Carbon Availability Transcriptomics

Chlamydomonas reinhardtii CC-1690 was grown in bioreactors. When the cells reached a non-stationary density, the acetate supply was stopped and the medium was kept at 25°C (control) or heated to 35 °C and 40 °C. Additionally to a preheat sample, four further samples were taken as triplicates after 2h, 4h, 8h, and 24h of heat treatment (nopump samples). The samples were analysed with NGS. In Zhang et al 2022 samples were taken at the same time points for 35°C and 40°C but with constant nutrient supply (TAP samples).

Table 1: Sampling schema for Transcriptomics analysis. Three biological replicates were measured. An x indicates triplicates measured at the respective time points.

Comparative analysis of both experiments were performed within this ARC. These include

1. combined normalization

2. statistical analysis

3. enrichment studies

4. visualization of all transcript signals and combined functional terms (Fig. 1)

Figure 1: Example of generated figures. (A) Exemplary visualization of the normalized counts of the HSP70C transcript. TAP 25°C samples are missing for 4 h and 8 h time points. A clear separation of the different temperature kinetics is visible. While the initial level is comparable for all time courses after heat onset of 35°C transcript counts increase strongly while the 25°C signal seems constant. Transcript counts in both 40°C experiments decreased during the first 4 hours of treatment. After 8 hours of heat stress the behaviour of temperature-regulated signals change to medium specific effects. Cells living in low-acetate media show distinct reduction of HSP70C transcripts while TAP-samples settle approximately at prior-heat levels. (B) Heatmap representation of (A). (C) Transcript signals that belong to the functional term "intraflagellar transport.IFT particle protein.complex B" are visualized as z-scores. Since there was no measurement for TAP-25°C this panel remains empty (lower left). Because z-scores may distort signals that did not change at all, a ANOVA was performed to separate constant transcript from transcripts that showed differential expression within their time courses. Red shadings indicate a global change of the transcript counts. Blue/grey signals can be considered as constant and have less relevance for the shown kinetics. As seen most transcripts show no response at 25°C but show distinct patterns for 35°C and 40°C respectively. The response shape is solely dependent on the applied temperature and is not influenced by the media the cells are grown in. While for 35°C the transcripts show a strong decrease within the first two hours and a strong increase during the last period, for 40°C samples the transcripts remain constant for the first 2 hours and show no strong increase for the last time point.

Methods

RNASeq

[sample preparation]

Raw sequencing data for constant-acetate experiments at 35°C and 40°C were obtained from (Zhang et al., 2022). Raw sequencing data for constant-acetate experiment at 25°C and acetate-depleting experiments at 25°C, 35°C, and 40°C were collected in this study and are available at JGI_XXX. RNA libraries were prepared and sequenced by the Joint Genome Institute (JGI, Community Science Program) using the NovaSeq platform generating 150-nt paired-end reads. Samples were quality control filtered using the JGI BBDuk and BBMap pipelines (Bushnell et al.). Samples were quality assessed using FastQC (Andrews) and mapped to the Chlamydomonas reinhardtii v5.6 genome (Merchant et al. 2007) using HISAT2 version 2.2.0 (Kim et al., 2015). Reads per feature were counted via featureCounts (Liao et al., 2014). The count matrix was combined with existing data (Zhang 2022), resulting in a 16,403x101 count matrix. Prior to imputation of two missing time points in the constant-acetate 25°C experiment, transcripts were filtered to have nonzero counts in at least 90 % of the samples.

Imputation

Measurements for time points 4h and 8h are missing for the constant-acetate 25°C time course. Beside preheat, 2h, and 24h samples, additional samples were taken for the constant-acetate 25°C time course at time points 0h, 0.5h, 1h, 26h, and 48h. After ensuring a high correlation (corr >= 0.993), the two missing time points were imputed as follows: For each transcript, two of the additional timepoint-triplicates were sampled randomly. It was ensured that not both missing timepoints were imputed with the same count data. The final count matrix consists of 14,893 transcripts in 90 samples (5 time points measured at 6 conditions as triplicates).

