Plant Molecular Biology 44: 699709, 2000.

2000 Kluwer Academic Publishers. Printed in the Netherlands.

699

Requirement for cytoplasmic protein synthesis during circadian peaks of

transcription of chloroplast-encoded genes in Chlamydomonas

Ryo Kawazoe, Seongbin Hwang1 and David L. Herrin

Section of Molecular Cell and Developmental Biology (Mail code A6700) and Institute for Cellular and Mole-
cular Biology, University of Texas at Austin, Austin, TX 78712, USA ( author for correspondence; e-mail:
dherrin@utxvms.cc utexas.edu); 1 present address: Dept. of Bioscience and Biotechnology, Sejong University,
Kwangjin-Gu Kunja-Dong 98, Seoul, Korea

Received 29 December 1999; accepted in revised form 28 April 2000

Key words: Chlamydomonas, chloroplast, circadian rhythms, cycloheximide-resistant mutant, nuclear-organelle

interactions, transcriptional regulation

Abstract
Cycloheximide, an inhibitor of cytoplasmic translation, induced a rapid reduction of 7080% in levels of mRNA
for the chloroplast elongation factor Tu (tufA) in asynchronously growing Chlamydomonas. This effect was shown
to be mainly transcriptional, and not restricted to tufA, as transcription of other chloroplast-encoded genes were
cycloheximide-sensitive, although not all equally (psbA showed no more than 40% inhibition). Confirmatory
evidence that the inhibition of chloroplast transcription was mainly due to blocking cytoplasmic translation was
obtained with the cycloheximide-resistant mutant act1, and by using another translation inhibitor, anisomycin.
In synchronously growing Chlamydomonas, chloroplast transcription is regulated by the circadian clock, with the
daily peak occurring during the early light period. When cycloheximide was added during this period, transcription
was inhibited, but not when it was added during the trough period (late light to early dark). Moreover, in synchro-
nized cells switched to continuous light, the drug blocked the scheduled increase in tufA mRNA, but did not remove
the pre-existing mRNA. These experiments define two functionally different types of chloroplast transcription in
Chlamydomonas, basal (cycloheximide-insensitive) and clock-induced (cycloheximide-sensitive), and indicate that
the relative contribution of each type to the overall transcription of a given gene are not identical for all genes. The
results also provide evidence for nuclear regulation of chloroplast transcription, thereby obviating the need for an
organellar clock, at least for these rhythms.

