Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 30.
Published in final edited form as: J Drug Deliv Sci Technol. 2022 May 10;72:103404. doi: 10.1016/j.jddst.2022.103404

Intercellular delivery of therapeutic oligonucleotides

Virginijus Valiunas 1, Chris Gordon 1, Laima Valiuniene 1, Daniel Devine 1,*, Richard Z Lin 1, Ira S Cohen 1, Peter R Brink 1
PMCID: PMC9886232  NIHMSID: NIHMS1865008  PMID: 36721641

Abstract

One promising approach to cancer therapeutics is to induce changes in gene expression that either reduce cancer cell proliferation or induce cancer cell death. Therefore, delivering oligonucleotides (siRNA/miRNA) that target specific genes or gene programs might have a potential therapeutic benefit. The aim of this study was to examine the potential of cell-based delivery of oligonucleotides to cancer cells via two naturally occurring intercellular pathways: gap junctions and vesicular/exosomal traffic. We utilized human mesenchymal stem cells (hMSCs) as delivery cells and chose to deliver in vitro two synthetic oligonucleotides, AllStars HS Cell Death siRNA and miR-16 mimic, as toxic (therapeutic) oligonucleotides targeting three cancer cell lines: prostate (PC3), pancreatic (PANC1) and cervical (HeLa). Both oligonucleotides dramatically reduced cell proliferation and/or induced cell death when transfected directly into target cells and delivery hMSCs. The delivery and target cells we chose express gap junction connexin 43 (Cx43) endogenously (PC3, PANC1, hMSC) or via stable transfection (HeLaCx43). Co-culture of hMSCs (transfected with either toxic oligonucleotide) with any of Cx43 expressing cancer cells induced target cell death (~20% surviving) or senescence (~85% proliferation reduction) over 96 hours. We eliminated gap junction-mediated delivery by using connexin deficient HeLaWT cells or knocking out endogenous Cx43 in PANC1 and PC3 cells via CRISPR/Cas9. Subsequently, all Cx43 deficient target cells co-cultured with the same toxic oligonucleotide loaded hMSCs proliferated, albeit at significantly slower rates, with cell number increasing on average ~2.2-fold (30% of control cells) over 96 hours. Our results show that both gap junction and vesicular/exosomal intercellular delivery pathways from hMSCs to target cancer cells deliver oligonucleotides and function to either induce cell death or significantly reduce their proliferation. Thus, hMSC-based cellular delivery is an effective method of delivering synthetic oligonucleotides that can significantly reduce tumor cell growth and should be further investigated as a possible approach to cancer therapy.

Keywords: gap junction, siRNA, miRNA delivery, extracellular delivery, intercellular delivery

Graphical Abstract

graphic file with name nihms-1865008-f0001.jpg

INTRODUCTION

RNA interference (RNAi) reduces the expression of a single gene product and so potentially could be the most targeted therapeutic available. RNAi is mediated by two kinds of regulatory RNAs, small interfering RNA (siRNA) and microRNA (miRNA), both of which can have profound effects on cell function [1–4]. While siRNA oligonucleotides can be designed to silence the synthesis of any protein in a cell, naturally occurring miRNA can turn multiple gene programs on and off [5, 6] by inhibiting the synthesis of more than one mRNA. Widespread use of RNAi technology has been hampered by less than effective delivery strategies with multiple toxicities which include side effects associated with the high circulating siRNA concentrations necessary to achieve therapeutic concentration levels in a target tissue [7]. Despite multiple trials, clinical use of oligonucleotide technology has had limited success [8, 9]. One approach that potentially bypasses these major negative side effects is cellular delivery of siRNA and/or miRNA to target cells and tissues, thus avoiding the presence of oligonucleotides in systemic circulation.

This study set out to evaluate the therapeutic potential of a cellular delivery system in an in vitro setting, as a prelude to considering it as a potential tool for delivery in vivo. We and others have demonstrated that certain specific oligonucleotides can move from one cell to another via the two intrinsic intercellular pathways, gap junctions [10, 11] and vesicular transfer via exocytosis of exosomes [11–13]. A cellular delivery system is reliant on three factors: 1) the best choice for delivery/source cell type, 2) the ability to load/transfect the chosen source cell type with a toxic/therapeutic oligonucleotide, and 3) demonstration of delivery of the toxic/therapeutic oligonucleotide to target cells from source cells via either or both intercellular pathways.

In this study, we used multiple cancer cell models for cell proliferation and investigated the effects of toxic/therapeutic oligonucleotides on single cancer cell populations and co-cultured populations of source and target cells. We utilized three human cancer cell lines: prostate (PC3), pancreatic (PANC1), and cervical (HeLa) as target cells. Our delivery (source) cells were human mesenchymal stem cells (hMSCs) transfected with synthetic oligonucleotides AllStars HS Cell Death siRNA (AllStars siRNA) or miR-16 mimic, since both have been shown to induce apoptosis or senescence [5, 14–16]. Our results demonstrate that miRNA/siRNA delivery occurs via both intercellular pathways. Further, the intercellular delivery of the toxic oligonucleotides led to target cancer cell death and/or reduced proliferation.

MATERIALS AND METHODS

Cells and culture conditions

Experiments were carried out using three human cancer lines: PC3, PANC1, and HeLa. All three cancer lines express Cx43 either endogenously (PC3 and PANC1) or were stably transfected with wild-type Cx43 (HeLaCx43). Human mesenchymal stem cells (hMSCs) endogenously expressing Cx43 [17] were used as delivery vehicles (source cells). Cx43 expression was demonstrated with Western blots and immunohistochemistry.

HeLa, PANC1 and PC3 cells were cultivated in Dulbecco’s modified Eagle medium (DMEM, Gibco BRL, Life Technologies) supplemented with 10% FBS (HyClone), 2mM L-glutamine, 100U/ml penicillin (Gibco BRL, Life Technologies), and 100μg/ml streptomycin (Gibco BRL, Life Technologies). 100μg/ml hygromycin (Sigma) was used to select the stably transfected HeLaCx43 cells. Human mesenchymal stem cells (hMSCs; mesenchymal stem cells, human bone marrow; Poietics™) were purchased from Clonetics/BioWhittaker (Walkersville, MD, USA) and cultured in mesenchymal stem cell (MSC) growth medium (Poietics™MSCGM; BioWhittaker).

To achieve heterologous cell to cell coupling, hMSCs were co-cultured with HeLa, PANC1 or PC3. For the purposes of identification, some cells were transfected with red (RFP) or green (GFP) fluorescent proteins (pIRES2-DsRed2 or pIRES2-EGFP, Clontech Laboratories, Inc.). Characterization of these cells, culture conditions, and staining methods for the identification of specific cells, have been previously described [18–22]. Oligonucleotide transfer experiments were carried out on cells cultured for 24 to 96 hours.

