Characterization of DNA topoisomerase I in three SN-38 resistant human colon cancer cell lines reveals a new pair of resistance-associated mutations

Background DNA topoisomerase I (Top1) is a DNA unwinding protein and the specific target of the camptothecin class of chemotherapeutic drugs. One of these, irinotecan, acting through its active metabolite SN-38, is used in the treatment of metastatic colorectal cancer. However, resistance to irinotecan represents a major clinical problem. Since molecular alterations in Top1 may result in resistance to irinotecan, we characterized Top1 in three human colon cancer cell lines with acquired resistance to SN-38. Methods Three SN-38 resistant (20–67 fold increased resistance) cell lines were generated and compared to wild-type parental cells with regards to: TOP1 gene copy number and gene sequence, Top1 expression (mRNA and protein), Top1 enzymatic activity in the absence and presence of drug, and Top1-DNA cleavage complexes in drug treated cells. TOP1 mutations were validated by PCR using mutant specific primers. Furthermore, cross-resistance to two indenoisoquinoline Top1-targeting drugs (NSC 725776 and NSC 743400) and two Top2-targeting drugs (epirubicin and etoposide) was investigated. Results Two of three SN-38 resistant cell lines carried TOP1 gene copy number aberrations: A TOP1 gene copy gain and a loss of chromosome 20, respectively. One resistant cell line harbored a pair of yet unreported TOP1 mutations (R364K and G717R) in close proximity to the drug binding site. Mutant TOP1 was expressed at a markedly higher level than wild-type TOP1. None or very small reductions were observed in Top1 expression or Top1 activity in the absence of drug. In all three SN-38 resistant cell lines Top1 activity was maintained in the presence of high concentrations of SN-38. None or only partial cross-resistance were observed for etoposide and epirubicin, respectively. SN-38 resistant cells with wild-type TOP1 remained sensitive to NSC 743400, while cells with mutant TOP1 was fully cross-resistant to both indenoisoquinolines. Top1-DNA cleavage complex formation following drug treatment supported the other findings. Conclusions This study adds to the growing knowledge about resistance mechanisms for Top1-targeting chemotherapeutic drugs. Importantly, two yet unreported TOP1 mutations were identified, and it was underlined that cross-resistance to the new indenoisoquinoline drugs depends on the specific underlying molecular mechanism of resistance to SN-38.


Background
In metastatic colorectal cancer (mCRC), current chemotherapeutic treatment options consist of either 5fluorouracil and folinic acid with oxaliplatin (FOLFOX) or 5-fluorouracil and folinic acid with irinotecan (FOLFIRI) [1,2]. However, the response rate to these regimens is only in the range of about 30-55 %. Resistance to chemotherapy and drug-induced side effects are major limitations [3,4], and the 5-year survival rate is less than 10 % [5]. One way to improve therapeutic efficacy is to introduce biomarkers to identify patients with a high likelihood of benefiting from drug treatment (see e.g. [6,7]).
Irinotecan belongs to the camptothecin class of chemotherapeutic drugs and is a pro-drug of the active metabolite SN-38. Camptothecins selectively target DNA topoisomerase I (Top1)-DNA cleavage complexes which form in the vicinity of replication and transcription complexes to unwind DNA [8][9][10]. Top1 is a 765 amino acid residue protein encoded by the 21-exon gene, TOP1 located on the long arm (q) of chromosome 20. Top1 binds supercoiled DNA, nicks a DNA strand allowing its rotation around the intact strand, and then religates the DNA [8,11]. Camptothecins bind and stabilize the Top1-DNA cleavage complexes, thus leading to DNA damage when replication or transcription occurs [8]. Similarly, DNA topoisomerase II (Top2) is the target of other classes of chemotherapeutic drugs, including the anthracylines and etoposide [8,12,13].
As Top1 is the direct target of SN-38, the active metabolite of irinotecan, it has been extensively studied as a possible mediator of resistance or as a predictive marker in mCRC. Top1 can be examined in several different ways; gene copy number aberrations and genetic mutations, mRNA and protein expression levels, and enzyme activity levels (see e.g. [6]). Studies have been performed both at the pre-clinical cellular level (e.g. [14][15][16][17]) and using clinical tumor samples (e.g. [17][18][19][20]). Positive correlation between Top1 protein level and gene copy number or mRNA level has been observed in several studies [14,21,22]. In cell-based studies, high Top1 expression and enzyme activity have generally been associated with sensitivity to camptothecins, whereas low Top1 is a common resistance mechanism [15,16,[23][24][25][26]. In addition, mutations or DNA methylation of the TOP1 gene have been associated with resistance to camptothecins [27,28]. Most mutations have been identified in cultured cells [27], and rarely in clinical patient material [29]. The largest clinical study investigating Top1 as a predictive marker of irinotecan treatment in mCRC to date is the UK FOCUS trial [18,30]. High tumor Top1 protein expression was found to correlate significantly with therapeutic benefit from irinotecan. However, a similar study, the Dutch CAIRO trial [31,32], was not able to replicate this finding.
In recent years, new classes of non-camptothecin Top1-targeting drugs have reached clinical development, e.g. the indenoisoquinolines, the dibenzonaphtyridinones and the indolocarbazoles [8,33,34]. Compared to camptothecins, indenoisoquinoline drugs are chemically stable, bind Top1-DNA cleavage complexes at other DNA sequences, form less reversible drug-Top1-DNA cleavage complexes and are not substrates of common multi-drug resistance efflux pumps [34,35].
In the present study we undertook a thorough investigation of the Top1 status in three human colon cancer cell lines with acquired resistance to SN-38 developed through approximately 9 months of drug exposure [36]. We investigated the Top1 gene copy number, genetic sequence, mRNA expression level, protein expression level, enzyme activity and formation of Top1-DNA cleavage complexes following drug treatment. In addition we tested the crossresistance to two non-camptothecin Top1-targeting drugs as well as two drugs targeting Top2.

