Medulla oblongata transcriptome changes during presymptomatic natural scrapie and their association with prion-related lesions

Background The pathogenesis of natural scrapie and other prion diseases is still poorly understood. Determining the variations in the transcriptome in the early phases of the disease might clarify some of the molecular mechanisms of the prion-induced pathology and allow for the development of new biomarkers for diagnosis and therapy. This study is the first to focus on the identification of genes regulated during the preclinical phases of natural scrapie in the ovine medulla oblongata (MO) and the association of these genes with prion deposition, astrocytosis and spongiosis. Results A custom microarray platform revealed that 86 significant probes had expression changes greater than 2-fold. From these probes, we identified 32 genes with known function; the highest number of regulated genes was included in the phosphoprotein-encoding group. Genes encoding extracellular marker proteins and those involved in the immune response and apoptosis were also differentially expressed. In addition, we investigated the relationship between the gene expression profiles and the appearance of the main scrapie-associated brain lesions. Quantitative Real-time PCR was used to validate the expression of some of the regulated genes, thus showing the reliability of the microarray hybridization technology. Conclusions Genes involved in protein and metal binding and oxidoreductase activity were associated with prion deposition. The expression of glial fibrillary acidic protein (GFAP) was associated with changes in the expression of genes encoding proteins with oxidoreductase and phosphatase activity, and the expression of spongiosis was related to genes encoding extracellular matrix components or transmembrane transporters. This is the first genome-wide expression study performed in naturally infected sheep with preclinical scrapie. As in previous studies, our findings confirm the close relationship between scrapie and other neurodegenerative diseases.


Background
Scrapie is a prion-associated encephalopathy that occurs naturally in sheep and goats. It is characterized by the accumulation of a pathological agent, the prion protein (PrP Sc ), mainly in the central nervous system [1]. PrP Sc differs from the endogenous normal form (PrP c ) in its conformation, partial resistance to proteolytic degradation and insolubility in the presence of detergents [2,3].
The incubation period of the disease is long and asymptomatic. PrP Sc can be detected in VRQ/VRQ sheep, genotype for the PRNP gene, two months after infection [7]. Three to six months after infection, the pathological agent is detected in the lymphoid formations associated with the gastrointestinal tract [8,9]. From six to nine months, the secondary lymphoid organs are also infected, and finally, at the tenth month after infection, the central nervous system is affected [10][11][12].
The neuropathological events in prion diseases occur at different times depending on the disease. High levels of PrP Sc exist without clinical disease in Gerstmann-Sträussler syndrome [13]; conversely, PrP Sc is present in very low levels in fatal familiar insomnia [14]. The degree of prion accumulation in specific brain regions does not correlate with the clinical features (reviewed in [15]). In addition to prion deposition, other molecular mechanisms act early during the disease. For example, the brain undergoes oxidative stress in the early stages of prion invasion into the brain and may predispose the brain to neurodegenerative mechanisms [16]. Genomic analysis confirmed the induction of cellular stress (oxidative stress and ER stress) and the activation of other molecular pathways in a murine model of prion disease [17]. Other functional genomic studies performed in animal models of scrapie infection have indicated that several genes are misregulated in the early phases of the infection [18][19][20][21][22][23][24].
To date, very few genomic approaches have focused on the analysis of the early molecular events in prion diseases and, to a lesser extent, studies dealing with the natural disease. The identification of the genes involved in the preclinical changes of the disease can help in the discovery of new biomarkers and targets for future diagnosis tests or treatments. In an earlier published work, we presented the differentially expressed genes in the brains of scrapie-symptomatic sheep and the relationship between scrapie-related neuropathological changes and the transcriptional activities of the identified genes [25]. The objectives of the present study were to identify the genes that are differentially expressed during natural preclinical scrapie infection in sheep using a CVI custom designed 4x44K ovine microarray and to determine the relationship between their expression patterns and prion-associated lesions. In this way, we discuss the variation in gene expression and its association with scrapie neuropathology during the progression of the disease.

Ethics statement
This study was performed in strict accordance with the recommendations for the care and use of experimental animals of the University of Zaragoza, in accordance with the law (R.D. 1201/2005). The protocol was approved by the Committee on the Ethics of Animal Experiments (Permit Number: PI02/08).

