Calreticulin–melatonin An unexpected relationship Manuel Macı ´ as 1 , Germaine Escames 1 , Josefa Leon 1 , Ana Coto 3 , Younes Sbihi 2 , Antonio Osuna 2 and Darı ´ o Acun ˜ a-Castroviejo 1 1 Departamento de Fisiologı ´ a and 2 Instituto de Biotecnologı ´ a, Universidad de Granada, Spain; 3 Departamento de Morfologı ´ a y Biologı ´ a Celular, Facultad de Medicina, Universidad de Oviedo, Spain Increasing evidence suggests that melatonin can exert some effect at nuclear level. Previous experiments using binding techniques clearly showed the existence of specific melatonin binding sites in cell nucleus of rat liver. To further identify these sites, nuclear extracts from rat hepatocytes were treated with different percentages of ammonium sulfate and purified by affinity chromatography. Subsequent ligand blot analysis shows the presence of two polypeptides of  60 and  74 kDa that bind specifically to melatonin. N-Terminal sequence analysis showed that the 60 kDa protein shares a high homology with rat calreticulin, whereas the 74 kDa protein shows no homology with any known protein. The binding of melatonin to calreticulin was further charac- terized incubating 2-[ 125 I]melatonin with recombinant calreticulin. Binding kinetics show a K d ¼ 1.08 ± 0.2 n M and B max ¼ 290 ± 34 fmolÆmg protein )1 , compatible with other binding sites of melatonin in the cell. The presence of calreticulin was further identified by Western blot analysis, and the lack of endoplasmic reticulum contamination in our material was assessed by Western blot and immunostaining with anti-calnexin Ig. The results suggest that calreticulin may represent a new class of high-affinity melatonin binding sites involved in some functions of the indoleamine including genomic regulation. Keywords: affinity chromatography; calreticulin; melatonin; nuclear receptor purification; receptor binding. Melatonin is a highly preserved molecule throughout phylogeny. It appears in very ancient unicellular organisms [1], remaining unchanged in multicellular species including humans [2]. In mammals, the circadian rhythm of melatonin is produced through a photoperiodic-dependent synthesis by the pineal gland [3]. In turn, melatonin translates photoperiodic information from clock and calendar mes- sages, acting as an endogenous synchronizer of several endocrine and nonendocrine rhythms [3]. This indoleamine is also produced by a variety of other tissues [4]. Melatonin exerts important regulatory influences on reproduction [5], and on neuroendocrine [6] and immune systems [7]. Moreover, it also controls cellular proliferation through regulatory effects on cell cycle kinetics [8], and prevents apoptosis in several tissues [9]. Recent studies have also focused on the antioxidant and free radical scavenging properties of melatonin [10–13]. Except for the antioxidant, nonreceptor-mediated effects of melatonin, the actions of the indoleamine suggest the existence of specific receptors in the cell. Three related, but distinct high affinity G i -protein-coupled melatonin receptor subtypes have been cloned [14–17]. Membrane receptors for melatonin are now classified as mt 1 ,MT 2 and MT 3 .In addition, biochemical and immunocytochemical studies in different mammalian tissues have shown the presence and accumulation of melatonin in the cell nuclei [18]. This nuclear localization of melatonin can be related to its described genomic effects including the regulation of the mRNA levels for antioxidant enzymes and the inducible isoform of nitric oxide synthase (iNOS) [19,20]. So far, no responding gene could be directly linked to the activation of membrane receptors by melatonin. However, a synergistic effect of S 20098 and CGP 52608, two selective agonists of the membrane and nuclear melatonin receptors, respect- ively, on interleukin (IL)-6 production by human mononu- clear cells has been shown [21]. Thus, to explain the nuclear actions of the indoleamine it is reasonable to assume the existence of a receptor in the nucleus of the cell. Previous studies with [ 3 H]melatonin showed the existence of specific nuclear binding sites for melatonin [22]. Using 2-[ 125 I]melatonin, the nuclear receptors for melatonin were fully biochemical and pharmacologically characterized [23,24]. The identification of melatonin as a ligand for the ROR receptors [25,26] allowed its classification as a nuclear effector. This viewpoint was further supported when it was found that the nuclear receptor for melatonin represses 5-lipoxygenase gene expression in human B lymphocytes [27]. Thus, it is reasonable to assume that nuclear melatonin Correspondence to D. Acun ˜ a-Castroviejo, Departamento de Fisiologı ´ a, Avenida de Madrid 11, E-18012 Granada, Spain. Fax: + 34 958 246295, Tel.: + 34 958 246631, E-mail: dacuna@ugr.es Abbreviations: ER, endoplasmic reticulum; ERa, estrogen receptor alpha; IL, interleukin; iNOS, inducible isoform of nitric oxide synthase; NAS, N-acetylserotonin; PAP, peroxidase– antiperoxidase; 4-P-PDOT, 4-phenyl 2-propionamidotetraline; GST, glutathione S-transferase. (Received 4 September 2002, revised 19 November 2002, accepted 16 December 2002) Eur. J. Biochem. 270, 832–840 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03430.x signaling is a basic mechanism for the various control functions of the indoleamine. The present work describes the purification and charac- terization of two proteins from nuclei extracts of rat hepatocytes that may represent a new class of melatonin receptors. These polypeptides, with molecular masses of 74 and 60 kDa, were purified by ammonium sulfate precipi- tation and affinity chromatography; characterized by SDS/ PAGE and Western blotting, and identified by their N-terminal amino acid sequence. The search for sequence similarities in protein databanks showed that the 60 kDa protein is highly homologous to calreticulin, whereas the 74 kDa protein is a novel, unidentified protein. Materials and methods Materials All reagents were of the highest purity available. Antibody against melatonin (G/S/7048483) was obtained from Stock- grand Ltd. (UK). Tris, sucrose, deoxyribonuclease (DNase), ribonuclease (RNase), ammonium sulfate, EDTA/Na 2 , phenylmethylsufonyl fluoride, leupeptin, pepstatin A, Triton X-100, Tween 20, melatonin (N-acetyl-5-methoxy- tryptamine), 6-hydroxymelatonin, N-acetylserotonin (NAS), goat anti-rabbit immunoglobulin, peroxidase–antiperoxi- dase complex, 3,3¢-diaminobenzidine tetrahydrochloride, and all other chemicals and dyes were purchased from Sigma-Aldrich Quı ´ mica (Madrid, Spain). 