The intracellular domain of amyloid precursor protein interacts with FKBP12

BBRC Biochemical and Biophysical Research Communications 350 (2006) 472–477 www.elsevier.com/locate/ybbrc

The intracellular domain of amyloid precursor protein interacts with FKBP12 Fan-Lun Liu 1, Pei-Hsueh Liu 1, Hsien-Wei Shao 1, Fan-Lu Kung

*

School of Pharmacy, National Taiwan University, Taipei, 10051, Taiwan, ROC Received 8 September 2006 Available online 25 September 2006

Abstract To elucidate the roles of the APP intracellular domain (AICD) in the development of Alzheimer’s disease, a yeast two-hybrid system was used to screen for AICD-interacting proteins. Our result revealed that FKBP12, an immunophilin with a peptidyl–prolyl cis–trans isomerase (PPIase) activity, may interact with AICD. This interaction was confirmed by coimmunoprecipitation studies. FKBP12 has been shown to be expressed at a higher level in areas of pathology of patients with neurodegenerative diseases. In addition, Pin1, a member of another PPIase family, has been suggested to be involved in the amyloidogenic APP processing and Ab production. The interaction between FKBP12 and AICD might hint at a possible role FKBP12 plays, probably in a fashion similar to Pin1, in the amyloidogenesis of APP. We also found that the interaction was interfered, in a dose-dependent manner, by FK506, whose neuroprotective effect has been suggested to be correlated with its PPIase inhibitory activity.  2006 Elsevier Inc. All rights reserved. Keywords: FKBPs; Alzheimer’s disease; AICD (APP intracellular domain); FK506

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder often associated with elderliness. One of its pathohistological characteristics is the formation of insoluble extracellular amyloid plaques [1]. The major constituent of amyloid plaque, the Ab peptide, is derived from the amyloid precursor protein (APP) by two sequential proteolytic cleavages. In addition to Ab, the aforementioned proteolytic processing also generates the APP intracellular domain (AICD), the extreme C-terminus of APP composed of 57–59 amino acid residues. In an attempt to elucidate the roles of AICD in the pathogenic pathway of Alzheimer’s disease, a yeast two-hybrid system was used to identify proteins that may interact with AICD. Our preliminary result revealed that FKBP12, a member of the FK506binding protein (FKBP) family, was one of the potential interaction partners of AICD. *

1

Corresponding author. Fax: +886 2 2391 9098. E-mail address: fl[email protected] (F.-L. Kung). These authors contributed equally to this work.

0006-291X/$ - see front matter  2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.09.073

FKBP12 is ubiquitous in nerve systems, and several studies have demonstrated that FKBP12 can form a complex with FK506 [2–4]. Like other immunophilins such as cyclophilins and other FKBPs, FKBP12 has a peptidyl– prolyl cis–trans isomerase (PPIase) activity, which is important in protein assembly, folding, and transportation, and has been suggested to be participating in many physiological pathways [5]. This PPIase activity is known to be inhibited by FK506, and the inhibitory effect of FK506 to FKBP12 PPIase has been shown to be correlated with the neurotrophic activity of FK506 [6–8], which may be separated from its immunosuppressive effect mediated via calcineurin inhibition [9–10]. The exact mechanism(s) of the neuroprotective activity of FK506, however, still remains unclear. In this study, the interaction of AICD with FKBP12 was confirmed by coimmunoprecipitation experiments. We also demonstrated that FK506 was able to block this interaction in a dose-dependent manner. The expression of FKBP12 has been shown to be higher in areas of pathology

