The autonomic nervous system regulates the secretion of bioactive proteins and peptides from salivary glands that can be important in systemic physiological responses. The prohormone submandibular rat-1, which is highly expressed in rat submandibular glands, can be cleaved to produce polypeptides with analgesic and anti-inflammatory activities. Human genes related to submandibular rat-1 have conserved biological functions and are potentially important in pain suppression, erectile function, and inflammation. In this study we describe the differential expression and posttranslational modification of submandibular rat-1 protein in salivary glands, the urogenital tract, lung, blood, and saliva in male Sprague-Dawley and Brown Norway rats. Submandibular rat-1 protein is secreted into saliva after the administration of β-adrenergic or cholinergic agonists. Removal of the sympathetic ganglion that innervates the salivary glands results in increased levels of submandibular rat-1 protein in salivary glands. The secretion of submandibular rat-1 in response to physiological stress may provide a large pool of submandibular rat-1-derived peptide products that can promote analgesia and decrease inflammation locally and systemically. This pathway may be conserved among mammals and may constitute an important anti-inflammatory and analgesic response to stress.
- sympathetic nervous system
- submandibular rat-1
- submandibular gland
autonomic nervous system regulation of endocrine and exocrine functions of glands is an important response to physiological insults and stress. The major salivary glands of mammals contain proteins and polypeptides that are released into blood or saliva in response to autonomic stimulation and mediate systemic responses (reviewed in Refs. 1, 7–9, 24, 34). In rats, sympathetic nerve modulation of the submandibular gland (SMG) regulates the severity of pulmonary inflammation and anaphylaxis (12, 13). Two peptides with anti-inflammatory and antishock activities were identified from SMG extracts and are cleavage products of the prohormone submandibular rat-1 (SMR1) (17).
The prohormone SMR1 can be cleaved to form at least five biologically active peptide products that have at least three distinct biological activities: analgesia, erectile function, and anti-inflammation (3, 5, 17, 33). The mature pentapeptide, sialorphin (QHNPR), is secreted into the bloodstream in response to acute stress and adrenergic stimulation (35) and is selectively taken up by peripheral targets (37). Sialorphin binds to membrane-anchored neutral endopeptidase (NEP) and competitively inhibits the enzymatic breakdown of substance P and Met-enkephalin, thus perpetuating the activation of μ- and δ-opioid receptors and producing analgesia (33). Additionally, sialorphin regulates sexual behavior (23) and erectile function in male rats. The Vcsa1 gene that encodes SMR1 was one of the most downregulated genes in the corpora of rats in three distinct models of erectile dysfunction, and gene transfer of plasmids expressing Vcsa1 or injection of sialorphin into the corpora of aging rats restored erectile function (3).
Two different regions of the SMR1 protein (SGEGV near the NH2 terminus and TDIFEGG near the COOH terminus) have anti-inflammatory activities (5, 17, 21). TDIFEGG (submandibular gland peptide-T or SGP-T), FEG, and metabolically stable feG are anti-inflammatory in all species studied to date, including mice, rats, cats, and sheep, and in human neutrophils in models of allergic pulmonary inflammation (5), spinal cord injury (10), and acute pancreatitis (28). Additionally, these peptides protect against endotoxic shock and anaphylaxis (14, 15, 17, 18, 22, 25, 39). Although the exact mechanism of action of TDIFEGG is not fully understood, it involves prevention of the influx of inflammatory cells to sites of inflammation. SGP-T decreases the migration and adhesion of neutrophils (5, 16, 19, 25) and the ability of neutrophils to generate intracellular reactive oxygen species (20).
Although there is no human homolog of SMR1, some of the protein's biological activities have recently been ascribed to human members of the variable coding sequence (VCS) gene family to which Vcsa1, which encodes SMR1, belongs. The human gene PROL1 encodes a protein that contains opiorphin (QRFSR), a peptide that inhibits two human enkephalin-catabolizing ectoenzymes and has analgesic activity analogous to rat sialorphin (32, 41). Opiorphin was isolated from human saliva and is predicted to be the product of cleavage events similar to those that liberate rat sialorphin (32, 41). The human PROL5/SMR3A gene, another VCS family member, has activity analogous to rat Vcsa1 in erectile function. There is a >10-fold decrease in hSMR3A transcripts in the corpora of human patients with erectile dysfunction (38), and gene transfer of hSMR3A into the corpora of aging rats restored erectile function to an extent similar to rat Vsca1 (38). These recent discoveries confirm that SMR1-like activities are conserved in related proteins in humans that are expressed in salivary glands and likely undergo regulation similar to that of SMR1.
Given the diverse biologically important functions of the peptide products of SMR1 and the evidence that homologous pathways exist in humans, it is of great significance to understand the regulation of the SMR1 protein and its role in biology and disease. Although the secretion of sialorphin has been studied, little is known about the expression, tissue distribution, processing, or secretion of the intact SMR1 protein or portions of the SMR1 protein containing the anti-inflammatory TDIFEGG peptide.
In this study, we describe the expression and posttranslational modification of SMR1 protein in the major salivary glands, urogenital tract, lung, saliva, and blood. Additionally, we demonstrate that secretion of TDIFEGG-containing species of the SMR1 protein from all major salivary glands is regulated by acute administration of sympathetic and parasympathetic mimetics and sympathetic ganglionectomy.
