A novel FK506 binding protein can mediate the immunosuppressive effects of FK506 and is associated with the cardiac ryanodine receptor.

FK506, an immunosuppressant that prolongs allograft survival, is a co-drug with its intracellular receptor, FKBP12. The FKBP12•FK506 complex inhibits calcineurin, a critical signaling molecule during T-cell activation. FKBP12 was, until recently, the sole FKBP known to mediate calcineurin inhibition at clinically relevant FK506 concentrations. The best characterized cellular function of FKBP12 is the modulation of ryanodine receptor isoform-1, a component of the calcium release channel of skeletal muscle sarcoplasmic reticulum. Recently, a novel protein, FKBP12.6, was found to inhibit calcineurin at clinically relevant FK506 concentrations. We have cloned the cDNA encoding human FKBP12.6 and characterized the protein. In transfected Jurkat cells, FKBP12.6 is equivalent to FKBP12 at mediating the inhibitory effects of FK506. Upon binding rapamycin, FKBP12.6 complexes with the 288-kDa mammalian target of rapamycin. In contrast to FKBP12, FKBP12.6 is not associated with ryanodine receptor isoform-1 but with the distinct ryanodine receptor isoform-2 in cardiac muscle sarcoplasmic reticulum. Our results suggest that FKBP12.6 has both a unique physiological role in excitation-contraction coupling in cardiac muscle and the potential to contribute to the immunosuppressive and toxic effects of FK506 and rapamycin.

FK506, an immunosuppressant that prolongs allograft survival, is a co-drug with its intracellular receptor, FKBP12. The FKBP12⅐FK506 complex inhibits calcineurin, a critical signaling molecule during T-cell activation. FKBP12 was, until recently, the sole FKBP known to mediate calcineurin inhibition at clinically relevant FK506 concentrations. The best characterized cellular function of FKBP12 is the modulation of ryanodine receptor isoform-1, a component of the calcium release channel of skeletal muscle sarcoplasmic reticulum.
Recently, a novel protein, FKBP12.6, was found to inhibit calcineurin at clinically relevant FK506 concentrations. We have cloned the cDNA encoding human FKBP12.6 and characterized the protein. In transfected Jurkat cells, FKBP12.6 is equivalent to FKBP12 at mediating the inhibitory effects of FK506. Upon binding rapamycin, FKBP12.6 complexes with the 288-kDa mammalian target of rapamycin. In contrast to FKBP12, FKBP12.6 is not associated with ryanodine receptor isoform-1 but with the distinct ryanodine receptor isoform-2 in cardiac muscle sarcoplasmic reticulum. Our results suggest that FKBP12.6 has both a unique physiological role in excitation-contraction coupling in cardiac muscle and the potential to contribute to the immunosuppressive and toxic effects of FK506 and rapamycin.
FK506 (tacrolimus) is a powerful immunosuppressive drug for treating graft rejection and autoimmune disorders. Rapamycin (RAP, 1 sirolimus) is an immunosuppressant structural-ly-related to FK506 but with a distinct mechanism of action. Both drugs bind to a family of intracellular receptors, the FK506 binding proteins (FKBPs), whose members include FK-BPs 12, 12.6, 13, 25, 51, and 52 (for review, see Ref. 1). All FKBPs are peptidyl-prolyl isomerases, catalyzing the cis-trans isomerization of peptidyl-prolyl bonds in peptides and proteins, an activity inhibited by both FK506 and RAP.
Peptidyl-prolyl isomerase inhibition is unrelated to immunosuppression. FK506 and RAP gain function upon binding FKBP12. The FKBP12⅐FK506 and FKBP12⅐RAP complexes are the actual immunosuppressive species whose targets are calcineurin (CaN) and the mammalian target of RAP (mTOR), respectively (for review, see Refs. 1 and 2). CaN is a Ca 2ϩ -dependent, serine-threonine phosphatase required during the commitment phase (G 0 3 G 1 ) of T-cell activation (3). Inhibition of CaN blocks the nuclear translocation of transcription factors such as nuclear factor of activated T-cells and NF-B, controlling the expression of cytokine genes whose products are required for immune response coordination (for review, see Ref. 2). RAP, unlike FK506, does not block lymphokine production but inhibits the T-cell proliferative response to cytokines by blocking G 1 3 S-phase progression. The function of mTOR, a 288-kDa protein related to phosphatidylinositol kinases, is unknown.
CaN is a ubiquitous protein, and its inhibition at unwanted sites is most responsible for the toxicity associated with FK506 therapy (4). That immunosuppression and toxicity are mechanistically linked through inhibition of CaN has been documented using the nonimmunosuppressive and nontoxic FK506 analog, L-685,818 (Ј818; see Fig. 1). The observations that Ј818 binds tightly to FKBP12, that the human FKBP12⅐Ј818 complex has little affinity for CaN, and that Ј818 reverses FK506 toxicity have demonstrated that CaN inhibition, not FKBP binding, is responsible for the toxicity profile of FK506 (4,5).
The cellular and pharmacologic functions of FKBP12 are unrelated. Physiologically, FKBP12 regulates the ryanodine receptor (RyR-1), an intracellular Ca 2ϩ -release channel (CRC) required for excitation-contraction coupling in skeletal muscle. The native CRC, isolated from the terminal cisternae (TC) of skeletal muscle sarcoplasmic reticulum (SR), is composed of four 565-kDa RyR-1 protomers and four FKBP12 molecules (6). FKBP-stripped CRC differs functionally from normal channels.
It is activated by lower concentrations of caffeine (6,7) or Ca 2ϩ (8,9), higher Mg 2ϩ concentrations are required for inactivation (9), and it has a greater open probability and displays longer mean open times in the full conductance state (8). These effects, reversed upon rebinding FKBP12, indicate that FKBP12 stabilizes a closed conformation of the channel. Cloned RyR-1, expressed in insect cells, also exhibits channel properties functionally different from those of the native CRC (7). Without FKBP12, the channel flickers among subconductance states, while co-expression of FKBP12 and RyR-1 generates channels opening to the full-conductance state (7), suggesting that FKBP12 insures cooperativity among RyR-1 protomers. FKBP12 may asymmetrically regulate ion flow through the channel, promoting the flow of Ca 2ϩ unidirectionally from the lumen of the SR to the cytoplasm during channel activation (10).
Until recently, FKBP12 was the sole FKBP believed to be relevant to FK506-mediated immunosuppression or toxicity. It had been the only FKBP known to be a potent mediator of FK506's inhibition of CaN in vitro (11) and signal transduction in Jurkat cells (12). Recently, a novel FKBP, FKBP12.6, was purified, sequenced, and characterized biochemically (13). Closely related to FKBP12, it has the same number of amino acids and 18 mostly conservative amino acid substitutions (13). The most striking substitution is that of a phenylalanine for a highly conserved tryptophan (13) forming the base of the drugbinding cavity. FKBP12.6 and FKBP12 have equal affinities for FK506 and are equipotent mediators of CaN inhibition by FK506 (13). Thus, FKBP12.6 has the potential to mediate the immunosuppression or toxicity associated with FK506 therapy.
We have cloned and expressed the cDNA encoding human FKBP12. 6. The characterization of human FKBP12. 6 in the presence and absence of its drug ligands is the subject of this report.

