Male Sprague Dawley rats (250–300 g, Envigo Correzzana, Italy) were assembled four per cage (38 cm × 60 cm × 20 cm). Rats (n = 88), were taken under precise conditions (24°C, 60% humidity, reversed 12 h light/dark cycle, with lights off from 08:00 to 20:00 h), with water and standard food ad libitum, and then a group (n = 80) were used for the experiments, performed between 10:00–14:00 h accordingly to the guidelines of the European Communities Directive of September 22, 2010 (2010/63/EU) and the Italian Legislation (D.L. March 4, 2014, n. 26). The protocol has been approved by the Italian Ministry of Health (Authorization n.1237/2015-PR to MRM), and by the Ethical Committee for Animal Experimentation of the University of Cagliari, Italy.

In order to validate the assumption that PCP treatment was effective in inducing behavioral changes resembling schizophrenia-like symptoms in our experimental conditions, i.e., it was effective in modifying PPI, of the acoustic startle response, 80 male rats were randomly assigned to five experimental conditions (n = 16/condition), each one with a different treatment regimen. Within each condition, half of the rats (n = 8) received PCP (purchased from R&D Systems Europe Ltd., Abingdon, UK 2.5 mg/kg, SC) or the corresponding volume of saline (1 mL/kg of rat body weight, SC; n = 8). The main conditions were as follows: acute treatment (animals were treated once); sub-chronic treatment (animals were treated once a day for eight consecutive days); chronic treatment (animals were treated once a day for 21 consecutive days). The acute treatment comprised animals that underwent the behavioral test (PPI assessment) after 10–15 min from the PCP injection. The sub-chronic and chronic treatment conditions comprised animals, which underwent the behavioral test (PPI assessment) 30 min after the last PCP injection (no wash out condition) and animals, which underwent the behavioral test 24 h after the last PCP injection (wash out condition). Treatments were carried out between 10:00 h and 13:00 h counterbalancing the sequence of PCP- and saline- treated rats. The day of the test, the rats in their home-cages were transferred during the dark phase of the cycle in a testing room under controlled environmental conditions (a sound proofed room lit by a dim red light). Each testing session was conducted with the same number of PCP treated rats and their matched saline treated controls. PPI testing was performed with minor modifications as already described (Frau et al., 2016). The apparatus used for detection of startle reflexes (Med Associates, St Albans, VT, USA) consisted of four standard cages placed in sound-attenuated chambers with fan ventilation. Each cage consisted of a plexiglas cylinder of 9 cm diameter, mounted on a piezoelectric accelerometric platform connected to an analog-digital converter. Two separate speakers conveyed background noise and acoustic bursts. Both speakers and startle cages were connected to a main computer, which detected and analyzed all chamber variables with specific software. Before each testing session, acoustic stimuli and mechanical responses were calibrated. The test began after acclimatization phase (5 min), with a 70 dB background white noise, continuing for the entire session. This period was followed by three blocks, comprising five pulse-alone trials of 115 dB (the first and the third), while the second includes a pseudorandom sequence of 50 trials, containing 12 pulse-alone trials, 30 trials of pulse preceded by 74, 78, or 82 dB pre-pulses (10 for each pre-pulse loudness level), and eight no-pulse trials, where exclusively the background noise was delivered. Time between two consecutive trials (inter-trial intervals) was arbitrarily selected between 10 s and 15 s. The % PPI was measured through the formula: % PPI = (100—mean startle amplitude for prepulse trials/mean startle for pulse alone trial) × 100.

Behavioral data from PPI experiments were analyzed by two way ANOVA. Although a trend toward a higher inhibition was observed with increasing tone intensities, we detected no significant differences in PPI responses to the three different prepulse intensities, therefore data from the three different prepulse intensities were put together for each rat to calculate individual mean values which were used for statistical analyses conducted by a two way ANOVA with the duration and the treatment as between subject factors. Post hoc pair wise contrasts on statistically significant main effects and/or interactions revealed by ANOVA analysis were obtained through Bonferroni test. Statistical analyses were applied using PRISM, Graph Pad 5 Software (San Diego, SA, USA) and the significance level were always p < 0.05.

