Natural variation of respiration-related traits in plants

Abstract

Plant respiration is one of the greatest global metabolic fluxes, but rates of respiration vary massively both within different cell types as well as between different individuals and different species. Whilst this is well known, few studies have detailed population-level variation of respiration until recently. The last 20 years have seen a renaissance in studies of natural variance. In this review, we describe how experimental breeding populations and collections of large populations of accessions can be used to determine the genetic architecture of plant traits. We further detail how these approaches have been used to study the rate of respiration per se as well as traits that are intimately associated with respiration. The review highlights specific breakthroughs in these areas but also concludes that the approach should be more widely adopted in the study of respiration per se as opposed to the more frequently studied respiration-related traits.

Figure 1

Schematic illustration of most commonly used QTL mapping populations. Laborious generated populations can mainly be distinguished by bi-parental (left) and multi-parental (mid, right) origin. Several different bi-parental populations can be generated, depending on the trait to assess and offspring self-compatibility, including near isogenic lines (NILs), introgression lines (ILs), backcrossed inbred lines (BILs), recombinant inbred lines (RILs) heterogeneous inbred family (HIF), and double haploids (DH). On the contrary, multi-parental lines, like multi-parental advanced generation inter-cross (MAGIC) populations are recombinant inbreds of multiple inter-crossed donors or NAM populations in which the donors are crossed to one “nested” genetic background.

Figure 2

Schematic overview of accessing natural variation for GWAS. Fundamental for these kinds of studies is the presence of a population as diverse as possible to cover diversity in their genomic variants, like SNPs and structural variance (SV), as well as their phenotypic response, like metabolite levels and gas exchange rate. Both components are integrated to perform association mapping resulting in potential QTL detection. These QTL are constrained by determining the linkage disequilibrium for cosegregation and harbor in general candidate genes explaining the variance in the phenotypic response by comparing arithmetic mean over their haplotype distribution.