(Received for publication, September 23, 1996, and in revised form, January 29, 1997)
From the Department of Chemistry, University of Nebraska, Lincoln, Nebraska 68588-0304
The regulation of protein synthesis was studied in
KRC-7 cells (rat hepatoma) grown in complete medium, during serum
starvation, and mitogen activation. Upon serum starvation, the cells
lost almost completely p67 mRNA, p67 protein, and protein
synthesis activity. After phorbol 12-myristate 13-acetate addition, the same serum-starved cells regained p67 mRNA, p67 protein, and
protein synthesis activity. Also, the extracts from the serum-starved cells phosphorylated the eukaryotic initiation factor-2 (eIF-2) -subunit. This eIF-2
-subunit phosphorylation was not observed when the extracts from either the cells grown in complete medium or
mitogen-activated cells were used (Gupta, S., Wu, S., Chatterjee, N.,
Ilan, J., Ilan, J., Osterman, J. C., and Gupta, N. K. (1995) Gene
Expr. 5, 113-122). We now report the following. 1) The eIF-2 kinase activity was the same in the cells grown in complete medium, after serum starvation, and subsequent mitogen stimulation. However, the eIF-2 kinase in the cells grown in complete medium and also after
mitogen activation of the serum-starved cells cannot phosphorylate eIF-2
-subunit as these cells contain p67. After removal of
endogenous p67 by p67 antibodies, the extracts from all these cells
similarly phosphorylated exogenously added eIF-2. 2) None of the cell
extracts showed p67 deglycosylase activity. 3) The p67 mRNA was
synthesized in serum-starved cells by expression of a p67 cDNA. The
appearance of p67 mRNA in the serum-starved cells was accompanied
by the appearance of p67 protein. Also, the rates of protein synthesis in the serum-starved cells were restored nearly to the level observed in the confluent cells. The expression of p67 cDNA also
significantly increased protein synthesis rates in the cells grown in
complete medium and in mitogen-activated cells. These results show that the loss of protein synthesis activity in serum-starved cells was due
to loss of p67 mRNA. The expressed p67 mRNA was stable in
serum-starved cells. These results, therefore, suggest that the loss of
p67 mRNA in serum-starved cells is due to loss of p67
transcription. The p67 transcription regulates translation.
An important regulatory mechanism in animal cells involves
phosphorylation of the -subunit of the
eIF-21 by eIF-2 kinases, heme-regulated protein
synthesis inhibitor, and double-stranded RNA-dependent
protein kinase. This inactivates eIF-2 activity and inhibits protein
synthesis (1). Numerous reports have indicated that the animal cells
widely use this phosphorylation mechanism to regulate protein synthesis
under different physiological conditions include serum starvation
(2-3), heat shock (3-6), and viral infection (7).
We have previously reported that a 67-kDa glycoprotein (p67) protects
the eIF-2 -subunit from inhibitory phosphorylation by eIF-2
kinase(s). This promotes protein synthesis in the presence of active
eIF-2 kinase(s) in the cells (8-10). The p67 contains multiple
O-linked GlcNac residues (9). Similar O-linked
GlcNac residues are also present on numerous regulatory proteins
including transcriptional activators (11). An important characteristic of p67 is that this protein is rapidly degraded and is induced under
certain physiological conditions (10). Also, the level of this protein
correlates directly with the protein synthesis activity of the cells.
These observations suggest that p67 plays a critical role in regulation
of protein synthesis.
Our previous reports have indicated that p67 levels in cells are regulated by two independent mechanisms. One of them is protein deglycosylation. During heme deficiency, in reticulocyte lysate (12), and also during vaccinia viral infection of animal cells (13-14) a latent p67 deglycosylase is activated. The activated deglycosylase deglycosylates p67 and inactivates it. The second one is variations of p67 mRNA levels. Using a Northern blot procedure, we provided evidence (15) that p67 mRNA is present in the KRC-7 cells grown in complete medium. This mRNA disappears almost completely from serum-starved cells. However, when PMA was added to the serum-starved cells, p67 mRNA appeared in increasing quantities. The level of p67 mRNA in the cells correlated directly with the p67 protein level and also with the protein synthesis activity of the cells.
