Proteasome inhibition and mechanism of resistance to a synthetic, library-based hexapeptide

Summary Background The hexapeptide 4A6 (Ac-Thr(tBu)-His(Bzl)-Thr(Bzl)-Nle-Glu(OtBu)-Gly-Bza) was isolated from a peptide library constructed to identify peptide-based transport inhibitors of multidrug resistance (MDR) efflux pumps including P-glycoprotein and Multidrug Resistance-associated Protein 1. 4A6 proved to be a substrate but not an inhibitor of these MDR efflux transporters. In fact, 4A6 and related peptides displayed potent cytotoxic activity via an unknown mechanism. Objective To decipher the mode of cytotoxic activity of 4A6. Methods Screening of 4A6 activity was performed against the NCI60 panel of cancer cell lines. Possible interactions of 4A6 with the 26S proteasome were assessed via proteasome activity and affinity labeling, and cell growth inhibition studies with leukemic cells resistant to the proteasome inhibitor bortezomib (BTZ). Results The NCI60 panel COMPARE analysis revealed that 4A6 had an activity profile overlapping with BTZ. Consistently, 4A6 proved to be a selective and reversible inhibitor of β5 subunit (PSMB5)-associated chymotrypsin-like activity of the 26S proteasome. This conclusion is supported by several lines of evidence: (i) inhibition of chymotrypsin-like proteasome activity by 4A6 and related peptides correlated with their cell growth inhibition potencies; (ii) 4A6 reversibly inhibited functional β5 active site labeling with the affinity probe BodipyFL-Ahx3L3VS; and (iii) human myeloid THP1 cells with acquired BTZ resistance due to mutated PSMB5 were highly (up to 287-fold) cross-resistant to 4A6 and its related peptides. Conclusion 4A6 is a novel specific inhibitor of the β5 subunit-associated chymotrypsin-like proteasome activity. Further exploration of 4A6 as a lead compound for development as a novel proteasome-targeted drug is warranted.

Screening of 4A6 using the NCI60 tumor cell line panel The NCI 60 human tumor cell line screen was used to assess the activity profile of 4A6 against a panel of tumor cell lines of various cell lineage [47]. Concentrations of 4A6 eliciting 50% growth inhibition (GI50) were determined after 48 h drug exposure. 4A6 sensitivity for each individual cell line is depicted relative to the mean GI50 of the total cell line panel.
4A6 cleavage assay Proteasome was purified from bovine liver as described previously [48]. For digestion assays, 1 μg proteasome was incubated with 1 μg 4A6 in 50 μl of 50 mM Tris-HCl buffer pH 8.5 at 45°C for 16 h. Subsequently, the reaction mixture was lyophilized and peptides purified using reversed-phase ZipTip® C18 tips (Millipore). The purified peptide mixture was mixed in a 1:1 ratio with 10 mg/ml 2,5dihydroxybenzoic acid (DHB, Bruker Daltonik) matrix solution in 0.1% TFA and spotted onto a MALDI (matrix assisted laser desorption/ ionization) target plate. MALDI-  Results depicted are the mean of at least 3 separate experiments ± S.D.
ND Not determined, CsA cyclosporin A # Data from Oerlemans et al. [46] *Solubility of peptide in medium is limited to a concentration of 50 μM TOF analysis was performed on an Autoflex, linear MALDI-TOF-MS (Bruker Daltonik GmbH, Bremen, Germany). Spectra were analyzed with flexAnalysis software (Bruker Daltonik).
Growth inhibition assays Evaluation of drug sensitivity was carried out as described before [49]. Cells were seeded at an initial density of 1.25 × 10 5 cells/ml in individual wells of a 24-well plate containing up to 50 μl of drug solutions. Inhibition of cell growth was determined after 72 h of incubation at 37°C by determining the number of viable cells viable cells using trypan blue exclusion. The drug concentration required to inhibit cell growth by 50% compared to untreated controls was defined as the IC 50 .
Western blot analysis (ubiquitinated proteins/proteasome subunits) Western blot analysis to determine protein levels of (i) β1, β2 and β5 proteasome subunits and (ii) the accumulation of ubiquitinated proteins after treatment with 4A6 was performed essentially as described previously [46,49]. Cells were harvested in the mid-log phase of growth and washed 3 times with ice-cold buffered saline pH 7.4. Total cell lysates of 5 × 10 6 cells were prepared by resuspension in 500 μl lysis buffer containing: 50 mM Tris-HCl (pH 7.6), 5 mM dithiotreitol, 20 μl PIC (Protease Inhibitor Cocktail; 1 tablet/ml H 2 O), 20% glycerol and 0.5% NP-40. The suspension was sonicated (MSE sonicator, amplitude 7, for 3 × 5 s with 20 s time intervals at 4°C) and centrifuged in an Eppendorf micro centrifuge (5 min, 12,000 rpm, 4°C). Protein content of the supernatant was determined by the Bio-Rad protein assay. 20-30 μg of total cell lysates were fractionated on a 10% polyacrylamide gel containing SDS and transferred onto a PVDF membrane. The membranes were pre-incubated overnight at 4°C in blocking buffer (5% Bio-Rad Blocker in TBS-T; 10 mM Tris-HCl, pH 8.0, 0.15 M NaCl, 0.1% Tween-20) to prevent non-specific antibody binding. After blocking, the membranes were incubated for 1 h at room temperature with primary antibodies for proteasome subunit β1 (1:1000, PW8140), β2 (1:1000, PW8145) and β5 (1:1000, PW8895) or ubiquitin (1:1000, Santa-Cruz, SC-8017). An antibody to α-tubulin was used (1:1000, Santa Cruz, sc-8035) to check and normalize for any loading differences. After 3 washing steps with TBS-T, the membranes were incubated for 1 h with HRP-labelled donkey-anti-rabbit (1:6000, Amersham, UK) or goat-anti-mouse (1:6000, Dako, Glostrup, Denmark) as secondary antibody. Detection of antibody binding was followed by chemoluminescence using Supersignal (Pierce Biotechnology, Rockford, USA) according to the manufacturers' instructions. Digital Image acquisition was performed using the Versadoc Imaging System (Biorad Lab., Veenendaal, The Netherlands). The signal intensity was determined densitometrically using Quantity One software (Bio-Rad) and was expressed relative to the intensity of the α-tubulin signal.

