Identification of flotillin-1 as an interacting protein for antisecretory factor

Available online at www.sciencedirect.com

Regulatory Peptides 146 (2008) 303 – 309 www.elsevier.com/locate/regpep

Identification of flotillin-1 as an interacting protein for antisecretory factor Ewa Johansson a , Ingela Jonson a , Mattias Bosaeus a , Eva Jennische b,⁎ a

Institute of Biomedicine, Department of Infectious Diseases, Section of Clinical Bacteriology, Göteborg University, Göteborg, Sweden b Department of Medical Chemistry and Cell Biology, Göteborg University, Göteborg, Sweden Received 5 April 2007; received in revised form 16 November 2007; accepted 19 November 2007 Available online 4 December 2007

Abstract Antisecretory factor (AF) also named S5a/Rpn10 was originally identified through its capacity to inhibit intestinal hypersecretion and was later shown to be a component in the proteasome complex. AF is also a potent anti-inflammatory agent and can act as a neuromodulator. In this study we used yeast two-hybrid screens, with yeast strain PJ692A transformed with the bait vector pGBKT7 (AF aa 1–105) against yeast strain Y187 pretransformed with human brain or placenta cDNA libraries, to identify AF-binding proteins. Flotillin-1 was identified as a specific interacting factor with AF. Immunohistochemistry showed co-localization of AF and flotillin-1 in nervous tissue. Flotillin-1 is an integral membrane protein and a component of lipid rafts, a membrane specialization involved in transport processes. Intracellular AF may affect secretory processes by regulating the localization of signal proteins to lipid rafts. © 2007 Elsevier B.V. All rights reserved. Keywords: S5a/Rpn10; Two-yeast hybrid; Lipid raft; Dot blot; Immunofluorescence

1. Introduction Antisecretory Factor (AF), a 41 kDa protein, was originally identified and cloned through its capacity to inhibit intestinal hypersecretion [1,2]. AF was subsequently discovered to be a subunit in the proteasome complex and it is therefore also known as S5a/Rpn10 [3,4]. AF is expressed in most tissues and can appear associated to the 26 S proteasome or free in the cytoplasm as well as in the nucleus [5–7]. AF is a potent anti-inflammatory agent and can act as a neuromodulator [8–10]. The antisecretory and anti-inflammatory active site of mammalian AF, an 8 amino acid (aa) sequence (35-IVCHSKTR-42), is localized in the Nterminal part of the protein [11], and two ubiquitin binding sites have been identified in the C-terminal part of the protein, aa 211– 230 and aa 282– 301 respectively [4]. AF is phylogenetically well preserved and it appears to be a unique protein since no family of AF-like proteins has been identified [14]. ⁎ Corresponding author. Institute of Biomedicine, Göteborg University, Department of Medical Chemistry and Cell Biology, P.O.B. 420, SE-405 30 Göteborg, Sweden. Tel.: +46 31 773 33 74; fax: +46 31 41 61 08. E-mail address: [email protected] (E. Jennische). 0167-0115/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2007.11.005

The antisecretory and anti-inflammatory activity of endogenous AF can be enhanced by intestinal exposure to bacterial toxins or defined dietary compounds [8,12]. An increase in AF activity is probably a part of the natural defence against diarrhoeal disease and contributes to termination of the illness. The mechanisms for regulation of AF activity are unknown. Previous studies using epitope-specific antibodies indicate that AF can appear in several conformational variants in nervous tissue [13]. These variants could reflect different functional states of the protein in which the active N-terminal part is either exposed or hidden depending on the conditions. In order to understand the mechanisms of actions for AF it is important to identify interactions with other intracellular proteins. When the AF sequence was analyzed using bioinformatic models, residues 5–188, which contains the antisecretory sequence, were predicted to contain a von Willenbrand-like motif (vWm) [14]. The presence of a vWm indicates that AF binds to other proteins via its N-terminal part, since vWm proteins are known proteinbinders [15]. A down regulator of the neddylation system, NUB1, has been reported to interact with AF/S5a [16]. However, NUB1 predominantly localizes in the nucleus and is therefore probably not mediating the antisecretory activity.

