Research Article Received: 30 January 2014
Revised: 24 March 2014
Accepted: 31 March 2014
Published online in Wiley Online Library: 13 May 2014
(wileyonlinelibrary.com) DOI 10.1002/psc.2642
Dissecting a role of a charge and conformation of Tat2 peptide in allosteric regulation of 20S proteasome Julia Witkowska,a Przemysław Karpowicz,a Maria Gaczynska,b Pawel A. Osmulskib and Elżbieta Jankowskaa* Proteasome is a ‘proteolytic factory’ that constitutes an essential part of the ubiquitin-proteasome pathway. The involvement of proteasome in regulation of all major aspects of cellular physiology makes it an attractive drug target. So far, only inhibitors of the proteasome entered the clinic as anti-cancer drugs. However, proteasome regulators may also be useful for treatment of inflammatory and neurodegenerative diseases. We established in our previous studies that the peptide Tat2, comprising the basic domain of HIV-1 Tat protein: R49KKRRQRR56, supplemented with Q66DPI69 fragment, inhibits the 20S proteasome in a noncompetitive manner. Mechanism of Tat2 likely involves allosteric regulation because it competes with the proteasome natural 11S activator for binding to the enzyme noncatalytic subunits. In this study, we performed alanine walking coupled with biological activity measurements and FTIR and CD spectroscopy to dissect contribution of a charge and conformation of Tat2 to its capability to influence peptidase activity of the proteasome. In solution, Tat2 and most of its analogs with a single Ala substitution preferentially adopted a conformation containing PPII/turn structural motifs. Replacing either Asp10 or two or more adjacent Arg/Lys residues induced a random coil conformation, probably by disrupting ionic interactions responsible for stabilization of the peptides ordered structure. The random coil Tat2 analogs lost their capability to activate the latent 20S proteasome. In contrast, inhibitory properties of the peptides more significantly depended on their positive charge. The data provide valuable clues for the future optimization of the Tat2-based proteasome regulators. Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. Keywords: proteasome; allosteric inhibition/activation; alanine scan; circular dichroism; Fourier transform infrared spectroscopy
Introduction
J. Pept. Sci. 2014; 20: 649–656
* Correspondence to: Elżbieta Jankowska, Department of Medicinal Chemistry, Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80-308 Gdańsk, Poland. E-mail: [email protected] a Department of Medicinal Chemistry, Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80-308 Gdańsk, Poland b Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, San Antonio, TX, USA Abbreviations: ChT-L, chymotrypsin-like; CD, circular dichroism; CP, core particle (20S proteasome); FTIR, Fourier transform infrared; HIV-1 Tat, human immunodeficiency virus type 1 transactivator protein; MW, molecular weight; PA, proteasome activator; PAN, proteasome activator nucleotide-activated; PGPH, post-glutamyl peptide hydrolyzing; PPII, polyproline II; TFA, trifluoroacetic acid; T-L, trypsin-like; UPS, ubiquitin-proteasome system.
Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd.
649
Allosteric regulation of enzymes enables modulation of their activity through binding an effector to the site distant from the active site and inducing structural changes, which can propagate to the active center, modifying conformation around it and, as a result, the enzyme affinity toward natural substrates [1]. The fact that binding of an allosteric effector proceeds far away from the classic orthosteric binding sites has important implications. First of all, active site topologies are often strongly conserved between distinct members of the enzyme family, so they are difficult to address in a selective manner by low MW compounds. Allosteric sites are usually less conserved; therefore, they may constitute better drug targets [2,3]. Second, binding an orthosteric ligand blocks the active site leading to the enzyme inhibition. In the case of allosteric regulators, not interacting directly with the active sites, more diverse effects on the enzyme properties, such as inhibition, activation, or specificity modulation, are achievable [2,4]. Accordingly, there is an increasing interest in utilizing the allosteric modulation phenomena in drug design and introducing allosteric drugs to the clinic [5–7]. The proteasome is multicatalytic and multisubunit protease complex amenable to allosteric regulation [8–10]. The 20S CP is formed by four heptameric rings arranged in αββα manner and creating together a chamber, which harbors the catalytic sites [11]. The active centers located in β5/5′, β2/2′, and β1/1′ subunits are responsible for the ChT-L, T-L, and PGPH proteasomal
peptidases, respectively [12,13]. Access of proteins and peptides to the catalytic chamber is restricted by the N-termini of α subunits, which form a regulated gate [14]. Three classes of activators: 11S, PA200, and 19S facilitate passage of substrates to the proteasome interior by binding to the α subunits and repositioning structural motifs blocking the enzyme entrance pore [15,16]. The 11S activator (PA28/REG, Trypanosoma brucei analog – PA26) is a heptameric protein that binds to the proteasome by inserting its seven C-terminal tails to the seven pockets located between the α subunits. To reorganize the α-tails and stabilize the opened conformation of the gate, 11S utilizes an
WITKOWSKA ET AL.
650
internal ‘activation loop’ [17]. Two other types of activators: PA200 (Blm10 in Saccharomyces cerevisiae) and 19S/PA700 (PAN in Archaea) contain a conserved C-terminal hydrophobictyrosine-X motif sufficient to open the gate upon binding within the pockets between neighboring α subunits [18]. Unlike the oligomeric 11S and 19S/PAN activators, which use multiple C-termini to bind to the proteasome α ring, Blm10 and its mammalian homolog PA200 are single chain proteins and can interact with only one pocket between α subunits, what results in a partially opened disordered conformation of the gate [19]. The UPS is a major contributor to the intracellular protein degradation machinery in mammalian cells. By degradation of short-lived regulatory proteins, the enzyme has an impact on many vital processes such as a cell cycle, DNA repair, sodium channel function, regulation of immune and inflammatory responses, and many others. Moreover, UPS is also responsible for disposal of any abnormal proteins – oxidized, unstructured, misfolded, or viral. Malfunction of the UPS can lead to serious disorders including cardiovascular diseases, certain cancers, muscular dystrophy, neurodegeneration, and systemic autoimmunity and makes the proteasome an important target for pharmacological intervention [20,21]. Numerous small competitive inhibitors of the proteasome have been already developed and several of them are currently under clinical evaluation [22,23]. Two inhibitors, bortezomib and carfilzomib, are approved for use with relapsed/refractory multiple myeloma and mantle cell lymphoma [24,25]. Although both compounds demonstrated efficacy in the treatment of these hematological malignancies, a significant number of patients do not respond to these drugs or develop the drug resistance [26]. We postulate that allosteric modulators may be a promising alternative to classic orthosteric ligands [27–29]. Because of its distinct mode of action, allosteric inhibitors, working alone or in combination with other drugs, may improve efficacy of the treatment, limit off-target effects, and allow to overcome resistance to the orthosteric drugs. Importantly, because allosteric effectors bind at noncatalytic sites that usually are under much lower evolutionary pressure, they themselves exhibit substantially decreased potential to quickly induce resistance. Recently, several allosteric inhibitors have been identified and extensively characterized. 5-amino8-hydroxyquinoline has been shown by NMR to bind to the α-subunits inside the proteasome inner chamber, and it was able to overcome some forms of bortezomib resistance in vitro [30]. Other promising noncompetitive inhibitors are imidazoline scaffolds with the best compound displaying submicromolar inhibitory capacity [31]. In addition to their canonical mTOR target (mammalian target of rapamycin protein), rapamycin and its double or single domain analogs in vitro also productively interact with the α face of 20S proteasome leading to inhibition of its peptidase and proteinase activity [28]. These compounds change the abundance of conformers with the open gate and compete with the 19S cap for the binding sites on the α face [28]. Imidazoline, quinoline, and rapamycin based compounds may overcome resistance of selected cancer cells to bortezomib. The outcome of allosteric interactions is not limited to inhibition but also extends to activation of the proteasome, and this property may be utilized to increase degradation of damaged (for instance, oxidized as a result of oxidative stress) proteins and reduce their aggregation [32,33]. However, small molecules that can activate or enhance the activity of the 20S CP are scarce and not well studied. The compounds already recognized as being able to activate the 20S proteasome are SDS, certain lipids, and polycations [34].
Unfortunately, these substances are either too weak or lack druglike properties to be considered as potential therapeutics or leading structures. In addition, SDS while activating the proteasome at low concentrations, at higher ones becomes its irreversible inhibitor [35]. More promising activators seem to be fatty acids, betulinic acid, and lithocholic acid derivatives [35–37]. They display some specificity toward the proteasome peptidases: oleic, linoleic, and linolenic acids activated ChT-L and PGPH peptidases of 20S proteasome while inhibiting its T-L activity, betulinic acid preferentially activated the ChT-L [35,36]. During our search for small synthetic compounds that can serve as allosteric regulators of the proteasome activity, we focused on sequence fragments of HIV-1 Tat protein. HIV-1 Tat protein inhibits the activity of the 20S proteasome, competes with the 11S activator in binding to the proteasome α face, and markedly influences major histocompatibility complex class I-associated antigen presentation [30,30]. Tat2 peptide has a sequence R49KKRRQRR56Q66DPI69, which is derived from the basic domain of HIV-1 Tat, supplemented by the QDPI fragment. The peptide sequence was developed to comprise residues Lys51, Arg52, and Asp67, relevant to Glu235, Lys236, and Lys239 in the PA28 activator and postulated to constitute the proteasome binding region [30]. Our previous studies demonstrated that Tat2 noncompetitively inhibits the activated CP [29]. In addition, Tat2 stimulates ChT-L activity of the latent CP in a relatively narrow range of the peptide concentrations, reaching the maximum effect at about 1 μM concentration. We determined that Tat2 is not degraded by the proteasome and, likewise to the full-length HIV-1 Tat protein, competes with the PA28 activator for binding sites on the proteasome α face [29]. To optimize the structure of Tat2-based proteasome regulators, we aimed at identification of Tat2 proteasome-targeting pharmacophore. To assign relative functional importance to each amino acid residues within the peptide, we performed alanine scan, substituting residues either one by one or in blocks. Subsequently, we characterized the structural properties of the resulting peptides by CD and FTIR spectroscopy and tested the effects of treatments with the peptides on the proteasome peptidase activity. The results suggest that the inhibitory and activating capabilities of Tat2 peptides are separately modulated by charge and particular conformation, respectively.
