(Received for publication, July 25, 1996, and in revised form, October 11, 1996)
From the Department of Pharmacology, Duke University Medical Center, Durham, North Carolina 27710
The unique gene for Ca2+/calmodulin-dependent protein kinase (CaMK) has been shown to be essential in Aspergillus nidulans. Disruption of the gene prevents entry of spores into the nuclear division cycle. Here we show that expression of a constitutively active form of CaMK also prevents spores from entering the first S phase in response to a germinating stimulus. Expression of the constitutively active kinase induces premature activation of NIMEcyclin B/NIMXcdc2 in G0/G1. As NIMXcdc2 is present in spores, the elevation of maturation promotion factor activity may be secondary to the early production of NIMEcyclin B or post-translation modification of maturation promotion factor. The expression of the constitutively active CaMK also results in the appearance of NIMA kinase activity within 1 h of the germinating signal. These results support the contention that the activities of maturation promotion factor and NIMA are coincidentally regulated in A. nidulans and suggest that the unscheduled appearance of one or both of these activities may be sufficient to prevent A. nidulans spores from entering into DNA synthesis.
Both extracellular Ca2+ and intracellular calmodulin (CaM)1 are essential for the nuclear division cycle and growth of Aspergillus nidulans (1, 2). Disruption of the unique CaM gene is lethal, and cells arrest with one or two nuclei and with absent or very short germ tubes (2). Analysis of a strain made conditional for the expression of CaM revealed a requirement for this regulatory protein for both G1/S and G2/M transitions (1). Two Ca2+-dependent targets of CaM have also been shown to be essential in A. nidulans, namely the catalytic or A subunit of the protein phosphatase 2B (3), also called calcineurin, and a multifunctional protein kinase (CaMK) (4). Disruption of the CaMK gene arrested spores with a single nucleus and with no germ tube extension, whereas when calcineurin is held low, spores proceed through the first DNA synthesis before arresting and hyphal growth as well as septation occur in the absence of nuclear division. A strain conditional for the expression of CaMK was created, and, whereas the nuclear division cycle and growth were markedly slowed, both proceeded (4). This was found to be due to a low amount of expression of the CaMK gene in repressing medium that resulted in about 10% the normal level of CaMK. Thus, it was difficult to determine the mechanism responsible for the markedly slowed growth.
Insight into the potential role of CaMK in cell cycle progression in
other systems has been obtained from the expression of constitutively
active forms of the enzyme. The transient expression of a
constitutively active CaMK using the mouse metallothionein viral
promoter caused a G2 arrest in C127 mouse cells (5). When
samples from arrested cells were assayed for histone H1 kinase (H1
kinase) activity following p13suc1 affinity purification, the
arrest was found to be associated with elevated H1 kinase activity.
Similarly, a constitutive form of rat brain
Ca2+/CaM-dependent kinase II (CaMKII
)
expressed in Schizosaccharomyces pombe causes a
G2 arrest, although H1 kinase activity was reported to
remain low in this case (6).
Because the disruption of the CaMK gene in A. nidulans arrests spores prior to nuclear division and germ tube formation, the present studies were undertaken in order to examine the potential role of CaMK in spore germination. We have demonstrated that truncation of the A. nidulans CaMK at Ile-292 generates a constitutively active Ca2+/CaM-independent enzyme. The expression of this enzyme in germinating spores causes arrest before entry into the first S phase prior to germ tube extension and this arrest is completely reversible up to 9 h. The arrest is accompanied by prematurely elevated H1 kinase activity and NIMA kinase activity. In contrast the expression of a truncated inactive kinase has no phenotypic consequences. The data suggest a role for CaMK-dependent phosphorylation and appropriate dephosphorylation upon germination for entry into S phase.
Strains were R153 (wA3; pyroA4), GR5 (A773; pyrG89; wA2; pyroA4), SO26 (nimT23; pyrG89; wA2; biA1; pabaA1), SO6 (nimA5; pyrG89; wA2; cnxE16; sC12; yA2; choA1; chaA1), and SO7 (wA3; nimA5), as well as SWJ-32 (nimG10; pyrG89; nicA2, chaA1) and SWJ-310 (nimR21;pabaA1) (both kindly provided by Steven W. James of Gettysburg College, Gettysburg, PA).
