Intracellular Peptides as Natural Regulators of Cell Signaling*

Protein degradation by the ubiquitin proteasome system releases large amounts of oligopeptides within cells. To investigate possible functions for these intracellularly generated oligopeptides, we fused them to a cationic transactivator peptide sequence using reversible disulfide bonds, introduced them into cells, and analyzed their effect on G protein-coupled receptor (GPCR) signal transduction. A mixture containing four of these peptides (20–80 μm) significantly inhibited the increase in the extracellular acidification response triggered by angiotensin II (ang II) in CHO-S cells transfected with the ang II type 1 receptor (AT1R-CHO-S). Subsequently, either alone or in a mixture, these peptides increased luciferase gene transcription in AT1R CHO-S cells stimulated with ang II and in HEK293 cells treated with isoproterenol. These peptides without transactivator failed to affect GPCR cellular responses. All four functional peptides were shown in vitro to competitively inhibit the degradation of a synthetic substrate by thimet oligopeptidase. Overexpression of thimet oligopeptidase in both CHO-S and HEK293 cells was sufficient to reduce luciferase activation triggered by a specific GPCR agonist. Moreover, using individual peptides as baits in affinity columns, several proteins involved in GPCR signaling were identified, including α-adaptin A and dynamin 1. These results suggest that before their complete degradation, intracellular peptides similar to those generated by proteasomes can actively affect cell signaling, probably representing additional bioactive molecules within cells.

Intracellular protein turnover is a crucial process for normal cell function; an excess of aged proteins usually leads to the formation of insoluble aggregates within the cell, causing severe pathologies (1). The concomitant action of proteasomes and other extralysosomal proteolytic systems (2,3) at different intracellular locations suggests that there is a continuous formation and release of free peptides within eukaryotic cells (4,5). The proteasome exists as a 2.4-MDa ATP-dependent complex containing proteolytic subunits that possess cleavage activity specific for hydrophobic, basic, and acidic amino acids (6). As expected from such a broad catalytic specificity, protein degradation by the proteasome produces oligopeptides containing 2-20 amino acids within cell nuclei and the cytosol (7). Although some of the peptides produced by the proteasome are known to escape degradation and are presented at the cell surface as antigens, the majority of these proteasomal products are thought to be rapidly broken down further into free amino acids (4,8).
However, the demonstration that free peptides added to the intracellular milieu can regulate cellular functions mediated by protein interactions suggests new putative roles for these molecules in gene regulation, metabolism, cell signaling, and protein targeting. Such interactions frequently involve specific consensus amino acid sequences that can be predicted based on similarities in domain composition. For example, a rationally designed peptide of eight residues was shown to inhibit ␦-protein kinase C (PKC) 4 translocation and function, as evidenced by its protective effect in a model of myocardial infarction (9). The c-Jun NH 2 -terminal kinase, a member of the stress-activated group of mitogen-activated protein kinases, is inhibited by a cell membrane-permeable peptide that decreases intracellular c-Jun NH 2 -terminal kinase signaling and confers long term protection to pancreatic ␤-cells against IL-1␤-induced apoptosis (10). Therefore, we have previously suggested that the natural peptides generated by proteasomes could interfere with intracellular protein interactions before their complete degradation, thereby affecting cell signaling, gene regulation, metabolism, and protein targeting (11).
Thimet oligopeptidase (EC 3.4.24.15; EP24.15) was first described as a neuropeptide-degrading enzyme present in the soluble fraction of rat brain homogenates (12,13). The primarily intracellular location (e.g. cytosolic and nuclear) of oligopeptidase EP24. 15 suggests that this enzyme has other functions in addition to degrading extracellular neuropeptides and hormones (11,14). EP24. 15 substrate specificity has been shown to depend predominantly on the peptide size with no clear preference for amino acid sequence (15)(16)(17)(18). Interestingly, peptides released by the proteasome (19 -21) are in the same size range of EP24.15 substrates (16,22,23). In fact, EP24.15 has recently been shown to play an intracellular role in either degrading or protecting major histocompatibility complex class I antigenic peptides produced by the proteasome (24 -27). To further explore the role of oligopeptidases in the metabolism of intracellular peptides, we recently developed an experimental approach employing a recombinant, catalytically inactive mutant form of EP24.15 to "capture" endogenous peptide substrates of this enzyme (22). The smallest peptide isolated in these assays contained 5 amino acids, and the largest had 16 amino acids (11,22,23), which is within the size range previously reported for natural and synthetic substrates of EP24.15 (15, 16, 28 -30). Although the peptides captured using this approach are mainly intracellular protein fragments (distinct from neuropeptides) that efficiently interacted (K i of ϳ1-10 M) with EP24.15 without being degraded, some others are typical substrates generally cleaved at more than one peptide bond to produce multiple shorter peptides (22). Additionally, the majority of the peptides identified using the inactive EP24. 15 were shown to contain a putative protein kinase phosphorylation site (31). Actually, substrate phosphorylation changes the kinetic parameters of structurally related oligopeptidases, including EP24.15, neurolysin (EC 3.4.24. 16), and angiotensinconverting enzyme (EC 3.4.15.1) (31). Phosphorylation of peptides leads to reduced degradation, whereas phosphorylation of peptides that interact as competitive inhibitors of these enzymes alters only the K i values (31). These data suggest that substrate phosphorylation could be one of the mechanisms whereby some intracellular peptides can escape degradation, thus becoming available to interfere, for example, in cell signaling and protein network composition (11,31).
