Harmful Algal Bloom-Forming Organism Responds to Nutrient Stress Distinctly from Model 2 Phytoplankton 3 4 Craig Mclean1,2,*, Sheean T
Total Page:16
File Type:pdf, Size:1020Kb
bioRxiv preprint doi: https://doi.org/10.1101/2021.02.08.430350; this version posted February 10, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. 1 Harmful Algal Bloom-Forming Organism Responds to Nutrient Stress Distinctly From Model 2 Phytoplankton 3 4 Craig McLean1,2,*, Sheean T. Haley3, Gretchen J. Swarr1, Melissa C. Kido Soule1, Sonya T. 5 Dyhrman3,4, and Elizabeth B. Kujawinski1 6 1 Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, 7 Woods Hole, Massachusetts 02543, United States 8 2 MIT/WHOI Joint Program in Oceanography/Applied Ocean Science and Engineering, 9 Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, 10 Woods Hole, Massachusetts 02543, United States 11 3 Biology and Paleo Environment Division, Lamont-Doherty Earth Observatory, Columbia 12 University, Palisades, NY, 10964 USA 13 4 Department of Earth and Environmental Sciences, Columbia University, Palisades, NY, 10964 14 USA 15 16 * - corresponding author – [email protected] 17 Total Word Count: 6057 18 Introduction Word Count: 611 19 Methods Word Count: 1945 20 Results & Discussion Word Count: 3298 21 Conclusion Word Count: 197 22 Color Figs: Figs 1-5 1 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.08.430350; this version posted February 10, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. 23 Summary: 24 • Resources such as nitrogen (N) and phosphorus (P) play an important role in primary 25 production and constraining phytoplankton bloom dynamics. Models to predict bloom 26 dynamics require mechanistic knowledge of algal metabolic shifts in response to resource 27 limitation. For well-studied model phytoplankton like diatoms, this information is plentiful. 28 However, for less-studied groups such as the raphidophytes, there remain significant gaps 29 in understanding metabolic changes associated with nutrient limitation. 30 • Using a novel combination of metabolomics and transcriptomics, we examined how the 31 harmful algal bloom-forming raphidophyte Heterosigma akashiwo shifts its metabolism 32 under N- and P-stress. We chose H. akashiwo because of its ubiquity within estuarine 33 environments worldwide, where bloom dynamics are influenced by N and P availability. 34 • Our results show that each stress phenotype is distinct in both the allocation of carbon and 35 the recycling of macromolecules. Further, we identified biomarkers of N- and P-stress that 36 may be applied in situ to help modelers and stakeholders manage, predict, and prevent 37 future blooms. 38 • These findings provide a mechanistic foundation to model the metabolic traits and trade- 39 offs associated with N- and P-stress in H. akashiwo, and evaluate the extent to which these 40 metabolic responses can be inferred in other phytoplankton groups. 41 42 Key words: Blooms, Metabolism, Nutrient Stress, Physiology, Phytoplankton, Raphidophyte, 43 Recycling 2 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.08.430350; this version posted February 10, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. 44 Introduction: 45 46 The scarcity of nitrogen (N) and phosphorus (P) limits primary production across aquatic 47 ecosystems (Moore et al., 2013; Ardyna & Arrigo, 2020) by regulating the growth and structure 48 of phytoplankton communities (Arrigo, 2005; Follows et al., 2007; Harke et al., 2016). These 49 communities consist of an extremely diverse group of phylogenetically and physiologically 50 distinct phytoplankton (Keeling et al., 2014; Caron et al., 2017). The abundance of individual 51 phytoplankton groups can vary widely over space and time (de Vargas et al., 2015), due in large 52 part to their group-specific evolutionary and ecological strategies for managing resource 53 limitation. These strategies vary from reallocation of intracellular nutrients (Van Mooy et al., 54 2009; Allen et al., 2011; Kujawinski et al., 2017), reduction of nutrient quotas needed for 55 growth (Grzymski & Dussaq, 2012; Read et al., 2017), or increased extracellular scavenging 56 (Haley et al., 2017; Harke et al., 2017; Kujawinski et al., 2017). 