M1 and M2 macrophages derived from THP-1 cells differentially modulate the response of cancer cells to etoposide

If you open up a solid tumor and look closely, you’ll find a lot more than cancer cells. One of the loudest presences is the macrophage — those shape-shifting immune cells that can act as a warrior or a medic depending on the signals they detect. In shorthand, M1 is the fighter: pro-inflammatory, cytotoxic, good at killing pathogens and, sometimes, tumor cells. M2 is the healer: anti-inflammatory, great at repairing tissue and building blood vessels, but also very effective at helping tumors evade trouble. It’s a simplified view that masks a spectrum, but it’s still a useful lens. In cancer, that lens matters because tumor-associated macrophages repeatedly appear when you ask why some tumors shrug off chemotherapy. Earlier work in breast cancer, for example, demonstrated that cathepsin-rich macrophages can literally rescue cancer cells from drugs like etoposide. So the hunch has been around for a while: the macrophages in the neighborhood might be adjusting chemotherapy sensitivity. Genin and colleagues decided to test that idea head-on, in a way that allows them to separate signal from noise. They took a human monocytic cell line, THP-1, nudged it into macrophages, and then polarized those macrophages into either M1-like or M2-like states. The differentiation step was clear: a day of phorbol myristate acetate at 150 nanomolar, followed by a day to settle and mature. The cells adhered, adopted macrophage morphology, and expressed familiar lineage markers — CD68, CD71, CD36 — while CD14 diminished as expected in this model. This gave them a standardized human macrophage backdrop to work with, similar to setting a baseline tone before the music changes. The first key choice was how to generate a strong M1 signal without harming the cells. They used interferon gamma with lipopolysaccharide, or LPS, the classic one-two for pro-inflammatory activation. However, LPS is a double-edged sword. In their case, high doses reduced viability — roughly two-thirds of the cells remained at twenty-four hours. So they dialed it down to a whisper: ten picograms per milliliter with interferon gamma, which preserved high viability — about ninety-three percent — while still polarizing macrophages into an M1-like state. Interferon gamma alone didn’t hurt, but it didn’t ignite the M1 fire either. On the flip side, pushing THP-1 macrophages toward M2 required a different approach. Interleukin-4 and interleukin-13, each at twenty nanograms per milliliter, set the direction, but achieving the full M2 signature took time. Some markers changed at twenty-four hours; many needed longer. This is a theme in this system: THP-1-derived macrophages are slower to settle into an M2 state than primary cells, so the incubation period extends — out to seventy-two hours — and the pattern sharpens. How do you confirm these are the macrophages they think they are? They checked at both the transcript and protein level. Under the low-dose LPS plus interferon gamma, the cells secreted and expressed the pro-inflammatory markers they were looking for: tumor necrosis factor alpha, or TNF-alpha, interleukin-1 beta, or IL-1 beta, interleukin-6, or IL-6, and the CXCL10 chemokine, with surface receptors like CD80 and HLA-DR increased. Enzyme-linked immunoassays confirmed the cytokine surges — IL-6 and CXCL10 weren’t just transcribed; they were pouring into the medium. The M2 story was different. With interleukin-4 and interleukin-13, CD206 — the mannose receptor — appeared early and climbed higher with time. Interleukin-10, or IL-10, increased. Fibronectin levels rose. Given an extra day or two, slower markers like CD163 and the chemokines CCL18 and CCL22 joined in, with enzyme-linked immunosorbent assay readouts for IL-10 and CCL18 rising in tandem. Importantly, the M2-polarized cells didn’t co-express the M1 repertoire at seventy-two hours. One more practical detail if you’re familiar with murine systems: human macrophages don’t alter nitric oxide production with polarization the way mouse cells do, so that readout won’t assist you here. With those macrophage states established, they posed the question to the cancer cells. Does living near M1 versus M2 macrophages affect how a tumor cell absorbs the impact from a chemotherapeutic? To test this without the interference of direct cell-to-cell contact, they used Transwell inserts with zero point four micrometre pores. Macrophages sit above while cancer cells are below; molecules flow, but cells don’t. They maintained a ratio of one macrophage to one tumor cell, and they stripped serum to avoid growth factors that could obscure results. Then they exposed the tumor cells — HepG2 hepatoma or A549 lung adenocarcinoma — to etoposide at fifty micromolar. The timing varied slightly by line, with HepG2 spending longer on the drug and A549 spending longer pre-incubating with macrophage signals, but the key idea remained: let the soluble conversation start, then introduce the stress. How do you detect an apoptotic collapse in action? You listen to the caspases and watch the membranes. In HepG2, M1 macrophages amplified etoposide’s impact. Cleaved caspase-3 levels were elevated. Cleavage of the PARP-1 protein spiked. The enzymatic readout for caspase-3 and caspase-7 activity rose, not as a blip but as a consistent increase across experiments. Flow cytometry reinforced the same story with differing accents: more Annexin V positivity, more late apoptotic or necrotic cells when M1 was present. M2 macrophages, by contrast, played defense. Those cleaved proteins were less abundant. Caspase activity decreased. Annexin V and propidium iodide labeled fewer cells. Even in the absence of the drug, the M1 co-cultures pushed a small increase in apoptotic signaling in HepG2, while M2 co-cultures suppressed it, which