Introduction mechanisms by which these different developmental

Chloroplast-encoded genes have been shown to be
regulated at the transcriptional, and several post-
transcriptional levels (Gruissem and Tonkyn, 1993;
Mayfield et al., 1995; Goldschmidt-Clermont, 1998).
Transcriptional regulation has been documented dur-
ing development, in response to light, and in daily
rhythms (Herrin et al., 1986; Leu et al., 1990;
Schrubar et al., 1990; Mullet, 1993; Salvador et al.,
1993; Hwang et al., 1996). In addition, both global
and gene-specific regulation of chloroplast transcrip-
tion has been observed (Schrubar et al., 1990; Sexton
et al., 1990; Mullet, 1993). However, the precise
and environmental signals regulate transcription of
chloroplast genes are poorly understood.
Global and/or large-scale changes in chloroplast
transcription during development may be mediated,
at least in part, by the differential expression of two
different, plastid-localized RNA polymerases (Green-
berg et al., 1984; Mullet, 1990; Pfannschmidt and
Link, 1994). One of these appears to be mainly
nuclear-encoded and phage-like (Lerbs-Mache, 1993;
Allison et al., 1995; Hedtke et al., 1997), whereas
the other is mainly plastid-encoded and bacteria-like
(Greenberg et al., 1984; Gruissem and Tonkyn, 1993;
Allison et al., 1995). It should be noted, however,
700
that the putative sigma factors for the bacterial-like
polymerase are themselves nuclear-encoded (Liu and
Troxler, 1996; Tanaka et al., 1996; Isono et al.,
1997; Kestermann et al., 1998; Tozawa et al., 1998).
Thus, this situation provides the nucleus with the
potential for global control of chloroplast transcrip-
tion by changing expression of the nuclear-encoded
polymerase subunits. However, there is, as of yet,
little direct evidence for nuclear regulation of global
chloroplast transcription (Klein, 1991).
In the green alga Chlamydomonas reinhardtii, the
predominant chloroplast RNA polymerase seems to
be of the bacterial type. This is based on the fact
that rpo-like genes are encoded in the chloroplast
genome (Fong and Surzycki, 1991), and that the bulk
of chloroplast transcription is sensitive to rifampicin
(Surzycki, 1969; Fong and Surzycki, 1991). There
is no clear evidence for a nuclear-encoded chloro-
plast RNA polymerase in Chlamydomonas; however,
its absence cannot be ruled out. Indeed, Jahn (1992)
partially purified an RNA polymerase that transcribes
a chloroplast tRNAglu gene, although the subcellular
origin of this enzyme is not known.
We reported previously that transcription of several
and possibly most chloroplast-encoded genes are reg-
ulated by a circadian clock in C. reinhardtii. Circadian
clock control may also be important to vascular plants
as the chloroplast psbD gene in wheat was shown to
be under circadian control (Nakahira et al., 1998). In-
terestingly, modern relatives of the putative ancestor
to green plant chloroplasts, a cyanobacteria-like or-
ganism, also have a circadian clock that controls the
transcription of many genes, including the photosys-
tem II gene psbA (Johnson et al., 1996). This raises
the possibility that there could be an ancient circadian
clock still residing in the chloroplast, directly regulat-
ing transcription. Alternatively, an extrachloroplastic
clock could regulate chloroplast transcription, for ex-
ample, by directing the production of a regulatory
protein(s) in the cytoplasm, such as those discussed
above. In this report, evidence is presented which
supports the latter hypothesis, namely that a nuclear-
encoded factor(s) is responsible for the daily peaks of
chloroplast gene transcription. Moreover, this putative
factor is relatively short-lived.
Materials and methods
Cell strains and culture conditions
The wild type 2137 mt+ (CC-1021) and cycloheximi-
de resistant mutant act1 mt+ (CC-1589) were grown
mixotrophically in flasks of Tris/acetate/phosphate
medium (Harris, 1989) with shaking (175 rpm) at
23 C. To obtain asynchronous cultures, cells were
grown for 2 days in continuous light to a density of 2
106 cells/ml (exponential phase); the light intensity
was ca. 40 E m2 s1 . To obtain synchronous cul-
tures, the cells were grown for three days under 12 h
light/12 h dark cycles to a density of 1 106 cells/ml
(using the same light intensity). The synchrony of the
cultures was checked by monitoring cell number and
cell division (Hwang and Herrin, 1994). For circa-
dian experiments, synchronously grown cultures were
transferred to continuous light using the same intensity
as above. To perform inhibitor treatments, the drugs
were added under ongoing growth conditions. Other
details are given in the text or figure legends.
RNA isolation and northern blot hybridization
The isolation of total RNA, northern blot hybridiza-
tion, and 32 P labeling of DNA have been described
(Hwang and Herrin, 1994). The blots were stained
with methylene blue prior to hybridization (Herrin and
Schmidt, 1988) to check for equal loading and transfer.
The DNA probe was the PstI-EcoRI fragment of the
tufA gene (Hwang et al., 1996), which was labeled
to a specific activity of ca. 1 109 dpm/g. The
autoradiographs, exposed to obtain signals within the
linear response range, were scanned and quantified as
previously described (Hwang et al., 1996).
Run-on transcription in permeabilized cells
Run-on transcription in toluene-permeabilized cells
was performed as described previously (Hwang et al.,
1996). The transcription mixtures contained 5 107
cells (2 108 cells/ml), 15 Ci of [-32 P]UTP (spe-
cific activity ca. 3000 Ci/mmol) as label, and were
incubated at 25 C for 15 min. The specific radioactiv-
ity of the RNA was measured by using DE-81 paper
(Sambrook et al., 1989). The transcriptional activity
of individual genes was assessed by hybridizing equal
amounts of the pulse-labeled RNA (10 g/ml in hy-
bridization solution) to DNA dot-blots (duplicate dots
of 2.5 g and 0.5 g) for 72 h. DNA probes for five
genes were used: pB10 for exon 1 of psaA1; pEC23 for
 701

Figure 1. Effect of cycloheximide on levels of tufA mRNA. An asynchronously growing culture of wild-type C. reinhardtii was subdivided
into two, cycloheximide was added to one, and both were returned to ongoing growth conditions. RNA was extracted at the indicated times (in
h) from drug-treated (+CHX) and untreated (Control) cells, and analyzed by northern blot hybridization with a tufA probe. A. Autoradiograph
of a representative northern blot; the blot was exposed to X-ray film for ca. 6 h. B. Quantification of the blot in A. The data were normalized
relative to the zero-time value, which was arbitrarily set to 100%.