Regulatory regions in proteins are frequently produced in a cassette-like fashion from domains that mediate molecular interactions or have enzymatic activity. Several interaction domains are present in hundreds of copies in the human proteome, and these are used repeatedly to regulate distinct aspects of cellular organization. These interaction domains can target proteins to a specific subcellular location, provide a means of recognition for protein post-translational modifications or chemical second messengers, nucleate the formation of multiprotein signaling complexes, and control the conformation, activity, and substrate specificity of enzymes (32,33). In signal transduction, enzymes (e.g. kinases) often generate modified amino acids on their substrates that are then recognized by specific modules. The cell therefore uses various combinations of a limited set of interaction domains to direct the actions of complex regulatory systems. Given the broad repertoire of protein-protein interactions involved in cell signaling, the direct approach of interfering with signaling cascades is of great value for the understanding and control of cell function. Moreover, the majority of peptide-based drugs that have achieved marketplace status are of natural rather than combinatorial library origin (34), and peptides that alter intracellular protein interactions have yielded lead compounds with in vivo activity (34). Here, we have evaluated the intracellular effect of peptides isolated from rat brain homogenates using the inactive EP24.15 "substrate capture" assay that contains a PKC phosphorylation site. Our results show that such peptides can efficiently interfere with GPCR signal transduction in CHO-S cells expressing AT1 receptors (AT1R-CHO-S) stimulated with angiotensin II and in HEK293 cells stimulated with isoproterenol. Thus, these data suggest the existence of an entirely new group of bioactive molecules having pharmacological effects within cells.

EXPERIMENTAL PROCEDURES
Peptide Extraction-Crude peptide extracts from rat brains and CHO-S and HEK293 cells were prepared as previously described with some modifications (35). Briefly, male Wistar rats were killed by decapitation, and the entire head was subjected to 10 s of microwave radiation in order to inactivate protein and peptide degradation (36). Tissues from five rats were added to hot (70°C) 10 mM HCl and homogenized (Polytron; Brinkmann). Homogenates were maintained at 70°C for 15 min and centrifuged at 1,500 ϫ g for 40 min at 4°C. CHO-S or HEK293 cells (8 ϫ 10 7 cells) were first washed three times with phosphate-buffered saline by centrifugation at 800 ϫ g for 5 min. The pellet was resuspended in 10 ml of H 2 O and was heated in a water bath for 20 min at 80°C to inactivate proteases. After cooling on ice, 20 l of 5 M HCl was added to give a final concentration of 10 mM. The cells were sonicated three times with 20 pulses (4 Hz). The homogenate was centrifuged at 1,500 ϫ g for 40 min at 4°C. After this point, the respective supernatants from either tissue or cells were collected in plastic ultracentrifuge tubes and centrifuged at 100,000 ϫ g for 30 min at 4°C. The supernatant was again collected and filtered through a Millipore centrifugal filter unit with a molecular weight cut-off of 5,000 Da. Peptides contained in the samples were further purified and concentrated with C18-like Oasis columns (Waters) and dried in a vacuum centrifuge.
Peptide Quantification-Peptide concentration in the peptide extracts described above was determined at pH 6.8 using fluorescamine, as previously described (37,38). The reaction was performed at pH 6.8 to ensure that only the amino groups of peptides and not those of free amino acids reacted with fluorescamine (38). Briefly, 2.5 l of sample was mixed with 25 l of 0.2 M phosphate buffer (pH 6.8) and 12.5 l of a 0.3 mg/ml acetone fluorescamine solution. After vortexing for 1 min, 110 l of water was added, and fluorescence was measured with a SpectraMax M2 e plate reader (Molecular Devices) at an excitation wavelength of 370 nm and an emission wavelength of 480 nm. A peptide mixture of known composition and concentration was used as the standard reference for determining the peptide concentration in HEK293 extracts (data not shown).
Inactive EP24.15 "Substrate Capture" Assays-Mutated, catalytically inactive EP24.15 (E474A; 90 g) was incubated with 50 g of the rat brain peptide extracts as described above and in 200 l of buffer (25 mM Tris-HCl, pH 7.5, containing 125 mM NaCl and 0.1% of bovine serum albumin) for 30 min at room temperature. At the end of this period, the reaction mixture was layered onto a dried Sephadex G-25 column (previously washed and equilibrated with Tris-buffered saline followed by centrifugation to remove the buffer) and centrifuged at 1,000 ϫ g for 2 min. The flow-through (ϳ200 l) was collected, and the peptide content was analyzed by high performance liquid chromatography (HPLC) using a Chromolith performance column (4.6 ϫ 100 mm; Merck), having a linear gradient of 5-35% acetonitrile in 0.1% trifluoroacetic acid, for 20 min at a flow rate of 1 ml/min. Specific peptide peaks were manually collected according to previous reports (22,23) and identified as described below.