Materials and Methods Peptide synthesis and purification All peptides were synthesized on TentaGel R PHB resin with loading capacity of 0.2 mmol/g (Rapp Polymere). Syntheses were performed using Fmoc strategy on a Millipore model 9050 Plus peptide synthesizer with continuous-flow methodology or using microwave radiation (Plazmatronika RM800; formerly Plazmatronika, now Ertec, Wroc∤aw, Poland) under nitrogen atmosphere. The first amino acids were attached to the resin using symmetrical anhydride method. The efficiency of the coupling was determined by the measurement of the resulting fulvene-piperidine adduct at λ = 301 nm. A N2 line was attached to a microwave vessel for continuous agitation of the resin. Coupling reactions were performed in the following conditions: power 24 W, maximum temperature 80 °C, and irradiation time 4.5 min. Every deprotection step was repeated twice using power 16 W, maximum temperature 60 °C, and irradiation time 5 and 3.5 min, respectively. The actual
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION temperature of the reactions was measured and found to be 71–82 °C for the coupling cycles and 53–63 °C for the deprotection steps. Crude peptides were purified by RP-HPLC using a C8 semipreparative Luna column (21.2 × 250 mm, 5 μm; Phenomenex Torrance, CA, USA). A linear gradient of acetonitrile in 0.1% aqueous TFA or acetonitrile in triethylamine phosphate containing buffers at pH 3 were used as a mobile phase. The MWs of the peptides were ascertained by MALDI TOF or electrospray ionization ion trap TOF liquid chromatography MS with C12 Jupiter Proteo column (150 × 2 mm, 4 μ, 90 Å; Phenomenex). The purity of the synthesized compounds was confirmed by analytical RP-HPLC using a C8 Kromasil column (4.6 × 250 mm, 5 μm Phenomenex) and a 30 min linear gradient of 5–80% acetonitrile in 0.1% aqueous TFA. UV absorption was observed at λ = 223 nm. FTIR spectroscopy The peptides were dissolved in H2O to a final concentration of 20 mg/ml. TFA solutions were prepared from 2% aqueous stock solution by consecutive dilutions allowing to obtain 1.5%, 1.25%, 1.0%, 0.8%, 0.4%, 0.2%, 0.1%, 0.05%, 0.025%, and 0.0125% TFA concentration. Spectra of water, as well as aqueous solutions of TFA and the peptides, were obtained by spotting 15 μl of each sample onto a CaF2 window, which was then assembled into the infrared cell (the thickness of a spacer was 0.006 mm). FTIR spectra were collected with a Bruker IFS66 spectrophotometer (Bruker Optic GmbH, Bremen, Germany) with a Deuterated Tri Glycine Sulfate detector. During data collection, the spectrometer was continuously purged with dry air free of CO2. Typically, 160 scans at 4 cm 1 resolution were averaged for each spectrum. At least three independent measurements were performed for each sample. Data processing was executed with GRAMS 5.0 (Galactic Enterprises). The water spectrum, recorded with the same accessory and instrument conditions, was subtracted from the spectra of aqueous solutions of TFA and the peptides. The appropriate spectrum of pure TFA (deprived of water contribution) was then subtracted from each peptide spectrum. The resulted spectra were baseline-corrected. The spectral second derivative was calculated using Sawitzky–Golay algorithm with 13-point smoothing function [38]. The removal of TFA and water contributions from the spectra was accomplished in the following steps:
For CD measurements, 1 mm path length quartz cuvettes were used. Protein samples were dissolved in H2O or in 50 mM Tris/HCl, pH 8.0 to a final concentration of 0.2 mg/ml. Spectra were acquired at 25 °C on a Jasco J-815 spectrometer in a range 185–260 nm for aqueous solutions and 195–260 nm for solutions of peptides in the buffer. Each spectrum is a result of three independent measurements. Biochemical studies Influence of the peptides on the catalytic properties of the 20S proteasome was tested with housekeeping CP purified from human erythrocytes (Enzo Life Sciences). The latent CP or CP activated with 0.005% SDS at a final concentration of 1.5 nM was utilized. The model fluorogenic peptide substrate succinyl-LeuLeuValTyr-4-methylcoumarin-7-amide (SucLLVY-MCA, Bachem) was employed to determine the proteasome ChT-L activity. Stock solutions of the substrate and the tested peptides were prepared in DMSO. The content of DMSO never exceeded 3% of the final reaction volume. All assays were performed in the 96-well plate format using a reaction volume of 100 μl. Tests were carried out in 50 mM Tris/HCl (pH 8.0) reaction buffer. Substrate was added at 100 μM final concentration (1% of the volume). The release of an aminomethylcoumarin (AMC) group was followed by measuring fluorescence emission at 460 nm in 2-min intervals for up to 60 min at 37 °C (Fluoroskan Ascent). The peptidolytic activity was calculated as nanomoles of the released AMC product per milligram of CP per second.
Results Peptides Syntheses To dissect regions/structural features, which influence Tat2 peptide interactions with the proteasome, we have performed a systematic alanine scan by replacing consecutive residues in the peptide sequence by alanine residues. The primary structures of the synthesized analogs are listed in Table 1. Structural Studies FTIR. Subtraction strategy of TFA contribution. Peptides synthesized on solid phase are obtained as TFA salts because they are cleaved from the resin with TFA treatment, and furthermore, they are purified in the presence of TFA. The main TFA band, associated with vibration of COO group, appears in FTIR spectra at about 1672 cm 1 [43,44], overlapping with the amide I band (1600–1700 cm 1) that corresponds to a peptide bond C=O stretching vibration, which is usually utilized to determine the secondary structure of peptides and proteins. To numerically remove contribution of TFA from the peptides’ spectra, we utilized one of the procedures proposed by Valenti et al. [42]. They suggested subtraction of TFA and water spectra separately as a method of choice only for the spectra collected as independent measurements or for different buffer solutions. In other cases, the simplified methodology of one step subtraction of water and TFA contribution was proposed. We found, however, that for different peptides, containing different amount of water in their lyophilized form, as well as different amount of TFA counter ions due to differences in the sequence, it is impossible to apply the same scaling factor as a subtraction coefficient for
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
651
1. Spectra of TFA in water were individually collected for several TFA concentrations in the range of 0.0125–2%. 2. Water contribution was removed by subtracting the scaled water spectrum from the spectra recorded for aqueous solution of TFA until the flat baseline was obtained between 1750 and 1950 cm 1 [39–41]. The scaling factor was determined individually for each case. 3. Water contribution from the peptides spectra was removed exactly as it was described earlier for TFA producing the buffer-corrected spectra of the peptides. 4. The spectra of pure TFA of appropriate concentration were subtracted from the buffer-corrected spectra of the peptides. Selection of the most appropriate TFA spectrum was carried out to meet the criterion of a complete removal of TFA contribution. Correct subtraction of TFA contribution was assessed by disappearance of the TFA bands at 1200 and 1147 cm 1 arising from the C-F stretching vibration [42].
CD spectroscopy
WITKOWSKA ET AL. Table 1. Synthesized Tat2 analogs Name Tat2 Tat2A1 Tat2A2 Tat2A3 Tat2A4 Tat2A5 Tat2A6 Tat2A7 Tat2A8 Tat2A9 Tat2A10 Tat2A11 Tat2A12 Tat2A2-5 Tat2A4-5 Tat2A5-7 Tat2A7-8
Sequence
Retention time (min)
MW (g/mol)
Molecular ion (m/z)
11.5 11.6 11.2 11.5 11.7 11.2 11.6 11.4 12.2 11.5 11.5 11.3 10.3 12.4 12.2 11.5 12.4
1636.9 1551.8 1579.8 1579.8 1551.8 1551.8 1579.9 1551.8 1551.8 1579.9 1592.9 1610.9 1594.8 1352.5 1466.7 1409.7 1466.7
1636.1 1551.2* 1579.2 1579.5 1551.2 1550.7 1579.8* 1551.4 1551.0 1579.6 1592.6 1611.8 1593.8 1351.8 1466.9 1409.4 1466.4
H-RKKRRQRRQDPI-OH H-AKKRRQRRQDPI-OH H-RAKRRQRRQDPI-OH H-RKARRQRRQDPI-OH H-RKKARQRRQDPI-OH H-RKKRAQRRQDPI-OH H-RKKRRARRQDPI-OH H-RKKRRQARQDPI-OH H-RKKRRQRAQDPI-OH H-RKKRRQRRADPI-OH H-RKKRRQRRQAPI-OH H-RKKRRQRRQDAI-OH H-RKKRRQRRQDPA-OH H-RAAAAQRRQDPI-OH H-RKKAAQRRQDPI-OH H-RKKRAAARQDPI-OH H-RKKRRQAAQDPI-OH
Analytical RP-HPLC analyses were performed using a C8 Kromasil column (4.6 × 250 mm, 5 μm) and a 30 min linear gradient of 5–80% acetonitrile in 0.1% aqueous TFA. The molecular weights of the peptides were ascertained by MALDI TOF or ESI IT TOF LC mass spectrometry with C12 Jupiter Proteo column (150 × 2 mm, 4 μ, 90 Å; Phenomenex). * Molecular ion detected by MALDI MS.
TFA and water removal from each spectrum. Despite the fact that the spectra of our peptides were collected with identically prepared samples, the simplified methodology was either not able to ensure the complete removal of the buffer contribution or caused over-subtraction in some regions, introducing artifacts to the spectrum. We determined that, under our experimental conditions, the most effective was the procedure in which water and TFA contributions were removed from the spectra separately and the order of subtraction was not important. The detailed description of the procedure is provided in Materials and Methods section. Structure of the Peptides
652
Fourier transform infrared spectroscopy. Almost all Tat2 analogs with a single Ala substitution displayed similar features of their FTIR spectra. Second derivative of the spectra showed two deep minima – around 1620 and 1680 cm 1 (Figure 1A). The first minimum suggests hydrated PPII conformation, whereas minima between 1660 and 1685 cm 1 are usually attributed to turns [45,46]. The only analog, which did not fit to this PPII-turn pattern, was Tat2A10, in which only one minimum, although with some minor features at various wavenumbers, was observed (Figure 1B). Location of the minimum at ~1645 cm 1 suggests that this peptide in liquid does not acquire preferentially any distinct conformation but rather represents a random coil structure [46]. Because in this peptide, the Asp residue was substituted with Ala, we suspect that this residue may be important for stabilization of Tat2 peptide conformation. Peptides Tat2A2-5, Tat2A4-5, Tat2A5-7, and Tat2A7-8, in which two or more consecutive residues were replaced with Ala, displayed FTIR spectra similar to Tat2A10 (Figure 1C). Their second derivative spectra exhibited only one main minimum located at about 1645 cm 1, what again suggests random coil conformation. In these Tat2 analogs, substitutions diminished the number of basic residues, while in the Tat2A10 variant,
Figure 1. Second derivative of FTIR spectra of the investigated Tat2 analogs: (A) peptides with single Ala substitutions, (B) analog with Asp10 residue exchanged for Ala, and (C) peptides with multiple Ala substitutions. Peptides concentration: 20 mg/ml.
alanine residue replaced an acidic aspartic acid. Loss of the preferred conformation in both types of peptides underlines the role of basic and acidic residues in stabilization of the peptides conformation. Most probably Asp and Arg/Lys residues create a salt bridge, which imposes a more rigid structure of the peptide.