psbA; pR03R141 for tufA; pBC7 for atpB; and p228
for 16S rRNA (Hwang et al., 1996). After stringent
washing and exposure to X-ray film, the films were
quantified as above. Hybridization to the vectors were
negligible under these conditions.
Results
The effect of cycloheximide on tufA mRNA levels
The wild-type strain of C. reinhardtii, growing under
continuous light, was treated with 10 g/ml cyclohex-
imide to stop cytoplasmic translation (Hoober, 1970;
Herrin and Schmidt, 1987) (at this concentration syn-
thesis of major proteins in the cytoplasm is inhibited
immediately by >95%; Herrin and Schmidt, 1987),
and then levels of the 1.7 kb tufA mRNA were fol-
lowed by northern blot analysis. As Figure 1 shows,
the tufA mRNA level began decreasing ca. 1 h after
the drug was added, declining rapidly to a low level
(ca. 10% of untreated) by 4 h. This residual level of
tufA mRNA persisted for at least another 4 h (data not
shown). The half-life of tufA mRNA is relatively short,
ca. 0.51.5 h, compared to genes for photosynthetic
proteins (Pfannschmidt and Link, 1994; Hwang et al.,
1996). Hence, the mRNA data could be explained
by a rapid decrease in tufA transcription, although a
drop in the stability of the mRNA was also possible.
A rifampicin chase experiment (Hwang et al., 1996)
was performed with cycloheximide-treated and con-
trol cells, and the results indicated that the half-life
of tufA mRNA was unaffected by cycloheximide (data
not shown).
Run-on transcription assays of tufA and other
chloroplast-encoded genes
To determine directly if a change in transcriptional ac-
tivity was responsible for the decrease in tufA mRNA
levels, run-on transcription in permeabilized cells was
used to estimate transcription rates of tufA and four
other chloroplast genes (rrn, psbA, psaA exon-1, and
atpB). Also, overall plastid transcription activity could
be estimated since, with this system, more than 95%
of the [32 P]UTP incorporation is chloroplast-derived
(Guertin and Bellemare, 1979). Figure 2 shows that
total and gene-specific transcriptional activity declined
702

Figure 2. Effect of cycloheximide on transcription of tufA and other chloroplast-encoded genes. An asynchronously growing culture of
wild-type C. reinhardtii was subdivided into two aliquots, cycloheximide was added to one (+CHX), and both were returned to ongoing
growth conditions. At the indicated intervals, cells were harvested and used for run-on transcription with [32 P]UTP. The 32 P-labeled RNAs
were hybridized to duplicate dots of 0.5 g (1X) and 2 g (4X), respectively, of several gene probes. A. Representative autoradiographs. The
32 P-labeled RNAs were from cells at the time of culture separation (0 time), and 4 h later (4 h control and 4 h +CHX). B. Time courses of
total and gene-specific transcription activity. Total incorporation of [32 P]UTP was determined as described in Materials and methods. The 4X
dots were scanned and each point represents an average of duplicates, which varied no more than 15%. The data were normalized relative to
the maximum values, which were arbitrarily set to 100%.
 703

Figure 3. Effect of anisomycin and cycloheximide on tufA mRNA levels and transcription rates in wild-type and the cycloheximide-resistant
mutant, act1. Asynchronously growing cells of each strain were treated with cycloheximide (+CHX), or anisomycin (+ANI), or left untreated
(Control), and then harvested at the indicated times. Portions were extracted for northern analysis or used for run-on transcription assays. A.
Time course of tufA mRNA levels in wild-type (WT) and act1 cells. The data were normalized relative to the highest value, which was set
to 100%. B. Time course of tufA transcription rates in wild-type (WT) and act1 cells. Pulse-labeled RNA (10 g) from each time point was
hybridized to duplicate dots (2 g) of the tufA gene. Each point represents an average of duplicates, which varied less than 15%. The data were
normalized relative to the highest value, which was set to 100%.

during the 4 h of cycloheximide treatment. The inhi-
bition started almost immediately after the drug was
added, and for most of the genes, transcription had
decreased to ca. 20% of its initial value by 23 h.
The psbA gene, in contrast, showed a weaker response
to the drug; inhibition was never more than 40%
throughout the time course. It should also be noted
that transcription was still present in all cases, even
after 4 h of drug treatment. In conclusion, these data
indicate that the decline in tufA mRNA induced by
cycloheximide is due to reduced transcription. They
also show that this inhibition is not unique to tufA,
but probably affects most chloroplast-encoded genes,
although not all to the same degree.
Effects of cycloheximide and anisomycin on the
cycloheximide-resistant mutant act1
Drugs can often have unexpected secondary effects;
hence, we wanted to look closer at the effect of
cycloheximide on chloroplast transcription by per-
forming two types of experiments. In the first, a
cycloheximide-resistant mutant, act1 (Sager, 1972),
was used. Cytoplasmic ribosomes from this mutant
are substantially resistant to cycloheximide (Fleming
et al., 1987), although the precise nature of the muta-
tion is unknown. Act1 cells were treated with 10 g/ml
cycloheximide and the expression of tufA analyzed
over a 4-h period. Figure 3A shows that the level of
tufA mRNA in cycloheximide-treated act1 decreased
transiently, but had mostly recovered by 4 h (ca. 90%
of the untreated level). In contrast, the level of tufA
704