MATERIALS AND METHODS
New Zealand White rabbits (female, 8 wk old) were purchased from Vandermeer Rabbitry (Edmonton, AB, Canada). Male 10-wk-old Sprague-Dawley rats were from Charles River Laboratories (St-Constant, PQ, Canada), and male 10-wk-old Brown Norway rats were from Harlan Sprague-Dawley (Indianapolis, IN). Animals were maintained in the animal facilities at the University of Alberta or the University of Calgary on a 12:12-h light-dark cycle with access to water and food ad libitum. All studies were approved by the University of Alberta and University of Calgary animal ethics committees in accordance with guidelines of the Canadian Council for Animal Care. Ganglionectomy and sham surgeries were performed at the University of Calgary as described previously (27).
Antibody design and production.
Synthetic peptides that correspond to different regions of SMR1 conjugated to keyhole limpet hemocyanin (KLH) were synthesized (Alberta Peptide Institute, Edmonton, AB, Canada). Preimmune serum was collected from each rabbit before immunization. One hundred micrograms of peptide 216 or of each of peptides 219 (1–3) was diluted into 0.5 ml of PBS, mixed with 0.5 ml of Freund's complete adjuvant, and injected into four subcutaneous sites. Three boosts (100 μg) of each of the antigens in Freund's incomplete adjuvant were performed every 4 wk after the primary injection. Blood was collected 7 days after each injection, and the IgG fraction from serum from terminal bleeds was isolated with the Melon Gel IgG Purification Kit (Pierce, Rockford, IL) according to the manufacturer's directions.
Direct enzyme-linked immunosorbent assay.
For determination of antibody specificity, peptides (100 ng) were bound to a 96-well enzyme-linked immunosorbent assay (ELISA) plate (MaxiSorp, Nalge Nunc International, Rochester, NY), blocked in blocking buffer [1× Tris-buffered saline, 0.1% Tween 20 with 5% (wt/vol) nonfat dry milk], and incubated with immune serum from each rabbit diluted in blocking buffer (1:5,000) and then with secondary anti-rabbit horseradish peroxidase (HRP) (1:5,000; Invitrogen). For analysis of SMG and saliva samples, lysates or saliva were fractionated with Microcon YM-3,000 molecular weight exclusion filters (Millipore, Billerica, MA) according to the manufacturer's directions. Samples were serially diluted in bicarbonate/carbonate coating buffer (100 mM), bound to 96-well plates, blocked with 5% milk in PBS, and incubated with anti-SMR1(216) at 1:500 in PBS containing 0.05% Tween 20 and anti-rabbit HRP at 1:500 in PBS containing 0.05% Tween 20. Tetramethylbenzidine substrate was added for 30 min, the reaction was stopped (1 M H2SO4), and absorbance was read at 450 nm. Background was subtracted, and an absorbance of 1.0 was considered a positive signal.
Tissue preparation and deglycosylation.
Rats were euthanized, and tissues were rapidly excised and frozen on dry ice. Glands were homogenized in RIPA lysis buffer (150 mM NaCl, 50 mM Tris pH 8.0, 1% NP-40, 0.5% deoxycholic acid, 0.1% SDS) containing 1% protease inhibitor cocktail (Sigma, St. Louis, MO). Protein concentrations were determined with a bicinchoninic acid assay (Pierce) according to the manufacturer's directions. Denaturing deglycosylation reactions were performed with the PRO-LINK kit (Prozyme, San Leandro, CA) according to the manufacturer's directions.
Saliva and blood collection.
Rats were starved for 16 h to reduce variability in saliva production and anesthetized with intraperitoneal (IP) injections of ketamine (60 mg/kg) and xylazine (6 mg/kg). The β1- and β2-adrenergic receptor agonist isoproterenol (12.5 mg/kg) and the nonselective muscarinic receptor agonist pilocarpine (0.1 mg/kg) in physiological saline were administered by IP injection. Saliva was collected from the mouth with a pipette and placed into tubes on ice containing 5 μl of protease inhibitor cocktail. Cells and debris were removed by centrifugation, and saliva was frozen until further use. At indicated times, rats were killed by exsanguination and blood was collected into heparinized tubes. Plasma was separated and concentrated with Microcon YM-100,000 and YM-3,000 molecular weight exclusion filters (Millipore) according to the manufacturer's directions.
Gel electrophoresis and Western blotting.
