Inositol 1,4,5-trisphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear calcium signaling by inositol 1,4,5-trisphosphate. Discrete distribution of inositol phosphate receptors to inner and outer nuclear membranes.

Transient rise in nuclear calcium concentration is implicated in the regulation of events controlling gene expression. Mechanism by which calcium is transported to the nucleus is vehemently debated. Inositol 1,4,5trisphosphate (InsP3) and inositol-1,3,4,5-tetrakisphosphate (InsP4) receptors have been located to the nucleus and their role in nuclear calcium signaling has been proposed. Outer nuclear membrane was separated from the inner membrane. The two membrane preparations were, as best as possible, devoid of cross contamination as attested by marker enzyme activity, Western blotting with antilamin antibody, and electron microscopy. InsP4 receptor and Ca -ATPase were located to the outer nuclear membrane. InsP3 receptor was located to the inner nuclear membrane. ATP or InsP4 induced nuclear calcium uptake. External free calcium concentration, in the medium bathing the nuclei, determined the choice for ATP or InsP4-mediated calcium transport. We present a mechanistic model for nuclear calcium transport. According to this model, calcium can reach the nucleus envelope either by the action of ATP or InsP4. However, the calcium release from the nucleus envelope to the nucleoplasm is mediated by InsP3 through the activation of InsP3 receptor, which is located to the inner nuclear membrane. The action of InsP3 in this process was instantaneous and transient and was sensitive to heparin.


Transient rise in nuclear calcium concentration is implicated in the regulation of events controlling gene expression. Mechanism by which calcium is transported to the nucleus is vehemently debated. Inositol 1,4,5trisphosphate (InsP 3 ) and inositol-1,3,4,5-tetrakisphosphate (InsP 4 ) receptors have been located to the nucleus and their role in nuclear calcium signaling has been
proposed. Outer nuclear membrane was separated from the inner membrane. The two membrane preparations were, as best as possible, devoid of cross contamination as attested by marker enzyme activity, Western blotting with antilamin antibody, and electron microscopy. InsP 4 receptor and Ca 2؉ -ATPase were located to the outer nuclear membrane. InsP 3 receptor was located to the inner nuclear membrane. ATP or InsP 4 induced nuclear calcium uptake. External free calcium concentration, in the medium bathing the nuclei, determined the choice for ATP or InsP 4 -mediated calcium transport. We present a mechanistic model for nuclear calcium transport. According to this model, calcium can reach the nucleus envelope either by the action of ATP or InsP 4 . However, the calcium release from the nucleus envelope to the nucleoplasm is mediated by InsP 3 through the activation of InsP 3 receptor, which is located to the inner nuclear membrane. The action of InsP 3 in this process was instantaneous and transient and was sensitive to heparin.
The mechanism of regulation of nuclear calcium signaling is vehemently debated currently (1). Nuclear calcium signals control a variety of nuclear functions (2,3), including gene transcription (4,5), DNA synthesis, DNA repair, and nuclear envelope breakdown (6). It has been shown that nuclear vesicle fusion requires nuclear calcium mobilization (7). There are indications that nuclear and cytosolic calcium signals are differentially regulated and are independent of each other. For examples, nuclear calcium concentrations are higher than cytosolic ones in a number of cell systems (8 -12) or lower in smooth muscle (13) and neuronal cells (14). The source of nuclear calcium signal has been proposed to be cytosolic calcium stores (15), whereas clear evidence has come forward favoring a nucleocytoplasmic barrier to calcium (16) or the nucleus being insulated from large cytosolic calcium changes (14).
The nucleus contains an endoplasmic reticulum-type calcium pump ATPase (17), which attests to the ATP-mediated nuclear calcium uptake (18). The nucleus is endowed with functional inositol trisphosphate receptors (19,20) and the necessary machinery for InsP 3 1 production (21). The nucleus also contains inositol 1,3,4,5-tetrakisphosphate receptors (22), which mediates nuclear calcium entry in the presence of InsP 4 (23). All this information has been gathered on the basis of work on isolated rat liver nuclei.