57 The impact of distinct nutrient response strategies is most evident when considering 58 harmful algal blooms (HABs). HABs may occur when a set of physiological traits allow a single 59 phytoplankton group to out-compete its neighbors (Wilson et al., 2019). Such blooms have 60 increased in frequency over the past 40 years with climate change and eutrophication (Wells et 61 al., 2015), causing hundreds of millions of dollars in economic damages to fisheries and public 62 health (Hoagland et al., 2002). Many blooms have been linked to resource availability 63 (Anderson et al., 2008), and as a result, knowledge of nutrient response mechanisms and their 64 associated trade-offs towards fitness holds tremendous promise in managing HABs (Sharma & 65 Steuer, 2019). The mechanics of how nutrients influence metabolic response strategies have 66 been identified in only a few model phytoplankton (e.g., diatoms (Brembu et al., 2017), 67 coccolithophores (Rokitta et al., 2014; Rokitta et al., 2016; Wördenweber et al., 2018)), leaving 68 significant gaps in our understanding of nutrient responses in HAB-forming groups. 69 We do not know the extent to which metabolism data from model phytoplankton 70 reflects that of non-model phytoplankton. Although core metabolic functional redundancies 71 occur in all primary producers (Louca et al., 2018), physiological studies suggest that 72 phytoplankton groups vary widely in their capabilities beyond carbon processing (Smayda, 3 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.08.430350; this version posted February 10, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. 73 1997). For example, elemental stoichiometry studies observe that phytoplankton groups differ 74 in their intracellular macromolecular pool composition (Bonachela et al., 2016), and 75 transcriptome studies show that phytoplankton respond differently to environmental 76 perturbations (Alexander et al., 2015b). Therefore, direct evaluations of non-model 77 phytoplankton metabolism, specifically under conditions of acute shortages of essential 78 nutrients (stress), are needed to evaluate the extent to which metabolic knowledge from model 79 phytoplankton can describe other groups. 80 One factor driving the knowledge discrepancy between model and non-model 81 phytoplankton groups is the paucity of fully sequenced eukaryotic genomes for most 82 phytoplankton, which are required to build genome-scale models of metabolism (Thiele & 83 Palsson, 2010; Levitan et al., 2014; Smith et al., 2019). Due to this constraint, investigators have 84 used ‘omics techniques like transcriptomics or metabolomics to characterize metabolic 85 responses of phytoplankton to nutrient stress (Wurch et al., 2011; Rokitta et al., 2014; Alipanah 86 et al., 2015; Rokitta et al., 2016; Brembu et al., 2017; Alipanah et al., 2018; Hennon & Dyhrman, 87 2019; Wurch et al., 2019). While each of these methods offers substantial insights on 88 physiological differences among phytoplankton (Alexander et al., 2015a), they fail to capture 89 the system-level dynamics driving changes in metabolism when used in isolation 90 (Wördenweber et al., 2018). For example, while metabolomic approaches provide evidence of 91 biochemical reactions, predicting the mechanism driving the activity is challenging. By contrast, 92 transcriptomic techniques reveal pathway level changes, yet such changes may be inhibited by 93 post-translational regulatory processes beyond the scope of the data. Applying metabolomic 94 and transcriptomic methods in tandem circumvents these issues. However, computational 95 challenges have typically limited these multi-‘omic efforts to targeted analyses of specific 96 pathways rather than system-level changes (Kujawinski et al., 2017). Combining ‘omics 97 techniques has tremendous promise for understanding the diversity of phytoplankton 98 responses to N- and P-stress. 99 In this study, we examined N- and P-stress metabolism using a combination of 100 metabolomics and transcriptomics data for the HAB-forming raphidophyte, Heterosigma 101 akashiwo. H. akashiwo populations are distributed ubiquitously within subtropical 4 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.08.430350; this version posted February 10, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. 102 environments (Taylor & Haigh, 1993; Smayda, 1998), and their blooms have caused significant 103 economic losses (Rensel, 2007). Both N- and P-stress are known to be important