Differential display implicates cyclophilin A in adult cortical plasticity

European Journal of Neuroscience, Vol. 18, pp. 61±75, 2003

ß Federation of European Neuroscience Societies

Differential display implicates cyclophilin A in adult cortical plasticity Lutgarde Arckens,1 Estel Van der Gucht,1 Gert Van den Bergh,1 Ann Massie,1 Inge Leysen,1 E. Vandenbussche,2,y Ulf T. Eysel,3 Roger Huybrechts4 and Frans Vandesande1 1

Laboratory of Neuroplasticity and Neuroproteomics, Katholieke Universiteit Leuven, Naamsestraat 59, B-3000 Leuven, Belgium Laboratory of Psychophysiology, Medical School, Katholieke Universiteit Leuven, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium 3 Department of Neurophysiology, Medical School, Ruhr-UniversitaÈt Bochum, D-447780 Bochum, Germany 4 Laboratory of Developmental Physiology and Molecular Biology, Katholieke Universiteit Leuven, Naamsestraat 59, B-3000 Leuven, Belgium 2

Keywords: cat, dLGN, immunocytochemistry, in situ hybridization, quantitative analysis, sensory deprivation, topographical organization, visual cortex

Abstract Removal of retinal input from a restricted region of adult cat visual cortex leads to a substantial reorganization of the retinotopy within the sensory-deprived cortical zone. Little is known about the molecular mechanisms underlying this reorganization. We used differential mRNA display (DDRT-PCR) to compare gene expression patterns between normal control and reorganizing visual cortex (area 17±18), 3 days after induction of central retinal lesions. Systematic screening revealed a decrease in the mRNA encoding cyclophilin A in lesion-affected cortex. In situ hybridization and competitive PCR con®rmed the decreased cyclophilin A mRNA levels in reorganizing cortex and extended this ®nding to longer postlesion survival times as well. Western blotting and immunocytochemistry extended these data to the protein level. In situ hybridization and immunocytochemistry further demonstrated that cyclophilin A mRNA and protein are present in neurons. To exclude the possibility that differences in neuronal activity per se can induce alterations in cyclophilin A mRNA and protein expression, we analyzed cyclophilin A expression in the dorsal lateral geniculate nucleus (dLGN) of retinally lesioned cats and in area 17 and the dLGN of isolated hemisphere cats. In these control experiments cyclophilin A mRNA and protein were distributed as in normal control subjects indicating that the decreased cyclophilin A levels, as observed in sensory-deprived area 17 of retinal lesion cats, are not merely a re¯ection of changes in neuronal activity. Instead our ®ndings identify cyclophilin A, classically considered a housekeeping gene, as a gene with a brain plasticity-related expression in the central nervous system.

Introduction The adult mammalian brain has the capacity to reorganize dramatically in answer to large-scale peripheral deafferentations. The loss of sensory input due to restricted deafferentations or injury to the sensory surface typically results in signi®cant shifts in the topography of the functional representations. The part of sensory cortex that normally represents the lesioned portion of the sensory surface will become occupied by the representation of adjacent sensory surface (Merzenich et al., 1983; Robertson & Irvine, 1989; Kaas et al., 1990; Gilbert & Wiesel, 1992). A common mechanism for different forms of adult plasticity is the potentiation of previously existing subthreshold inputs through local changes in synaptic weight (Wall, 1977; Bear et al., 1987). In addition, the central nervous system has the capacity for axonal and dendritic growth and synapse turnover in order to adapt the neuronal circuitry. New synaptic contacts would become the new source of activation for

Correspondence: Dr Lutgarde Arckens, as above. E-mail: [email protected] y

Deceased, December 11, 2002

Received 29 January 2003, revised 7 April 2003, accepted 7 April 2003 doi:10.1046/j.1460-9568.2003.02726.x

sensory-deprived neurons. Both phenomena could occur at the cortical and subcortical level in order to contribute to the cortical reorganization. However, in the visual system, electrophysiological experiments and tracing studies did not reveal a signi®cant contribution of the geniculate nucleus to cortical reorganization after retinal lesions (Eysel et al., 1980, 1981; Eysel, 1982; Darian-Smith & Gilbert, 1995). Therefore the long-range horizontal connections are recognized as the prime candidates for effecting reorganization in visual cortex (Darian-Smith & Gilbert, 1994, 1995; Chino et al., 1995; Chino, 1995; Das & Gilbert, 1995; Donoghue, 1995). Studies of synaptic plasticity in the mature CNS have demonstrated that changes in gene expression are required for the consolidation of long-lasting changes in synaptic strength (Frank & Greenberg, 1994; Nguyen et al., 1994; Tully, 1997). Several immediate early genes, neurotrophins, neurotransmitters and their receptors have already been implicated in adult brain plasticity (Garraghty et al., 1991; Rosier et al., 1995; Conti et al., 1996; Arckens et al., 1998, 2000a,b; Kilgard & Merzenich, 1998; Gierdalski et al., 1999; Obata et al., 1999). Nevertheless, the molecular cascade responsible for representational plasticity remains illusive. Differential display (DDRT-PCR) is a mRNA ®ngerprinting technique allowing the systematic screening for differences in gene

62 L. Arckens et al. expression between experimental conditions (Liang & Pardee, 1992). The possibility to detect up-regulated and down-regulated genes makes it an attractive screening method for identifying genes involved in cortical plasticity. Our strategy included the systematic comparison of the mRNA expression pattern with DDRT-PCR between the sensory-deprived and the normal portion of visual areas 17 and 18 three days after the induction of homonymous central retinal lesions. We identi®ed one of the down-regulated genes as cyclophilin A. We used in situ hybridization and immunocytochemistry to con®rm the differential expression and to reveal the regional and cellular distribution of cyclophilin A. These techniques also allowed the investigation of the effect of postlesion survival time. Competitive PCR, Western blotting and immunocytochemistry permitted quanti®cation of the differences. We further demonstrated that sensory deprivation is not enough to induce decreased cyclophilin A expression in visual cortex as observed after retinal lesions to ascertain a role for cyclophilin A in adult brain plasticity.

Materials and methods Animals and tissue preparation All animal experiments were carried out in accordance with the standards of the Institutional Ethical Committee of the Katholieke Universiteit Leuven, Belgium and the European Communities Council Directive of 24 November 1986 (86/609/EEC). All efforts were made to reduce the number of animals used. Forty-three adult cats (Animal facilities, Katholieke Universiteit Leuven, Belgium) were used. Most of these animals have been used in previous investigations (Zhang et al., 1995; Arckens et al., 1998, 2000a). Eight animals served as normal controls. In 20 animals, the part of the retinas representing the area centralis of the visual ®eld was photocoagulated. Binocular lesions of 108 or 208 were induced under ketalar/xylacine anaesthesia (0.5 mL Ketalar, 0.2 mL Xylacine, i.m.) using a LOG-2 Xenon light photocoagulator (Clinitex Inc., Denver, CO). Before and after the induction of the lesions a fundal photograph was taken of both eyes as an indication of the position and size of the lesion (Vandenbussche et al., 1990). The animals were maintained in a normal light environment (10 h light : 14 h dark) for different survival times [3 days (n ˆ 4), 1 week (n ˆ 2), 2 weeks (n ˆ 4), 1 month (n ˆ 3), 3 months (n ˆ 3), 4.5 months (n ˆ 1), 6 months (n ˆ 1), 12 months (n ˆ 2)]. In 15 cats, visual input was con®ned to one hemisphere by sectioning the left optic tract and the corpus callosum as described by Vanduffel et al. (1995). Both operations were carried out under sterile conditions using a Zeiss operating microscope and following deep general anaesthesia with ketamine hydrochloride (Ketalar: 10 mg/kg i.m.) and pentobarbital (Nembutal: 60 mg/kg i.v.). Completeness of the optic tract section was con®rmed behaviourally in all animals using a visual ®eld perimetry test (Sprague & Meikle, 1965; Sherman, 1973; Vanduffel et al., 1995). Post-operation survival time varied between 3 days and one year [3 days (n ˆ 2), 2 weeks (n ˆ 5), 1 month (n ˆ 3), 3 months (n ˆ 1), 4.5 months (n ˆ 2), 8 months (n ˆ 1), 12 months (n ˆ 1)]. We will refer to these animals as `isolated hemisphere' cats. These cats were used as control animals to identify those genes that are already affected by loss of neuronal activity alone without an accompanying shift in cortical representational maps. All animals were killed with an overdose of Nembutal (60 mg/kg, i.v.) under ketamine hydrochloride anaesthesia (10 mg/kg i.m.). When applied for mRNA analysis or Western blotting, the brains were rapidly removed and immediately frozen by immersion in liquid-nitrogen cooled isopentane and stored at 70 8C. Frontal sections of 15 and 200 mm were cut on a cryostat. The 15 mm sections were mounted on

