are said to be at the compensation point, and those leaves are hypothesized to be targeted to senesce [204].
In this regard, the level of photosynthates (or other nutrients), evaluated at the leaf phloem, could act di-
rectly or indirectly as a signal to trigger the senescence-associated dismantling of the leaf. Metabolic im-
balances caused by restricted photosynthetic activity [6,205], but also by the strong nitrogen demand from
growing organs [206], have been proposed to act as signals triggering the senescence program in old
leaves. Nevertheless, the signaling mechanisms by which these imbalances are supposed to be perceived
are currently indeterminate. The metabolic signal hypothesis is supported by several indirect experimen-
tal observations. Victorin, a fungal toxin whose only demonstrated effect at the molecular level is the in-
hibition of a photorespiratory enzyme, is known to induce a decay process in oat leaves displaying all typ-
ical symptoms of senescence (chlorophyll loss, Rubisco degradation, and chromatin fragmentation, as
well as ethylene and Ca^2 signaling) [207]. Moreover, in agreement with the hypothesis of a photosyn-
thate sensor, overexpression of Arabidopsishexokinase, a key regulatory enzyme in sugar metabolism,
induces rapid senescence in transgenic tomato plants [208].
Adverse environmental conditions caused by both biotic and abiotic factors (such as viral infections
[209], ozone exposure [210] or continuous darkness [199]) have been shown to induce senescence. Be-
sides, stress can advance or accelerate natural senescence to different degrees depending on the type and
intensity of the stress and the developmental stage of the plants [211]. Stress processes may have a direct
influence on the onset of leaf senescence through the putative photosynthate sensor signaling because
they usually cause a decline in photosynthetic efficiency. However, overproduction of ROIs (a common
consequence of stress processes) may also work as a signal triggering secondary responses inside the cell,
especially at the membrane level. This is particularly true in case of oxidative processes such as the stress
caused by ozone pollution (reviewed in Ref. 212).
Most plant growth regulators (phytohormones) are known to influence senescence, either promoting
(ethylene, abscisic acid, and jasmonates) or inhibiting (cytokinins, auxins, and gibberellins) it. However,
cytokinins and ethylene have been demonstrated to exert the greatest influence on plant senescence pro-
cesses. A decline of cytokinins is observed during natural leaf senescence, and the exogenous application
of cytokinins to excised leaves prevents senescence [213]. Moreover, delayed leaf senescence in tobacco
plants expressing the maize homeobox gene knotted1under control of the promoter of a senescence-reg-
ulated gene (SAG12) is accompanied by increased cytokinin content [214]. Perhaps the most conclusive
experiment on the effect of cytokinins in senescence has been the transformation of tobacco plants with
an enzyme of cytokinin synthesis (the Agrobacterium tumefaciensisopentenyl transferase gene) under the
SAG 12 promoter. These plants show an autoregulatory production of cytokinins (thereby avoiding the
deleterious effects of overproduction) and a concomitant delay of plant leaf senescence without any other
phenotypic alterations [17].
Ethylene accelerates the onset of senescence in some plant species. The Arabidopsismutant etr1-1,
which is insensitive to ethylene because of an inactive receptor, shows a delay of several days in chloro-
phyll loss [215]. Moreover, leaf senescence is transiently delayed in transgenic tomato plants with
blocked ethylene biosynthesis due to antisense expression of the 1-amino cyclopropane-1-carboxylic acid
(ACC) oxidase gene, but once the senescence process is started, the expression pattern of SAGs does not
differ from that of wild-type plants [216]. These results indicate that ethylene influences senescence tim-
ing and, because the ethylene concentration increases during stress conditions, this could be a mechanism
for adjusting the speed of the plant response to the environment.
In the last instance, control of senescence has to be exerted through regulation of gene expression
acting at both the nuclear and the chloroplast genome. Some of the most abundant proteins in the chloro-
plast, such as Rubisco and chlorophyll a/bbinding proteins, are composed of different subunits encoded
in both the nuclear and the chloroplastic genome. A certain degree of coordination in the expression of
the genomes must exist during senescence. Indeed, nuclear control of cell senescence has been postulated
according to several facts [2]. First, there are mutations in the nuclear genome that alter the senescence
syndrome. Second, chloroplast senescence is prevented by enucleation. Third, selective inhibitors of nu-
clear RNA synthesis inhibit senescence-related processes. In contrast, specific inhibitors of organelle
RNA polymerases do not inhibit senescence. Something similar happens with the inhibitors of protein
synthesis: cycloheximide (an inhibitor of the 80S cytoplasmic ribosomes) blocks a variety of senescence-
related changes, whereas chloramphenicol (an inhibitor of the 70S organella ribosomes) does not delay
senescence. It has been suggested that the control of the nuclear-encoded chloroplast RNA polymerase
could be the key step in this coordination [2].
SENESCENCE IN PLANTS AND CROPS 193