304

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309

The aim of the present study was to identify AF interaction molecules. With the aid of the yeast two-hybrid method, using the AF amino acid sequence 1–105 as bait, we identified a number of potential binding proteins, and of these flotillin-1 was considered most interesting in relation to our previous knowledge of AF effects and was chosen for further studies. 2. Materials and methods 2.1. Computer generated model of AF Templates for threading were obtained from the GenThreader database (http://bioinf.cs.ucl.ac.uk). Two templates were used, the I domain from integrin alpha2-beta1 (PDB = 1AOX) and the human von Willebrand factor A3 domain (PDB = 1ATZ). Both contain a von Willebrand A motif. AF sequence was aligned to the template in SwissPDBViewer 3.7 using the results from the database search. The modelling project was sent to Swiss-Model (www.expasy.org) and a PDB structure was obtained. 2.2. Plasmid constructs 2.2.1. pGBKT7(AF aa1-105) The AF region coding for aa 1–105 was amplified by PCR using a human brain Quick-clone cDNA (Clontech) and oligonucleotides containing EcoR1 restriction site (italics), (5'GGCGAATTCATGGTGTTGGAAAGCAC3') and (5'GAC GAATTCTGATTCTTGCCTTGTCGG3'). The resulting PCR fragment was cut with EcoRI and inserted into the yeast expression vector pGBKT7 (Clontech) to generate a construct of truncated AF cDNA fused in-frame to the GAL4 DNA-binding domain of pGBKT7 as a bait for two-hybrid screening. 2.2.2. pGEX-flotillin-1 A PCR fragment encoding full length flotillin-1 was amplified from human brain cDNA (Clontech) using primer pair (5′CGGGAATTCATGTTTTTCACTTGTGGCCCAAATG3′) and (5'GTCGAATTCAGGCTCAGGCTGTTCTCAAAGG'3), which have EcoR1 sites (italics) in-frame with the pGEX-1λT vector (Amersham Biosciences), to produce the fusion protein GST-flotillin-1. The sequences amplified by PCR were confirmed by sequencing using BigDye terminator cycle sequencing (Applied Biosystems) and ABI Prism 310 genetic analyzer (Applied Biosystems).

whole brain from a 37-year-old Caucasian man, cause of death trauma, second library: human placenta). Approximately 2 × 106 clones from the human brain or the human placenta libraries were screened. Positive transformants were selected on medium lacking leucine, tryptophan, and histidine, synthetic dropout (SD). Yeast colonies were inoculated on new dropout plates lacking leucine, tryptophan, histidine, and adenine (SD-4 plates) and a colony lift filter assay was performed to test β-galactosidase activity, in order to observe the expression of LacZ reporter gene and so confirm the specificity of the interaction. cDNA inserts of the positive clones were amplified by PCR using primers complementary to the sequence of pACT2 vector. The total DNA was recovered from positive clones and introduced into Escherichia coli strain JM109. The plasmid from the libraries was recovered from the clones independently grown on the Luria–Bertani plate containing ampicillin, and analyzed by restriction digests. The positive inserts were further characterized by using the BigDye terminator cycle sequencing (Applied Biosystems) and ABI Prism 310 genetic analyzer (Applied Biosystems), searching for gene sequence similarity in the GenBank™ database with the program BLAST. In a second round of testing, the bait and prey plasmids were co-transformed into yeast strain AH109 (Clontech). The transformants were assayed again for growth on SD-4 plates and for β-galactosidase in a filter assay. 2.4. Expression and purification of GST-flotillin-1 Human flotillin-1 cDNA was cloned into the pGEX-1λT vector (Amersham Biosciences) to allow expression of an inframe fusion protein GST-flotillin-1 (see previous section). This plasmid (pGEX-flotillin-1) was transformed into E. coli strain BL21, and overexpression of GST-flotillin-1 was induced by adding 0.1 mM isopropyl-β-D-thiogalactopyranoside. The E. coli cells were harvested 4 h after induction. After sonication and centrifugation, the supernatant containing the expressed fusion protein was purified by passing the lysates through glutathione-agarose (Amersham Biosciences). The recombinant protein was either eluted with mild elution buffer or digested with thrombin to elute pure full length flotillin-1 protein without the GST affinity tail. The purity of the flotillin proteins was assayed by SDS-PAGE followed by staining with Coomassie Brilliant Blue or by immunoblotting using an anti-flotillin-1 monoclonal antibody (BD, Transduction Lab, USA). The blots were developed with alkaline phosphatase-conjugated goat antimouse immunoglobulins followed by 5-bromo-4-chloro-3-indolylphosphate and 4-nitro blue tetrazolium (Roche Diagnostics).

