suggesting the likelihood of parallel functions of 294 both these genes in differentiating hPSCs. To validate the gene expression profiles observed in our EB 295 assay, we asked if the observed expression differences of these genes that critically regulate endoderm 296 specification were able to predict the propensity of the hPSC lines to differentiate toward the definitive 297 endodermal fates. Consequently, we differentiated the panel of hPSCs toward definitive endoderm using 298 an established protocol (Fig. S5B), and consistent with the findings of Kojima et al [47], the expression 299 profiles of endoderm specifiers (MIXL1 and EOMES in our case) in our EB assay closely matched the 300 propensity of the hPSC lines to induce SOX17 expression upon directed differentiation toward the definitive 301 endoderm fate (Fig. S5C,D). The differential expression profiles of MIXL1 and EOMES in the EB assay 302 were also able to predict the induction efficiency of mature endodermal fates. Specifically, lines from the 303 MIXL1/EOMES-Strong cluster outperformed candidate lines from the MIXL1/EOMES-Weak cluster in the 304 induction of pancreatic progenitors as marked by the co-expression of PDX1 and NKX6.1 (Fig. S5E). These 305 data provide protein level phenotypic validation of the variable gene expression observed in our EB assay. 306 Given that geometrically confined hPSC colonies are able to induce organized fate patterning[24-26], we 307 asked if subjecting geometrically-confined hPSC colonies to defined endodermal differentiation conditions 308 could be used as an assay to predict the differentiation propensity of hPSC lines. We selected three hPSC 309 lines -H9-1, HES3-2, and HES3-1to represent each MIXL1/EOMES induction compartment defined in 310 the EB assay ( Fig. 3Bii, S5A), and differentiated them as geometrically-confined colonies in defined 311 endodermal induction conditions (Fig. S6A). Interestingly, we found that the relative efficiency of SOX17 312 and FOXA2 double-positive expression under these experimental conditions closely matched the 313 endodermal lineage-bias of the lines as predicted by the EB assay ( Fig. S5A-D, Fig. S6B,C). Taken  314 together, these data validate the differential gene expression observed between the panel of hPSC lines 315 by demonstrating congruence between MIXL1 and EOMES temporal dynamics and endoderm lineage bias 316 of hPSC lines and provide proof-of-concept data that the defined differentiation protocols in geometrically-317 confined hPSC colonies can be used as quick assays to assess lineage bias of hPSC lines. 318 hereafter 'SB') which antagonizes Nodal signalling (Fig. 3Di) recapitulated the observed switch in emergent 332 gene expression. After a three-day induction, we observed that colonies grown in the presence of SB 333 upregulated genes associated with neurulation whereas those grown in the absence of SB upregulated 334 genes associated with gastrulation ( Fig. 3Dii, Fig. S7). These results are consistent with our hypothesis 335 that Nodal signalling distinguishes gastrulation and neurulation-associated gene expression profiles in 336 differentiating hPSCs. Given that differentiating hPSCs in the absence of Nodal signalling upregulated 337 neurulation associated genes, we next set to investigate if BMP4 treatment of geometrically confined hPSC 338 colonies in presence of SB gave rise to early neurulation-associated spatially patterned fate allocation. 339 340

342
In a recent study, we demonstrated that the peri-gastrulation-like fate patterning in geometrically confined 343 hPSC colonies occurs via a stepwise process of RD and PI where a BMP4-Noggin RD network self-344 organizes a phosphorylated SMAD1 (pSMAD1) signalling gradient within the colonies, resulting in the peri-345 gastrulation-like fates being patterned in a manner consistent with the PI paradigm[24]. We set out to 346 investigate if a conserved mechanism would give rise to neurulation-associated fate patterning. As a first 347 step, we asked if a BMP4-Noggin RD network governed pSMAD1 self-organization within the geometrically-348 confined hPSC colonies treated with BMP4 and SB. Consistent with the presence of a BMP4-Noggin RD 349 network[24], we observed an upregulation of both BMP4 and Noggin upon BMP4 treatment of hPSCs in 350 the presence of SB (Fig. 4A). We next asked if BMP4 treatment of geometrically-confined hPSC colonies 351 in the presence of SB would result in the self-organized gradient of nuclear localized pSMAD1. Indeed, 352 pSMAD1 activity within the colonies rapidly self-organized into a radial gradient under these experimental 353 conditions (Fig. 4B-D). We next queried the importance of Noggin in the formation of the pSMAD1 gradient 354 by generating two homozygous knock-outs of Noggin ('C1', and 'C7') using Crispr/CAS9 (characterization 355 of lines shown in Fig. S8) and tested whether the absence of Noggin compromised the self-organization of 356 pSMAD1. Consistent with our hypothesis that a BMP4-Noggin RD network was underlying the pSMAD1 357 self-organization, the formation of the pSMAD1 signalling gradient was significantly compromised in Noggin 358 knockout lines C1 and C7 compared to the wildtype control ( Fig. 