Adaptive DNA amplification of synthetic gene circuit opens a way to overcome cancer chemoresistance

Abstract

Drug resistance continues to impede the success of cancer treatments, creating a need for experimental model systems that are broad, yet simple, to allow the identification of mechanisms and novel countermeasures applicable to many cancer types. To address these needs, we investigated a set of engineered mammalian cell lines with synthetic gene circuits integrated into their genome that evolved resistance to Puromycin. We identified DNA amplification as the mechanism underlying drug resistance in 4 out of 6 replicate populations. Triplex-forming oligonucleotide (TFO) treatment combined with Puromycin could efficiently suppress the growth of cell populations with DNA amplification. Similar observations in human cancer cell lines suggest that TFOs could be broadly applicable to mitigate drug resistance, one of the major difficulties in treating cancer.

Keywords: DNA amplification; RNA sequencing; drug resistance; evolution; synthetic biology.

Fig. 1.

Schematic illustration of how the cell populations in this study were derived. Schematic illustration of the mNF gene circuit regulating the Puromycin resistance gene PuroR and the fluorescent reporter eGFP that was chromosomally integrated into monoclonal CHO cells and then evolved to develop stable Puromycin resistance. Each of the six replicate populations evolved independently in the same constant Puromycin concentration of 35 μg/mL until it fully recovered to confluence. We cultured and passaged each replicate subsequently in the absence of Puromycin, with or without 0.05 ng/mL Dox, with samples frozen right after evolution, and then repeatedly during a “drug holiday”. Sanger sequencing revealed hTetR protein-coding mutations in replicate Evo2, enhancer mutations in Evo1 and Evo3, while Evo4, Evo5, and Evo6 were devoid of gene circuit mutations. Replicates Evo3 and Evo4 representing the last two scenarios were selected for mRNA sequencing, in addition to ancestral replicate Anc5 used for control purposes.

Fig. 2.

Transcriptomic changes after adaptation to Puromycin in replicates Evo3 and Evo4. (A and B) RNA sequencing shows more prominently elevated expression of RNAs mapping to gene circuit regions compared to native genes in Puromycin-resistant mNF populations Evo3 (A) and Evo4 (B) compared to the untreated control Anc5. (C) Normalized RNA sequencing coverage over the mNF gene circuit in the untreated control Anc5, as well as the Puromycin-resistant Evo3 and Evo4 replicate populations cultured with or without Dox. (D) Fold-changes of RNA expression levels for the native (circles) and circuit (diamonds) genes in the Puromycin-resistant Evo3 and Evo4 replicate populations versus the untreated control. Blue and black colors indicate significantly up-regulated and down-regulated circuit genes and native genes, respectively. (E) Overexpression of metallothionein 1 (Mt1, Left) and glutathione S-transferase Y1 (LOC100766772, Right) in the Puromycin-resistant Evo3 and Evo4 replicate populations cultured with or without Dox, compared to ancestral control Anc5. (F) Enriched KEGG pathways of the upregulated (Left) and downregulated (Right) genes. The sizes of the dots indicate the number of overlapping genes, while the depth of the colors indicates the ratio of the overlapping genes.

Fig. 3.

The four mutation-free evolved replicates have mNF gene circuit DNA amplification. Copy number analysis of PuroR (Top Right), hTetR (Top Left), eGFP (Bottom Right), and HygR (Bottom Left) genes shows consistently elevated DNA content in experimental replicates Evo3 through Evo6 compared to untreated mNF control population Anc5. The HygR gene amplification is not presented in Evo4, which is consistent with the RNA sequencing data. Evo2 lacks DNA amplification as expected due to the hTetR coding mutation. The survival mechanism of Evo1 that also lacks DNA amplification is unclear. The DNA copy number of gene circuit components is normalized to internal reference gene Vinculin, Mean ± SD of three independent repeats.

Fig. 4.

