Array-CNCC: precise aggregation and arrayed plating facilitate quantitative phenotyping of human cranial neural crest cells and craniofacial disease modelling

  1. Ewa Ozga
  2. Katarzyna M Milto
  3. Martina Demurtas
  4. Lawrence E Bates
  5. Graeme Grimes
  6. Takuya Azami
  7. Jing Su
  8. Carlo De Angelis
  9. Marco Trizzino
  10. Jennifer Nichols
  11. Hannah K Long

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Facial development is highly sensitive to genetic and environmental perturbation, with craniofacial malformation associated with over one-third of congenital birth defects. The face arises during an early and largely inaccessible window of embryonic development, with a large contribution from transient and multipotent cranial neural crest cells (CNCCs). Assessment of the molecular and cellular mechanisms driving normal and disordered human facial development therefore relies greatly on the use of in vitro cellular models. Here, we adapted a neurosphere-based CNCC differentiation protocol to facilitate robust quantification of early specification and migration events. Introduction of single-cell aggregation with arrayed plating enabled standardisation of neurosphere size, growth and patterning. Inclusion of fibronectin coating enhanced the efficiency of neurosphere attachment and synchronicity of CNCC migration timing. To demonstrate application of the Array-CNCC method, we developed a strategy for mosaic co-culture, which can facilitate differentiation of wildtype untreated cells directly alongside cells exposed to distinct drug treatments or genetic alterations. Finally, we present a screening approach which we use to test the impact of distinct extracellular matrix components on neurosphere morphology, CNCC migration and gene expression. Together, the Array-CNCC method is highly amenable to quantitative phenotyping and screening approaches, enabling enhanced craniofacial disease modelling with both cellular and molecular readouts.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    The authors do not wish to provide a response at this time.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Summary:

    Ozga, et al. develop a more reproducible method for cranial neural crest cell (CNCC) differentiation from human pluripotent stem cells, which is compatible with arrayed screening techniques such as automated microscopy. The method is based on a prior neurosphere aggregation protocol, but optimized here for cell number and attachment conditions.

    Major comments:

    The key conclusions are overall well-supported for CNCC specification and migration, but I have three major comments that could be reasonably addressed without much additional work:

    1. The gene expression analysis lacks statistical power, and does not enable a comprehensive transcriptome-wide comparison with previously published differentiation methods. A minimum of 3-4 technical replicates are necessary and standard for bulk RNA-seq analyses with DESeq2. This explains why the number of differentially expressed genes and enriched GSEA terms is so low. While single-cell analysis is likely beyond the scope of this methods paper, I would have liked to see a comprehensive expression analysis that enables readers to understand how this method compares to prior methods (not only the standard method included here).
    2. H9 hESCs are biased towards neural differentiation. Although the authors include an unrelated hiPSC line (SV20) to demonstrate CNCC differentiation by immunofluorescence (IF) and qRT-PCR, these cells were not included in the bulk RNA-seq study, and IF only included passage1 (Fig. S6). The homogeneity of H9-derived CNCCs at passage 4 is intriguing (Fig. 5BC), and I would like to see the same staining for SV20-derived CNCCs at passage 4, to learn whether this method is broadly applicable to other human pluripotent stem cells.
    3. It is unclear whether the different substrates bias CNCC toward specific terminal cell fates. Whether true CNCC multipotency is maintained is not addressed. I would have liked to see stainings for terminal CNCC-derived cell types to rule out that some matrices impact terminal differentiation. The DESeq2 analysis does not address this, since it's underpowered with only 2 replicates per condition. Also in the IF analysis there were more SOX9-TWIST+ cells with FN; do these cells differentiate more quickly in the presence of FN?

    Minor comments:

    When CNCCs are passaged, is the core left behind during accutase treatment or removed prior to accutase? If it's not removed, how is passage 4 so homogenous? Do SOX2+SOX9- cells become SOX9+ upon accutase passage, or are they lost from culture? Fig. 5C: the legend is clever but not that helpful to the reader - the supplementary figure has a simpler legend. Fig. 5D: This should be a boxplot, as it is difficult for the reader to discern color gradients on a log-scale. But the authors need at least 1-2 more RNA-seq replicates for a boxplot.

