Transcriptional coregulation in cis around a contact insulation site revealed by single-molecule microscopy

Reviewer #1 (Public review):

In this manuscript, Kerlin et al. introduce a novel and conceptually important framework for analyzing allelic transcriptional heterogeneity using single-molecule microscopy. The authors aim to distinguish regulatory interactions occurring in cis-between genes on the same allele-from those in trans, between alleles, thereby extending classical models of transcriptional noise into the spatial and allelic domain. They apply this approach to three genes within the FOS locus in MCF7 cells, under both basal and estrogen-induced conditions, and report distinct patterns of transcriptional coordination that depend on gene proximity and chromatin insulation.

A major strength of this work lies in its innovative methodology and the clarity with which the analytical framework is described. The authors effectively build on foundational ideas in gene expression variability and adapt them to resolve a previously underexplored question - how nearby genes on the same allele may influence each other's transcriptional activity. The imaging data are of high quality, the mathematical derivation is comprehensive, and the overall presentation is strong. The study makes a compelling argument for the value of allele-resolved analysis, highlighting that failure to account for allelic and chromatin context may lead to inaccurate or incomplete interpretations of regulatory mechanisms.

That said, the scope of the data is currently limited to a single locus in one cell type. As such, some of the general conclusions, particularly those in the abstract and discussion, may be overstated. The evidence supports the findings within the FOS locus, but it remains unclear whether the observed patterns apply broadly across the genome. The utility and generality of the method would be significantly strengthened by additional validation.

One specific area where the analysis could be improved is through the inclusion of randomized control comparisons. For example, the results presented in Figure 2D and analyzed in Figure 3 could be compared against randomized datasets to establish a baseline of what would be expected by chance. This would help determine the significance of the observed correlations and strengthen confidence in the model's specificity.

Additionally, the framework should be tested on simulated datasets with a known ground truth to evaluate the robustness of its assumptions and the reliability of its outputs. Testing the approach against existing allele-specific single-cell datasets from other studies would also help assess its generalizability. While the authors suggest the framework could be extended to transcriptomics and spatial omics, these possibilities are not explored in the current study, and future work in this direction should be clearly marked as such.

In summary, this manuscript presents a methodologically rigorous and biologically significant advance in the study of gene regulation. The approach fills an important gap by enabling allele-resolved, locus-specific analysis of transcriptional coordination, with implications for both basic science and clinical applications. The conclusions are well supported within the studied context, but further validation - particularly through randomized data comparison, simulations, and broader application - would be valuable in assessing the broader utility of the framework.

Reviewer #2 (Public review):

Summary:

I am not familiar with mathematical modeling of gene expression, so I will evaluate this manuscript solely from a biological point of view.

Kerlin et al. combined single-molecule RNA FISH and mathematical modeling approaches to quantitatively characterize changes in the transcriptional dynamics of three neighboring genes at the FOS locus in response to estradiol (E2) stimulation. They showed that the neighboring JDP2 and BATF genes, located on the same side of the TAD boundary, exhibit highly coordinated bursting dynamics. While FOS and JDP2/BATF are strongly insulated (~7:1 intra-to-inter-domain contact ratio) by the TAD boundary, correlated bursting dynamics were still observed between these gene pairs, suggesting that enhancers can bypass strong insulation sites. The authors proposed that burst co-occurrence arises from the activity of ERα-bound enhancers at the locus. They also proposed that the burst size correlation between two neighboring genes located on the same side of the TAD boundary results from local spreading of histone marks.

Strengths:

The direct visualization of coordinated transcriptional bursting across a strong insulation site is novel. This finding was carefully analyzed using the mathematical framework developed by the authors.

Weaknesses:

Several models were proposed based on single-molecule RNA FISH analysis of the FOS locus, but the generality of these findings remains uncertain. The proposed models were not directly tested through follow-up experiments, leaving the authors' conclusions largely speculative.

Reviewer #3 (Public review):

Summary

Kerlin et.al combined single-molecule RNA FISH with oligonucleotide-based DNA FISH to directly examine the transcriptional activities of three adjacent genes at individual alleles in MCF7 cells. Importantly, they provided quantitative methods to resolve allele-specific (cis) and cell-to-cell (trans) variation and quantified the contribution of burst co-occurrence and burst size, which may help to more accurately analyze transcription coregulation. They found that transcriptional variability is largely gene-autonomous, and by disentangling burst co-occurrence and burst size after E2 induction, they proposed two distinct mechanisms of local gene regulation.

Strengths:

(1) Innovative Research Methods: Successfully integrates single-molecule RNA FISH with oligonucleotide-based DNA FISH to directly image the transcriptional activities of three adjacent genes at individual alleles. This enables the observation of transcriptional dynamics more precisely and provides a powerful tool for studying gene regulation.

(2) Novel Data Analysis Approaches: Develops two new analysis methods to dissect the sources of gene activity (co)variation. One approach separates allele-extrinsic, allele-intrinsic, and gene-autonomous components, and the other quantifies the contributions of burst co-occurrence and burst size correlations. These methods help to more accurately analyze transcriptional correlations between genes and reveal potential regulatory mechanisms.

Weaknesses:

Biological Insights: The findings challenge the traditional view of contact insulation sites as strict regulators of gene coregulation and suggest two distinct coregulatory mechanisms influenced by local chromosome folding. However, expression activity of multiple genes is differentially correlated at the population-level or cell-level versus single-allele-level. More in-depth analysis is needed for further biological insights.