Elena
Sacco
a,
Sherwin J.
Abraham
b,
Alessandro
Palmioli
a,
Gaetana
Damore
a,
Anna
Bargna
a,
Elisa
Mazzoleni
a,
Vadim
Gaponenko
*b,
Marco
Vanoni
a and
Francesco
Peri
*a
aDepartment of Biotechnology and Biosciences, University of Milano-Bicocca, P.zza della Scienza, 2, 20126, Milano, Italy. E-mail: francesco.peri@unimib.it; Fax: (+39)0264483565; Tel: (+39)0264483453
bDepartment of Biochemistry and Molecular Genetics, University of Illinois at Chicago, 900 S. Ashland Ave., Chicago, IL 60607. E-mail: vadim@uic.edu; Fax: +1 (312) 413-0353; Tel: +1 (312) 355-4965
First published on 7th March 2011
Since mutations of Ras genes have a great incidence in human tumours, Ras oncoproteins are a major clinical target for the development of anticancer agents. We have developed synthetic molecules able to inhibit Ras activation. Here we present new, water-soluble Ras inhibitors composed by an aromatic pharmacofore moiety covalently linked to different sugars. New glycosylated compounds bind to Switch 2 region of Ras, also involved in effector binding, inhibit GEF-catalyzed nucleotide exchange on Ras in vitro, and reduce Ras-dependent proliferation of murine fibroblasts. The influence of the sugar unit on Ras binding affinity and on the biological activity of Ras inhibitors has been investigated.
Oncogenic Ras mutants are found in 30% of human tumors, with great incidence in colon (50%), and pancreatic (90%) adenocarcinomas.7 Furthermore, aberrant Ras signalling contributes to pathogenesis of some human developmental disorders, including Neuro-Cardio-Facio-Cutaneous syndromes.8 Thus, Ras proteins are considered a major clinical target for the above diseases, and a significant effort has been recently devoted to the development of small molecules that are able to bind to Ras and inhibit its activation.
We synthesized small molecules that bind to Ras and inhibit GEF-mediated nucleotide exchange in vitro.9–12 Some of these synthetic Ras inhibitors are only composed by aromatic pharmacophores,9,12 others contain aromatic moieties covalently linked to sugar scaffolds derived from COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundglucose and COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundarabinose.10,11 While showing affinity to Ras in vitro10,11 and selective toxicity towards colorectal cancer cells expressing the oncogenic K-RasG13D mutant,13 synthetic Ras ligands previously developed by our group suffered from low COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundwater solubility and therefore poor bioavailability. Recently we developed a small molecule (here named compound 2) showing interesting in vitro activity associated with good COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundwater solubility.14 Such molecule, containing an aromatic pharmacophore (here named compound 1) conjugated to COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundglucose, is able to bind to the β-3/α-2 region of H-Ras-GDP that is involved in GEF binding, as demonstrated by NMR experiments. By MD simulations and docking refinement of such NMR binding data two most representative low-energy binding modes between 2 and Ras were obtained. In one only the aromatic part of 2 interacts with Ras, while the glucose unit stays external to the binding cavity, in the other one, the sugar too is involved in binding interactions with Ras. Here we defined the role of glucose moiety in Ras binding and in inhibitory efficacy of compound 2 and of other differently glycosylated compounds.
![]() | ||
Fig. 1 (A) Structure of compounds 1–4. (B) Normalized HN chemical shift perturbations resulting from NMR titration of H-Ras-GDP with compounds 1–4. The graphs are generated based on HSQC spectra of H-Ras-GDP at the protein:compound ratio of 1![]() ![]() |
NMR titration experiments also provided an opportunity to determine dissociation constants for compounds 1–4 binding to H-Ras-GDP. The chemical shift changes caused by compound binding were used in a fitting procedure to determine the Kd values. Example fits for compounds 1–4 interacting with the dominant binding site in the β-3/α-2 region of H-Ras-GDP are shown in ESI (Fig. S1†). Compounds 2, 3 and 4 displayed good affinities with dissociation constants of 9.5 ± 1.2 μM, 5.5 ± 2.9 μM and 7.8 ± 2.9 μM, respectively. Compound 1 bound H-Ras-GDP weakly with a dissociation constant of 21.4 ± 7.6 μM.
