Umeå University's logo

Gene co-expression networks (GCNs) are a powerful approach for exploring transcriptional regulation by identifying functionally related genes through their expression patterns across various conditions. The inference of GCNs can be achieved by various computational algorithms, each with distinct merits and limitations. The choice of algorithm can influence the network structure and the biological interpretation derived from it. By using a combination of different methods, biases can be minimised providing more robust and complementary insights. These methodologies are particularly valuable for non-model species, a challenge exemplified by Norway spruce and Scots pine. With ongoing climate change, drought and cold stresses are becoming increasingly important factors shaping the survival of these boreal conifers. Boreal regions are experiencing more frequent and prolonged drought periods, alongside greater variability in early spring, including sudden freeze-thaw events and episodes of extreme cold. Understanding the genetic regulation through which species such as Norway spruce and Scots pine, perceive, respond to, and potentially recover from drought and cold is therefore of high importance. 

In this thesis I have used an extensive collection of transcriptomic data generated from boreal tree species under abiotic stress conditions to infer GCNs to reveal coordinated patterns of gene expression responses to environmental challenges. In addition, comparative analyses of GCNs enabled the systematic assessment of conservation and divergence of co-expression among these species, identifying both shared regulatory circuits and species-specific adaptations. Analyses uncovered down-regulated modules of developmental processes, up-regulated modules of abiotic stress response, and several candidate transcription factors directly connected to these stress-responsive pathways. Comparison with boreal angiosperms revealed divergent responses in core cold-regulatory processes, most notably in the regulation and representation of C- repeat Binding Factor (CBF) transcription factors. The abiotic stress response patterns of both cold and drought were largely shared between the two conifer species, indicating a high degree of conservation in their transcriptional responses. This conservation extended to the organisation of topologically associated domains, where a subset of highly conserved co-expressed orthologs were found at the same location in the genomes of these conifers. 

Together, these analyses demonstrated the utility of comparative co-expression networks as a tool for understanding both conserved and diverged regulatory mechanisms, while offering new perspectives on the resilience of conifers in the context of environmental change.

Samuttrycksnätverk (GCNs) är ett kraftfullt verktyg för att utforska transkriptionsreglering genom att identifiera funktionellt relaterade gener via deras uttrycksmönster under olika förhållanden. GCNs kan konstrueras med hjälp av olika beräkningsalgoritmer, som var och en har styrkor och begränsningar. Valet av algoritm kan påverka nätverkets struktur och den biologiska tolkning som görs utifrån det. Genom att använda en kombination av olika metoder kan bias minimeras, vilket ger mer robusta och kompletterande insikter. Dessa metoder är särskilt värdefulla för icke-modellorganismer, vilket illustreras av gran och tall. I takt med det pågående klimatförändringarna blir torka och kyla allt viktigare faktorer som påverkar överlevnaden hos dessa boreala barrträd. Boreala regioner upplever nu mer frekventa och långvariga torkperioder, tillsammans med större variationer under tidig vår, inklusive plötsliga frysnings-tiningscykler och perioder av extrem kyla. Att förstå den genetiska regleringen genom vilken arter som gran och tall uppfattar, reagerar på och potentiellt återhämtar sig från torka och kyla är därför av stor betydelse.

I denna avhandling har jag använt en omfattande samling transkriptomdata genererad från boreala trädslag under abiotiska stressförhållanden för att konstruera GCNs och därigenom avslöja koordinerade uttrycksmönster som svar på miljöutmaningar. Dessutom möjliggjorde jämförande analyser av GCNs en systematisk bedömning av bevarande och divergens i samuttryck mellan arterna, vilket identifierade både delade regulatoriska nätverk och artspecifika anpassningar. Analyserna avslöjade nedreglerade moduler kopplade till utvecklingsprocesser, uppreglerade moduler involverade i abiotisk stressrespons, och flera kandidattranskriptionsfaktorer som var direkt kopplade till reglering av dessa stressrespons. Jämförelser mellan boreala angiospermer visade divergerande response i centrala processer reglerade av kyla, mest noterbart i regleringen och representationen av C-repeat Binding Factor (CBF)-transkriptionsfaktorer. Responsmönstren för både kyla och torka var i stor utsträckning delade mellan de två barrträdsarterna, vilket tyder på en hög grad av bevarande i deras transkriptionella svar. Detta bevarande sträckte sig även till organisationen av topologiskt associerade domäner, där en undergrupp av starkt bevarade samuttryckta ortologer hittades på samma position i båda arternas genom.

