A Specific Primed Immune Response in Drosophila Is Dependent on Phagocytes

A Drosophila melanogaster fly has no antibodies, no T cells, no B cells, and no immunological memory — at least, none that immunology textbooks will acknowledge. And yet, Pham and colleagues showed that a fly exposed to a near-deadly dose of Streptococcus pneumoniae, then challenged weeks later with a dose that kills every naïve fly in the cohort, survives. It does not survive because it was vaccinated in any classical sense or because it has adaptive machinery. It survives because something about that first encounter changed it. If there's no memory system, how does it remember? The accepted wisdom in immunology has always been definitive: adaptive immunity — specific, long-lasting, and capable of distinguishing one pathogen from another — requires T and B cells and therefore belongs exclusively to vertebrates. Invertebrates get innate immunity, which is powerful but stereotyped. Innate immunity recognizes broad molecular patterns, not specific strains, and produces the same effector outputs every time it is activated. Drosophila's innate toolkit is compact and well-mapped: a humoral arm that secretes antimicrobial peptides into the fly's blood-like hemolymph, a melanization response that deposits toxic melanin at wound and parasite surfaces, and a cellular arm of phagocytic hemocytes — the fly's equivalent of macrophages — that engulf and kill invaders. Two signaling pathways, Toll and imd, coordinate these responses depending on the pathogen detected. That is the entire system. The logic of the field stated that this system could not produce memory. Pham and colleagues designed a clean two-step test of that logic. Wild-type Oregon R flies received a sublethal priming injection of either 250 live colony-forming units of S. pneumoniae or a heat-killed preparation. One week later, those same flies were challenged with a lethal dose — a dose that reliably kills naïve flies within two days. The result was dramatic and reproducible across cohorts of 158 to 228 flies per condition, repeated at least three times. Naïve flies died quickly. Primed flies survived at rates that were statistically extraordinary, with p-values below 0.0001 in multiple assays. Survival mapped directly to bacterial clearance: when primed flies were challenged with 400 colony-forming units — the lowest dose lethal to naïve animals — they had killed almost all of the S. pneumoniae within a single day. Naïve flies challenged the same way still harbored bacteria at that point. Then came the persistence data. Flies primed on day zero were challenged on days 1, 3, 7, 10, and 14. At every single time point, primed flies died significantly more slowly than naïve controls. Day 1 showed a p-value below 0.0001. Day 7 showed a p-value below 0.0001. Day 14 showed a p-value of 0.0003. The authors noted they could not go further — by week three, flies are five weeks old and begin dying from wounding stress alone — but within the limits of the assay, protection was detectable within 24 hours and persisted for the life of the fly. These are not genetically engineered animals. These are wild-type flies, carrying only the innate machinery described above, showing a persistent, organism-level change from a single prior exposure. The next question was whether this protection was general or specific. The answer is: neither, exactly. It is coarse. Priming with S. pneumoniae protected against S. pneumoniae, but not against Salmonella typhimurium, Listeria monocytogenes, or Mycobacterium marinum. However, the effect was not limited to a single bacterial strain — Beauveria bassiana, a natural fungal pathogen of flies, could also prime a similar species-specific protective response against subsequent B. bassiana challenge. Critically, S. pneumoniae-primed flies showed no improved clearance of Escherichia coli when exposed to 6,000 colony-forming units of that bacterium. The protection is targeted, not a global increase in immune response. The specificity sharpened into a more pointed result when Pham and colleagues tested whether strong, broad activation of both signaling pathways could substitute for S. pneumoniae priming. They injected flies with a mixture of heat-killed E. coli, Micrococcus luteus, and B. bassiana — a cocktail designed to maximally activate both Toll and imd — and then challenged them with lethal S. pneumoniae. No protection was observed. When dead S. pneumoniae was added to that same mixture, protection was restored. The implication is stark: general immune pathway activation is not sufficient. Something about the S. pneumoniae cue itself is required. The mechanism is not about increasing immunity broadly; it is about something specific to the encounter. So the team went looking for the mechanism, adopting an approach to eliminate possible suspects. The first suspect was antimicrobial peptides, the canonical output of Toll and imd signaling. S. pneumoniae is a relatively weak inducer of antimicrobial peptide transcription to begin with, and transcript levels of defensin, attacin, and diptericin were not elevated one week after priming — by the time of challenge, there was no circulating peptide signal to explain the protection. The broad-pathway activation experiment already showed that Toll and imd activation is insufficient. Thus, circulating antimicrobial peptides are off the table. The second suspect was the imd pathway itself. Flies carrying loss-of-function alleles of imd and dTak1 — key components of that signaling branch — were primed and challenged at a reduced dose of 100 colony-forming units, because these mutants are more sensitive. Even in those mutants, primed flies survived at significantly higher rates than naïve mutant flies, with p-values below 0.0001. The imd pathway is not necessary. The Toll pathway is required — Pham and colleagues are clear about that — but it isn't sufficient. Toll activation without the specific S. pneumoniae priming signal does not produce the effect. Thus, the signaling is necessary for establishing something, but the effector doing the actual work must lie downstream. That left phagocytes. To test them, the team used a bead-blockade method: injecting flies with tiny 0.2 micrometre polystyrene beads to saturate phagocytic capacity and confirm inhibition with fluorescently labeled bacteria. The effect on baseline sensitivity was striking — normally, it takes three thousand colony-forming units to kill a fly, but bead-injected flies were killed by just twenty. Phagocytes are clearly doing significant work even during an ordinary infection. Then came the critical test: bead-block primed flies challenged with lethal S. pneumoniae, to see if the priming benefit persists. It did not. Primed flies with blocked phagocytosis died at the same rate as naïve flies with blocked phagocytosis. The priming advantage was lost. Both bead-treated groups died significantly faster than mock-treated primed flies, with p-values below 0.0001. The protection, whatever its molecular basis, resides in the phagocytes. Remove their function, and the memory disappears with it. The architecture of the finding, then, is this: a sublethal encounter with S. pneumoniae leaves a durable trace in the fly's phagocytic cells. When the same pathogen appears again, those cells clear it faster and more completely. The Toll pathway is necessary to establish this state, but it operates through the cellular arm, not through circulating peptides. The enhancement is specific — the same phagocytes do not clear E. coli more effectively after S. pneumoniae priming. And this state lasts the entire life of the fly. What Pham and colleagues are pointing out is something the field may have overlooked for a long time. The vertebrate-centric understanding of adaptive immunity — built around T cells, B cells, and antibodies — may have caused researchers to miss memory-like phenomena in animals that lack those tools. If a fly can generate a specific, persistent, phagocyte-dependent primed response using only innate machinery, the question isn't whether this is a shadow of "real" adaptive immunity. The question is what is happening inside the hemocyte the second time around. What molecular change in that cell makes it faster, more effective, and more targeted against a pathogen it encountered weeks earlier? Drosophila, with its genetic tractability and its freedom from the overlapping complexity of T and B cell responses, is now poised to answer that. The notion that insect immune responses cannot adapt wasn't just incomplete; it may have been diverting researchers away from a phenomenon that turns out to be far more widespread than anyone suspected. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.