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION Circular dichroism. Circular dichroism studies of Tat2 analogs were performed in both water and Tris buffer, to test the influence of the buffer, in which all biological studies were executed, on the peptides conformation. The spectra collected under both conditions were almost identical, with the exception that the spectra in the buffer (not shown) were recorded only to λ = 195 nm because of interference of Tris with the polarized light below this wavelength. The CD spectra of unordered peptides containing more than 10 residues are generally marked by a relatively strong negative band near 200 nm, which may be accompanied by a weak negative band or shoulder at about 220 nm [47]. Spectra of Tat2 analogs with both single and multiple Ala substitutions are characterized by a negative band at 195–198 nm, but interestingly, they present a positive instead of the expected negative band around 220 nm (Figure 2). Because the sequences of the peptides do not contain aromatic residues, which may produce such a positive band, the most probable explanation for this type of a negative–positive sign pattern of the ππ* and nπ* bands, respectively, is predominance of the PPII conformation [48]. PPII is a special type of secondary structure, which is stabilized only by hydrogen bonds with the solvent molecules. It displays characteristic temperature dependence of CD spectra, what was demonstrated for the original Tat2 peptide [29]. In contrast to the FTIR results, there are no distinct signs of any type of a turn conformation in the recorded CD spectra. It was established, however, that, because of usually low intensity of a turn signal in CD spectrum, midsize peptides without constraints are rather poor models of turns [47].
proteasome. Activating capacity was probed with the latent enzyme, whereas inhibitory abilities were investigated using the proteasome already activated by the treatment with 0.005% SDS. Alanine scan with single alanine substitutions Tat2 peptide is a modest activator of the ChT-L peptidase of the latent 20S proteasome, enhancing degradation of the model peptide up to twofold (Figure 3A). Substitution of a single amino acid residue in Tat2 sequence did not reduce the activating capacity of the resulted analogs toward ChT-L peptidase. To the contrary, most of alanine analogs were more efficient activators than the parent Tat2 peptide (Figure 3A). Only the replacement of Asp10 residue diminished capability of the resulting Tat2A10 peptide to activate the latent proteasome. The SDS-activated 20S proteasome responded to changes introduced to the sequence of Tat2 peptide in a more diverse manner. The original Tat2 at 1 μM concentration inhibited the ChT-L activity of the proteasome to less than 20% of the control value (Figure 3B). Ala substitutions that decreased the peptide charge by a single basic residue – Arg or Lys, located either in the N-terminal part (Ta2A1, Tat2A2, Tat2A3, Tat2A4, and Tat2A5) or in the middle of the sequence (Tat2A7 and Tat2A8) – caused almost complete loss of the peptides inhibitory capacity (Figure 3B). In contrast, substitutions of residues from the C-terminal part (Gln9, Asp10, Pro11, and Ile12) and the only nonbasic residue in the middle of the peptide sequence (Gln6) had no effect on the peptides capability to inhibit the proteasome.
Biochemical studies In our previous studies [29], we have established that Tat2 interacts with the CP as a classic noncompetitive inhibitor as only the value of Vmax was lowered in the presence of the peptide, whereas the Km value remained unchanged. We have postulated an allosteric mode of action of the compound. We proposed that it binds to the site other than the active site, and by inducing conformational changes propagating toward the active center, it is able to modulate the proteasome activity. Such actions may result in diverse effects on the enzyme peptidases, not only inhibition but also stimulation, the outcomes that we had observed before with the Tat2 peptide [29]. Therefore, we tested the ability of Tat2 analogs to both activate and inhibit the
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
653
Figure 2. Far-UV CD spectra of Tat2 analogs dissolved in water (peptides concentration 0.2 mg/ml). The results are shown in molar ellipticity units.
Figure 3. Activating (A) and inhibitory (B) effects of Tat2 analogs (concentration: 1 μM) toward the 20S proteasome, with root mean square errors added to each bar. In all graphs, 100% of the relative effect (x-axis) refers to activity (nanomoles of released AMC product per milligram of CP protein per second) recorded with a solvent (DMSO) without the peptides added to a reaction mixture.
WITKOWSKA ET AL. Alanine scan with multiple alanine substitutions The peptides with multiple alanine substitutions, in which from two to four consecutive residues are replaced by Ala, had completely abrogated capability to inhibit the SDS-activated 20S proteasome (Figure 4). Such peptides were also not able to activate the latent 20S to a significant extent (at most 150% of the control; results not shown). Because in peptides Tat2A2-5, Tat2A4-5, Tat2A5-7, and Tat2A7-8, the substitutions substantially diminish the number of basic residues in the sequence and strongly affect the effective charge of the peptides, the crucial role of a positive charge in allosteric interactions of the peptides with the proteasome is implied. This result is also consistent with the outcome of the inhibition profiles of Tat2 peptides with single Ala substitutions. In these cases, less efficient inhibition was observed only for the peptides with the basic residues replaced with Ala (Figure 3B). On the other hand, random coil conformation of the peptides with multiple Ala substitutions may explain the lack of stimulating capabilities of the peptides toward the 20S peptidases.
Discussion In our search for novel and effective proteasome modulators, we focused on established protein ligands affecting activity of the 20S proteasome. HIV-1 Tat protein is one of such ligands [49,50]. We have previously discovered that the peptide RKKRRQRRQDPI (Tat2), consisting of two fragments of Tat protein (RKKRRQRR and QDPI), activated the latent 20S proteasome and noncompetitively inhibited CP activated by 0.005% SDS treatment [29]. The structure-activity studies of Tat2 analogs with single or multiple alanine substitutions allowed us to pinpoint features that affect the potential of the peptides to interact with the proteasome. Circular dichroism spectra do not allow for a detailed analysis of the peptides conformation. All the studied compounds are characterized by similar CD spectra, corresponding to PPII conformation. There are no features typical for turns in the recorded spectra. Their absence is probably linked to a low intensity of bands produced by turns, which is typical for midsize peptides with only weak structural constrains [47].
654
Figure 4. Inhibitory capacity of Tat2 analogs with multiple adjacent residues replaced with Ala. One hundred percent of the relative effect (y-axis) refers to activity (nanomoles of released AMC product per milligram of CP protein per second) recorded with a solvent (DMSO) without the peptides added to a reaction mixture.
Fourier transform infrared spectroscopy delivered much more distinctive information. Studies of the peptides structure by FTIR show that substitution of a single amino acid residue does not significantly change the peptide conformation unless it is the Asp residue. In this case, the spectral features typical for PPII and a turn structure (minima in the second derivative trace at about 1620 and 1680 cm 1, respectively) disappear from the spectrum, and instead, a minimum characteristic for the random coil conformation appears (Figure 1A and B). It seems that Asp residue is actively engaged in stabilization of Tat2 conformation, probably by formation of a salt bridge with the basic residues present in the peptide sequence. Lack of conformational changes associated with substitution of any single Arg/Lys residue indicates that this salt bridge is not very specific and that neighboring residues are able to substitute the exchanged amino acid to create the ionic interactions with Asp. However, when two or more neighboring basic residues are replaced by Ala, the salt bridge stabilizing the ordered conformation of the peptide cannot be formed and the peptides adopt a random coil structure (Figure 1C). Inability to maintain the ordered PPII/turn conformation results in a loss of the peptides activating capabilities toward the ChT-L peptidase of the latent 20S proteasome. None of the compounds with multiple amino acids replaced by Ala and characterized by random coil conformation was able to activate the proteasome. Concomitantly, single Ala substitutions, which do not change the structure of the peptides, do not diminish the peptides power to activate the latent CP. Most of single Ala scan peptides were even better activators than Tat2 itself. Only the peptide Tat2A10, in which Asp residue was replaced with Ala, probably leading to disruption of the salt bridge stabilizing the peptide ordered conformation, was not able to significantly activate the 20S proteasome. These results indicate that there is a correlation between the structure of the peptides and their activating abilities toward the latent proteasome. Likely, PPII and/or turn motifs are important for the peptide interactions with the binding sites, which are able to send allosteric signals enhancing the proteasome activity. Furthermore, we noticed that activating capacity depends on the peptide concentration: compounds, which were able to activate the latent CP at low concentrations (0.5–1.0 μM), at higher concentrations (10 μM) lost their activating capacity or even became the proteasome inhibitors (results not shown). It may indicate that there is a limited number of ‘activating’ spots on the proteasome surface. When they are all saturated, the peptides may bind nonspecifically to other sites impeding conformational changes necessary for the enzyme activation. Inhibitory capacity of Tat2 peptides was very sensitive to the exchange of basic amino acids – Arg and Lys residues. In the Tat2 sequence, there are seven such residues: Arg in positions 1, 4, 5, 7, and 8, and Lys in positions 2 and 3. Substitution of any of them significantly diminished or even completely abolished the ability of the affected peptide to inhibit the SDSactivated 20S proteasome (Figure 3B). On the other hand, introducing Ala in the position of the nonbasic residues, Gln6, Gln9, Asp10, Pro11, or Ile12, did not influence the peptides inhibitory activity. These results suggest that the charge of the peptide is the most important feature for allosteric inhibition of the proteasome. Although our studies aimed to elucidate the mechanism of inhibition/activation of the 20S proteasome by Tat2 and its analogs are still under way, it is reasonable to suspect that these
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION compounds act as allosteric modulators. They likely bind outside the catalytic pocket and send a conformational signal changing the performance of the active sites. Henceforth, Tat2 peptides would join the other already described low-molecular mass noncompetitive regulators of the proteasome. First, PR39 inhibitor was proposed to bind to the α7 subunit and cause a unique destabilizing effect on the motions of the proteasomal α ring [51,52]. Second, TROSY NMR spectroscopy experiments revealed that binding of a small-molecule allosteric inhibitor chloroquine results in a shift in populations of the 20S proteasome conformations [8]. These conformations differ in contiguous regions that connect the α-ring with the active sites. Changes in their populations result in alteration of the substrates proteolysis pattern. Third, rapamycin and its analogs were found to noncompetitively inhibit the 20S proteasome and perturb its conformational equilibrium, studied by atomic force microscopy [28]. When it comes to activators, peptides based on the Cterminal sequence of ATPase complex PAN, the particle homologous to 19S and allosterically regulating the activity of 20S in Archaea, were found in cryo-electron microscopy experiments as residing in the pockets formed between the αsubunits [18]. Homologous peptides derived from the C-termini of eukaryotic 19S ATPases were expected to open the gate of the CP and facilitate degradation of substrates [18]. Finally, hydrophobic peptides were postulated to open the gate and activate the 20S core [53]. Binding of modulators in the distance to the strictly conserved active sites opens an important avenue in the development of more selective therapeutics intended for proteasome inhibition as well as enables the proteasome activation.
Conclusions In summary, free Tat2 and majority of its analogs with a single Ala substitution in liquid preferentially adopt a conformation containing PPII/turn structural motifs. Replacing either Asp10 or two or more adjacent Arg/Lys seems to disrupt ionic interactions responsible for stabilization of the ordered structure of the peptides. The random coil Tat2 analogs lose their ability to activate the latent proteasome. Apparently, the conformation as determined for the free peptides is not indicative of their inhibitory capability. To the contrary, the strong positive charge of the peptides is necessary for them to become proteasome inhibitors. The established structure-activity dependences will be utilized to optimize the structure of peptidomimetics serving as effective allosteric regulators of the 20S proteasome. Acknowledgement This work was supported by the National Science Centre grant no. 2011/01/B/ST5/06616 and Polish Ministry of Science and Higher Education grant no. DS 530-8440-D379-14, William and Ella Owens Medical Research Foundation grant (M.G.), and IIMS Pilot grant (P.A.O.).
References
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
655
1 Laskowski RA, Gerick F, Thornton JM. The structural basis of allosteric regulation in proteins. FEBS Lett. 2009; 583: 1692–1698. 2 Merdanovic M, Mönig T, Ehrmann M, Kaiser M. Diversity of allosteric regulation in proteases. ACS Chem. Biol. 2013; 8: 19–26. 3 Hauske P, Ottmann C, Meltzer M, Ehrmann M, Kaiser M. Allosteric regulation of proteases. Chem. Bio. Chem. 2008; 9: 2929–2928.
