TROP2 expression in tumors from LNCaP xenografts treated with flutamide and docetaxel

In view of the poorer prognosis exhibited by TROP2-high low-grade tumors, we aimed to determine whether TROP2 expression is associated with chemoresistance in androgen-sensitive tumor cells. To do so androgen-sensitive LNCaP cells were xenografted into NOD/SCID mice, and recurrence after treatment with docetaxel chemotherapy (10 mg/kg/week) and/or with the anti-androgen compound flutamide (33 mg/kg/day) was investigated. Treatment commenced when tumors reached 100mm³ and tumor volume was measured to determine the growth rate of tumors under different treatment conditions. Efficacy of docetaxel and flutamide treatments was corroborated by their respective impact on testes and prostate weight compared to control mice (Supplementary Figure S2).

Flutamide treatment did not induce a regression of tumor volume but slowed down tumor growth (day 0: 114.8 ± 2.5mm³; day 49: 603.5 ± 60.1mm³) compared to the control group (day 0: 108.5±0.4 mm³, day 49: 1438.0 ± 609.7mm³) (P = 0.0009) (Figure ​(Figure2A).2A). Docetaxel alone or in combination with flutamide decreased tumor volume (Figure ​(Figure2B),2B), and post-treatment tumor recurrence was delayed in mice treated with docetaxel alone compared to those who received a combination of flutamide and docetaxel (P = 0.0406). At day 21 (end of docetaxel treatment), mice from both groups had similar tumor volumes (docetaxel: 45.6 ± 13.9mm³; combined: 51.8 ± 13.5mm³ P = 0.3512), but after docetaxel treatment had finished, tumor volume in mice treated with docetaxel alone continued to decrease (day 35: 24.9 ± 5.0mm³) whereas mice from the combination treatment group maintained a steady tumor size (day 35: 52.0 ± 18.6mm³) (P = 0.0454). RT-qPCR was used to quantify the expression of TROP2 mRNA as well as that of putative markers for basal/cancer stem cells (CD49f, POU5F1 (Oct4), DCLK1, ALDH1A3) or luminal stem cells (NKX3.1) [17] in residual tumors (Figure ​(Figure2C).2C). Expression of TROP2 and DCLK1 mRNAs was robustly up-regulated in tumors following relapse from the combination treatment of flutamide and docetaxel, while recurring tumors showed no change in CD49f, Oct4, NKX3.1 or ALDH1A3 (Figure ​(Figure2C).2C). High extracellular TROP2 expression was also detected by immunohistochemistry in tumors that underwent combination treatment (Figure ​(Figure2D).2D). These findings demonstrate that TROP2 is enriched in tumor cells that recur after treatment of androgen-sensitive xenografts with docetaxel/androgen ablation combination therapy.

Open in a separate window

Figure 2

A–B. Tumor growth in NOD/SCID mice treated with vehicle (control), 33mg/kg/day flutamide (FLT), 10mg/kg/week docetaxel (DTX) or a combination of docetaxel and flutamide (DTX + FLT), as indicated. Data was analyzed by two-way repeated measures ANOVA and a P value < 0.05 for the interaction between treatments and time was considered to be statistically significant (n = 5 mice/group). C. mRNA expression levels in prostate tumors derived from LNCaP cell xenografts treated as above. *, P<0.05 and ****, P<0.0001, significantly different from control and flutamide treatment groups. ##, P<0.01 and ####, P<0.0001, significantly different from docetaxel treatment group. D. Immunohistochemical staining of TROP2 on tumors derived from xenografts treated as indicated in (A) and (B). Scale bar=100μm.

Androgen-sensitive TROP2^high prostate cancer cells display higher resistance to high dose chemotherapy and androgen deprivation

