Targeting G-quadruplexes as novel anti-breast cancer strategies

 ABSTRACT

The realization that G-quadruplex (G4) DNA structures are involved in transcriptional regulation and the maintenance of genome integrity has led to growing interest in their potential as cancer therapeutic targets. Herein, two separate strategies were undertaken with the aim of specifically targeting G4s to eventually regulate the transcriptional output of cancer-related genes. The first assessed the chromatin remodelling activity in a breast cancer cell line of three G4-stabilizing small molecules, GTC-365 and GTC-260 shown to have high specificity for the hTERT G4 and GQC-05 shown to have high specificity for the cMYC G4. The assay for transposase-accessible chromatin with sequencing (ATAC-Seq) was used to generate a library of tagmented genomic DNA following compound treatment in order to generate genome-wide maps of nucleosome depleted regions. Given that nucleosome positioning determines the ability of transcriptional machinery to access DNA, any chromatin remodelling activity by these small molecules could alter transcription. For the second G4-targeting anti-cancer strategy, a CRISPR-Cas9 system was designed with the aim of resetting the methylation status of G4s in the promoter region of specific oncogene and tumour suppressor genes in breast cancer cells. Alterations to genomic methylation patterns has been shown to influence the formation of G4s, and as these structures are important in gene regulation, changes to their formation could also impact the expression of genes.

INTRODUCTION

Breast cancer, the principal cause of cancer death among females globally (Torre et al., 2015), represents a major obstacle to public health. Despite survival rates doubling in the past 40 years in the UK, the fact that approximately 31 individuals die every day in the UK as a result of the disease (Cancer Research UK, 2016) demonstrates the acute need for more effective treatments. The pursuit of such novel treatments has led to the development of targeted therapies, including those that aim to target specific DNA regions or structures to prevent cancerous growth.

 

The commonly recognised structure of DNA is that of a duplex molecule in which two antiparallel strands are held together by Watson-Crick base pairing. However, it has been realised that DNA can undergo alternative base-pairing to form non-canonical secondary structures. Indeed, in certain guanine-rich regions of DNA it is possible for the formation of G-quadruplexes (G4s), four-stranded structures consisting of guanine bases stabilised by a central cation, to occur. Figure 1 depicts the G4 structure. The realization that these G4s are involved in transcriptional regulation(Gray et al., 2014) and the maintenance of genome integrity (Bochman, Paeschke and Zakian, 2012) has led to growing interest in their potential as cancer therapeutic targets (Monchaud and Teulade-Fichou, 2008; Yang and Okamoto, 2010; Vy Thi Le et al., 2012; Bidzinska et al., 2013; Ohnmacht et al., 2015). It is particularly their high abundance in regulatory, promoter and nucleosome depleted regions of the genome (Hänsel-Hertsch et al., 2016) that makes them such valuable targets for selectively reprogramming gene expression as an anticancer strategy.

Although numerous small molecules with high affinity for G4s have been reported and the molecular mechanisms of their G4-binding activity investigated, there is a need for greater understanding of the endogenous effects of these ligands within a cellular context. Indeed, it is thought that small molecule interactions with the genome are dependent on a dynamic interaction between the DNA code and chromatin states (Rodriguez and Miller, 2014). If G4-binding ligands are to achieve any clinical relevance, it is crucial that more is understood about the way in which they interact and impact the chromatin landscape of the human genome. Herein, the chromatin remodelling activity in the MCF-7 breast cancer genome of three G4-stabilizing small molecules, GTC-365 and GTC-260 shown to have high specificity for the hTERT G4⁸ and GQC-05 shown to have high specificity for the cMYC G4⁹, was measured.

It has also been realised that changes in DNA cytosine methylation (5-mC) can impact the formation of G4s, and that G4s in the oncogenic promoters of cancerous cells demonstrate altered methylation profiles. Therefore separately within this project, a CRISPR system was developed with the aim of resetting the methylation status of G4s in the promoters of specific oncogene and tumour suppressor genes in breast cancer cells. This was done in order to evaluate whether such changes in methylation could lead to altered gene expression, ultimately with the aim of halting the growth pathways and activating the death pathways of breast cancer cells.

