Neville Sanjana
@nevillesanjana.bsky.social
Scientist at the New York Genome Center & NYU.
http://sanjanalab.org
http://sanjanalab.org
Getting back to the MYC story, we next wondered whether the CREs we identified bind specific TFs. We looked for those TFs whose expression was correlated with MYC expression and where binding sites could be identified in the CREs.
March 3, 2025 at 2:57 AM
Getting back to the MYC story, we next wondered whether the CREs we identified bind specific TFs. We looked for those TFs whose expression was correlated with MYC expression and where binding sites could be identified in the CREs.
There is likely a combination of effects on growth driven by both a genomic enhancer and ncRNA transcript since expressing the same ncRNA in trans partially (but not completely) rescues the phenotype!
March 3, 2025 at 2:57 AM
There is likely a combination of effects on growth driven by both a genomic enhancer and ncRNA transcript since expressing the same ncRNA in trans partially (but not completely) rescues the phenotype!
Although CCAT1-MYC looping was the largest change upon CCAT1 transcript loss, this wasn’t just a local change: Many DNA contacts were impacted outside of the MYC TAD — on the same chromosome and genome-wide!
March 3, 2025 at 2:57 AM
Although CCAT1-MYC looping was the largest change upon CCAT1 transcript loss, this wasn’t just a local change: Many DNA contacts were impacted outside of the MYC TAD — on the same chromosome and genome-wide!
How could a lncRNA impact gene expression?
We found that loss of the CCAT1 transcript led to DNA conformation changes; the physical contact between the DNA was impacted by the RNA.
We found that loss of the CCAT1 transcript led to DNA conformation changes; the physical contact between the DNA was impacted by the RNA.
March 3, 2025 at 2:57 AM
How could a lncRNA impact gene expression?
We found that loss of the CCAT1 transcript led to DNA conformation changes; the physical contact between the DNA was impacted by the RNA.
We found that loss of the CCAT1 transcript led to DNA conformation changes; the physical contact between the DNA was impacted by the RNA.
One of the CREs identified in 3 cell lines is located at the promoter for the lncRNA CCAT1 (we can’t escape lncRNAs these days...!)
We found that cell growth decreased when the DNA was silenced AND when the CCAT1 transcript was targeted using a RNA-targeting CRISPR (Cas13).
We found that cell growth decreased when the DNA was silenced AND when the CCAT1 transcript was targeted using a RNA-targeting CRISPR (Cas13).
March 3, 2025 at 2:43 AM
One of the CREs identified in 3 cell lines is located at the promoter for the lncRNA CCAT1 (we can’t escape lncRNAs these days...!)
We found that cell growth decreased when the DNA was silenced AND when the CCAT1 transcript was targeted using a RNA-targeting CRISPR (Cas13).
We found that cell growth decreased when the DNA was silenced AND when the CCAT1 transcript was targeted using a RNA-targeting CRISPR (Cas13).
How are these CREs regulating MYC?
Using H3K27ac HiChIP we found that all but one of the CREs is in physical contact with the MYC promoter. ➰
However, contact 👉👈 is not enough for regulation: less than 10% of all contacts to the MYC promoter were identified as CREs.
Using H3K27ac HiChIP we found that all but one of the CREs is in physical contact with the MYC promoter. ➰
However, contact 👉👈 is not enough for regulation: less than 10% of all contacts to the MYC promoter were identified as CREs.
March 3, 2025 at 2:43 AM
How are these CREs regulating MYC?
Using H3K27ac HiChIP we found that all but one of the CREs is in physical contact with the MYC promoter. ➰
However, contact 👉👈 is not enough for regulation: less than 10% of all contacts to the MYC promoter were identified as CREs.
Using H3K27ac HiChIP we found that all but one of the CREs is in physical contact with the MYC promoter. ➰
However, contact 👉👈 is not enough for regulation: less than 10% of all contacts to the MYC promoter were identified as CREs.
We perturbed these CREs in 1 or 2 cell lines were the CRE is present (from the CRISPRi screens) and 1 where it wasn’t. Indeed, we see decreased cell growth and decreased MYC expression in the cell lines with the CRE and no change in cells lines without the CRE.
March 3, 2025 at 2:43 AM
We perturbed these CREs in 1 or 2 cell lines were the CRE is present (from the CRISPRi screens) and 1 where it wasn’t. Indeed, we see decreased cell growth and decreased MYC expression in the cell lines with the CRE and no change in cells lines without the CRE.