4 Zorn JA, Wells JA. Turning enzymes ON with small molecules. Nat. Chem. Biol. 2010; 6: 179–188. 5 Melancon BJ, Hopkins CR, Wood MR, Emmitte KA, Niswender CM, Christopoulos A, Conn PJ, Lindsley CW. Allosteric modulation of seven transmembrane spanning receptors: theory, practice and opportunities for central nervous system drug discovery. J. Med. Chem. 2012; 55: 1445–1464. 6 Nussinov R, Tsai C-J. Allostery in disease and in drug discovery. Cell 2013; 153: 293–305. 7 Mohr K, Schmitz J, Schrage R, Tränkle C, Holzgrabe U. Molecular alliance – from orthosteric and allosteric ligands to dualsteric/bitopic agonists at G protein coupled receptors. Angew. Chem. Int. Ed. 2013; 52: 508–516. 8 Ruschak AM, Kay LE. Proteasome allostery as a population shift between interchanging conformers. Proc. Natl. Acad. Sci. 2012; 109: E3454–E3462. 9 Osmulski PA, Hochstrasser M, Gaczynska M. A tetrahedral transition state at the active sites of the 20S proteasome is coupled to opening of the α-ring channel. Structure 2009; 17: 1137–1147. 10 Śledź P, Förster F, Baumeister W. Allosteric effects in the regulation of 26S proteasome activities. J. Mol. Biol. 2013; 425: 1415–1423. 11 Groll M, Ditzel L, Löwe J, Stock D, Bochtler M, Bartunik HD, Huber R. Structure of 20S proteasome from yeast at 2.4 Å resolution. Nature 1997; 386: 463–471. 12 Dick TP, Nussbaum AK, Deeg M, Heinemeyer W, Groll M, Schirle M, Keilholz W, Stevanović S, Wolf DH, Huber R, Rammensee HG, Schild H. Contribution of proteasomal beta-subunits to the cleavage of peptide substrates analyzed with yeast mutants. J. Biol. Chem. 1998; 273: 25637–46. 13 Jung T, Catalgol B, Grune T. The proteasomal system. Mol. Aspects Med. 2009; 30: 191–296. 14 Groll M, Bajorek M, Kohler A, Moroder L, Rubin DM, Huber R, Glickman MH, Finley D. A gated channel into the proteasome core particle. Nat. Struct. Biol. 2000; 7: 1062–1067. 15 Bajorek M, Glickman MH. Keepers at the final gates: regulatory complexes and gating of the proteasome channel. Cell. Mol. Life Sci. 2004; 61: 1579–1588. 16 Stadtmueller BM, Hill CP. Proteasome activators. Mol. Cell 2011; 41: 8–19. 17 Whitby FG, Masters EI, Kramer L, Knowlton JR, Yao Y, Wang CC, Hill CP. Structural basis for the activation of 20S proteasome by 11S regulators. Nature 2000; 408: 115–120. 18 Rabl J, Smith DM, Yu Y, Chang S-C, Goldberg AL, Cheng Y. Mechanism of gate opening in the 20S proteasome by the proteasomal ATPases. Mol. Cell 2008; 30: 360–368. 19 Sadre-Bazzaz K, Whitby FG, Robinson H, Formosa T, Hill CP. Structure of a Blm10 complex reveals common mechanism for proteasome binding and gate opening. Mol. Cell 2010; 37: 728–735. 20 Jankowska E, Stoj J, Karpowicz P, Osmulski PA, Gaczynska M. The proteasome in health and disease. Curr. Pharm. Des. 2013; 19: 1010–1028. 21 Nalepa G, Rolfe M, Harper JW. Drug discovery in the ubiquitinproteasome system. Nat. Rev. Drug Discov. 2006; 5: 596–613. 22 Groll M, Huber R, Moroder L. The persisting challenge of selective and specific proteasome inhibition. J. Pept. Sci. 2009; 15: 58–66. 23 Kisselev AF, van der Linden WA, Overkleeft HS. Proteasome inhibitors: an expanding army attacking a unique target. Chem. Biol. 2012; 19: 99–115. 24 Driscoll JJ, Burris J, Annunziata CM. Targeting the proteasome with bortezomib in multiple myeloma: update on therapeutic benefit as an upfront single agent, induction regimen for stemcell transplantation and as maintenance therapy. Am. J. Ther. 2012; 19: 133–144. 25 Kortuem KM, Stewart AK. Carfilzomib. Blood 2013; 121: 893–897. 26 Ruschak AM, Slassi M, Kay LE, Schimmer AD. Novel proteasome inhibitors to overcome bortezomib resistance. J. Natl. Cancer Inst. 2011; 103: 1–11. 27 Gaczynska M, Osmulski PA, Gao Y, Post MJ, Simons M. Proline and arginine-rich peptides constitute a novel class of allosteric inhibitors of proteasome activity. Biochemistry 2003; 42: 8663–8670. 28 Gaczynska M, Osmulski PA. Rapamycin allosterically inhibits the proteasome. Mol. Pharmacol. 2013; 84: 104–113. 29 Jankowska E, Gaczynska M, Osmulski P, Sikorska E, Rostankowski R, Madabhushi S, Tokmina-Lukaszewska M, Kasprzykowski F. Potential allosteric inhibitors of the proteasome activity. Biopolymers 2010; 93: 481–495.
WITKOWSKA ET AL. 30 Li X, Wood TE, Sprangers R, Jansen G, Franke NE, Mao X, Wang X, Zhang Y, Verbrugge SE, Adomat H, Li ZH, Trudel S, Chen C, Religa TL, Jamal N, Messner H, Cloos J, Rose DR, Navon A, Guns E, Batey RA, Kay LE, Schimmer AD. Effect of noncompetitive proteasome inhibition on bortezomib resistance. J. Natl. Cancer Inst. 2010; 102: 1069–1082. 31 Azevedo LM, Landsdell TA, Ludwig JR, Mosey RA, Woloch DK, Cognan DP, Patten GP, Kuszpit MR, Fisk JS, Tepe JJ. Inhibition of the human proteasome by imidazoline scaffolds. J. Med. Chem. 2013; 56: 5974–5978. 32 Seo H, Sonntag K-C, Kim W, Cattaneo E, Isacson O. Proteasome activator enhances survival of Huntington’s disease neuronal model cells. PLoS One 2007; e238: 1–7. 33 Dal Vechio FH, Cerqueira F, Augusto O, Lopes R, Demasi M. Peptides that activate the 20S proteasome by gate opening increased oxidized protein removal and reduced protein aggregation. Free Rad. Biol. Med. 2014; 67: 304–313. 34 Wilk S, Chen W-E, Wójcik C, Magnusson RP. In Proteasomes: The World of Regulatory Proteolysis, Hilt W, Wolf DH, Landes Bioscience, Georgetown, 2000; pp. 143–150. 35 Watanabe N, Yamada S. Activation of 20S proteasomes from spinach leaves by fatty acids. Plant Cell Physiol. 1996; 37: 147–151. 36 Huang L, Ho P, Chen C-H. Activation and inhibition of proteasomes by betulinic acid and its derivatives. FEBS Lett. 2007; 581: 4955–4959. 37 Dang Z, Jung K, Qian K, Lee K-H, Huang L, Chen C-H. Synthesis of lithocholic acid derivatives as proteasome regulators. ACS Med. Chem. Lett. 2012; 3: 925–930. 38 Savitzky A, Golay A. Soothing and differentiation of data by simplified least squares procedures. Anal. Chem. 1964; 36: 1627–1639. 39 Wang S-L, Wei Y-S, Lin S-Y. Subtractive similarity method used to study the infrared spectra of proteins in aqueous solution. Vibr. Spectrosc. 2003; 31: 313–319. 40 Chittur KK. FTIR and protein structure at interfaces. BMES Bull. 1999; 23: 3–9. 41 Smith BC. Fundamentals of Fourier Transform Infrared Spectroscopy. CRC Press: Boca Raton London New York, 2011; pp. 56–62. 42 Valenti LE, Paci MB, De Pauli CP, Giacomelli CE. Infrared study of trifluoroacetic acid unpurified synthetic peptides in aqueous solution: trifluoroacetic acid removal and band assignment. Anal. Biochem. 2011; 410: 118–123.
43 Surewicz WK, Mantsch HH, Chapman D. Determination of protein secondary structure by Fourier transform infrared spectroscopy: a critical assessment. Biochemistry 1993; 32: 389–394. 44 Andrushchenko VV, Vogel HJ, Prenner E. Optimization of hydrochloric acid concentration used for trifluoroacetate removal from synthetic peptides. J. Pept. Sci. 2007; 13: 37–43. 45 Elangovan S, Margolis HC, Oppenheim FG, Beniash E. Conformational changes in salivary proline-rich protein 1 upon adsorption to calcium phosphate crystals. Langmuir 2007; 23: 11200–11205. 46 Kong J, Yu S. Fourier transform infrared spectroscopic analysis of protein secondary structures. Acta Biochim. Biophys. Sin. 2007; 39: 549–559. 47 Perczel A, Hollósi M. In Circular Dichroism and the Conformational Analysis of Biomolecules, Fasman GD. Plenum Press: New York and London, 1996; pp. 285–380. 48 Sreerama N, Woody NW. Poly(Pro)II helices in globular proteins: Identification and circular dichroic analysis. Biochemistry 1994; 33: 10022–10025. 49 Huang X, Seifert U, Salzmann U, Henklein P, Preissner R, Henke W, Sijts AJ, Kloetzel P-M, Dubiel W. The RTP site shared by the HIV-1 Tat protein and the 11S regulator subunit α is crucial for their effects on proteasome function including antigen processing. J. Mol. Biol. 2002; 323: 771–782. 50 Seeger M, Ferrel K, Frank R, Dubiel W. HIV-1 Tat inhibits the 20S proteasome and its 11S regulator-mediated activation. J. Biol. Chem. 1997; 272: 8145–8148. 51 Gao YH, Lecker S, Post MJ, Hietaranta AJ, Li J, Volk R, Li M, Sato K, Saluja AK, Steer ML, Goldberg AL, Simons M. Inhibition of ubiquitin-proteasome pathway-mediated IκBα degradation by a naturally occurring antibacterial peptide. J. Clin. Invest. 2000; 106: 439–448. 52 Gaczynska M, Osmulski PA, Gao Y, Post MJ, Simons M. Proline- and arginine-rich peptides constitute a novel class of allosteric inhibitors of proteasome activity. Biochemistry 2003; 42: 8663–8670. 53 Kisselev AF, Kaganovich D, Goldberg AL. Binding of hydrophobic peptides to several non-catalytic sites promotes peptide hydrolysis by all active sites of 20 S proteasomes – evidence for peptideinduced channel opening in the alpha-rings. J. Biol. Chem. 2002; 277: 22260–22270.