Nuclear envelope (24) consists of an inner membrane and an outer membrane interrupted by nuclear pores (25). The outer nuclear membrane constitutes a continuum with the endoplasmic reticulum (26), whereas the inner nuclear membrane forms a membrane envelope for the nucleoplasm and is lined by nuclear lamina, which are the attachment sites for nuclear lamins (27). In the study presented in this paper, we isolated the two membranes of the nuclear envelope, as best as possible, free from cross contaminations from one and the other. The criteria for the purity of the inner and the outer membrane were the marker enzyme activity, electron microscopy, and Western blotting with antilamin antibodies. We document data demonstrating that the inner nuclear membrane preparation is right-side-in vesicular structure, which is the site for InsP 3 R location. In contrast, the outer nuclear membrane, as isolated here, consists of membranous sheaths with enriched mannose-6-phosphatase activity, devoid of immunoreactivity with antilamin antibodies, and is the site for location of high affinity InsP 4 R and Ca 2ϩ -pump ATPase. We also provide evidence that depending upon the external free calcium concentrations, ATP or InsP 4 mediates calcium transport to the nuclear envelope. Purification of Inner and Outer Nuclear Membranes of Nuclear Envelope-Rat liver nuclei were isolated as described earlier (28). Briefly, small pieces of liver were homogenized in 8 volumes of a medium containing 1.3 M sucrose, 1.0 mM MgCl 2 , and 10 mM potassium phosphate buffer, pH 6.8. The homogenate, after filtration through four layers of cheesecloth, was centrifuged for 15 min at 1000 ϫ g, and the resulting pellet was suspended in a minimum volume of homogenization medium. This suspension was mixed with a medium containing 2.4 M sucrose, 1.0 mM MgCl 2 , and 10 mM potassium phosphate buffer, pH 6.8, to give a final 2.2 M sucrose concentration and centrifuged for 1 h at 100,000 ϫ g. The resulting nuclear pellet was suspended in a medium containing 0.25 M sucrose, 4.0 mM MgCl 2 , and 20 mM Tris-HCl, pH 7.5, and centrifuged for 15 min at 1000 ϫ g. The pellet constituted the final nuclear preparation devoid of any microsomal, mitochondrial or plasma membrane contaminants (19,22,28) and was the starting material for the separation of inner and outer membrane of the nuclear envelope. Nuclei were suspended in a medium containing 0.25 M sucrose, 10 mM MgCl 2 , 1 mM dithiothreitol, 10 g/ml leupeptin, 2 mM phenylmethylsulfonyl fluoride, and 50 mM Tris-HCl, pH 7.4 (buffer A) at a protein concentration of 2 mg/ml. Sodium citrate, 1% (w/v), was added to this nuclear suspension, and it was incubated for 30 min on ice while stirring gently and centrifuged for 15 min at 500 ϫ g. The supernatant consisted of outer nuclear membrane, while the pellet contained the inner nuclear membrane. Earlier reports (29,30) have shown that treatment of nuclei with sodium citrate (1%) preferentially removed the outer membrane of the nuclear envelope without damaging the inner nuclear membrane. The pellet was suspended in buffer A at a protein concentration of 5 mg/ml and was digested with DNase 1 (250 g/ml final concentration) for 14 h at 4°C. The digested material was separated on a sucrose gradient (0.25, 1.6, and 2.4 M) after centrifugation for 2 h at 10,000 ϫ g in a SW 28 swinging bucket rotor. The inner membrane was recovered at the 0.25-1.6 M sucrose interface, recentrifuged for 20 min at 100,000 ϫ g and finally suspended in buffer A. The material recovered at the 1.6 -2.4 M sucrose interface constituted nucleolar and undigested nuclear material. The supernatant obtained after sodium citrate treatment (above) was centrifuged at 100,000 ϫ g for 20 min, resuspended in buffer A, layered over 1.6 M sucrose medium, and centrifuged as above. The membrane fraction collected at the 0.25-1.6 M sucrose interface was outer nuclear membrane. Both inner and outer nuclear membrane preparations were subjected to NAD pyrophosphorylase (31,32), mannose-6-phosphatase (30), and NADPH cytochrome c reductase activity (33).
Electron Microscopic Studies-Each of the inner and outer nuclear membrane preparation was suspended in a medium containing 0.12 M cacodylate buffer, pH 7.4, and 2.5% glutaraldehyde, incubated for 1 h at room temperature, and centrifuged for 10 min at 12,000 ϫ g. The pellet was post-fixed in 1% osmium tetraoxide in 0.12 M cacodylate buffer, pH 7.4 for 1 h, dehydrated with serial ethanol concentrations and propylene oxide, embedded in Spurr resin (Fluka, Switzerland) followed by thin sectioning.
