The Transcriptome and Proteome of the Diatom Thalassiosira pseudonana Reveal a Diverse Phosphorus Stress Response

Picture a single diatom — a microscopic glass-caged phytoplankton, smaller than the width of a human hair — drifting through a patch of ocean where phosphorus has nearly run out. It can't swim away. It can't wait for a delivery. What it can do is run one of the most sophisticated nutrient-scavenging programs in the living world. Dyhrman and colleagues went inside that program, and what they found reshapes how we think about ocean chemistry. Phosphorus is a quiet regulator of ocean life. It limits marine primary production across significant stretches of ocean — the subtropical North Atlantic is a well-documented example. Because diatoms account for roughly forty percent of all ocean primary production, what happens to diatoms under phosphorus stress has consequences for the entire marine carbon cycle. Diatoms don't just fix carbon; they also export phosphorus to the seafloor, where it ends up locked in apatite-containing sediments. So the link runs both ways: phosphorus controls diatom growth, and diatom physiology controls where phosphorus ends up. Despite this centrality, the molecular details of how diatoms respond to phosphorus scarcity were largely unknown before this study. Iron stress and nitrogen stress had been characterized in diatoms, but phosphorus stress had not. Dyhrman and colleagues set out to close that gap using the model diatom Thalassiosira pseudonana. The experimental setup was clean and paired. Cultures of Thalassiosira pseudonana were grown under two phosphorus regimes: phosphorus-replete at thirty-six micromolar phosphate and phosphorus-deficient at just zero-point-four micromolar. Then the team applied two molecular lenses simultaneously. The first was tag-sequencing, a form of deep transcriptomics where short twenty-one base-pair sequence tags represent individual transcripts. The second was quantitative shotgun proteomics — liquid chromatography coupled to mass spectrometry, or LC-MS/MS — which counts peptide fragments to estimate protein abundance. Using both matters, because transcripts don't always predict protein levels. A gene can be loudly transcribed while its protein is quietly degraded, or vice versa. The scale of the findings was substantial: three hundred and eighteen transcripts were differentially regulated, and one hundred and thirty-six proteins were differentially abundant. The two layers were positively but imperfectly correlated — the log-fold change in protein abundance scaled at roughly half the log-fold change in transcript abundance, with an r-squared of approximately half. Coordination, yes. Lockstep, no. The most conceptually clean response to phosphorus scarcity is to go get more of it. Thalassiosira pseudonana does this aggressively. A phosphate transporter — protein ID 24435 — was upregulated in both the transcriptome and the proteome, and its transcript had the highest absolute tag counts in the entire phosphorus-deficient dataset. Earlier physiological work, cited by the authors, showed that phosphorus-deficient Thalassiosira pseudonana dramatically increases its maximal uptake rate without changing its affinity for phosphate — exactly what you'd expect if the cell is adding more transporter copies to its surface rather than tinkering with individual transporter efficiency. More transporters mean more throughput. But there's a ceiling on how much inorganic phosphate a cell can pull from water that barely has any. So the diatom also reaches for a different phosphorus pool — the dissolved organic phosphorus, or DOP, that typically exceeds inorganic phosphate in the upper ocean. The key enzymes here are alkaline phosphatases, metalloenzymes that cleave phosphate groups off organic molecules, releasing inorganic phosphate that the cell can absorb. Dyhrman and colleagues found at least four putative alkaline phosphatase genes upregulated at both the transcript and protein levels under phosphorus deficiency. They confirmed the biochemical activity directly and visualized it on whole cells using enzyme-labeled fluorescence — phosphorus-deficient cells lit up; phosphorus-replete cells did not. Two of the phosphatases appear to carry signal peptides consistent with secretion, suggesting they sit on the cell surface, positioned to hydrolyze DOP right before uptake. The cell is, in effect, processing its food at the door. Alongside those phosphatases, a putative glycerophosphoryl diester phosphodiesterase — protein ID 23858 — showed one of the most dramatic inductions in the entire dataset. In the phosphorus-replete transcriptome, it registered just five tags per million. In the phosphorus-deficient transcriptome, it registered five thousand five hundred and sixty tags per million. Its protein was roughly twenty-fold more abundant under