Chemicals
Sodium cyanide (NaCN CAS No. 143-33-9, 97.2% purity) and cyanate (NaOCN CAS No. 917-61-3, 96% purity) were bought from Sigma–Aldrich (St. Louis, MO) and stored at room temperature. All other laboratory reagents were of analytical or molecular biology grades.
Animals
Young adult male heterozygous rats (Crl: NIH-Fox1 rnu/Fox 1+, 6–8 weeks old) (N = 52), weighing 140-210 g upon arrival were donated by Professor Neuwelt, Department of Neurology, Oregon Health & Science University (OHSU). They were caged in an animal room maintained on a 12/12 h light/dark cycle in the Center for Research on Occupational and Environmental Toxicology (CROET). Food and water were given ad libitum. Experimental protocols were approved by the OHSU institutional animal care and use committee (IACUC).
Diet and dosing regimens
Diet
Custom-synthesized isonitrogenous rodent diet with either all amino acids (AAA-diet; code TD09460) or lacking 75% of the SAA-content relative to the control diet (SAA-deficient diet; code TD09463) were purchased from Harlan (Madison, WI) and stored at 4°C until use. We chose a 75%-deficient but not free SAA to allow for animal survival and cyanide detoxification since previous experimentation showed animals fed with SAA-free chow had dramatic weight loss and muscle weakness (Tor-Agbidye et al. 1999a).
Dosing regimens
Animals were first acclimated for a 5-day period on a diet consisting of 4:1 portions of normal rodent chow (PMI Nutrition International, NJ) with either AAA-diet (N = 28) or 75%-SAA deficient diet (N = 24). On the 6th day, they were assigned to experimental groups (N = 7-10/group) and treated intraperitoneally (one injection per day) for up to 6 weeks as follows: (Banea-Mayambu et al. 1997) AAA-diet, 2.5 mg/kg body weight (bw) NaCN; (Berg et al. 2013) AAA-diet, 50 mg/kg bw NaOCN; (Boivin et al. 2013) AAA-diet, equivalent amount of vehicle (1 μl/g bw saline); (Chabwine et al. 2011) SAA-deficient diet, 2.5 mg/kg bw NaCN; (Cliff et al. 2011) SAA-deficient diet, 50 mg/kg bw NaOCN; (Crist et al. 1973) SAA-deficient diet, equivalent amount of vehicle (1 μl/g bw saline). Dose selection for cyanide was informed by previous findings of sublethal and/or lethal cyanide poisoning prior to konzo outbreaks (Banea-Mayambu et al. 1997). Cyanate was given at doses similar to doses known to induce neuropathy in humans (Ohnishi et al. 1975).
Specific protocols
Animal observations
Rodents were examined daily in an open field for physical signs including walking difficulties, tremors and hind limb extension reflexes, which were elicited when the animal was gently raised by the tail. Motor function was assessed by animal performance on an accelerating rotating rod. Animals were individually placed on rotating rods in a software-driven rotarod apparatus (AccuScan Instruments, Inc., Columbus, OH) set in an accelerating mode. The rotation speed was gradually increased from 5 to 25 r.p.m. The apparatus had an automatic system for fall detection via photobeams. Rotarod performance was analyzed to estimate how odds of remaining on the rod for at least 60 seconds changed over time. Testing was carried out only in animals maintained on SAA diet to determine whether SAA dietary restriction had an impact on the toxicity of cyanide, our main experimental paradigm of interest.
Tissue preparation for proteomics
Serum samples were obtained on day 14 and 28 of the study period. Blood was collected from the salphenous vein using vacutainer tubes with no anticoagulants from all rats and kept overnight at 4°C. The serum was then collected in sterile tubes and centrifuged at 15,000 rpm for 15 min at 4°C. Serum was aliquoted in cryotubes and stored at -80°C till later LC-MS/MS studies. At study termination, rats were deeply anesthetized with 4% isofluorane (1 l oxygen/min) and transcardially perfused with saline through the ascending aorta to remove the remaining blood. The spinal cords were dissected out, flash-frozen in liquid nitrogen, and stored at -80°C. Prior to LC-MS/MS studies, the frozen spinal cord tissue was thawed by adding 500 μl of 100 mM ammonium bicarbonate, and the tissue was uniformly dispersed by 3X15 sec of sonication (Sonic Dismembranator, Model 60, Fisher Scientific). Samples were then centrifuged at 12,000 g for 30 min at 4°C, the supernatant removed, the pellet re-suspended and centrifugated, and the two supernatants pooled. The pellet was then re-suspended in 250 μl 100 mM ammonium bicarbonate by 5 sec of sonication as before. The protein content of the spinal cord soluble fraction, suspended pellet, and serum samples were then determined using the BCA assay (Pierce Chemical, Rockford, IL). The equivalent of 50 μg portions of the two spinal cord fractions and serum were then dried by vacuum centrifugation in preparation for trypsin digestion.
