Chinese Chemical Letters  2020, Vol. 31 Issue (2): 418-422   PDF    
Discovery of [1, 2, 4]triazolo[1, 5-a]pyrimidine derivatives as new bromodomain-containing protein 4 (BRD4) inhibitors
Shuai Wang1, Dandan Shen1, Lijie Zhao1, Xiaohan Yuan, Jialing Cheng, Bin Yu*, Yichao Zheng*, Hongmin Liu*     
Key Laboratory of Technology of Drug Preparation, Ministry of Education of China & School of Pharmaceutical Sciences, Zhengzhou University, Zhengzhou 450001, China
Abstract: Targeting bromodomain-containing protein 4 (BRD4) has been proved to be an effective strategy for cancer therapy. To date, numerous BRD4 inhibitors and degraders have been identified, some of which have advanced into clinical trials. In this work, a focused library of new [1, 2, 4]triazolo[1, 5-a]pyrimidine derivatives were discovered to be able to inhibit BRD4. WS-722 inactivated BRD4 (BD1/BD2), BRD2 (BD1/BD2) and BRD3 (BD1/BD2) broadly with the IC50 values less than 5 μmol/L. Besides, WS-722 inhibited growth of THP-1 cells with an IC50 value of 3.86 μmol/L. Like (+)-JQ1, WS-722 inhibited BRD4 in a reversible manner and enhanced protein stability. Docking studies showed that WS-722 occupied the central acetyl-lysine (Kac) binding cavity and formed a hydrogen bond with Asn140. In THP-1 cells, WS-722 showed target engagement to BRD4. Cellular effects of WS-722 on THP-1 cells were also examined, showing that WS-722 could block c-MYC expression, induce G0/G1 phase arrest and p21 up-regulation, and promote differentiation of THP-1 cells. BRD4 inhibition by WS-722 resulted in cell apoptosis and upregulated expression of cleaved caspased-3/7 and PARP in THP-1 cell lines. The [1, 2, 4]triazolo[1, 5-a] pyrimidine is a new template for the development of new BRD4 inhibitors.
Keywords: Epigenetic regulation    BRD4 inhibitor    [1, 2, 4]triazolo[1, 5-a]pyrimidine    THP-1 cells    AML treatment    

As histone acetylation can be written by histone acetyltransferases (HATs) and erased by histone deacetylases (HDACs), the reading process of histone acetylation marks is much less studied until the discovery of potent and selective inhibitors of bromo and extra-terminal (BET) domain proteins [1, 2]. As a unique protein domain, BET domain, a highly conserved domain composed of about 110 amino acids, can recognize the histone acetylation modification and mediate interaction of proteins. Proteins with BET domain can be divided into nine protein families due to the sequence similarity, of which BRD2, BRD3, BRD4 and BRDT are the largest and most studied. All members of BET family have two BRD active domains BD1 and BD2 and an extra-terminal (ET) region, which bind to acetylated lysine located in H3 and H4 as well as some other non-histone proteins [3, 4]. Among the BET family members, BRD4, abnormally expressed in many tumor cancer cells, has been reported to be associated with various diseases [5-10], which makes it an important epigenetic target for disease treatment. To date, a large number of small-molecule inhibitors and degraders targeting BET proteins have been reported [11-19], some of which such as (+)-JQ1, OTX-015, TEN-010, I-BET762, CPI-0610, I-BET151, PLX51107, ABBV-075, AZD5153, BMS-986158 and INCB0543294 are currently undergoing clinical assessment at different phases for cancer therapy [20, 21].

Our previous work has showed that the bicyclic triazole fused pyrimidines have possessed interesting biological activities [22-33]. Following our previous work, herein a focused library of [1, 2, 4]triazolo[1, 5-a]pyrimidine-based BRD4 inhibitors were synthesized and evaluated for their inhibitory activity against BRD4. Among these compounds, WS-722 effectively and reversibly inhibited BRD4 (BD1) and BRD4 (BD2) with the IC50 values of 2.15 and 4.36 μmol/L, respectively. Cellular target engagement in THP-1 cells was confirmed by the thermal shift assay. Cellular effects were also examined in THP-1 cells, indicating that WS-722 induced differentiation of THP-1 cells, arrested cell cycle at G0/G1 phase and resulted in obvious cell apoptosis. The results suggest that the [1, 2, 4]triazolo[1, 5-a]pyrimidine is a new starting point for the development of BRD4 inhibitors.