MATERIALS AND METHODS
Isolation of the cDNA Encoding Human FKBP12.6 -Using the bovine amino acid sequence, nested polymerase chain reaction (PCR) was used to isolate a cDNA fragment encoding human FKBP12.6. The primary sense primers corresponding to the amino-terminal six amino acids, MGVEIE, were ATGGGNGTNGARATAGA, ATGGGNGTNGA-RATCGA, and ATGGGNGTNGARATTGA. The primary antisense primers, corresponding to amino acids 81-86, VAYGAT, were GTNGC-NCCRTANGCGAC, GTNGCNCCRTANGCAAC, GTNGCNCCRTANG-CTAC, and GTNGCNCCRTANGCCAC. Using Taq Ultima polymerase (Perkin-Elmer), the primers were used in all 12 possible combinations in reactions performed according to the manufacturer's instructions.
The secondary sense primers, corresponding to amino acids 14 -19, RTFPKK, were (A/C)GNACNTTYCCNAAGAA and (A/C)GNACNTTY-CCNAAAAA. The secondary antisense primers, corresponding to amino acids 60 -65, FEEGAA, were GCNGCNCCYTCYTCGA and GCNGC-NCCYTCYTCAA. One l of a 1:10 dilution of each primary PCR reaction mixture was used in each of the four possible secondary PCR reactions (48 reactions total). The combination of the (A/C)GNACNT-TYCCNAAGAA and GCNGCNCCYTCYTCAA primers gave the greatest amount of 156-base pair product. These two primers, resynthesized with EcoRI linkers attached, were used to generate a product that was digested with EcoRI, subcloned into the EcoRI site of pUC19, and sequenced to confirm that it encoded a fragment of hFKBP12.6.
The remainder of the cDNA encoding hFKBP12.6 was cloned using the rapid amplification of cDNA ends) (RACE) technique. To obtain the 3Ј end of the cDNA, a sense primer, RCCTTTCAAGTTCAGAA, corresponding to a specific sequence obtained from the partial hFKBP12.6 cDNA clone obtained above, was used. The first antisense primer, GACTCGAGTCGACATCGATTTTTTTTTTTTTTTTT, anneals to poly(A) tracts. The PCR reactions were performed as described previously (14), and 1 l of the 10-fold diluted product was used in a secondary PCR reaction. In the second PCR reaction, the sense primer (corresponding to a specific nucleotide sequence in the original partial cDNA) was AAACAGGAAGTCATCAA, and the antisense primer was GACTCGAGTCGACATCG. The PCR reactions were performed as described for the primary amplification; the products were purified by agarose gel electrophoresis, and the major 800-base pair product was reamplified with the same set of secondary primers except that the sense primer contained an EcoRI linker. The product was cloned between the EcoRI and SalI sites of pUC19.
To obtain the 5Ј end of the gene, human brain 5Ј-RACE-Ready cDNA (Clontech) was used as the template. A 5Ј anchor primer was supplied by the manufacturer. The first antisense primer (corresponding to a nucleotide sequence in the original partial cDNA) used was TTGAT-GACTTCCTGTTTGCCAATTC. The PCR conditions were as described for the 3Ј RACE reactions. The products of the first reaction were diluted 10-fold, and those used in a second PCR reaction with the manufacturer's anchor primer and a second antisense primer, GAAAG-GYTTGTTTCTGTCTCTGGAT. The product of the secondary reaction was reamplified using an EcoRI-linkered antisense primer, and the product was subcloned into the EcoRI site of pUC19 (the Race-Ready cDNA contains an EcoRI site at the 5Ј end) and sequenced. Alignment of the 5Ј RACE product, the original PCR fragment, and the 3Ј RACE product generated a contiguous DNA sequence. To ensure that the product of the alignment represented one contiguous cDNA, EcoRIlinkered primers corresponding to the extreme 5Ј and 3Ј ends of the sequence were used to PCR the cDNA in one piece from human brain cDNA. The PCR product was subcloned into the EcoRI site of pUC19 and sequenced.
Bacterial Expression and Purification of hFKBP12. 6 -The open reading frame (ORF) of the hFKBP12.6 cDNA was subcloned from the complete cDNA by PCR using GAATTCCCATGGGCGTGGAGATCGAG as the sense primer and TTGGATCCTCACTCTAAGTTGAGCAG as the antisense primer. The PCR product was digested with NcoI and BamHI and subcloned between the NcoI and BamHI sites of the bacterial expression vector pET3d (Novagen), and the plasmid was transformed into BL21(DE3) cells. Expression of FKBP12.6 and production of the bacterial lysate were as described for FKBP12 (15). The lysate was dialyzed overnight against CM (5 mM sodium phosphate (pH 6.8), 1 mM EDTA, and 5 mM ␤-mercaptoethanol) buffer (13). The dialyzed protein was purified on a TosoHaas CM-3SW HPLC column (21.5 mm ϫ 15 cm) as described previously (13). Up to 75 mg of pure hFKBP12.6 were obtained per liter of Escherichia coli. Purified hFKBP12.6 may be stored at Ϫ70 or at 4°C.
Construction and Expression of a GST-hFKBP12.6 Fusion Gene-Sense (AAGCTTGGATCCGGCGTGGAGATCGAGACC) and antisense (AAGCTTTTGGATCCTCACTCTAAGTTGAGCAG) oligonucleotides were used in a PCR reaction to generate, from the hFKBP12.6 cDNA, a BamHI-linkered DNA fragment containing the ORF of hFKBP12.6. This fragment was digested with BamHI and subcloned into pGEX2T (Pharmacia Biotech) at the BamHI site. The plasmid was transformed into BL21(DE3) cells, and the GST-hFKBP12.6 fusion protein was expressed as described previously (16) with two modifications. First, induction with isopropyl-1-thio-␤-D-galactopyranoside proceeded overnight. Second, after elution from the glutathione-agarose affinity column, the protein-containing fractions were applied at a flow rate of 6 ml/min to a TosoHaas DEAE 3SW HPLC column (21.5 mm ϫ 15 cm) equilibrated in buffer containing 5 mM Tris (pH 7.8). The GST-FKBP12.6 fusion protein was eluted with a linear gradient of 0 -500 mM NaCl in 5 mM Tris (pH 7.8) buffer over a period of 1 h at a flow rate of 6 ml/min. Fractions (6 ml) were collected, and one protein peak was eluted at about 200 mM NaCl. About 250 mg of GST-hFKBP12.6 were obtained per liter of E. coli. The homogeneous fusion protein can be stored at 4°C or at Ϫ70°C.