The inner nuclear membrane was incubated with antilamin antibodies (1:100 dilution) for 3 h at room temperature. This was centrifuged for 15 min at 12,000 ϫ g. The resulting pellet was washed three times with a phosphate buffer saline (PBS). The final pellet was suspended in PBS, incubated with secondary antibody (anti-IgG labeled with colloi-dal gold, particle size 12 nm) at a 1:50 dilution, and incubated for 3 h at room temperature, washed three times as before, fixed, embedded, and thinly sectioned. In order to render antilamin antibody accessible to the site of antigen i.e. interior to the inner nuclear membranes, it was necessary to suspend the inner nuclear membrane in a low salt medium containing 10 mM HEPES, pH 7.5, 10 mM KCl, and 10 mM MgCl 2 . This was centrifuged and the pellet was treated as above. Electron microscopic examination was carried out with an EM 420 Phillips instrument.

H]InsP 3 and [ 3 H]InsP 4 Binding
Assays-Respective nuclear membranes were suspended in a medium containing 50 mM Tris-HCl, pH 8.0, and 2 mM EDTA for [ 3 H]InsP 3 (20 -60 Ci/mmol specific activity) binding. Standard [ 3 H]InsP 4 (21 Ci/nmol) specific activity) binding assay medium contained 10 mM MES-KOH, pH 6.5, 2 mM EDTA, 20 mM NaCl, and 120 mM KCl. The binding was done in an Eppendorf tube at 0°C (on ice) for 10 min in a 400-l final volume, and the protein concentration was maintained at 0.1 mg/assay. Bound and free radioligands were separated by centrifugation (Beckman Microfuge) for 5 min at 12,000 rpm, and the supernatant was removed by aspiration. Any remaining fluid was removed with tissue, and the pellet was suspended in 1 ml of tissue solubilizer (Soluene 350; Packard) plus 70 l of acetic acid and transferred to a 5 ml of Biofluor liquid scintillator, and the radioactivity was determined. Nonspecific binding was determined in the presence of 10 M InsP 3 or InsP 4

45
Ca 2ϩ Uptake Studies-Isolated nuclei or inner nuclear membranes were suspended in a medium containing 0.25 M sucrose, 2 mM EGTA, 2 mM EDTA, 4 mM K 2 HPO 4 , 4 mM MgCl 2 , 50 mM Tris-HCl, pH 8.0, 2.08 mM CaCl 2 in the absence of ATP and 2.14 mM in the presence of ATP (the free calcium concentration as calculated (34) was 1 M). Traces of 45 Ca 2ϩ were added (2 Ci/ml; final). 45 Ca 2ϩ movement was monitored at 37°C at indicated times and was terminated by filtration under vacuum over GF/B (Whatman) glass fiber filters, followed by rapid rinsing with 3 ml of ice-cold medium containing 0.25 M sucrose, 2 mM EGTA, 2 mM EDTA, 4 mM K 2 HPO 4 , 4 mM MgCl 2 , 50 mM Tris-HCl, pH 8.0. The filters were placed in 5 ml of Biofluor liquid scintillator, and the amount of 45 Ca 2ϩ trapped was determined. The amount of 45 Ca 2ϩ transported or released after a given time was defined as the radioactivity at time 0 minus the radioactivity trapped on the filters.
Western Blotting-Nuclear membrane preparations were electrophoresed employing SDS-polyacrylamide gel and transferred to nitrocellulose membrane (Amersham Hybond C-super) using a Transblot semi-dry transfer cell (Bio-Rad). Nitrocellulose membranes were incubated with a medium containing 50 mM potassium phosphate, pH 7.4, 0.2 M NaCl (PBS), and 3% bovine serum albumin for 2 h at room temperature. Subsequently, nitrocellulose strips were washed three times with PBS ϩ 0.05% Tween 20 and incubated with desired antiserum at appropriate dilution for 2 h at room temperature. The strips were washed three times with PBS ϩ 0.05% Tween 20 and incubated with a second antibody conjugated with alkaline phosphatase. After multiple washings with PBS ϩ Tween medium, the immunocomplex was revealed by alkaline phosphatase Misty Purple reagent (Amersham Corp.).