2.3. Yeast two-hybrid system 2.5. Expression of AF (aa 1–105) in vitro The yeast two-hybrid screen was carried out using a GAL4based yeast two-hybrid system (MATCHMAKER Two-Hybrid System 3; Clontech). The bait vector, pGBKT7 (AF aa 1–105), described above, was introduced into Saccharomyces cerevisiae strain PJ692A. Next, the transformed yeast was mated with the yeast strain Y187, pretransformed with an expression vector containing the cDNA library fused to its DNA activation domain. We used pretransformed MATCHMAKER libraries in the pACT2 vector from Clontech (mRNA source—first library: normal,

To generate [35S]-labelled AF (aa 1–105), pGBKT7 (AF aa 1– 105) was expressed using the TNT Quick-coupled transcription/ translation kit (Promega) and Redivue™ L-[35S]methionine (Amersham Biosciences) according to the manufacturer's instructions, using 1 μg of plasmid DNA. Reaction (50 μl) was carried out for 1.5 h at 30 °C, after which a small aliquot was removed and AF (aa 1–105) protein translation was monitored by SDS-PAGE (see above) followed by autoradiography. The polyacrylamide gel was

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309

305

were expressed using the system described above resulting in radioactively labelled c-Myc tag and SV40 large T-antigen together with a HA epitope respectively. 2.6. Co-immunoprecipitation

Fig. 1. Computer generated model describing the structure of AF (aa 15–151). The biologically active peptide (aa 36–51) is marked in black.

placed in fixing solution (50% methanol, 10% glacial acetic acid, 40% water) for 30 min followed by soaking in 1% glycerol for 5 min before completely drying on a sheet of filter paper at 80 °C for 1.5 h under vacuum. The gel was exposed on Kodak X-ray film for 15 h at room temperature before development to visualize AF (aa 1–105) together with the c-Myc epitope tag. As control vectors pGBKT7 without inserted fragment and pGADT7-T (Clontech)

To confirm the identified interaction, an in vitro co-immunoprecipitation was performed using the Matchmaker Co-IP Kit (Clontech). For each immunoprecipitation sample, 10 µl of radiolabelled in vitro translated AF (aa 1–105), including c-Myc tag, was combined with recombinant expressed flotillin-1 (with the GST tail) and incubated at room temperature for 1 h. The proteins were then mixed with 1 μg of c-Myc monoclonal antibody or 1 μg anti-flotillin-1 monoclonal antibody (BD, Transduction Lab, USA) for 1 h before the prepared Protein A beads were added to the reaction tubes and rotated for 1 h. The beads were washed five times with buffer 1 and twice with buffer 2 (Clontech). Bound proteins were eluted by boiling at 80 °C for 5 min in SDS-loading buffer and separated on a 15% SDS-PAGE gel followed by autoradiography or immunoblotting with antiflotillin-1 monoclonal antibody as described above. As controls radioactively labelled AF (aa 1–105) was mixed with anti-flotillin1, and flotillin-1 was mixed with c-Myc antibody. Flotillin-1 was also combined with the expressed pGBKT7 before incubation with c-Myc antibody or with anti-flotillin-1. Labelled AF (aa 1–105) was incubated with the irrelevant protein SV40 large T-antigen

Fig. 2. Co-immunoprecipitation assay performed in order to examine the in vitro interaction between [35S]-AF (aa 1–105) and recombinant expressed flotillin-1. The mixed proteins were resolved on a SDS-PAGE gel after incubation with c-Myc antibody, HA-Tag antibody or with flotillin-1 antibody, as indicated. The precipitates were analyzed by autoradiography showing [35S]-labelled products or by immunoblotting using anti-flotillin-1 antibody to detect flotillin-1 fusion protein. Results from the transparent film and the blotted filter were combined on the figure (A and B). To the left is the molecular weight standard applied (R). A). The association between AF (aa 1–105) and flotillin-1. Lane 1: AF (aa 1–105)+c-Myc antibody; Lane 2: flotillin-1+flotillin-1 antibody; Lane 3: AF (aa 1–105)+flotillin-1+c-Myc antibody; Lane 4: AF (aa 1–105) +flotillin-1+flotillin1 antibody. B). Control assay testing the specificity of the antibodies and the interaction between AF (aa 1–105) and flotillin-1. Lane 5: AF (aa 1–105)+flotillin-1 antibody; Lane 6: flotillin-1+c-Myc antibody; Lane 7: c-Myc tag+flotillin-1+c-Myc antibody; Lane 8: c-Myc tag+flotillin-1 +flotillin-1 antibody; Lane 9: AF (aa 1–105)+SV40 large T-antigen+c-Myc antibody; Lane 10: AF (aa 1–105) +SV40 large T-antigen+HA-Tag antibody.