CRISPR/Cas9 Cx43 KO

To knock out endogenous Cx43 expression in PANC1 and PC3 cell lines, we used a Cx43 CRISPR/Cas9 gene KO Plasmid (Santa Cruz) according to the manufacturer’s protocol. Briefly, cells were plated on a 35mm dish (Falcon) with medium lacking antibiotics at ~60% of cell density at the time of the forward transfection. The transfection was performed using Lipofectamine 3000 reagent (Invitrogen) in OPTI-MEM medium (Gibco), connexin-43 CRISPR/Cas9 KO Plasmid (h) (sc-400241, Santa Cruz), connexin-43 HDR Plasmid (h) (sc-400241-HDR, Santa Cruz), and P3000 reagent (Invitrogen). The diluted components were then combined and the complex was added directly to the culture dish and incubated at 37°C 5% CO2 for ~48 hours. Successful transfection was visually confirmed via detection of the green fluorescent protein (GFP) while successful co-transfection of the CRISPR/Cas9 KO and HDR Plasmid was confirmed by detecting red fluorescent protein (RFP) using fluorescent microscopy. The successfully co-transfected cells were selected and placed in a new culture medium containing puromycin (Sigma, P9620-10) at 10μg/ml. After many rounds of cell division, the cells were transferred from the 35mm dish to a larger vessel and puromycin selection continued. Due to a mixed population of RFP expression, flow cytometry was performed for RFP cell selection. The sorted cells were maintained under puromycin control. The Cx43 KO was confirmed with Western blot analysis.

Cell preparation for oligonucleotide delivery

As a positive control for oligonucleotide (siRNA/miRNA) delivery, we used AllStars siRNA (AllStars HS Cell Death Control siRNA (Qiagen)), which according to the manufacturer’s description, is a blend of siRNAs that target human genes essential for cell survival and induce cell apoptosis [14, 15]. Cy5-labeled and non-labeled 22-nucleotide long MISSION microRNA mimics (hsa-miR-16) were obtained from Sigma-Aldrich. As a negative control for non-specific effects of oligonucleotide delivery, we used custom non-targeting (scrambled) control siRNA (Eurofins Operon). Cells were transfected with 100 nM of AllStars siRNA, miR-16 mimic or negative control siRNA using Lipofectamine RNAiMAX (Invitrogen) as directed by the manufacturer’s protocols.

For cellular oligonucleotide delivery, hMSC source cells were transfected with miR-16 mimic or AllStars siRNA 24 hours before they were co-cultured with the recipient HeLa, PANC1 or PC3 cells. After 24 hours, the transfection medium from the source hMSCs was aspirated, the cells were repeatedly (3 times) rinsed with PBS and washed with fresh medium following exchange of the complete medium. The cells were seeded at a 1:1 to 1:5 hMSCs to target cells ratio and were co-cultured together for up to 96 hours.

Camera imaging and cell counting approach

Cell growth/proliferation rates were assessed by cell counting methods. Transfected cells were seeded or co-cultured with recipient cells in 35 mm Petri dishes with marked bottoms. The cells were imaged by light and epifluorescence microscopy using a 14-bit, 16364-pixel, gray-scale digital CCD-camera (Axiocam, Zeiss) connected to an inverted Olympus IX 71 epifluorescence microscope. Cells were visualized and counted using Axiovision (Zeiss) and/or ImageJ software in designated/marked areas of interest 24 hours after transfection/passage. This number was used to normalize the cell numbers in that designated area at every subsequent 24-hour time mark up to 96 hours.

Electrophysiological measurements and cell to cell dye transfer

Electrophysiological and fluorescence recordings were carried out on cell pairs cultured for 1 to 4 days. Glass cover slips with adherent cells were transferred to an experimental chamber mounted on the stage of an inverted Olympus IX 71 epifluorescence microscope. Cells were perfused at room temperature (~22 °C) with bath solution containing (in mM): 140 NaCl, 1 MgCl2, 5 KCl, 2 CaCl2, 2 CsCl, 2BaCl2, 5 HEPES (pH 7.4), and 10 glucose. Functional gap junction coupling in parental and Cx43 KO cells was tested via double patch clamp and/or cell to cell Lucifer Yellow (LY) fluorescent dye transfer [23]. Patch pipettes were pulled from glass capillaries (code GC150F; Harvard Apparatus) with a horizontal puller (DMZ-Universal, Zeitz-Instrumente) and were filled with saline containing (in mM): 120 K-aspartate, 10 NaCl, 3 MgATP, 5 HEPES (pH 7.2), and 10 EGTA (pCa ~8). Filled patch pipettes typically had a resistance of 2 to 5 MΩ.

Signal recording and analysis

Voltage and current signals were recorded using patch clamp amplifiers (Axopatch 200). The current signals were digitized with a 16-bit A/D Converter (Digidata 1440A) and stored within a personal computer. Data acquisition and analysis were performed with pClamp10 software (Molecular Devices). Curve fitting and statistical analysis were performed using SigmaPlot and SigmaStat, respectively (Jandel Scientific). The t-test was used for all cases unless otherwise noted. The results are presented as mean ± SE (P<0.05 was considered significant).

RESULTS

Cx43 KO cells exhibit lack of gap junction-mediated communication

We have previously demonstrated that both gap junctions and extracellular-mediated (exocytotic/endocytotic) pathways have potential in delivering siRNA [11]. To further investigate the effectiveness of these intercellular pathways in delivering toxic (therapeutic) oligonucleotides to target cancer cells, we proceeded to eliminate gap junction-mediated delivery. We used CRISPR/Cas9 gene editing to knock out endogenous Cx43 expression in two target cancer cell lines, PC3(Cx43 KO) and PANC1(Cx43 KO) [24–28]. Likewise, for this same purpose we also used connexin deficient HeLa parental (wild-type) cells [29]. Western blot analysis shown in Figure 1A indicates Cx43 expression in PC3 cells. Functional gap junctions in PC3 cells were confirmed by junctional current recordings obtained by dual whole cell patch clamp (right panel in Figure 1A) and cell to cell LY dye spread (Figure 1B). CRISPR/Cas9 modified PC3(Cx43 KO) cells showed no expression of Cx43 (Figure 1C). In addition, a dual whole cell patch clamp demonstrates that there is no measurable junctional conductance (right panel in Figure 1C) and the lack of LY dye transfer between the pairs of modified cells reinforces the absence of functional coupling and permeability between PC3(Cx43 KO) cells (Figure 1D).

Figure 1. Cx43 expression and functional gap junctions.

Figure 1.

