Nucleolin Aptamer N6L Reprograms the Translational Machinery and Acts Synergistically with mTORi to Inhibit Pancreatic Cancer Proliferation

Simple Summary Pancreatic cancer is an aggressive disease characterized by its invasiveness, rapid progression, and resistance to conventional therapy. There is a need to identify new molecules to improve current therapies. The aim of this study was to analyze how the pancreatic cancer cells react to the treatment with an inhibitor of nucleolin, N6L. To this end, we analyzed how the translation was affected in the cells during the treatment. We discovered that in response to N6L, a signaling pathway called the mTOR pathway was activated and was involved in the activation of translation of a subset of mRNA that could be involved in the resistance of the cells to the treatment. Indeed, we showed that the combined action of inhibitors of the mTOR pathway with N6L synergistically inhibited the cancer cells’ proliferation. We propose that this new combination of molecules could be a novel therapeutic option for pancreatic cancer. Abstract We previously showed that N6L, a pseudopeptide that targets nucleolin, impairs pancreatic ductal adenocarcinoma (PDAC) growth and normalizes tumor vessels in animal models. In this study, we analyzed the translatome of PDAC cells treated with N6L to identify the pathways that were either repressed or activated. We observed a strong decrease in global protein synthesis. However, about 6% of the mRNAs were enriched in the polysomes. We identified a 5′TOP motif in many of these mRNAs and demonstrated that a chimeric RNA bearing a 5‘TOP motif was up-regulated by N6L. We demonstrated that N6L activates the mTOR pathway, which is required for the translation of these mRNAs. An inhibitory synergistic effect in PDAC cell lines, including patient-derived xenografts and tumor-derived organoids, was observed when N6L was combined with mTOR inhibitors. In conclusion, N6L reduces pancreatic cells proliferation, which then undergoes translational reprogramming through activation of the mTOR pathway. N6L and mTOR inhibitors act synergistically to inhibit the proliferation of PDAC and human PDX cell lines. This combotherapy of N6L and mTOR inhibitors could constitute a promising alternative to treat pancreatic cancer.


Introduction
Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive disease, for which mortality closely parallels incidence with a 5-year survival rate of less than 6% [1]. It is predicted to be the second leading cause of cancer-related death by 2030 [2]. The biology of PDAC contributes to early recurrence and metastasis, resistance to chemotherapy and radiotherapy, because of its complexity at the genomic, epigenetic, and metabolic

Spheroid and Organoid Formation
For the spheroid assay, PDAC cell lines were cultured in 96-well ultra-low attachment (ULA) plate and allowed to form a spheroid overnight. The next day, PDAC spheroids were treated or not with different concentrations of N6L alone or in combination with mTORi, RAPA, and INK128 in 4 replicates. Spheroid growth was monitored by IncuCyte ZOOM ® (Michigan, USA) live-cell imaging. Over 96 h, the spheroid area was calculated using ImageJ software, and data were normalized to control.
For organoid formation, mPDAC cells (1 × 10 3 cells/mouse in 50 µL PBS) were orthotopically injected into the pancreas of FVB/n syngeneic mice, as previously described [9]. Mice were sacrificed after 4 weeks, and tumor-derived organoids were established by following Tuveson's laboratory protocol [22]. Briefly, tumors were digested with 0.012% (w/v) collagenase XI (Sigma (Burlington, MA, USA) and 0.012% (w/v) dispase (Gibco (Waltham, MA, USA) in DMEM media containing 1% FBS, and the resulting cell suspension was incorporated into growth factor-reduced matrigel (Corning (Corning, NY, USA)) to obtain a 3D culture. All in vivo experiments were carried out with the approval of the appropriate ethical committee and under conditions established by the European Union.

Viability Assay (MTS) and Apoptosis Analysis
Cell viability was assessed by MTS assay (CellTiter 96 AQueous MTS Reagent, Promega (Madison, WI, USA). Briefly, cell medium was supplemented with MTS reagent, incubated for 2 h, and then the absorbance at 450 nm was recorded on TECAN microplate reader (Sunrise ® (Basel, Switzerland)). Viability for treated cells was normalized to non-treated cells on the same plate.
Apoptosis analysis was performed on the IncuCyte ZOOM ® live-imaging with Caspase-3/7 Green Reagent for Apoptosis (Essen BioScience (Ann Arbor, MI, USA)). Cells were treated or not with increasing concentrations of N6L in 96-well plates in triplicate, and Caspase-3/7 reagent was added to the cells. IncuCyte ZOOM ® live-imaging software was used to count the fluorescent object numbers per mm 3 in each well, which reflects the caspase 3/7 activation. Staurosporine (1 µM) was used as a positive control for inducing apoptosis.

