Introduction

Colorectal cancer (CRC) is one of the most common cancers worldwide, accounting for approximately 9 % of cancer-related deaths [1], mostly due to metastasis that is a result of resistance to immuno/chemotherapy [2]. Most CRC patients are diagnosed with localized disease amenable to curative surgical resection [3]. However, approximately 20 % of patients are diagnosed with distant metastases [3], highlighting the need for more advanced therapeutic regimens.

Neoadjuvant chemotherapy has assumed increasing importance in the treatment of CRC, replacing surgery as the sole therapy of choice [3]. The KRAS gene has been shown to have inverse prognostic power in predicting whether a CRC patient will respond to such neoadjuvant therapies [4]. Close to 35–45 % of patients with CRC has mutant KRAS [5–11]. Ninety-five percent of KRAS mutation occur in exons 12 and 13, with the former alone contributing to 80 % of detected mutations [4]. It has been observed that patients harboring mutant KRAS have a tendency to develop resistance to neoadjuvant chemotherapy and hence suffer from poor response rate to chemotherapy.

MicroRNAs (miRNAs), small noncoding endogenously produced RNAs, can function as both tumor suppressors and oncogenes by either causing target mRNA degradation or inhibiting their translation [12, 13]. Mir-143 has been shown to target KRAS in both CRC [14] and prostate cancer [15]. The human let-7 family of tumor suppressor miRNAs also targets KRAS [16] and has been shown to be frequently lost in human cancers [17, 18]. Even though replenishment of let-7 and miR-143 seem to be a viable treatment strategy, let-7 replenishment failed to arrest proliferation and induce cell death in KRAS mutant tumor cells, severely limiting its clinical use [19, 20]. Thus, the objective of the current study was to determine if let-7b or miR-143 replenishment would function as a chemosensitizer for conventional chemotherapeutic agent like paclitaxel (PTX) in KRAS mutant and wild-type tumor cells.

Materials and methods

Cell culture

Human colon cancer cell lines SW48 (wild-type KRAS, BRAF, PIK3CA, PTEN, and TP53) and LoVo (mutant KRAS-G13D, A14V, wild-type BRAF, PIK3CA, PTEN, and TP53) [21] were obtained from the ATCC (Manassas, VA, USA) and maintained at 37 °C in a CO2 incubator in Dulbecco’s modified Eagle’s media (DMEM) containing 10 % fetal bovine serum (FBS) and 1 % penicillin-streptomycin. This article does not contain any studies with human participants or animals performed by any of the authors.

Cell transfection and treatments

Hsa-miR-143 mimic, let-7b-5p mimic, and the non-targeting (scramble oligonucleotides) control (SCR) (Life Technologies, Beijing, China). Cells (4 × 104) were transiently transfected with 50 nM of the mimics or scramble control using Lipofectamine LTX (Life Technologies) as per the manufacturer’s guidelines. Twelve hours after transfection, the cells were subjected to treatment with indicated concentrations of PTX (Sigma-Aldrich, Beijing, China) for 36 h.

Analysis of combination effects

CompuSyn software (ComboSyn, Paramus, NJ) [21] was used to calculate the combination index (CI) between PTX and let-7b or miR-143. Synergistic, additive, and antagonistic effects were defined as CI < 0.9, CI = 0.9–1.1, or CI > 1.1, respectively [22].

RNA and miRNA extraction and qRT-PCR

TRIzol reagent was used to isolate total RNA. First-strand cDNA was synthesized using the First Strand cDNA synthesis Kit (Life Technology), which was then used for qRT-PCR using TaqMan Gene Expression probes (Life Technology). TBP (TaqMan Assay ID Hs00427620_m1) was used as an internal control for assessing KRAS (TaqMan Assay ID Hs00364284_g1) transcript level. Data was normalized to TBP expression and analyzed by the −ΔΔCt method. MiRNA from tissues and cells was extracted using the mirVana miRNA isolation kit (Life Technology), and TaqMan miRNA assays (Life Technology (TaqMan Assay IDs 002146, 002619, and 001093, respectively) were used to quantify the expression levels of hsa-miR-143, let-7b, and RNU6B. Data was normalized to RNU6B expression and analyzed by the −ΔΔCt method.