Open in a new tab

(A) Left: Western blot analysis of Cx43 in prostate cancer cell line PC3 and in HeLa cells stably expressing Cx43 (HeLaCx43). Right: Junctional current recording from a PC3 cell pair. (B) Multi-cell preparation where the patch electrode on the right contains Lucifer Yellow (LY). Fluorescence emission monitored over time (7- and 15-min recordings shown) revealed LY transfer from the injected cell to the two neighboring PC3 cells. (C) Left: RFP expression indicates CRISPR/Cas9 modified PC3(Cx43 KO) cells with disrupted Cx43 expression as shown by Western blots (middle panel). Zero junctional current (right panel C) and lack of LY transfer (D) indicates the absence of functional coupling in PC3(Cx43 KO) cells. Scale bar, 10μm.

Figure 2 illustrates similar results obtained with pancreatic cancer cells PANC1. When compared to parental PANC1 cells, PANC1(Cx43 KO) cells also exhibited a loss of Cx43 expression (Figure 2A) and junctional communication (Figure 2D). LY transfer was not detected in PANC1(Cx43 KO) cell pairs, but was detected in parental PANC1 cell pairs (Figures 2C–D). Western blots in Figure 2B show that HeLaCx43 cells abundantly express Cx43, but connexin deficient HeLaWT cells do not. Conductance and permeability properties of HeLaCx43 gap junctions have been studied extensively [19, 23, 30], as has the lack of communication in HeLaWT cells [18, 29]. hMSCs also showed expression of Cx43 (Figure 2B) and cell to cell LY dye spread consistent with previous reports [17, 23].

Figure 2. Cx43 expression in cell lines.

Figure 2.

Open in a new tab

(A) Western blot analysis: Cx43 is expressed in pancreatic cancer cell line PANC1 but not in modified PANC1 CRISPR/Cas9 Cx43 KO cells. (B) Western blot analysis of Cx43 in hMSCs, HeLaCx43, and HeLaWT cells. HeLaWT cells do not express Cx43. (C-E) Cell to cell LY transfer via gap junctions detected in PANC1 (C) and hMSCs (E) but not in PANC1(Cx43 KO) cells (D). Scale bar, 10μm.

Delivery of oligonucleotides to target cells

The effects of miR-16 mimic and AllStars siRNA on target cancer cells’ proliferation were studied on single cell populations and co-cultured populations of source cells and target cells. Our delivery (source) cells are hMSCs transfected with either the miRNA mimic or the AllStars siRNA.

Our experimental design is illustrated in Figure 3. Both paradigms, the direct transfection and co-culture of transfected hMSCs are shown in Figure 3A and were used in cell counting/cell proliferation measurements. Figure 3B shows that virtually every source hMSC contains Cy5-tagged miR-16 mimic 24 hours after direct transfection with 100 nM of miR-16 mimic Cy5.

Figure 3. An assay for cellular delivery of miRNA mimic/siRNA.

Figure 3.

Open in a new tab

(A) Schematics of experimental conditions. (B) hMSCs 24 hours after transfection with 100 nM of miR-16 mimic Cy5 (red) using lipofectamine RNAiMAX. Scale bar, 10μm.

Cellular delivery of miR-16 mimic is illustrated in Figure 4. Figures 4A–C are fluorescence images of a 24-hour co-culture of cell pairs composed of one source cell (hMSC) transfected with red-tagged miR-16 mimic Cy5 and a target cell (PC3, PANC1 or HeLaCx43) expressing GFP that demonstrate miR-16 mimic Cy5 transfer from source to target cell. In these cases, both intercellular pathways can participate in the delivery of miR-16 mimic from hMSCs to target cells [11].

Figure 4. Detection of cell to cell transfer of oligonucleotides through Cx43.

Figure 4.

Open in a new tab

Confocal microscopy images of a co-culture of hMSCs pre-loaded with miR-16 mimic Cy5 and PC3 (A), PANC1 (B) and HeLaCx43 (C) cells, all expressing GFP. In all cases, after 24 hours in co-culture, red-labeled miR-16 mimic Cy5 can be detected in the recipient cells (green) of the pairs. Scale bars are 10μm.

Direct transfection of oligonucleotides

As a form of control to test the accuracy of our cell proliferation assay, we used AllStars siRNA that induces apoptosis in cells [14, 15]. We directly transfected 100 nM AllStars siRNA for 24 hours to determine the mixture’s effect on inhibiting cell proliferation rates of PANC1, PC3, HeLaCx43 and hMSC populations.

Figure 5 illustrates the ability of AllStars siRNA to reduce cell proliferation. In all instances, direct transfection of 100 nM AllStars siRNA (Figure 5A) triggered total cell death in 96 hours with almost no survivors, as shown in the example for PANC1 cells in the lower panel of Figure 5B. Figure 5C summarizes the proliferation rates for AllStars siRNA treated cells. For all cell types investigated, cells were counted in 4 different designated areas (dishes) at every 24 hours up to 96 hours after transfection/passage.

Figure 5. Direct transfection of 100 nM AllStars siRNA inhibits proliferation and induces cell death in tumor cells.

Figure 5.

Open in a new tab

(A) Schematic of experiment. (B) Control (upper panel) and siRNA transfected (lower panel) PANC1 cells 24 and 96 hours after passage/transfection. Cells were counted (yellow numeral) in designated areas 24 hours after transfection/passage and this number was used to normalize the cell numbers in that designated area at every subsequent 24-hour mark. (C) Plots of tumor cell growth for control (green symbols) PANC1 cells and hMSCs, and siRNA transfected (red symbols) PC3, PANC1, and HeLaCx43 cells. Data represent means ± S.E. obtained from a total of n=4 cell counts in each cell group. Scale bars, 50μm.

Examples of normal growth rates for control PANC1 cells (upper panel Figure 5B) and hMSCs are plotted to illustrate typical proliferation rates along with the effects of AllStars siRNA for all cell types (Figure 5C). After 96 hours in culture, the number of PANC1 and hMSC cells increased 6.3 ± 0.2- and 3.3 ± 0.2-fold, respectively. In contrast, all AllStars siRNA transfected cells (PANC1, PC3, HeLaCx43 and hMSC) exhibited a dramatic cell number decrease (P<0.001) over the 96 hours after transfection. The AllStars siRNA data is consistent with previous observations [14, 15, 31] and justifies the cell proliferation rate/cell counting methods we are using to assess target cell proliferation.

A similar experiment is shown in Figure 6 where a miR-16 mimic was directly transfected for 24 hours into PANC1, PC3 and hMSCs. Figure 6B compares cultured PC3 control cells and miR-16 mimic treated PC3 cells 24 and 96 hours after transfection. Figure 6C summarizes proliferation rates for miR-16 mimic transfected PC3, PANC1 and hMSC cells, as well as control PC3 and PANC1 cells. At 96hours, PC3 (n=4) and PANC1 (n=4) cells exhibited a 7.3 ± 0.4- and 5.3 ± 0.2-fold cell number increase, respectively. In all cases, miR-16 mimic inhibited the normal proliferation rate of all cell types but did not completely eliminate cells. The cell number increased only marginally over 96 hours (1.3 ± 0.1-fold) for both PANC1 cells and hMSCs. The number of miR-16 mimic treated PC3 cells declined (0.5 ± 0.1-fold change) over 96 hours. These proliferation rates for all miR-16 mimic treated cells were significantly different than those of not transfected control cells (P<0.001).