Cell culture
The cell lines HCT116 and HT29 were obtained from the NCI/Development Therapeutics Program, while LoVo was obtained from the American Tissue Culture Collection. Cells were maintained at 37°C, 5 % CO 2 in RPMI 1640 + Glutamax growth medium (Invitrogen, Naerum, Denmark) supplemented with 10 % fetal calf serum (Invitrogen). SN-38 resistant cell lines were generated in our laboratory by exposing three colon cancer cell lines to gradually increasing drug concentrations for 8-10 months [36]. The cells were maintained in drug-free growth medium for at least 1 week and at most 4 weeks prior to any experiments.

Chemotherapeutic drugs
SN-38 (Sigma-Aldrich, Copenhagen, Denmark) was purchased and dissolved in dimethyl sulfoxide (DMSO) at a concentration of 10 mM and stored at -20°C. The indenoisoquinoline drugs NSC 725776 (LMP776) and NSC 743400 (LMP400), provided by the laboratory of Dr. Yves Pommier, were dissolved in DMSO at a concentration of 5 mM and stored at -20°C. Epirubicin (2 mg/ml, Actavis Nordic A/S, Gentofte, Denmark) and etoposide (20 mg/ml, Pfizer, New York, USA) were purchased and stored at -20°C. Drugs were diluted in growth medium immediately prior to use.

Drug sensitivity MTT assay
In vitro drug sensitivity was determined using the MTT (methylthiazolyldiphenyl-tetrazolium bromide) assay. Cells were seeded in 96-well plates, and a range of drug concentrations was added the following day. Following 48 h of drug exposure, the medium was discarded and the plates were incubated with medium containing MTT (0.5 mg/ml, Sigma-Aldrich) for 3 h. Acidified (0.02 M HCl) sodium dodecyl sulphate (20 %, Sigma-Aldrich) was added to dissolve the formed formazan. Optical density at 570 nm (and 670 nm for background) was measured, and the cell viability was calculated in percent compared to untreated cells. Experiments were repeated three times and the mean IC50-value ± standard deviation was determined. Relative resistance for each resistant cell line was calculated by dividing the mean IC50-value of the resistant cell line by the mean IC50-value of the corresponding parental cell line.

RNA purification and mRNA analysis
The RNA purification and TOP1 mRNA analysis is previously described [36]. Briefly, RNA was harvested from each cell line in triplicate using TRIzol Reagent (Invitrogen) and quantified using a Nano-Drop ND-1000 (Thermo Scientific, Waltham, USA). The Top1 mRNA level was obtained from a gene expression microarray analysis (Human Gene Expression Microarrays G4112F, Agilent Technologies, Santa Clara, USA) done in triplicate calculating mean ± standard deviation.