Animals
A total of 10 Rasa Aragonesa female sheep aged 1-5 years were included in this study. Six of them were selected from flocks located in areas free of scrapie and were used as controls.
Four of the animals exhibited preclinical signs of scrapie, and the diagnoses were made by third eyelid biopsies [26] and confirmed using the rapid test (TeEsE, Bio-Rad) and immunohistochemistry to detect PrP Sc using the 6 H4 monoclonal antibody [27]. This characterization was performed considering the presence of the clinical signs associated with the disease as per previously reported criteria [26]. All of the animals belonged to flocks that had been previously characterized as scrapieaffected flocks and were located in different geographical areas. The animals were genotyped for PRNP polymorphisms via full Open Reading Frame sequencing as previously reported [28], and the sheep chosen for this study displayed the ARQ/ARQ genotype without other coding mutations outside the 136, 154 and 171 codons, which is the most susceptible genotype in this ovine breed [28]. The presence of the prion protein was confirmed by immunohistochemical methods and western blotting [27].

Tissue collection and RNA isolation
Animals were sacrificed by intravenous injection of sodium pentobarbital and exsanguination. Necropsy was performed immediately, and the physical examination of the scrapie-infected and control animals did not reveal any additional pathological signs. The samples were rapidly preserved and processed according to established guidelines regarding safety. The lesion pattern in scrapie is bilateral; therefore, one-half of the caudal medulla oblongata, including the obex, was snap-frozen in liquid nitrogen prior to long-term storage at −80°C until RNA extraction. The other half was formalin-fixed and paraffin-embedded for further histopathological analysis. Total RNA was isolated from a tissuemizer-disrupted medulla oblongata in duplicate using TRIzol W (Invitrogen AG) followed by a phenol and chloroform extraction and subjected to a purification step with the NucleoSpin W RNA clean-up kit RNAII (Macherey-Nagel GmbH & Co. KG). The quality of the total RNA was assessed using the RNA 6000 Nano Assay kit and the 2100 Bioanalyzer (Agilent Technologies). The RNA integrity number (RIN) index for each sample was estimated using the Agilent 2100 Expert software. The RIN provides a numerical assessment of the integrity of RNA that facilitates the standardization of the quality interpretation. Only high quality RNA samples with an RIN number equal to or higher than 7 were further processed for microarray analysis.

Histology and prion immunohistochemical detection
A histopathological study of the medulla oblongata at the level of the obex was performed in HE-stained slices (one from each individual control and each positive animal).
Immunohistochemical (IHC) studies were performed on adjacent sections. For every antibody, positive and negative controls (the omission of primary antibodies from the control and scrapie slides) were performed.
Detection of the prion proteins was performed following pretreatment as previously described [29]. Briefly, sections were pretreated with 98% formic acid, hydrated and then autoclaved to enhance antigen retrieval. To block endogenous peroxidase activity, the sections were incubated with blocking reagent (DAKO) for 10 min after proteinase K digestion (Roche, 4 g/ml). Next, the sections were incubated with the monoclonal primary antibody L42 (R-Biopharm, dilution 1:500) at RT for 30 min. Endogenous peroxidase blocking was used to process sections. The enzyme-conjugated polymer Envision (DAKO, 30 min) was used as the visualization system and DAB (DAKO, 10 min) as the chromogen. The sections were counterstained with hematoxylin.
Astrocytosis was evaluated based on glial fibrillary acidic protein (GFAP) immunostaining, as previously described [30,31]. Briefly, after heat-induced epitope retrieval pretreatment with citrate buffer (pH 6.0), the sections were incubated for 1 h at RT with a rabbit polyclonal anti-GFAP antibody (DAKO, dilution 1:400). The omission of the primary antibodies from the control and scrapie slides served as negative controls in the routine immunoreactions.
The preparations were examined with a Zeiss Axioskop 40 optical microscope (Carl Zeiss AG) and a 40 × −magnification objective lens (Carl Zeiss AG). The images were captured with a digital camera (AxioCam MRc5, Zeiss AG) that was coupled to the microscope and a computer and were analyzed using the ImageJ 1.4.3.67 image-analysis software package (Psion Image, NIH) to determine the areas occupied by PrP Sc deposition, astrocytosis and spongiosis. For the evaluation of the IHC and HE slides, captured images were opened in NIH Image/ImageJ using the area method to evaluate the indices of positivity. The total area occupied by brown markers (PrP and GFAP) or by white spaces (spongiosis) was estimated by setting a "threshold" using the thresholding tool for the selection of these areas, and the positive IHC/HE index for that image was calculated. Using the Student´s t test, significant differences between the control and scrapie groups were detected.

