Karenia brevis allelopathy compromises the lipidome, membrane integrity, and photosynthesis of competitors

Abstract

The formation, propagation, and maintenance of harmful algal blooms are of interest due to their negative effects on marine life and human health. Some bloom-forming algae utilize allelopathy, the release of compounds that inhibit competitors, to exclude other species dependent on a common pool of limiting resources. Allelopathy is hypothesized to affect bloom dynamics and is well established in the red tide dinoflagellate Karenia brevis. K. brevis typically suppresses competitor growth rather than being acutely toxic to other algae. When we investigated the effects of allelopathy on two competitors, Asterionellopsis glacialis and Thalassiosira pseudonana, using nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS)-based metabolomics, we found that the lipidomes of both species were significantly altered. However, A. glacialis maintained a more robust metabolism in response to K. brevis allelopathy whereas T. pseudonana exhibited significant alterations in lipid synthesis, cell membrane integrity, and photosynthesis. Membrane-associated lipids were significantly suppressed for T. pseudonana exposed to allelopathy such that membranes of living cells became permeable. K. brevis allelopathy appears to target lipid biosynthesis affecting multiple physiological pathways suggesting that exuded compounds have the ability to significantly alter competitor physiology, giving K. brevis an edge over sensitive species.

Figure 1

oPLS-DA models reveal that lipidomes of Thalassiosira pseudonana and Asterionellopsis glacialis are disrupted by Karenia brevis allelopathy. Filled symbols represent lipidomes of algae exposed to K. brevis through molecule-permeable but cell impermeable membranes, empty symbols represent lipidomes from unexposed algae (controls). oPLS-DA model generated from (A) ¹H NMR spectral data and (B) from UHPLC/MS metabolic features from lipidomes of T. pseudonana (blue squares; variance captured along each latent variable is stated in parentheses). oPLS-DA model generated from (C) ¹H NMR spectral data and (D) from UHPLC/MS metabolic features from lipidomes of A. glacialis (yellow circles).

Figure 2

Volcano plot summarizes the differences in the lipdome of T. pseudonana (blue squares) and A. glacialis (yellow circles) when exposed vs. not exposed to K. brevis allelopathy. The relative abundances of 80 metabolites were significantly different (p < 0.05 after Bonferroni correction, see SI Materials and Methods) in T. pseudonana upon exposure to K. brevis allelopathy. Red lines indicate log₂ fold difference of ±1. Six metabolites with concentrations that were significantly different in A. glacialis when exposed to K. brevis were also significantly different in concentration when T. pseudonana was exposed to K. brevis.

Figure 3

Exposure of T. pseudonana to K. brevis led to cell membrane damage. (A) K. brevis allelopathy significantly decreased T. pseudonana membrane integrity as indicated by permeability of live T. pseudonana cells measured by co-staining of SYTOX Green (stains nucleus of cells with permeable cell membranes) and Neutral Red (stains cytoplasm of live cells only via vacuole uptake; N = 5; asterisk denotes statistically significant difference between treatment and dilute media controls via Bonferroni-corrected t-test; p = 0.0164 for caged K. brevis, p = 0.0113 for uncaged K. brevis; α = 0.0167). (B) Presence of K. brevis, caged or uncaged, significantly decreased growth of T. pseudonana while exposure to caged T. pseudonana had no effect relative to dilute media controls (asterisk denotes statistical difference between control and treatment; p < 0.0001 for caged K. brevis, p = 0.0005 for uncaged K. brevis).