Figure 6. miR-16 mimic inhibits tumor cell growth.

Figure 6.

Open in a new tab

(A) Schematic of direct transfection of miR-16 mimic to cells. (B) PC3 cells transfected with 100 nM miR-16 mimic (upper panel) and control cells (lower panel) after 24 and 96 hours in culture. Cells were counted (yellow numeral) in designated areas 24 hours after transfection/passage and this number was used to normalize the cell numbers in that designated area at every subsequent 24-hour mark. (C) Plots of normalized cell number (means ± S.E., n=4 for each cell type) vs. time for control (green symbols) and miR-16 mimic transfected (red symbols) PC3, PANC1 and hMSC cells. Scale bars, 50μm.

This result is consistent with PANC1, PC3, and hMSCs becoming senescent and/or where cell proliferation rate approximates the rate of cell death.

An important issue associated with direct transfection of toxic/therapeutic oligonucleotides is controlling whether the reduction in cellular proliferation is not from transfection toxicity itself. In one series of experiments, we performed a mock transfection (no toxic oligonucleotides) of target cells using the adequate volume of transfection agent lipofectamine RNAiMAX.

Figure 7A (upper panel) shows cultured PANC1 cells 24 and 96 hours after mock transfection, exhibiting substantial cell proliferation. Plots of proliferation rates of control and mock transfected PANC1 cells are compared in Figure 7B. Over 96 hours, the number of mock transfected (RNAiMAX) PANC1 and control PANC1 cells increased on average 5.8 ± 0.2- and 5.5 ± 0.3-fold, respectively. However, the difference in proliferation rates between these two cell populations was not statistically significant (P=0.381). These data show that lipofectamine RNAiMAX does not affect PANC1 cell proliferation, thus ruling out transfection toxicity in our experimental conditions. Similar results were obtained with PC3 and HeLaCx43 cells (data not shown).

Figure 7. Comparison of transfection effects on cell proliferation.

Figure 7.

Open in a new tab

(A) PANC1 cells transfected with only lipofectamine RNAiMAX (upper panel) or negative control siRNA (lower panel) after 24 and 96 hours in culture show an increase in cell number over time. Yellow numerals in the bottom left corners correspond to the cell count in each designated area. (B) Plots of normalized cell number (means ± S.E., n=4 in each group) vs. time for control (Inline graphic), mock transfected (RNAiMAX) (Inline graphic), and control siRNA transfected (Inline graphic) PANC1 cells. Scale bars, 50μm.

In another series of control experiments, PANC1 cells were directly transfected with a negative control (scrambled) siRNA. Control siRNA neither induced cell death nor inhibited cell growth, as illustrated in the lower panel of Figure 7A. However, a significant decrease in the proliferation rate of transfected cells was observed. PANC1 cells treated with negative control siRNA exhibited 3.7 ± 0.2-fold cell number increases over a 96-hour time period, which was significantly lower (P=0.002) than that of the control cells (Figure 7B). Nonetheless, the PANC1 cell line transfected with control siRNAs still proliferated significantly (P<0.001) when compared to AllStars siRNA or miR-16 mimic transfected cells (Figures 5 and 6). This confirms that the cell death and dramatically reduced or fully inhibited cell proliferation shown in Figures 5 and 6 is almost completely induced by the toxicity of the therapeutic oligonucleotides (AllStar siRNA or miR-16 mimic) and is only marginally related to the transfection toxicity.

Co-culture of target cells with toxic oligonucleotide loaded hMSCs

To better understand the intercellular delivery of miR-16 mimic, hMSCs transfected with miR-16 mimic were co-cultured with PC3 or PANC1 wild-type cells expressing Cx43 and also with Cx43 knockout cells (PC3(Cx43 KO) or PANC1(Cx43 KO)) for various time intervals (24, 48, 72 and 96 hours). A comparison of proliferation rates for Cx43 expressing target cells and for Cx43 KO target cells is shown in Figure 8. Examples of images of PC3 cells in culture are shown in Figures 8A (when Cx43 is expressed) and 8B (Cx43 KO). Summary plots of proliferation rates are presented in Figure 8C. The growth rate for control Cx43 expressing PC3 and PANC1 cells increased 7.6 ± 0.5- and 6.1 ± 0.1-fold over 96 hours, respectively (left and right panels Figure 8C). In the presence of miR-16 mimic transfected hMSCs, wild-type PC3 and PANC1 cells (red symbols in Figure 8C) did not proliferate, behaving similarly to PC3 and PANC1 cells that had been directly transfected with miR-16 mimic (as shown in Figure 6). The marginal 1.2 ± 0.1- and 1.02 ± 0.05-fold cell number increases, corresponding to an ~85% proliferation reduction, were significantly different from the proliferation rates of control PC3 and PANC1 cells (P<0.001 in both cases). CRISPR/Cas9 modified PC3(Cx43 KO) and PANC1(Cx43 KO) cells (gray symbols) did proliferate when co-cultured with miR-16 transfected hMSCs, albeit at a slower rate (2.0 ± 0.3- and 2.6 ± 0.3-fold cell number increases, respectively). At the 96-hour mark, when co-cultured with miR-16 mimic transfected hMSCs, PC3(Cx43 KO) cell number is reduced ~3.8-fold relative to control cells (P<0.001), a ratio of cell numbers between wild-type control PC3 (green symbols) and PC3(Cx43 KO) (gray symbols) cells (left panel Figure 8C). When wild-type PC3 cells expressing Cx43 gap junctions (red symbols) are co-cultured with miR-16 mimic transfected hMSCs, their cell number is reduced ~6.3-fold in comparison to control PC3 cells (P<0.001). This suggests a ~2.7-fold effect attributable to gap junctions. Note that hMSCs do not appreciably proliferate (Figure 6C) when transfected with miR-16 mimic, thus the proliferation rates shown in Figure 8 are principally due to PC3 cells. For PANC1 cells (right panel Figure 8C), the cell number is reduced ~6.0-fold for wild-type and ~2.3-fold for PANC1(Cx43 KO) cells. This suggests a similar ~2.7-fold effect attributable to Cx43 gap junctions.

Figure 8. hMSCs containing miR-16 mimic reduce cancer cell proliferation.

Figure 8.