Brain tissues from rats that did not receive any treatment nor were exposed to the PPI procedure were also used both for immunohistochemistry (n = 4) as well as for the determination of amino acid sequence by HPLC-MS (n = 4). Irrespective of the fact that rats underwent or not the PPI procedure and/or the PCP treatment, for ELISA at sacrifice, rats underwent deep diethyl ether anesthesia, blood (approximately 2 ml) was drawn by cardiac puncture, hence rats were rapidly decapitated for tissue extraction. The brain slide samples (coronal sections of 2 mm) were obtained using a cooled rat brain matrix through razor blades. The samples from different brain areas (hypothalamus, ventral and dorsal hippocampus, amigdaloid and accumbens nuclei, prefrontal cortex) were dissected from the brain slides using punches of 3, 4 and 5 mm dimensions as appropriate, following the coordinates of the Paxinos Atlas of the rat brain (Paxinos and Watson, 2007). Tissue samples were extracted (~10 ml/g wet tissue) in ice-cold phosphate buffer saline (PBS: 0.01 mol/L, pH 7.4) containing protease inhibitor cocktail (PIC, 5 m/ml in PO4 buffer, 0.01–0.05 M, pH 7.2–7.4 + NaCl 0.15 MP8340, Sigma-Aldrich, Schnelldorf, Germany), homogenized for 3 min using a micro sample pestle, hence tubes were heated for 10 min, and centrifuged (3000 rpm, 15 min). Supernatants were stored frozen until use (−30°C). Blood samples were collected in a tube containing ethylenediaminetetraacetic (EDTA: 1.78 mg/ml), rapidly centrifuged (11,000 rpm, 5 min), and plasma was stored frozen (−30°C) until analysis. For HPLC-ESI-MS, brain samples pooled from four untreated male rats were extracted in PBS containing PIC, homogenized by an Ultra-Turrax Homogenizer (Ika-Werke, Staufen, Germany), heated in a boiling water bath (10–15 min), and centrifuged (3000 rpm, 10–15 min). The supernatant was transferred on centrifugal filters with the molecular weight cut-off of 10 kDa and 3 kDa (Amicon Ultra device, Merck Millipore, Tullagreen Carrigtwohill Co. Cork, Ireland) and the resulting filtrate was lyophilized and stored frozen. Differently from ELISA and HPLC-MS, for immunohistochemistry, rats were anesthetized using chloral hydrate (400 mg/kg i.p.) and perfused-fixed with 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2–7.4.

The VGF N- and C- terminus, TLQP, NERP-3 and NERP-4 (also named NAPPE) antisera were produced as previously reported (Ferri et al., 1995; Brancia et al., 2005; Cocco et al., 2010; D’Amato et al., 2015), while in order to obtain the VGF_375–407 antiserum, the N-terminal decapeptide (GGGEDEVGEE), was conjugated with bovine thyroglobulin, via an additional C-terminal cysteine and used for immunizations. Specificity of each antiserum was achieved through ELISA (Table 1). It is relevant to point out here that our VGF antibodies, raised against truncated sequences at their N- or C- terminus, are expected to cross-react also with extended peptides. We assume that TLQP-62 will react in the TLQP assay much as TLQP-21 and in a way comparable to the VGF C-terminus standard in the relevant assay. Such assumption appears to fit well with our previous studies of TLQP-peptides, based on gel chromatography coupled with ELISA (Brancia et al., 2010, 2016; Noli et al., 2014, 2015). In some tissues, a large MW form bona fide corresponding to VGF (found in the void volume upon sephadex chromatography) was recognized by both the VGF N-terminus and C-terminus assays (Brancia et al., 2010; Cocco et al., 2010; Noli et al., 2014).

Competitive ELISA was performed as previously reported (D’Amato et al., 2008; Cocco et al., 2010). For each assay characterization (Table 1), a standard curve was obtained using a range of concentrations of each of the peptides listed. The peptide corresponding to the immunogen was treated as “reference”, with a certain amount of such peptide producing a 50% inhibition of the maximum signal (50% inhibitory concentration), and showing 100% reactivity in the assay. The other peptides listed resulted in a standard curve parallel to the “reference” one, but variably shifted to the right (more peptide required to produce 50% inhibition, hence lower reactivity in the assay, i.e., <100% reactivity), or to the left (less peptide produced 50% inhibition, hence higher reactivity, >100% vs. reference peptide). Synthetic peptides were used for plate coating (Nunc, Milan, Italy) for 3 h at room temperature, then the wells were treated with PBS (containing 9% normal serum from the secondary antibody donor species, 20 nM aprotinin, and 1 mg/mL EDTA) for 2 h. Primary incubations, with the VGF antibodies, (dilution: 1:5000–1:12000) were carried out in duplicate, including serial standard dilutions in parallel with rat samples (brain or plasma; 3 h at room temperature). Biotinylated secondary antibodies (1:10000; Jackson, West Grove, PA, USA), streptavidin-peroxidase conjugate (Biospa, Milan, Italy), and tetramethylbenzidine (X-tra Kem-En-Tec, Taastrup, Denmank) as substrate were used to reveal the positive labeling. The reaction was stopped with hydrogen cloride (1 mol/L) and optical density was measured at 450 nm using a multilabel plate reader (Chameleon: Hidex, Turku, Finland). Recovery of synthetic peptide/s added to plasma, or to tissue samples at extraction was >85% for all assays used.