Analysis of Synergy/Antagonism from Combination Studies
The effects of nucleolin antagonist (N6L) combined with mTOR inhibitors (mTORi) on PDAC cell growth were analyzed using the freely Combenefit software from CRUCK Cambridge Institut [23], which enables to determine synergy/antagonism using the published Loewe model and attributes a positive score indicates synergy, a score of 0 for additive effect, and a negative score for antagonism. The "3D surface", "contour", and "matrix" views were selected as graphical outputs for the synergy distribution and are represented in the Results/Figures section.

Dual-Luciferase Reporter Assay for 5 TOP Constructs
NCL-depleted cells or N6L-treated cells or not were seeded in 96-well plates over 24 h. For the 5 TOP activity assay, cells were co-transfected with a monocistronic reporter containing 5 TOP motif of Rps3 or Rpl36 genes (GenScript (Piscataway, NJ, USA) and pRL construct constitutively expressing the Renilla luciferase gene as an internal control for transfection efficiency. Luciferase assays were performed 2 h after transfection with the reporter plasmids using X-tremeGENE 9 reagent (Roche (Basel, Switzerland). Dual-luciferase assays were performed using the Dual-Glo luciferase reagent (Promega (Madison, WI, USA) according to the manufacturer's instructions and a Tecan M1000 plate reader. The relative luciferase activity was calculated by firefly-luciferase activity/Renilla luciferase activity.

SUnSET for Global Protein Synthesis, Western Blots
Global protein synthesis analysis by puromycylation followed by puromycin detection was performed [24]. Briefly, puromycin (1 µg/mL) was added to the N6L-treated or nontreated cell medium for 2 h at 37 • C. Cells were harvested, lysed in Laemmli buffer, and loaded onto a polyacrylamide gel. Puromycin incorporation was detected by Western blot on whole-cell protein extracts.
Twenty micrograms of total protein lysates were run on a 10-15% SDS polyacrylamide gel and transferred onto a nitrocellulose membrane. The membrane was blocked with 3% nonfat milk in TBST. The antibodies (listed in Table S2) were incubated for 1 h 30 min in 3% milk-TBST. Proteins were detected by chemiluminescence using an anti-rabbit or anti-mouse HRP-conjugated antibody and ECL substrate (Covalab (Bron France). Images were collected on a ChemiDoc XRS+ (Bio-Rad (Hercules, CA, USA), and the signal was analyzed using the Bio-Rad ImageLab software. The original western blots can be found in Figure S6.

Polysome Profiling and RNA Extraction and Sequencing
Cells were cultured in 15 cm dishes and were treated or not for 48 with 50 µM N6L. Before cell lysis, cells were incubated with emetine then washed with cold PBS as previously published by our lab [25]. For cell fractionation, 10 × 10 6 cells were resuspended in lysis hypotonic buffer for 10 min at 4 • C. Cell lysates were successively centrifuged at 700 g and 1200 g to eliminate the nuclei and the mitochondria, respectively. Cytoplasmic lysates (1-2 mg of proteins) were loaded onto 15-47% sucrose density gradients and centrifuged at 217,000 g for 2 h at 4 • C. Gradients were fractionated into 19 fractions, and the OD at 254 nm was continuously recorded using an ISCO fractionator (Teledyne ISCO (Lincoln, NE, USA). RNA from cytosolic and polysomal pooled fractions was extracted and purified using TRIzol reagent (Invitrogen, (Waltham, MA, USA)).

