ABSTRACT
The effects of changing growth rates on the levels of
40S
pre-rRNA and two r-protein mRNAs were examined to gain insight into the coordinate
transcriptional regulation of ribosomal genes in the ascomycete fungus
Neurospora crassa
. Growth rates were varied either by altering carbon nutritional conditions, or
by subjecting the isolates to inositol-limiting conditions. During carbon up- or down-shifts, r-protein mRNA levels were stoichiometrically coordinated.
Changes in
40S
pre-rRNA levels paralleled those of the r-protein mRNAs but in a non-stoichiometric manner. Comparison of
crp-2
mRNA levels with those of a
crp-2::qa-2
fusion gene indicated no major effect from changes in
crp-2
mRNA stability.
Crp-2
promoter mutagenesis experiments revealed that two elements of the
crp-2
promoter, -95 to -83 bp (Dde box) and -74 to -66 bp (CG repeat) important for transcription under
constant growth conditions, are also critical for transcriptional regulation by
a carbon source. Ribosomal protein mRNA and rRNA levels were unaffected by
changes in growth rates when the cultures were grown under inositol-limiting conditions, suggesting that, under these conditions,
transcription of the ribosomal genes in
N.crassa
was regulated independently of growth rate.
Ribosomes are essential for viability and their synthesis requires a major
metabolic effort on the part of the cell (
1
,
2
). Consequently, it is presumably advantageous for the cell to closely regulate
the synthesis of their components. Regulation of ribosome synthesis typically
occurs at two levels: (i) control of the number of ribosomes in response to
varying cellular demands, and (ii) regulation of the synthesis of the ribosomal
components to ensure that each component is present in sufficient and equimolar
amounts.
In
Escherichia coli
, changing growth rates due to changing environmental or nutritional conditions
results in a rapid adjustment in the number of ribosomes and is paralleled by
changes in rRNA and r-protein levels (
3
-
5
). The regulation of ribosomal gene expression appears to occur primarily at the
translational level through an autogenous feedback mechanism (
6
).
In contrast, in
Saccharomyces cerevisiae
, transcriptional control plays a major role in the regulation of ribosomal gene
expression during changes in carbon (
7
-
9
) or nitrogen conditions (
10
), or during amino acid starvation (
11
). During carbon shifts, changes in r-protein mRNA levels in yeast are stoichiometrically coordinated with
changing 35S rRNA levels (
7
,
12
). Expression of most yeast r-protein genes is coordinated at the transcriptional level by virtue of a
conserved upstream activator sequence (UAS
rpg
) found in the promoters of most genes (
7
,
8
,
13
,
14
). A repressor-activator protein (RAP1), (
15
) also called TUF (
16
) or GRF (
17
), binds the UASrpg sequence and plays a role in coordinating transcription of r-protein genes (
13
,
18
). A 45 bp sequence at the 3'-end of the rRNA enhancer region is known to be both necessary and
sufficient for transcriptional activation during a carbon upshift from ethanol
to glucose (
9
). The regulatory factor(s) which binds to this region have not yet been
identified, and consequently its role, if any, in coordinating transcription by
RNA polymerases I and II is unknown. Although regulation of r-protein gene expression in yeast occurs primarily at the transcriptional
level, post-transcriptional control mechanisms such as mRNA splicing (
19
) and protein degradation (
20
,
21
) are also operative in some cases.
In higher eukaryotes, the regulation of ribosomal gene expression is not as well
understood as in yeast. In mouse, the promoter elements required for
transcription of three r-protein genes (
rpL30, rpL32
and
rpL16
) have been identified, but nothing is known about their possible role in
coordinating the transcription (
22
-
24
). In other organisms such as
Xenopus
(
25
) or
Drosophila
(
26
), translational control plays a major role in regulating r-protein expression during embryogenesis or differentiation.
In
N.crassa
, rRNA and r-protein levels change rapidly in response to changing carbon nutrition
sources (
27
,
28
). In contrast, tRNA, DNA and total cellular protein levels do not change until
rRNA and r-protein levels have stabilized at levels characteristic of the new media (
1
). More recent work has demonstrated that the mRNA levels of four r-proteins decrease rapidly and coordinately to 2-3% of preshift levels following a carbon downshift from sucrose to
quinic acid (
29
). Together, these results suggest that the expression of r-protein genes in
N.crassa
may be regulated at the transcriptional level. They also emphasize the
particular sensitivity of ribosomal gene expression to growth conditions.