Even for CREs that were previously identified in human or mouse studies (mostly using large deletions), we were able to increase the resolution of the CRE boundaries. 🎛️🗺️ Up to 10,000-fold in one case!
March 3, 2025 at 2:43 AM
Even for CREs that were previously identified in human or mouse studies (mostly using large deletions), we were able to increase the resolution of the CRE boundaries. 🎛️🗺️ Up to 10,000-fold in one case!
We identified 32 cis-regulatory elements (CREs) across the cancer types — many of which were new ones. Most of these CREs are unique to a cancer/cell type, but some are shared.
March 3, 2025 at 2:43 AM
We identified 32 cis-regulatory elements (CREs) across the cancer types — many of which were new ones. Most of these CREs are unique to a cancer/cell type, but some are shared.
It's amazing data to see 🌟: Precise "spikes" 📍at sites that overlap cis-regulatory elements (CREs) — enhancers, repressors and promoters — all based on functional phenotype (growth). 📈
March 3, 2025 at 2:40 AM
It's amazing data to see 🌟: Precise "spikes" 📍at sites that overlap cis-regulatory elements (CREs) — enhancers, repressors and promoters — all based on functional phenotype (growth). 📈
We designed ~110,000 CRISPR guide RNAs to tile the MYC TAD in 6 cancer cell lines (5 different types of cancer). At each of these thousands of locations, we use CRISPR inhibition to shut down (repress) any enhancers that might be present.
March 3, 2025 at 2:40 AM
We designed ~110,000 CRISPR guide RNAs to tile the MYC TAD in 6 cancer cell lines (5 different types of cancer). At each of these thousands of locations, we use CRISPR inhibition to shut down (repress) any enhancers that might be present.
Which noncoding region to pick? We first looked at the regulatory landscape around key oncogenes as given by hallmarks of enhancer activity. We found that the oncogene MYC had the highest regulatory diversity.
March 3, 2025 at 2:40 AM
Which noncoding region to pick? We first looked at the regulatory landscape around key oncogenes as given by hallmarks of enhancer activity. We found that the oncogene MYC had the highest regulatory diversity.
A major motivation for this work is that there are many studies and resources exploring protein-coding genes across different human cell lines....
... but, relatively few that target the SAME noncoding region ACROSS different cell lines.
... but, relatively few that target the SAME noncoding region ACROSS different cell lines.
March 3, 2025 at 2:40 AM
A major motivation for this work is that there are many studies and resources exploring protein-coding genes across different human cell lines....
... but, relatively few that target the SAME noncoding region ACROSS different cell lines.
... but, relatively few that target the SAME noncoding region ACROSS different cell lines.
Similar thoughts from Francis Collins — who co-led the sequencing of the first human genome (a civilization-scale achievement!) — and resigned today from the NIH:
www.nytimes.com/2025/03/01/u...
www.nytimes.com/2025/03/01/u...
March 1, 2025 at 7:23 PM
Similar thoughts from Francis Collins — who co-led the sequencing of the first human genome (a civilization-scale achievement!) — and resigned today from the NIH:
www.nytimes.com/2025/03/01/u...
www.nytimes.com/2025/03/01/u...
Indirect costs fund vital resources for biomedical scientists:
March 1, 2025 at 7:23 PM
Indirect costs fund vital resources for biomedical scientists:
Wise words from Tom Maniatis — who helped launch the biotech revolution, which brought SO MANY health & tech advances in the ~50 years since the advent of programmable gene manipulation.
www.cell.com/cell/fulltex...
www.cell.com/cell/fulltex...
March 1, 2025 at 7:23 PM
Wise words from Tom Maniatis — who helped launch the biotech revolution, which brought SO MANY health & tech advances in the ~50 years since the advent of programmable gene manipulation.
www.cell.com/cell/fulltex...
www.cell.com/cell/fulltex...
YES! We worked w/ @nygenome.org lab neighbor @gamzeandgursoy.bsky.social (& #ConorWalker in her lab) to use cancer genomic datasets to answer this. Some PD-L1 trans-regulators are mutated in pancreatic cancers, alter PD-L1 in vivo, and lead to differences in survival after immunotherapy!