The compounds tested here share the same aromatic moiety and different covalently linked sugar moieties. Because the compounds involve very similar interfaces when interacting with H-Ras-GDP we hypothesized that the compounds bind H-Ras-GDP through their aromatic ring structure. To test this hypothesis we performed a saturation transfer difference (STD) experiment that is designed to delineate the moieties in small compounds that are directly involved in binding large proteins (Fig. 2A–C). Compound 1 bound H-Ras-GDP significantly weaker than the other compounds displaying low intensity STD signals in the aromatic region. This observation is in agreement with our hypothesis and with the lower binding affinity for compound 1 measured by NMR titration. Compounds 3 and 4 (Fig. 2B and 2C respectively) exhibited larger STD signals in the aromatic region than compound 1 (Fig. 2A). The 1H signals belonging to the sugar moieties expected to be observed in the region roughly ranging from 2 to 5 ppm of 1H NMR spectrum were very weak. The only significant STD signals in the aliphatic region belonged to the buffer components. Together, this supports our hypothesis that the aromatic ring structure that is common in the tested compounds provides the dominant binding site to H-Ras-GDP. However, there is still a question of what causes differences in compound binding specificity and affinity. One possible explanation of this phenomenon is structural modulation of the ring structure by the attached sugar moiety. To test this idea we compared one-dimensional NMR spectra of compounds 1, 3, 4 (Fig. 2D–F). Indeed, the 1H signals belonging to common aromatic ring structure in compounds 1, 3, 4 do not overlay with compound 1 (Fig. 2D) exhibiting more significant differences from compounds 3 and 4 (Fig. 2E and 2F respectively). Although the 1H signals in the aromatic region of compounds 3 and 4 are more similar to each other, clear differences are also discernable. Therefore the attached sugar moieties slightly alter the relative orientation of the aromatic rings and thus affect compound binding specificity and affinity.
![]() | ||
Fig. 2 STD spectra of compound 1 (A), compound 3 (B), and compound 4 (C) bound to H-Ras-GDP. The aromatic 1H signals belonging to the tested compounds and the 1H signals belonging to the buffer components are marked. One-dimensional 1H-NMR spectra of compounds 1 (D), 3 (E), and 4 (F). The dashed lines indicate differences in chemical shifts for 1H signals in the aromatic region of compounds 1, 3, and 4. |
![]() | ||
Fig. 3 (A) Dose-response curves of compounds 1–4 dissolved in DMSO-containing buffer on GEF-catalyzed exchange in Ras, obtained by plotting the exchange rate as a function of compound concentration. The exchange rate of control reaction (without any synthetic compound) was normalized to 100. (B) Inhibitory efficacy of compounds 1–4 on GEF-catalyzed nucleotide exchange on Ras as derived from curves in Figure S2,† expressed as IC50 values. |
The effects of compounds 2–4 on the Ras-dependent proliferation of normal (NIH3T3) and KRAS-transformed (NIH-RAS) murine fibroblasts were then investigated. Only COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundwater soluble compounds 2–4 were assayed, while compound 1 was not tested because COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundDMSO required for its solubilization has a cytotoxic effect at concentrations higher than 0.1% (v/v).17Cells were plated at a density of 3000 cells cm−2, left to adhere to the wells for 18 h, and then treated with compounds 2–4 supplemented in the growth medium, and counted at 24, 48 and 72 h after treatment. Growth curves of normal and transformed fibroblasts treated with increasing concentrations of compound 3 are reported in Fig. 4A and 4B, respectively. These data highlight a dose-dependent inhibitory effect of compound 3 on cell proliferation. Similar results were obtained with compounds 2 and 4 (data not shown). Growth curves of normal (Fig. 4C) and transformed (Fig. 4D) fibroblasts treated with 100 μM of each compound show that compounds 2–4 had a similar inhibitory efficacy on proliferation.