0.1% poly L-lysine (Sigma-Aldrich, St. Louis, MO) coated slides and stored at 20 8C until processed for in situ hybridization for zif268. We have shown previously that in situ detection of the inducible immediate early gene zif268 is a reliable tool for outlining the lesion-affected portions of visual areas following partial sensory deprivation (Arckens et al., 2000a). The 200 mm sections were collected on RNase free slides and were used to either isolate total RNA or to extract proteins from brain tissue punches collected separately from sensory-deprived and normal portions of areas 17 and 18 with the zif268 autoradiograms as a guide to discriminate these two cortical regions (Fig. 3B, boxes indicate the two regions from which tissue was sampled). Total RNA was used in differential display and competitive PCR. The protein extracts were used in Western blotting experiments. For the cyclophilin A in situ hybridization experiments, brains were rapidly removed, immediately frozen by immersion in liquid-nitrogen cooled isopentane, stored at 70 8C and cut into 15 mm frontal sections on a cryostat. For cyclophilin A immunocytochemistry, the animals were perfused transcardially with cold 4% paraformaldehyde in phosphate buffer (PBS, 0.15 M, pH 7.4). The brains were post®xed for 24 h in the same ®xative and then rinsed in water for another 24 h. Frontal sections (50 mm) were cut on a vibratome, collected in 24-well plates and processed for immunocytochemistry. Differential display (DDRT-PCR) DDRT-PCR has been performed using the DeltaTM Differential display kit (BD Biosciences Clontech, Palo Alto, CA). Total RNA was prepared from the tissue punches of the lesion-affected and normal portions of area 17 and 18 of a retinal lesion cat, 3 days postlesion and using the RNeasy kit (Qiagen, Westburg, Leusden, The Netherlands) according to the manufacturer's instructions. The RNA preparations were subjected to DNase I digestion in the presence of a ribonuclease inhibitor for 30 min. RNA was extracted with phenol and precipitated in ethanol. Equal amounts of RNA (2 mg) were transcribed to cDNA in 10 mL reactions containing 1 mM dNTP, 0.1 mM oligo dT primer, 200 units Moloney murine leukaemia virus (MMLV) reverse transcriptase and 2 mL 5  First strand buffer (250 mM Tris, pH 8.3, 30 mM MgCl2, 375 mM KCl). Reverse transcription was performed for 1 h at 42 8C and followed by a termination step at 75 8C for 10 min. The cDNA was diluted and stored at 20 8C until used. Aliquots of 1 mL cDNA were subjected to a polymerase chain reaction (PCR) employing different combinations of the T and P display primers (1 mM each), along with 50 mM dNTP, 50 nM 33PdATP (1000±3000 Ci/mmol, NENTM Life Science, Zaventem, Belgium) and 1 U Advantage KlenTaq polymerase (BD Biosciences Clontech) in a 20 mL ®nal volume. PCR conditions were as described by the manufacturer: one cycle at 94 8C for 5 min for denaturing, cooling to 40 8C for low stringency annealing of the primers for 5 min, heating up to 68 8C for extension during 5 min. This round was followed by two cycles of 94 8C for 2 min, 40 8C for 5 min, and 68 8C for 5 min. This was followed by 25 high stringency cycles: 94 8C for 1 min, 60 8C for 1 min and 68 8C for 1 min. One ®nal step of 68 8C for 7 min was added to the last cycle. The differential display reactions were either stored at 20 8C or examined immediately in electrophoresis. Each PCR sample was separated on a 0.4-mm thick 5% polyacrylamide/8 mM urea sequencing gel in a S2 sequencing system (Life Technologies, Merelbeke, Belgium) at 2700 V. Gels were dried on Whatman 3MM paper at 80 8C for 1 h and were exposed to X-ray ®lms (b-Max Hyper®lm, Amersham Biosciences, Roosendaal, The Netherlands) for 24 h. Differential bands were excised from the gel, boiled in 40 mL sterile H2O for 5 min and the cDNA was separated form the gel by centri-

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity fugation. The obtained preparations were used for re-ampli®cation by 30 high stringency cycles in 50 mL PCR mixtures containing the corresponding primer pairs as used in the original display reaction under identical conditions with the exception of the initial round for nonspeci®c annealing and omitting the alfa 33P-dATP (NENTM. Life Science). Re-ampli®ed cDNAs were puri®ed by agarose gel electrophoresis and gel extraction using the `Qiaquick gel extraction kit' (Qiagen) and were ligated into pCR1IITM vector, used to transform competent Escherichia coli INValfaF0 cells according to the protocol of the `TA cloning kit' (Invitrogen, Life Technologies, Merelbeke, Belgium). Cloned cDNA fragments were sequenced using the SequenaseTM Version 2.0 DNA sequencing kit (Amersham Biosciences) using SP6-derived and T7-derived sequencing primers (Invitrogen). The identity of the cDNAs was determined by sequence homology searching using the gene database (EMBL and GenBank). Rapid ampli®cation of cDNA ends (RACE) technology was used to obtain additional sequence information to resolve the complete coding sequence. We performed 50 and 30 RACE PCR following the protocol of the `MarathonTM cDNA Ampli®cation Kit' (BD Biosciences Clontech). Double stranded cDNA was synthesized starting from total RNA of cat visual cortex. Following creation of blunt-ends with T4 DNA polymerase, both ends of the double stranded cDNA were ligated to the specially designed Marathon cDNA Adaptor by T4 DNA ligase. The 50 and 30 RACE reactions were performed using the Marathon adaptor primer, AP1, in combination with gene-speci®c primers (Fig. 1A, sequences in bold). The resulting PCR products were puri®ed and ligated into pCR1IITM vector and used to transform competent E. coli INValfaF0 cells according to the protocol of the `TA cloning kit' (Invitrogen). Clones with inserts of the expected size, as identi®ed in electrophoresis, were sequenced as described for the display clones.

63

The cat cyclophilin A mRNA sequence is available from GenBank (accession number AY029366). In situ hybridization Cyclophilin A mRNA expression was analyzed in normal controls (n ˆ 3), in cats with retinal lesions [3 days (n ˆ 2), 1 week (n ˆ 2), 1 month (n ˆ 2), 3 months (n ˆ 2), 4.5 months (n ˆ 1), 6 months (n ˆ 1), 12 months (n ˆ 2)] and in isolated hemisphere cats [(3 days (n ˆ 2), 2 weeks (n ˆ 2), 1 month (n ˆ 2), 3 months (n ˆ 1), 4.5 months (n ˆ 2), 8 months (n ˆ 1), 12 months (n ˆ 1)]. Labelling and puri®cation of the oligonucleotide probes as well as the detection of the speci®c signal following in situ hybridization were performed as described earlier (Arckens et al., 1995, 1998, 2000a). Brie¯y, oligonucleotides were 30 -end-labelled with 35S-dATP (NENTM Life Science) and terminal deoxynucleotidyl transferase (Life Technologies) and separated from unincorporated nucleotides with Nensorb-20 columns (NENTM Life Science). Cryostat sections were post®xed in 4% paraformaldehyde in PBS (0.1 M, pH 7.4), dehydrated and delipidated. The sections were incubated overnight at 37 8C with approximately 8  105 counts/min probe in 500 mL hybridization buffer (50% formamide, 4  SSC, 1  Denhardt's solution, 10% dextran sulphate, 100 mg/mL salmon sperm DNA, 250 mg/mL tRNA, 60 mM DTT, 1% N-lauryl-sarcosine, 26 mM NaHPO4, pH 7.4). The next day the sections were washed four times 15 min in 1  SSC at 42 8C, dried and exposed to autoradiographic ®lm (b-Max Hyper®lm, Amersham Biosciences) for 2±4 weeks or dipped in emulsion (LM-1, Amersham Biosciences) and exposed for 6 weeks to reveal regional and cellular localization of the probes, respectively. The ®lm and emulsion were developed in Kodak D19 (Kodak, Belgium) and transferred to Ilford Hypam (Kodak) rapid ®x. The emulsion-dipped sections were counter-

Fig. 1. Nucleotide and deduced amino acid sequence of the cDNA encoding cat cyclophilin A. One continuous open reading frame of 492 nucleotides encodes a protein of 164 amino acids. The poly adenylation signal is underlined. The positions of the P9 display primers are indicated in black, the 30 and 50 RACE primers are indicated in bold, the oligonucleotide probe for in situ hybridization is boxed and the gene-speci®c primers used in competitive PCR are indicated in grey (GENBANK accession number: AY029366). ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

64 L. Arckens et al. stained with thionin to visualize the cell bodies and the lamination pattern of the dLGN and the visual cortex. The sections were dehydrated and coverslipped and examined microscopically with bright®eld and dark-®eld optics using a Leitz DM RBE microscope (Leica, Leitz Instruments, Heidelberg, Germany). Four control experiments were performed to ascertain the speci®city of the antisense oligonucleotide probes used for the in situ hybridizations. Sets of tissue sections were hybridized with the sense oligonucleotide in an identical manner to the antisense oligonucleotide probe. A competition experiment included the prehybridization of sets of tissue sections with a 50-fold excess of relevant or irrelevant unlabelled probe for 1 h prior to hybridization with labelled antisense probe. Some sections were incubated with RNase containing buffer (0.005%) (Sigma-Aldrich) in 0.1 M PBS (pH 7.4) for 1 h at 37 8C prior to hybridization with labelled probe. An oligonucleotide probe (Fig. 1, boxed sequence) complementary to the nucleotides 325±372 of the cat cyclophilin A sequence was used for the detection of cyclophilin A mRNA in cat brain. Zif268 in situ hybridization has been performed with a probe complementary to the nucleotides encoding amino acids 2±16 of the rat zif268 gene (Wisden et al., 1990). This oligonucleotide probe has already been used successfully for the analysis of zif268 mRNA expression in cat visual system in previous publications (Zhang et al., 1994, 1995; Arckens et al., 2000a).

cDNA and 2 mL of the DNA competitor at a known concentration. The amount of competitor DNA in each reaction tube varied from 10 to 10 4 attomoles. The PCR pro®le was: denaturation at 94 8C for 10 min, annealing at 50 8C for 1 min and extension at 72 8C for 1 min for 29 cycles in a DNA thermal cycler (Applied Biosystems). PCR products were visualized on 4% Nusieve GTG agarose TAE gels (FMC Bioproducts, SanverTECH, Boechout, Belgium) containing 10 mg/mL ethidium bromide (Sigma-Aldrich) and were analyzed densitometrically using the Image Master 1D Elite version 3.01 (Amersham Biosciences). Variation in ¯uorescence intensity due to the molecular weight difference between target (231 bp) and competitor DNA (303 bp) were corrected. The ratio of the ampli®ed cyclophilin A DNA product to the competitor DNA PCR product was calculated as [target ¯uorescence/mimic ¯uorescence]  [mimic size (bp)/target size (bp)]. For each individual competitive PCR reaction the log10 of the ratio (y-axis) was plotted against the corresponding log10 of the initial known competitor concentration in attomoles/mL added at each ampli®cation (x-axis). The initial amount of target cyclophilin A was calculated by determining the point of the curve at which the ratio of the target to the competitor equals 1 (when y ˆ 0). The Wilcoxon Matched-Pairs Signed-Ranks Test was used for statistical analysis with P < 0.05 considered signi®cant.

