A bioavailable cathepsin S nitrile inhibitor abrogates tumor development

Raji (human B cell) (ATCC), MC38 (murine colorectal carcinoma) (ATCC) and MCF7 (human breast adenocarcinoma) (ATCC) were cultured in media containing 10 % FCS (Sigma Aldrich, UK) and 50 U/mL Penicillin and 50 μg/ml streptomycin (Life Technologies, UK). HUVEC (Cellworks) primary cells were cultured in large vessel endothelial cell growth medium (Cellworks) containing human large vessel endothelial cell growth supplement (KC1016) and antibiotic supplement (KC1019). Cells were maintained in 5 % CO₂ at 37 °C. MCF7 cell line was validated by STR profiling by LGC Standards in May 2011. All subsequent cells have only been taken from these validated stocks.

To a solution of Boc-Cys-OMe (0.88g, 3.74 mmol) in DMF (30 mL) at 0 °C was added K₂CO₃ (0.51 g, 3.74 mmol) and MeI (0.23 mL, 3.74 mmol) warmed to room temperature and stirred for 20 h. The reaction was diluted with a 9:1 water/NH₄Cl (2 M aq. Solu.) and extracted with EtOAc (×2). The organic extracts were washed with 5 % brine, dried over Na₂SO₄ and concentrated. Dissolved in MeOH (10 mL) and added to a 2 N methanolic-HCl solution and stirred at room temperature for 3 h. The mixture was concentrated and the solid was washed with MTBE (50 mL ×3) to afford (S)-methyl 2-amino-3-(methylthio)propanoate hydrochloride) as a clear light yellow oil (0.62 g, 90 %). Dissolved in MeOH (8 mL) and cooled to -78 °C, added 2,2,2,4′-trifluoroacetophenone (0.60 g, 3.35 mmol) and MeOK (0.47 g, 6.70 mmol). The reaction was slowly warm to room temperature whilst stirring for 20 h – solution A. To a solution of NaBH₄ (0.51 g, 13.4 mmol) in DME (10 mL) at 0 °C was added a 2 M solution of ZnCl₂ in Et₂O (8.03 mL). The mixture was allowed to age for 20 h –solution B. Solution A was cooled to -40 °C, diluted with MeCN (28 mL) and slowly added solution B over 20 min, then stirred at -40 °C for 2.5 h. The reaction was quenched with acetone (40 mL) over 20 min and then warmed to room temperature. The mixture was poured into an ice/water mix (250 mL), the pH was adjusted to 5 with 1 N HCl and extracted with EtOAc (200 mL ×3). The organic extracts were washed with brine (400 mL), dried over Na₂SO₄ and concentrated. Dissolved in DMF (15 mL), added 1-amino-1-cyclopropanecarbonitrile hydrochloride (0.39 g, 3.35 mmol), HBTU (1.59 g, 4.20 mmol) and NMM (0.54 mL, 4.00 mmol) and stirred at room temperature for 20 h. Diluted the reaction with water (120 mL) and extracted with EtOAc (100 mL ×2). The combined organic extracts were washed with 5 % brine (120 mL ×2) and brine (140 mL), dried over Na₂SO₄ and concentrated. The residue was purified by column chromatography (40 g column) using 0 to 75 % EtOAc in hexanes to afford (S)-N-(1-cyanocyclopropyl)-3-(methylthio)-2-(((S)-2,2,2-trifluoro-1-(4-fluorophenyl)ethyl)amino)propanoate as a clear colourless oil (0.36 g, 29 %). To a solution of (S)-N-(1-cyanocyclopropyl)-3-(methylthio)-2-(((S)-2,2,2-trifluoro-1-(4-fluorophenyl)ethyl)amino)propanoate (0.34 g, 0.91 mmol) in EtOAc (8.78 mL) was added NaTg•2H₂O (3.56 mg, 0.01 mmol) and TBAH (15.6 mg, 0.05 mmol) followed by addition of H₂O₂ (30 % w/w in water, 0.23 mL). The reaction was stirred at room temperature for 3 h. Diluted the reaction mixture with EtOAc (70 mL) and washed with a solution of 2 M Na₂S₂O₃ (100 mL ×2) and brine (100 mL), dried over Na₂SO₄ and concentrated to afford (S)-N-(1-cyanocyclopropyl)-3-(methylsulfonyl)-2-(((S)-2,2,2-trifluoro-1-(4-fluorophenyl)ethyl)amino) propanoate (compound 6) as a white powder (0.29 g, 79 %); ¹H-NMR (400 MHz, d ₆ -DMSO) δ 9.05 (s, 1H), 7.45 (dd, J = 8.0, 6.0 Hz, 2H), 7.25 (dd, J = 8.0, 6.0 Hz, 2H), 4.37 (quintet, J = 8.0 Hz, 1H), 3.74-3.66 (m, 1H), 3.54-3.46 (m, 2H), 3.17 (dd, J = 16.0, 4.0 Hz, 1H), 3.12 (s, 3H), 1.42-1.33 (m, 2H), 1.05-0.98 (m, 1H), 0.79-0.74 (m, 1H); LC-MS > 98 %, m/z = 408.12 [M + H].