Expression and Purification of hFKBP12, yFKBP12, hFKBP13-His6, hFKBP25, hFKBP52-His6, and hCyPA-Recombinant hFKBP12 was expressed and purified as described previously (15). The ORFs encoding yeast FKBP12 (yFKBP12) and human FKBP25 (hFKBP25) were generated by PCR, cloned into pET3d between the NcoI and BamHI sites, and transformed into BL21(DE3) cells. Both FKBPs were expressed as described for hFKBP12 (15). Whereas, yFKBP12 localizes to the periplasm and was purified as described for hFKBP12 (15), hFKBP25 localizes to the cytosol. The cell pellet from 1 liter of cells expressing hFKBP25 was resuspended in 10 ml of CM buffer and frozen in liquid nitrogen. After thawing, 1% (v/v) of Triton X-100, 2 mM phenylmethylsulfonyl fluoride, and 2 g/ml each of aprotinin, leupeptin, pepstatin, and trypsin inhibitor were added to the suspension. Cells were lysed in an ice bath using a Kontes microultrasonic cell disrupter (4 ϫ 15 s; maximum power) and debris was removed by centrifugation (35,000 ϫ g for 30 min). The extract was applied to a TosoHaas CM-3SW HPLC column (21.5 mm ϫ 15 cm) and equilibrated in CM buffer at a 6 ml/min flow rate. hFKBP25 was eluted from the column with a linear gradient of 0 -300 mM NaCl in CM buffer over a period of 1 h at a flow rate of 6 ml/min. hFKBP25 elutes between 100 and 150 mM NaCl and is homo-geneous. The purified material was fully active as determined by its complete binding to RAP-Sepharose and by its peptidyl-prolyl isomerase activity, which agreed with published values (for review, see Ref. 17). Purified hFKBP25 was stored at Ϫ70°C.
The ORF encoding the processed form of human FKBP13 (hFKBP13) fused to 10 histidine residues was generated by PCR using the sense oligonucleotide AGATATACCATGGGCCATCATCATCATCATCATC-ATCATCATCACAGCAGCGGCCATATCGACGACGACGACAAGACG-GGGGCCGAGGGCAAAAGG and the antisense oligonucleotide AAC-CTTGGATCCTTACAGCTCAGTTCGTCGCTC. The PCR product was digested with NcoI and BamHI and cloned between the NcoI and BamHI sites of pET3d, and the resulting plasmid was transformed into BL21(DE3) cells. Protein expression and cell lysis was as described for hFKBP25. hFKBP13 was applied to a nickel nitrilotriacetic acid (Qiagen) affinity column (1 cm ϫ 5 cm) and washed, and the pure protein was eluted with 250 mM imidazole (pH 7.0) according to the manufacturer's instructions. The ORF encoding human FKBP52 (hFKBP52) was generated by PCR using the sense oligonucleotide AATTGTCGAC-CATATGACAGCCGAGGAGATGAAGGCG and the antisense oligonucleotide AATTCTCGAGCTATGCTTCTGTCTCCACCTGAGA. The product was digested with NdeI and XhoI and cloned into pET15b (Novagen), a polyhistidine fusion vector. hFKBP52 was expressed and purified as described for hFKBP13 above. That recombinant hFKBP13 and hFKBP52 are fully active was confirmed by their complete binding to FK506-Sepharose and by their peptidyl-prolyl isomerase activities, which were in accord with published values (for review, see Ref. 17). hCyPA was expressed and purified to homogeneity as described previously (18). hFKBP13 and hFKBP52 were stored at Ϫ70°C, and hCyPA was stored at 4°C.
Protein Determinations-Protein determinations were performed according to the method of Bradford (19) using bovine plasma albumin (Bio-Rad) as the standard. The protein reagent concentrate was purchased from Bio-Rad.
FK506 Binding, CaN Phosphatase, and Peptidyl-prolyl Isomerase Assays-The LH-20 binding assay was performed as described previously (20) with the modifications noted (21). The CaN phosphatase assay has been described previously (22). When immunophilins were titrated, reaction mixtures (60 l) contained 40 mM Tris (pH 8), 100 mM NaCl, 6 mM Mg(OAc) 2 , 0.1 mM CaCl 2 , 0.1 mg/ml bovine serum albumin, 0.5 mM dithiothreitol, 190 nM bovine brain calmodulin (Sigma), 3 nM bovine brain CaN (Sigma), 50 M drug (FK506, CsA, or Ј818), and 40 M [ 33 P]RII peptide (Peptides International; 600 cpm/pmol). When drugs were titrated, the reaction mixtures were identical except that 50 M immunophilin was substituted for 50 M drug. The RII peptide, DLD-VPIPGRFDRRVSVAAE, was phosphorylated at the serine residue as described previously (22). The CaN phosphatase reaction mixtures were incubated for 30 min at 30°C, and dephosphorylation was initiated by peptide addition and allowed to proceed for 10 min at 30°C. Termination of the reaction and separation of free phosphate from phosphorylated peptide were performed as described previously (22). The peptidylprolyl isomerase assay was performed as described previously (13) on peptide substrates obtained from Bachem.
Expression of Transfected Plasmids-To confirm that transfected DNAs were expressed, 20 l of the supernatant from mock-transfected cells and from transfected cells were subjected to SDS-PAGE on a 16% denaturing gel (Novex) in parallel with concentration standards. Proteins were transferred to a membrane and analyzed by Western blotting as described previously (13). To detect hFKBP12, hFKBP12.6, or yFKBP12, the membrane was probed with a 1:10,000 dilution of an antipeptide antibody (R2565) directed against the sequence DVELLKLE. To detect hFKBP13 or hFKBP25, the membrane was probed with a 1:10,000 dilution of an antibody (R2816 or R2819, respectively) derived from injection of rabbits with the recombinantly produced purified protein. To detect CaN␣, the membrane was probed with a 1:10,000 dilution of an antipeptide antibody (R2930) specific for the sequence LSNSSNIQ. Peptide synthesis, chemistry for coupling peptides to thyroglobulin, and antiserum production have been described previously (13). Antibodies to detect hCyPA and hFKBP52 (Affinity Bioreagents) were used at a dilution of 1:1,000.
Preparation of Cardiac and Skeletal Muscle Sarcoplasmic Reticulum and Terminal Cisternae-The TC of skeletal muscle SR were isolated from dog or rabbit skeletal muscle as described previously (28). Cardiac microsomes were isolated from heart as described previously (29). Cardiac junctional SR was isolated from cardiac microsomes by sucrose density gradient centrifugation following Ca 2ϩ -phosphate loading as described previously (30).
[  (20) using the modifications described previously (6). Samples (3 l) of the solublized vesicles were incubated for 30 min at 37°C in binding mixture (volume, 60 l) containing between 1.0 and 30.0 nM [ 3 H]dihydro-FK506 (55,000 cpm/pmol). Nonspecific binding was determined by the addition of 1 M unlabeled L-683,590 (Ј590, Fig. 1), a potent FK506 analog (4). Following the incubation, 50 l of the sample (containing 5 g of cardiac SR or 1.0 g of skeletal muscle TC) were applied to a 2-ml Sephadex LH-20 column equilibrated in LH-20 column buffer (binding mixture without bovine serum albumin present) to separate free from bound ligand.
The high affinity [ 3 H]ryanodine binding site density in cardiac and skeletal muscle SR was determined by Scatchard analysis of [ 3 H]ryanodine binding isotherms as described previously (6). Because the native CRC contains one high affinity binding site for ryanodine, the B max value is proportional to the concentration of CRC present in SR. The stoichiometry of FKBP per CRC was calculated directly from the ratio of B max values for [ 3 H]dihydro-FK506 and [ 3 H]ryanodine binding as described previously (6).