Metaphase preparation
Metaphase preparation has previously been described [37]. Briefly, upon reaching a confluence of approximately 70 %, colcemid (Invitrogen) was added to cell cultures. After 2 h at 37°C, cells were harvested and a hypotonic treatment was performed (0.075 M KCl) for 10 min. Cells were fixed (fixative: 3:1 vol/vol absolute methanol and glacial acetic acid) and the suspension was dripped onto glass slides.

Fluorescence-in-situ-hybridization (FISH) gene copy number analysis
The TOP1/Centromere-20 (CEN-20) probe combination and relevant protocol has previously been described [17]. FISH reagents were from the Cytology FISH Accessory Kit (K5499) and the Histology FISH Accessory Kit (K5799) (Dako A/S, Glostrup, Denmark). Metaphase specimens were fixed in 3.7 % formaldehyde, washed, dehydrated and air dried. Once dry, FISH probe was loaded onto slide, denatured and hybridized overnight. Excess probe was removed by washing in stringency buffer. Slides were washed, dehydrated, air dried and mounted. To determine the presence and mechanism of TOP1 copy number alteration in cell lines, signal locations and numbers were noted for 50 metaphases for each cell line at 1000x magnification.

PCR validation of TOP1 mutations
Primers with similar melting temperatures specific for wild-type (wt) or mutant (mt) sequences were designed: 5′-GGACTTTTCCGTGGCC-3′ (364wt, forward); 5′-TGGACTTTTCCGTGGCT-3′ (364mt, forward); 5′-GATAATTGAGTTTGGAGGTTCC-3′ (717wt, reverse); 5′-CAGATAATTGAGTTTGGAGGTTCT-3′ (717mt, reverse). 364 and 717 refers to the amino acid residue positions of the mutations. The mutated nucleotides are underlined in the primer sequences. The primers were purchased from Eurofins MWG Operon. cDNA from cell lines were prepared as described above, and PCR products were amplified as described above using combinations of the wild-type and mutation specific primers. PCR products (10 μl per sample) were loaded on a 2 % agarose gel along with a DNA ladder (MassRuler DNA Ladder, Low Range, Fermentas, Thermo Scientific) and negative controls (PCR reactions run on RNAse-free water). Gel electrophoresis was done as described above and the gel was photographed under UV illumination.

Protein purification and Western blotting
Cells were trypsinized, resuspended and washed in cold PBS and pellet frozen at -80°C. Cell pellets were resuspended in lysis buffer [1 % SDS, 10 mmol/L Tris (pH 7.4), 40 μL of 25x protease inhibitors (Roche, Basel, Switzerland), 10 μL phosphatase inhibitors (Sigma), 1 mL water]. Samples were sonicated (15 s), kept for 5 min on ice and heated for 5 min at 95°C. Total protein was determined using the Lowry method, i.e. absorbance measurement and using a standard curve of bovine serum albumin in sample buffer. Samples were diluted and 10 μg of total protein per sample were applied on a 4-20 % gradient gel (Novex, Invitrogen) together with a marker (SeaBlue Plus2 Pre-Stained Standard, Invitrogen), and subjected to electrophoresis (120 V, 1.5-2 h). Gel was transferred to a polyvinylidene difluoride (PVDF) membrane by semi-dry blotting 5 V overnight. Membrane was blocked for 1 h at room temperature in 5 % milk in PBS-Tween20 buffer. Subsequently, the membrane was stained with primary antibody, either anti-Top1 (C21, 1:1000, BD Biosciences Pharmingen, Franklin Lakes, NJ, USA) or anti-actin (1:5000, ab3280, Abcam, Cambridge, UK) (loading control), by incubation overnight at 4°C. Membrane was washed three times, and incubated with HRP-conjugated secondary antibody (1:10000, sheep anti-mouse, GE Healthcare, Little Chalfont, UK) for 1 h at 30°C. Subsequently, the membrane was washed three times and substrate (SuperSignal West Pico Chemiluminescent Substrate, Pierce, Thermo Scientific) was added according to the manufacturer's instructions. Signal was developed in the dark room using a photographic film. Western blots were replicated.

