My latest publication, Physiology governing diatom vs. dinoflagellate bloom and decline in coastal Santa Monica Bay, was published in Frontiers in Microbiology November 2023. It’s exploratory work that enabled me to learn more about biology and an opportunity to teach myself bioinformatics and statistics related to metatranscriptomics. The take-home message is in the title: I detailed the metabolic and environmental processes that may facilitate on the Southern California coast these two distinct algal blooms.
I think that coastal algal blooms are like any annually occurring team sport. Football is the cleanest analogy. In football, the exact same 32 teams participate each year on a regular cycle. We watch them battle for dominance during football season until a winner emerges following the big game, The Super Bowl. During the off-season, the teams aren’t dead or gone. They instead, quietly exist in a more dormant state.
In a way, an algal bloom off the California coast during the spring is like the result of the Super Bowl. That big green (or brown or yellow or red or bioluminescent blue) film you see atop the water marks the victory of one (or three) dominant species. And this victorious species is a photosynthetic microbe that has been selected from a pool of photosynthetic competitors that probably persist (dormant/ at low abundances) throughout the year. And like the victorious football team, the dominant blooming alga is not generally “better” than the losers; it’s entirely possible that the champs this year may suck next year and that the losers might win under subtle differences in the environment. But this year, the dominant species has capitalized on chance occurrences that have allowed it to outcompete others following the commencement of the on-season in Southern California coastal ecosystems: Springtime, because there is ample sunlight from longer days and there is nutrition from spring wind-driven upwelling of nutrient-rich water.
Strong wind gusts moving south along the California coast during spring months results in the offshore transport of nutrient depleted surface water, and the upwelling of nutrient rich bottom water into the sun-lit layers where phytoplankton live in ample sunlight, poised to respond to the nutrient pulse.
The results of the Super Bowl clearly announces the winning team, but not the plays or the choices that resulted in the winner’s dominance. By comparison, I can easily tell you the dominant species using a microscope and/or genetics. But to examine the metabolic responses to change, the game winning metabolic strategies and plays, I look to shifts in metabolic and physiological priorities during the blooms by examining the whole community’s gene expression: Metatranscriptomics, which is the collection and sequencing of mRNA from an entire population.
In my most recent paper, published November 2023, I compared the metabolic plays (i.e. how they responded metabolically to changing environmental cues) of numerically dominant algal species from two contrasting guilds. Diatoms vs dinoflagellates are analogous to any rivalry you can think of. How I did this is very simple. I examined the pool expressed genes during transitions across four key stages of their respective bloom—diatoms bloomed during 2018; dinoflagellates in 2019.
Blooms have four key stages: pre-bloom, algal bloom, peak bloom magnitude, bloom decline. For most of the year, nutrient concentrations in the surface ocean are relatively low. So Algal species in the water during pre-bloom are not blooming and at low concentrations because they don’t have sufficient nutrients. For example, Nitrogen is a common limiting factor in the ocean. But during spring when nitrogen-rich water is upwelled, algae begin compete for dominance. The dominant competitor species begins to rapidly proliferate into an algal bloom as they rapidly consume the upwelled nutrients. At some point nutrients run out and this, nutrient deficiency puts a cap on the magnitude of the bloom. The maximum magnitude therefore represents the beginning of stress due to depleted nutrients. Once their cell abundances begin to decline, (or chlorophyll a concentrations) the writing is on the wall. The reign of this dynasty has reached its end.
So why am I measuring expressed genes? Well, the ultimate goal of everything living is to not die (and later, to continue to exist through reproducing). And expressing genes is the precursor to manufacturing proteins that a cell needs to survive the current situation. Not only that, gene expression is VERY (metabolically) expensive, consuming much ATP that could have been otherwise used for other important, life-saving processes. Because an expressed gene costs so much, its presence in a cell at a given moment likely represents the metabolic priorities at a given moment. For example, I saw that diatoms preferentially expressed nitrogen transport genes during peak abundance after nutrients were depleted. I hypothesize that this represents a switch from passive diffusion under high nutrient conditions to actively sucking, gasping for nutrients, once Nitrogen was depleted.
The full publication is published open access HERE in Frontiers in Microbiology, but I’ll hilight a few of my favorite take homes below.