TABLE I Marker enzyme activity
Outer and inner nuclear membranes were separated from the nuclear envelope as described under "Experimental Procedures." NAD pyrophosphorylase activity was determined by monitoring spectrophotometrically the formation of NADH at 340 nm. Mannose-6-phosphatase activity was determined by measuring the absorbance at 730 nm. NADPH cytochrome c reductase was determined by monitoring the reduction of cytochrome c at 550 nm. Figures in parentheses represent the percentage. NAD pyrophosphorylase activity with the inner membrane was enriched coupled with decreased NADPH cytochrome c reductase activity as compared with the intact nuclei. Similarly, mannose-6-phosphate activity was enriched with the outer membrane as compared with the nuclei. This demonstrates the successful separation of the two membranes with as little as possible cross-contamination. These data are means of two independent preparations with replicates varying by Ͻ 10%.  (31), and NADPH cytochrome c reductase activity was measured by the reduction of cytochrome c as described (33). Protein was determined according to Bradford (35) with bovine serum albumin as a standard.

RESULTS
The use of a high molarity sucrose in liver homogenization medium and maintaining still higher sucrose concentration in subsequent centrifugation step resulted in a nuclear preparation that was devoid of any plasma membrane or microsomal contaminants. The question of microsomal-nuclear association during nuclei purification process was adequately checked by comparing nuclear preparation and final postnuclear materials from two sets of starting liver homogenate, i.e. normal liver homogenate from x number of rats and the homogenate supplemented with microsomes prepared from the same number of rats (for details, see Ref. 19). Fig. 2 illustrates well defined nuclear structure with intact nuclear pores (Fig. 2B).
Purity and Integrity of Inner and Outer Nuclear Membranes-The two nuclear membranes were scrutinized for marker enzyme activity. The specific activity of NAD pyrophosphorylase, a specific marker for nuclei (32), was enhanced in the inner nuclear membrane preparation (Table I). This indicated that NAD pyrophosphorylase activity may serve as inner nuclear membrane marker enzyme. Activity of NADPH cytochrome c reductase, a microsomal marker enzyme, was less than 2% of the total homogenate activity in the isolated nuclei (Table I). This shows that the isolated nuclei were not contaminated with microsomes. Hence, it may be argued that the outer nuclear membrane, as purified here, did not contain microsomes. Mannose-6-phosphatase activity has been reported to be present, although in different percentages, in both microsomes and nuclei (30). The specific activity of mannose-6-phosphatase was higher with the outer nuclear membrane preparation (Table I), which suggested that this enzyme is mainly associated with the outer membrane. In fact the inner nuclear membrane showed a marked decrease in mannose-6-phosphatase specific activity.
Western blotting with antilamin antibody revealed immunoreactive protein band exclusive to the inner nuclear membrane FIG. 2. Electron microscopic examination of nuclear membranes. A, the intact nuclei with nuclear envelope, nucleolus and nuclear material. B, nuclear pores indicated by arrow. C, outer nuclear membranous sheaths obtained after sodium citrate treatment followed by sucrose gradient centrifugation. D, the inner nuclear membrane obtained after sucrose gradient centrifugation of DNasedigested nuclear material. The vesicular nature of inner nuclear membrane is preserved (E) as in D, at higher magnification. F and G, inner nuclear membrane suspended in a low salt medium causing rupture of the membrane and accessibility of antilamin antibody to its target. The lamin on nuclear lamina was labeled with colloidal gold (dark points). Bar  Electron microscopic examination of inner nuclear membrane showed its vesicular nature (Fig. 1, D and E), whereas the outer membrane was represented by membranous sheath (Fig. 2C). The sidedness of the inner nuclear membrane was verified by immunoelectron microscopy using colloidal gold and antilamin antibody (Fig. 2, F and G). Lamin was spotted only on the inner side of the inner membrane (Fig. 2F) and only on one side of the membrane preparation (Fig. 2G). This tempted us to propose that the inner nuclear membranes, as purified, were right-side-in orientation. When the inner nuclear membrane was not treated with a low salt medium, antilamin antibody could not be accessible to the lamin lying on the nuclear lamina. No lamin was seen by colloidal gold immunochemistry, and the intact inner nuclear membrane vesicles as represented in Fig. 2E remained intact (data not shown) without low salt treatment. This provided additional support for the hypothesis that the inner membranes were vesicular in nature and were right-side-in.
InsP 3 R Located to the Inner Nuclear Membrane-Western blotting (Fig. 3) was performed with the inner and the outer nuclear membrane preparations. The 220-kDa immunoreactive protein was present only in the inner membrane and was not identified with the outer nuclear membrane preparation.
