The mRNA encoding the soybean rbcS gene, SRS4, is degraded into a set
of discrete lower-molecular-weight products in light-grown soybean see
dlings and in transgenic petunia leaves. The 5'-proximal products have
intact 5' ends, lack poly(A) tails, lack various amounts of 3'-end se
quences, and are found at higher concentrations in the polysomal fract
ion. To study the mechanisms of SRS4 mRNA decay more closely, we devel
oped a cell-free RNA degradation system based on a polysomal fraction
isolated from soybean seedlings or mature petunia leaves. In the soybe
an in vitro degradation system, endogenous SRS4 mRNA and proximal prod
uct levels decreased over a 6-h time course. When full-length in vitro
-synthesized SRS4 RNAs were added to either in vitro degradation syste
m, the RNAs were degraded into the expected set of proximal products,
such as those observed for total endogenous RNA samples. When exogenou
sly added SRS4 RNAs already truncated at their 3' ends were added to e
ither system, they too were degraded into the expected subset of proxi
mal products. A set of distal fragments containing intact 3' ends and
lacking various portions of 5'-end sequences were identified in vivo w
hen the heterogeneous 3' ends of the SRS4 RNAs were removed by oligonu
cleotide-directed RNase H cleavage. Significant amounts of distal frag
ments which comigrated with the in vivo products were also observed wh
en exogenous SRS4 RNAs were degraded in either in vitro system. These
proximal and distal products lacking various portions of their 3' and
5' sequences, respectively, were generated in essentially a random ord
er, a result supporting a nonprocessive mechanism. Tagging of the in v
itro-synthesized RNAs on their 5' and 3' ends with plasmid vector sequ
ences or truncation of the 3' end had no apparent effect on the degrad
ation pattern. Therefore, RNA sequences and/or structures in the immed
iate vicinity of each 3' end point may be important in the degradation
machinery. Together, these data suggest that SRS4 mRNA is degraded by
a stochastic mechanism and that endonucleolytic cleavage may be the i
nitial event. These plant in vitro systems should be useful in identif
ying the cis- and trans-acting factors involved in the degradation of
mRNAs.