Top1 ELISA assay
Cells were trypsinized, resuspended and washed in cold PBS and pellet frozen at -80°C. The Top1 ELISA was performed as previously described [39]. Cell pellets were sonicated in lysis buffer and protein concentration was determined by BCA assay. Mouse anti-Top1 monoclonal antibody clone C21.1 (BD Biosciences Pharmingen, 1:1000), was used as the capture antibody. Pure rTop1 (EMD Biosciences, Inc.) was used as to make the standards. Samples and standards were diluted in PBS-casein and incubated overnight at 2°C to 8°C. Rabbit anti-Top1 polyclonal antibody Ab28432 (Abcam, 1:500 in PBS-casein) was used as the probe followed by the addition of extra serumabsorbed goat-anti-rabbit horseradish peroxidase conjugate (KPL, 1:1000). Probe antibody and HRP-conjugate were pre-incubated with mouse serum (Sigma Aldrich, 1:1000) to lower background signal. Finally, Pico-ELISA substrate (Thermo Scientific Pierce) was added and chemiluminescence was measured on an Infinite 200 M (Tecan Group Ltd.). Top1 levels were normalized to 1 μg protein load.

Top1 enzyme activity assay
Cells were trypsinized, resuspended and counted, and for each cell line 1 million cells were pipetted to each of three eppendorf tubes on ice. Cells were pelleted (5 min centrifugation, 300 g, 4°C) and snap-frozen in dry ice and ethanol and stored at -80°C until analysis. Nuclear extracts were prepared essentially as previously described [40], and Top1 activity measured in titration experiments with or without added SN-38 (at concentrations stated in the text) using the standard Rolling circle Enhanced Enzyme Activity Detection (REEAD) protocol as previously described [40,41]. The activity was calculated in terms of numbers of Top1 specific signals relative to the amount of signals resulting from the addition a known concentration of control circles, as previously described [40].
Drug treatment and detection of Top1-DNA cleavage complexes using alkaline elution Top1-DNA cleavage complexes (DPCs) were detected using alkaline elution as previously described [42]. In brief, cells seeded in flasks were radiolabeled overnight with 0.02 μCi/ml [ 14 C]thymidine and chased with radioisotope-free medium 4 h before drug treatment. Cells were treated with either SN-38 (10, 1 or 0.1 μM) or NSC 743400 (1 μM) for 1 h. Untreated cells were included as controls. Cells were harvested by scraping and quickly pipetted to tubes on ice. Cell aliquots were placed in ice-cold PBS and irradiated with 3000 rad to break the DNA. Cells were layered onto polyvinylchloride-acrylic copolymer (protein adsorbing) filters and lysed with LS-10 (2 M NaCl, 0.2 % sarkosyl, and 0.04 M disodium EDTA, pH 10). DNA was eluted from the filters with tetrapropylammonium hydroxide-EDTA (pH 12.15). After elution, filters were incubated for 1 h at 65°C with 1 M HCl and an additional hour at room temperature in the presence of 0.4 M NaOH. Radioactivity in fractions and filters was measured with a liquid scintillation analyzer (2200A Tri Carb Scintillation Analyzer, Packard Instruments, Meridien, USA) and the fraction of DNA retained on the filter at each time point was calculated. The results were converted to rad-equivalent (a measure of DNAprotein crosslinks). Experiments were done twice each with two technical replicates.

Statistical analyses
mRNA expression values in triplicate was compared between corresponding parental and resistant cell lines using unpaired Student's t-tests. Formation of DNA crosslinks in different cell lines following drug treatment (either SN-38 or NSC 743400; 1 μM) was compared using paired Student's t-tests. A p-value less than 0.05 was considered significant.

Bioinformatics analysis of identified TOP1 mutations
The evolutionary conservation of selected amino acids in the Top1 protein was analyzed across seven different species (human, rhesus macaque, mouse, cow, frog, zebrafish and arabidopsis) using multiple sequence alignment (Clustal Omega tool, EMBL-EBI, www.ebi.ac.uk).

