Identification and characterization of a novel glutaminase inhibitor

Abstract

In humans, there are two forms of glutaminase (GLS), designated GLS1 and GLS2. These enzymes catalyse the conversion of glutamine to glutamate. GLS1 exists as two isozymes: kidney glutaminase (KGA) and glutaminase C (GAC). Several GLS inhibitors have been identified, of which DON (6-diazo-5-oxonorleucine), BPTES (bis-2-(5-phenylacetamido-1, 3, 4-thiadiazol-2-yl) ethyl sulphide), 968 (5-(3-Bromo-4-(dimethylamino)phenyl)-2,2-dimethyl-2,3,5,6-tetrahydrobenzo[a]phenanthridin-4(1H)-one) and CB839 (Telaglenastat) are the most widely used. However, these inhibitors have variable efficacy, specificity and bioavailability in research and clinical settings, implying the need for novel and improved GLS inhibitors. Based on this need, a diverse library of 28,000 compounds from Enamine was screened for inhibition of recombinant, purified GAC. From this library, one inhibitor designated compound 19 (C19) was identified with kinetic features revealing allosteric inhibition of GAC in the µm range. Moreover, C19 inhibits anti-CD3/CD28-induced CD4+ T-cell proliferation and cytokine production with similar or greater potency as compared to BPTES. Taken together, our data suggest that C19 has the potential to modulate GLS1 activity and alter metabolic activity of T cells.

Keywords: BPTES; CB839; CD4+ T cells; GAC; GLS inhibitor; high-throughput screening.

Fig. 1

Cloning, expression, purification and functional testing of recombinant human glutaminase Δ₁₂₉GAC. (A) Schematic representation of the cDNA of Δ₁₆GAC, Δ₇₂GAC and Δ₁₂₉GAC. (B) Δ₁₆GAC, Δ₇₂GAC and Δ₁₂₉GAC were PCR amplified from His‐tagged hGAC inserted in the PET24b vector. (C) Expressed recombinant His‐tagged Δ₁₂₉GAC was purified using a HisTrap 1 mL Ni²⁺‐NTA affinity column. Protein purity analysed by SDS/PAGE shows a ~ 50 kDa His‐tag Δ₁₂₉GAC protein. The SDS/PAGE image shown in the figure was cropped from same part of the gel. Full‐length gel image is presented in Fig. S1. (D) Δ₁₂₉GAC activity was measured in a coupled enzyme assay where glutaminase activity was measured indirectly by monitoring Δ₁₂₉GAC activity as the production of NADH from the conversion of glutamate to α‐ketoglutarate by GDH (see Material and Methods for details). (E) Δ₁₂₉GAC activity was measured in a coupled enzyme assay determining NADH production as a function of glutamine conversion to glutamate by Δ₁₂₉GAC and in turn glutamate to α‐KG with NADH by GDH. Optimal concentration of glutamine was determined by titration of incremental concentrations (3.2–100 mm) and measuring NADH at 340 nm after 5, 10 and 15 min. Glutamine concentrations above 3.5 mm led to substrate inhibition. Measured absorbance from enzyme activity was adjusted by a negative control containing only buffer, protein and substrate. (F) The optimal level of Pi in the assay was determined in the presence of 3.125–100 mm Pi. 50 mm Pi was determined as the optimal concentration for the assay. Measured absorbance from enzyme activity was adjusted by a negative control containing only buffer, protein and substrate. (G) The linearity response of the assay was determined by plotting the raw data subtracted by the blank readings to validate the assay for kinetic studies. Linearity (R ² = 0.98) was determined in the presence of 4 mm glutamine up to 5 min, implying a robust assay. Data shown are means ± S.D (n = 3).