Proteasome activity in cell lysates and intact cells
Chymotrypsin-like, trypsin-like and caspase-like proteolytic activities of the proteasome were determined in freshly prepared cell lysates as described previously [21,46]. Five million untreated or bortezomib-exposed THP1 cells were washed 3 times with ice-cold PBS and pelleted by centrifugation (5 min, 12,000 RPM, 4°C). Cell pellets were then resuspended in an ATP-containing lysis buffer; 10 mM Tris-HCl buffer (pH 7.8) containing 5 mM ATP, 0.5 mM DTT and 5 mM MgCl 2 , and kept on ice for 10 min. For complete lysis, cells were sonicated (MSE sonicator, amplitude 7, for 3 × 5 s with 20 s time intervals at 4°C) followed by centrifugation (5 min, 12,000 RPM, 4°C) to remove cell debris. The supernatant was collected and protein concentration was determined using the Bio-Rad protein assay. Fluorogenic substrates to measure the chymotrypsin-like, trypsin-like and caspaselike activity were Suc-Leu-Leu-Val-Tyr-amc, Ac-Arg-Leu-Arg-amc and Z-Leu-Leu-Glu-amc, respectively, all at a final concentrations of 100 μM. The substrates were incubated with 20 μg of total cell protein extract in the presence or absence of specific inhibitors (bortezomib for chymotrypsin-like activity, Ac-APnLD-H for caspase-like activity and leupeptin-for trypsin-like activity) in a total assay volume of 200 μl. The release of amc (7-amino-4-methyl-coumarin) was monitored online over a 2-h time period at 37°C with 5 min intervals. Fluorescence was measured on a Tecan SpectraFluor apparatus (Giessen, The Netherlands) using excitation and emission wavelengths of 360 and 465 nm, respectively. Proteolytic activity was calculated from the slopes of the linear portion of the curves. All results were expressed as percentage relative to untreated THP1/WT cells (100%). Inhibition of chymotrypsin-like activity in intact cells was measured by the Proteasome-Glo™ cell-based assay (Promega, Leiden, The Netherlands), using Suc-LLVY-aminoluciferin as a substrate, according to the manufacturer's instructions.

