Download Coastal Ocean Institute - Final Project Report

Coastal Ocean Institute - Final Project Report
A transcriptome-enabled study of nutrient and CO2 responses for three harmful algae
Principal Investigator: Sonya Dyhrman
What were the primary questions you were trying to address with this research? (Or, if more
appropriate, was there a hypothesis or theory that you were trying to prove or disprove?)
Harmful algal blooms (HABs) can be detrimental to coastal ecosystem services, aquaculture, and public health.
Many of these blooms are related to nutrient availability, particularly in coastal and estuarine settings. Despite
decades of study, the ability to predict how nutrients and CO2 influence the growth of different coastal HABforming phytoplankton species is still limited. Recent advances in DNA/RNA sequencing make it possible to study
the physiological response of HAB species to nutrient availability with unprecedented resolution. The goal of this
project was to grow several HAB species under low nitrogen, low phosphorus, nutrient replete (control), and high
CO2 conditions. RNA from each growth condition was sequenced for the purpose of 1) identifying the genes that are
expressed, or “turned on”, and 2) quantifying the number of each expressed gene across the different treatments. In
this way responses to these conditions can be identified along with their mechanistic underpinnings.
What have you discovered or learned that you didn't know before you started this work?
We looked at the transcriptomes of several different HAB species, including two dinoflagellates, Prorocentrum
minimum and Alexandrium monilatum, the raphidophyte Heterosigma akashiwo, and the brown-tide forming alga
Aureococcus anophagefferens. There was a tremendous amount of data generated for these four species. For
example, the Heterosigma and Aureococcus transcriptomes yielded upregulated transcripts related to the transport
and metabolism of nitrogen in the low nitrogen treatment and upregulated transcripts encoding enzymes that
hydrolyze organic phosphorus in the low phosphorus treatment (Figure 1). Aureococcus also had upregulated
transcripts encoding enzymes to relieve arsenic toxicity. Heterosigma upregulated transcripts related to lightharvesting and carbon concentrating mechanisms under high CO2, including two sodium bicarbonate co-transporters
(SLC4A1 and SLC4A10). While these genes were found in all conditions, they were most highly expressed under
high CO2. Work is ongoing to try to understand how these genes might be functioning in the cell under high CO2.
What is the significance of your findings for others working in this field of inquiry and for the
broader scientific community?
The transcriptome profiles for these species grown under these different growth conditions have not, to our
knowledge, been cross-compared before. These profiles are a powerful tool that will help the community of HAB
scientists investigate the extent to which HAB species share common strategies in their nutritional physiology that
contribute to their success at bloom formation.
What is the significance of this research for society?
Recent work indicates that the HAB species we studied will grow faster with the increased CO2 and temperatures
predicted in many coastal environments, yielding the potential for more severe and economically costly algal
blooms. Understanding how key HAB species, particularly those found in coastal environments, respond to
different nutrient and CO2 perturbations will aid in monitoring and management strategies.
What were the most unusual or unexpected results and opportunities in this investigation?
A comparison of the different species’ transcriptomes yielded conserved, or similar, nutrient-regulated responses.
These responses suggest that phytoplankton that form dense, coastal blooms may undergo similar physiological
shifts in response to changes in their environment. Investigating these similarities will aid in our understanding of
what drives the ecological success of these species in situ.
What were the greatest challenges and difficulties?
Sequencing of the different growth conditions for each species generated between 20 – 40 million reads, or small
fragments. These reads had to be edited to remove any errors and trimmed to select for fragments of a certain size.
These trimmed reads were then assembled to generate transcripts. The number of transcripts sequenced for each
species ranged between 31,000 – 41,000. These transcripts were then mapped to existing genomes, when available,
or annotated using other sequence databases and processed through a software program to identify gene expression
patterns (Figure 2). While these large datasets provided, with unprecedented resolution, a unique window into the
capacity of these organisms to acquire nutrients, and the physiological costs and tradeoffs associated with different
strategies under conditions of variable nutrient and CO2 supply, they also presented huge computational and
analytical challenges.
When and where was this investigation conducted? (For instance, did you conduct new field
research, or was this a new analysis of existing data?)
This experimental portion of this work was conducted in the laboratory as large-scale phytoplankton growth studies
in 2011 – 2012. The analysis of the transcriptomes began in 2013.
What were the key tools or instruments you used to conduct this research?
This project relied on the implementation of an analysis pipeline we developed to process sequence data provided by
the National Center for Genome Resources for the Marine Microbial Eukaryote Transcriptome Sequencing Project,
a Gordon and Betty Moore Foundation initiative.
Is this research part of a larger project or program?
Yes – this is part of ongoing efforts to understand the physiological ecology of coastal bloom forming phytoplankton
species.
What are your next steps?
Analysis of individual transcriptomes will continue to tease out additional genes of interest related to nutritional
physiology of HABs. Work will also continue to survey shared genes and metabolic pathways across different
phytoplankton species, as well as within a single species across different growth conditions. As an example, this
project has provided us with the ability to grouping the expression of Heterosigma transcripts by metabolic pathway
(Figure 2). These maps allow for a broad overview of physiological differences between growth conditions and
across species. Several publications are also in preparation.
Have you published findings or web pages related to this research? Please provide a citation,
reprint, and web link (when available).
Research and methods developed in this grant were a part of the following publications:
Frischkorn, K. R., Harke, M. J., Gobler, C. J. and Dyhrman, S. T. 2014. De novo assembly of Aureococcus
anophagefferens transcriptomes reveals diverse responses to the low nutrient and low light conditions present during
blooms. Frontiers in Microbiology. DOI: 10.3389/fmicb.2014.00375.
Wurch, L. L., Gobler, C. J., and Dyhrman, S. T. 2014. Expression of a xanthine permease and phosphate transporter
in cultures and field populations of the harmful alga Aureococcus anophagefferens: tracking nutritional deficiency
during brown tides. Environmental Microbiology: DOI: 10.1111/1462-2920.12374.
Please provide photographs, illustrations, tables/charts, and web links that can help illustrate
your research.
200
+P
-P
150
100
50
Ph
os
ph
at
as
e
rte
po
ns
tra
P
5'
N
uc
le
ot
id
as
e
5
4
3
2
1
0
r
Transcripts per million (TPM)
Heterosigma akashiwo CCMP2393
2500
2000
1500
1000
500
+P
-P
Ph
os
ph
at
as
e
rte
po
ns
tra
P
5'
N
uc
le
ot
id
as
e
50
40
30
20
10
0
r
Transcripts per million (TPM)
Aureococcus anophagefferens CCMP1850
Figure 1. Upregulated transcripts found in both HAB species investigated in
this study in response to low phosphorus. The number of transcripts was
normalized to the total number of transcripts (transcripts per million).
Control -N
-P
+CO2
Figure 2. Heatmap of the relative expression of KEGG (Kyoto Encyclopedia of Genes
and Genomes) modules associated with different metabolic pathways in Heterosigma
across the different growth conditions (Control, -N, -P and +CO2). Color indicates the
proportion of total transcripts mapping to each KEGG module relative to all KEGG
annotated transcripts. Grouping the expression of transcripts by metabolic pathway
allows for a broad overview of physiological differences between growth conditions and
across species.