PREreview del Chromosome-arm-specific telomere length governs dual modes of structural genome evolution in IDH-mutant astrocytoma

This review reflects comments and contributions by Joseph Biggane, Nitya Khetarpal, Yuka Gliday, Sarah Gilmour, Diptarup Mallick, Chinyere Mary-Cynthia Ikele, Timothy En Haw Chan, Haofu Huang, Briana Pinales, and Debraj Manna. Review synthesized by Joseph Biggane.

This study investigated the association between the alternative lengthening of telomeres pathway and structural genome instability in IDH-mutant astrocytomas. The authors find that shorter telomeres are associated with breakage-related events, whereas longer ALT-maintained telomeres are linked to ecDNA formation and amplification.

Crowd Review Comments:

• Pertaining to the Title:

  • Several reviewers recommended softening the language in the title. In particular, the reviewers found that while the results are interesting, and potentially important, the research demonstrates association rather than causation.

• Pertaining to the Introduction:

  • In response to “...raises the possibility that chromosome-specific telomere length directly influences the structural fate of individual chromosome arms.”, a reviewer suggested the addition of another sentence and additional references. While it is clear that the telomere length is heterogeneous in these cell types, it is not made clear how or why this would influence "structural fate". 

• Pertaining to the Results:

  • In subsection ‘Astrocytoma genomes are defined by segmental aneuploidy and variable telomere length dynamics’...

    • In response to “Telomere content in astrocytoma was higher on average and more dispersed than in glioblastoma or oligodendroglioma…”, The authors’ claim that telomere length is more dispersed in astrocytoma seems to be based in Figure S1B. While the points may be more dispersed in astrocytoma, there also appears to be more data points than in other tumor types, and no N is given in this figure. Information about sample number should be added and statistical support for increased dispersion should be provided for this claim.

  • In subsection ‘Long-read sequencing identifies abundant complex structural variants in a cohort of astrocytomas’...

    • In response to Figure 2C, the reviewers noted that panel C is very difficult to interpret given that almost all SVs identified are BNDs. Further, the breakdown of those SVs presented in Figure S2C should be incorporated into this figure.

    • In response to Figure 2D, the reviewers questioned how the authors established the manual edits, and whether their classifications have been compared to other datasets, or measure them orthogonally, to assess consistency.

    • In response to Figure 2D, the reviewers noted that the vertical axis label is obscured by the image.

    • In response to Figure 2H, the reviewers noted that size data was not presented in this panel, so size data was inappropriate for this figure legend. They further noted that size data was presented in Figure S5.

  • In subsection ‘Extrachromosomal DNA is pervasive in IDH-mutant astrocytomas and associates with tumor grade’...

    • In response to Figure 3D (Left), the reviewers noted that the authors reported that telomere length amplification association was only significant for the selected subset of multi-region intrachromosomal ecDNA, but not for all canonical ecDNA. The authors should consider explaining the rationale for focusing on this specific ecDNA subclass because this justification is unclear in the presented data.

    • In response to ‘Recurrently amplified loci harbored oncogenes such as EZH2, KDM5A, ZEB1, and SOX9 (Figure S6C, Supplementary Table 6) and overlapped a repeat-rich landscape dominated by retroelements (Figure 3C)’, The reviewers note that while the authors demonstrate that HGs do contain repeats and retroelements, the human genome is very rich in retroelements. It is not clear from the data presented whether or not these elements represent a significant enrichment of repeats over the baseline. It would be beneficial to include size-matched control regions from random segments of the human genome.

  • In subsection ‘Chromosome-arm-specific telomere length analysis reveals heterogeneity consistent with ALT and dual modes of genomic instability’...

    • In response to ‘Arm-level TL median values spanned a broad range (0.1–13 kb)...’, The reviewers suggested softening the tone in this statement. Specifically, the authors discuss that this estimate is based on sequencing data with an N50 length of 12kb. Thus, it seems unlikely that the reads fully span the telomeres. Perhaps the authors could compare this data with similar N50/coverage from another tumor type with lower ALT activity (as in Figure 1) to assess whether the distribution is different with less ALT activity.

  • In subsection ‘Structural variant breakpoints are enriched at telomeric and centromeric regions’...

    • In response to Figure 5A, the reviewers suggest that the authors might consider an additional metaplot with all chromosomes centered around the pericentromere region. Because all the data is centered around telomeres it is unclear if there is any pericentromeric SV enrichment.

    • In response to ‘This structural fragility likely reflects the repetitive sequence composition, late replication timing, and recombination-prone structure of subtelomeric and pericentromeric DNA...’, The reviewers note that in classical recombination studies, the opposite is true- subtelomeres and pericentromeres are nearly universally considered coldspots for recombination. The following alternative interpretation is offered: the elevated breakpoints here likely represent relaxed selection against SVs in this area rather than a recombinogenic structure. The lack of references for these claims was further noted.

    • In response to ‘heterochromatin-driven fragility’, the reviewers note that no evidence is presented which links heterochromatin and the SV pattern.

    • The reviewers also suggest softening the tone/claims in the final paragraph of this results subsection for similar reasons to the comments made about the title.

• Pertaining to the Discussion:

  • In response to ‘...recombination-prone architecture of subtelomeric and pericentromeric DNA...’, the reviewers again note that the literature does not necessarily support the notion that subtelomeres and pericentromeres are recombination prone.

  • The reviewers applaud the authors’ inclusion of the limitations of this study in their discussion, especially the lack of matched germline sequencing.

• Pertaining to the Methods:

  • In subsection ‘Genome alignment’... 

    • In response to ‘For four samples sequenced across two flowcells, reads from both runs were merged prior to alignment’, the reviewers sought clarity on whether the reads were merged per sample, in case of batch effects across flow cells.

  • In subsection ‘Arm-level aneuploidy analysis’...

    • The reviewers took note of the in-house analysis pipeline and the fact that the claims using this analysis cannot be validated by other researchers.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.