Precise and scalable self-organization in mammalian pseudo-embryos

TL;DR

This study reveals intrinsic reproducibility in the self-organization of gastruloids, encompassing growth dynamics and gene expression patterns, suggesting that these phenomena might constitute fundamental features of multicellularity.

Abstract

Gene expression is inherently noisy, posing a challenge to understanding how precise and reproducible patterns of gene expression emerge in mammals. We investigate this phenomenon using gastruloids, an in vitro model for early mammalian development. Our study reveals intrinsic reproducibility in the self-organization of gastruloids, encompassing growth dynamics and gene expression patterns. We observe a remarkable degree of control over gene expression along the main body axis, with pattern boundaries positioned at single-cell precision. Furthermore, as gastruloids grow, both their physical proportions and gene expression patterns scale proportionally with system size. Notably, these properties emerge spontaneously in self-organizing cell aggregates, distinct from many in vivo systems constrained by fixed boundary conditions. Our findings shed light on the intricacies of developmental precision, reproducibility, and size scaling within a mammalian system, suggesting that these phenomena might constitute fundamental features of multicellularity.

Figures (15)

• Figure 1: Reproducible gastruloid growth, scaled to system size. A: Gastruloid midline length variation as a function of time. Curves shown for 57 gastruloids followed individually over time (blue) and mean (black). Percent variation around the mean is reported for each time point. Spread of initial number of seeded mESCs is $N_0=305\pm28$ cells (Fig. S1A). (For an equivalent relationship for volume see Fig. S2A.) Inset shows a brightfield image of a gastruloid at $120\,{\rm h}$, overlaid with its midline ranging from anterior (A) to posterior (P) pole (red, top) and sliced evenly for volume reconstruction (yellow, bottom); scalebar is $100\,$µm; also see Fig. S1B-D. B: Gastruloid volume and total cell count (inset and Fig. S1E) as a function of time. Volumes are normalized by the average number of initial seed cells $\overline{N}_0$ at time zero (color code). Each line represents the mean of on average 15 gastruloids with the same $N_0$. Percentages correspond to residual variations within which normalized volumes collapse for 17 different values of $\overline{N}_0$. Similar collapse for normalized gastruloid cell counts for four values of $\overline{N}_0$ (inset). C: Scatter plot of total cell count versus the measured volume for 492 individual gastruloids at different time points (color code) and with varying $\overline{N}_0$ (symbol); Pearson correlation coefficient is $r=0.99$. Inset shows correlation ($r=0.78$) of variability for $N$ and $V$ for sets of gastruloids with identical age and $N_0$. This is evidence that the independent methods for measuring $N$ and $V$ are accurate estimates of gastruloid growth. D: Cell count $\overline{N}(t)/\overline{N}_{300}(t)$ as a function of the initial seed cell count $\overline{N}_0/300$ in units of the reference seed at $\overline{N}_0=300$. Time is encoded by color (see legend). Black diagonal (slope = 1) represents perfect scaling of gastruloid size at time $t$ upon changes in $\overline{N}_0$ ranging over $50\le \overline{N}_0 \le 1100$. Dashed line estimates expected deviations from perfect scaling at $120\,{\rm h}$ due to fluctuations in $\overline{N}_0/300$ and in the doubling time $t_{\rm D}$ given a simple exponential growth model (Fig. S2E and Methods). Detailed representations for individual time points can be found in Fig. S2G. Inset shows the same relationship centered around $\overline{N}_0=300$ where the regression slope is statistically indistinguishable from one at all time points (see Table S1).
• Figure 2: Reproducibility of gene expression patterns in gastruloids. A: Maximum projections of four confocal image stacks of $120\,{\rm h}$ old gastruloids stained by immunofluorescence for SOX2, CDX2, BRA, FOXC1. AP-axis is in a left--right orientation. Scalebar is $100\,$µm. B: n individual raw gene expression profiles (light color) for the four markers in A and the corresponding average profile (dark bold) projected on the midline and reported relative to gastruloid length $L$. C: Variability ($\sigma_I/\overline{I}$) of the respective gene expression patterns from B as a function of relative position along the midline $x/L$. Error bars are obtained by bootstrapping. Dashed lines represent the average variability in the region where genes are most highly expressed (values in B, see Fig. S5B).