Competitive PCR

Tissue punches collected separately from the lesion-affected and normal portions of area 17 of retinal lesion cat (2 weeks, n ˆ 3) and isolated hemisphere cat (2 weeks, n ˆ 2), and of corresponding regions in normal controls (n ˆ 2) were homogenized in lysis buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol) using a Te¯on glass homogenizer. The homogenates were boiled (5 min), sonicated (15 s) and centrifuged at 10 000 g for 10 min. The supernatants were taken and the pellets discarded. Protein concentrations were determined using the modi®ed Bradford method described by Qu et al. (1997). From each sample 1, 0.5 and 0.1 microgram protein, together with a multicoloured molecular weight marker (Novex, Invitrogen), were separated on a 4±12% Bis Tris gel (Novex, Invitrogen) and transferred to nitrocellulose membrane (Amersham Biosciences). Non-speci®c binding was blocked by preincubating the membrane for 1 h in 5% dry nonfat milk in Tris stock buffer (0.05 M Tris, pH 7.6). After a short rinse in Tris-saline (0.01 M Tris, 0.1% Triton X-100, pH 7.6) the membrane was incubated with the polyclonal rabbit anti-human cyclophilin A serum (ABR, 10 P's, Zandhoven, Belgium, lot # PA3021) at 1 : 5000 in Tris-saline, overnight at room temperature. According to the manufacturer's instructions, this antibody is cyclophilin A speci®c but may also recognize cyclophilin B and cyclophilin C because of the high degree of homology with cyclophilin A. The membrane was washed 3  5 min with the same buffer and incubated with a secondary goat anti-rabbit-peroxidase conjugate (1 : 1000, Dako, Glostrup, Denmark) in Tris-saline for 30 min. After three 5min rinses in Tris-saline and one 5-min rinse in Tris stock, the peroxidase activity was detected with the enhanced chemiluminescence system (ECL-Plus) and Hyper®lmTM ECL (Amersham Biosciences). For densitometry, the resulting Western blot images were scanned on a Sharp JX-330 scanner. Care was taken that no intensity values exceeded the dynamic range of the autoradiographic ®lm or our scanner. These images were then analyzed using Image Master 1D Elite software version 3.01 (Amersham Biosciences). Lanes and bands were de®ned, background was subtracted (rubber band) and band volumes were measured. The Wilcoxon Matched-Pairs Signed-Ranks Test was used for statistical analysis with P < 0.05 considered sig-

Total RNA was isolated from tissue punches of lesion-affected and normal area 17 of retinal lesion cat (2-weeks postlesion, n ˆ 2) and of normal and visually deprived cortex of an isolated hemisphere cat (2 weeks postlesion, n ˆ 1) using the `RNeasy kit' (Qiagen). Total RNA concentrations were determined spectrophotometrically at 260 nm. The synthesis of cDNA out of total RNA (20 ng) was performed in a 20-mL reaction volume with random hexamer primers and MMLV transcriptase from the `RNA PCR Core kit' according to the manufacturer's instructions (Applied Biosystems, Nieuwekerk a/d Ysel, The Netherlands). The reaction mixture was incubated at 42 8C for 45 min, 90 8C for 5 min and 4 8C for 10 min. The cDNA products were stored at 20 8C until use. Expression levels of mRNA encoding cyclophilin A were quanti®ed by competitive PCR (Van der Gucht et al., 2003). In a competitive PCR, the competitor and the target template are ampli®ed using the same primer pair in the same reaction. This competitor is a DNA fragment containing the same primer template sequences as the target but a completely different intervening sequence and is constructed by two successive PCR ampli®cations according to the protocol of the `PCR MIMIC Construction kit' (BD Biosciences Clontech). The competitor was puri®ed on a chromatographic Chroma-spin ‡ TE100 column (BD Biosciences Clontech). The concentration of the puri®ed DNA competitor was determined spectrophotometrically at 260 nm and by densitometric analysis with a DNA mass marker (fX174/HaeIII, New England Biolabs Inc., Westburg) containing known quantities of DNA. The reaction mixture of each competitive PCR ampli®cation consisted of 2 mL MgCl2 (25 mM), 5 mL 10  PCR buffer Gold (Applied Biosystems), 1 mL of each dNTP (0.4 mM), 0.25 mL Ampli Taq Gold1 DNA polymerase (5 U/mL; Applied Biosystems), 0.5 mM of the upstream cat-speci®c cyclophilin A primer 50 GGCAAGTCCATCTACGGGG30 (Eurogentec, Seraing, Belgium) complementary to nucleotides 223±241 and 0.5 mM of the downstream cat-speci®c cyclophilin A primer 50 CTTGCGTTCCGGGACCC30 (Eurogentec) spanning nucleotides 536±553 (Fig. 1, grey sequences) and 24.7 mL distilled water. To this reaction mixture was added 2 mL of the sample

Western blotting

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity

65

ni®cant. Bars represent the log10 of the mean of the band volume ratios of central/peripheral cyclophilin A for retinal lesion cats and of left/ right hemisphere area 17 for isolated hemisphere cats.

cats and of the left/right hemisphere area 17 ratio for isolated hemisphere cats.

Immunocytochemical procedure

Results

Vibratome sections of control subjects (n ˆ 3), retinal lesion cats [3 days (n ˆ 2), 14 days (n ˆ 2), 1 month (n ˆ 1), 3 months (n ˆ 1)] and isolated hemisphere cats [14 day cat (n ˆ 1), 1 month cat (n ˆ 1)] were immunostained for cyclophilin A. The vibratome sections were preincubated with normal goat serum (1 : 5, 45 min). For the detection of cyclophilin A, we used the same polyclonal rabbit antiserum against human cyclophilin A (ABR) as applied in Western blotting experiments at a dilution of 1 : 2000. The primary antiserum was detected with biotinylated goat anti-rabbit IgGs (1 : 600, 30 min, Dako) and peroxidase-conjugated streptavidin (1 : 500, 30 min, Dako). All the dilutions were made in Tris-saline (0.01 M, pH 7.4) and all incubations were performed at room temperature under gentle agitation. The reaction product was visualized as a black precipitate using the glucose oxidase-DAB-nickel method (Shu et al., 1988). Sections were mounted on gelatin-coated slides, dehydrated, and coverslipped. Results of the immunocytochemical experiments were analyzed using a Leitz DM RBE microscope (Leica) equipped with a colour video camera (Optronics Engeneering, Goleta, CA) attached to a computer-aided image analysing system (Bioquant, R&M Biometrics, Nashville, TN). Areal densities were determined at a magni®cation of 40 following the procedures of Leuba et al. (1998). The thickness of the vibratome sections (50 mm) made the determination of the laminar boundaries, as well as the recognition of cell types, easy (O'Leary, 1941; Lund et al., 1979). We quanti®ed the number of neuronal pro®les in frontal sections (n ˆ 3 for each animal) taken between anterior-posterior level P9 and P4. Immunoreactive neuronal pro®les were selected using a semiautomatic threshold procedure, based on the optical density of the neuronal pro®les. Criteria of object inclusion were the shape and the size of the neuronal perikarya and a strong immunoreactive status. The numbers of selected neuronal pro®les were used to determine the mean number per square millimetre cortex within cortical layers II± III and VI within each cortical region (lesion-affected vs. control). The Wilcoxon Matched-Pairs Signed-Ranks test was performed for statistical analysis with P < 0.05 considered signi®cant. Bar graphs represent the log10 of the central/peripheral ratio for retinal lesion

Differential display revealed the up- and down-regulated expression of several genes in primary visual cortex three days after the induction of homonymous retinal lesions. The use of 95 primer pair combinations resulted in approximately 100 clones containing candidate gene fragments with a role in cortical plasticity. Sequencing of the cDNA fragments identi®ed ®ve genes with a higher expression and seven genes with a lower expression in lesion-affected compared to control peripheral visual cortex of the same animal (Table 1). As a ®rst high throughput screening method, synthetic oligonucleotides against these genes were applied in in situ hybridization experiments. Only for clone P4P91, which showed high homology to monkey cyclophilin A based on sequencing information (Table 1), the hybridization signal con®rmed the decreased expression in lesion-affected cortex as observed in the differential display reaction. At the time we identi®ed cyclophilin A as a potential plasticityrelated gene in mammalian brain, our laboratory was actively optimizing a competitive PCR assay for the quanti®cation of the immediate early gene c-fos mRNA expression levels in retinal lesion cats, with cyclophilin A and glyceraldehyde-3-phosphate-dehydrogenase as candidates for internal experimental control to normalize mRNA levels between the different samples (Van der Gucht et al., 2003). We therefore devoted this study completely to the investigation of cyclophilin A. The gene fragment of clone P4P91 was ampli®ed between two P9 primers as indicated in Fig. 1 (black sequences). To obtain the fulllength coding sequence of cat cyclophilin A a 50 RACE reaction was performed. This resulted in a cDNA sequence of 624 nucleotides containing a continuous open reading frame of 492 bp. A 30 RACE reaction revealed an additional 114 nucleotides of the 30 noncoding region containing the consensus polyadenylation signal AATAAA at position 685±690, 11 nucleotides before the poly(A) tract (Fig. 1, underlined). The cyclophilin A cDNA encodes a protein of 164 amino acids. Comparison of the nucleotide sequence and deduced amino acid sequence with those of known mammalian cyclophilin con®rmed that this cDNA actually corresponds to cat cyclophilin A. Alignment of the cat, human, monkey, rat and mouse cyclophilin A sequence,