Peptide Sequencing and Bioinformatic Analysis-Peptides peaks manually collected as described above were sequenced by positive nanoelectrospray ionization using peptide-containing aliquots collected during HPLC. Typical conditions were a capillary voltage of 1 kV, a cone voltage of 30 V, and a dessolvation gas temperature of 100°C. The protonated peptides were subjected to collision-induced dissociation with argon in the 15-45 eV collision energy range. All of the mass spectrometry experiments were done with a Q-tof mass spectrometer (Micromass, UK) in a Qq-orthogonal time-of-flight configuration. Peptide sequences were determined manually from the ESI-MS/MS product ion mass spectra with the help of the Pep-Seq software package (Micromass, UK). To identify the putative protein precursors of the peptides sequenced by ESI-MS/ MS, a protein data base (available on the World Wide Web) was searched for short, nearly exact matches (rodentia origin), as previously described (39). The prediction of post-translational modification sites in sequenced peptides was done with the ExPASy Proteomics Server on the World Wide Web.
Peptide Selection and Synthesis-Peptides identified as described above were further selected for functional analysis on GPCR signal transduction based on their ability to both interact with EP24.15 (either as substrates or competitive inhibitors only) and on possessing a predicted PKC post-translational modification site. These peptides (supplemental Table 2) were synthesized with or without a fluorescein-labeled peptide (NH 2 -terminal) either coupled to the TAT peptide (CRKKRRQRRR) sequence through reversible disulfide bonds (40) or as a regular peptide without any labeling or additional modification (AnaSpec Inc., San Jose, CA).
EP24.15 Enzyme Assays-The enzymatic activity of EP24.15 was determined in triplicate in a continuous assay using the quenched fluorescent substrate (QFS) (7-methoxycoumarin-4acetyl-Pro-Leu-Gly-Pro-D-Lys-(2,4-dinitrophenyl), as previously described (41,42). Apparent inhibition constants (K i(app) ) of the regular synthesized peptides and determination of peptide bond cleavage were determined as previously described (22). Analysis of Peptide Entry and Stability within Cells-The entry of peptides into CHO-S cells was analyzed by HPLC coupled to a fluorescence detector. CHO-S cells grown under standard conditions (43) were incubated with the studied peptides (20 M) for 30 min at 37°C. Cells were then scraped from 6-well plates, suspended in 6 ml of phosphate-buffered saline, and centrifuged at 800 ϫ g for 10 min at 4°C. The cell pellet was resuspended in milli-Q water plus protease inhibitor mixture (Sigma), and cell lysis was carried out using three freeze-thaw cycles. Cell lysates were centrifuged at 20,200 ϫ g for 15 min at 4°C. The resulting supernatant was acidified, and peptide content was separated by HPLC on a C-18 chromatographic column (5 m, 4.6 ϫ 100 mm; Waters XTerra TM RP18), having a linear gradient from 20 to 80% acetonitrile plus 0.1% trifluoroacetic acid for 20 min at a flow rate of 1 ml/min. To enable highly sensitive detection of fluorescein-labeled peptides despite acidic chromatography conditions, the column eluate was mixed on-line with an aqueous ammonia solution (NH 4 OH/water, 2.5:100, v/v) with a flow rate of 0.75 ml/min (44). Peak identification and quantification was made possible by using standards comprising the peptide in the TAT-containing form and the TAT-lacking form, which was achieved by incubating the peptides in thiol-reducing dithiothreitol (10 mM). Fluorescence was monitored at an excitation wavelength of 490 nm and an emission wavelength of 520 nm.
Measurement of the Extracellular Acidification Rates-Extracellular acidification rates (ECAR) in CHO-S cells were evaluated as an index of metabolic activity by the application of a cytosensor microphysiometer (Molecular Devices) and a computer work station. AT1R-CHO-S cells were subcultured onto a polycarbonate membrane with a 3-m pore size in transwell capsules at a density of 2 ϫ 10 5 cells. On the day of the experiment, a spacer ring and an insert cup were placed into each transwell, and the assembled units were transferred into the microphysiometer chambers. Low buffered Dulbecco's modified Eagle's medium enriched with NaCl (2.26 g/liter) was used as the running medium in the microphysiometer, with a perfusion rate of 100 l/min. The extracellular acidification rate was measured as the change in pH over time, which was determined by 15-s potentiometer rate measurements (V/s; pump-off cycle) after a 40-s pump cycle with a 20-s delay (60 s total cycle time). Acidification rates (V/s) were normalized as a percentage of the base line, since the acidification rate data are dependent on cell seeding density and uniformity. Experiments were carried out as follows: cells were allowed to stabilize for 60 min, having been perfused with running medium in the cytosensor chambers. Initially, a 15-s stimulation with ang II (5 nM) was performed. Alteration in ECAR following ang II stimulation was monitored for 10 min, after which running medium was replaced by running medium plus peptides; this mixture was used to perfuse the cells for 30 min before the second stimulation with AT1 receptor agonist. In this way, each well was used as its own control, and results are expressed as the percentage of the first response.
