Most cancer biology starts with a dish. The line you choose fixes the genetic background, the differentiation state, and the immune context in which your question is asked, so the catalogue below is organised the way experimentalists actually think about it: by organ system, then by subtype. For each group I note what makes the line useful and where it can mislead. The emphasis is on background and caveats rather than protocols, and none of this substitutes for authenticating a line and confirming its identity before you rely on it.
Lung
Lung cancer lines are unusually well characterised because the driver landscape is clear: KRAS, EGFR, ALK, and TP53 alterations partition the disease, and the non-small-cell (NSCLC) versus small-cell (SCLC) split changes almost everything downstream.
- A549 — the workhorse NSCLC line, derived from a lung adenocarcinoma and carrying wild-type TP53 with a KRAS G12S mutation. It is adherent, fast-growing, and easy to transfect, which is exactly why it is oversubscribed: it is a poor model of primary tumour heterogeneity and is notoriously aneuploid. It is also a hypo-triploid line whose karyotype is a common source of copy-number artefacts.
- NCI-H1299 — a p53-null NSCLC line, valuable precisely because it lacks functional p53; useful for reconstitution and stress-response studies, but its unusually large NRAS-amplified background makes it a special case.
- NCI-H1975 — a EGFR L858R/T790M double-mutant adenocarcinoma line, the canonical model for third-generation EGFR inhibitor work and acquired TKI resistance.
- NCI-H358 / NCI-H460 / NCI-H1650 — KRAS-mutant (H358, H460) and EGFR-mutant (H1650) adenocarcinoma backgrounds, all with differing TP53 status; useful as a panel rather than individually.
- NCI-H520 / NCI-H226 / NCI-H1703 — squamous-cell carcinoma lines, historically the harder NSCLC subtype to model, so these lines are correspondingly valuable for squamous-specific questions.
- NCI-H69 / NCI-H446 — small-cell lung cancer lines that grow as suspension aggregates; they model the neuroendocrine SCLC phenotype but are slow and awkward to work with.
- LLC (Lewis lung carcinoma) — not human: a mouse line used for syngeneic in vivo lung tumour and metastasis models, selected because an intact immune system is in the experiment.
Digestive system
The digestive tract spans liver, stomach, colorectum, and pancreas, each with its own canonical lines. The recurring caveat is differentiation: many of these lines are dedifferentiated relative to the tissue they claim to represent.
- HepG2 — a hepatoblastoma-derived line widely used as a "liver" model; it is not a true hepatocellular carcinoma, retains substantial hepatocyte-like metabolic capacity, and is the standard for drug metabolism and toxicity assays.
- Huh-7 — a differentiated hepatocellular carcinoma line; permissive for hepatitis C virus replication and a staple of antiviral and lipid-metabolism work.
- Hep3B — an HBV-integrated hepatocellular carcinoma line that secretes hepatitis B surface antigen; useful for viral and liver-cancer studies.
- HCCLM3 — a highly metastatic hepatocellular carcinoma line selected for in vivo invasiveness; valuable for metastasis biology but far from a typical clinical tumour.
- BEL-7402 — a long-established Chinese hepatocellular carcinoma line; still used, though its provenance and drift warrant authentication.
- MGC-803 / SGC-7901 / AGS — gastric lines: MGC-803 and SGC-7901 are poorly differentiated and widely used in the Chinese literature; AGS is a KRAS-mutant, microsatellite-stable line and the best-characterised of the three.
- HCT-116 / HT-29 / SW480 / SW620 / Caco-2 — colorectal lines that together cover the microsatellite-instability and chromosomal-instability routes: HCT-116 is MSI-high with a KRAS G13D and intact wild-type BRAF; HT-29 is BRAF V600E and MSI; SW480 and its lymph-node-derived sibling SW620 are a matched primary/metastasis pair; Caco-2 differentiates spontaneously into a polarised enterocyte-like monolayer and is the standard intestinal permeability model.