Proteasome affinity labelling
Proteasome activity profiling assays were performed as described [50,51]  was used to measure protein content. Proteins were denatured by boiling in reducing sample buffer and analyzed by 12% SDS-PAGE using NuPAGE pre-cast gels (Invitrogen). Gels were then scanned for fluorescence emission using a ProXPRESS 2D Proteomic imaging system (Perkin Elmer). Images were analyzed using Totallab analysis software (Nonlinear Dynamics, Newcastle upon Tyne, UK). Sypro staining served as a loading control.
Apoptosis assay Induction of apoptosis was analyzed by flow cytometry using APOPTEST™-FITC A700 (VPS Diagnostics, Hoeven, the Netherlands) according to the instructions of the manufacturer. In short, induction of apoptosis was determined after 24 h' drug exposure. One million cells were harvested and washed 3 times with ice-cold PBS. The cell pellet was incubated for 30 min with 7-Amino-actinomycin D (7-AAD) on ice followed by incubation with Annexin-V according to the instructions of the manufacturer. Annexin-V (early apoptosis) and 7-AAD (late apoptosis) staining was measured by flow cytometry (Beckton & Dickinson, FACScalibur) and analysed using FCSexpress V3 software (Denovo software, Thornhill, Canada).
Statistics Statistical analysis was performed using Analysis of Variance between groups (ANOVA) in Graphpad prism version 6.0. P values <0.05 were considered to be statistically significant.

4A6 vs bortezomib activity against NCI60 panel of tumor cell lines
In order to get an initial insight regarding the cytotoxic activity of 4A6, we first tested 4A6 in the NCI60 tumor cell line panel that is composed of 60 malignant cell lines of distinct tissue lineage [47]. 4A6 showed remarkable activity towards a panel of leukemia, breast cancer, melanoma, and to some extent colon cancer cells (Fig. 2) Cells with acquired resistance to peptide-based proteasome inhibitor bortezomib are cross-resistant to the cytotoxic peptides 4A6 and 4E11 Because of the overlapping activity profile of 4A6 and BTZ in the NCI60 panel, we tested 4A6 in human THP1 cell lines with acquired resistance to BTZ. These cell lines displayed crossresistance to other known peptide-based proteasome inhibitors (e.g. ALLN, MG132), but also to the linear cytotoxic hexapeptide 4A6, the latter of which has an unknown mechanism of action [46]. To further explore the molecular basis of this observation, THP1 cells with various levels of BTZresistance were screened for their sensitivity to 4A6, a dimer form of 4A6, another linear cytotoxic hexapeptide 4E11 (Fig.  1), and the cyclic cytotoxic decapeptide cyclosporin A (  Fig. 3) and >60-fold cross resistance to the 4A6dimer. With respect to the peptide 4E11, a consistently higher IC50 value compared to 4A6 (Table 1) along with limitations in solubility of peptides above a concentration of 50 μM, allowed for the assessment of relatively low level (>13-fold) cross-resistance to 4E11. No cross-resistance of bortezomibresistant cells was observed for cyclosporin A. Collectively, these results indicate that the peptides 4A6 and 4E11 share properties with known inhibitors of the ubiquitin-proteasome system, including BTZ.
To address whether or not 4A6 is also capable of inhibiting one or both of the other protease activities harbored by the proteasome, chymotrypsin-, caspase-and trypsin-like activities were measured in THP1 cell extracts in the absence or presence of 4A6. Consistent with results shown in Fig. 4a, 4A6 elicited  50 for each individual tumor cell line and as GI 50 relative to the mean GI 50 of all cell lines tested potent inhibitory effects (84-93%) on chymotrypsin-like proteasome activity, but had no inhibitory effect on caspase-and trypsin-like activity over a wide concentration range of 1-100 μM (Fig. 4b). These results demonstrate that 4A6 is a potent and selective inhibitor of chymotrypsin-like proteasome activity.