Luciferase Reporter Gene Assay-To analyze the possible effects of the peptides on GPCR intracellular signaling, a reporter gene construct that drives luciferase expression in response to G i , G s , or G q stimulation and that was made up of three copies of the IL-6 multiple response element in tandem with a cAMP-responsive element (CRE) was constructed as previously described (45). The DNA sequences of the constructs were verified by automated sequence analysis (Mega-BACE 1000 DNA analysis system; Amersham Biosciences). To assess promoter activity, a dual luciferase reporter assay system (Promega) was used. In brief, CHO-S cells (Invitrogen) were plated in 6-well plates (0.5 ϫ 10 6 cells/well) overnight and were transiently transfected with pEGFP coding for the AT1 receptor (1 g), the pGL3-CRE-MRE promoter luciferase construct (1 g), and the pRL-CMV plasmid (16 ng) (a plasmid that encodes for Renilla luciferase and is used to normalize transfection efficacy) using Lipofectamine 2000 TM reagent (Invitrogen) for 6 h. Transfected cells were incubated in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum overnight, being subsequently seeded in 96-well plates at a density of 0.15 ϫ 10 6 cells/well. Forty-eight hours after transfection, CHO-S cells were treated either with isolated peptides or a mixture of peptides 15 min before and immediately before ang II (1 M) stimulation. Peptides were subsequently added to the cells at 30-min intervals for the first 1 h and at 60-min intervals until the end of the assay. After incubation for a total of 5 h at 37°C, firefly and Renilla luciferase activities in cell lysates were measured using an LMax luminometer (Molecular Devices), according to the manufacturer's instructions. The relative luciferase activity was calculated by normalizing firefly luciferase activity to that of Renilla luciferase. This same protocol was also used to analyze the effect of peptides in another signaling pathway, that of the ␤-adrenergic receptor, using HEK293 cells stimulated with isoproterenol (10 M).
EP24. 15 Overexpression in CHO-S and HEK293 Cells-In order to investigate the effect of EP24.15 overexpression on the signal transduction pathways of AT1 and ␤-adrenergic receptors, CHO-S and HEK293 cells were transfected with the set of luciferase vectors described above plus 1 g of either the vector coding for wild type EP24.15 (WT-EP24.15) or the "mock" control vector (43). After 48 h, cells were stimulated with ang II or isoproterenol, and luciferase activity was subsequently measured after 5 h, as detailed above. To measure the level of EP24.15 overexpression, an aliquot of the same HEK293 and CHO-S cells employed in the luciferase assays was used to measure EP24.15 activity in triplicate in a continuous assay using the QFS substrate (41, 42), as described above.
Peptide Affinity Columns and Target Identification-In an attempt to uncover putative intracellular targets of the studied peptides, peptide affinity columns were constructed. The columns were obtained through the covalent binding of each peptide through its NH 2 terminus to the gel of a HiTrap NHSactivated HP column (1 ml), according to the manufacturer's instructions (GE Healthcare). Rat brain cytosol extracts were incubated with individual peptide columns for 1.5 h at 4°C. Columns were extensively washed with 10 ml of phosphatebuffered saline and then with 1 ml of 0.1 M glycine, pH 2.7. This first milliliter of glycine wash (corresponding to three void vol-umes of the column) was discarded, since it was expected to contain many nonspecific/lower affinity bound proteins. Proteins that remained attached were eluted with the next 9 ml of 0.1 M glycine, pH 2.7, wash, concentrated in centrifugal filter devices (Amicon Ultra; Millipore), and subjected to separation by SDS-PAGE and staining with Coomassie Blue. Specific protein bands were visually identified, and excised bands were subjected to in-gel reduction, alkylation, and enzymatic digestion (Roche Applied Science) in a HEPA-filtered hood to reduce keratin background. LC-MS/MS analysis was performed on the in-gel digest extracts using an Agilent (Santa Clara, CA) 1100 binary pump directly coupled to a mass spectrometer at the Proteomics and Mass Spectrometry Facility at Tufts University School of Medicine (Boston, MA), according to standard protocols. Briefly, 2-8 l of sample was injected into a column using an LC Packings FAMOS autosampler (Sunnyvale, CA). Nanobore electrospray columns were constructed from 360-mm outer diameter, 75-mm inner diameter fused silica capillaries with the column tips tapered to a 15-mm opening (New Objective, Woburn, MA). The columns were packed with 200 Å of 5-m C18 beads (Michrom BioResources, Auburn, CA) and a reverse-phase packing material to a length of 10 cm. The flow through the column was split precolumn to achieve a flow rate of 300 nl/min. The mobile phase used for gradient elution consisted of 0.3% acetic acid, 99.7% water (A) and 0.3% acetic acid, 99.7% acetonitrile (B). Tandem mass spectra (LC-MS/MS) were acquired on a Thermo LTQ ion trap mass spectrometer (Thermo Corp., San Jose, CA). Needle voltage was set to 3 kV, isolation width was 3 Da, relative collision energy was 30%, and dynamic exclusion was used to exclude recurring ions. Ion signals above a predetermined threshold automatically triggered the instrument to switch from MS to MS/MS mode for generating fragmentation spectra. The MS/MS spectra were searched against the NCBI nonredundant protein sequence data base using the SEQUEST computer algorithm (46).