Analysis of variance of ELISA data was carried out by one-way ANOVA, followed by post hoc multiple comparison tests (Student-Newman-Keul test), or by two-tailed Student’s t-test as appropriate by means of the StatistiXL software.

Freeze-dried material (5 μg) were re-suspended in 50 μl of aqueous formic acid (0.1% v/v) and 10 μl of the solution analyzed by an Ultimate 3000 RSLCnano system coupled with an Orbitrap Elite mass spectrometer (ThermoFisher, San Jose, CA, USA), using an EASY-Spray Column PepMap®RSLC C18 (3 μm particle diameter; 75 μm ID × 15 cm). Eluent solutions were constituted by: 0.1% (v/v) aqueous formic acid and acetonitrile with 0.1% (v/v) aqueous formic acid. The used gradient was linear from 0% to 55% of B in 35 min, at a flow rate of 300 nL/min. The Elite-Orbitrap mass spectrometer operated in data-dependent mode. Each full MS scan (60,000 resolving power) was followed by five MS/MS scans applied on the first five multiple-charged ions dynamically selected and disintegrated through collision-induced dissociation at a normalized collision energy of 35%. Thermo Proteome Discoverer 1.4 software was utilized to analyze Tandem mass spectra, while the SEQUEST cluster (University of Washington, Seattle, WA, USA, licensed by Thermo Electron Corp) was useful to search engine against UniProtKB Rattus norvegicus proteome (release 2016-03). The limits used for peptide matching were Xcorr scores greater than 1.5 for singly charged peptide ions while 2.0 for doubly charged ions and 2.5 for triply charged ions. Precursor mass search tolerance was set to 10 ppm, and fragment mass tolerance to 0.02 Da. The N-terminal modification (Gln → pyro-Glu) and the C-terminal amidation were selected as dynamic modifications. A false discovery rate (FDR) below 1% was applied. Peptide sequences and sites of covalent modifications were confirmed by manual spectra annotation by matching experimental MS/MS spectra with the theoretical ones generated by the MS-Product program available on Protein Prospector website¹. The match was assumed as positive when: (1) all the experimental m/z values were higher than 5% in the theoretical fragmentation spectrum; and (2) experimental and theoretical values differed less than ±0.03 m/z.

After fixation time, brains were included in an embedding medium (Cocco et al., 2003) and cut (at 10 μm) using a HM-560 cryomicrotome (Microm; Walldorf, Germany; Ferri et al., 2002). Sections comprehensive of the prefrontal cortex and the nucleus accumbens were incubated overnight in a humid chamber mixing the anti VGF antibodies (dilutions: 1:600–1:4000) with different neurotransmitters/enzymes/neuromodulators (Table 2), all diluted in PBS containing 30 ml/L of normal donkey serum, 30 ml/l of normal rat serum and 0.02 g/L NaN3. Relevant species-specific donkey secondary antibodies conjugated with Cy₃ or Cy₂ (Jackson Immunoresearch Laboratories, West Grove, PA, USA) were used (1:400). Slides were covered with PBS-glycerol (40%), observed and photographed using BX41 and BX51 fluorescence microscopes (Olympus, Milan, Italy) equipped with the Fuji S2 and S3 Pro digital cameras (Fujifilm, Milan, Italy). Routine controls included substitution of each antibody, in turn, with PBS, the use of pre-immune or non-immune sera, and the testing of each secondary antibody with their respective non-relevant primary antibodies.

In the normal rats, the VGF peptides were expressed in all the brain areas examined (hypothalamus, prefrontal cortex, nucleus accumbens, dorsal and ventral hippocampus, amygdaloid nucleus) and also in the plasma (Figure 1). VGF peptide levels ranged between approximately 4 pmol/g and 900 pmol/g in the brain and between 6 pmol/ml and 40 pmol/ml in plasma. Among the brain areas studied, the richest in the VGF peptides was the amygdaloid nucleus containing about 900 pmol/g, revealed with the N-terminus antibody. Analysis of each brain area for the content of the VGF peptides, revealed the following: the hypothalamus, dorsal hippocampus and prefrontal cortex contained more NERP-4, the accumbens and amygdaloid nuclei were more reactive to the N-terminus antibody, while the ventral hippocampus contained more TLQP peptides compared to the other VGF peptides. Conversely, in plasma, the TLQP peptides were the most represented.