The general synthetic route of compounds D1-D29 is shown in Scheme S1 (Supporting information). The synthesized compounds are shown in Fig. 1 with the core scaffold highlighted in red. Clearly, modifications focused on the variations of substituents, generating a focused library of [1, 2, 4]triazolo[1, 5-a]pyrimidine derivatives.

Download:
Fig. 1. Modifications based on the score scaffold for SARs studies.

Based on the established TR-FRET protocol (Fig. S1 in Supporting information), we first screened the synthesized compounds for their inhibitory activity against BRD4 (BD1) using (+)-JQ1 as a positive control and the results are shown in Fig. S2 (Supporting information). The compounds showed varied inhibitory activity against BRD4, indicating the importance of substituents for the activity. Particularly, nine of these compounds (Fig. S2 in Supporting information) exhibited acceptable inhibitory activity against BRD4 (BD1) at 10 μmol/L with the inhibitory rates over 39%. In view of their biochemical potency against BRD4 (BD1), these nine compounds were then further evaluated against BRD4 (BD2), the IC50 values of these compounds were shown in Table 1. Compound D7 (also named as WS-722) displayed the best potency against BRD4 (BD1/BD2) with the IC50 values of 2.15 and 4.36 μmol/L, respectively. Compound D13 bearing the p-bromobenzyl group had comparable inhibitory activity with D7 against BRD4. In contrast, compounds D1, D12, D14, D15, D18 and D19 showed decreased inhibitory activity against BRD4. Compared to D12, introduction of the R group to the core scaffold did not significantly improve the inhibitory activity. D24 inhibited BRD4 (BD1) and BRD4 (BD2) moderately with the IC50 values of 11.21 and 8.64 μmol/L, respectively.

Table 1
Inhibitory activity of selected compounds against BRD4.

In view of the acceptable potency of WS-722 toward BRD4 (Fig. 2A), we further evaluated its selectivity against BRD2 and BRD3. As shown in Figs. 2B and C, WS-722 also moderately inhibited BRD2 and BRD3, indicating that WS-722 is a pan-BRD inhibitor in vitro. A dilution assay was then used to test the reversibility using (+)-JQ1 as the control compound. We found that 100-fold dilution of the BRD4/WS-722 mixture resulted in the recovery of BRD4 activity (Figs. 2D and E). Similarly, the BRD4 activity of (+)-JQ1 can also be recovered after dilution (Figs. 2D and E). These results indicate that like (+)-JQ1, WS-722 reversibly bound to BRD4 (BD1) and BRD4 (BD2) (Figs. 2D and E). The melting temperature (Tm) is defined as the temperature at which half of the protein is native and the other half denatured, and could be used to describe the stability of a protein. The positive changes in Tm upon ligand binding indicate thermal stabilization. To further evaluate the binding property of WS-722 to BRD4, the protein thermal shift assay [34] was used to detect the △Tm between the Tm (BRD4 + 25 μmol/L WS-722) value and the reference Tm (BRD4 + DMSO) using the (+)-JQ1 as the control compound. As indicated in Fig. 2F, for BRD4 (BD1) and BRD4 (BD2), the △Tm increased upon WS-722 and (+)-JQ1 treatment. The results suggest that (+)-JQ1 and WS-722 could enhance BRD4 protein stability.

Download:
Fig. 2. Biochemical characterization of WS-722. (A, B, C) Inhibitory activity of WS-722 against BRD2/3/4 (BD1) and BRD2/3/4 (BD2); (D) Dilution assay of WS-722 to BRD4 (BD1) with TR-FRET assay, (+)-JQ1 was used as the positive control; (E) Dilution assay of WS-722 to BRD4 (BD2) with TR-FRET assay, (+)-JQ1 was used as the positive control; (F) BRD4 (BD1) and BRD4 (BD2) recombinant protein thermal shift assays were performed when treated with WS-722 and (+)-JQ1. d(Fluorescence)/dT and △Tm was analyzed. Data are the mean ± SD. Each experiment was repeated three times. *** P < 0.001.