NCL Targeting by N6L Impairs Protein Synthesis and Induces a Translational Reprogramming in mPDAC Cells
mRNA translation is a central cellular process that regulates growth and metabolism [31]. To evaluate the impact of N6L treatment on mPDAC translatome, we performed polysome profiling analysis on cells treated or not with N6L for 48 h (Figure 1).
Time-course analysis of mPDAC treated with different concentrations of N6L shows that at 30-50 µM, a strong reduction in cell proliferation is observed ( Figure S1A) while metabolic activity, which reflected cell viability, assessed by MTS test, is reduced by 56% ( Figure S1B).  (15-47%). Following ultracentrifugation, 40S and 60S ribosome subunits, the 80S monosomes, and polysomes are separated. RNAs were extracted from polysome pooled fractions (mRNA-associated polysomes) and cytoplasmic fractions (cytoplasmic mRNA). The RNA quality and integrity were verified using a Bioanalyser 2100 before sequencing and bioinformatics analysis. B,C-Polysome profiles of mPDAC cells non-treated (mPDAC NT) or treated To evaluate the impact of N6L treatment on mPDAC translatome, treated and nontreated mPDAC cells were harvested, and RNAs were extracted from an aliquot of the cytoplasmic fractions while the remaining fraction was analyzed through sucrose gradient to perform polysome profiling. RNAs were then extracted from the pooled polysomal fractions. The extracted RNAs (from cytoplasmic and polysomal pooled fractions) were submitted to deep sequencing (workflow shown in Figure 1A).
The polysome profile of mPDAC non-treated cells ( Figure 1B) shows the classical profile of ribosome distribution. In contrast, when cells were treated with N6L ( Figure 1C), a strong decrease in polysome peaks is observed with a reduced level of 40S and 60S subunits and an important increase in the 80S peak in comparison to non-treated cells.
In order to evaluate the ability of mPDAC cells to synthesize proteins following N6L treatment, we used the surface sensing of translation (SUnSET) technique [24], which is based on the incorporation of puromycin to newly synthesized proteins and its detection with anti-puromycin antibodies ( Figure 1). As shown in fig1.D-E, mPDAC-treated cells displayed a significant decrease (26%) in the amount of puromycin-labeled peptides, compared to non-treated mPDAC as soon as after 24 h of N6L treatment (lane 4) with a maximum decrease of 52% after 48 h (lane 6) ( Figure 1D,E).
These results indicated a decrease in protein synthesis in N6L-treated cells and agree with the decrease in the mPDAC translation capacity upon N6L treatment.
To determine how the translatome was affected by N6L treatment, results from the deep sequencing of cytoplasmic and polysomal fractions were analyzed ( Figure 2). The RNA-sequencing analysis using the package DeSeq2 showed that among 14,384 mRNAs present in mPDAC, 39% (5610) were significantly expressed in both non-treated and N6Ltreated cells (p-adj < 0.05) ( Figure 2A).
These significantly expressed genes in non-treated and treated mPDAC cells were used to identify bona fide alterations in translational efficiency in response to N6L by calculating the translational index (TI) ( Figure 2D). TI reflects the translation efficiency for each expressed mRNA by measuring the percentage of mRNA engaged in the polysomal (poly) fractions to a total amount of this mRNA found in the cytoplasmic (cyto) fractions and comparing these percentages between treated (T) and non-treated (NT) mPDAC cells (TI = X1/X2, with X1 and X2 representing the ratio between polysomal-to-cytoplasmic fraction in mPDAC-treated and non-treated cells, respectively).
By applying a TI cut-off = 1.5 and p-adj < 0.05, we identified 886 genes whose TI was significantly changed in response to N6L over 48 h ( Figure 2C), indicating that N6L treatment altered the translation efficiency of 886 mRNAs out of the 14,384 analyzed. The range of TI varying from −2 to +2.5 ( Figure 2D). The translation of 673 (75.96%) of these mRNAs was down-regulated, whereas that of 213 (24.04%) mRNAs was up-regulated ( Figure 2C). Therefore, down-regulation of translation efficiency was more frequent than up-regulation (genes list in Table S1).  in both NT and T conditions (A). About 6% (886) of these 5610 genes were significantly deregulated at translational levels (TI cut-off = 1.5, p < 0.05) (B). Among these 886 translationally deregulated genes, 24.04% were up-regulated, and 75.96% were down-regulated (C). Distribution of the Translational Index (TI) (TI = X1/X2), with X1 and X2 expressing the ratio between polysomal-to-cytoplasmic fraction in mPDAC-treated and non-treated cells, respectively (D) (complete gene list in Table S1). (E,F)-Ontology analysis was determined using the functional annotation clustering tools from EnrichR software. Translationally down-regulated genes are involved in the cell cycle, Fanconi anemia and RNA transport (E). While translationally up-regulated genes are mainly involved in ribosome/translation and metabolism (F).
To determine the main functions of the genes whose mRNAs recruitment within polysomes was altered in response to N6L treatment, we performed a gene ontology analysis using the functional annotation clustering analytic modules of EnrichR bioinformatics resources that provides a rank classification of enriched functions based on the determination of P-values and enrichment scores [32]. Gene ontology analysis revealed functional pathways corresponding to cell cycle, Fanconi anemia, and RNA transport for the 673 mRNAs translationally repressed ( Figure 2E) and functional pathways corresponding to ribosome/translation and oxidative phosphorylation for the 213 mRNAs translationally up-regulated ( Figure 2F).
Interestingly, among these 673 translationally down-regulated mRNAs (shown in Figure S2A), 315 were transcriptionally stable (46.8%), encoding notably proteins implicated in phosphatidylinositol biosynthesis (purple panel), and 340 were transcriptionally up-regulated (51%) encoding proteins implicated in mitotic division (green panel) ( Figure S2C,D). In addition, among the 213 translationally up-regulated genes (shown in Figure S2B), 73 were transcriptionally stable (34.3%) encoding proteins implicated in translation/ribosomal proteins (blue panel), and 131 were transcriptionally down-regulated (61.5%) encoding proteins implicated in the regulation of cholesterol storage (red panel) ( Figure S2E,F). Altogether, these data suggest that the recruitment of these mRNAs into polysomes was controlled upon N6L treatment.