The RP-HPLC high-resolution ESI-MS analysis allowed the detection of 13 mass values attributed to different fragments of proVGF. Table 3 reports the experimental monoisotopic m/z values of the mono-charged peptides of proVGF, the post-translational modifications and the sequences determined by MS/MS experiments. Fragments VGF_24–54 and VGF_24–60 derived from the proVGF N-terminus, and fragments VGF_587–600, VGF_588–600, derived from the C-terminus portion. Moreover, the following peptides belonging to the NERP family were identified: NERP-3 (VGF_180–209), with the N-terminal Gln converted to Pyro-Glu, NERP1 (VGF_285–309), amidated at the C-terminus and two fragments encompassing NERP-4 (VGF_489–507 and VGF_487–507). We unfortunately failed to identify the TLQP and GGGE peptides, however, we detected and characterized the same VGF_353–372 fragment previously identified in the striatum (Karlsson et al., 2013), and its shorter form (VGF_360–372).

As shown in Figure 2, PCP treatment induced an impairment of PPI response after acute, sub-chronic and chronic treatment, with some differences in the entity of the effects due to the duration of the treatment regimen. Accordingly, two ways ANOVA revealed a significant effect of the drug (F_(1,110) = 107.2, p < 0.001), of the treatment regimen (F_(4,110) = 17.92, p < 0.001), and of interaction between the two factors (F_(4,110) = 24.46, p < 0.001). Post hoc comparisons revealed that acute, sub-chronic and chronic PCP treatments without wash out, all induced PPI impairment (24.33%, −2.62% and −1.60% for acute, sub-chronic and chronic treatment, respectively, compared to values between 57.68% and 66.49% of saline treated rats, all p < 0.001); in contrast, PCP-treated animals which underwent the 24 h wash out displayed PPI values indistinguishable from those of their saline matched controls (64.07% and 63.51% for sub-chronically and chronically PCP treated rats, respectively; all p > 0.05 when comparing PCP-washed out animals to the corresponding saline-treated rats, and all p < 0.001 when comparing PCP-washed out animals with the corresponding PCP not washed out rats). Finally, post hoc comparisons revealed also a significant difference in the entity of PPI impairment between acutely (24.33%) and both no wash out sub-chronically (−2.62%) and chronically (−1.60%) PCP-treated rats (both p < 0.001), with no difference in the entity of PPI impairment between no washed out sub-chronically and chronically treated rats (p > 0.05).

VGF peptide levels were measured in the prefrontal cortex, nucleus accumbens, dorsal and ventral hippocampus, amygdaloid nucleus and hypothalamus as well as in plasma of saline- and PCP- treated rats that underwent the PPI test under the acute, sub-acute and chronic treatment condition, with the aim of unraveling possible changes in the levels of one or more VGF peptides (Figure 3). We found changes in the content of VGF peptides exclusively within the prefrontal cortex and nucleus accumbens but not in the other brain areas studied or in plasma (not shown). In the prefrontal cortex, we found an increase in VGF C-terminus after sub-chronic and chronic PCP-treated rats, as well as for TLQP peptides exclusively in the sub-chronic group when compared to the respective saline treated rats (p < 0.01). Conversely, in the nucleus accumbens, PCP treatment produced a decrease in VGF C-terminus levels after all types of treatment (acute, sub-chronic and chronic), while N-terminus- and GGGE-peptides were reduced after chronic PCP treatment when compared to the respective groups of saline-treated rats (p < 0.01). In both the prefrontal cortex and nucleus accumbens, the amount of all peptides changed after PCP-treatment, returned to similar levels to those of saline-treated controls after 1 day from the last PCP treatment (sub-chronic and chronic wash out groups).

The localizations of the VGF peptides, whose levels were changed by ELISA in the prefrontal cortex and nucleus accumbens after PCP treatment, were studied in the same areas of normal rats, to characterize the neuronal circuits containing the VGF peptides. We used VGF antibodies together with antibodies against the relevant neurotransmitters, enzymes or neuromodulators (Table 2). In the prefrontal cortex, the ~57% of the VGF C-terminus immunoreactivity (Figure 4A) was found within the majority of parvoalbumin perikarya (Figure 4B), while the remained positive labeling stayed uncharacterized. The majority of the TLQP immunoreactivity (Figure 4C) was found in axons containing tyrosine hydroxylase (TH; Figure 4D) while roughly half of the TH positive nerve axons contained the TLQP peptides. In the nucleus accumbens, GGGE- (not shown) and VGF C-terminus- (Figures 4E,F) immunoreactivities were revealed in ~30%–40% of neuron terminals containing glutamic acid decarboxylase (GAD; and vice-versa) inside the shell. The N-terminus immunoreactivity was almost wholly found in all structures containing somatostatin (axons, perikarya and nerve terminals; Figures 4G,H) and in a small percentage (~5%) of NO synthase (NOS) containing cells (not shown).