Regulation of ornithine decarboxylase expression by anisosmotic shock in alpha-difluoromethylornithine-resistant L1210 cells.

Ornithine decarboxylase (ODC) activity is known to be strongly enhanced in mammalian cells by a sudden reduction in ambient osmolality. The effect of hypoosmotic shock on the regulation of ODC protein and mRNA levels was studied in a variant L1210 mouse leukemia cell line (D-R cells) which expresses ODC at greater than or equal to 100-fold higher levels than the parental cells. Hypoosmotic stress increased ODC activity in proportion with the osmotic gradient imposed to both D-R cells and their normal counterparts. A 60% decrease in medium osmolality increased ODC activity and the amount of immunoreactive ODC protein from 20- to 30-fold after 4 h without any detectable change in ODC mRNA contents in D-R cells. ODC induction was sustained up to 48 h after hypoosmotic shock, with maximal activity levels being observed at 24 h. Hypotonic shock dramatically increased (up to 36-fold) the rate of ODC synthesis as measured by 10-min pulses with 35S-labeled methionine, in agreement with kinetic constants predicted from the changes observed for the enzyme activity. Moreover, hypoosmotic stress extended the half-life of ODC activity from 35 +/- 10 to 212 +/- 67 min and blocked any degradation of the radiolabeled immunoreactive protein, which had a half-life of 28 +/- 6 min under isotonic conditions, for at least 120 min after addition of cycloheximide. The induction of ODC by hypoosmotic stress was quickly reversed by a sudden upshift of osmolality through a very rapid inhibition of ODC biosynthesis and an increase in the rate of enzyme degradation. Thus, hypoosmotic stress activates the expression of ODC exclusively through post-transcriptional mechanisms in D-R cells. The osmotically induced accumulation of ODC molecules is quite unique as shown by the fact that ODC is the major protein (approximately 25% of total) synthesized during the first 4 h following a 60% hypotonic shock, despite a 30-50% reduction of the rate of labeled precursor incorporation into soluble proteins.

mammalian ornithine decarboxylase (ODC),' together with S-adenosylmethionine decarboxylase one of the two key enzymes involved in the regulation of the de nouo biosynthesis of polyamines, is subject to tight control by the osmotic and ionic characteristics of the cellular environment. For instance, relatively small alterations in the Na+, K', or Mg*+ concentrations of tissue culture media have been reported to modulate the induction of ODC consecutive to the addition of fresh medium to stationary cell cultures (1). Moreover, ODC activity rapidly and dramatically increases following hypotonic shock in proportion to the osmotic challenge over a wide range of osmolalities in several mammalian cell types in culture (Z-6). This rapid induction of ODC activity by hypoosmotic stress is accompanied by an often dramatic accumulation of putrescine with little, if any, change in spermidine and spermine contents (2,4,5). A striking stimulation of the rate of putrescine (4,5) transport has also been reported, much in analogy with the phenomenon observed in Escherichiu coli (7,8). However, the mechanism responsible for the increase in ODC activity consecutive to hypotonic treatment has not been elucidated yet.
We have recently isolated a variant L1210 cell subline which stably expresses ODC to at least loo-fold higher levels than the parental cell line by selection for its resistance to growth inhibition by cY-difluoromethylornithine (DFMO), an enzyme-activated, irreversible inhibitor of ODC (9). This DFMO-resistant (D-R) cell line offers a distinct advantage for the study of the molecular biology of ODC expression because of its extraordinarily high ODC mRNA and enzyme contents. Moreover, ODC mRNA and protein present in these cells have biochemical characteristics undistinguishable from those of the original strain, and ODC expression is subject to the same post-transcriptional regulation by polyamines in D-R cells as in their normal counterparts (9,10). This model was thus chosen in order to elucidate the molecular mechanisms responsible for the effect of ambient osmolality on ODC expression. We report here that despite the exceptionally high basal levels of ODC present in these cells, striking amounts of enzymatically active ODC are nevertheless accumulated in a highly specific and reversible manner following a hypoosmotic shock through a marked stimulation of the rate of enzyme synthesis at a constant level of ODC gene transcripts and through a decreased rate of turnover.  (12) in an atmosphere of 5% CO,, 95% air at 95% relative humidity.
