Thank you for visiting Nature. The browser version you are using has limited support for CSS. For the best experience, we recommend that you use a newer version of the browser (or turn off the compatibility mode in Internet Explorer). At the same time, to ensure continued support, we will display sites without styles and JavaScript.

Scientific Reports Volume 5, Article Number: 12974 (2015) Cite this article

The constitutively active tyrosine kinase BCR-ABL is the underlying cause of chronic myeloid leukemia (CML). Current CML treatments rely on long-term use of tyrosine kinase inhibitors (TKI), which target the ATP binding site of BCR-ABL. During treatment, 20-30% of CML patients will develop TKI resistance, which is usually attributed to point mutations in the drug binding region. We designed a new class of peptide inhibitors to target the substrate binding site of BCR-ABL by transplanting a sequence derived from abltide (the best substrate for Abl kinase) onto the cell-penetrating cyclic peptide MCoTI-II point. Using the novel kinase inhibition test, the three grafted cyclic peptides showed significant Abl kinase inhibition in vitro in the low micromolar range. Our work also showed that the redesigned MCoTI-II with abltide sequence grafted in loop 1 and loop 6 inhibited the activity of [T315I]Abl, a mutant Abl kinase containing a "gatekeeper" mutation, in vitro. This mutation is notorious for being multi-drug resistant. The results of serum stability and cell internalization studies confirmed that the MCoTI-II scaffold provides enzyme stability and cell permeability properties for the lead molecule. In conclusion, our study emphasizes that the redesigned cyclic peptide containing abltide-derived sequences is a promising substrate-competitive inhibitor of Abl kinase and T315I mutants.

Chronic myeloid leukemia (CML) is a disease of the hematopoietic system characterized by increased and unregulated growth of myeloid cells. From the onset of onset, CML usually goes through three clinical phases: chronic phase, accelerated phase, and terminal phase. The Philadelphia chromosome (Ph) is a cytogenetic marker of CML, derived from the mutual translocation of chromosomes 9 and 22. This fusion between the breakpoint cluster (BCR) gene on chromosome 22 and the Abelson (ABL) tyrosine kinase gene on chromosome 91,2 forms the BCR-ABL oncogene. The BCR-ABL gene encodes a consistently high level of cytoplasmic and constitutively active BCR-ABL tyrosine kinase, which is detected in >90% of CML patients and 25% of adult patients with acute lymphoblastic leukemia (ALL)3. Animal studies also provide evidence to support BCR-ABL as the carcinogenic cause of CML because it has been shown to induce myeloproliferative syndrome that is very similar to the chronic phase of human CML4.

Philadelphia chromosome positive (Ph) patients in the chronic phase of CML rely on continuous administration of small molecule tyrosine kinase inhibitors (TKI). The first-line therapy is imatinib mesylate (IM, also known as STI-571 or Gleevec®), which is a TKI that binds to the inactive form of BCR-ABL in the ATP gap and prevents the conformation required for kinase activation Change 5. The clinical resistance of TKI therapy is an important issue in the treatment of patients with advanced CML1,6, mainly because the induction of point mutations in the kinase domain of BCR-ABL impairs the interaction between IM and ATP binding cracks7. Two second-generation TKIs, Dasatinib 8, 9 and Nilotinib 9, and a third-generation TKI, Bosutinib 10, 11, 12, were developed to overcome IM-resistant BCR-ABL mutations However, none of them showed significant activity against T315I-this is the most problematic mutant because it is resistant to multiple TKIs. In 2012, ponatinib13 (AP24534, Iclusig™) was approved by the US Food and Drug Administration (FDA) for the treatment of CML or ALL Ph patients with T315I mutation. Although ponatinib has shown effective inhibition of all clinically important BCR-ABL single mutants including T315I, compound mutants containing T315I mutations are highly resistant to this TKI13,14,15. Therefore, overcoming BCR-ABL-dependent resistance to current CML therapies remains a major challenge for drug design.

In addition to the ATP cleft, the catalytic domain of BCR-ABL (Figure 1a) also includes a second different site: the substrate binding site. The contact area between the kinase substrate and the kinase domain is larger than that of ATP, and the substrate binding site is specific to each kinase, which indicates that compared with TKI, inhibitors targeting this site are less affected by mutations . Therefore, peptide inhibitors that target the substrate binding site are an alternative strategy that can be used to inhibit BCR-ABL with a higher specificity than small molecule TKIs.

In this study, the three-dimensional structure of Abl kinase and MCoTI-II and the amino acid sequence of MCoTI-II variants were considered.

(a) Abl kinase, the substrate-ATP conjugate binds to the catalytic site (PDB ID: 2g2f). The substrate (abltide, magenta) binds in the gap between the N lobe and the C lobe; the phosphorylation site faces the ATP binding pocket in the N lobe. (b) The three-dimensional structure and amino acid sequence of natural MCoTI-II (PDB ID: lib9). Cysteine-rich peptides have a unique cyclic cystine knot (CCK) motif, including a cyclic backbone and three interlocking disulfides (shown in yellow). The starting point of the peptide sequence (G1) is connected to the corresponding position on the ribbon structure with a dotted line. Six cysteine ​​residues divide the backbone into six loops. Cycles 1 and 6 replaced by foreign sequences in this study are highlighted in red and blue, respectively. (c) Sequence alignment of natural MCoTI-II and MTAbl peptides. The six cysteines are highlighted in yellow and are numbered using Roman numerals (I–VI). The foreign sequence containing the Abl kinase recognition motif inserted into loop 1 or 6 is colored in red and blue, respectively. Phosphorylated tyrosine is shown in bold, and phosphorylated tyrosine residues are marked with an asterisk. Cys I-IV, II-V, and III-VI disulfide bonds are shown with dark gray lines. MCoTI-II and all MTAbl peptides are cyclized head-to-tail and are represented by light gray lines. The affinity of MTAbl00 and MTAbl08 to Abl kinase was only assessed using molecular modeling (marked with superscript M).

Substrate-based kinase inhibitors are usually designed using knowledge about a series of peptide substrates17,18. Using 2.5 billion synthetic peptides and 9 tyrosine kinases19,20 large-scale studies of kinase specificity led to the identification of the required common motifs Ile/Val/Leu-Tyr-Xaa-Xaa-Pro/Phe (among them Xaa is any amino acid) Abl kinase substrate recognition. Since Abl kinase has the same characteristics as the catalytic domain of BCR-ABL, which is essential for its carcinogenic activity, abltide (EAIYAAPFAKKK) is the best substrate for Abl kinase containing a common motif and can be used as a rationally designed The starting point is a substrate-based oncogenic BCR-ABL inhibitor.

Although peptides have the characteristics of high target specificity and low toxicity, their development as therapeutic agents is hindered by their low stability and restriction of access to intracellular spaces 21. The discovery of cyclic peptides, peptides showing a head-to-tail cyclic backbone and a cystine knot motif (Figure 1b), conferred high stability against physical and enzymatic degradation22, opening up new possibilities for the development of peptide-based drug candidates Some cyclic peptides, such as MCoTI-I and MCoTI-II (coyote trypsin inhibitor-I and II)23, have also been reported to penetrate cells 24, 25 (subsequently classified as cyclic cell penetrating cells). Permeable peptide 26), indicating that they can also be used to deliver biologically active peptides into cells. In fact, they have been successfully used as scaffolds for designing stable peptide therapy with non-natural biological activity and/or for intracellular delivery 27, 28, 29, 30, 31, 32. Examples include: transplantation of biologically active sequences that are active against foot-and-mouth disease virus 3C protease into loop 1 of MCoTI-II32, effective pro-angiogenic peptides inserted into loop 6 of MCoTI-II28, and introduction of the terminal fragment of N-p53 into MCoTI-I The ring 6 produces an anti-tumor compound 31.

In the current study, MCoTI-II was selected as the framework to stabilize linear peptide sequences that are active on Abl kinase and deliver them to cells. In this way, we have developed new peptide-based inhibitors with prolonged serum stability and cell permeability properties, which target the substrate binding site of BCR-ABL and may prevent the binding of endogenous substrates.

Eight kinds of MTAbl peptides with exogenous sequences grafted into MCoTI-II loop 6 or loop 1 were designed using molecular models. Six of these peptides and several variants designed to increase cellular uptake were synthesized. The inhibitory activity against natural Abl kinase and [T315I] Abl mutants, as well as serum stability, cytotoxicity, and cellular uptake of lead compounds were evaluated.