4-P-PDOT (4-phenyl 2-propionamidotetraline) and luzindole (2-benzyl- N-acetyptryptoamine) were obtained from Tocris Cookson Ltd (Bristol, UK). Isolation of liver nuclei and nuclear protein fractionation Nuclei were isolated from rat liver by the procedure described elsewhere [28], with some modifications [24]. Briefly, rat livers (0.9–1.2 g) were individually homogenized in 3 mL of buffer A(10m M Tris/HCl, 0.3 M sucrose, pH 7.4) with a glass/ Teflon homogenizer (10 strokes) and layered over 3 mL of buffer A containing 0.4 M sucrose. The samples were then centrifuged at 2500 g for 10 min. The pellet was gently resuspended without vortexing in 1 mL of buffer B (50 m M Tris/HCl, pH 7.4) and centrifuged again. All procedures were carried out at 4 °C. The pellet containing pure, intact nuclei was checked by electron microscopy for the quality of the isolated nuclei (data not shown) [47]. The purified nuclei were resuspended in 1 mL of buffer B and disrupted with a Polytron homogenizer (7 s, set point 9) to obtain a crude nuclear extract. The crude nuclear extract was further purified by resuspension in buffer B containing protease inhibitors (5 m M EDTA-Na 2 ,0.1m M phenyl- methylsufonyl fluoride, 20 l M leupeptin and 2 l M pepstatin A) and 0.1% Triton X-100. The homogenate was incubated with DNase and RNase (25 lgÆmL )1 ), for 1 h at 37 °Cand centrifuged to 48 000 g for 20 min at 4 °C. The proteins in the supernatant (Triton X-100 soluble fraction) were preci- pitated with different concentrations of ammonium sulfate solutions (0–25, 25–45, and 45–65%) [29]. After gentle stirring in an ice bath for 30 min, precipitate was collected by centrifugation at 48 000 g for 30 min and resuspended in buffer B. The resuspended sample was desalted in a Sephadex G-25 gel filtration column (Amersham Pharmacia Biotech Europe GmbH, Barcelona, Spain) pre-equilibrated with buffer B containing 50% glycerol, and stored at ) 20 °C. Affinity chromatography by melatonin–agarose The hydroxyl group of 6-hydroxymelatonin was coupled to the epoxide group of epoxy-activated Sepharose 6B (Amer- sham Pharmacia Biotech Europe GmbH) [30] to yield a resin, designated melatonin–agarose (Fig. 1). Briefly, epoxy- activated Sepharose 6B was swelled and washed extensively with deionized water, and 6-hydroxymelatonin was dis- solved in freshly prepared 50% dioxane/50% 0.1 M sodium phosphate buffer, pH 8.7 (at higher pH values, 6-hydroxy- melatonin was not stable). The 6-hydroxymelatonin solu- tion was then mixed with the activated resin and sealed under nitrogen. The mixture was agitated in darkness for 48–72 h at 32 °C. Ligand excess was eliminated washing the gel with 50% dioxane, followed by bicarbonate (0.1 M , pH 8.0) and acetate (0.1 M , pH 4.0) buffers. Unreacted epoxide groups were then blocked by incubating the resin with 1 M ethanolamine for 16 h at 30 °C. The resin was extensively washed with deionized water, and then with bicarbonate and acetate buffers containing 500 m M NaCl and 0.05% digitonin. The resin was resuspended in an equal volume of 0.05% digitonin and stored at 4 °C. Crude nuclear extracts and solubilized material were separately incubated with melatonin–agarose (15 : 1 v/v) for 16–24 h at 4 °C under gentle agitation (100 r.p.m. on an orbital shaker). The resin was then pelleted (1500 g)and washed with 10 volumes of 0.05% digitonin until protein was no longer detected in the final washed solution. The resin was specifically eluted by incubating it with melatonin (10 l M ; one volume) for 6 h to 4 °C with moderate agitation. To remove excess of ligand, the eluate was put over Sephadex G-25 columns pre-equilibrated with 0.05% digitonin. The final eluate was lyophilized and stored to )20 °C until electrophoresis analysis. Protein content in each step of purification was measured using a Bio-Rad protein assay reagent, with bovine serum albumin as protein standard [31]. SDS/PAGE and ligand blotting Samples obtained from affinity column were electropho- resed on 12.5% SDS/PAGE [32] using the PhastSystem (Amersham Pharmacia Biotech GmbH Europe). The gels for protein profiles were stained with silver according to Fig. 1. Schematic representation of the synthesis of melatonin–agarose resin. The hydroxyl group of 6-hydroxymelatonin was coupled to the epoxide group of epoxy-activated Sepharose 6B to yield melatonin– agarose. Ó FEBS 2003 Melatonin–calreticulin relationship (Eur. J. Biochem. 270) 833 Heukeshoven and Dernick [33]. Proteins separated by SDS/ PAGE were transferred to nitrocellulose using the Phar- macia Semi-dry Transfer kit [34]. Briefly, the nitrocellulose strips were incubated for 1 h at room temperature in blocking buffer [0.4% gelatin in PBST (150 m M phosphate- buffered saline, pH 7.4, containing 0.1% (v/v) Tween 20)], followed by incubation with PBST buffer containing 10 l M melatonin for 1 h at 4 °C. Ligand blot analysis was carried out using 1 : 800 dilution in PBST buffer of a specific polyclonal antibody against melatonin (G/S/704–8483; Stockgrand Ltd, Guildford, UK) for 2 h at room tempera- ture. After washing three times in PBST, the nitrocellulose membranes were incubated for 1 h with sheep peroxidase conjugated secondary antibody (1 : 800) in PBST and then washed again as above. The blots were finally developed with 3,3¢-diaminobenzidine. Gels and blots were digitized and processed by QuantiScan software (Biosoft, UK) and the molecular masses of the polypeptides were calculated according to their R f values. Western blotting Proteins obtained from both crude nuclear extract and solubilized material were loaded and separated by 12.5% SDS/PAGE, and Western blot analysis was performed according to the procedure of Towbin et al.