The DFMO-resistant (D-R) subline derived from the L1210 cell line (9)  Total radiolabel incorporation into cytosol proteins was measured as described previously (10). Osmolality of incubation media was determined by cryoscopy using a freezing point depression osmometer (Advanced Instruments).

RESULTS
Effect of Hypoosmotic Stress on ODC Activity-In all experiments described in the present study, serum, a well-known inducer of ODC (22), was omitted from the medium formulation to avoid repeated centrifugation of the cell suspensions necessary for subsequent protein analysis, since cells were found to become exceedingly fragile under hypoosmotic conditions. However, other medium components known to modulate ODC activity independently of the external osmolality such as amino acids using the A transport system (6,(23)(24)(25)(26) and glucose (27) were kept constant, the osmolality being manipulated by varying only the NaCl concentration.
The effect of hypotonic shock on ODC activity in L1210 cells and their DFMO-resistant counterparts was examined following a 4-h incubation in RPM1 1640 medium with osmolality adjusted from 130 to 325 mosm/kg with NaCl. As shown in Fig. 1, the level of ODC activity was inversely correlated with the osmolality in both L1210 and D-R cells. Thus, as osmolality was reduced from 325 to 130 mosm/kg, ODC activity measured at 4 h was increased by a factor of up to 40-and 20-fold in L1210 and D-R cells, respectively. The effect of diluting the medium with water to produce an identical decrease in osmolality yielded qualitatively similar results, although the induction was only about 5-fold in both cell lines (data not shown).
The time course of the induction of ODC activity by a 60% hypotonic shock in D-R cells is presented in Fig. 2. The most rapid phase of induction of enzyme activity occurred in the first 4 h and resulted in a g-fold increase over the initial ODC levels, with 50% of the increase being observed within 120 min following the onset of hypoosmotic stress. The differential effect of hypotonic treatment was, however, even larger (about 30-fold), because ODC activity decreased exponentially Cells were suspended in serum-free RPM1 1640 medium at 325 (0,O) or 130 mosm/kg (e, W), incubated for the indicated time period, and harvested for the determination of ODC activity. Two hours following the onset of hypoosmotic stress, NaCl (dissolved in serum-free medium) was added (arrow) to parallel dishes to bring the osmolality to 520 (0) or 325 (m) mosm/kg in cells preincubated at 325 (0)  by about 70% during that period in cells incubated under isotonic conditions. This decrease in enzyme activity observed in control cells was prevented by reintroducing the serum component (data not shown). Following this initial period of rapid induction, hypoosmotic conditions further sustained a slower increase in ODC activity up to about 24 h after the downshock. The enzyme activity (40,637 f 937 units/mg) measured at this stage represented a 270-fold difference with isosmotic conditions and a l7-fold increase over the initial ODC activity levels. More prolonged incubation (up to 48 h) of D-R cells under serum-free conditions led to the appearance of gross vacuolation and cell lysis under both osmotic conditions. Nevertheless, ODC activity was still induced 5-fold over the initial levels in D-R cells incubated for 48 h at 130 mosm/ kg. Furthermore, a sudden reversal of osmolality from 130 to 325 mosm/kg (Fig. 2, arrow) induced a rapid decrease in ODC activity in cells incubated previously for 2 h under hypoosmotic conditions. Enzyme activity levels had reached control values 6 h following the hypertonic shock. Likewise, ODC activity decayed even more rapidly in cells maintained for 2 h in isotonic medium and then exposed to an increase in osmolality of the same absolute magnitude (from 325 to 520 mosm/kg).
The rate of incorporation of L-[Yllmethionine into immunoreactive ODC was also dramatically increased in D-R cells incubated under hypoosmotic conditions, with a pattern similar to that seen for the enzyme activity in the initial phase of the induction (Fig. 3). Radiolabeling of ODC typically reached a plateau after 2 h of incubation in cells subjected to hypotonic treatment, probably as a result of a general decrease in the rate of intracellular precursor incorporation (CL Fig.  8A). Thus, the induction of ODC activity by hypotonic shock in D-R cells results from the rapid accumulation of de nouo synthesized enzyme molecules. Taken together, the present results further indicate that the intracellular levels of enzymatically active ODC are rapidly modulated in a sustained but reversible manner by osmotic stress, the direction of the observed variations being inversely related to changes in medium osmolality.