December 26, 2024 at 3:18 AM
YES! We worked w/ @nygenome.org lab neighbor @gamzeandgursoy.bsky.social (& #ConorWalker in her lab) to use cancer genomic datasets to answer this. Some PD-L1 trans-regulators are mutated in pancreatic cancers, alter PD-L1 in vivo, and lead to differences in survival after immunotherapy!
And lest you think these CREs have only slight effects: At this same CRE with the validated TF binding, we find that ablation of EITHER the CRE OR the TFs enhances T cell immunotherapy using primary human T cells engineered with mesothelin CARs. 🗡️☠️
December 26, 2024 at 3:18 AM
And lest you think these CREs have only slight effects: At this same CRE with the validated TF binding, we find that ablation of EITHER the CRE OR the TFs enhances T cell immunotherapy using primary human T cells engineered with mesothelin CARs. 🗡️☠️
But these TF-CRE pairings are just predictions. We need to test our predictions! Xinhe (w/ stellar PhD student #XiaoWang) decided to do just that using CUT&RUN to measure TF binding. 🔎 They found that 2 predicted TFs (SRF & BPTF) indeed bind to one of the PD-L1 CREs.
December 26, 2024 at 3:18 AM
But these TF-CRE pairings are just predictions. We need to test our predictions! Xinhe (w/ stellar PhD student #XiaoWang) decided to do just that using CUT&RUN to measure TF binding. 🔎 They found that 2 predicted TFs (SRF & BPTF) indeed bind to one of the PD-L1 CREs.
One idea💡: Look for overlap between TF motifs found in each CRE and highly significant TFs from the screen. Below, the x-axis indicates the TF motif enrichment and the y-axis indicates TF significance. In total, we link 13 validated PD-L1 CREs to at least one trans-regulator!
December 26, 2024 at 3:18 AM
One idea💡: Look for overlap between TF motifs found in each CRE and highly significant TFs from the screen. Below, the x-axis indicates the TF motif enrichment and the y-axis indicates TF significance. In total, we link 13 validated PD-L1 CREs to at least one trans-regulator!
In the TFome-wide CRISPR screens, Xinhe identified both known and novel PD-L1 trans-regulators. For example, KMT2D and SP1, previously reported PD-L1 activators, rank highly in the PD-L1 low cell populations. Nice!! 🎯
December 26, 2024 at 3:18 AM
In the TFome-wide CRISPR screens, Xinhe identified both known and novel PD-L1 trans-regulators. For example, KMT2D and SP1, previously reported PD-L1 activators, rank highly in the PD-L1 low cell populations. Nice!! 🎯
Using this TFome-wide library, Xinhe knocked out all TFs and again flow sorted cells by PD-L1 expression — but now perturbing trans regulators instead of cis elements! 💡💡💡
December 26, 2024 at 3:09 AM
Using this TFome-wide library, Xinhe knocked out all TFs and again flow sorted cells by PD-L1 expression — but now perturbing trans regulators instead of cis elements! 💡💡💡
To test this, Xinhe & lab 3D genome guru 👩🔬 #ChristinaCaragine performed enhancer HiChIP and found that many of the PD-L1 CREs that we validated individually contact the promoter of PD-L1 or the promoter of a nearby PD-L1 regulator (JAK2).
December 26, 2024 at 3:09 AM
To test this, Xinhe & lab 3D genome guru 👩🔬 #ChristinaCaragine performed enhancer HiChIP and found that many of the PD-L1 CREs that we validated individually contact the promoter of PD-L1 or the promoter of a nearby PD-L1 regulator (JAK2).
But how do CREs modulate gene expression? These CREs, even distant ones, may contact the promoter of gene via 3D DNA looping. Any bound TF at the CRE would then be close to the PD-L1 promoter.
December 26, 2024 at 3:09 AM
But how do CREs modulate gene expression? These CREs, even distant ones, may contact the promoter of gene via 3D DNA looping. Any bound TF at the CRE would then be close to the PD-L1 promoter.
You'll notice in the plot (and diagram) above that there are both basal and interferon (IFNγ) treated cells. That's because PD-L1 expression often increases after IFNγ (to put the brakes on T cells/immune responses). And that's true of the pancreatic cancer cells we used too!
December 26, 2024 at 3:09 AM
You'll notice in the plot (and diagram) above that there are both basal and interferon (IFNγ) treated cells. That's because PD-L1 expression often increases after IFNγ (to put the brakes on T cells/immune responses). And that's true of the pancreatic cancer cells we used too!