[34].Briefly, separated proteins were transferred onto polyvinylidene difluoride membranes and blocked for 1 h at room temperature in 0.4% gelatin in PBST. Then, the gels were incubated for 1 h at room temperature with rabbit polyclonal antisera to calnexin and calreticulin (Sigma- Aldrich, Spain) at 1 : 1000 in PBST. Blots were washed three times in PBST, exposed to horseradish peroxidase- coupled antirabbit immunoglobulin, and detected by ECL according to the manufacturer’s protocol (Amersham Pharmacia Biotech, Spain). Anti-calnexin immunohistochemistry Isolated nuclei and microsomes were used for the immu- nohistochemical localization of calnexin in endoplasmic reticulum (ER). The homogenates were fixed by immersion in Formaline’s fixative (4%). After dehydration in graded alcohol, the homogenates were embedded in paraffin and cut into 10-lm sections. The sections were mounted on gelatin-coated slides and processed by the peroxidase– antiperoxidase (PAP) technique. After three 10-min rinses in phosphate buffer, endogenous peroxidase within the homogenates was blocked by a solution of 0.3% hydrogen peroxidase in NaCl/P i at room temperature for 30 min. This step was followed by three washes in NaCl/P i and incuba- tion for 30 min in a solution of 1 : 30 goat serum in NaCl/ P i . Sections were treated with rabbit polyclonal antisera to calnexin (Stressgen Biotechnologies) in serial dilutions from 1 : 200 to 1 : 5000 in NaCl/P i and incubated for 24 h at room temperature in a humid atmosphere. The slides were then rinsed in NaCl/P i and incubated with goat antirabbit immunoglobulin diluted 1 : 100 in NaCl/P i for 1 h and then with PAP complex diluted 1 : 100 for 1 h. The sites of peroxidase attachment were demonstrated by incubation in 0.005% 3,3¢-diaminobenzidine tetrahydrochloride solution in Tris/HCl buffer (50 m M , pH 7.6) containing 0.025% hydrogen peroxide. Finally the sections were rinsed in water, dehydrated and coverslipped. To ensure method specificity, the usual controls were performed. N-Terminal sequence analysis Proteins separated by SDS/PAGE were transferred to polyvinylidene difluoride and stained with Coomassie blue. The polypeptides of 60 and 74 kDa were excised and the first 15 N-terminal amino acid residues of each protein were sequenced by the Protein/Peptide Micro Analytical Labor- atory (California Institute of Technology, USA). Protein sequences comparisons were carried out using the FASTA program [35]. Swiss-Prot databanks were accessed though the GeneBank online service. Expression and purification of fusion protein Recombinant fusion protein glutathione S-transferase– calreticulin (GST–calreticulin) or GST were expressed in Escherichia coli strain BL21 (DE3)pysL (Stratagene, La Jolla, CA, USA) using a plasmid encoding rabbit calreti- culin provided by M. Michalak (Alberta University, Edmonton, Canada) [36]. Rabbit and rat calreticulin share 92.57% homology. Cells were grown to late log phase and induced to express the fusion proteins by addition of 0.25 m M isopropyl-1-thio- D -galactopyranoside for 4 h. The cells were harvested and lysed in lysis buffer (50 m M Tris, pH 7.8, 0.4 M NaCl,10%glycerol,0.5 m M EDTA, complete protease inhibitor, 0.1% Nonidet P-40) containing 1% Triton X-100 and 350 lgÆmL )1 lysozyme. Soluble proteins were separated from the inclusion bodies and bacterial debris by centrifugation at 10 000 g for 20 min at 4 °C. The recombinant proteins were purified from the supernatant by glutathione-Sepharose (Amersham Pharmacia Biotech Europe GmbH) and extensively washed with NaCl/P i . Matrix bound protein was used for binding assay. 2-[ 125 I]Melatonin binding assay An aliquot of 20 lL of GST-calreticulin or GST proteins weremixedwith50m M Tris-HCl, pH 7.4 (6.96 m M CaCl 2 , 100 l M dithiothreitol) in a total volume of 100 lL, yielding a final protein concentration of 200 lgÆmL )1 . The mixture was incubated at 37 °C for 2 h in the presence of 100 p M radiolabeled ligand (2-[ 125 I]melatonin, 81.4 TbqÆmmol )1 ) and the incubation was stopped adding 100 lLofcold 100% trichloroacetic acid followed by centrifugation at 45 000 g for 15 min at 4 °C. The supernatant was discarded by aspiration, and the radioactivity on the pellets was determined in a gamma-counter. Nonspecific binding, cal- culated in the presence of 10 l M melatonin, was 12–16% of the total binding. Kinetic parameters (K D and B max )and IC 50 values were measured from the displacement curves with the LIGAND-PC program (KELL software, Biosoft, UK). Protein content was determined as described [31]. Statistics Data are expressed as means ± SEM. Comparisons among groups were made by a one-way analysis of variance ( ANOVA ) followed by Student’s t-test. 834 M. Macı ´ as et al. (Eur. J. Biochem. 270) Ó FEBS 2003 Results Purification by affinity chromatography The results obtained after purification of the melatonin receptor present in rat liver nuclei by affinity chromato- graphy are shown in Table 1. Affinity chromatography of crude nuclear extract (6000–8000 g) gave 0.14% of protein (8.4–11.2 lg). The yield of the melatonin–agarose step significantly increased when solubilized material was used as source of melatonin receptor. A higher percentage of purified protein was obtained from the 45 to 65% ammo- nium sulfate fraction. This fraction yields 0.45% of protein (2.7–3.6 lg protein) after its purification by melatonin– agarose. Electrophoresis analysis Aliquots of both crude nuclear extract and solubilized material purified by affinity chromatography were analyzed by SDS/PAGE gel electrophoresis followed by silver staining (Fig. 2). Crude nuclear extract shows a wide stain ranging from 94 to 14 kDa (lanes 1 and 2). The protein pattern was similar in samples either untreated (lane 1) or treated with b-mercaptoethanol (lane 2). The number of protein bands decreased with ammonium sulfate treat- ment in solubilized