Data Analysis-Results are expressed as the mean Ϯ S.E. Statistical comparisons were done using unpaired t tests or analysis of variance followed by the Tukey test. A value of p Ͻ 0.05 indicated a significant difference. Statistical analyses of data were generated using GraphPad Prism, version 4.02 (GraphPad).

RESULTS
We have previously shown that catalytically inactive oligopeptidase EP24.15 is suitable for the isolation of novel bioactive peptides (22,47). Most of these peptides have 5-16 amino acid residues, are fragments of intracellular proteins, and contain a putative protein kinase post-translational modification site (22,23,31). In the present work, a total of nine novel peptides were identified in rat brain extracts using this inactive enzyme "substrate capture" assay (not shown). Three of these novel peptides (FE2, FE3, and FE4; supplemental Table 1) and a fourth peptide previously identified in mouse adipose tissue (5A) (23) all contained a PKC phosphorylation motif and were subsequently used to investigate the effect of intracellular peptides in signal transduction triggered by specific GPCR agonists. It is noteworthy that the cerebellin peptide precursor protein is known to contain a signal peptide sequence for entry in the secretory pathway. However, it has been previously demonstrated that misfolded proteins from the secretory pathway are transferred to the cytosol and are also degraded by the proteasome (48). A random peptide without known homology to any protein sequence previously described in the public protein data base (available on the World Wide Web) was used as a control peptide (RD) during these analyses (supplemental Table 2).
The results presented in Table 1 confirm that all four peptides isolated with the catalytically inactive oligopeptidase EP24.15 and chosen for functional analyses in GPCR signal transduction competitively interact with EP24.15, inhibiting the degradation of its QFS. As expected, the RD peptide had no effect upon QFS cleavage. HPLC and LC-MS analysis of the incubation reactions indicated that peptides 5A and FE2 were cleaved by EP24.15, whereas peptides FE3 and FE4 remained intact (Table 1). Whereas the longer peptide FE2, containing 15 amino acid residues, is cleaved at one site, peptide 5A, containing only 7 residues, had three distinct peptide bonds cleaved by EP24.15 (Table 1). These distinct patterns of substrate cleavage by EP24.15 suggest an intricate processing mechanism for peptides degraded by this peptidase within cells, in addition to possible competitive inhibition by peptides that interact with but are not efficiently degraded by EP24. 15.
The stability of these peptides within CHO-S cells was analyzed before evaluating their possible effect on GPCR signal transduction. The peptides were synthesized fused to a cationic TAT-derived peptide sequence (CRKKRRQRRR) as a means of introducing them into living cells. This sequence is known to promote the internalization of cargo bonded to it (40,49). However, some previous studies have shown that the TAT-derived sequence is also known to alter the behavior of its cargo (e.g. in the case of short kinase substrates) (50). It has also been shown to strongly influence the cargo's intracellular localization and/or accessibility to other molecules (40). To overcome these limitations, disulfide bonds were used to link the investigated peptide to the TAT peptide sequence, as previously described (40). These peptides were fluorescently labeled at the NH 2 terminus (supplemental Table 2) to allow quantification of fairly low concentrations and follow their intracellular release from the TAT sequence. Fig. 1 shows a typical chromatogram obtained from CHO-S cell extracts after incubation with TATbonded peptides; it is possible to observe a clear fluorescent peak corresponding to the peptide-free form of the TAT-derived sequence, among other unidentified peaks corresponding to degradation products or modified peptides. After 30 min, the intracellular concentration of these peptides within CHO-S cells varied from 1.61% (FE3) to 100% (RD) of the initial extracellular concentration ( Table 2), indicating that distinctive peptides have particular degradation ratios after entering cells (e.g. FE3 Ն FE2 Ն 5A Ͼ FE4 Ͼ RD; Table 2).
Peptide concentrations in rat brain, CHO-S, and HEK293 cytosol extracts were determined by the fluorescamine assay. The average molar concentration of peptides within the rat brain cytosol was 60.5 M (121.8 Ϯ 5.3 g/brain; n ϭ 4), assuming an average rat brain volume of 2 ml and having a molecular mass of 1000 Da/peptide. In CHO-S and HEK293 cells, the peptide concentration was even higher, reaching 0.3 mM in CHO-S and 0.6 mM in HEK293 cells, considering the same mass of 1000 Da/peptide and 1 pl as the average cell volume. Therefore, it seemed reasonable to evaluate the putative intracellular effect of peptides (supplemental Table 1) at concentrations of 5, 20, and 80 M, which is below the natural concentration of peptides in both CHO-S and HEK293 cells lines, for the following functional experiments.