Usain Bolt is a fast runner but he is equally fast off the starting block. Diatoms are like this. The respond quickly to nutrients and also grow quickly because of their surface area and genetically encoded strategies for nutrient utilization.
Diatoms are first off the starting block and Fast runners:
Diatoms are superior competitors under high nutrient environments, and they bloomed immediately following nutrient pulse from upwelling. Like the starting gun of a race, clues from the gene expression (what they were doing metabolically) showed me that diatoms metabolically responded quickly to this nutrient cue. There was an uptick in expressed genes associated with nitrogen assimilation, a decisive move that likely facilitated the other cellular processes: cell division, photosynthesis, respiration all showed increase at this moment. Diatoms are poised to utilize inorganic nitrogen, and their fast response and employment epitomized a universal diatom: “The early bird catches the worm”, the first one off the block stands a better chance of winning the race.
Like protein and carbohydrates, excess fat can be broken down and utilized when food consumption is not possible. Chitin is a nitrogenous compound found in some diatoms that may act as a nutrition storage molecule analogous to body fat.
Slimming down: Chitin as nutrient storage fat:
In addition to expressing genes to actively pull in inorganic nitrogen after it was depleted, diatoms also began to strip proteins for parts. My favorite example of this involved the expression of chitinase—which is an enzyme designed break up nitrogen-rich molecules called chitin—during peak bloom magnitude after nitrogen depletion. Diatoms can’t swim, and if they sink they die (yes, there are exceptions). Some control buoyancy while also storing nitrogen by making long spicules made of chitin, although the nitrogen storage aspect may be a convenient side-effect. The dominant diatom species, thalassiosira just so happens to make chitin.
Weird food.
Beggars can’t be choosers: The wide nitritional pallet of dinoflagellates.
On the flip side, dinoflagellates may not out compete diatoms for prime sources of (new) Nitrogen—inorganic Nitrogen species such as NO3-, NO2-, NH4- —especially during high-nutrient circumstances like spring upwelling. But, dinoflagellates have grit. They survive and dominate in environments where inorganic nutrients are extremely low because they have evolved a few unorthodox strategies. For example, whereas diatoms don’t swim, dinos do; whereas diatoms are generally large, dinos are small (think surface area and Michaelis-Menten kinetics); whereas diatoms prefer inorganic nutrients and are purely photosynthetic, most dinos are less picky eaters that will scavenge for dissolved organic nutrients, steal chloroplasts from others, and/or consume large and small prey: all in the name of survival.
During spring 2019, dinoflagellates formed a massive bloom. For comparison, the diatom blooms that occurred in both 2018 and 2019 peaked at a chlorophyll a concentration of 18 µg/L. The dinoflagellate bloom during 2019 peaked at 64 µg/L! But that’s not all. It bloomed to such a high magnitude after inorganic nitrogen had been depleted and after diatoms had appeared to have tapped out as a result (I measured extremely low concentrations of nitrate+nitrite and transcripts from diatoms were extremely rare). How?
I’m still not entirely sure how dinoflagellates bloomed to such a high magnitude during such low inorganic nitrogen concentrations. And their stone cold responses to change didn’t make this a open and shut case. But their gene expression points to their utilization of organic sources of nutrition including the consumption of prey. I think a speculative, but sexy idea is that zoosporic parasites facilitated the blooming of dinoflagellates that could benefit from parasite presence (directly as food) and/or from their impact (indirectly consuming the guts of lysed, large diatom prey). Akashiwo and Margalefidinium—the blooming dinoflagellate taxa during 2019—are both able to consume prey and use organics.
Shaolin vs Wu-Tang: Diatoms and dinoflagellates have different physiological fighting styles.
I’ve been practicing martial arts since I was a child. One element of martial arts is the many flavors, the diversity of styles. Even within a group, you can find diversity. For example, two Chinese Kung Fu styles, Shaolin vs Wu Tang. Japanese Karate, Korean Karate, American Boxing, Afro-American 52 Blocks, Afro-Brazilian Capoeira, African NGolo, French Savat, Israeli Krav Maga, Russian Sanbo, African etc. etc. If you know anything about some of these, you can feel the similarity and dissimilarity among them. For example the two Kung Fu styles are distinct from one another, but are more similar to each other than either are to any form of Japanese Karate. A way to quantify this feeling is with NMDS or PCA. For example, see the toy example of an NMDS below.