F.-L. Liu et al. / Biochemical and Biophysical Research Communications 350 (2006) 472–477

in the brains of patients with neurodegenerative diseases [11], suggesting its involvement in the pathogenesis of those diseases. In addition, it has recently been proposed that Pin1, which belongs to another family of PPIase (the Parvulin PPIase family), is involved in the regulation of APP processing via its isomerization activity on T668–P669 of APP [12]. Our results support the hypothesis that FKBP12 may participate in the regulation of APP processing in a fashion similar to Pin1, and FK506 may partially induce a neurotrophic activity either by hindering the interaction between FKBP12 and its substrate, AICD in this case, or by inhibiting the PPIase activity of FKBP12 directly. Materials and methods Yeast two-hybrid screening. The yeast two-hybrid screening was conducted using Matchmaker GAL4 Two-Hybrid System 3 (Clontech, Palo Alto, CA, USA) following manufacturer’s instructions. pGBKT7-AICD was used as a bait in the two-hybrid screening as previously described [13]. The identities of the positive library plasmids were verified by DNA sequencing (Core Facility, College of Medicine, National Taiwan University), and the sequencing results were compared with the GenBank non-redundant database using the BLAST programs from NCBI. Yeast protein extraction and Western blotting. Yeast strain PJ69-4 [14] harboring both pGBKT7-AICD and the AD/library plasmid of interest was grown in 50 mL of SD/-Trp/-Leu medium overnight at 30 C. Cells were harvested by centrifugation (3000g for 5 min at 4 C). Protein extraction was conducted using the Y-PER yeast protein extraction reagent (Pierce, Rockford, IL, USA) according to manufacturer’s suggestion. For Western blotting analysis, proteins on the polyacrylamide gel were transferred to a PVDF membrane (Pall, East Hills, NY, USA), and probed with either the anti-APP C-terminal polyclonal antibody (171610, Calbiochem, San Diego, CA, USA) or the anti-HA monoclonal antibody (sc-7392, Santa Cruz Biotechnology, Santa Cruz, CA, USA). Anti-HA antibody was used to detect GAL4-AD fusion proteins since the pACT2 vector was specially designed to express HA epitope-tagged fusion protein. The anti-FKBP12 polyclonal antibody (sc-6174, Santa Cruz Biotechnology) used later for the coimmunoprecipitation assay was not suitable for the detection of the GAL4-AD-FKBP1233–108 encoded by pACT2FKBP1233–108 since the epitope recognized by this antibody is lost in this fusion protein. The membrane was then incubated with anti-rabbit or antimouse IgG HRP-conjugated antibody (Amersham Biosciences). Proteins of interest (i.e., GAL4-DBD-AICD and GAL4-AD-FKBP1233–108) were visualized with ECL Western Blotting Detection Reagent (Amersham Bioscience) by enhanced chemiluminescence method. Construction of pACT2-FKBP1242–108 and yeast two-hybrid assay. The further-truncated FKBP12 construct (pACT2-FKBP1242–108) was generated by XmaI digestion/mung bean nuclease treatment/re-ligation of pACT2-FKBP1233–108, which was identified by yeast two-hybrid screening as described in the previous section. The sequence of the candidate plasmid was verified by DNA sequencing (Core Facility, College of Medicine, National Taiwan University). pACT2-FKBP1242–108 obtained was then cotransformed with pGBKT7-AICD into yeast strain PJ69-4. The cell growth of the cotransformants was checked on dropout medium (SD/Trp/-Leu/-Ade and SD/-Trp/-Leu/-Ade/-His/3-AT) as previously described. Again, PJ69-4 cotransformed with pGBKT7/pCL1 and pGBKT7-53/ pGADT7-T were used as positive controls, whereas the ones cotransformed with pGADT7/pGBKT7 and pGBKT7-Lam/pGADT7-T were used as negative controls. Immunoprecipitation and Western blotting. SH-SY5Y cells were grown in a 1:1 mixture of Ham’s F12 nutrients and minimal essential medium (MEM, GibcoBRL/Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS), 1 mM sodium pyruvate, 100 IU/mL penicillin, and 100 lg/mL streptomycin at 37 C, 95% air, and 5% CO2, and replaced with fresh media every two to three days until cells became