656 wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
Revised: 24 March 2014
Accepted: 31 March 2014
Published online in Wiley Online Library: 13 May 2014
(wileyonlinelibrary.com) DOI 10.1002/psc.2642
Dissecting a role of a charge and conformation of Tat2 peptide in allosteric regulation of 20S proteasome Julia Witkowska,a Przemysław Karpowicz,a Maria Gaczynska,b Pawel A. Osmulskib and Elżbieta Jankowskaa* Proteasome is a ‘proteolytic factory’ that constitutes an essential part of the ubiquitin-proteasome pathway. The involvement of proteasome in regulation of all major aspects of cellular physiology makes it an attractive drug target. So far, only inhibitors of the proteasome entered the clinic as anti-cancer drugs. However, proteasome regulators may also be useful for treatment of inflammatory and neurodegenerative diseases. We established in our previous studies that the peptide Tat2, comprising the basic domain of HIV-1 Tat protein: R49KKRRQRR56, supplemented with Q66DPI69 fragment, inhibits the 20S proteasome in a noncompetitive manner. Mechanism of Tat2 likely involves allosteric regulation because it competes with the proteasome natural 11S activator for binding to the enzyme noncatalytic subunits. In this study, we performed alanine walking coupled with biological activity measurements and FTIR and CD spectroscopy to dissect contribution of a charge and conformation of Tat2 to its capability to influence peptidase activity of the proteasome. In solution, Tat2 and most of its analogs with a single Ala substitution preferentially adopted a conformation containing PPII/turn structural motifs. Replacing either Asp10 or two or more adjacent Arg/Lys residues induced a random coil conformation, probably by disrupting ionic interactions responsible for stabilization of the peptides ordered structure. The random coil Tat2 analogs lost their capability to activate the latent 20S proteasome. In contrast, inhibitory properties of the peptides more significantly depended on their positive charge. The data provide valuable clues for the future optimization of the Tat2-based proteasome regulators. Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. Keywords: proteasome; allosteric inhibition/activation; alanine scan; circular dichroism; Fourier transform infrared spectroscopy
Introduction
J. Pept. Sci. 2014; 20: 649–656
* Correspondence to: Elżbieta Jankowska, Department of Medicinal Chemistry, Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80-308 Gdańsk, Poland. E-mail: [email protected] a Department of Medicinal Chemistry, Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80-308 Gdańsk, Poland b Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, San Antonio, TX, USA Abbreviations: ChT-L, chymotrypsin-like; CD, circular dichroism; CP, core particle (20S proteasome); FTIR, Fourier transform infrared; HIV-1 Tat, human immunodeficiency virus type 1 transactivator protein; MW, molecular weight; PA, proteasome activator; PAN, proteasome activator nucleotide-activated; PGPH, post-glutamyl peptide hydrolyzing; PPII, polyproline II; TFA, trifluoroacetic acid; T-L, trypsin-like; UPS, ubiquitin-proteasome system.
Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd.
649
Allosteric regulation of enzymes enables modulation of their activity through binding an effector to the site distant from the active site and inducing structural changes, which can propagate to the active center, modifying conformation around it and, as a result, the enzyme affinity toward natural substrates [1]. The fact that binding of an allosteric effector proceeds far away from the classic orthosteric binding sites has important implications. First of all, active site topologies are often strongly conserved between distinct members of the enzyme family, so they are difficult to address in a selective manner by low MW compounds. Allosteric sites are usually less conserved; therefore, they may constitute better drug targets [2,3]. Second, binding an orthosteric ligand blocks the active site leading to the enzyme inhibition. In the case of allosteric regulators, not interacting directly with the active sites, more diverse effects on the enzyme properties, such as inhibition, activation, or specificity modulation, are achievable [2,4]. Accordingly, there is an increasing interest in utilizing the allosteric modulation phenomena in drug design and introducing allosteric drugs to the clinic [5–7]. The proteasome is multicatalytic and multisubunit protease complex amenable to allosteric regulation [8–10]. The 20S CP is formed by four heptameric rings arranged in αββα manner and creating together a chamber, which harbors the catalytic sites [11]. The active centers located in β5/5′, β2/2′, and β1/1′ subunits are responsible for the ChT-L, T-L, and PGPH proteasomal
peptidases, respectively [12,13]. Access of proteins and peptides to the catalytic chamber is restricted by the N-termini of α subunits, which form a regulated gate [14]. Three classes of activators: 11S, PA200, and 19S facilitate passage of substrates to the proteasome interior by binding to the α subunits and repositioning structural motifs blocking the enzyme entrance pore [15,16]. The 11S activator (PA28/REG, Trypanosoma brucei analog – PA26) is a heptameric protein that binds to the proteasome by inserting its seven C-terminal tails to the seven pockets located between the α subunits. To reorganize the α-tails and stabilize the opened conformation of the gate, 11S utilizes an
WITKOWSKA ET AL.
650
internal ‘activation loop’ [17]. Two other types of activators: PA200 (Blm10 in Saccharomyces cerevisiae) and 19S/PA700 (PAN in Archaea) contain a conserved C-terminal hydrophobictyrosine-X motif sufficient to open the gate upon binding within the pockets between neighboring α subunits [18]. Unlike the oligomeric 11S and 19S/PAN activators, which use multiple C-termini to bind to the proteasome α ring, Blm10 and its mammalian homolog PA200 are single chain proteins and can interact with only one pocket between α subunits, what results in a partially opened disordered conformation of the gate [19]. The UPS is a major contributor to the intracellular protein degradation machinery in mammalian cells. By degradation of short-lived regulatory proteins, the enzyme has an impact on many vital processes such as a cell cycle, DNA repair, sodium channel function, regulation of immune and inflammatory responses, and many others. Moreover, UPS is also responsible for disposal of any abnormal proteins – oxidized, unstructured, misfolded, or viral. Malfunction of the UPS can lead to serious disorders including cardiovascular diseases, certain cancers, muscular dystrophy, neurodegeneration, and systemic autoimmunity and makes the proteasome an important target for pharmacological intervention [20,21]. Numerous small competitive inhibitors of the proteasome have been already developed and several of them are currently under clinical evaluation [22,23]. Two inhibitors, bortezomib and carfilzomib, are approved for use with relapsed/refractory multiple myeloma and mantle cell lymphoma [24,25]. Although both compounds demonstrated efficacy in the treatment of these hematological malignancies, a significant number of patients do not respond to these drugs or develop the drug resistance [26]. We postulate that allosteric modulators may be a promising alternative to classic orthosteric ligands [27–29]. Because of its distinct mode of action, allosteric inhibitors, working alone or in combination with other drugs, may improve efficacy of the treatment, limit off-target effects, and allow to overcome resistance to the orthosteric drugs. Importantly, because allosteric effectors bind at noncatalytic sites that usually are under much lower evolutionary pressure, they themselves exhibit substantially decreased potential to quickly induce resistance. Recently, several allosteric inhibitors have been identified and extensively characterized. 5-amino8-hydroxyquinoline has been shown by NMR to bind to the α-subunits inside the proteasome inner chamber, and it was able to overcome some forms of bortezomib resistance in vitro [30]. Other promising noncompetitive inhibitors are imidazoline scaffolds with the best compound displaying submicromolar inhibitory capacity [31]. In addition to their canonical mTOR target (mammalian target of rapamycin protein), rapamycin and its double or single domain analogs in vitro also productively interact with the α face of 20S proteasome leading to inhibition of its peptidase and proteinase activity [28]. These compounds change the abundance of conformers with the open gate and compete with the 19S cap for the binding sites on the α face [28]. Imidazoline, quinoline, and rapamycin based compounds may overcome resistance of selected cancer cells to bortezomib. The outcome of allosteric interactions is not limited to inhibition but also extends to activation of the proteasome, and this property may be utilized to increase degradation of damaged (for instance, oxidized as a result of oxidative stress) proteins and reduce their aggregation [32,33]. However, small molecules that can activate or enhance the activity of the 20S CP are scarce and not well studied. The compounds already recognized as being able to activate the 20S proteasome are SDS, certain lipids, and polycations [34].
Unfortunately, these substances are either too weak or lack druglike properties to be considered as potential therapeutics or leading structures. In addition, SDS while activating the proteasome at low concentrations, at higher ones becomes its irreversible inhibitor [35]. More promising activators seem to be fatty acids, betulinic acid, and lithocholic acid derivatives [35–37]. They display some specificity toward the proteasome peptidases: oleic, linoleic, and linolenic acids activated ChT-L and PGPH peptidases of 20S proteasome while inhibiting its T-L activity, betulinic acid preferentially activated the ChT-L [35,36]. During our search for small synthetic compounds that can serve as allosteric regulators of the proteasome activity, we focused on sequence fragments of HIV-1 Tat protein. HIV-1 Tat protein inhibits the activity of the 20S proteasome, competes with the 11S activator in binding to the proteasome α face, and markedly influences major histocompatibility complex class I-associated antigen presentation [30,30]. Tat2 peptide has a sequence R49KKRRQRR56Q66DPI69, which is derived from the basic domain of HIV-1 Tat, supplemented by the QDPI fragment. The peptide sequence was developed to comprise residues Lys51, Arg52, and Asp67, relevant to Glu235, Lys236, and Lys239 in the PA28 activator and postulated to constitute the proteasome binding region [30]. Our previous studies demonstrated that Tat2 noncompetitively inhibits the activated CP [29]. In addition, Tat2 stimulates ChT-L activity of the latent CP in a relatively narrow range of the peptide concentrations, reaching the maximum effect at about 1 μM concentration. We determined that Tat2 is not degraded by the proteasome and, likewise to the full-length HIV-1 Tat protein, competes with the PA28 activator for binding sites on the proteasome α face [29]. To optimize the structure of Tat2-based proteasome regulators, we aimed at identification of Tat2 proteasome-targeting pharmacophore. To assign relative functional importance to each amino acid residues within the peptide, we performed alanine scan, substituting residues either one by one or in blocks. Subsequently, we characterized the structural properties of the resulting peptides by CD and FTIR spectroscopy and tested the effects of treatments with the peptides on the proteasome peptidase activity. The results suggest that the inhibitory and activating capabilities of Tat2 peptides are separately modulated by charge and particular conformation, respectively.