Site-directed MutagenesisThe plasmid p4b-11, which
contains the full-length A. nidulans CaMK cDNA, was
kindly provided by Dr. Diana C. Bartelt of St. Johns University (7).
The cDNA fragment was introduced into the EcoRI site of
M13mp8 and used for mutagenesis by the method of Olsen and Eckstein (8)
following the Amersham protocol. The inactive kinase was made by
mutating Lys-50 to Met using the oligonucleotide
5-CAGGATAATCATTACCGCAAC-3
. The truncated kinase was created by
introducing a stop codon at Ile-292 using the oligonucleotide 5
-GCAATGTACGCTCAGATCTCGGG-3
.
The wild type and mutant CaMK cDNAs were introduced
into pGEM3Zf() (Promega) between the KpnI and
SalI sites. One µg of each purified plasmid was linearized
with SalI and transcribed by the method of Melton et
al. (9) using T7 RNA polymerase. In vitro translation
reactions were performed with 1 µg of CaMK transcript in a 50-µl
reaction containing 33 µl of reticulocyte lysate and 20 µCi of
[35S]Met according to the Promega protocol. In order to
correct for the efficiency of translation of the different cDNAs,
the reaction mix was subjected to SDS-polyacrylamide gel
electrophoresis and the radiolabeled bands were excised from the gel
and counted in a liquid scintillation counter. Endogenous kinase
activity in reticulocyte lysate was subtracted from each sample, and
the activity was normalized for the efficiency of translation using 1 µCi of translation product for each Met residue in the various
protein sequences. Kinase activity was measured in the presence
of 35 mM Na-Hepes, pH 7.8, 10 mM
Mg(OAc)2, 50 µM ATP, 1 µCi/tube
[
-32P]ATP (4 Ci/mmol), 50 µM GS-10
peptide (PLRRTLSVAA) containing either 1 mM
CaCl2 and 1 µM A. nidulans CaM or
2 mM EGTA in a final volume of 50 µl. The reaction
mixture was incubated for 2 min at 30 °C, 20 µl of the mixture was
spotted onto P-81 filter paper and immediately washed with 75 mM H3PO4, and radioactivity
incorporated into the peptide was quantified using a liquid
scintillation counter (10).
A. nidulans strains were grown in minimal medium (10 mM urea, 7 mM KCl, 2 mM MgSO4, and 1 × trace elements and vitamins) (11) containing 50 mM glycerol (MMG), 50 mM glycerol plus 100 mM threonine (MMG/T) or 50 mM dextrose (MMD). Two percent agar was added for solid medium, 5 mM uridine and 10 mM uracil were added for pyrG89 strains, and 5 µM pyridoxine for pyroA4 strains. The CaM inhibitor W7 was used at a concentration of 250 µM and purchased from Sigma. The SO26 strain was grown in enriched medium (YG) containing 0.5% yeast extract and 2% glucose.
Transformation of A. nidulans with CaMK Expression VectorsThe mutant A. nidulans CaMK cDNAs were
introduced into pAL5, an A. nidulans expression vector
containing the alcA promoter, the 3-untranslated region of
the nonessential histone H2A gene and the pyr4 nutritional
marker, using the KpnI (5
) and BamHI (3
)
restriction sites (12). Transcription from the alcA promoter is regulated by the carbon source in the medium; glucose represses, threonine induces, and glycerol allows a basal level of transcription. The histone H2A gene targets the homologous recombination to the nonessential H2A locus. The pyr4 nutritional marker
complements the pyrG89 mutation and allows growth in the
absence of uridine and uracil. The GR5 strain was transformed as
described by Lu and Means (11). The transformants were selected on YG
plates without uridine or uracil, and cloned by streaking to single
colonies three times.
The germlings were harvested on Miracloth, washed with water, and frozen in liquid nitrogen. One hundred µg of frozen mycelia were ground with a mortar and pestle for 5 min on ice in 1 ml of homogenizing buffer (35 mM HEPES, pH 7.8, 10% glycerol, 50 mM NaCl, 1 mM MgCl2, 5 mM EGTA, 1 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride, 2 µg/ml leupeptin, and 10 µg/ml aprotinin). The homogenate was clarified by centrifugation at 16,000 × g and 4 °C for 10 min. The protein concentration was determined as described by Bradford (13). CaMK activity was determined as described above using 10 µg of total cell extract and 50 µM autocamtide-2 (KKALRRQETVDAL) as a peptide substrate.
