Rapid, high-throughput phenotypic profiling of endosymbiotic dinoflagellates (Symbiodiniaceae) using benchtop flow cytometry

Abstract

Endosymbiotic dinoflagellates (Family Symbiodiniaceae) are the primary producer of energy for many cnidarians, including corals. The intricate coral-dinoflagellate symbiotic relationship is becoming increasingly important under climate change, as its breakdown leads to mass coral bleaching and often mortality. Despite methodological progress, assessing the phenotypic traits of Symbiodiniaceae in-hospite remains a complex task. Bio-optics, biochemistry, or "-omics" techniques are expensive, often inaccessible to investigators, or lack the resolution required to understand single-cell phenotypic states within endosymbiotic dinoflagellate assemblages. To help address this issue, we developed a protocol that collects information on cell autofluorescence, shape, and size to simultaneously generate phenotypic profiles for thousands of Symbiodiniaceae cells, thus revealing phenotypic variance of the Symbiodiniaceae assemblage to the resolution of single cells. As flow cytometry is adopted as a robust and efficient method for cell counting, integration of our protocol into existing workflows allows researchers to acquire a new level of resolution for studies examining the acclimation and adaptation strategies of Symbiodiniaceae assemblages.

Fig 1

A) Sample point cloud visualizing all cytometry-detected particles within a well-replicate, and demonstrating how Symbiodiniaceae cells are identified based on high red fluorescence (red circle). B, C) Excitation/emission spectra employed by the flow cytometry protocol presented in this contribution. Excitation laser wavelengths are indicated by arrows. Lines depict absorption spectra, while lines with shaded areas underneath illustrate the intensity of emission wavelengths based on the corresponding excitation laser. B) The blue laser (488 nm) simultaneously excites the two dominant light harvesting complexes (LHC) of Symbiodiniaceae (PCP and acpPC) and a set of antioxidant associated pigments: FbFPs, diadinoxanthin, and beta-carotene (C). The two groups fluoresce differently upon excitation, which are differentiated with two emission filters that detect red (695/50 nm) (B) and green (525/30 nm) (C) spectra (grey boxes).

Fig 2. Samples prepared under different light (Dark vs Light) and temperature (Ice vs No Ice) conditions.

Samples processed on ice within roughly two hours (136 minutes) of airbrushing yielded the most consistent results (dashed boxes), as informed by the statistical groups (p < 0.05) that are represented by letters A) Red fluorescence degraded quickly. B) Green fluorescence and forward scatter (cell size) (C) remained relatively stable over time. D) Side scatter (cell roughness), increased with time E) Cell concentrations were relatively consistent results when processed within 136 minutes.

Fig 3. Six dilutions (50x, 20x, 10x, 5x, 2x, and 1x) were tested 16 times each across a 96-well plate.

The cytometry run took ~6 hours estimating ~20 minutes between each of the 16 replicates (Time). Ten-fold and five-fold dilutions were the most consistent for half of the plate (dashed boxes) suggesting that runs should be limited to ≤ 48 wells (half of a standard 96-well plate). Over- or under-dilution of samples had a large effect on resulting parameter estimates. A) Red fluorescence degraded over time and had more variation with 50x and 20x dilutions. B) Green fluorescence increased over time, presumably due to heat generated by the flow cytometer. C) Cell size (FSC) did not change over time, but highly concentrated samples led to an overestimation of cell sizes, likely due to cell clumping. D) Cell shape (SSC) was the most stable parameter, suggesting low rates of cell lysis. E) Cell concentrations were heavily affected by dilution and time spent in the flow cytometer.

Fig 4. Phenotypic profiles built from red fluorescence, green fluorescence, forward scatter, and side scatter for upside-down jellyfish processed with our flow cytometry protocol (S1 File) after three weeks of exposure to different light conditions (Dark, Light, and Variable).

A) Principal component analysis demonstrating divergence of phenotypic profiles associated with jellyfish of different experimental light conditions. B-I) Distributions for each phenotypic trait used to generate a phenotypic profile. Lines with asterisk indicate pairwise comparisons tested with Post-Hoc Dunn’s tests ranging producing three categories for p-values: p > 0.05 (ns), p < 0.05 (*), and p < 0.001 (***). B-E) Each distribution represents 16,000–20,000 observations, representing the true phenotypic variance within a Symbiodiniaceae assemblage, which allowed for discriminating between treatments, even with small effect size. F-I) Phenotypic measurements averaged to the jellyfish host to better resolve conditional effect, despite decreased statistical power.

Fig 5. Genetic diversity of Symbiodiniaceae assemblages in-hospite may reveal either low complexity (one dominant clade) or high complexity (more than one dominant clade) communities [3]; autofluorescence profiles may be either convergent or divergent, as shown here by PCAs consisting of the discussed phenotypic metrics.

As indicated by arrows, the combination of genetic diversity and phenotypic profiles has the potential to allow for the classification of Symbiodiniaceae assemblage functional strategies into homogenous, plastic, redundant, or diverse categories upon implementation into other relevant workflows. PCA datasets are theoretical and were generated and analyzed using R.