Open in a new tab

(A) hMSCs pre-loaded with miR-16 mimic in co-culture with PC3 cells for 24 and 96 hours. Fluorescence indicates GFP expression in PC3 cells. (B) Co-culture of hMSCs pre-loaded with miR-16 mimic and PC3(Cx43 KO) cells at 48 and 96 hours. Fluorescence indicates RFP expression in PC3(Cx43 KO) cells. (C) Summary plots of normalized cell number vs. time for co-cultured cells. Left panel: hMSC(miR-16 mimic) with PC3 (Inline graphic), hMSC(miR-16 mimic) with PC3(Cx43 KO) (Inline graphic), and control PC3 cells (Inline graphic). Right panel: hMSC(miR-16 mimic) with PANC1(Inline graphic), hMSC(miR-16 mimic) with PANC1(Cx43 KO) (Inline graphic), and control PANC1 cells (Inline graphic). Cells were counted and plotted (mean ± S.E., n=4) according to the same procedure as in Figure 5. Scale bars, 50μm.

The experimental data showed that miR-16 mimic can be delivered to target cells by both intercellular delivery pathways. In both cases, when delivered via the vesicular/exosomal pathway or gap junctions, miR-16 mimic reduced the number of target cells. The reduced proliferation rates for PC3(Cx43 KO) and PANC1(Cx43 KO) cells suggest that target cells are in a senescent state or the proliferation rate approximates the cell death rate.

As shown in Figure 5, the effects of AllStars siRNA when directly transfected resulted in the death of all cells. Figure 9 illustrates the ability of AllStars siRNA to reduce cancer cell proliferation in co-culture. Proliferation of HeLa wild-type cells (no Cx43 expression) and HeLaCx43 cells co-cultured with AllStars siRNA loaded hMSCs is compared in Figures 9A and B. The images taken at 24 and 96 hours clearly indicate that hMSCs loaded with AllStars siRNA induced cell death in HeLaCx43 but not in HeLaWT cells in co-cultures. Figure 9C shows summarized plots of the data collected in all co-culture experiments with various combinations of hMSCs and target cells. At the 96-hour mark, when co-cultured with AllStars siRNA transfected hMSCs, HeLaWT cell number is reduced ~4.5-fold, whereas when gap junction expressing HeLaCx43 cells are co-cultured with AllStars siRNA transfected hMSCs, their cell number is reduced ~57-fold in comparison to control HeLaWT cells (P<0.001), with ~15% of cells surviving (left panel Figure 9C). Co-culture of hMSCs transfected with AllStars siRNA and Cx43 expressing PC3 or PANC1 cells similarly resulted in zero proliferation of hMSCs and near complete cell death of the target cancer cells, with ~25% surviving at 96 hours (middle and left panels in Figure 9C). All three Cx43 expressing cell types were unable to proliferate and exhibited declining cell numbers significantly different than those of control cells: HeLaCx43 (0.15 ± 0.1-fold, n=4, P<0.001), PC3 (0.27 ± 0.02-fold, n=4, P<0.001), and PANC1 (0.25 ± 0.03-fold, n=4, P<0.001).

Figure 9. hMSCs loaded with AllStars siRNA inhibit tumor cell growth.

Figure 9.

Open in a new tab

(A) Co-culture of hMSCs pre-loaded with 100 nM AllStars siRNA and HeLaCx43 cells at 24- and 96-hour time marks showing cell number decreases over time. (B) Co-culture of hMSCs pre-loaded with 100 nM AllStars siRNA and HeLaWT cells at 24- and 96-hour marks shows the cell number increases over time. (C) Summary plots of normalized cell number (mean ± S.E., n=4) vs. time for co-cultured cells. Left panel: hMSC(siRNA) with HeLaCx43 (Inline graphic), hMSC(siRNA) with HeLaWT (Inline graphic), and control HeLaWT cells (Inline graphic). Middle panel: hMSC(siRNA) with PC3 (Inline graphic), hMSC(siRNA) with PC3(Cx43 KO) (Inline graphic), and control PC3 cells (Inline graphic). Right panel: hMSC(siRNA) with PANC1 (Inline graphic), hMSC(siRNA) with PANC1(Cx43 KO) (Inline graphic), and control PANC1 cells (Inline graphic). Cells were counted in designated areas 24 hours after co-culture (yellow numerals in (A) and (B)) and this number was used to normalize the cell numbers in that designated area at every subsequent 24-hour mark. Scale bars, 50μm.

Interestingly, co-culture of AllStars siRNA loaded hMSCs with PC3(Cx43 KO), PANC1(Cx43 KO) and HeLaWT cells resulted in the survival of a population of PC3(Cx43 KO), PANC1(Cx43 KO) and HeLaWT cells, which is the opposite of the effects of direct transfection with AllStars siRNA. For all three cell lines devoid of Cx43, the proliferation rate was reduced ~3–4.5-fold compared to controls (P<0.001), but in all cases cell number increased even in the presence of AllStars siRNA (HeLaWT: 1.9 ± 0.2-fold, PC3(Cx43 KO): 2.1 ± 0.07-fold and PANC1(Cx43 KO): 1.86 ± 0.05-fold). Note that hMSCs do not appreciably proliferate when transfected with AllStars siRNA (Figure 5C). These data show that both intercellular pathways are viable as active delivery systems and can be effectively used to reduce target cell proliferation.

Multiple reports have suggested that hMSCs derived from various sources may have varying inhibitory and stimulatory effects on tumor development [32–35]. To test whether hMSCs exhibit any anti- or pro-proliferative effects in our conditions, we performed control experiments co-culturing hMSCs not loaded with toxic oligonucleotides with the target PC3, PANC1 and HeLaCx43 cells. Examples of images illustrating GFP-expressing PC3 cell proliferation in co-culture with hMSCs are shown in Figure 10A. We compared proliferation rates at the 96-hour mark for different types of target cells obtained from cell monocultures (PC3, PANC1 and HeLaCx43) with those rates obtained from co-culture with hMSCs (PC3:hMSC, PANC1:hMSC and HeLaCx43:hMSC) as shown in Figure 10B. Analysis of cell number increase showed that hMSCs did not significantly affect proliferation for all types of target cells: control PC3 vs. PC3:hMSC (6.9 ± 0.3-fold vs. 6.3 ± 0.2-fold, P=0.082); control PANC1 vs. PANC1:hMSC (6.3 ± 0.2-fold vs. 5.8 ± 0.2-fold, P=0.1742); control HeLaCx43 vs. HeLaCx43:hMSC (10.3 ± 0.8-fold vs. 11.6 ± 1.1-fold, P=0.377).

Figure 10. Quantification of proliferation of target cells co-cultured with hMSCs.

Figure 10.

Open in a new tab

(A) Co-culture of hMSCs and PC3 cells at 24- and 96-hour time marks demonstrating target cell proliferation. PC3 cell number (yellow numerals) increased over time. Fluorescence indicates GFP expression in PC3 cells. (C) Summary plots of normalized cell number increase at 96 hours for control target cells (PC3, PANC1, HeLaCx43) vs. analogous target cells co-cultured with hMSCs (PC3:hMSC; PANC1:hMSC; HeLaCx43:hMSC). Bars represent means ± S.E. The numerals on the bars correspond to the number of designated areas (dishes) of counted cells. No significant differences between groups including the same cell types were observed (P=0.082; P=0.1742; P=0.377).