MATERIALS AND METHODS

Assessing chromatin remodelling activity

Cell culture

MCF-7 cells were cultured in Minimum Essential Media  (Thermo Fisher, MEM , nucleosides, cat. no. 12571063) supplemented with 10% FBS (Thermo Fisher), in a humidified incubator at 37C and 5% CO₂. Cells were passaged every three days.

IC₅₀ values were calculated using the MTS Assay (CellTiter 96 AQ_ueous Cell Proliferation Assay, Promega). 2000 cells were seeded into each well of a 96 well plate and after overnight adhering, incubated for 72 h with 100 µL G4-stabilising compound at ten different concentrations ranging from 0.01 µM to 100 µM. 40 µL of tetrazolium dye was then added to each well and after 2 h incubation, optical density read at 490nm using the 2300 EnSpire Multimode Plate Reader. Background absorbance was accounted for by the use of triplicate ‘blank’ wells that did not contain cells but media and tetrazolium dye only. The results are shown as the concentration at which cell growth was prevented after the 72 h incubation period, expressed as a percentage of the controls. The IC₅₀ was calculated using linear regression analysis (GraphPad Prism).

According to the calculated IC₅₀ values, drug dilutions of appropriate concentrations were made up to a total of 10 ml in MCF-7 complete media. GQC-05 was diluted in complete media to a 0.3 µM solution, GTC-260 to a 3 µM solution and GTC-365 to a 20 µM solution. MCF-7 cells were seeded into 15 mm plates and allowed to adhere overnight. The following day, complete media was removed and for the control plate, replaced with 10 ml fresh complete media, whilst for each treatment plate replaced with 10 ml drug dilution. Plates were left for 72 hr to allow cells to reach 80% confluency.

After the drug incubation period, media was removed and all cell monolayers washed twice with 3 ml Phosphate Buffered Saline (PBS). Cell detachment was performed by addition of 3 ml TrypLE Select 1X (Thermo Fisher, cat no. 12563029) to each plate; these were incubated for 5 min at 37C. Cells were collected by centrifugation for 5 min at 200g and subsequently counted using a haemocytometer.

ATAC-seq

50,000 cells per reaction were used as input for the assay for Transposase-Accessible Chromatin with high-throughput sequencing (ATAC-seq). The assay was performed on each treatment type, including untreated controls, in duplicate according to Buenrostro et al (2015), but centrifuging samples for 5 min only at 500 x g and 4C once the cell pellet had been resuspended in 50 l cold lysis buffer. The quality of purified libraries was assessed using the Agilent 2200 TapeStation and quantified using the KAPA Library Quantification Kit for Illumina platforms (Kapa Biosystems, kit code KK4824) on the LightCycler 480 Real-Time PCR System, according to manufacturer instructions. Libraries were sequenced using 50-bp paired end reads on the HiSeq2500 platform.

RNA-seq

Total RNA extraction from cells treated with each drug, and from untreated control cells, was performed using the Qiagen RNeasy Mini Kit (cat. no 74106), according to manufacturer instructions. In order to determine if the extracted RNA was sufficient for library preparation, samples were quantified using the Invitrogen Qubit 3 Fluorometer, according to manufacturer instructions. RNA-seq libraries were then prepared using the Illumia TruSeq Stranded Total RNA LT with Ribo-Zero Human Sample Preparation Kit (cat. no 20020597). Libraries were sequenced using 50-bp single reads on the HiSeq2500 platform.

Bioinformatics

Bioinformatic analysis of ATAC-seq data is currently being completed at the Garvan Institute of Medical Research. That of RNA-seq data is being sequenced by AGRF Bioinformatics. Image analysis is being carried out in real time using HiSeq Control Software v2.2.68 and Real Time Analysis v1.18.66.3. This performs real time base calling using the HiSeq instrument computer and the Illumina bcl2fastq 2.20.0.422 pipeline is used to produce sequence data.