Effect of Hypoosmotic Stress on ODC mRNA Levels-A 60% hypotonic shock did not result in any detectable change in mRNA content at the end of a l-(results not shown) or 3-h incubation (Fig. 4), i.e. at periods of nearly maximal rate of increase in ODC activity levels (cf. Fig. 2). Thus, the increase in ODC activity induced by hypotonic shock in D-R cells is likely resulting from changes in the control of the enzyme expression at a post-transcriptional level.
Effect of Hypoosmotic Stress on the Rate of ODC Biosynthesis-The rate of ODC synthesis was determined during the rapid initial phase of induction of ODC by hypotonic shock by measuring immunoreactive ODC radioactivity at the end of lo-min pulses with L-[35S]methionine. Since the period of labeling was short as compared with the half-life of ODC under the present experimental conditions (cf. Figs. 6 and 7), the rate of incorporation of the labeled amino acid likely reflected mostly the rate of de rwuo synthesis. Although the relative degree of increase of the rate of ODC synthesis measured by this short pulse method was found to differ between experiments, biosynthesis of ODC was indeed strongly accelerated by hypoosmotic shock, the differential effect being greater at earlier time points following transfer to low osmolality ( Fig. 5A and Table I). A sudden reversal of osmolality from 130 to 325 mosm/kg decreased the apparent rate of enzyme synthesis to near control values within 10 min, whereas the transition from 325 to 520 mosm/kg practically abolished ODC synthesis (Fig. 5B). It must be noted, however, A, D-R cells were incubated for 3 h in serum-free RPM1 1640 medium at 325 (C, control) or 130 mosm/kg (H, hypoosmotic) and total RNA was isolated as described in text. Twenty micrograms of RNA was then analyzed by agarose gel electrophoresis, electroblotted on nitrocellulose, and hybridized with a 32P-labeled ODC cRNA probe. Arrowheads indicate the mobility of standard RNA of the given length (in kilobase). B, the given amounts of total RNA (isolated as described above) were blotted and UV-cross-linked on nitrocellulose and hybridized with a Y&labeled ODC cRNA probe as described in text. C, quantitation by liquid scintillation spectrometry of hybrid radioactivity from the dot blot shown in B. A, V, Analysis of RNA from cells incubated at 325 and 130 mosm/kg, respectively. that the rate of general incorporation of L-[35S]methionine into soluble proteins in these cells was even more severely inhibited by hypertonic (by 70 and 60% for cells initially at 325 and 130 mosm/kg, respectively) than hypotonic shock (29%) in these short pulse experiments. Nevertheless, strong indication that the high rate of ODC synthesis in hypotonitally treated D-R cells is indeed drastically suppressed upon hypertonic shock comes from the fact that the reduced halflife of the enzyme under those conditions was barely affected by the addition of cycloheximide (uide infra). Thus, the effect of a decrease in the ambient osmolality on the rate of ODC biosynthesis is extremely rapid and, apparently, quickly re- versible, thus reinforcing the notion that regulation of ODC expression by changes in osmolality is exerted at post-transcriptional steps.
Effect of Hypoosmotic Stress on ODC Degradation-The possible contribution of changes in the rate of ODC turnover in the accumulation of ODC in hypotonically treated D-R cells was investigated using two different approaches. First, cells were prelabeled for 15 min in isosmotic medium containing 0.16 pM of L-[35S]methionine as the sole source of the amino acid and then exposed to iso-or hypotonic conditions in the presence of 2 mM unlabeled L-methionine. Although radioactive methionine was not removed from the medium, such conditions closely approached those of a pulse-chase experiment as shown by the rapid disappearance (tw = 31 min) of immunoreactive radiolabeled ODC in control cells (Fig. 6A). On the other hand, the decay of ODC labeling over  9.2 a 4-h period had a half-life of 164 min in cells transferred to low osmolality medium, as measured by this method. However, the radioactivity of total soluble proteins also decreased at a much slower rate in pulse-labeled cells transferred to 130 mosm/kg (data not shown).