First, a molecular model of linear abltide combined with Abl kinase was constructed to facilitate the design of grafted MCoTI-II analogs. The model is established by homology with the crystal structure of a variant of abltide, which has an inactive form of Abl kinase 5 and an active form of protein kinase A with peptide substrates, ATP and two magnesium ions (PDB ID: 1atp) combined. The molecular dynamics (MD) simulation of this model is stable within 40 ns, and the stem root mean square deviation (rmsd) from the crystal structure conformation is only ~2 Å (Supplementary Figure S1). The conformation of abltide is almost the same as the crystal structure of modified abltide, and the main chain rmsd is ~1.5 Å.

Then use abltide as an anchor point to "docking/graft" the MCoTI-II scaffold into the Abl/abltide complex to generate the peptide MTAbl00-03 and 08-11 models complexed with Abl kinase. Using three repetitions of 20 ns MD simulation to study each system, the binding postures obtained indicate that the tested MTAbl peptide mainly interacts with the C lobe of Abl kinase (Supplementary Figure S2). A total of four MTAbl peptides grafted in ring 6 and four MTAbl peptides grafted in ring 1 are considered for molecular modeling, as detailed below.

Design a loop with graft in loop 6 by replacing the natural loop with the first 9 residues of abltide (MTAbl00) and replacing an extra glycine residue at the N-terminus (MTAbl01), C-terminus (MTAbl03), or both Peptide terminal (MTAbl02). After fitting the complex structure to the kinase backbone, the grafted peptide backbone rmsd (graft rmsd) was used to evaluate the stability of the grafted sequence from its conformation in the linear abltide/Abl complex. The graft rmsd of MTAbl02 is ~1.5 Å, as shown in Figure 2, which is comparable to the skeleton rmsd measured during the simulation of the Abl and linear abltide complex (Supplementary Figure S3). Therefore, the simulation shows that the abltide grafted in MTAbl02 can adopt almost the same conformation as the linear abltide when it is compounded with Abl. In contrast, simulations of the other three MTAbl peptides show that the conformation of their grafted abltide is different from that of linear abltide. The grafted rmsd values ​​of MTAbl00, MTAbl01, and MTAbl03 are ~2.0 Å, ~2.5 Å, and ~3.0-4.0 Å, respectively (Figure 2). This model shows that the C-terminus of the transplanted abltide sequence is located in the negatively charged region generated by the Abl kinase. Therefore, several grafted analogs MTAbl04, MTAbl05, MTAbl06 and MTAbl07 in loop 6 were designed to incorporate additional basic residues at the C-terminus of the abltide sequence.

A molecular model of the interaction between Abl kinase and MTAbl00, MTAbl01, MTAbl02 and MTAbl03 peptides, including the graft in loop 6.

Above: The superposition of the binding modes observed after 20 ns molecular dynamics (MD) simulation of four MTAbl peptides and 50 ns MD simulation of abltide. Five MD frameworks are mounted on the kinase backbone. The molecular surface of Abl kinase is shown in white. The grafted peptide and linear abltide are displayed using a ribbon-like representation, and their side chains are displayed using a bar-like representation. The backbone of the MTAbl peptide is shown as a curve. Bottom: In the MD simulation of MTAbl peptide and Abl kinase complex, the evolution of the root mean square deviation of grafted abltide (graft rmsd). The grafted rmsd is calculated as the backbone rmsd between the grafted abltide and the corresponding sequence in abltide, after fitting the skeleton of Abl kinase.

The MCoTI-II project grafted into loop 1 initially focused on optimizing the length of the graft, because loop 1 of MCoTI-II has only 6 residues, while linear abltide has 12 residues. The designed peptide incorporates the first seven (MTAbl08), eight (MTAbl09) or ten (MTAbl10) residues of abltide into the scaffold. A nine-residue graft containing the first eight residues of abltide and preceded by a glycine residue (MTAbl11) was also considered. The molecular model study of the graft in ring 6 shows that adding glycine at the C-terminus is not conducive to binding, and adding a spacer to the N-terminus of the grafted peptide in MTAbl11 can make the peptide adopt and bind to the interface. The results of MD simulation show that the grafted fragment of MTAbl09 adopts almost the same conformation as the linear abltide at the interface (grafted rmsd ~1.5 Å). The conformation of the grafted fragments in the other three peptides MTAbl08, MTAbl10 and MTAbl11 deviated from the optimal abltide conformation, as shown by their grafted rmsds of 2.5 Å or higher (Supplementary Figure S4). It was found that the simulation of MTAb108 was particularly unstable, resulting in a very high transplant rmsd value. The three simulations of MTAbl09 converged to similar binding modes, but the simulations of MTAbl10 and MTAbl11 did not, indicating that the scaffold did not form a strong or specific interaction with the kinase.

14 kinds of mutant peptides (MTAbl01-07, 09-15) grafted with abltide analogs in ring 1, ring 6 or ring 1 and ring 6 were synthesized (Figure 1c). The one-dimensional 1 H NMR spectra of all MTAbl peptides show that the peaks in the amide region are well dispersed and aligned with the MCoTI-II spectra, indicating that these peptides are folded in natural form (data not shown). Two-dimensional homonuclear NMR was used to determine the Hα chemical shifts of three representative MTAbl peptides, namely MTAbl07, MTAbl09 and MTAbl13. The comparison of the secondary Hα chemical shifts of these grafted peptides and MCoTI-II is shown in Figure 3. The secondary Hα chemical shift of the scaffold position (the gray shaded area in Figure 3) is equivalent to the equivalent position of MCoTI-II, indicating that the three-dimensional structures of MTAbl07, MTAbl09 and MTAbl13 are similar to the parent peptide MCoTI-II. As expected, major differences in Hα chemical shifts between MCoTI-II and MTAbl mutants were observed in the mutant loops (ie loop 1 and/or loop 6) or the region near these two loops.

Secondary Hα chemical shift analysis of MCoTI-II and selected graft analogs MTAbl07, MTAbl09 and MTAbl13.

The color used for each peptide is shown in the key in the upper right corner. The dotted line represents the secondary chemical shift values ​​of -0.1 and 0.1 ppm. The sequence alignment of the four peptides is shown below the graph. The secondary Hα chemical shifts of regions with the same sequence are shown in gray shading. The yellow circle marks the position of the cysteine ​​residue.

The kinase inhibitory effect of grafted peptides was evaluated using the BacKin assay 33, which quantified the phosphorylation of abltide displayed on the surface of E. coli cells. After incubation with biotinylated anti-phosphotyrosine antibody and subsequent labeling with streptavidin-conjugated phycoerythrin, flow cytometry was used to measure the percentage of abltide phosphorylated by Abl kinase. Unlike typical kinase inhibition methods, such as the ADP-Glo​​™ Kinase Assay (Promega), which evaluates kinase activity by quantifying the amount of ADP formed by the kinase reaction, the BacKin assay provides an alternative method to quantify kinase inhibitory activity . The level of phosphorylated substrate on the solid support. This new assay was developed to evaluate the inhibitory activity of kinase substrates, which cannot be fairly evaluated using ADP-Glo​​™.

Abltide has inhibitory activity against Abl kinase at low micromolar concentrations (IC50: 5.7 μM), as measured by the BacKin test, and interestingly, phosphorylated abltide (p-abltide) has similar activity. This similarity indicates that abltide, whether non-phosphorylated or phosphorylated, can bind to Abl kinase and inhibit its activity. Through the synthesis of a set of 17 abltide analogs, the importance of phosphorylated tyrosine was further studied, in which tyrosine was replaced by alanine, phenylalanine or one of 15 unnatural phenylalanine residues Replace (see Supplementary Figure S5). At concentrations up to 64 μM, none of these peptides showed significant inhibition of Abl kinase; therefore, the tyrosine residue at position 4 of abltide seems to be the best choice for binding and inhibiting Abl kinase. In addition, the results indicate that the phosphorylated substrate can also be used as an inhibitor, indicating that abltide can be used as an inhibitor of Abl kinase without replacing its phosphorylated tyrosine.