    Significance

    The authors framed the significance of the work appropriately, provided the authors can address the major comments that describe what caveats currently limit the impact of this study.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    A study by Ozga et al. presents a new protocol and methodology for neurosphere-based differentiation of cranial neural crest cells (CNCCs) from human embryonic stem cells (hESCs), designed to enable robust quantification of early developmental processes, including specification, delamination, and migration. The method utilizes single-cell aggregation combined with arrayed plating to standardize neurosphere size, adherence, growth, and CNCC formation. The authors demonstrate that this approach yields a more robust and experimentally tractable in vitro CNCC model for quantitative phenotyping and screening compared with standard protocols.

    The study further shows that this in vitro system can be used to evaluate the effects of distinct extracellular matrix (ECM) components on CNCC behavior. The authors also propose that the model could be applied to investigate CNCC migratory patterns, gene expression changes, drug responses, and genetic perturbations. Overall, the manuscript is well written and logically organized, and it introduces a promising in vitro platform that may facilitate investigations into human birth defects associated with abnormalities in CNCC development and differentiation, which contribute to the formation of orofacial structures and the peripheral nervous system.

    Major Comments

    1. The authors provide transcriptional profiling and immunofluorescence staining of key transcription factors to support their claim that the differentiated cells derived from hESCs possess CNCC characteristics. However, a defining property of CNCCs is their multipotency and their ability to differentiate into multiple lineages, including osteoblasts, chondrocytes, neurons, and smooth muscle cells. It would be important to clarify whether the differentiation potential of the derived CNCCs was experimentally evaluated. If such assays were not performed, the authors should discuss how their current data support the multipotent identity of the cells and their capacity to generate these distinct lineages.
    2. The neural crest differentiation medium and the neural crest long-term maintenance medium contain BMP2 and the Wnt agonist CHIR-99021. The manuscript does not provide sufficient justification for the inclusion of these factors. Additional explanation should be provided regarding why these components were selected and which downstream signaling pathways or transcriptional programs they are expected to activate during CNCC differentiation and maintenance.
    3. The authors state that the large number of differentially expressed genes observed across different ECM conditions likely reflects signaling events downstream of ECM binding. However, no direct evidence is presented to support this claim. To strengthen this conclusion, the authors should identify enriched or upregulated signaling ligands and receptors (signaling pathways) associated with ECM-cell interactions in the ECM comparison datasets and in the passage-4 CNCCs maintained in the long-term maintenance medium.
    4. In Figure 7H, additional explanation is needed for why genes associated with early neural tube identity and patterning (OTX2, PAX6, WNT8A/B, ZIC1, and MEIS1) are enriched in cells cultured on vitronectin-coated plates. The authors should clarify whether this observation reflects a shift in cell identity or whether it could be due to contamination by non-CNCC populations originating from the neurospheres.
    5. The authors report that 10,000-cell aggregates frequently developed necrotic centers and occasionally dissociated into smaller clusters, whereas 8,000-cell aggregates did not display these features. Additional clarification would be helpful to explain what distinguishes the 10,000-cell aggregates from the smaller aggregates. For example, differences in oxygen or nutrient diffusion, mechanical stability, or signaling gradients could potentially account for this observation.
    6. In Figure 5B (ARRAY-FN condition), the level of TWIST1 protein appears reduced in CNCCs compared with other culturing conditions. However, the transcriptomic data presented in the heatmap (Figure 5D) do not indicate a corresponding difference at the mRNA level. The authors should comment on this apparent discrepancy and discuss whether post-transcriptional regulation, protein stability, or methodological differences between assays might explain the observation.

    Minor Comments.

    -None

    Referee cross-commenting

    In line with the reviewer's comment, I suggest performing statistical analysis and significance testing for all quantitative measurements.