CRISPR

Design of CRISPR-Cas9 constructs

A CRISPR-Cas9 system was developed to target alterations in DNA cytosine methylation present in breast cancer cells. Genes were chosen that contain a G4 demethylated region (G4-DMR) in their promoter and display abnormal transcriptional activity in breast cancer cells. This selection was achieved using MCF-7 whole genome maps of G4-DMR. The G4-DMR in the core promoter of the human telomerase reverse transcriptase catalytic subunit (hTERT) and that in the PDGFR gene were chosen. For each gene, potential guide RNA (gRNA) sequences of 17-18 base pairs (bp) in the target promoter were found using the crispr.mit.edu site and run on BLAT to ensure specificity to the target. The 60mer oligos of the sequences found to be specific to the target gene promoter were ordered from Integrated DNA Technologies (IDT) and extended at either end to produce 100bp double stranded DNA using Q5 Hot Start High-Fidelity 2X Master Mix and the following cycling conditions:

Step	Temperature	Time	10 cycles
Initial Denaturation	980C	2 minutes
Denaturation	980C	15 seconds
Annealing*	500C	20 seconds
Extension	720C	15 seconds

For AMPure purification of PCR products, twice the reaction volume of AMPure beads was added to PCR product and the solution vortexed then incubated for 5 min at room temperature (RT). Tubes were placed on a magnetic rack for 5 min and the supernatant was removed. Beads were washed twice for 30 s with 200 µL of 70% ethanol, ensuring complete removal of all ethanol each time. Tubes were microfuged and any residual ethanol removed. Beads were soaked for 1 min in 20 µL TE/10, and thoroughly mixed using a pipette. Tubes were incubated for 3 min at RT and then placed on a magnetic rack for 3 min. 19 µL of the supernatant was transferred to a new tube, and this diluted to 250 fmol in TE/10.  1 µg of gRNA cloning vector was digested with 0.3 µL of AflII and 20 units Shrimp Alkaline Phosphatase in a 50 μL reaction for 1 h. Tubes were incubated for 20 min at 65 ^oC to inactivate both enzymes. AMPure purification was carried out using one reaction volume of beads and eluting in 20 µL of TE/10. Nanodrop was used to quantitate the digestion. 250 fmol of gRNA, 10 fmol of the digested gRNA vector, and 1 X Gibson Assembly Master Mix were added to a microfuge tube and incubated for 1 hr at 50 ^oC. AMPure purification was performed again using 1 volume of beads but eluting in 10 µL of dH₂O. The top 10 competent cells were transformed using Gibson assembly product and plated onto LB Amp agar. A minimum of 4 transformants were selected for miniprep, and diagnostic restriction digest carried out on these.

Transfection of MCF-7 cells

The day before transfection, MCF-7 cells were seeded in complete media at 2.5 x 10⁵ per well. The next day, plasmid DNA (pDNA) stock was diluted in Milli-Q water to a 100 ng µL^-1 stock. 10 µL diluted pDNA was diluted in OptiMEM to a total 100 µL solution. 3 µL FuGENE HD was added directly into this solution and briefly vortexed. The tube was incubated for 15 min at RT and the solution then spread dropwise over the media. The plate was gently rocked back and forth to mix and incubated overnight at 37 ^oC and 5% CO_2. The following day, media and transfection mix were removed and replaced with complete media. The next day, EGFP expression was assessed using fluorescence-activated cell sorting (FACS).

 

Bisulphite conversion of transfected cells, amplification and sequencing

In order to establish the methylation profiles of DNA targets, bisulphite conversion was carried out using the Zymo Research EZ DNA Methylation-Direct Kit (cat no. D5020), according to manufacturer instructions and using transfected MCF-7 cells as the input material.

Prior to amplification and sequencing of the bisulphite converted DNA, four primer pairs were designed manually for both the hTERT and PDGFRβ DNA targets. These primers had to be in a 1000 bp region close to the guide target and, in order to amplify the bisulphite converted versions of both the demethylated and methylated sequences, had to avoid CpGs. Primers were designed and first tested on untransfected, but bisulphite converted, MCF-7 DNA in order to check for successful amplification. Successful primer pairs were used to amplify the transfected DNA.

qPCR analysis

Due to time constraints, qPCR analysis of the amplified DNA is now being undertaken by others. This analysis will assess the relative mRNA levels of the hTERT and PDGFRβ target genes was assessed using quantitative PCR, comparing transcription levels before and after transfection.