Custom sheep oligo-DNA microarray
The custom CVI 4x44K microarrays contained custom eArray-designed 60-mer probes on previously sequenced normalized and subtracted cDNA libraries of ovine Peyer's Patch, obex and tonsil, supplemented by the publicly available Ovis aries transcripts from the NCBI/EBI databases and by the Agilent O. aries transcript catalog. All of the arrays were printed using Sureprint technology (Agilent Technologies).

Preparation of the labeled cDNA and microarray hybridization
All of the procedures for the preparation of the labeled cRNA probes and subsequent Genechip hybridizations were performed according to the Agilent Technologies One-Color Microarray-Based Gene Expression Analysis guidelines. First, cDNA was synthesized using 1 μg total RNA as a template and the T7 Promoter Primer of Agilent One-Color RNA Spike-In (Quick Amp Kit, One-Color, Agilent Technologies). The cDNA was then transcribed and labeled using T7 RNA Polymerase and cyanine 3-CTP. Finally, the labeled cRNAs were cleaned up using Qiagen RNeasy mini spin columns.
The samples were then hybridized to custom CVI-Agilent 4x44K chips for 17 h at 65°C and 6 rpm. Following the manufacturer's protocol, the chips were then washed and incubated with wash buffers and scanned using the GenePix 4200AL Scanner (Axon Instruments) in conjunction with GenePix Pro 6.0 software.
The hybridizations of each sample were performed in duplicate, resulting in 8 microarrays for the preclinical scrapie animals and 12 for the negative control animals.

Microarray data analysis
The hybridization data were extracted with the Agilent Feature Extraction, version 9.5.3.1, image analysis application (Agilent Technologies) before processing with GeneSpring GX 10.0.2 (Agilent Technologies). Using the 75th percentile method intensity, the chip values were normalized, and the expression values were calculated. The global medulla oblongata gene expression profiles from the clinical scrapie-infected animals were compared to the negative controls, through a linear model that accounts for both technical (random animal effects) and biological replicates (disease effects). In addition, a multiple testing correction proposed by Benjamini-Hochberg was applied. Further, only genes with a P-value ≤ 0.05 for the difference between healthy and preclinical individuals and a 2-fold change (FC) as the lower limit were selected. Although 2-FC is used as cutoff value, according to our experience, we consider that the major conclusion of our work is not changing even if we would have used a different threshold, but always higher than 1.5. These genes were clustered by their Euclidean distance coefficient using the Permut-Matrix software [32]. A BLAST search of the GenBank database was performed to identify the genes that were similar to the differentially expressed probes. The molecular functions of the genes were classified according to Gene Ontology (GO) using an on-line functional annotation of DAVID Bioinformatics Resources 2008 [33,34] (NIAID/NIH, USA).

The relationship between neuropathology and gene expression
Using a Mixed Model Analysis under a Bayesian approach by the Gene Expression Analysis with Mixed Models (GEAMM) software [35], the relationship between neuropathological lesions and gene expression was analyzed. The statistical model assumed the following Bayesian likelihood of logarithm of gene expression data provided by the oligo-DNA microarray: , where a is the array effect, b is the vector regression slope associated with the numerical valuation of the neuropathological changes (c: prion deposition, spongiosis or astrocytosis), X is the incidence matrix that relates the array effects to the logarithm of gene expression data (y), and R is the matrix of residual (co) variances with probe-specific residual variance and null residual covariances. Prior distributions were assumed to be flat for a, b and R. A more detailed description of the statistical procedure was described by Casellas et al. (2008) [35].
The Bayesian analysis was performed using a Gibbs sampler approach [36] with a single chain of 500,000 iterations after discarding the first 50,000. The results with a posterior probability below 0.01 for a regression slope associated with a neuropathological lesion greater (or lower) than zero were selected.