For two-dimensional (2D) electrophoresis, proteins were precipitated with a 2D clean-up kit (Bio-Rad Laboratories, Mississauga ON, Canada) according to the manufacturer's directions, rehydrated, loaded onto 7-cm isoelectric focusing (IEF) strips [isoelectric point (pI) 4–7] (Bio-Rad), and focused in a Protean IEF cell (Bio-Rad). Proteins were separated on 13.5% polyacrylamide gels and transferred to 0.45-μm nitrocellulose membranes. Prestained protein standards (Bio-Rad) were run on each gel. Membranes were blocked with Odyssey blocking buffer (Li-Cor Biosciences, Lincoln, NE). Primary antibodies were preimmune serum from rabbits 216 or 219 (1:1,000) anti-rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA; 1:1,000), IgG-purified anti-SMR1 216 (1:1,000; 2.8 μg/ml), IgG-purified anti-SMR1 219 (1:1,000; 2.8 μg/ml), anti-phospho-threonine (Santa Cruz Biotechnology, sc-5267; 1:200), anti-renin (H105) (Santa Cruz Biotechnology, sc-22752; 1:150), and anti-prekallikrein/kallikrein (Abcam, Cambridge, MA, ab1006; 1:1,000). Secondary antibodies were goat anti-rabbit 800 (Rockland Immunochemicals, Gilbertsville, PA; 1:10,000) and goat anti-mouse 680 (Invitrogen Canada, Burlington, ON, Canada; 1:10,000). Blots were visualized with an Odyssey imager (Li-Cor) by scanning simultaneously at 700 and 800 nm. Odyssey software was used for molecular weight determination and quantitation of Western blots.
To quantitate SMR1 protein levels, 15 μg of total protein from a single saliva sample designated to be the standard for all studies was loaded on each gel and compared with all saliva fractions and lysates from each tissue. The most prominent bands that were immunoreactive with anti-SMR1(216) (12, 16, 19, 21, and 23 kDa) were quantitated, and their intensities were totaled. The intensity of each sample was divided by the intensity of the standard sample, which was assigned a value of 1. On the basis of the protein concentration of the lysate and the mass of tissue homogenized relative to the mass of the whole tissue, the amount of each tissue equal to 1 unit of SMR1 (the intensity of the 15 μg of saliva standard) was calculated. One unit is approximately equal to the amount of SMR1 in 0.45 mg of SMG (0.15% of the gland), 0.55 mg of sublingual gland (SLG) (1.1% of the gland), or 0.54 mg of parotid gland (PG) (0.54% of the gland) from an average untreated Sprague-Dawley rat.
RNA was collected, and contaminating DNA was removed with the RNAqueous-4PCR kit (Ambion, Austin, TX) according to the manufacturer's directions. Reverse transcription was performed with Superscript II (Invitrogen), oligo(dT) (Invitrogen), and RNaseout (Invitrogen) according to the manufacturer's directions. Negative controls lacking RNA or Superscript II were performed with each reverse transcription reaction. PCR was performed on a Rotor-Gene machine (Corbett Life Science, Sydney, Australia). Each reaction contained each primer at 0.2 M, 100 ng of cDNA from the reverse transcription reaction, and SYBR Green PCR master mixture (Invitrogen). Standard curves were generated for each primer pair, and the slope and efficiency calculated from the curves were used to determine SMR1 mRNA levels relative to the housekeeping gene cyclophilin A. Melting curves were performed with each analysis to confirm the presence of a single band, and amplified products were run out in 1% agarose and sequenced to confirm that the product was SMR1. An annealing temperature of 58°C was used for all primer pairs. Primers for PCR were as follows: SMR1 GTTCTTGTTAATCTTCCCGGTTT (forward) and TTTTGGGTAGTCGCAGTAGTGTT (reverse) and cyclophilin A CACCGTGTTCTTCGACATCAC (forward) and CCAGTGCTCAGAGCTCGAAAG (reverse).
Computational and statistical analyses.
Predictions of glycosylation sites were performed with the NetNGlyc server (http://www.cbs.dtu.dk/services/NetNglyc/) (9a) and the NetOGlyc server (http://www.cbs.dtu.dk/services/NetOGlyc/) (11). Prediction of phosphorylation sites was performed with the NetPhos server (http://www.cbs.dtu.dk/services/NetPhos/) (2). The expression of SMR1 in lung was predicted based on data deposited in the Gene Expression Omnibus (National Institutes of Health) (http://www.ncbi.nlm.nih.gov/geo/gds/gds_browse.cgi?gds=64) (6). Western blots were digitized with UN-SCAN-IT Gel Digitizing Software v.6.1 (Silk Scientific). Results are presented as means ± SD or means ± SE for deglycosylation experiments. Means were compared with a two-tailed t-test or one-way ANOVA followed by a Newman-Keuls multiple-comparison test using a 95% confidence interval.
RESULTS AND DISCUSSION
Production of antibodies to SMR1.
To evaluate the expression and processing of the SMR1 protein and in particular the anti-inflammatory TDIFEGG portion of the protein, we produced polyclonal antibodies by immunizing rabbits with synthetic peptides that correspond to distinct sequences in SMR1 (Fig. 1A). The immunoreactivity and specificity of the antibodies was evaluated with a semiquantitative ELISA. Serum from rabbit 216 that was immunized with the anti-inflammatory TDIFEGG sequence was only immunoreactive with the same synthetic peptide (Fig. 1B). Serum from rabbit 219 was immunoreactive with only one of the three peptides used in the immunization and showed no cross-reactivity (Fig. 1B). Antibodies consisting of IgG purified fractions of the serum from 216 and 219 were used at 2.8 ng/ml for Western blotting in subsequent experiments.
Characterization of SMR1 expression and posttranslational modification in rat submandibular glands.