TABLE 2 Cytosolic concentration of peptides within CHO-S cells
Peptides coupled to a TAT-derived peptide sequence were added to cells at a concentration of 20 M and incubated during 30 min at 37°C before cytosol extraction. Fluorescent peptide concentration in these cytosol extracts (see Fig. 1) was calculated based on the HPLC automatic generation of peak areas, and retention time was based on standard peptides. The volume of one cell was estimated as 1 pl for molar calculations of peptide intracellular concentrations. Results are the average of two independent assays. All of the peptides chosen in this work had a putative PKC phosphorylation site, which was previously shown to be the most abundant post-translational modification motif found in peptides so far isolated using the inactive EP24.15 (31). Ang II activates the AT1 receptor through a G␣ q/11 -mediated activation of phospholipase C-␤, generating diacylglycerol, which then activates PKC. Thus, the AT1 receptor was initially used in microphysiometry experiments to investigate the intracellular effects of these peptides on GPCR signal transduction. Stimulation of CHO-S cells expressing AT1 receptors (AT1R-CHO-S) with ang II increased, in a concentration-dependent manner, the rate of extracellular acidification as measured by the cytosensor microphysiometer, producing a typical dose-dependent response for ang II with an EC 50 of 2 nM ( Fig. 2A). A 5 nM concentration of ang II was used to stimulate cells, since this yielded a consistent alteration in extracellular acidification rates (ECAR) while being at the same time close to the EC 50 concentration. The experimental protocol for the cytosensor assays was designed to utilize each transwell as its own control. To do this, cells were stimulated for a first time with ang II (5 nM) in the absence of any peptide treatment, and the alteration in ECAR was recorded and considered as the first response. Next, the same transwell was perfused for 30 min with either a 5 or 20 M concentration of each TATbonded peptide (5A, FE2, FE3, and FE4), and a second ang II stimulation was performed, this time in the presence of analyzed peptides. Under these conditions, no statistically significant changes in the ang II-induced increase in ECAR were observed (Fig. 2, B-E). However, an accentuated and concentration-dependent inhibition of the increase in ECAR was observed when this same protocol was applied to AT1R-CHO-S cells treated with a mixture of the four TAT-bonded peptides (5A, FE2, FE3, FE4; 5 or 20 M each peptide; Fig. 2F). The linearity coefficient (r 2 ) of ECAR relative values was 0.962, which indicates an excellent correlation between concentration and effect (Fig. 2F). TATbonded RD peptide had no effect on ang II-induced increase in ECAR at concentrations of up to 80 M (Fig. 2, B-F). Peptides lacking the TAT sequence had no effect on ECAR (data not shown).

Peptide Concentration Percentage of initial concentration
A luciferase gene reporter assay was subsequently employed to further analyze the intracellular effects of peptides in GPCR signal transduction. Stimulation of AT1R-CHO-S cells with ang II (1 M) efficiently induced luciferase activity (Fig. 3). Treatment of cells with a mixture of the same four peptides (20 M/each) efficiently augmented the rate of gene reporter transcription (Fig. 3A). None of the individual peptides when assayed at 20 M produced changes in luciferase activity (Fig.  3B). However, when assayed at 80 M (Fig. 3C), which corresponds to the overall peptide concentration in the peptide mixture, all peptides were able to increase luciferase activity (Fig.  3C). Neither TAT peptide alone nor TAT-bonded RD peptide were able to alter the luciferase activity triggered by ang II in any of the above conditions (Fig. 3, B-F). Peptides without the TAT sequence were also deprived of any effect on ang II-induced gene reporter transcription (Fig. 3A), indicating that these peptides must be located within cells in order to modulate luciferase gene reporter activity. A possible effect exerted by the fluorescent group positioned at the NH 2 terminus of the peptide was dismissed as unlikely given that peptide 5A lacking 5-carboxyflurescein was also shown to affect ang IIinduced luciferase transcription in this system (Fig. 3D), thus showing that peptides with a free NH 2 terminus can also affect GPCR signal transduction. In order to investigate peptide effects in a signaling pathway that does not involve, at least directly, PKC activation, luciferase activity was measured in HEK293 cells treated with the nonspecific ␤-adrenergic agonist isoproterenol. This signaling system differs substantially from the ang II AT1R-CHO-S cells in several ways. The first is that HEK293 cells were not transiently transfected with GPCR receptors, since functional ␤ 2 -adrenergic receptors are endogenously expressed in this cell line (51). The second difference refers to the signaling pathway employed by ␤-adrenergic receptors. It is well established that these receptors are coupled to G s proteins that, when stimulated, activate adenylate cyclase, raising intracellular cAMP levels and activating protein kinase A as well as other downstream molecules (52). Similar to ang II in AT1R-CHO-S cells, all peptides, except for the RD peptide, were able to increase isoproterenol-induced luciferase activity in HEK293 cells (Fig. 3E), despite the differences between the AT1R-CHO-S and HEK293 intracellular signaling pathways. These data suggest that these peptides have a broader functional potential to affect GPCR signal transduction.
We then investigated possible mechanisms related to the above-observed effects of intracellular peptides on GPCR signal transduction. To verify if the predicted PKC phosphorylation motifs present in these active peptides (supplemental Table 2) were in fact functional, the peptides were incubated with PKC in vitro. Peptides 5A, FE2, and FE4 were efficiently phosphorylated by PKC (Fig. 4, A, B, and D), whereas peptides FE3 and RD were not (Fig. 4C and data not shown, respectively). A second set of assays were performed with TAT-bonded peptides (supplemental Table 2), producing similar results (data not shown). Therefore, the presence of a PKC motif in these peptides does not seem to be relevant to their biological activity.