Western Blot Analysis of Sarcoplasmic Reticulum Fractions-Muscle SR fractions were loaded onto SDS-PAGE gels separately or co-loaded with 30 ng of human recombinant FKBP12 or FKBP12.6. Proteins were separated by electrophoresis on a 12.5% polyacrylamide gel (12.5% acrylamide containing 2.6% cross-linker polymerized with 0.25% TEMED and 0.05% ammonium persulfate) at 90 V for 2 h. Western blotting
Purification of FKBP12.6 from Cardiac RyR-The cardiac RyR (RyR-2) was purified as described previously (33). FKBP12.6 was isolated from RyR-2 by dissociation of the FKBP⅐drug complex from the RyR essentially as described for isolation of FKBP12 from the skeletal muscle RyR (32). Briefly, 100 g of RyR-2 were incubated in Superose 6B column buffer (20 mM Tris-Cl (pH 7.4), 0.5 M KCl, 0.5% CHAPS, 2 mM dithiothreitol, and 1 g/ml leupeptin) containing 12 M Ј590 (volume, 650 l) for 1 h at room temperature. The RyR was then adsorbed to hydroxyapatite (0.1 ml) by incubation at 4°C for 1 h. The supernatant, containing the dissociated FKBP12.6⅐Ј590 complex, was obtained by low speed sedimentation of the hydroxyapatite resin in a 1.5-ml microcentrifuge tube.
Sequencing of the FKBP Associated with the Cardiac RyR-The FKBP associated with the cardiac RyR was subjected to automated Edman degradation to determine the amino-terminal amino acid sequence. 250 l of a 0.66 g/ml solution of the purified protein with FK506 bound in 0.5% CHAPS, 0.5 M KCl, 2 mM dithiothreitol, and 2 g/ml leupeptin were concentrated to 80 l in a Speed Vac concentrator. 70 l of the concentrated protein were applied to a treated 13-mm "peptide" glass fiber filter membrane (Porton Instruments, 19101) in three equal aliquots. The filter was allowed to dry between each protein application. After the last application, the filter was rinsed briefly in methanol to remove excess detergent. The filter was then installed in an Applied Biosystems 475A pulsed liquid protein sequencer and subjected to 14 cycles of automated degradation following the manufacturer's recommended program for use of the "peptide" support on the instrument. Automated analysis of phenylthiohydantoin-derivatives was performed with an on-line Applied Biosystems 120A narrow bore high performance liquid chromatography, and data were collected and analyzed on an Applied Biosystems 900A data station as described previously (34). Amino acid sequence data was obtained in 10 of the first 11 cycles. The calculated initial yield was 43%, based upon the amount of protein present in the original solution. Approximately 50 pmol of FKBP12 from skeletal muscle RyR were sequenced separately as a control and gave the expected amino acid sequence for 14 cycles of automated Edman degradation. Analysis of the data indicated an initial yield of 100% and a repetitive yield of 93%.
Affinity Purification of mTOR and CaN From Rat Brain-Preparation of rat brain extracts and affinity purification of CaN and mTOR with the GST-hFKBP12⅐FK506 and the GST-hFKBP12⅐RAP complexes, respectively, have been described previously (16). In some experiments, a GST-FKBP12.6 fusion protein was substituted for the GST-FKBP12 fusion protein with no other changes to the procedure. To detect mTOR by Western blotting, a rabbit antiserum developed against the yeast TOR2 peptide HDLELAVPG (amino acids 2074 -2082) was prepared by methods described previously (13). This antiserum recognizes the corresponding sequence, RDLELAVPG (amino acids 2134 -2142), in rat mTOR. Western blots were performed as described previously (13), and the blot was probed with a 1:7500 dilution of the anti-mTOR peptide antiserum. To detect CaN␣, rabbit antisera were developed against the human CaN␣ gene 1 peptide SNSSNIQ (amino acids 515-521), the human CaN␣ gene 2 peptide TGNHTAQ (amino acids 518 -524), and the human CaN␣ gene 3 peptide QGKKAHS (amino acids 496 -502). These sequences are well conserved in the corresponding rat CaNs. The antibodies were diluted 1:10,000 and used to probe Western blots.

RESULTS
Cloning of Human FKBP12.6 -Reasoning that bovine and human FKBP12.6 (hFKBP12.6) would be highly conserved, PCR primers based upon the bovine sequence (13) were used to isolate an 880-base pair human brain cDNA (Fig. 2) encoding a 108-amino acid protein identical in sequence to that of the bovine protein (13). After we had obtained the clone, the same cDNA, obtained by random sequencing, was reported (35). Northern blotting of oligo(dT)-purified RNA from a variety of human tissues shows that steady-state hFKBP12.6 mRNA levels are highest in brain and thymus (Fig. 3, panels A and B). Because FK506 has significant adverse neurologic side effects (36), the steady-state hFKBP12 and hFKBP12.6 mRNA levels in anatomically distinct regions of the brain were compared (Fig. 3, panels E and F). Relative to hFKBP12 message levels, hFKBP12.6 message levels in the brain are lower overall (note the different exposure times). However, the steady-state levels of the hFKBP12 and hFKBP12.6 mRNAs in each section of the brain parallel one another, with the highest levels of both transcripts located in the caudate nucleus and the lowest levels located in the corpus callosum and substantia nigra.
Characteristics of Recombinant Human FKBP12.6 -The cDNA encoding hFKBP12.6 was expressed in E. coli. The protein localizes to the periplasm and is purified to homogeneity (Fig. 4) in one chromatography step on a CM column. Recombinant hFKBP12.6 binds FK506 with the same affinity (K d ϭ 0.5 nM) as both purified bovine brain FKBP12.6 (13) and recombinant hFKBP12 (data not shown). The specific binding activities, measured using a modified LH-20 assay, of recombinant yFKBP12, hFKBP12, and hFKBP12.6 are approximately 30 ng of [ 3 H]dihydro-FK506/g of protein, a value in good agreement with that measured for FKBP12 (21). Because the Bradford assay used to measure protein concentration is known to overestimate FKBP concentration by 2.5-fold (21), the binding activities are close to the theoretical maximum expected for a 1:1 molar complex between FKBP and FK506. Despite a calculated molecular weight slightly less than that of hFKBP12, recombinant hFKBP12.6, like purified bovine brain FKBP12.6 (13), migrates more slowly on denaturing gels than hFKBP12 (Fig. 4).
The catalytic efficiency (k cat /K m ) of hFKBP12 toward peptidyl-prolyl substrates correlates strongly with the hydrophobicity of the amino acid immediately preceding the proline (37). This contrasts with the promiscuous peptidyl-prolyl isomerase substrate specificity observed with CyPA, the binding protein for the structurally unrelated immunosuppressive drug, CsA. We compared the abilities of purified recombinant hFKBP12.6 and hFKBP12 to catalyze the isomerization to the trans form of tetrapeptides of the general structure N-succinyl-Ala-Xaa-cis-Pro-Phe-p-nitroanilide where Xaa is any one of 12 amino acids (Table I). hFKBP12.6 exhibits substrate preferences similar, but not identical, to those observed for hFKBP12. As with hFKBP12, substrates in which a hydrophobic amino acid precedes proline are greatly preferred by hFKBP12.6. For both FKBPs, the most reactive substrates have Leu, Ile, Phe, or Nle at the Xaa position, while the least reactive substrate has Gly at the Xaa position. With most of the tetrapeptide substrates tested, the catalytic efficiency of hFKBP12.6 is roughly 2-fold lower than that observed with hFKBP12. When Xaa is Val or Nle, the catalytic efficiencies of hFKBP12 and hFKBP12.6 are about equal. Only with the His-Pro substrate does hFKBP12.6 exhibit more reactivity than hFKBP12.