We have now studied the detailed mechanism of regulation of p67 mRNA, p67 protein, and protein synthesis activity in KRC-7 cells during serum starvation and subsequent mitogen activation. Several questions remain unanswered. The mechanism of disappearance of p67 mRNA and protein in serum-starved cells remains unanswered. This disappearance could be due to shut-off of p67 transcription and subsequent slow decay of the p67 transcript. Alternatively, nuclease activation in serum-starved cells may rapidly degrade p67 mRNA. In the absence of new p67 mRNA synthesis, p67 protein may slowly decay. Alternatively, a latent p67 deglycosylase may be activated in the serum-starved cells. This activated deglycosylase may rapidly deglycosylate newly synthesized p67. The deglycosylated p67 will then degrade. The requirement of p67 mRNA and p67 protein in protein synthesis is another aspect that needs to be investigated. Upon serum starvation, both p67 mRNA and p67 protein disappear. This accompanies loss of protein synthesis activity of the cells. These results, however, do not provide a conclusive evidence for the single requirement of p67 mRNA or p67 protein in overall protein synthesis in the cells. Other protein factors, besides p67, may also be lost after serum deprivation. These factor(s) may also be essential for protein synthesis. In the present work, we have expressed p67 cDNA in transformed KRC-7 cells grown in complete medium, during serum starvation, and subsequent mitogen activation. We have studied the effects of expression of this p67 cDNA on p67 mRNA and p67 protein synthesis.
The cloned cell line KRC-7, derived from Reuber H35 rat hepatoma cells, was kindly provided by John Koontz (University of Tennessee, Knoxville). The cells were grown as described previously (15). The serum-starved cells were prepared by washing cell monolayers with Hank's balanced solution followed by culturing in serum-free Dulbecco's modified Eagle's medium for an additional 3 days. A part of these cells were then activated by addition of 1.5 µM phorbol 12-myristate 13-acetate (PMA) (Sigma) for 2 h. For analysis, 6 × 105 cells were seeded onto 100-mm tissue culture dishes containing 10 ml of the medium.
Plasmids and OligonucleotidesThe pOPRSVI-CAT vector (Stratagene) is an episomal mammalian expression vector. It was used to express the p67 cDNA. It has also a neomycin resistance gene as a selection marker. For preparation of pOPRSVI-p67, the p67 gene was excised from pGEM-p67 (16) using NotI. The CAT gene was similarly removed from pOPRSVI CAT by NotI digestion. The 5467-bp linearized pOPRSVI and the 1.9-kb p67 cDNA were gel-purified using GeneClean® II kit (Bio 101, Inc.). The fragments were then ligated using standard protocol to make the pOPRSVI-p67 construct. The pOPRSVI-p67 construct was then amplified by cell transformation into Epicurian Coli® XL1-Blue MR competent cells (Stratagene). The positive colonies were checked by standard miniprep method and NotI digestion to ensure the presence of the intact insert.
Characterization and Detection of pOPRSVI-p67 DNA in KRC-7 cellsTo ensure proper orientation, the plasmid was digested with HindIII. Three DNA fragments (5.6, 1.5, and 0.267 kb) were produced as expected (data not shown). The correct orientation was also confirmed through sequencing (data not shown). The positive colonies were then amplified and pOPRSVI-p67 was purified. The existence of pOPRSVI-p67 DNA in KRC-7 cells was confirmed by Southern blot of SacI-digested episomal DNA (purified by Hirt extraction protocol) (17). The probe used was random primer labeled 300-bp fragment from p67 cDNA (positions 400-700 bp). A single fragment at the position 7.4-kb confirmed the presence of pOPRSVI-p67 plasmid (data not shown). This result indicates that POPRSVI-p67 vector DNA was maintained as an episome inside the cells. In each case, the plasmids were isolated from 2 × 107 cells. The plasmids were SacI-digested and then quantitated by using the random primer-labeled probe containing the p67 cDNA sequence (300 bp (400-700 bp from p67 cDNA)). Based on a molecular mass of 4662 kDa of pOPRSVI-p67 construct, it was calculated that there were approximately 300 copy numbers/cell.