In
N.crassa,
expression of unlinked structural genes involved in complex metabolic processes
are often coordinated at the transcriptional level through the presence of
conserved promoter elements (
30
-
34
). The promoters of six cytoplasmic ribosomal protein (
crp
) genes and a translation elongation factor (
tef-1
) (
35
) in
N.crassa
have been sequenced to date (
29
,
36
-
39
; I. de la Serna and B. Tyler, unpublished). Several conserved elements are
present in some or all of the promoters of these genes, as well as in the
transcriptional regulatory regions of the 5S and
40S
rRNA genes. These include the CG repeat (
29
), the Taq box (
29
), the Ribo box element (
29
,
40
) and a Dde box sequence (Fig.
6
). The Ribo box is required for transcription of the 5S rRNA genes
in vitro
(
40
) and
in vivo
(
41
). The Ribo box and Dde box are required for transcription from the
40S
promoter
in vitro
(
42
), but have not been tested
in vivo
. Conserved elements within the ribosomal gene promoters have the potential to
play roles in coordinating transcription by the three classes of RNA
polymerases.
Mutational analysis of the
crp-2
promoter has identified six sequences important for optimal transcription
in vivo
during exponential growth under constant nutritional conditions (Cujec and
Tyler, submitted to
Mol. Gen. Genet
.). Optimal transcriptional efficiency is dependent upon unidentified elements
between -245 and -189 bp and between -48 and +9 bp, as well as on sequences from -153 to -147 bp (includes most of the -152 Dde box) and on sequences from -95 to -83 bp (includes all of
the -97 Dde box sequences) (Fig.
1
A). In addition, the -74 CG repeat is also critical for optimal transcription and is redundant
to an upstream element (probably another CG repeat) between -189 and -154 bp.
DNA restrictions and ligations were done as described (
43
). DNA probes (100 ng) were 5'-end-labeled using 50 [mu]Ci (Amersham) [[gamma]-
32
P]ATP and T
4
polynucleotide kinase. DNA probes were 3'-end-labeled by incubation with the Klenow fragment (4 U) in the
absence of dNTPs for 10 min before adding [[alpha]-
32
P]dCTP (50 [mu]Ci, Amersham) and the remaining dNTPs (0.2 mM).
Construction of the
crp-2
promoter mutations (Fig.
1
A and B) used in this study and their targeting to the
qa-2
locus were as described (
41
; Cujec and Tyler, submitted). Transformant 246(pQaRp4-WT) contains 1.45 kbp of
crp-2
promoter and leader sequences fused to the ATG codon of 355 bp of 5'
qa-2
(catabolic dehydroquinase) coding sequences. Transformants, 246(pQaRp4[Delta]-95), 246(pQaRp4[Delta]-82) and 246(pQaRp4[Delta]-65) contain 5' deletion mutations in the
crp-2
promoter sequences of
crp-2::qa-2
.
Crp-2::qa-2
mRNA levels in 246(pQaRp4[Delta]-95), 246(pQaRp4[Delta]-82) and 246(pQaRp4[Delta]-65) were 19%, 4% and 3% of wild-type levels [in 246(pQaRp4-WT)] respectively, during
steady-state growth in sucrose media (Cujec and Tyler, submitted to
Mol. Gen. Genet
.). Transformants 246(pQaRp4-2RB) and 246(pQaRp4-3CG) contain substitution mutations in either the conserved Ribo box
or CG repeat sequences of the
crp-2::qa-2
fusion. The double Ribo box mutation reduced transcription to 69% of wild-type levels, while the triple CG repeat increased transcription slightly
(136%) during growth in sucrose (
41
). Transformants 246(pQaRp4[Delta]-95-89RB) and 246(pQaRp4[Delta]-95-74CG) contain substitution mutations in
either the -89 Ribo box or the -74 CG repeat element in the context of a 5' deletion to -95 bp. These mutations decreased transcription to 17% [246(pQaRp4[Delta]- 95-89RB)], or to 2% [246(pQaRp4[Delta]-95-74CG)] of wild-type
levels (
41
) during exponential growth in sucrose media.