Since the effect of hypotonic shock on apparent general proteolysis suggested that the rate of dilution of the intracellular free pool of L-[35S]methionine might be impaired under hypoosmotic conditions, a second type of measurement of ODC degradation was obtained by exposing cells continuously labeled with L-[35S]methionine to 200 pM cycloheximide. Parallel determination of the rate of ODC degradation with cycloheximide yielded tXh values of 106 + 2 and 60 + 6 min at 130 and 325 mosm/kg, respectively (mean + S.D. of triplicate determinations from two independent experiments) when the drug was added at zero time of the incubation (Fig. 6B). On the other hand, when cycloheximide was added 120 min after osmotic shock, the intracellular content of radiolabeled ODC (Fig. 7A) as well as ODC activity levels (Fig. 7B) were remarkably stable. ODC activity had a half-life of 212 f 67 min in seven independent cell incubations at 130 mosm/kg, and extending the period of incubation with cycloheximide up to 3 h did not affect the measured half-life of the enzyme activity (tM = 224 min). No parallel degradation of the immunoreactive protein was measurable at 130 mosm/kg up to 120 min after addition of cycloheximide. During the same time interval, ODC had t*,> values at 28 + 6 min (n = two experiments) and 35 f 10 min (n = 7 experiments) under isosmotic conditions for the radiolabeled protein and enzymatic activity, respectively. A sudden hypertonic reversal to normosmotic conditions accelerated the decay rate of ODC activity and immunoreactive protein so that tl/, was now 69 and 160 min, respectively, in cycloheximide-treated cells preincubated for 2 h at 130 mosm/kg (Fig. 7, A and B). In the absence of cycloheximide, ODC activity decayed with a halflife of 73 min in cells submitted to reversal of hypoosmotic conditions (data not shown). Thus, the rapid decrease of ODC activity in cells shifted from 130 to 325 mosm/kg was likely caused by an acceleration of the degradation rate, in addition to a strong inhibition of ODC biosynthesis. However, the decay rate of ODC activity and immunoreactive labeled protein was not affected in cells shifted from 325 to 520 mosm/ kg.
These data thus suggest that a reversible stabilization of ODC against proteolytic mechanisms is a major factor contributing to its accumulation upon exposure of D-R cells to hypoosmotic shock. This decrease in ODC degradation rate is more important at later stages in the induction of the enzyme. Furthermore, an increase in ODC proteolysis also plays an important role in the suppression of the enzyme induction in cells exposed previously to hypotonic conditions and shifted to normosmotic medium. However, an increase in the degradation rate of ODC does not appear to participate, at least at early stages, in the dramatic decrease of the enzyme activity observed in cells subjected to a hypertonic shock when previously kept under isosmotic conditions. General Protein Synthesis-As shown in Fig. 8A, the average rate of L-[""S]methionine incorporation into total soluble proteins was reduced by up to 50% by a 4-h exposure of D-R cells to hypoosmotic conditions. This inhibition became mainly apparent in the second half of the incubation period and was also observed when lo-min pulses with L-[""S]methionine were performed at different time points (data not shown). The fact that the total radioactive ODC content also reached a plateau after about 2 h of hypoosmotic incubation (cf. Fig. 3) despite a demonstrable stimulation of the rate of  its biosynthesis at that stage ( Fig. 5 and Table II) thus suggests that hypotonic treatment leads to a reduced pool of labeled methionine available for protein synthesis. The densitometric pattern of cytosol proteins at the end of a 4-h period of labeling with L-[35S]methionine revealed a general decrease of about 30% in radiolabel incorporation with the exception of the ODC band, which represented 0.35 and 25% of total radiolabeled soluble proteins in control and hypotonically treated D-R cells, respectively (Fig. 8B). Thus, notwithstanding the fact that one-dimensional gel electrophoresis might leave minor proteins with similar properties undetected, these results indicate that the accumulation of ODC protein consecutive to hypotonic shock appears to be quite unique and that ODC becomes the major protein synthesized in D-R cells under hypoosmotic stress.