RESULTS

Figure 3 shows the cell viabilities of MCF-7 cells incubated separately with each of the three G4 stabilising ligands. The inhibitory activity of each of the ligands varied, as demonstrated by the IC₅₀ values.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unfortunately, sequencing data for the chromatin remodelling project is currently unavailable as these results are still undergoing analysis. The CRISPR project is also being completed by others; qPCR is yet to be finished and therefore the final results cannot be reported here.

 

DISCUSSION

Although considerable research exists aiming to create new classes of G4-binding ligands or elucidate their G4-binding mechanisms, it is crucial to understand the chromatin-level changes induced by such ligands before their relevance in any clinical setting can be achieved. Whilst analysed sequencing data is not yet available, it is important to evaluate the methods used in this project and the challenges that were faced.

An important step prior to sequencing of ATAC-seq libraries was the evaluation of DNA quality in order to avoid wasted time and resources on degraded libraries that would not yield results. The 2200 TapeStation was chosen for this purpose as a more efficient method of quality control than the labour intensive traditional process of gel electrophoresis. Indeed, the TapeStation very rapidly analyses samples, reducing the time for the entire experiment to be carried out. Given that ATAC-seq was partially selected due to the fact that the protocol can be completed much faster than other methods that can provide information on chromatin accessibility, such as DNase I hypersensitive sites sequencing (DNase-seq), it was desirable to keep the time taken for downstream processes to a minimum.

After loading DNA libraries onto the 2200 TapeStation, the analysis software provides a gel image. It is thought that gels demonstrating a laddering pattern with periodicity of approximately 200 bp are indicative of intact chromatin that has undergone sufficient but not excessive tagmentation (Milani et al., 2016). The profiles in figure 4 exhibit such a pattern, which was used as evidence that the DNA from all samples was intact and that chromatin had been successfully tagmented. This, along with results from qPCR analysis that the quantity of DNA in each sample was adequate, led to the decision that all samples should be sequenced. Notably, fragments smaller than 100 bp are likely to be mitochondrial in origin and therefore were not considered in quality analysis.

Although these high-quality libraries were eventually produced, it should be realised that optimal tagmentation was not achieved immediately. Repeated ATAC-seq attempts led to either under- or over-tagmentation of DNA, with all such samples having to be abandoned. In order to optimise experiments, the ATAC-seq protocol was attempted with both fewer (25,000) and greater (75,000) cell numbers, but these changes did not improve results. However, when the experiment was carried out with fewer cells, under-tagmentation was observed in which fragments were overly long, whereas increasing cell numbers led to over-tagmentation in which the majority of fragments were too short and gel banding patterns were no longer visible. Lowering cell numbers did, nonetheless, provide the opportunity to optimise lysis conditions; centrifuging samples for 5 min only at 500 x g and 4C once the cell pellet had been resuspended in 50 l cold lysis buffer, rather than the suggested 10 min, led to the creation of high-quality libraries when the protocol was again attempted with 50,000 cells.

The other breast cancer treatment project, involving use of the CRISPR-Cas9 system, required the production of primers for the amplification and sequencing of bisulphite converted DNA. This was a particular challenge as the experiment aimed to target the hTERT and PDGFRβ promoter G4s, and promoter regions are characterised by containing a very high density of CpGs. It was therefore crucial to avoid all CpG regions when designing these primers, as they were intended for the amplification of bisulphite converted DNA; coverage of any CpGs would result in the methylated or unmethylated sequence being selectively amplified. Indeed, depending on whether a T or a C was selected in the primer where the CpG occurred, a 0 or 100% methylation result would be the output. Such skewed data would render any results obsolete.