4A6 is a reversible inhibitor of chymotrypsin-like proteasome activity
Activity probing of constitutive and immunoproteasome βsubunits in EL4 cells that were pre-exposed to 4A6 for 2-24 h revealed a marked and specific inhibition of the β5 subunit with half maximal inhibition at 4A6 concentrations between 0.1 and 0.5 μM and complete inhibition at concentrations >5 μM (Fig. 5a). We next assessed whether this inhibition of β5-subunit could be recovered after removal of 4A6. Data shown in Fig. 5b illustrate that 4A6 is a reversible inhibitor of β5-subunit activity as initial recovery of activity could be observed already after 15 min of 4A6 drug removal and almost complete recovery after 2 h of 4A6 withdrawal. For comparison, the proteasome inhibitor MG132 blocked activity probing of all β-subunits, with a recovery 2 h after drug withdrawal (Fig. 5b). BTZ predominantly inhibited β5-subunit probing  but affinity labeling was fully recovered within 2 h after drug withdrawal (Fig. 5b). We finally explored whether 4A6 remained intact as a peptide or could be subject to proteolytic cleavage when exposed to purified proteasomes. Comparison of mass spectra of the intact peptide (Fig. 5c) and the peptide after proteasomal digestion (Fig. 5d) showed that next to 4A6 (m/z 1080.6), one main additional peak appeared after digestion at m/z 749.5, corresponding to the 4-mer peptide Ac-Thr(tBu)-His(Bzl)-Thr(Bzl)-Nle-OH. A smaller peak appeared at m/z 934.6, corresponding to 5-mer peptide Ac-Thr(tBu)-His(Bzl)-Thr(Bzl)-Nle-Glu(OtBu)-OH. This indicates that 4A6 is predominantly cleaved at the P4-P5 position and to a lesser extent at the P5-P6 position. The main 4A6 proteasomal cleavage product, Ac-Thr(tBu)-His(Bzl)-Thr(Bzl)-Nle-OH was synthesized, but did not show any proteasome inhibitory effect or cell growth inhibitory potential (data not shown). Hence, these results suggest that 4A6 is a dual substrate and reversible inhibitor of proteasome subunit β5.

Cellular exposure to 4A6 induces accumulation of ubiquitinated proteins and apoptosis but displays properties distinct of bortezomib
One hallmark of proteasome inhibition is the accumulation of ubiquitinated proteins, which are toxic to cells and induce apoptosis [53,54]. Exposure of THP1/WT cells to 4A6 and 4E11 for 24 h resulted, just as for the known proteasome inhibitor BTZ, in a marked accumulation of ubiquitinated proteins, illustrated by a characteristic smear upon Western blot probed with an anti-ubiquitin antibody (Fig. 6a). In contrast, the same concentrations of 4A6 and 4E11 did not provoke any accumulation of ubiquitinated proteins in bortezomib-resistant cells. Consistent with these observations was the efficient induction of apoptosis by 4A6 in parental THP1/WT cells but none by 4A6 (over a concentration range of 0-25 μM) in THP1/BTZ 200 cells (Fig. 6b and c). For comparison, the anti-cancer drug and topoisomerase II inhibitor etoposide (VP16) was equally effective in inducing apoptosis in THP1/WT and THP1/BTZ 200 cells (not shown).
To explore whether 4A6 shares properties with the known proteasome inhibitor BTZ, we investigated the ability of 4A6 to mimic a reported feature of BTZ, the disregulation of intracellular calcium homeostasis that triggers caspase activation and apoptosis [55]. This process could be counteracted by inhibitors of the mitochondrial calcium uniporter (e.g. Ruthenium Red), thereby providing a protective effect against BTZ [55]. While a marked abrogation of BTZ activity could be obtained by Ruthenium Red, no effect of this compound was observed with respect to 4A6 activity (Fig. 6d). These results suggest that 4A6 has no apparent impact on mitochondrial calcium homeostasis.

4A6 provokes proteasome β5 subunit induction
Given the specific targeting of 4A6 of the β5 subunit of the proteasome, we explored whether exposure to 4A6 had an effect on the expression of the β5 subunit as compared to the other catalytic subunits β1 and β2. To this end, THP1/ WT cells and the bortezomib-resistant cell lines THP1/BTZ 100 and THP1/BTZ (−100) , the latter being a subline of THP1/ BTZ 100 that was grown in the absence of BTZ for 6 months, were exposed to a concentration range of 4A6 (0.1-10 μM) for 24 h (Fig. 7a). No significant effects of 4A6 exposure were observed regarding expression of the β1 and β2 proteasome subunits. In contrast, a dose-dependent increase in proteasome β5 subunit expression was noted in both THP1/WT sublines with relatively low basal levels of β5 expression and the two BTZ-resistant cell lines, including THP1/BTZ (−100) cells that retained a level of cross-resistance to 4A6 similar as THP1/ BTZ 100 cells (Fig. 7a). Densitometric analysis showed a 3-5 fold increase in β5 subunit induction in all three cell lines upon exposure to 10 μM 4A6 exposure as compared to drug-free controls (Fig. 7b). This result implies that induction of proteasome β5 subunit expression constitutes a rapid adaptive response upon targeting of this subunit by the inhibitor 4A6.

