Differences between winter oilseed rape (Brassica napus L.) cultivars in nitrogen starvation-induced leaf senescence are governed by leaf-inherent rather than root-derived signals

Nitrogen (N) efficiency of winter oilseed rape (Brassica napus L.) line-cultivars (cvs.), defined as high grain yield under N limitation, has been primarily attributed to maintained N uptake during reproductive growth (N uptake efficiency) in combination with delayed senescence of the older leaves accompanied with maintained photosynthetic capacity (functional stay-green). However, it is not clear whether genotypic variation in N starvation-induced leaf senescence is due to leaf-inherent factors and/or governed by root-mediated signals. Therefore, the N-efficient and stay-green cvs. NPZ-1 and Apex were reciprocally grafted with the N-inefficient and early-senescing cvs. NPZ-2 and Capitol, respectively and grown in hydroponics. The senescence status of older leaves after 12 days of N starvation assessed by SPAD, photosynthesis and the expression of the senescence-specific cysteine protease gene SAG12-1 revealed that the stay-green phenotype of the cvs. NPZ-1 and Apex under N starvation was primarily under the control of leaf-inherent factors. The same four cultivars were submitted to N starvation for up to 12 days in a time-course experiment. The specific leaf contents of biologically active and inactive cytokinins (CKs) and the expression of genes involved in CK homeostasis revealed that under N starvation leaves of early-senescing cultivars were characterized by inactivation of biologically active CKs, whereas in stay-green cultivars synthesis, activation, binding of and response to biologically active CKs were favoured. These results suggest that the homeostasis of biologically active CKs was the predominant leaf-inherent factor for cultivar differences in N starvation-induced leaf senescence and thus N efficiency