SMR1 mRNA is expressed at high levels in the SMG of male rats (30), and several peptide products of the SMR1 protein are secreted from the SMG into saliva and blood (35). We evaluated SMR1 protein expression in SMG with our novel antibodies. Proteins from SMG of male Sprague-Dawley or Brown Norway rats were separated by SDS-PAGE and blotted with anti-SMR1 antibodies. SMR1 protein was detected in several bands that ranged in molecular mass from 12 to 35 kDa, with the most strongly immunoreactive species at 19, 21, and 23 kDa (Fig. 2A). Anti-SMR1(216) and anti-SMR1(219) antibodies produced similar banding patterns at molecular mass of 12, 16–19, 21–23, and 30–35 kDa when blots were overexposed (not shown). Anti-SMR1(216) consistently showed greater immunoreactivity with bands at molecular mass of 16, while anti-SMR1(219) consistently showed greater immunoreactivity with bands at molecular mass of 12 (Fig. 2A). Normal anti-rabbit IgG antibody and preimmune serum taken from rabbits 216 and 219 did not have immunoreactivity with rat SMG proteins, demonstrating that the immunostaining was specific (Fig. 2A). We also compared our antibodies with an antibody produced by C. Rougeot (Laboratoire de Pharmacologie des régulations neuroendocrines, Institut Pasteur, Paris, France; firstname.lastname@example.org) and used previously (40). This antibody was produced with the SNPPTQLLSTEAQANTK peptide sequence [similar to anti-SMR1 219 (peptide 3), Fig. 1], and it produced a banding pattern on Western blots indistinguishable from that of anti-SMR1(219) (not shown). The pattern of immunoreactive proteins was not different between Sprague-Dawley and Brown Norway rats (Fig. 2A). There was a small amount of interrat variation in the relative abundance and intensity of SMR1 bands, but all rats evaluated showed a similar pattern of bands and similar expression level in the SMG (n = 12) (Fig. 2D).
To further characterize the many protein species detected, 2D SDS-PAGE was used to separate SMG proteins and Western blots were performed. Anti-SMR1(216) detected at least 36 spots with molecular mass of 12–25 kDa and pI ranging from 4 to 7 (Fig. 2B). Anti-SMR1(219) detected the same protein pattern as anti-SMR1(216) (not shown), whereas normal anti-rabbit IgG and preimmune serum did not detect any protein spots under the same conditions (not shown).
Full-length SMR1 has a predicted molecular mass of 16 kDa and a reported 12-kDa cleavage product extending from Arg70 to the COOH terminus (36). The presence of higher-molecular mass bands in addition to bands at 16 and 12 kDa suggests that SMR1 is posttranslationally modified. SMR1 has two potential N-linked glycosylation sites at residues Asn129 and Asn136 (9a, 29) and several potential O-linked glycosylation sites (11) in addition to several threonine residues that could be phosphorylated (2). There was no significant overlap between anti-phospho-threonine antibodies and anti-SMR1 antibodies when analyzed with simultaneous two-color scanning (not shown), indicating that SMR1 is likely not constitutively phosphorylated in rat SMG.
We evaluated the effect of treatment with N-glycanase, O-glycanase, and Sialidase A on the banding pattern of SMR1 protein. Incubation with N-glycanase, but not O-glycanase or Sialidase A, decreased the apparent molecular mass of a large proportion of the SMR1 protein (Fig. 2C). After treatment with N-glycanase, the 23- and 21-kDa bands accounted for only 4% and 7% of total SMR1, respectively, while 45% and 15% of the SMR1 protein migrated near the predicted molecular mass of 16 kDa and at the predicted fragment mass of 12 kDa (Fig. 2, C and D). These results suggest that full-length and cleaved SMR1 are N-glycosylated in the SMG. These modifications may account for many of the different protein species observed on 2D Western blots.
We were unable to perform mass spectrometry analysis to identify the different SMR1 protein species because SMR1 was too low in abundance relative to other proteins of similar molecular mass to be identified on silver-stained gels, and our antibodies were not able to immunoprecipitate sufficient quantities of SMR1 protein. We believe that it is unlikely that a protein other than SMR1 would be recognized by two different polyclonal SMR1 antibodies and not recognized by control antibodies. Therefore, we suggest that bands detected by both anti-SMR1(216) and anti-SMR1(219) and not detected by preimmune serum controls are almost certainly species of the SMR1 protein. Since anti-SMR1(216) and anti-SMR1(219) consistently detect the same pattern of bands on Western blots, we present blots produced with the anti-SMR1(216) antibody that recognizes epitopes in the anti-inflammatory TDIFEGG sequence.
SMR1 is expressed in sublingual and parotid glands.