aligns with their broader functions. Non-polarized M0 macrophages mostly observed from the sidelines. Switch to A549, and the main finding remains, but the details become smaller. The M1 condition still enhanced caspase-3 and caspase-7 activity during etoposide treatment, and Annexin V and propidium iodide profiles shifted toward apoptosis. On Western blots, cleaved caspase-3 and PARP-1 were visible in some runs but not all — two out of three blots showed a clear increase — and the enhancement magnitude was more modest than in HepG2. The M2 condition, however, consistently provided protection across readouts, resulting in a flatter apoptotic response to the same chemotherapeutic exposure. There’s an easy alternative explanation to rule out: perhaps etoposide itself is scrambling macrophage identity, pushing them toward or away from M1, and what you’re observing is an artifact of repolarization. Genin and colleagues investigated this. They profiled classical M1 and M2 transcripts and the secreted cytokines under co-culture with the drug. Etoposide didn’t convert M1 into M2 or vice versa. IL-6, an M1-associated cytokine, increased in the presence of the drug, and some M2-associated signals like CCL18 decreased under etoposide, but the overall polarization programs remained consistent. The same held true when comparing monocultures to co-cultures under identical conditions. So the story they tell shows that it's not macrophages changing identity; it's tumor cells changing vulnerability because of the factors those macrophages secrete. Let’s summarize. In a human, reproducible system where only soluble factors can travel, M1-polarized THP-1 macrophages make HepG2 and A549 cells more sensitive to etoposide-induced apoptosis. M2-polarized macrophages make them less sensitive. The assays align — biochemical cleavage of caspase-3 and the PARP-1 protein, functional enzyme activity, and membrane changes detected by Annexin V and propidium iodide — and the qualitative effect is consistent across both liver-derived and lung-derived lines. The magnitude varies with the biology of the line, which is exactly as expected. Why does this matter? Because it reframes that old cartoon of M1 versus M2 and gives it a clear, chemotherapy-relevant perspective. If a tumor microenvironment leans toward M2, it’s not just muted in its immune response — it may actively shield tumor cells from drugs that are meant to kill them. If you can steer those macrophages back toward an M1-like program, you might not need a new drug to get a better response. You might just need a different neighborhood. A note about the model itself, because it’s more than a tool. THP-1-derived macrophages aren't primary cells, and that’s both a limitation and a benefit. You lose some in vivo nuances — the vascular cues, the stromal interactions, the heterogeneous development of tissue macrophages — and you’re employing indirect co-culture rather than allowing direct cell contact. On the other hand, you gain control and reproducibility. You can set the polarization with defined cytokines, confirm it with markers, and then pose a precise question about soluble signals and drug response. Genin and colleagues are transparent about the adjustments needed: low LPS to maintain viability, longer incubation to establish M2 markers, and the fact that readouts like nitric oxide are less helpful in human macrophages than they are in mouse systems. There’s also a conceptual guardrail in their experiments. Because etoposide didn’t repolarize the macrophages, the direction of causality is clearer. The shift in apoptosis arises from how the tumor cells perceive and integrate the factors derived from macrophages, not from the macrophages changing their roles mid-experiment. This doesn’t indicate which molecules are doing the work — is it IL-10’s anti-inflammatory effects, altered growth factor profiles, metabolic by-products, something else? — but it does clarify that the effect exists in the interaction between cells, not within the macrophage's identity switch. So where does this leave us? First, with a concrete takeaway: macrophage polarity can alter chemotherapy response in a controlled human model. In HepG2, the effect is pronounced, with clear increases in cleaved caspase-3 and the PARP-1 protein under M1 conditions and an equally clear reduction under M2. In A549, the effect is subtler, but present, showing up where it counts — in caspase activity and in the fraction of cells entering apoptosis. Second, this provides a plausible explanation for clinical observations that tumors rich in M2-like macrophages tend to be more resistant to drugs. The microenvironment is not just background. It’s a modifier of dosage, a living buffer or amplifier. If you’re considering what comes next, keep it grounded. The next logical step is to identify the soluble players. The same co-culture design can be combined with cytokine neutralization or fractionated conditioned media to determine which signals influence caspase activation. Another avenue is to explore whether this polarity-dependent change applies to other chemotherapeutics with different mechanisms — such as a platinum or a taxane — and to see if the pattern aligns with DNA damage signaling or mitochondrial priming. Then, judiciously, bring it closer to the clinic: primary human macrophages from patient blood, organoid co-cultures that provide stromal context, or ex vivo tumor slices where you can alter macrophage programs and observe drug sensitivity changes in real time. But even without those subsequent steps, the message is actionable. As Genin and colleagues demonstrate, you don’t have to alter the tumor cell’s genome to change how it dies. You can influence its neighbors. And with macrophages, the distinction between a fighter and a healer could mean the difference between a drug that works and a drug that doesn’t.