The specificity of InsP 3 receptor located to the inner nuclear membrane was analyzed by competitive displacement of [ 3 H]InsP 3 by various inositol phosphates (Fig. 4). InsP 3 at 6.5 nM inhibited 50% of the [ 3 H]InsP 3 binding and was the most potent inhibitor of binding (Fig. 4). Binding was also inhibited, although to a lesser extent, by the 2,4,5 InsP 3 isomer, InsP 4 (1,3,4,5), InsP 4 (3,4,5,6), and InsP 5 . The specific binding increased linearly with the amount of inner membrane protein added in the binding assay and heparin at 10 g/ml completely abolished [ 3 H]InsP 3 binding (data not shown). Scatchard plots (Fig. 5) of saturation isotherms showed a single class of binding sites with K d ϭ 5.5 nM and B max ϭ 1205 fmol/mg protein (Fig.  5A). K d (7.0 nM) and B max (385 fmol/mg protein) were found with the intact nuclear preparation (Fig. 5B). Almost no change in the K d values of InsP 3 R located to the inner nuclear membrane preparation as compared with the intact nucleus attested to the functional efficiency of the InsP 3 R in the inner nuclear membrane preparation. Moreover, increased B max values in the case of inner membrane indicated a rise in receptor density upon membrane purification. No meaningful [ 3 H]IP 3 binding activity was found with the outer membrane (data not shown).

FIG. 4. [ 3 H]InsP 3 binding to purified inner nuclear membrane.
[ 3 H]InsP 3 binding was carried out in an Eppendorf tube at 0°C (on ice) for 10 min in a medium containing 50 mM Tris-HCl, pH 8.0, and 2 mM EDTA. The inner membrane protein was 0.1 mg, and the total volume of assay medium was 400 l. The binding was terminated by centrifugation (Beckman Microfuge) at 12,000 rpm for 5 min, and the supernatant was removed by aspiration. The concentration of [ 3 H]InsP 3 was 1 nM plus various concentrations of competing ligands. Each assay was carried out in quadruplicate. Numbers in parentheses on the panel denote the respective isomer of various inositol phosphate derivatives. ing was sensitive to pH, pH 6.5 being optimum as observed earlier (22). Scatchard analysis of saturation isotherms (Fig. 7) revealed a high affinity binding site associated with the outer membrane. The outer membrane showed K d of 7.06 nM and B max of 3.6 pmol/mg protein (Fig. 7A), whereas in the inner membrane preparation the K d was 60.6 nM and B max ϭ 13.3 pmol/mg protein. These data indicate that the outer nuclear membrane was the site for high affinity InsP 4 binding, while the low affinity binding site was associated with the inner nuclear membrane (23).

FIG. 5. Scatchard plots of [ 3 H]InsP 3 binding to the inner nuclear membrane (A) and to the intact nuclei (B). The K d and
Ca 2ϩ -ATPase Immunoreactivity with the Outer Nuclear Membrane-A single immunoreactive (100 kDa) protein band (Fig. 8) was identified with the outer nuclear membrane showing the location of Ca 2ϩ -ATPase with this membrane preparation.
ATP Versus InsP 4 -mediated 45 Ca 2ϩ Transport to the Nucle-us-Isolated intact nuclei were incubated in a medium supplemented with calcium chloride so as to give indicated free calcium concentrations (34), and a trace amount of 45 Ca 2ϩ was added. ATP-mediated nuclear calcium transport was distinguished from the InsP 4 -mediated phenomenon depending upon the external free calcium concentration bathing the isolated nuclei. Free calcium levels in the vicinity of 0.5 M favored ATP-mediated uptake. The maximum uptake was around 1.0 M free calcium concentration. In sharp contrast, the action of InsP 4 eliciting nuclear 45 Ca 2ϩ uptake became evident only above 0.5 M free calcium levels, and higher free calcium levels (5-10 M) were more favorable for the nuclear calcium transporting process induced by InsP 4 (Fig. 10). 45 Ca 2ϩ Movement to the Inner Nuclear Membrane-To assess the functional relevance of InsP 3 R located to the inner nuclear membrane, the role of InsP 3 on 45 Ca 2ϩ movement was investigated. InsP 3 mediated 45 Ca 2ϩ transport to the inner nuclear membrane (Fig. 11A) within seconds after InsP 3 addition. The addition of ionomycin (Fig. 11B) in the medium bathing the  inner nuclear membrane preparation caused 45 Ca 2ϩ uptake into the inner membrane vesicles. The effect of ionomycin was 5-6-fold greater than that of InsP 3 . In the absence of InsP 3 or ionomycin, no meaningful calcium transport was seen and the action of InsP 3 was sensitive to heparin. ATP (18) or InsP 4 (23) were not able to transport 45 Ca 2ϩ to the inner nuclear membrane vesicles (data not shown).