Results
Colon cancer cell lines resistant to SN-38 SN-38 resistant human colon cancer cell lines were generated from the three human colon cancer cell lines HCT116, HT29 and LoVo by approximately 9-10 months exposure to increasing concentrations of SN-38 [36]. Resistance to SN-38 in these cell lines were 67-, 55-and 20-fold, respectively, when comparing IC50-values for HCT116-SN38, HT29-SN38 and LoVo-SN38 to IC50values of their the corresponding parental cell lines HCT116-Wt, HT29-Wt and LoVo-Wt [36].

TOP1 gene copy number aberrations in SN-38 resistant cells identified by FISH
TOP1 gene copy number was determined using a TOP1/ CEN-20 FISH probe combination, where CEN-20 is used as a reference (TOP1 is located at 20q12-q13.1), in both the parental and resistant cells (see Table 1 and Fig. 1). LoVo-Wt and HCT116-Wt, both had two normal copies of chromosome 20, generating two TOP1 and CEN-20 signals. In HCT116-SN38, 46 % of scored metaphases harbored two normal copies of chromosome 20, whereas in the remaining 54 %, an additional TOP1 signal was observed on a chromosome that did not contain a CEN-20 signal. No aberrations were detected in the LoVo-SN38 cell line. In HT29-Wt, three normal copies of chromosome 20 were detected, along with four TOP1 signals surrounding CEN-20, indicative of 20q isochromosome formation. A similar pattern was observed in HT29-SN38; however only two copies of chromosome 20 were detected, indicating a chromosome 20 loss. Additionally, a CEN-20 signal was detected in a chromosome without TOP1 signal. It should be noted that this signal was observed outside of the centromere on this unknown chromosome (see Fig. 1).

TOP1 sequencing reveals a yet unreported pair of highly expressed mutations in SN-38 resistant cells
The DNA sequence of the full coding region of TOP1 in parental and SN-38 resistant cells were obtained by PCR amplification using six primer sets followed by bidirectional sequencing of the PCR products. Sequences of corresponding parental and resistant cells were compared to identify acquired mutations in TOP1 (see Table 2). No mutations were detected in HT29-SN38 or LoVo-SN38. In HCT116-SN38 two non-synonymous, heterozygous mutations were detected, namely c.1336C > T (corresponding to R364K, arginine to lysine change) and c.2395G > A (corresponding to G717R, glycine to arginine change) (see Fig. 2a for sequencing chromatograms). These mutations were confirmed by re-sequencing. Multiple sequence alignment demonstrated the mutations to correspond to evolutionarily conserved residues (data not shown). R364 was conserved in seven of seven analyzed species, while G717 was conserved in six of seven species (all except zebrafish; S717).
The identified TOP1 mutations were validated using PCR primers designed to be specific for wild-type (wt) or mutant (mt) sequences at amino acid residue positions 364 and 717 in TOP1. A fragment of TOP1 was amplified by PCR using combinations of primers and cDNA (made from mRNA) from parental or resistant cells (see Fig. 2b). Parental HCT116-Wt cells highly expressed wild-type TOP1 (using wild-type primers at both positions 364 and 717), while no fragment was amplified using the combination of mutant primers. A relatively weak band was seen using 364wt-717mt primers, which might represent unspecific primer binding or a weak expression of mutant G717R TOP1 in parental cells. On the other hand, SN-38-resistant HCT116-SN38 cells only weakly expressed wild-type TOP1, while 364mt-717mt and the wt-mt combinations of TOP1 were highly expressed. Single-mutant (wt-mt or mt-wt combinations of primers) TOP1 appeared to be more highly expressed than double-mutant (mt-mt) TOP1.
Top1 mRNA and protein expression levels show no difference between parental and SN-38 resistant cells TOP1 mRNA expression was determined by microarray analysis [36] (see Fig. 3a). No significant difference in mRNA level between corresponding parental and resistant cells was observed. In addition, Top1 protein expression levels, determined by Western blotting and ELISA (see Fig. 3b and c) were similar in the corresponding parental and SN-38 resistant cells.