306

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309

expressed from vector pGADT7-T and thereafter mixed with cMyc antibody or with HA-Tag polyclonal antibody (Clontech). 2.7. Dot-blot analysis of AF binding to flotillin-1 Pure protein of flotillin-1 expressed in E. coli was dotted on nitrocellulose membranes in 1:3 dilutions (starting with100 ng/ dot) and air dried. Non-specific binding sites on the nitrocellulose membranes were blocked by incubation in 1% BSA in PBS at room temperature for 1 h and then incubated with different AF peptides of the AF sequence (1–105, 1–51, or 316–340) at the concentration of 1.7 nM for 2 h. The peptides were made by organic chemical synthesis on solid phase by Innovagen AB (Lund, Sweden). The membranes were then treated for 1.5 h with characterized AF-antibodies, directed against different positions of the AF sequence, 1–105 (a105) or 316–340 (aM2), and preimmune-serum was used as control [13]. The blots were developed as described above. To verify the specificity of the AF-antibodies and to exclude cross-reactivity between the AF-antibodies and flotillin-1, the various AF-sequences (1–105, 1–51, and 316–340) and full length flotillin-1 were dotted on nitrocellulose membranes. The membranes were then incubated with preimmune-serum, a105, aM2 and anti-flotillin-1 and developed as described. 2.8. Immunohistochemistry Sections of 4 µm were prepared from formaldehyde fixed paraffin embedded rat CNS. After antigen retrieval by heating

the sections in a microwave oven in 0.01 M citrate buffer, pH 6.0, the sections were incubated with a monoclonal antibody against flotillin-1 and a polyclonal antibody against AF (a105). TRITC-conjugated goat anti-mouse antibodies and FITCconjugated goat anti-rabbit antibodies (Jackson Lab.) were used as secondary reagents. 3. Results 3.1. Computer generated model of AF We performed a protein structure prediction on the AF sequence (aa 15–151) using the GenTHREADER method [17]. As can be seen in Fig. 1, the generated globular protein has the active part of AF (aa 36–51) exposed outwards and consequently is able to be a reactive domain in the AF molecule. 3.2. Identification of flotillin-1 as an AF-binding protein using the yeast two-hybrid assay To identify new potential AF-interacting proteins in vivo, yeast two-hybrid screens were performed using an AF construct (aa 1– 105) as bait and cDNA libraries from human brain and placenta. Fourteen brain and fifteen placenta clones were isolated that scored positive for the reporter genes (trp1, leu2, ade2, his3, and lacZ). The cDNA plasmids were isolated, and duplicates were eliminated by restriction digests. Next, a total of 7 and 13 unique clones, respectively, were retested for specificity of β-galactosidase expression; 3 clones were identified as identical after both

Fig. 3. A. Dot blot showing specificity of antibodies. Nitrocellulose membranes were dotted with serial dilutions of AF peptides (1 =aa 1–105; 2= aa 1–51; 3=aa 316–340) and 4: full length flotillin-1. The membranes were probed with a105, aM2, anti-flotillin-1 and preimmune-serum as indicated. B. Nitrocellulose filter-binding assay for analysis of AF binding to flotillin-1. Recombinant produced flotillin-1 was immobilized onto nitrocellulose at the indicated concentrations and probed from left to right with different AF peptides (aa 1–105, 1–51, or 316–340), following which the membranes were treated with (A) preimmune-serum, (B) AF-antiserum a105, (C) AF-antiserum aM2.