Quantitative real-time PCR
Quantitative real-time PCR (qRT-PCR) was performed to confirm the expression of the 12 genes/sequences involved in the mechanisms related to neurodegenerative or reparation processes and/or had a high level of differential expression in the scrapie group compared to the controls in the oligo-DNA microarray expression analysis. Eight of these genes also displayed the highest significance in the Mixed Model Analysis. The PCR primer sequences used for the quantification of the genes encoding amyloid beta (A4) precursor (APP), aquaporin 4 (AQP4), calcineurin-like phosphoesterase domain-containing 1 (CPPED1), golgi golgin subfamily 4 (GOLGA4), maguk p55 subfamily member 7 (MPP7), nell2 (NELL2), CD3 gamma chain (CD3G), granulysin (GNLY), lysosomal protein transmembrane 4 beta (LAPTM4B) and serine/arginine-rich splicing factor 3 (SRSF3) and the two ovine scrapie related sequences (OSRS1) and (OSRS2) are shown in Table 1. RNA samples used for qRT-PCR were the same used for microarray experiments, the qRT-PCR assays were designed with Primer Express 2.0 software (Applied Biosystems) to select appropriate primer sequences from known sheep or bovine sequences. Whenever possible, the exon-exon border was included to prevent the amplification of genomic DNA in the PCR reaction. Complementary DNA (cDNA) was synthesized from 1 μg RNA using random hexamer primers with the Superscript First Standard Synthesis System for RT-PCR (Invitrogen). To confirm the elimination of any remaining DNA, reverse transcription with and without the enzyme was performed.
qRT-PCR was performed using SYBR W Green (PE Applied Biosystems) assays. PCR amplification was performed in an ABI-Prism fast 7500 Sequence Detection System (PE Applied Biosystems). All qRT-PCR reactions were run in triplicate in total reaction volumes of 10 μl with 10-20 ng of cDNA as the template and a 300 nM final primer concentration. Universal conditions were used with an initial 10 min activation and denaturation step at 95°C, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. The baseline and threshold for the Ct calculations were set automatically with the ABI-Prism 7500 software Version 2.0.1. The levels of gene expression were determined using the comparative Ct method.
To improve the normalization accuracy, the geometric mean of three housekeeping genes was used to calculate the normalization factor (NF), which was used to normalize the expression level of each gene in each sample [37]. The NF was calculated from the GAPDH, G6PDH and RPL32 expression data. These are the three most stable reference housekeeping genes in the sheep medulla oblongata, and they have been used as internal references for expression studies in scrapie [38]. The primers and PCR conditions for the amplification of these housekeeping genes have been described previously [38,39].
The quantitative results obtained from the qRT-PCR assays were expressed as the fold-change. Student's t test analyses were used to determine if the differences observed between the groups were statistically significant (P < 0.05).

Preclinical scrapie-related lesions
The neuropathological features of scrapie were evaluated in the medulla oblongata tissues of 6 control and 4 preclinical scrapie-infected sheep. Spongiosis, PrP Sc deposition and GFAP immunoreactivity were consistent with the features of classical scrapie [40]. PrP Sc deposition and spongiosis were only detected in the affected animals ( Figure 1). Particular medullary areas in the obex, such as the nucleus dorsal motor of the vagus, the spinal tract of the trigeminal nerve and the solitary tract nucleus, were severely affected in the infected group. Even with the high variability observed in the scrapie group, the differences between the groups were statistically significant (P < 0.01 and P < 0.05).
Identification of the genes in the medulla oblongata that are differentially expressed in natural scrapie A total of 86 probe sets displayed statistically significant differences between the control and scrapie preclinical groups that were equal to or greater than a 2-fold change. The genes from Ovis aries are relatively poorly annotated, but BLAST searches against publicly available databases allowed the identification of a set of 44 known genes from the complete set of 86 differentially expressed genes. The microarray data were deposited in the array express and are accessible through accession number E-MTAB-866. To determine the gene ontology (GO) categories of the deregulated genes in scrapie, we used DAVID Bioinformatics Resources 2008 [33,34] (NIAID/NIH, USA). Based on the GO analysis, 35 genes had known functions (Table 2), of which 3 were upregulated (5.7%) and 32 were downregulated (94.3%). The functional group with the highest number of regulated genes (18) was that of the phosphoprotein-encoding genes ( Table 2). In addition, downregulated genes were included in GO groups encoding for proteins located in the lumen of organelles or the extracellular matrix and involved in the immune response and apoptosis. After the clustering analysis, the animals were grouped according to their disease condition ( Figure 2).