Preparation of whole cell lysates and immunoblot analysis

Cells were lysed in lysis buffer (25 mM Tris–HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 % NP-40 and 5 % glycerol) supplemented with protease inhibitor cocktail (Roche Diagnostics, Indianapolis, USA). Lysates were resolved by SDS-PAGE and probed with KRAS and PARP antibody (Abcam, Cambridge, USA). The blot was probed for β-actin (Cell Signaling, Danvers, USA) to confirm equal loading.

Cell viability assay

Cell proliferation was quantitated using a mitochondrial colorimetric assay (MTT assay, Sigma-Aldrich, St. Louis, MO) as per the manufacturer’s recommendations. The absorbance was measured at 570 nm. The results, expressed in terms of relative optical density (OD), as mean ± standard deviation.

Apoptosis assay

Apoptosis was assessed by using the Annexin V and PI staining kit (BD Biosciences, San Jose, CA, USA) as per the manufacturer’s recommendations.

Cell proliferation assay

Apoptosis was assessed by using the CFSE staining kit (LifeTechnology, Carlsbad, CA, USA) as per the manufacturer’s recommendations.

In vitro transwell migration and invasion assays

For in vitro migration assays, indicated cells treated with mitomycin C were serum starved, trypsinized, and introduced into the upper chamber (1 × 105) of the transwell (8-μm pore size). Migration and invasion analysis was done using Culturex 96-well cell migration and Culturex 96-well BME cell invasion assay kits (R&D Systems, Beijing, China), respectively. Data obtained from both sets of experiments were used to analyze percent migration and invasion and were expressed as mean ± standard error of mean (SEM).

Statistical analyses

Statistical analyses were performed using SPSS version 20.0 (IBM Corporation, NY). A P < 0.05 was considered statistically significant.

Results

MiR-143, but not let-7b, replenishment selectively enhances the chemosensitivity of KRAS mutant tumor cells

To determine the therapeutic potential of miR-143 and let-7b as a chemosensitizer, we evaluated the effect of miR-143 and let-7b replenishment on the cytotoxicity of PTX in KRAS wild-type SW48 (Fig. 1a) and mutant LoVo (Fig. 1b) cells. PTX is routinely used for treatment of solid tumors, and hence, we chose the same. Transfection of miR-143 mimic significantly chemosensitized LoVo cells, which harbor KRAS mutation, but did not affected KRAS wild-type SW48 cells (Fig. 1a). No effect with let-7b mimic was seen in either cell line (Fig. 1).

Fig. 1
figure 1

MiR-143, but not let-7b, repletion chemosensitizes KRAS mutant, but not wild-type colon cancer tumor cells. Colon cancer cells SW48 (KRAS wild-type) (a) and LoVo (KRAS mutant) (b) were transfected with either 50 nM miR-143 mimic, let-7b mimic, or a scramble control (SCR) for 12 h. The cells were then treated with paclitaxel (PTX, 0.1–50 nM) for 36 h. Cell viability was assessed by the crystal violet assay. Each experiment was carried out at least three times. Data represent mean ± SEM

Full size image

MiR-143 mimic decreased the IC50 of PTX from 23 ± 3 to 6 ± 1 nM in LoVo cells, while the reduction was less than 1 % in KRAS wild-type SW48 cells (Fig. 2). As a negative control, the scramble control minimally influenced the cytotoxicity of PTX in either cell line. To determine whether miR-143/PTX combination resulted in synergistic effect, the combination index was calculated from the proliferation data generated in each cell line. This analysis showed that the combination indices of miR-143/PTX scored 0.53 in KRAS mutant LoVo cells compared to 1.09 in SW48 cells, signifying a synergistic response in cells containing mutant KRAS (Fig. 2). Let-7b had a combination index >1 in both cell lines (Fig. 2). Cumulatively, these results indicate that miR-143 replenishment selectively sensitizes KRAS mutant cells to the cytotoxicity of PTX.

