An inositol 1,4,5-trisphosphate-sensitive Ca2+ pool in liver nuclei

Proc. Nati. Acad. Sci. USA Vol. 87, pp. 6858-6862, September 1990 Biochemistry An inositol 1,4,5-trisphosphate-sensitive Ca2+ pool in liver nuclei PIERLUIGI NICOTERA*t, STEN ORRENIUS*, THOMAS NILSSONt, AND PER-OLOF BERGGRENt Departments of *Toxicology and tEndocrinology, Karolinska Institutet, Box 60400, S-104 01 Stockholm, Sweden Communicated by Rolf Luft, May 23, 1990 ABSTRACT Recent studies in our laboratory have re- vealed the existence of an ATP- and calmodulin-dependent Ca2+ uptake system in rat liver nuclei that can promote increases in the free Ca2+ concentration in the nuclear matrix. In the present investigation we show that liver nuclei possess a marked ability to sequester and buffer Ca2+, suggesting a potential role for the nucleus in the regulation of the cytosolic free Ca2+ concentration. In addition, we demonstrate that the intracellular messenger, inositol 1,4,5-trisphosphate [Ins- (1,4,5)P3], stimulates the release of a fraction of the nuclear Ca2' and transiently lowers the intranuclear free Ca2+ con- centration. The Ins(1,4,5)P3-stimulated Ca2+ release is fol- lowed by Ca2+ reuptake into an inositol phosphate-insensitive nuclear compartment. Together, these results demonstrate that liver nuclei contain, at least, two Ca2+ pools, one of which is releasable by Ins(1,4,5)P3. These rmdings are consistent with a role for the nucleus in the modulation of the cytosolic free Ca2+ level by agonists and suggest that the control of the nuclear Ca2+ load by second messengers may participate in the regulation of intranuclear Ca2+-dependent processes by hor- mones and other agents. Ca2" signals generated by hormones, neurotransmitters, and growth factors are known to regulate many cellular functions. When cells interact with agents that stimulate the hydrolysis of inositol lipids, Ca2+ is mobilized from intracellular stores that are sensitive to inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] (1). The intracellular localization of the Ins(1,4,5)P3-sensitive Ca2+ pool is still debated. However, immunocytochemical studies, using an antibody that recognizes the Ins(1,4,5)P3 receptor, have revealed that this receptor is present in the perinuclear fraction of the endoplasmic reticulum (ER) and on the nuclear envelope (2). The intracellular distribution of Ca2+ after hormone stimu- lation, or after excitation of nerve or muscle cells, depends not only on the site of Ca2+ release but also on the seques- tration of Ca2+ into intracellular pools and on the opening of membrane channels (1). Thus, recent studies using video- imaging systems (3, 4) have shown that Ca2+ signals within single cells are not uniform and that, after stimulation, large responses are generated by the release of Ca2+ near or within the nucleus. We have recently reported (5) that, in liver nuclei, in- creases in the intranuclear free Ca2+ concentration are me- diated by a Ca2+ pump that is distinct from previously described Ca2+ translocases. The presence of an ATP- dependent Ca2+ uptake system in the nuclear envelope suggests that the intranuclear Ca2+ level can be regulated independently of cytosolic Ca2+ fluctuations and provides a basis for the potential modulation of intranuclear Ca2+ by second messengers known to affect Ca2+-dependent pro- cesses within the nucleus. Here, we report that liver nuclei have a high capacity to sequester Ca2+ and that intracellular messengers, such as Ins(1,4,5)P3, can release part of the Ca2+ accumulated by the nuclei, suggesting the existence of, at least, two nuclear Ca2l pools, one of which is sensitive to inositol phosphates. MATERIALS AND METHODS Isolation of Nuclei and Measurement of Ca2+ Sequestration. Nuclei were isolated from rat liver using a technique that yields a nuclear fraction virtually free of contamination by microsomal, mitochondrial, and plasma membranes (5). Af- ter isolation, nuclei were resuspended