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309

307

screens. Subsequent sequence analysis showed that two clones were coding for a partial sequence of human flotillin-1 cDNA, ranging from the codon of either aa 37 or aa 39 to the stop codon. The interaction was verified by re-transformation. Other identified interacting clones were coding for several proteasome related proteins of potential interest. However, since the aim of the study was to try to identify binding proteins which could be involved in the antisecretory activity of AF, further studies were concentrated on flotillin-1. Flotillin-1 is a component of lipid rafts, a membrane area involved in transport processes and flotillin-1 is thus an interesting candidate for AFinteractions.

the co-immunoprecipitation assay. The full length fusion protein GST-flotillin-1, was mixed with in vitro translated [35S]-labelled AF (aa 1–105) followed by anti-c-Myc or anti-flotillin-1 incubations. After extensive washing, the precipitated proteins bound to the Protein A beads were eluted and analyzed by autoradiography to detect the radioactively labelled AF (aa 1–105) or by immunoblotting to demonstrate flotillin-1. As shown in Fig. 2A, flotillin-1 was co-immunopurified with AF (aa 1–105), indicating a direct physical association between these proteins. All the control assays testing the antibody specificity or the interaction between AF (aa 1–105) and the irrelevant protein SV40 large T-antigen were negative (Fig. 2B).

3.3. In vitro interaction of AF with flotillin-1

3.3.2. Dot-blot Control experiments showed that the used AF-antibodies (a105 and aM2) were specific against the AF-sequences, (1– 105, and 316–340) respectively and that there was no crossreactivity against flotillin-1. (Fig. 3A)

3.3.1. Co-immunoprecipitation The protein interaction identified between AF (aa 1–105) and flotillin-1 in the yeast two-hybrid screens was verified in vitro by

Fig. 4. Sections from rat brain showing AF-immunoreactivity (A and B) green fluorescence/ FITC; Flotillin-1-immunoreactivity (C and D) red fluorescence/ TRITC, and overlay (E, F) showing co-localization yellow. Left panel: Fourth ventricle with choroid plexus. The epithelial cells in the choroid plexus express both markers, the ependyma lining the ventricle wall express only AF. Right panel: Section from the brain stem showing co-localization of AF and flotillin-1 in most neurons. Bars = 100 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

308

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309

In a complementary approach, dot-blot analysis was performed to confirm AF binding to flotillin-1. The proteins are not denaturated in the dot-blot method as in Western blot after SDS polyacrylamide electrophoresis, and consequently it is possible to analyze native protein-binding. The membranes dotted with flotillin-1 reacted positively after incubation with AF peptides aa 1–105 and aa 1–51, but not with AF peptide aa 316–340, visualized after incubation with antiserum a105, nor after treatment with antiserum aM2 or the preimmune-serum (Fig. 3B). The stringency of the test was high, since no specific binding of the irrelevant AF peptide, aa 316–340, or the antiserum aM2 was shown. The dot-blot assay shows that the binding to flotillin-1 occurs in the N-terminal of the AF sequence by an epitope located between aa 1–51. 3.4. Immunohistochemistry Sections from rat brain double labelled with antibodies against AF and flotillin-1 showed that there was a high degree of co-localization between AF and flotillin-1 in neurons and in the epithelial cells of the choroid plexus (Fig. 4). 4. Discussion In this study we present evidence that flotillin-1 can be a binding protein for Antisecretory Factor (AF). Although there can be problems with using the yeast two-hybrid system to identify cell membrane proteins, positive interactions coding for the integral membrane protein flotillin-1 were identified when screening both the brain and placenta libraries. Flotillin-1 is known as a key structural component and a marker of lipid rafts [18,19]. Lipid rafts are specialized areas in the cell membrane in which there is a high concentration of receptors, ion channels, and cytoskeletal contacts. These specialized areas are involved in many cellular processes, such as membrane trafficking, molecular sorting, and signal transduction [18,19]. Flotillin-1 belongs to a large family of integral membrane proteins that carry an evolutionarily-conserved domain called the prohibitin homology (PHB) domain; proteins carrying this domain are thought to have affinity for lipid rafts [19]. The dot-blot experiments showed that flotillin-1 interacts/ binds to the N-terminal part of AF, carrying the anti-inflammatory and antisecretory sequence, but not to the C-terminal part, carrying the ubiquitin binding sites. This indicates that the interaction between AF and flotillin-1 is probably not involved in the regulation of the proteolytic degradation in the proteasome, but rather in the regulation of the antisecretory and/or anti-inflammatory activity of AF. AF mRNA is expressed in many tissues [5] and AF appears to be produced in excess of other proteasome components [13]. A part of the synthesized AF is secreted, since it can be detected in plasma, but it is likely that AF can also act locally interacting with other intracellular proteins. Previous electrophysiological studies indicate the AF modulates GABAA receptor-mediated responses, but it does not appear to be a general GABA antagonist or agonist [10]. In these experiments evoked glutamatergic and GABAergic synaptic transmissions were investigated in hippocampus slices