Another characteristic of all four bioactive peptides investigated here was their ability to competitively interact with EP24.15, inhibiting the degradation of QFS with relatively low K i values (Table 1). Therefore, intracellular competitive inhibition of EP24.15 could be a mechanism related to the intracellular biological function of these peptides. In fact, overexpression of EP24.15 reduced both ang II-and isoproterenol-induced luciferase activity in CHO-S and HEK293 cells, respectively (Fig. 5). These results are in agreement with those observed by the addition of intracellular peptides to cells, resulting in increased luciferase expression in similar gene reporter assays (Fig. 4).
To further characterize the mechanism of action of these peptides, possible protein targets were searched in rat brain cytoplasmic extracts that were passed through peptide affinity columns. The peptide coupling efficiency to the column's resin varied from 100% (FE2), 58% (FE3), 46% (RD), and 20% (5A) to 16.5% (FE4). No proteins were observed in eluates from columns made with peptides 5A and FE4, which could be related to their consistently lower coupling efficiency. On the other hand, a different set of proteins was found to interact with peptides FE2, FE3, or RD (supplemental Table 3). In fact, 67 different proteins were identified in the eluate from the FE2 column, 19% of which are known to interfere with GPCR signal transduction (Table 3). In the FE3 column eluate, 14 different proteins were identified, with 42% previously described as being involved in GPCR signaling (Table  3). It is worth mentioning that not all of these proteins are expected to interact directly with these peptides; thus, secondary protein interactions could be responsible for such a large number of proteins identified.

DISCUSSION
In the present study, oligopeptides isolated from rat brain tissue using the inactive EP24.15 (22) were introduced into CHO-S and HEK293 cells and were found capable of altering GPCR signal transduction. EP24.15 overexpression itself changed both ang II and isoproterenol signal transduction, suggesting a physiological significance for these original findings. The mechanism by which these peptides modulate GPCR cell signaling seems to involve their ability to interact with EP24.15 and other specific proteins identified herein. Therefore, our present data suggest that EP24.15 and intracellular peptides are natural regulators of GPCR signal transduction.
To investigate the possible intracellular function of these peptides, fragments of intracellular proteins, isolated using the inactive EP24.15, were first fused to a TAT-derived peptide sequence using disulfide bonds, as previously described (40). To   analyze the effect of peptides on GPCR signal transduction, two distinct models were used. Initially, the use of a microphysiometer was employed to study signal transduction pathways, since activation or inhibition of a wide variety of effectors of transduction processes is known to affect the metabolic rate (53). Energy metabolism is tightly coupled to cellular ATP consumption, which generates acidic products. Changes in cell physiology, such as receptor activation, alter the rate of energy use and, hence, the rate of extracellular acidification, which is subsequently measured by the microphysiometer (54). The increase in extracellular acidification rate induced by ang II in CHO-S cells was significantly inhibited by specific peptide mixtures with an excellent correlation (linearity coefficient of 0.962) between concentration and effect. It is worth mentioning that ang II activates the AT1 receptor through a G␣ q/11 -mediated activation of phospholipase C-␤, generating diacylglycerol and thereby activating PKC (55,56). Thus, it was particularly interesting to exclude the functional relevance of the PKC phosphorylation site frequently found in peptides isolated using the inactive EP24.15 (31). Subsequently, a gene reporter assay that drives luciferase expression in response to stimulation of GPCRs (45) was used to further investigate the intracellular effect of peptides in signal transduction. With the exception of the random peptide, all peptides investigated here significantly increased the luciferase activity induced by ang II in AT1R-CHO-S, a result repeated in HEK293 cells stimulated by isoproterenol, which activates ␤-adrenergic receptor signaling through G s proteins (57). Whereas it is clear that the "EP24.15-captured peptides" investigated here are able to alter signal transduction of distinctive GPCRs, they decreased the response of ang II as measured by the microphysiometer and increased the response of both ang II and isoproterenol in the luciferase gene reporter assay. One of the possible explanations for these opposite effects is the difference in the period of agonist stimulation in each assay. Although in the microphysiometry assays, ang II stimulation was short (15 s), minimizing AT1 desensitization (58), in the luciferase assays, agonist stimulation lasted 5 h, a situation expected to induce receptor desensitization, internalization, and different signaling components/feedback loops.