Fig. 2

HTS of chemical compounds inhibiting Δ₁₂₉GAC activity. (A) Incremental doses of the GLS1 inhibitor CB839 (1–1800 nm) were used as a negative control for Δ₁₂₉GAC enzyme inhibition yielding an IC₅₀‐value 60 nm by nonlinear curve fitting. Data are presented as mean ± S.D (n = 3). (B) GDH activity was measured after 20 min in the presence of 30 nm CB839. CB839 showed no inhibition effect on secondary enzyme (GDH) in the HTS assay. Data are presented as mean ± S.D (n = 3). (C) Validation of the primary screening assay. A test run of the primary screen consisting of 192 wells for both negative control (10 µm CB839, blue) and positive control (normal assay, red) was used and measured after a 20‐min time point to calculate the statistical coefficient, Z‐prime (Z′)‐factor and signal to background noise (S/B) to validate the primary HTS. (D) 28,000 compounds were measured in a total of 80 plates for Δ₁₂₉GAC specificity. Individual Z′‐factor for each plate is plotted, and after 80 plate readings, it was determined to be 0.78 ± 0.07. (E) (I) through (V) show the numbers of compounds that inhibited Δ₁₂₉GAC. This shows that 400 compounds exhibit ≥ 30% inhibitory effect in the primary screen and were submitted for counter screen (II). The rescreen (III) was performed with triplicates of 37 compounds after counter screen (step II). From the rescreen (III), 3 compounds were validated by dose response (10 points in duplicates) resulting in 3 positive compounds (IV). After characterization of 3 compounds, C19 was chosen with ≥ 90% inhibition (Step V). (F) Chemical structures of compound C2, C15 and C19, respectively. (G) Characterization of C2, C15 and C19 in the rescreen was performed by 10‐point dose response with twofold serial dilution (0.39–200 µm). IC₅₀ values were determined for C2 to be 2.3 µm, C15 to be 5.7 µm and C19 to be 6.8 µm by nonlinear curve fitting. C2 and 15 in contrast to compound 19 did not inhibit Δ₁₂₉GAC activity by 100%. Data are mean ± S.D (n = 3).

Fig. 3

Mode of inhibition of C19 and its quantitative interaction with Δ₁₂₉GAC. (A) Steady‐state kinetics were employed to determine the mode of inhibition of C19. Inhibitory effects were measured as enzymatic activity of Δ₁₂₉GAC (6 nm) using the enzyme assay and in the presence of 0, 16, 25 and 30 µm of compound. K_m and V _max were calculated and indicate that C19 exhibits noncompetitive inhibition as V _max, but not K_m , is altered by C19. The respective levels of NADH were measured after 5 min of incubation using the extinction coefficient 6220 M⁻¹·cm⁻¹ at 340 nm. Data points shown are mean ± SD (n = 5), and the curves were generated from 12 different substrate concentrations. For calculated K_m , V _max and K _cat values, check Table 1. (B) Quantitative interaction between CB839 and nonlabelled Δ₁₂₉GAC was performed by MST analysis. Incremental concentrations of CB839 (0–1000 nm) were tested with 100 nm Δ₁₂₉GAC. Interaction of CB839 and Δ₁₂₉GAC yielded a K_d ‐value of 25.6 ± 8.4 nm. (C) Using NT‐647‐labelled Δ₁₂₉GAC (25 nm) yielded a K_d ‐value of 14.74 ± 3.83 µm. All K_d values were determined using nonlinear curve fitting. Data are means ± SD (n = 3). (D) Incremental concentrations of C19 (0–1000 µm) were tested with 100 nm Δ₁₂₉GAC yielding a K_d ‐value of 20.0 ± 4.8 µm. Experimental conditions were 20% MST power and 20% excitation power at 25 °C.

Fig. 4

C19 inhibits anti‐CD3/CD28‐induced CD4+ T‐cell proliferation and cytokine secretion superior to BPTES. (A) Activated CD4+ T cells, MD‐468 and MD‐231 TNBC cell lines were treated with control compound CB839, and thymidine incorporation was assayed after 72 h of proliferation. Data are presented as mean ± S.D (n = 3). (B) CFSE‐labelled 1 × 10⁶ purified human CD4+ T cells were treated with C19 (25 µm) and BPTES (25 µm) followed by stimulation with anti‐CD3/CD28 beads for 96 h. At 96 h, cells were subjected to flow cytometry. Representative flow cytometry histograms for a single donor demonstrating proliferation of CFSE‐labelled CD4+ T cells at 96 h in culture with anti‐CD3/CD28 beads, BPTES (25 µm) and C19 (25 µm). Data are presented as the mean of 2 independent experiments ± SD run with triplicates. Measurement bars and numbers represent the percentage of proliferating cells. (C) Levels of TNF‐α, IL‐10, IL‐2, IL‐6, INF‐γ, IL‐17A and IL‐4 were measured in cell culture supernatants on day 3 of α‐CD3/CD28‐stimulated CD4+ T cells in the presence of either BPTES (25 µm) or C19 (25 µm); and data were normalized to stimulated control. All the cytokines showed significant decrease in concentration in the presence of C19 compared to control. Data are means ± SD (n = 9; ****P < 0.001, ***P < 0.005, **P < 0.01, *P < 0.05).