(Received for publication, November 11, 1996, and in revised form, December 6, 1996)
From the Department of Biochemistry, Tohoku University School of Medicine, Sendai 980-77, Miyagi, Japan
Cyclic ADP-ribose (cADPR) is a second messenger for Ca2+ mobilization via the ryanodine receptor (RyR) from islet microsomes for insulin secretion (Takasawa, S., Nata, K., Yonekura, H., and Okamoto, H. (1993) Science 259, 370-373). In the present study, FK506, an immunosuppressant that prolongs allograft survival, as well as cADPR were found to induce the release of Ca2+ from islet microsomes. After islet microsomes were treated with FK506, the Ca2+ release by cADPR from microsomes was reduced. cADPR as well as FK506 bound to FK506-binding protein 12.6 (FKBP12.6), which we also found occurs naturally in islet microsomes. When islet microsomes were treated with cADPR, FKBP12.6 dissociated from the microsomes and moved to the supernatant, releasing Ca2+ from the intracellular stores. The microsomes that were then devoid of FKBP12.6 did not show Ca2+ release by cADPR. These results strongly suggest that cADPR may be the ligand for FKBP12.6 in islet RyR and that the binding of cADPR to FKBP12.6 frees the RyR from FKBP12.6, causing it to release Ca2+.
Glucose is the primary stimulus of insulin secretion and synthesis in the pancreatic islets of Langerhans (1-3). Cyclic ADP-ribose (cADPR)1 is generated in pancreatic islets by glucose stimulation, serving as a second messenger for Ca2+ mobilization in the endoplasmic reticulum to secrete insulin (4-6). cADPR activates the ryanodine receptor (RyR) of a variety of cells to release Ca2+ from the intracellular stores (4, 6-16). RyRs have been purified from both skeletal and cardiac muscle (17, 18), and FK506-binding protein 12 (FKBP12) and FK506-binding protein 12.6 (FKBP12.6) were copurified with type 1 RyR from striated muscle and with type 2 RyR from cardiac muscle, respectively (19, 20). FKBP12 and FKBP12.6 were shown to bind selectively to type 1 and type 2 RyR, respectively (21). It was reported that the type 1 RyR was activated by dissociation of FKBP12 from the RyR by the addition of FK506 to release Ca2+ (22).
In the present study, we show that cADPR binds to FKBP12.6, that FKBP12.6 is present in islet microsomes, and that when cADPR was added to islet microsomes, FKBP12.6 was dissociated from the microsomes to release Ca2+.
Calmodulin was purchased from Calbiochem. Fluo 3 was obtained from Molecular Probes. 17-Allyl-1,14-dihydroxy-12-[2-(4-hydroxy-3-methoxycyclohexyl)-1-methylvinyl]-23,25-dimethoxy-13,19, 21,27-tetramethyl-11,28-dioxa-4-azatricyclo-[22.3.1.04,9]octacos-18-ene-2,3,10,16-tetrone (FK506) was kindly provided by Fujisawa Pharmaceutical Co., Ltd. (Osaka, Japan). A sheep anti-mouse antibody labeled with horseradish peroxidase and ECL reagents were from Amersham Corp., a goat anti-rabbit antibody labeled with horseradish peroxidase was from Zymed (San Francisco, CA), Immobilon-P was from Millipore (Bedford, MA), pMALc2 vector and amylose resin were from New England Biolabs (Beverly, MA), [proryl-3H]dihydro-FK506 and [3H]NAD+ were from DuPont NEN, NAD+ was from Boeringer, ATP and ADP-ribose were from Sigma, nicotinamide was from Wako Pure Chemical Industries, Ltd. (Osaka, Japan), and D-myo-inositol 1,4,5-trisphosphate and ryanodine were from Biomol Research (Natick, MA). A monoclonal antibody against human FKBP12 (clone 3F4-70) was kindly provided by Kazuyuki Otsuka and Dr. Masakazu Kobayashi ( Fujisawa Pharmaceutical Co., Ltd.), and an anti-FKBP12.6 antiserum, which also reacts with rat FKBP12, was kindly provided by Gou Ichien (Esai Co., Ltd., Tokyo).
Calcium Release AssayMicrosomes were prepared as described
previously (4, 6). In brief, 2,000 islets from Wistar male rats
(240-280 g) were homogenized with a Pellet mixer (Treff, Degersheim,
Switzerland) in 0.2 ml of acetate intracellular medium composed of 250 mM potassium acetate, 250 mM
N-methylglucamine, 1 mM MgCl2, and
20 mM Hepes (pH 7.2) supplemented with 0.5 mM
ATP, 4 mM phosphocreatine, creatine phosphokinase (2 units/ml), 2.5 mM benzamidine, and 0.5 mM
phenylmethylsulfonyl fluoride. After the homogenates had been
centrifuged for 45 s at 13,000 × g, the
microsomes were prepared by Percoll density gradient centrifugation at
20,000 × g for 40 min at 10 °C. Release of
Ca2+ was monitored in 0.6 ml of intracellular medium
composed of 250 mM potassium gluconate, 250 mM
N-methylglucamine, 1 mM MgCl2, and
20 mM Hepes (pH 7.2) supplemented with 1 mM
ATP, 4 mM phosphocreatine, creatine phosphokinase (2 units/ml), 2.5 mM benzamidine, 0.5 mM phenylmethylsulfonyl fluoride, 7 µg/ml bovine brain calmodulin (6),
and 3 µM Fluo 3 with the addition of 30 µl of the islet microsome fraction (10 µg of protein) (4, 6). FK506 or cADPR (4, 6)
was added into the incubation, and Fluo 3 fluorescence was measured at
490 nm excitation and 535 nm emission with a JASCO CAF-110
intracellular ion analyzer (Tokyo, Japan) at 37 °C (6). Total
accumulated Ca2+ in islet microsomes was estimated by the
increase of Fluo 3 fluorescence caused by the addition of 200 nM ionomycin (Sigma) to the Ca2+ release medium
containing the microsomes, and the ambient free Ca2+
concentration ([Ca2+]) was calculated using the following
equation, as described previously (6): [Ca2+] = Kd × (F Fmin)/(Fmax
F), where Kd = 400 nM. In
response to 100 nM cADPR, islet microsomes exhibited
18 ± 2 nmol Ca2+ release/mg protein, which
corresponded to 31 ± 4% of the total accumulated
Ca2+.
1 µl of rat islet cDNA library
(2 × 106 plaque-forming unit) (23, 24) was used as a
template for PCR (23-25). The sequences of sense and antisense primers
were 5-GGAATTCCGCGTCCTTTTCCTCCTCCT-3
and
5
-GGAATTCCTTGAGGTTTATGGCATATAGTT-3
for the isolation of rat
FKBP12 cDNA, which corresponded to nucleotide sequences 48-68 and
636-658 of mouse FKBP12 mRNA (26), and 5
-ATGGGCGTGGAGATCGAGA-3
and 5
-CAAGCTTGGAAGGACATTCCCCAAGAA-3
for the isolation of rat FKBP12.6
cDNA, which corresponded to nucleotide sequences 1-19 and 306-326
of human FKBP12.6 mRNA (27, 28). Nucleotide sequences were
determined as described (23, 25).
RT-PCR was performed as described previously (23,
24). The oligonucleotides corresponding to 48-68
(5-CCCGAAGCGCGGCCAGACCTG-3
) and 515-536
(5
-TAGGTCAACACACATACAGAAG-3
) of rat FKBP12 mRNA, 37-57
(5
-GGAAGGACATTCCCTAAGAAG-3
) and 237-257
(5
-GTAGCTCCATATGCCACATCA-3
) of rat FKBP12.6 mRNA, and 135-155
and 951-971 of rat glyceraldehyde-3-phosphate dehydrogenase mRNA
(29) were used as PCR primers.
Rat islet microsomes (100 µg of
protein) were chromatographed on 20% SDS-polyacrylamide gel
electrophoresis (30) and transferred to Immobilon P. The membrane was
incubated with a monoclonal antibody against human FKBP12 or with an
anti-FKBP12.6 antiserum, which also reacts with rat FKBP12 (see Fig.
4A). The monoclonal antibody was diluted at 0.2 µg/ml, and
the antiserum was diluted 2000 times with 5% milk powder. After
rinsing, the membrane was incubated with a sheep anti-mouse antibody
labeled with horseradish peroxidase or with a goat anti-rabbit antibody
labeled with horseradish peroxidase and developed using the ECL
reagents as described (6, 25).
cADPR Binding Assay
The cDNA encoding rat FKBP12.6 was subcloned into a pMALc2 vector and recombinant maltose binding protein (MBP)-FKBP12.6 fusion protein was expressed in the cytoplasm of Escherichia coli. The fusion protein was purified to homogeneity by amylose resin chromatography. MBP-FKBP12.6 fusion protein (100 nM) was incubated with the indicated concentrations of FK506 or cADPR in the presence of [proryl-3H]dihydro-FK506 (0.5 µCi) or [3H]cADPR (0.5 µCi) that was prepared from [3H]NAD+ by Aplysia kurodai ADP-ribosyl cyclase (4, 6) in Dulbecco's phosphate-buffered saline at 25 °C for 30 min. MBP-FKBP12.6 was precipitated with amylose resin, and the radioactivity of the precipitate was measured with a Beckman scintillation spectrometer (LS-6500). Kd values were calculated using DeltaGraph® Pro 3 (DeltaPoint, Inc.). The binding specificity of FKBP12.6 was examined as described above using 100 nM of [proryl-3H]dihydro-FK506 or [3H]cADPR with 100 µM of cADPR, FK506, NAD+, ATP, ADP-ribose, nicotinamide, D-myo-inositol 1,4,5-trisphosphate, and ryanodine as competitors.