• Figure 3: Single-cell pattern boundary precision in gastruloids. A: Close-up of Fig. 2B for five SOX2 expression profiles as a function of position along the midline (green), with pattern boundary positions of individual profiles marked at the half-maximal expression value (EC50, blue crosses). Mean profile of gastruloid midlines in dark green (n=44). Dashed line is at mean position for these five profiles. B: Distribution of SOX2 pattern boundary positions from Fig. 2B. The mean defines the pattern boundary position $x_B$ (n=44); the standard deviation of this distribution (blue bar in A) of $2.4\%$ defines the positional error for pattern boundary establishment. C: Positional error directly calculated from the standard deviation of intensity values across the individual expression profiles in A, $\sigma_I(x/L)$. For each position $x/L$, this expression error is propagated into an error in position, $\sigma_{x/L}$ (see Methods). Color code as in Fig. 2; gray areas correspond to one and two effective cell diameters $d_c$, respectively, including measurement errors (Fig. S7 and S8).
• Figure 4: Scaling of AP gene expression patterns in gastruloids. A: Confocal images of gastruloids immunofluorescently stained for four different genes each representing a different initial seed number $\overline{N}_0$ (in white, also Fig. S10). AP-axis from bottom to top. Scalebars are $100\,$µm. B: Normalized mean expression profiles for sets of gastruloids with the same $\overline{N}_0$ (color code in C) as a function of the relative position $x/L$ along the average midline of the respective set (n=15--50 gastruloids per $\overline{N}_0$). AP-axis is in a left--right orientation. Inset shows normalized mean expression profiles as in B as a function of average position in absolute units of the respective set. C: Boundary position $x_{B}$ in absolute units of individual gastruloids seeded with varying $\overline{N}_0$ (same color code as in B) as a function of absolute individual gastruloid length $L$. Bold diagonal line indicates gastruloid length ($x_{B} = L$). Dashed line shows linear fit with intercept at zero. Perfect scaling would imply $R^{2}=1$, meaning that 100% of the observed boundary position variance is related to gastruloid length. Here, for the genes SOX2, CDX2, and BRA, the scaling relationship with respect to gastruloid length explains 96--97% of the boundary position variance.
• Figure S1: Experimental detail, protocols, and analysis. A: Gastruloid protocol as described before with a Chi-pulse on day three Beccari2018a. Initial seeding either done by manual multi-pipetting or using Fluorescence-activated Cell Sorting (FACS) VandenBrink2020, implying a different variability in the initial number of seeded cells $N_0$; 10% vs. 2%, respectively. Blue arrows indicate addition of Chiron and change of medium. B: Discarding all gastruloids grown in outer wells for increasing reproducibility. Empirical observation determined largely from different behaviors for gastruloids grown in inner versus outer wells Mansoury2021. C: Image analysis steps include the definition of a smooth contour (I), drawing the midline (II), and slicing along this midline using an equidistant positioning of two sets of equal-number points on each side of the contour (III). For III, the points in left half (light blue) and in right half (dark blue) are equidistant along the contour, respectively. Gastruloid volume is reconstructed by assuming each slice is rotationally symmetric (i.e., a truncated cone). Scalebar is $100\,$µm. D: Gastruloids imaged with brightfield microscopy. Gastruloid elongation efficiency is 97% for multi-pipetting and 99% for FACS seeding, for $\overline{N}_0$. The remaining gastruloids have multiple poles (e.g., red framed image). Scalebar is $100\,$µm. E: Schematic of the protocol to measure the volume and cell count of individual gastruloids. Brightfield images of gastruloids are acquired before chemical dissociation, left; fluorescent images of all individual cells composing the gastruloid are acquired after dissociation using confocal microscopy (see Methods).