Top1 enzyme activity in the presence of drug is highly affected in SN-38 resistant cells
The enzymatic activity of Top1 in nuclear extracts from parental and resistant cells was determined by the Rolling circle Enhanced Enzyme Activity (REEAD) assay previously described by Stougaard et al. [41] and Andersen et al. [40]. Top1 activity in the absence of drug determined at different dilutions of nuclear extract was found to be similar or slightly lower in the SN-38 resistant and corresponding parental cells (see Fig. 4a). Next, Top1 activity was measured in the presence of increasing concentrations of SN-38 (see Fig. 4b). In the parental cells, Top1 activity diminished with increasing concentrations of SN-38, while the Top1 activity in the SN-38 resistant cells was largely maintained with increasing SN-38 concentrations.    Table 3, including IC50-values for SN-38 for comparison. HCT116-SN38 displayed a strong resistance to NSC 725776 and NSC 743400, but was sensitive to epirubicin and etoposide. HT29-SN38 showed a moderate or strong resistance to NSC 725776 and epirubicin, respectively, but was relatively sensitive to NSC 743400 and etoposide. LoVo-SN38 displayed a moderate resistance to NSC 725776 and epirubicin, but was sensitive to NSC 743400 and etoposide.

Top1-DNA cleavage complexes formed after drug treatment is altered in SN-38 resistant cells
Parental and SN-38 resistant cells were treated with SN-38 or NSC 743400, and Top1-DNA cleavage complexes were assessed by alkaline elution (see Fig. 6). Firstly, cells were exposed for 1 h to SN-38. All three parental cells displayed a significantly higher number of Top1-DNA cleavage     complexes than their corresponding SN-38 resistant cells, in line with the drug sensitivity data. Secondly, cells were exposed for 1 h to NSC 743400, as this drug showed strong cross-resistance in HCT116-SN38, but not in HT29-SN38 and LoVo-SN38. In line with this, significantly fewer cleavage complexes were formed in HCT116-SN38 cells compared to HCT116-Wt cells, whereas cleavage complexes were comparable in parental and resistant HT29 and LoVo cell lines.

