Let-7 Sensitizes KRAS Mutant Tumor Cells to Chemotherapy

KRAS is the most commonly mutated oncogene in human cancers and is associated with poor prognosis and drug resistance. Let-7 is a family of tumor suppressor microRNAs that are frequently suppressed in solid tumors, where KRAS mutations are highly prevalent. In this study, we investigated the potential use of let-7 as a chemosensitizer. We found that let-7b repletion selectively sensitized KRAS mutant tumor cells to the cytotoxicity of paclitaxel and gemcitabine. Transfection of let-7b mimic downregulated the expression of mutant but not wild-type KRAS. Combination of let-7b mimic with paclitaxel or gemcitabine diminished MEK/ERK and PI3K/AKT signaling concurrently, triggered the onset of apoptosis, and reverted the epithelial-mesenchymal transition in KRAS mutant tumor cells. In addition, let-7b repletion downregulated the expression of β-tubulin III and ribonucleotide reductase subunit M2, two proteins known to mediate tumor resistance to paclitaxel and gemcitabine, respectively. Let-7 may represent a new class of chemosensitizer for the treatment of KRAS mutant tumors.


Introduction
KRAS mutations occur in approximately 20% of all human cancers and are particularly prevalent in pancreatic ductal adenocarcinoma (PDAC,~90%), non-small cell lung cancer (NSCLC, 25%) and colorectal cancer (~40%) [1]. As a driver oncogene, constitutively activated KRAS transduces cascades of parallel phosphorylation reactions in RAF/MEK/ERK and PI3K/AKT/ mTOR pathways among others, culminating with uncontrolled cell proliferation, evasion of apoptosis, and metastasis. Consequently, intensive effort has been made to target mutant RAS for the treatment of cancer. A number of strategies have been pursued in order to thwart KRAS-driven oncogenesis, including the combination of kinase inhibitors (such as PI3K and MEK inhibitors) to simultaneously repress downstream RAS effectors [2,3], the inhibitors that directly attenuate the kinase activity of mutant KRAS [4,5], and the small interference RNAs (siRNAs) that specifically knockdown mutant KRAS [6]. To date, none of these approaches has been approved for clinical uses. Treatment of KRAS mutant tumors still largely relies on conventional chemotherapy, which commonly results in poor response rate and development of drug resistance in cancer patients.
MicroRNAs (miRNAs) are endogenous, single-stranded noncoding RNAs (~22 nucleotides in length) that control gene expression at the post-transcriptional level [7]. By imperfectly base-pairing with the 3'-untranslated region (3'-UTR) of the target messenger RNAs (mRNAs), miRNAs suppress protein translation by either impeding the translation initiation or accelerating the degradation of mRNAs. It is estimated that 60% of all human protein-coding genes are the direct targets of miRNAs [8]. Since the first discovery over two decades ago, more than 1,000 human miRNAs have been identified, many of which are aberrantly expressed in tumor cells and play causative roles in tumorigenesis and tumor progression.
Consisted of 13 related miRNAs located on 9 different chromosomes, the human let-7 microRNA family is crucially involved in cell differentiation and proliferation during development by directly targeting a large number of oncogenes including RAS, HMGA2 and MYC [9]. Let-7 family members are tumor suppressor miRNAs abundantly expressed in differentiated normal tissues, but are frequently lost in human cancers including NSCLC and PDAC [10,11]. Notably, the expression of let-7a and let-7f is inversely correlated with the survival of NSCLC patients [11]. In NSCLC cells, transfection of let-7a, let-7b, let-7c, let-7d and let-7g all potently reduced KRAS expression and impaired cell proliferation, reflecting the similar biological functions of the let-7 family members [12]. Ectopic expression of let-7b and let-7g has been shown to repress tumor initiation in KRAS mutant NSCLC in mouse models [13,14]. However, let-7 restoration does not trigger apoptosis in KRAS mutant tumor cells, limiting its anticancer potency against pre-established tumors [15,16].