We have previously shown that islet microsomes release
Ca2+ in response to cADPR (4). In the present study, we
found that FK506, one of the most widely used immunosuppressive agents,
induced the release of Ca2+ from islet microsomes. The
dose-response curve of FK506 on the Ca2+ release from islet
microsomes was concentration-dependent, with half-maximal
release occurring at 2 µM, and maximal Ca2+
release occurring at 10 µM (Fig.
1A). In addition, as shown in Fig.
1B, after the islet microsomes were treated with FK506, the Ca2+ release by cADPR from the microsomes was reduced
depending on the concentration of FK506; the maximal reduction was seen
at 5-25 µM FK506. Because cADPR and FK506 appear to
induce the release of Ca2+ by a common mediator, we next
tried to determine if this occurs by a targeting of the same ligand.
The cellular target for FK506 is thought to be FKBP12 and FKBP12.6.
Therefore, we isolated FKBP12 and FKBP12.6 cDNAs from a rat islet
cDNA library. As shown in Fig. 2A, rat
FKBP12 is composed of 108 amino acids and highly conserved with human
(31, 32), mouse (26), bovine (33), and rabbit (19) FKBP12. Rat FKBP12.6
is also a 108-amino acid protein and completely conserved with human
(27, 28) and bovine (33) FKBP12.6. RT-PCR analyses revealed that FKBP12
and FKBP12.6 mRNAs were ubiquitously expressed in rat tissues
including pancreatic islets and streptozotocin/nicotinamide-induced
insulinomas (Fig. 2B); FKBP12 mRNA was detected in
RINm5F cells, but FKBP12.6 mRNA was not. RINm5F cells, rat
insulinoma-derived immortal cells, which do not release
Ca2+ in response to cADPR (34), synthesize and secrete very
little insulin and show negligible sensitivity to glucose (35).
Streptozotocin/nicotinamide-induced insulinomas contain as much insulin
mRNA as normal islets (36-38) and retain the sensitivity to
glucose (39). Therefore, it is FKBP12.6 rather than FKBP12 that plays a
role in the CD38-cADPR signaling (5) in insulin secretion by glucose in
islets, because CD38, which catalyzes the synthesis and degradation of
cADPR (23, 40, 41), a second messenger for Ca2+
mobilization in glucose-induced insulin secretion in islets (4, 5), is
expressed in islets and streptozotocin/nicotinamide-induced insulinomas
but not in RINm5F cells (23, 24). We then isolated microsomes from rat
islets and carried out immunoblot analyses. As shown in Fig. 2
(C and D), although islet microsomes did not contain FKBP12 (Fig. 2C, lane 4), the microsomes
contained FKBP12.6 (Fig. 2D, lanes 5 and
6), suggesting that FKBP12.6 is the target for FK506 and/or
cADPR to release Ca2+ from islet microsomes.
Next, we examined the binding of FKBP12.6 to cADPR. The recombinant rat
FKBP12.6 bound to FK506 at a Kd value of 32 nM. As shown in Fig. 3A, cADPR
was found to bind to FKBP12.6 at a Kd value of 35 nM. The cADPR binding was inhibited by FK506 and neither
structurally nor functionally related analogues of cADPR inhibited the
cADPR binding to FKBP12.6 (Fig. 3B). These results indicate
that FKBP12.6 acts as a cADPR-binding protein and strongly suggest that
cADPR is the actual ligand for FKBP12.6 because FK506 does not normally
exist in mammalian cells.
It was reported that FKBP12/12.6 bound to RyR tightly and that by the addition of FK506, FKBP12/12.6 were dissociated from RyR to form FK506-FKBP12/12.6 complexes (19, 22, 33). In addition, the open probability of the type 1 RyR Ca2+ channel was greatly increased when FKBP12 was released from the RyR by the addition of FK506 (22). As shown in Fig. 4, after treatment of islet microsomes with cADPR, FKBP12.6 was not detected in the microsomes but was recovered in the supernatant (Fig. 4A), and Ca2+ release from the microsome treated by cADPR or FK506 was reduced. FK506 as well as cADPR then had almost no effect in releasing Ca2+ from the 1 µM cADPR-pretreated microsomes (Fig. 4B). Our recent experiment indicated that type 2 RyR is expressed in rat islets.2 From these results, it is strongly suggested that when cADPR binds to FKBP12.6 in islet microsome RyR and causes the dissociation of FKBP12.6 from the RyR to form FKBP12.6-cADPR complex, the channel activity of the RyR is thereby increased to release Ca2+ from the endoplasmic reticulum. As described previously (6), the RyR can also be activated by Ca2+/calmodulin-dependent protein kinase II. The interaction between the dissociation of FKBP12.6 from RyR and the phosphorylation of RyR by Ca2+/calmodulin-dependent protein kinase II remains to be elucidated.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) D86641[GenBank] and D86642[GenBank].
We are grateful to Kazuyuki Otsuka and Masakazu Kobayashi (Fujisawa Pharmaceutical Co., Ltd.) for providing the antibody against FKBP12, Gou Ichien (Esai Co., Ltd., Tokyo) for providing the antiserum against FKBP12.6, Hideo Kumagai for technical assistance, and Brent Bell for valuable assistance in preparing the manuscript for publication.