in an ice-cold TKM solution (50 mM Tris HCI, pH 7.5/25 mM KCI/4 mM MgCl2) and sedimented at 1000 x g for 5 min. The highly purified pellet was resuspended in incubation medium [125 mM KC1/2 mM K2HPO4/25 mM Hepes/4 mM MgCI2/2 mM EGTA, pH 7.0 (adjusted with KOH)]. The appropriate concentrations of Ca2' and EGTA required to achieve free Ca2+ concentrations ranging from 0.1 ,uM to 10 ,uM were determined and verified as described (5). To measure nuclear Ca2+ sequestration, 2 Al of the nuclear suspension (1.3 ,ug of dry weight per ml) was added to 25 IlI of incubation buffer (free Ca2+ concentration was between 2 and 10 ,M) supplemented with an ATP-regenerating system, consisting of 2 mM MgATP, 10 mM phosphocreatine, and 20 units of creatine kinase per ml. Although the preparation was virtually free of contaminating microsomes, as a further precaution, all experiments were performed in the presence of 2,5-di(tert-butyl)-1,4-benzohydroquinone (tBuBHQ), a po- tent and selective inhibitor of microsomal Ca2+ sequestration (6); tBuBHQ specifically inhibits the ATP-dependent seques- tration of Ca2+ into the ER and releases Ca2+ from this pool (7, 8), without affecting the nuclear Ca2+ uptake, which is mediated by a distinct, tBuBHQ-insensitive Ca2+ pump (5). Nuclei were maintained in suspension by a magnetic stirrer and Ca2+ concentration was measured by a Ca2+-selective minielectrode (9). Experiments were performed at room temperature and additions were made from 100 times con- centrated stock solutions, using constant volume pipettes (10). Calibration ofthe electrode was performed prior to each experiment. Fluorescence Measurements of Intranuclear Free Ca21 Con- centration. These measurements were carried out as de- scribed (5). Briefly, nuclei were preloaded with 7 puM of the fluorescent Ca21 indicator, fura-2 AM. Loading was per- formed at 4°C for 45 min. The nuclear fraction was then washed and resuspended in the incubation medium supple- mented with the appropriate amounts of EGTA and Ca2+ to give a free Ca2+ concentration of 100 nM (5). ATP (1 mM) was then added and fluorescence was monitored in a dual- wavelength fluorimeter (ZFP 22; Sigma), using the excitation pair 336-366 nm and the emission cutoff at 500 nm. Details of the fura-2 loading and the calculation of the intranuclear free Ca2+ concentration were given elsewhere (5). The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 6858 Abbreviations: Ins(1,4,5)P3, inositol 1,4,5-trisphosphate; ER, endo- plasmic reticulum; tBuBHQ, 2,5-di(tert-butyl)-1,4-benzohydroqui- none. tTo whom reprint requests should be addressed. Proc. Natl. Acad. Sci. USA 87 (1990) 6859 Fluorescence Microscopy. Fluorescence images of nuclei stained with ethidium bromide were obtained using a Nikon Diaphot microscope equipped with a fluorescence light source. The excitation filter was at 360 nm and the emission barrier filter was at 590 nm. Nuclei were photographed using a Nikon Fx35Va camera loaded with either Kodak Ekta- chrome 160 ISO (for fluorescence microscopy images) or Kodak T-Max 400 ASA (for phase-contrast images). Materials. Ins(1,4,5)P3 was purchased from Amersham. Inositol 1,3,4-trisphosphate and inositol 1,3,4,5-tetrakisphos- phate were kindly provided by Robin Irvine (Cambridge, U.K.). Fura-2 AM, ATP, phosphocreatine, and hexokinase were obtained from Sigma. All other reagents were of the highest grade ofpurity available and were obtained from local commercial sources. RESULTS AND DISCUSSION Nuclear Ca2+ Sequestration. We have previously described a method to isolate from rat liver a highly purified nuclear fraction that can transport Ca2+ (5). Phase-contrast imaging of the fraction showed that the preparation was free of whole cells (Fig. 1A). Staining with ethidium bromide further re- vealed that the isolated nuclei appeared intact and that the medium did not contain DNA fragments (Fig. 1B). Addition of the nuclear suspension to incubation medium, containing 