473

confluent. Cells were washed with phosphate-buffered saline (PBS, 2.7 mM KCl, 137 mM NaCl, 10 mM Na2HPO4, and 1.4 mM KH2PO4, pH 7.4), trypsinized with 0.05% trypsin/0.53 mM EDTA (GibcoBRL), and lysed by brief sonication in 100 lL lysis buffer (50 mM Tris–HCl, pH 7.4, 100 mM NaCl, 1% Triton X-100, 0.5% NP-40, 1 mM DTT, 1 mM PMSF, and 1· protease inhibitor mixture). The cell lysate was obtained by centrifugation at 15,000g for 30 min at 4 C. Approximately 0.5 lL of rabbit anti-APP C-terminal antibody (171610, Calbiochem, San Diego, CA, USA) or anti-FKBP12 polyclonal antibody (sc-6174, Santa Cruz Biotechnology) was incubated with 50 lL of 50% protein A–Sepharose CL-4B bead slurry (Amersham Biosciences) for 1 h at room temperature. The cross-linking reaction was initiated by the addition of 6.5 lL of 5 mM disuccinimidyl suberate (DSS) and the reaction mixture was incubated for 1 h at room temperature with gentle agitation. Fifty microliters of TBS (25 mM Tris–HCl, pH 7.2, 150 mM NaCl) was added to stop the reaction and wash off excess DSS. Unbound antibodies were removed with 50 lL of 0.1 M glycine (pH 2.8). An appropriate amount of the SH-SY5Y protein extract prepared as described was incubated with above antibody- conjugated protein A–Sepharose CL-4B beads for 1 h at 4 C with gentle agitation. The beads were washed with PBS. Proteins bound to the beads were eluted by boiling in 1· Tris– Tricine sample buffer (0.05 M Tris–HCl, pH 6.8, 12% glycerol, 4% SDS, 0.1 M DTT, and 0.01% coomassie R250) for 5 min and were separated by 16.5% Tris–Tricine SDS–PAGE. For Western blotting analysis, proteins transferred to a PVDF membrane were probed with either the anti-APP C-terminal polyclonal antibody (171610, Calbiochem) or the antiFKBP12 polyclonal antibody (sc-6174, Santa Cruz Biotechnology) and were visualized with ECL Western blotting detection reagent (Amersham Bioscience) by enhanced chemiluminescence method. b-Galactosidase liquid assay. Yeast PJ69-4 strains of interest were grown in 5 mL of SD/-Trp/-Leu medium overnight at 30 C. The overnight cultures were added into 8 mL of YPD medium and incubated for an additional 8 h until the OD600 of the cells reached approximately 0.8. Varying concentrations of FK506 (courtesy of Fujisawa Pharmaceutical Co., Ltd., Osaka, Japan) were added at the time of dilution if necessary. Cells were harvested by centrifugation (10,000g for 30 s), lysed by freezethawing with liquid nitrogen, and then mixed with appropriate amount of Z buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, and 1 mM MgSO4, pH 7.0) supplemented with 0.27% b-mercaptoethanol. Reaction was initiated by adding 2.2 mM o-nitrophenyl-b-D-galactoside (ONPG) into the mixture and quenched with sodium carbonate after incubation at 30 C for 30 min or until the yellow color developed. The OD420 is then read to determine the b-galactosidase unit using the following equation: bgalactosidase units = (OD420 · 1000)/(OD600 · t · v), where v is the volume of the sample (in mL) used in the assay and t is the time of incubation (in min) of the reaction at 30 C [15].

Results and discussion Identification of FKBP12 as a potential AICD-interacting protein To elucidate the roles the intracellular domain of amyloid precursor protein (AICD) might play in the pathogenic pathway of Alzheimer’s disease, a yeast two-hybrid system was used to screen for proteins that may directly interact with AICD [13]. As has been previously reported, 53 potential positive colonies were obtained from a human brain cDNA library (in pACT2) of approximately 3.5 · 106 independent clones [13]. AD/library plasmids isolated from these candidate colonies were characterized by restriction enzyme digestion and sequence analysis, and were reintroduced into pGBKT7-AICD- or pGBKT7-containing PJ69-4 cells to confirm the interactions between bait and

474

F.-L. Liu et al. / Biochemical and Biophysical Research Communications 350 (2006) 472–477

