Article pubs.acs.org/jpr
Neurotoxin-Induced Neuropeptide Perturbations in Striatum of Neonatal Rats Oskar Karlsson,*,†,‡,# Kim Kultima,#,†,§ Henrik Wadensten,† Anna Nilsson,† Erika Roman,† Per E. Andrén,† and Eva B. Brittebo† †
Department of Pharmaceutical Biosciences, Uppsala University, SE-751 24 Uppsala, Sweden Department of Environmental Toxicology, Uppsala University, SE-752 36 Uppsala, Sweden § Department of Medical Sciences, Cancer Pharmacology and Computational Medicine, Uppsala University, SE-751 85 Uppsala, Sweden ‡
S Supporting Information *
ABSTRACT: The cyanobacterial toxin β-N-methylamino-Lalanine (BMAA) is suggested to play a role in neurodegenerative disease. We have previously shown that although the selective uptake of BMAA in the rodent neonatal striatum does not cause neuronal cell death, exposure during the neonatal development leads to cognitive impairments in adult rats. The aim of the present study was to characterize the changes in the striatal neuropeptide systems of male and female rat pups treated neonatally (postnatal days 9−10) with BMAA (40−460 mg/kg). The label-free quantification of the relative levels of endogenous neuropeptides using mass spectrometry revealed that 25 peptides from 13 neuropeptide precursors were significantly changed in the rat neonatal striatum. The exposure to noncytotoxic doses of BMAA induced a dose-dependent increase of neurosecretory protein VGF-derived peptides, and changes in the relative levels of cholecystokinin, chromogranin, secretogranin, MCH, somatostatin and cortistatin-derived peptides were observed at the highest dose. In addition, the results revealed a sex-dependent increase in the relative level of peptides derived from the proenkephalin-A and protachykinin-1 precursors, including substance P and neurokinin A, in female pups. Because several of these peptides play a critical role in the development and survival of neurons, the observed neuropeptide changes might be possible mediators of BMAA-induced behavioral changes. Moreover, some neuropeptide changes suggest potential sex-related differences in susceptibility toward this neurotoxin. The present study also suggests that neuropeptide profiling might provide a sensitive characterization of the BMAA-induced noncytotoxic effects on the developing brain. KEYWORDS: neurotoxin, BMAA, neonatal striatum, neuropeptide, ALS/PDC, cyanobacteria, sex differences, cortistatin
■
INTRODUCTION Cyanobacteria exist as free-living or symbiotic organisms in terrestrial and aquatic environments worldwide. These organisms produce several cyanotoxins and are capable of massive proliferation, with large cyanobacterial blooms in various types of waters. Most cyanobacteria produce the neurotoxic nonprotein amino acid β-N-methylamino-L-alanine (BMAA).1,2 BMAA has been detected in several water systems, including temperate aquatic ecosystems, and in mollusks and fish used for human consumption, suggesting that BMAA bioaccumulates in aquatic food chains.3,4 Exposure to BMAA has been implicated in the etiology of Amyotrophic lateral sclerosis/Parkinsonism-dementia complex (ALS/PDC) on the island of Guam5,6 and in ALS and Alzheimer’s disease in North America.7 BMAA is an ionotropic and metabotropic glutamate receptor agonist that induces neuronal degeneration via excitotoxic mechanisms, although other mechanisms of toxicity might also © 2013 American Chemical Society
be involved, such as oxidative stress or the misincorporation of the beta-amino acid into protein.8−11 At low concentrations, BMAA is not considered to be acutely neurotoxic to adult rodents12,13 and the access of BMAA to the adult rodent brain is reported to be limited.14,15 In contrast, autoradiographic studies revealed that 3H-BMAA-derived radioactivity is transferred across the blood−brain barrier in neonatal mice, with a distinct localization in specific brain regions, such as the striatum and the hippocampus.16 BMAA treatment of neonatal rats during development induced transient behavioral changes, such as disturbed motor function and hyperactivity in neonates16 and long-term cognitive impairments and changes in neuronal protein expression in adults.17 A high dose (460 mg/kg) of BMAA induced acute neuronal cell death in the neonatal Received: October 30, 2012 Published: February 15, 2013 1678
dx.doi.org/10.1021/pr3010265 | J. Proteome Res. 2013, 12, 1678−1690
Journal of Proteome Research
Article
The Uppsala animal ethical committee approved all animal experiments, which were performed in accordance with the guidelines of the Swedish legislation on animal experimentation (Animal Welfare Act SFS1998:56) and European Union legislation (Convention ETS123 and Directive 86/609/EEC).