DISCUSSION
It has been known for some time that calcium plays a crucial role in nuclear function; however, precise nature of nuclear calcium signaling has remained enigmatic. The existence of a separate and discrete control mechanism for calcium transport across the nuclear membrane has mainly relied on studies on isolated nuclei (18,19,23,36,37). Isolation of nuclei free from mitochondrial, plasma membrane, and microsomal contaminations has been carried out in many laboratories (23,25,36,37,38). InsP 3 R (19), InsP 4 R (23), and Ca 2ϩ -ATPase (17) have been located to the nucleus, but the distinct site of their location to the nuclear membranes has remained elusive. Nuclear envelope (27) consists of an outer membrane and an inner membrane, and the two nuclear membranes, although continuous, remain nonidentical in their surroundings, density, and protein compositions (30). We have succeeded in separating the two membranes of nuclear envelope, as best as possible, with-out cross-contamination. Three lines of investigation substantially favor this claim. First, the increase in NAD pyrophosphorylase specific activity (Table I) was associated with the inner nuclear membrane as compared with the intact nuclei, whereas specific activity of mannose-6-phosphatase and NADPH cytochrome c reductase was enhanced in the case of purified outer nuclear membrane. Considering the percentage of total activity of NAD pyrophosphorylase, about 26% of liver homogenate activity was seen with isolated nuclei and about 6% was found with the inner membrane (Table I). NADPH cytochrome c reductase activity (a marker for microsomes) was less than 2% with the intact nuclei and was almost negligible with the inner nuclear membrane. A differential distribution of the three marker enzymes associated with the purified inner and outer nuclear membranes attest to their separate entity. This was further confirmed by Western blotting analysis with antilamin antiserum (Fig. 1). The outer nuclear membrane showed no immunoreactive protein bands with antilamin antiserum. Immunoreactivity with antilamin antibody was only manifested with the inner nuclear membrane. A third line of evidence for the observed difference between inner and outer membrane is attested by electron microscopic investigation (Fig. 2). The outer nuclear membrane constituted membranous sheath like structures (Fig. 2C), whereas the inner nuclear membrane was well defined vesicular structures. Based on marker enzyme activity, Western blot analysis, and electron microscopic studies, we have reasons to propose that the inner and the outer membranes of the nuclear envelope were separated and purified, as best as experimentally feasible, free from contamina- FIG. 9. Inositol phosphate selectivity for nuclear 45 Ca 2؉ uptake. Isolated nuclei were incubated at 37°C for 5 min in the presence of various inositol phosphates at their indicated concentrations. Calcium chloride was supplemented into the medium bathing nuclei so as to give 1 M free calcium concentration according to Fabiato (34). Traces of 45 Ca 2ϩ (2.0 Ci/ml; 1 Ci ϭ 37 GBq) were present. Calcium uptake was terminated by filtering under vacuum over GF/B Whatman glass fiber filters, followed by rapid rinsing with 3 ml of ice cold medium. Filters were transferred to 5-ml Biofluor liquid scintillation vials, and the 45 Ca 2ϩ trapped on the filters was determined by spectrometry. Each experiment was carried out in quadruplicate. Inositol phosphates were: tions from one membrane to the other. We are also tempted to postulate that NAD pyrophosphorylase activity, initially reported as a nuclear marker enzyme activity (22,32), may serve as a marker for the inner nuclear membrane.
Concerning the sidedness of the inner nuclear membrane vesicles, the immunoelectron microscopic investigations were carried out employing antilamin antiserum with colloidal gold. Lamins were observed toward the inner side of the inner membrane, indicating that the inner nuclear membrane, as isolated here, were right-side-in. Another line of support for the intactness as well as right-side-in inner membrane vesicles is derived from the observation that colloidal gold immunoreactivity was not seen with the intact membrane preparation. This may be attributed to inaccessibility of antibody to the lamins lying on nuclear lamina. The lamins could only be identified when inner membranes were treated with low salt medium, which rendered the antilamin antibody accessible to nuclear lamina, a site for lamin attachment.