Determination of Cell Cycle KineticsThe number of nuclei per germling was determined after staining samples with the fluorescent dyes, DAPI or H33825 (11). DNA synthesis was quantified by [3H]adenine incorporation according to the method of Bergen et al. (14). The flow cytometry was performed as described by James et al. (15).
Histone H1 and NIMA Kinase AssaysExtracts were prepared
from frozen mycelia by grinding in a mortar and pestle in H buffer (25 mM Tris, pH 7.5, 10 mM -glycerophosphate, 1 mM EGTA, 0.1% Nonidet P-40, 5 mM
p-nitrophenyl phosphate, 1 mM sodium
orthovanadate, 0.1 mM dithiothreitol, 2 µg/ml aprotinin, 2 µg/ml leupeptin, 5 mM benzamidine, and 1 mg/ml Pefabloc
(Boehringer Mannheim) for 5 min on ice. The H1 kinase assays were
performed following p13suc1 affinity purification with p13
agarose (Oncogene Science, Uniondale, NY) from 100 µg of cell extract
as described by Booher et al. (16). Briefly,
affinity-purified samples were washed three times in H buffer and once
in KAB buffer (50 mM Tris, pH 7.5, 10 mM MgCl2, 1 mM dithiothreitol). The samples were
assayed at 30 °C for 10 min in KAB buffer containing 125 µg/ml H1
histone, 80 µM ATP, and 2.5 µCi/tube
[
-32P]ATP. Reactions were stopped by addition of
5 × Laemmli sample buffer and separated on a 12% acrylamide gel
containing SDS. NIMA kinase activity was determined as described by Lu
and Means (17) with the following modifications. 1) NIMA protein was
immunoprecipitated from 200 µg of extract protein using an antibody
generated to the COOH-terminal 17 amino acids of A. nidulans
NIMA. The antibody was generously provided by Kun Ping Lu and Tony
Hunter (The Salk Institute, La Jolla, CA). 2) Assays contained 100 nM microcystin (Calbiochem) and 10 µM protein
kinase A inhibitor (Sigma). In addition, 100 µM PLM-(54-72) (DEEEGTFRSSIRRLSTRRR) was used as a
peptide substrate.
Western blots were performed by standard methods (18). A. nidulans CaMK was measured using affinity-purified (19) rabbit polyclonal antibody to CaMK (4), followed by incubation with 125I-protein A. Bands were quantified using a Molecular Dynamics PhosphorImager equipped with a densitometer.
Since
our previous work had demonstrated that CaMK is essential for
germination (4), we examined changes in CaMK protein expression in
germinating spores. Wild type GR5 spores were isolated and grown for
0-6 h in MMD/A, and, at the times indicated in Fig. 1A, aliquots were frozen. Extracts were
prepared and CaMK protein quantified by Western blot analysis. Fig.
1A demonstrates that CaMK protein is present at low levels
in A. nidulans spores and accumulates as the spores
germinate. Fig. 1B shows the densitometric quantitation of
the autoradiograph from Fig. 1A, and a 15-fold increase in
CaMK over the 6-h time course. Germination of spores in minimal medium
requires 3-4 h for entry into the first S phase (4, 14, 20) and 5-6 h
for completion of the first mitosis. This suggests that CaMK increases
as spores germinate in a manner similar to that found for CaM, which
increases 4-fold within 4 h of germination (2). In contrast, once
the cells are exponentially growing, there is no cell
cycle-dependent change in CaMK protein levels as shown in
Fig. 1C. SO26 (nimT23ts) cells were
grown 6 h in enriched medium (YG) at 32 °C. The cells were
arrested in G2 at the non-permissive temperature for 4 h and released into fresh YG at the permissive temperature. Aliquots were harvested at 5- or 10-min intervals for 90 min. In YG medium one
complete cell cycle occurs within 95 min (14). In addition, these
extracts were assayed for CaMK activity in the presence of excess
Ca2+/CaM and no changes were
observed.2 Equivalent CaMK protein
expression was also observed throughout a 150-min time course following
release from the SO6 (nimA5ts) G2
arrest in YG medium. Similar results were obtained with release of the
GR5 strain from an hydroxyurea (S phase) arrest.2
Collectively these experiments demonstrate that CaMK protein is present
at a constant level throughout the normal cell cycle. Since CaM levels
vary during the A. nidulans cell cycle (21), the
physiologically relevant CaMK activity is likely to be regulated by
changes in the Ca2+ and/or CaM concentration, rather than
at the level of transcription or translation of the kinase.