mRNA was reduced by ca. 70% in similarly treated

wild-type cells. Run-on transcription analysis of the
tufA gene gave a similar pattern (Figure 3B), verify-
ing that the maintenance of the tufA mRNA levels in
act1 was due to transcriptional resistance. In addition
to tufA, transcription of the other genes used in Fig-
ure 2 were similarly resistant to cycloheximide in act1
(data not shown). These data support the view that the
effect of cycloheximide on chloroplast transcription is
mainly due to its ability to inhibit translation on cy-
toplasmic ribosomes. The transient, partial inhibition
by cycloheximide in act1 may be explained by the fact
that such strains (and wild-type) are more sensitive to
the drug in liquid culture (Fleming et al., 1987) than
on solid media (where the maximum level of antibiotic
resistance was shown to be 20 g/ml; Fleming et al.,
1987).
The other type of experiment that was performed
was to see if a structurally unrelated cytoplasmic
translation inhibitor, anisomycin (Chua and Gilham,
1977), would give the same result as cycloheximide.
Figure 3A shows that anisomycin produced a strong
decline in the tufA mRNA level in both wild type and
act1, and with decay kinetics identical to that induced Figure 4. Effect of cycloheximide on tufA mRNA levels in syn-
by cycloheximide. Figure 3A also shows that tufA chronized cells at peak (ca. L2) versus trough periods (ca. L10) of
mRNA is sensitive to anisomycin in the act1 mutant, tufA mRNA accumulation. Cells were grown for three L-D cycles,
as expected. aliquots were removed at L2 and L10 of the fourth cycle, and treated
with cycloheximide under ongoing growth conditions. Total RNA
was extracted at the indicated times from untreated (Control) and
Effects of inhibiting cytoplasmic translation under drug-treated cells (+CHX), and used for northern blot analysis with
diurnal and circadian conditions the tufA probe. The data were normalized relative to control cells at
L2 and L10, which were set to 100%. L and D refer to light and dark
The preceding data indicated that chloroplast tran- periods, respectively.
scription is tightly coupled to cytoplasmic translation
in C. reinhardtii. We hypothesized that the principal Similarly, transcription of tufA was inhibited
purpose for this interdependency would be regulatory, strongly by cycloheximide added at L2, but only
and sought evidence to support this contention. When weakly when added at L10 (Figure 5). Figure 5
wild-type C. reinhardtii is grown under 12 h light/12 h also shows that the transcription of several other
dark cycles, tufA transcription rates and mRNA levels chloroplast-encoded genes was strongly inhibited by
fluctuate daily, with the peak occurring in the early cycloheximide added at L2, but not at L10. As seen
light period (L1L4), and the trough in the late-light before with asynchronous cells (see Figure 2), there
to early-dark phase (L10D2) (Hwang et al., 1996). was less inhibition of psbA transcription than with
If the expression peak was under tight nuclear con- the other genes. Finally, it is important to note that
trol, then we might expect to see a difference in the the transcription rates at the different time points can
response to cycloheximide during peak versus trough be compared directly, since they were obtained from
periods. To test this, cycloheximide was added to the same culture. It is significant that the transcrip-
synchronized cells at either L2 or L10, and tufA tran- tion rates for most of the trough period (L10D2) are
scription rates and mRNA levels followed for 4 h. similar to the transcription rates obtained during the
Figure 4 shows that the tufA mRNA level declined peak (L2L6) after 30 min of cycloheximide treat-
substantially (ca. 80%) when the drug was added at ment. In other words, cycloheximide drops the peak
L2, but that there was only a small effect (1020%) transcription rates down to the trough or basal rates.
when the drug was added at L10.
 705

Figure 5. Effect of cycloheximide on the transcription of tufA and other chloroplast-encoded genes in synchronized cells at peak (ca. L2) versus
trough (ca. L10) periods. A culture was entrained by growth under three L-D cycles (12 h/12 h), then aliquots were removed at L2 and L10 of
the fourth cycle, and treated with cycloheximide under ongoing growth conditions. Cells were harvested at the indicated times from untreated
(Control) and drug-treated (+CHX) cultures, and used for run-on transcription analysis as in Figure 2. The data for each gene were normalized
relative to the transcription rate at L2, which was set to 100%. It should be noted that the relative transcription rates at different times for a
given gene can be compared directly. L and D refer to light and dark periods respectively.
706

We also performed a synchronization-shift exper-

iment to verify that the tufA mRNA peak is sensitive
to cycloheximide under continuous (or circadian) con-
ditions. Cells were synchronized by growth for three
light-dark cycles, and then shifted to continuous light
(LL) at the end of the third dark period. After 24 h in
LL (i.e. L0 of Day 2 in LL), and immediately prior
to the second scheduled peak of tufA mRNA, cyclo-
heximide was added to the culture and tufA mRNA
levels followed. Figure 6 shows that cycloheximide
blocked the scheduled increase in tufA mRNA, but
did not affect the pre-existing level of tufA mRNA.
Thus, cytoplasmic protein synthesis is required for the
induction of tufA mRNA under circadian conditions.
Since we have previously shown that the stability of
tufA mRNA does not fluctuate under these conditions
(Hwang et al., 1996), these data also support the
view that cytoplasmic protein synthesis is required for
circadian clock control of tufA transcription.