Specific [ 3 H]InsP 3 binding was associated with the inner nuclear membrane, which was sensitive to heparin, an accredited inhibitor of InsP 3 R (39). Of the various inositol phosphates examined, InsP 3 -(1,4,5) isomer was the most effective inhibitor of [ 3 H]InsP 3 binding to the inner nuclear membrane (Fig. 4). Scatchard analysis of saturation isotherms indicated a density of 1205 fmol/mg protein and a high affinity (K d 5.5 nM) for these sites (Fig. 4A). The receptor density in the inner nuclear membrane was 4 times higher as compared with the intact nucleus (B max , 385 fmol/mg protein) without altered K d (7.0 nM). This further attests to the efficiency of the procedure adopted for inner nuclear membrane separation from the outer membrane. The [ 3 H]InsP 3 binding data were further supported by Western blot analysis employing antibodies raised against rat brain type I InsP 3 R (Fig. 3). A distinct 220-kDa immunoreactive protein band was identified with the inner nuclear membrane, which was absent from the outer membrane preparation. The 220-kDa protein band was earlier identified as nuclear InsP 3 R and was a target for nuclear protein kinase C phosphorylation (20). It may be recalled here that the same antibody recognized 260-kDa protein band in rat cerebellar extract or rat microsomal fraction (20). The location of InsP 3 R to the nuclear membranes has been disputed. Patch-clamp studies indicate its location to the outer nuclear membrane (40). Data documented in this paper are in agreement with the inner membrane location of InsP 3 R in a confocal microscopic study (37). The inner membrane location of nuclear InsP 3 R also found support from the study demonstrating inhibition of nuclear vesicle fusion by blocking InsP 3 R activation (41). Lack of [ 3 H]InsP 3 binding to the outer nuclear membrane (data not shown) coupled with the absence of InsP 3 R immunoreactivity (Fig. 3) provides strong argument favoring that the InsP 3 R is located to the inner nuclear membrane.
Out of various inositol phosphates tested, InsP 4 (1,3,4,5) was the most effective and selective inhibitor of [ 3 H]InsP 4 binding to the outer (Fig. 6A) or inner nuclear membrane (Fig. 6B) preparation. The rank order of displacement of radioligand with various inositol phosphates was: InsP 4 (1,3,4,5), InsP 5 , InsP 6 , InsP 4 (3,4,5,6), and InsP 3 . The Scatchard analysis of saturation isotherms of the InsP 4 binding sites located to the outer membrane (Fig. 7A) gave K d of 7.06 nM and B max of 3.6 pmol/mg protein, corresponding to the high affinity binding site observed with the intact nuclei (23). The K d value was a bit higher than the K d (1.55 nM) observed with the intact nucleus. This may be attributed to the sodium citrate treatment and consequent formation of membranous sheathlike preparation. Likewise, the Scatchard analysis of the InsP 4 binding sites to the inner nuclear membrane gave a K d of 60.6 nM and B max value of 13.3 pmol/mg protein. The K d value obtained with the inner nuclear membrane corresponded with the low InsP 4 binding site observed with the intact nuclei (23). These results show that [ 3 H]InsP 4 binding sites are located to the outer as well as to the inner nuclear membranes, but they are not identical either qualitatively or quantitatively. The high affinity putative InsP 4 R seems to be located to the outer nuclear membrane. We have purified 2 nuclear high affinity InsP 4 R protein, which has a molecular mass of 74 kDa on silver-stained SDS-PAGE. The purified nuclear InsP 4 R gave a K d of 2.3 nM and actual B max of 11.2 nmol/mg protein. Proteins of multiple molecular mass have been purified as InsP 4 R from pig cerebellum (42), pig platelets (43), and rat cerebellum (44). Functionally, the outer nuclear membrane InsP 4 binding protein intervenes the action of InsP 4 in nuclear calcium transport (discussed below), whereas InsP 4 binding site located to the inner nuclear membrane does not seem to be involved in the nuclear calcium movement.