4E-F), indicating an integral involvement 359 of BMP inhibitors like Noggin in the self-organization of the pSMAD1 signalling gradient. In our previous 360 study using the peri-gastrulation-like model, we showed that a BMP4-Noggin RD computational model 361 predicts the experimentally observed responses of a pSMAD1 self-organized gradient at the periphery and 362 the center of the colonies to perturbations to the BMP4 dose in the induction medium and size of the 363 geometrically-confined hPSC colony [24]. Specifically, we showed that reducing the BMP4 dose while 364 maintaining the colony size reduces the levels of pSMAD1 at the periphery, and reducing the colony size 365 while maintaining a constant BMP4 dose in the induction medium results in an increase of pSMAD1 levels 366 at the center of the colonies [24]. We reasoned that a conserved mechanism underlying the pSMAD1 self-367 organization would result in identical responses to these perturbations. Consistent with our anticipated 368 results, reducing the BMP4 dose in the induction medium while maintaining the colony size resulted in a 369 reduction of the detected immunofluorescent levels of nuclear localized pSMAD1 at the colony periphery 370 (Fig. S9A-C). Furthermore, reducing the colony size while maintaining the BMP4 dose in the induction 371 medium increased the detected immunofluorescent levels of nuclear localization of pSMAD1 at the colony 372 centers (Fig. S9D-E). Taken together, these data demonstrate that in absence of Nodal signalling, pSMAD1 373 activity in the geometrically-confined hPSC colonies self-organizes into a signalling gradient and suggest 374 that a BMP4-Noggin RD system governs this observation (Fig. 4G) consistent with our previous study [24]. 375 376 Nodal signalling contributes to the shape of the self-organized pSMAD1 gradient 377 378 Our data indicate that the pSMAD1 signalling gradient self-organizes via an RD network present in the BMP 379 signalling pathway where Noggin functions as an important inhibitor (Fig. 4). Given that Nodal signalling 380 targets include multiple BMP antagonists such as Cer1, GDF3, Follistatin (FST), etc.
[52], we asked if Nodal 381 signalling contributed to the formation of the pSMAD1 signalling gradient in BMP4-treated geometrically 382 confined hPSC colonies. To probe the role of Nodal in the observed pSMAD1 self-organization, we 383 compared the formation of the pSMAD1 gradient in geometrically confined hPSC colonies of 500μm 384 diameter treated with BMP4 for 24h where the induction media either contained Nodal or SB (Fig. 5A). The 385 pSMAD1 signalling gradients formed in the presence and absence of Nodal signalling were significantly 386 different from each other when treated with either 25ng/ml of 50ng/ml of BMP4 in the induction medium 387 demonstrating the involvement of Nodal signalling in the formation of the self-organized signalling gradient 388 ( Fig. 5B-D, S10A-C). The results from these studies provided a few notable observations. First, in 500μm 389 diameter colonies treated for 24h with BMP4 in presence of Nodal ligands, we observed two prominent 390 peaks of pSMAD1 expressionone peak was observed at the periphery as expected from previous 391 reports [24][25][26], and another on at the colony center that has not been previously reported in colonies of 392 this size (Fig. 5B-D). This observation provides further support to the proposition that the self-organization 393 of pSMAD1 arises via an RD mechanism which can result in spatial oscillations of morphogen activity[24], 394 while providing evidence of crosstalk between BMP and Nodal signalling pathways in establishing the 395 pSMAD1 signalling gradient. Another notable observation was that the pSMAD1 expression that declined 396 from the periphery of the colonies dropped rapidly in the presence of Nodal ligands, but the decline was far 397 more gradual in the presence of SB ( Fig. S10A-C). Cells expressing discernible levels of pSMAD1 were 398 observed much farther into the colony from the periphery than observed in the condition when Nodal ligands 399 were present in the induction medium. Finally, the level of pSMAD1-associated immunofluorescence 400 detected at the colony periphery in the presence of SB was significantly higher than the levels detected in 401 the presence of Nodal ( Fig. 5B-D, S10A-C). Taken together these data provides further justification for the 402 hypothesis that the pSMAD1 signalling gradient arises via an RD mechanism, and demonstrate that Nodal 403 signalling contributes to the shape of the self-organized pSMAD1 signalling gradient. 