Heart Muscle RyR Is Associated with FKBP12.6 -It is wellestablished that FKBP12 is associated with the skeletal muscle RyR (RyR-1) (6,31,38), stabilizing calcium flux through the CRC (6 -8). RyR-2 in heart muscle is an isoform distinct from RyR-1 in skeletal muscle and associates with a novel, uncharacterized FKBP, termed FKBP-Cardiac (FKBP-C) (32). Like both bovine (13) and human FKBP12.6 (Fig. 4), FKBP-C migrates slightly slower than hFKBP12 on SDS gels (32), suggesting that FKBP-C and FKBP12.6 are the same protein. To confirm their identity, TC preparations from canine skeletal and heart muscle SR fractions were analyzed by Western blotting using an antibody that recognizes both FKBPs (Fig. 5). That the immunoreactive bands in the skeletal and cardiac muscle fractions are FKBP12 and FKBP12.6, respectively, is confirmed by co-loading the fractions with recombinant hFKBP12 or hFKBP12.6. In skeletal muscle TC, the immunoreactive band (lane 1) has the same mobility as the FKBP12 standard (lane 7). When co-loaded in the same well with either hFKBP12 (lane 2) or hFKBP12.6 (lane 3), the skeletal muscle band co-migrates with hFKBP12, whereas hFKBP12.6 is well separated. Therefore, as observed with the rabbit skeletal muscle RyR (6,31), the canine skeletal muscle RyR is associated with FKBP12. In contrast, in cardiac SR fractions (lane 4), the antibody detects a band with somewhat slower mobility. Because the cardiac SR fractions were isolated in the presence of 0.6 M KCl, the immunoreactive bands have broadened (lanes 4-6) and have slightly reduced mobilities relative to the standards (lanes 7 and 8). Nevertheless, when co-loaded in the same well with either hFKBP12 (lane 5) or hFKBP12.6 (lane 6), it is apparent that the immunoreactive band in cardiac SR migrates with hFKBP12.6, whereas hFKBP12 is well separated. These results indicate that the FKBP associated with the canine heart RyR, previously called FKBP-C, is FKBP12.6.  Confirmation that FKBP12.6 is associated with the cardiac muscle RyR was obtained by amino-terminal sequencing of FKBP-C obtained from purified canine cardiac RyR preparations. FKBP-C was stripped from purified cardiac RyR with FK506 and separated from the RyR by hydroxyapatite chromatography. Amino-terminal sequencing of the purified protein gave the eleven amino acid sequence GVEIETISXGD, identical to the amino-terminal sequence of both bovine and human FKBP12.6 and different in two amino acids from the 11 aminoterminal amino acids of both bovine and human FKBP12, GVQVETISPGD. The observations that the RyR purified from canine heart is associated with a protein that co-migrates with FKBP12.6 on denaturing gels and has the same amino-terminal sequence as both bovine and human FKBP12.6 indicates that FKBP12.6 is specifically associated with the canine heart RyR. In the cytosol of dog heart and in canine skeletal muscle TC, only FKBP12 has been detected (32), indicating that the interaction between FKBP12.6 and the heart RyR is specific and not due to the absence of FKBP12 in heart muscle.
There Are Four FKBP12. 6 Molecules per Heart Muscle RyR-The binding isotherm of [ 3 H]dihydro-FK506 to canine cardiac muscle TC is a simple hyperbola with Scatchard analysis yielding a straight line indicative of a single class of FK506 binding site (data not shown). The binding parameters, obtained from five different TC preparations, give a dissociation constant (K d ) of 13.2 Ϯ 4.8 nM and a B max of 25.1 Ϯ 5.5 pmol/mg of protein (Table II). The affinity of the interaction between FKBP12 and FK506 is lower than reported (21) (0.4 -0.8 nM) due to the presence of 0.5% CHAPS in the assay (6). The CRC contains a single high affinity ryanodine binding site/homotetramer (33,39). Therefore, the ratio of [ 3 H]dihydro-FK506 binding to ryanodine binding is a measure of the FKBP12.6:RyR-2 protomer stoichiometry. Table II compares the number of [ 3 H]dihydro-FK506 binding sites to ryanodine binding sites in several SR and TC preparations from rabbit and dog. The stoichiometry is approximately 4 mol of FKBP12.6/mol of canine cardiac muscle RyR homotetramer. This ratio is equivalent to the FKBP12: RyR-1 ratio observed in skeletal muscle (Table II) (6). Thus, the structure of the native CRC in canine heart muscle SR can be represented as (FKBP12.6) 4 (RyR-2 protomer) 4 .
The observation that FKBP12.6 associates specifically with the cardiac CRC makes it likely that FKBP12.6 modulates channel gating of the cardiac isoform in a manner similar to that observed for modulation of the skeletal muscle RyR-1 by FKBP12. Thus, the few amino acid differences between FKBP12 and FKBP12.6 have important consequences for channel binding specificity. We have expanded our characterization of hFKBP12.6 and have performed a pharmacological comparison of hFKBP12.6 and hFKBP12 both in vitro and in Jurkat cells in an effort to uncover differences between the two molecules that might help to explain their apparently different physiological roles.
The hFKBP12.6⅐FK506 Complex is a Potent CaN Inhibitor In Vitro-We have compared the abilities of all known human FKBPs (hFKBPs), yeast FKBP12 (yFKBP12), and human CyPA (hCyPA) to mediate inhibition of CaN phosphatase activity by FK506 in vitro. All FKBPs and hCyPA were expressed in E. coli and purified to homogeneity. hFKBP13 and hFKBP52 were histidine-tagged to aid in their purification while all other FKBPs, as well as hCyPA, contained only native sequences.
The phosphatase assays were designed to measure CaN inhibition by the various immunophilin-drug complexes and to minimize any effect made by the equilibrium between the complex and the free immunophilin and drug molecules. Therefore, immunophilin-drug complex formation at a particular immunophilin or drug concentration was maximized by having an excess of one component. In one set of assays (Fig. 6A)  values (legend to Fig. 6) obtained for CaN inhibition by a particular immunophilin-drug complex are similar to one another. Both the drug and immunophilin titrations (Fig. 6, A and B, respectively) demonstrate that the FK506 complexes with hFKBP12.6 and hFKBP12 are equipotent to one another and to the yFKBP12⅐FK506 complex as CaN inhibitors. As a control, and to further validate the assay, the ability of the hCyPA⅐CsA complex to inhibit CaN was measured. In agreement with observations that, by several criteria, CsA is 10 -100-fold less potent than FK506 (for review, see Ref. 40), the hCyPA⅐CsA complex was about 15-fold less active than the hFKBP12⅐FK506 complex as a CaN inhibitor (Fig. 6, A and B).