As appropriate, 4 * 10
7
conidia or 5-10 g wet mycelia were inoculated into 2 l flasks containing 400 ml
minimal Vogel's liquid media (
44
) and shaken (300 r.p.m.) at 25oC in the dark. For the carbon shift experiments, the media were
supplemented with inositol (50 [mu]g/ml) and either sucrose (1.5%) or glycerol (1.0% w/v) as carbon source. In
the inositol depletion experiments, minimal Vogel's media contained sucrose
(1.5%) and either 0.7, 5.0 or 50 [mu]g/ml of filter-sterilized inositol.
In the carbon downshift experiments, conidia were inoculated into sucrose media
and grown to exponential phase (16 h). The mycelia were harvested by filtration
through miracloth (Calbiochem), rinsed with several volumes of distilled water,
and divided into three aliquots. One aliquot (2-3 g) was returned to sucrose media for 2 h, another aliquot was
transferred to glycerol media for 2 h, while the remaining aliquot was grown in
glycerol for 8 h. Harvested mycelia was frozen in liquid N
2
immediately and stored at -80oC. In the carbon upshift experiments, conidia were initially
inoculated into sucrose media to permit spore germination and mycelial
outgrowth. After 16 h, the mycelia were harvested, washed with 2-3 vol water, and transferred to glycerol. After 16 h in glycerol, the
mycelia were harvested again and divided into three aliquots. One aliquot was
returned to glycerol for 2 h, another was transferred to sucrose for 2 h, and
the remaining mycelia were grown in sucrose for 8 h.
In the inositol deprivation experiments, conidia from the appropriate
transformants were grown in minimal media supplemented with inositol (5 [mu]g/ml) for 20 h to permit mycelial growth. The mycelia (40 g wet
weight/flask) were then harvested and 10 g wet mycelia were inoculated into
culture media (400 ml) containing either 0.7 or 50 [mu]g/ml inositol. In all cases, the inositol was filter sterilized before
addition to the autoclaved media. At 6 h intervals for the next 12 h, half the
media (including mycelia) was removed and 200 ml fresh media containing the
appropriate inositol concentration was added to each flask. After the final
media replenishment (t = 12 h), 20 ml of media was removed and the wet and dry
weights of the collected mycelia determined. In order to determine the growth
rates at different inositol concentrations, additional 20 ml aliquots were
obtained at t = 15, 16.5, 18, 19.5 and 21 h. At 16.5 and 21 h, 150 ml media was
removed from the flasks and the mycelia stored for subsequent RNA extraction.
Total RNA was extracted from frozen mycelia and poly(A
+
) RNA purified using a oligo-dT-cellulose column essentially as described by Patel
et al.
(
45
). The S1 nuclease hybridization assay was used to quantify the mRNAs of
interest (
46
). The DNA probes used were a 567 bp
Bam
HI-
Sty
I (3'-end-labeled) fragment for the [beta]-tubulin gene (
47
), a 580 bp
Bst
BI-
Eco
NI (5'-end-labeled) fragment for
crp-1
(
36
), a 213 bp
Bst
YI-
Eco
RI (5'-end-labeled) fragment for
crp-2
(
40
) and a 112 bp
Xho
I-
Rsa
I (5'-end-labeled) fragment for the
40S
pre-rRNA (
42
). The
qa-2
(
30
) probe was obtained by cloning 336 bp of
qa-2
sequences (
Ava
I-
Sph
I) present on a
Ava
I-
Sal
I fragment into pUC18 (
Ava
I-
Sal
I) and then digesting with
Ava
I and
Nde
I (680 bp, 3'-end-labeled). The
40S
probe spans the transcription initiation site and hybridizes to sequences in
the external transcribed spacer, upstream of an efficient RNA processing site.