404 Since an RD network results in the morphogen gradient as a consequence of the expression of both 405 activators and inhibitors of the morphogen[11,12], we hypothesized that the likely reasons for this 406 observation could be due to either a change in the amount of activator (change in BMP4 levels) or the 407 amount of inhibitor (change in the level of BMP antagonists) in the system. When we tested gene expression 408 of activators and inhibitors after 24h of Vehicle versus Nodal or SB treatment on hPSCs which either allowed 409 Nodal expression or dramatically downregulated it (Fig. S10D), SB treatment provoked an increased 410 positive feedback as indicated by increased detected levels of BMP4 transcripts (Fig. 5E); and a reduced 411 negative feedback as indicated by significantly reduced transcript levels of BMP antagonists like CERL, 412 GDF3, and FST (Fig. 5F). Taken together, these data suggest that Nodal signalling can contribute to the 413 RD-mediated self-organization of the pSMAD1 signalling gradient; and that this contribution might occur 414 due to a change in the levels of activators and antagonists of BMP signalling. Having established that 415 pSMAD1 activity in the geometrically confined hPSC colonies treated with BMP4 and SB self-organizes 416 into a signalling gradient, we next focused on investigating if this gradient induced the expression of fates 417 associated with the differentiating ectoderm. 418 419 Pre-neurulation-like fate patterning arises in a manner consistent with PI 420 421 In the presence of Nodal signalling, BMP4 treatment of geometrically-confined hPSC colonies results in 422 self-organized pSMAD1 gradient and a spatially patterned acquisition of gastrulation-associated fates [24]. 423 Although we observe the formation of the pSMAD1 gradient when geometrically-confined hPSC colonies 424 are treated with BMP4 and SB ( Fig. 4-5), we did not observe expression of key gastrulation associated 425 markers like BRA, EOMES, SOX17, and GATA6 ( Fig. S11A-B). These observations are consistent with 426 the need of Nodal in inducing gastrulation-associated fates [24]. Since we observed that differentiating 427 hPSCs in the absence of Nodal signalling upregulate a neurulation-associated gene profile (Fig. 3D) with concomitant expression of TFAP2A (NN, NPB), and SIX1 (panplacodal competent NPB) (Fig. S11C). 443 We define this fate patterning as 'pre-neurulation-like' and using SOX2, and GATA3 as the markers of the 444 PN, and NN tissues, we set out of test if the fate patterning arose in a manner consistent with the positional 445 information (PI) paradigm. 446 Given that the PI paradigm posits that developmental fates arise due to thresholds of morphogen levels, 447 and we asked if the perturbations of pSMAD1 levels at the colony periphery ( Fig. S9A-C) and the colony 448 center (Fig. S9D-E) resulted in pSMAD1 threshold mediated changes in expression of GATA3 (NN), and 449 SOX2 (PN) fates respectively. Consistent with the idea of a pSMAD1 threshold dependent patterning of the 450 PN and the NN tissues marked by GATA3, and SOX2, we find that reducing the pSMAD1 levels at the 451 colony periphery (Fig. S12A) significantly reduced the GATA3 expression at the colony periphery ( Fig.  452 S12B-C) and increasing the pSMAD1 levels at the colony center (Fig. S12D) dramatically reduced the 453 SOX2 expression (Fig. S12E-F). These data indicate that thresholds of pSMAD1 regulated the patterning 454 of the SOX2 and GATA3 within the geometrically confined hPSC colonies. However, the formalization of 455 the PI paradigm has been updated to include time as a critical parameter that patterns the developmental 456 cell fates. Specifically, fate patterning mediated by PI is known to arise as a function of the morphogen 457 concentration and time of induction [17,18,24,56]. Consequently, we tested four different doses of BMP4 458 (3.125ng/ml, 6.25ng/ml, 12.5ng/ml, and 25ng/ml) in the induction medium for four different induction times 459 (12h, 24h, 36h, and 48h) and measured the levels of SOX2 and GATA3 detected. We observed that the 460 fate patterning of GATA3 arose as a function of both the concentration of BMP4 in the induction medium 461 and the time of induction (Fig. 6A-B, C(i), Fig. S13) indicating that the patterning within the geometrically 462 confined colonies arises in a manner consistent with PI ( Fig. 6C(ii)). Thus far, our data indicate that the pSMAD1 gradient was enforced outside-in within the geometrically 467 confined hPSC colonies via a BMP4-Noggin RD network, and the pre-neurulation-like fates arose in a 468 manner consistent with PI. In agreement with this idea, perturbing the shapes of the geometrically confined 469 hPSC colonies did not result in fate patterning that deviated from the expected results (Fig. S14) oscillations in accordance with the RD paradigm. Surprisingly, when we tested the response of pSMAD1 475 spatial signalling dynamics in 3mm diameter colonies after BMP4 and SB treatment for 24h, we did not 476 observe any obvious additional foci at either 50ng/ml (Fig. S15) or 200ng/ml (Fig. S16)  Methods for composition), a 24h BMP4 and SB treatment of hPSC colonies of 3mm diameter resulted in 484 rudimentary peaks of PSMAD1 activity at a BMP4 dose of 50ng/ml (Fig. S17) and prominent peaks of 485 pSMAD1 activity at a dose of 200ng/ml (Figs. S18-S19). In addition, after 48h of BMP4 and SB treatment, 486 although we did not note any additional foci of GATA3 in SR medium at BMP4 doses of either 50ng/ml 487 (data not shown) or 200ng/ml (Fig. S20), in an N2B27 basal medium supplemented with SB, rudimentary 488 peaks were observed at 50ng/ml of BMP4 (Fig. S21), and robust peaks were noted at 200ng/ml of BMP4 489 (Figs. 6D, S22). These observations are consistent with the proposition that an RD network self-organizes 