Transfection of pOPRSVI Constructs in KRC-7 Cell LineThe pOPRSVI-p67 constructs with p67 cDNA (pOPRSVI-p67) and without p67 cDNA (pOPRSVI-0) were used to transfect the rat hepatoma cell line KRC-7 using Lipofectin reagent (Life Technologies, Inc.) (15). 50 µl of Lipofectin reagent and 10 µg of the construct were used. The cells were then selected in geneticin (800 µg/ml). The selection procedure took approximately 3 weeks.
Determination of RNA LevelsThe Northern blot technique was the same as described (15). The cells were grown under different growth conditions (the cells grown in complete medium (50-70%), serum-starved, and PMA-induced) and were then harvested. The total RNA was isolated using guanidium isothiocyanate method (18) and was analyzed (15).
Immunoblot AnalysisThe immunoblot analysis was used to measure p67protein levels in the cell extract (6 × 105 cells) at different physiological condition (10). The wild type cells and also the KRC-7 cells transformed with pOPRSVI-p67 vectors with p67 cDNA were grown to 50-70% confluency in complete medium. The cells were then starved in serum-free medium for 3 days. The starved cells were then incubated with PMA (1.5 µM) for 2 h. The cell extracts were prepared as described (10). The equal concentrations of proteins (100 µg) from each cell extract were used. The proteins were separated by 15% SDS-PAGE and were then transferred to Biotrace NT membranes. The proteins were immunoblotted using p67 mono- or p67 polyclonal antibodies (10). As reported (19), the p67 monoclonal antibodies recognize specifically the glycosyl residues of p67. These antibodies were routinely used for studies of p67 deglycosylation.
eIF-2The reaction conditions
for eIF-2 -subunit phosphorylation by eIF-2 kinases using cell-free
extracts were the same as described (10). The radioactively labeled
eIF-2
(P) formed during incubation was immunoprecipitated using
eIF-2
-polyclonal antibodies and Protein A-agarose. The
immunoprecipitates were then analyzed by 15% SDS-PAGE followed by
autoradiography (10).
Cells were grown in complete Dulbecco's modified Eagle's medium to 50-70% confluency. A part of these cells were then starved in serum-free medium for 3 days. After 3 days of serum starvation, the cells were treated with PMA for 2 h. All these cells were washed twice with phosphate-buffered saline (Life Technologies, Inc.) and were incubated in short term labeling medium (serum-free Dulbecco's modified Eagle's medium (Life Technologies, Inc.) lacking methionine) at 37 °C with 10% CO2 in an incubator for 15 min. The cells (1 × 106) were then washed and incubated in short term labeling medium for 10, 20, and 30 min in the presence of 0.1 mCi/ml of [35S]methionine. The cells were then harvested and lysed in lysis buffer (10). [35S]Methionine-labeled proteins (10 µg) in the cell extracts were separated by 15% SDS-PAGE and were detected by autoradiography using Kodak X-Omat AR film.
The extracts
from the cells grown in complete medium, serum-starved, and PMA-induced
KRC-7 cells were analyzed for eIF-2 -subunit phosphorylation. Excess
exogenous eIF-2 was added. The eIF-2
-subunit phosphorylation was
measured both before and after removal of endogenous p67 by treatment
of the cell extracts with p67 polyclonal antibodies. Without p67
antibody treatment, the extent of eIF-2
-subunit phosphorylation
should be the result of the combined action of both endogenous p67 and
eIF-2 kinase(s). If the p67 level is high, p67 will protect eIF-2
-subunit from eIF-2 kinase(s). After p67 antibody treatment, the
endogenous eIF-2 kinase(s) will phosphorylate eIF-2
-subunit. The
extent of this phosphorylation should be a measure of the eIF-2 kinase
activity of the cells. Our previous report has indicated that p67
antibody treatment increased eIF-2
-subunit phosphorylation in
hemin-supplemented reticulocyte lysate presumably because these
antibodies inactivated endogenous p67 (10, 12).