hippocampus, retrosplenial and cingulate cortices, but no neuronal cell death was observed in the neonatal striatum.18 The striatum is important for several cognitive processes19,20 and even subtle effects during the development of the brain region might contribute to the observed cognitive impairments in learning and memory in adult animals following neonatal exposure to BMAA.18,21 To elucidate the mechanisms of BMAA-induced changes following neonatal exposure, more detailed studies on the early effects of BMAA in the developing brain are needed. The high and selective uptake of 3H-BMAA in the striatum, which is important for motor function and cognitive tasks,19,22 makes this brain region an interesting candidate for studies of BMAAinduced effects during development. The aim of the present study is to examine the effects of BMAA on neuropeptide levels in the neonatal striatum using a label-free mass spectrometrybased approach. The results showed that neonatal exposure to BMAA induced a dose-dependent increase of VGF-derived peptides associated with important developmental changes in the brain. We also showed that neonatal exposure to BMAA increased the relative levels of the protachykinin 1-derived C-terminal-flanking peptide, substance P and neurokinin A in female pups. Because several of these peptides play a critical role in the differentiation and survival of neurons, the observed neuropeptide changes in the neonatal striatum might mediate BMAA-induced behavioral effects.16,18,21
■
Tissue Extraction and Sample Preparation
The frozen brain samples were denatured at 95 °C for 30−40 s using a rapid heat transfer inactivation instrument (Stabilizor, Denator AB, Gothenburg, Sweden) to completely prevent protein degradation.24 The tissue samples were subsequently transferred to low-retention Eppendorf tubes, suspended in prechilled extraction solution (7.5 μL 0.25% (v/v) aqueous acetic acid/mg tissue) and homogenized by sonication (Vibra cell 750, Sonics & Materials, Inc., Newtown, CT) in an ice-bath for 30 s.25 Each sample suspension was centrifuged at 14 000g for 40 min at 4 °C to remove insoluble material. The supernatant was transferred to Microcon 10 kDa cutoff spin columns (YM-10, Millipore, Bedford, MA) and centrifuged at 14 000g for 90 min at 4 °C. The resulting peptide filtrate was frozen at −80 °C until further analysis. Randomized Block Design of Mass Spectrometry (MS) Analysis
For quantitative MS analyses, the samples were run in a restricted randomized block design. Thus, one biological sample from each treatment group (six groups) was randomly selected and consecutively analyzed (as one block of samples). This procedure was repeated until all samples were analyzed. The order in which the samples within each block were analyzed was restricted (i.e., the order in which the samples from different treatment groups were injected was different in all blocks). A restricted randomized block design was used to avoid systematic bias due to flow changes in the LC system and minimize the effect of a potential carryover effects between samples affecting the relative peptide calculations. Prior to each block of samples, a blank run (0.25% (v/v) aqueous acetic acid) was performed and a reference sample of pooled material was analyzed as every seventh sample for internal quality control. After all forty-eight biological samples were analyzed, the procedure was repeated, constructing new blocks of samples and analyzing all samples a second time. This procedure was undertaken to minimize the number of missing values in each sample and to facilitate the estimation of both technical and biological variations. All samples were analyzed without changing the analytical column.