material, and only two bands were clearly identified in 45–65% ammonium sulfate samples (lane 5). These bands, corresponding to polypeptides of 60 and 74 kDa, were also present in 25–45% (lane 4) and 0–25% (lane 3) ammonium sulfate fractions and in crude nuclear extract samples (lanes 1 and 2). An aliquot of molecular mass marker solution was applied as reference (M r lane). N-Terminal sequence of the p74 and p60 proteins To analyze the primary structure of 74 and 60 kDa polypeptides, protein microsequencing was carried out. In order to transfer a sequenceable quantity of protein onto the polyvinylidene difluoride, several nanomoles of protein were loaded onto the gel prior to electrophoresis. Electroblotted polypeptides were located by staining of the polyvinylidene difluoride with Coomassie blue. The sequences obtained were subjected to similarity searches in the database network. As no sequence identity was found for the 74 kDa protein (SFLEEDRNDQPVEI) after comparing with known protein sequences, this protein is suggested to be a novel protein. Searching the sequence database of the National Center of Biotechnology Information (NCBI), the protein–protein blast program showed no similarity of the 74 kDa protein with any other known protein to date. Besides, there is no information in a GeneBank that may help to identify this protein. Interestingly, the N-terminal sequence of the first 12 residues of the 60 kDa protein (DPAIYFKEQFLDGFA) was 100% identity to that of rat calreticulin. Western blotting and anticalnexin immunohistochemistry To identify the presence of calreticulin and to discard a possible contamination of our preparation with ER, Western blot analyses were performed using the antibodies anticalreticulin and anticalnexin (Fig. 3). Lanes 1 and 5 correspond to crude nuclear extract and lanes 2–4 corres- pond to solubilized material treated with 0–25%, 25–45% and 45–65% ammonium sulfate, respectively. Lanes 1–4 were incubated in the presence of calreticulin antibody and lane 5 with calnexin antibody. A positive control for calnexin antibody is shown in lane 6, that correspond to the microsome sample. Only one band corresponding to the polypeptides of 60 kDa was recognized in the presence of anticalreticulin (lanes 1–4), and this band is enriched by 25–65% ammonium sulfate precipitation. A lack of immu- noreactive band is apparent in lane 5. The results confirm the identity of the 60 kDa protein as calreticulin, and exclude a contamination from ER. To further assess whether calreticulin was not being copurified from ER during the purification procedure, anticalnexin immunohistochemistry was carried out in preparations from nuclei homogenate. A fraction contain- ing ER was also used for control purposes. The results show a lack of immunoreactivity in the nuclei homogenates, whereas positive immunoreactivity for anticalnexin was found in homogenates from microsomes (data not shown). Ligand blotting Figure 4 demonstrates that the 74 and 60 kDa proteins specifically bind melatonin as determined by ligand blotting using anti-melatonin Ig. Lane 3 corresponds to crude nuclear material and lanes 4–6 correspond to solubilized material treated with 0–25, 25–45 and 45–65% ammonium sulfate, respectively. Lanes 3, 4, 5 and 6 were incubated with 10 l M melatonin. Only the bands corresponding to the Table 1. Purification of the melatonin receptor. Starting material refers to the amount of protein present in samples of crude nuclear extract and in samples of solubilized material treated with different concentrations of ammonium sulfate. Affinity chromatography data shows the amount of protein obtained after incubation of these samples with melatonin agarose. The percentage of purified proteins in relation to the protein content in the starting material is shown in brackets. Sample Starting material (lg protein) Affinity chromatography (lg protein) Crude nuclear extract 6000–8000 8.4–11.2 (0.14%) Solubilized material (Triton X-100) 0–25% ammonium sulfate precipitate 900–1200 2.2–2.9 (0.24%) 25) 45% ammonium sulfate precipitate 800–900 2.7–3.1 (0.34%) 45–65% ammonium sulfate precipitate 600–800 2.7–3.6 (0.45%) Ó FEBS 2003 Melatonin–calreticulin relationship (Eur. J. Biochem. 270) 835 polypeptides of 74 and 60 kDa were recognized in the presence of melatonin, thus suggesting that these proteins bind melatonin selectively. Lanes corresponding to crude nuclear extract or solubilized material incubated with melatonin in the absence of anti-melatonin Ig (lane 1) or incubated with anti-melatonin Ig in the absence of ligand (lane 2) served as controls. Standards stained with Coo- massie blue R were included (M r lane). These molecules do not correspond to proteins such as histones which are associated with DNA, as was further assessed by amino acid sequencing, because they were not recognized by antibodies against histones (data not shown). Binding experiments To further assess the characteristics of melatonin–calreti- culin binding, a series of competitive experiments were carried out. Figure 5 (left) shows a typical displacement curve for 2-[ 125 I]melatonin performed with bacterially produced GST-calreticulin fusion proteins. The IC 50 value was 0.97 ± 0.3 n M . Scatchard transformation was carried out from competition experiments after recalculating the specific activities (Fig. 5, left inset). Kinetic analysis of these data yielded a K d of 1.08 ± 0.2 n M and a B max of 290 ± 34 fmolÆmg protein )1 . Specific binding was undetectable in the absence of Ca 2+ . Binding experiments in the presence of different doses (1 n M to 10 m M ) of NAS, 4-P-PDOT and luzindole showed no competition with the radioligand. Moreover, the binding was specific for calreticulin, as no specific binding to bacterially produced GST protein was detected. Discussion The objective of this work was to purify the nuclear melatonin receptor elsewhere characterized in liver nuclei [23,24], by classical protein purification approaches. From this work, two polypeptides of 74 and 60 kDa that specifically bind to melatonin