These data show a negligible effect of hMSCs on target cell proliferation in our experimental conditions. Consequently, that confirms that the toxic oligonucleotides delivered intercellularly by hMSCs have a major impact on target cell proliferation and cell death as shown in Figures 8 and 9.

DISCUSSION

Oligonucleotide therapeutics have an enormous beneficial potential for many diseases including cancers. However, their implementation has been limited by the high concentrations of the oligonucleotide that must be used because of their limited adsorption and rapid excretion. These high concentrations can cause systemic toxicity [36, 37]. However, to avoid these systemic problems, oligonucleotides could be delivered to spaces isolated from the systemic circulation. One example that the FDA has approved is an antisense product named SPINRAZA marketed by IONIS and delivered directly to the CSF to treat spinal muscular atrophy [38]. Improved systemic delivery has been achieved progressively with 1) nanoparticles containing many oligonucleotide molecules which were internalized more effectively, 2) targeting these nanoparticles to specific molecules on the target cell membrane surface, and 3) including lipid molecules to enhance cell permeation [38].

In this article we propose another potential improvement to systemic delivery, using source cells loaded with the therapeutic oligonucleotides to deliver to the target cells. Substantially larger concentrations of oligonucleotides can be loaded into a source cell than in alternative approaches to systemic delivery. The latest improvements in oligonucleotide chemistry can be employed and even a plasmid carrying the therapeutic oligonucleotide can be transfected into the source cell along with the oligonucleotide load for continuous production of the therapeutic nucleotide. Our results in this study provide evidence for the efficacy of this cell-based approach using hMSCs as the source cell in vitro. At this stage, the cell-based delivery of oligonucleotides has limited therapeutic use, such as being placed in the vicinity of remaining target cells following the surgical removal of a tumor. However, for this approach to achieve major success, the hMSCs must express a targeting molecule on their surface to avoid side effects due to non-target cell delivery [39].

Our results demonstrate a cell-based approach which can deliver toxic/therapeutic oligonucleotides in the form of siRNA and miRNA mimic via two naturally occurring intercellular pathways, namely gap junction channels and exocytosis/endocytosis/exosome [10–13, 40–42]. Furthermore, the effectiveness of both intercellular pathways was confirmed in delivering a toxic package which killed and/or reduced proliferation of malignant cells.

Exocytosis/endocytosis/exosomes represent an inherent property of all cells [13, 43, 44]. We previously showed that significant alteration of these cellular processes can be achieved pharmacologically, but the resultant side effects associated with long term experiments become problematic [11]. Thus, in this study, in order to differentiate between transfer via the two intercellular pathways, we eliminated gap junction mediated delivery by knocking out Cx43 expression with CRISPR/Cas9 gene editing. We have focused on Cx43 because hMSCs dominantly express Cx43 with little to no Cx40 [17]. Two of the target cancer cell lines, PC3 and PANC1, endogenously express Cx43 [24–28] while HeLa parental (wild-type) cells are connexin deficient [29], but were stably transfected with Cx43 (HeLaCx43) [19, 23].

Both of the toxic oligonucleotide types used in this study were chosen because of their potential to effectively inhibit cancer cell proliferation. According to the manufacturer (Qiagen), AllStars siRNA is designed to target essential cell survival genes. miR-16 also has also been shown to have multiple gene targets which cause cell cycle arrest, reduction in proliferation and migration, and trigger apoptosis [5]. Multiple genes encoding for cyclin D1 and cyclin E, which affect cell cycle and proliferation, have been shown to be down regulated by miR-16 [45–48]. Another one of the many target genes regulated by miR-16 is Bcl-2. Bcl-2 proteins regulate cell death (apoptosis) and are overexpressed in many types of human cancers [49]. Expression of miR-16 is inversely correlated with levels of Bcl-2, and Bcl-2 repression induces apoptosis in a leukemic cell line model [50]. Altered miR-16 expression has been also observed in prostate cancer [45, 51]. However, the investigation of specific oligonucleotides for a cancer cure is outside the scope of our study. Rather, our intent is a demonstration of the utility of cell-based delivery using both a single target and a multi-target approach to negatively impact cancer cell proliferation.

Direct transfection with AllStars siRNA induced apoptosis in all cancer cell lines and hMSC delivery cells, as is consistent with previous studies [14, 15]. Direct transfection of miR-16 mimic resulted in senescence with no net growth nor cell number decline for all target cell types as well as delivery hMSCs. Direct transfection is analogous to vesicular/exosomal mediated delivery and thus represents a control for the effectiveness of both toxic oligonucleotides (miR-16 mimic and AllStars siRNA) in the absence of gap junction-mediated delivery.

Intercellular delivery of miR-16 mimic or AllStars siRNA was effective in inhibiting the proliferation rates of PC3, PANC1, PC3(Cx43 KO), PANC1(Cx43 KO), as well as wild-type HeLa and Cx43 expressing HeLaCx43 cells, resulting in a decline in cell number for all target cell types.

Co-culture of hMSCs transfected with miR-16 mimic with any of the target cells expressing Cx43 triggered senescence in all three cancer cell lines. The efficiency of miR-16 mimic was reduced when target cells did not express Cx43. Although, the proliferation rate of connexin deficient target cells was significantly decreased as compared with control cells, this reduction was not as pronounced relative to Cx43 expressing target cells. Co-culture of AllStars siRNA transfected hMSCs with target cells returned the same result as direct transfection of all cell types. In this case, as hMSCs die, AllStars siRNA is released into the extracellular media. In effect, the hMSCs are acting like the Lipofectamine vehicle used for direct transfection. The net result for all Cx43 expressing cells after 96 hours is zero cellular proliferation and a reduction in cell number.

We compared cell-based delivery of oligonucleotides in wild-type Cx43 expressing cells and CRISPR/Cas9 Cx43 knockouts of those same cell types. Such a comparison allowed for the determination of vesicular/exosomal delivery independent of gap junction-mediated delivery. Particularly significant is the fact that Cx43 KO cells and wild-type HeLa were able to proliferate, although at a much more moderate rate in the presence of AllStars siRNA, suggesting two possibilities. One is that the lack of Cx43 expression has an effect on transcription/translation of anti-apoptotic activity [52, 53]. The other possibility is that one or more of the AllStars siRNAs is more dependent on an aqueous intercellular pathway exclusive of the extracellular space, resulting in incomplete triggering of apoptosis [54, 55].