There have been no reports of SMR1 protein expression in SLG or PG, so we looked for SMR1 expression in these other major salivary glands. Anti-SMR1(216) detected immunoreactive bands in rat SLG in a banding pattern and intensity similar to that in SMG (Fig. 3, A and C). After treatment with N-glycanase, the majority of SLG SMR1 protein migrated at 16 and 12 kDa (Fig. 3C). SMR1 immunoreactivity was also detected in rat PG (Fig. 3, A and C). SMR1 in PG had a different banding pattern, with the major species (55%) at 21 kDa and weaker bands at 26, 28, and 12 kDa (Fig. 3, A and C). Treatment with N-glycanase caused a reduction in the amount of SMR1 at 26 and 21 kDa and a concurrent increase in the amount of PG SMR1 at 16 kDa from 3% to 30% (Fig. 3C). Anti-SMR1(219) showed the same banding patterns as anti-SMR1(216), and preimmune serum showed no immunoreactivity for SLG and PG proteins (not shown). O-glycanase and Sialidase A did not alter the migration of SLG or PG SMR1 (not shown). Anti-SMR1(216) detected similar banding patterns for protein from SLG and PG of Brown Norway rats (not shown).
SMR1 is expressed in urogenital tissues and lung in male rats.
SMR1 mRNA expression in the prostate and penis (31, 40) and protein expression in the penis (40) have been described previously. We investigated the expression of SMR1 in male urogenital tract tissues, where the protein is known to have a role in sexual behavior and erectile function. The banding pattern observed for SMR1 immunoreactivity in prostate glands was somewhat similar to that in SMG and SLG, with major bands at 23, 21, and 19 kDa and a weak band at 16 kDa, but additional bands at ∼35 and ∼25 kDa were observed. This suggests that some similar posttranslational modifications of the protein may occur in salivary and prostate glands. SMR1 was also expressed in penile tissue as reported previously (40). In accordance with previous reports, SMR1 immunoreactivity was detected at higher molecular masses in penile tissue than observed in salivary glands, predominantly in bands at ∼25, 23, and 21 kDa and in weaker bands at 16 and 12 kDa (Fig. 3, A and C) (40). N-glycanase treatment resulted in a decrease in SMR1 immunoreactivity at 23 and 21 kDa and an increase in immunoreactivity at 16 kDa (Fig. 3C). SMR1 was also expressed in testis. Anti-SMR1(216) detected a variety of bands in testis of molecular mass 12, 16, 21, 25, 27, 33, 41, and 65 kDa, with the most prominent species occurring at 25 kDa (Fig. 3A). Anti-SMR1(219) detected bands at 12, 16, 21, 25, and 41 kDa (not shown). N-glycanase treatment caused a modest increase in the intensity of the band at 12 kDa and a modest decrease in the intensity of the band at 25 kDa (Fig. 3C).
The TDIFEGG peptide at the COOH terminus of SMR1 decreases pulmonary inflammation in models of allergy (4, 5). Recent microarray data indicate that SMR1 mRNA may be expressed in the lung (6). We used our antibodies to determine whether SMR1 protein is detectable in rat lung tissue. SMR1 immunoreactivity was detected in the lungs of male rats at levels less than those in salivary glands but similar to levels in penis and testis (Fig. 3, A and B). The most highly immunoreactive band detected by anti-SMR1(216) was at 26 kDa, and other immunoreactive bands were detected at 12, 23, 32, 35, and 42 kDa. Anti-SMR1(219) also detected bands in lung at 12, 23, 26, 32, 35, and 42 kDa (not shown). Preimmune serum was not immunoreactive with any bands in rat lung (not shown). N-glycanase did not change the banding pattern or relative abundance of any bands detected with anti-SMR1 antibodies (Fig. 3C).
SMR1 protein is present at comparable levels in SMG and SLG and slightly lower levels in PG, where it is N-glycosylated (Fig. 3, B and C). SMR1 is also present, but at lower levels, in the prostate, penis, testis, and lung and is N-glycanase sensitive to some extent in all tissues except lung (Fig. 3, B and C). In some tissues, bands were detected at high molecular masses. One explanation for this is that SMR1 may be aggregated or bound to other proteins, despite the denaturing and reducing conditions used in Western blotting. We attribute the high-molecular-mass bands in testis and lung tissue to SMR1, since both antibodies to different regions of the protein showed a very similar banding pattern and detected bands at the same molecular masses.
SMR1 is constitutively present in plasma.
We investigated whether the SMR1 protein is present in the bloodstream of rats under resting conditions. Blood was collected by cardiac puncture, and plasma was isolated. SMR1 protein could not be detected in whole plasma containing 400 μg of total protein, the maximum amount that could be loaded on a gel. However, when the 3- to 100-kDa fraction of plasma was concentrated 10-fold, SMR1 immunoreactivity was detected on Western blots (Fig. 3A). Anti-SMR1(216) was most immunoreactive, with a band at 26 kDa and more weakly immunoreactive with a band at 12 kDa (Fig. 3, A and C). A large band at 60–70 kDa was detected with anti-SMR1(216) and preimmune serum and may have been nonspecific (Fig. 3A). Treatment with N-glycanase did not change the migration of plasma SMR1, indicating that SMR1 in the bloodstream is N-glycanase insensitive (Fig. 3C).
Surprisingly, the apparent molecular mass and N-glycanase sensitivity of the majority of SMR1 in blood were different from those in salivary glands. This could suggest that the salivary glands are not the major source of circulating SMR1, or it could suggest that circulating SMR1 is immediately processed or complexed with other proteins. The N-glycanase insensitivity could be a result of the protein not being N-glycosylated, or it may be protected from digestion by glycanase enzymes by virtue of its conformation or interactions with other proteins.