Discussion
As Top1 is the sole target of camptothecins [8,43], Top1 alterations play a critical role in mediating resistance to irinotecan and as a possible predictive biomarker of response to irinotecan in mCRC. In the present study, we characterized the Top1 status in three human colon cancer cell lines with acquired resistance to the irinotecan's metabolite SN-38, investigating Top1 gene copy number, genetic sequence, mRNA and protein expression, enzyme activity and formation of Top1-DNA cleavage complexes following drug treatment of cells. In addition, we investigated the cross-resistance to two non-camptothecin Top1-targeting drugs and two Top2targeting drugs. Compared to their sensitive parental counterparts, the SN-38 resistant cells did not show noticeable changes in the expression level of Top1, neither looking at mRNA or protein expression. There appeared to be a relatively good correlation between Top1 protein expression measured by Western blotting and ELISA. Furthermore, we found the Top1 activity in SN-38-resistant cells, in the absence of any drug treatment, to be either unchanged or slightly reduced compared to parental cells. In previous studies, camptothecin response has been associated with high Top1 protein expression [16,24,26] or high Top1 enzyme activity [15,[23][24][25]. Moreover, two studies found a high Top1 mRNA expression to be associated with camptothecin sensitivity [25,44], while two others did not [15,45]. Looking at the TOP1 gene copy number in the SN-38 resistant cells using FISH, we identified changes in two of the three resistant cell lines (HCT116-SN38 and HT29-SN38). The first was a TOP1 gene copy gain in a subpopulation of cells (about half of the cells) independently of centromere-20, a marker for the chromosome bearing TOP1 (HCT116-SN38), while the other was a loss of one copy of chromosome 20 (including one copy of TOP1) (HT29-SN38). Two cell line studies by McLeod and Keith [21] and Romer et al. [17] demonstrated a positive correlation between TOP1 gene copy number and sensitivity to SN-38 and between gene copy number and protein expression, respectively. The positive association between TOP1 amplification and expression of Top1 mRNA and protein was confirmed by another study [22], and Top1 mRNA and protein expression have been shown to correlate well in the NCI-60 cancer cell line panel [14].
Using primers covering the full coding region of TOP1 we did a mutational analysis of the gene in the SN-38 resistant and parental cells. Two of three resistant cell lines did not harbor any mutations, while the third (HCT116-SN38) harbored two mutations. Both identified mutations were non-synonymous and heterozygous; one located at amino acid residue position 364 (c.1336C > T, R364K, arginine to lysine change) in the Top1 core domain and the other at position 717 (c.2395G > A, G717R, glycine to arginine change) in the Cterminal domain [27,46]. Both were shown by multiple sequence alignment to be located at highly evolutionarily conserved positions in the Top1 protein. Using TOP1 wildtype and mutant specific primers, we showed by PCR that the mutated cells (HCT116-SN38) expressed mutant TOP1 mRNA at a markedly higher level than wild-type TOP1 mRNA, even though the sequencing analysis suggested heterozygosity of the mutations. TOP1 containing a single mutation (R364K or G717R) appeared to be expressed at a higher level than double-mutant TOP1, suggesting that the mutations are present on separate alleles or in separate subpopulations in the HCT116-SN38 cells, each of them conferring resistance. This is in accordance with the FISH analysis, which detected two subpopulations in the HCT116-SN38 cell line. In line with these findings, we demonstrated that Top1 activity in nuclear extract from the HCT116-SN38 cell line was largely unaffected in the presence of high doses of SN-38, which completely eliminated the Top1 activity in parental cells, while the Top1 activity in drug absence was the same in HCT116-SN38 and parental cells. Furthermore, we measured the degree of formation of Top1-DNA cleavage complexes by alkaline elution following SN-38 treatment of cells, and showed that far fewer complexes were formed in HCT116-SN38 cells compared to parental cells. These findings suggest that SN-38 binding to Top1-DNA is hindered in the mutationharboring cells. In previous studies, identified TOP1 mutations have clustered in regions close to the structural site where Top1 binds DNA and camptothecin, i.e. the regions 361-364 (DNA minor groove), 503-533 (minor groove) and 717-729 (major groove) [11,47]. Some mutations have been shown to hinder binding of drug, while others destabilize the drug-Top1-DNA cleavage complex or enhance Top1 DNA religation. Other mutations have also been identified in relationship to the linker region of Top1 [38,48]. In a study by Li et al. [49] amino acids in the 361-364 region was demonstrated to be involved both in enzyme catalysis and camptothecin resistance. More specifically, experimental substitution of amino acids in the 361-364 range, i.e. R362L (arginine to leucine) and R364G (arginine to glycine) was shown to affect the catalytic activity of Top1, however R364G only slightly reduced the activity compared to wild-type enzyme [49]. Furthermore, Li et al. [49] showed that R364G Top1 was able to bind DNA with the same affinity as wild-type enzyme, however camptothecin was largely unable to bind R364G Top1-DNA cleavage complexes and cause DNA breaks. In addition, a R364H (arginine to histidine change) mutation was previously described in two camptothecin resistant prostate cancer cell lines [50]. The R364H mutation did not affect the catalytic activity, but rendered the cells resistant to camptothecin [50]. As lysine is a large positively charged amino acid, similar to histidine, it is very likely that R364K Top1 is functionally similar to R364H Top1. Several other mutations associated with camptothecin resistance have previously been reported in the amino acid region 361-365 of Top1 [51][52][53][54][55]. The C-terminal domain of Top1 is known to be involved in both enzyme catalytic activity and drug binding [27], and several camptothecin resistance-associated mutations have been reported in the 717-737 region [56][57][58][59][60][61][62], including a pair of mutations in a tumor sample from a cisplatin/irinotecan treated lung cancer patient [29]. A camptothecin-associated mutation at position 717 (G717V, glycine to valine change) has previously been reported together with the mutation T729I [56]. The authors showed that the mutation-harboring cells displayed similar Top1 catalytic activity as wild-type cells, and that each mutation on its own rendered yeast cells resistant to camptothecin [56]. Losasso et al. [62] investigated various amino acid substitutions in the 729 position, and suggested that this position is part of a hydrophobic pocket important for drug sensitivity [62]. Recently, mutations have also been reported in the linker region (amino acid residues 636-712), between the core and C-terminal domains of Top1 [38,48]. Losasso et al. [48] investigated a resistance-associated mutation at position 653 and suggested that altered Top1 linker flexibility is a likely mechanism of resistance [48]. Gongora et al. [38] have subsequently identified other mutations in the linker region, which could confer resistance by this mechanism [38]. One of the linker region mutations found by Gongora et al. (Glu710Gly) was further analyzed in S. cerevisiae and the data indicated that a fully functional linker region of Top1 is important to confer camtotethecin sensitivity [63]. We did not detect any TOP1 mutations in the two other SN-38 resistant cell lines (HT29-SN38 and LoVo-SN38). However, even in these cells Top1 activity was unaffected by the presence of large concentrations of SN-38, and thus other mechanisms must be responsible for this finding.
Lastly, we assessed the three SN-38 resistant cell lines for cross-resistance to two non-camptothecin Top1-targeting drugs in clinical trials (indenoisoquinolines: NSC 725776/ LMP776/indimitecan and NSC 743400/LMP400/indotecan) and two clinical Top2-targeting drugs (epirubicin and etoposide). All three SN-38 resistant cell lines displayed no or very little cross-resistance to etoposide, a specific Top2 inhibitor [12,13,46], while showing more but still only partly cross-resistance to epirubicin, a DNA intercalating Top2 inhibitor [64]. Previous studies have demonstrated that reduced activity of Top1 can be compensated by increased activity of Top2 and thus increased sensitivity to etoposide [65,66], and furthermore that camptothecinresistant cells retain sensitivity to Top2-targeting drugs [50]. In the three SN-38 resistant cell lines, the indenoisoquinolines displayed an interesting pattern of resistance, from full cross-resistance to no cross-resistance. The two indenoisoquinolines are currently in clinical development [67]. In the present study, the cell line harboring the R364K-G717R mutations (HCT116-SN38) showed very strong cross-resistance to LMP400 and LMP776. The two SN-38 resistant cell lines, which did not carry any mutations in TOP1 (HT29-SN38 and LoVo-SN38), were partly cross-resistant to LMP776, while showing either no or only small cross-resistance to LMP400. Studies in our laboratory [36] showed that HT29-SN38 and LoVo-SN38 both strongly upregulated expression (mRNA; 25-and 60-fold, respectively) of the well-known drug-efflux pump ABCG2 (BCRP) [68,69], while HCT116-SN38 did not (data not shown). This suggests that LMP400 can remain active in cancers resistant to SN-38, which display upregulation of this multidrug resistance protein, if TOP1 is wild-type. This is in line with a previous study by Antony et al. [35], where LMP776, but not LMP400, was shown to be a weak substrate of the ABCG2 pump. LMP400 was thus effective in cells overexpressing ABCG2, which displayed a 46-fold resistance to SN-38 [35]. Our experiment measuring Top1-DNA cleavage complexes following treatment with LMP400 supported the cross-resistance findings. These results highlight how the underlying molecular mechanism of camptothecin resistance in cancer cells determines their resistance-profile to new classes of drug such as the indenoisoquinolines.