2-10 p.M Ca2 , 10 A.M tBuBHQ, and an ATP-regenerating system, resulted in a rapid decrease in the extranuclear Ca2+ concentration to 120-400 nM (Fig. 2A). The ability of the nuclear fraction to sequester Ca2+ was considerable (1.7-2 pmol of Ca2+ per pug of dry weight) and was saturated after 10 mnu. For comparison, the amount of Ca2+ sequestered by FIG. 1. Phase-contrast and fluorescence microscopy of isolated liver nuclei stained with ethidium bromide. Isolated liver nuclei were resuspended in incubation medium supplemented with ethidium bromide (5 ,ug/ml). (A) Phase-contrast image of seemingly intact nuclei with granular structure (chromatin). (B) Fluorescence microscopy of the same field, demonstrating intranuclear fluorescence and the absence of nuclear fragments in the preparation. (x 100.) Biochemistry: Nicotera et al. 6860 Biochemistry: Nicotera et al. 500 E ~~~~~~~~~~~~~350- 6.0- + +~~~~~~~~~ (~~~~~)~~~~Hexokinasie L j6.51-_ 0 ~ 4mm100 L 1mm FIG. 2. Ca2+ uptake by the liver nuclear fraction and increase in the intranuclear free Ca2+ concentration. The isolated nuclear fraction was washed and resuspended in incubation buffer. (A) Two microliters of the nuclear suspension was added to 25 gl of the same incubation buffer supplemented with an ATP-regenerating system and 10 gM tBuBHQ. Nuclei were maintained in suspension by a magnetic stirrer and the Ca2+ concentration was measured by a Ca2+-selective minielectrode. The trace shown is typical of three different experiments, each performed on at least two different nuclear preparations. Glucose (30 mM) and hexokinase (20 units/ml) were added to remove ATP at the end of the experiment. (B) Nuclei were preloaded with the fluorescent Ca2+ indicator, fura-2 AM. The nuclear fraction was then washed and resuspended in the above buffer supplemented with 2 mM EGTA and the appropriate amount of Ca2+ to give a final free Ca2+ concentration of 100 nM. ATP (1 mM) was then added and the changes in fluorescence ratio were monitored. liver microsomes, in the absence of oxalate, is about 10-12 pmol/,ug of dry weight (11). Notably, studies in situ by Somlyo et al. (12) have shown that the average Ca2+ content in liver microsomes and nuclei is 5 pmol/,ug ofdry weight and 0.8 pmol/,ug of dry weight, respectively. Addition of glucose plus hexokinase to rapidly remove ATP from the incubation medium resulted in the release of the sequestered Ca2+ (Fig. 2A). Comparison of the time course of ATP-dependent Ca2+ sequestration by the nuclear fraction and the increase in intranuclear free Ca2" concen- tration, measured in nuclei loaded with fura-2, revealed that the latter occurred more rapidly and was saturated well before the capacity of the nuclear fraction to sequester Ca2+ had been exhausted (Fig. 2B). This suggests that most of the Ca2+ sequestered by the nuclear fraction was bound to intranuclear constituents, and that liver nuclei have a high Ca2+-buffering capacity. Furthermore, the observation that addition of glucose plus hexokinase to the incubation caused the release of most of the sequestered Ca2+ indicates that the bound Ca2+ can be easily mobilized and may therefore contribute to the regulation of the cytosolic free Ca2+ con- centration. Ins(1,4,5)P3-Stimulated Ca2l Release. The observation that Ca2' does not passively diffuse across the nuclear envelope (5) suggested that regulatory mechanisms exist to modulate intranuclear Ca2+ levels. Since it appeared unlikely that the availability of ATP should be the sole regulator of nuclear Ca2+ translocation [half-maximal affinity for ATP of the nuclear Ca2` uptake system being 75 ,uM (5)], we investigated whether other mechanisms were involved in modulating nuclear Ca2+ transport. The role of second messengers, such as Ins(1,4,5)P3, in the regulation of cytosolic Ca2+ concen- tration is well established, and it appears that a similar regulation of the intranuclear Ca2+ concentration may medi- ate the effects of