Recent studies have shown that BRD4 has played key roles in the maintenance of aberrant chromatin states in AML, acute lymphoblastic leukemia (ALL), myeloma and lymphoma, and treatment with BRD4 inhibitors could recapitulate anti-leukemic effects in several AML cell lines [35-37]. Initially, the antiproliferative activity of WS-722 was evaluated against THP-1 cells. As shown in Fig. 3A, after treatment for 7 days, WS-722 moderately inhibited growth of THP-1 cells with an IC50 value of 3.86 μmol/L. To confirm whether WS-722 could abrogate BRD4 activity in acute leukemia cell lines, we used the cellular thermal shift assay to study thermal stability of BRD4 upon WS-722 treatment in THP-1 cell line. THP-1 cells were treated with WS-722 and then heated for 3 min at 50 ℃. After freezing in liquid nitrogen and thawing on ice, equal amounts of supernatant were removed and blotted with the BRD4 antibody. Our results suggested that WS-722 stabilized BRD4 in a concentration-dependent manner, suggesting cellular target engagement of WS-722 to BRD4 in THP-1 cells (Fig. 3B). It has been reported that BET-bromodomain inhibition potently suppresses MYC gene expression in leukemia by blocking its transcription [38]. Hence, to further explore the biological role of our BRD4 inhibitor WS-722 in THP-1 cells, THP-1 cells were treated with WS-722 and then subjected to western blotting analysis. As shown in Fig. 3C, treatment with WS-722 dose-dependently decreased expression of c-MYC, without impact on BRD4 expression (Fig. 3C). To examine higher-order influences over biological networks regulated by c-MYC, which can influence cell cycle and p21 (also known as cyclin-dependent kinase inhibitor 1 or CDK-interacting protein 1) expression [39]. Expression of cell cycle regulatory protein p21 was also examined after treatment with WS-722 for 48 h. As shown in Fig. 3D, WS-722 treatment concentration-dependently induced accumulation of p21 in THP-1 cells. p21 is an universal cell-cycle inhibitor directly controlled by p53 and p53-independent pathways [40, 41]. Then, the cell cycle of THP-1 cells was analyzed with flow cytometry after treatment with WS-722. As shown in Fig. 3E, WS-722 arrested cell cycle of THP-1 cells at G0/G1 phase in a concentration-dependent manner, which was consistent with previous reports [42-44]. All these results confirmed that pharmacological inhibition of BRD4 could block c-MYC expression and then induce G0/G1 phase arrest and p21 up-regulation.

Download:
Fig. 3. WS-722 treatment decreases c-MYC and induces cell cycle in THP-1 cell line. (A) The effect of compound WS-722 on the viability of THP-1 cells for 7 days; (B) Enhancement of thermal stability of BRD4 in THP-1 cells; (C) BRD4 and c-MYC expression in THP-1 cells when treated with WS-722 for 24 h. GAPDH was used as the loading control; (D) Expression of p21 when THP-1 cells were treated with WS-722 for 48 h. GAPDH was used as the loading control; (E) Effect of WS-722 on the cell cycle of THP-1 cells. * P < 0.05; ** P < 0.01; *** P < 0.001.

To study the effect of WS-722 on cell apoptosis, THP-1 cells were treated with WS-722 at doses from 0.3 μmol/L to 10 μmol/L for 48 h. We found WS-722 induced apoptosis at high concentration (3.0 μmol/L) as detected by annexin V staining and PI uptake (Fig. 4A). Then samples were taken for the induction of apoptosis by measuring caspase-3, caspase-7 and PARP activity. As shown in Fig. 4B, after treatment of THP-1 cells with WS-722 for 48 h, cleaved caspase 3, caspase 7 and PARP increased in a concentration-dependent manner. It has been reported that MYC (a family of regulator genes and proto-oncogenes that code for transcription factors) arrests the differentiation of embryonic stem cells and various neoplasms [45]. So, we assessed whether WS-722 could promote differentiation of THP-1 cells by quantifying the expression of two differentiation related markers: CD14 and CD11b. As shown in Figs. 4C and D, WS-722 induced expression of CD11b and CD14 in THP-1 cells, which promoted differentiation of leukemia cells dose-dependently. To conclude, these results suggest that WS-722 down-regulated expression of c-MYC, induced cell cycle arrest, and promoted differentiation of THP-1 cells.