deficiency. Diesterases hydrolyze a different class of organic phosphorus compounds than alkaline phosphatases do, so their co-induction suggests the cell is casting a wide net across the DOP pool, not just targeting one molecule type. While all that scavenging ramps up externally, Thalassiosira pseudonana is simultaneously reducing its internal phosphorus demand — and the mechanism is striking. Cell membranes are normally built from phospholipids, molecules that require phosphorus in their backbone. Under phosphorus stress, the diatom replaces those phospholipids with two non-phosphorus alternatives: sulfolipids, specifically sulphoquinovosyldiacylglycerol, and betaine lipids. Both perform the same structural role in a membrane without requiring a single phosphate group. The sulfolipid biosynthesis gene UDP-sulfoquinovose synthase, protein ID 269393, was upregulated in both the transcriptome and the proteome, providing direct molecular evidence for the substitution. This is a proactive conservation move: the cell is not just hunting for more phosphorus; it is rebuilding itself to need less. The phosphorus freed from membranes can be redirected elsewhere. Interestingly, the Thalassiosira pseudonana genome lacks clear homologs of the bacterial betaine lipid synthase genes, so the authors flag four SAM methyltransferase-domain transcripts as candidates for further study. There's also a storage dimension. The team identified vacuolar transporter chaperone genes — Vtc4, protein ID 43150, and a second Vtc homolog — that were upregulated in the phosphorus-deficient transcriptome, and the VTC4 protein itself was more abundant under deficiency. These proteins are linked to polyphosphate synthesis, and solid-state phosphorus-31 nuclear magnetic resonance directly confirmed increased cellular polyphosphate in the deficient cultures. Polyphosphate is a linear chain of phosphate molecules that cells can stockpile. The upregulation of synthesis genes during deficiency, rather than during phosphorus abundance, suggests the cell is actively reallocating its remaining phosphorus into storage form — a counter-intuitive move that may buffer the cell against further depletion. Beyond these targeted responses, phosphorus deficiency reorganizes the cell's core metabolism. Several glycolytic enzymes shift in a coordinated fashion across both transcript and protein datasets, consistent with a rerouting of hexose-phosphate metabolism to bypass phosphate-requiring steps and recycle phosphate internally. Translation machinery is dialed back: ribosomal protein genes are downregulated in both datasets. That makes sense when you remember that RNA synthesis can be a cell's largest single phosphorus expenditure — in the cyanobacterium Prochlorococcus, it accounts for roughly half of total phosphorus uptake. Reducing ribosomes means reducing the phosphorus locked in ribosomal RNA. Supporting the idea that translation is actively regulated rather than passively slowing, a PUF-family RNA-binding protein — protein ID 31875 — was among the most highly induced transcripts, and PUF proteins are known to repress translation by binding messenger RNA. Crucially, this transcript was not differentially expressed under nitrogen, iron, or silicon limitation — suggesting its induction is specific to phosphorus stress. What does this all add up to? Thalassiosira pseudonana runs at least five coordinated strategies simultaneously: storing phosphorus as polyphosphate, ramping up transport, hydrolyzing dissolved organic phosphorus via alkaline phosphatases and a diesterase, rebuilding membranes without phospholipids, and dialing back phosphorus-intensive biosynthesis. Each strategy is visible at the molecular level, and many show coordination between transcript and protein — not perfect mirroring, but consistent directionality. This is not a single-switch stress response. It is a system-level reorganization. For biogeochemical modeling, that distinction matters. Models that treat phosphorus as a single fixed resource, or that infer dissolved organic phosphorus processing only from alkaline phosphatase measurements, will underestimate what a diatom can actually do in a low-phosphorus patch of ocean. Diatoms that can reach into the dissolved organic pool via diesterases, rebuild their membranes on the fly, and compress their phosphorus demand at the level of ribosomes are considerably more resilient than simpler models assume. The open question that Dyhrman and colleagues leave on the table is the right one: how widespread are these strategies across other diatom species, and how will they scale as ocean phosphorus distributions shift? For now, the organism that started this story — one single-celled diatom navigating a phosphorus desert — turns out to be running a biochemical program of remarkable sophistication. 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.