Identification of neuropathology-related genes
We identified many genes with known functions whose expression was related to PrP Sc deposition (1,011), although few genes were related to astrocytosis (21) and spongiosis (66). The expression of the genes that displayed a high probability of regression with scrapie lesions changed less than 2-fold ( Figure 3). The gene ontology analysis revealed that genes associated with prion deposition encoded for proteins involved in protein and ion binding, oxidoreductase activity and transcription. Genes encoding for proteins involved in metal ion binding showed a positive association with both GFAP and spongiosis. In addition, genes encoding for proteins with oxidoreductase and phosphatase activity were associated with GFAP expression, and genes coding for extracellular matrix components or transmembrane transporters were associated with spongiosis. A list of known genes whose expression was highly correlated with PrP Sc deposition, GFAP expression and spongiosis is shown in Figure 3. Only genes with a high probability of a positive (or negative) slope of regression are presented (P < 3.5x10 -3 ).  Validation of gene expression profiling by quantitative RT-PCR To confirm the results of the microarray, we performed qRT-PCR using SYBR Green on a selected number of targets. For validation, we chose 4 genes from the expression study and 6 genes and 2 sequences from the association analysis. Eight of the genes/sequences were upregulated in the microarray (APP, AQP4, CPPED1, GOLGA4, MPP7, NELL2, OSRS1 and OSRS2) and four displayed downregulation (CD3G, GNLY, LAPTM4B and SRSF3). In most cases, the selection of genes was based on previous reports showing their associations with prionrelated and other neurodegenerative diseases [41][42][43][44] but was also due to their potential involvement in the mechanisms involved in neurodegeneration. The qRT-PCR analyses confirmed the microarray expression results (Figure 4). The differences between the control and scrapie groups were statistically significant for each of the 12 genes analyzed (P < 0.05).

Discussion
Transmissible spongiform encephalopathies, or prion diseases, are fatal neurodegenerative diseases with characteristic spongiform lesions, neuronal cell loss, astrocytosis and the accumulation of the pathological form of the prion protein [45]. The precise mechanisms regulating these processes remain unknown. Genomic approaches are a potential tool to understand the molecular basis of complex mechanisms; in addition, they allow the discovery of new disease biomarkers.
The analysis of gene expression profiling can elucidate the molecular basis of this pathology. Several studies have focused on genomic analyses of brain tissue from animal models of prion diseases, including CJD, scrapie and BSE [19][20][21][22][23][24]. However, there are fewer studies involving the mRNA profiles of "natural" human CJD [46], bovine BSE [47] or ovine scrapie [48]. We previously reported a genomic analysis performed in tissues obtained from sheep naturally infected with scrapie in terminal stages [25]. However, the molecular changes can occur before the onset of the disease; for example, PrP Sc accumulation occurs early in the disease in both the central nervous system [49] and peripheral tissues [50]. In mouse models, genomic expression profiles revealed the induction of oxidative and endoplasmic reticulum (ER) stress, activated ER and mitochondrial apoptosis pathways, and activated cholesterol biosynthesis in the central nervous system of preclinical mice [17]. We report here the first transcriptome study of the central nervous system (CNS) in sheep naturally infected with scrapie in preclinal stages that associated the variations in the expression profile with the features of scrapie neuropathology.