The Abl kinase inhibitory activity of the MCoTI-II loop 6 variant MTAbl01-03 (Figure 1c) was evaluated using the BacKin assay at a peptide concentration of 32 μM (Figure 4a). MTAbl01, MTAbl02 and MTAbl03 inhibited the kinase activity by 75%, 90% and 50%, respectively. These values ​​are related to the similar conformation of the grafted fragment and the conformation of the linear abltide bound to Abl kinase, as shown in the molecular model study (Figure 2); therefore, the conformation of the grafted fragment at the interface seems to affect the activity.

Inhibition of Abl kinase activity induced by MTAbl peptide measured using the BacKin test.

(a) The percentage of Abl kinase (0.2 U/mL) inhibited for 30 minutes at 37°C in the presence of 32 μM of each peptide. MTAbl peptides that are significantly different in inhibitory efficacy compared to abltide are marked with an asterisk (****p <0.0001; ***p <0.001, as assessed by one-way analysis of variance). (b) Increase the inhibition of Abl kinase activity induced by the concentration of IM, MTAbl13 or MTAbl14. MTAbl13* corresponds to the inhibition of [T315I]Abl kinase. The data was fitted to a sigmoidal dose response curve with a variable slope model, and GraphPad Prism 6 was used for analysis. The results shown here are the mean ± SEM from two independent experiments.

Based on our preliminary screening and molecular modeling, three analogues, MTAbl04, MTAbl05 and MTAbl06, were synthesized, aiming to have better charge complementarity with Abl kinase and to test their inhibitory activity. MTAbl04 was derived from MTAbl03 by substituting a lysine residue for the C-terminal glycine residue in ring 6, while MTAbl05 and MTAbl06 were derived from MTAbl03 by substituting a lysine or arginine residue for the C-terminal glycine residue in ring 6, respectively. It is derived from MTAbl02. MTAbl04 shows similar activity to MTAbl03, inducing ~50% reduction in kinase activity, while MTAbl05 and MTAbl06 have similar activities as MTAbl02, inhibiting ~90% of kinase activity. Therefore, adding a basic residue to the C-terminus will not result in a significant difference in activity at a peptide concentration of 32 μM. In order to understand the subtle differences between the variants, the inhibitor dose response curve was evaluated and the IC50 value was calculated (Supplementary Table S1). The difference between MTAbl05 and MTAbl06 is only in the C-terminal residues of loop 6, which are lysine and arginine residues, respectively, but the activity has a 3-fold difference (IC50: 15.2±7.0μM versus 5.5±1.6μM). In order to evaluate the contribution of the extra positive charge in loop 6 to kinase inhibition, the dose-dependent activity of MTAbl07, which has three arginine residues at the C-terminus of loop 6, was also evaluated. MTAbl07 has an activity comparable to MTAbl06 (IC50: 4.1 μM), indicating that the extra arginine residue in loop 6 does not significantly increase the binding to Abl kinase. Overall, the extra basic residue at the C-terminus of loop 6 only resulted in a small improvement in activity.

Among the three loop 1 variants (MTAbl09-11), only MTAbl09 achieved 90% inhibition at 32 μM, which is equivalent to the activity of the best inhibitor with loop 6 grafts, namely MTAbl02, MTAbl05, and MTAbl06. This result is consistent with our molecular model study, indicating that MTAbl09 is the only analogue grafted with abltide in loop 1, which can completely maintain the conformation adopted from linear abltide at the interface with Abl kinase. The incorporation of recognition motifs in loop 1 and loop 6 (MTAbl13) increased the inhibitory activity, compared to loop 1 (MTAbl09, IC50: 18.3 μM) or loop 6 (MTAbl06, IC50: 5.5 μM). Although MTAbl13 was found to be less potent than IM (IC50: 0.3 μM), its potency is comparable to other substrate-based kinase inhibitors in previous studies34,35. Finally, the activities of MTAbl12 and MTAbl14 (the phosphorylated versions of MTAbl06 and MTAbl13, respectively) were evaluated, and they were found to have comparable IC50 values ​​to non-phosphorylated peptides. This result is consistent with the measurement result of the inhibitory activity of phosphorylated abltide, indicating that the cyclic peptide incorporated into non-phosphorylated abltide can act as a kinase inhibitor.

Ph CML patients carrying the T315I mutation in BCR-ABL do not respond to most drugs currently available. Therefore, the MTAbl peptide has the greatest correlation with the inhibitory activity of [T315I]Abl kinase and has been studied in vitro. The catalytic activity of [T315I]Abl kinase was first checked by using LC-MS to compare the phosphorylation of 30 μM abltide catalyzed by 0.2 U/mL wild-type or [T315I]Abl kinase. The time required for wild-type and [T315I]Abl kinase to phosphorylate half of the abltide molecule was 24.2 ± 0.7 and 50.0 ± 1.6 minutes, respectively (Supplementary Figure S6). The dose response curve of IM, MTAbl13 and its phosphorylated form MTAbl14 to wild-type Abl kinase is shown in Figure 4b. The IC50 values ​​of IM, MTAbl13 and MTAbl14 are 0.3, 1.3 and 2.6μM, respectively. The inhibitory activity of MTAbl13 on [T315I]Abl kinase was 12.2 μM, while IM had no inhibitory activity within 256 μM (Figure 4b and Supplementary Table S1).

The proteolytic stability of MTAbl06, MTAbl07, MTAbl09, MTAbl12 and MTAbl13 as well as natural MCoTI-II and abltide in 100% human serum was evaluated for up to 24 hours (Figure 5). Except for MTAbl12, all tested MTAbl peptides are stable, with more than 75% peptide remaining after 24 hours of incubation. The observed lower stability of MTAbl12 is due to the dephosphorylation of p-Tyr residues, as indicated by RP-HPLC spectroscopy and mass spectrometry analysis, where the retention time and quality of the degradation products are the same as MTAbl06 (ie dephosphorylated MTAbl12) Same. According to one-way analysis of variance, MTAbl09 and MTAbl13 are statistically more stable than MCoTI-II (p <0.05 and p <0.001, respectively), while MTAbl07 is as stable as MCoTI-II. Compared with cyclic peptides, linear abltide is completely degraded in human serum within 1 hour. Compared with the control sample in PBS solution, the binding of the test compound to serum protein or Eppendorf Tubes® is negligible (<5%).

Stability of MTAbl peptide in human serum.

The time course of the percentage of remaining peptides when incubated in 100% human serum at 37°C for up to 24 hours. A one-way analysis of variance was used to compare the percentage of remaining MTAbl peptide with the percentage of MCoTI-II at the 24-hour time point. After 24 hours of incubation in human serum, MTAbl09 and MTAbl13 were significantly more stable than MCoTI-II (*p <0.05 and ***p <0.001, respectively). The result is the mean ± SEM of three replicates.

Since BCR-ABL is mainly located in the cytoplasm, it is important to evaluate whether the designed peptide can be internalized into the cell. MTAbl06, MTAbl07, MTAbl09, MTAbl13, and MTAbl15 were fluorescently labeled with an AlexaFluor® 488 molecule (A488), and their cell penetration characteristics were compared with the labeled MCoTI-II using flow cytometry26,36. TAT is a well-studied cell penetrating peptide 37 that serves as a positive control. Initially, we confirmed that in the resazurin-based cell viability assay, none of these peptides showed cytotoxicity to HeLa cells at concentrations up to 64 μM for 2 hours (data not shown). Then HeLa cells were treated with five concentrations of A488-labeled MTAbl peptide alone for 1 hour, and the average fluorescence emission intensity of the treated cells measured at 530/30 nm using flow cytometry is shown in Figure 6a. The Alexa-labeled peptide enters the cell in a dose-dependent manner, and the uptake efficiency level is: TAT> MTAbl15> MTAbl07> MTAbl06 ≥ MCoTI-II. The peptide uptake comparison of A488-MCoTI-II relative to 8 μM is shown in Figure 6b. A488-MTAbl07 and A488-MTAbl15 showed statistically higher cellular uptake than A488-MCoTI-II (p <0.05 and p <0.001, respectively). In contrast, compared with the same concentration of MCoTI-II, A488-MTAbl09 and A488-MTAbl13 showed a 25% and 10% reduction in cell uptake, respectively. No significant decrease in fluorescence intensity was observed after the addition of Trypan Blue. This is a non-permeable quencher that can quench the fluorescence of non-internalized/membrane-bound peptides (Figure 6b), indicating that the monitored fluorescence is not from the membrane Binding but derived from internalization peptide.