490 BMP signalling activity (Fig. 6E) and the patterned fates arise in a manner consistent with PI (Fig. 6C) colonies to give rise to these fates, we treated the colonies with BMP4 and SB for 24h, then tested three 504 different treatments. Specifically, we either treated the colonies for a further 48h with BMP and SB and 505 stained for keratins using a pan-keratin antibody and DLX5 (markers of NNE); or CHIR99021 ('CHIR'a 506 wnt agonist) and SB and stained for SOX10 (a marker of the NC fate); or for a period of 72h with Noggin 507 and SB and stained for PAX6 (Fig. 7A). Consistent with our expected results, we observed that sustained 508 BMP4 and SB treatment resulted in robust expression of DLX5 and showed clear staining of a pan-keratin 509 antibody, indicating acquisition of an NNE identity (Fig. 7B). Furthermore, robust SOX10 staining was 510 observed in colonies treated with CHIR and SB (Fig. 7C); and the colonies treated with Noggin and SB 511 expressed PAX6a bona fide marker of the NP (Fig. 7D). 512 Taken together, our data are consistent with our hypothesis that a RD network in BMP signalling self-513 organizes the pSMAD1 gradient in geometrically confined hPSC colonies and Nodal signalling dissects 514 peri-gastrulation-associated and pre-neurulation-associated fates (Fig. 8)  derived from human cells while maintaining aspects of the structure and organization of the native tissues. 526 Although the field has taken impressive strides towards making organoids for a variety of different organs, 527 achieving a reproducible response between each organoid remains problematic. Furthermore, quantitative 528 image analysis of immunofluorescent data from high-content organoid-based screening studies is currently 529 challenging. An alternative approach for harnessing the potential of employing appropriately organized 530 tissues in screening studies is to start with 2-dimensional cultures of the specific stem/progenitor cells with 531 controlled geometries and allow them to self-organize into 'organoid-like' tissue surrogates. Of late, 532 numerous studies have employed this approach to derive developmentally-relevant tissue organization 533 [24-26,66,67]. Importantly, the response between individual organoid-like structuresfor specific cell lines 534 and medium conditionsis far more reproducible than what can currently be achieved in 3D organoids. 535 Furthermore, given that these organoid-like structures are secured in position for the assay-duration, they 536 are far more amenable to high-content image analysis than their 3D counterparts. The high-throughput 537 platform we report enables robust geometric confinement of a variety of cell types and can be employed 538 for numerous applications. Indeed, we have employed it to micro-pattern human PSCs, mouse PSCs, 539 Retinal Pigmented Epithelial cells, human hemogenic, and pancreatic progenitors, human cardiomyocytes, 540 human keratinocytes, mouse embryonic fibroblasts (MEFs), among others. This platform is poised to be 541 employed for high-throughput drug screens of organoid-like surrogates of a variety of tissues. 542 As proof-of-principle for one such application, we employed the peri-gastrulation-like patterning to assess 543 the differentiation propensity of hPSC lines. signalling to not only orchestrate the self-organization of pSMAD1 activity into a signalling gradient in a 569 manner consistent with RD, but also coordinates the interpretation of the gradient into either peri-570 gastrulation-like, or pre-neurulation-like fate patterns in accordance with PI. 571

Reaction diffusion mechanisms 572 573
In a recent study, we proposed that the pSMAD1 signalling gradient self-organizes under the regulation of 574 a BMP4-Noggin RD system [24]. Our data in this study demonstrates that the pSMAD1 gradient formed in 575 the presence and absence of Nodal signalling is significantly different (Fig. 5, S10) (Fig. 5F), Gremlin family proteins [26], and possibly others can act 585 as the inhibitors that enforce the pSMAD1 signalling gradient; and in conditions permissive of Nodal 586 signalling, CERL and GDF3 and possibly others can further antagonize BMP signalling. In support of this 587 notion, a previous study has reported that siRNA mediated inhibition of CERL and Lefty in experimental 588 conditions permissive of Nodal signalling dramatically compromises the formation of peri-gastrulation-589 associated patterns (Fig. 6 in ref. 25). Taken together, the topology of the RD network in BMP signalling 590 needs to incorporate the role played by Nodal and potentially multiple BMP4 antagonists in addition to 591 Noggin (like CERL GDF3, FST among others). A deeper and more comprehensive understanding of the 592 RD network in the BMP pathway requires further careful studies and computational platforms that enable 593 studying multiple nodes in RD networks will be very valuable [14]. 