Lamins

Counting Lamin A Across Nearly 500 Tissue Regions Showed It Falls With Prostate Cancer's Progression

Undergraduate researcher — immunohistochemistry + image quantificationMcCord Genome Lab, with Dr. Rebeca San Martinwith Priyojit Das, Dr. Rachel Patton McCordMay – Dec 2021Published — AACR abstract, Cancer Res 2023Immunohistochemistry, tissue microarray scoring, Excel

497 ROIs scored for Lamin A/C384 ROIs scored for Lamin B24 diagnostic categories comparedPublished — AACR, Cancer Res 2023 (Abstract A061)

Context

What the system looked like before.

Prostate cancer is one of the leading causes of cancer death in men. On a cellular basis, the protein Lamin A is known for being linked to prostate cancer and promoting its growth. Most men who are diagnosed have overexpressed Lamin A/C in their tissue, contributing to the growth of prostate cancer. Lamin A and Lamin C are structural, filamentous proteins that support the nuclear envelope, aiding chromatin organization, DNA replication and transcription. They're derived from the LMNA gene through alternative splicing.

Almost 1 in 8 men are diagnosed with prostate cancer every year. Understanding how Lamin A/C expression tracks disease progression could point toward alternative, earlier treatment targets.

Question

Studies have shown high-risk cancers or tumor growths have increased Lamin A/C expression in cell lines — does that hold at the tissue level, and can it be quantified directly from patient histology?

Approach & rationale

I worked with Dr. Rebeca San Martin on her research investigating lamin protein expression in prostate cancer. In my project, I analyzed histological sections of patient cancer tumors and found that as cancer worsens, the expression of Lamin A goes down while Lamin B1 remains relatively constant.

The fuller design, from the published methods: immunostaining, ChIP-seq, and Hi-C together characterized Lamin A, B1, and B2 expression across a panel of cell lines modeling progression in vitro — RWPE (normal epithelium), LNCaP (adenocarcinoma), VCaP and MDAPCa2a/b (bone metastatic) — then checked whether the same pattern held in patient tissue microarrays via immunohistochemistry, which is the piece I scored.

Methodology & iteration

Including the paths that did not hold.

Approach A — score LMNA-stained regions directly

TriedWith a brief understanding of immunohistochemistry and a drive full of epithelial tissue stains, I began quantifying nuclear morphology, navigating gaps between gland lumens the way you'd map the solid parts and avoid the holes in a slice of Swiss cheese.

AssumedThat counting Lamin A-positive nuclei region by region would give a reliable percentage per diagnostic category on its own.

Fell shortSome ROIs were overcounted — nuclei were called positive that weren't. It didn't show up until the same tissue was scored a second way.

Approach B — cross-check against Lamin B2

TriedThe same tissue regions were independently scored for Lamin B2, using the same ROI-based counting approach, this time avoiding stromal tissue.

Held upComparing LMNA counts to LMNB2 counts on the same slides surfaced exactly which regions the LMNA pass had over- or under-counted, since the two markers are stained on the same cells and shouldn't diverge for reasons unrelated to biology.

Final approach

TriedCounts were compared side by side across both markers, discrepancies resolved region by region, and the corrected LMNA counts carried into the final analysis — 497 ROIs — alongside the LMNB2 counts, 384 ROIs, across prostatitis, hyperplasia, adenocarcinoma II, and adenocarcinoma III tissue.

Numbered ROI overlay on Lamin A/C-stained tissue — Numbered ROI overlay on LMNA-stained tissue — the counting pass.

Numbered ROI overlay on Lamin B2-stained tissue — The same tissue, scored for Lamin B2 — the cross-check pass.

In short

We identified that Lamin A content is significantly lower in bone metastatic cell lines.

Results

Numbers first, not buried in prose.

We identified that Lamin A content is significantly lower in bone metastatic cell lines. While VCaP cells lack Lamin A deposition, MDAPCa2a cells have a heterogeneous distribution: cells that grow in contact with the cell culture surface express Lamin A while cells that grow atop 3D clusters do not. In contrast, Lamin B1 and B2 content is consistent across cell lines. Analysis of tissue microarrays showed congruent results to the in vitro findings: patient samples derived from higher-grade adenocarcinoma show a decreased number of Lamin A-positive cells.

The hand-counted data backs this directly — percent Lamin A-positive nuclei drops across prostatitis → hyperplasia → adenocarcinoma II → adenocarcinoma III, with every pairwise comparison significant at p < 0.0001.

My part of this AACR abstract was the tissue microarray scoring above — the fuller abstract also covers ChIP-seq, Hi-C, and radial chromosome positioning run by my collaborators.

Violin plot of percent Lamin A-positive nuclei by diagnosis — Percent Lamin A-positive nuclei by diagnosis, every pairwise comparison significant at p < 0.0001.

Significance

Where this went.

My counts were shown across a modeled continuum from normal epithelium to metastatic prostate cancer, leading to my name being a third author in an AACR abstract.

San Martin R, Das P, Hossain S, McCord RP. "Nuclear lamin content as a biomarker for prostate cancer progression." Cancer Res 2023;83(11 Suppl):Abstract A061.

Published AACR abstract page — The published abstract — AACR, Cancer Res 2023, Abstract A061.

Reflection

This was the first project I contributed to in the lab, with a postdoc, Dr. Rebeca San Martin. Her work focused on 3D genome organization — specifically with lamins and LADs — on epithelial-to-metastatic progression, using prostate cancer cell lines across normal, adenocarcinoma, and metastasized stages. My job was to count nuclei in squamous epithelial cells — before I knew it, my entire summer was spent staring and hand-counting nuclei. The images were beautiful; the staining made it look like kaleidoscopic art. How something so beautiful could be so deadly was fascinating and somehow mesmerizing.

It felt strange to be considered an author when all I had done was count and could barely grasp the entirety of the project at the time. But working with Dr. San Martin reminded me of Dr. McCord — the way they both exhibited such strong expertise that they could translate it to me taught me to look at intelligence in a completely different light.

The hierarchical nature of DNA folding inside a cell's nucleus — understanding lamins as the dense, fibrous meshwork the B compartment attaches to — became even more fascinating once I connected it to compartment switching itself. It bridged my time under Dr. San Martin to what I was later doing with Heng Li: CTCF and cohesin as chromatin structural proteins, and lamins as the structural proteins that connect chromatin to the inner nuclear membrane.

BackAll work NextCTCF →