The remaining hFKBP⅐FK506 complexes are very poor CaN inhibitors. hFKBP25 is unable to inhibit CaN at even the highest drug and immunophilin concentrations tested. The hFKBP13⅐FK506 and hFKBP52⅐FK506 complexes are very weak inhibitors of CaN activity. Phosphatase inhibition by the latter two complexes is observed at immunophilin-drug concentrations unlikely to be attained within most cells. Thus, the hFKBP13⅐FK506 and hFKBP52⅐FK506 complexes may not make significant contributions to CaN-dependent immunosuppression or toxicity. The FKBP12.6⅐Ј818 Complex Is a Weak CaN Inhibitor in Vitro-Despite having only minor differences from FK506 at two positions, C-18 and C-21 (Fig. 1), Ј818 is a powerful in vivo antagonist of the immunosuppressive and toxic side effects of FK506 (4). The C-18 hydroxyl group is responsible for the antagonism exerted by Ј818 since Ј590 (Fig. 1), identical to Ј818 except for the C-18 hydroxyl group, is almost equivalent to FK506 with respect to immunosuppressive potency and toxicity (4). Ј818 binds tightly to hFKBP12, displacing FK506 (K d values are 0.7 and 0.4 nM for Ј818 and FK506, respectively), but the resulting hFKBP12⅐Ј818 complex is a poor CaN inhibitor (4,5). These observations suggest that the C-18 position of FK506, on the solvent-exposed face of the FKBP12⅐drug complex (41), contacts CaN, since the added hydroxyl group in Ј818 abolishes CaN binding, either by forcing the hydrophilic hydroxyl group into an unfavorable hydrophobic environment or through steric hindrance.
In noteworthy contrast to hFKBP12, the yFKBP12 complex with Ј818 is a potent CaN inhibitor (5). Surrounding FK506 on the solvent-exposed surface of the complex are approximately 26 amino acids that are likely to be close to CaN. Throughout these 26 residues, there are 10 differences between yeast and human FKBP12. Because the three-dimensional structures of the yeast and human FKBP12⅐Ј818 complexes are almost identical, among the 10 changes in yFKBP12 there must be amino acids that directly interact with CaN, thereby compensating for Ј818 (42). These amino acids in yFKBP12 in some way neutralize the effects of the hydroxyl group at C-18 (5). In the absence of a crystal structure for the FKBP12⅐FK506⅐CaN complex, Ј818 is a useful pharmacological probe for helping to identify FKBP residues that might interact with CaN.
Of the 18 amino acid differences between hFKBP12.6 and hFKBP12, two residue changes in hFKBP12.6 (Arg 50 and Val 91 in Fig. 2), are among those 26 surface residues that surround the drug on the face of the complex. Because Ј818 uncovered differences between yeast and human FKBP12 not observed with FK506, we believed it might also uncover differences between hFKBP12 and hFKBP12.6. Therefore, the FKBP12.6⅐Ј818 complex as well as the Ј818 complexes with all other hFKBPs and yFKBP12 were tested for CaN inhibition. As before, two versions of the assay, titrating drug (Fig. 6C) and titrating immunophilin (Fig. 6D), were performed with both assays giving similar IC 50 values (legend to Fig. 6). As described previously, yFKBP12 is a surprisingly potent CaN inhibitor when complexed with Ј818 (5), albeit 24-fold less potent than when complexed with FK506 (Fig. 6). When sufficiently high concentrations (M) of immunophilin are present, we find that the hFKBP12⅐Ј818 complex can inhibit CaN, in contrast to results from previous assays (4,5) where the submicromolar hFKBP12 concentrations used were insufficient to bind the micromolar amounts of Ј818 required for detectable inhibitory activity. The Ј818 complex with hFKBP12.6 is equipotent to the hFKBP12⅐Ј818 complex as a CaN inhibitor. However, both complexes are 200 -400-fold poorer CaN inhibitors than the FK506 versions of the complexes. Therefore, unlike yFKBP12, none of the amino acid differences in hFKBP12.6 relative to hFKBP12 are able to compensate for the hydroxyl group of Ј818. In the presence of Ј818, none of the other FKBPs can inhibit CaN, corroborating the observations made with FK506.
FKBP12. 6 Can Mediate FK506-Sensitivity in a T-Cell Line-CaN inhibition in vitro does not always correspond to inhibition of CaN-dependent signaling pathways in vivo. For example, although the cyclophilin C⅐CsA complex inhibits CaN in vitro, cyclophilin C does not mediate CsA-sensitivity in Jurkat cells (12). The ability of hFKBP12.6 to mediate FK506-sensitivity in T-cells was therefore examined. Previously, the effects of overexpressing hFKBP12, hFKBP13, hFKBP25, and three of the human cyclophilins on the IC 50 of FK506 and CsA were measured (12). Overexpression of an immunophilin will result, at equilibrium, in a greater concentration of the immunophilin⅐ drug⅐CaN complex at a particular drug concentration. The previous experiments showed that overexpression of hFKBP12 rendered Jurkat cells 2-3-fold more sensitive to FK506, whereas overexpression of hFKBP13 and of hFKBP25 had no effect (12). This indicated that, among the three FKBPs tested, only hFKBP12 can mediate the inhibitory effects of FK506 in cells.
We performed an assay similar to the one described (12), but to increase sensitivity, we incorporated an important modification. Overexpression of the catalytic subunit of CaN (CaN␣) is known to render activated Jurkat cells 4 -5-fold less sensitive to the effects of FK506 and CsA (3,43). Our modification was to overexpress the cDNAs encoding both the catalytic (CaN␣) and regulatory (CaN␤) subunits of CaN by transient transfection in Jurkat cells using the mammalian expression vector pcDL-SR␣296 (SR␣) (23). Protein overexpression was confirmed by Western blot analysis comparing extracts from transfected and nontransfected cells (Fig. 7), and drug sensitivity was quantitated by measuring ␤-galactosidase production from a co-transfected reporter plasmid, pIL2.Gal, containing the IL-2 promoter fused to the ␤-galactosidase reporter gene. Overexpression of both CaN␣ and CaN␤ rendered the cells insensitive to the effects of up to 10 nM FK506, almost 1000 times the normal IC 50 , 0.012 nM (Fig. 8A). There are two possible explanations for the drug insensitivity generated by CaN overexpression. Either the CaN is ectopically expressed in a subcellular compartment, separating it from the hFKBP12⅐FK506 complex, or the CaN levels have exceeded the FKBP12 levels such that, even at the highest drug concentrations, there is sufficient free CaN available to participate in the signaling pathway. If the second explanation is correct, then co-expression of hFKBP12 will revert the cells to drug sensitivity and we will have generated an assay with a greatly amplified readout relative to the 2-3-fold shift in IC 50 observed in the previous assay (12). To test between these alternatives, Jurkat cells overexpressing CaN␣ and CaN␤ were co-transfected with expression constructs encoding the FKBPs tested in the previous assay (hFKBP12, hFKBP13, and hFKBP25) (12). Overexpression of hFKBP12, confirmed by Western blotting (Fig. 7), reestablishes the FK506-sensitivity of the cells (Fig.  8A). Reflecting the greatly increased amount of CaN that must be inhibited, the IC 50 (0.26 nM) has shifted 20-fold relative to nontransfected cells (0.012 nM). This result indicates that CaNoverexpression is cytosolic and that hFKBP12 mediates FK506-sensitivity in a T-cell line, thereby validating our assay. In contrast, sensitivity to FK506 is not recovered upon hFKBP25 overexpression (Fig. 8A), confirming the results of Bram et al. (12) and our result that the hFKBP25⅐FK506 complex cannot inhibit CaN in vitro. We find that hFKBP13 has some ability to mediate the inhibitory effects of FK506, consistent with our observation (see Fig. 6) and with the observations of others (44,45) that the hFKBP13⅐FK506 complex can inhibit, albeit weakly, CaN phosphatase activity.
