Lack of Correlation between Extensive Accumulation of Bisnucleoside Polyphosphates and the Heat-Shock Response in Eukaryotic Cells*

The accumulation in large amounts of bisnucleoside polyphosphates (Ap4X) after heat shock in Xenopus Zuevis oocytes or cultured hepatoma cells (HTC cells) is observed after exposure to temperatures of 45 OC or higher. The accumulation is a transient phenomenon, with the collapse in cellular ATP concentration se- verely affecting the rate of synthesis of Ap4X, allowing degrading activities to empty the pool of these compounds under prolonged heat shock. This accumulation of Ap4X to high levels, compared to the basic content, is only observed under conditions leading to irreversible damage, ultimately resulting in the death of the cell. It is shown that the increase in Ap4X after hy- perthermia is due to the partial or almost complete inhibition of their degradation pathways, rather than to a stimulation of their rate of synthesis. Finally, the synthesis of heat-shock proteins could be observed un- der conditions which do not lead to important accumulation of Ap4X, therefore ruling out the possibility that these adenylylated nucleotides would behave as chemical signals ("alarmones") triggering the synthesis of heat-shock proteins. Nevertheless, on the basis of our earlier results (Guklon, G., Sovia, D., Ebel, J. P., Befort, D., and Remy, P. (1985) Embo J. 4, 3743-3749), it cannot be excluded that Ap4X might play

The accumulation in large amounts of bisnucleoside polyphosphates (Ap4X) after heat shock in Xenopus Zuevis oocytes or cultured hepatoma cells (HTC cells) is observed after exposure to temperatures of 45 O C or higher. The accumulation is a transient phenomenon, with the collapse in cellular ATP concentration severely affecting the rate of synthesis of Ap4X, allowing degrading activities to empty the pool of these compounds under prolonged heat shock. This accumulation of Ap4X to high levels, compared to the basic content, is only observed under conditions leading to irreversible damage, ultimately resulting in the death of the cell. It is shown that the increase in Ap4X after hyperthermia is due to the partial or almost complete inhibition of their degradation pathways, rather than to a stimulation of their rate of synthesis. Finally, the synthesis of heat-shock proteins could be observed under conditions which do not lead to important accumulation of Ap4X, therefore ruling out the possibility that these adenylylated nucleotides would behave as chemical signals ("alarmones") triggering the synthesis of heat-shock proteins. Nevertheless, on the basis of our earlier results (Guklon, G., Sovia, D., Ebel, J. P., Befort, D., and Remy, P. (1985) Embo J. 4, 3743-3749), it cannot be excluded that Ap4X might play a role in the regulation of the heat-shock response; this would, however, rely on variations in Ap4X concentrations which do not exceed a factor of 2.
Bisnucleoside polyphosphates, including &adenosine tetraphosphate (Ap4A),' have been shown to accumulate in prokaryotic cells following heat shock or oxidative stresses (Lee et al., 1983;Bochner et al., 1984). These compounds were proposed to be "alarmones" signaling to the cell the onset of a thermal or oxidative stress and triggering the heat-shock response (Lee et al., 1983). Recently, we showed that microinjection of Ap4A into Xenopus laevis oocytes does not lead to the synthesis of any heat-shock protein (hsp). On the contrary, Ap4A microinjection results in a strong or even complete inhibition of the synthesis of hsps after thermal stress, with the exception of the 70-kDa hsp (Guidon et al., 1985). These results led us to reinvestigate the relation between heat shock and accumulation of bisnucleoside polyphosphates in Xeno-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom reprint requests should be addressed. The abbreviations used are: Ap,A and Ap,A, 5',5'"-P1,P-diadenosine tetra-and triphosphate, respectively; Ap,X and Ap,X, 5',5'"bisnucleoside tetra-and triphosphates of the symmetrical or asymmetrical type, respectively; hsp, heat-shock protein; Hepes, 4-(2-hydroxyethy1)-l-piperazineethanesulfonic acid. pus oocytes as well as HTC cells (hepatoma tissue culture). This paper will show that an extensive accumulation of bisnucleoside polyphosphates in oocytes is only observed above 45 "C, whereas the synthesis of hsps has been shown to be optimal at 33 "C (Bienz and Gurdon, 1982). Likewise, Ap4X accumulation in HTC cells requires temperatures above 45 "C, whereas the heat-shock response is already observed at 43 "C. Furthermore, we observed that Ap4X accumulation after a severe heat shock results mainly from the irreversible denaturation of a specific hydrolase. Finally, Ap,A steady-state concentration in Xenopus oocytes was shown to depend strongly on internal ATP concentration. The biological meaning of variations in the pool of Ap4X (or Ap3X) should therefore be cautiously examined in the light of the variations in the level of their respective precursors, the nucleoside trior diphosphates.