Inhibition of cathepsin activity using compound 6 Recombinant cathepsin activity: Analysis of recombinant cathepsin activity was performed in a 96-well plate. All assays were performed in triplicate in the presence of sodium acetate assay buffer (sodium acetate 100 mM, EDTA 1 mM, Brij 0.1 %, and Dithiothreitol 2 mM, pH 5.5). Recombinant CTSS (4 nM), K (4.25 nM), V (4 nM), L (4 nM) and B (3.5 nM) (Calbiochem, UK) was incubated with compound 6 at a range of concentrations. The concentration of recombinant protein used in these assays was assumed to be equivalent to the active enzyme concentrations. Cathepsin activity was monitored using peptidyl fluorescent substrates; Cbz-VVR-AMC (20 μM, CTSS), Cbz-FR-AMC (20 μM, Cathepsins K, V and at 5 μM for cathepsin L) and Cbz-RR-AMC (20 μM, cathepsin B). Protease activity was monitored over the period of 1 h using a fluorometer (Flurostar Optima) with excitation at 390 nm and emission at 460 nm. Progress curve data points generated by compound 6 were fitted to equation 1 using GraphFit software.

Where, \( v \) _o is the steady state rate, \( E \) is the enzyme concentration and \( I \) is the inhibitor concentration. Using this equation values of \( v \) were generated for each concentration of inhibitor. Using these values, Morrison Plots were subsequently produced (\( v \) versus [I]), allowing determination of K _{i (app)} values. To account for any completing substrate these values were corrected using equation 2, allowing the generation of true K _i values for compound 6 versus cathepsins S, K, V, L and B.

Western blotting was carried out as previously described [29] using the following primary antibodies: mouse anti-human CD74 (1:400) (sc-47741, Santa Cruz, USA [30]), rat anti-mouse CD74 (1:1000) (555317, BD Biosciences, USA [31]) and rat anti-α-tubulin (1:10000) (ab6160, Abcam, UK [32]). The membrane was subsequently incubated with appropriate secondary antibody; goat anti-mouse HRP conjugate (1:10000) (172-1011, BioRad, UK [29]) or rabbit anti-rat HRP conjugate (1:40000) (ab102199, Abcam, UK [29]). Proteins were detected by chemiluminescence protocol and exposed using the BioRad Molecular Imager ChemiDoc XRS+ Imaging System (BioRad, USA).

A 24-well transwell plate (Corning, UK) containing 8.0 μm polycarbonate membrane was coated with Matrigel (1 mg/mL) (BD Biosciences). MC38 and MCF7 cells were seeded into the upper well at a density of 2.5x10⁵ per well in serum-free media in the presence of compound 6 at a range of concentrations and invasiveness evaluated as previously described [24]. Each condition was performed in duplicate, with 9 images taken per membrane at X20 magnification. Images were analysed using ImageJ and figures were generated using GraphPad Prism. Results demonstrate the mean number of cells per field of view per group ± SEM. Statistical analysis carried out by analysis of variance with Tukey test comparing all conditions.