Differential gene expression in preclinical scrapie
Our microarray hybridization analysis identified 86 probes with changes in expression greater than 2-fold compared with the controls. Using the same platform, our previous study of sheep clinically affected with scrapie [25] revealed 350 differentially regulated probes, indicating that at the early stages of the disease, fewer genes are active or the expression changes are not high enough to be detected by microarray. From the 86 probes, 40 were also differentially regulated in clinical animals [25] and 46 were identified only in preclinical sheep (some of these probes are shown in Table 2).
In vitro studies have shown variations in the phosphoproteome of N2a cells after PrP Sc infection [51] and specific inflammatory profiles in microglia [52]. In addition, genes involved in defense, the immune response or encoding for secreted extracellular proteins are differentially regulated in murine models during prion infection [20]. In the clinical phase of classical scrapie, the genes involved in ion-binding and transport, nucleotide binding and the immune system are differentially expressed [25]. In accordance with the in vitro and murine models and with our previous results in classical scrapie at the late stages of the disease, the differentially regulated genes identified in this study encoded phospho-proteins, extracellular matrix organization proteins and immune response-related proteins. Therefore, these mechanisms seem to be involved in the neuropathology of scrapie from the early phases of the disease.
The microarray analysis was validated using qRT-PCR for 4 genes; their differential regulation was confirmed in all cases. The high adjustment between the fold changes obtained with the microarray and the qRT-PCR of the genes selected for confirmation reflects the high credibility of the microarray and the gene alignment

NSR
Shown are the Blast results of the clones with significant alterations in gene expression (P ≤ 0.05 and ≥ 2-fold change). FC of significant alteration during clinical stage is given in the last column (NSR: no significant regulation). The GO was determined using on-line functional annotation of DAVID Bioinformatics Resources 2008 (NIAID/NIH, USA). Only the genes with a known GO are shown. *Genes chosen for validation by quantitative RT-PCR. **FC values obtained in a previous study developed by our group comparing scrapie-infected animals in a clinical stage and controls [25].
analysis. Three of these genes were downregulated in scrapie tissues. Two of them, CD3G and GNLY, are involved in the immune response. CD3G is part of the T-cell receptor-CD3 complex, which plays an important role in coupling antigen recognition to several intracellular signal-transduction pathways [53]. The CD3 T cells accumulate near or around blood vessels and in the CNS parenchyma of mice inoculated with scrapie [54], suggesting the infiltration of T cells in the brain. In contrast, the downregulation of CD3G observed in our natural model suggests that these cells are decreased in preclinical scrapie. Similarly, we observed a significant decrease in the expression of GNLY, which is a potent antimicrobial protein contained within the granules of CTL and NK cells [55]. Taking together, our results suggest a decline of immune activity in prion diseases, as described for other neurodegenerative diseases such as Alzheimer's disease (AD) [56]. Another gene that was downregulated in the preclinical medullae encodes the lysosomal protein transmembrane 4 protein (LAPTM4). The endosomal and lysosomal compartments are implicated in trafficking, recycling and the final degradation of prions [57]. It has been proposed that autophagy might play a protective role in prion diseases, leading to the degradation of prions. Galectin-3 knockout mice express low levels of lysosomal activation marker (LAMP-2) and autophagy markers, suggesting that endosomal/lysosomal dysfunction in combination with reduced autophagy may contribute to the development of prion diseases [58]. The downregulation of LAPTM4 is in accordance with these results and might indicate a dysfunction of the lysosomal-endosomal pathway in preclinical scrapie.
One of the genes upregulated in the microarray hybridization analysis was Maguk p55 subfamily member 7 [MPP7]. The membrane-associated guanylate kinase homologues (MAGUKs) are a family of peripheral membrane proteins that form multiprotein complexes containing distinct sets of transmembrane, cytoskeletal, and cytoplasmic signaling proteins. MPP7 acts as an important adapter that promotes epithelial cell polarity, tight junction formation via its interaction with DLG1 and is involved in the assembly of protein complexes at sites of cell-cell contact [59]. The cellular prion protein PrP c is also located at cell-cell adhesion sites in polarized/differentiated enterocytes and interacts with desmosomal proteins and with actin and actin-binding proteins at cell-cell junctions [60]. Moreover, in the CNS, the PrP c is located in the microvascular endothelium and at intercellular junctions of cultured brain endothelial cells of mouse, rat and human origin [61]. We report here for the first time the upregulation of the gene encoding the MPP7 protein in preclinical scrapie and its positive association with PrP Sc deposition, suggesting a possible alteration of cell-cell adhesion the early stages of the disease.