N6L Induces a Global Decrease in mRNA Translation Correlated with a Decrease in EiF3 mRNAs Translation
Because the polysome profiling was indicative of a variation of global translational activity of the cell, we investigated the expression of the mRNAs coding for essential proteins involved in translation regulation that were subjected themselves to a translational regulation as revealed by our functional gene ontology analysis ( Figure 2E). Among the 673 mRNA that are down-regulated, several mRNA encode for proteins involved in translation initiation ( Figure 2E).
EiF3a mRNA and EiF3c mRNA, coding for the essential components of the translation initiation EiF3 complex, were strongly translationally repressed under N6L treatment (TI = −0.99 and −0.54, respectively) while mRNA encoding the other members (EiF3e-j) were mostly unchanged ( Figure 3A). Using Western blotting analysis, EiF3a and EiF3c protein levels were significantly reduced ( Figure 3B,C) while the protein level of the other EiF3 members (EiF3e, h, and k) did not change ( Figure 3D,E) upon N6L treatment. Using quantitative PCR, no change was observed in the expression of the EiF3a and 3c mRNAs ( Figure 3F) in agreement with the transcriptomic analysis ( Figure S2). Altogether, these results revealed that N6L decreases the EiF3a and EiF3c expression at the translational level that contributes to the down-regulation of global mRNA translation.
protein levels were significantly reduced ( Figure 3B,C) while the protein level of the other EiF3 members (EiF3e, h, and k) did not change ( Figure 3D,E) upon N6L treatment. Using quantitative PCR, no change was observed in the expression of the EiF3a and 3c mRNAs ( Figure 3F) in agreement with the transcriptomic analysis ( Figure S2). Altogether, these results revealed that N6L decreases the EiF3a and EiF3c expression at the translational level that contributes to the down-regulation of global mRNA translation.