- PANC-1 / MIAPaCa-2 / BxPC-3 — pancreatic ductal adenocarcinoma lines: all KRAS-mutant and TP53-altered, with a dense desmoplastic phenotype that fails to reproduce in a dish; MIAPaCa-2 is unusually poorly differentiated and fast-growing.
Breast and gynaecologic
Breast cancer lines map cleanly onto the clinical receptor subtypes, which is why they remain the field's reference set. The gynaecologic lines below cover cervical and ovarian disease, where viral aetiology and platinum response dominate the biology.
- MCF-7 / T47D — oestrogen-receptor-positive, luminal lines; MCF-7 is the classic ER-positive model with an intact p53 pathway and oestrogen-dependent growth, while T47D carries a PIK3CA mutation. Both are excellent for endocrine-response work but grow slowly and are prone to losing receptor expression with poor handling.
- MDA-MB-231 / MDA-MB-468 — triple-negative lines (ER, PR, and HER2 negative); MDA-MB-231 is highly invasive and mesenchymal and is the workhorse metastasis model, while MDA-MB-468 has EGFR amplification. Neither captures the full heterogeneity of clinical triple-negative disease.
- SK-BR-3 / BT-474 — HER2-amplified lines, the standard models for trastuzumab and HER2-pathway research; SK-BR-3 is PIK3CA-mutant and more proliferative, BT-474 more epithelial and hormone-responsive.
- HeLa / SiHa / CaSki — cervical lines: HeLa is the immortal HPV-18-positive line that underpins modern cell culture; SiHa and CaSki are HPV-16-positive, with CaSki carrying integrated HPV genomes at high copy number.
- A2780 / SK-OV-3 — ovarian lines: A2780 is a platinum-sensitive, TP53-mutant line and the standard for chemoresistance work (its cisplatin-resistant derivative A2780cis is equally common); SK-OV-3 is a clear-cell-like, HER2-expressing line with a PIK3CA mutation.
Urologic
Prostate lines separate by androgen-receptor status, which determines whether they are models of hormone-dependent or castration-resistant disease. Bladder and renal lines add urothelial and clear-cell contexts.
- LNCaP / 22Rv1 — androgen-receptor-positive prostate lines; LNCaP is the canonical androgen-dependent model with a TP53 mutation and a functional AR, while 22Rv1 is castration-resistant and expresses a constitutively active AR splice variant. Both are useful for AR-axis and anti-androgen work.
- PC-3 / DU145 — androgen-receptor-negative, castration-resistant lines; PC-3 is PTEN-null and highly invasive, DU145 is PTEN-wild-type with a TP53 mutation. They model AR-independent resistance but cannot answer AR-directed questions.
- 5637 / T24 — bladder urothelial lines: T24 is the classic HRAS-mutant, TP53-altered line and the standard bladder-cancer model; 5637 is also HRAS-mutant and more differentiated.
- 786-O / ACHN / Caki-1 — renal lines: 786-O is a VHL-deleted clear-cell line, the reference model for hypoxia-signalling and VHL biology; ACHN is a papillary-like renal cell line; Caki-1 is a clear-cell line derived from a skin metastasis and used for renal metastasis work.
Neuro, skin, and bone
Here the lines are grouped by lineage rather than organ, because glioma, neural crest, melanoma, osteosarcoma, and squamous oral lines share developmental and signalling themes even though the tissues differ.
- U87MG / U251 — glioblastoma lines that are the field's standard in vitro glioma models; U87MG is highly tumorigenic in mice and easy to engineer, but its identity has drifted and it is often misreported, and neither line reproduces the invasive and immunosuppressive glioma microenvironment convincingly.
- SH-SY5Y / SK-N-SH / Neuro-2a — neuroblastoma-derived neuronal models: SH-SY5Y and its parental SK-N-SH differentiate on retinoic acid or neurotrophin treatment and are used for neuronal differentiation and neurotoxicity; Neuro-2a is a mouse neuroblastoma line with a faster growth rate.