In the current work, we explored the therapeutic potential of let-7 as a chemosensitizer for conventional chemotherapy. Our study reveals that let-7b repletion selectively downregulates mutant KRAS expression and potentiates the anticancer activity of paclitaxel and gemcitabine in KRAS mutant tumor cells, which is accompanied by attenuated cell proliferation, enhanced apoptosis and the reversal of the epithelial-mesenchymal transition (EMT) phenotype in tumor cells.

Cell transfection and treatments
Let-7b-5p mimic and the non-targeting (scramble oligonucleotides) control were purchased from Bioneer (Alameda, CA). Transfection was carried out using Lipofectamine 2000 (Life Technologies) according to the manufacturer's procedures. In brief, cells were 60-80% confluent at the time of transfection. Let-7b mimic or the scramble control was mixed gently with Lipofectamine 2000 diluted in the Opti-MEM reduced serum medium (Life Technologies), and incubated at room temperature for 25 minutes before treating the cells at a final concentration of 50 nM. The transfected cells were incubated for 6 hours, the medium was then replaced with fresh growth medium supplemented with 2% serum but no antibiotics. Twenty four hours post transfection, the cells were treated with varying concentrations of paclitaxel or gemcitabine for 48 hours.

Cell viability assay and analysis of combination effects
Cells were seeded at a density of 5,000-8,000 cells/well in 96-well plates, and were treated in triplicates with varying concentrations of paclitaxel (0.1-50 nM) or gemcitabine (5-150 nM or 5-150 μM) individually, or in combination with let-7b mimic (50 nM) or the scramble control (50 nM) as described above. Cells were then fixed with 1% glutaraldehyde, stained with 0.1% crystal violet, and dissolved in 10% acetic acid. The absorbance was quantified at 595 nm on a plate reader. The relative cell viability was calculated as the percentage of the absorbance of the treated vs. the untreated wells. IC 50 of paclitaxel or gemcitabine was calculated by GraphPad Prism 5 (GraphPad Software, La Jolla, CA).
RNA extraction, cDNA synthesis, and quantitative real-time PCR (qRT-PCR) Cells were transfected with let-7b mimic or the scramble control (50 nM) for 24 hours. The cells were then incubated with paclitaxel (20 nM) or gemcitabine (50 nM in NIH-H1975, A549 and BxPC-3 cells, 40 μM in Panc-1 cells) for 48 hours. Total RNA of cells was extracted using TRIzol (Life Technologies). The PrimeScript RT Reagent Kit (Clontech, Mountain View, CA) was used for the reverse transcription of cDNA according to the manufacturer's procedures. The reactions were incubated first at 16°C for 30 minutes, at 42°C for 30 minutes, and then inactivated by incubation at 85°C for 5 minutes.
The expression of let-7b was examined by qRT-PCR using TaqMan miRNA probes (Applied Biosystems, Foster City, CA) and Premix Ex Taq kit (Clontech) according to the manufacturer's procedures. In these experiments, the relative expression level of let-7b in the cells was normalized to that of U6 snRNA, a ubiquitously expressed small nuclear RNA. The expression of mRNAs was examined by qRT-PCR using the forward and reverse primers (Eurofins, Ebersberg, Germany) and EvaGreen dye (Biotium, Hayward, CA) according to the manufacturer's protocol, and was normalized by that of β-actin in each sample. Both EvaGreen and TaqMan qPCRs were conducted using the ABI 7500 RT-PCR system (Applied Biosystems) and all of the reactions were run in triplicate. The reactions were incubated at 95°C for 5 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. A comparative threshold cycle (ΔCT) method was used to compare each treatment with the internal control, and the values are expressed as 2 -ΔΔCT . The sequences of the primers are listed in S1 Table. Apoptosis assay Apoptosis was assessed by detecting the externalization of phosphatidyl serine using Annexin V (BD Biosciences, San Jose, CA). Following miRNA transfection and drug treatment as described above, A549 or Panc-1 cells were harvested and stained with Annexin V-FITC and propidium iodide (PI). After incubation for 15 minutes at room temperature, the samples were analyzed using an Accuri C6 Flow Cytometer System (BD Biosciences).