agonists on intranuclear Ca2+-dependent processes (4, 13). As shown in Fig. 3A, the addition of Ins(1,4,5)P3 to Ca2+-loaded nuclei stimulated Ca2+ release (0.30-0.34 pmol of Ca2+ per ,ug of dry weight-i.e., 17-20%o of the sequestered Ca2+); this release was followed by Ca2+ reuptake. Repetitive additions of Ins(1,4,5)P3 did not result in further increases in the extranuclear Ca2+ concentration, showing that the effect of Ins(1,4,5)P3 was not due to Ca2+ contamination. Control experiments performed in the ab- sence of nuclei excluded the possibility that the Ca2+ fluc- tuations were due to interference of Ins(1,4,5)P3 with the Ca2+ microelectrode. To rule out the possibility that residual ER membranes, not fully inhibited by tBuBHQ, could con- tribute to the Ins(1,4,5)P3 effect, tBuBHQ was omitted from the incubation medium in some experiments. Under these conditions Ins(1,4,5)P3-mediated Ca2l release was identical to that measured in the presence of tBuBHQ (not shown), further strengthening the evidence that the source of the Ca2" release was entirely nuclear. Addition of Ins(1,4,5)P3 to fura-2- and Ca2+-loaded nuclei caused a small decrease in the intranuclear free Ca2+ concentration, followed by a rapid recovery (Fig. 3B). To investigate whether Ins(1,4,5)P3- stimulated Ca2+ release was dependent upon interaction with the Ins(1,4,5)P3 receptor, the nuclear fraction was pretreated with heparin, a known inhibitor of Ins(1,4,5)P3 binding to its receptor and of Ins(1,4,5)P3-induced Ca2+ release (14-16). As shown in Fig. 3D, heparin prevented the Ins(1,4,5)P3- stimulated Ca2+ release from the nuclear fraction, suggesting that a mechanism similar to that responsible for the release of Ca2' from the ER may be involved. Fig. 3C shows the control trace in the absence of heparin. The relationship between the Ins(1,4,5)P3 concentration and Ca2+ release is illustrated in Fig. 4. Half-maximal stimulation of Ca2' release was ob- served at 0.75-1 ,uM Ins(1,4,5)P3, whereas maximal stimu- lation was seen at 5 ,M Ins(1,4,5)P3. A characteristic property of the Ins(1,4,5)P3 receptor pres- ent in other subcellular fractions is that it does not undergo desensitization (1). Thus, the decrease in Ca2' release ob- served in this study may have been due to Ins(1,4,5)P3 metabolism (17) or Ca2' reuptake into a nuclear compartment that is insensitive to Ins(1,4,5)P3. Although we cannot ex- clude that rapid metabolism of Ins(1,4,5)P3 was responsible for the decline of the Ca2+ release, this seems unlikely since rapid, repetitive additions of Ins(1,4,5)P3 did not modify the reuptake of Ca2' by the nuclear fraction (cf. Fig. 3A). Further, the observation that only a fraction of the nuclear Ca2+ could be released by Ins(1,4,5)P3 suggests the existence of at least two pools of Ca2+ in the nuclei, one of which is insensitive to Ins(1,4,5)P3. To investigate the specificity for Ins(1,4,5)P3, we used inositol 1,3,4-trisphosphate, which has a low affinity for the Ins(1,4,5)P3 receptor and lacks Ca2+-mobilizing properties (18), and inositol 1,3,4,5-tetrakisphosphate. Both inositol 1,3,4-trisphosphate and inositol 1,3,4,5-tetrakisphosphate were unable to mobilize Ca2+ from the nuclear fraction (not shown). The mobilization ofCa2' by Ins(1,4,5)P3 was not due to contaminating microsomes or other organelles. Two lines of evidence support this contention. (i) The activities of mitochondrial, microsomal, and plasma membrane marker enzymes found in the preparation were negligible. (ii) The Proc. Natl. Acad. Sci. USA 87 (1990) Proc. Natl. Acad. Sci. USA 87 (1990) 6861 6.5 I E_____ ~~~~~~~~~~~~~100L,o 4 min ki 4mm ~~~ ~~~~~~~ATP1m N Nuclei acj~ Ca2 6.0 eNucle parin Ca2+concntraion as timuatedby 1mM AP.ns(1,4,5)P 3 5,M a de ttetieidctdb h ro. C oto rc o h 6.5- 1ns(1 ,4,5)P., 4min 4min FIG. 3. Ins(1,4,5)P3-induced Ca2+ release in liver nuclei. (A) Experimental procedures were identical to those illustrated in the legend to Fig. 2. When indicated by