5

6 1 2 4 3

SD/-Trp/-Leu GAL4-DBD fusions

GAL4-AD fusions

1

pGBKT7

pGADT7

2

pGBKT7-53

pGADT7-T

3

pGBKT7-Lam

pGADT7-T

4

pGBKT7-AICD

pGADT7

5

pGBKT7-AICD

pACT2-FKBP33-108

6

pGBKT7

pACT2-FKBP33-108

SD/-Trp/-Leu/-Ade/-His/3-AT

AICD interacts with FKBP12 The interaction was further verified by coimmunoprecipitation assays of lysate from SH-SY5Y cells expressing APP and FKBP12 endogenously. The cell lysate was

Mr(kD) 34 26

cotransformed PJ69-4

prey proteins (Fig. 1). Partial sequences of 24 such plasmids were found to be identical to portions of the open reading frames of 9 different known proteins (data not shown). Among those, one plasmid contained a fragment encoding the C-terminal 76 amino acids of human FKBP12 (FK506-binding protein 1A, GenBank Accession Nos. NP_000792, NP_463460), which was in the same reading frame as that of GAL4-AD and should therefore be able to express an in-frame GAL4-AD-FKBP1233–108 fusion protein. Western blotting analysis was performed to verify that both GAL4-DBD-AICD and GAL4-AD-FKBP1233–108 fusion proteins were expressed in the PJ69-4 cells harboring both pGBKT7-AICD and pACT2-FKBP1233–108 (Fig. 2). Fig. 2 shows that the bait and prey fusion proteins were expressed at the expected sizes (27.6 and 26.9 kD, respectively) in the cotransformants but not in the control cells (i.e., untransformed PJ69-4). Yeast two-hybrid results from Fig. 1 suggest that FKBP12 is a potential AICD-interacting protein. Since neither PJ69-4 harboring both pGBKT7-AICD and pGADT7 nor that cotransformed with pGBKT7 and pACT2-FKBP1233–108 was able to grow on selective medium, the interaction occurred only in the presence of both AICD and FKBP1233–108 in the form of fusion proteins.

control

Fig. 1. Yeast two-hybrid analysis of the interaction of AICD with FKBP12. Yeast PJ69-4 cells were cotransformed with indicated plasmids (left). Proteinprotein interactions were checked by the growing conditions of the cotansformants on selective media (SD/-Trp/-Leu, middle; SD/-Trp/-Leu/-Ade/-His/3AT, right). pGBKT7-53 and pGADT7-T encode fusions between the GAL4-DBD and murine p53 and GAL4-AD and SV40 large T-antigen, respectively. pGBKT7-Lam encodes a fusion of GAL4-DBD and human lamin C. PJ69-4 cotransformed with pGBKT7-53/pGADT7-T was used as a positive control since p53 is known to interact with SV40 large T-antigen, whereas the ones cotransformed with pGADT7/pGBKT7 (both are empty vectors) and pGBKT7-Lam/pGADT7-T were used as negative controls.

GAL4-AD-FKBP1233-108 WB: anti-HA

34

GAL4-DBD-AICD

26 WB: anti-APP-C-terminal Fig. 2. Expression of the GAL4-DBD-AICD and GAL4-ADFKBP1233–108 fusion proteins in yeast strain PJ69-4 cotransformed with both pGBKT7-AICD and pACT2-FKBP1233–108. Protein extracts from the cotransformants were prepared as previously described and separated by 10% SDS–PAGE. The predicted sizes of GAL4-DBD-AICD and GAL4-AD-FKBP1233–108 are 27.6 and 26.9 kD, respectively. For Western blotting analysis, proteins on the polyacrylamide gel were transferred to a PVDF membrane, and probed with either the anti-APP C-terminal polyclonal antibody (171610, Calbiochem) (lower panel) or the anti-HA monoclonal antibody (sc-7392, Santa Cruz Biotechnology) (upper panel). Cell extract from untransformed PJ69-4 was used as a control.

