T4 polynucleotide kinase, Phi29 DNA polymerase, T4 DNA ligase, exonuclease I (ExoI), and exonuclease III (ExoIII) were obtained from Fisher Scientific (Slangerup, Danmark). All oligonucleotides were obtained from DNA Technology A/S (Aarhus, Denmark). CodeLink Activated Slides came from SurModics (Eden Prairie, MN, USA), and Vectashield was from Vector Laboratories (Peterborough, UK). Pap Pen was purchased from Dako (Glostrup, Denmark), CPT was from Sigma-Aldrich (Broenby, Denmark). Cell culture media (Minimum Essential Medium and McCoy 5A medium), Fetal Bovine Serum (FBS), 0.25% Trypsin-EDTA (25200-056), Non-Essential Amino Acid (11140-050) and PenStrep (15140-122) stock were obtained from Invitrogen (Naerum, Denmark).

The substrate for hTopI, S(hTopI)Id16, had the sequence 5′-AGA AAA ATT TTT AAA AAA ACT GTG AAG ATC GCT TAT TTT TTT AAA AAT TTT TCT AAG TCT TTT AGA TCC CTC AAT GCT GCT GCT GTA CTA CGA TCT AAA AGA CTT AGA-3′, the positive control substrate, S(PosC)Id33, had the sequence 5′-p-AGA AAA ATT TTT AAA AAA ACT GTG AAG ATC GCT TAT TTT TTT AAA AAT TTT TCT AAG TCT TTT AGA TCC CTC AAT GCA CAT GTT TGG CTC CGA TCT AAA AGA CTT-3′, the fluorescently labeled detection oligonucleotides, ID16-TAMRA and ID33-6FAM, had the sequences 5′-TAMRA-CCT CAA TGC TGC TGC TGT ACT AC-3′ and 5′-6FAM-CCT CAA TGC ACA TGT TTG GCT CC-3′ respectively. The Rolling Circle Amplification (RCA) primer used for Rolling Circle Enhanced Enzyme Activity Detection (REEAD) had the sequence 5′-C6amine-CCA ACC AAC CAA CCA AAT AAG CGA TCT TCA CAG T-3′. The TDP1-biosensor had the sequence 5′-ATTO488-AAA GCA GGC TTC AAC GCA ACT GTG AAG ATC GCT TGG GTG CGT TGA AGC CTG CTT T-BHQ1-3′.

Caco2 cells were grown in minimum essential medium (MEM) supplemented with 20% FBS, 1% non-essential amino acids, 1% PenStrep. HT29 cells were grown in McCoy 5A medium supplemented with 10% FBS, 1% PenStrep. The cell cultures were maintained in a humidified incubator (5% CO₂/95% air atmosphere at 37 °C). Cells were harvested by trypsin treatment (0.25% Trypsin-EDTA solution). The trypsin was inactivated with FBS-containing media, followed by two consecutive washes with 1× PBS. Cells were counted, aliquoted into tubes each containing 5 × 10⁵ cells, and stored at −80 °C until further analysis. Preparation of whole cell extracts was done by mixing 5 × 10⁵ cells with 500 μL, 250 μL, or 100 μL lysis buffer (10 mM Tris-HCl, pH 7.5, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 1 mM DTT) for hTopI activity measurement, hTopI activity measurement in the presence of CPT, and TDP1 activity measurement, respectively. Tubes with cells and lysis buffer were incubated on ice for 10 min before the extract was used.

Cell extracts from Caco2 or HT29 cells were analyzed by 10% SDS-polyacrylamide gel-electrophoresis followed by western blotting onto a Nitrocellulose Protran BA85 (GE Healthcare Life Sciences, Broenby, Denmark) membrane and antibody incubation using a polyclonal hTopI antibody (Topogen, Port Orange, FL, USA) and a Anti-TATA binding protein TBP [1TBP18] antibody (Abcam, Cambridge, UK) as loading control.