Materials and Methods Peptide synthesis and purification All peptides were synthesized on TentaGel R PHB resin with loading capacity of 0.2 mmol/g (Rapp Polymere). Syntheses were performed using Fmoc strategy on a Millipore model 9050 Plus peptide synthesizer with continuous-flow methodology or using microwave radiation (Plazmatronika RM800; formerly Plazmatronika, now Ertec, Wroc∤aw, Poland) under nitrogen atmosphere. The first amino acids were attached to the resin using symmetrical anhydride method. The efficiency of the coupling was determined by the measurement of the resulting fulvene-piperidine adduct at λ = 301 nm. A N2 line was attached to a microwave vessel for continuous agitation of the resin. Coupling reactions were performed in the following conditions: power 24 W, maximum temperature 80 °C, and irradiation time 4.5 min. Every deprotection step was repeated twice using power 16 W, maximum temperature 60 °C, and irradiation time 5 and 3.5 min, respectively. The actual
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION temperature of the reactions was measured and found to be 71–82 °C for the coupling cycles and 53–63 °C for the deprotection steps. Crude peptides were purified by RP-HPLC using a C8 semipreparative Luna column (21.2 × 250 mm, 5 μm; Phenomenex Torrance, CA, USA). A linear gradient of acetonitrile in 0.1% aqueous TFA or acetonitrile in triethylamine phosphate containing buffers at pH 3 were used as a mobile phase. The MWs of the peptides were ascertained by MALDI TOF or electrospray ionization ion trap TOF liquid chromatography MS with C12 Jupiter Proteo column (150 × 2 mm, 4 μ, 90 Å; Phenomenex). The purity of the synthesized compounds was confirmed by analytical RP-HPLC using a C8 Kromasil column (4.6 × 250 mm, 5 μm Phenomenex) and a 30 min linear gradient of 5–80% acetonitrile in 0.1% aqueous TFA. UV absorption was observed at λ = 223 nm. FTIR spectroscopy The peptides were dissolved in H2O to a final concentration of 20 mg/ml. TFA solutions were prepared from 2% aqueous stock solution by consecutive dilutions allowing to obtain 1.5%, 1.25%, 1.0%, 0.8%, 0.4%, 0.2%, 0.1%, 0.05%, 0.025%, and 0.0125% TFA concentration. Spectra of water, as well as aqueous solutions of TFA and the peptides, were obtained by spotting 15 μl of each sample onto a CaF2 window, which was then assembled into the infrared cell (the thickness of a spacer was 0.006 mm). FTIR spectra were collected with a Bruker IFS66 spectrophotometer (Bruker Optic GmbH, Bremen, Germany) with a Deuterated Tri Glycine Sulfate detector. During data collection, the spectrometer was continuously purged with dry air free of CO2. Typically, 160 scans at 4 cm 1 resolution were averaged for each spectrum. At least three independent measurements were performed for each sample. Data processing was executed with GRAMS 5.0 (Galactic Enterprises). The water spectrum, recorded with the same accessory and instrument conditions, was subtracted from the spectra of aqueous solutions of TFA and the peptides. The appropriate spectrum of pure TFA (deprived of water contribution) was then subtracted from each peptide spectrum. The resulted spectra were baseline-corrected. The spectral second derivative was calculated using Sawitzky–Golay algorithm with 13-point smoothing function [38]. The removal of TFA and water contributions from the spectra was accomplished in the following steps:
For CD measurements, 1 mm path length quartz cuvettes were used. Protein samples were dissolved in H2O or in 50 mM Tris/HCl, pH 8.0 to a final concentration of 0.2 mg/ml. Spectra were acquired at 25 °C on a Jasco J-815 spectrometer in a range 185–260 nm for aqueous solutions and 195–260 nm for solutions of peptides in the buffer. Each spectrum is a result of three independent measurements. Biochemical studies Influence of the peptides on the catalytic properties of the 20S proteasome was tested with housekeeping CP purified from human erythrocytes (Enzo Life Sciences). The latent CP or CP activated with 0.005% SDS at a final concentration of 1.5 nM was utilized. The model fluorogenic peptide substrate succinyl-LeuLeuValTyr-4-methylcoumarin-7-amide (SucLLVY-MCA, Bachem) was employed to determine the proteasome ChT-L activity. Stock solutions of the substrate and the tested peptides were prepared in DMSO. The content of DMSO never exceeded 3% of the final reaction volume. All assays were performed in the 96-well plate format using a reaction volume of 100 μl. Tests were carried out in 50 mM Tris/HCl (pH 8.0) reaction buffer. Substrate was added at 100 μM final concentration (1% of the volume). The release of an aminomethylcoumarin (AMC) group was followed by measuring fluorescence emission at 460 nm in 2-min intervals for up to 60 min at 37 °C (Fluoroskan Ascent). The peptidolytic activity was calculated as nanomoles of the released AMC product per milligram of CP per second.
Results Peptides Syntheses To dissect regions/structural features, which influence Tat2 peptide interactions with the proteasome, we have performed a systematic alanine scan by replacing consecutive residues in the peptide sequence by alanine residues. The primary structures of the synthesized analogs are listed in Table 1. Structural Studies FTIR. Subtraction strategy of TFA contribution. Peptides synthesized on solid phase are obtained as TFA salts because they are cleaved from the resin with TFA treatment, and furthermore, they are purified in the presence of TFA. The main TFA band, associated with vibration of COO group, appears in FTIR spectra at about 1672 cm 1 [43,44], overlapping with the amide I band (1600–1700 cm 1) that corresponds to a peptide bond C=O stretching vibration, which is usually utilized to determine the secondary structure of peptides and proteins. To numerically remove contribution of TFA from the peptides’ spectra, we utilized one of the procedures proposed by Valenti et al. [42]. They suggested subtraction of TFA and water spectra separately as a method of choice only for the spectra collected as independent measurements or for different buffer solutions. In other cases, the simplified methodology of one step subtraction of water and TFA contribution was proposed. We found, however, that for different peptides, containing different amount of water in their lyophilized form, as well as different amount of TFA counter ions due to differences in the sequence, it is impossible to apply the same scaling factor as a subtraction coefficient for
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
651
1. Spectra of TFA in water were individually collected for several TFA concentrations in the range of 0.0125–2%. 2. Water contribution was removed by subtracting the scaled water spectrum from the spectra recorded for aqueous solution of TFA until the flat baseline was obtained between 1750 and 1950 cm 1 [39–41]. The scaling factor was determined individually for each case. 3. Water contribution from the peptides spectra was removed exactly as it was described earlier for TFA producing the buffer-corrected spectra of the peptides. 4. The spectra of pure TFA of appropriate concentration were subtracted from the buffer-corrected spectra of the peptides. Selection of the most appropriate TFA spectrum was carried out to meet the criterion of a complete removal of TFA contribution. Correct subtraction of TFA contribution was assessed by disappearance of the TFA bands at 1200 and 1147 cm 1 arising from the C-F stretching vibration [42].
CD spectroscopy
WITKOWSKA ET AL. Table 1. Synthesized Tat2 analogs Name Tat2 Tat2A1 Tat2A2 Tat2A3 Tat2A4 Tat2A5 Tat2A6 Tat2A7 Tat2A8 Tat2A9 Tat2A10 Tat2A11 Tat2A12 Tat2A2-5 Tat2A4-5 Tat2A5-7 Tat2A7-8
Sequence
Retention time (min)
MW (g/mol)
Molecular ion (m/z)
11.5 11.6 11.2 11.5 11.7 11.2 11.6 11.4 12.2 11.5 11.5 11.3 10.3 12.4 12.2 11.5 12.4
1636.9 1551.8 1579.8 1579.8 1551.8 1551.8 1579.9 1551.8 1551.8 1579.9 1592.9 1610.9 1594.8 1352.5 1466.7 1409.7 1466.7
1636.1 1551.2* 1579.2 1579.5 1551.2 1550.7 1579.8* 1551.4 1551.0 1579.6 1592.6 1611.8 1593.8 1351.8 1466.9 1409.4 1466.4
H-RKKRRQRRQDPI-OH H-AKKRRQRRQDPI-OH H-RAKRRQRRQDPI-OH H-RKARRQRRQDPI-OH H-RKKARQRRQDPI-OH H-RKKRAQRRQDPI-OH H-RKKRRARRQDPI-OH H-RKKRRQARQDPI-OH H-RKKRRQRAQDPI-OH H-RKKRRQRRADPI-OH H-RKKRRQRRQAPI-OH H-RKKRRQRRQDAI-OH H-RKKRRQRRQDPA-OH H-RAAAAQRRQDPI-OH H-RKKAAQRRQDPI-OH H-RKKRAAARQDPI-OH H-RKKRRQAAQDPI-OH
Analytical RP-HPLC analyses were performed using a C8 Kromasil column (4.6 × 250 mm, 5 μm) and a 30 min linear gradient of 5–80% acetonitrile in 0.1% aqueous TFA. The molecular weights of the peptides were ascertained by MALDI TOF or ESI IT TOF LC mass spectrometry with C12 Jupiter Proteo column (150 × 2 mm, 4 μ, 90 Å; Phenomenex). * Molecular ion detected by MALDI MS.
TFA and water removal from each spectrum. Despite the fact that the spectra of our peptides were collected with identically prepared samples, the simplified methodology was either not able to ensure the complete removal of the buffer contribution or caused over-subtraction in some regions, introducing artifacts to the spectrum. We determined that, under our experimental conditions, the most effective was the procedure in which water and TFA contributions were removed from the spectra separately and the order of subtraction was not important. The detailed description of the procedure is provided in Materials and Methods section. Structure of the Peptides
652
Fourier transform infrared spectroscopy. Almost all Tat2 analogs with a single Ala substitution displayed similar features of their FTIR spectra. Second derivative of the spectra showed two deep minima – around 1620 and 1680 cm 1 (Figure 1A). The first minimum suggests hydrated PPII conformation, whereas minima between 1660 and 1685 cm 1 are usually attributed to turns [45,46]. The only analog, which did not fit to this PPII-turn pattern, was Tat2A10, in which only one minimum, although with some minor features at various wavenumbers, was observed (Figure 1B). Location of the minimum at ~1645 cm 1 suggests that this peptide in liquid does not acquire preferentially any distinct conformation but rather represents a random coil structure [46]. Because in this peptide, the Asp residue was substituted with Ala, we suspect that this residue may be important for stabilization of Tat2 peptide conformation. Peptides Tat2A2-5, Tat2A4-5, Tat2A5-7, and Tat2A7-8, in which two or more consecutive residues were replaced with Ala, displayed FTIR spectra similar to Tat2A10 (Figure 1C). Their second derivative spectra exhibited only one main minimum located at about 1645 cm 1, what again suggests random coil conformation. In these Tat2 analogs, substitutions diminished the number of basic residues, while in the Tat2A10 variant,
Figure 1. Second derivative of FTIR spectra of the investigated Tat2 analogs: (A) peptides with single Ala substitutions, (B) analog with Asp10 residue exchanged for Ala, and (C) peptides with multiple Ala substitutions. Peptides concentration: 20 mg/ml.
alanine residue replaced an acidic aspartic acid. Loss of the preferred conformation in both types of peptides underlines the role of basic and acidic residues in stabilization of the peptides conformation. Most probably Asp and Arg/Lys residues create a salt bridge, which imposes a more rigid structure of the peptide.