MCoTI-II, MTAbl06, MTAbl07, MTAbl09, MTAbl13 and MTAbl15 are internalized into HeLa cells.

(a) The peptide is conjugated to an Alexa Fluor® 488 molecule. A flow cytometer with an excitation wavelength of 488 nm and an emission wavelength of 530 nm (30 nm bandpass) was used to analyze the average fluorescence intensity (au) of HeLa cells treated with different concentrations of peptides at 37°C for 1 hour. Include cell penetrating peptide TAT as a positive control. The values ​​plotted in the graph are obtained after adding trypan blue (TB), a water-soluble quencher, which quenches the fluorescence of non-internalized peptides and membrane-damaged cells. (B) The relative cellular uptake of MCoTI-II by cells treated with TAT or five transplanted peptides before and after the addition of TB. The internalization of MTAbl peptide into HeLa cells was compared with MCoTI-II by using one-way analysis of variance after adding TB to compare the average fluorescence emission intensity (*p <0.05; *** p <0.001). The results shown here are the mean ± SEM from three independent experiments.

MTAbl13 contains a graft in its two loops and has the most effective inhibitory activity among all MTAbl peptides; therefore, its tertiary structure in solution was elucidated by nuclear magnetic resonance spectroscopy. The distance and dihedral angle limits are derived from one-dimensional and two-dimensional homonuclear 1 H NMR experiments and are used for structural calculations (Supplementary Table S2 and Supplementary Figure S7). The detailed statistics are shown in Supplementary Table S2, and the structure coordinates have been stored in the protein database (PDB ID: 2mt8). As shown in Figure 7, loops 2, 3, 4, and 5 of MTAbl13 are conformationally similar to the corresponding loops of MCoTI-II, which is consistent with the secondary Hα chemical shift analysis presented in Figure 3. The connected disulfide bond found that ring 1 and ring 6 adopt a different conformation from MCoTI-II. When contacted with Abl kinase, this change may result in a slight change in the conformation of the grafted abltide, and we hypothesized that this may be the reason for the increased activity observed compared to the single grafted peptides MTAbl06 and MTAbl09.

The superposition of the NMR solution structure of MCoTI-II and MTAbl13.

The skeletons of MCoTI-II (blue, PDB ID 2ib9) and MTAbl13 (purple) are displayed using stick representation. The disulfide bonds forming the cystine knot motif are yellow bars.

The secondary Hα chemical shift analysis shows that ring 1 and ring 6 of MTAbl13 are usually disordered. The superposition of the 20 lowest energy conformations calculated from the structure further shows that ring 1 and ring 6 are flexible compared to the rest of the molecule (Supplementary Figure S8). Compared with loop 6, loop 1 adopts a more constrained conformation (Supplementary Figure S8), but both loops protrude from the cystine knot core and may be inserted into the substrate binding pocket of Abl kinase. None of the NMR models of the more constrained conformation of loop 1 correspond to the kinase-bound conformation. In contrast, loop 6 is more flexible and may therefore be able to adopt a bound conformation.

A resazurin-based cell viability assay was used to evaluate the effect of MTAbl peptide on the growth and apoptosis of K562 cells. K562 is an immortalized cell line derived from a 53-year-old female CML patient, 38 has been widely used as a model for CML research during the explosion crisis. Although MTAbl peptides showed potent activity against Abl kinase in vitro, MTAbl06, MTAbl07, MTAbl13, and MTAbl15 did not inhibit the growth of K562 by up to 64 μM (Supplementary Table S1).

Although current first-line therapies are initially effective in inhibiting cancer progression, mutations in the ATP binding site render these drugs ineffective, and drug resistance may develop after several years of continuous treatment. All human kinases have similar ATP binding pockets, so it is difficult to design highly specific drugs for these sites. On the other hand, due to the functional requirements of targeting specific substrates, the ligand binding site of kinases is more obvious than the ATP binding cleft, and it is less prone to mutation 39,40, which indicates that substrate-based inhibitors have the potential to be developed as TKI with high specificity.

We tried to convert abltide from a substrate to an inhibitor by replacing the phosphorylated tyrosine with natural or unnatural amino acids, but the peptides produced were inactive. Then the BacKin inhibition test showed that phosphorylated abltide (p-abltide) is as active as abltide, which indicates that the substrate of Abl kinase can act as an inhibitor after being phosphorylated. Similarly, the phosphorylated MTAbl peptide showed similar activity to its non-phosphorylated form, namely MTAbl12 vs. MTAbl06 and MTAbl14 vs. MTAbl13. Certain aspects of phosphorylated substrates that inhibit Abl kinase are still unclear; for example, the state of the kinase to which phosphorylated peptides bind is unknown. In fact, it is impossible for phosphorylated peptides to bind to the inactive conformational state of Abl kinase because the activation loop adopts a conformation that is incompatible with substrate binding. Because the charged phosphate group of phosphorylated tyrosine will repel ATP molecules in space and charge, another state of the kinase is that the activation loop is in active form but the ATP crack is empty or occupied by ADP molecules. It may be the state of phosphorylated peptide targeting.

A total of 14 synthetic cyclic peptides were designed, which contained one or two grafted abltides, and generally showed moderate to high inhibitory activity against Abl kinase. The most effective peptides, namely MTAbl06, MTAbl07 and MTAbl13, exhibit low micromolar IC50 values, as tested using the BacKin assay. The IC50 values ​​determined using this assay may underestimate the ability of MTAbl peptides to inhibit Abl kinase, because these values ​​are measured based on competition with abltide, which is the best substrate for Abl kinase and has a higher affinity for Abl than its physiology Substrate. In fact, the Michaelis-Menten constant Km of Abl phosphorylated abltide is between 4 μM20 and 21 μM41, but it has a flanking sequence of phosphorylated tyrosine (Tyr-207) derived from the classical substrate CRKL The phosphorylation of the peptide has a Km of 134 μM41. The Km of abltide is in the same order of magnitude as the optimal IC50 value of the MTAbl peptide, indicating that the correspondingly designed peptide may have reached the maximum potential of abltide. Molecular modeling studies have shown that although the scaffold shows a different orientation relative to the kinase, the relative efficacy of the MTAbl peptide is mainly determined by the ability of the transplanted abltides to adopt the optimal conformation. This result indicates that the natural MCoTI-II scaffold does not significantly contribute to the affinity with kinases.

The MTAbl peptide grafted in loop 1 (for example, MTAbl09) or loop 6 (for example, MTAbl06) resulted in effective inhibition of Abl kinase. Additional basic residues at the C-terminus of the graft, such as MTAbl06 or MTAbl07, also showed a slight increase in kinase inhibitory activity. Compared with the corresponding single grafted peptides MTAbl09 and MTAbl06, the introduction of abltide-derived sequences into loops 1 and 6 of MTAbl13 increased the inhibitory activity by 14-fold and 3-fold, respectively. The NMR solution structure of MTAbl13 shows a small relative displacement of the cysteines anchored by loops 1 and 6 compared with MCoTI-II, and is presumed to be compared with the single grafted analog. This substitution may result in different presentation of the grafted peptide to the kinase. If both grafted abltides remain active, the increase in local concentration is a possible explanation for the increased activity. As another recent study proposed, grafting two copies of the same active peptide into the backbone ring scaffold also Lead to an increase in activity42.

MTAbl13 showed inhibitory activity against wild-type and [T315I] Abl kinase. Nevertheless, the IC50 of MTAbl13 to [T315I]Abl is 9 times higher than that of Abl kinase. Although MTAbl13 showed reduced efficacy on [T315I] mutants compared with wild-type kinases, IM had no effect on [T315I] mutants up to 256 μM, which is> 800 times. The order of magnitude difference in the inhibitory activity of MTAbl13 against wild-type and mutant kinases may be due to the reported reduced kinase activity of [T315I]Abl on the classic BCR-ABL substrate.