594 In Nodal-permissive experimental conditions, BMP4 treatment of geometrically confined hPSC colonies 595 results in a gradient that downregulates sharply (Fig. S10B-C)[24-26]. This has led to a proposition that in 596 BMP4 treated geometrically confined hPSC colonies, BMP signalling is active exclusively at the colony 597 periphery and inactive everywhere elsea spatial profile that can be modelled as a step-function along the 598 colony radius [71]. These authors claim that this apparent step-function-like response in the pSMAD1 activity 599 occurs due negative feedback enforced on BMP signalling at two different levels. First, by a BMP signalling 600 mediated upregulation of BMP inhibitors [24,26], and second, due to a cell density mediated re-localization 601 of BMP receptors from being present apically to becoming localized at basolateral regions, rendering them 602 inaccessible for ligand mediated activation, everywhere in the colony except for the periphery [26]. 603 Consistent with their proposed model, we also identify a valuable role of BMP inhibitors in orchestrating the 604 pSMAD1 signalling gradient. However, our data indicate that the underlying mechanism regulating the 605 pSMAD1 self-organization is inconsistent with a wide-spread dampening of BMP signalling owing to 606 inaccessible BMP receptors. In this study, in addition to the unambiguous RD-like spatial expression 607 patterns of pSMAD1 and differentiation markers like GATA3, in large colonies (3mm diameter), we identify 608 experimental conditions that result in prominent expression of pSMAD1 at the periphery and the center of 609 hPSC colonies of 500µm diameter when treated with 25ng/ml of BMP4 for 24hobservations consistent 610 with RD-like behaviour (Fig. 5D). Neither of these observed expression profiles would arise in conditions 611 where BMP receptors were inaccessible everywhere except for the colony periphery. Additional insight has 612 been provided by Xue et al [27] where it was demonstrated that when treated with BMP4 for an extended 613 period of time (4 days), all cells in the hPSC colony expressed nuclear-localized pSMAD1 (Fig 5A-B  with an increase in levels of BMP antagonists like Noggin and Gremlin family proteins owing to more cells 620 secreting these inhibitors, there would likely be a dramatic increase in the levels of BMP antagonists like 621 CERL, GDF3, FST which would upregulate due to pronounced levels of a community-effect mediated 622 increase in endogenous Nodal. Abrogation of the fate patterning in this system in response to siRNA 623 mediated inhibition of CERL and LEFTY (Fig 6 in ref 24)an experimental condition that in principle should 624 neither interfere with Noggin expression or change colony density, also supports our interpretation. Finally, 625 our study provides direct evidence for the involvement of Nodal signalling in the apparent step-function-like 626 response in pSMAD1 expression along the colony radius. When colonies of 500µm diameter were treated 627 with 25ng/ml of BMP4 and SB, the signalling gradient formed showed no step-like response in the spatial 628 pSMAD1 profile. Instead the signalling gradient gradually decreased in strength (as indicated by pSMAD1 629 fluorescence levels) from the colony periphery to the colony center (Fig 5A-D). However, in the presence 630 of Nodal signalling, the pSMAD1 signalling gradient, indeed, downregulated sharplyas expected from a 631 step-function-like response. This is consistent with the proposition that in regions within the differentiating 632 geometrically-confined hPSC colonies that have active Nodal signalling, which have been shown to be 633 immediately interior to the peripheral cells [26], there would be heightened levels of BMP antagonists like 634 CERL causing a dramatic reduction of BMP signalling. 635 Notably, in our study, we report different aspects that can result in variability in experimental results when 636 studying the stereotypic RD-like periodic response in BMP signalling in the hPSC context. One such source of variability is the level of endogenous Nodal signalling between different hPSC lines (Fig. 3C). Given the 638 role that Nodal signalling plays in the formation of the pSMAD1 gradient (Fig. 5, S10), and the critical role 639 it plays in ensuring the peri-gastrulation-associated fate patterning[24], the variability in endogenous levels 640 of Nodal signalling can cause inconsistent responses between different cell lines and culture conditions. 641 Importantly, a recent study has also shown drastically different responses in endogenous Nodal activation 642 within the same hPSC line when cultured under different conditions for routine maintenance[74] -643 highlighting that even culture conditions for routine hPSC maintenance can have an effect in the response 644 in the peri-gastrulation-like assay. Secondly, we observed that when geometrically confined hPSC colonies 645 of 3mm diameter were treated with a high dose of BMP4 in SR medium, the stereotypical RD-like spatial 646 periodicity of either pSMAD1 or GATA3 were not readily observed. However, changing the medium to an 647 N2B27 based medium rescued these periodic responses. A key component of SR medium is Knockout 648 Serum Replacement (KSR) which is known to contain lipid associated proteins like lysophosphatidic acid 649 (LPA), and although the mechanism of action remains unclear, molecules like LPA have been shown to 650 have an inhibitory effect on hPSC differentiation[57,58]. Given the above caveats associated with in vitro 651 experiments, studies aimed at investigating the details of the RD network in BMP signalling in the hPSC 652 contextespecially those directed toward investigating the specifics of the spatial periodicity of morphogen 653 activity and fate patterning, would benefit from removing these sources of variability between hPSC lines. 654 Employing basal medium like N2B27 which is devoid of components like AlbmaxII and LPA, avoiding 655 undefined media like those conditioned on MEFs, and removing Nodal signalling from their system by SB 656 supplementation represent experimental conditions better suited for these studies. 657