Sammantaget visar dessa analyser nyttan av att jämföra samuttrycksnätverk som ett verktyg för att förstå både bevarade och divergerade regulatoriska mekanismer, och ger samtidigt nya perspektiv på barrträdens resiliens i en föränderlig miljö.

Stem bending in trees induces flexure wood but its properties and development are poorly understood. Here, we investigated the effects of low-intensity multidirectional stem flexing on growth and wood properties of hybrid aspen, and on its transcriptomic and hormonal responses.

Glasshouse-grown trees were either kept stationary or subjected to several daily shakes for 5 wk, after which the transcriptomes and hormones were analyzed in the cambial region and developing wood tissues, and the wood properties were analyzed by physical, chemical and microscopy techniques.

Shaking increased primary and secondary growth and altered wood differentiation by stimulating gelatinous-fiber formation, reducing secondary wall thickness, changing matrix polysaccharides and increasing cellulose, G- and H-lignin contents, cell wall porosity and saccharification yields. Wood-forming tissues exhibited elevated jasmonate, polyamine, ethylene and brassinosteroids and reduced abscisic acid and gibberellin signaling. Transcriptional responses resembled those during tension wood formation but not opposite wood formation and revealed several thigmomorphogenesis-related genes as well as novel gene networks including FLA and XTH genes encoding plasma membrane-bound proteins.

Low-intensity stem flexing stimulates growth and induces wood having improved biorefinery properties through molecular and hormonal pathways similar to thigmomorphogenesis in herbaceous plants and largely overlapping with the tension wood program of hardwoods.

Gene regulatory and gene co-expression networks are powerful research tools for identifying biological signal within high-dimensional gene expression data. In recent years, research has focused on addressing shortcomings of these techniques with regard to the low signal-to-noise ratio, non-linear interactions and dataset dependent biases of published methods. Furthermore, it has been shown that aggregating networks from multiple methods provides improved results. Despite this, few useable and scalable software tools have been implemented to perform such best-practice analyses. Here, we present Seidr (stylized Seiðr), a software toolkit designed to assist scientists in gene regulatory and gene co-expression network inference. Seidr creates community networks to reduce algorithmic bias and utilizes noise corrected network backboning to prune noisy edges in the networks.

Using benchmarks in real-world conditions across three eukaryotic model organisms, Saccharomyces cerevisiae, Drosophila melanogaster, and Arabidopsis thaliana, we show that individual algorithms are biased toward functional evidence for certain gene-gene interactions. We further demonstrate that the community network is less biased, providing robust performance across different standards and comparisons for the model organisms.

Finally, we apply Seidr to a network of drought stress in Norway spruce (Picea abies (L.) H. Krast) as an example application in a non-model species. We demonstrate the use of a network inferred using Seidr for identifying key components, communities and suggesting gene function for non-annotated genes.

In-depth genome characterization is still lacking for most of biofuel crops, especially for centromeres, which play a fundamental role during nuclear division and in the maintenance of genome stability. This study applied long-read sequencing technologies to assemble a highly contiguous genome for yellowhorn (Xanthoceras sorbifolium), an oil-producing tree, and conducted extensive comparative analyses to understand centromere structure and evolution, and fatty acid biosynthesis. We produced a reference-level genome of yellowhorn, ∼470 Mb in length with ∼95% of contigs anchored onto 15 chromosomes. Genome annotation identified 22,049 protein-coding genes and 65.7% of the genome sequence as repetitive elements. Long terminal repeat retrotransposons (LTR-RTs) account for ∼30% of the yellowhorn genome, which is maintained by a moderate birth rate and a low removal rate. We identified the centromeric regions on each chromosome and found enrichment of centromere-specific retrotransposons of LINE1 and Gypsy in these regions, which have evolved recently (∼0.7 MYA). We compared the genomes of three cultivars and found frequent inversions. We analyzed the transcriptomes from different tissues and identified the candidate genes involved in very-long-chain fatty acid biosynthesis and their expression profiles. Collinear block analysis showed that yellowhorn shared the gamma (γ) hexaploidy event with Vitis vinifera but did not undergo any further whole-genome duplication. This study provides excellent genomic resources for understanding centromere structure and evolution and for functional studies in this important oil-producing plant.