- A375 / B16-F10 — melanoma: A375 is a human BRAF V600E line and the reference for targeted BRAF/MEK inhibitor studies; B16-F10 is a mouse line selected for high lung-colonisation, the classic syngeneic metastasis model.
- U-2 OS / MG-63 / Saos-2 — bone: U-2 OS is an osteosarcoma line retaining osteoblastic features and a usable model for bone differentiation; MG-63 is a poorly differentiated osteosarcoma line with a TP53 rearrangement; Saos-2 lacks functional p53 and is widely used as a p53-null background.
- CAL-27 — an oral squamous-cell carcinoma line used for head-and-neck cancer biology, with a TP53 mutation and moderate invasiveness.
Endocrine
Endocrine lines are fewer but well chosen: thyroid carcinoma lines model the follicular and papillary subtypes, and islet lines model insulin secretion and beta-cell biology.
- CAL-62 / B-CPAP / TPC-1 — thyroid carcinoma lines: B-CPAP and TPC-1 carry the BRAF V600E mutation and model papillary thyroid carcinoma; CAL-62 is KRAS-mutant and used for anaplastic or poorly differentiated thyroid disease. All three grow slowly relative to more common lines.
- INS-1 / MIN6 — rodent insulinoma-derived beta-cell lines that retain glucose-stimulated insulin secretion, making them the standard in vitro models for beta-cell function and diabetes research; MIN6 secretion degrades with passage number.
Using these lines well
The catalogue is a starting point, not a guarantee. Three habits keep tumour-line work trustworthy. First, authenticate and re-authenticate: STR profiling of human lines is the minimum standard, and the published mismatch rates for misidentified lines justify routine checks. Second, treat a line as a model of a subtype, not of a disease — most of the lines above are aneuploid, harbour decades of unrecorded drift, and represent a single patient's tumour, so a conclusion drawn in one line should be confirmed in at least a second. Third, match the model to the question: syngeneic and immunocompetent models (such as B16-F10 and the mouse colon lines) are indispensable when the immune system is part of the hypothesis, whereas human lines are best used for cell-intrinsic molecular work. Choose deliberately, report the exact source and passage, and the rest of the catalogue will serve you well.
References
- Giard DJ, Aaronson SA, Todaro GJ, et al. In vitro cultivation of human tumors: establishment of cell lines derived from a series of solid tumors. J Natl Cancer Inst. 1973;51(5):1417–1423. doi:10.1093/jnci/51.5.1417
- Soule HD, Vazguez J, Long A, Albert S, Brennan M. A human cell line from a pleural effusion derived from a breast carcinoma. J Natl Cancer Inst. 1973;51(5):1409–1416. doi:10.1093/jnci/51.5.1409
- Fogh J, Fogh JM, Orfeo T. One hundred and twenty-seven cultured human tumor cell lines producing tumors in nude mice. J Natl Cancer Inst. 1977;59(1):221–226. doi:10.1093/jnci/59.1.221
- Fidler IJ. Selection of successive tumour lines for metastasis. Nat New Biol. 1973;242(118):148–149. doi:10.1038/newbio242148a0
- Spengler BA, Biedler JL, Ross RA. A corrected karyotype for the SH-SY5Y human neuroblastoma cell line. Cancer Genet Cytogenet. 2002;138(2):177–178. doi:10.1016/s0165-4608(02)00523-x
- Bairoch A. The Cellosaurus, a cell-line knowledge resource. J Biomol Tech. 2018;29(2):25–38. doi:10.7171/jbt.18-2902-002
- Masters JR, Thomson JA, Daly-Burns B, et al. Short tandem repeat profiling provides an international reference standard for human cell lines. Proc Natl Acad Sci U S A. 2001;98(14):8012–8017. doi:10.1073/pnas.121616198