NCL Targeting by N6L Increases 5 TOP mRNA Translation by Activating the mTOR Pathway
Although the global translation was overall reduced upon N6L treatment, the translatome analysis identified a subset of mRNAs (24.04%) that were enriched in the polysomal fractions ( Figure 2C-F). Functional gene ontology analysis of these 213 mRNAs identified genes encoding proteins involved in ribosome biogenesis, translation, or energy metabolism pathways ( Figure 2F). In particular, it was very striking to find many mRNAs coding for ribosomal proteins ( Figure 4A). latome analysis identified a subset of mRNAs (24.04%) that were enriched in the polysomal fractions ( Figure 2C-F). Functional gene ontology analysis of these 213 mRNAs identified genes encoding proteins involved in ribosome biogenesis, translation, or energy metabolism pathways ( Figure 2F). In particular, it was very striking to find many mRNAs coding for ribosomal proteins ( Figure 4A).  Ribosomal protein transcripts are characterized by the presence of a specific motif in their 5 UTR called 5 terminal oligopyrimidine (5 TOP) sequence immediately adjacent to the cap structure [34]. This family of 5 TOP mRNAs generally encodes proteins involved in translation initiation, elongation, and termination.
By comparing the mRNAs enriched in polysomes after N6L treatment mRNA to those classified as 5 TOP mRNAs [33], out of these 213 genes enriched in polysome fractions, we identified 53 TOP mRNAs (24.2%) ( Figure 4B, Table S1). Not surprisingly, gene ontology analysis of these 53 mRNAs reveals that there are all involved in pathways related to translation, cytoplasmic ribosomal protein, or Cap-dependent translation initiation. In contrast, the set of non-5 TOP is exclusively involved in pathways related to mitochondrial metabolism, including mitochondrial translation, ATP synthesis, or oxidative phosphorylation.
Most of these 53 5 TOP translationally up-regulated mRNAs (81.1%) were transcriptionally down-regulated ( Figure 4C). This means that, upon N6L treatment, enrichment in the polysome fraction is not a consequence of a higher accumulation of these mRNAs, but it is rather due to an increase in the recruitment of these mRNAs in the polysomes.
Collectively, these data suggest that NCL targeting by N6L promotes selective recruitment to polysomes of some mRNAs bearing 5 TOP motifs in their 5 UTR. This recruitment maintains translation as seen for RPL36 and RPS3 using Western-blotting analysis ( Figure 4D) despite the decreased accumulation of these mRNAs upon N6L treatment.
To determine if the 5 TOP motif plays an active role in this recruitment, we insert the 5 TOP motif of Rpl36 and Rps3 mRNAs upstream the firefly gene reporters and downstream the CMV promoter ( Figure 4E). These plasmids, pFL-5 TOP, were co-transfected with pRL (for normalization) in mPDAC cells treated with increasing doses of N6L. After 24 h, the luciferase activity was quantified to determine if the 5 UTR was required for the increased translation of the chimeric mRNA. Indeed, in the presence of the increasing amount of N6L, the presence of these 5 UTR sequences provided an increasing amount of luciferase activity, suggesting that the presence of this motif increased the recruitment of these mRNAs in the polysomes upon N6L treatment ( Figure 4F).
As the 5 TOP motifs are mainly regulated by the mTOR pathway [35], our data suggested that the N6L treatment could activate the mTOR pathway in mPDAC.
Therefore, to determine if the mTOR pathway was activated in mPDAC cells upon N6L treatment, we analyzed the expression/activation of the three primary downstream effectors of mTOR 4E-BP1, RPS6 (a downstream target of S6K), and AKT by performing Western blotting and quantification of the phosphorylated-to-total form ratios of RPS6, 4EBP1 and AKT accumulations ( Figure 5A-D).
We showed that the N6L treatment induces strong phosphorylation of RPS6, 4EBP1 (γ and β forms), and AKT, compared to non-treated mPDAC cells ( Figure 5A; lanes 1,2) while the expression of NCL protein is decreased. We then determined whether NCL expression modulates the mTOR pathway. Indeed, the activation of mTOR, via the hyper-phosphorylation of RPS6, AKT, and 4EBP1, was also observed when NCL was down-regulated by siRNA ( Figure 5A, lanes 3,4). In addition, the phosphorylation of RPS6 and AKT occurred in an N6L dose-dependent manner as revealed by Western blotting ( Figure 5B) and immunofluorescence analysis of phosphorylated RPS6 ( Figure 5C).
Collectively, these results showed that NCL inhibition, either through siRNA or with N6L treatment, causes activation of the mTOR pathway in mPDAC cells.
The translation of 5 TOP mRNAs is regulated by the binding of La-related protein 1 (LARP1) to the cap and TOP sequence [33,36,37]. Indeed, Western-blotting results showed an accumulation of LARP1 in an N6L dose-dependent manner ( Figure 5D).
In sum, these data demonstrate that the N6L treatment modified the translatome of treated PDAC cells through the activation of the mTOR pathway, thereby allowing the translation of a specific set of mRNAs involved in the translational and metabolic pathways.
In sum, these data demonstrate that the N6L treatment modified the translatome of treated PDAC cells through the activation of the mTOR pathway, thereby allowing the translation of a specific set of mRNAs involved in the translational and metabolic pathways.  (lanes 3,4). and represent hyper-and hypo-phosphorylated forms of 4EBP1.  (lanes 3,4). γ and β represent hyper-and α hypo-phosphorylated forms of 4EBP1. The phosphorylated-to-total protein ratio was quantified for RPS6, AKT, and 4EBP1. (B,C)-mTOR pathway activation was analyzed in mPDAC, non-treated (NT), or treated (T) with N6L at increasing doses using Western blotting (B) and immunofluorescence staining (C) by anti-phosphorylated RPS6 (red) and DAPI for nuclei (blue). (D)-Western-blotting analysis of LARP1, RPS6, 4EBP1, and EiF4E expression in mPDAC treated or not with N6L at 30 and 50 µM for 48 h.