DISCUSSION
The present study provides evidence that in an ODCoverproducing cell line, hypoosmotic conditions induce a dramatic, sustained accumulation of enzymatically active ODC despite the presence of already high basal ODC levels. This increase is brought about by the combination of increased enzyme synthesis from a constant number of transcripts and decreased degradation of the enzyme. The lack of effect of hypotonic conditions on ODC mRNA content is consistent with the fact that actinomycin D did not affect the initial rate of induction of ODC activity elicited by hypoosmotic shock in other mammalian systems (2,4). At the peak of enzyme activity induced by a 60% decrease in osmolality (-40 pmol/ h/mg protein) in D-R cells, which is by far the highest level ever reported in a mammalian system, ODC represents about 1.3% of total soluble proteins, based on the estimated catalytic activity of homogeneous mouse kidney ODC (13). Although decreasing the NaCl concentration to lower the osmolality may affect several aspects of the cellular response by changing the electrochemical gradients, the accumulation of ODC induced by that treatment most likely results from an osmotic effect. In experiments not presented here, ODC induction was abolished for at least 4 h when the osmolality of media with a reduced NaCl concentration was kept constant by the addition of "impermeant" osmolytes such as mannitol, sucrose, or choline chloride, as found in other mammalian systems (2)(3)(4)(5)(6).
The stimulatory effect of hypoosmotic stress on translational activity appears to be restricted to the expression of ODC protein. We cannot rule out the possibility that an increased rate of ODC biosynthesis might result from the sparing of ODC mRNA from a general inhibition of the translatability of other mRNAs, assuming that the basal rate of ODC mRNA translation is largely limited by its competition with other messages for a common step in protein biosynthesis such as polysome formation (29). However, the time-dependent decrease of the rate of L-[35S]methionine incorporation into ODC as well as total soluble protein induced by hypoosmotic stress strongly suggests that the rate of entry of the precursor into the pool available for protein synthesis was progressively impaired. Indeed, the incorporation of other amino acids such as glycine and leucine into trichloroacetic acid-insoluble material is also reduced by hypoosmotic stress in Ehrlich ascites tumor cells (30) and cultured mouse mammary glands (4). These effects are most likely resulting from the decrease in the influx rate and increase of the efflux rate of amino acids observed following hypotonic treatment (31). It is likely that such effects on the kinetics of methionine transport had minimal effects on the initial rate of ODC labeling in the short pulse periods used for its determination.
Furthermore, any decrease in the transport rate of methionine would only underestimate the stimulation of the rate of ODC biosynthesis induced by hypoosmotic stress.
Measurements of the rate of ODC biosynthesis by the short pulse method are consistent with the kinetic analysis of changes in enzyme activity during the first 4 h of the time course (Table II). The rate of ODC degradation was seen to exceed the rate of its biosynthesis in cells kept under isotonic conditions, as expected from the gradual decay of the enzyme activity (Fig. 2). It is quite clear that the approximately 23fold increase in the rate of ODC synthesis was the main factor responsible for the enzyme accumulation induced by hypotonic shock. The rate of ODC biosynthesis at 120 min was only slightly lower than that measured at the onset of hypoosmotic stress, indicating that the contribution of a reduction in the rate of ODC proteolysis was mostly important at later stages of the induction of ODC activity. Furthermore, the estimated rate of ODC synthesis became extremely low in cells shifted from 325 to 520 mosm/kg, as confirmed by the pulse labeling experiments (Fig. 5B). On the other hand, since the absolute rate of decay of ODC activity (i.e. k' PO) was very high and apparently unaffected by cycloheximide as noted above, no valid calculation of the low rate of ODC synthesis could be performed from the activity data. Nevertheless, it is quite clear that a hypertonic shock quickly reduces ODC activity levels through a dramatic repression of ODC synthesis, in conjunction with a 3-fold increase in the absolute rate of degradation of the enzyme. The evidence presented here that hypotonic shock also results in the prolongation of the half-life of ODC in D-R cells is in accordance with previous reports in other mammalian systems (2,4). Since a significant, albeit submaximal increase in ODC half-life took place immediately following hypotonic shock (cf Fig. 6, B and C), the mechanism responsible for decreasing the rate of ODC degradation does not have an absolute requirement for new protein synthesis and is also effective in protecting ODC molecules pre-existing before hypotonic treatment against subsequent proteolysis. The slow reversibility, together with the time-dependent increase in the extent of stabilization of ODC by hypotonic shock, could indicate a partial requirement for the additional synthesis of unknown factor(s) involved in the control of ODC degradation.