Table 1. Summary of differentially expressed clones identi®ed by DDRT-PCR Gene

Identification

Higher mRNA expression in lesion-affected central cortex Known Glucose-6 phosphate dehydrogenase Cu/ZnSOD MEF2A EAAT-2 Similar to other ESTs

STS EST 127954

Higher mRNA expression in peripheral control cortex Known Neurofibromin SCIP Thymosin beta-4 Beta-adaptin RET II GST1-GTP binding protein Cyclophilin A

Accession no

Identity

Species

M24470 K00065 NM005587 U03505

72% (216/296) 74.5% (180/241) 79% (780/987) 98% (225/230)

Human Human Human Human

AC055731

69% (135/193)

Human

M89914 M72711 X16053 M34175 X15786 X17644 AF023860 AF023961

88% (152/173) 89% (100/112) 87% (141/161) 75% (169/226) 85% (207/241) 82.5% (80/97) 87.5% (311/355)

Human Rat Mouse Human Human Human Vervet monkey Macaque monkey



The identity obtained by nucleotide database search (numbers in parenthesis, number of identical nucleotides over total length of the cDNA sequence). Species, the species with which sequence similarity was highest; EST, expressed sequencing tag. ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

66 L. Arckens et al.

Fig. 2. Distribution of cyclophilin A mRNA in the visual cortex and the dLGN of normal cat. (A) In area 17 and 18 cyclophilin A mRNA is distributed homogeneously along the cortex. Cortical layers II to VI express cyclophilin A mRNA. Layers III and VI contain slightly higher cyclophilin A mRNA levels compared to the other layers. (B) Cyclophilin A mRNA is homogeneously distributed over lamina A and A1 of the dLGN. The C laminae express lower levels of cyclophilin A mRNA. CA, CA ®elds of hippocampus; WM. white matter; A, A1 and C, the layers of the dLGN; OT, optic tract. Scale bar, 2 mm.

respectively, showed 92.3%, 92.5%, 91% and 90.3% nucleic acid identity and 97.6%, 97.6%, 93.4%, 94% amino acid identity. To con®rm the differential expression observed for the cyclophilin A differential display fragment, a cat-speci®c oligonucleotide (Fig. 1, boxed sequence) was chosen to perform in situ hybridization experiments on brain sections of retinal lesion cats and normal control animals. In normal subjects, cyclophilin A mRNA is widely distributed throughout the brain. Investigation of visual thalamic and cortical structures revealed a rather homogenous distribution of cyclophilin A mRNA over the dLGN and area 17 and 18 of control animals (Fig. 2). Laminar differences were small. Cortical layers III and VI of area 17 and 18 contained slightly higher levels of cyclophilin A mRNA (Fig. 2A). In the dLGN, both A laminae clearly displayed higher cyclophilin A mRNA levels compared to the C laminae (Fig. 2B). Retinal lesions induced decreased cyclophilin A mRNA expression in distinct portions of area 17 and 18 (Fig. 3). This lesion-induced reduction was minimal 3 days postlesion but became more apparent after longer survival times although the difference remained small (Fig. 3A and C). Comparison of the position and size of the area 17 and 18 regions with decreased cyclophilin A mRNA levels with the retinotopic maps of Rosenquist (1985) indicated that they corresponded to the lesions. This was also veri®ed by comparing adjacent sections hybridized for cyclophilin A and zif268 (Fig. 3A±D). We have shown previously that in situ detection of the inducible immediate early gene zif268 is a reliable tool to outline the lesion-affected portions of visual areas following partial sensory deprivation (Arckens et al., 2000a). Different zif268 expression levels re¯ect different levels of neuronal activity. As shown in Fig. 3, the portions of area 17 and 18 with low cyclophilin A mRNA levels (Fig. 3A and C) matched those of low zif268 mRNA expression (Fig. 3B and D). These observations demonstrate that cyclophilin A mRNA expression was decreased in those cortical regions that represent the lesioned retina. Analysis of emulsion coated slides revealed that cyclophilin A mRNA was mainly expressed by neurons. The silver grains were found predominantly over neurons while the signal over glial cells and neuropil was low (Fig. 4A and B). At the cellular level, comparison of the cyclophilin A mRNA levels in the lesion-affected portion of area

17 and its normal counterpart again revealed decreased cyclophilin A mRNA expression within the lesion projection zone. The number of silver grains per cell was clearly lower in the central lesion-affected portion (Figs 4C and D, and 5B) than in the peripheral normal portion of area 17 (Figs 4A and B, and 5A). This effect was observed in supragranular and infragranular layers and was most pronounced in layer VI (Figs 4 and 5). Competitive PCR was used to con®rm quantitatively the ®ndings obtained by differential mRNA display and in situ hybridization analyses and to estimate the extent of the observed decrease in cyclophilin A mRNA expression. Quanti®cation of competitive PCR products of cyclophilin A 2-weeks postlesion revealed a 63% decrease of the cyclophilin A mRNA level in lesion-affected area 17 compared to peripheral cortical counterparts (Figs 6 and 7A; P < 0.05). Immunocytochemistry and Western blotting for cyclophilin A extended our ®ndings to the protein level. We ®rst determined the speci®city of the polyclonal rabbit cyclophilin A antibody by immunoblot analysis. Cyclophilin A is known to be a soluble, cytoplasmic protein. We therefore prepared a soluble brain protein fraction by homogenizing brain tissue punches, pelleting the membranes by centrifugation and using the supernatant representing the cytosolic and soluble secreted protein fraction. After the electrophoretic separation of the proteins and the transfer to nitrocellulose membrane the cyclophilin A antibody revealed a single strong band of the appropriate molecular weight of approximately 17 kDa (Fig. 8, GoÈldner & Patrick, 1996). The antibody did not cross-react with any other isoforms of cyclophilin such as cyclophilin B (25 kDa, Price et al., 1991) or cyclophilin C (22±25 kDa, Schneider et al., 1994). The antibody used in this study therefore speci®cally recognized cyclophilin A in cat brain. We then compared cyclophilin A protein levels in the lesion-affected and normal cortex of retinal lesion animals by analysing dilution series of the protein extracts for cyclophilin A content. As shown in Fig. 8A and B, cyclophilin A was hardly detectable in a 0.5-mg protein sample in lesion-affected visual cortex (Fig. 8B), while in the peripheral normal cortex of the same animal cyclophilin A was clearly detectable under the same conditions (Fig. 8A). Quantitative analysis of the Western blots revealed signi®cantly lower cyclophilin A protein levels in lesion-affected area 17 compared to normal control counterparts (Fig. 7B; P < 0.05). The

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity

67

Fig. 3. Distribution of cyclophilin A and zif268 mRNA in area 17 of retinal lesion cat. (A and C) In retinal lesion cat we can distinguish two regions in area 17 based on the expression of cyclophilin A. The upper portion of area 17 shows low cyclophilin A mRNA levels while the lower portion of area 17 contains higher cyclophilin A mRNA levels, especially in layer VI. The difference in intensity between the two cortical portions is small even after long postlesion survival times (A, 4.5 months; C, 12 months). Comparison of adjacent sections hybridized for cyclophilin A (A and C) and zif268 (B and D) reveals that the region of low cyclophilin A mRNA expression corresponds to the lesion-affected portion of area 17 characterized by low zif268 mRNA expression. Arrows indicate the border between the upper, lesion-affected portion of area 17 and the lower normal portion of area 17, characterized by low and high mRNA expression for zif268 and cyclophilin A. The boxes in panel B indicate where tissue punches were taken for RNA and protein extraction. Note that peripheral area 17 samples are taken several millimetres away from the border (arrow) of the lesion-affected area 17 portion. 17, area 17; 18, area 18; 19, area 19; 20a, area 20a. Scale bar, 2 mm.

Western blot data therefore argue that the cyclophilin A protein levels are decreased similarly to the cyclophilin A mRNA levels after retinal lesioning. We used immunocytochemistry for cyclophilin A to determine the distribution and the cell type immunoreactive for cyclophilin A in cat striate cortex (Fig. 9). In normal animals, cyclophilin A-immunoreactive neurons were present in all cortical layers of area 17. Both the nucleus and the cytoplasm of neurons were stained. Most cyclophilin A-positive neurons were found in cortical layers II±III and V±VI. Layer IV contained only few cyclophilin A-immunoreactive neurons. We did not detect cyclophilin A-positive glial cells. Investigation of

area 17 of retinal lesion cats revealed an effect of the lesion on cyclophilin A immunoreactivity (Fig. 9A and B). Like in in situ hybridization, the effect was weak three days postlesion. Two to 12 weeks postlesion, a clear decrease in the number of cyclophilin Apositive cells was observed in the lesion projection zone of area 17 (Fig. 9B). The lesion-affected portion of area 17 displayed less cyclophilin A-positive cells than the normal portion in layer II±III (Fig. 9C and D), and even more pronounced in layer VI (Fig. 9E and F). Quantitative analysis con®rmed these qualitative observations (Fig. 7C). Whereas for normal subjects the central/peripheral ratio of the mean number of cyclophilin A-immunoreactive cells per square