The stability of the MTAbl peptide in human serum was evaluated to provide guidance on its potential stability in vivo. Time course studies show that the stability of MTAbl peptide is equal to or higher than that of MCoTI-II. Interestingly, despite the high sequence similarity, MTAbl07 is not as stable as MTAbl06, and the slight decrease in stability of MTAbl07 may be due to the two additional basic residues (Arg) contained in loop 6. Compared with MTAbl13, the serum stability of MTAbl13 was significantly increased (p <0.001). MCoTI-II after 24 hours of incubation was observed, which can be explained by the fact that there are fewer positively charged residues in MTAbl13 compared to MCoTI-II (Supplementary Table S1). Overall, the results clearly show that despite the sequence variation, the cyclic cystine knot motif confers excellent enzyme stability to the MTAbl peptide (half-life> 24 hours).

The cell uptake efficiency of the substrate-based inhibitor showed that the internalization efficiency of A488-MTAbl07 into HeLa cells was 2.2 times higher than that of A488-MCoTI-II. This finding is consistent with previous studies, demonstrating that the cellular uptake of MCoTI-II can be enhanced by increasing the overall positive charge of the peptide44,45. A488-MTAbl15 is an analog of A488-MTAbl07, in which the first and second lysine residues in loop 1 are replaced by Arg, and its cellular uptake is 3.5 times higher than that of A488-MCoTI-II, which indicates that MCoTI-II The cell internalization of the scaffold can be further improved by replacing Lys with Arg in loop 1 of the molecule.

Although it has high inhibitory activity on Abl kinase, when administered at a concentration of up to 64 μM, the transplanted peptides MTAbl06, MTAbl07, MTAbl13, and MTAbl15 have no significant effect on the growth of K562 cells. One possible explanation for their unexpectedly low potency on K562 cells is that only a limited number can reach BCR-ABL in the cytoplasm. Although we showed that the transplanted peptides can penetrate cells, and in fact, compared with MCoTI-II, MTAbl07 and MTAbl15 have improved internalization efficiency, but a certain number of transplanted peptides may have been trapped in the cell compartment (such as endosome )middle. Our future work will focus on optimizing the internalization of MTAbl peptides to increase the number of peptides reaching the cytoplasm, thereby increasing their effective concentration at the protein-protein interaction interface involved in the BCR-ABL signaling network. Molecular dynamics simulation studies proposed a potential strategy to increase the inhibitory activity by generating specific interactions between the MCoTI-II scaffold and Abl kinase and by capturing the transplanted abltide into a conformation compatible with Abl kinase binding. Whether the substrate-based Abl kinase inhibitor designed in this study can actively inhibit BCR-ABL kinase, whether it is wild-type or a T315I mutation in the cytoplasm, remains to be determined; however, kinase inhibitory activity is related to serum stability and increased cellular uptake The combination of efficiency makes these peptides promising candidates for next-generation TKIs.

Natural MCoTI-II was isolated from the seed extract of M. cochinchinensis, as described above26.

The initial molecular model of the binding of abltide to active Abl kinase, in which an ATP molecule and two magnesium ions bound in the ATP crevice were established by homology and improved using 50 ns MD simulation. An anchor-based method was used in Modeller to generate a model of the complex between the grafted cyclic peptide and Abl kinase, and the 20 ns MD simulation was used to improve it. The initial molecular model of abltide binding to active (DFG-in) Abl kinase, in which one ATP molecule and two magnesium ions are combined in the ATP cracks, based on three templates established through homology: inactive (DFG-out) crystals Structure) Abl kinase combined with chemically modified abltide (PDB ID 2g2f), active (DFG-in) Abl kinase combined with ADP molecule (PDB ID 2g2i) crystal structure and ATP combined with active cAMP kinase crystal structure and two Magnesium ion (PDB ID 1atp). Then use ProPKA46 to calculate the protonation state of all side chains and ends. The three-dimensional RISM47,48 implemented in AmberTools 1349 and placevent50 was used to calculate the position of bound water molecules. According to NACCESS (V2.1.1)51 measurement, only water molecules with accessible surface area less than 20 A2 are retained. The system was then placed in a dodecahedron box and solvated with a total of approximately 20,000 SPC water molecules. The system uses a 10,000-step steepest descent algorithm to minimize energy. Water molecules are balanced in a 1 ns molecular dynamics simulation, and non-water molecules are strongly confined to their initial positions. Then the system was submitted to a 50 ns molecular dynamics simulation, using gromos 54a7 force field 52, a relative permittivity ε of 6153 reactive field electrostatics, a v-rescale temperature coupling set to 300 K, and a Berendsen pressure coupling set to 1. Atm55, the Coulomb radius is 1.4 Å, the Coulomb radius switch is 0.8 Å, and the Verlet cutoff scheme for calculating van der Waals interaction. All bonds are subject to the LINCS algorithm. Use Gromacs 4.6.557 for energy minimization and molecular dynamics simulations. These three simulations were stable in the past 40 ns and tended to a similar conformation. Then use the final model of MCoTI-II (PDB ID: 1ib9) and the NMR solution structure as a template, and use Modeller 9v12 to construct a model combining Abl kinase and MTAbl peptides (MTAbl00, MTAbl01, MTAbl02, MTAbl03, MTAbl08, MTAbl09), MTAbl10, MTAbl11) uses the bound peptide as an anchor. Then, using the same molecular modeling scheme described for the linear abltide/Abl system, eight complexes were prepared and studied by 20 ns molecular dynamics simulation. Use VMD58 and gromacs tools to analyze molecular dynamics simulations. Use APBS software to calculate the electrostatic potential generated by Abl kinase.

As mentioned earlier, the linear precursors of 14 MTAbl peptides containing N-terminal Cys residues were manually synthesized using solid-phase peptide synthesis and Boc chemistry. In short, the precursor peptide is constructed on PAM-Gly-Boc resin with S-trityl mercaptopropionic acid as a thioester linker and cleaved from the resin with hydrogen fluoride. Through sulfur heterocyclic reaction, linear precursor peptides with multiple Cys and C-terminal thioesters undergo a series of intramolecular acyl rearrangements to form α-aminothiolactone, and finally irreversible S and N acyl migration occurs. The amide bond formed between the end 60,61. The cyclization and oxidation of the MCoTI-II mutant was performed in 0.1 M ammonium bicarbonate (pH 8.2) at a peptide concentration of 0.1 mg/mL. The peptide mixture was stirred at room temperature for 24 hours, and purified by reverse phase high performance liquid chromatography (RP-HPLC) on a preparative Phenomenex C18 column. Solvent B (90% v/v acetonitrile/10% v/v H2O/0.045% v/v TFA (trifluoroacetic acid)) vs. Solvent A (100% v/v H2O/0.05% v/) 1%/min V TFA on gradient RP-HPLC was used to separate peptides.

Abltide, [Y4A]abltide, [Y4F]abltide, and p-abltide contain phosphorylated tyrosine in the fourth position, synthesized with 15 other abltide analogs (abltide1-15), and the same position is phenylalanine Derivative substitution, as shown in Supplementary Figure S5. The Phe derivatives used in this study are: 1. 4-fluoro-L-phenylalanine, 2. 4-chloro-L-phenylalanine, 3. 4-methyl-L-phenylalanine, 4. 4-nitro-L-phenylalanine, 5. 4-amino-L-phenylalanine, 6. L-homophenylalanine, 7. 3,4-difluoro-L-phenylalanine Acid, 8. 3,4-Dimethoxy-L-phenylalanine, 9. Pentafluoro-L-phenylalanine, 10. p-trifluoromethyl-L-phenylalanine, 11 . 4-Benzoyl-L-phenylalanine, 12. p-phenyl-L-phenylalanine, 13. 4-iodo-L-phenylalanine, 14. 4-cyano-L-benzene Alanine and 15.4-tert-butyl-L-phenylalanine. Abltide and its 18 analogs were synthesized on an automatic peptide synthesizer (Symphony®, Protein Technologies) using solid-phase peptide synthesis and Fmoc chemistry. In short, by successively adding Fmoc-protected amino acids, the peptide chain is assembled on 2-chlorotrityl chloride (2-CTC) resin. The peptide was then removed from the resin by treatment with TFA/triisopropylsilane/H2O (95:2.5:2.5, v/v) for 2 hours. Rotary evaporation was used to remove TFA from the mixture, and the remaining solution was mixed with cold ether and solvent A/B (1:1, v/v). The aqueous layer containing crude linear peptides was collected and purified to 95% purity by RP-HPLC on a preparative C18 column. The quality of the peptide was confirmed by ESI-MS, and the purity was checked by analytical HPLC.