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION Circular dichroism. Circular dichroism studies of Tat2 analogs were performed in both water and Tris buffer, to test the influence of the buffer, in which all biological studies were executed, on the peptides conformation. The spectra collected under both conditions were almost identical, with the exception that the spectra in the buffer (not shown) were recorded only to λ = 195 nm because of interference of Tris with the polarized light below this wavelength. The CD spectra of unordered peptides containing more than 10 residues are generally marked by a relatively strong negative band near 200 nm, which may be accompanied by a weak negative band or shoulder at about 220 nm [47]. Spectra of Tat2 analogs with both single and multiple Ala substitutions are characterized by a negative band at 195–198 nm, but interestingly, they present a positive instead of the expected negative band around 220 nm (Figure 2). Because the sequences of the peptides do not contain aromatic residues, which may produce such a positive band, the most probable explanation for this type of a negative–positive sign pattern of the ππ* and nπ* bands, respectively, is predominance of the PPII conformation [48]. PPII is a special type of secondary structure, which is stabilized only by hydrogen bonds with the solvent molecules. It displays characteristic temperature dependence of CD spectra, what was demonstrated for the original Tat2 peptide [29]. In contrast to the FTIR results, there are no distinct signs of any type of a turn conformation in the recorded CD spectra. It was established, however, that, because of usually low intensity of a turn signal in CD spectrum, midsize peptides without constraints are rather poor models of turns [47].
proteasome. Activating capacity was probed with the latent enzyme, whereas inhibitory abilities were investigated using the proteasome already activated by the treatment with 0.005% SDS. Alanine scan with single alanine substitutions Tat2 peptide is a modest activator of the ChT-L peptidase of the latent 20S proteasome, enhancing degradation of the model peptide up to twofold (Figure 3A). Substitution of a single amino acid residue in Tat2 sequence did not reduce the activating capacity of the resulted analogs toward ChT-L peptidase. To the contrary, most of alanine analogs were more efficient activators than the parent Tat2 peptide (Figure 3A). Only the replacement of Asp10 residue diminished capability of the resulting Tat2A10 peptide to activate the latent proteasome. The SDS-activated 20S proteasome responded to changes introduced to the sequence of Tat2 peptide in a more diverse manner. The original Tat2 at 1 μM concentration inhibited the ChT-L activity of the proteasome to less than 20% of the control value (Figure 3B). Ala substitutions that decreased the peptide charge by a single basic residue – Arg or Lys, located either in the N-terminal part (Ta2A1, Tat2A2, Tat2A3, Tat2A4, and Tat2A5) or in the middle of the sequence (Tat2A7 and Tat2A8) – caused almost complete loss of the peptides inhibitory capacity (Figure 3B). In contrast, substitutions of residues from the C-terminal part (Gln9, Asp10, Pro11, and Ile12) and the only nonbasic residue in the middle of the peptide sequence (Gln6) had no effect on the peptides capability to inhibit the proteasome.
Biochemical studies In our previous studies [29], we have established that Tat2 interacts with the CP as a classic noncompetitive inhibitor as only the value of Vmax was lowered in the presence of the peptide, whereas the Km value remained unchanged. We have postulated an allosteric mode of action of the compound. We proposed that it binds to the site other than the active site, and by inducing conformational changes propagating toward the active center, it is able to modulate the proteasome activity. Such actions may result in diverse effects on the enzyme peptidases, not only inhibition but also stimulation, the outcomes that we had observed before with the Tat2 peptide [29]. Therefore, we tested the ability of Tat2 analogs to both activate and inhibit the
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
653
Figure 2. Far-UV CD spectra of Tat2 analogs dissolved in water (peptides concentration 0.2 mg/ml). The results are shown in molar ellipticity units.
Figure 3. Activating (A) and inhibitory (B) effects of Tat2 analogs (concentration: 1 μM) toward the 20S proteasome, with root mean square errors added to each bar. In all graphs, 100% of the relative effect (x-axis) refers to activity (nanomoles of released AMC product per milligram of CP protein per second) recorded with a solvent (DMSO) without the peptides added to a reaction mixture.
WITKOWSKA ET AL. Alanine scan with multiple alanine substitutions The peptides with multiple alanine substitutions, in which from two to four consecutive residues are replaced by Ala, had completely abrogated capability to inhibit the SDS-activated 20S proteasome (Figure 4). Such peptides were also not able to activate the latent 20S to a significant extent (at most 150% of the control; results not shown). Because in peptides Tat2A2-5, Tat2A4-5, Tat2A5-7, and Tat2A7-8, the substitutions substantially diminish the number of basic residues in the sequence and strongly affect the effective charge of the peptides, the crucial role of a positive charge in allosteric interactions of the peptides with the proteasome is implied. This result is also consistent with the outcome of the inhibition profiles of Tat2 peptides with single Ala substitutions. In these cases, less efficient inhibition was observed only for the peptides with the basic residues replaced with Ala (Figure 3B). On the other hand, random coil conformation of the peptides with multiple Ala substitutions may explain the lack of stimulating capabilities of the peptides toward the 20S peptidases.
Discussion In our search for novel and effective proteasome modulators, we focused on established protein ligands affecting activity of the 20S proteasome. HIV-1 Tat protein is one of such ligands [49,50]. We have previously discovered that the peptide RKKRRQRRQDPI (Tat2), consisting of two fragments of Tat protein (RKKRRQRR and QDPI), activated the latent 20S proteasome and noncompetitively inhibited CP activated by 0.005% SDS treatment [29]. The structure-activity studies of Tat2 analogs with single or multiple alanine substitutions allowed us to pinpoint features that affect the potential of the peptides to interact with the proteasome. Circular dichroism spectra do not allow for a detailed analysis of the peptides conformation. All the studied compounds are characterized by similar CD spectra, corresponding to PPII conformation. There are no features typical for turns in the recorded spectra. Their absence is probably linked to a low intensity of bands produced by turns, which is typical for midsize peptides with only weak structural constrains [47].
654
Figure 4. Inhibitory capacity of Tat2 analogs with multiple adjacent residues replaced with Ala. One hundred percent of the relative effect (y-axis) refers to activity (nanomoles of released AMC product per milligram of CP protein per second) recorded with a solvent (DMSO) without the peptides added to a reaction mixture.
Fourier transform infrared spectroscopy delivered much more distinctive information. Studies of the peptides structure by FTIR show that substitution of a single amino acid residue does not significantly change the peptide conformation unless it is the Asp residue. In this case, the spectral features typical for PPII and a turn structure (minima in the second derivative trace at about 1620 and 1680 cm 1, respectively) disappear from the spectrum, and instead, a minimum characteristic for the random coil conformation appears (Figure 1A and B). It seems that Asp residue is actively engaged in stabilization of Tat2 conformation, probably by formation of a salt bridge with the basic residues present in the peptide sequence. Lack of conformational changes associated with substitution of any single Arg/Lys residue indicates that this salt bridge is not very specific and that neighboring residues are able to substitute the exchanged amino acid to create the ionic interactions with Asp. However, when two or more neighboring basic residues are replaced by Ala, the salt bridge stabilizing the ordered conformation of the peptide cannot be formed and the peptides adopt a random coil structure (Figure 1C). Inability to maintain the ordered PPII/turn conformation results in a loss of the peptides activating capabilities toward the ChT-L peptidase of the latent 20S proteasome. None of the compounds with multiple amino acids replaced by Ala and characterized by random coil conformation was able to activate the proteasome. Concomitantly, single Ala substitutions, which do not change the structure of the peptides, do not diminish the peptides power to activate the latent CP. Most of single Ala scan peptides were even better activators than Tat2 itself. Only the peptide Tat2A10, in which Asp residue was replaced with Ala, probably leading to disruption of the salt bridge stabilizing the peptide ordered conformation, was not able to significantly activate the 20S proteasome. These results indicate that there is a correlation between the structure of the peptides and their activating abilities toward the latent proteasome. Likely, PPII and/or turn motifs are important for the peptide interactions with the binding sites, which are able to send allosteric signals enhancing the proteasome activity. Furthermore, we noticed that activating capacity depends on the peptide concentration: compounds, which were able to activate the latent CP at low concentrations (0.5–1.0 μM), at higher concentrations (10 μM) lost their activating capacity or even became the proteasome inhibitors (results not shown). It may indicate that there is a limited number of ‘activating’ spots on the proteasome surface. When they are all saturated, the peptides may bind nonspecifically to other sites impeding conformational changes necessary for the enzyme activation. Inhibitory capacity of Tat2 peptides was very sensitive to the exchange of basic amino acids – Arg and Lys residues. In the Tat2 sequence, there are seven such residues: Arg in positions 1, 4, 5, 7, and 8, and Lys in positions 2 and 3. Substitution of any of them significantly diminished or even completely abolished the ability of the affected peptide to inhibit the SDSactivated 20S proteasome (Figure 3B). On the other hand, introducing Ala in the position of the nonbasic residues, Gln6, Gln9, Asp10, Pro11, or Ile12, did not influence the peptides inhibitory activity. These results suggest that the charge of the peptide is the most important feature for allosteric inhibition of the proteasome. Although our studies aimed to elucidate the mechanism of inhibition/activation of the 20S proteasome by Tat2 and its analogs are still under way, it is reasonable to suspect that these
wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
ROLE OF TAT2 CHARGE/STRUCTURE IN PROTEASOME ALLOSTERIC MODULATION compounds act as allosteric modulators. They likely bind outside the catalytic pocket and send a conformational signal changing the performance of the active sites. Henceforth, Tat2 peptides would join the other already described low-molecular mass noncompetitive regulators of the proteasome. First, PR39 inhibitor was proposed to bind to the α7 subunit and cause a unique destabilizing effect on the motions of the proteasomal α ring [51,52]. Second, TROSY NMR spectroscopy experiments revealed that binding of a small-molecule allosteric inhibitor chloroquine results in a shift in populations of the 20S proteasome conformations [8]. These conformations differ in contiguous regions that connect the α-ring with the active sites. Changes in their populations result in alteration of the substrates proteolysis pattern. Third, rapamycin and its analogs were found to noncompetitively inhibit the 20S proteasome and perturb its conformational equilibrium, studied by atomic force microscopy [28]. When it comes to activators, peptides based on the Cterminal sequence of ATPase complex PAN, the particle homologous to 19S and allosterically regulating the activity of 20S in Archaea, were found in cryo-electron microscopy experiments as residing in the pockets formed between the αsubunits [18]. Homologous peptides derived from the C-termini of eukaryotic 19S ATPases were expected to open the gate of the CP and facilitate degradation of substrates [18]. Finally, hydrophobic peptides were postulated to open the gate and activate the 20S core [53]. Binding of modulators in the distance to the strictly conserved active sites opens an important avenue in the development of more selective therapeutics intended for proteasome inhibition as well as enables the proteasome activation.
Conclusions In summary, free Tat2 and majority of its analogs with a single Ala substitution in liquid preferentially adopt a conformation containing PPII/turn structural motifs. Replacing either Asp10 or two or more adjacent Arg/Lys seems to disrupt ionic interactions responsible for stabilization of the ordered structure of the peptides. The random coil Tat2 analogs lose their ability to activate the latent proteasome. Apparently, the conformation as determined for the free peptides is not indicative of their inhibitory capability. To the contrary, the strong positive charge of the peptides is necessary for them to become proteasome inhibitors. The established structure-activity dependences will be utilized to optimize the structure of peptidomimetics serving as effective allosteric regulators of the 20S proteasome. Acknowledgement This work was supported by the National Science Centre grant no. 2011/01/B/ST5/06616 and Polish Ministry of Science and Higher Education grant no. DS 530-8440-D379-14, William and Ella Owens Medical Research Foundation grant (M.G.), and IIMS Pilot grant (P.A.O.).
References
J. Pept. Sci. 2014; 20: 649–656 Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci
655
1 Laskowski RA, Gerick F, Thornton JM. The structural basis of allosteric regulation in proteins. FEBS Lett. 2009; 583: 1692–1698. 2 Merdanovic M, Mönig T, Ehrmann M, Kaiser M. Diversity of allosteric regulation in proteases. ACS Chem. Biol. 2013; 8: 19–26. 3 Hauske P, Ottmann C, Meltzer M, Ehrmann M, Kaiser M. Allosteric regulation of proteases. Chem. Bio. Chem. 2008; 9: 2929–2928.