We used two different protein concentrations in the cell extracts (100 and 200 µg). The results without p67 antibody treatment are shown in
Fig. 1, panel A. The extracts from the cells
grown in complete medium (lanes 1 and 2) and the
mitogen (PMA)-activated (lanes 5 and 6) cells did
not show detectable phosphorylation. However, the extracts from the
serum-starved cells significantly phosphorylated eIF-2 (lane
3), and the extent of this phosphorylation increased with
increasing concentration of the proteins used in the experiments
(lane 4). The experiment described in panel B was
carried out under similar conditions as described in panel A
except that these cell extracts were preincubated with excess p67
antibodies. The extent of eIF-2 -subunit phosphorylation in all the
cell extracts significantly increased and were nearly the same at equal
protein concentration in each cell extract. The eIF-2
-subunit
intensities of the phosphorylation bands in panel B were
determined. They were as follows: lane 1, 1.0; lane 2, 1.8; lane 3, 0.98; lane 4, 1.6;
lane 5, 1.0; and lane 6, 1.9.
As mentioned, the above experiments were performed with exogenously
added eIF-2 (Fig. 2). Similar experiments were also
performed without exogenous eIF-2 (data not shown). These results were
the same (also see Fig. 5). These results indicate that the eIF-2 kinase and not eIF-2 is limiting in the cell extracts. These results also indicate that an eIF-2 kinase remains in active form and at the
same level under different growth conditions. The p67 level in the
cells varies under different growth conditions. This p67 level
correlates with the ability of the cell extracts to phosphorylate eIF-2
-subunit.
Analysis of p67 Deglycosylase Activity
The p67 deglycosylase activity in KRC-7 cells was assayed under different growth conditions (Fig. 2). As reported (13-14), the KRC-7 cells contain a latent p67 deglycosylase. This p67 deglycosylase is activated immediately after vaccinia viral infection. In our experiments, we have used as a control an extract from the vaccinia viral infected KRC-7 cells. The p67 deglycosylation of the exogenously added p67 was measured by immunoblotting using p67 monoclonal (panels A-C, lanes 1-4) and polyclonal (panels A-C, lanes 5-8) antibodies. As expected, the p67 monoclonal antibody-reactive materials rapidly disappeared in the extracts form the vaccinia viral infected cells (panel A, lanes 1-4), and the polyclonal reactive materials remained relatively stable (panel A, lanes 5-8). Under similar experimental conditions, both the p67 mono- and polyclonal reactive materials remained essentially unchanged in the extracts from the cells grown in complete medium (panel B, lanes 1-4, p67 monoclonal antibodies; lanes 5-8, p67 polyclonal antibodies) and the serum-starved cells (panel C, lanes 1-4, p67 monoclonal antibodies; lanes 5-8, p67 polyclonal antibodies).
These results show that although KRC-7 cells contain a latent p67 deglycosylase, this p67 deglycosylase is not activated in serum-starved cells and is not responsible for p67 disappearance in these cells.
Northern Blot Analysis of p67-mRNAThe p67 mRNA was
determined in wild type and in pOPRSVI-p67 transformed KRC-7 cells
under different growth conditions (Fig. 3). The standard
Northern blotting procedure was used (15). An equal amount of RNA (15 µg) in each experiment was used. This RNA showed similar 28 S rRNA
bands when analyzed by gel electrophoresis followed by staining using
ethidium bromide (data not shown).
As shown (Fig. 3, panel A), the p67 mRNA was present in
the cells grown in complete medium (lane 1), and the level
of this mRNA increased slightly when PMA was added to the cells
grown in complete medium (lane 2). Also, the p67 mRNA
level in the cells transformed with pOPRSVI-p67 increased significantly
(lane 3). This level remained unchanged upon PMA addition
(lane 4). Upon serum starvation for 3 days, the p67 mRNA
almost completely disappeared from the wild type KRC-7 cells
(lane 5). This level appeared in increased quantities after
2 h of PMA addition (lane 6). By expression of p67
cDNA in the serum-starved transformed cells, the p67 mRNA was
restored to nearly the same level as in the the cells grown in complete
medium (lane 7). Upon addition of PMA in these cells, the
p67 mRNA appeared in increased quantities after 2 h in the transformed serum-starved cells (lane 8). The size of the
p67 mRNA (1.9 kb) synthesized in the presence of p67 cDNA was
similar to the endogenous p67 mRNA (2 kb, (16)). The levels of a
controlled RNA, glyceraldehyde-3-phosphate dehydrogenase RNA remained
essentially unchanged during different growth conditions (Fig.