Discussion
Here we have shown that the cytotoxic hexameric 4A6 peptide elicits its pharmacological activity via selective and reversible inhibition of the chymotrypsin-like proteasome activity. The specific targeting of the chymotrypsin-like proteasome activity by 4A6 was further corroborated by upregulation of the expression of the β5 subunit of the proteasome. Moreover, cells harboring mutations in the β5 subunit which confer resistance to BTZ [21], displayed a marked cross-resistance to 4A6. Most peptide-based proteasome inhibitors contain tri-or tetrapeptide moieties that dock into one or more of the active site pockets of the proteasome [16,56]. However, peptides extended with N-terminally linked spacers and specific caps can also retain their proteasome inhibitory potential [50,57]. Notwithstanding this fact, 4A6, as well as another hexameric peptide (4E11) exhibited a motif and mode of action distinct from known peptide-based proteasome inhibitors. The linear hydrophobic nature of 4A6 likely facilitates its interactions with the β5-subunit of the proteasome that preferentially cleaves after hydrophobic amino acid residues [12,14]. To this end, we explored whether or not the interaction of 4A6 with the proteasome involves mere steric occlusion of the β5 active site or alternatively, that the 4A6 peptide serves as a cleavage substrate of the proteasome. Consistent with this notion may be the fact that the dimeric form of 4A6, which contains the same amino acid sequence as 4A6 and is therefore also likely to be cleaved by the proteasome, is almost equally effective in inhibiting β5-associated proteasome activity ( Fig. 4a and Table 1). In this context, it is important to note that replacement of Thr(Bzl) at the P3 position by Lys(Z) or Ala in 4A6 abolished the cytotoxic effect by 4A6 [32], suggesting that this residue is essential for effective proteasome binding and inhibition.
The marked level of cross-resistance to 4A6 of cells resistant to the proteasome inhibitor BTZ (Table 1, Fig. 3) supports the conclusion that 4A6 and BTZ share a common mode of interaction with the β5 active site. In fact, studies from our laboratory revealed that the molecular basis of BTZ resistance in these cells involved a point mutation in the PSMB5 gene that introduced a single amino acid change (Ala → Thr) at position 49 of the PSMB5 protein [46]. Since the Ala49 position resides in the BTZ binding pocket of PSMB5 and is involved in the interaction with BTZ [16,30,56,58,59], the Ala49Thr mutation is likely to underlie loss of BTZ binding and acquisition of bortezomib resistance [46]. The even higher levels of cross-resistance to 4A6 than resistance levels to BTZ suggest that Ala49 is even more critical in binding the 4A6 peptide than BTZ. In this respect, it was interesting to note that exposure of BTZ-resistant cells to 4A6 provoked a marked upregulation of mutant PSMB5 protein (Fig. 7), presumably as a compensatory mechanism to counteract loss of proteasome activity due to inhibition by 4A6.
Although the indicated PSMB5 mutation may be the dominant factor in conferring drug resistance to 4A6, it was previously reported that cellular extrusion by the MDR efflux transporters P-gp (ABCB1) and MRP1 (ABCC1) could also confer resistance to 4A6 [32]. This was further illustrated herein in the activity profile of 4A6 in the NCI60 panel of tumor cell lines where cells with a consistent MDR phenotype (mainly P-gp) were markedly less sensitive to 4A6 (Fig. 2). In contrast, such a MDR phenotype had relatively a marginal impact on the activity of BTZ. The presence of the boron group in BTZ most likely abolishes the ability of this compound to serve as a proficient substrate for MDR transporters as compared to other small peptides [60,61]. Although Fig. 2 demonstrated an overlap in activities against some tumor types (leukemia/breast cancer), the current study indicates (Fig. 6d) that at least one mode of action of 4A6 was distinct from BTZ by not inducing apoptosis/ growth inhibition via disregulation of mitochondrial calcium homeostasis [55]. Consistent with this study we showed here that inhibition of the mitochondrial uniporter with ruthenium red abrogated the growth inhibitory effects of BTZ, but had no effect on 4A6 activity. Collectively, this study reported on 2 novel hexameric peptide-based proteasome inhibitors with several properties distinct from currently identified proteasome inhibitors, including BTZ. One of these peptides, 4A6, may serve as a lead compound for drug development by further optimization of its selective proteasome β5 subunit targeting against leukemia and breast cancer cells. The notion that 4A6 is a bona fide P-gp and MRP1 substrate may on the one hand compromise some of its activity against tumor cell expressing this drug efflux transporter, but on the other hand it may underlie a different, possibly more favorable toxicity profile than BTZ.