Algae-Produced Pfs25 Elicits Antibodies That Inhibit Malaria Transmission

Malaria kills hundreds of thousands of people every year, and every single one of those deaths passes through a mosquito. Not through a person — through a mosquito. There is a brief window, inside the insect's gut, where the parasite is completely exposed. If you could put the right antibody into that blood meal, you could shut down transmission before the next bite ever happens. That's the vaccine Gregory and colleagues set out to build. The surprising part is where they grew it. The proteins at the center of this story are called Pfs25 and Pfs28 — surface proteins expressed on the sexual stage of Plasmodium falciparum, the parasite responsible for the deadliest form of malaria. These proteins appear on gametocytes, gametes, and zygotes inside the mosquito midgut. They are not expressed in the human host in any meaningful way. That's the key to the whole strategy: a vaccinated person carries antibodies in their blood that do nothing for them personally. However, when a mosquito takes that blood meal, those antibodies travel into the midgut and attack the parasite at its most exposed moment. You're not vaccinating people to protect themselves. You're vaccinating them to protect everyone around them. Prior work by Duffy and Kaslow showed that Pfs25 and Pfs28 can act synergistically as transmission-blocking antigens, and Saxena and colleagues characterized them as essential mosquito-stage proteins. The challenge was never the concept. The challenge was making the proteins. Both Pfs25 and Pfs28 are structurally unforgiving. Each is built from four tandem epidermal growth factor-like domains, or EGF-like domains, and each domain contains multiple cysteines that must pair into exactly the right disulfide bonds. Get those bonds wrong and the protein misfolds, presenting the wrong surface to the immune system. Antibodies raised against a misfolded protein won't recognize the real thing on the parasite. That requirement for precise folding killed every obvious production approach. Bacterial systems like Escherichia coli couldn't form the correct disulfide network at all. Yeast could fold the disulfide bonds, but yeast adds glycans, or sugar chains, to proteins as a post-translational modification, and those glycans change the protein's shape. In human clinical testing, yeast-produced Pfs25 triggered an allergic reaction. Plant-based transient expression systems showed some promise, but they too required mutating the protein sequence just to prevent glycosylation. Every standard path either failed to fold the protein or required altering it to survive the production system. The protein the field needed was the one no standard system could reliably make. Enter a single-celled green alga called Chlamydomonas reinhardtii, and specifically its chloroplast. The chloroplast is the photosynthetic organelle inside the algal cell, and it turns out to be an oddly perfect match for this problem. Chloroplasts can fold complex, cysteine-rich eukaryotic proteins — they have the machinery to form disulfide bonds. But they do not glycosylate proteins. The glycosylation machinery simply isn't there. So the chloroplast threads the needle: it folds; it doesn't modify. The algae also carry none of the inflammatory contaminants, endotoxins, or known human pathogens that complicate other expression systems, and the U.S. Food and Drug Administration classifies them as generally regarded as safe. To get the malaria proteins into the chloroplast, the team synthesized codon-optimized versions of the Pfs25 and Pfs28 genes. Codon optimization means rewriting the DNA sequence so the chloroplast's own translation machinery reads it efficiently, without changing the amino acid sequence of the resulting protein. These genes were cloned into a chloroplast expression cassette and transformed directly into the C. reinhardtii chloroplast genome. Each protein carried a C-terminal FLAG tag so it could be recovered by affinity chromatography. The proteins expressed. Pfs25 accumulated to half a percent of total soluble protein; Pfs28 to two-tenths of a percent. On sodium dodecyl sulfate polyacrylamide gel electrophoresis, or SDS-PAGE, the dominant bands migrated right where you'd expect: near 21.4 kilodaltons for Pfs25 and 20.2 kilodaltons for Pfs28. Under non-reducing and native conditions, both proteins appeared as larger species, consistent with dimerization, which is characteristic of the native proteins. However, accumulating at the right size is not the same as folding correctly. To test structure, the team used circular dichroism spectroscopy — a method that measures how a protein differentially absorbs left versus