Initially, several primer pairs were designed with the aim of amplifying a fragment of maximum 300 bp. However, it proved impossible to locate enough regions without CpGs, and therefore an approach that aimed to amplify one large fragment of approximately 1000 bp was used. This enabled searching for primers within a larger DNA region, which was advantageous because regions further from the G4-forming region are slightly less rich in CpGs. The primer design still had to be carried out by hand rather than using the online sites that tend to be routine, because online sites tended to fail to find any potential primer sites due to the density of CpGs in the area. Eventually, enough locations that were sparse in CpGs were located and primers were ordered and tested. These were tested initially on untransfected MCF-7 cells to prevent wastage of resources, and the two most successful primer pairs were chosen for further use.

This project entailed the broader problem of specificity of the CRISPR-Cas9 construct. Although off-target sequences were tested to check whether the construct could indeed bind to sites away from the intended hTERT or PDGFRβ G4 regions, it remains a possibility that untested sites could be affected by the system. Indeed, although next generation sequencing (NGS) can be used to assess the specificity of the CRISPR system, this cannot distinguish all DSBs as the local chromatin environment can greatly impact the efficiency of the non-homologous end joining (NHEJ) process (Zhang et al., 2014). Given that methylation is a prominent mechanism by which the genome regulates transcription, the idea that this CRISPR-Cas9 construct could even possibly alter the methylation status of unidentified similar sequenced sites is one that must be extensively addressed in future. Such research could include attempting to use a CRISPR construct with the Cpf1 nuclease, which has been discovered to be more specific, causing double strand breaks (DSBs) at fewer off-target sites than Cas9 (Kim et al., 2016; Kleinstiver et al., 2016).

 

The cost of breast cancer on society today, both financially and in human loss of life, necessitates the discovery of novel treatments that can produce fewer harmful side effects and better clinical outcomes. Research into the use of G4 stabilising ligands and a CRISPR-Cas9 system are clearly at their very early stages, but demonstrate efforts that will hopefully, in the future, be able to provide exactly this.

 

REFERENCES

Bidzinska, J., Cimino-Reale, G., Zaffaroni, N. and Folini, M., 2013. G-Quadruplex Structures in the Human Genome as Novel Therapeutic Targets. Molecules, 18(10), pp.12368–12395.

Bochman, M.L., Paeschke, K. and Zakian, V.A., 2012. DNA secondary structures: stability and function of G-quadruplex structures. Nature Reviews Genetics, 13(11), pp.770–780.

Cancer Research UK, 2016. Breast Cancer – UK Incidence Statistics [Online]. Available from: https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/breast-cancer#heading-Two [Accessed 2 September 2018].

Gray, L.T., Vallur, A.C., Eddy, J. and Maizels, N., 2014. G quadruplexes are genomewide targets of transcriptional helicases XPB and XPD. Nature Chemical Biology, 10(4), pp.313–318.

Hänsel-Hertsch, R., Beraldi, D., Lensing, S.V., Marsico, G., Zyner, K., Parry, A., Marco Di Antonio, Pike, J., Kimura, H., Narita, M., Tannahill, D. and Balasubramanian, S., 2016. G-quadruplex structures mark human regulatory chromatin. Nature Genetics, 48(10), pp.1267–1272.

Kim, D., Kim, J., Hur, J.K., Been, K.W., Yoon, S. and Kim, J.-S., 2016. Genome-wide analysis reveals specificities of Cpf1 endonucleases in human cells. Nature Biotechnology, 34(8), pp.863–868.

Kleinstiver, B.P., Tsai, S.Q., Prew, M.S., Nguyen, N.T., Welch, M.M., Lopez, J.M., McCaw, Z.R., Aryee, M.J. and Joung, J.K., 2016. Genome-wide specificities of CRISPR-Cas Cpf1 nucleases in human cells. Nature Biotechnology, 34(8), pp.869–874.

Milani, P., Escalante-Chong, R., Shelley, B.C., Patel-Murray, N.L., Xin, X., Adam, M., Mandefro, B., Sareen, D., Svendsen, C.N. and Fraenkel, E., 2016. Cell freezing protocol optimized for ATAC-Seq on motor neurons derived from human induced pluripotent stem cells. Scientific Reports, 5(6), pp.25474.

Monchaud, D. and Teulade-Fichou, M.-P., 2008. A hitchhiker’s guide to G-quadruplex ligands. Organic & Biomolecular Chemistry, 6(4), pp.627–636.

