studies can now be aided by the use of the endogenous aryl sulfatase (Ars) gene, which can serve as a re-
porter gene and has an inducible promoter [29].
Another molecular tool added to the box for Volvoxresearch is the characterization of a transposable
element, waggishly named Jordanfor its first-class jumping ability [30]. This allows the techniques of
transposon tagging, which have been of great utility in Arabidopsisand other plants [31,32], to be em-
ployed in Volvox. Transposition of Jordancan be induced by temperature shock [1,30], and activity of a
tagged gene is frequently restored when it is excised [30]. Its disadvantage is that it is present in roughly
50 copies/genome, so identification of a gene that has been tagged by Jordanis somewhat problematic.
However, this problem can be overcome, and both the regAand a glsgene have been cloned via transpo-
son tagging with Jordan[33,34].
The cloned glsgene is now referred to as glsA; different letters will be assigned to other genes whose
disruption causes the gonidialess phenotype as they are cloned [34]. Its identity as a glsgene was con-
vincingly demonstrated by its ability to rescue well-established mutants when introduced via the trans-
formation system outlined before. Consistent with the proposed role of glsgenes in controlling the asym-
metric division that gives rise to the differentiation of germ and somatic cells (Figure 2), glsAmessenger
RNA (mRNA) is expressed during cleavage, with the highest level seen at the asymmetric cleavage itself.
It is not expressed at all in mature somatic cells. The glsA protein has been localized to the mitotic spin-
dle, and a model has been proposed for its involvement in the asymmetric cleavages that lead to the germ-
soma dichotomy.
Much to the gratification of all, the sequence and expression pattern of the regAgene also turn out
to be consistent with its proposed role (Figure 2) as an inhibitor of gonidia-specific genes [33]. RegA
mRNA can first be detected in very young somatic cells, where it increases, and is not detectable in go-
nidial cells or precursors. Although its amino acid sequence does not show direct homology to any known
protein, it contains several features (helix-loop-helix domain, nuclear localization signal, a region very
rich in Glu, Ala, and Pro) that are compatible with its being a transcriptional repressor. Whether or not
theregAlocus’ hypermutability is a consequence of programmed DNA rearrangement [21] is a target of
current investigation.
A surprising finding related to the regAgene, however, has been the genes that are the presumed tar-
gets of its repression: the cloning and identification of complementary DNAs (cDNAs) that are specifi-
cally expressed in gonidial cells [35] have revealed that a substantial majority are nuclear-encoded genes
that are utilized in the chloroplast for photosynthesis [33,36]. This has led to the hypothesis that somatic
cells are unable to enter into a reproductive program largely due to an inability to photosynthesize; they
senesce at least partly because they run out of metabolic reserves [33,37]. An appealing feature of this hy-
pothesis is that it can fairly simply explain the evolutionary origin of the germ-soma dichotomy: because
many photosynthetic genes in higher plants are coordinately regulated via common cis-acting elements
in their promoters [38], it is not difficult to imagine that the regAgene could have evolved from a preex-
isting gene that encodes a transcription factor that carries out this regulation. Further examination of regA
and its homologues in different volvocalean species will probably shed light on this hypothesis. Its rele-
vance to basic mechanisms of cell differentiation in higher organisms should also prove very interesting.
III. MOSSES
Bryophytes are clearly much more closely related to higher plants than are the Volvocales, both bio-
chemically and functionally. Mosses respond to many of the same growth regulators as higher plants (re-
viewed in Ref. 39)—in fact, one of the first direct demonstrations of the effect of a cytokinin on plant de-
velopment was its stimulation of bud development in a moss [40]. Like higher plants, bryophytes have a
multicellular, differentiated sporophyte, and there is no separation early in development of discrete germ
line and somatic tissue. Thus, the study of developmental and physiological processes in mosses is quite
relevant to higher plants.
As with Volvox, genetic analysis of bryophytes, especially the generation of mutants, is greatly aided
by the fact that the dominant stage of the life cycle is haploid. Unlike the situation for Volvox, however,
most differentiated cells of mosses, like those of many higher plants, are totipotent, remaining capable of
redifferentiating and thence giving rise to an entire new plant. Thus, mutations that produce a sterile phe-
notype can still be maintained and studied. They can even be further mutagenized, because techniques of
808 WALLACE