Cell cycle analysis
Following miRNA transfection and drug treatment as described above, A549 or Panc-1 cells were harvested. The treated cells (1 × 10 6 ) were fixed in 70% ethanol at 4°C overnight, washed with cold PBS, resuspended in a staining solution (0.1% Triton-X-100, 50 μg/ml PI and 1 mg/ml RNase A) for 30 minutes at room temperature, and analyzed by fluorescence cytometry using a BD Accuri C6 Flow Cytometer System. The DNA content distribution was analyzed using the FlowJo 9.3.1 software (Tree Star, Ashland, OR).

Colony formation assay
Following miRNA transfection and drug treatment as described above, A549 or Panc-1 cells were harvested and re-seeded in triplicate in 12-well plates at a density of 100 cells/well. The plates were incubated at 37°C for 14 days. The adherent cell colonies were fixed with 1% glutaraldehyde and stained with 0.1% crystal violet at room temperature for 30 minutes.

Scratch wound healing assay
Following miRNA transfection and drug treatment as described above, A549 or Panc-1 cells were harvested and re-seeded to grow to full confluence in 6-well culture plates. The cell monolayers were scratched with 200-μL pipette tips. The wounded monolayers were cultured in 5% FBS media for 24 hours. Closure of the wounded areas was observed under an inverted microscope (Olympus, Tokyo, Japan) at 0, 12 and 24 hours, and quantified using ImageJ software (NIH, Bethesda, MD).

Invasion Assay
Cell invasion was evaluated using a Boyden chamber system with a polycarbonate membrane (8-μm pore size; Corning, New York, NY). For the invasion assay, the chamber inserts were pre-coated with 250 μg/mL matrigel (BD Biosciences). Following miRNA transfection and drug treatment as described above, A549 or Panc-1 cells (5×10 4 ) were suspended in serum-free DMEM and seeded in the chamber inserts, and the lower chamber was filled with DMEM supplemented with 10% FBS. The cells were incubated for 40 hours. The invading cells on the bottom of the chamber inserts were fixed with 1% glutaraldehyde and stained with 0.1% crystal violet, and photographed under an inverted microscope.

Statistical analysis
Data were presented as the mean ± SEM from at least 3 independent experiments. The Student's t-test was used to compare the statistical difference between the two tested groups. P value of less than 0.05 was considered statistically significant.

Let-7b restoration selectively enhances the chemosensitivity of KRAS mutant tumor cells
To explore the therapeutic potential of let-7 as a chemosensitizer, we assessed the effect of let-7b repletion on the cytotoxicity of paclitaxel and gemcitabine in NSCLC cells (A549 and NIH-H1975), PDAC cells (Panc-1 and BxPC-3) and breast cancer cells (MDA-MB-231 and MCF-7). The effect in the normal lung fibroblast cell line MRC-5 and the normal breast epithelial cell line MCF-10A were also evaluated. We chose to study paclitaxel and gemcitabine because of their frequent use as monotherapies and in combination therapies for the treatment of solid tumors. We found that transfection of let-7b mimic pronouncedly chemosensitized A549, Panc-1 and MDA-MB-231 cells, all of which harbor KRAS mutations, but only slightly affected KRAS wild-type NIH-H1975, BxPC-3, MCR-5, MCF-7 and MCF-10A cells (Fig 1 and S1 Fig).
Transfection of let-7b mimic decreased the IC 50 of paclitaxel from 21 nM to 4 nM in A549 cells, from 25 nM to 8 nM in Panc-1 cells, and from 10 nM to 4 nM in MDA-MB-231 cells, while the reduction was less than 30% in KRAS wild-type cells (Table 1 and S2 Table). Likewise, co-treatment with let-7b mimic reduced the IC 50 of gemcitabine to a much greater extent in KRAS mutant tumor cells than in KRAS wild-type cells. As a single agent, let-7b mimic caused less than 15% inhibition in cell proliferation irrespective of KRAS status (Fig 1 and S1 Fig). As a negative control, the scramble control minimally influenced the cytotoxicity of paclitaxel or gemcitabine in all cell lines.