Construction and Characterization of Constitutively Active and Inactive Forms of A. nidulans CaMK
Fig.
2A is a schematic representation of the
domain structure of A. nidulans CaMK compared to rat brain
CaMKII. The latter enzyme can be made independent of
Ca2+/CaM by truncation at Leu-290 or rendered inactive by
mutation of Lys-42, which is involved in ATP binding (22). By analogy a
constitutive form of A. nidulans CaMK was created by
removing the putative CaM-binding domain by truncation at amino acid
Ile-292 (CaMKct). An inactive construct was created by changing Lys-50 to Met (K50M). This mutation was made in both the wild type and truncated enzymes as depicted in Fig. 2A. The mutants were
expressed in an in vitro transcription and translation
system to verify that the K50M mutation produced an inactive enzyme and
the amino acid 292 truncation produced a constitutive enzyme. The
activity of the translated proteins was determined using the GS-10
peptide as a substrate as shown in Fig. 2B. The assay was
carried out in the presence of Ca2+/CaM or in the presence
of EGTA. The translation mixture from wild type CaMK (CaMKwt) cDNA
showed Ca2+/CaM-dependent phosphorylation of
GS-10. The reaction velocity was 3.74 × 10
8 mol of
phosphate/ml of lysate/min in the presence of Ca2+/CaM and
0.14 × 10
8 mol/ml of lysate/min in the presence of
EGTA. However, the translation mixture of CaMKct contained similar
kinase activity either in the presence of Ca2+/CaM or EGTA
(Ca2+/CaM, 5.01 × 10
8 mol/ml of
lysate/min; EGTA, 6.80 × 10
8 mol/ml of lysate/min).
Thus the specific activity of the Ca2+/CaM- independent
form was at least as high as the
Ca2+/CaM-dependent activity of the wild type
enzyme. The translation mixture containing CaMKwtK50M or CaMKctK50M did
not show significant kinase activity above that in the translation
mixture without added mRNA. These results demonstrate that
truncation of A. nidulans CaMK converted the enzyme to a
Ca2+/CaM-independent form, whereas the point mutation,
K50M, abolished kinase activity whether in the context of the
full-length or truncated protein.
Generation of A. nidulans Strains Conditional for the Expression of Mutated CaMK cDNAs
In order to be able to regulate the
expression the mutant CaMK cDNAs in A. nidulans, the
cDNAs for CaMKct and CaMKctK50M were independently subcloned into
the A. nidulans expression vector, pAL5 (12). The pAL5
vector contains the alcA promoter and the histone H2A
3-terminator to target integration of the construct to a non-essential
but homologous site and allow the expression of heterologous cDNAs.