Accumulation of Bisnucleoside Polyphosphates during Heat
Shock- Fig. 1 (upper) shows the variations of Ap4X content, as a function of time, in Xenopus oocytes submitted to heat shock at variable temperatures. As can be seen, the accumulation of bisnucleoside polyphosphates in large amounts only occurs above 45 "C. Up to 41 "C, the Ap4X concentration does not vary by more than a factor of 2; but as the mean variation in the measurements from sample to sample is pretty large, this factor may well be within the limits of the experimental uncertainty. When heat shock is carried out at 46 "C, a 8-10fold accumulation of Ap4X can be observed. If the temperature is further elevated to 50 "C, no significant accumulation of Ap4X can be observed any longer, most likely because the enzymatic system responsible for the synthesis of Ap4X is rapidly heat-inactivated. Likewise, the accumulation of AplX in large amounts in HTC cells requires an exposure of the cells at temperatures higher than 45 "C, as shown in Fig. 1  (lower). Even a l-h exposure at 43 "C only leads to a modification of Ap4X concentration by a factor close to 2.
Relations between Ap& and ATP Concentratwns-The variation in Ap4X concentration after heat shock is clearly biphasic, exhibiting an increase in the first 30 min (oocytes) or 60 min (HTC cells) of the thermal stress, followed by a decrease for longer incubations. This biphasic behavior probably arises from the collapse of the ATP cellular content, as which for ATP usually lie in the millimolar range. With ATP cellular concentration decreasing during heat shock from roughly 1 mM (in the oocyte) or several millimolar (in HTC cells) to a few tens of PM, it can easily be understood that the rate of synthesis of Ap,X will meanwhile drop markedly. Since the size of the Ap4X pool results from BR equilibrium between synthesis and degradation reactions (see below), a marked decrease in the rate of synthesis should lead to a decrease in the size of the pool. It should be noticed that the decrease in AprX concentration in HTC cells is only observed for much more severe conditions (1 h at 50 "C) than in the oocyte system, suggesting that the ATP-generating system is much more thermostable in the HTC cells, perhaps reflecting substantially different "working" temperatures (20 "C for the oocytes and 37 "C for HTC cells).
The fact that the Ap4X pool is dependent upon ATP concentration is further illustrated in Fig. 3, which shows that microinjection of the oocytes with ATP (2 mM final concentration) prior to hyperthermic treatment leads to the accumulation of Ap4X at higher levels than in control oocytes. It should be noted that even in the absence of heat shock (time 0 in Fig. 31, the Ap4X concentration is 5-6-fold higher when the oocytes are microinjected with ATP. Of course this dependent relationship between Ap4X and ATP pools requires that the biological meaning of variations in bisnucleoside polyphosphates be interpreted in light of ., mock-injected control oocytes. prolonged over 5 min, conditions which ensure the accumulation of Ap,X, irreversible damage starts to occur, as shown by the partial or almost complete lack of recovery of the ATP pool upon return to 18 "C. HTC cells behave similarly since 30 min of heat shock at 43 "C (allowing an efficient hsp synthesis) are accompanied by a 40% lethality, whereas 30-min exposures to 48 or 50 "C (favoring the Ap4X accumulation) lead to an almost complete lethality (Fig. 5 ) . This is also reflected by the complete collapse of [36S]methionine incorporation at 37 "C in newly synthesized proteins following heat shock at temperatures higher than 45 "C, as illustrated in Fig. 6.
We also observed similar results when studying the accumulation of AplX in Saccharomyces cereuisiae upon hyperthermia or Cd2+ treatment (Baltzinger et al., 1986). These results suggest that the accumulation of bisnucleoside polyphosphates in eukaryotic cells, in response to stress, is only observed under conditions irreversibly affecting the viability of the cells.
Does Accumulation of Bisnuckoside Polyphosphates Arise from Increase in Their Rate of Synthesis?-The recent observation that bisnucleoside polyphosphates could be accumulated in prokaryotes not only under hyperthermia, but also following exposure to a variety of oxidizing agents led Bochner et al. (1984) to propose that those compounds were alarmones signaling to the cell the existence of an oxidative stress. Aminoacyl-tRNA synthetases (or tRNAs) were proposed to be potential sensors ensuring coupling between the onset of oxidative conditions and an increased production of bisnucleoside polyphosphates. It was therefore of interest to investigate whether the accumulation of Ap4X results from an increased rate of synthesis or from the inhibition of the degradation pathway since degradation enzymes specific for Ap4X were described in a variety of cells: prokaryotic Plateau et al., 1985), as well as eukaryotic ( Barnes and Culver, 1982;Hohn et al., 1982;Ogilvie and Antl, 1983;Cameselle et al., 1984;Costas et al., 1984). As already stated, the rates of synthesis could not be measured by microinjection of radioactive ATP into the oocytes because the steady state of Ap4X labeling was reached within the time needed for microinjection. The experiments were thus carried out on oocytes extracts prepared as described under "Materials and Methods." Fig. 7 shows that no increase in the rate of synthesis of Ap4A can be detected following hyperthermic treatment of the oocyte, neither in the absence nor in the presence of reducing and complexing agents in the cellular extract. On the contrary, the rate of synthesis is slightly decreased (roughly by a factor of 2) upon prolonged incubation at 45 "C. This allows one to rule out a triggering mechanism based either upon oxidative modification of the competent aminoacyl-tRNA synthetases or upon the liberation of activating cations in the cytoplasm. It could be argued that we do not measure the true rate of synthesis of Ap,A due to the existence of both synthesizing and degrading activities in the cellular extract. This is not the case since the degrading activity in Xenopus oocytes is strongly inhibited after severe heat shock, as shown in Fig. 8; the upper panel shows the fate of radioactive Ap,A microinjected in oocytes at a final concentration of 2 mM. Obviously, Ap4A is degraded rapidly, the main degradation product formed being ATP, with the radioactivity then appearing more slowly in ADP and AMP, probably because of secondary reactions. As shown in the lower panel, this hydrolytic activity is not significantly affected by a 60-min treatment at 35 "C, conditions which allow an efficient heat-shock response (see below) but which do not lead to the accumulation of bisnucleoside polyphosphates (see above). On the contrary, incubation at 45 "C, which favors this accumulation, results in a marked inhibition of the Ap4A hydrolyzing activity since 15 or 45 min of exposure at this temperature lead to a 2-or 10fold reduction in this activity, respectively. It should be emphasized that this residual 10% activity appears to be rather thermostable and could correspond to nonspecific hydrolysis (by a phosphodiesterase-type enzyme).
The above observations show that the accumulation of bisnucleoside polyphosphates is the consequence of inhibition of their degradation pathway, rather than the result of stimulation of their biosynthesis.
Relations between Heat-Shock Response and Accumulation of Bisnucleoside Polyphosphutes-We therefore observed that the accumulation of Ap4X in large amounts in Xenopus oocytes as well as in HTC cells requires exposure to temperatures of 45 "C or higher. It is worthwhile to compare the above results to the temperature dependence of heat-shock protein synthesis. Bienz and Gurdon (1982) showed that hsp synthesis in the oocyte exhibits a sharp temperature optimum at 33 "C, an increase to 37 "C already leading to a drop by a factor of 2 in the rate of synthesis of the hsps. In the meantime, the overall protein synthesis drops from 50% to less than 10% of its maximum efficiency. Very similarly, Fig. 9 shows that the synthesis of hsps in HTC cells is already efficient at 43 "C.
Although the difference between the temperatures required to trigger the heat-shock response and the accumulation of Ap4X is less in HTC cells (2 "C) than in the oocytes system (12 "C), it is clear that high concentrations of Ap,X are not a prerequisite to hsp synthesis. The higher sensitivity of oocytes to temperature with regard to hsp synthesis is probably due to the fact that these cells have a normal working temperature around 20 "C, which means that exposure to temperatures above 30 "C may represent a pretty severe thermal stress.
The above results allow one to rule out the hypothesis that Ap4X could be chemical messengers triggering the synthesis of hsps as postulated by Lee e t al. (1983) on the basis that an Escherichia coli htpR mutant which is unable to synthesize hsps still accumulates Ap4X.