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

68 L. Arckens et al.

Fig. 4. Cellular resolution of the cyclophilin A mRNA distribution in layer VI of the visual cortex of retinal lesion and isolated hemisphere cat. Paired dark-®eld and bright-®eld photographs of cyclophilin A mRNA expression in the unaffected control (A and B) and lesion-affected (C and D) portion of layer VI of area 17 of retinal lesion cat (1 month postlesion) show that the decreased cyclophilin A mRNA levels in the affected region of area 17 are caused by a decrease in the number of silver grains per cell, indicating the presence of less cyclophilin A mRNA in lesion-affected neurons of striate cortex. The silver grains are grouped over neurons (arrowhead). The signal on glial cells (arrow) and neuropil is much lower. Dark ®eld photographs of cyclophilin A mRNA in the left (E) and right (F) hemisphere of an isolated hemisphere cat (4.5 months postsurgery) illustrate equal cyclophilin A mRNA expression levels in both hemispheres, comparable to that in the unaffected control region of area 17 of retinal lesion cat (B). Scale bar, 20 mm. ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity

69

Fig. 5. Cellular resolution of the cyclophilin A mRNA distribution in layer II±III of the visual cortex of retinal lesion cat and isolated hemisphere cat. Dark-®eld photographs of cyclophilin A mRNA expression in the unaffected control (A) and lesion-affected (B) portion of layer II±III of area 17 of retinal lesion cat (1 month postlesion) show that the decreased cyclophilin A mRNA levels in the affected region of area 17 are caused by a decrease in the number of silver grains per cell, indicating the presence of less cyclophilin A mRNA in lesion-affected neurons of striate cortex. Dark-®eld photographs of cyclophilin A mRNA in the right (C) and left (D) hemisphere of an isolated hemisphere cat (4.5 months postsurgery) illustrate equal cyclophilin A mRNA expression levels in both hemispheres, comparable to that in the unaffected control region of area 17 of retinal lesion cat (A). Scale bar, 20 mm.

millimetre cortex equalled 1 in the supragranular and the infragranular layers of area 17 this was not the case for cats with retinal lesions. As illustrated in Fig. 7C, this ratio was clearly affected by the lesions in both supragranular and infragranular layers of area 17, with the strongest effect again in layer VI. To link the decreased expression of cyclophilin A mRNA and protein to cortical plasticity, it was necessary to certify that sensory deprivation per se is not enough to induce similar changes in cyclophilin A expression. To that purpose two control experiments were designed. A ®rst set of experiments involved the analysis of cyclophilin A mRNA and protein expression in the visual cortex of isolated hemisphere cats. Unilateral visual deafferentation by sectioning the left optic tract and the corpus callosum had no effect on cyclophilin A mRNA levels (Figs 4E and F, and 5C.D). In situ hybridization revealed similar cyclophilin A mRNA levels in the supragranular and infragranular layers of the left visually deprived hemisphere and the right control hemisphere, which keeps on receiving normal visual input, independent of the survival time postsurgery.

Competitive PCR also suggested the expression of equal amounts of cyclophilin A mRNA over the two hemispheres (Fig. 7A). Immunocytochemistry and Western blotting for cyclophilin A con®rmed the hybridization experiments. As with normal subjects, Western blotting revealed equal amounts of cyclophilin A in both hemispheres of isolated hemisphere cats (Figs 7B, and 8C and D). Cyclophilin A immunocytochemistry also revealed similar numbers of cyclophilin A-immunoreactive neurons in both hemispheres of isolated hemisphere cats for the supragranular and infragranular layers (Fig. 7C). In a second control experiment we analyzed the cyclophilin A mRNA levels in the dLGN of retinally lesioned and isolated hemisphere cats. If sensory deafferentation would be capable of inducing decreased cyclophilin A mRNA levels, we would expect to see a lesion-affected portion in the left and right dLGN of retinal lesion cats vs. a difference in signal intensity between the sensory-deprived (left) and nondeprived (right) dLGN of the unilaterally deafferented cats. As illustrated in Fig. 10, this was not the case. Independent of the deafferentation period we always observed a homogenous signal of

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

70 L. Arckens et al.

Fig. 6. Quanti®cation of cyclophilin A mRNA by competitive PCR in area 17 of retinal lesion cat. (A and B) Representative agarose gel pro®les and standard curves from peripheral, normal (A) and lesion-affected cortex (B) of a 14-day-retinal lesion cat. A serial ten-fold dilution of competitor DNA (lane 1±6: 10±10 4 attomoles/ mL) was coampli®ed with a constant amount of sample cDNA. Signal densities of bands corresponding to cyclophilin A target (231 bp) and competitor (303 bp) were determined by densitometry. The logarithm of the corrected ¯uorescence ratio (y-axis) was plotted against the logarithm of the initial mimic concentration (x-axis). A line was drawn from a linear regression analysis of the data points and the amount of target cyclophilin A was calculated by determining the x-intercept for the point of the curve where the ratio of target to mimic equals one (when y ˆ 0).

similar intensity over the entire left and right dLGN of all experimental animals.

Discussion Previous investigations comparing the cyclophilin A content of various tissues showed that cyclophilin A is present in mammalian brain in higher concentrations than in any other tissue and may perform critical neuronal functions (Handschumacher et al., 1984; Ryffel et al., 1991; GoÈldner & Patrick, 1996; Hovland et al., 1999). The present study revealed a consistent distribution of cyclophilin A mRNA and protein over the nucleus and the cytoplasm of neurons throughout the cat brain, in agreement with earlier studies showing a prominent neuronal distribution of cyclophilin A in the rat brain (Marks et al., 1991; Lad et al., 1991; Ryffel et al., 1991; Dawson et al., 1994; GoÈldner & Patrick, 1996). We never observed cyclophilin A-immunoreactive cells with processes radiating in all directions, a characteristic morphological feature of astrocytes. Furthermore, the absence of a notable staining in white matter makes a signi®cant presence of cyclophilin A in oligodendrocytes unlikely (GoÈldner & Patrick, 1996). Several studies investigating the possible effects of altered physiological conditions on gene expression in the brain refer to cyclophilin A as a housekeeping gene (Sahin et al., 1995). Cyclophilin A is therefore often used as an internal experimental control to normalize levels of individual mRNAs (Chen et al., 1999; Hager et al., 1999; Harrison et al., 2000; Bond et al., 2002). In a study investigating the effect of monocular deprivation on gene expression in the visual cortex of adult monkey, cyclophilin A mRNA was found not to be in¯uenced by the loss of neuronal activity due to monocular eye lid suture or TTX injection in contrast to other genes including Ca2‡/calmodulin-dependent protein kinase II alfa (Benson et al., 1991). These results are in agreement with our ®ndings in showing unaltered cyclophilin A

expression unless the loss of neuronal activity was accompanied by plastic phenemona. Indeed, our data revealed decreased mRNA and protein levels for cyclophilin A in cat visual cortex only during the reorganization of its topography in answer to restricted lesions of the retina. Unilateral deafferentation of the visual cortex by surgery could not induce such changes in cyclophilin A expression in the hemisphere with reduced neuronal activity. Taken together these results suggest an inhibitory role for cyclophilin A in cortical plasticity. The absence of a clear effect on cyclophilin A expression in the lateral geniculate nucleus would then suggest that thalamic reorganization is either nonexistent or at least of a much smaller impact. In line with these ®ndings, tracing studies in animals with retinal lesions could also not reveal a signi®cant participation of dLGN connections to visual cortical plasticity (Darian-Smith & Gilbert, 1995), whereas electrophysiological experiments revealed topographic map reorganization limited to no more than a 100±200 mm border region of the deafferented geniculate portion (Eysel, 1982). Yount & collaborators (1992) reported the up-regulation of cyclophilin A expression in the rat hippocampus after electrolytically induced seizures, while Porter et al. (1996) linked decreased cyclophilin A levels in the hippocampus to seizure activity induced by electroconvulsive shock. Together with the neuron-speci®c localization of cyclophilin A, these ®ndings challenge the proposed role of cyclophilin A as a housekeeping gene expressed at constant relative abundance in all cells and all tissues. Instead, our results link a decreased cyclophilin A expression to the capacity of adult cortex to reorganize dramatically in answer to the loss of sensory input. This down-regulation of cyclophilin A might modulate or gate other cellular processes underlying the functional and/or structural changes that occur during cortical plasticity. We have illustrated unambiguously that the retinal lesions provoke cyclophilin A changes in both supragranular and infragranular layers of sensory-deprived area 17. The observation that the effect is different

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity

71

Fig. 8. Western blot analysis of soluble cat brain protein from area 17 of retinal lesion cat (A and B) and isolated hemisphere cat (C and D) probed with cyclophilin A antibody. A dilution series of soluble cat area 17 protein (1, 0.5, 0.1 mg in lanes 1, 2 and 3, respectively) was subjected to 4±12% Bis Tris gelelectrophoresis, transferred to nitrocellulose membrane and probed with cyclophilin A antibody. A single cyclophilin A immunoreactive band of expected size of 17 kDa is detected, indicating the speci®city of the antibody for cat cyclophilin A. No cross-reaction for cyclophilin B (25 kDa) or cyclophilin C (22±25 kDa) is observed. In the unaffected control (A) portion of area 17 of a two-week-retinal lesion cat cyclophilin A is clearly detectable in lanes 1 and 2 and just visible in lane 3. In contrast, in lesion-affected (B) area 17 of the same animal, cyclophilin A is hardly detectable in lane 2 and completely absent in lane 3, indicating that the cyclophilin A protein levels in lesion-affected cortex are lower compared to normal visual cortex. The nondeprived (C) and visually deprived (D) hemisphere of isolated hemisphere cat (two weeks postsurgery) show identical cyclophilin A concentrations, comparable to the normal portion of area 17 of retinal lesion cat.