Using liquid chromatography-mass spectrometry, we analyzed 0.2 U/mL human active Abl kinase (1067 U/mg, EMD Millipore) and [T315I]Abl (Abl(T315I) protein, EMD Millipore) phosphorylated 30 μM abltide Time course spectroscopy (LC/MS, Shimadzu LCMS-2020). Prepare Abltide and Abl kinases and mix them in kinase buffer (50 mM Tris-HCl, 10 mM MgCl2, 0.1 mM EDTA, 2 mM dithiothreitol and 0.01% Brij 35 (v/v), pH 7.5; for protein Kinase NEBuffer, New England Biolabs) and trigger the reaction by adding 500 μM ATP (adenosine 5'-triphosphate disodium salt hydrate, Sigma-Aldrich). Incubate the mixture with gentle shaking at 37 °C and stop with 10% 20 mg/mL dihydroxybenzoic acid at nine time points (0, 5, 10, 20, 30, 45, 60, 90 and 120 minutes) reaction. The samples were injected into a C18 analytical column (Jupiter 300 5 μm 300 Å, 150 × 2.0 mm, Phenomex), and MS was used to evaluate the quality of each component. The percentage of phosphorylation of abltide was determined by comparing the peak area of ​​phosphorylated abltide (1417 Da) with the peak area of ​​30 μM abltide from LC (1336.6 Da). The experiment was repeated three times in three independent days.

BacKin is a method to evaluate the inhibition of Abl kinase activity by using bacterial surface display substrates, as described above. In short, subculture E. coli (MC1061) cells (cells expressing abltide) showing abltide sequence on the surface and grow them at 37°C with continuous shaking at 225 rpm for 2 hours or until the optical density of the culture reaches 600 nm. 0.6. Induce the culture with 0.04% (w/v) l-arabinose and incubate for another hour at 37°C. Then the cells expressing Abltide were centrifuged at 3000 g for 4 minutes, washed with cold phosphate buffered saline (PBS) and resuspended in 0.2 U/mL Abl kinase or [T315I] Abl kinase buffer (NEBuffer, New England Biolabs). The abltide expressing cells were incubated with the designed peptide in a concentration range of 64 μM to 0.5 μM, and then 500 μM ATP was added to evaluate the inhibition of Abl kinase activity. The measurement is performed in duplicate for 30 minutes on a shaker at 37°C at a speed of 180 rpm. The reaction was terminated by centrifugation at 3000 g for 4 minutes at 4°C. The cells were washed twice with ice-cold PBS, and then incubated with 1 μg/mL biotinylated antibody anti-phosphotyrosine 4G10 (Merck Millipore) at 4°C on a shaker for 45 minutes. Before adding 5 μg/mL streptavidin-R-phycoerythrin (SAPE, Invitrogen), the cells were centrifuged and the antibody-containing supernatant was removed. The cells were incubated for another 45 minutes at 4°C, after which the remaining SAPE was removed. Flow cytometry (BD Canto II) is used to monitor the fluorescence intensity of each sample using 488 nm excitation and a 530/30 bandpass filter.

The correct folding of the MTAbl peptide was confirmed using 1H NMR. Dissolve the peptide in H2O/D2O (9:1, v/v) to a concentration of 1–2 mg/mL. The NMR spectrum was recorded on a Bruker Avance-600 MHz NMR spectrometer at a temperature of 298 K. The mixing time of TOCSY and NOESY experiments are 80 ms and 100-300 ms, respectively. Use Sparky to analyze spectra. The internal reference concentration of the spectrum is 0.00 ppm 2,2-dimethyl-2-silylpentane-5-sulfonic acid (DSS).

The structure of MTAbl13 was calculated as previously described for other cyclic disulfide-rich peptides42. In short, samples of MTAbl13 are prepared in 90% H2O/10% D2O (9:1, v/v) or 99.96% D2O (Cambridge Isotope Laboratories, Inc.) at ~1 mM and pH~3. Use Avance-500 or Avance-600 MHz spectrometer (Bruker) to obtain 1D and 2D NMR spectra (1H, 1H TOCSY, NOESY, DQF-COSY and ECOSY and 1H, 13C HSQC). The spectra were processed with TopSpin (Bruker) and analyzed by CCPNMR62. Calculate the distance constraint between protons based on the relative intensity of the NOE cross peak. The 3JHN-Hα coupling constant is measured from the one-dimensional spectrum or from the reversed cross-peak splitting in the DQF-COSY spectrum, and the 3JHα-Hβ coupling constant is measured from the ECOSY spectrum. The framework amide protons involved in intramolecular hydrogen bonds are identified by their temperature sensitivity and deuterium exchange rate (Supplementary Figure 7). Use CYANA 3.063 for preliminary structural calculations, and then use CNS64 for further structural calculations and refinements in the water shell. The set of 20 structures with the lowest energy was selected, and detailed statistics are provided in Supplementary Table S2.

The stability of the abltide graft analogues MTAbl06, MTAbl07, MTAbl12, MTAbl13 and natural MCoTI-II were evaluated in 100% human male AB serum (Sigma) using a method modified from previous studies. All peptides were tested in 100% human serum at a final concentration of 30 μM. The peptides were incubated in human serum at 37°C for 0, 1, 2, 3, 5, 8, 11 and 24 hours. The reaction was stopped at the prescribed time, and the serum protein in each sample was denatured and precipitated, and then centrifuged at 17,000 g for 10 minutes to separate the serum protein from the peptide sample. Load 100 μL of the supernatant on an analytical column (150 × 2.0 mm) and run on RP-HPLC, using a linear 1%/min gradient of 0-40% solvent B at a flow rate of 0.3 mL/min. The same amount of peptide was incubated with PBS and treated in parallel with the serum-treated sample at 0 and 24 hours as a negative control. The elution time of each peptide is determined by the PBS control at time 0. Use the height of the serum-treated peptide peak from time 0 as the 100% recovery rate to calculate the percentage of remaining peptide at each time point.

Natural MCoTI-II, MTAbl06, MTAbl07, MTAbl09, MTAbl13, and abltide were labeled with Alexa Fluor® 488 (abbreviated as A488) using the method described by Cascales et al.26. In short, a peptide solution prepared in 0.1 M sodium bicarbonate (pH 8.3) was incubated with a 2-fold excess (molar ratio) of Alexa Fluor® 488 5-SDP ester (Molecular Probes, Life Technologies) at room temperature 2 Hour. Drum mixer. The conjugated peptide was purified using RP-HPLC, and the quality was confirmed by ESI-MS. A peptide labeled with an A488 molecule was collected and used for cell uptake studies. TAT (NH2-YGRKKRRQRRRPPQG-COOH) is labeled with the same fluorophore and used as a positive control in the internalization assay.

Human cervical cancer cells (HeLa) were inoculated in a 175 cm2 tissue culture flask (Falcon), and in Dulbecco's Modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum, 100 U/mL penicillin and 100 mg/mL streptomycin. ) Until 80% confluence. K562 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 2 mM l-glutamine, 10% fetal bovine serum, 100 U/mL penicillin and 100 mg/mL streptomycin, and Grow to 80% confluence.

The cytotoxicity of MTAbl peptide to HeLa cells was evaluated before the cell internalization assay. The cytotoxicity test of mammalian cells was carried out using the method described earlier. HeLa cells were seeded in 96-well tissue culture plates (2.5 × 103 cells/well) the day before the assay. Dissolve and dilute peptides TAT, MCoTI-II, MTAbl06, 07, 09, 13, and 15 in sterile water (dilute 2 times starting from 640 μM). The peptide solution was diluted 10-fold with serum-free medium and incubated with the cells in triplicate for 2 hours. Controls containing H2O or 0.01% (v/v) Triton X have 100% and 0% cell viability, respectively. After incubation, the peptide solution was removed, and 10 μL aliquots of 0.02% (w/v) sterile resazurin (resazurin sodium salt, Sigma) were added to each well containing 100 μL of fresh medium. The cells were cultured for 22 hours under standard conditions of 37 °C and 5% CO2. The absorbance of the plate was measured on a plate reader at 540 and 620 nm. The cytotoxicity of peptides abltide, MCoTI-II, MTAbl06, 07, 13, and 15 on K562 cells was evaluated using a resazurin-based assay at 5 × 103 cells per well.