Combinations of N6L and mTORi Are Synergistic on PDAC Cell Growth and Viability Inhibition
The activation of the mTOR pathway by N6L induces an increase in translation of 5 TOP mRNAs, which may contribute to a decrease in efficiency of the N6L treatment on cell viability and proliferation. We tested the effects of the combination of the mTOR inhibitors with N6L in human and murine PDAC models.
Two concentrations of RAPA, AZD2014, or INK128 were tested alone or in combination with N6L for mPDAC confluence/growth assay over 72 h using IncuCyte ZOOM ® liveimaging (Figures 6 and S3). As previously described in Figure S1, N6L induced a decrease in cell growth in a dose-dependent manner, whereas mTORi alone had no or modest effects on mPDAC cell growth at the tested concentrations (10 and 25 nM) ( Figure S3A-C). However, the combination of both mTOR inhibitors and N6L had strong inhibitory effects on mPDAC cell growth ( Figure 6A-C), suggesting an additive or synergistic effect between the two molecules. Using the Combenefit ® software [23] based on the LOEWE method, we demonstrated that the combination of N6L (5 or 10 µM) and mTORi had a strong synergistic effect on the mPDAC growth inhibition ( Figure 6G; upper panels).
We then further validated these combinations in patient-derived xenograft (PDX) models and other preclinical cell lines. Whereas mTORi alone has no or modest effects on PDX cell growth at the tested concentrations ( Figure S3D-F), except for INK128 at a high concentration of 100 nM), the combination of RAPA, AZD2014, or INK128 with N6L was superior to single-agent treatment ( Figure 6D-F), and this effect on PDX growth inhibition was synergistic ( Figure 6G, lower panels).
As N6L treatment impairs the cell viability of mPDAC cells ( Figure S6A), we wanted to determine if the death of the cells was due to caspase-dependent apoptosis as previously described [9,17]. Apoptosis induction assay was performed using the IncuCyte ZOOM ® live-imaging. A fluorescent reagent, activated when effector caspases 3/7 were activated, allowed the device to count the number of apoptotic cells per mm 2 every 2 h during 72 h. We observed a strong and significant increase in caspase-dependent apoptosis in mPDACtreated cells in a dose-dependent manner compared to non-treated cells ( Figure S6B). In addition, the morphology of the N6L-treated cells appears different from that of nontreated cells.
Upon N6L and mTORi dual treatment, we also observed a strong increase in caspasedependent apoptosis in mPDAC-treated cells with N6L combined to the dual mTORi INK128 ( Figure 6H).
To accurately predict the efficacy of these new drug combinations for PDAC therapy, we tested them on three-dimensional culture models by establishing spheroids and PDAC tumor-derived organoid cultures. To this end, we allowed PDAC and PDX cells to form spheroid, as previously described [40]. Then, we treated them with different drugs concentrations of N6L or INK128, alone or in combinations over 96 h, using IncuCyte ZOOM ® live-imaging ( Figure 6I-L). To evaluate the impact of the drug combination on spheroid growth, their area was calculated using ImageJ software. The results showed that the spheroid area decreased by 22% and 21%, upon N6L or INK128 alone, respectively ( Figure 6I-K), while when INK128 is combined with N6L treatment, the mPDAC spheroid area decreased more than 41% (p = 0.006). The same results were obtained for PDX spheroids ( Figure 6J), MiaPaca-2 ( Figure S4D,E), and Panc-1 ( Figure S5D,E) cell lines. In PDX models, the combination of N6L and INK128 significantly decreased the PDX spheroid area by more than 60% (p = 0.003) comparing to non-treated conditions ( Figure 6L).
We also tested the combination of N6L and INK128 on mPDAC tumor-derived organoid viability ( Figure 6M,N). The organoid cultures were established from mPDAC tumors generated in mice as previously described [22,41]. An analysis of MTS data by the LOEWE method in the PDAC-derived organoids models demonstrated that the combination of N6L and INK128 had an inhibitory synergistic effect on the cell viability ( Figure 6M,N).  Mechanistically, we confirmed by Western-blotting analysis the inactivation of mTOR pathway upon the N6L and mTORi dual treatment ( Figure 7A,B), where a strong decrease in RPS6 and 4EBP1 phosphorylation in mPDAC and PDX cells is observed. We obtained the same results on the other human PDAC cell lines MiaPaca and Panc ( Figure S5F and Figure S6F, respectively). Of note, the dual mTORC1/mTORC2 INK128 inhibitor seemed to have a strong effect in comparison to the other mTOR inhibitors when combined with N6L.
Collectively, these results demonstrate that NCL targeting by N6L treatment and inhibition of the TOR pathway act synergistically to inhibit PDAC cell growth.