In a recent study on the mouse brain peptidome, the relative levels of peptide fragments of intracellular proteins were shown to change from mouse to mouse, whereas other protein fragments were common to all brain extracts and showed low variability (59). Previously, we have also identified several oligopeptides from rat and mice tissues with sequences matching specific regions of intracellular proteins (11,22,23). Here, we have found high peptide concentra-  tions (60 -600 M) in cells (CHO-S and HEK293) and in rat brain extracts. Additionally, in the present report, we have shown that EP24.15 overexpression in CHO-S and HEK293 cells is sufficient to reduce activation of the luciferase reporter induced by either ang II or isoproterenol in a gene reporter assay, suggesting that an increased intracellular degradation of peptides affects GPCR signal transduction and can alter cell homeostasis. Altogether, the above evidence is contrary to previous assumptions that intracellular peptides generated from protein turnover are rapidly converted into free amino acids and escape degradation, only participating in antigen presentation (8,21,60). The mechanisms responsible for the stability of intracellular peptides escaping immediate degradation has been suggested to include binding to specific proteins that prevent further proteolytic degradation (8) and structural alterations, such as phosphorylation, that in vitro has been shown to reduce substrate degradation by EP24.15, EP24.16 (31), and prolyl-oligopeptidase (61). Moreover, EP24.15 has been shown to be a rate-limiting enzyme for the degradation of proteasomal products (37) and, since its activity is regulated by glutathionylation, could also delay peptide degradation within cells (62). Other possibilities also exist and together may contribute to the relatively high levels of intracellular peptides determined herein.
EP24.15 is ubiquitously distributed in mammalian cells and tissues (63,64) thought to be involved in extracellular neuropeptide metabolism (65). However, additional intracellular functions were suggested for EP24.15 (11,22) based on its predominant nuclear localization in the rat brain (66,14) and on the demonstration that EP24.15 participates in antigen presentation through the major histocompatibility class I (24,26,27). To investigate possible novel intracellular functions for this enzyme, we developed an experimental "substrate capture" assay that uses mutated, catalytically inactive EP24.15 and found its endogenous substrates to be fragments of intracellular proteins containing from 5 to 17 amino acids (22). Using this "substrate capture" assay, we found that engineered mice containing one, two, or three copies of the angiotensin-converting enzyme gene have a distinctive oligopeptidase-arrested peptide profile in their adipose tissue; in fact, mice containing three copies of the angiotensin-converting enzyme gene were shown to have smaller body weight and periepididymal adipose tissue accumulation compared with one-copy mice (23). Two of these peptides containing putative post-translational modification sites (LVVYPWTQRY and VVYPWTQRY) were shown in vitro to inhibit phosphorylation of a synthetic standard substrate of protein kinase C (23). Based on these data, it was suggested that intracellular peptides could be natural regulators of protein interactions interfering with cell signaling (11,23). Corroborating these suggestions, here we have shown that not only does the introduction of EP24.15 intracellular "captured substrates" within cells alter GPCR signal transduction, but EP24.15 overexpression in both CHO-S and HEK293 cells is enough to reduce the activation of a luciferase reporter stimulated by specific GPCR agonists. Together, these data suggest a physiological role for intracellular peptide substrates of EP24.15 in cell signaling. We also identified several other proteins that individ-ually or in complex bind to these peptides in affinity chromatography assays; binding of these peptides to proteins other than EP24.15 could also account for their biological effects on cell signaling. Thus, it is exciting to speculate that intracellular peptides could be a part of the homeostatic mechanism that maintains protein network composition and that alterations of this steady state either by adding peptides to the intracellular environment or by overexpressing EP24.15 could modulate cell functioning in ways such as those shown here for GPCR signal transduction.
The proteins identified to bind to at least two of the functional peptides investigated here (FE2 and FE3) have been shown to be involved in GPCR signaling. For example, dynamin participates in GPCR sequestration from the plasma membrane, whereas 14-3-3s are scaffold proteins that act in distinctive intracellular signaling complexes (67,68). It is known that adaptor protein complex-2, found to interact with both FE2 and FE3 peptides, is the most abundant adaptor protein in clathrin-coated pits and functions directly in GPCR internalization (69). In this way, the interaction between the adaptor protein complex-2 and the peptides investigated herein could impair to some degree the normal endocytic process of activated AT1 or ␤-adrenergic receptors, allowing them to signal for a longer time and increasing luciferase expression in the gene reporter assays. Although the exact mechanism by which each of these peptides affects GPCR signal transduction in the two models investigated herein still needs further attention, the differences observed in comparing their effects with those of the random control peptide allows for the conclusion that the actions of the peptides were not due to the TAT carrier or to the NH 2 -terminally attached fluorescent group, as highlighted under "Results." The latter is particularly important, since most of the naturally generated peptides have a free NH 2 terminus. Furthermore, peptides not having the capacity to enter cells (i.e. peptides without TAT) did not have an effect, leading to the conclusion that the effects of these peptides on GPCR signal transduction are not due their extracellular interactions. Therefore, these data suggest that intracellular peptides can alter GPCR signaling, directly binding to proteins from the signaling pathway and indirectly altering EP24.15 activity.
In conclusion, our present results suggest that EP24.15 is physiologically relevant in degrading intracellular peptides that help to control GPCR signal transduction. In certain pathological conditions, a significant change in the intracellular peptide composition could disturb cell homeostasis, since GPCRs constitute the largest class of cell surface receptors in humans, mediating key processes, including cognition, mood, appetite, pain, and synaptic transmission (70). Therefore, we believe that intracellular peptides should be considered as key elements for the proper understanding of cellular complexity and function.