The plasmid also contains the pyr4 gene from
Neurospora crassa to provide a nutritional marker, which complements the pyrG89 mutation and allows growth in the
absence of exogenous pyrimidines (11). The plasmids were used for the transformation of the GR5 strain carrying the pyrG89
mutation. The pure haploid strains were analyzed by Southern blot
analysis. The strains were designated as follows: alcCaMKct-4,
conditional for the expression of CaMKct cDNA; and alcCaMKctK50M-4,
conditional for the expression of CaMKctK50M cDNA. The number of
copies of the cDNA integrated into the genome was assessed by
measuring the ratio of the intensity of the 0.83-kb band of the
transgene and the 2-kb band of the wild type.2 As there is
only one CaMK gene in A. nidulans (23), the signal from the
2.0-kb endogenous CaMK gene was used as an internal standard for a
single copy signal. The estimated copy number predicted in this manner
is indicated with a hyphen after the name of the strain. In inducing
medium (minimal medium containing threonine and glycerol, or MMG/T)
CaMKct mRNA rapidly increases 10-fold within 1 h in alcCaMKct
mycelia. In de-repressing conditions (minimal medium containing
glycerol, or MMG), a moderate 2.5-fold increase of CaMKct mRNA, was
detected between 30 min and 1 h. In contrast, in repressing
conditions (minimal medium containing glucose, or MMD), there was no
significant CaMKct mRNA production. These changes in mRNA are
not shown but were measured as a prerequisite to examining changes in
CaMK activity. Fig. 2C shows the inducible constitutive CaMK
activity in alcCaMKct-4 cell lysates using autocamtide-2 as a peptide
substrate, because it was found in preliminary studies to be the best
and most specific substrate for A. nidulans CaMK. The
alcCaMKct-4 spores were inoculated into MMD and grown for 8-12 h. The
germlings were transferred to fresh MMD or MMG/T and harvested at the
times indicated in Fig. 2C. Extracts were prepared and
assayed in the presence of EGTA to detect only the CaMKct activity or
in the presence of Ca2+/CaM to detect the total CaMK
activity. As shown in Fig. 2C, switching from repressing to
inducing medium produced a 3-fold increase in autocamtide-2
phosphorylation above the normal level of
Ca2+/CaM-dependent enzyme activity within
1 h that is maintained for at least 4 h (open
circles compared to closed triangles). Addition of
Ca2+/CaM to the extract of alcCaMKct-4 produced even
greater CaM kinase activity (open triangles compared to
open circles). Since we do not know the level of
Ca2+ in vivo, it is safe to presume that the
endogenous level of CaM kinase activity is something between these
values. Therefore, the activity produced represents an overexpression
of CaM kinase activity. As expected, there is no significant
Ca2+/CaM-independent kinase activity in cell lysates from
cells continuously grown in MMD (closed circles).
Furthermore, no CaMK activity is induced in the alcCaMKctK50M-4 strain,
although the induction of the CaMKct and CaMKctK50M proteins was
verified as early as 1 h after transfer to inducing medium by
Western blot analysis.2
In order to assess the effect of overexpression of CaMKct
cDNAs, transformant spores were germinated on MMD and MMG/T solid medium. The wild type strain R153 and the GR5 and alcCaMKctK50M-4 strains grew equally well on inducing or repressing plates (data not
shown). In contrast the alcCaMKct-4 strain did not grow on inducing
plates. Fig. 3 shows representative germlings from GR5, alcCaMKct-4, and alcCaMKctK50M-4 after 12 h of growth in inducing or repressing liquid cultures. Induction of CaMKctK50M has no phenotypic effect on either nuclear division or germ tube extension (Fig. 3F) compared to the GR5 control (Fig. 3B).
On the other hand, expression of CaMKct inhibits both nuclear division
and germ tube extension (Fig. 3D). These results indicate
that expression of Ca2+/CaM-independent CaMK severely
inhibits the growth of A. nidulans in a manner that requires
an active protein kinase. The effect of repression or overexpression of
CaMKct on A. nidulans growth was also evaluated by
monitoring the changes in total dry weight. There was no difference in
the dry weight of GR5 and alcCaMKct-4 strains grown in repressing
medium, confirming the results seen in Fig. 3. However, in inducing
medium, there was virtually no increase in the dry weight of alcCaMKct,
whereas the dry weight of GR5 showed a linear increase with
time.2 Fig. 3G demonstrates that the effect of
induction of CaMKct for up to 9 h is readily reversible upon
transfer of the cells to repressing medium. After a 2-h lag, the cells
have re-entered the nuclear division cycle.
Overexpression of CaMKct Inhibits Entry into S Phase
In order
to determine more specifically the point of arrest due to
overexpression of CaMKct, DNA synthesis was measured by quantifying the
incorporation of [3H]adenine into DNA. The alcCaMKct-4
spores were germinated in MMD or MMG/T in the presence of
[3H]adenine. After the indicated times, germlings were
harvested and [3H]adenine incorporation in the DNA
fraction was determined (Fig. 4). Arrowheads
have been inserted above the graph to show approximate times of
successive S phases and mitoses based on quantifying the number of
nuclei per cell at each time point. In MMD, the first increase in
incorporation of [3H]adenine into DNA in the alcCaMKct-4
strain occurred between 3 and 4 h after germination, which
corresponds to entry into the first S phase. This is consistent with
the nuclear division kinetics showing an increase from 1 to 2 nuclei/cell between 3 and 6 h. Subsequent increases were observed
between 5 and 6 h, as well as between 9 and 10 h; the
increases are also consistent with the timing of the next rounds of
nuclear division. However, when the alcCaMKct-4 strain was germinated
in MMG/T, there was no significant [3H]adenine
incorporation into the DNA fraction. These results demonstrate that the
overexpression of CaMKct upon germination of quiescent spores prevents
entry into the first S phase.