A 48-well plate was coated with Matrigel (10 mg/mL) and incubated at 37 °C for 1 h. 1×10⁵ HUVECs were seeded per well in triplicate, treated with compound 6 (100 nM/10000 nM) and incubated for 18 h. Imaging was performed using a Nikon Eclipse microscope with six images taken from triplicate wells and analysis carried out using ImageJ. Figures were generated using Graphpad Prism displaying the average total tubule length per field of view. The result was expressed as the mean total tubule length per group ± SD. Statistical analysis carried out by analysis of variance with Tukey test comparing all conditions.

CTSS knock down construct sh835 was purchased from Sigma Aldrich, UK and were transduced into MC38 cells using lentiviral particles as previously described [6]. MC38 cells containing knockdown construct were continuously selected for using puromycin (6 μg/mL). Successful knock-down of CTSS was confirmed by RT-PCR.

RNA was extracted using STAT - 60 (Biogenesis, UK) according to manufacturer’s protocol and cDNA generated using Improm-II reverse transcriptase kit (Promega, UK) as previously described [29]. For amplification of the CTSS and GAPDH cDNA, the following primer sets were used: CTSS forward primer 5′ GGGATCTCTGGAAGAAAACCC’3 and reverse primer 3′- TTCGGAGACTGTCGGGGAAT’5. GAPDH forward primers 5′-AAGGTCATCCCAGAGCTGAA-3′ and reverse primer 3′CTGCTTCACCACCTTCTTGA-5′.

All mice used in these experiments were supplied with housing and subsequent experimentation was carried out in accordance with the Animal (Scientific Procedures) Act 1986, following UKCCCR guidelines and approved by the Ethical Review Committee with Queen’s University Belfast. 20 week old C57BL/6 mice were intraperitoneally injected with either compound 6 (100 mg/kg) or control. Spleens were harvested after 18 h and homogenized in RMPI1640 using nylon mesh. Red blood cells were lysed using ammonium chloride potassium (ACK) lysis buffer. Whole splenocytes were then lysed and protein analysed by western blot using rat anti-mouse CD74 antibody, as described previously.

Six to eight week old C57BL/6 mice were subcutaneously injected with 5.0×10⁶ MC38 cells resuspended in growth factor reduced matrigel (4 mg/ml, diluted in sterile PBS) into the right flank on day 0. Upon tumors reaching 100 mm³, mice were then treated with compound 6 (2 mg compound 6 in 4 % DMSO: 96 % peanut oil; 100 mg/kg) or vehicle control (4 % DMSO: 96 % peanut oil) via intraperitoneal injection every 2 to 3 days. Tumor volumes were calculated as Volume = length × breadth × π/6 (n = 5/group). Data was presented as mean tumor volume per group ± SEM. Statistical analysis carried out by student’s t test. Blood serums were sampled at day 13 from mice and subjected to CCL2 ELISA (R&D Systems, UK) as previously described [6].

Six to seven week old BALB/c nude mice were subcutaneously injected with an oestrogen pellet (approx 0.5 mg). 10 days later, mice were subcutaneously injected with 3.0x10⁶ MCF7 cells suspended in Matrigel (4 mg/ml, diluted in PBS). Upon tumors reaching 100 mm³, mice were then treated with compound 6 (2 mg compound 6 in 4 % DMSO: 96 % peanut oil; 100 mg/kg) or vehicle control (4 % DMSO: 96 % peanut oil) via intraperitoneal injection every 2 to 3 days. Tumor volume was calculated as described previously (n = 6/group). Data was presented as the mean tumor volume per group ± SEM. Statistical analysis carried out by student’s t test.

MCF7 tumor sections (6 µm) were formalin fixed, paraffin-embedded and subjected to immuno-staining as previously described [16]. Sections (n = 4/group) were incubated in rabbit anti-human Ki67 primary antibody (1:300) (ab155580, Abcam, UK [16]). Slides were subsequently stained with biotinylated goat anti-rabbit (1:300) (BA-1000, Vector, UK [16]) secondary antibody. The number of Ki67 positive cells was assessed by random selection of 10 fields of view per tumor under X20 magnification using a Leica DM5500B microscope and AL software. The number of Ki67 positive cells was presented as the mean Ki67 positive cells ± SEM. Statistical analysis carried out by student’s t test.