Discussion
Nucleolin (NCL) overexpression is associated with a poor prognostic for many cancers, including pancreatic cancer. PDAC remains one of the most aggressive and lethal diseases because of the limited therapeutic options. The nucleolin aptamer N6L shows antitumor effects on breast cancer [17], glioblastoma [19], non-small cell lung carcinoma [20], and PDAC [9,21].

Discussion
Nucleolin (NCL) overexpression is associated with a poor prognostic for many cancers, including pancreatic cancer. PDAC remains one of the most aggressive and lethal diseases because of the limited therapeutic options. The nucleolin aptamer N6L shows antitumor effects on breast cancer [17], glioblastoma [19], non-small cell lung carcinoma [20], and PDAC [9,21].
NCL is a multifunctional protein that is required for cell growth and proliferation. The inhibition of NCL with aptamers such as N6L, AS1411, or through silencing with siRNA induces a profound remodeling of the cell (and nucleus) morphology with a strong effect on the cell cycle progression and on genes expressions [10,14]. Our study shows that PDAC cells treated with NCL inhibitors activate pathways that allow the production of proteins involved in translation, despite the global translational repression indicating that the cells are trying to overcome the inhibition of NCL function. The activation of the mTOR signaling pathway is part of this cellular response to NCL inhibition to help the cell to survive despite the inhibition of many of the vital functions of NCL. The synergistic effect of NCL and mTOR inhibitors shows that the activation of the mTOR pathway is important for the survival of the cells when NCL is inhibited.
We found that although N6L strongly inhibits global translation, some mRNAs escape this global translational shutdown ( Figure 2). The expression of these mRNAs that continue to be recruited in polysomes in N6L-treated cells was either stable or lower ( Figure S2), indicating that this increased recruitment in polysomes was not the consequence of an increase in their transcription but rather the consequence of regulation at the translational level. These translationally up-regulated mRNAs are mostly implicated in cell metabolism and translation machinery.
Many of these mRNAs recruited in the polysomes under N6L treatment contains a 5 -Terminal oligopyrimidine (5 TOP) in their 5 UTR and encodes for ribosomal proteins and for factors regulating translation.
The 5 TOP motif begins with a C nucleotide directly adjacent to the cap structure, followed by a series of approximately 4-15 pyrimidines, often followed by a G-rich region [34]. The 5 TOP motif is highly conserved and is found in all human ribosomal proteins as well as non-ribosomal proteins involved in translation [42,43]. This shared TOP motif allows cells to quickly modulate the expression of proteins involved in ribosome production and protein synthesis in response to changes in cellular homeostasis. We hypothesize that the cells activate the translation of these 5 TOP mRNAs to counteract the negative effect of N6L on cell survival. Mechanistically, the translation of these 5 UTR motifs is regulated by the mTOR pathway. mTOR pathway is crucial to control cell growth and survival in physiological as well as pathological conditions. It acts in two distinct mTOR complexes, mTORC1 and mTORC2, which differ in associated proteins and by sensitivity to rapamycin and its derivatives [44]. The major function of mTORC1 is the promotion of cellular growth and proliferation via increasing protein synthesis and inhibition of autophagy. Functions of mTORC2 are less well studied and include organization of the actin cytoskeleton, control of cellular metabolism, and anti-apoptotic properties via stimulation of the AKT-FOXO pathway.
The inhibition of the mTORC1 pathway leads to a strong decrease in the translation of transcripts containing the 5 TOP motif [35,45]. However, how mTOR regulates 5 TOP mRNA translation remains not completely understood.