were obtained. Luzindole and 4-P-PDOT, two selective antagonists for the mt 1 /MT 2 subtypes of the melatonin membrane receptors [37] do not compete with 2-[ 125 I]melatonin binding to calreticulin, the 60 kDa protein. These results, together with the sequencing Fig. 4. Ligand blotting (SDS/PAGE) of nuclear extracts purified by affinity chromatography. Crude nuclear extract (lane 3) and solubilized material pretreated with 0–25, 25–45 and 45–65% (lanes 4, 5 and 6, respectively) ammonium sulfate. The blots were preincubated with 10 l M melatonin (lane 3, 4, 5 and 6) as described in Material and methods. Lane 1 and 2 served as controls. The crude nuclear extract was incubated with 10 l M melatonin in absence of anti-melatonin Ig (lane 1), or incubated with the primary antibody in absence of ligand (lane 2). Molecular mass markers stained with Coomassie blue R are indicatedintheM r lane. Fig. 2. Silver-stained SDS/PAGE gels (12.5% homogenous media) of purified material by affinity chromatography. The gels show the dena- tured polypeptide composition of crude nuclear extract either untreated (lane 1) or treated with 1-mercaptoethanol (lane 2), and solubilized material treated with ammonium sulfate at 0–25% (lane 3), 25–45% (lane 4) and 45–65% (lane 5). Molecular mass markers are indicated in the M r lane. Fig. 3. Western blotting of crude nuclear extract (lanes 1 and 5) and solubilized material pretreated with 0–25, 25–45 and 45–65% (lanes 2, 3 and 4, respectively) ammonium sulfate. The blots were incubated with anti-calreticulin (lanes 1–4) and anti-calnexin Igs (lane 5) as described in Material and methods. Lane 6 corresponds to a positive control for calnexin antibody in microsomes. 836 M. Macı ´ as et al. (Eur. J. Biochem. 270) Ó FEBS 2003 data, suggest that the proteins purified in this study represent a new class of specific binding sites for melatonin. The proteins present in the crude nuclear extract, precipitated with ammonium sulfate, allowed us to increase the efficiency of the receptor purification by decreasing the number of proteins present in the sample. We found that maximal calreticulin activity was obtained in samples precipitated with 45–65% ammonium sulfate, i.e. in the similar precipitating fraction as previously described [38]. However, the fraction obtained with 25–45% ammonium sulfate precipitation also has high content of calreticulin and so, the largest yield of melatonin receptors (0.79%) was obtained after affinity chromatography of samples precipi- tated with 25–65% ammonium sulfate. In addition, the number of proteins separated by SDS/PAGE and stained with silver nitrate was significantly reduced after ammonium sulfate treatment. The results suggest that these sample pretreatments before affinity chromatography not only increased the percentage of receptor obtained but also removed a large number of proteins that might interfere with the purification procedure itself. The development of the melatonin–agarose resin allowed us to improve the purification procedure of the melatonin nuclear receptor. Criteria used for the validation of other affinity resins [39] strongly suggest that melatonin–agarose interacts with solubilized receptors in a specific manner. Purification achieved with melatonin–agarose is similar to that achieved with affinity resins developed for the purifi- cation of other receptors [40]. The purified proteins obtained after affinity chromatography show pharmacological pro- perties compatible with a receptor of melatonin. These data support the utility of the melatonin–agarose resin to purify the melatonin nuclear receptor and suggest that large-scale purification may be now feasible. Electrophoresis and Western blot analysis of the samples purified by affinity chromatography revealed the presence of two proteins corresponding to molecular weights of 74 and 60 kDa. The N-terminal amino acid sequence of the purified proteins was searched in protein databases for homology identity. Regarding the 74 kDa protein, no proteins with a similar sequence were found, suggesting that this polypep- tide is a novel, formerly unknown protein. It was, however, an unexpected finding that the N-terminal amino acid sequence of the 60 kDa protein showed considerable sequence homology with rat calreticulin. It is interesting that calreticulin, a highly acidic protein, moves at about 60–65 kDa on SDS/PAGE [41], although the deduced molecular mass from the amino acids is 46 kDa. A 60-kDa polypeptide obtained after SDS/PAGE of a 55–70% ammonium sulfate precipitate of the HeLa cell cytosol was also identified as calreticulin by mass spectrometry [38]. Thus, the similarity of our purification methodology compared with that used by these authors, the molecular mass of the polypeptide obtained by SDS/PAGE, and the sequence analysis results, strongly suggest that the 60 kDa protein purified by us is calreticulin. The melatonin binding to calreticulin is highly specific and displays nanomolar affinity. The specificity of this biding is also supported because NAS, the metabolic precursor of melatonin with certain degree of affinity for other melatonin binding sites [24], does not interfere with the melatonin binding. Moreover, the presence of 4-P-PDOT and luzindole does not interfere with the binding of melatonin to calreticulin. These results suggest a highly specific binding of melatonin to calreticulin, although further experiments with saturation studies should be carried out to assess that these binding sites can be saturated, thus confirming the existence of functional binding of melatonin to calreticulin. The data strongly suggests a new class of protein binding site for melatonin and suggests that calreticulin is a target for the intracellular action of melatonin. Calreticulin is a ubiquitous and highly conserved Ca 2+ - binding protein of the ER that could be regulated through intracellular signaling pathways involving Ca 2+ binding [42]. The protein is multifunctional and may play an important role in the modulation of a variety of cellular processes. These functions include chaperon activity, con- trol of intracellular Ca 2+ homeostasis, and regulation of cell adhesiveness by interacting with the integrins at the cytoplasmatic site of the plasma membrane. Surprisingly, calreticulin controls the steroid-sensitive gene expression Fig. 5. Binding of 2-[ 125 I]melatonin to calreti- culin. (A) Competition experiments were per- formed with recombinant calreticulin (GST– calreticulin) and increasing concentrations of nonlabeled melatonin. Results were expressed as the percentage of specifically bound 2-[ 125 I]melatonin. Inset: Scatchard plot of the data in (A) showing the binding kinetics of melatonin to GST–calreticulin. (B) Values of B max of melatonin binding to GST–calreti- culin in the presence (CRT + Ca 2+ )and absence(CRT)ofCa 2+ , and to GST protein. Melatonin, d;NAS,s; 4-P-PDOT, .; luzindole, h.