Paraffin-embedded MCF7 tumor sections (6 μm) were stained for apoptosis using terminal deoxynucleotidyl tranferase-mediated dUDP-nick-end labelling (TUNEL)-based TumorTACS™ In Situ Apoptosis Detection Kit (R&D Systems, UK [16]) as per the manufacturer’s instructions. TUNEL staining was assessed by random selection of 10 fields of view per tumor under a X20 magnification using a Leica DM5500B microscope and AL software. The number of TUNEL positive cells were counted and presented as the mean TUNEL positive cells ± SEM. Statistical analysis carried out by student’s t test.

Effects on cell proliferation following compound 6 treatment or presence of knock-down construct sh835 were assessed by MTT assays as previously described [6]. Absorbance values were measured at 570 nM and results presented as mean ± SD.

To assess the inhibition of both intra- and extracellular CTSS in our models we synthesized a dipeptide nitrile based inhibitor (compound 6), previously reported as a potent and selective inhibitor of human and murine CTSS (Fig. 1a) [26]. This compound has been previously used to investigate the therapeutic potential of CTSS in atherosclerosis [27]. The compound was synthesized in-house and its effectiveness against a panel of human cathepsin recombinant proteases was assessed.

Analysis of fluorimetric substrate turnover by CTSS demonstrated that compound 6 blocked CTSS activity in a concentration dependent manner (Fig. 1b). Furthermore compound 6 inhibited the activity of recombinant cathepsins K, V and B to varying degrees with limited activity against L (Fig. 1c-f). We then determined the K _i of compound 6 against each of these enzymes and subjected to non-linear analysis and fitted to Morrison kinetics [33]. This demonstrated that there was a high preference of compound 6 towards human recombinant CTSS (2.71 nM) over the other closely related proteases (CTSK 155 nM; CTSV 784 nM; CTSL 7870 nM; CTSB 1940 nM) (Additional file 1: Figure S1), in agreement with previously reported IC₅₀ selectivities [26].

To assess the effect of compound 6 on CTSS activity in a more complex milieu, we firstly incubated the inhibitor with murine MC38 colorectal carcinoma cell lysates, and measured turnover of a fluorogenic substrate Cbz-VVR-AMC. Addition of compound 6 to MC38 lysates demonstrated a concentration dependent reduction in CTSS-like activity (Fig. 2a).

Next we assessed the effectiveness of the inhibitor towards a physiologically relevant substrate. The invariant chain (CD74) is a known intracellular substrate of CTSS and when CTSS activity is ablated through inhibition or deletion, it leads to accumulation of a 10 kDa invariant chain fragment (lip10) [34, 35]. Thus, monitoring of this fragment in antigen presenting cells can inform the ability of CTSS inhibitors function within the cell. Raji cells were treated with compound 6 and the accumulation of the lip10 fragment was analysed by western blot (Fig. 2b). Semi-quantitative analysis of these bands by densitometry revealed compound 6 to have an IC₅₀ of 2.9 μM (Fig. 2c).

Previously we have demonstrated the effect of blocking cell invasion through CTSS shRNA mediated depletion using MC38 cells [16]. Here we found that treatment with compound 6 resulted in a dose dependent inhibition of MC38 cell invasion (Fig. 3a). In parallel to these experiments, we assessed potential cytotoxic effect of compound 6 on MC38 cells at these concentrations, finding no significant effect on cell viability (Additional file 2: Figure S2). Taken together this indicated that compound 6 inhibited migration of the cells through the artificial ECM layer by blocking invasion as opposed to an anti-proliferative or cytotoxic effect.

To further evaluate the selectivity of compound 6, we decided to utilise an MC38 CTSS knock-down cell line, achieved through stable transfection of sh835 [6], to evaluate the effect versus a non-targeting control (NTC) shRNA. The sh835 CTSS knock-down line was validated by RT-PCR (Fig. 3b). The sh835 knock-down cells proved less invasive as anticipated versus the NTC cells (Fig. 3c). Furthermore, treatment of the NTC cells with compound 6 resulted in a reduced level of cell invasion comparable to the sh835 cells. Importantly, treatment of the sh835 cells with compound 6 demonstrated no additional reduction in the invasion, strongly indicating selectivity of compound 6 towards CTSS. MTT analysis indicated no effect of compound 6 on cell viability on either the MC38 NTC or sh835 cell lines (Additional file 3: Figure S3).

