C. Protein Synthesis Is a Prerequisite for Abscission
Anatomists often observed that protein accumulated in separating abscission zone cells [3] and EM
observations of the cytoplasm showed it to be rich in organelles, particularly Golgi and rough
endoplasmic reticulum (RER) [70,132]. Studies of the incorporation of labeled amino acids and nu-
cleotides demonstrated very active synthesis of proteins and RNA in the AZ [123,126,133]. Some of
the proteins synthesized during abscission are thought to play an essential role in the process, as it has
been widely demonstrated that both transcriptional and translational protein synthesis inhibitors block
abscission [3].
Both qualitative and quantitative changes in the protein profiles of separating AZ cells have been re-
ported [2,134,135]. Complementary abscission-related changes in mRNA populations have also been ob-
served [137,138]. In addition, Poovaiah et al. [136] demonstrated that the pattern of protein phosphory-
lation is altered in abscising zones.
D. Cell Wall Hydrolases and Their Control
After the initial observations that cellulase increased in separation layers, there was some confusion be-
cause it was not appreciated that more than one form of cellulase was found in AZs. Lewis and Varner
[139] concluded that a cellulase isoenzyme with an alkaline isoelectric point (9.5 cellulase) was specif-
ically involved in bean leaf abscission. It was formed de novo in AZs and its activity seemed correlated
with weakening [139] (Figure 6). Antibodies raised against 9.5 cellulase were used to discriminate be-
tween it and the other isoenzymes involved in normal growth [140]. The 9.5 cellulase was localized in
the separation layer and adjacent stele of bean AZs [140,141], and its production was accelerated by
ethylene and inhibited by IAA [142] (Figure 6). This increase in specific cellulase isoforms has now
been reported in many different abscission systems, including leaves [140], flowers [4], and fruit
[16,143].
Despite its familiar name, 9.5 cellulase will not attack crystalline cellulose. It is assayed by its abil-
ity to break down soluble carboxymethylcellulose and is really a Cx-cellulase [144]. Its natural substrate
is not known, but is likely to be a -1,4-glucan in the wall matrix [145]. By itself, bean 9.5 cellulase will
not cause cells to separate, but it acts synergistically with pectinase [146].
On the basis of the anatomical observations, one might expect polygalacturonases (PGs) to be in-
volved in breakdown of the middle lamella. PG increases have been reported in Impatiens,Sambucus,
tomato, orange, and peach [5,143,147]. As with cellulase, there are abscission-specific isoforms of PG.
Transgenic plants have been used to show that the PG associated with ripening in tomato is not the same
as that involved in abscission [148]. Both exo- and endo-cleaving PGs increase, although the endocleav-
ing enzyme is probably more important.
The cellulase from bean and Sambucusabscission zones have been cloned [138,149,150]. The bean
cDNA has been sequenced and has 64% identically matched nucleotides to the cellulase from avocado
fruit [150]. The partial sequence of the SambucuscDNA shows that it is very similar [138]. They share
consensus sequences with a series of E2-type cellulase from microbial and other plant sources [144].
Bonghi et al. [143] have made use of this similarity by employing avocado fruit cellulase cDNA as a het-
erologous probe. It hybridized to 1.8- and a 2.2-kb mRNAs, which accumulated in ethylene-treated peach
abscission zones.
Tucker et al. [149] have studied the expression of bean 9.5 cellulase. In situ hybridization showed
that the cellulase mRNA was confined to the separation layer and the adjacent stele [151] (Figure 11).
Northern blot analysis indicated that cellulase mRNA was virtually absent from uninduced abscission
zones but increased as they weakened in ethylene [149] (Figure 12). This increase was dependent,
at least in the short term, on the presence of ethylene. Indoleacetic acid suppressed the increase even in
an ethylene atmosphere. Removal of ethylene after cellulase mRNA had started to accumulate, in con-
junction with inhibition of any endogenous ethylene with norbornadiene, caused the cellulase mRNA
levels to decline to very low levels [149]. Indoleacetic acid administered to the fracture surfaces after
abscission had occurred inhibited further accumulation of cellulase mRNA in the fracture surface cells
even in the presence of ethylene [71]. This suggests that expression of 9.5 cellulase is under the joint
control of both IAA and ethylene. The cellulase gene complete with its upstream sequences has now
been cloned [71].
ABSCISSION 219