Fig. 7. Quantitative analysis of relative changes in cyclophilin A mRNA and protein expression levels by competitive PCR (A), Western blotting (B) and immunocytochemistry (C) in retinal lesion cat (RL) and isolated hemisphere cat (IH). Bar graphs represent the central/peripheral (C/P) area 17 ratio for retinal lesion cats and the left/right hemisphere area 17 ratio for isolated hemisphere cats, corresponding to a lesion-affected/control ratio in both animal models. (A) The C/P ratio for retinal lesion cats clearly illustrates the lower expression of cyclophilin A mRNA in central, lesion-affected area 17 compared to peripheral, control cortical counterparts (P < 0,05). In isolate hemisphere cats the left/right hemisphere ratio equals 1, indicating that a mere loss of neuronal activity is not enough to induce changes in cyclopilin A mRNA expression levels in visual cortex. (B). Western blotting revealed similar ®ndings at the protein level: the C/ P ratio for retinal lesion cats clearly illustrates the lower expression of cyclophilin A protein in central, lesion-affected area 17 compared to peripheral, control cortical counterparts (P < 0,05). In isolate hemisphere cats the left/right hemisphere ratio again equals 1. (C) Cell counts of cyclophilin A-immunoreactive neurons revealed a statistically signi®cant lower number of cells in supragranular (P < 0.05) and infragranular (P < 0.01) layers in lesion-affected area 17 but not in the left hemisphere of cats in which visual input was con®ned to the right hemisphere. The effect was more pronounced in layer VI, but clearly present also in cortical layer II±III.

between cortical layers in that it seems more profound in infragranular layers is in agreement with previous investigations for glutamate and the immediate early genes c-fos and zif268 (Arckens et al., 2000a,b). Also there we observed a more pronounced infragranular effect regarding the decreased expression of glutamate, c-fos and zif268 in sensory-deprived area 17 of retinal lesion cats. To what extend these cortical layers would contribute differently to cortical plasticity remains an open question. But evidence that cortical layers participate differently in brain plasticity already exists and layer-speci®c differences in lateral intracortical connections are generally held responsible (Gilbert et al., 1990; Diamond et al., 1994; Obata et al., 1999; Trachtenberg et al., 2000; Beaver et al., 2001). Although not investigated in full detail, the cyclophilin A mRNA expression seemed to remain different from the normal situation for up to one year. Likewise, Obata et al. (1999) have evidenced elevated expression, within the cortical scotoma, for nerve growth factors, BDNF, NT-3, synaptic vesicle associated proteins such as synapsin I, and CREB, MAP2 and CaMKII. Whereas some of these molecular events already occurred very rapidly (within 3 days), some came about later, but more strikingly, many of these molecular changes also remained for more than one year indicating that the fullest extent of the recovery takes many months. Cyclophilin A was originally identi®ed as a target protein involved in cyclosporin A-derived immunosuppresion (Handschumacher et al., 1984). Subsequently, two independent groups reported that cyclophilin A and peptidyl-prolyl cis-trans-isomerase were identical proteins (Fischer et al., 1989; Takahashi et al., 1989). Cyclophilin A is the ®rst identi®ed member of a family of cyclophilins. The ubiquitous expression and evolutionary conservation of this protein family suggest that cyclophilins have a fundamental role in cellular metabolism. They catalyse the cis-trans isomerization of a peptidyl-prolyl bond and

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

72 L. Arckens et al.

Fig. 9. Cellular distribution of cyclophilin A in area 17 of retinal lesion cat. In the peripheral normal visual cortex of a two-week-retinal lesion cat (A) cyclophilin Aimmunoreactive neurons are present in layers II to VI. Layer II±III and VI contain more cyclophilin A-positive neurons compared to layer IV. The lesion-affected visual cortex of a two-week retinal lesion cat (B) displays a similar distribution pattern but the number of cyclophilin A-immunoreactive neurons is clearly decreased in layer VI and also in layer II±III. Panels C, D and E, F show an enlargement of layers II±III and VI from normal and lesion-affected cortex, respectively. Cyclophilin A is always present in the cytoplasm and nucleus of neurons. No cyclophilin A immunoreactive glial cells are detected. The strong but diffuse immunostaining in layer I may represent an edge effect. Scale bar, 100 mm. ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity

73

Fig. 10. Distribution of cyclophilin A mRNA in the dLGN of retinal lesion cat and isolated hemisphere cat. Cyclophilin A mRNA is homogeneously distributed over the dLGN of retinal lesion cat (A) and isolated hemisphere cat (B). We do not detect a lesion-affected portion in the dLGN after retinal lesions (A). We never observe a difference in signal intensity for the sensory deprived dLGN after sectioning the left optic tract and the corpus callosum (B). Cyclophilin A mRNA levels always appear normal in the dLGN of all experimental animals. The arrow indicates the border between the deafferented and the normal portion of the dLGN as determined based on the retinotopic maps of Sanderson (1971) and the induced retinal lesion size (Arckens et al., 1998). Scale bar, 2 mm.

accelerate the folding of both nascent and denatured proteins such as the calcium regulating hormone calcitonin, carbonic anhydrase, actin and tubulin (Schmid et al., 1986; Freskgard et al., 1992). Cyclophilin A has also been reported as a requirement for the functional insertion of heterologously expressed neurotransmitter gated ion channels into the membrane of Xenopus oocytes (Helekar et al., 1994). Cyclophilin A has therefore been proposed to play a role in the maturation of neuron-speci®c membrane proteins (Shieh et al., 1989; Helekar et al., 1994). As a molecular chaperone, cyclophilin A might be involved in mechanisms of synaptic plasticity as shown for heat shock proteins. Hsp 70 and Hsc 70, both recognized to show molecular chaperone activity, are considered candidate plasticity genes based on their induced expression in hippocampus following injection with kainate, a treatment that induces seizures and LTP-like potentiations (Nedivi et al., 1993; Ohtsuka & Suzuki 2000). Cyclophilins and FK506-binding protein family (FKBPs) are generally referred to as immunophilins and are recognized for their immunosuppressive action when bound to cyclosporin A and FK506, respectively (Marks, 1996). Drug-immunophilin complexes inhibit the phosphatase activity of calcineurin. It is noteworthy that in the brain the distribution of cyclophilin A and FKBP12 closely resembles that of calcineurin, a Ca2‡/calmodulin-activated protein phosphatase (Dawson et al., 1994). By modifying the phosphorylation state of substrates like the NMDA and IP3 receptors, GAP-43, nNOS, synapsin I, dynamin I and CREB, calcineurin has been suggested to play a role in the dynamics of growth cones, neurite extension, neurotransmitter release, synaptic ef®cacy, long-term depression and potentiation (Nichols et al., 1994; Chang et al., 1995; Bito et al., 1996; Guerini, 1997; Hamilton & Steiner, 1998; Mansuy et al., 1998; Norris et al., 1998; Snyder et al., 1998). Although it is well established that ligand-immunophilin complexes bind and inhibit calcineurin, it is somewhat more controversial as to whether similar interactions occur in the absence of ligands. At least for FKBP12,

Cardenas et al. (1995) showed functional effects on calcineurin in vivo in the absence of ligand. Despite the lack of knowledge of endogenous immunophilin ligands, we can speculate that immunophilins, including cyclophilin A, can modulate or gate the above-mentioned neuronal and synaptic processes by acting on calcineurin. Moreover, just recently, small molecule ligands of cyclophilins have been evidenced to stimulate neurite growth in vitro (Steiner et al., 2002). Although initially molecular studies of gene expression during synaptic plasticity focused on the identi®cation of positive regulators, more recent work has established the additional requirement of the removal of inhibitory constraints. MHC I, known as a mediator of cell± cell interactions in the immune system, was recently discovered to be down-regulated in the developing geniculate nucleus following intracranial infusion of TTX using osmotic minipumps. MHC I has been suggested to play a role in structural and synaptic re-modelling in the developing and mature CNS (Corriveau et al., 1998; Huh et al., 2000). LTP, a classical model for synaptic plasticity, has also been shown to elicit decreases in the amount of speci®c proteins. LTP could either trigger the degradation of key proteins or activate factors that recognize negative transcription elements, which could decrease the expression of speci®c genes (Mayford et al., 1992; Silva & Giese, 1994; Abel et al., 1998). Likewise, we have shown that a reduction in cyclophilin A correlates with plastic rearrangements of adult mammalian neocortex and that this altered expression is not due to decreased neuronal activity. In conclusion, in the past few years, the use of large-scale screening techniques such as differential mRNA display, DNA chip and serial analysis of gene expression technology have led to the identi®cation of several proteins involved in different forms of brain plasticity. Nevertheless, many questions remain about the identity and the role of plasticity genes and their physiological effects. Only the identi®cation of more positive and negative mediators of adult brain plasticity can eventually lead to the elucidation of the underlying molecular cascade

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

74 L. Arckens et al. and to the development of new pharmaceuticals useful in the treatments for sensory loss and brain damage.

Acknowledgements We thank Ria Vanlaer and Heidi Allemeersch for expert technical assistance. This work was supported by grants of the Queen Elisabeth Medical Foundation, the Fund for Scienti®c Research ± Flanders, Belgium (FWO) and the Research Fund of the Katholieke Universiteit Leuven (OT 01/22), Belgium. Gert Van den Bergh and Ann Massie are research assistants and Lutgarde Arckens is a postdoctoral fellow of the FWO-Flanders (Belgium). Estel Van der Gucht is a postdoctoral fellow of the Research Council of the Katholieke Universiteit Leuven.

Abbreviations DDRT-PCR, differential mRNA display; dLGN, dorsal lateral geniculate nucleus; MMLV, Moloney murine leukaemia virus; PCR, polymerase chain reaction; RACE, rapid ampli®cation of cDNA ends.