SMR1 is secreted into saliva in response to autonomic stimulation.
Sialorphin is secreted into the bloodstream and saliva in response to acute stress and adrenergic stimulation (35). However, secretion of the SMR1 protein and fragments of the protein that contain the COOH-terminal anti-inflammatory TDIFEGG sequence has not been investigated. In light of our observations that SMR1 is extensively posttranslationally modified, we wanted to determine whether there is selective secretion of particular SMR1 protein species. We compared the effects of acute adrenergic and cholinergic stimulation on secretion of SMR1 into saliva and determined whether there was differential secretion of SMR1 species. Anesthetized male Sprague-Dawley rats were administered saline, the sympathomimetic β1- and β2-adrenergic agonist isoproterenol (12.5 mg/kg), or the parasympathomimetic nonspecific cholinergic agonist pilocarpine (0.1 mg/kg) by IP injection, and saliva was collected for 1 h. The saliva produced in response to each autonomimetic (a combination of secretions from all salivary glands) was evaluated for SMR1 expression.
Rats treated with saline did not produce a measurable quantity of saliva. Rats treated with isoproterenol produced 8.0 ± 1.4 μl/min of saliva containing 488 ± 82 μg of protein (n = 9) (Fig. 4, A and B). Seven of nine rats treated with pilocarpine produced 18.2 ± 5.9 μl/min of saliva containing 31 ± 10 μg of protein (Fig. 4, A and B). Two rats treated with pilocarpine produced much less saliva than the other seven and were not included in the analyses (not shown). Thus, on average, rats treated with isoproterenol produced approximately half the volume of saliva containing 16-fold more protein than rats treated with pilocarpine. In response to both agonists, saliva production began 3–10 min after injection, was maximal between 10 and 30 min, and slowed to nearly negligible secretion 60 min after injection (Fig. 4A) even though the rats were hydrated with physiological saline to compensate for lost volume.
Western blots demonstrated that saliva elicited by both isoproterenol and pilocarpine contained SMR1 protein (Fig. 4, C–E). Rats stimulated with isoproterenol secreted 46-fold more SMR1 protein into saliva than rats treated with pilocarpine during the 60 min of secretion (Fig. 4C). Isoproterenol also elicited ∼2.8-fold more SMR1 secretion per mass of total protein than pilocarpine (Fig. 4D). One-dimensional (1D) and 2D Western blots with anti-SMR1(216) indicated that saliva contained multiple species of SMR1 protein between pI 4 and 7 and between molecular mass 12 and 25 kDa in similar proportion to the pattern in SMG (Fig. 4, D–G). Saliva collected at all times from 5 min to 60 min and elicited by either agonist contained similar ratios of the different species of SMR1 protein as evaluated by 2D Western blots (not shown). Treatment of saliva with N-glycanase caused the majority of SMR1 to migrate at 16 and 12 kDa (Fig. 4, E and F). These results suggest that full-length, fully glycosylated SMR1 protein is secreted into saliva in significant quantities, and that there is little selectivity in the secretion of SMR1 species over the range of molecular mass detectable with these methods.
We also looked for the presence of the salivary gland proteins kallikrein and renin in saliva by Western blot. In contrast to SMR1, kallikrein secretion was proportional to total protein secretion. That is, isoproterenol caused 16-fold more protein secretion and 16-fold more kallikrein secretion than pilocarpine (not shown). Isoproterenol caused 300-fold more renin secretion into saliva than pilocarpine (not shown). These results suggest that the secretion of SMR1, kallikrein, and renin may be regulated by different mechanisms. We collected total saliva, which was the additive product of secretions from all salivary glands, and the proteins studied may not have been secreted by each gland in the same proportion. SMR1 from isoproterenol-elicited saliva was strongly detected on Western blots in 0.3 μl of saliva; however, 5 μl of saliva was required for equivalent detection of renin or kallikrein. Although this difference could be partly due to differences in antibody affinities, it also suggests that a relatively large amount of SMR1 is secreted into saliva compared with other secreted proteins of biological significance. The large amount of SMR1 secreted into saliva may be further processed by the multitude of proteolytic enzymes secreted into saliva in parallel, or may function as an intact protein to modulate pain perception and inflammation following stressful events. SMR1 could act locally to modulate inflammation in the oral cavity or gastrointestinal tract or on tissues directly exposed to saliva (e.g., licked wounds). Additionally, fragments of SMR1 could be reabsorbed and act systemically if sufficient quantities were swallowed.
In addition to measuring SMR1 protein secretion into saliva, we evaluated whether small peptide fragments derived from the COOH terminus of SMR1 were present in saliva by direct ELISA with the anti-SMR1(216) antibody, which allowed us to detect anti-SMR1(216) immunoreactivity nonquantitatively. With this assay, anti-SMR1(216), which recognizes epitopes within the TDIFEGG peptide, showed immunoreactivity toward synthetic TDIFEGGG peptides complexed with a 100-fold excess of BSA and toward SMG and saliva proteins. Control anti-rabbit IgG and preimmune serum were not immunoreactive, demonstrating the specificity of the ELISA (not shown). We were able to obtain a linear relationship over a limited range with plated sample or synthetic peptide and immunoreactivity for all samples; however, we were unable to establish reproducible quantitation with this method, so we present results as positive or negative for immunoreactivity.