Statistical testing

Imputed count data were tested for differential expression using DESeq2 v1.38.3 (Love et al., 2014). The following tests were performed: (i) constant-acetate 25°C vs constant-acetate 35°C; (ii) constant-acetate 25°C vs constant-acetate 40°C; (iii) acetate-depleting 25°C vs acetate-depleting 35°C; (iv) acetate-depleting 25°C vs acetate-depleting 40°C; (v) (constant-acetate 35°C and acetate-depleting 35°C) vs (constant-acetate 40°C and acetate-depleting 40°C); (vi) (constant-acetate 35°C and constant-acetate 40°C) vs (acetate-depleting 35°C and acetate-depleting 40°C); (vii) interaction of v and vi.

Sample normalization

The normalization of the count matrix was conducted using the median of ratios method (Love et al., 2014).

Correlation analysis

The natural logarithm was determined for all normed counts before averaging triplicates. Pearson correlation coefficients were calculated for all sample pairs. Correlation coefficients varied from 0.71 to 1.0. All constant-acetate 25°C samples highly correlated with each other (Sup. Fig. X) with the lowest correlation of 0.993 between the preheat and 48h time point.

PCA

PCA was performed on averaged triplicates using FSharp.Statsv0.4.11 on transcripts that were nonzero in at least 26 of the 30 samples (Venn et al., 2022a).

Functional set figures

Functional descriptions were determined for each transcript (Merchant et al., 2007; Usadel et al., 2009; Venn and Muehlhaus, 2022b). Transcripts were grouped according to their functional description and time courses of their averaged normed counts were visualized as z score and log2 fold change respectively. An ANOVA was performed for each transcript at each treatment to elucidate whether a transcript underwent a relevant change during its time course (Venn et al., 2022a).

Visualization

Heatmaps, PCA, and functional sets figures were created using Plotly.NETv4.0.0 (Schneider et al., 2022).

References:

• Benedikt Venn, Lukas Weil, Kevin Schneider, David Zimmer & Timo Mühlhaus. (2022a). fslaborg/FSharp.Stats. Zenodo. https://doi.org/10.5281/zenodo.6337056
• Benedikt Venn & Timo Mühlhaus (2022b), CSBiology/OntologyEnrichment: Release 0.0.1 (0.0.1). Zenodo. https://doi.org/10.5281/zenodo.6340412
• Love, M.I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014). https://doi.org/10.1186/s13059-014-0550-8
• Andrews, S. (n.d.). FastQC A Quality Control tool for High Throughput Sequence Data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/
• BBDuk: https://sourceforge.net/projects/bbmap/
• Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nature Methods, 2015 Mar 9. doi: 10.1038/nmeth.3317 .
• Ramirez F, Dundar F, Diehl S, Gruning BA, Manke T. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 2014 Jul 1;42:W187-W191.
• Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014 Apr 1;30(7):923-30.
• Schneider K, Venn B and Mühlhaus T. Plotly.NET: A fully featured charting library for .NET programming languages [version 1; peer review: awaiting peer review]. F1000Research 2022, 11:1094 (https://doi.org/10.12688/f1000research.123971.1)
• Merchant, S. S., Prochnik, S. E., Vallon, O., Harris, E. H., Karpowicz, S. J., Witman, G. B., … Grossman, A. R. (2007). The Chlamydomonas Genome Reveals the Evolution of Key Animal and Plant Functions. Science, 318(5848), 245–250. https://doi.org/10.1126/science.1143609
• Zhang, N., Mattoon, E.M., McHargue, W. et al. Systems-wide analysis revealed shared and unique responses to moderate and acute high temperatures in the green alga Chlamydomonas reinhardtii. Commun Biol 5, 460 (2022). https://doi.org/10.1038/s42003-022-03359-z
• featureCounts: Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014 Apr 1;30(7):923-30.
• Usadel B, Poree F, Nagel A, Lohse M, Czedik-Eysenberg A, Stitt M (2009) A guide to using MapMan to visualize and compare Omics data in plants: a case study in the crop species, Maize. Plant Cell Environment, 32: 1211-1229