Ohnmacht, S.A., Marchetti, C., Gunaratnam, M., Besser, R.J., Haider, S.M., Di Vita, G., Lowe, H.L., Mellinas-Gomez, M., Diocou, S., Robson, M., Šponer, J., Islam, B., Barbara Pedley, R., Hartley, J.A. and Neidle, S., 2015. A G-quadruplex-binding compound showing anti-tumour activity in an in vivo model for pancreatic cancer. Scientific Reports, 16(5), pp.11385.

Rodriguez, R. and Miller, K.M., 2014. Unravelling the genomic targets of small molecules using high-throughput sequencing. Nature Reviews Genetics, 15(12), pp.783–796.

Torre, L.A., Bray, F., Siegel, R.L., Ferlay, J., Lortet-Tieulent, J. and Jemal, A., 2015. Global cancer statistics, 2012: Global Cancer Statistics, 2012. CA: A Cancer Journal for Clinicians, 65(2), pp.87–108.

Vy Thi Le, T., Han, S., Chae, J. and Park, H.-J., 2012. G-quadruplex binding ligands: from naturally occurring to rationally designed molecules. Current Pharmaceutical Design, 18(14), pp.1948–1972.

Yang, D. and Okamoto, K., 2010. Structural insights into G-quadruplexes: towards new anticancer drugs. Future Medicinal Chemistry, 2(4), pp.619–646.

Zhang, J.-H., Pandey, M., Kahler, J.F., Loshakov, A., Harris, B., Dagur, P.K., Mo, Y.-Y. and Simonds, W.F., 2014. Improving the specificity and efficacy of CRISPR/CAS9 and gRNA through target specific DNA reporter. Journal of Biotechnology, 189, pp.1–8.

ACKNOWLEDGEMENTS

This research was funded by the National Breast Cancer Foundation and the Cancer Council WA. Dr Nicole Smith, Prof Killugudi Swaminatha Iyer and Arnold Ou provided invaluable intellectual input. Many thanks to The University of Western Australia for the use of laboratory space that made this research possible.

 

APPENDICES

Appendix A

Personal Objectives and Learning Outcomes 1 (POLO 1)

Brief overview of the company/institute and department you are placed in:

I am working in the School of Molecular Sciences at the University of Western Australia, conducting research in Professor Killugudi Swaminatha Iyer’s laboratory.

Summary of your role and responsibilities:

As a placement student, I have been assigned research on two different projects, both revolving around the development of novel breast cancer therapeutics. The first project involves research into the use of small molecules to target G-quadruplexes that form in the promoter of genes involved in the development of breast cancer. In particular, we are currently working towards resolving the crystal structure of a particular G-quadruplex with an associated ligand. The second project is aiming to design a CRISPR system using TET-1 to demethylate specific DNA targets as a potential cancer therapeutic.

Summary of your current and upcoming work, projects, assignments:

Currently, I am undertaking multiple inductions so that I am able to independently use the laboratory’s equipment. For example, I have been inducted into the tissue culture lab, learning how to split and seed MCF-7 and MCF-10a cells as these are the cells that I will be working with. I am learning protocols for various processes such as qPCR, DNA extraction and quantification and ATAC-Seq. I will soon be inducted on the use of the confocal microscope, as my work will involve BG4 antibody imaging.

What are your overall expectations of the placement?

I expect that I will be required to quickly build on my repertoire of microbiological techniques so that I am able to provide meaningful data for both of my projects.

Which aspects of your degree do you hope to be able to use on your placement?

Both of my projects are genetics related, and so I am enjoying building on my existing genetics knowledge and improving my understanding of epigenetics. Having undertaken laboratory work during my degree, I am looking forward to using the techniques that I have learnt with greater independence and in a more ‘real’ scientific environment where results are genuinely unknown (despite the inevitable disappointment that this can entail!).

Biology and Biochemistry Principles, Practice & Achievement: The ability to apply an understanding of theoretical, methodological, empirical and practical knowledge and skill, to the solution of problems.  The ability to select and successfully apply appropriate principles, methods and techniques to placement tasks and reflect on their application.

Level of capability at start of placement: