Bacterial capture and lysis on diatom spines reveal a suspension-feeding strategy
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (Arcadia Science)
Abstract
Phytoplankton contribute nearly half of global primary production and play a central role in ocean biogeochemistry. Although traditionally viewed as phototrophs, a growing body of evidence indicates that many phytoplankton are mixotrophic, supplementing photosynthesis with heterotrophic acquisition of nutrients and organic matter. Yet, the mechanisms enabling heterotrophic feeding in diatoms remain poorly understood. Here, we show that the chitinous spines of the diatom Conticribra weissflogii can serve as attachment sites for the bacterium Marinobacter adhaerens . Cryo-electron tomography and viability staining revealed that a substantial fraction of spine-associated M. adhaerens undergoes lysis. Consistent with bacterial lysis on the spines, Raman spectroscopy showed that diatom cells bearing spine-associated bacteria acquired more bacteria-derived nitrogen than uncolonized cells. Although microfluidic assays revealed a low bacterial affinity for the spines, our results suggest that attachment can be promoted by fucoidan deposits identified on the spines — a compound for which bacteria showed significantly higher binding affinity than chitin — as well as by their geometry, which encounter rate calculations show to be particularly efficient at intercepting bacteria. Together, these findings reveal a previously undescribed form of suspension-feeding in phytoplankton, reminiscent of the role of zooplankton pseudopods in prey capture. Diatom spines intercept bacteria from the surrounding water, retain them near the cell surface, and facilitate subsequent nutrient acquisition, probably through bacterial lysis. Our results expand the functional repertoire of diatom spines beyond defense and buoyancy, identifying spine-mediated bacterial capture as a previously overlooked pathway contributing to phytoplankton mixotrophy.
Significance statement
Diatoms produce long chitinous spines whose functions have been linked to buoyancy control and defense against grazers. Here, we reveal a previously unrecognized function of these appendages: the spines of the diatom Conticribra weissflogii serve as attachment sites for the bacterium Marinobacter adhaerens , a substantial fraction of which undergoes lysis on the spine surface. Using 15 N isotopic labelling and Raman spectroscopy, we demonstrate that diatoms harboring spine-associated bacteria acquire significantly more bacterial-derived nitrogen than uncolonized cells, consistent with spine-mediated bacterial capture contributing to diatom nutrient acquisition. We further show that C. weissflogii spines are coated with fucoidan, which likely promotes bacterial attachment. This suspension-feeding strategy — in which spines coated with fucoidan, intercept bacteria from the surrounding water, analogous to the slime-coated pseudopods of radiolarians and foraminiferans — has never previously been documented in phytoplankton. Our findings expand the known functional repertoire of diatom spines and identify a previously overlooked pathway of mixotrophic nutrient acquisition in phytoplankton with potentially broad implications for marine nutrient cycling.
Article activity feed
-
and this observation was confirmed by scanning electron microscopy (SEM; Fig. 1C, Fig. S1)
The panel labels in Figure 1 do not line up with the images or with the callouts in the text, and it is hard to follow the colonization claim until they do.
As printed, panel A looks like Calcofluor White fluorescence of the spines, which matches the legend, but that panel is never cited. Calcofluor staining is only pointed to later, as Fig. S4. Panels B and C look like bright-field or epifluorescence of bacteria on spines, not SEM. Panels D and E look like the SEM overview and the pili close-up. The quantitative panels then appear to be F-H.
The main text cites a different map. It calls 1B bright-field and 1C SEM, then later cites SEM as 1B. It sends the day-6 / day-14 colonization percentages to 1D, the "one to three bacteria per cell" count …
and this observation was confirmed by scanning electron microscopy (SEM; Fig. 1C, Fig. S1)
The panel labels in Figure 1 do not line up with the images or with the callouts in the text, and it is hard to follow the colonization claim until they do.
As printed, panel A looks like Calcofluor White fluorescence of the spines, which matches the legend, but that panel is never cited. Calcofluor staining is only pointed to later, as Fig. S4. Panels B and C look like bright-field or epifluorescence of bacteria on spines, not SEM. Panels D and E look like the SEM overview and the pili close-up. The quantitative panels then appear to be F-H.
The main text cites a different map. It calls 1B bright-field and 1C SEM, then later cites SEM as 1B. It sends the day-6 / day-14 colonization percentages to 1D, the "one to three bacteria per cell" count to 1E, and the cell-wall attachment image to 1F. In the legend those letters are the SEM panels and the first quantification plot. The same letters are then reused: 1E is also cited for exclusive spine attachment versus the cell body (legend: H), and 1F-G later for the low number of spine-associated bacteria (legend: the counts), after 1F has already been used for the cell-wall image.
A single consistent key in the figure, the legend, and the Results would make the attachment observation much easier to trust. Right now a reader cannot tell which image is meant to support which sentence.
-
. None of these stress-associated features were observed in the 8 tomograms of M. adhaerens cells from a diatom-free culture, suggesting that they were specifically associated with the diatom–bacterium interaction.
The cryo-ET argument in Figure 2 would be easier to follow with a healthy M. adhaerens cell in the same figure. The text says none of the stress features (vesicles, lost outer membrane, enlarged periplasm, loss of integrity) were seen in eight tomograms from a diatom-free culture, but Figure 2 only shows attached cells that already look compromised. A matched tomographic slice of an unattached cell, collected and displayed the same way as 2A–C, would give the reader a baseline for what you are calling damage.
-
Bacteria (indicated by yellow arrowheads) attached to diatom spines.
Is the yellow blob the Dsred signal from the bacteria? It'd be helpful to clarify that (and that its not artificial highlighting)
-