MATERIALS AND METHODS
Chemicals
Unless otherwise stated, all chemicals were obtained from SigmaAldrich Co. (St. Louis, MO). β-N-Methylamino-L-alanine (LBMAA) hydrochloride (≥97%) was used. Animal Experiments
Pregnant outbred Wistar rats were obtained from Taconic (Ejby, Denmark). Each dam was housed alone in a standard cage (59 × 38 × 20 cm) containing wood-chip bedding and nesting material. The animals were maintained on standard pellet food (R36 Labfor; Lantmännen, Kimstad, Sweden) and water ad libitum, and housed in a temperature- and humidity-controlled environment with a 12-h light/dark cycle (lights on at 6 a.m.). After the day of birth (postnatal day (PND) 0), all male and female pups in each litter were randomly assigned to the control or to one of the BMAA treatment groups. The male pups were given one daily subcutaneous injection (20 μL/g) of BMAA at 40 mg/kg (M40; corresponding to 50 mg/kg BMAA HCl; n = 8), 150 mg/kg (M150; corresponding to 200 mg/kg BMAA HCl; n = 8), or 460 mg/kg (M460; corresponding to 600 mg/kg BMAA HCl; n = 8) freshly dissolved in Hanks’ balanced salt solution, or vehicle (MC; n = 8), for 2 days (PNDs 9 and 10). The female pups were treated with 150 mg/kg BMAA (F150; n = 8), or vehicle (FC; n = 8), for 2 days (PNDs 9 and 10). The rat pups were sacrificed by decapitation at 24 h after the last BMAA treatment. The brains were dissected on ice and sliced manually in a cooled rat brain matrix (coronal sections, 1 mm slots; ASI Instruments, Inc., Warren, MI) using razor blades. The caudate putamen (striatum) was dissected from the sections with the guidance of a neonatal rat brain atlas.23 Striatum was collected from sections between approximately 6.2 and 5.3 mm, but not beyond the anterior commissure. The tissues were immediately frozen on dry ice and stored at −80 °C until further use for peptide analysis.
Liquid Chromatography (LC)
An aliquot (5 μL) of the peptide filtrate was obtained from each animal and analyzed on a nano-LC system (Ettan MDLC, GE Healthcare, Uppsala) using a nano-electrospray ionization (nano-ESI) interface coupled with an Q-Tof-2 (Waters, Manchester, U.K.) or a linear ion trap (LTQ; Thermo Electron, San Jose, CA) mass spectrometer for quantification and identification, respectively. The sample was injected and desalted on a precolumn (300 μm inner diameter (i.d.) × 5 mm, C18 PepMapT, 5 μm, 100 Å, LC Packings, Amsterdam, The Netherlands) at a flow rate of 10 μL/min for 10 min. A 15-cm fused silica emitter with a 75 μm i.d. and a 375 μm outer diameter (New Objective; Woburn, MA) was used as the analytical column. The analytical column was packed in-house with reversephase Reprosil-Pur C18-AQ 3-μm, 120 Å, resin (Dr. Maisch, GmbH; Ammerbuch-Entringen, Germany) using a pressurized packing device (Proxeon Biosystems; Odense, Denmark). Buffer A 1679
dx.doi.org/10.1021/pr3010265 | J. Proteome Res. 2013, 12, 1678−1690
Journal of Proteome Research
Article
(2% (v/v) acetonitrile in 0.1% (v/v) aqueous formic acid) and Buffer B (0.1% (v/v) formic acid in acetonitrile) were used. The samples were injected and desalted on a precolumn. The analytical column was a Biopshere C18 120A, 360/75-μm column, with 15-cm length. For the separation and elution of the peptides, a 90-min gradient from 8 to 40% Buffer B at a flow rate of approximately 200 nL/min was applied. Tandem MS data were acquired in a data-dependent manner, with continuous switching between full MS (m/z 300−2000) and zoom scans, followed by 5 MS/MS scans, where the most intense peak is selected before inclusion on an exclusion list during a 60 s period. The analysis confirmed the identity of 6 peptides that were vaguely identified in the main run. The MS/MS spectra for these peptides are provided in the Supporting Information.