The formation of a complex between ODC and its polyamine-induced antizyme has been postulated as a control point in ODC proteolysis (32). As discussed in earlier reports (9, lo), the induction of antizyme is unlikely to be a limiting step in ODC degradation in D-R cells since, as the half-life of ODC in these cells is close to that observed for the parental cell line, ODC amplification would then be expected to have occurred in concert with that of the antizyme itself. A more plausible regulatory step in the control of ODC degradation would be post-translational modifications of the enzyme such as phosphorylation (33)(34)(35). Alternatively, the capacity of the proteolytic system responsible for the degradation of ODC could approach saturation at the very high amounts of enzyme accumulated at later stages of ODC induction. Moreover, a possible stimulation of the overall rate of intracellular protein degradation resulting from a shift in substrate susceptibility to proteolysis induced by hypotonic treatment might lead to a decreased rate of ODC degradation if ODC is spared as a target by such a hypothetical effect. However, no evidence could be found for an increased rate of decay of radiolabeled, total soluble protein contents under hypoosmotic conditions ( Fig. 8A and data not shown). The reason for the discrepancy observed here between the decay rate of ODC activity and that of the immunoreactive labeled protein is not clear. The intriguing possibility that ODC activity might be affected by post-translational modifications in the absence of proteolytic attack, deserves further consideration.
The mechanism underlying the highly specific effect of hypoosmotic shock on the regulation of ODC enzyme levels is at present not understood.
However, there are several noteworthy similarities between this phenomenon and the effect of polyamines on the expression of ODC. It is now well established that spermidine, spermine, and, to a much lesser degree, putrescine specifically repress the translation of ODC mRNA (10,(36)(37)(38)(39)(40)(41)(42) and accelerate the degradation of the enzyme (10,(36)(37)(38)(39)(40)43) without detectable changes in the level of ODC gene transcripts (10,(36)(37)(38)(39)(40)(41)43). It is unclear how the hypoosmotically induced increase in ODC levels could be related to changes in polyamine contents. In D-R cells,' as well as in several other systems (2,4,5), there is little, if any, variation in the intracellular contents of spermidine and spermine following hypotonic treatment, whereas putrescine levels are dramatically increased as a consequence of ODC induction.
It should be pointed out, however, that relatively minor changes in polyamine content have been shown to trigger profound effects on the translation of ODC mRNA and on turnover of the enzyme in intact L1210 cells (43). As most vertebrate cell types possess some capacity for osmoregulatory volume decrease, mainly through the efflux of KC1 (reviewed in Refs. 44 and 45), the lower ionic strength conditions prevailing in cells recovering from a hypotonic shock are likely to thermodynamically affect polyamine activities so as to decrease their actual free concentrations.
Such a shift in intracellular polyamine activities might be evaluated by comparing the dependence of the rates of ODC mRNA translation and ODC enzyme degradation on the concentration of exogenously added polyamines under iso-and hypotonic conditions.
A most interesting parallelism can also be drawn between the induction of ODC by hypotonic shock and that caused by amino acids transported via the Na'-dependent A and N systems such as asparagine and glutamine (6,(22)(23)(24)(25). In rat hepatocytes, supraphysiological amounts of asparagine induce the accumulation of enzymatically active ODC through a severalfold stimulation of its synthesis and inhibition of its degradation without any change in ODC mRNA content (46). It is noteworthy that millimolar concentrations of amino acids transported through a Na'-dependent process characteristically induce cell swelling, consistent with their internalization, followed by a regulatory volume decrease through an enhanced rate of K' efflux (47)(48)(49)(50). In fact, it has been argued that under physiological conditions, the main function of osmoregulatory volume decrease in most mammalian cells is to antagonize cell swelling which would inevitably ensue rapid transport of solutes, especially when accumulation is driven by the electrochemical Na' gradient (47)(48)(49)(50). The rapidity of the effect of hypotonic shock on the post-transcriptional events controlling ODC levels demonstrated in this report and that described in the case of Na+-activated amino acid ' R. Poulin and A. E. Pegg, unpublished observations. transport, and possibly sugar transport (27), thus point to a mechanistic relationship, the most likely being the involvement of ionoregulatory and osmoregulatory phenomena in the reversible modulation of ODC expression.