All samples containing proteins >3 kDa molecular mass from saliva or SMG (including full-length SMR1) were positive for SMR1 immunoreactivity (Table 1). We also detected immunoreactivity with anti-SMR1(216) in saliva fractions containing peptides <3 kDa from 11 of 13 isoproterenol-stimulated rats and 10 of 13 pilocarpine-stimulated rats (Table 1). When Western blots were performed, no immunoreactivity could be detected above 10 kDa in the <3 kDa molecular mass filtrate fractions (not shown). Therefore, we do not believe that contamination with full-size SMR1 can account for the immunoreactivity detected in these low-molecular-mass fractions. These results indicate that saliva from the majority of rats contains peptides smaller than 3 kDa that contain epitopes in the TDIFEGG peptide. These are likely to be biologically active as anti-inflammatory agents.
We also assayed for the presence of small peptide fragments of SMR1 in SMG lysates to determine whether liberation of the small peptides occurs before or after secretion. Small peptide fragments of SMR1 were detected in the SMG of eight of nine uninjected rats (Table 1). This suggests that SMR1 is cleaved into fragments smaller than 3 kDa in the gland in unstimulated conditions. We were not able to detect small SMR1-derived peptides in SMG from isoproterenol-treated rats (Table 1), suggesting that all peptide fragments from these SMG were secreted in the hour after isoproterenol injection. Small peptide fragments of SMR1 were present in ∼50% of pilocarpine-treated rats (Table 1). Interestingly, the five rats that showed an absence of immunoreactive small peptides in saliva also showed an absence of immunoreactive small peptides in SMG. Our results suggest that some small peptides containing epitopes in the TDIFEGG motif are generated from SMR1 in the SMG and subsequently secreted under sympathetic nervous system control. It is possible that SMR1 is cleaved both in salivary glands and in the saliva after secretion, but since we were unable to quantify the small peptides with our techniques we cannot comment on the relative contribution of each pathway.
There are no detectable changes in SMR1 levels in blood after autonomic stimulation.
The major salivary glands of mammals secrete proteins in both an exocrine (into saliva) and an endocrine (into bloodstream) manner. It has been demonstrated that sialorphin, a mature peptide product of SMR1, is secreted in an endocrine fashion from SMG directly into the bloodstream in response to acute stress (35). We were unable to detect changes in the amount of SMR1 in the bloodstream 30 or 60 min after isoproterenol or pilocarpine treatment (not shown). Since we collected systemic blood by cardiac puncture 30 or 60 min after stimulation, if local endocrine release and subsequent uptake of SMR1 occurred it may not have been detected by our methods. It is also possible that the intact prohormone is not secreted into the bloodstream after sympathetic stimulation but its smaller peptide products are (35). We did not detect any immunoreactive peptides <3 kDa in plasma with semiquantitative ELISAs; however, if TDIFEGG was bound to carrier proteins or aggregated, it may not be present in the <3 kDa fraction. More comprehensive studies will be required to determine whether there is endocrine secretion of SMR1 or TDIFEGG.
SMR1 protein is depleted from salivary glands after treatment with isoproterenol.
Rats were injected with isoproterenol, pilocarpine, or saline, and tissues were harvested after 30 and 60 min. Injection of either autonomimetic did not change the mass of any of the salivary glands or the amount of total protein per gland mass (not shown). There was a 70% decrease in the amount of SMR1 protein in the SMG 60 min after isoproterenol treatment as determined by quantitative Western blot (Fig. 5, A and B). SMR1 protein levels were also decreased by 45% in SLG and by 80% in PG 60 min after isoproterenol injection (Fig. 5, C and D). Pilocarpine injection did not change the amount of SMR1 in SMG, SLG, or PG (Fig. 5). The decrease in SMR1 protein in the glands after isoproterenol treatment is likely the result of the robust secretion of the protein into saliva. Although isoproterenol caused the secretion of kallikrein and renin into saliva, there was no detectable change in the amount of kallikrein or renin in SMG (Fig. 5A) or other salivary glands (not shown) 60 min after isoproterenol injection. If kallikrein and renin are more abundant proteins than SMR1, a decrease in their protein levels may not be detected even if equivalent amounts of all proteins were secreted.