To determine whether let-7b/paclitaxel or let-7b/gemcitabine 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 let-7b/paclitaxel and let-7b/ gemcitabine scored well below 1.0 in KRAS mutant A549, Panc-1 and MDA-MB-231 cells, signifying a synergistic response (Table 1 and S2 Table). By contrast, in KRAS wild-type NIH-H1975, BxPC-3, MCR-5, MCF-7 and MCF-10A cells, the combination indices were around 1.0, pointing to an additive response. Together, these results indicate that let-7b repletion selectively sensitizes KRAS mutant cells to the cytotoxicity of paclitaxel and gemcitabine.

Let-7b selectively downregulates mutant KRAS expression
To understand the differential effect of let-7b on the wild-type vs. mutant KRAS tumor cells in response to paclitaxel and gemcitabine, we first compared the endogenous levels of let-7b and KRAS in these cell lines. As shown in Fig 2A, the expression of let-7b was highest in NIH-H1975 and lowest in Panc-1 cells. Inversely, KRAS mRNA and protein levels were lowest in NIH-H1975 and highest in Panc-1 cells (Fig 2B and 2C). A549 and BxPC-3 cells expressed comparable levels of endogenous let-7b and KRAS. In each tissue type, there was an overall trend of lower let-7b and higher KRAS expression associated with mutant KRAS, which is consistent with the literature [19,20]. Transfection of let-7b mimic restored intracellular let-7b to a similar level in all four cell lines regardless of KRAS mutational status or drug exposure ( Fig  2D).
Next, we examined KRAS expression in response to let-7b/cytotoxin treatment. We found that irrespective of paclitaxel or gemcitabine exposure, transfection of let-7b mimic markedly reduced KRAS expression at both the protein and mRNA levels in KRAS mutant A549 and Panc-1 cells (Fig 2E and 2F), whereas the effect was insignificant in KRAS wild-type NIH-H1975 and BxPC-3 cells. These results strongly suggest that mutant KRAS is more susceptible to the negative modulation by let-7b than the wild-type KRAS.
Combination of let-7b with cytotoxins potentiates the blockage in MEK/ ERK and PI3K/AKT signaling and promotes apoptosis in KRAS mutant tumor 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 [21,22]. To functionally validate the downregulation of mutant KRAS observed above, the effect of let-7b/cytotoxin treatment on MEK/ERK and AKT signaling was investigated. Consistent with the decreased mutant KRAS protein in A549 and Panc-1 cells (Fig 2E), transfection of let-7b mimic reduced the phosphorylation of MEK and ERK1/2, two RAS downstream effectors (Fig 3A). Importantly, combination of let-7b mimic and paclitaxel or gemcitabine diminished the phosphorylation of MEK, ERK1/2 and AKT to a much greater extent than let-7b mimic or either drug individually (Fig 3A and 3B). By contrast, let-7b mimic, paclitaxel or gemcitabine either alone or in combination failed to appreciably inhibit MEK/ERK or PI3K/AKT signaling in KRAS wildtype NIH-H1975 and BxPC-3 cells (S2 Fig). Furthermore, two apoptotic markers, the cleaved caspase-3 and PARP, were more strongly induced in the cells co-treated with let-7b/paclitaxel or let-7b/gemcitabine than either agent alone, which was accompanied by the reduced expression of BCL-2, an anti-apoptotic protein (Fig 3C and 3D). In line with the Western blotting data, the results from Annexin V/PI-staining showed that combination of let-7b mimic with paclitaxel or gemcitabine nearly doubled the apoptotic cell population (~20%) in both A549 and Panc-1 cells, compared to those treated with either drug alone (Fig 3C and 3D). As a single agent, let-7b mimic induced a minor surge in apoptotic cells (~5%), in agreement with its modest cytotoxicity observed in cell proliferation assay (Fig 1). Collectively, these data indicate that combination of let-7b mimic with paclitaxel or gemcitabine potently blocks MEK/ERK and PI3K/AKT signaling, leading to increased apoptosis in KRAS mutant cells.