Positional information mechanisms 658
In this study, we report that differentiating hPSCs (either culturing EBs in an FBS containing medium, or 659 treatment of geometrically confined colonies with BMP4) upregulate a gastrulation-associated expression 660 profile when endogenous Nodal signalling is active. However, in the case where Nodal signalling is 661 downregulated, the same differentiation pulse upregulates a neurulation-associated gene expression 662 profile. In addition, we demonstrated that perturbing Nodal signalling during BMP4 treatment of 663 geometrically-confined hPSC colonies can result in a PI mediated interpretation of the emergent pSMAD1 664 signalling gradient into peri-gastrulation-associated or pre-neurulation-associated fate patterning. Notably, 665 inducing the pre-neurulation-like fates did not require an initial differentiation toward the ectodermal lineage 666 prior to inducing a pSMAD1 signalling gradient within the colonies. This observation, although apparently 667 contradictory from a developmental point of view, is consistent with previous in vitro studies. For instance, 668 the neuromesodermal precursorsa developmental population that arises during late gastrulation and 669 resides in the node-streak border, caudal lateral epiblast, and provides further evidence to the proposition that the fate patterning occurs in a manner consistent with PI 713 as at reduced levels of BMP signalling, the fate patterning would be predicted to arise after longer durations. 714 In addition, under these conditions where BMP signalling has not activated at sufficiently high levels, 715 consistent with the PI model, the authors did not observe fates associated with the non-neural ectodermal 716 identity. 717 718

719
In conclusion, we report characterization of a high-throughput microtiter plate that enables robust geometric 720 confinement of a variety of cell types. We employ this platform to screen hPSC lines for their ability to induce 721 gastrulation-associated fate patterning and observe a Nodal-dependent response in the efficiency of BRA 722 (a gastrulation-associated fate) induction, thereby providing a proof of principle of the ability of this platform 723 to be employed for high-throughput screening experiments. In addition, we identify that differentiating 724 hPSCs upregulate either gastrulation, or neurulation associated gene profiles in a Nodal signalling 725 dependent manner. Further, we demonstrate that in BMP4 treated geometrically-confined hPSC colonies, 726 Nodal signalling can affect the RD mediated self-organization of pSMAD1the downstream effector of 727 BMP signalling; and that it also regulates the switch between peri-gastrulation-like and pre- with 20 ng ml−1 FGF-2 (PeproTech). H1, H7, and H9-3 cells were passaged 1:6 every 4-5 days and were 751 disassociated into small clumps using 0.1% collagenase IV (Invitrogen). HES3-1, HES3-2 were passaged 752 1:24 every 4-5 days and dissociated using TryplE Express (Invitrogen). All cell lines were confirmed 753 negative for mycoplasma contamination. 754 755 Platform set up, and XPS studies were performed using 22mmx22mm borosilicate coverslips (Fisher 756 Scientific), and the 96-well plate platform was developed using custom sized (110mmx74mm) Nexterion-D 757

Preparation of PEG plates
Borosilicate thin glass coverslips (SCHOTT). The glass coverslips were activated in a plasma cleaner 758 (Herrick Plasma) for 3 minutes at 700 mTorr and incubated with 1 ml of Poly-L-Lysine-grafted-Polyethylene 759 Glycol (PLL-g-PEG(5KD), SUSOS,) at a concentration of 1 mg/ml at 37°C overnight. The glass slides were 760 then rinsed with ddH2O and dried. The desired patterns were transferred to the surface of the PEG-coated 761 side of the coverslip by photo-oxidizing select regions of the substrate using Deep UV exposure for 10 762 minutes through a Quartz photomask in a UV-Ozone cleaner (Jelight). Bottomless 96-well plates were 763 plasma treated for 3 minutes at 700 mTorr and the patterned slides were glued to the bottomless plates to 764 produce micro-titer plates with patterned cell culture surfaces. Adhesives validated for biocompatibility 765 standards ISO10993, and USP Class VI were utilized for the assembly of the plates. Prior to seeding cells 766 onto the plates, the wells were activated with N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide 767 hydrochloride (Sigma) and N-Hydroxysuccinimide (Sigma) for 20 minutes. The plates were thoroughly 768 washed three times with ddH2O, and incubated with Geltrex (diluted 1:150) for 4h at room temperature on 769 an orbital shaker. After incubation, the plate was washed with Phosphate Buffered Saline (PBS) at least 770 three times to get rid of any passively adsorbed extracellular matrix (ECM) and seeded with cells to develop 771 micro-patterned hPSC colonies. 