Unique KOPN-8 Resistance Mechanisms

Characterization of acquired EPZ-5676 resistance in cell line models of MLL rearranged leukemia Scott R. Daigle†, Carly T. Campbell†, Nigel J. Waters, Edward J. Olhava, Robert A. Copeland, Stephen J. Blakemore, Roy M. Pollock, Jesse J. Smith Epizyme Inc., 400 Technology Square, Cambridge MA 02139, USA. † presenting author

DOT1L inhibitor EPZ-5676 is currently under Phase 1 clinical trial investigation in relapsed/refractory patients with acute leukemia, including those with an MLL-rearrangement (MLL-r). Early clinical results, including complete remissions, support ongoing clinical development and preclinical investigation into mechanisms precipitating EPZ-5676 treatment induced resistance. MLL-r cell lines KOPN-8 (MLL-ENL) and NOMO-1 (MLL-AF9) were exposed to an EPZ-5676 concentration above the pre-determined 14 day proliferation assay IC90. Initial treatment of the cell lines led to the expected inhibition of H3K79 dimethylation (H3K79me2) and MLL-r target genes HOXA9 and MEIS1 as outlined in previous work (1). Resistance to EPZ-5676 in both cell lines emerged following three weeks of continued treatment with EPZ-5676 and was defined by increased growth rates in the presence of inhibitor. Mechanisms of resistance for both cell lines were investigated using RNA-seq and ChIP-seq on parental and resistant cell line pools. Our analysis identified common characteristics between the resistant cell lines, but mechanisms by which they became resistant differed. Global H3K79me2 reduction was maintained in both refractory cell lines, yet ChIP-seq analysis of resistant pools identified specific loci with H3K79me2 recovery in KOPN-8 cells. In resistant KOPN-8 cells recovery of H3K79me2 was concentrated at the HOXA locus and other MLL-r target genes (e.g. MEIS1 and RUNX2), with the remainder of actively transcribed genes maintaining H3K79me2 inhibition at levels observed in parental cells. In contrast, resistant NOMO-1 cells did not recover H3K79me2 at any actively transcribed genes, including those of the MLL-r signature. Only resistant KOPN-8 cells regained expression of the MLL-r target genes HOXA9 and MEIS1. Of note both resistant NOMO-1 and KOPN-8 cell lines had 3 and 100 fold upregulation of the ABCB1 (MDR1, P-gp) transporter, respectively, when compared to a matched control cell line. To explore the role of drug efflux transporter ABCB1 on resistance, we treated cells with Valspodar, a known inhibitor of ABCB1. Following treatment with 1 µM Valspodar, only KOPN-8 cells showed decreased cell growth similar to the naïve control cell line, supporting the presence of an ABCB1-independent mechanism in NOMO-1 cells in addition to ABCB1 upregulation. Detailed gene expression and pathway analysis was completed using Selventa’s Discovery Platform and unique resistance mechanisms were identified in both cell models. In summary, we have identified two mechanisms of EPZ-5676 resistance in MLL-r cell lines, one mechanism dependent on and the other independent of ABCB1. Further refinement of these mechanisms will aid in providing hypotheses for testing mechanisms of EPZ-5676 treatment emergent resistance in patients and may support designing future rational clinical combinations.

Abstract Results

Conclusions

www.epizyme.com

# 2701

References [1] Daigle, S.R. et al . (2011) Selective Killing of Mixed Lineage Leukemia Cells by a Potent Small-Molecule DOT1L Inhibitor. Cancer Cell, 20, 53-65 [2] Daigle, S. R. et al. (2013) Potent inhibition of DOT1L as treatment of MLL-fusion leukemia. Blood, 122, 1017-1025. [3] Krivstov, A.V. and Armstrong S.A. (2007) MLL Translocations, histone modifications, and leukaemia stem-cell development. Nat Rev Cancer, 7, 823-833 [4] Krivstov, A.V. (2008) H3K79 methylation profiles define murine and human MLL-AF4 leukemias. Cancer Cell, 14(5), 355-368

• Continuous exposure of two MLL-r cell lines to EPZ-5676 lead to development of a resistant phenotype despite maintenance of global H3K79me2 inhibition

• The mechanism of altered cell growth potential induced by EPZ-5676 was distinct in each cell line:

-KOPN-8 demonstrated upregulation of EPZ-5676-efflux pump ABCB1 and selective recovery of H3K79 methylation at loci associated with MLL-r activation -NOMO-1 mechanism of resistance was independent of ABCB1 expression

• Pathway analysis of ChIP-seq and RNA-seq data revealed multiple additional potential mechanisms of resistance unique and common to both cell lines.

Disclosures: All Authors: Epizyme Employment, Equity Ownership

Figure 2: Growth of KOPN-8 and NOMO-1 cells during development of resistance to 4.5 µM EPZ-5676. As indicated by arrows, after approximately 21 days of dosing EPZ-5676 treated cells regained growth rates similar to vehicle control. Viable cells were counted and split every 3 to 4 days in the presence of EPZ-5676 or DMSO vehicle control and split adjusted results plotted on a logarithmic scale.