*P < 0.001. Ó FEBS 2003 Melatonin–calreticulin relationship (Eur. J. Biochem. 270) 837 [43,44]. This was an unexpected finding, as calreticulin is an ER-resident protein [45] and steroid receptors are found either in the cytoplasm or in the nucleus. Several pieces of evidence suggest a parallel between calreticulin and calmodulin in nuclear melatonin signaling. Melatonin binds both calmodulin and calreticulin only in the presence of Ca 2+ [46]. Calreticulin mediates nuclear export of the glucocorticoid receptor, and overexpression of calreticulin antagonizes nuclear receptor-dependent tran- scriptional activation [38]. Melatonin protects against glucocorticoid-induced apoptosis regulating glucocorticoid receptor expression [47]. Calreticulin specifically interacts with the first zinc finger of different nuclear receptors. Based on this feature, it was shown that calreticulin interacts with amino acids 206–211 of the DNA binding domain region of estrogen receptor alpha (ERa), reversing ERa inhibition of invasion in vitro [48]. Calmodulin also modulates ERa interacting with amino acids 290–310 of this receptor [49], whereas melatonin–calmodulin interaction blocks the acti- vation of estrogen receptor for DNA binding [50]. Thus, the oncostatic effects of melatonin against ERa activation may depend on its binding to calmodulin and calreticulin, preventing both the binding of ERa to DNA and its proliferative effects. The problem of calreticulin localization into the cell continues. Calreticulin-like immunoreactivity was detected in the nucleus of some cells, although it seems that it is not a nuclear resident protein [45,51]. An explanation for these contradictions may depend on the purification methodo- logy. In fact, resident nuclear proteins are associated with the Triton-insoluble nuclear fractions, whereas the Triton- soluble fractions contain proteins of ER origin [52]. Michalak [45] identified 60-kDa calreticulin in the Triton X-100 soluble fraction of purified nuclei, which includes the solubilized outer nuclear membrane containing proteins of the ER, but not in the Triton-insoluble fraction containing nuclear material surrounded by the inner nuclear mem- brane. These results suggest that calreticulin is not a resident nuclear protein. Other reports failed to identify calreticulin in the cytosol, suggesting that the ER, but not the cytosol form of calreticulin is responsible for inhibition of gluco- corticoid receptor-mediated gene expression [44,45,53,54], a proposed function for this protein. However, in vitro DNA- binding assays indicated that recombinant calreticulin could inhibit DNA binding by steroid receptors, suggesting that the effect of calreticulin on nuclear hormone receptor transactivation might be direct [48]. Experimental evidence exists supporting both nuclear [53] and cytosolic [38,54,55] localization of calreticulin. These data provide evidence for two pools of calreticulin, the first contained within the lumen of the ER, and the second contained within the cytosol. Our data show that nuclei homogenates are lacking calnexin, a marker for ER [56], but that they do contain calreticulin. Therefore, our procedure for nuclear protein purification started from a material lacking ER contamin- ation, suggesting that the calreticulin found in our material does not come from this localization and that it was associated with the nuclei. It seems that, although it is not a nuclear resident protein, calreticulin may localize in the nucleus. The mechanism(s) by which calreticulin molecules are imported into, and retained in, the nucleus are unknown. It is unclear whether nuclear localization of calreticulin is determined by its simple exclusion from the ER, possibly due to elimination of its N-terminal signal sequence, or by its retrotranslocation from the endoplasmic reticulum to the cytoplasm. The classical view of strict protein compartmentalization has now been challenged, and it is thought that proteins may shuttle between the nuclear and cytoplasmic compartments [57]. Calreticulin may have a similar behavior to localize in distinct subcellular compartments, acting independently on compartment-specific targets. These data may suggest that the interaction between melatonin and calreticulin (and calmodulin) could be of physiological importance in regulating the activity of a broad spectrum of nuclear receptors [38]. This hypothesis is further supported because the high melatonin-calreticulin binding affinity, which correlates well with the melatonin concentration in nucleus [18,24]. Therefore, melatonin might be a mechanism involved in importing and/or retaining calreticulin in the nucleus. Melatonin–calreticulin interaction also can be related to the balance of ligand- induced import and calreticulin-dependent export, provi- ding the cell with a nuclear transport-based mechanism. Based on these findings, it is necessary to re-evaluate our current understanding of the molecular pathways of mela- tonin actions. Acknowledgements We thank Dr M. Michalak for providing the plasmid which encode GST-calreticulin and Dr M. Martı ´ n for the binding experiments. We also thank Dr C. Carlberg for helpful suggestions. This work was supported by the CICYT grant SAF98 : 0156 and Junta de Andalucı ´ a (CTS-101). M. Macı ´ as is a fellow from the Programa de Formacion de Personal Investigador, Ministerio de Educacion y Cultura, Spain. References 1. Kondo, T., Mori, T., Lebedeva, N.V., Aoki, S., Ishiura, M. & Golden, S.S. (1997) Circadian rhythms in rapidly dividing cya- nobacteria. Science 275, 224–227. 