Because Tumor-Initiating Cells (TICs) are thought to play a role in enhanced treatment resistance, we then investigated whether cells demonstrating high levels of membrane TROP2 expression exhibit TIC properties in androgen-sensitive human prostate cancer cells. LNCaP cells that highly expressed extracellular TROP2 did not stand out as a separate population. Instead, cells presented with a continuum of low to high extracellular TROP2 levels, and the top 2% of cells with the brightest fluorescence after labeling with a selective antibody were chosen as TROP2^high cells (Figure ​(Figure3A).3A). Baseline growth was similar in TROP2^high, TROP2^low and ungated LNCaP cells (Supplementary Figure S2). Extreme Limiting Dilution Analysis showed that there was no difference in sphere-forming efficiency (Table ​(Table11 and Supplementary Figure S2) between TROP2^high, TROP2^low and ungated LNCaP cells. This result indicates that TROP2^high cells do not display enhanced self-renewal abilities compared to other cell populations, despite the mRNA enrichment for the putative tumor-initiating cell marker ALDH1A3 in TROP2^high cells (Figure ​(Figure3B).3B). Similarly, no significant difference was detected between the sphere-forming efficiency of TROP2^high and TROP2^low and ungated cells from another androgen-sensitive cell line, 22Rv1, as determined using ELDA (Table ​(Table11 and Supplementary Figure S2). These results suggest that High TROP2 expression does not selectively enrich pre-existing cell populations with high self-renewal ability in androgen-sensitive prostate cancer cells.

Open in a separate window

Figure 3

A. FACS profile of LNCaP cells according to their surface TROP2 expression level. Black line: cells were stained mouse isotype control; red line: cells were stained with TROP2 antibody. B. RT-qPCR quantitation of TACSTD2 (TROP2), POU5F1 (Oct4), NKX3.1, and ALDH1A3 mRNA expression. DCLK1 was undetectable in these cells C. Percentage of viable TROP2^high, TROP2^low and ungated LNCaP cells at the end of a 5-day treatment with docetaxel alone (DTX, top) or in combination with flutamide (DTX + FLT, bottom). Data represents the mean percentage of surviving cells after exposure to the three highest concentrations of docetaxel (1nM, 10nM, 100nM). D. Percentage of viable TROP2^high, TROP2^low and ungated LNCaP cells after a 5-day treatment with docetaxel alone (DTX, top) or in combination with flutamide (DTX + FLT, bottom) followed by a 7-day recovery phase in docetaxel-free medium. Data represents the mean percentage of surviving cells after exposure to the three highest concentrations of docetaxel (1nM, 10nM, 100nM). In C and D: *, P < 0.05, **, P < 0.01, significantly different from TROP2^low cells, one-way ANOVA with Bonferroni post-hoc, n=3.

In order to analyze their response to therapeutic agents in vitro, TROP2^high, TROP2^low and ungated LNCaP cells were sorted by FACS and treated for 5 days with docetaxel alone or in combination with 10μM flutamide. Cell viability was assessed immediately after treatment cessation (Figure ​(Figure3C)3C) or 7 days after withdrawal of docetaxel (Figure ​(Figure3D3D).

The number of residual live cells was significantly lower after a 7-day treatment withdrawal compared to straight after treatment cessation regardless of the cell population (Figure 3C–3D). Since docetaxel was the only compound present under all conditions this indicates that docetaxel-induced cell death continued well after the removal of this compound from the culture medium.

Dose-response curve analysis indicated that docetaxel decreased the viability of TROP2^high, TROP^low and ungated LNCaP cells with similar IC₅₀ (Supplementary Figure S3), suggesting that cytotoxic treatment kills these different cell populations with a similar potency.

However, when we analyzed the maximal effect of docetaxel (defined here as the mean percentage of viable cells that remain after exposure to the three highest concentrations of docetaxel – 1nM, 10nM and 100nM), we found that a significantly higher percentage of TROP2^high cells remained alive compared to TROP2^low cells, both a the end of the high dose docetaxel treatment (DTX) or following a 7-day relapse (Figure 3C–3D). Our finding demonstrates that docetaxel kills TROP2^high cells with a lower maximal efficacy compared to TROP2^low cells. Similar results were also found for androgen-sensitive 22Rv1 cells (Supplementary Figures S3 and S4).

When androgen-sensitive cells were incubated with docetaxel in the presence of flutamide (DTX + FLT), TROP2^high 22Rv1 cells were again able to recover more efficiently from exposure to high docetaxel concentrations (Supplementary Figure S4), while this was not detected in LNCaP cells (Figure ​(Figure3D),3D), where the recovery of all cell population was similar after a 7-day treatment withdrawal (Figure ​(Figure3D3D).