Download:
Fig. 4. Effect of WS-722 on cell apoptosis and differentiation in THP-1 cell line. (A) THP-1 cell apoptosis when treated with WS-722 at doses from 0.3 μmol/L to 3.0 μmol/L for 48 h. (B) Cleaved caspase-3, caspase-7 and PARP expression when treated with WS-722 for 48 h. (C and D) Flow cytometry analysis of CD11b (C) and CD14 (D) in THP-1 cells treated with WS-722. Data are the mean ± SD. Each experiment was repeated three times. * P < 0.05; ** P < 0.01; *** P < 0.001.

(+)-JQ1 features a novel thieno-triazolo-1, 4-diazepine scaffold and is a well characterized cell permeable BET family inhibitor. (+)-JQ1 strongly blocks binding of a tetra-acetylated histone H4 peptide to BRD4 and potently inhibits the first and second bromodomains (BD1 and BD2) of BRD4 with IC50 values of 77 and 33 nmol/L, respectively [46]. To explain the observed potency of compound WS-722 against BRD4, an in silico docking simulation was carried out using the MOE 2015.10 software package. The crystal structure of the first bromodomain (BD1) of human BRD4 in complex with (+)-JQ1 (PDB code: 3MXF) was therefore used as a docking receptor for computational studies. Similar to interactions observed in acetyl-lysine (Kac) complexes [47], (+)-JQ1 (Fig. S3A in Supporting information) was fitted well into the hydrophobic Kac binding site with an extraordinary shape complementarity [46]. As revealed by the co-crystal structure of BRD4 (BD1)/(+)-JQ1 complex (Fig. S3C in Supporting information), the triazole ring of (+)-JQ1 formed a hydrogen bond with surrounding residue Asn140 in BRD4 (BD1), and the conserved BET residues stabilized the binding of (+)-JQ1 through hydrophobic interactions. As depicted in Figs. S3A and S3B (Supporting information), like (+)-JQ1, WS-722 occupied the central acetyl-lysine (Kac) cavity and formed a hydrogen bond with the key residue Asn140 as well (Fig. S3D in Supporting information). The binding models may be responsible for the observed potency of WS-722 against BRD4 (BD1) (IC50 = 2.15 μmol/L). We also found that WS-722 did not fit into the regions occupied by the bulky t-butyl ester group and p-chlorophenyl group (Fig. S3A), the bulky t-butyl ester group is predicted to mitigate binding to the central benzodiazepine receptor [48]. Additionally, the hydrophobic propargyl group was directed to a hydrophilic region (Fig. 3SB), which is not occupied by (+)-JQ1. These observations may explain the weaker potency of WS-722 against BRD4 (BD1) than (+)-JQ1. As demonstrated in Fig. S3B, the phenyl ring in WS-722 was oriented to the solvent region, suggesting that further structural modifications at this site may be allowed for improved potency and/or PD/PK properties. We believe that the structural basis may facilitate further structure-based drug design (SBDD) for more potent [1, 2, 4]triazolo[1, 5-a]pyrimidine-based BRD4 inhibitors.

In summary, a focused library of new [1, 2, 4]triazolo[1, 5-a]-pyrimidine derivatives were synthesized and evaluated for their inhibitory activity against BRD4. The shortlisted compound WS-722 broadly inactivated BRD4 (BD1/BD2), BRD2 (BD1/BD2) and BRD3 (BD1/BD2) with the IC50 values less than 5 μmol/L. Besides, WS-722 inhibited growth of THP-1 cells with an IC50 value of 3.86 μmol/L. Like (+)-JQ1, WS-722 inhibited BRD4 in a reversible manner and enhanced protein stability as indicated in the protein thermal shift assay. Docking studies showed that WS-722 had a similar binding model with (+)-JQ1, occupying the central acetyllysine (Kac) binding cavity and forming a hydrogen bond with the key residue Asn140. In THP-1 cells, WS-722 showed cellular target engagement to BRD4. Cellular effects of WS-722 in THP-1 cells were also examined, showing that WS-722 could block c-MYC expression, induce G0/G1 phase arrest and p21 up-regulation, and promote differentiation of leukemia cells. What is more, inhibition of BRD4 by WS-722 in THP-1 cell lines resulted in obvious cell apoptosis and up-regulated the expression levels of cleaved caspased-3/7 and PARP. The [1, 2, 4]triazole[1, 5-a]pyrimidine may be served as a new scaffold for the development of novel BRD4 inhibitors.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Nos. 81703326, 81773562, 81602961 and 81430085), Scientific Program of Henan Province (No. 182102310123), China Postdoctoral Science Foundation (Nos. 2018M630840 and 2019T120641).