Fig. 2
figure 2

Effect of miR-143 or let-7b repletion on the cytotoxicity of paclitaxel in the KRAS mutant LoVo and wild-type SW48 cell lines. Each experiment was carried out at least three times

Full size image

MiR-143 selectively downregulates mutant KRAS expression

To understand the differential effect of miR-143 and let-7b on the wild-type versus mutant KRAS tumor cells in response to PTX, we first compared the endogenous levels of miR-143 and let-7b and KRAS in these two cell lines. The expression of miR-143 was 25-fold higher in SW48 compared to LoVo cells (Fig. 3a). Conversely, KRAS mRNA (Fig. 3b) and protein (Fig. 3c) expression was significantly higher in LoVo cells compared to SW48 cells. Both SW48 and LoVo cells expressed comparable levels of let-7b. Next, we examined KRAS expression in response to miR-143/PTX treatment. Transfection of miR-143 mimic restored intracellular miR-143 to a similar level in both cell lines regardless of KRAS mutational status or drug exposure (Fig. 4a). We found that irrespective of PTX exposure, transfection of miR-143 mimic markedly reduced KRAS mRNA expression in KRAS mutant LoVo cells (Fig. 4b), whereas the effect was insignificant in KRAS wild-type SW48 cells. These results are indicative of the fact that mutant KRAS is more susceptible to the negative modulation by miR-143 than the wild-type KRAS.

Fig. 3
figure 3

Detection of steady state expression levels of miR-143, let-7b, and KRAS mRNA and protein in the KRAS mutant LoVo and wild-type SW48 cell lines. a The endogenous levels of miR-143 and let-7b, normalized to RNU6B in SW48 and LoVo cell lines. The endogenous levels of KRAS mRNA normalized to β-actin (ACTB) expression (b), and KRAS protein (c) in SW48 and LoVo cells. The Western blot was re-probed with β-actin to serve as a loading control. Each experiment was carried out at least three times. Data represent the mean ± SEM. *P < 0.05

Full size image
Fig. 4
figure 4

Mir-143 selectively downregulates mutant KRAS expression. The expression of miR-143 (a) and KRAS mRNA (b) in the SW48 and LoVo cells transfected with scrambled control (SCR) or miR-143 mimic alone or in combination with paclitaxel (PTX). Each experiment was carried out at least three times. Data represent the mean ± SEM. *P < 0.05

Full size image

Combination of miR-143 with PTX induces apoptosis in KRAS mutant LoVo cells

Mutant KRAS constitutively activates MEK/ERK and PI3K/AKT signaling pathways, both of which are pivotal to the survival and proliferation of tumor cells [23, 24]. Thus, we next investigated if combination PTX and miR-143 mimic treatment induced apoptosis in the mutant KRAS harboring LoVo cells. Annexin V/PI-staining showed that combination of miR-143 mimic with PTX decreased percent live cells by 70 % as compared to 40 % with either reagent alone (Fig. 5a, b). The apoptotic markers, PARP, were more strongly induced in the cells co-treated with miR-143/PTX than either agent alone (Fig. 5c). Furthermore, the induction of apoptosis was not accompanied by an inhibition of proliferation as determined by similar CFSE staining pattern in PTX and combined miR-143 mimic and PTX treated LoVo cells over 72 h (P = 0.893) (Fig. 5d).

Fig. 5
figure 5

MiR-143 mimic and paclitaxel (PTX) combination blocks mutant KRAS signaling and induces apoptosis. The effect of scrambled control (SCR), miR-143 mimic, PTX, or miR-143 mimic + PTX on cell death in mutant KRAS LoVo cells. a The apoptotic cells were detected by flow cytometry using Annexin V-FITC and PI dual staining and Annexin V and PI double-negative cells were plotted as percent live cells. Data represent the mean ± SEM. *P < 0.05. b The apoptotic protein marker (poly(ADP-ribose) polymerase (PARP)) was assessed by Western blotting. The blot was re-probed with β-actin to serve as a loading control. Each experiment was carried out at least three times

Full size image

Combination of miR-143 with PTX markedly reduces migration and invasion of KRAS mutant tumor cells

LoVo cells are well characterized for their high motility [25], which is reflective of the metastatic nature of CRC. We next investigated the effect of miR-143/PTX combination on in vitro migration and invasion of LoVo cells as it will be a direct indicator of the potential of this combination therapy for in vivo usage. As shown in Fig. 6, treatment of PTX or miR-143 alone caused 20–30 % reduction in both in vitro migration and invasion, whereas around 50–55 % decrease were observed when both miR-143 and PTX were used (P < 0.05).