These sequences are rapidly degraded during processing of the
40S
pre-RNA transcript (
42
). Consequently, the amount of
40S
pre-rRNA hybridizing to the DNA probe more accurately reflects rRNA
transcription rates than the level of the stable mature rRNAs. All probes were
gel purified prior to hybridization. The [beta]-tubulin,
qa-2
,
crp-1
and
crp-2
probes were hybridized simultaneously to 5 [mu]g poly(A
+
) RNA for 16 h at 58oC, while the
40S
pre-rRNA, [beta]-tubulin,
crp-1
and
crp-2
probes were hybridized to 100 [mu]g of total RNA overnight at 56oC. After digestion with 50 U of S1 nuclease (Boehringer Mannheim), the
products were separated on a 5% polyacrylamide-7 M urea gel. Protected DNA fragments were visualized by autoradiography
and bands quantified either by the AMBIS radioanalytical system (San Diego, CA)
or by phosphoimager analysis (Sunnyvale, CA).
The
crp-2
protein in
N.crassa
is homologous to S11 in
E.coli
(
48
) and rp59 (CRY1) in
S.cerevisiae
(
2
) and is assumed to be essential for viability. In order to assay different
crp-2
promoter mutations, and to judge effects on mRNA stability, the promoter of the
qa-2
gene (catabolic dehydroquinase) was replaced with wild-type or mutant
crp-2
promoters, using gene targeting (Cujec and Tyler, submitted to
Mol. Gen. Genet
.). A transformant containing wild-type
crp-2
promoter sequences integrated at the
qa-2
locus, 246(pQaRp4-WT), was grown first in sucrose containing medium, then transferred to glycerol containing medium (carbon
downshifts). The S1 nuclease assay was used to simultaneously quantify
crp-1,
crp-2
,
crp-2::qa-2
and [beta]-tubulin mRNA levels at various times during the shifts. In some
experiments, a DNA probe hybridizing to processed regions of the
40S
pre-rRNA was also used in order to determine the effects of changing carbon
nutritional sources on
40S
transcription levels.
Crp-2::qa-2
,
crp-1
,
crp-2
and
40S
pre-RNA levels were standardized to those of [beta]-tubulin and are expressed as a fraction of transcript levels
at 2 h sucrose (Figs
2
and
3
).
The mRNA levels of the
crp-1
,
crp-2
and
crp-2::qa-2
genes in 246(pQaRp4-WT) were closely coordinated during carbon upshifts (Figs
2
and
3
B). Two hours after transfer to sucrose to glycerol, mRNA levels were 1.5-2.0-fold higher than those of cultures in glycerol. After 8 h in
sucrose, the
crp-1
,
crp-2
and
crp-2::qa-2
mRNA levels decreased to levels 15-50% higher than those prior to the shift. In contrast,
40S
pre-rRNA levels were 7-fold greater than preshift levels after 2 h in sucrose and remained
5-fold greater after 8 h in sucrose (Fig.
3
B). These results demonstrate that r-protein mRNA levels change coordinately during changes in carbon sources.
Under these conditions,
40S
pre-rRNA levels are also coordinated with r-protein mRNA levels, however, the coordination is not
stoichiometric.
In order to identify the carbon-responsive element(s) in the
crp-2
promoter, transformants containing 5' deletions in the
crp-2
sequences integrated at the
qa-2
locus were subjected to carbon shifts. Initial experiments demonstrated that a
promoter deletion to -153 bp, which reduced transcription in sucrose media to 45% of wild-type, did not affect the transcriptional regulation of the
crp-2::qa-2
gene during either a carbon up or downshift (data not shown). A 5' deletion to -95 bp reduced
crp-2::qa-2
mRNA levels to 20% of wild-type levels during steady growth in sucrose media but did not affect the
transcriptional regulation of the
crp-2::qa-2
gene during carbon shifts (Fig.
4
A). In the downshift experiments,
crp-2::qa-2
mRNA levels in 246(pQaRp4[Delta]-95) dropped to 11% of sucrose levels after 2 h in glycerol and then
returned to 68% of preshift levels after 8 h in glycerol (Fig.
4
A). Following an upshift to sucrose media,
crp-2::qa-2
mRNA levels in 246(pQaRp4[Delta]-95) increased to 212% of pre-shift levels after 2 h in sucrose before decreasing to 123%
of pre-shift levels after 8 h in sucrose (Fig.