Finally, the ability of hFKBP12.6 to restore FK506 sensitivity in CaN-overexpressing Jurkat cells was examined. We also tested hFKBP52 and yFKBP12 since the FK506 complexes with these FKBPs have been characterized for CaN inhibition, in vitro (5,46), but have not been tested in Jurkat cells. Transfection of the CaN-overexpressing Jurkat cells with the cDNA encoding hFKBP12.6 restored FK506-sensitivity (Fig. 8B). The IC 50 of FK506 in the cells overexpressing hFKBP12.6 is 0.18 nM, demonstrating that hFKBP12.6 is equipotent to hFKBP12 at mediating the inhibitory effects of FK506 upon CaN-dependent signaling events in Jurkat cells and corroborating our result that the hFKBP12.6 and hFKBP12 complexes with FK506 are equipotent CaN inhibitors in vitro. Also in agreement with the CaN phosphatase assay, yFKBP12 is equipotent to both hFKBP12 and hFKBP12.6 at mediating the inhibitory effects of FK506 in Jurkat cells (Fig. 8B). In contrast, the FK506 complex with hFKBP52 is only a weak inhibitor of signaling since, even at 10 M drug concentrations, it is unable to completely block IL-2 promoter activation (Fig. 8B).
T-Cells Can Be Made Sensitive to Ј818 by Overexpression of hFKBP12.6 -To confirm the efficacy of the various FKBPs at mediating the inhibitory effects of FK506, we used the antagonist and close FK506 relative, Ј818, which cannot block IL-2 promoter stimulation in activated Jurkat cells containing wildtype levels of CaN (Fig. 8C). We have previously demonstrated (Fig. 6, C and D) that, at high hFKBP12 or hFKBP12.6 concentrations, Ј818 can inhibit CaN in vitro. Therefore, we reasoned that by overexpressing FKBPs relevant to the inhibitory effects of FK506, the IL-2 promoter could be made sensitive to Ј818, thereby converting it from an antagonist to an agonist. Because Ј818 is such a weak agonist, its use in this assay provides a more stringent test of the abilities of the various FKBPs to mediate inhibition of FK506-sensitive signaling pathways in the cell. Moreover, it allows distinctions to be made between FKBPs, such as between yFKBP12 and hFKBP12, that cannot observed when FK506 is used. Therefore, expression plasmids encoding each of the known human FKBPs (as well as yFKBP12) were co-transfected into Jurkat cells with the pIL2.Gal reporter plasmid activated in the presence of Ј818 and the amount of ␤-galactosidase produced measured.
Overexpression of hFKBP12.6 renders activated Jurkat cells sensitive to Ј818 (Fig. 8C). Moreover, hFKBP12.6 and hFKBP12 are equipotent in this assay, in agreement with their equal abilities to mediate FK506-sensitivity in vivo and in accord with their abilities, when complexed with Ј818, to inhibit CaN in vitro. Further validation of these transfection experiments as an accurate measure of the ability of FKBPs to mediate drug sensitivity was obtained when the yFKBP12 expression plasmid was transfected into Jurkat cells. IL-2 promoter activity in activated cells is strikingly sensitive to the yFKBP12⅐Ј818 complex (Fig. 8C), again correlating with the results obtained in the CaN phosphatase assay. Overexpression of hFKBPs 13, 25, or 52 had no effect on the Ј818 sensitivity of the IL-2 promoter, confirming that these FKBP complexes with FK506-like drugs are ineffective inhibitors of the signaling pathway to IL-2 gene transcription.
The CsA sensitivity of the IL-2 promoter in activated Jurkat cells was measured as a control. Transfections with the cDNAs encoding the CaN␣ and CaN␤ subunits rendered the promoter less sensitive but not insensitive to CsA (Fig. 8D), which differs from the results obtained with FK506. The inability to make the cells insensitive to CsA upon CaN overexpression is due to naturally high CyPA levels in Jurkat cells (Fig. 7, row G). Scanning densitometry (data not shown) of the bands in Fig. 7 shows that the molar concentration of endogenous CyPA in Jurkat cells is still 7-fold greater than CaN␣, even when the latter protein is overexpressed (Fig. 7, compare lanes F and G). cells cannot be rendered completely insensitive to CsA. Nevertheless, the CaN-overexpressing cells are less sensitive to CsA because there is more CaN to inhibit (Fig. 8D). The decreased sensitivity to CsA is reversed 9-fold in the CaN-overexpressing cells by co-transfection with the cDNA encoding hCyPA (Fig.  8D) thereby confirming previous results (12) that hCyPA can mediate the inhibitory effects of CsA. The FKBPs are specific for mediating the inhibitory effects of FK506 and not of CsA because FKBP overexpression does not alter the CsA-sensitivity of the IL-2 promoter (Fig. 8D).
The hFKBP12.6⅐RAP Complex Binds mTOR-The hFKBP12⅐ RAP complex binds to a 288-kDa protein, mTOR, that has been isolated from rat brain (16,47), bovine thymus (48), and a human T-leukemic cell line (49). To determine if the hFKBP12.6⅐RAP complex also binds mTOR, purified GST-hFKBP12.6 or GST-hFKBP12 fusion protein was incubated with a rat brain extract in the absence or presence of 10 M RAP, and protein complexes were precipitated with glutathione-agarose beads. After washing the beads to remove proteins binding nonspecifically, proteins were eluted from the beads with reducing SDS-sample buffer (16), subjected to electrophoresis on denaturing gels, and analyzed by Western blotting using antipeptide antibodies that recognize mTOR or CaN␣. The RAP complex with hFKBP12.6, as described previously for hFKBP12, binds specifically to a high molecular weight protein previously identified as mTOR (Fig. 9, lanes 5 and 8). This experiment, repeated 4 times, reproducibly shows that the RAP complex with GST-FKBP12.6 precipitates less mTOR than the GST-FKBP12 complex. As observed with FKBP12, FKBP12.6 binding to mTOR is dependent upon RAP and is not observed in the presence of FK506. In the presence of FK506, the GST-FKBP complexes, as expected, precipitate CaN (Fig. 9, lanes 6  and 9). DISCUSSION We have cloned the cDNA encoding human FKBP12.6 and have characterized the expressed protein pharmacologically and physiologically. Physiologically, FKBP12.6 has a role distinct from that of FKBP12. FKBP12 is associated with RyR-1 of skeletal muscle SR, whereas FKBP12.6 is specifically associated with RyR-2 of cardiac muscle SR. Pharmacologically, FKBP12.6 is almost indistinguishable from FKBP12. FKBP12.6 is the only other FKBP family member equipotent to FKBP12 at inhibiting CaN in vitro and at mediating the FK506-sensitivity of a CaN-dependent signal transduction pathway. Moreover, when complexed with RAP, FKBP12.6, like FKBP12, binds mTOR.