Figure 3: Global and loci specific H3K79me2 analysis in KOPN-8 and NOMO-1 cells following treatment with 4.5 µM EPZ-5676 at indicated time points. (A) ELISA analysis of global H3K79me2 levels, plotted as a percentage of vehicle control. (B) Assessment of H3K79me2 at HOXA9 and MEIS1 gene locus using ChIP-seq. The difference between the sensitive and resistant cell lines is shown as the FDR corrected p-value (q-value). Significant recovery (q ≤ 0.05) of H3K79me2 was observed in both HOXA9 and MEIS1 gene locus as the KOPN-8 cells developed resistance to EPZ-5676. No significant recovery of H3K79me2 was observed in NOMO-1 cells at any loci.

A

B

Figure 2: MLL-r Cell Lines Develop Resistance to EPZ-5676 Following Extended Dosing

Figure 3: H3K79me2 Remains Globally Inhibited in Resistant Cells: Methylation Returns at MLL-r Target Genes Loci in KOPN-8 Only

KOPN-8 (MLL-ENL) NOMO-1 (MLL-AF9)

Figure 4: Gene Expression Changes Associated with EPZ-5676 Resistance

Acknowledgements

H3K79me2 ELISA (KOPN-8) H3K79me2 ELISA (NOMO-1)

Figure 4: RNA-seq analysis in KOPN-8 and NOMO-1 vehicle control, EPZ-5676 sensitive, and EPZ-5676 resistant cells. Relative mRNA expression levels are plotted as a percentage of those in vehicle-treated control cells. (A) HOXA9 and MEIS1 gene expression begins to recover in resistant KOPN-8 cells while no recovery of target gene expression is observed in resistant NOMO-1 cells. (B) ABCB1 expression is dramatically upregulated (100 fold) in resistant KOPN-8 cells, and moderately upregulated in resistant NOMO-1 cells (3 fold).

KOPN-8

NOMO-1

ABCB1 Expression

H3K79me2 ChIP-seq (KOPN-8)

MLL-r Target Gene Expression

Background

No MLL Fusion

Aberrant gene activation (e.g. HOXA9, MEIS1)

Normal gene expression

Aberrant H3K79 methylation

Normal H3K79 methylation

Leukemogenesis

Differentiation

Figure 1: Relevance of DOT1L Mediated Histone H3K79 Methylation in MLL-Rearranged Leukemia

Figure 6: Analysis of ChIP-seq and RNA-seq Data Based on Inferred Mechanisms

H3K79me2 ChIP-seq (NOMO-1)

Figure 6: Detailed gene expression and pathway analysis was performed on ChIP-seq and RNA-seq data, yielding potential resistance mechanisms unique to each cell line, as well as a shared set.

MLL Fusion

MLL Normal

HOXA

9

Sensitive – Day 14

Resistant – Day 53

Vehicle

2.1 fold decrease (FDR=0.2)

MEI

S1

Sensitive – Day 14

Resistant – Day 53

Vehicle

2 fold decrease (FDR=0.3)

50 reads

50 reads MEI

S1

Sensitive – Day 10

Resistant – Day 28

Vehicle

1.3 fold increase (FDR=1.6E-3)

HOXA

9

Sensitive – Day 10

Resistant – Day 28

Vehicle

3.5 fold increase (FDR=1.1E-5)

80 reads

80 reads

The authors would like to thank David Drubin, Ty Thompson, Mike Maria, Louis Latino and Renee Deehan Kenney of Selventa (www.selventa.com) for using their Discovery Platform approach on our ChIP-seq and RNA-seq data.

Figure 1: MLL fusion partner recruits DOT1L resulting in ectopic H3K79 methylation and leads to enhanced expression of leukemogenic genes, including HOXA9 and MEIS1 (3, 4).

A B

C Unique NOMO-1 Resistance Mechanisms Resistance Mechanisms Shared by

KOPN-8 and NOMO-1

A

B Courtesy, Dr. P. Ho and Dr. K. Maloney

47 genes reached significance (q ≤ 0.05) 0 genes reached significance (q ≤ 0.05)

100x 3x 3x 1.5x

Figure 5: ABCB1 Inhibitor Valspodar Resensitizes KOPN-8 to EPZ-5676 With No Effect in NOMO-1

KOPN-8 NOMO-1

Figure 5: Growth of KOPN-8 and NOMO-1 resistant and DMSO vehicle control cells when treated with 1 µM ABCB1 inhibitor Valspodar for 21 days. Resistant cells were cultured in media containing 4.5 µM EPZ-5676. Viable cells were counted and split every 3 to 4 days and split adjusted results plotted on a logarithmic scale.