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi: https://doi.org/10.1016/j.cclet.2019.08.029.

References
[1]
P. Filippakopoulos, J. Qi, S. Picaud, et al., Nature 468 (2010) 1067-1073. DOI:10.1038/nature09504
[2]
E. Nicodeme, K.L. Jeffrey, U. Schaefer, et al., Nature 468 (2010) 1119-1123. DOI:10.1038/nature09589
[3]
P. Filippakopoulos, S. Knapp, Nat. Rev. Drug Discov. 13 (2014) 337-356. DOI:10.1038/nrd4286
[4]
J. Shi, C.R. Vakoc, Mol. Cell 54 (2014) 728-736.
[5]
A. Wyce, Y. Degenhardt, Y. Bai, et al., Oncotarget 4 (2013) 2419-2429.
[6]
V. Sahai, K. Kumar, L.M. Knab, et al., Mol. Cancer Ther. 13 (2014) 1907-1917. DOI:10.1158/1535-7163.MCT-13-0925
[7]
C. Pastori, M. Daniel, C. Penas, et al., Epigenetics 9 (2014) 611-620. DOI:10.4161/epi.27906
[8]
D. Da Costa, A. Agathanggelou, T. Perry, et al., Blood Cancer J. 3 (2013) e126. DOI:10.1038/bcj.2013.24
[9]
Z. Cheng, Y. Gong, Y. Ma, et al., Clin. Cancer Res. 19 (2013) 1748-1759. DOI:10.1158/1078-0432.CCR-12-3066
[10]
M. Pervaiz, P. Mishra, S. Gunther, Chem. Rec. 18 (2018) 1808-1817. DOI:10.1002/tcr.201800074
[11]
Y. Duan, Y. Guan, W. Qin, et al., MedChemComm 9 (2018) 1779-1802. DOI:10.1039/C8MD00198G
[12]
Z. Liu, B. Tian, H. Chen, et al., Eur. J. Med. Chem. 151 (2018) 450-461. DOI:10.1016/j.ejmech.2018.04.006
[13]
M. Pérez-Salvia, M. Esteller, Epigenetics 12 (2016) 323-339. DOI:10.1080/15592294.2016.1265710
[14]
J. Lu, Y. Qian, K. Raina, et al., Blood 126 (2015) 755-763. DOI:10.1182/blood.V126.23.755.755
[15]
G.M. Matthews, S. Gandolfi, J. Bruggentheis, et al., Blood 128 (2016) 1062. DOI:10.1182/blood.V128.22.1062.1062
[16]
G.E. Winter, D.L. Buckley, J. Paulk, et al., Science 348 (2015) 1376-1381. DOI:10.1126/science.aab1433
[17]
S. Crunkhorn, Nat. Rev. Drug Discov. 14 (2015) 459. DOI:10.1038/nrd4670
[18]
L. Ouyang, L. Zhang, J. Liu, et al., J. Med. Chem. 60 (2017) 9990-10012. DOI:10.1021/acs.jmedchem.7b00275
[19]
Z. Liu, P. Wang, H. Chen, et al., J. Med. Chem. 60 (2017) 4533-4558. DOI:10.1021/acs.jmedchem.6b01761
[20]
G. Andrieu, A.C. Belkina, G.V. Denis, Drug Discov. Today 19 (2016) 45-50. DOI:10.1016/j.ddtec.2016.06.004
[21]
R. Vazquez, M.E. Riveiro, L. Astorgues-Xerri, et al., Oncotarget 8 (2017) 7598-7613. DOI:10.18632/oncotarget.13814
[22]
L. Chang, M. Xiao, L. Yang, et al., Bioorg. Med. Chem. 26 (2018) 5006-5017. DOI:10.1016/j.bmc.2018.08.021
[23]
Z.H. Li, X.Q. Liu, T.Q. Zhao, et al., Eur. J. Med. Chem. 139 (2017) 741-749. DOI:10.1016/j.ejmech.2017.08.042
[24]
Z.H. Li, X.Q. Liu, T.Q. Zhao, et al., Bioorg. Med. Chem. Lett. 27 (2017) 4377-4382. DOI:10.1016/j.bmcl.2017.08.021