IP

Mr(kD)

input

F.-L. Liu et al. / Biochemical and Biophysical Research Communications 350 (2006) 472–477

APP CTF

15 10

IP: anti-FKBP12 WB: anti-APP-C-terminal 15 10 Lane

1

2

FKBP12 IP: anti-APP-C-terminal WB: anti-FKBP12

Fig. 3. Coimmunoprecipitation of APP C-terminal fragment(s) (APP CTF) and FKBP12. Cell lysate from SH-SY5Y cells was immunoprecipitated with either anti-APP-C-terminal antibody or anti-FKBP12 antibody (Lane 2, IP). Proteins bound to the antibody-conjugated protein A–Sepharose CL-4B beads (Amersham Biosciences) were separated by 16.5% Tris–Tricine SDS–PAGE and analyzed by Western blotting using either the anti-FKBP12 goat polyclonal antibody (sc-6174, Santa Cruz) (lower panel) or the anti-APP-C-terminal rabbit polyclonal antibody (171610, Calbiochem) (upper panel). Aliquots of cell lysate were also loaded directly onto the gels and analyzed by Western blotting (Lane 1, input).

immunoprecipitated with either the anti-APP-C-terminal antibody or the anti-FKBP12 antibody and analyzed by Western blotting. Results shown in Fig. 3 reveal that FKBP12 could be co-precipitated with APP C-terminal fragment(s) and vice versa. AICD itself was still undetectable, just as previously observed [13,16]. Although it is not yet clear whether other parts of APP interact with FKBP12 or not, the experiments described above suggest the association of APP C-terminal fragment with FKBP12 in intact cells, and it is very likely that this interaction is direct. Truncated FKBP12 fusion protein lacking residues 1-41 does not interact with GAL4-DBD-AICD As their names imply, FKBPs have long been known to be the cellular receptors of the immunosuppressants FK506 and rapamycin [2–3,17]. Several studies have demonstrated that FK506 can form a complex with FKBP12 [2–4], which then may further interact with other proteins. Amino acid residues of FKBP12 important for the FKBP12-FK506 interaction include the five residues which form hydrogen bonds with FK506 (Tyr83, Asp38, Glu55, Ile57, and Gly87), the indole of Trp60 and the phenyl ring of Phe47 positioned in the hydrophobic binding pocket, and Tyr27, Phe37, and Phe100 in the carbonyl binding pocket [18–19]. Other residues, such as Gln4, Glu32, Asp33 [20], as well as a stretch of residues from Asp33 to Lys48 (including Asp42, Arg43, Pro46, Phe37, Gly90, Ile/Leu91, Pro93, Glu55, and the loop around Gly20, Gln21, and Glu108) [19,21], have been suggested to be participating in the interaction with other binding partners of either