The cellular uptake of fluorescently labeled peptides was checked in HeLa cells using the method previously described. In short, 105 cells/well were seeded in a 24-well plate and incubated overnight at 37°C in a humidified atmosphere (5% CO2). A peptide labeled with Alexa Fluor® 488 is dissolved in sterile water from 80 μM in 2-fold serial dilutions. In the absence or presence of labeled peptides at a final concentration of 0.5, 1, 2, 4, or 8 μM, add 100 μL of serum-free medium to each well and incubate with the cells for 1 hour at 37 °C . After incubating for 1 hour, the peptides were removed and the cells were washed with Dulbecco's Phosphate Buffered Saline (DPBS, Life Technologies). After that, the cells in the 24-well plate were treated with 0.25% Trypsin-EDTA (Gibco®, Life Technologies) for 3 minutes, washed off with DPBS and transferred to Eppendorf Tubes® individually. The cells were centrifuged at 500 g for 3 minutes to remove the supernatant and resuspended in 1 mL DPBS. FACS flow cytometry (BD Canto II) was used to measure the fluorescence intensity of cells excited at 488 nm using a 530/30nm bandpass filter.

The results are shown as the mean ± SEM from two or three replicates, as shown in the figure. A one-way analysis of variance and Bonferroni's intra-group multiple comparison test were used to determine statistical significance.

How to cite this article: Huang, YH, etc. Design of substrate-based BCR-ABL kinase inhibitors using cyclic peptide scaffolds. science. Rep. 5, 12974; doi: 10.1038/srep12974 (2015).

An, X. etc. BCR-ABL tyrosine kinase inhibitor in the treatment of Philadelphia chromosome-positive chronic myeloid leukemia: a review. Leukemia Research 34, 1255–1268 (2010).

Lugo, TG, Pendergast, AM, Muller, AJ & Witte, ON Tyrosine kinase activity and conversion efficiency of bcr-abl oncogene products. science. 247, 1079–1082 (1990).

CAS ADS article Google Scholar 

Quintas-Cardama, A., Kantarjian, H., and Cortes, J. Flying under the radar: a new wave of BCR-ABL inhibitors. Nat Rev drug discovery. 6, 834–848 (2007).

Daley, GQ, Van Etten, RA and Baltimore, D. Philadelphia chromosome P210bcr/abl gene induces chronic myeloid leukemia in mice. science. 247, 824–830 (1990).

CAS ADS article Google Scholar 

Levinson, New Mexico et al. Src-like inactive conformation in the abl tyrosine kinase domain. Public Science Library Biology. 4. e144 (2006).

Gore, I'll wait. Clinical resistance to STI-571 cancer treatment caused by mutation or amplification of the BCR-ABL gene. science. 293, 876–880 (2001).

Soverini, S. etc. ABL mutations in patients with advanced chronic phase chronic myelogenous leukemia who have pre-cytogenetic resistance to imatinib are associated with a greater likelihood of progression to a blast crisis and shorter survival: GIMEMA Chronic Myeloid Leukemia Working Group A study. J Clinical Oncology. 23, 4100–4109 (2005).

Talpas, M. et al. Dasatinib treats Imatinib-resistant Philadelphia chromosome-positive leukemia. N Engl J Med. 354, 2531–2541 (2006).

Kantarjian, H. et al. The application of nilotinib in imatinib-resistant CML and Philadelphia chromosome-positive ALL. N Engl J Med. 354, 2542–2551 (2006).

Weisberg, E. et al. Characterization of AMN107, a selective inhibitor of natural and mutant Bcr-Abl. cancer cell. 7, 129–141 (2005).

Shah, NP etc. Overcoming imatinib resistance with new ABL kinase inhibitors. science. 305, 399–401 (2004).

CAS ADS article Google Scholar 

Golas, JM, etc. SKI-606 is a dual 4-anilino-3-quinolinecarbonitrile inhibitor of Src and Abl kinase. Degeneration of K562 xenografts. Cancer research. 63, 375–381 (2003).

O'Hare, T. et al. AP24534 is a pan-BCR-ABL inhibitor for the treatment of chronic myeloid leukemia. It can effectively inhibit the T315I mutant and overcome mutation-based resistance. cancer cell. 16, 401–412 (2009).

O'Hare, T., Zabriskie, MS, Eiring, AM & Deininger, MW push the limits of targeted therapy for chronic myeloid leukemia. Nat Rev cancer. 12, 513–526 (2012).

Zabriskie, MS etc. The BCR-ABL1 compound mutation combined with the key kinase domain position confers clinical resistance to prnatinib in patients with Ph chromosome-positive leukemia. cancer cell. 26, 428–442 (2014).

Levitzki, A. & Mishani, E. Tyrphostins and other tyrosine kinase inhibitors. Annu Rev Biochem. 75, 93–109 (2006).

Chapelat, J. et al. Screening method for substrate activity of receptor tyrosine kinases: a proof-of-concept study. Eur J Med Chem. 57, 1-9 (2012).

Eldar-Finkelman, H. & Eisenstein, M. Peptide inhibitors targeting protein kinases. Curr Pharm Des. 15, 2463–2470 (2009).

Songyang, Z. etc. Use directed peptide libraries to determine the best substrate for protein kinases. Curry creatures. 4, 973–982 (1994).

Songyang, Z. etc. The catalytic specificity of protein tyrosine kinases is essential for selective signal transduction. nature. 373, 536–539 (1995).

CAS ADS article Google Scholar 

Sato, AK, Viswanathan, M., Kent, RB & Wood, CR Therapeutic peptides: technological progress promotes the development of peptides. Curr Opin Biotechnol. 17, 638–642 (2006).

Colgrave, ML & Craik, DJ Thermal, chemical and enzymatic stability of the cyclic peptide kalata B1: the importance of cyclic cystine knots. Biochemistry. 43, 5965–5975 (2004).

Felizmenio-Quimio, ME, Daly, NL & Craik, DJ Cyclic proteins in plants: The solution structure of a novel macrocyclic trypsin inhibitor from Momordica chinensis. J Biochemistry. 276, 22875-22882 (2001).

Greenwood, KP, Daly, NL, Brown, DL, Stow, JL & Craik, DJ Cyclic cystine knot small protein MCoTI-II is internalized into cells through macropinocytosis. Int J Biochem Cell Biol. 39, 2252–2264 (2007).

Contreras, J., Elnagar, AY, Hamm-Alvarez, SF & Camarero, JA The cellular uptake of the cyclic peptide MCoTI-I follows multiple endocytic pathways. J Control release. 155, 134–143 (2011).

Cascales, L. etc. Identification and characterization of a new family of cell penetrating peptides: cyclic cell penetrating peptides. J Biochemistry. 286, 36932–36943 (2011).

Gunasekera, S. etc. Engineering stable vascular endothelial growth factor-A antagonist: synthesis, structural characterization and biological activity of cyclic peptide graft analogs. J Medical Chemistry. 51, 7697–7704 (2008).

Chen, Ly et al. Use a stable, disulfide-rich ring stent to design pro-angiogenic peptides. blood. 118, 6709–6717 (2011).

Aboye, TL, etc. Design of a novel cyclic peptide-based CXCR4 antagonist with anti-human immunodeficiency virus (HIV)-1 activity. J Medical Chemistry. 55, 10729–10734 (2012).

Huang, CT etc. An oral active peptide bradykinin B1 receptor antagonist designed from a cyclic peptide scaffold for the treatment of inflammatory pain. Anger Wander Chemical. 51, 5620–5624 (2012).

Ji, Y. etc. The p53 tumor suppressor pathway is activated in vivo by engineered cyclic peptides. J Am Chem Soc. 135, 11623–11633 (2013).

Thongyoo, P., Roque-Rosell, N., Leatherbarrow, RJ & Tate, EW Chemical and bionic fully synthetic natural and engineered MCoTI cyclized products. Tissue Biomolecular Chemistry. 6, 1462–1470 (2008).

Enriquez, ST etc. A novel quantitative kinase assay using bacterial surface display and flow cytometry. Public Science Library One. 8. e80474 (2013).

Parang, K. etc. Mechanism-based protein kinase inhibitor design. Nat Structural Biology. 8, 37–41 (2001).

Ward, BC, Kavalukas, S., Brugnano, J., Barbul, A. & Panitch, A. Peptide inhibitors of MK2 are expected to inhibit abdominal adhesions. J Surgical Research. 169, e27–36 (2011).