Although CaMKct does not require CaM, we performed a series of FACS
analyses to elucidate the normal order of requirements for CaM during
spore germination. Panel 1 of Fig. 5 shows
the DNA histogram of SWJ-310 (nimRts) spores
arrested at the non-permissive temperature in G1 (14) with
a 1 N DNA content. The identity of NIMR has not been
determined; however, it serves as a control for a G1 DNA
content. Panel 2 shows the DNA histogram of SO7
(nimAts) spores arrested at the non-permissive
temperature in G2 with a 2 N DNA content.
Panel 3 shows the SWJ-32 (nimGts)
strain arrested in G1; however, this arrest is slightly
leaky as some cells contain a 2 N and a 4 N DNA
content. The identity of NIMG has been shown to be the A. nidulans cyclin required for both G1/S and
G2/M transitions.3 Panels
4-6 show the DNA histograms when the germlings in panels 1-3, respectively, are released into the permissive temperature for 2 h in the presence of the CaM antagonist, W7. It has
previously been shown that CaM is required in A. nidulans
for the progression from G1 to S and from G2 to
M, and for completion of mitosis (1). Thus panel 4 demonstrates that the requirement for CaM in the G1/S
transition is distal to the nimR arrest point, as the cells do not obtain a 2 N DNA content. Interestingly in
panel 6 the DNA content of the SWJ-32
(nimGts) strain released into W7 at the
permissive temperature does not change compared to that in panel
3 in the absence of W7 at the non-permissive temperature. These
results suggest that CaM is required after the
NIMGcyclin B temperature-sensitive
arrest point for entry into S phase. Panel 5 verifies that
progression from a G2 arrest through mitosis and into a new
round of DNA synthesis also requires CaM. As controls, panels
7-9 show the cycling of the nimR, nimA, and
nimG germlings, respectively, 2 h after release from
the non-permissive temperature.
Overexpression of CaMKct in A. nidulans Spores Causes Premature Activation of Histone H1 Kinase and NIMA Kinase Activities
In
order to investigate the mechanism responsible for the complete arrest
of spores prior to entry into the first S phase, which results from the
overexpression of CaMKct, we evaluated changes in two protein kinases
known to be required for progression through the nuclear division cycle
in A. nidulans. First, the H1 kinase activity in spores and
mycelia grown for up to 6 h in MMG/T was determined following
p13suc1 affinity purification. As can be seen in Fig.
6A, the presence of the constitutive CaMK
caused a premature elevation in H1 kinase activity as early as 1 h
after exposure to inducing medium. This increase in H1 kinase activity
was similar to that seen at 4 h in the alcCaMKctK50M-4 strain,
which corresponds to the time of entry into S phase (Fig. 4). In
addition, when alcCaMKct-4 spores are grown in MMD, the increase in H1
kinase activity is very similar to the activity seen in the
alcCaMKctK50M-4 strain grown in MMD or MMG/T at each time point between
0 and 6 h.2 This demonstrates that the increase in H1
kinase activity is the result of the expression of CaMKct. Similar
results were obtained using an anti-NIMXcdc2 antibody to
precipitate NIMXcdc2, the A. nidulans homologue of
p34cdc2, followed by the kinase assay.2 The
antibody was kindly provided by Steve Osmani of the Weis Research
Institute, Danville, PA. (24). Thus the increase in H1 kinase activity
caused by overexpression of CaMKct appears to be due to unscheduled
activation of NIMXcdc2, which is required for both
G1 and G2 in A. nidulans (24).