772 Comparison between PEG plates with μCP plates 773 PEG plates (as described above) and μCP plates (as reported previously[30]) were generated with 774 patterned islands of 200μm in diameter with 500μm separation between adjacent colonies. A single cell 775 suspension of CA1s was generated by incubating in 1ml of TryplE (Invitrogen) per well for 3 minutes at  776 37°C. The TryplE was blocked using in equal volume SR medium (see 'Human pluripotent stem cell culture' 777 section above for composition) and the cells were dissociated by pipetting to generate a single cell 778 suspension. The cells were centrifuged into a pellet and the supernatant aspirated to remove any residual 779 TryplE. A single cell suspension was then generated in SR medium supplemented with 10μl of ROCKi and 780 20ng/ml of bFGF at a cell density of 500,000 cells/ml and 100μl of the suspension was plated onto the PEG 781 and μCP plates for a period of 2-3h till robust cell attachment was observed. The cells were then left to 782 make a confluent colony overnight (~12h The hPSC lines that were cultured in feeder-dependent techniques for routine maintenance were first feeder 799 depleted by passaging the cells at 1:3 on geltrex and cultured on Nutristem. To seed cells onto ECM-800 immobilized PEG-UV 96-well plates, a single cell suspension of the hPSC lines was generated as described 801 above. The cells were centrifuged and re-suspended at a concentration of 1 x 10 6 cells/ml in SR medium 802 supplemented with 20ng/ml bFGF (R&D) and 10µM ROCK inhibitor Y-27632. Wells were seeded in the 803 PEG-patterned 96 well plates at a density of 60,000 cells/well for plates with colonies of 500μm diameter, 804 80,000 cells/well for colonies of 1mm diameter, and at 120,000 cells/well for plates with colonies of 3mm 805 diameter and incubated for 2-3h at 37°C. After 2-3h, the medium was changed to SR without ROCKi. When 806 confluent colonies were observed (12-18h after seeding), the peri-gastrulation-like induction or pre-807 neurulation-like induction was initiated as follows. A) Peri-gastrulation-like induction (Fig. 2) was performed 808 in SR medium supplemented with 100ng/ml of bFGF (R&D) and 50ng/ml of BMP4. B) Unless otherwise 809 stated, pre-neurulation-like induction with 500μm colonies was performed with SR medium (see 'Human 810 pluripotent stem cell culture' section above for composition) supplemented with 100ng/ml of bFGF with 811 25ng/ml of BMP4, and 10μM SB431542 ('SB'). C) Endoderm fingerprinting assay (Fig. S6) was performed 812 with N2B27 medium supplemented with 25ng/ml of Wnt3A and 50ng/ml of ActivinA. D) RD-like periodic 813 pattern induction of pSMAD1 activity, and pre-neurulation-like fates was tested in both SR, and N2B27 814 mediums. In the case of SR, the medium was supplemented with 10μM SB, 100ng/ml of bFGF, and either 815 50ng/ml or 200ng/ml of BMP4. In the case of N2B27, the medium was supplemented with 10μM SB, 816 10ng/ml of bFGF, and either 50ng/ml of 200ng/ml of BMP4. 817 Embryoid body differentiation assay 818 The differentiation media for the EB assay contained 76% DMEM, 20% Fetal Bovine Serum (FBS), 1% 819 Penicillin/Streptomycin, 1% non-essential amino acids, 0.1mM β-mercaptoethanol, 1% Glutamax, (all 820 Invitrogen individually or applied together to produce Noggin knock-out. Latter leads to a deletion of a DNA fragment 838 of 112 bp and to a predetermined stop codon (Fig. S8A). 839 Transfection and evaluation of cutting efficiency: We first evaluated the cutting efficiency of SpCas9 for 840 each individual gRNA on a population level. For this, we seeded CAI hESC into 24-well plates such that 841 they are 50-60 % confluent at the day of transfection (approx. 24h after seeding). CmgRNA were generated 842 by mixing 1μM AltR CRISPR crRNA (IDT, custom oligo entry) with 1μM AltR CRISPR tacrRNA (IDT, Cat. 843 1073189), annealed at 95°C for 5 min and cooled down at room temperature. GeneArtTMPlatinumTM Cas9 844 Nuclease (Invitrogen, B25641) was diluted to 1μM using Opti-MEM (Thermo Fisher Scientific, 31985062). 845 Cas9 and cmgRNA were mixed at a concentration of 0.3μM each in 25μl of OptiMEM. After incubation at 846 room temperature for 5 minutes, 1μl of EditProTM Stem (MTI Globalstem) diluted in 25μl of Opti-MEM was 847 added to the Cas9/cmgRNA complex and incubated for 15 minutes at room temperature. Before adding 848 the reagent -Cas9/cmgRNA mix to the cells, medium was replaced with 500μl / 24 well of fresh mTeSR. 849 Medium was replaced 24h after transfection. 