right circularly polarized light, giving a fingerprint of the protein's secondary structure. Deconvolution of the spectra showed both algae-produced proteins were composed of roughly 60 percent beta-strands and turns, about 10 percent alpha-helix, and 30 percent unordered peptide. That profile is consistent with the known crystal structure of Pvs25, the closely related homologue from P. vivax. Mass spectrometry covering about 64 percent of the Pfs25 sequence detected disulfide linkages numbered 1, 4, and 6 as intact, while linkages 2, 7, 8, 10, and 11 may not have been completely formed. Some bonds, not all. Partially but meaningfully correct. The most stringent test of folding came from the conformational monoclonal antibodies. Anti-Pfs25 monoclonal 4B7 and anti-Pfs28 monoclonal 2D8 are antibodies that only bind correctly folded epitopes. If the protein is unfolded or reduced, they don't recognize it. Both monoclonals bound their respective algae-produced proteins in non-reduced samples and failed to bind in reduced samples. The conformational epitopes were present. The algae had made something that looked, structurally, like the real thing. Then the team immunized mice. Balb/c mice received purified Pfs25 or Pfs28 with complete Freund's adjuvant as a prime and incomplete Freund's for boosters — a standard immunization protocol. Pooled sera showed high enzyme-linked immunosorbent assay titers against the algal antigens while pre-immune sera did not. More importantly, antisera from Pfs25-immunized mice recognized native P. falciparum sexual-stage Pfs25 in parasite lysates: a 25 kilodalton band under reducing conditions and an approximately 40 kilodalton band under non-reducing conditions that co-migrated with the 4B7 epitope. The antibodies raised against the algae-made protein crossed over and recognized the real parasite protein. Pfs28 antisera also recognized a parasite band at around 22 kilodaltons, but detection required ten times more lysate, and surface staining was weak. That asymmetry between the two proteins becomes the central result of the whole study. The team tested transmission-blocking activity using the standard membrane-feeding assay, or SMFA. In this assay, P. falciparum NF54 gametocytes are mixed with heat-inactivated mouse antisera and fed to female Anopheles stephensi mosquitoes. After nine days, the mosquitoes are dissected, the midguts are stained with mercurochrome, and oocysts are counted. Oocysts are the developmental stage the parasite must reach inside the mosquito to continue the transmission cycle. No oocysts mean no transmission. Antibodies from Pfs25-immunized mice completely blocked transmission. There were no oocysts in any of the dissected mosquitoes that received the Pfs25 immune sera. Antibodies from Pfs28-immunized mice reduced oocyst counts, but the reduction did not reach statistical significance by the Wilcoxon nonparametric test the team applied. The immunofluorescence data tracked with this: Pfs25 antisera produced surface staining on cultured sexual-stage parasites identical to the staining pattern of the conformational monoclonal 4B7, while Pfs28 antisera produced weak staining that was often indistinguishable from background. The antibodies that recognized the surface worked. The ones that didn't, didn't block. Gregory and colleagues position these results as a proof of concept with direct implications for cost and access. Subunit vaccines are the most expensive category of vaccine to produce — the paper cites U.S. prices of ten to one hundred and twenty dollars per dose. For a disease concentrated in low-income countries, that cost structure is a barrier before a single dose is delivered. Algae change the economics. They are cheap to grow, fast to divide, easily scaled, and free of the contamination risks that require expensive downstream processing in other systems. The paper also notes that coupling algal vaccine production with algal biofuel co-products could reduce costs further. It also explicitly raises the possibility of fusing mucosal adjuvants to vaccine candidates for oral delivery — which would eliminate injections and sidestep cold-chain storage requirements entirely. That last point is speculative, a direction rather than a finding. What is a finding is this: algae are the first recombinant system to produce an unmodified, aglycosylated version of Pfs25 that folds correctly, is recognized by conformational antibodies, and elicits an immune response that completely stops malaria transmission in a mosquito feeding assay. The protein the field couldn't make, made in a pond organism, working exactly as hoped. The proof of concept held. 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.