Panel B of Fig. 6 shows the changes in the activity of the mitotic Ser/Thr protein kinase, NIMA (25, 26), in the same extracts. NIMA was immunoprecipitated with an antibody to the COOH-terminal 17 amino acids of NIMA. The immunoprecipitates were assayed using PLM (DEEEGTFRSSIRRLSTRRR) for the substrate peptide. This peptide is the best substrate characterized for NIMA (27) and is a very poor substrate for A. nidulans CaMK. The data demonstrate that NIMA activity is low in spores and increases between 4 and 6 h as cells enter mitosis in the alcCaMKctK50M-4 strain. However, in the alcCaMKct-4 strain NIMA, activity is elevated by 1 h and remains elevated over the 6-h time course. Thus as was the case for NIMXcdc2, the expression of CaMKct results in premature activation of NIMA to the level usually restricted to the G2 phase of the nuclear division cycle.
A. nidulans CaMK like other multifunctional CaMKs can be made Ca2+/CaM-independent (CaMKct) by truncation prior to the regulatory domain. A. nidulans CaMKct expressed in rabbit reticulocyte lysate had similar activity in the presence of Ca2+/CaM or EGTA to the expressed wild type CaMK in the presence of Ca2+/CaM. The truncated CaMKct was expressed in A. nidulans with the inducible alcA promoter and caused germinating spores to arrest prior to S phase. Previous studies have documented that the overexpression of a constitutive CaMK in vertebrate (5) or fission yeast cells (6) causes a G2 arrest. We have confirmed a G2 arrest in A. nidulans caused by induction of CaMKct in cycling cells.2 In addition, we have demonstrated that overexpression of a constitutive CaMK prevents germination or entry into the proliferative cycle from the dormant state of conidia. This arrest is accompanied by elevated histone H1 and NIMA kinase activities. The H1 kinase activity is at least partly due to NIMXcdc2 as increased activity is immunoprecipitable with an antibody to the NIMXcdc2 protein (24).
The mechanism explaining the arrest with elevated H1 kinase activity is
not clear. It is possible that the presence of CaMKct leads to the
phosphorylation of a component of the NIMXcdc2/cyclin B
complex, resulting in activation of the cyclin-dependent kinase (CDK). Alternatively CaMKct may prevent an obligatory
dephosphorylation step resulting in the induction of a checkpoint
irrespective of the state of CDK activity. Finally CaMKct may lead to
phosphorylation of NIMXcdc2 on an inappropriate site resulting
in a non-productive activation of H1 kinase. It has been shown that
decreased extracellular Ca2+ or intracellular CaM causes a
G2 arrest and prevents activation of NIMXcdc2 and
NIMA (20). Thus some Ca2+/CaM-dependent event,
perhaps phosphorylation by CaMK, is important for the activation of
NIMXcdc2 and NIMA in A. nidulans. One possibility to
consider is the fact that the highly conserved activating
phosphorylation site of p34cdc2, Thr-161, exists in the
A. nidulans NIMXcdc2 sequence and is located in a
consensus CaMK phosphorylation site, making it a potential target for
CaMK (24). However, our preliminary data indicate that CaMK does not
phosphorylate NIMXcdc2 immunoprecipitated from A. nidulans extracts. Another model for activation of p34cdc2
has been suggested (28), in which the target for CaMK may be the Tyr
phosphatase Cdc25, which is responsible for the activation of
p34cdc2/cyclin B by dephosphorylation of Thr-14 and Tyr-15. In
theory initial phosphorylation of Cdc25 causes a low level of
activation of the Cdc25 phosphatase, which can then dephosphorylate and
activate p34cdc2/cyclin B. The activation of
p34cdc2/cyclin B causes a positive feedback to further
phosphorylate and activate Cdc25. This model is supported by the recent
observation that CaMKII can phosphorylate and increase the activity
of Cdc25 in vitro.4 Thus
NIMTcdc25, the A. nidulans Cdc25 homologue, is
perhaps a more likely target for CaMK.
One explanation for the premature increase of NIMA activity is that the elevated NIMXcdc2 activity causes the activation or stabilization of this protein kinase. The carboxyl-terminal non-catalytic domain of NIMA has multiple p34cdc2 consensus phosphorylation sites. O'Connell et al. (29) have shown that this phosphorylation is not required for the activity of NIMA; however, they suggest that the phosphorylation of these sites may stabilize or help localize NIMA. Perhaps the elevated NIMA activity is simply a consequence of the G1/S arrest and prematurely elevated NIMXcdc2 activity. Consistent with this possibility, either a G2 arrest or a mitotic arrest has been shown to result in elevated NIMA kinase activity (30-32).