4, panel B). As control, we also used KRC-7 cells
transformed with pOPRSVI-0. The p67 mRNA remained essentially the
same as in the wild type KRC-7 cells (data not shown). These results
provide evidence that p67 cDNA in the construct synthesizes p67
mRNA in the serum-starved cells. Using the Image Quant software,
the relative intensities of the Northern blot bands in panel
A were determined. They were as follows: lane 1, 1.0;
lane 2, 1.4; lane 3, 4.4; lane 4, 4.2; lane 5, 0.3; lane 6, 9.7; lane 7, 1.0, and lane 8, 12.78.
Immunoblot Analysis of p67 Protein
The p67 protein levels in the wild type and in the transformed cells were determined using standard immunoblot procedures (Fig. 4). The p67 protein is present in the KRC-7 cells grown in complete medium (lane 1). This protein level remained the same when these cells were treated with PMA (lane 2). The p67 protein level also remained essentially the same in the KRC-7 cells transformed with pORSVI-p67 without (lane 3) and with PMA (lane 4) addition. The p67 protein disappeared from the serum-starved KRC-7 cells (lane 5). However, this protein was present in the serum-starved cells transformed with pORSVI-p67 at a level comparable to that observed in the cells grown in complete medium (lane 7). Upon PMA addition, the p67 protein appeared in both wild type (lane 6) and the serum-starved cells transformed with pORSVI-p67 (lane 8). The level in both cases was essentially the same.
These results provide evidence that p67 cDNA synthesized p67 mRNA in serum-starved cells. This p67 mRNA synthesized p67 protein.
eIF-2The
cell extracts from the KRC-7 cells transformed with pORSVI-p67 under
different growth conditions were analyzed for eIF-2 -subunit
phosphorylation (Fig. 5). The same experimental conditions as described in Fig. 1 were used except that exogenous eIF-2 was not
added. As before, the eIF-2
-subunit phosphorylation was measured
using two different protein concentrations (100 and 200 µg). As shown
in panel A, upon expression of p67 cDNA in the
transformed cells, none of the cell extracts (the cells grown in
complete medium, lanes 1 and 2; serum-starved,
lanes 3 and 4; and mitogen-activated, lanes
5 and 6) showed any detectable eIF-2
-subunit
phosphorylation. The experiments described in panel B were
carried out under similar conditions as described in panel A
except that these cell extracts were preincubated with excess p67
antibodies to remove endogenous p67. The extent of eIF-2
-subunit
phosphorylation in all the cell extracts significantly increased and
were nearly the same at equal protein concentrations from each cell
extract. The eIF-2
-subunit phosphorylation in each case increased
similarly with increasing protein concentrations in each cell extract
(compare lanes 1, 3, 5 with 2, 4, 6).
These results show that the p67 protein synthesized in the
serum-starved cells by expression of p67 cDNA protected eIF-2
-subunit from eIF-2 kinase(s)-catalyzed phosphorylation.
We determined
the rates of protein synthesis in the wild type and in the transformed
cells under different growth conditions by in vivo labeling
with [35S]methionine (Fig. 6) for different
intervals as follows: 10 min (panel A), 20 min (panel
B), and 30 min (panel C). The radioactively labeled
proteins were then analyzed by SDS-PAGE and autoradiography. In all
cases, the rate of protein synthesis increased with increasing times.
The rates were nearly the same in the wild type cells (all panels, lane 1) and in the cells transformed with pOPRSVI-0
(all panels, lane 2). This rate increased significantly with
pOPRSVI-p67 (all panels, lane 3) grown in complete media.
Upon serum starvation the rates of protein synthesis of wild type cells
(all panels, lane 4) and the cells transformed with
pOPRSVI-0 (all panels, lane 5) were drastically reduced.
Upon expression of p67 cDNA in the serum-starved cells, this rate
was restored (all panels, lane 6) to the level observed in
the confluent cells (all panels, lane 1). Most of the
proteins synthesized in the wild type cells grown in complete medium
(all panels, lane 1) were now synthesized (all panels,
lane 6). Upon PMA addition to the serum-starved KRC-7 cells, the
protein synthesis rate was similarly restored in the wild type cells
(all panels, lane 7) and in the cells transformed with
pOPRSVI-0 (all panels, lane 8). This rate was significantly more increased in the cells transformed with pOPRSVI-p67 (all panels, lane 9). Using Image Quant software, the relative
intensities of a 38-kDA band in panel C were determined.