Let-7b cooperates with cytotoxins to block cell cycle progression and inhibit colony formation of KRAS mutant tumor cells
The let-7 family is reported to impact on cell cycle progression and proliferation through negatively regulating multiple oncogenes [23]. To further assess the concerted effect of let-7b and cytotoxins on cell proliferation, we performed DNA content analysis by flow cytometry. As shown in Fig 4A and 4B, transfection of let-7b mimic in A549 and Panc-1 cells caused an accumulation in G 1 phase and a corresponding reduction in S and G 2 /M phases, which is consistent with the role of let-7 as a negative regulator of G 1 -to-S phase transition. When combined with paclitaxel, a microtubule inhibitor that caused potent G 2 /M phase arrest, the treatment elicited even more severe G 2 /M phase arrest. On the other hand, combination of let-7b mimic with gemcitabine, a nucleoside analog that caused S phase accumulation, arrested the cell cycle at G 1 phase. In addition, combination of let-7b mimic with paclitaxel or gemcitabine notably increased the apoptotic sub-G 1 population in both cell lines, consistent with increased apoptosis observed in Fig 3C and 3D. Next, the colony formation assay was carried out to further characterize the effect of let-7b/ cytotoxin combination on the proliferative capacity of tumor cells. As shown in Fig 4C and 4D, the number and size of colonies formed by A549 and Panc-1 cells were drastically decreased following the treatment of let-7b/paclitaxel or let-7b/gemcitabine, compared to the individual agents. Taken together, these results support the notion that let-7b synergizes with cytotoxins to arrest cell cycle and inhibit the proliferation of KRAS mutant tumor cells.

Combination of let-7b with cytotoxins markedly reduces migration and invasion of KRAS mutant tumor cells
Both A549 and Panc-1 cell lines are well characterized for their high motility [24,25], which is reflective of the metastatic nature of NSCLC and PDAC. To investigate the effect of let-7b/cytotoxin combination on migration, we performed the scratch wound healing assay. A scratch wound was created at 48 hours post let-7b/cytotoxin treatment as indicated by the black lines, whereas the wound closure was measured based on the gap area filled by the migrating tumor cells. As shown in Fig 5A, A549 and Panc-1 cells treated with the scramble control had over 90% wound closure by 24 hours. Treatment of paclitaxel or gemcitabine alone caused 30-40% reduction in the wound closure, whereas over 50-60% decrease in the wound closure was observed in the cells treated with let-7b/paclitaxel or let-7b/gemcitabine combination (p < 0.05). To further evaluate the effect of let-7b/cytotoxin treatment on the invasiveness of tumor cells, the matrigel transwell invasion assay was performed. We found that the invasion of A549 cells was ablated when paclitaxel or gemcitabine was combined with let-7b mimic (Fig 5B). By contrast, either agent individually only moderately reduced the invading cell population. Similar results were observed in Panc-1 cells.
EMT is an important process during tumor cell migration and invasion, which converts adherent epithelial cells to motile mesenchymal cells [26]. Let-7 is known to directly target HMGA2, an important transcription factor that regulates EMT [27]. We found that let-7b repletion repressed HMGA2 protein in A549 and Panc-1 cells, which was accompanied by an increase in the expression of epithelial marker E-cadherin and the coordinated decrease in the mesenchymal markers Snail 1 and vimentin (Fig 5C). It is worth noting that the combined treatment of let-7b mimic with paclitaxel or gemcitabine reduced vimentin level more drastically than either agent individually. Collectively, these results indicate that combination of let-7b repletion with paclitaxel or gemcitabine greatly compromises the migratory ability and invasiveness of KRAS mutant tumor cells and reverts the EMT phenotype. Chemosensitization of KRAS mutant cells by let-7b is associated with the suppression of TUBB3 and RRM2 TUBB3 is one of the β-tubulin subtypes that has low abundance in most normal tissues but is highly expressed in several solid tumors including NSCLC and PDAC [28,29]. Overexpression of TUBB3 has been correlated with tumor resistance to taxane chemotherapy [30]. To investigate the role of TUBB3 in mediating the sensitization of paclitaxel by let-7b mimic, we evaluated the mRNA and protein levels of TUBB3 in the cells exposed to let-7b/paclitaxel treatment. We found that TUBB3 was highly expressed in KRAS mutant A549 and Panc-1 cells, but was barely detectable in KRAS wild-type NIH-H1975 and BxPC-3 cells (Fig 6A). Transfection of let-7b mimic reduced TUBB3 at both the transcriptional and translational levels in A549 and Panc-1 cells (Fig 6A and 6B), which paralleled the decline in KRAS protein caused by let-7b restoration (Fig 2E). These findings are consistent with the notion that TUBB3 expression is upregulated by mutant but not wild-type KRAS [29]. Not surprisingly, TUBB3 expression in NIH-H1975 and BxPC-3 cells was not altered by let-7b repletion (Fig 6A and 6B). These results suggest that let-7b sensitizes KRAS mutant tumor cells to the cytotoxicity of paclitaxel in part through the suppression of TUBB3.
Overexpression of RRM2 has been linked to gemcitabine resistance in PDAC. Knockdown of RRM2 using siRNA has been shown to chemosensitize PDAC cells to gemcitabine [31]. Based on the microarray analysis and the presence of let-7 complementary sites in the 3'-UTR of the gene, RRM2 is considered to be a putative target of let-7 [12]. To study the involvement of RRM2 in the sensitization of gemcitabine by let-7b restoration, we analyzed RRM2 expression in response to let-7b/gemcitabine treatment. Curiously, we found that RRM2 protein level was elevated (~2.5-fold) in all gemcitabine-treated cells ( Fig 6C). As shown in Fig 6D, the transcription of RRM2 was markedly induced by gemcitabine treatment, and the induction was more robust in the cells harboring mutant KRAS (~6-fold) than those with wild-type KRAS (2 .5-fold). Transfection of let-7b mimic clearly attenuated gemcitabine-induced RRM2 expression, even though RRM2 protein and mRNA levels were still above the basal levels found in the untreated cells (Fig 6C and 6D). Given the critical role of RRM2 in mitigating gemcitabine cytotoxicity and the robust induction of RRM2 expression upon gemcitabine treatment, these results suggest that the downregulation of RRM2 by let-7b may participate in the sensitization of gemcitabine in KRAS mutant cells.

Discussion
Restoration of let-7 family members has been previously shown to reduce chemoresistance in tumor cells. For instance, resistance to cisplatin in glioblastoma cells has been associated with low let-7b level. Reconstitution of let-7b re-sensitized glioblastoma cells to cisplatin by abrogating cyclin D1 [32]. In cisplatin-resistant esophageal squamous cells, transfection of let-7c restored the sensitivity to cisplatin by inactivating IL-6/STAT3 pathway [33]. In docetaxelresistant NSCLC cells, the forced expression of let-7c increased the in vitro and in vivo sensitivity to docetaxel via targeting BCL-xL, which was accompanied by partial reversal of the EMT phenotype [34]. In a recent study by Boyerinas et al., the authors showed that let-7g Let-7b/cytotoxin combination reduces migration and invasion of KRAS mutant tumor cells. A, the effect of let-7b mimic, paclitaxel or gemcitabine on cell migration was evaluated in A549 and Panc-1 cells at 0, 12 and 24 hours following the scratch wound (images, left panel). The area of wound healing was quantified by ImageJ (right panel). B, the matrigel invasion assay was conducted to evaluate the effect of let-7b mimic, paclitaxel or gemcitabine on the invasiveness of A549 and Panc-1 cells. Images were acquired using an inverted microscope at × 40 magnification. C, Protein levels of HMGA2, Snail 1, E-cadherin and vimentin in A549 and Panc-1 cells were assessed by Western blotting. Each experiment was carried out at least 3 times. Data represent the mean ± SEM. *, p < 0.05; **, p < 0.01. doi:10.1371/journal.pone.0126653.g005 replacement re-sensitized MDR1-overexpressing ovarian cancer cells to the cytotoxicity of paclitaxel via IMP-1-mediated reduction of MDR1 [35]. These studies provide strong evidence that let-7 repletion can re-sensitize drug-induced chemoresistance, although the sensitization by let-7 appears to be highly dependent on the cellular context in tumor cells.
In the current study, we uncovered a new role of let-7b as a chemosensitizer in drug-naive KRAS mutant tumor cells. Given that let-7 binds to 3'-UTR of KRAS mRNA independent of its mutational status in the coding region, it is intriguing that transfection of let-7b mimic only diminished the expression of mutant but not wild-type KRAS mRNA (Fig 2F). While the molecular mechanism accounting for this selectivity remains to be fully elucidated, it is plausible that the downregulation of KRAS expression by let-7b is dependent on the stoichiometry between KRAS mRNA and let-7b in tumor cells. Compared to KRAS wild-type cells, KRAS mutant tumor cells of the same tissue origin were shown to express notably higher KRAS mRNA and lower let-7b levels (Fig 2A and 2B). Suppression of KRAS is thus likely rate-limited by low let-7b level in KRAS mutant cells, which becomes pronouncedly accelerated upon let-7b repletion and results in decreased KRAS mRNA and protein levels. In contrast, targeting of KRAS mRNA by let-7b may already operate at full capacity in KRAS wild-type cells and is not subjected to further enhancement in the presence of ectopic let-7b level. Nevertheless, the downregulation of mutant KRAS by let-7b alone was insufficient to block the proliferation of KRAS mutant cells nor did it induce appreciable apoptosis, consistent with the findings that knockdown of mutant KRAS using siRNAs had limited antitumor effect in vitro and in vivo [19,36]. This is because even though depletion of mutant KRAS inhibits MEK/ERK signaling, other pathways such as PI3K/AKT signaling can remain activated through RAS-independent mechanisms. In fact, MEK inhibition alone by small-molecule inhibitors was ineffective in suppressing KRAS mutant tumors. The combination of PI3K and MEK inhibitors on the other hand drastically improved the tumor response in preclinical models [2,3]. The MEK/ERK and PI3K/ AKT signaling each promotes cell growth and survival via their own distinct downstream effectors, while they both converge on the BH3 family of proteins that regulate apoptosis [37]. Combined MEK/ERK and PI3K/AKT inhibition is required to effectively trigger apoptosis in KRAS mutant tumor cells [2]. We found that combination of let-7b repletion with paclitaxel or gemcitabine diminished both MEK/ERK and PI3K/AKT signaling in KRAS mutant tumor cells, leading to substantial increase in apoptosis. Concomitant blockage of these two key pathways is thus largely accountable for enhanced chemosensitivity observed in let-7b-transfected KRAS mutant tumor cells.
Overexpression of HMGA2 is closely correlated with the malignant phenotype and poor prognosis of NSCLC and PDAC [38,39]. HMGA2 has been shown to drive EMT and metastasis of epithelial tumors in vivo [40]. The EMT phenotype of tumor cells is associated with drug resistance to conventional chemotherapy including gemcitabine and paclitaxel, as well as to the molecularly targeted therapies [41][42][43]. As a transcription factor, HMGA2 cooperates with transforming growth factor β in inducing Snail 1 expression, a zinc-finger transcription factor crucially involved in EMT and tumor progression [44]. By directly binding to the E-boxes of the human E-cadherin promoter, Snail 1 represses the transcription of E-cadherin, a key adhesion protein that is typically lost in epithelial tumor cells during EMT [45]. While HMGA2 is a direct target of let-7, ectopic expression of HMGA2 partially rescued the growth inhibition by let-7 [27]. Interestingly, HMGA2 mRNA also functions as a competing endogenous RNA that impedes let-7 regulation of other oncogene targets [46]. In addition, HMGA2 is involved in maintaining cancer stem cells in the undifferentiated state, which are resistant to most of the cytotoxic drugs [47]. Our results confirm that let-7b repletion downregulates HMGA2 and its major downstream effector Snail 1, resulting in the concomitant increase in E-cadherin and decrease in vimentin expression. We show that the combined treatment of let-7b mimic and paclitaxel or gemcitabine markedly suppressed the migration and invasion of KRAS mutant cells. The reduced HMGA2 expression and the reversal of the EMT phenotype observed in this study likely contribute to the chemosensitization of tumor cells.
Paclitaxel and gemcitabine inhibit cell proliferation via distinct mechanisms. The finding that let-7b repletion sensitized KRAS mutant tumor cells to both drugs suggests that in addition to the shared molecular targets, let-7 possibly also modulates other signaling molecules that synergize specifically with each drug. In NSCLC patients, high TUBB3 expression is associated with poor prognosis, which is regulated by mutant KRAS signaling [28]. Knockdown of TUBB3 using siRNA has been shown to significantly increase the cytotoxicity of paclitaxel and induce apoptosis in KRAS mutant tumor cells [48]. We observed a clear decrease in TUBB3 expression at both the transcriptional and translational levels in let-7b-transfected KRAS mutant tumor cells, paralleling with the decline of mutant KRAS in these cells. It is plausible that TUBB3 downregulation by let-7b may play a causative role in sensitizing KRAS mutant cells to paclitaxel. Notably, TUBB3 is a direct target of miRNA-200c, which is an important tumor suppressor that induces epithelial differentiation and reverts EMT in tumor cells [49]. Restoration of miRNA-200c markedly sensitizes tumor cells to the microtubule-targeting drugs including paclitaxel.
Ribonucleotide reductase is one of the key determinants of gemcitabine resistance in human cancers. Activity of this enzyme, which catalyzes the reduction of ribonucleoside diphosphates to the corresponding 2'-deoxyribonucleotides (dNTP), the building blocks for DNA synthesis, is modulated by its M2 subunit RRM2 [50]. Overexpression of RRM2 expands the dNTP pool and confers gemcitabine resistance to tumor cells [51]. In gemcitabine-resistant tumor cells, RRM2 mRNA is found to be elevated nearly 10-fold relative to that of the parental cells [52]. The RRM2 mRNA level is inversely correlated with the response rate in cancer patients treated with gemcitabine [53]. In the current study, we found that RRM2 expression was rapidly induced by gemcitabine, implying that the upregulation of RRM2 expression in tumor cells is an early event in acquiring resistance to gemcitabine. Attenuation of RRM2 expression by let-7b repletion is potentially a promising approach to chemosensitize KRAS mutant tumor cells to gemcitabine.
In summary, we have shown that let-7b selectively sensitized KRAS mutant tumor cells to the cytotoxicity of paclitaxel and gemcitabine. This was accompanied by diminished MEK/ ERK and PI3K/AKT signaling, the reversal of the EMT phenotype, and the downregulation of HMGA2, TUBB3 and RRM2. These results have important implications for the treatment of KRAS mutant tumors. Currently, the response rate of paclitaxel in NSCLC patients is about 26%, while gemcitabine is effective in about 10% of the PDAC patients [54,55]. Increasing the sensitivity of these tumors to chemotherapy should significantly improve the therapeutic outcome and counteract drug resistance. Therapeutic exploitation of let-7 as a general chemosensitizer for KRAS mutant tumors warrants further investigation.