CONCLUSION
Our earlier results (Guedon et al., 1985) have shown that Ap4A does not trigger the synthesis of heat-shock proteins when microinjected in Xenopus oocytes. The present report shows that, in the two types of eukaryotic cells tested, hsp synthesis, in response to hyperthermic treatment, can occur under conditions where the intracellular content in Ap4X does not vary by more than a factor of 2. The accumulation of bisnucleoside polyphosphates at high levels, which has already been reported in prokaryotic cells under a large variety of stresses, occurs only under severe heat-shock conditions, apparently resulting in irreversible damage, ultimately leading to cell death.
Furthermore, the transient accumulation of Ap4X in the cell does not rely on an increased rate of synthesis, but on the inhibition of a specific degradation pathway.
These results do not support the idea that bisnucleoside polyphosphates could play a role of alarmones in the heatshock response, the synthesis of which would be enhanced upon oxidative damage to the competent aminoacyl-tRNA synthetases (or tRNAs), at least in eukaryotes.
The accumulation of Ap4X to large extents (compared to the basic levels) appears rather as a nonphysiological byphenomenon due to the irreversible thermal (or oxidative) inactivation of a degradation enzyme specific for bisnucleoside tetraphosphates. Our results do not allow one to completely rule out participation of bisnucleoside polyphosphates in the regulation of the response to an external stress since we showed that microinjection of Ap4A into Xenopus oocytes can lead to a decrease of hsp synthesis, with the exception of the 70-kDa hsp (Gubdon et dl., 1985). If they do play such a role, this will occur for cellular concentrations which, most likely, do not differ from the basal level by more than a factor of 2.  was r e c o w r e d and centrifuged for 5 minuter at 12000 ' pm i n a Janetlky bench top centrifuge.

MO
The supernatant was kept a t -80°C.