prepared from rats in which an increased endogenous AF activity had been induced by cholera toxin or a specific feed [10]. The GABAA-mediated synaptic transmission was suppressed in these slices compared to slices prepared from control animals. Since AF and flotillin-1 are co-localized in the CNS the registered effect could be due to an interaction between intracellular AF and flotillin-1. GABAA receptors have been shown to be localized to discrete lipid raft microdomain clusters in rat neurons [20], and lipid rafts may contribute to the clustering of membrane proteins at distinct functional localizations on the cell surface. Such clustering could occur after interaction of a ligand e.g. AF with flotillin-1. It is thus possible that AF exerts its effects by influencing a number of receptors through binding to flotillin-1. Several other proteins have recently been shown to regulate the localization of signalling proteins to lipid rafts by binding to flotillin-1, for example, the intracellular domain of amyloid precursor protein [21], myocilin [22], proteins carrying the sorbin homology domain [23], and neuroglobin [24]. AF is a potent inhibitor of the intestinal fluid secretion induced by cholera toxin [1,8,25]. Cholera toxin attaches to mucosal cells by binding to the ganglioside GM1. Enzyme-linked immunosorbent assays performed in our laboratory show that AF does not bind to GM1 (unpublished). GM1 has been shown to be co-localized with flotillin-1 [26] and it is thus possible that AF can modulate the signal pathway of cholera toxin by interaction with flotillin-1. In conclusion, we have identified and verified flotillin-1 as an AF-binding protein. By interacting with flotillin-1 intracellular AF may affect secretory processes by regulating the localization of signal proteins to lipid rafts. However, in many systems the effects of endogenous AF can be reproduced by exogenous addition of AF or AF peptides, indicating the presence of interacting structures exposed on the external part of the plasma membrane. Further studies are needed to identify these structures. Acknowledgements This work was supported by grants from the Frimurare-Barnhusdirektionen, Adlerbertska Research Foundation, the Magnus Bergvall Foundation, the Swedish animal welfare agency, and the Swedish federal government under the LUA/ALF agreement, grant 7157. References [1] Johansson E, Lönnroth I, Lange S, Jennische E, Jonson I, Lönnroth C. Molecular cloning and expression of a pituitary gland protein modulating intestinal fluid secretion. J Biol Chem 1995;270:20615–20. [2] Tateishi K, Misumi Y, Ikehara Y, Miyasaka K, Funakosh A. Molecular cloning and expression of rat antisecretory factor and its intracellular localization. Biochem Cell Biol 1999;77:223–8. [3] Ferrell K, Deveraux Q, van Nocker S, Rechsteiner M. Molecular cloning and expression of a multiubiquitin chain binding subunit of the human 26S protease. FEBS Lett 1996;381:143–8. [4] Young P, Deveraux Q, Beal RE, Pickart CM, Rechsteiner M. Characterization of two polyubiquitin binding sites in the 26 S protease subunit 5a. J Biol Chem 1998;273:5461–7.