4 Zorn JA, Wells JA. Turning enzymes ON with small molecules. Nat. Chem. Biol. 2010; 6: 179–188. 5 Melancon BJ, Hopkins CR, Wood MR, Emmitte KA, Niswender CM, Christopoulos A, Conn PJ, Lindsley CW. Allosteric modulation of seven transmembrane spanning receptors: theory, practice and opportunities for central nervous system drug discovery. J. Med. Chem. 2012; 55: 1445–1464. 6 Nussinov R, Tsai C-J. Allostery in disease and in drug discovery. Cell 2013; 153: 293–305. 7 Mohr K, Schmitz J, Schrage R, Tränkle C, Holzgrabe U. Molecular alliance – from orthosteric and allosteric ligands to dualsteric/bitopic agonists at G protein coupled receptors. Angew. Chem. Int. Ed. 2013; 52: 508–516. 8 Ruschak AM, Kay LE. Proteasome allostery as a population shift between interchanging conformers. Proc. Natl. Acad. Sci. 2012; 109: E3454–E3462. 9 Osmulski PA, Hochstrasser M, Gaczynska M. A tetrahedral transition state at the active sites of the 20S proteasome is coupled to opening of the α-ring channel. Structure 2009; 17: 1137–1147. 10 Śledź P, Förster F, Baumeister W. Allosteric effects in the regulation of 26S proteasome activities. J. Mol. Biol. 2013; 425: 1415–1423. 11 Groll M, Ditzel L, Löwe J, Stock D, Bochtler M, Bartunik HD, Huber R. Structure of 20S proteasome from yeast at 2.4 Å resolution. Nature 1997; 386: 463–471. 12 Dick TP, Nussbaum AK, Deeg M, Heinemeyer W, Groll M, Schirle M, Keilholz W, Stevanović S, Wolf DH, Huber R, Rammensee HG, Schild H. Contribution of proteasomal beta-subunits to the cleavage of peptide substrates analyzed with yeast mutants. J. Biol. Chem. 1998; 273: 25637–46. 13 Jung T, Catalgol B, Grune T. The proteasomal system. Mol. Aspects Med. 2009; 30: 191–296. 14 Groll M, Bajorek M, Kohler A, Moroder L, Rubin DM, Huber R, Glickman MH, Finley D. A gated channel into the proteasome core particle. Nat. Struct. Biol. 2000; 7: 1062–1067. 15 Bajorek M, Glickman MH. Keepers at the final gates: regulatory complexes and gating of the proteasome channel. Cell. Mol. Life Sci. 2004; 61: 1579–1588. 16 Stadtmueller BM, Hill CP. Proteasome activators. Mol. Cell 2011; 41: 8–19. 17 Whitby FG, Masters EI, Kramer L, Knowlton JR, Yao Y, Wang CC, Hill CP. Structural basis for the activation of 20S proteasome by 11S regulators. Nature 2000; 408: 115–120. 18 Rabl J, Smith DM, Yu Y, Chang S-C, Goldberg AL, Cheng Y. Mechanism of gate opening in the 20S proteasome by the proteasomal ATPases. Mol. Cell 2008; 30: 360–368. 19 Sadre-Bazzaz K, Whitby FG, Robinson H, Formosa T, Hill CP. Structure of a Blm10 complex reveals common mechanism for proteasome binding and gate opening. Mol. Cell 2010; 37: 728–735. 20 Jankowska E, Stoj J, Karpowicz P, Osmulski PA, Gaczynska M. The proteasome in health and disease. Curr. Pharm. Des. 2013; 19: 1010–1028. 21 Nalepa G, Rolfe M, Harper JW. Drug discovery in the ubiquitinproteasome system. Nat. Rev. Drug Discov. 2006; 5: 596–613. 22 Groll M, Huber R, Moroder L. The persisting challenge of selective and specific proteasome inhibition. J. Pept. Sci. 2009; 15: 58–66. 23 Kisselev AF, van der Linden WA, Overkleeft HS. Proteasome inhibitors: an expanding army attacking a unique target. Chem. Biol. 2012; 19: 99–115. 24 Driscoll JJ, Burris J, Annunziata CM. Targeting the proteasome with bortezomib in multiple myeloma: update on therapeutic benefit as an upfront single agent, induction regimen for stemcell transplantation and as maintenance therapy. Am. J. Ther. 2012; 19: 133–144. 25 Kortuem KM, Stewart AK. Carfilzomib. Blood 2013; 121: 893–897. 26 Ruschak AM, Slassi M, Kay LE, Schimmer AD. Novel proteasome inhibitors to overcome bortezomib resistance. J. Natl. Cancer Inst. 2011; 103: 1–11. 27 Gaczynska M, Osmulski PA, Gao Y, Post MJ, Simons M. Proline and arginine-rich peptides constitute a novel class of allosteric inhibitors of proteasome activity. Biochemistry 2003; 42: 8663–8670. 28 Gaczynska M, Osmulski PA. Rapamycin allosterically inhibits the proteasome. Mol. Pharmacol. 2013; 84: 104–113. 29 Jankowska E, Gaczynska M, Osmulski P, Sikorska E, Rostankowski R, Madabhushi S, Tokmina-Lukaszewska M, Kasprzykowski F. Potential allosteric inhibitors of the proteasome activity. Biopolymers 2010; 93: 481–495.
WITKOWSKA ET AL. 30 Li X, Wood TE, Sprangers R, Jansen G, Franke NE, Mao X, Wang X, Zhang Y, Verbrugge SE, Adomat H, Li ZH, Trudel S, Chen C, Religa TL, Jamal N, Messner H, Cloos J, Rose DR, Navon A, Guns E, Batey RA, Kay LE, Schimmer AD. Effect of noncompetitive proteasome inhibition on bortezomib resistance. J. Natl. Cancer Inst. 2010; 102: 1069–1082. 31 Azevedo LM, Landsdell TA, Ludwig JR, Mosey RA, Woloch DK, Cognan DP, Patten GP, Kuszpit MR, Fisk JS, Tepe JJ. Inhibition of the human proteasome by imidazoline scaffolds. J. Med. Chem. 2013; 56: 5974–5978. 32 Seo H, Sonntag K-C, Kim W, Cattaneo E, Isacson O. Proteasome activator enhances survival of Huntington’s disease neuronal model cells. PLoS One 2007; e238: 1–7. 33 Dal Vechio FH, Cerqueira F, Augusto O, Lopes R, Demasi M. Peptides that activate the 20S proteasome by gate opening increased oxidized protein removal and reduced protein aggregation. Free Rad. Biol. Med. 2014; 67: 304–313. 34 Wilk S, Chen W-E, Wójcik C, Magnusson RP. In Proteasomes: The World of Regulatory Proteolysis, Hilt W, Wolf DH, Landes Bioscience, Georgetown, 2000; pp. 143–150. 35 Watanabe N, Yamada S. Activation of 20S proteasomes from spinach leaves by fatty acids. Plant Cell Physiol. 1996; 37: 147–151. 36 Huang L, Ho P, Chen C-H. Activation and inhibition of proteasomes by betulinic acid and its derivatives. FEBS Lett. 2007; 581: 4955–4959. 37 Dang Z, Jung K, Qian K, Lee K-H, Huang L, Chen C-H. Synthesis of lithocholic acid derivatives as proteasome regulators. ACS Med. Chem. Lett. 2012; 3: 925–930. 38 Savitzky A, Golay A. Soothing and differentiation of data by simplified least squares procedures. Anal. Chem. 1964; 36: 1627–1639. 39 Wang S-L, Wei Y-S, Lin S-Y. Subtractive similarity method used to study the infrared spectra of proteins in aqueous solution. Vibr. Spectrosc. 2003; 31: 313–319. 40 Chittur KK. FTIR and protein structure at interfaces. BMES Bull. 1999; 23: 3–9. 41 Smith BC. Fundamentals of Fourier Transform Infrared Spectroscopy. CRC Press: Boca Raton London New York, 2011; pp. 56–62. 42 Valenti LE, Paci MB, De Pauli CP, Giacomelli CE. Infrared study of trifluoroacetic acid unpurified synthetic peptides in aqueous solution: trifluoroacetic acid removal and band assignment. Anal. Biochem. 2011; 410: 118–123.
43 Surewicz WK, Mantsch HH, Chapman D. Determination of protein secondary structure by Fourier transform infrared spectroscopy: a critical assessment. Biochemistry 1993; 32: 389–394. 44 Andrushchenko VV, Vogel HJ, Prenner E. Optimization of hydrochloric acid concentration used for trifluoroacetate removal from synthetic peptides. J. Pept. Sci. 2007; 13: 37–43. 45 Elangovan S, Margolis HC, Oppenheim FG, Beniash E. Conformational changes in salivary proline-rich protein 1 upon adsorption to calcium phosphate crystals. Langmuir 2007; 23: 11200–11205. 46 Kong J, Yu S. Fourier transform infrared spectroscopic analysis of protein secondary structures. Acta Biochim. Biophys. Sin. 2007; 39: 549–559. 47 Perczel A, Hollósi M. In Circular Dichroism and the Conformational Analysis of Biomolecules, Fasman GD. Plenum Press: New York and London, 1996; pp. 285–380. 48 Sreerama N, Woody NW. Poly(Pro)II helices in globular proteins: Identification and circular dichroic analysis. Biochemistry 1994; 33: 10022–10025. 49 Huang X, Seifert U, Salzmann U, Henklein P, Preissner R, Henke W, Sijts AJ, Kloetzel P-M, Dubiel W. The RTP site shared by the HIV-1 Tat protein and the 11S regulator subunit α is crucial for their effects on proteasome function including antigen processing. J. Mol. Biol. 2002; 323: 771–782. 50 Seeger M, Ferrel K, Frank R, Dubiel W. HIV-1 Tat inhibits the 20S proteasome and its 11S regulator-mediated activation. J. Biol. Chem. 1997; 272: 8145–8148. 51 Gao YH, Lecker S, Post MJ, Hietaranta AJ, Li J, Volk R, Li M, Sato K, Saluja AK, Steer ML, Goldberg AL, Simons M. Inhibition of ubiquitin-proteasome pathway-mediated IκBα degradation by a naturally occurring antibacterial peptide. J. Clin. Invest. 2000; 106: 439–448. 52 Gaczynska M, Osmulski PA, Gao Y, Post MJ, Simons M. Proline- and arginine-rich peptides constitute a novel class of allosteric inhibitors of proteasome activity. Biochemistry 2003; 42: 8663–8670. 53 Kisselev AF, Kaganovich D, Goldberg AL. Binding of hydrophobic peptides to several non-catalytic sites promotes peptide hydrolysis by all active sites of 20 S proteasomes – evidence for peptideinduced channel opening in the alpha-rings. J. Biol. Chem. 2002; 277: 22260–22270.
656 wileyonlinelibrary.com/journal/jpepsci Copyright © 2014 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2014; 20: 649–656