![]() | ||
Fig. 4 Growth curves of normal NIH3T3 (A) and K-RAS transformed NIH-RAS (B) murine fibroblasts grown in media supplemented with different concentration of compound 3: 0 μM (black diamonds and dotted lines) 50μM (black squares) 100 μM (black triangles) and 250μM (black circles). Growth curves of NIH3T3 (C) and NIH-RAS (D) cells treated with 100 μM of compound 2 (open squares), 3 (open triangles) and 4 (open circle). Control curves (with 0μM compound) are showed as open diamond and dotted lines. Plotted data are mean ± s.d. computed from at least two independent experiments performed in triplicates. |
Specificity of the action of the compounds was confirmed by assaying the activation state of p42/p44 MAPKs, well known Ras effectors, that act downstream of growth-factor activated pathway. NIH-RAScells treated for 24 h with 250μM compound 3 show reduced phosphoMAPK levels if compared with cells treated with buffer (Fig. 5A). The Ras-specific GEF involved in the growth-factor activated pathway is hSos1 that shares with RasGRF1 the general structure of the catalytic core, despite some different regulation of catalytic activity. To confirm that our compounds can inhibit hSos1-mediated exchange, as well as GRF1-mediated exchange we performed hSos1-catalyzed nucleotide exchange assay on Ras in presence of different concentrations of compound 3 (Fig. S3, ESI). Fig. 5B shows that compound 3 inhibits in a dose dependent manner the hSos1-catalyzed nucleotide exchange on H-Ras, with a calculated IC50 that is very similar to that observed for RasGRF1-catalyzed exchange, indicating that the inhibirory effect of compounds is not limited to a specific GEF protein, but is extended likely to all as-GEFs that share the same Ras-binding site.
![]() | ||
Fig. 5 (A)Phosphorylated MAPK level in cell lysates from NIH3T3, NIH-RAS cells not treated or treated for 24 h with 250mM Cmp3, or buffer. Data are normalized on the Phospho-MAPK level in NIH-RAS cells taken equal to 100. Data shown are mean ± standard deviation of two independent experiments, each performed in triplicate. Single and double asterisk above histograms indicates a statistical significance of 95% and 99% respectively, calculated by Student's t-test in comparison to cells treated with buffer. (B) Dose–response curves of compound 3 on GEF-catalyzed exchange of Ras. The initial exchange rate of each exchange reaction (mean of at least three independent experiments) was plotted as a function of the compound concentration (dose–response curves). The exchange rate of control reaction (performed without compound) was normalized to 100. |
All compounds specifically bind to β-3/α-2 region (Switch 2) of H-Ras-GDP, also involved in binding with effectors, including GEF. This supports the hypothesis that such compounds can function as Ras inhibitors by affecting Ras-GEF binding. Accordingly, all compounds are able to inhibit the GEF-catalyzed nucleotide exchange on Ras in vitro. In this assay, glycosylated compounds were more potent in inhibiting Ras activation than compound 1. This result is in agreement with the lower affinity of such compound for Ras and also could be due to the fact that it, unlike compounds 2–4, does not bind residue R68 (Fig. 1C intersection among compounds 2, 3 and 4) that plays a critical role in conformational changes occurring in Ras activation and in binding to GEFs.18 In fact, in the crystal structure of Ras in complex with the catalytic domain of the GEF hSos1 (PDB code 1BKD),19 R68 of Ras directly interacts with E1002 of hSos1. The inability to interact with R68 could therefore account for the weakest biological activity of 1.
In conclusion the sugar moiety of Ras inhibitors is not only involved in increasing their COMPOUND LINKS
Read more about this on ChemSpider
Download mol file of compoundwater solubility, but plays a key role in orientating the pharmacophore groups in a way that improve the Ras affinity and the Ras inhibitory efficacy. Preparation of larger libraries of glycol-randomized synthetic Ras inhibitors in which aromatic hydroxylamine-containing pharmacophores are linked to a variety of sugars or other natural compounds20 is in progress. Their characterization will allow to define in more detail structure/activity relationships in this class of compounds.
Footnote |
† Electronic supplementary information (ESI) available: Synthetic methods, NMR, biology, Figure S1, Figure S2, Figure S3. See DOI: 10.1039/c0md00264j |
This journal is © The Royal Society of Chemistry 2011 |