E. Johansson et al. / Regulatory Peptides 146 (2008) 303–309 [5] Lange S, Jennische E, Johansson E, Lönnroth I. The Antisecretory Factor— synthesis and intracellular localization in porcine tissues. Cell Tissue Res 1999;296:607–17. [6] Davidson TS, Hickey WF. Distribution and immunoregulatory properties of antisecretory factor. Lab Invest 2004;84:307–19. [7] Kiss P, Szabo A, Hunyadi-Gulyas E, Medzihradszky KF, Lipinszki Z, Pal M, et al. Zn2+-induced reversible dissociation of subunit Rpn10/p54 of the Drosophila 26 S proteasome. Biochem J 2005;391:301–10. [8] Lönnroth I, Lange S, Jennische E, Johansson E, Jonson I, Torres J. Cholera toxin protects against action by Clostridium difficile toxin A. The role of antisecretory factor in intestinal secretion and inflammation in rat. APMIS 2003;111:969–77. [9] Davidson TS, Hickey WF. Antisecretory factor is regulated by inflammatory mediators and influences the severity of experimental autoimmune encephamyelitis. J Leukoc Biol 2004;74:835–44. [10] Kim M, Wasling P, Xiao MY, Jennische E, Lange S, Hanse E. Antisecretory factor modulates GABAergic transmission in the rat hippocampus. Regul Pept 2005;129:109–18. [11] Johansson E, Lange S, Lönnroth I. Identification of an active site in the antisecretory factor protein. Biochim Biophys Acta 1997;1362:177–82. [12] Björck S, Bosaeus I, Ek E, Jennische E, Lönnroth I, Johansson E, Lange S. Food-induced stimulation of the antisecretory factor improve recurrence from human colonic inflammatory disease. A study of a concept. GUT 2000;46:824–9. [13] Jennische E, Johansson E, Hansson HA, Jonson I. Immunohistochemical staining patterns using epitope-specific antibodies indicate conformation variants of antisecretory factor/S5a in the CNS. APMIS 2006;114:529–38. [14] Lange S, Lönnroth I. The Antisecretory Factor: synthesis, anatomical and cellular distribution and biological action in experimental and clinical studies. Int Rev Cyt 2001;210:39–75. [15] Edwards YJ, Perkins SJ. Assessment of protein fold predictions from sequence information: the predicted alpha/beta double wound fold of the von Willebrand factor type A domain is similar to its crystal structure. J Mol Biol 1996;260:277–85.

309

[16] Tanji K, Tanaka T, Kamitani T. Interaction of NUB1 with the proteasome subunit S5a. Biochem Biophys Res Commun 2005;337:116–20. [17] Jones DT. GenTHREADER: an efficient and reliable protein fold recognition method for genomic sequences. J Mol Biol 1999;287:797–815. [18] Bickel PE, Scherer PE, Schnitzer JE, Oh P, Lisanti MP, Lodish HF. Flotillin and epidermal surface antigen define a new family of caveolae-associated integral membrane proteins. J Biol Chem 1997;272:13793–802. [19] Morrow IC, Parton RG. Flotillins and the PHB domain protein family: rafts, worms and anaesthetics. Traffic 2005;6:725–40. [20] Dalskov SM, Immerdal L, Niels-Christiansen LL, Hansen GH, Schousboe A, Danielsen EM. Lipid raft localization of GABA-A receptor and Na+, K+ATPase in discrete microdomain clusters in rat cerebellar granule cells. Neurochem Int 2005;46:489–99. [21] Chen TY, Liu PH, Ruan CT, Chiu L, Kung FL. The intracellular domain of amyloid precursor protein interacts with flotillin-1, a lipid raft protein. Biochem Biophys Res Comm 2006;342:266–72. [22] Joe MK, Sohn S, Choi YR, Park H, Kee C. Identification of flotillin-1 as a protein interacting with myocilin: implications for the pathogenesis of primary open-angle glaucoma. Biochem Biophys Res Commun 2005;336:1201–6. [23] Kimura A, Baumann CA, Chiang CH, Saltiel AR. The sorbin homology domain: a motif for the targeting of proteins to lipid rafts. Proc Natl Acad Sci USA 2001;98:9098–103. [24] Wagasugi K, Nakano T, Kitatsuji C, Morishima I. Human neuroglobin interacts with flotillin-1, a lipid raft microdomain-associated protein. Biochem Biophys Res Commun 2004;318:453–60. [25] Grondahl ML, Sorensen H, Unmack MA, Holm A, Skadhauge E. Neuronal involvement in the effect of an antisecretory factor-derived peptide on induced secretion in the porcine small intestine. J Comp Physiol, A Neuroethol Sens Neural Behav Physiol 2002;188:589–94. [26] Li N, Mak A, Richards DP, Naber C, Keller BO, Li L, Shaw ARE. Monocyte lipid rafts contain proteins implicated in vesicular trafficking and phagosome formation. Proteomics 2003;3:536–48.