475

FKBP12 or the FKBP12-FK506 complex. To see whether any of those aforementioned amino acid residues of FKBP12, particularly residues 33–48 since many of those have been suggested to be involved in the interaction with other molecules, is important for its interaction with AICD, a further truncated FKBP12 construct, pACT2FKBP1242–108, was generated from pACT2-FKBP1233–108 and cotransformed with pGBKT7-AICD into PJ69-4 cells for yeast two-hybrid assay. Growth selection results indicate that GAL4-AD-FKBP1242–108 was unable to interact with GAL4-DBD-AICD, suggesting that the fragment containing residues 33–41 of FKBP12 was required for the interaction (data not shown). Since this fragment contains two residues, Phe37 and Asp38, that are important for FK506 binding as just mentioned, we were intrigued to further investigate the effect of the addition of FK506 on the interaction between AICD and FKBP12. FK506 interferes with the interaction between GAL4-DBDAICD and GAL4-AD-FKBP1233–108 in a dose-dependent manner To see whether the interaction between AICD and FKBP12 (or more precisely, between GAL4-DBD-AICD and various truncated forms of GAL4-AD-FKBP12) is interfered by FK506, b-galactosidase levels expressed from a GAL7-LacZ reporter carried in the yeast strain PJ69-4 cotransformed with pGBKT7-AICD and pACT2FKBP1233–108 (or pACT2-FKBP1242–108) in the presence or absence of FK506 were analyzed. Results from this assay reveal a significant interaction between GAL4-DBD-AICD and GAL4-AD-FKBP1233–108, though not as strong as the positive control (p53 and SV40 large T-antigen) (Fig. 4A). The interaction between GAL4-DBD-AICD and GAL4AD-FKBP1242–108, on the other hand, was almost undetectable (Fig. 4A). These results are consistent with what was observed earlier in the yeast two-hybrid assay (discussed in the previous section). It was also clearly shown in Fig. 4 that b-galactosidase activity in cells expressing both GAL4DBD-AICD and GAL4-AD-FKBP1233–108 was reduced when treated with FK506. This inhibition is dose-dependent (Fig. 4B) and specific to this interacting pair since no inhibition was observed for the known interacting pair p53 and SV40 large T-antigen (Fig. 4A, pGBKT7-53 + pGADT7-T). As previously mentioned, FK506 has been shown to have various pharmacological properties, including immunosuppressive, neuroprotective, and neuroregenerative effects [6–10]. Two possible general mechanisms can be proposed if FKBP(s) is involved in the processes. On the one hand, it has been suggested that the formation of the FKBP12-FK506 complex and the interaction between the complex and its interacting partner(s) may contribute to certain pharmacological effect(s). One extensively studied example is the calcineurin-mediated immunosuppressive effect. Previous studies have demonstrated that upon forming a complex with FKBP12, FK506 is able to block the phosphatase activity of calcineurin [9,22–23]. It has

476

F.-L. Liu et al. / Biochemical and Biophysical Research Communications 350 (2006) 472–477

Fig. 4. Analysis of interaction between GAL4-DBD-AICD and various forms of GAL4-AD-FKBP12. b-Galactosidase assay was conducted as described in detail under Materials and methods. Data shown are means ± SD (shown as error bars) for at least three independent experiments performed in duplicate or triplicate. (A) b-Galactosidase activity produced by each plasmid combination in the absence or presence of FK506 (25 lg/mL). pCL1 encodes the wild-type GAL4. PJ69-4 cotransformed with pGBKT7/pCL1 or pGBKT7-53/pGADT7-T were used as positive controls, whereas the one cotransformed with pGADT7/pGBKT7 (both are empty vectors) was used as a negative control. The level of b-galactosidase activity of the lysate from pCL1-containing strain is at least 5-fold higher than the other positive control strain expressing GAL4-DBD-p53 and GAL4-AD-T-antigen, and is more than one hundred-fold higher than the background level of the negative control (shown in the inset). (B) Dose-dependent inhibition of reporter gene expression by FK506 treatment, as determined by b-galactosidase activity. Here the cells cotransformed with pGBKT7-AICD/pACT2-FKBP1233–108 or pGBKT7-AICD/pACT2-FKBP1242–108 were treated with increasing amounts of FK506 as indicated for 8 h. Results are shown as percent inhibition.

been proposed that the binding of FKBP12-FK506 complex, particularly at a place apart from the active site of calcineurin, may generate subtle alternation in the geometry of calcineurin active site and therefore a non-competitive inhibitory effect [21]. On the other hand, it is also possible that the pharmacological effects are the results of the FK506-mediated dissociation of FKBP12 from its interacting partners. This could be due to the competition between FK506 and the FKBP12-interacting partner for the same binding site on FKBP12, the steric hindrance caused by FK506 binding, or the conformational change of FKBP12 upon FK506

binding. Therefore, it is not necessary for FK506 to have a binding site on FKBP12 overlapped with the interacting interface between FKBP12 and other proteins to trigger a dissociation reaction, as exemplified in the cases of type I TGFb receptor (TbR-I)-FKBP12 [24] and RyR1–FKBP12 complexes [20]. While the interface between the two proteins in the complex is overlapped with the FK506 binding site in the former, the FK506 binding site is exposed when FKBP12 forms a complex with RyR1 in the latter case [24,20]. The neuroprotective activity of FK506, which has been suggested to be correlated with the inhibitory effect of