The genomic association with scrapie-related lesions
Studies of the associations between gene expression and the intensity of scrapie lesions have been shown to be a powerful tool to detect genes potentially involved in the development of these lesions [25]. In the present study, we found a relatively low number of genes with differential expression in the preclinical tissues; however, the association study allowed the identification of genes that had slight changes (FC < 2) in the microarray hybridization analysis but whose expression was strongly related to the development of the scrapie-related lesions. The genes that displayed a positive regression with prion deposition were involved in several cellular mechanisms, the most frequent of which being protein, metal and ion binding, oxidoreductase activity and transcription factors. The possible role of PrP c to protect cells from oxidative stress is well documented [62], as is the capacity of the prion protein to bind Cu 2+ [63,64]. In addition, many other roles have been attributed to the prion protein, such as transmembrane signaling or cell adhesion [65]. Genes involved in these mechanisms have been shown to be associated with prion diseases, astrocytosis or spongiosis, thereby corroborating the reliability of our association study.
The expression of 6 genes and 2 sequences was validated using qRT-PCR. These genes were chosen because of their known role in brain metabolism and/or neurodegeneration, and the sequences displayed the most significant degrees of association with scrapie lesions. We found a slight up-regulation of the amyloid beta (A4) precursor protein (APP). APP and PrP are both cellsurface proteins residing in cholesterol-rich lipid rafts of the cell membrane and play an important role in the development of AD and prion diseases, respectively. These diseases share a number of clinical, pathological and biochemical characteristics [66]. The brains of CJD patients display the pathology of both prion diseases and AD [67]. The overexpression of APP in other genetic expression profiling studies in scrapie murine models has been previously reported [24]. In addition, the overexpression of APP facilitates the rapid development of artificial (See figure on previous page.) Figure 2 Condition trees of the clustering analysis. The hierarchical cluster analysis (Euclidean distance clustering algorithm) was performed using PermutMatrix [32], and 86 clones/genes differed significantly. Each colored bar represents a gene. The color represents the level of expression, and the sample information is listed across the top. The names of the known genes are indicated. Note the distinct patterns of altered gene expression between the positive and control animals. scrapie [68]. The overexpression of APP in the preclinical naturally infected animals found in our study are in accordance with these previous studies and highlight the possible early interaction between APP and PrP.
Water metabolism is of major importance in a number of physiological processes in the CNS. Alterations in the distribution of water and cerebrospinal fluid in the brain are a common occurrence in multiple neuropathological conditions, including brain edema, brain tumors, stroke, hyponatremia, head injuries and hydrocephalus. Aquaporin 4 (AQP4) is most likely expressed by activated glial cells, and an increase in its level is indicative of ongoing astrocytosis [69]. The increase in the expression levels of AQP4 has been reported in Creutzfeldt-Jakob disease, bovine spongiform encephalopathy and scrapie-infected transgenic mice [22,43,44,70]. Our work confirms the upregulation of AQP4 in the preclinical phases of natural ovine scrapie.
Genes that have never been associated with prion diseases or other neurodegenerative diseases were shown to be significantly regulated. Our microarray data indicated an increase in the expression of a gene similar to calcineurin-like phosphoesterase domain-containing 1 [CPPED1] that was confirmed by quantitative RT-PCR. The CPPED1 protein has hydrolase and metal ionbinding activities. To date, no studies have reported the differential regulation of this gene in neurodegenerative diseases. However, PrP c interacts with a range of divalent metal ions and maintains the their homeostasis, and the conformational change that occurs in the formation of PrP Sc is induced by the interaction with ions (see [71] for review). This gene has a positive association with prion deposition (GEAMM analysis), suggesting a possible role in early scrapie development. However, further analysis will be essential to confirm this conclusion.
