surprisingly, some show a positive gravitropism, the opposite of normal. Unlike the phototropic mutants,
however, gametophores in the gtrmutants isolated thus far display a normal gravitropic response. Inter-
actions between the pathways for light- and gravity-mediated responses are beginning to be sorted out
[63].
D. Molecular Approaches
As with Volvox, many of the tools and techniques of molecular genetics have become readily available
for moss research only in the last 10 years or so (reviewed in Ref. 64). However, they have done so
with such success that Physcomitrellais now sometimes referred to as a “green yeast,” some of its
properties paralleling those of Saccharomyces cerevisiae[4]. It is very easily and efficiently trans-
formed, sometimes generating hundreds of transformed plants from a microgram of DNA [5,6] (J. Wal-
lace, unpublished results). Its genome size is estimated as a tractable 480 Mbp, about three times that
ofArabidopsis[65]. Higher plant promoters function in moss (e.g., Ref. 47), and codon usage is very
similar in both groups [48,66]. The tetracycline induction-repression system also functions well in
moss, so introduced transgenes can be turned on and off at will [67]. The moss Ceratodon purpureus
has also been successfully transformed [68], so Physcomitrellamay soon have some competition on the
molecular front.
Perhaps of most importance to the utility of Physcomitrellais that it is the only known land plant in
which a transgene will homologously recombine with its genomic counterpart with high efficiency [7].
This is allowing the approaches of targeted gene knockout and allele replacement, often referred to as “re-
verse genetics,” to be exploited in a plant as they have been so profitably exploited in fungal and animal
systems. Among the genes whose targeted disruption have been reported so far are a Cab gene [69]; a
gene encoding a delta 6-acyl-group desaturase [70]; FtsZ, an ancestral tubulin gene involved in plastid di-
vision [71]; and the multiubiquitin chain binding subunit of the 26S proteasome [72].
The disruption of the proteasome gene [72] is especially interesting, as it affects the early moss de-
velopmental pathway described earlier. The proteasome recognizes and degrades proteins that are
ubiquinated, either because they are aberrant or because their degradation is programmed (reviewed in
Ref. 73). Moss plants in which this gene has been knocked out behave as Budmutants; that is, they are
unable to make the switch from protonemal (filamentous) to three-dimensional growth. Thus, it appears
as though specific protein turnover is part of this basic differentiation process. Surprisingly, however, the
developmental block can be overcome by treatment with auxin and cytokinin. This suggests that the
switch in cellular development is mediated by at least these two processes—targeted proteolysis and
changes due to hormone signal transduction—and that increasing the signal from one can overcome a
deficit in the other. It seems likely that basic cell differentiations are usually effected by several things:
targeted proteolysis, hormones, mRNA degradation, phosphorylations, calcium, cAMP, etc. These are
probably all involved and interrelated in changing cell fate. If the example from Physcomitrellacan be
generalized, a deficiency in one of these mechanisms (e.g., specific protein degradation) can be overcome
if one of the others (e.g., hormone response) is increased.
A final feature of Physcomitrellathat makes it somewhat like a green yeast is the phenomenon of un-
stable transformants. When moss is transformed and placed on selective medium, three types of resistant
colonies emerge: transient expressers of the transgene, which die after a few weeks; stable transformants;
and unstables, which continue to grow (albeit slowly) and express the transgene as long as they are on se-
lective medium [74,75]. If an unstable transformant is grown without selection pressure for a couple of
weeks, it loses its ability to express the transgene and dies if replaced on selection. It is believed that in
unstable transformants the transforming plasmid is maintained and replicates extrachromosomally, and in
many cases it can be recovered from the moss by transforming the moss DNA into E.coli, where the plas-
mid will still confer prokaryotic antibiotic resistance (C. Knight and J. Wallace, unpublished results).
Thus, the possibility exists of generating a “shuttle vector” that can be transferred from moss to bacteria
and back.
In this era of comparative genomics and proteomics, the discovery that so many genes and proteins
in higher organisms have homologues in simpler ones has proved to be of great benefit because proper-
ties and functions of the genes are generally much more easily explored in the latter. Thus, it is no sur-
prise that studies to generate expressed sequence tag (EST) libraries in Physcomitrellahave already be-
gun [48,66]. One can expect that sequencing of the entire moss genome is not too far off.
814 WALLACE