Fig. 6
figure 6

MiR-143 mimic and paclitaxel (PTX) combination inhibits migration and invasion of KRAS mutant LoVo, but not wild-type SW48 cells. a The effect of miR-143 mimic or PTX alone and in combination on cell migration was evaluated in SW48 and LoVo cells for 8 h. b Matrigel invasion assay was conducted to evaluate the effect of miR-143 mimic or PTX alone and in combination on the invasiveness of LoVo and SW48 cells. Each experiment was carried out at least three times. Data represent the mean ± SEM. *P < 0.05

Full size image

Discussion

Restoration of miR-143 alone or in combination with miR-145 has been previously shown to reduce tumorigenic potential in pancreatic [26], bladder [27], and colorectal [28] cancers. However, there has been no study reported looking at the potential of miR-143 replenishment as a sensitizer to reduce chemoresistance in cancer cells, inclusive of CRC. On the other hand, let-7 family members can also attenuate chemoresistance [29–33]. We however did not observe any let-7b-mediated re-sensitization to drug-induced chemoresistance.

In the present study, our results show that miR-143 can function as a chemosensitizer in drug-naive KRAS mutant cancer cells. Given that miR-143 has cognate binding site in the KRAS mRNA’s 3′-untranslated region (UTR), irrespective of any coding region mutation, it is surprising that transfection of miR-143 mimic only had effect on diminishing the expression of mutant and not wild-type KRAS mRNA (Fig. 4b). This can perhaps be explained if suppression of KRAS expression by miR-143 is dictated by the stoichiometry between KRAS mRNA and miR-143 in tumor cells and not the binding capacity to the KRAS mRNA per se. In support of this hypothesis is the fact that compared to cells with wild-type KRAS, cells harboring mutant KRAS had higher steady-state expression of KRAS mRNA and protein coupled with decreased miR-143 expression (Fig. 3). Hence, in KRAS mutant cells, the low expression of miR-143 is rate limiting, which will also explain the significant suppression of KRAS mRNA level induced by miR-143 mimic. The high levels of miR-143 in cells with wild-type KRAS will thus mean optimal targeting of the KRAS mRNA, which is not further potentiated by miR-143 mimic. Of note, though that suppression of mutant KRAS by miR-143 alone could not attenuate cell proliferation or induce cell death, which corroborates previous findings that RNAi-mediated downregulation of mutant KRAS has limited antitumorigenic potential [33]. The later can be explained by the fact that RAS-independent signaling pathways would still be operant when mutant KRAS are suppressed by RNAi.

As shown in Fig. 4a, transfection of miR-143 mimic replenished the low endogenous levels of miR-143 in LoVo cells to the abundance level in parental or miR-143 mimic transfected SW480 cells. Now, if our hypothesis is correct, it would mean that miR-143 being at a stoichiometric excess post-transfection of the mimic in LoVo cells should be able to successfully target KRAS transcript and suppress its expression. Now, referring to Fig. 4b showed us whereas scrambled transfected LoVo cells (low miR-143 stoichiometry) had high KRAS expression, miR-143 mimic transfected LoVo cells (high miR-143 stoichiometry) had low KRAS expression. In fact, the mimic transfected LoVo cells suppressed KRAS more than in the SW480 cells. This suppression was even more potentiated by co-treatment with paclitaxel. It will be potentially interesting to evaluate of the reverse experiment using miR-143 antagomir in SW480 cells would increase KRAS expression in a similar fashion.

In conclusion, our results show that miR-143 can potentially sensitize tumor cells harboring mutant KRAS to treatment with PTX, which has obvious clinical benefits for treating tumors with mutant KRAS. Given that the response rate to PTX therapy alone is normally less than 30 %, use of miR-143 mimic should significantly improve the therapeutic outcome and mitigate drug resistance, which calls for more clinical trials involving the same hypothesis.