Enhanced self-renewal in TROP2^high cells from androgen-independent prostate cancer cell lines

To determine whether high TROP2 expression also characterizes a sub-population of cells with enhanced resistance to high dose chemotherapy in androgen-independent cells, we then characterized the TROP2^high population in the PC3 and DU145 cell lines. Approximately 2.7% of PC3 cells were found to highly express extracellular TROP2 (Figure ​(Figure4).4). TROP2 mRNA was not enriched in TROP2^high PC3 cells compared to other populations (Figure ​(Figure4B),4B), but we found that the amount of TROP2 targeted to the extracellular membrane was indeed much higher in TROP2^high cells than in any other populations (Figure ​(Figure4C).4C). Quantification using ELDA indicated that TROP2^high PC3 cells had higher spheroid forming capacity than ungated cells and TROP2^low cells (Table ​(Table22 and Supplementary Figure S5). Similar results were obtained with TROP2^high cells isolated from the androgen-independent cell line DU145 (Table ​(Table22 and Supplementary Figure S5). In addition, TROP2^high PC3 cells showed high mRNA levels of cancer stem cell markers such as OCT4, ALDH1A3 and DCLK1.

Open in a separate window

Figure 4

A. FACS profile of PC3 cells according to their surface TROP2 expression level. Black line: cells were stained mouse isotype control; red line: cells were stained with TROP2 antibody. B. RT-qPCR quantitation of TACSTD2 (TROP2), POU5F1 (Oct4), NKX3.1, ALDH1A3 and DCLK1 mRNA expression. C. IF staining for TROP2 on PC3 cells sorted as in (A). Scale bar = 50μM. D. Percentage of viable TROP2^high, TROP2^low and ungated PC3 cells at the end of a 5-day treatment with docetaxel (top) or following a 7-day recovery phase in docetaxel-free medium (bottom). Data represents the mean percentage of surviving cells after exposure to the three highest concentrations of docetaxel (1nM, 10nM, 100nM). E. Percentage of viable TROP2^high, TROP2^low and ungated PC3 cells at the end of a 5-day treatment with docetaxel (left) or following a 7-day recovery phase in docetaxel-free medium (right). Data represents the mean percentage of surviving cells after exposure to the three highest concentrations of docetaxel (1nM, 10nM, 100nM). In D and E: *, P < 0.05, ***, P < 0.001, one-way ANOVA with Bonferroni post-hoc, n = 3);

Since the expression of TROP2 was previously shown to overlap with ALDH activity - a putative TIC marker of clinical relevance [18], we performed additional ELDAs to assess whether the high self-renewal ability of TROP2^high PC3 cells was associated with their ALDH activity. However, we found that extracellular TROP2 expression predicted the spheroid-forming frequency of PC3 sub-populations irrespective of their ALDH activity. Indeed, TROP2^highALDH^low and TROP2^highALDH^high cells showed a similarly high sphere-forming frequency, indicating that their self-renewal ability is 5 to 10 times higher compared to all other cell sub-populations within this cell line (Table ​(Table3).3). This suggests that TROP2 expression is a better predictor of tumor-initiating properties than ALDH in these androgen-independent cells.

To investigate whether TROP2 expression is also associated with differential chemoresistance abilities in androgen-independent cells, we exposed PC3 and DU145 cells to docetaxel for 5 days, followed or not by a 7-day post-treatment withdrawal. Similar to what was found for androgen-sensitive cells, significantly lower numbers of viable cells were detected after post-treatment withdrawal compared to right after treatment cessation (Figure ​(Figure4).4). No significant survival difference in cell viability between TROP2^high and other cell populations was observed immediately after a 5-day treatment with the chemotherapeutic drug docetaxel for either of the two cell lines (Figure 4D–4E and Supplementary Figure S6–S7). Seven days after withdrawal of this compound, the proportion of surviving TROP2^low PC3 cells was lower than those in ungated and TROP2^high PC3 cells (Figure ​(Figure4D4D and Supplementary Figure S6), but this was not the case for DU145 cells (Figure ​(Figure4E4E and Supplementary Figure S7), suggesting that this may not be an overall characteristic of androgen-independent TROP2^high cells. Taken together, these results indicate that the TROP2^high population displays a stronger self-renewal potential than other cells in androgen-independent prostate cancer cells, but does not identify a cell sub-population with enhanced chemo-resistance ability.

Tumor xenografts

Slide-mounted cryo-cut tumor sections were fixed with 4% paraformaldehyde for 10 minutes and stained with mouse anti-human TROP2 antibody (BD Bioscience 551317, 1/200) overnight at 4°C followed by incubation with fluorescein horse anti-mouse IgG (Vector laboratories, 1/200) for 1 hour at room temperature. Hoechst 33342 (Invitrogen) was used to stain nuclei. Fluorescent images were taken with a Nikon digital camera attached to a Nikon A1R confocal microscope (Nikon).

SurvExpress data analysis

The correlation between TROP2 expression and disease outcome was analyzed using the SurvExpress bioresource [14] (http://bioinformatica.mty.itesm.mx:8080/Biomatec/SurvivaX.jsp). Selected datasets were originally from the Gene Expression Omnibus (GSE16560, 281 patients [16], and GSE40272, 98 patients [15]). Selection criteria included minimal patient number per study (>80), presence of TACSTD2 (TROP2)-specific probes, and availability of survival and/or recurrence-free survival data. Presented data summarizes TROP2 expression levels (median, 5-95 percentile) and Kaplan-Meier analysis (Hazard ratio) of survival or recurrence-free survival as indicated in high and low risk patient subgroups. Optimized risk scores were determined as described in [14].

LNCaP xenografts

NOD/SCID mice were obtained from the Monash Animal Research Platform and ethical approval for in vivo experiments was obtained from the Monash University Standing Committee on Ethics in Animal Experimentation (Ethics numbers: VCPA.2010.33 and MIPS.2012.15).

LNCaP cells (1×10⁶) were injected subcutaneously into the left flank of NOD/SCID mice and treatments were initiated once tumors reached a volume of 100mm³. The timeline for the experiment is provided in Supplementary Figure S1, with the different treatment groups as follows: The control group received a slow-releasing pellet containing placebo and an intra-venous injection of vehicle; the flutamide group (chemical androgen ablation) received 33mg/kg/day flutamide and an intra-venous injection of vehicle once a week for four weeks; the docetaxel chemotherapy group received a placebo pellet and an intra-venous injection of 10mg/kg docetaxel once a week for four weeks; The Combined group (flutamide + docetaxel) received a combination of the flutamide and docetaxel doses indicated above. Flutamide treatment was performed by implanting a slow-release flutamide pellet on the lateral side of the neck of the mouse. An incision equal in diameter to that of the pellet (approximately 6mm) was made using sterile surgical scissors. A pocket horizontal to the cut was made using blunt forceps about 2cm beyond the incision site. A slow-releasing flutamide pellet (100mg for 90-days releasing 33mg/kg/day, Innovative Research of America) or placebo pellet (Innovative Research of America) was placed into the pocket with forceps and angled to the right on the lateral side of the neck between the ear and shoulder where there is maximal space between the skin and muscle. The incision site was then sealed with 9mm BBL™ AUTOCLIP wound clips (BD Diagnostics).

Stock solutions of 20mg/ml docetaxel (Tocris) were prepared in 100% ethanol. A final dilution was made just prior to injection in a vehicle of Tween 80 (Sigma) and 5% dextrose (5/5/90 vol/vol/vol).

Mice treated with docetaxel or combination therapy were culled when tumor relapse had occurred, defined by an increase in tumor size in two consecutive measurements following the end of treatment. Due to slight variations in treatment response among individual mice, individual animals were culled at slightly different time points. In control and androgen ablation groups, mice were culled 49 days after the beginning of treatment when the first mouse in the control group had to be culled when tumors had grown to a diameter of >12mm, as per ethical requirements (n=5).

Cell culture

Human prostate cancer cell lines PC3 (CRL-1435) and LNCaP (CRL-1740), DU145 (HTB-81) were obtained from American Type Culture Collection (ATCC) and cultured in Ham's F-12K (Invitrogen) or RPMI medium 1640 (Invitrogen). The 22Rv1 cell line (CRL-2505) was obtained from Prof. Martin Lackmann (Department of Biochemistry, Monash University, Melbourne, Australia). Media for all cell lines was supplemented with 10% Fetal Bovine Serum (FBS) (Invitrogen 10100-147). The usual androgen concentration within FBS (0.1–1 nM Di Hydro Testosterone) is sufficient to efficiently maintain androgen-sensitive cells such as LNCAP and 22RV1 [36].

For spheroid culture, cells were grown in serum-free DMEM/F-12 medium (Invitrogen 10565) supplemented with 1% N2 (Invitrogen 17502-048), 2% B27 (Invitrogen 17504-044), 20 ng/mL human epidermal growth factor (EGF Invitrogen PHG0311), 20ng/mL human fibroblast growth factor-basic (FGF-b Invitrogen PHG0026) and 100 units of Penicillin-Streptomycin (Invitrogen) in ultra-low attachment plates (Costar).

FACS analysis and sorting

Cells were stained with Mouse anti-human TROP2 antibody (BD Bioscience 551317, 1/200) in antibody staining buffer (0.9% (w/v), sodium azide and 2% FBS in PBS) for 1 hour. Following three washes in PBS, Alexa Fluor® 647 Goat Anti-Mouse IgG secondary antibody (Invitrogen A21236, diluted 1 in 1000 in antibody staining buffer) was incubated for one hour at 4°C. Stained cells were then washed three times in antibody staining buffer prior to FACS sorting and analyzing.

Cell sorting was performed using a MoFlo®Astrios^TM (Beckman Coulter). The gating strategy to define negative populations for ALDH and TROP2 staining was designed so as to include at least 99.9% of cells in the DEAB and isotype negative control samples. A distinct population highly expressing TROP2 was identified and sorted from PC3 cells (~2.75% of total viable cells). The same gating strategy was applied to LNCaP, DU145 and 22Rv1 cells. Since no individualized TROP^high population was detected in these cell lines, the TROP2^high population was defined as the top 2% of TROP2-expressing cells. FACS analysis was performed using a FACSCanto™ II (BD Science).

RT-qPCR

mRNA levels for ten genes that were previously reported to act as prostate stem cell or TIC markers or that are involved in cancer progression and metastasis (CD44, ITGA6 (aka CD49f), TACSTD2 (TROP2), POU5F1 (OCT3-4), NKX3.1, CTNNB1 (beta-catenin), DCLK1, DICER, AKT, ALDH1A3) were detected by qPCR in both sorted cells and tumor tissues by RT-qPCR. RNA from both tissue and cells was extracted using the RNeasy Plus Micro kit (Qiagen) following the manufacturer's instructions. Real-time RT-qPCR was carried out with pre-designed human TAQMAN® probes (Invitrogen) with iScript™ One-Step RT-PCR Kit (Bio-Rad) in a CFX96™ Real-Time PCR Detection System (Bio-Rad)., For both PC3 and LNCaP cell lines, glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a housekeeping gene, while, the ribosomal proteins RPLP0 and RPLR30 were used for tumors derived from PC3 and LNCaP xenografts respectively.

Extreme limiting dilution assay (ELDA)

Cells were sorted and seeded in 3 different concentrations: 10 cells/well, 30 cells/well and 100 cells/well in low-attachment 96-well plates in spheroid culture medium. For each individual experiment, 120 replicates were performed for each concentration of cells. The presence or absence of spheroids (included in the count when size >50μm diameter) in each well was determined at day 14 using a conventional light microscope (Nikon). When run under these conditions, ELDA provides a robust estimate of the tumor-initiating cell frequency, calculated and analyzed using the ELDA webtool (http://bioinf.wehi.edu.au/software/elda)[37].

In vitro toxicity assays

Cells were treated with various concentrations of docetaxel for 5 days and washed three times with PBS at the end of treatment. Cells were then cultured in their ordinary culture medium without docetaxel for 7 days before being tested for cell viability. For the combination treatment, 10μM of flutamide remained in the medium after the withdrawal of docetaxel. Cell viability was assessed using the CellTiter®-Blue cell viability assay (Promega) according to the manufacturer's instructions.

siRNA transfection

Both PC3 and LNCaP cells were transfected with TROP2-target siRNA (Taqman Silencer® Select siRNA, Thermofisher, https://www.thermofisher.com/order/genome-database/details/sirna/s8366, 50nM final concentration) for 24 hours. Cells were then detached and seeded in 96-well plates 24 hours prior to treatment.

The authors are very grateful to Prof Chris Hovens and to Dr. Anne Nguyen for their invaluable support to obtain patient samples and perform immunostaining experiments. We also thank Prof. Colin Pouton for helpful discussions, Dr. Corina Kim-Fuchs for her advice on mouse surgery techniques, and Swati Irani for technical help with androgen receptor staining.