When CD55 goes where it shouldn't: the nucleus
This page collects a presentation I put together on a 2024 study of CD55 in high-grade serous ovarian cancer. It's the mirror image of the CHAPLE syndrome page: instead of a patient with no working CD55, this is a cancer cell with too much of it — some of which ends up somewhere CD55 was never expected to go.
High-grade serous ovarian cancer, and why it comes back
HGSOC is the most common form of ovarian cancer, and its standard treatment has a predictable failure mode: a small population of chemoresistant cells is usually what's left behind.
High-grade serous ovarian cancer (HGSOC) accounts for roughly 75% of ovarian cancer cases and originates in the epithelium of the ovaries or fallopian tubes. The standard of care is platinum-based chemotherapy (such as cisplatin), and it usually works well at first — most cells in a tumor are platinum-sensitive and die off. But every HGSOC tumor also contains a small population of ovarian cancer stem cells (CSCs) that are inherently resistant to chemotherapy. Kill off the sensitive majority, and what's left is enriched for exactly the cells most likely to drive recurrence.
Surface CD55: a shield against complement
Before this study, CD55's role in ovarian cancer was understood the same way it's understood everywhere else on this site — as a surface protein blocking complement.
Ovarian cancer cells overexpress CD55 at the cell surface. There, it works exactly as described on the CD55 explorer and the CHAPLE syndrome page: it binds C3b and C4b and accelerates decay of the complement convertases, so the membrane attack complex never forms. For a tumor, that's a shield — it blunts one of the immune system's main ways of killing abnormal cells, and was the leading explanation for why CD55-high tumors are harder to clear.
CD55 blocks complement at the surface
C3b and C4b are decay-accelerated before the cascade can escalate — no membrane attack complex, no complement-mediated killing of the cancer cell. This alone was long thought to be the whole story.
It doesn't explain chemoresistance
Complement evasion explains why tumors survive immune attack — it doesn't explain why a subset of cells specifically resist platinum chemotherapy and behave like stem cells. That gap is what this study addresses.
A pool of CD55 doesn't stay on the surface
Using immunofluorescence and subcellular fractionation, the authors found CD55 in the nucleus of ovarian cancer cells — not just the membrane.
Nuclear CD55 turns out to be glycosylated and derived from the same pool of protein expressed at the cell surface — it traffics inward rather than being made separately. The trafficking is driven by a specific signal: CD55's serine/threonine (S/T) domain, the same flexible stalk that (as covered on the CD55 explorer) normally just links the complement-binding CCP domains to the membrane. Deleting that S/T domain is enough to block nuclear entry — and, critically, to resensitize otherwise chemoresistant cells to cisplatin. Nuclear CD55 was also found enriched in ascites from chemoresistant patients and in a subset of patient tumor specimens, tying the lab finding back to real disease behavior.
In the nucleus, CD55 silences a tumor suppressor
Once inside, CD55 doesn't just sit on chromatin — it interferes with the protein that normally keeps stem-cell programs switched off.
Nuclear CD55 binds and suppresses ZMYND8, an epigenetic regulator and tumor suppressor that normally represses stemness and chemoresistance programs. With ZMYND8 attenuated, the Polycomb Repressive Complex 2 (PRC2) — EZH2, SUZ12, and EED — becomes more active, and H3K27 trimethylation (H3K27me3) spreads across chromatin. That repressive histone mark silences tumor-suppressive genes, and the net effect is self-renewal and cisplatin resistance.
Normal chromatin state
- ZMYND8 represses stemness and resistance programs
- PRC2 (EZH2, SUZ12, EED) stays in check
- No abnormal spread of H3K27 trimethylation
With nuclear CD55
- CD55 binds and attenuates ZMYND8
- PRC2 activity increases
- H3K27me3 spreads, silencing tumor-suppressive genes
- Result: self-renewal and cisplatin resistance
The target might be trafficking, not just CD55 itself
Because deleting CD55's S/T domain blocks nuclear entry without necessarily removing CD55 from the surface, it separates two jobs that were previously bundled into one protein: complement evasion at the membrane, and epigenetic reprogramming in the nucleus. That distinction matters therapeutically — a strategy that blocks nuclear trafficking specifically could resensitize resistant tumors to platinum chemotherapy without giving up whatever surface function CD55 still has.
How the finding was established
The authors used immunofluorescence and subcellular fractionation to localize CD55 at both the cell surface and the nucleus of ovarian cancer cells, and compared the stability (half-life) of each pool. To find the trafficking signal, they generated a series of CD55 deletion mutants and tested nuclear localization, cisplatin sensitivity, and stem-cell frequency for each — in vitro and in mouse models. Binding partners of nuclear CD55 were identified by immunoprecipitation followed by mass spectrometry, and downstream transcriptional consequences were mapped with RNA sequencing.
As reported in the study: nuclear CD55 was found in a subset of ovarian cancer specimens and in ascites from chemoresistant patients, and its nuclear localization is driven by a trafficking code containing the CD55 serine/threonine domain — deletion of that domain was sufficient to sensitize chemoresistant ovarian cancer cells to cisplatin.[1]
Two diseases, one gene, opposite problems
CHAPLE syndrome is what happens when CD55 is completely absent. This study is close to the opposite: CD55 isn't just overexpressed, some of it is mislocalized — trafficking into a compartment where it wasn't meant to act at all, using a domain that normally just anchors it in the membrane. Both are reminders that CD55's biology isn't fully captured by "on" or "off" at the gene level — location and protein-level regulation matter too. That's a useful caution for Misha's own PhD work: silencing CD55 with ZFPoff or CRISPRoff removes the gene's output entirely, but it can't speak to what happens to protein that's already been made and is on the move.
Sources
-
[1]
Bharti R, Dey G, Khan D, et al., "Cell surface CD55 traffics to the nucleus leading to cisplatin resistance and stemness by inducing PRC2 and H3K27 trimethylation on chromatin in ovarian cancer" — Molecular Cancer, 2024