Both naturally occurring intercellular pathways are also potentially useful in vivo, and for that reason it is crucial to determine their relative strengths and weaknesses. We showed that for in vitro delivery to be successful, source cells containing the toxic oligonucleotide payload must survive for some period of time. Thus, when subsequently placed with the target cells, they remain able to suppress cancer cell growth. With vesicular delivery, dilution and ubiquitous endocytosis in non-target cells [11] is a significant weakness in an in vivo setting unless a homing mechanism such as a CAR-T approach is used [56, 57]. Even in the presence of a homing mechanism, the vesicles can diffuse away from their target cells, limiting their effectiveness. For gap junction-mediated delivery, a different dilemma exists in vivo, as delivery cells expressing Cx43 would be able to couple with non-target Cx43 expressing cells, potentially producing negative side effects. In both cases, a homing approach would significantly improve the application of intercellular delivery by cells such as allogenic hMSCs, which are immune stealth for some time after placement in vivo.

CONCLUSION

hMSCs carrying therapeutic oligonucleotides can deliver their toxic package to target cancer cells in co-culture via both intercellular pathways (gap junctions and exocytosis/endocytosis/exosome) resulting in reduced cell proliferation or apoptotic cell death, with a lesser effect observed in Cx43 knock out target cells.

Although this suggests that cell-based delivery via hMSCs is a promising approach for cancer therapy, these cells must either be delivered into the vicinity of the cancer or have the capability to home to such locations in order for their therapeutic use to be optimized.

ACKNOWLEDGEMENTS

This study was supported by the National Institutes of General Medical Sciences grant R01GM088181 (to VV). This project was also partially supported by Mesoblast Ltd.

Footnotes

DECLARATION OF INTERESTS

The authors declare that they have no conflicts of interest.