Discussion

These results demonstrate that ongoing cytoplas-

mic protein synthesis is required for the circadian
peaks of transcription of several, and probably most,
chloroplast-encoded genes. This conclusion relies
mainly on the differential effect of cycloheximide, a
potent inhibitor of cytoplasmic translation in C. rein-
hardtii, on transcription of chloroplast-encoded genes.
Although the use of drugs always carries the pos-
sibility of unknown side effects, there are several
observations that support the vailidity of this approach.
First, translation in C. reinhardtii is quite sensitive to
cycloheximide, allowing relatively low concentrations
of the drug (10 g/ml) to be used. Second, since the
response was fairly rapid, the drug treatments could
be kept to a reasonably short time frame (4 h).
Third, the inhibition of tufA (and other chloroplast-
encoded gene) transcription by cycloheximide was
substantially reduced in the cycloheximide-resistant
mutant, act1. Fourth, a structurally unrelated transla-
tion inhibitor, anisomycin, induced a decline in tufA
transcription with kinetics similar to cycloheximide.
Fifth, and possibly most compelling, there was little
or no effect of cycloheximide on transcription of these
genes during the trough period of the circadian cycle
(i.e., late-light to early-dark period). The inevitable
implication of these results is that one or more nuclear-
encoded factors are responsible for the daily peaks of
chloroplast transcription, thereby obviating the need
Figure 6. Effect of cycloheximide on the circadian peak of tufA
mRNA. Wild-type cells were entrained by three L-D cycles, and
then shifted to continuous light (LL) for 2 days. At L0 (subjective)
of the second 24 h in LL, an aliquot was removed and treated with
cycloheximide under ongoing growth conditions. Total RNA was
extracted at the indicated times from the drug-treated (+CHX) and
untreated (Control) cells, and subjected to northern blot analysis
(10 g/lane). A. Autoradiograph of the northern blot hybridized
with the tufA gene probe. Time is indicated as hours of the sub-
jective light period. B. Quantification of the northern blot in A. The
tufA mRNA level at subjective L3 in the control cells was used to
normalize the data; it was set to 100%.
for a circadian clock in the organelle, at least for these
rhythms.
These experiments also separate chloroplast tran-
scription into two types, which are distinguishable
by their short-term sensitivity to cytoplasmic trans-
lation inhibitors. We refer to these types as basal
and inducible (i.e. clock-induced in this case); basal
transcription is insensitive to short-term treatments
with cycloheximide, whereas induced transcription is
rapidly inhibited. It should be understood that these
are operational definitions, and there could be condi-
tions (e.g. stress) that change basal transcription rates,
or there could be other factors besides the clock that
modulate the inducible phase of transcription. For ex-
ample, there is evidence of a direct effect of light
on transcription of the psbA and rrn genes (Desh-

pande et al., 1997) that is superimposed on the clock-
controlled rhythm. This is analogous to the regulation
of the nuclear-encoded light-harvesting chlorophyll
a/b-binding protein genes in Chlamydomonas, which
are light- and clock-regulated (Kindle, 1987; Hwang
and Herrin, 1994; Jacobshagen et al., 1996). It would
be interesting to know if light-stimulated transcrip-
tion of psbA and rrn is also sensitive to cytoplasmic
translation inhibitors.
The relative contribution of basal versus induced
transcription is not the same for all genes that we ex-
amined. The psbA gene was inhibited by only 3040%
by cycloheximide, compared to 7080% for the oth-
ers. This is interesting because the circadian rhythm of
psbA transcription showed a lower amplitude than the
other genes (Hwang et al., 1996). The low amplitude
can now be explained by a relatively greater contribu-
tion of the cycloheximide-resistant, basal transcription
to the daily transcription pattern of psbA. In contrast,
transcription of the rrn genes showed a strong cir-
cadian fluctuation, and correspondingly strong cyclo-
heximide sensitivity. This seems somewhat surprising
for RNAs that are needed in such large amounts. The
reason for such tight control over rrn transcription
is not known with certainty, although one possibility
is that the synthesis of these RNAs must be tightly
coordinated with synthesis of the ribosomal proteins
(Schmidt et al., 1983).
This study raises several important questions.
What is the molecular nature of the cytoplasmically
synthesized factor(s) responsible for the circadian
peaks of chloroplast transcription? What is the mole-
cular nature of the factors that determine basal tran-
scription? To what extent do these findings apply to
other systems? The core RNA polymerase responsi-
ble for the two transcription modes may be the same
bacterial-type polymerase, based on the fact that ri-
fampicin inhibits close to 100% of total chloroplast
transcription in asynchronous cells (Sirdivant et al.,
1985), and has been shown to block tufA transcription
in both trough and peak periods in synchronized cells
(Hwang et al., 1996). Thus, an obvious candidate for
the clock-induced factor is a sigma-like subunit of the
RNA polymerase.
In bacteria, sigma factors are necessary for tran-
scription initiation at the correct site, and are central
to regulation (Doi and Wang, 1986). Moreover, at
least one bacterial sigma factor has been shown to
be an unstable protein that is regulated by proteoly-
sis (Takayanagi et al., 1994), and Tsinoremas et al.
(1996) have demonstrated a role for a sigma factor
707
(rpoD2) in circadian oscillations of transcription in
Synechococcus. Sigma-like factors have been identi-
fied in C. reinhardtii (Surzycki, 1969; Tozawa et al.,
1998), although they have not been well characterized
(or cloned). Sigma-like factors from a few vascular
plants and a red alga have been cloned, and found to
consist of a small gene family (Liu and Troxler, 1996;
Tanaka et al., 1996; Isono et al., 1997; Kestermann
et al., 1998; Tozawa et al., 1998). Although the indi-
vidual roles for these proteins have not been resolved,
if the C. reinhardtii sigma factors are also a small gene
family, one could imagine one sigma factor for basal
transcription (that is also very stable), and another,
short-lived sigma factor for clock-induced transcrip-
tion. The cloning of sigma factors from C. reinhardtii
will enable this hypothesis to be tested.
A variation on this idea (which could also explain
these results) would be differential phosphorylation of
a major sigma factor(s). Phosphorylation of sigma-like
factors from mustard chloroplasts has been demon-
strated, and shown to affect their activity (Tiller and
Link, 1993; Baginsky et al., 1997). The application of
this hypothesis would require that an essential compo-
nent of the signaling/phosphorylation system be: (1)
synthesized in the cytoplasm, (2) relatively unstable,
and (3) controlled by the circadian clock.
Another possible means of controlling large-scale
changes in chloroplast transcription could involve
modulation of DNA topology. Changes in DNA topol-
ogy have been shown to affect chloroplast transcrip-
tion in vivo and in vitro (Stirdivant et al., 1985; Lam
and Chua, 1987; Thompson and Mosig, 1987). In
C. reinhardtii, changes in chloroplast DNA topology
have been shown to correlate with circadian changes
in transcription rates (Salvador et al., 1998). Inter-
estingly, light-induced alterations in the topology of
certain regions of the chloroplast genome in C rein-
hardtii were found to be blocked by cycloheximide
(Thompson and Mosig, 1990). Thus, it seems possible
that the nuclear-encoded factor(s) that we have pro-
posed as a global transcription regulator could work
indirectly by modulating DNA topology.
Results such as these, which implicate such a
short-lived factor in the control of chloroplast tran-
scription in C. reinhardtii, have not been reported
previously. Studies of light-activated chloroplast tran-
scription in dark-grown barley have shown that cy-
cloheximide blocks the photo-induction of transcrip-
tion of several chloroplast genes (Gamble and Mullet,
1989; Klein, 1991). Klein (1991) also found that
cycloheximide, added after 5 h of light, led to a de-
708
crease of total chloroplast transcription activity of 60%
within 4 h. Those data are reminiscent of the results
presented here with C. reinhardtii, and suggest that
a labile regulatory factor may be a conserved feature
of chloroplast transcriptional regulation. It should be
mentioned that Kapoor et al. (1997) have also re-
ported an inhibitory effect of cycloheximide on atpB-E
transcripts in tobacco, but this study only evaluated
steady-state RNA levels, making it difficult to know
when transcription was turned off. It will be interesting
to learn if the recently described circadian control of
psbD transcription in wheat (Nakahira et al., 1998) is
also mediated by a labile, cytoplasmically synthesized
factor.
Acknowledgements
This research was supported by grants from the USDA
(NRICGP 96-35301-3420 and 1999-01512) and the
R.A. Welch Foundation (F-1164).
References
Allison, L.A., Simon, L.D. and Maliga, P. 1995. Deletion of rpoB
reveals a second distinct transcription system in plastids of higher
plants. EMBO J. 11: 28022809.
Baginsky, S., Tiller, K. and Link, G. 1997. Transcription factor
phosphorylation by a protein kinase associated with chloroplast
RNA polymerase from mustard (Sinapis alba). Plant Mol. Biol.
34: 181189.
Chua, N.H. and Gillham, N.W. 1977. The sites of synthesis of the
principal thylakoid membrane polypeptides in Chlamydomonas
reinhardtii. J. Cell Biol. 74: 441452.
Deshpande, N.N., Bao, Y. and Herrin, D.L. 1997. Evidence for
light/redox-regulated splicing of psbA pre-RNAs in Chlamy-
domonas chloroplasts. RNA 3: 3748.
Doi, R.H. and Wang, L.-F. 1986. Multiple prokaryotic ribonucleic
acid polymerase sigma factors. Microbiol. Rev. 50: 227243.
Fleming, G.H., Boynton, J.E. and Gillham, N.W. 1987. The
cytoplasmic ribosomes of Chlamydomonas reinhardtii: charac-
terization of antibiotic sensitivity and cycloheximide-resistant
mutants. Mol. Gen. Genet. 210: 419428.
Fong, S. and Surzycki, S. 1992. Chloroplast RNA polymerase genes
of Chlamydomonas reinhardtii exhibit an unusual structure and
arrangement. Curr. Genet. 21: 485497.
Gamble, P. and Mullet, J.E. 1989. Blue light regulates the accumula-
tion of two psbD-psbC transcripts in barley chloroplasts. EMBO
J. 8: 27852794.
Greenberg, B.M., Narita, J.O., DeLuca-Flaherty, C., Gruissem, W.,
Rushlow, K.A. and Hallick, R.B. 1984. Evidence for two RNA
polymerase activities in Euglena gracilis chloroplasts. J. Biol.
Chem. 259: 1488014887.
Goldschmidt-Clermont, G. 1998. Coordination of nuclear and
chloroplast gene expression in plant cells. Int. Rev. Cytol. 177:
115180.
Gruissem, W. and Tonkyn, J.C. 1993. Control mechanisms of plastid
gene expression. Crit. Rev. Plant Sci. 12: 1957.
Guertin, M. and Bellemare, G. 1979. Synthesis of chloroplast ri-
bonucleic acid in Chlamydomonas reinhardtii toluene-treated
cells. Eur. J. Biochem. 96: 125129.
Harris, E. 1989. The Chlamydomonas Sourcebook. Academic Press,
San Diego, CA.
Hedtke, B., Brner, T. and Weihe, A. 1997. Mitochondrial
and chloroplast phage-type RNA polymerases in Arabidopsis.
Science277: 809811.
Herrin, D.L. and G.W. Schmidt. 1987. Chloroplast gene expression
in chloroplast ribosome-deficient mutants of Chlamydomonas
reinhardtii. In: J. Biggins (Ed.) Progress in Photosynthesis
Research, Vol. IV, Martinus Nijhoff, Dordrecht, Netherlands,
pp. 645648.
Herrin, D.L. and Schmidt, G.W. 1988. Rapid, reversible staining of
northern blots prior to hybridization. Biotechniques 6: 196200.
Herrin, D.L., Michaels, A.S. and Paul, A.-L. 1986. Regulation of
genes encoding the large subunit of ribulose-1,5-bisphosphate
carboxylase and the photosystem II polypeptides D-1 and D-2
during the cell cycle of Chlamydomonas reinhardtii. J. Cell Biol.
103: 18371845.
Hoober, J.K. 1970. Sites of synthesis of chloroplast membrane
polypeptides in Chlamydomonas reinhardii y-1. J. Biol. Chem.
245: 43274334.
Hwang, S. and Herrin, D.L. 1994. Circadian regulation of lhc gene
transcription in Chlamydomonas reinhardtii. Plant Mol. Biol. 26:
557569.
Hwang, S., Kawazoe, R. and Herrin, D.L. 1996. Transcription of
tufA and other chloroplast-encoded genes is regulated by a circa-
dian clock in Chlamydomonas. Proc. Natl. Acad. Sci. USA 93:
9961000.
Isono, K., Shimizu, M., Yoshimoto, K., Niwa, Y., Satoh, K., Yokota,
A. and Kobayashi, H. 1997. Leaf-specifically expressed genes
for polypeptides destined for chloroplasts with domains of 70
factors of bacterial RNA polymerases in Arabidopsis thaliana.
Proc. Natl. Acad. Sci. USA 94: 1498414953.
Jacobshagen, S., Kindle, K.L. and Johnson, C.H. 1996. Tran-
scription of CABII is regulated by the biological clock in
Chlamydomonas reinhardtii. Plant Mol. Biol. 31: 11731184.
Jahn, D. 1992. Expression of the Chlamydomonas reinhardtii
chloroplast tRNAglu gene in a homologous in vitro transcrip-
tion system is independent of upstream promoter elements. Arch.
Biochem. Biophys. 298: 505513.
Johnson, C.H., Golden, S.S., Ishiura, M. and Kondo, T. 1996.
Circadian clocks in prokaryotes. Mol. Microbiol. 21: 511.
Kapoor, S., Suzuki, J.Y. and Sugiura, M. 1997. Identification and
functional significance of a new class of non-consensus-type
plastid promoters. Plant J. 11: 327337.
Kestermann, M., Neukirchen, S., Kloppstech, K. and Link, G. 1998.
Sequence and expression characteristics of a nuclear-encoded
chloroplast sigma factor from mustard (Sinapis alba). Nucl.
Acids Res. 26: 27472753.
Kindle, K.L. 1987. Expression of a gene for a light-harvesting
chlorophyll a/b-binding protein in Chlamydomonas reinhardtii:
effect of light and acetate. Plant Mol. Biol. 9: 547563.
Klein R.R. 1991. Regulation of light-induced chloroplast tran-
scription and translation in eight-day-old dark-grown barley
seedlings. Plant Physiol 97: 335342.
Lam, E. and Chua, N.-H. 1987. Chloroplast DNA gyrase and in vitro
regulation of transcription by template topology and novobiocin.
Plant Mol. Biol. 8: 415424.
Lerbs-Mache, S. 1993. The 110-kDa polypeptide of spinach plas-
tid DNA-dependent RNA polymerase: single-subunit enzyme

or catalytic core of multimeric enzyme complexes. Proc. Natl.
Acad. Sci. USA 90: 55095513.
Leu, S., White, D. and Michaels, A. 1990. Cell cycle-dependent
transcriptional and post-transcriptional regulation of chloroplast
gene expression in Chlamydomonas reinhardtii. Biochim. Bio-
phys. Acta 1049: 311317.
Liu, B. and Troxler, R.F. 1996. Molecular characterization of a
positively photoregulated nuclear gene for a chloroplast RNA
polymerase factor in Cyanidium caldarium. Proc. Natl. Acad.
Sci. USA 93: 33133318.
Mayfield, S.P., Yohn, C.B., Cohen, A. and Danon, A. 1995. Regu-
lation of chloroplast gene expression. Annu. Rev. Plant Physiol.
Plant Mol. Biol. 46: 147166.
Mullet, J.E. 1993. Dynamic regulation of chloroplast transcription.
Plant Physiol. 103: 309313.
Nakahira, Y., Baba, K., Yoneda, A., Shiina, T. and Toyoshima,
Y. 1998. Circadian-regulated transcription of the psbD light-
responsive promoter in wheat chloroplasts. Plant Physiol. 118:
10791088.
Pfannschmidt, T. and Link, G. 1994. Separation of two classes of
plastid DNA-dependent RNA polymerases that are differentially
expressed in mustard (Sinapis alba L.) seedlings. Plant Mol.
Biol. 25: 6981.
Sager, R. 1972. Cytoplasmic Genes and Organelles. Academic
Press, New York.
Salvador, M.L., Klein, U. and Bogorad, L. 1993. Light-regulated
and endogenous fluctuations of chloroplast transcript levels in
Chlamydomonas. Regulation by transcription and RNA degrada-
tion. Plant J. 3: 213219.
Salvador, M.L., Klein, U. and Bogorad, L. 1998. Endogenous fluc-
tuations of DNA topology in the chloroplast of Chlamydomonas
reinhardtii. Mol. Cell Biol. 18: 72357242.
Sambrook, J., Fritsch, E.F. and Maniatis, T. 1989. Molecular
Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor
Laboratory Press, Plainview, NY.
Schmidt, R.J., Richardson, C.B., Gillham, N.W. and Boynton, J.E.
1983. Sites of synthesis of chloroplast ribosomal proteins in
Chlamydomonas. J. Cell Biol. 96: 14511463.
Schrubar, H., Wanner, G. and Westhoff, P. 1990. Transcriptional
control of plastid gene expression in greening sorghum seedlings.
Planta 183: 101111.
709
Sexton, T.B., Christopher, D.A. and Mullet, J.E. 1990. Light-
induced switch in barley psbD-psbC promoter utilization: a novel
mechanism regulating chloroplast gene expression. EMBO J. 9:
44854494.
Stirdivant, S.M., Crossland, L.D. and Bogorad, L. 1985. DNA su-
percoiling affects in vitro transcription of two maize chloroplast
genes differently. Proc. Natl. Acad. Sci. USA 82: 48864890.
Surzycki, S. 1969. Genetic functions of the chloroplast of Chlamy-
domonas reinhardi: effect of rifampin on chloroplast DNA-
dependent RNA polymerase. Proc. Natl. Acad. Sci. USA 63:
13271334.
Surzycki, S.J. and Shellenbarger, D.L. 1976. Purification and char-
acterization of a putative sigma factor from Chlamydomonas
reinhardi. Proc. Natl. Acad. Sci. USA 73: 39613965.
Takayanagi, Y., Tanaka, K. and Takahashi, H. 1994. Structure of
the 50 upstream region and the regulation of the rpoS gene of
Escherichia coli. Mol. Gen. Genet. 243: 525531.
Tanaka, K., Oikawa, K., Ohta, N., Kuroiwa, H., Kuriowa, T. and
Takahashi, H. 1996. Nuclear encoding of a chloroplast RNA
polymerase sigma subunit in a red alga. Science 272: 19321935.
Thompson, R.J. and Mosig, G. 1987. Stimulation of a chloroplast
promoter by novobiocin in situ and in E. coli implies regulation
by torsional stress in the chloroplast DNA. Cell 48: 281287.
Thompson, R.J. and Mosig G. 1990. Light affects the structure of
Chlamydomonas chloroplast chromosomes. Nucl. Acids Res. 18:
26252631.
Tiller, K. and Link, G. 1993. Phosphorylation and dephosphoryla-
tion affect functional characteristics of chloroplast and etioplast
transcription systems from mustard (Sinapis alba L.). EMBO J.
12: 17451753.
Tozawa, Y., Tanaka, H. and Wakasa, K. 1998. Nuclear encoding
of a plastid factor in rice and its tissue- and light-dependent
expression. Nucl. Acids Res. 26: 415419.
Troxler, R.F., Zhang, Z.F., Hu, J. and Bogorad L. 1994. Evidence
that factors are components of chloroplast RNA polymerase.
Plant Physiol. 104: 753759.
Tsinoremas, N.F., Ishiura, M., Kondo, T., Andersson, C.R., Tanaka,
K., Takahashi, H., Johnson, C.H. and Golden S.S. 1996. A sigma
factor that modifies the circadian expression of a subset of genes
in cyanobacteria. EMBO J. 15: 24882495.