The cardiac CRC (RyR-2) is a 565-kDa protomer 64% identical to RyR-1 (50,51). The hydropathy profiles and predicted secondary structures of the cardiac and skeletal isoforms are virtually identical (51). Both are activated by Ca 2ϩ , ATP, and caffeine; both are inactivated by Mg 2ϩ and ruthenium red; and both contain one high affinity and several low affinity ryanodine binding sites (52). Although morphologically and functionally similar, the channels are not identical (52). We have shown that FKBP-C (32), isolated from the canine cardiac RyR, comigrates with hFKBP12.6 on SDS-PAGE gels and has the same 11-amino acid amino-terminal sequence as both bovine and human FKBP12.6. Our finding that there are four FK506 binding sites per high affinity ryanodine binding site in cardiac SR suggests that the structure of the cardiac CRC can be represented as (RyR-2) 4 (FKBP12.6) 4 , analogous to the structure of the skeletal muscle CRC, (RyR-1) 4 (FKBP12) 4 . Thus, the native CRC isoforms in heart and skeletal muscle SR are further distinguished from one another in that different FKBP isoforms comprise a portion of their structures. The structural and functional similarities between FKBP12.6 and FKBP12 and between the cardiac and skeletal muscle RyR isoforms, suggests that the role of FKBP12.6 in the native cardiac CRC is similar to the role of FKBP12 in the skeletal muscle CRC.
The stoichiometry of four molecules of FKBP12.6 per native tetrameric CRC obtained by [ 3 H]dihydro-FK506 binding isotherms relies on the assumption that the ryanodine receptor is the predominant or only SR protein that binds FKBP12.6. Recent studies confirm that this is the case. Endogenous FKBP of cardiac SR was exchanged with the GST-FKBP12.6 fusion protein using exchange methodology developed for the skeletal muscle RyR (9). The TC was then solublized with CHAPS, and protein complexes with the GST-FKBP12.6 fusion protein were affinity purified on a GST-Sepharose affinity column. RyR-2 was the predominant protein in the SR that was tightly bound to GST-FKBP12.6. 2 In the presence of drug, the abilities of several human FK-BPs to inhibit CaN in vitro and a CaN-dependent signaling pathway in cells have been compared. The abilities of the immunophilin-drug complexes to inhibit CaN and their abilities to block IL-2 transcription correlate precisely. The FK506 complexes with FKBP13 and FKBP52, weak CaN inhibitors in vitro, are weak inhibitors of IL-2 promoter activity when the proteins are overexpressed in Jurkat cells. This correlation extends to the yFKBP12⅐Ј818 complex, a more potent CaN inhibitor and a more potent inhibitor of IL-2 promoter stimulation than the Ј818 complexes with hFKBP12 or hFKBP12.6. Thus, our results support the proposed mechanism of action of FK506 in which CaN inhibition blocks IL-2 transcription, thereby preventing T-cell activation (54). The RAP complexes with hFKBP12.6 and hFKBP12 exhibit similar but not identi- cal properties. Qualitatively, the hFKBP12⅐RAP complex binds more mTOR than the hFKBP12.6⅐RAP complex, suggesting that hFKBP12 plays a greater role in mediating RAP's antiproliferative effects in human cells.
The adverse side effects of FK506 immunotherapy include nephrotoxicity, diabetogenicity, the development of lymphoproliferative disorders, expressive aphasia, seizures, coma, drowsiness, lethargy, tremors, and aggressiveness (36,55). Toxicities associated with RAP treatment in non-rodent mammals include vomiting, diarrhea, thrombocytopenia, and gastrointestinal ulceration (56). Because inhibition of IL-2 promoter activity in T-cells is a convenient measure of the in vivo potentials of FKBP⅐FK506 complexes to inhibit a CaN-dependent signaling pathway, our data reflect the impact that the various FKBP⅐FK506 complexes can have upon a physiological process involving CaN. The response of any single cell to FK506 or RAP will depend upon several factors including 1) the expression levels of the immunophilins mediating drug action; 2) the abilities of the immunophilin-drug complexes to interact with their immediate targets; 3) the concentration of FK506 or RAP that gets into the cell; 4) the concentration of CaN or mTOR in the cell; and 5) the function of the downstream substrates of CaN or mTOR and their importance to the cell. Thus, the therapeutic and toxic side effects of FK506 and RAP are likely due to the formation of multiple FKBP⅐drug complexes with varying affinities for their pharmacologic target proteins, CaN and mTOR. Although nothing is known about mTOR substrates, CaN substrates that are candidate proteins responsible for the deleterious effects of FK506 at recognized sites of toxicity include the Na ϩ ,K ϩ -ATPase in the brain and kidney (57) and nitric oxide synthase in the brain (58). There have been no reports of FK506 toxicity that would implicate the CRC. Because only a portion of the cellular FKBP pool is bound by FK506 at immunosuppressive doses, the cytosolic FKBP in cardiac and skeletal muscle may act as a buffer shielding the CRC from the effects of the drug.
The design of one of our assays (Fig. 8), reversing the effects of CaN-overexpression, underscores the importance of the cellular FKBP:CaN ratio in establishing FK506-sensitivity. FKBP12 is both ubiquitous and abundant (59) and a survey of 15 different rat tissues documented FKBP/CaN ratios ranging from 13 to 343 (60). The lowest FKBP/CaN ratios were found in seven anatomically distinct regions of the brain, a reflection of the great abundance (61) and probable importance, of CaN there. FKBP/CaN ratios in the thymus and spleen are among the highest, approximately 200 and 100, respectively (60). If the FKBP/CaN ratios in spleen and thymus reflect ratios to be found in lymphocytes, then the Law of Mass Action dictates that, in the presence of FK506, a greater proportion of CaN will be inhibited in lymphocytes than in brain cells (even assuming that the drug is distributed equally to the brain, which it is not). In both Jurkat and murine T-cells, the intracellular concentration of FKBP is 6 -7 M with only 3-5% of the FKBP pool bound by FK506 at the IC 50 for inhibition of cellular activation (53). To decrease neurotoxic effects associated with FK506 therapy, one strategy would be to design FK506 analogs with decreased affinity for FKBP, thereby allowing equilibrium to relieve toxicity. With decreased affinity for FKBP, less FKBP⅐drug complex would be formed to inhibit the high levels of CaN in brain. In lymphoid cells, the high FKBP/CaN ratio would compensate for the lower affinity of the novel analog for FKBP, thereby maintaining the immunosuppressive efficacy of the drug. Similar considerations apply to the toxic effects associated with RAP therapy.
The substrate preferences exhibited by FKBP12.6 and FKBP12 in the peptidyl-prolyl isomerase assay overlap almost completely. Therefore, our observation that FKBP12.6 and FKBP12 are physiologically associated with distinct RyR isoforms may be unrelated to peptidyl-prolyl isomerase activity. Both in the presence and absence of drug, the biochemical and cellular read-outs used in our study have demonstrated that FKBP12.6 is highly similar to FKBP12 and that, where it is abundant, FKBP12.6 will be an important mediator of the inhibitory effects of FK506 and RAP. Our inability to uncover any significant biochemical or pharmacological differences between the two immunophilins that might account for their overlapping, and yet distinct, physiological roles suggests further complexities among the FKBPs remaining to be understood.