The hTetR protein remains active in the evolved replicates. (A) Average eGFP intensity per cell versus DNA copy number of the eGFP ORF in each evolved replicate population. High DNA copy number is associated with higher eGFP intensity, with a linear trend for both Dox-treated and untreated cells (Pearson's r of 0.85 and 0.81 respectively) after excluding the Evo2 population with inactive hTetR. Dashed lines indicate the linear fit to the Dox-treated and untreated data points. (B) Schematic illustration of the hTetR activity-sensing plasmid. mCherry is driven by an hTetR-dependent promoter (CMV-D2i) whereas TagBFP is constitutively expressed and used as a plasmid copy number control in each transfected cell. (C) Average relative mCherry/BFP fluorescence values (mCherry expression normalized by BFP expression to estimate plasmid copies per cell. NF: Anc5 cells; Par: parental CHO cells without a gene circuit. Unpaired t test was performed for each comparison, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5.

Computational modeling suggests mechanism of elevated protein expression. (A) Simplified chemical reaction scheme used in simulations. The unbound promoter A (thick arrow) has a high transcription rate m. hTetR can reversibly bind to a single (R01) or both (R02) operator sites in the promoter. Repressed promoters can still be transcribed at a low rate, leading to transcriptional leakage. (B) Comparison of simulated and experimental (10, 12) dose response curves (eGFP steady state levels versus Dox concentration) of ancestral CHO mNF cells, and Evo2 cells bearing inactivating hTetR mutation. (C) Steady state levels of eGFP and hTetR mRNA and proteins increase with increasing DNA copy number, as we observe experimentally in Evo3 through Evo6 (Fig. 4A). (D) Fraction of promoter states (unbound promoter, singly bound promoter, and doubly bound promoter) at increasing DNA copy number suggests promoter leakage as the mechanism for increased protein expression.

Fig. 6.

TFOs as target-specific countermeasures of drug resistance due to DNA amplification. (A) List of selected TFO target sites and their location within the mNF gene circuit, with the corresponding purine stretches individually highlighted. (B) Combinatorial treatment 35 or 50 μg/mL Puromycin and 200 nM TFO2 versus non-targeting control oligonucleotide for all evolved replicates. We normalized the growth rate for each condition by the corresponding growth rate of the control population treated with control oligo and the same concentration of Puromycin. (C) Combinatorial treatment of 0 or 35 μg/mL Puromycin with a total 200 nM TFOcombo mix consisting of TFO1, TFO2, and TFO3 at 1:1:1 ratio versus non-targeting control oligo, for all evolved replicates. We normalized the growth rate for each condition by the corresponding control growth rate of the control population treated with control oligo and the same concentration of Puromycin. Unpaired t test was performed for each comparison, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7.

TFOs overcome DNA-amplification-dependent drug resistance in cancer cells. (A) Doxorubicin response curve in the Doxorubicin-resistant subline H69AR derived from H69 parental cells. We used serially diluted Doxorubicin concentrations ranging from 100 μM to 0.001 μM and normalized growth rates to the corresponding untreated controls. We determined the IC50 (Doxorubicin concentration producing 50% growth inhibition) mathematically, fitting the plot of percent growth versus logarithmic drug concentrations (n = 4 for each concentration). (B) ABCC1 gene DNA and mRNA amplification in H69AR cells with or without Doxorubicin compared to H69 parental cells. We gave the cells a drug holiday by removing Doxorubicin from the growth medium for 4 wk and then re-selecting them by re-adding 0.8 μM Doxorubicin for 2 wk before qPCR (n = 3). (C) The effect of combinatorial treatment using 0, 50, or 400 μM Doxorubicin together with 200 nM total concentration of TFOcombo mix consisting of ABCC1_TFO1, ABCC1_TFO2, ABCC1_TFO3, and ABCC1_TFO4 at 1:1:1:1 ratio versus non-targeting control oligo, for H69AR compared with H69 parental cells. We normalized the growth rate for each condition by the growth rate of the non-treated population of each cell line. We used an unpaired t test for each comparison, n = 4, “ns” P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.