In this study I classified the function of genes—what they do—expressed by the three dominant taxa: two diatom species and one dinoflagellate family. I calculated the gene expression levels—think of the many instruments in an orchestra, and the conductor instructing certain instruments to increase or decrease their volume according to the situation—at five time points during two different situations from different years. This gave me thirty samples with different functional compositions to compare.
Multi-variate statistical analysis is a great way to cluster similar objects with multiple attributes. For example, if you wanted to group major cities or humans (potential lovers, teams based on the composition of athletes, types of spenders at a supermarket) in some meaningful way based on multiple distinct variables.
I used two flavors of (multi-variate) statistical analysis (ordination techniques) to project the thirty samples into clusters of samples with highly similar compositions of genes. The result was a divide between Diatoms and Dinoflagellates, regardless of the year or bloom stage.
Non-metric, multi-dimensional scaling plot illustrating the groupings of Diatom (Grey=Pseudo-nitzschia, LakerGold=Thalassiosira) vs Dinoflagellate (LakerPurple) metabolic activity through both blooms (Diatom bloom in 2018=circles; Dino bloom in 2019=diamonds). Diatoms form a grouping separate from dinoflagellates which illustrates metabolic distinction, distinction in utilization of different strategies from their metabolic play book.
Diatoms are most physiologically similar during times of peace:
The same technique clustered the physiological profiles of distinct diatom species during diatom dominance in 2018, but these groupings were farther away for samples taken during 2019, a time when they had to struggle for survival. Compared to Thalassiosira, who’s expression profiles were not much different under the different situations, Pseudo-nitzschia’s expression profiles collected during the two different situations were greater distances apart. These patterns shows that distinct diatom species may be very similar during times of plenty, but divergent “when shit hits the fan”, nutritionally. Additionally, it shows that some species have a wider range of adaptive responses.
A complete blueprint of a species and what it is capable of is called a genome. A completed genome for Thalassiosira was published in 2004 by Virginia Armbrust and her team which informs us on exactly what this diatom can do. And despite the ability of the other diatom species, Pseudo-nitzschia, to produce neurotoxins during its bloom, it’s genome has yet to be sequenced. Although, several transcriptomic studies have been conducted, revealing bit by bit of this diatom’s capabilities. But the day we have the genome is the day we may elucidate the question of its comparative adaptive range.
Poker face: Unresponsive or alternative splicing?
In this study I showed that dinoflagellates exhibited comparatively muted differential gene expression across the three distinct bloom stages: proliferation, stagnation, and decline. This is not new. Many studies have also shown that Dinoflagellates appear to have a stone cold demeanor and lack of response amid crisis like a good poker face.
Differential expression (mean log2 fold change relative to Day 1) of 36 functional categories grouped into four broad physiological processes of numerically dominant diatoms (A) vs. dinoflagellates (B) during periods (x-axis) of their bloom in 2018 vs. 2019, respectively. Nitrate + nitrite (Nitrogen), phosphate, and chlorophyll concentrations are shown in the top three rows. The four physiological processes are color-coded in the left-hand vertical column.
However, in light of post-transcriptional gene regulation via alternative splicing—which is a hallmark of eukaryotes—I am not certain that dinoflagellates are truly transcriptionally unresponsive, keeping most gene expression and metabolic priorities the same despite the circumstances. For example, splicing in a “poison exon” would result in the destruction, not translation, of an mRNA transcript. This post-transcriptional mechanism of down regulating gene expression—so-called nonsense mediated decay—would be undetectable with standard (automated) alignment-based pipelines for functional classification.
Alternatively spliced genes can produce multiple end products with distinct fates in the cell. In this example, segment B is retained or disgarded and results in differences in the subsequent protein product. If segment B contained a stop codon, failing to splice it out would result in the destruction of this transcript. This mode of gene regulation by alternatively spliced “Poison exons” is called nonsense-mediated decay (Lareau et. al. 2007).
The dearth of dinoflagellate genomes makes it difficult to investigate the prevalence of alternative splicing in this guild, but sequencing technology is moving at break-neck speeds. Scott Roy used to say that “…the data is never perfect.” And just as Musashi pointed out the similarities between a carpenter, merchant, and soldier, my role of a biologist and bioinformatician is to keep his skills sharp, even and especially before the data is perfect. To be continued…