Our results show that spores contain a low level of NIMXcdc2; however, NIMEcyclin B was not detectable in spores by Western blot analysis.2 The H1 kinase activity increases severalfold between 3 and 4 h, the time that corresponds to the initiation of the first S phase. Therefore, it is possible that the synthesis of NIMEcyclin B could regulate the entry into DNA synthesis. Indeed, overexpression of NIMEcyclin B increases pre-maturation promotion factor levels in A. nidulans (33). Apparently, A. nidulans contains a single cyclin-dependent kinase (NIMXcdc2) that is required both for G1 and G2 progression (24). It may also be the case that a single B-type cyclin is required (NIMEcyclin B). The nimE gene was originally defined as a temperature-sensitive mutation that caused a block in G2 (14, 33, 34), whereas nimG defined a temperature-sensitive mutation that arrested cells in G1 (Fig. 5, panel 3). Both nimEts and nimGts have been shown to be the result of mutations in the same gene, that for the A. nidulans equivalent of cyclin B (33).3 The possibility that a single cyclin-CDK complex might regulate progression into both S and M phases in A. nidulans is strengthened by the intriguing observations that in S. pombe mutants have been generated in which a single B-type cyclin can regulate the unique p34cdc2 to promote both S phase and mitosis (35-37). Furthermore, analysis of DNA content by flow cytometry suggests that there is a CaM requirement for G1/S progression distal to the arrest point caused by the mutation in cyclin B present in the nimGts strain.
Why would premature H1 kinase activity lead to arrest prior to entry into S phase? The study of replication and the prevention of rereplication may shed some light on this issue. Screens for mutations which allow rereplication in S. pombe have resulted in the isolation of mutations in cdc2 and in cdc13 (the cyclin partner of cdc2 in S. pombe) (38, 39). This evidence makes a strong case for the role of the cyclin-CDK complex in preventing rereplication during G2. Su et al. (40) have proposed a model whereby active cyclin-CDK complexes, which are present throughout S and G2 phases, prevent reinitiation by preventing further assembly of functional replication origins. In mitosis cyclin is degraded and CDK inactivated, allowing entry into the G1 phase and the ability to assemble a replication origin for a new round of DNA synthesis. Therefore, rereplication is coupled to progression through mitosis. There is reason to believe this model also applies to other organisms in addition to S. pombe. For example mutations in Drosophila cyclin A led to endoreduplication in normally mitotic cells (41), and in cultured rat fibroblasts treatment with Tyr kinase inhibitors to inactivate cyclin B/CDK induces DNA rereplication (42). Finally CDKs have been shown to inhibit assembly of pre-replication centers in Xenopus oocyte extracts (43). If A. nidulans spores do not contain functional replication origins, this model would predict that the premature H1 kinase activity caused by CaMKct prevents assembly of origins of replication and therefore prevents DNA synthesis. Furthermore, this model would also predict that the effects of CaMKct should be reversible as we have shown to be the case. The removal of CaMKct would allow the inactivation of H1 kinase and permit the formation of origins of replication and subsequent entry into S phase. The fact that this occurs with a lag of 2-3 h would provide time for the destruction of CaMKct protein and assembly of replication complexes. Taken together it seems likely that the arrest occurs because some unknown protein(s) is phosphorylated by CaMKct and must be dephosphorylated to allow entry into the cell cycle. Further understanding of these experiments awaits the elucidation of the in vivo substrate(s) for A. nidulans CaMK and CaMKct.
We thank Danny Lew, Donna Crenshaw, and members of the Means laboratory for invaluable discussions during these studies and the preparation of this manuscript. We thank Steve Osmani and Aysha Osmani for antibodies to A. nidulans NIMEcyclin B and NIMXcdc2, K. P. Lu and Tony Hunter for the antibody to NIMA, Steve James for the SWJ-32 and SWJ-310 strains and for assistance with the flow cytometric analysis of A. nidulans, and Diana Bartelt for the cDNA for A. nidulans CaMK.