Under mTOR activation, 4EBP1 is phosphorylated, which prevents 4EBP1 from disrupting the interaction of eIF4E and the mRNA cap, and has been shown to selectively enhance 5 TOP mRNA translation [35,39]. The role of LARP1 in the regulation of TOP mRNAs remains controversial. While LARP1 was shown to regulate TOP mRNA stability, it has also been described as a positive or negative regulator of TOP mRNA translation, depending on the context and the energy status of the cells [47]. LARP1 interacts with the raptor and is phosphorylated by mTORC1, which is thought to modulate its mRNAbinding activity [45,48]. In our study, the phosphorylation of 4EBP1, RPS6, and AKT and the accumulation of LARP1 upon N6L treatment enhance the 5 TOP mRNA translation show that the mTOR pathway is activated and that LARP1 could act as a positive regulator.
The mTOR pathway is reported to be aberrantly active in several cancers, including PDAC [49,50], in part due to mutations in upstream regulatory molecules including PTEN, AKT, and TSC1/2. Most PDAC cancers have RAS mutations leading to activation of the MEK/ERK pathway, which can inactivate TSC1/2 and thereby activate TORC1. Therefore, mTOR is a relevant target for the treatment of PDAC [49]. Several types of mTOR inhibitors such as rapamycin, its rapalogs, and dual mTORC1/mTORC2 inhibitors have been examined in various cancer models, including PDAC [49,51]. However, the effects of mTOR inhibitors used as monotherapy in clinical trials have not shown significant effects and are sometimes dampened by several resistance mechanisms [52]. Combined therapies with mTOR inhibitors and other pathway inhibitors or conventional therapies are under investigation in preclinical and clinical trials in different tumor types [53]. Hence, novel therapeutic strategies based on mTOR inhibition still need to be developed. In PDAC, there is an increased need for new therapeutic strategies with better efficacy and less toxicity.
In our study, NCL targeting by N6L enhances the mTOR pathway-inducing reprogramming of the translatome by an increase in specific mRNAs (including 5 TOP mRNAs) in the polysomes while the translation of most of the mRNAs is strongly down-regulated. This specific recruitment of mRNAs coding for ribosomal proteins and for factors required for the regulation of translation could be involved in mechanisms put in place by the cell in reaction to N6L to try to better resist the inhibitory effects of the drugs.
Indeed, when we combined N6L with mTOR inhibitors, we observed a synergistic inhibitory effect in different 2D and 3D preclinical models including, PDX and organoidderived tumors that faithfully reproduce the molecular features of PDAC human tumors, display a high predictive value for clinical efficacy in patients and are critically needed for the development of new treatments. We used two classes of mTORi, allosteric mTORC1 inhibitor rapamycin (RAPA, Sirolimus) and dual mTORC1/mTORC2 inhibitors AZD2014 (Vistusertib) and INK128 (Sapanisertib) [38,39]. Furthermore, we showed that this synergistic effect of inhibiting cell proliferation and viability is maintained even when the PDAC cells are co-treated with low doses of N6L and mTORi (data not shown) that can help to prevent side effects due to mTORi toxicity in the clinic by reducing mTORi doses. In addition, we confirmed by Western-blotting analysis the inactivation of the mTOR pathway upon the N6L and mTORi dual treatment.

Conclusions
In this study, we show that the NCL aptamer's strong inhibition of PDAC cell growth is coupled with the reprogramming of the translational machinery through the activation of the mTOR signaling pathway. We further demonstrate that NCL targeting by N6L treatment sensitizes PDAC cells to mTOR inhibitors. We propose a novel therapeutic strategy based on mTOR inhibition combined with NCL targeting by N6L to treat PDAC.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/cancers13194957/s1, Table S1: RNAseq results of control and N6L-treated PDAC cells. Translational index calculation and analysis of 5 TOP-containing mRNAS. Table S2: list of antibodies used in the studies. Figure S1: NCL targeting by N6L decreases mPDAC cell proliferation by inducing apoptosis. Figure S2: NCL targeting by N6L dysregulates mPDAC cell translation. Figure S3: The impact of mTOR inhibitors treatment on mPDAC and PDX cell growth. Figure S4: Synergistic effects of N6L/mTORi combination on MiaPaca cells. Figure S5: Synergistic effects of N6L/mTORi combination on Panc cells Figure S6: Originals western blot images.