Conclusions
We generated three SN-38-resistant human colon cancer cell lines and investigated Top1. We detected no difference in the expression level of Top1 and no to very little reduction in Top1 activity in the absence of drug. A markedly increased activity of Top1 in the presence of SN-38 was seen in all three resistant cell lines. TOP1 gene aberrations were detected in two of three cell lines, and a not previously reported pair of mutations in TOP1 was identified in one cell line. The SN-38 resistant cells displayed an interesting pattern of cross-resistance to two indenoisoquinoline Top1-targeting drugs: SN-38 resistant cells with mutant TOP1 and no overexpression of drug-efflux pump ABCG2 were resistant to LMP400, while SN-38 resistant cells with wild-type TOP1 and overexpression of ABCG2 remained sensitive to LMP400. Furthermore, cross-resistance to Top2-targeting drugs was not existent or limited. Thus, this study adds to the growing knowledge about anti-cancer resistance mechanisms for camptothecins and the new class of indenoisoquinoline Top1-targeting drugs.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions NFJ established the SN-38 resistant cell lines, designed and carried out the studies, drafted the manuscript. KA participated in data acquisition and interpretation related to the alkaline elution assays. AR carried out the Top1 enzyme activity assay and data interpretation. DHS carried out the metaphase preparation and analyses. TDP and JHD established and assisted in the ELISA assays and data interpretation. MUR designed the MTT assay and helped in data analyses. HLZ designed primers for validation of mutations and interpretation of PCR results. BRK designed the Top1 enzyme activity assay and participated in interpretation of results. JS, NB and YP conceived the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.