The three-dimensional (3-D) network of the granum-stroma thylakoid assembly of vascular plant chloroplasts exhibits complex structural/functional heterogeneity. A complete understanding of the ultrastructure of this assembly is critical for our understanding of thylakoid function. The prevailing historical model of thylakoid structure, based on information derived from serial section analyses of electron microscopy (EM) images, suggests a helical arrangement of stroma membranes wound around the granum stacks. More recently, electron tomography has emerged as the leading method for the study of thylakoid ultrastructure, as it provides for higher resolution in the depth dimension. The first detailed 3D topological model derived from electron tomography was in disagreement with the helical model, whereas a more recent electron tomography study, conducted under somewhat different experimental conditions, suggested that basic features of the helical model are still valid. Here, we review the conventional EM data and present a critical discussion of the two electron tomography data sets in an attempt to establish a consensus model that accommodates all the information presently available.

The Arabidopsis AtCRK5 protein kinase is involved in the establishment of the proper auxin gradient in many developmental processes. Among others, the Atcrk5-1 mutant was reported to exhibit a delayed gravitropic response via compromised PIN2-mediated auxin transport at the root tip. Here, we report that this phenotype correlates with lower superoxide anion (O₂^•−) and hydrogen peroxide (H₂O₂) levels but a higher nitric oxide (NO) content in the mutant root tips in comparison to the wild type (AtCol-0). The oxidative stress inducer paraquat (PQ) triggering formation of O₂^•− (and consequently, H₂O₂) was able to rescue the gravitropic response of Atcrk5-1 roots. The direct application of H₂O₂ had the same effect. Under gravistimulation, correct auxin distribution was restored (at least partially) by PQ or H₂O₂ treatment in the mutant root tips. In agreement, the redistribution of the PIN2 auxin efflux carrier was similar in the gravistimulated PQ-treated mutant and untreated wild type roots. It was also found that PQ-treatment decreased the endogenous NO level at the root tip to normal levels. Furthermore, the mutant phenotype could be reverted by direct manipulation of the endogenous NO level using an NO scavenger (cPTIO). The potential involvement of AtCRK5 protein kinase in the control of auxin-ROS-NO-PIN2-auxin regulatory loop is discussed.