Based on the immunoreactivity employing endoplasmic reticulum anti-Ca 2ϩ -ATPase antiserum, it may be proposed that Ca 2ϩ -pump ATPase is located to the outer nuclear membrane and not to the inner nuclear membrane (Fig. 8). The exclusive location of Ca 2ϩ -pump ATPase to the outer nuclear membrane indicated that this enzyme is common to microsomes and nuclei. Lanini et al. (17) suggested that nuclear Ca 2ϩ -ATPase is identical to the endoplasmic reticulum Ca 2ϩ -ATPase. Recent studies (45,46,47) utilizing thapsigargin and 2,5-di-(tertbutyl)-1,4-benzohydroquinone have pointed out functional differences between nucleus and endoplasmic reticulum Ca 2ϩ -ATPase. Fig. 9 depicts selective nature of InsP 4 (1,3,4,5) isomer in mediating 45 Ca 2ϩ transport to the isolated intact nuclei. Other inositol phosphates, including InsP 5 , InsP 6 , or InsP 3 , did not elicit nuclear calcium transport. InsP 4 -or ATP-mediated calcium transport was only observed with the intact nucleus (23) and not with the inner nuclear membrane (data not shown). This has further provided support to the location and function of Ca 2ϩ -ATPase and InsP 4 R to the outer nuclear membrane.
Having established the location of Ca 2ϩ -ATPase and InsP 4 R to the outer nuclear membrane and accepting the role of ATP and InsP 4 in calcium transport to the isolated nuclei, the pertinent question addressed in this study was to distinguish between ATP-mediated and InsP 4 -mediated calcium transport to the nucleus (Fig. 10). According to the data depicted in Fig.  10, the external free calcium concentration in the medium bathing the nuclei is one of the factors that determines the mode of transport of calcium to the nucleus. For instance, free calcium concentration below 0.1 M may serve as a switch-on mechanism for the Ca 2ϩ -ATPase pathway by which ATP intervenes calcium entry into the nucleus. This pathway appears to be short circuited when the calcium levels exterior to the nucleus rises above 1 M. The InsP 4 -mediated nuclear calcium transport is favored above 0.8 M free calcium levels, and calcium levels above 1 M or up to 5-10 M seemed effective for InsP 4 -mediated calcium transport to the nucleus. Thus, it may be suggested that the mode of calcium transport to the nucleus depends on the calcium concentration in the external medium, which in the context of cell would account to cytosolic calcium variations (14,48,49). We do not know if InsP 4 R acts as an InsP 4 -gated calcium channel (50,51). We are tempted to speculate that InsP 4 may orient cytosolic calcium waves to the nucleus.
The inner nuclear membrane can serve as a model system to address the role of InsP 3 -mediated nuclear calcium transients since InsP 3 R is located to this membrane. In fact data are provided here demonstrating that InsP 3 -induced calcium release was primarily directed toward the inner nuclear space. The action of InsP 3 was instantaneous, transient, and sensitive to heparin (Fig. 11A). In the absence of InsP 3 , no meaningful calcium movement was observed, vindicating the idea the inner membrane may serve as a barrier to free calcium diffusion. Calcium was also transported to the inner nuclear space by the action of ionomycin. The action of ionomycin was also instantaneous and 5 times greater than that of InsP 3 . To quantify the concentration of calcium transported by InsP 3 or ionomycin, we determined the diameter (2.0 Ϯ 0.23 m), volume (4.18 Ϯ 1.6 m 3 ), and the number of inner membrane vesicles (100 ϫ 10 6 )/mg of protein. Based on these values, in our experimental conditions, 1000 nmol of calcium uptake/mg of protein was equivalent to 2.4 M (considering the added calcium concentration in the medium 2.08 mM). Accordingly, InsP 3 was able to deliver calcium into the inner nuclear space to a tune of 2 M. Likewise, ionomycin was capable of transporting 10 M calcium to the inner nucleus space. Here the action of InsP 3 was sensitive to heparin (39), which is additional evidence for the functional localization of InsP 3 R to the inner nuclear membrane. The transient nature of calcium movement into the inner nuclear space tempts us to speculate that the calcium diffuses out of nucleus through the nuclear pores.
In conclusion, we present a mechanistic model for calcium movement in and out of the nucleus. Ca 2ϩ -ATPase and InsP 4 R are located to the outer nuclear membrane and thus intervene in the action of ATP and InsP 4 , respectively, in eliciting calcium transport to the nuclear envelope. Calcium can be further released to the inner nuclear space by the activation of InsP 3 R (located to the inner membrane) by InsP 3 . Calcium can diffuse out from the nucleoplasm by the nuclear pores. This model presupposes the generation of InsP 3 within the nucleus, which awaits further studies, although the machinery to generate InsP 3 is present within the nucleus (21). Studies have begun to discern that InsP 3 is capable of controlling calcium within the nucleus (52).