(See figure on previous page.) Figure 3 Relationship between gene expression profiles and scrapie histopathological lesions. Proteins encoded by genes whose expression is associated with PrP Sc deposition, glial fibrillary acidic protein expression and spongiosis. Only the highly significant related genes are shown (P < 3.5x10 -3 ). The slope of regression between histopathological lesions and gene expression was obtained under a Mixed Model approach. Golgi golgin subfamily 4 [GOLGA4] may play a role in the delivery of transport vesicles containing GPI-linked proteins from the trans-Golgi network through its interaction with microtubule-actin crosslinking factor 1 (MACF1). The prion protein is attached to the outer leaflet of the plasma membrane by a glycosyl-phosphatidylinositol (GPI) anchor (reviewed in [72]). Our results demonstrate the upregulation of GOLGA4 is positively associated with PrP Sc deposition, suggesting that this protein might have a role in PrP trafficking.
We previously reported that anti-apoptotic genes are overexpressed in terminal scrapie, which suggested the activation of neuroprotective mechanisms during the disease [30]. In accordance with this, we found here that neural tissue-specific epidermal growth factor-like repeat domain-containing protein) NELL2 is overexpressed in preclinical scrapie. NELL2 is a secreted glycoprotein that is predominantly expressed in neural tissues and increases in vitro cell survival under cell deathinducing conditions [73]. In addition, NELL2 may play an important role in the maintenance of neural functions by regulating the intracellular machinery involving Ca 2+ signaling, synaptic transport and/or vesicle release [74]. In our study, NELL2 displayed a positive association with PrP Sc deposition and spongiosis, which suggests a possible role in the pathogenesis of the disease related to the role of PrP in Ca 2+ homeostasis [75].
Finally, we observed the downregulation of a serine/ arginine-rich splicing factor 3 [SRSF3], which seems to be involved in the differential splicing of the low-density lipoprotein receptor (LDLR), a major apolipoprotein E (APOE) receptor that has been associated with cholesterol homeostasis and, possibly, AD development [76]. This splicing factor is a proto-oncogene [77] and is antiapoptotic [78]. To our knowledge, our work is the first study to describe the differential regulation of this gene in prion diseases. The downregulation of SRSF3 is in accordance with its probable protective activity against neuronal cell death. Further studies will be necessary to investigate the possible role of SRSF3 in the disease.
In addition to the regulation of known genes, several non-annotated sequences were differentially expressed in the preclinical medullae and associated with scrapie lesions. We confirmed the upregulation of two sequences (OSRS1 and OSRS2) that were associated with astrocytosis and spongiosis, respectively. These sequences did not display homology with any known genes (Table 1), but they show a high homology with parts of two published bovine sequences (FQ482089.2 and NW_003104406.1, respectively). These sequences come from an ovine cDNA library generated from the brain and lymphoid tissue of scrapie-and control-infected sheep [25]. Although further analyses are necessary to confirm their differential regulation in a wider number of animals or in different prion animal models, these custom sequences can represent potential unknown biomarkers useful for the diagnosis of presymptomatic prion disease.

Conclusions
In summary, this is the first genome-wide expression study performed in naturally infected sheep with preclinical scrapie and shows the induction of a reduced number of genes compared with the changes shown in clinical scrapie sheep. Differentially regulated genes confirmed the involvement of the immune system, alterations in the extracellular matrix and changes in the ion binding in the neuropathology of prion diseases. In addition, changes in the levels of genes encoding for proteins related to cell-cell contact and trafficking and re-cycling pathways could play an important role in the development of the disease. The association of genomic changes with scrapie lesions allowed the identification of a higher number of candidate genes to be used as biomarkers and could be useful to develop biotools for the early diagnosis of the disease. In addition to their inclusion in the previous functional groups, the identified genes were related to water metabolism or apoptosis. As in previous studies, our findings confirm the close relationship between scrapie and other neurodegenerative diseases. Moreover, the reported association analysis contributes to the knowledge of the molecular mechanisms underlying the pathogenesis of prion diseases. Further studies are required to locate the cellular proteins encoded by these differentially regulated genes and to study the expression of the identified genes in other brain areas and, in this manner, contribute to the knowledge of their role in the disease. and coordination and helped to draft the manuscript. AB participated in the design of the microarray study and sequence alignment. RB conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.