Page 227 - Read Online

Zhang et al. J Cancer Metastasis Treat 2020;6:21 I http://dx.doi.org/10.20517/2394-4722.2020.40 Page 5 of 11

patterns correlated with improved prognosis. Our approach in this study was enabled by the availability of
high-quality, publicly available tumor gene expression datasets from large cohorts of myeloma patients that
included either extensive survival annotation or comparisons of healthy vs. tumor tissues (GSE4581 and
GSE6477) [21-25] . The schema depicted below illustrates the approach that we employed to understand the
[19]
mechanism of drug synergy between the HDAC inhibitor, entinostat and an Mtor inhibitor, everolimus .

The upstream predictors identified as “activated” from the drug combination by ingenuity pathway analysis
(IPA) from the 37 patient-survival associated genes included Cdkn2a (p16/p19), p53, and Rb. MYC, E2F,
and TBX2 were predicted as “inhibited” by the combination. The combination worked cooperatively to
[19]
lower MYC protein stability, partially through FBXW7-mediated degradation . The combination also
[20]
worked to increase the activity of the Rb1/Cdkn2a tumor suppressor pathways . The drug combination
enhanced the overall survival rate of tumor-bearing BALB-bclxl transgenic mice and lowered MYC
protein levels in tumors of these immunocompetent mice. Our studies in the NCI-60 cell line panel found
that most tumors, regardless of their tissue of origin, responded synergistically to the mTORi/HDACi
combination. In early molecular classification schemes of multiple myeloma patients based on heirachical
clustering of gene expression in myeloma samples, seven clusters were identifed as proliferation (PR),
low bone, multiple myeloma SET domain (MMSET), hyperdiploid, cyclin D1 (CD-1), cyclin D2 (CD-2)
and avian musculoaponeurotic fibrosarcoma (MAF) . Using gene expression data from samples within
[21]
these same subgroups, we determined a gene score for our 37 drug-responsive genes to predict how many
patients would be expected to benefit from combination treatment. Roughly 50% in most subgroups were
predicted to benefit; there were two exceptions: all patients in the PR group and only 17% in the CD-2
group were predicted to benefit from the drug combination based on their expression scores for the 37
[19]
gene signature . This is of course, hypothetical and would need to be tested in a clinical trial. In addition,
the drug combination did not have a direct effect on gene expression of genes involved in determining
[21]
Zhan et al. ’s proliferation index. Cells with mutations in MYC residues required for its degradation did
[19]
not respond to the drug combination .
TARGETING MYC TRANSCRIPTION AND DEGRADATION
Our systems analyses led us to explore a more direct approach to targeting MYC. We screened a small
molecule microarray library for binders of the G-quadruplex located in the NHEIII region of the MYC
promoter, and identified a benzofuran-containing molecule, D089, that could stabilize the G4 structure
[26]
and inhibit MYC transcription . We demonstrated that D089 inhibited MYC with greater affinity than
other G4-containing genes (e.g., RAS, VEGF, BCL2, and Rb1). The small molecule was relatively potent in
inhibiting multiple myeloma cell proliferation but was ineffective in tumor cell lines that had deleted the
portion of the MYC promoter containing the G4 sequence . In subsequent studies, we analyzed a series
[26]
of analogs to find one, DC-34, that was more potent in its activity against myeloma cells. In these studies,
we were able to define an nuclear magnetic resonance structure of the small molecule bound to the G4
[27]
structure which should allow structure-guided design of even more potent compounds . The discovery
of a drug that directly targets MYC has been elusive, and thus far there are no approved drugs for this
indication. The development of a more direct approach for inhibiting MYC activity seems warranted given
[28]
the overall importance of MYC to a wide range of tumor types .
Our early work involving retroviral induction studies of mouse PCTs suggest that inhibition of MYC alone
[29]
may not be curative . Early induction studies with retroviral vectors clearly showed that overexpression
of MYC alone could not induce PCTs; but when MYC was paired with RAS or RAF, high incidences of
[29]
plasma cell tumors could be induced even in PCT-resistant strains of mice . These data are in agreement
with recent studies indicating that MYC mutations are acquired secondary genetic events in myeloma
[30]
progression . A key aspect of the mTORi/HDACi combination is its ability to decrease MYC protein
stability; however, in some myeloma cell lines, we have observed a “compensatory” MYC mRNA increase