[25]
Z.H. Li, X.Q. Liu, P.F. Geng, et al., Eur. J. Med. Chem. 138 (2017) 1034-1041. DOI:10.1021/acsmedchemlett.6b00423
[26]
Z.H. Li, X.Q. Liu, P.F. Geng, et al., ACS Med. Chem. Lett. 8 (2017) 384-389. DOI:10.1021/acsmedchemlett.6b00423
[27]
B. Wang, B. Zhao, L.P. Pang, et al., Pharmacol. Res. 122 (2017) 66-77. DOI:10.1016/j.phrs.2017.05.025
[28]
B. Yu, X.J. Shi, Y.F. Zheng, et al., Eur. J. Med. Chem. 69 (2013) 323-330. DOI:10.1016/j.ejmech.2013.08.029
[29]
L.H. Huang, Y.F. Zheng, Y.Z. Lu, et al., Steroids 77 (2012) 710-715. DOI:10.1016/j.steroids.2012.03.002
[30]
S. Yuan, B. Yu, H.M. Liu, Adv. Synth. Catal. 361 (2019) 59-66. DOI:10.1002/adsc.201801226
[31]
S. Wang, L. Zhao, X.J. Shi, et al., J. Med. Chem. 62 (2019) 2772-2797. DOI:10.1021/acs.jmedchem.9b00113
[32]
Z. Li, L. Ding, Z. Li, et al., Acta Pharm. Sin. B 9 (2019) 794-808.
[33]
S. Yuan, S. Wang, M. Zhao, et al., Chin. Chem. Lett. (2019), doi: http://dx.doi.org/10.1016/j.cclet.2019.07.019.
[34]
K. Huynh, C.L. Partch, Curr. Protoc. Protein Sci 79 (2015) 28.9.1-28.9.14.
[35]
J. Zuber, J. Shi, E. Wang, et al., Blood 120 (2012) 2843-2852.
[36]
J. Zuber, J. Shi, E. Wang, et al., Nature 478 (2011) 524-528. DOI:10.1101/gad.17269211
[37]
J.E. Delmore, G.C. Issa, M.E. Lemieux, et al., Cell 146 (2011) 904-917. DOI:10.1016/j.cell.2011.08.017
[38]
J.A. Mertz, A.R. Conery, B.M. Bryant, et al., Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 16669-16674. DOI:10.1073/pnas.1108190108
[39]
Y.J. Jeong, H.S. Hoe, J. Cell. Biochem. 119 (2018) 2036-2047. DOI:10.1002/jcb.26366
[40]
W.S. El-Deiry, Cancer Res. 76 (2016) 5189-5191. DOI:10.1158/0008-5472.CAN-16-2055
[41]
A.L. Gartel, Mol. Cancer Ther. 5 (2006) 1385-1386. DOI:10.1158/1535-7163.MCT-06-0163
[42]
J.H. Jeong, S.S. Kang, K.K. Park, et al., Mol. Cancer Ther. 9 (2010) 2102-2113. DOI:10.1158/1535-7163.MCT-09-1159
[43]
D.M. Lepley, J.C. Pelling, Mol. Carcinog. 19 (1988) 74-82. DOI:10.1002/(sici)1098-2744(199707)19:2<74::aid-mc2>3.0.co;2-l
[44]
T. Gulappa, R.S. Reddy, S. Suman, A.M. Nyakeriga, C. Damodaran, Cancer Lett. 337 (2013) 177-183. DOI:10.1016/j.canlet.2013.05.014
[45]
A. Cieslar-Pobuda, V. Knoflach, M.V. Ringh, et al., Biochim. Biophys. Acta-Mol. Cell Res. 1864 (2017) 1359-1369. DOI:10.1016/j.bbamcr.2017.04.017
[46]
P. Filippakopoulos, J. Qi, S. Picaud, et al., Nature 468 (2010) 1067-1073. DOI:10.1038/nature09504
[47]
F. Vollmuth, W. Blankenfeldt, M. Geyer, J. Biol.Chem. 284 (2009) 36547-36556. DOI:10.1074/jbc.M109.033712
[48]
P.F. vonVoigtlander, R.N. Straw, Drug Develop. Res. 6 (1985) 1-12. DOI:10.1002/ddr.430060102