2. Hardeland, R. & Fuhrberg, B. (1996) Ubiquitous melatonin – presence and effects in unicells, plants and animals. Trends Comp. Biochem. 2, 25–44. 3. Reiter, R.J. (1991) Pineal gland. Interface between the photo- periodic environment and the endocrine system. Trends Edocrinol. Metab. 2, 13–19. 4. Huether, G. (1993) The contribution of extrapineal sites of mela- tonin synthesis to circulating melatonin levels in higher verte- brates. Experientia 49, 665–670. 5. Tamarkin, L., Baird, C.J. & Almeida, O.F.X. (1985) Melatonin: a coordinating signal for mammalian reproduction? Science 227, 714–720. 6. Reiter, R.J. (1995) Functional pleiotropy of the neurohormone melatonin: Antioxidant protection and neuroendocrine regula- tion. Front. Neuroendocrinol. 16, 383–415. 7. Liebmann, P.M., Wolfler, A., Felsner, P., Hofer, D. & Schauen- stein, K. (1997) Melatonin and the immune system. Int. Arch. Allergy Immunol. 112, 203–211. 8. Cos,S.,Fernandez,F.&Sa ´ nchez-Barcelo ´ , E.J. (1996) Melatonin inhibits DNA synthesis in MCF-7 human breast cancer cells in vitro. Life Sci. 58, 2447–2453. 9. Sainz, R.M., Mayo, J.C., Uria, H., Kotler, M., Antolin, I., Rodriguez, C. & Mene ´ ndez-Pelaez, A. (1995) The pineal 838 M. Macı ´ as et al. (Eur. J. Biochem. 270) Ó FEBS 2003 neurohormone melatonin prevents in vivo and in vitro apoptosis in thymocytes. J. Pineal Res. 19, 178–188. 10. Reiter, R.J., Melchiorri, D., Sewerynek, E., Poeggeler, B., Barlow- Walden,L.,Chuang,J.I.,Ortiz,G.G.&Acun ˜ a-Castroviejo, D. (1995) A review of the evidence supporting melatonin’s role as an antioxidant. J. Pineal Res. 18, 1–11. 11. Reiter, R.J. (1998) Oxidative damage in the central nervous sys- tem: protection by melatonin. Progr. Neurobiol. 56, 359–384. 12. Martı ´ n, M., Macı ´ as, M., Escames, G., Leo ´ n, J. & Acun ˜ a- Castroviejo, D. (2000) Melatonin but not vitamins C and E maintains glutathione homeostasis in t-butyl hydroperoxide- induced mitochondrial oxidative stress. FASEB J. 14, 1677–1679. 13. Acun ˜ a-Castroviejo, D., Martı ´ n, M., Macı ´ as, M., Escames, G., Leo ´ n, J., Khaldy, H. & Reiter, R.J. (2001) Melatonin, mitochondria and cellular bioenergetics. J. Pineal Res. 30, 65–74. 14. Ebisawa, T., Karne, S., Lerner, M.R. & Reppert, S.M. (1994) Expression cloning of a high-affinity melatonin receptor from Xenopus dermal melanophores. Proc. Natl Acad. Sci. USA 91, 6133–6137. 15. Reppert, S.M., Weaver, D.R. & Ebisawa, T. (1994) Cloning and characterization of a mammalian melatonin receptor that medi- ates reproductive and circadian responses. Neuron 13, 1177–1185. 16. Reppert, S.M., Godson, C., Mahle, C.D., Weaver, D.R. & Slau- genhaupt, S.A. (1995) Molecular characterization of a second melatonin receptor expressed in human retina and brain: the Mel 1b melatonin receptor. Proc.NatlAcad.Sci.USA92, 8734–8738. 17. Reppert, S.M., Weaver, D.R. & Godson, C. (1996) Melatonin receptors step into the light: cloning and classification of subtypes. Trends Pharmacol. Sci. 17, 100–102. 18. Mene ´ ndez-Pelaez, A. & Reiter, R.J. (1993) Distribution of mela- tonin in mammalian tissues: The relative importance of nuclear versus cytosolic localization. J. Pineal Res. 15, 59–69. 19. Antolin, I., Rodriguez, C., Sainz, R.M., Mayo, J.C., Uria, H., Kotler, M.L., Rodriguez-Colunga, M.J., Tolivia, D. & Mene ´ ndez- Pelaez, A. (1996) Neurohormone melatonin prevents cell damage: effect on gene expression for antioxidant enzymes. FASEB J. 10, 882–890. 20. Crespo, E., Macı ´ as,M.,Pozo,D.,Escames,G.,Martı ´ n, M., Vives, F., Guerrero, J.M. & Acun ˜ a-Castroviejo, D. (1999) Mela- tonin inhibits expression of the inducible NO synthase II in liver and lung and prevents endotoxemia in lipopolysaccharide-induced multiple organ dysfunction syndrome in rats. FASEB J. 13, 1537– 1546. 21. Garcı ´ a-Maurin ˜ o, S., Gonzalez-Haba, M.G., Calvo,J.R., Goberna, R. & Guerrero, J.M. (1998) Involvement of nuclear binding sites for melatonin in the regulation of IL-2 and IL-6 production by human blood mononuclear cells. J. Neuroimmunol. 92, 76–84. 22. Slominski, A. & Pruski, D. (1993) Melatonin inhibits proliferation and melanogenesis in rodent melanoma cells. Exp Cell Res. 206, 189–194. 23. Acun ˜ a-Castroviejo, D., Pablos, M.I., Mene ´ ndez-Pela ´ ez,A.& Reiter, R.J. (1993) Melatonin receptors in purified cell nuclei of liver. Res. Comm. Chem. Pathol. Pharmacol. 82, 253–256. 24. Acun ˜ a-Castroviejo, D., Reiter, R.J., Mene ´ ndez-Pelaez, A., Pablos, M.I. & Burgos, A. (1994) Characterization of high-affinity mela- tonin binding sites in purified cell nuclei of rat liver. J. Pineal Res. 16, 100–112. 25. Becker-Andre ´ , M., Wiesenberg, I., Schaeren-Wiemers, N., Andre ´ , E., Missbach, M., Saurat, J H. & Carlberg, C. (1994) Pineal gland hormone melatonin binds and activates an orphan of the nuclear receptor superfamily. J. Biol. Chem. 269, 28531–28534. 26. Wiesenberg, I., Missbach, M., Kahlen, J P., Schra ¨ der, M. & Carlberg, C. (1995) Transcriptional activation of the nuclear receptor RZR by the pineal gland hormone melatonin and iden- tification of CGP 52608 as a synthetic ligand. Nucleic Acids Res. 23, 327–333. 27. Steinhilber, D., Brungs, M., Werz, O., Wiesenberg, I., Danielsson, C., Kahlen, J P., Nayeri, S., Schra ¨ der, M. & Carlberg, C. (1995) The nuclear receptor for melatonin represses 5-lipoxygenase gene expression in human B lymphocytes. J. Biol. Chem. 270, 7037– 7040. 28. Blum, M., McEwens, B.S. & Roberts, J.L. (1987) Transcriptional analysis of tyrosine hydroxylase gene expression in the tuber- oinfundibular dopaminergic neurons of the rat arcuate nucleus after estrogen treatment. J. Biol. Chem. 262, 817–821. 29. Englard, S. & Seifter, S. (1990) Precipitation techniques. Methods Enzymol. 