arrows, 5 AtM Ins(1,4,5)P3 Was added. (B) Nuclei were preloaded with fura-2 AM and increase in the intranuclear free Ca2+ concentration was stimulated by 1 mM ATP. Ins(1,4,5)P3 (5 MLM) was added at the time indicated by the arrow. (C) Control trace for the trace in D. Arrows indicate the addition of 5 ,uM Ins(1,4,5)P3 and calibration with 0.125 nmol of Ca2+. (D) Heparin (100 Ag/ml) was added prior to the addition of 5 MAM Ins(1,4,5)P3. Calibration was identical to that shown in the trace in C. incubation medium contained 10 AM tBuBHQ, which selec- tively inhibits the microsomal Ca2+ pump (6) and the Ins(1,4,5)P3-stimulated Ca2+ release from the ER (7). In contrast, tBuBHQ does not inhibit nuclear Ca2+ uptake and does not promote Ca2+ release from isolated nuclei (ref. 5 and this study). Ins(1,4,5)P3 receptors were recently identified on the nu- clear membrane and in the ER in the perinuclear region of cerebellar Purkinje cells (2). Therefore, it was suggested that Ins(1,4,5)P3-induced Ca2+ release is not a property of a single organelle but is distributed in specialized regions of several intracellular membranes. The presence of Ins(1,4,5)P3 recep- tors on the nuclear envelope, together with our findings, support this idea. At the present time, it is not known where 100 E C E I+/ 0 ..-50 bi(I) bi 0 ~~~~~~2.55.0 Ins(1,4,5)P, (AM) FIG. 4. Relationship between Ins(1,4,5)P3 concentration and Ca2+ release. The nuclear suspension was incubated with increasing Ins(1,4,5)P3 concentrations, and Ca2+ release into the incubation medium was monitored as reported in the legends to Figs. 2 and 3. Results are typical of three replicates performed on at least two different nuclear preparations. Maximal Ca2+ release was 0.34 pmol/Ag of dry weight. the sites of Ca2l transport in and out of the nucleus are located. The only structures known to span the two nuclear membranes are the nuclear pores. Thus, it seems likely that at least the Ca2+ uptake system, which promotes increases in the intranuclear free Ca2 , is associated with the pore com- plex. Conversely, the observation that only part of the Ca2+ sequestered by the nuclei can be released by Ins(1,4,5)P3 suggests that the inositol phospholipid-sensitive Ca2+ pool may be restricted to the space between the nuclear mem- branes. According to this model, a recent study using con- focal microscopy has suggested that the Ca2' rise observed in the regions surrounding the nucleus during nerve cell stimulation is likely to originate from the nuclear envelope itself (4). If so, this may provide an explanation for the modest decrease in the intranuclear free Ca2+ concentration caused by Ins(1,4,5)P3, although we cannot yet exclude that the relatively small change is the result of rapid intranuclear Ca2+ buffering (due to release of Ca2+ from bound sites and/or Ca2' resequestration into the nucleus). It remains to be established whether the Ins(1,4,5)P3- stimulated release of nuclear Ca2+ may contribute to the elevation of cytosolic Ca2+ produced by Ca2+-mobilizing hormones and growth factors. The observation that Ins(1,4,5)P3-generating agonists produce a Ca2+ rise that originates in the proximity of the nucleus, where the respon- sive Ca2+ store seems to be associated with a 140-kDa Ca2+-ATPase-like protein (3), suggests that the nuclear Ca2+ released by Ins(1,4,5)P3 may be involved in the physiological response to hormones. Additionally, the Ins(1,4,5)P3- stimulated release of Ca2+ from the nucleus may function as a local regulatory signal to control Ca2+ load in the nucleus during cell activation. A possible role for nuclear Ca2+ in the regulation of agonist-induced cytosolic Ca2+ fluctuations remain to be elucidated by further studies. This study was supported by grants from the Swedish Medical Research Council (03X-2471 and 19X-00034), the Bank of Sweden Tercentenary Foundation, and Fondazione Clinica del Lavoro Isti- Biochemistry: Nicotera et al. 6862 Biochemistry: Nico