Current status and trends in forest genomics

Abstract

Forests are not only the most predominant of the Earth's terrestrial ecosystems, but are also the core supply for essential products for human use. However, global climate change and ongoing population explosion severely threatens the health of the forest ecosystem and aggravtes the deforestation and forest degradation. Forest genomics has great potential of increasing forest productivity and adaptation to the changing climate. In the last two decades, the field of forest genomics has advanced quickly owing to the advent of multiple high-throughput sequencing technologies, single cell RNA-seq, clustered regularly interspaced short palindromic repeats (CRISPR)-mediated genome editing, and spatial transcriptomes, as well as bioinformatics analysis technologies, which have led to the generation of multidimensional, multilayered, and spatiotemporal gene expression data. These technologies, together with basic technologies routinely used in plant biotechnology, enable us to tackle many important or unique issues in forest biology, and provide a panoramic view and an integrative elucidation of molecular regulatory mechanisms underlying phenotypic changes and variations. In this review, we recapitulated the advancement and current status of 12 research branches of forest genomics, and then provided future research directions and focuses for each area. Evidently, a shift from simple biotechnology-based research to advanced and integrative genomics research, and a setup for investigation and interpretation of many spatiotemporal development and differentiation issues in forest genomics have just begun to emerge.

Keywords: CRISPR-mediated genome editing; Forest genetic diversity and climate adaption; Genome assembly; Haploid induction; Perennial growth and seasonality regulation; QTL, association Studies, and genomic selection; Single cell RNA-seq; Systems biology and data analysis; Transformation and regeneration; Wood formation.

Figure 1

Light microscopy of 20-µM thick sections of 5-week-old Leucaena leucocephala nodules. (a) Induction of alkaline phosphatase activity in the Leucaena nodule formed by the midA::phoA mutant RUH128 of Rhizobium TAL1145. (b) Another mutant RUH129, in which the phoA insertion on midA was in the opposite orientation, was used as a negative control. Nodule sections were stained for phosphatase activity. RUH128 cells inside the nodule expressed phosphatase activity whereas RUH129 cells inside the nodule did not show any detectable phosphatase activity. Bar represents 400 µM. (c) Schematic of a Leucaena nodule occupied by a Rhizobium strain that degrades mimosine into pyruvate, formate and ammonia.

Figure 2

Roles of mimosine and free-living rhizobia in iron uptake by Leucaena. (1) Leucaena root exudates contain mimosine, which chelates Fe³⁺ to form Fe³⁺-mimosine complexes (2). Three types of Fe³⁺-mimosine complexes may be formed depending on the amount of mimosine available in the rhizosphere (box). The Fe³⁺-mimosine complexes are taken up by Leucaena roots using an oligopeptide-type of transporter (3). Fe³⁺-mimosine complexes can also be taken up by mimosine-degrading rhizobia in the rhizosphere using an ABC transporter (4). Rhizobia degrade Fe³⁺-mimosine complexes and release Fe²⁺ into the rhizosphere (5). Fe²⁺ is taken up by Leucaena through an IRT transporter (6).

Figure 3

(a), (b) Predictions of genomic offset to future climate change in Pinus densata using a full set of 47,612 SNPs in exome sequences. (c), (d) Subset of 2,025 significant GEA SNPs. (a) and (c) reflect scenario representative concentration pathway (RCP) 2.6 2070; (b) and (d) reflect scenario RCP 8.5 2070. Red and blue indicate high and low genomic offset, respectively (Adapted from Zhao et al.^[259]).

Figure 4

Clustering eight Pinus taeda full-sib families (colored circles) based on three different SNP marker sets. AgriSeq Targeted GBS panel (on the right) clustered the full-sib families similar to the same SNP markers selected from the Pinus taeda SNP array (in the middle). Trees not clustered are likely pedigree errors or the half-sibs (sharing one parent of another full-sib family).

Figure 5

Illustration of how to construct a multilayered hierarchical gene regulatory network (ML-hGRN) to encompass a given transcription factor (TF) using Top-down GGM and Bottom-up GGM Algorithms synergistically.