48h after transfection, cells were harvested by incubation in 850 Gentle Cell Dissociation Reagent (STEMCELL Technologies, 07174) for 7 min. Dissociation reagent was 851 removed and cells were resuspended in cultivation medium, pipetted to single cells and spin down for 5min 852 at 200g. Cells were resuspended in 25μl of Cell Lysis Buffer mixed with 1μl Protein Degrader, both from 853 the GeneArtTM Genomic Cleavage Detection Kit (Invitrogen, A24372). Cells were lysed at 68°C for 15min, 854 95°C for 10min and kept on ice. PCR was performed using Phusion High Fidelity DNA Polymerase (NEB, 855 M0530) according to manufactures protocol using 2μl of the cell lysate. Primer for the PCR were the 856 following: (fwd) 5'CTACGACCCAGGCTTCATGGC'3, (rev) 5'GACGGCTTGCACACCATGC3'. PCR 857 product of un-transfected and transfected samples were analyzed on 2.5% MetaPhore Agarose Gel (Lonza) 858 PCR products were analyzed using GeneArt Genomic Cleavage Detection Kit (Invitrogen, A24372) 859 according to manufacturer's protocol. The cleaved and un-cleaved samples were loaded on 2.5% 860 MetaPhore Agarose Gel (Lonza) and the bands were analyzed using ImageJ. Percentage of gene 861 modification was calculated as described in a previous report (Fig. S8B)[80]. Additionally, PCR products 862 were send for Sanger Sequencing. Chromatograms were analyzed using TIDE [81]. 863 Cell line generation: The cell line was generated using gRNA1 & 2 mixed with Cas9 at 0.3μM each. The 864 transfection was proceeded with exact same protocol as described above using 6 x 24 wells. After three 865 days, cells reached confluency and were seeded to 6 well plates at sufficiently low densities to achieve 866 clonal growth from single cells. Approximately 7 days after seeding, single clones were picked and 867 transferred to 96 well plates. 24 clones were expanded for 2 passages and PCR was performed on cell 868 lysates as described above (Fig. S8C). PCR products were send for Sanger Sequencing and aligned to 869 (NCBI ID9241) and to untransfected wildtype sequence (Fig. S8C). Clones with clear loss of function 870 mutations in both alleles (C1 and C7) were further characterized for their pluripotency marker expression 871 (Fig. S8D). 872 Quantitative PCR analysis 873 RNA extraction for all gene expression analysis studies was performed using Qiagen RNAeasy miniprep 874 columns according to the manufacturer's protocol, and the cDNA was generated using Superscript III 875 reverse transcriptase (Invitrogen) as per the manufacturer's instructions. The generated cDNA was mixed 876 with primers for the genes of interest and SYBR green mix (Roche, Sigma) and the samples were run on 877 an Applied Biosystems QuantStudio 6 flex real-time PCR machine. The relative expression of genes of 878 interest was determined by the delta-delta cycle threshold (∆∆Ct) method with the expression of GAPDH 879 as an internal reference. Primer sequences used are provided in Supplementary Information (Table S2). 880 Immunofluorescent staining, and image analysis 881 After the peri-gastrulation-like or the pre-neurulation-like induction was completed, the plates were fixed 882 with 3.7% paraformaldehyde for 20 min, rinsed three times with PBS and then permeabilized with 100% 883 methanol for 3 min. After permeabilization, the patterned colonies were blocked using 10% fetal bovine 884 serum (Invitrogen) in PBS overnight at 4°C. Primary antibodies were incubated at 4°C overnight (antibody 885 sources and concentrations are shown in Table S3). The following day, the primary antibodies were 886 removed, and the plates were washed three times with PBS followed by incubation with the secondary 887 antibodies and DAPI nuclear antibody at room temperature for 1 h. Single-cell data were acquired by 888 scanning the plates using the Cellomics Arrayscan VTI platform using the ' TargetActivation.V4' bioassay  889 algorithm. This algorithm utilizes the expression intensity in the DAPI channel to identify individual nuclei in 890 all fields imaged and acquires the associated intensity of proteins of interest localized within the identified 891 region. As previously described[24], single-cell data extracted from fluorescent images were exported into 892 our custom built software, ContextExplorer (Ostblom et al, unpublished), which classifies cells into colonies 893 via the DBSCAN algorithm. Cartesian coordinates relative to the colony centroid are computed for every 894 cell within a colony. Hexagonal binning is used to group cells from multiple colonies according to their 895 relative location within a colony. Average protein expression of cells within a bin is represented by the color 896 map, which is normalized to the lowest and highest expressing hexagonal bins. In the line plots of spatial 897 expression trends, cells are grouped in annular bins according to the Euclidean distance between a cell 898 and the colony centroid. For each colony, the mean expression of all cells within an annular bin is computed. 899 The average of all the colony means is displayed in the line plot together with the standard deviation and 900 the 95% confidence interval (CI).  protocol for transferring carboxyl-rich micro-patterns onto glass coverslips. B-C) Carbon 1s (C1s) spectra 905 acquired using X-Ray Photoelectron Spectroscopy. B) C1s spectra of glass coverslip incubated with PLL-906 g-PEG compared to blank glass coverslip. C) C1s spectra of PLL-g-PEG coated glass coverslips photo-907 exposed to Deep-UV light for different times of exposure. Dotted lines signify binding energies associated 908 with untreated glass (285.0eV), or presence of PEG (286.6eV). D) Line plot representation of detected