F.-L. Liu et al. / Biochemical and Biophysical Research Communications 350 (2006) 472–477

FK506 to FKBP12 PPIase [6–8], could be explained by either hypothesis, i.e., the binding of FK506 to FKBP12, which would probably affect the PPIase activity no matter what, could either facilitate the interaction between the complex and a not yet identified protein partner or block the interaction between FKBP12 and its binding protein. Pastorino et al. recently showed that Pin1, a peptidyl–prolyl isomerase belongs to the Parvulin PPIase family, preferentially bound to T668-phosphorylated APP and proposed that Pin1 was involved in the regulation of APP processing via its isomerization activity on T668–P669 of APP [12]. In their model, Pin1 promotes the cis to trans isomerization, which Pastorino et al. suggested to favor non-amyloidogenic APP processing [12]. The interaction between FKBP12 and AICD might hint at a possible role FKBP12 plays, probably in a fashion similar to Pin1, in the amyloidogenesis of APP. What FK506 does under this scenario is then more likely to promote the dissociation of the FKBP12-AICD (or FKBP12-APP) complex as suggested by the second mechanism. Functional studies are in progress to further clarify whether FKBP12 is indeed involved in the regulation of the localization and processing of APP. Acknowledgments We thank Dr. Jih-Hwa Guh for constructive comments and suggestions. We also wish to thank Ms. Ling Chiu for her excellent technical assistance. This work was supported in part by funding from the National Science Council of Taiwan (Grant Nos. NSC-90-2320-B-002-148 and NSC93-2320-B-002-124). References [1] D.J. Selkoe, Biochemistry of altered brain proteins in Alzheimer’s disease, Annu. Rev. Neurosci. 12 (1989) 463–490. [2] J.J. Siekierka, S.H.Y. Hung, M. Poe, C.S. Lin, N.H. Sigal, A cytosolic-binding protein for the immunosuppressant FK506 has isomerase activity but is distinct from cyclophilin, Nature 341 (1989) 755–757. [3] M.W. Harding, A. Galat, D.E. Uehling, S.L. Schreiber, A receptor for the immunosuppressant FK506 is a cis–trans peptidyl–prolyl isomerase, Nature 341 (1989) 758–760. [4] J.J. Siekierka, G. Wiederrecht, H. Greulich, D. Boulton, S.H.Y. Hung, J. Cryan, P.J. Hodges, N.H. Sigal, The cytosolic-binding protein for the immunosuppressant FK506 is both a ubiquitous and highly conserved peptidyl-prolyl cis–trans isomerase, J. Biol. Chem. 265 (1990) 21011–21015. [5] F. Edlich, G. Fischer, Pharmacological targeting of catalyzed protein folding: the example of peptide bond cis–trans isomerases, Handb. Exp. Pharmacol. 172 (2006) 359–404. [6] J.P. Steiner, M.A. Connolly, H.L. Valentine, G.S. Hamilton, T.M. Dawson, L. Hester, S.H. Snyder, Neurotrophic actions of nonimmunosuppressive analogues of immunosuppressive drugs FK506, rapamycin and cyclosporin A, Nat. Med. 3 (1997) 421–428.