We have demonstrated that secretion of SMR1 in response to sympathetic stimulation is not restricted to SMG of rats. SLG and PG also express significant levels of SMR1 protein and release a significant proportion of their stored prohormone into saliva in response to autonomimetics. On the basis of estimates from quantitative Western blots, there is approximately the same amount of SMR1 protein per total protein in SMG and SLG and ∼75% as much in PG. Taking into account the average masses of the glands (∼300 mg, ∼50 mg, and ∼100 mg for SMG, SLG, and PG, respectively), we estimate that a SMG contains 7.5-fold more SMR1 protein than a SLG and 3.5-fold more SMR1 protein than a PG. Given that 70%, 45%, and 80% of the SMR1 protein from SMG, SLG, and PG, respectively, are secreted after sympathetic stimulation, we predict that the SMR1 in saliva would be derived 71% from SMG, 6% from SLG, and 23% from PG. Therefore, it is not surprising that the banding pattern for SMR1 in saliva most closely resembles the banding pattern of SMR1 in SMG (Fig. 4, D and E). We compared the total amount of SMR1 protein that was collected in saliva over 60 min to the amount of SMR1 protein present in each salivary gland. On the basis of the decrease in SMR1 in glands after isoproterenol treatment, we would expect that 65% of the total SMR1 in the major salivary glands would have been secreted. Our saliva collection methods were crude (collecting saliva from the oral cavity by pipette), so a significant proportion of saliva was likely not collected. In fact, we estimate from quantitative Western blots that we recovered ∼30% of the SMR1 that we predict was secreted. We recovered ∼20% of the total amount of SMR1 protein in salivary glands in saliva after isoproterenol treatment and ∼0.4% of total glandular SMR1 in saliva after pilocarpine treatment. Given that 20–65% of the total SMR1 in salivary glands enters saliva in response to adrenergic stimulation and only 0.4–2% of the total SMR1 in salivary glands enters saliva in response to cholinergic stimulation, it is evident that sympathetic nervous system activation has significantly more potential than the parasympathetic nervous system to regulate systemic responses through the release of SMR1 and its fragments.
We cannot speculate on the relative contribution of each salivary gland to the kallikrein or renin released into saliva, since there were no detectable decreases in the amounts of these proteins in any glands after the injection of autonomimetics, possibly because of the higher abundance of these proteins.
Superior cervical ganglionectomy causes accumulation of SMR1 in salivary glands.
Bilateral decentralization or removal of the superior cervical ganglion attenuates allergen-induced pulmonary inflammation in sensitized rats (27). This effect is mediated by SMG anti-inflammatory factors, since prior removal of the SMG abolishes this effect (13). We hypothesize that SMR1 is one anti-inflammatory factor that is regulated by the superior cervical ganglion in this pathway. To determine whether innervation of the salivary glands by the sympathetic nervous system regulates SMR1 expression or processing, we performed superior cervical ganglionectomy or sham surgery on male rats. Tissues were harvested 8, 9, or 10 days after surgery and analyzed by Western blot. Rats in which the sympathetic ganglia were removed had 2-fold and 1.5-fold more SMR1 in SMG and PG, respectively, than sham-operated animals (Fig. 6). The increase in SMR1 protein was not the result of increased transcription, since real-time RT-PCR demonstrated that ganglionectomy did not increase SMR1 mRNA in SMG or PG compared with the housekeeping gene cyclophilin A (not shown). The increase in SMR1 protein in ganglionectomized rats was likely due to decreased secretion of the protein in response to tonic sympathetic stimulation or normal intermittent stress responses over the period of time that the ganglion was removed. The relative abundance of different species of SMR1 was not changed in ganglionectomized rats (Fig. 6A). Isoproterenol injection (8, 9, or 10 days after surgery) caused a similar degree of SMR1 secretion into saliva in sham-operated and ganglionectomized rats (not shown). This was expected, since isoproterenol would directly activate adrenergic receptors on the salivary glands, bypassing the requirement for an intact sympathetic nerve.
In conclusion, we report the expression of SMR1 protein in all major salivary glands, urogenital tissues, and lungs in male rats and demonstrate that SMR1 is extensively N-glycosylated in glandular tissues and differentially posttranslationally modified. Injection of autonomimetics stimulated the nonselective secretion of SMR1 from the major salivary glands into saliva. In response to β1- and β2-adrenergic stimulation, the majority of SMR1 in salivary glands was secreted into saliva in 60 min. We suggest that the secretion of SMR1 into saliva in response to sympathetic stimulation may provide a large pool of SMR1-derived peptide products that can produce analgesic and anti-inflammatory responses locally and systemically. This may constitute an important pathway by which mammals respond to acute stress. Additionally, factors such as emotions, drugs, diseases, and environmental changes regulate the sympathetic nervous system, and this pathway may be involved in systemic responses to a variety of physiological insults. It is likely that other mammals including humans utilize an analogous pathway, and that the protein products of human VCS gene family members are regulated similarly. This pathway may modulate pain responses and inflammation in humans. The role of similar pathways in other animals and the significance of the differential regulation of SMR1 and its expression in a variety of tissues warrant future investigation.
This work was funded by AllerGen NCE (project no. 07-B3D). Salary support for K. E. Morris, R. S. Hoeve, and P. Forsythe was provided by the Alberta Heritage Foundation for Medical Research. Salary support for R. D. Mathison, C. D. St. Laurent, and A. D. Befus was from AllerGen NCE, the Canadian Institutes of Health Research, and the AstraZeneca Canada Inc. Chair in Asthma Research, respectively.
The contributions of David Kirk at the University of Calgary, who performed the surgeries, and Sam Harirforoosh and Yokananth Sekar at the University of Alberta, who optimized the 2D gel systems, are gratefully appreciated. We are grateful to C. Rougeot at the Institut Pasteur for her gift of an anti-SMR1 antibody.
Present address of P. Forsythe: Dept. of Pathology and Molecular Medicine, Brain-Body Institute, McMaster University, Hamilton, ON, Canada L8N 4A6.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2009 the American Physiological Society