We have previously demonstrated in vitro that CTSS inhibition can attenuate endothelial tube formation, a process that underpins neoangiogenesis [36]. HUVEC cells were plated out on Matrigel coated plates and tube formation quantified after 18 h. Treatment of these cells with compound 6 elicited a mild but significant reduction in average tubule length at 100 nM (Fig. 3d). Taken together these in vitro studies demonstrate that compound 6 could inhibit CTSS to block pro-tumorigenic characteristics of tumor and endothelial cells, warranting further investigation with in vivo models.

To ensure that dosing of compound 6 in vivo would sufficiently block tumor CTSS, we first evaluated the ability of compound 6 to block lip10 degradation in C57BL/6 mice spleens upon administration. Mice were treated with compound 6 (100 mg/kg) and the level of CD74 degradation in their splenic lysates was evaluated. These experiments revealed a clear accumulation of lip10 fragment in mice (group of three) treated with compound 6 in comparison to mice treated with vehicle control (Fig. 4a). Extrapolation of band strength furthermore revealed lip10 band strength to be increased by ~13.13-fold, consistent with the effect observed in vitro using the Raji cells (Additional file 4: Figure S4).

We then examined the effect of compound 6 treatment on mice bearing MC38 syngeneic colorectal adenocarcinoma tumors. This model provides a fully immunocompetent microenvironment and we have used this model previously in our genetic ablation studies examining contribution of CTSS to tumor growth when derived from the tumor and/or the stoma [16]. We monitored MC38 tumor growth over a 16 day period with treatment of compound 6 or vehicle control delivered via intraperitoneal injection starting on day zero and following every 2 to 3 days. Dosing of compound 6 resulted in a reduction in tumor progression of approximately 50 % by day 16 when compared to vehicle-treated controls (Fig. 4b), an effect similar to what we had previously observed with CTSS depletion [16]. The weights of these animals were also monitored during the treatment timecourse, confirming no adverse effects of the inhibitor on mice weight and appearance (Additional file 5: Figure S5). During the course of this study, we also analysed serums from the tumor bearing animals and assessed CCL2 levels. Consistent with our previous report that CTSS inhibition can block CCL2 production [6], here we found a reduction in CCL2 levels at day 13, 24 h after the 5^th administration of compound 6 (Additional file 6: Figure S6).

Previously we have examined the inhibition of CTSS in human astrocytoma and colorectal tumors [24], but recently Sevenich and colleagues have shown that CTSS inhibition may also have utility in breast cancer [19]. MCF7 cells are oestrogen dependent adenocarcinomas that have previously been shown to express CTSS [37], and have previously been used as a model of tumor invasion in a number of other studies [38‐41]. Taking these findings together, we examined whether the anti-tumor effects of compound 6 translated to a more human relevant model. Indeed, evaluation of compound 6 on MCF7 cell invasion demonstrated a dose dependent reduction in tumor cell invasion (Fig. 5a); which was independent of any effects on cell viability (Additional file 7: Figure S7).

The impact of compound 6 in vivo was then evaluated in the MCF7 xenograft model. The treatment of the mice resulted in a 30 % reduction of tumor growth by day 29 compared to vehicle-treated controls (Fig. 5b). To determine any affect on the health of the mice, their weights were obtained at the beginning of the study and treatment with compound 6 did not affect the health of the mice (Additional file 8: Figure S8).

Upon completion of the xenograft study, the mice were sacrificed, tumors harvested and MCF7 tissue sections prepared and stained for proliferation marker Ki67. Evaluation of Ki67 positive cells revealed that cell proliferation was significantly reduced in the compound 6 treatment group (Fig. 5c). Evaluation of compound 6 effect on neo-angiogenesis was also carried out by CD34 staining, however, no observable differences could be confirmed between mice treated with compound 6 and control mice (Data not shown). Finally, the presence of apoptotic cells was evaluated using TUNEL staining. The results demonstrated an increased level of apoptosis in the compound 6 treated MCF7 tissue sections when compared with vehicle-treated controls (Fig. 5d). Taken together, CTSS inhibition was shown to block tumor progression, through decreased invasion, decreased proliferation and increased apoptosis, consistent with previous observations [16].