Transcriptional Mechanisms of Thermal Acclimation in Prochlorococcus

Abstract

Low temperature limits the growth and the distribution of the key oceanic primary producer Prochlorococcus, which does not proliferate above a latitude of ca. 40°. Yet, the molecular basis of thermal acclimation in this cyanobacterium remains unexplored. We analyzed the transcriptional response of the Prochlorococcus marinus strain MIT9301 in long-term acclimations and in natural Prochlorococcus populations along a temperature range enabling its growth (17 to 30°C). MIT9301 upregulated mechanisms of the global stress response at the temperature minimum (17°C) but maintained the expression levels of genes involved in essential metabolic pathways (e.g., ATP synthesis and carbon fixation) along the whole thermal niche. Notably, the declining growth of MIT9301 from the optimum to the minimum temperature was coincident with a transcriptional suppression of the photosynthetic apparatus and a dampening of its circadian expression patterns, indicating a loss in their regulatory capacity under cold conditions. Under warm conditions, the cellular transcript inventory of MIT9301 was strongly streamlined, which may also induce regulatory imbalances due to stochasticity in gene expression. The daytime transcriptional suppression of photosynthetic genes at low temperature was also observed in metatranscriptomic reads mapping to MIT9301 across the global ocean, implying that this molecular mechanism may be associated with the restricted distribution of Prochlorococcus to temperate zones. IMPORTANCE Prochlorococcus is a major marine primary producer with a global impact on atmospheric CO₂ fixation. This cyanobacterium is widely distributed across the temperate ocean, but virtually absent at latitudes above 40° for yet unknown reasons. Temperature has been suggested as a major limiting factor, but the exact mechanisms behind Prochlorococcus thermal growth restriction remain unexplored. This study brings us closer to understanding how Prochlorococcus functions under challenging temperature conditions, by focusing on its transcriptional response after long-term acclimation from its optimum to its thermal thresholds. Our results show that the drop in Prochlorococcus growth rate under cold conditions was paralleled by a transcriptional suppression of the photosynthetic machinery during daytime and a loss in the organism's regulatory capacity to maintain circadian expression patterns. Notably, warm temperature induced a marked shrinkage of the organism's cellular transcript inventory, which may also induce regulatory imbalances in the future functioning of this cyanobacterium.

Keywords: Prochlorococcus; cyanobacteria; marine; thermal acclimation; transcriptomics.

FIG 1

Growth rates, cell size, and number of mRNA transcripts per cell in samples of Prochlorococcus marinus MIT9301 collected after long-term thermal acclimation. Between 3 and 7 biological replicate samples are represented, depending on the temperature treatment, and error bars report the standard deviation between replicates. On the top of each plot, the minimum (T_min), optimum (T_opt), and maximum (T_max) temperatures are indicated, and additionally, in the upper left plot, temperature treatments where RNA samples were collected are shown with arrows. (A) Growth rate and size of MIT9301 cells along the thermal niche. (B) Estimates of mRNA transcripts per cell along the thermal niche in samples collected at daytime (left panel) and nighttime (right panel). Lowercase italic letters denote statistically significant differences (analysis of variance [ANOVA] and Tukey post hoc test; P < 0.001).

FIG 2

Clusters of Prochlorococcus marinus MIT9301 genes according to their pattern of daytime and nighttime expression along the thermal niche. Within each cluster, the right and left panels represent daytime and nighttime expression of the same genes, respectively. Dot colors indicate local density at each point of the scatterplot, with red circles indicating a high density of dots. Line colors within each SoftCluster indicate the membership value assigned by the fuzzy c-means soft clustering of each gene, ranging from 1 (red, high score) to 0.5 (blue, low score). Genes with a membership value lower than 0.5 are not plotted, and neither are they included in the total number of genes for each cluster (but they are included in Table S5). Below each cluster, the top 20 genes expressed are shown in rank order plots, and the expression levels (measured as [mRNA] and day/night log₂ fold expression ratio) of a selection of representative genes are shown. Daytime values appear as blue lines and nighttime values as black lines. Between 3 and 7 biological replicate samples are represented, depending on the temperature treatment and error bars report the standard deviation between replicates. Asterisks in the plots denote significant differences in transcript concentration along the thermal gradient in daytime (blue asterisks) or nighttime (black asterisks) according to the Kruskal-Wallis test (*, P< 0.05; **, P < 0.01; ***, P < 0.001). In the log₂ fold ratios, values above 0 represent preferential expression during daytime (highlighted in blue), while values below 0 represent preferential expression during nighttime.

FIG 3

Cellular expression levels (measured as [mRNA]) in Prochlorococcus marinus MIT9301 during daytime (blue lines) and nighttime (black lines) along the thermal niche of (A) RNA polymerase components, including sigma factors, (B) histidine kinases and other regulatory proteins, (C) genes involved in the stress response, and (D) nitrogen and phosphate acquisition genes. Asterisks denote significant differences in transcript concentration along the thermal gradient in daytime (blue asterisks) or nighttime (black asterisks) according to the Kruskal-Wallis test (*, P < 0.05; **, P < 0.01; ***, P < 0.001). The SoftCluster membership of each gene is shown only for those cases where the probability score was >0.80.

FIG 4

Schematic diagram showing some of the main components of the photosynthetic and carbon metabolism pathways in Prochlorococcus. Colors of genes (or their corresponding protein complexes) follow the same code as their respective SoftClusters in Fig. 2.

FIG 5

Expression patterns of a selection of Prochlorococcus photosynthetic genes of photosystem II (psbA and psbJ) and photosystem I (psaA, psaB, and psaF) along the thermal gradient 17 to 30°C in experimental acclimations (as analyzed in P. marinus MIT9301 by transcriptomics [left panel]) and in situ environmental conditions (as analyzed in reads mapping to MIT9301 identified in metatranscriptomes of the Tara Oceans data set [right panel]). In both cases, reads were normalized using DESeq2 and log transformed. Linear regression lines are shown for all genes with significant Spearman correlation coefficients at a P value of <0.05 (indicated in the plots).