References Abel, T., Martin, K.C., Bartsch, D. & Kandel, E.R. (1998) Memory suppressor genes: inhibitory constraints on the storage of long-term memory. Science, 279, 338±341. Arckens, L., Eysel, U.T., Vanderhaeghen, J.-J., Orban, G.A. & Vandesande, F. (1998) Effect of sensory deafferentation on the GABAergic circuitry of the adult cat visual system. Neuroscience, 83, 381±391. Arckens, L., Schweigart, G., Qu, Y., Wouters, G., Vandesande, F., Eysel, U.T. & Orban, G.A. (2000b) Cooperative changes in GABA, glutamate and activity levels: the missing link in cortical plasticity. Eur. J. Neurosci., 12, 4222±4232. Arckens, L., Van der Gucht, E., Eysel, U.T., Orban, G.A. & Vandesande, F. (2000a) Investigation of cortical reorganization in area 17 and nine extrastriate visual areas through the detection of changes in immediate early gene expression as induced by retinal lesions. J. Comp. Neurol., 425, 531±544. Arckens, L., Zhang, F., Vanduffel, W., Mailleux, P., Vanderhaeghen, J.-J., Orban, G.A. & Vandesande, F. (1995) Localization of the two protein kinase Cb-mRNA subtypes in cat visual system. J. Chem. Neuroanat., 8, 117±124. Bear, M.F., Cooper, L.N. & Ebner, F.F. (1987) A physiological basis for a theory of synapse modi®cation. Science, 237, 42±48. Beaver, C.J., Ji, Q. & Daw, N.W. (2001) Layer differences in the effect of monocular vision in light- and dark-reared kittens. Vis. Neurosci., 18, 811±820. Benson, D.L., Isackson, P.J. & Jones, E.G. (1991) Differential effects of monocular deprivation on glutamic acid decarboxylase and type II calcium-dependent protein kinase gene expression in the adult monkey cortex. J. Neurosci., 11, 31±47. Bito, H., Deisseroth, K. & Tsien, R.W. (1996) CREB phosphorylation and dephosphorylation: a Ca2‡- and stimulus duration-dependent switch for hippocampal gene expression. Cell, 87, 1203±1214. Bond, B.C., Virley, D.J., Cairns, N.J., Hunter, A.J., Moore, G.B., Moss, S.J., Mudge, A.W., Walsh, F.S., Jazin, E. & Preece, P. (2002) The quanti®cation of gene expression in animal model of brain ischemia using TaqMan real-time PCR. Brain Res. Mol. Brain Res., 106, 101±116. Cardenas, M.E., Hemenway, E., Muir, R.S., Ye, R., Fiorentino, D. & Heitman, J. (1995) Immunophilins interact with calcineurin in the absence of exogenous immunosuppressive ligands. EMBO J., 13, 5944±5957. Chang, H.Y., Takei, K., Sydor, A.M., Born, T., Rusnak, F. & Jay, D.G. (1995) Asymmetric retraction of growth cone ®lopodia following inactivation of calcineurin. Nature, 376, 686±690. Chen, L., Segal, D.M. & Mash, D.C. (1999) Semi-quantitative reversetranscriptase polymerase chain reaction: an approach for the measurement of target gene expression in human brain. Brain Res. Brain Res. Protoc., 4, 132±139. Chino, Y.M. (1995) Adult plasticity in the visual system. Can. J. Physiol. Pharmacol., 73, 1323±1338. Chino, Y.M., Smith, E.L., Kaas, J.H., Sasaki, Y. & Cheng, H. (1995) Receptive®eld properties of deafferentated visual cortical neurons after topographic map reorganization in adult cats. J. Neurosci., 15, 2417±2433. Conti, F., Minelli, A. & Pons, T.P. (1996) Changes in glutamate immunoreactivity in the somatic sensory cortex of adult monkeys induced by nerve cuts. J. Comp. Neurol., 368, 503±515.

Corriveau, R.A., Huh, G.S. & Shatz, C.J. (1998) Regulation of class I MHC gene expression in the developing and mature CNS by neuronal activity. Neuron, 21, 505±520. Darian-Smith, C. & Gilbert, C.D. (1994) Axonal sprouting accompanies functional reorganization in adult cat striate cortex. Nature, 368, 737±740. Darian-Smith, C. & Gilbert, C.D. (1995) Topographical reorganization in the striate cortex of the adult cat and monkey is cortically mediated. J. Neurosci., 15, 1631±1647. Das, A. & Gilbert, C.D. (1995) Long-range horizontal connections and their role in cortical reorganization revealed by optical recording of cat primary visual cortex. Nature, 375, 780±784. Dawson, T.W., Steiner, J.P., Lyons, W.E., Fotuhi, M., Blue, M. & Snyder, S.H. (1994) The immunophilins, FK506 binding protein and cyclophilin, are discretely localized in the brain: relationship to calcineurin. Neuroscience, 2, 569±580. Diamond, M.E., Huang, W. & Ebner, F.F. (1994) Laminar comparison of somatosensory cortical plasticity. Science, 265, 1885±1888. Donoghue, J.P. (1995) Plasticity of adult sensorimotor representations. Curr. Opin. Neurobiol., 5, 749±754. Eysel, U.T. (1982) Functional reconnections without new axonal growth in a partially denervated visual relay nucleus. Nature, 299, 442±444. Eysel, U.T., Gonzalez-Aguilar, F. & Mayer, U. (1980) A functional reorganization in the visual system of adult cats: lateral geniculate neurons with displaced receptive ®elds after lesions of the nasal retina. Brain Res., 181, 285±300. Eysel, U.T., Gonzalez-Aguilar, F. & Mayer, U. (1981) Time-dependent decrease in the extent of visual deafferentation in the lateral geniculate nucleus of adult cats with small retinal lesions. Exp. Brain Res., 41, 256±263. Fischer, G., Wittmann-Liebhold, B., Lang, K., Kiefhaber, T. & Schmid, F.X. (1989) Cyclophilin and peptidyl- prolyl cis-trans isomerase are probably identical proteins. Nature, 337, 476±478. Frank, D.A. & Greenberg, M.E. (1994) CREB: a mediator of long-term memory from molluscs to mammals. Cell, 79, 5±8. Freskgard, P.-O., Bergenhem, N., Jonsson, B.-H., Svensson, M. & Carlsson, U. (1992) Isomerase and chaperone activity of prolyl isomerase in the folding of carbonic anhydrase. Science, 258, 466±468. Garraghty, P.E., LaChica, E.A. & Kaas, J.H. (1991) Injury-induced reorganization of somatosensory cortex is accompanied by reductions in GABA staining. Somatosens. Mot. Res., 8, 347±354. Gierdalski, M., Jablonska, B., Smith, A., Skanfgiel-Kramska, J. & Kossut, M. (1999) Deafferentation induced changes in GAD67 and GluR2 mRNA expression in mouse somatosenory cortex. Brain Res. Mol. Brain Res., 71, 111±119. Gilbert, C.D., Hirsch, J.A. & Wiesel, T.N. (1990) Lateral interactions in visual cortex. Cold Spring Harb. Symp Quant. Biol., 55, 663±677. Gilbert, C.D. & Wiesel, T.N. (1992) Receptive ®eld dynamics in adult primary visual cortex. Nature, 356, 150±152. GoÈldner, F.M. & Patrick, J.W. (1996) Neuronal localization of the cyclophylin A protein in the adult rat brain. J. Comp. Neurol., 372, 283±293. Guerini, D. (1997) Calcineurin: not just a simple protein phosphatase. Biochem. Biophys. Res. Commun., 235, 271±275. Hager, G., Eckert, E. & Qchwaiger, F.W. (1999) Semiquantitative analysis of low levels of mRNA expression from small amounts of brain tissue by nonradioactive reverse transcriptase-polymerase chain reaction. J. Neurosci. Meth., 89, 141±149. Hamilton, G.S. & Steiner, J.P. (1998) Immunophilins: beyond immunosuppression. J. Med. Chem., 41, 5119±5143. Handschumacher, R.E., Harding, M.W., Rice, J. & Drugge, R.J. (1984) Cyclophilin: a speci®c cytosolic binding protein for cyclosporin A. Science, 226, 544±547. Harrison, D.C., Medhurst, A.D., Bond, B.C., Campbell, C.A., Davis, R.P. & Philpott, K.L. (2000) The use of quantitative RT-PCR to measure mRNA expression in a rat model of focal ischemia ± caspace-3 as a case study. Brain Res. Mol. Brain Res., 75, 143±149. Helekar, S.A., Neff, S. & Patrick, J. (1994) Prolyl isomerase requirement for the expression of functional homo-oligomeric ligand-gated ion channels. Neuron, 12, 179±189. Hovland, A.R., La Rosa, F.G., Hovland, P.G., Cole, W.C., Kumar, A., Prasad, J.E. & Prasad, K.N. (1999) Cycosporin A regulates the levels of cyclophilin A in neuroblastoma cells in culture. Neurochem. Int., 35, 229±235. Huh, G.S., Du Boulanger, L.M.H., Riquelme, P.A., Brotz, T.M. & Shatz, C.J. (2000) Functional requirement for class I MHC in CNS development and plasticity. Science, 290, 2155±2159. Kaas, J.H., Krubitzer, L.A., Chino, Y.M., Langston, A.L., Polley, E.H. & Blair, N. (1990) Reorganization of retinotopic cortical maps in adult mammals after lesions of the retina. Science, 248, 229±231.