The hTopI reaction was carried out in a 20 μL reaction volume containing a divalent cation depletion buffer (10 mM Tris-HCl, pH 7.5, 5 mM EDTA, 50 mM NaCl, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 mM DTT) supplemented with S(hTopI)Id16 DNA substrate to a final concentration of 0.5 μM. Reactions were initiated by the addition of 10 μL cell extract in the presence or absence of CPT as indicated in the figure legends. DMSO was added as solvent control of CPT. Incubation was continued at 37 °C for 60 min for the dilution experiments and 15 min for CPT sensitivity experiments, before the heat inactivation of the enzymes by incubation for 5 min at 95 °C. To allow quantification, 5 nM of ligated S(PosC)Id33 was added to the heat inactivated reaction mixture and to increase the efficiency of hybridization NaCl was added to a final concentration of 250 mM. Preparation of the ligated S(PosC)Id33 was done using the T4 DNA ligase, and the resulting ligated circles were exonuclease digested with ExoI and ExoIII to remove non-ligated S(PosC)Id33 before gel-purification using 8% polyacrylamide gel. The concentration of the obtained circles was determined by spectrophotometric measurement. The 5′-amine-coupled RCA primer was linked to CodeLink Activated Slides according to the manufacturer's protocol. Subsequently, reaction mixtures were hybridized to the surface coupled primer for 60 min at 37 °C. Slides were washed for 1 min at room temperature in wash buffer 1 (100 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.3% SDS) and subsequently for 1 min at room temperature in wash buffer 2 (100 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween-20). Finally, the slides were dehydrated in 99.9% ethanol for 1 min and air-dried. Rolling circle DNA synthesis was performed for 60 min at 37 °C in 1× Phi29 buffer supplemented with 0.2 μg/μL BSA, 250 μM dNTP, and 1 unit/μL Phi29 DNA polymerase. The reaction was stopped by washing in wash buffers 1 and 2 for 1 min each followed by dehydration in 99.9% ethanol for 1min. The RCA products were detected by hybridization of 0.2 μM of each of the detection probes ID16-TAMRA (red fluorescent labeled) and ID33-6FAM (green fluorescent labeled) in a buffer containing 20% formamide, 2× SSC, and 5% glycerol overnight at 37 °C. The slides were washed in wash buffer 1 for 10 min, in wash buffer 2 for 5 min, dehydrated in ethanol for 1 min, mounted with Vectashield, and visualized under a fluorescence microscope as described previously [24]. For quantitative depiction of the results, the number of signals originating from hTopI activity (red signals) or spike in control circles (green signals), respectively, were counted on 10 pictures and the results calculated as the ratio between the number of red (R) and green (G) signals (R/G) as described [24]. All experiments were done in triplicate

The cells were seeded at 6 × 10⁴ cells/cm² and allowed to adhere before CPT was added to the media at final concentrations of 0.1, 0.2, 0.5, and 1 μM. CPT-treatment was continued for 24 h at 5% CO₂/95% air atmosphere at 37 °C. After harvesting as described above, cells were resuspended in 1× PBS, and the proportions of dead and live cells were determined by trypan blue staining (0.4%) followed by cell counting in a hemocytometer. The viability was calculated using the formula (live cell count/total cell count) × 100.

A dilution series of 5 × 10⁵, 2.5 × 10⁵, and 1.25 × 10⁵ cells/100 μL was made using lysis buffer as diluent. The TDP1 activity assay was carried out in a black Corning 384 well microplate with a final volume of 25 μL containing 20 μL cell extract, 1× TDP-buffer (20 mM Tris-HCl pH 8, 100 mM KCl, 10 mM DTT, 10 mM EDTA and 0.05% Triton X-100), and 0.5 μM TDP1-biosensor. For the negative control lysis buffer was used instead of cell extract. Real-time fluorescence measurements were performed every 10 s at 37 °C using a FlexStation 3 set to an excitation wavelength at 494 nm, an emission wavelength at 518 nm and a cutoff at 515 nm. The data were transferred to Microsoft Excel, where the linear slope from the data measured from 15 to 20 min was calculated for all samples and thereafter visualized as an XY-plot using GraphPad Prism. The experiments were done in triplicate.

Although hTopI has been examined previously for use as a predictive biomarker of CPT sensitivity these studies have focused solely on the level of gene copy number, mRNA expression, protein amount, or activity rather than testing the CPT response of the endogenous expressed hTopI. The hTOPI gene copy number is frequently increased in colorectal cancer [25,26] and the two colon cancer cell lines used here, Caco2 and HT29 are known to have an approximate twofold difference in hTOPI gene copy number. Caco2 has eight copies of the hTOPI gene, whereas HT29 has five [27]. These cell lines were examined with regards to hTopI protein amount, hTopI activity, cellular CPT sensitivity, TDP1 activity, and CPT response of hTopI in whole cell extracts.

Similar to McLeod et al. [19], we found a correlation between the known hTOPI gene copy number and the hTopI protein amount based on western blot analysis of extract from the two cell lines. Thus, the Caco2 cell line, which has eight copies of the hTOPI gene, had around a twofold higher amount of hTopI protein compared to the HT29 cell line, which has five copies of the hTOPI gene (Figure 1A). This could indicate that there is a similar expression from the hTOPI genes in the two cell lines. However, hTOPI gene copy number, RNA level, and protein amount is not capable of predicting changes in activity due to posttranslational factors. This may be problematic since the activity of TopI is known to be influenced by posttranslational modifications [28,29] as well as protein-protein interactions [30].