4
B). The decrease at 8 h may be due to the cultures entering stationary phase. In
both the up- and down-shifts, changes in
crp-2::qa-2
mRNA levels in 246(pQaRp4[Delta]-95) were similar to those normally observed in the control
transformant 246(pQaRp4-WT).
In an attempt to determine whether the ribosomal genes respond to carbon source
or to growth rate directly, r-protein mRNA and
40S
pre-rRNA levels were compared in cultures growing in sucrose media at
different growth rates. Growth rates were varied by growing the transformants,
which are inositol auxotrophs, in media containing either sufficient (50 [mu]g/ml) or limiting (0.7 [mu]g ml) inositol concentrations.
Based on preliminary experiments, the growth rates of 246(pQaRp4-WT) and 246(pQaRp4[Delta]-95) were compared at regular intervals beginning 12 h after
transfer to media containing either 50 or 0.7 [mu]g/ml inositol. After 16.5 h in media containing 50 [mu]g/ml inositol, the growth rate of 246(pQaRp4-WT) was three times greater (5 mg/h) than in 0.7 [mu]g/ml inositol (1.6 mg/h) while the growth rate of 246(pQaRp4[Delta]-95) was six times greater (11 mg/h versus 1.7
mg/h) (Fig.
5
). After 21 h, both transformants had ceased to grow under inositol-limiting conditions, while their growth rates under sufficient inositol
conditions remained about the same as at 16.5 h.
In
N.crassa
,
40S
pre-rRNA and r-protein mRNA levels change rapidly following changes in carbon
nutritional conditions (
27
-
29
). This fact, and the presence of conserved elements in the transcription
regulatory regions of r-protein and rRNA genes suggest that transcriptional regulation may play an
important role in the expression of ribosomal genes in
N.crassa
(
29
,
40
,
42
). As a first step in gaining a better understanding of how ribosomal genes in
N.crassa
are regulated at the transcriptional level, rRNA and r-protein mRNA levels were compared in cultures having different growth
rates. Growth rates were varied either by shifting cultures from one carbon
source to another, or by subjecting them to inositol-limiting concentrations. In addition, elements required for
transcriptional regulation
in vivo
were identified using mutations in the promoter of the r-protein gene,
crp-2.
This study demonstrated that r-protein mRNA levels in
N.crassa
are stoichiometrically coordinated during a carbon upshift from glycerol to
sucrose, or a downshift from sucrose to glycerol. During the carbon shifts,
40S
pre-rRNA levels changed coordinately with r-protein mRNA levels. However,
40S
pre-rRNA levels appeared more responsive to the quality of the carbon nutrient
source than r-proteins genes. The lack of stoichiometry between r-protein mRNA and rRNA levels may reflect the nature of the assays
(transient pre rRNA versus mRNA), or the possibility that post-transcriptional mechanisms may also be important in balancing the
expression of ribosomal genes.
Transcription of the
crp-2::qa-2
fusion gene in 246(pQaRp4-WT) is entirely dependent upon the
crp-2
promoter during exponential growth in sucrose media (Cujec and Tyler, submitted
to
Mol. Gen. Genet
.). In the carbon up- and down-shift experiments described in this study,
crp-2::qa-2
mRNA levels in 246(pQaRp4-WT) were closely coordinated with those of the endogenous
crp-1
and
crp-2
genes. Thus, the 1.5 kbp of
crp-2
promoter sequences fused to the
qa-2
gene in 246(pQaRp4-WT) are sufficient to confer normal transcriptional regulation on the
qa-2
gene during carbon shifts. This observation also indicates that the observed
changes in
crp-2
mRNA levels are unlikely to be due to differences in mRNA stability. This is
supported by the observation that upstream promoter deletions, which do not
affect the mRNA, abolished regulation.
In order to identify the carbon-responsive element(s) in the
crp-2
promoter,
N.crassa
isolates containing mutant
crp-2
promoter sequences were subjected to carbon shifts and the effects of these
mutations on
qa-2
mRNA levels determined. Since the
crp-2
promoter is strong, it was possible to measure changes in mRNA levels even when
a mutant promoter retained only 3-4% activity.
Crp-2
promoter sequences from -95 to -83 bp and from -73 to -66 bp were identified as being required for
transcriptional regulation during carbon shifts. The region from -95 to -83 bp in the
crp-2
promoter matches the Dde box consensus sequence (-97 Dde box) while the region from -73 to -66 bp matches the -74 CG repeat consensus. Sequences homologous to the
Dde box consensus element are also present in another functional upstream
element in the
crp-2
promoter (-152 Dde box), as well as in the promoters of the six other
N
.
crassa
r-proteins sequenced to date, a translation elongation factor (
tef-1
) and in the 40S rRNA promoter (Fig.
6
) (
29
,
36
,
37
,
39
,
42
; I. de la Serna and B. Tyler, unpublished). Deletion of the -152 Dde box in the
crp-2
promoter resulted in a 2-fold decrease in transcription during growth in sucrose (Cujec and Tyler,
submitted to
Mol. Gen. Genet
.), while a substitution mutation in the Dde box of the 40S rRNA promoter caused
a 7-fold drop in transcription
in vitro
(
42
). Together, these results suggest that
crp-2
promoter sequences required for optimal transcription under constant growth
conditions (Dde box and CG repeat) are also required for transcriptional
regulation during changes in nutritional sources. Consequently, the
transcriptional regulation of the
crp-2
gene in
N
.
crassa
appears analogous to that of numerous r-protein genes in
S
.
cerevisiae
; the UAS
rpg
element found in the promoters of most yeast r-protein genes is required for transcription under optimal growth
conditions as well as following a nutritional downshift (
7
,
8
,
13
,
14
,
18
). Efforts are presently underway to identify the protein(s), which might bind
the -97 Dde box and the -74 CG repeat and to determine how they interact with the basal
transcription machinery to regulate transcription under constant, as well as
changing nutritional conditions.
In
S.cerevisiae
and
N.crassa
, changing carbon nutritional conditions affect the regulation of r-genes as well as genes involved in carbon metabolism (
7
,
12
,
27
,
28
). Different carbon sources also affect the growth rates of both organisms (
1
,
18
). Consequently, it is unclear whether changes in ribosomal gene transcript
levels following a carbon shift are due to changes in growth rate which in turn
alter transcription rates, or whether the signaling pathways which normally
detect changes in carbon status interact directly with the transcription
machinery responsible for r-gene expression. One way to address this question is to vary the growth
rates of
N.crassa
without altering the carbon status of the media. The cofactor, inositol is
involved in membrane structure and has no nutritional value (
49
). Consequently in these experiments, growth rates were varied by adjusting
inositol concentrations in the presence of a constant source of carbon (1.5%
sucrose).
Inositol starvation caused a 3-6-fold reduction in the growth rates of 246(pQaRp4-WT) and 246(pQaRp4[Delta]-95). However, despite the differences in growth
rates, the r-protein mRNA and
40S
pre-rRNA levels were similar. The unresponsiveness of r-protein mRNA and
40S
pre-rRNA levels to different growth rates suggests that transcription is
dependent upon the carbon source in the media and not on the growth rate of the
culture. This is further supported by the observation that sharp drops in r-protein and rRNA expression only occur during a downshift from sucrose to
glycerol, but not from glycerol to no carbon source (R. Ballica, I. de la Serna
and B. Tyler, unpublished). The fact that the difference in
crp-2::qa-2
mRNA levels between transformants 246(pQaRp4- WT) and 246(pQaRp4[Delta]-95) remained similar regardless of inositol concentrations
suggests that the
crp-2
promoter does not have a regulatory element upstream of -95 bp which is specifically required for transcription during slow
growth.
We thank Eva Steinberger for performing some of the oligonucleotide mutagenesis, Felipe Arredondo for oligonucleotides and assistance in screening transformants and Jeff Hall for photography. This work was supported by National Institutes of Health grant R01 GM42178.
+
Present address: Departments of Medicine, Microbiology and Immunology, Howard
Hughes Medical Institute, University of California, San Francisco,
CA 94143-0724, USA
REFERENCES
Return