(0.25% (v/v) aqueous acetic acid) and Buffer B (84% acetonitrile and 16% of 0.25% (v/v) aqueous acetic acid) were used as mobile phases. For the separation and elution of the peptides, a 40-min gradient from 3% to 65% Buffer B at a flow rate of approximately 200 nL/min was applied. Peptide Quantification
The quantitative analysis of the peptide samples was performed on the Q-Tof-2 instrument. The label-free quantification of relative peptide levels was conducted as described previously8−11 with minor modifications. Brifely, the mass spectrometer was calibrated according to the manufacturer’s recommendations using a PEG solution (Fluka, Switzerland). MS data were collected in a continuous mode in the m/z range 300−1000 for 45 min. The raw MS data were converted into ASCII files using the Data Bridge module available in the MassLynx software (V3.5, Micromass) and imported into DeCyder MS2.0 (GE Healthcare). Ion peaks were automatically detected (parameters: elution time 0.2 min, typical peak resolution 8000, accepted charge states 1−12, signal-to-noise cutoff of 4, background subtracted quantification (smooth surface). Thereafter, the DeCyder MS 2.0 time alignment function was applied, with the following parameters: max stretch/compress, 1 min; max leader, 10%; stretch/compress penalty, 0.1. The ion peaks of the time aligned intensity maps were subsequently matched with the following tolerances: time ±1 min and m/z ± 0.2 Da. Prior to data export, the identified peptides were manually controlled and the mismatched peaks were removed from subsequent data analyses.
Normalization and Data Analysis
Normalization was conducted on log2-transformed data exported from the DeCyderMS software. To correct for global intensity differences between peptide runs, the data were normalized in two steps.29 Briefly, a linear regression was fitted for each individual run to a median run constructed of all median peak values for ions matched in >10% of all runs. On the basis of the linear regression equation, new values were predicted for each run. A locally weighted polynomial regression (Lowess) was subsequently fitted for each matched peptide against the run order, and the mean value across all runs was added to retain the native intensity dimension. For each matched peptide, a proportion of 0.2 neighbors (runs), weighted by their distance to the measurement, were used to controll the smoothness of the fit. Two technical replicates and three biological samples were removed due to low peak intensity or high background, resulting in a small fraction of peaks successfully matched to all other samples. These quality measures generated the following samples per group: MC (n = 8), M40 (n = 7), M150 (n = 8), M460 (n = 7), FC (n = 7) and F150 (n = 8). All LC−MS analyses include missing peak values due to technical reasons. Consequently, only the peaks/peptides, which were identified, correctly aligned and matched in at least five samples in each group, were subjected to downstream data analyses, assuring that the number of observations in all groups is similar and excluding peaks/peptides with a large number of missing values. To examine differences in the peak expression between the groups three different linear models were employed. Different types of models were used to obtain the best estimates for the different biological questions considered. (i) In the first model, the summation of the group means for males (four groups) was considered as fixed effects and individual samples were defined as random effects. This model provides the best estimate of the treatment effect of BMAA in males only. (ii) In the second model, the summation of the group means for sex (two groups), the treatment groups in both sexes (M150 and F150) and the interaction between sex and treatment were considered as fixed effects. Individual samples were defined as random effects. This model considers the interaction effect between males and females and the effect of 150 mg/kg BMAA, which is the only dose that was measured in both males and females. In the case of a significant interaction effect, females and males respond differently to BMAA exposure. (iii) In the third model, the summation of the group means (six groups) for all different groups was considered as fixed effects and individual samples were defined as random effects to compare all six groups. Based on this model the variance estimates are presented for the biological samples and technical
Peptide Identification
Nano-LC LTQ MS/MS analysis was performed on each sample for the identification of peptides. The same nano-LC system and setup as used for the peptide quantification was applied. Tandem MS data were obtained in a data-dependent manner, with continuous switching between full MS (m/z 300−2000), zoom (most intense peak in full scan) and MS/MS scans, where the most intense peak is selected twice in a time window of 40 s before inclusion on an exclusion list during a 150-s period. The tandem mass spectra were converted into data files using Xcalibur (Version 2.0 SR2) and further compiled into mgf files using an in-house developed script. The obtained MS/MS spectra were searched against the following sequence collections from the SwePep database,26 SwePep precursors, SwePep peptides and SwePep prediction27 or against the UniProt Knowledgebase (UniProtKB) using X! Tandem28 for the specific identification of endogenous peptides. The following settings were used for the database search: peptide mass tolerance of +2 /−0.5 Da; fragment mass tolerance of ±0.5 Da; unspecific cleavage (SwePep precursor), respectively, no-cleavage (SwePep prediction); possible post-translational modifications (N-terminal acetylation, N-terminal pyro-glutamic acid derived from glutamine, C-terminal amidation, oxidation of methionine and tryptophan, phosphorylation of serine, tyrosine and threonine). A significance threshold of log