We have previously published a method for highly sensitive detection of hTopI using Rolling Circle Enhanced Enzyme Activity Detection (REEAD), capable of measuring hTopI cleavage-religation activity at the single molecule level (Figure 1B) [24]. This method, which is based on a DNA-sensor with a preferred DNA binding site of hTopI, can detect any changes in activity whether it is DNA-binding, -cleavage, or -ligation. Using this method we found an approximate twofold higher hTopI activity in the extract from the Caco2 cell line compared to the extract from the HT29 cell line (Figures 1C,D). In contrast to previous findings where hTopI protein amount and enzyme activity did not correlate [12,21], we found that the twofold higher hTopI activity in Caco2 correlated well with the hTOPI gene copy number and protein amount. This discrepancy between the two studies may likely be due to the use of different cell lines, which regulate hTopI activity differently at the posttranslational level, thus without affecting the hTopI protein amount.

To examine if the hTOPI gene copy number, protein amount, or enzyme activity could be correlated to cellular CPT sensitivity, the Caco2 and HT29 cultures were exposed to increasing concentration of CPT and the surviving cells were counted and scored as the percentage of the total number of cells. Despite the fact that the Caco2 cells have eight copies of the hTOPI gene, and the HT29 have five, twofold more hTopI protein, and twofold higher activity of hTopI, the Caco2 and HT29 cells had similar sensitivity to CPT (Figure 2A). This was unexpected since hTopI activity had been found to be the best determinant for CPT sensitivity, where high hTopI activity resulted in high cellular CPT sensitivity [21].

An explanation to this could be mutations in the hTOPI gene, which have been shown to be able to modify the cellular CPT sensitivity [22,31,32]. However, sequencing of the hTOPI genes from either culture did not reveal any mutations in the genes (data not shown). Decreased cellular CPT-sensitivity can be obtained also through upregulation of enzyme activities involved in repair of TopI-DNA complexes such as TDP1 [33,34]. We have previously published an optical DNA-based sensor (Figure 2B) capable of detecting the activity of TDP1 specifically even in crude cellular extract [35]. The TDP1-sensor was shown to be capable of precisely detecting changes in TDP1 activity, and since it has been shown that increased TDP1 activity can influence cellular toxicity of CPT, we measured the activity of TDP1 in extract from both Caco2 and HT29. In Figure 2C it can be seen that Caco2 cells did indeed have an approximate twofold higher TDP1 activity compared to HT29. The higher activity of TDP1 in Caco2 might contribute to the lower CPT sensitivity. Although Barthelmes et al. did not measure the TDP1 activity when they showed that overexpression of TDP1 could reduce CPT sensitivity, they reported a more than 100-fold increase in TDP1 protein [33]. Despite that it has been shown that the amount of TDP1 protein does not always correspond to TDP1 activity [36], a 100-fold increase in TDP1 protein is likely to cause much more than a twofold increase in activity. Thus, the twofold increased TDP1 activity in Caco2 compared to HT29 is likely to contribute to the overall CPT sensitivity of the cell line but might not be the only factor.

Since hTopI activity has been reported to be regulated both by posttranslational changes [37,38] and protein-protein interactions [30], it is possible that these factors may affect the drug response of hTopI directly. Such factors will not be revealed by measuring hTopI activity, -expression, -gene copy number, or by analyzing the hTOPI gene for mutations. Therefore, we speculated that since the activity of the endogenously expressed hTopI did not correlate with the CPT sensitivity of the cell lines, different posttranslational modifications or protein-protein interactions could affect hTopI differently in the two cell lines resulting in similar cellular sensitivity to CPT. We have previously stressed the importance of using sensor systems resembling the natural substrates for the targeted enzymes as much as possible, i.e., DNA-based sensor systems for DNA-modifying enzymes, since this will enable detection of differences in the DNA interaction of both enzymes and drugs [39]. In the REEAD assay such a DNA-based sensor system is exploited and is capable of detecting the cleavage-religation activity of hTopI at the single molecule level. Thus, CPT will prevent religation of the DNA-sensor leading to the generation of fewer closed DNA circles and thereby also fewer signals. Performing REEAD in the presence of CPT should therefore allow detection of any changes in the sensitivity of hTopI towards CPT. As seen in Figure 3 hTopI from the Caco2 cell line was less sensitive towards CPT than was hTopI from HT29 cells. The specificity of REEAD and the use of a divalent cation depleted buffer, leave most enzymes inactive due to lack of co-factors. Thus, it seems plausible that the different drug response observed in the two cell cultures could be a combination of increased TDP1 activity and different drug sensibility of the endogenously expressed hTopI.