182, 285–300. 30. Anonymous. (1988) Affinity Chromatography. Principles and Methods. Pharmacia LKB Biotechnology, Uppsala, Sweden. 31. Bradford, M. (1976) A rapid and sensitive method for the quan- titation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 32. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680– 685. 33. Heukeshoven, J. & Dernick, R. (1985) Simplied method for silver staining of proteins in polyacrylamide gels and mechanism of sil- ver staining. Electrophoresis 6, 103–112. 34. Towbin, H., Staeheling, T. & Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some apllications. Proc. Natl Acad. Sci. USA 76, 4354–4350. 35. Pearson, W.R. & Lipman, D.J. (1988) Improved tools for biolo- gical sequence comparison. Proc. Natl Acad. Sci. USA 85, 2444– 2448. 36. Baksh, S. & Michalak, M. (1991) Expression of calreticulin in Escherichia coli and identificatmon of its Ca2+ binding domains. J. Biol. Chem. 266, 21458–21465. 37. Dubocovich,M.L.,Yun,K.,Al-Ghoul,W.M.,Benloucif,S.& Masana, M.I. (1998) Selective MT2 melatonin receptor anta- gonists block melatonin-mediated phase advances of circadian rhythms. FASEB J. 12, 1211–1220. 38. Holaska, J.M., Black, B.E., Love, D.C., Hanover, J.A., Lesyk, J. & Paschal, B.M. (2001) Calreticulin is a receptor for nuclear export. J. Cell Biol. 152, 127–140. 39. Caron,M.G.,Srinivasan,Y.,Pitha,J.,Kociolek,K.&Lefkowitz, J.R. (1979) Affinity chromatography of the b-adrenergic receptor. J. Biol. Chem. 254, 2923–2927. 40. Lomasney, J.W., Leeb-Lundberg, L.M.F., Cotecchia, S., Regan, J.W., DeBernardis, J.F., Caron, M.G. & Lefkowitz, R.J. (1986) Mammalian a 1 -adrenergic receptor. Purification and characteri- zation of the native receptor ligand binding subunit. J. Biol. Chem. 261, 7710–7716. 41. Milner, R.E., Baksh, S., Shemanko, C., Carpenter, M.R., Smillie, L., Vance, J.E., Opas, M. & Michalak, M. (1991) Calre- ticulin, and not calsequestrin, is the major calcium binding protein of smooth muscle sarcoplasmic reticulum and liver endoplasmic reticulum. J. Biol. Chem. 266, 7155–7165. 42. Krause, K.H. & Michalak, M. (1997) Calreticulin. Cell 88, 439– 443. 43. Burns, K., Duggan, B., Atkinson, E.A., Famulski, K.S., Nemer, M., Bleackley, R.C. & Michalak, M. (1994) Modulation of gene expression by calreticulin binding to the glucocorticoid receptor. Nature 367, 476–480. 44. Dedhard, S., Rennie, P.S., Shago, M., Hagesteijn, C.Y.L., Yang, H., Filmus, J., Hawley, R.G., Bruchovsky, N., Cheng, H., Matusik, R.J. & Giguere, V. (1994) Inhibition of nucleare hor- mone receptor activity by calreticulin. Nature 367, 480–483. 45. Michalak, M., Burns, K., Andrin, C., Mesaeli, N., Jass, G.H., Busaan, J.L. & Opas, M. (1996) Endoplasmic reticulum form of calreticulin modulates glucocorticoid-sensitive gene expression. J. Biol. Chem. 271, 29436–29445. Ó FEBS 2003 Melatonin–calreticulin relationship (Eur. J. Biochem. 270) 839 46. Leo ´ n, J., Macı ´ as, M., Escames, G., Camacho, E., Khaldy, H., Martı ´ n,M.,Espinosa,A.,Gallo,M.A.&Acun ˜ a-Castroviejo, D. (2000) Structure-related inhibition of calmodulin-dependent nNOS activity by melatonin and synthetic kynurenines. Mol. Pharmacol. 58, 967–975. 47. Sainz, R.M., Mayo, J.C., Reiter, R.J., Antolin, I., Esteban, M.M. &Rodrı ´ guez, C. (1999) Melatonin regulates glucocorticoid receptor: An answer to its antiapoptotic action in thymus. FASEB J. 13, 1547–1556. 48. Platet, N., Cunat, S., Chalbos, D., Rochefort, H. & Garcia, M. (2000) Unliganded and liganded estrogen receptors protect against cancer invasion via different mechanisms. Mol. Endocrinol. 14, 999–1009. 49. Garcia Pedrero, J.M., Del Ri, B., Martinez-Campa, C., Muramatsu, M., Lazo, P.S. & Ramos, S. (2002) Calmodulin is a selective modulator of estrogen receptors. Mol. Endocrinol. 16, 947–960. 50. Garcia Rato, A., Garcia Pedrero, J.M., Martinez, A., Del Rio, B., Lazo, P.S. & Ramos, S. (1999) Melatonin blocks the activation of estrogen receptor for DNA binding. FASEB J. 13, 857–868. 51. Michalak, M., Corbett, E.F., Mesaeli, N., Nakamura, K. & Opas, M. (1999) Calreticulin: one protein, one gene, many func- tions. Biochem. J. 344, 281–292. 52. Blobel, G. & Potter, V.R. (1966) Nuclei from rat liver: isolation method that combines purity with high yield. Science 154, 1662– 1665. 53. Roderick, H.L., Campbell, A.K. & Llewellyn, D.H. (1997) Nuclear localisation of calreticulin in vivo is enhanced by its interaction with glucocorticoid receptors. FEBS Lett. 405, 181– 185. 54. Jethmalani, S.M., Henle, K.J., Gazitt, Y., Walker, P.D. & Wang, S.Y. (1997) Intracellular distribution of heat-induced stress gly- coproteins. J. Cell Biochem. 66, 98–111. 55. Yoon, G.S., Lee, H., Jung, Y., Yu, E., Moon, H.B., Song, K. & Lee, I. (2000) Nuclear matrix of calreticulin in hepatocellular carcinoma. Cancer Res. 60, 1117–1120. 56. Wada, I., Rindress, D., Cameron, P.H., Ou, W.J., Doherty, J.J., Louvard, D., Bell, A.W., Dignard, D., Thomas, D.Y. & Bergeron, J.J. (1991) SSR alpha and associated calnexin are major calcium binding proteins of the endoplasmic reticulum membrane. J. Biol. Chem. 266, 19599–19610. 57. Yang,L.,Guerrero,J.,Hong,H.,DeFranco,D.B.&Stallcup, M.R. (2000) Interaction of the tau2 transcriptional activation domain of glucocorticoid receptor with a novel steroid receptor coactivator, Hic-5, which localizes to both focal adhesions and the nuclear matrix. Mol. Biol. Cell 11, 2007–2018. 840 M. Macı ´ as et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . the antibodies anticalreticulin and anticalnexin (Fig. 3). Lanes 1 and 5 correspond to crude nuclear extract and lanes 2–4 corres- pond to solubilized material treated with 0–25%, 25–45% and. Calreticulin–melatonin An unexpected relationship Manuel Macı ´ as 1 , Germaine Escames 1 , Josefa Leon 1 , Ana Coto 3 , Younes Sbihi 2 , Antonio Osuna 2 and Darı ´ o Acun ˜ a-Castroviejo 1 1 Departamento. either untreated (lane 1) or treated with 1-mercaptoethanol (lane 2), and solubilized material treated with ammonium sulfate at 0–25% (lane 3), 25–45% (lane 4) and 45–65% (lane 5). Molecular mass