    Significance

    The authors demonstrate that this neurosphere-based Array-CNCC yields a more robust and experimentally tractable in vitro CNCC model for quantitative phenotyping and screening compared with standard protocols. The study introduces a promising in vitro platform that may facilitate investigations into human birth defects associated with abnormalities in CNCC development and differentiation, which contribute to the formation of orofacial structures and the peripheral nervous system.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Ozga et al present an adapted method of generating neurospheres derived from human cell lines in order to standardize CNCC generation. They utilize multi-well plates and single cell aggregation to standardize and synchronize neurosphere generation and maintenance, namely growth, patterning, attachment and migration. They also show examples of possible downstream experiments. Their technique allows for comparisons between environments, conditions, genetic disturbances and drug responses, to better our understanding of neural crest cell processes and associated diseases, while reducing sample variability. Minor suggestions are outlined below.

    Major Comments

    There needs to be some discussion between the array CNCC protocol under consideration and other differentiation protocols. In the Okuno et al protocol, (2017) hiPSC is cultured for 3d in ES Media to generate embryo bodies before switching to NC media to generate NCC. How would this new protocol be amenable to alternative methods of 3D neural crest cultures such as the one by Okuno et al, and others? Comparison of this work to other methods of NCC 3D culturing would be beneficial. In addition, Ozga et al., mention that the most standard method for NCC culturing is either 2D culturing or 3D culturing. The manuscript would be improved if the authors discuss how this proposed method compares to commonly used 2D culturing. Overall, additional discussion about how the array CNCC protocol improves other commonly used NCC differentiation techniques is necessary to better place the impact of this work.

    Minor comments:

    1. FIG 1: It is not clear which of the in vivo steps listed are aligned to stages of the in vitro culture system proposed. Could the authors add this comparison either in the figure or in the text that describes the technique?
    2. Pg7 LINE 175: Neurospheres generated from a larger number of cells (specifically 8000-10000) are described as having a necrotic core and figure 2 is referenced. Could this be pointed out on the image or a different image used to show this necrotic core?
    3. FIG S3b: It is not clear whether the n=2 is referring to 2 plates generated from one experiment or 2 separate experiments. Please be more explicit about what the n's mean throughout
    4. FIG S7A: Could the same picture style and quantification be done for the ones expressing mcherry, then perhaps a graph of distribution of cells (y axis 100%) expressing mcherry vs gfp (based on your principle both should be close to 50%)
    5. I suggest that the section describing the effects of different ECM (Pg14 LINE 438-523 and associated Figure) be moved to immediately after the section outlining fibronectin's role on morphology and gene expression (Pg 10 LINE 313-347) and before the section describing long-term passaging effects of fibronectin use (Pg 12 LINE 349-379). The wording of the introductory paragraph of this section can be modified and moved towards the end of the paragraph, stressing that this experiment was to determine the best ECM protein to use and to depict an example of a downstream application of this technique.
    6. Pg 17 LINE 539-40 seems to be referring to the wrong figure
    7. Pg 17 LINE 544-545, it was not presented in the results that there was an analysis of neurospheres generated from fewer than 1000 cells, though this is a logical conclusion. If this analysis was not done, please make it clear that this is a presumption.

    Referee cross-commenting

    In line with comments made by reviewers 2 and 3, an exploration into if the proposed method of culturing affects the multipotency of cNCCs would be beneficial. cNCCs are capable of differentiating into chondrocytes, osteoblasts, neurons etc. The impact of the different ECM proteins on capability to conduct terminal differentiation should be addressed, either experimentally or with logical conclusions from the available RNA-seq data and literatures.

    Significance

    This work highlights an advancement in techniques used to study neural crest specific processes. It allows for the standardization of samples that can be used for environmental, genetic and drug-based studies. It adapts and expands on techniques previously used in mouse embryonic stem cells and shows its use in human embryonic stem cells and IPSCs. Readers might be interested in downstream analyses made possible by this technique, such as mosaic co-culturing and array plating. Better contextualization of this work in the context of currently used differentiation techniques will add power to the overall message.

  5. Version published to 10.64898/2026.01.18.696654 on bioRxiv