477

[7] S.H. Snyder, M.M. Lai, P.E. Burnett, Immunophilins in the nervous system, Neuron 21 (1998) 283–294. [8] S. Brecht, K. Schwarze, V. Waetzig, C. Christner, S. Heiland, G. Fischer, K. Sartor, T. Herdegen, Changes in peptidyl–prolyl cis–trans isomerase activity and FK506 binding protein expression following neuroprotection by FK506 in the ischemic rat brain, Neuroscience 120 (2003) 1037–1048. [9] J. Liu, J.D. Farmer Jr., W.S. Lane, J. Friedman, I. Weissman, S.L. Schreiber, Calcineurin is a common target of cyclophilin–cyclosporin A and FKBP-FK506 complexes, Cell 66 (1991) 807–815. [10] J.P. Griffith, J.L. Kim, E.E. Kim, M.D. Sintchak, J.A. Thomson, M.J. Fitzgibbon, M.A. Fleming, P.R. Caron, K. Hsiao, M.A. Navia, X-ray structure of calcineurin inhibited by the immunophilin-immunosuppressant FKBP12-FK506 complex, Cell 82 (1995) 507–522. [11] M. Avramut, C.L. Achim, Immunophilins and their ligands: insights into survival and growth of human neurons, Physiol. Behav. 77 (2002) 463–468. [12] L. Pastorino, A. Sun, P.J. Lu, X.Z. Zhou, M. Balastik, G. Finn, G. Wulf, J. Lim, S.H. Li, X. Li, W. Xia, L.K. Nicholson, K.P. Lu, The prolyl isomerase Pin1 regulates amyloid precursor protein processing and amyloid-beta production, Nature 440 (2006) 528–534. [13] T.Y. Chen, P.H. Liu, C.T. Ruan, L. Chiu, F.L. Kung, The intracellular domain of amyloid precursor protein interacts with flotillin-1, a lipid raft protein, Biochem. Bioph. Res. Co. 342 (2006) 266–272. [14] P. James, J. Halladay, E.A. Craig, Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast, Genetics 144 (1996) 1425–1436. [15] J.H. Miller, Experiments in molecular genetics, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1972. [16] P. Cupers, I. Orlans, K. Craessaerts, W. Annaert, B. De Strooper, The amyloid precursor protein (APP)-cytoplasmic fragment generated by gamma-secretase is rapidly degraded but distributes partially in a nuclear fraction of neurones in culture, J. Neurochem. 78 (2001) 1168–1178. [17] B.E. Bierer, P.S. Mattila, R.F. Standaert, L.A. Herzenberg, S.J. Burakoff, G. Crabtree, S.L. Schreiber, Two distinct signal transmission pathways in T lymphocytes are inhibited by complexes formed between an immunophilin and either FK506 or rapamycin, Proc. Natl. Acad. Sci. USA 87 (1990) 9231–9235. [18] G.D. Van Duyne, R.F. Standaert, P.A. Karplus, S.L. Schreiber, J. Clardy, Atomic structure of FKBP-FK506, an immunophilin–immunosuppressant complex, Science 252 (1991) 839–842. [19] J.W. Becker, J. Rotonda, B.M. McKeever, H.K. Chan, A.I. Marcy, G. Wiederrecht, J.D. Hermes, J.P. Springer, FK-506-binding protein: three-dimensional structure of the complex with the antagonist L685,818, J. Biol. Chem. 268 (1993) 11335–11339. [20] M. Samso, X. Shen, P.D. Allen, Structural characterization of the RyR1–FKBP12 interaction, J. Mol. Biol. 356 (2006) 917–927. [21] C.R. Kissinger, H.E. Parge, D.R. Knighton, C.T. Lewis, L.A. Pelletier, A. Tempczyk, V.J. Kalish, K.D. Tucker, R.E. Showalter, E.W. Moomaw, L.N. Gastinel, N. Habuka, X. Chen, F. Maldonado, J.E. Barker, R. Bacquet, J.E. Villafranca, Crystal structures of human calcineurin and the human FKBP12-FK506-calcineurin complex, Nature 378 (1995) 641–644. [22] C.B. Klee, G.F. Draetta, M.J. Hubbard, Calcineurin, Adv. Enzymol. Relat. Areas Mol. Biol. 61 (1988) 149–200. [23] D.A. Fruman, C.B. Klee, B.E. Bierer, S.J. Burakoff, Calcineurin phosphatase activity in T lymphocytes is inhibited by FK506 and cyclosporin A, Proc. Natl. Acad. Sci. USA 89 (1992) 3686–3690. [24] M. Huse, Y.G. Chen, J. Massague, J. Kuriyan, Crystal structure of the cytoplasmic domain of the type I TGF beta receptor in complex with FKBP12, Cell 96 (1999) 425–436.