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75

Cyclophilin A and adult brain plasticity Kilgard, M.P. & Merzenich, M.M. (1998) Cortical map reorganization enabled by nucleus basalis acticity. Science, 279, 1714±1718. Lad, R.P., Smith, M.A. & Hilt, D.C. (1991) Molecular cloning and regional distribution of cyclophilin A. Brain Res. Mol. Brain Res., 9, 239±244. Leuba, G., Kraftsik, R. & Saini, K. (1998) Quantitative distribution of parvalbumin, calretinin, and calbindin D-28k immunoreactive neurons in the visual cortex of normal and Alzheimer cases. Exp. Neurol., 152, 278±291. Liang, P. & Pardee, A.B. (1992) Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science, 257, 967±970. Lund, J.S., Henry, G.H., MacQueen, C.L. & Harvey, A.R. (1979) Anatomical organization of the primary visual cortex (area 17) of the cat. A comparison with area 17 of the macaque monkey. J. Comp. Neurol., 184, 599±618. Mansuy, I.M., Mayford, M., Jacob, B., Kandel, E.R. & Bach, M.E. (1998) Restricted and regulated overexpression reveals calcineurin as a key component in the transition from short-term to longterm memory. Cell, 92, 39±49. Marks, A.R. (1996) Cellular functions of immunophilins. Physiol. Rev., 76, 631±649. Marks, W.H., Harding, M.W., Handschumacher, R., Marks, C. & Lorber, M.I. (1991) The immunocytochemical distribution of cyclophilin in normal mammalian tissues. Transplantation, 52, 340±345. Mayford, M., Barzilai, A., Keller, F., Schacher, S. & Kandel, E.R. (1992) Modulation of an NCAM-related adhesion molecule with long-term synaptic plasticity in Aplysia. Science, 256, 638±644. Merzenich, M.M., Kaas, J.H., Wall, J.T., Sur, M., Nelson, R.J. & Felleman, D. (1983) Topographic reorganization of somatosensory cortical areas 3b and 1 in adult monkeys following restricted deafferentation. Neuroscience, 8, 33±55. Nedivi, E., Hevroni, D., Israeli, D. & Citri, Y. (1993) Numerous candidate plasticity-related genes revealed by differential cDNA cloning. Nature, 363, 718±722. Nguyen, P.V., Abel, T. & Kandel, E.R. (1994) Requirement of a critical period of transcription for induction of a late phase of LTP. Science, 265, 1104± 1107. Nichols, R.A., Suplick, G.R. & Brown, J.M. (1994) Calcineurin-mediated protein dephosphorylation in brain nerve terminals regulates the release of glutamate. J. Biol. Chem., 269, 23817±23823. Norris, C.M., Halpain, S. & Foster, T.C. (1998) Alterations in the balance of protein kinase/phosphatase activities parallel reduced synaptic strength during aging. J. Neurophysiol., 80, 1567±1570. O'Leary, J.L. (1941) Structure of the area striata of the cat. J. Comp. Neurol., 75, 131±164. Obata, S., Obata, J., Das, A. & Gilbert, C.D. (1999) Molecular correlates of topographic reorganization in primary visual cortex following retinal lesions. Cereb. Cortex, 9, 238±248. Ohtsuka, K. & Suzuki, T. (2000) Roles of molecular chaperones in the nervous system. Brain Res. Bull., 53, 141±146. Porter, R.H.P., Burnet, P.W.J., Eastwood, S.L. & Harrison, P.J. (1996) Contrasting effects of electroconvulsive shock on mRNAs encoding the high af®nity kainate receptor subunits (KA1 and KA2) and cyclophilin in the rat. Brain Res., 710, 97±102. Price, E.R., Zydowski, L.D., Baker, C.H., McKeon, F.D. & Walsh, C.T. (1991) Human cyclophilin B: a second cyclophilin gene encodes a peptidyl-prolyl isomerase with a signal sequence. Proc. Natl Acad. Sci. USA, 88, 1903±1907. Qu, Y., Moons, L. & Vandesande, F. (1997) Determination of serotonin, catecholamines and their metabolites by direct injection of supernatants from chicken brain tissue homogenate using liquid chromatography with electrochemical detection. J. Chromatogr. B. Biomed. Sci. Appl., 704, 351±358. Robertson, D. & Irvine, D.R.F. (1989) Plasticity of frequency organization in auditory cortex of guinea pigs with partial unilateral deafness. J. Comp. Neurol., 282, 456±471. Rosenquist, A.C. (1985) Connections of Visual Cortical Areas in the Cat. In Peters, A., & &.Jones, E.G., (Eds), Cerebral cortex Vol. 3. Plenum, New York, pp. 81±117. Rosier, A., Arckens, L., Demeulemeester, H., Orban, G.A., Eysel, U.T., Wu, Y.J. & Vandesande, F. (1995) Effect of sensory deafferentation on immunoreactivity of GABAergic cells and on GABA receptors in the adult cat visual cortex. J. Comp. Neurol., 359, 476±489.

75

Ryffel, B., Woerly, G., Greiner, B., Haendler, B., Mitasch, M.J. & Foxwell, B.M. (1991) Distribution of the cyclosporin binding protein cyclophilin in human tissues. Immunology, 72, 399±404. Sahin, M., Dowling, J.J. & Hock®eld, S. (1995) Seven protein tyrosine phosphatases are differentially expressed in the developing rat brain. J. Comp. Neurol., 351, 617±631. Sanderson, K.J. (1971) The projection of the visual ®eld to the lateral geniculate nucleus and medial interlaminar nuclei in the cat. J. Comp. Neurol., 143, 101±118. Schmid, F.X., Gra¯, R., Wrba, A. & Beintema, J.J. (1986) Role of proline peptide bond isomerization in unfolding and refolding of ribonuclease. Proc. Natl Acad. Sci. USA, 83, 872±876. Schneider, H., Charara, N., Schmitz, R., Wehrli, S., Mikol, V., Zurini, M.G., Quesniaux, V. & Movva, N.R. (1994) Human cyclophilin C: primary structure, distribution, and determination of binding speci®city for cyclosporins. Biochemistry, 33, 8218±8224. Sherman, S.M. (1973) Visual ®eld defects in monocularly and binocularly deprived cats. Brain Res., 49, 25±45. Shieh, B.-W., Stamnes, M.A., Seavello, S., Harris, G.L. & Zuker, C.S. (1989) The ninaA gene required for visual transduction in Drosophila encodes a homologue of cyclosporin A-binding protein. Nature, 338, 67±70. Shu, S., Ju, G. & Fan, L. (1988) The glucose oxidase-DAB-nickel method in peroxidase histochemistry of the nervous system. Neurosci. Lett., 84, 169±171. Silva, A.J. & Giese, K.P. (1994) Plastic genes are in!. Curr. Opin. Neurobiol., 4, 413±420. Snyder, S.H., Sabatini, D.M., Lai, M.M., Steiner, J.P., Hamilton, G.S. & Suzdak, P.D. (1998) Neural actions of immunophilin ligands. Trends Pharmacol. Sci., 19, 21±26. Sprague, J.M. & Meikle, T.H. (1965) The role of the superior colliculus in visually guided behaviour. Exp. Neurol., 11, 115±146. Steiner, J.P., Conolly, M.A., Moeller, T.A., Wu, Y. & Hamilton, G.S. (2002) Small molecule cyclophilin ligands stimulate neurite outgrowth in sensory ganglion cultures and protect cells from excitotoxic cell death in neuronal cultures. Soc. Neurosci. Abstr., 28, 628.12. Takahashi, N., Hayano, T. & Suzuki, M. (1989) Peptidyl-prolyl cis-trans isomerase is the cyclosporin A-binding protein cyclophilin. Nature, 337, 473±475. Trachtenberg, J.T., Trepel, C. & Stryker, M.P. (2000) Rapid extragranular plasticity in the absence of thalamocortical plasticity in the developing visual cortex. Science, 287, 2029±2032. Tully, T. (1997) Regulation of gene expression and its role in long-term memory and synaptic plasticity. Proc. Natl Acad. Sci. USA, 94, 4239±4241. Van der Gucht, E., Massie, A., De Klerck, B., Peeters, K., Winters, K., Gerets, H.H.J., Clerens, S., Vandesande, F. & Arckens, L. (2003) Molecular cloning and differential expression of the immediate early gene c-fos. Brain Res. Mol. Brain Res., 111, 198±210. Vandenbussche, E., Eysel, U.T. & Orban, G.A. (1990) In¯uence of retinal lesions on grating acuity of the cat. Neurosci. Lett., 110, 249±255. Vanduffel, W., Vandenbussche, E., Singer, W. & Orban, G.A. (1995) Metabolic mapping of the visual areas in he behaving cat: a quantitative [14C] 2deoxyglucose study. J. Comp. Neurol., 352, 1±20. Wall, P.D. (1977) The presence of ineffective synapses and the circumstances which unmask them. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 278, 361±372. Wisden, W., Errington, M.L., Williams, S., Dunnett, S.B., Waters, C., Hitchcock, D., Evan, G., Bliss, T.V.P. & Hunt, S.P. (1990) Differential expression of immediate early genes in the hippocampus and the spinal cord. Neuron, 4, 603±614. Yount, G.L., Gall, C.M. & White, J.D. (1992) Limbic seizures increase cyclophilin mRNA levels in rat hippocampus. Brain Res. Mol. Brain Res., 14, 139±142. Zhang, F., Halleux, P., Arckens, L., Vanduffel, W., Van BreÂe, L., Mailleux, P., Vandesande, F., Orban, G.A. & Vanderhaeghen, J.-J. (1994) Distribution of immediate early gene zif-268, c-fos, c-jun and jun-D mRNA in the adult cat with special reference to brain region related to vision. Neurosci. Lett., 176, 137±141. Zhang, F., Vanduffel, W., Shiffmann, S.N., Mailleux, P., Arckens, L., Vandesande, F., Orban, G.A. & Vanderhaeghen, J.-J. (1995) Decrease of zif-268 and c-fos and increase of c-jun mRNA in the cat area 17, 18 and 19 following complete visual deafferentation. Eur. J. Neurosci., 7, 1292±1296.

ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 61±75