Torcato, IM, etc. The antibacterial activity of Sub3 depends on membrane binding and cell penetration. Studentization. 14, 2013-2022 (2013).

Madani, F., Lindberg, S., Langel, U., Futaki, S. & Graslund, A. The cellular uptake mechanism of cell penetrating peptides. J Biophysics. In 2011, 414729 (2011).

Lozzio, CB & Lozzio, BB Philadelphia chromosome-positive human chronic myeloid leukemia cell line. blood. 45, 321–334 (1975).

Avrahami, L., Licht-Murava, A., Eisenstein, M., and Eldar-Finkelman, H. GSK-3 inhibition: high selectivity to achieve moderate efficacy. Journal of Biochemistry and Biophysics. 1834, 1410-1414 (2013).

Licht-Murava, A. & Eldar-Finkelman, H. Use substrate recognition to selectively inhibit protein kinases. Curr Pharm Des. 18, 2914–2920 (2012).

Hantschel, O. etc. BCR-ABL decouples the classic JAK2-STAT5 signaling in chronic myeloid leukemia. Nat. Chemical biology. 8, 285–293 (2012).

Conibear, AC etc. The cyclic cystine ladder of theta-defensin as a stable bifunctional scaffold: a proof-of-concept study using integrins to bind the RGD motif. Studentization. 15, 451–459 (2014).

Skaggs, BJ, etc. Phosphorylation of the ATP binding ring guides the carcinogenicity of resistant BCR-ABL mutants. Proc Natl Acad Sci USA. 103, 19466–19471 (2006).

CAS ADS article Google Scholar 

D'Souza, C., Henriques, ST, Wang, CK & Craik, DJ The structural parameters that regulate the cellular uptake of disulfide-rich cyclic cell penetrating peptides: MCoTI-II and SFTI-1. Eur J Med Chem. (2014).

Huang, YH, Chaousis, S., Cheneval, O., Craik, DJ & Henriques, ST Optimized the cyclic peptide backbone to improve cell permeability. Former pharmacist. 6, 17, 10.3389/fphar.2015.00017 (2015).

Article PubMed PubMed Central Google Scholar 

Olsson, MHM, Søndergaard, CR, Rostkowski, M. & Jensen, JH PROPKA3: Treat internal and surface residues consistently in empirical pKa predictions. J Chemical theoretical calculations. 7, 525–537 (2011).

Beglov, D. & Roux, B. Describe the integral equation for the dissolution of polar molecules in liquid water. J Physical Chemistry. B. 101, 7821–7826 (1997).

Kovalenko, A. & Hirata, F. The average potential between two molecular ions in a polar molecular solvent: a study by a three-dimensional reference interaction site model. J Physical Chemistry. B. 103, 7942–7957 (1999).

Case, DA etc. Amber 13. (University of California, San Francisco, 2012).

Sindhikara, DJ, Yoshida, N. & Hirata, F. Placevent: An algorithm for predicting the distribution of explicit solvent atoms-applied to HIV-1 protease and F-ATP synthase. J Computational Chemistry. 33, 1536–1543 (2012).

Hubbard, SJ & Thornton, JM NACCESS, computer program. (Department of Biochemistry and Molecular Biology, University College London, 1993).

Schmid, N. et al. Definition and testing of GROMOS force field versions 54A7 and 54B7. Eur Biophys J. 40, 843–856 (2011).

Heinz, TN, Gunsteren, WFV & Hünenberger, PH Comparison of four methods for calculating the dielectric constant of liquids through molecular dynamics simulation. J Chemical Physics. 115, 1125–1136 (2001).

CAS ADS article Google Scholar 

Bussi, G., Donadio, D. and Parrinello, M. Standardize sampling through speed rescaling. J Chemical Physics. 126, 014101, 10.1063/1.2408420 (2007).

CAS ADS article PubMed PubMed Central Google Scholar 

Berendsen, HJC, Postma, JPM, Van Gunsteren, WF, DiNola, A. & Haak, JR Molecular dynamics coupled with an external bath. J Chemical Physics. 81, 3684–3690 (1984).

CAS ADS article Google Scholar 

Hess, B., Bekker, H., Berendsen, HJC & Fraaije, JGEM LINCS: Linear constraint solver for molecular simulation. J Computational Chemistry. 18, 1463–1472 (1997).

Pronk, S. etc. GROMACS 4.5: A high-throughput, highly parallel open source molecular simulation toolkit. Bioinformatics (Oxford, England). 29, 845–854 (2013).

Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual Molecular Dynamics. J Moore diagram. 14, 33–38 27-28 (1996).

Baker, NA, Sept, D., Joseph, S., Holst, MJ & McCammon, JA Electrostatics of nanosystems: applications in microtubules and ribosomes. Proc Natl Acad Sci USA. 98, 10037–10041 (2001).

CAS ADS article Google Scholar 

Dawson, PE, Muir, TW, Clark-Lewis, I. & Kent, SB synthesize proteins through natural chemical linkage. science. 266, 776–779 (1994).

CAS ADS article Google Scholar 

Tam, JP & Lu, YA Biomimetic strategies for the synthesis and fragmentation of cyclic proteins. Protein science. 7, 1583–1592 (1998).

Franken, WF and others. CCPN data model of nuclear magnetic resonance spectroscopy: development of software pipeline. protein. 59, 687–696 (2005).

Guntert, P., Mumenthaler, C. and Wuthrich, K. used the new program DYANA to perform torsion angle dynamics for NMR structural calculations. J Molecular Biology. 273, 283–298 (1997).

Brünger, AT, etc. Crystallography and Nuclear Magnetic Resonance System: A new software suite for structure determination of macromolecules. Acta Crystallogr D Biol Crystallogr. 54, 905–921 (1998).

Troeira Henriques, S., Huang, YH, Chaousis, S., Wang, CK & Craik, DJ The anticancer and toxicity properties of cyclic peptides depend on phosphatidylethanolamine phospholipid targeting. Studentization. 15, 1956–1965 (2014).

This work was supported by a grant (APP1060225) from the National Health and Medical Research Council (NHMRC). DJC is an NHMRC professor-level researcher (APP1026501), STH received the support of the Discovery Early Career Research Award (DE120103152), and CKW received the support of the NHMRC Early Career Scholarship (536578). We thank you for using the facilities of the Queensland Nuclear Magnetic Resonance Network. We thank Stijn M. Agten, Phillip Walsh, Philip Sunderland and Olivier Cheneval for their assistance in peptide synthesis and Stephanie Chaousis for technical assistance in internalization research.

Current address: AITHM Biological Discovery and Therapeutic Molecular Development Center, James Cook University, Queensland, Australia, 4870

Institute of Molecular Biological Sciences, University of Queensland, Brisbane, 4072, Queensland, Australia

Yen-Hua Huang, Sónia T. Henriques, Conan K. Wang, Louise Thorstholm, Norelle L. Daly, Quentin Kaas, and David J. Craik

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

YHH designed and executed the main experiment. STH, CKW and QK designed and conducted additional experiments. LT and NLD contributed to the design of the peptides used in this study. YHH, STH, CKW and QK wrote and edited the manuscript. DJC initiated and supervised the research and edited the manuscript. All authors reviewed the manuscript.

The author declares that there are no competing economic interests.

This work has been licensed under the Creative Commons Attribution 4.0 International License Agreement. The images or other third-party materials in this article are included in the Creative Commons license of the article, unless otherwise stated in the credit line; if the material is not included under the Creative Commons license, the user will need permission from the license holder to copy The material. To view a copy of this license, please visit http://creativecommons.org/licenses/by/4.0/

Huang, YH., Henriques, S., Wang, C. etc. Design of substrate-based BCR-ABL kinase inhibitors using cyclic peptide scaffolds. Scientific Report 5, 12974 (2015). https://doi.org/10.1038/srep12974

DOI: https://doi.org/10.1038/srep12974

Anyone you share the following link with can read this content:

Sorry, there is currently no shareable link in this article.

Provided by Springer Nature SharedIt content sharing program

International Journal of Peptide Research and Therapy (2019)

By submitting a comment, you agree to abide by our terms and community guidelines. If you find content that is abusive or does not comply with our terms or guidelines, please mark it as inappropriate.

Scientific Report (Sci Rep) ISSN 2045-2322 (online)