They were as follows: lane 1, 1.0; lane 2, 0.93;
lane 3, 1.24; lane 4, 0.05; lane 5, 0.04; lane 6, 0.95; lane 7, 1.46; lane
8, 1.3; and lane 9, 1.8.
Nutritional deprivation leads to a loss in protein synthesis
activity of the cells. Previous reports (2, 6) have indicated an
increased eIF-2 -subunit phosphorylation in cells grown in nutritionally deprived medium. We now provide evidence that an eIF-2
kinase remains in active form and essentially at the same level in the
cells grown in complete medium and in the serum-starved and
mitogen-activated cells. In the cells grown in complete medium and
mitogen-activated cells, this eIF-2 kinase does not phosphorylate eIF-2
as these cells contain p67. When the extracts were pretreated with p67
antibodies to remove endogenous p67, all the cell extracts similarly
phosphorylated eIF-2
-subunit (Fig. 1). Our results thus suggest
that the changing p67 levels in the cells are responsible for variable
phosphorylations observed with different cell extracts. The
serum-starved cells are mostly devoid of p67. Therefore, the extracts
from these cells efficiently phosphorylated eIF-2.
We investigated the mechanism of changes in p67 protein level. During heme deficiency in reticulocyte lysate (12), and during vaccinia viral infection of animal cells (13-14), the changes in p67 level involves activation of a latent p67 deglycosylase. The activated p67 deglycosylase deglycosylates p67 and inactivates it. In this paper, we provide evidence that although the p67 deglycosylase remains in latent form in KRC-7 cells, this p67 deglycosylase is not activated during serum starvation and is not responsible for p67 disappearance (Fig. 2).
We have previously reported (15) that both p67 mRNA and p67 protein almost completely disappear from the serum-starved cells. We now provide evidence that the disappearance of p67 mRNA is responsible for the loss in p67 protein and accompanying loss in protein synthesis activity in the serum-starved cells. By expressing a p67 cDNA, we synthesized p67 mRNA in the serum-starved cells (Fig. 3). The appearance of p67 mRNA restored p67 protein level in the serum-starved transformed cells (Fig. 4). Also, the protein synthesis rates of the serum-starved cells transformed with pOPRSVI-p67 was significantly restored. Several proteins almost completely disappeared from the serum-starved wild type cells (Fig. 6, lane 4). By expressing p67 cDNA in the same serum-starved cells, the synthesis of almost all of these proteins was restored (Fig. 6, lane 6). These results suggest that the loss in p67 mRNA in the serum-starved cells may be the main cause for the loss in protein synthesis activity. Transfection with pOPRSVI-p67 significantly increased protein synthesis rates in the confluent (Fig. 6, all panels, lane 3) and also in the PMA-activated serum-starved cells (Fig. 6, all panels, lane 9). We provide evidence that this increased rate is due to expression of p67 cDNA and was not observed when the cells were transfected with pOPRSVI-0.
Two possible mechanisms for the loss of p67 mRNA in the serum-starved cells can be envisioned. 1) Shut-off of p67 transcription and 2) activation of a nuclease. In our experiments, we synthesized p67 mRNA in the presence of p67 cDNA in the the cells grown in complete medium and then grew the cells in the serum-starved media. The p67 mRNA in the cells transformed with pOPRSVI-p67 increased significantly in the cells grown in complete medium and at different intervals after serum deprivation (data not shown). If serum starvation led to activation of a nuclease, it is expected that this nuclease would also have degraded p67 mRNA synthesized in the presence of p67 cDNA in the serum-starved cells. These results, therefore, suggest that the disappearance of p67 mRNA is due to shut-off of p67 transcription. The p67 transcription regulates the translation. The regulation of p67 transcription is the topic of the accompanying article (20).
We thank Prof. John W. B. Hershey (Davis, CA) for many valuable suggestion during the preparation of this manuscript.