Protein digestion
Fifty μg portions of dried serum, soluble and insoluble spinal cord fractions were dissolved in 50 μl of 100 mM ammonium bicarbonate containing 1 mg/ml of Rapigest detergent (Waters Corp., Milford, MA). Eight μl of 50 mM dithioerythritol (DTT) solution was then added and the samples incubated for 60 min at 60º C, followed by addition of 8 μl of 150 mM iodoacetamide solution and incubation at room temperature for 30 min. An additional 16 μl of 50 mM DTT was then added, samples incubated at room temperature for an additional 15 min, and 20 μl of 0.1 μg/μl proteomics grade trypsin added (Sigma Chemical Co., St Louis, MO). Following an overnight incubation at 37°C, samples were acidified by addition of 0.5% trifluoroacetic acid (final concentration), incubation at 37°C for 45 min, centrifugation at 13,000 g for 10 min, and the supernatant frozen prior to analysis of peptides by mass spectrometry.
Mass spectrometric analysis
Ten μg samples of peptide digests were injected onto a trap cartridge (Michrom Bioresources, Auburn, CA) at 20 μl/min, samples were washed for 5 min with 2% acetonitrile, 0.1% formic acid, and the trap cartridge placed in-line with a 0.5 × 250 mm SB-C18 reverse phase column (Agilent Technologies, Santa Clara, CA). Peptides were then eluted using a 200 min gradient of 7-30% acetonitrile containing 0.1% formic acid at a flow rate of 10 μl/min. Peptide m/z values (MS spectra) and fragment ions (MS/MS spectra) were collected using an LTQ linear ion trap mass spectrometer (Thermo Scientific, San Jose, CA) with an ion max electrospray source fitted with a 34Ga metal needle. Each survey MS spectrum from 400–2000 m/z was followed by 3 data-dependant MS/MS spectra on the 3 most intense ions found in each survey MS spectrum. The instrument used the dynamic exclusion feature of its control software to ignore previously analyzed ions (repeat count of 1, maximum exclusion list of 50, and 30 sec exclusion time), a tune file configured with one μscan, a maximum ion accumulation time of 200 m/sec, and AGC targets of 30,000 and 10,000, respectively, for MS and MS/MS spectra.
MS/MS spectra from each analysis were converted to dta files using Bioworks 3.3 (Thermo Scientific) and each MS/MS spectrum matched to peptide sequences in a database using the program Sequest (Version 27, rev. 12, Thermo Scientific). Searches were performed with a static modification of +57 on cysteines for carboxyimidomethylation, and differential modifications of +16 for oxidation of methionines, and +43 for carbamoylation of lysines. The search also used trypsin specificity. A database of rat sequences was created from the Uniprot database (Swiss Bioinformatics Institute, Geneva, Switzerland) downloaded in Feb. 2011 (15,524 entries), the sequences of common contaminants added, and the database amended with sequence reversed entries to estimate peptide false discovery rates. Sequest results were then processed using the program Scaffold (version 4.0.5, Proteome Software, Portland, OR) to produce a list of identified peptides and their carbamoylation sites. Peptide and protein confidence levels of 95% and 99% were used, and a minimum of 2 unique peptides matching each protein entry was required.
Since few carbamoylated peptides were detected in the soluble fraction of spinal cord, further analysis to test for differences in the level of carbamoylation between different treatment groups was only performed for serum and spinal cord insoluble fractions. Differences in levels of carbamoylation in various treatment groups was determined by tabulating the numbers of MS/MS spectra assigned to carbamoylated peptides from albumin in serum digests, and the proteins myelin basic protein, neurofilament light polypeptide, 2’3’-cyclic-nucleotide 3’-phosphodiesterase, myelin proteolipid protein, and glial fibrillary acidic protein in digests of water-insoluble proteins from spinal cord. Inclusion lists for each of the identified carbamolyated peptides was created by specifying the m/z value for each in the instrument’s control software. This maximized the collection of MS/MS spectra for carbamoylated peptides during the quantitative analysis so that a shorter data collection period was required. This mass spectrometric analysis was identical to those described above. However they used a 60 min gradient, inclusion lists specifying the m/z values of the targeted carbamoylated peptides, and the dynamic exclusion feature of the instrument’s control software was disabled. Differences in the level of carbamoylation in various sample groups was then calculated by summing the numbers of MS/MS scans (spectral counts) assigned to carbamoylated peptides from the target proteins in each sample. The MS/MS scans collected during the analysis were again searched using Sequest using the same database and differential searches for modified peptides as before. However, due to the size of the result files, the Sequest results were filtered using PAW software (Wilmarth et al. 2009), which controlled peptide false discovery using numbers of matches to the sequence reversed entries. Peptides identified as being carbamoylated were separately filtered from unmodified peptides to maintain their false discovery rate below 5%. The linearity of the carbamoylation detection was tested by mixing protein digests from cyanate and vehicle treated animals from the sulfur deficient diet group (Figure 2). The numbers of identified carbamoylated peptides as a function of percent cyanate treated serum or spinal cord protein was determined following Sequest searches and filtering of MS/MS data using Scaffold as described above.
Statistical analysis
The rotorod performance was analyzed using generalized estimating equations (GEE) to estimate the odds of remaining on the accelerating rotating rod for at least 60s changed over time (per each additional day of treatment). At study termination, a secondary analysis of the performance on the rotorod was conducted. An overall (log-rank) test on the time needed to fall off the rotorod among the three groups was carried out. Numbers of carbamoylated sites per protein were assumed to follow a Poisson distribution with the mean number of sites dependent upon diet, treatment, or the interaction between the two. GEE was again used to account for the duplicate samples collected within each animal. Statistical significance was set at 0.05 (0.10 for interaction effects); all p-values are two-sided and confidence intervals set to 95% coverage.