REFERENCES

  • [1].Elbashir SM, Lendeckel W, Tuschl T, RNA interference is mediated by 21- and 22-nucleotide RNAs, Genes Dev. 15 (2001) 188–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Caplen NJ, Mousses S, Short interfering RNA (siRNA)-mediated RNA interference (RNAi) in human cells, Ann N Y Acad Sci. 1002 (2003) 56–62. [DOI] [PubMed] [Google Scholar]
  • [3].Miller H, Grollman AP, DNA repair investigations using siRNA, DNA Repair (Amst). 2 (2003) 759–763. [DOI] [PubMed] [Google Scholar]
  • [4].Zeng Y, Yi R, Cullen BR, MicroRNAs and small interfering RNAs can inhibit mRNA expression by similar mechanisms, Proc Natl Acad Sci U S A. 100 (2003) 9779–9784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Aqeilan RI, Calin GA, Croce CM, miR-15a and miR-16–1 in cancer: discovery, function and future perspectives, Cell Death Differ. 17 (2010) 215–220. [DOI] [PubMed] [Google Scholar]
  • [6].Lam JK, Chow MY, Zhang Y, Leung SW, siRNA Versus miRNA as Therapeutics for Gene Silencing, Mol Ther Nucleic Acids. 4 (2015) e252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Rettig GR, Behlke MA, Progress toward in vivo use of siRNAs-II, Mol Ther. 20 (2012) 483–512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Stein CA, Castanotto D, FDA-Approved Oligonucleotide Therapies in 2017, Mol Ther. 25 (2017) 1069–1075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Shen H, Sun T, Ferrari M, Nanovector delivery of siRNA for cancer therapy, Cancer Gene Ther. 19 (2012) 367–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Valiunas V, Polosina YY, Miller H, Potapova IA, Valiuniene L, Doronin S, Mathias RT, Robinson RB, Rosen MR, Cohen IS, Brink PR, Connexin-specific cell-to-cell transfer of short interfering RNA by gap junctions, J Physiol. 568 (2005) 459–468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Valiunas V, Wang HZ, Li L, Gordon C, Valiuniene L, Cohen IS, Brink PR, A comparison of two cellular delivery mechanisms for small interfering RNA, Physiol Rep. 3 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Alvarez-Erviti L, Seow Y, Yin H, Betts C, Lakhal S, Wood MJ, Delivery of siRNA to the mouse brain by systemic injection of targeted exosomes, Nat Biotechnol. 29 (2011) 341–345. [DOI] [PubMed] [Google Scholar]
  • [13].Chairoungdua A, Smith DL, Pochard P, Hull M, Caplan MJ, Exosome release of beta-catenin: a novel mechanism that antagonizes Wnt signaling, J Cell Biol. 190 (2010) 1079–1091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Lee JY, Crake C, Teo B, Carugo D, de Saint Victor M, Seth A, Stride E, Ultrasound-Enhanced siRNA Delivery Using Magnetic Nanoparticle-Loaded Chitosan-Deoxycholic Acid Nanodroplets, Adv Healthc Mater. 6 (2017). [DOI] [PubMed] [Google Scholar]
  • [15].Imani R, Shao W, Taherkhani S, Emami SH, Prakash S, Faghihi S, Dual-functionalized graphene oxide for enhanced siRNA delivery to breast cancer cells, Colloids Surf B Biointerfaces. 147 (2016) 315–325. [DOI] [PubMed] [Google Scholar]
  • [16].Kitadate A, Ikeda S, Teshima K, Ito M, Toyota I, Hasunuma N, Takahashi N, Miyagaki T, Sugaya M, Tagawa H, MicroRNA-16 mediates the regulation of a senescence-apoptosis switch in cutaneous T-cell and other non-Hodgkin lymphomas, Oncogene. 35 (2016) 3692–3704. [DOI] [PubMed] [Google Scholar]
  • [17].Valiunas V, Doronin S, Valiuniene L, Potapova I, Zuckerman J, Walcott B, Robinson RB, Rosen MR, Brink PR, Cohen IS, Human mesenchymal stem cells make cardiac connexins and form functional gap junctions, J Physiol. 555 (2004) 617–626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].Valiunas V, Weingart R, Brink PR, Formation of heterotypic gap junction channels by connexins 40 and 43, Circ Res. 86 (2000) E42–49. [DOI] [PubMed] [Google Scholar]
  • [19].Valiunas V, Gemel J, Brink PR, Beyer EC, Gap junction channels formed by coexpressed connexin40 and connexin43, Am J Physiol Heart Circ Physiol. 281 (2001) H1675–1689. [DOI] [PubMed] [Google Scholar]
  • [20].Valiunas V, Kanaporis G, Valiuniene L, Gordon C, Wang HZ, Li L, Robinson RB, Rosen MR, Cohen IS, Brink PR, Coupling an HCN2-expressing cell to a myocyte creates a two-cell pacing unit, J Physiol. 587 (2009) 5211–5226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Gemel J, Valiunas V, Brink PR, Beyer EC, Connexin43 and connexin26 form gap junctions, but not heteromeric channels in co-expressing cells, J.Cell Sci. 117 (2004) 2469–2480. [DOI] [PubMed] [Google Scholar]
  • [22].Potapova I, Plotnikov A, Lu Z, Danilo P Jr., Valiunas V, Qu J, Doronin S, Zuckerman J, Shlapakova IN, Gao J, Pan Z, Herron AJ, Robinson RB, Brink PR, Rosen MR, Cohen IS, Human mesenchymal stem cells as a gene delivery system to create cardiac pacemakers, Circ.Res. 94 (2004) 952–959. [DOI] [PubMed] [Google Scholar]
  • [23].Valiunas V, Beyer EC, Brink PR, Cardiac gap junction channels show quantitative differences in selectivity, Circ.Res. 91 (2002) 104–111. [DOI] [PubMed] [Google Scholar]
  • [24].Mehta PP, Perez-Stable C, Nadji M, Mian M, Asotra K, Roos BA, Suppression of human prostate cancer cell growth by forced expression of connexin genes, Dev.Genet. 24 (1999) 91–110. [DOI] [PubMed] [Google Scholar]
  • [25].Lamiche C, Clarhaut J, Strale PO, Crespin S, Pedretti N, Bernard FX, Naus CC, Chen VC, Foster LJ, Defamie N, Mesnil M, Debiais F, Cronier L, The gap junction protein Cx43 is involved in the bone-targeted metastatic behaviour of human prostate cancer cells, Clin Exp Metastasis. 29 (2012) 111–122. [DOI] [PubMed] [Google Scholar]
  • [26].Hou X, Khan MRA, Turmaine M, Thrasivoulou C, Becker DL, Ahmed A, Wnt signaling regulates cytosolic translocation of connexin 43, Am J Physiol Regul Integr Comp Physiol. 317 (2019) R248–R261. [DOI] [PubMed] [Google Scholar]
  • [27].Dovmark TH, Saccomano M, Hulikova A, Alves F, Swietach P, Connexin-43 channels are a pathway for discharging lactate from glycolytic pancreatic ductal adenocarcinoma cells, Oncogene. 36 (2017) 4538–4550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Kawasaki Y, Tsuchida A, Sasaki T, Yamasaki S, Kuwada Y, Murakami M, Chayama K, Irsogladine malate up-regulates gap junctional intercellular communication between pancreatic cancer cells via PKA pathway, Pancreas. 25 (2002) 373–377. [DOI] [PubMed] [Google Scholar]
  • [29].Mesnil M, Krutovskikh V, Piccoli C, Elfgang C, Traub O, Willecke K, Yamasaki H, Negative growth control of HeLa cells by connexin genes: connexin species specificity, Cancer Res. 55 (1995) 629–639. [PubMed] [Google Scholar]
  • [30].Valiunas V, White TW, Connexin43 and connexin50 channels exhibit different permeability to the second messenger inositol triphosphate, Sci Rep. 10 (2020) 8744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Huang H, Zhu Y, Li S, MicroRNA-122 mimic transfection contributes to apoptosis in HepG2 cells, Mol Med Rep. 12 (2015) 6918–6924. [DOI] [PubMed] [Google Scholar]
  • [32].Lee HY, Hong IS, Double-edged sword of mesenchymal stem cells: Cancer-promoting versus therapeutic potential, Cancer Sci. 108 (2017) 1939–1946. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Khalil C, Moussa M, Azar A, Tawk J, Habbouche J, Salameh R, Ibrahim A, Alaaeddine N, Anti-proliferative effects of mesenchymal stem cells (MSCs) derived from multiple sources on ovarian cancer cell lines: an in-vitro experimental study, J Ovarian Res. 12 (2019) 70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Li X, Li Z, Effects of human umbilical cord mesenchymal stem cells on co-cultured ovarian carcinoma cells, Microsc Res Tech. 82 (2019) 898–902. [DOI] [PubMed] [Google Scholar]
  • [35].Rhodes LV, Antoon JW, Muir SE, Elliott S, Beckman BS, Burow ME, Effects of human mesenchymal stem cells on ER-positive human breast carcinoma cells mediated through ER-SDF-1/CXCR4 crosstalk, Mol Cancer. 9 (2010) 295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Lorenzer C, Dirin M, Winkler AM, Baumann V, Winkler J, Going beyond the liver: progress and challenges of targeted delivery of siRNA therapeutics, J Control Release. 203 (2015) 1–15. [DOI] [PubMed] [Google Scholar]
  • [37].Zhang MM, Bahal R, Rasmussen TP, Manautou JE, Zhong XB, The growth of siRNA-based therapeutics: Updated clinical studies, Biochem Pharmacol. 189 (2021) 114432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Juliano RL, The delivery of therapeutic oligonucleotides, Nucleic Acids Res. 44 (2016) 6518–6548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Damase TR, Sukhovershin R, Boada C, Taraballi F, Pettigrew RI, Cooke JP, The Limitless Future of RNA Therapeutics, Front Bioeng Biotechnol. 9 (2021) 628137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Lim PK, Bliss SA, Patel SA, Taborga M, Dave MA, Gregory LA, Greco SJ, Bryan M, Patel PS, Rameshwar P, Gap junction-mediated import of microRNA from bone marrow stromal cells can elicit cell cycle quiescence in breast cancer cells, Cancer Res. 71 (2011) 1550–1560. [DOI] [PubMed] [Google Scholar]
  • [41].Thuringer D, Jego G, Berthenet K, Hammann A, Solary E, Garrido C, Gap junction-mediated transfer of miR-145–5p from microvascular endothelial cells to colon cancer cells inhibits angiogenesis, Oncotarget. 7 (2016) 28160–28168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Mittelbrunn M, Sanchez-Madrid F, Intercellular communication: diverse structures for exchange of genetic information, Nat Rev Mol Cell Biol. 13 (2012) 328–335. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Sudhof TC, Rizo J, Synaptic vesicle exocytosis, Cold Spring Harb Perspect Biol. 3 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [44].Raposo G, Stoorvogel W, Extracellular vesicles: exosomes, microvesicles, and friends, J Cell Biol. 200 (2013) 373–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Bonci D, Coppola V, Musumeci M, Addario A, Giuffrida R, Memeo L, D’Urso L, Pagliuca A, Biffoni M, Labbaye C, Bartucci M, Muto G, Peschle C, De Maria R, The miR-15a-miR-16–1 cluster controls prostate cancer by targeting multiple oncogenic activities, Nat Med. 14 (2008) 1271–1277. [DOI] [PubMed] [Google Scholar]
  • [46].Wang F, Fu XD, Zhou Y, Zhang Y, Down-regulation of the cyclin E1 oncogene expression by microRNA-16–1 induces cell cycle arrest in human cancer cells, BMB Rep. 42 (2009) 725–730. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Minella AC, Loeb KR, Knecht A, Welcker M, Varnum-Finney BJ, Bernstein ID, Roberts JM, Clurman BE, Cyclin E phosphorylation regulates cell proliferation in hematopoietic and epithelial lineages in vivo, Genes Dev. 22 (2008) 1677–1689. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].Rivas MA, Venturutti L, Huang Y-W, Schillaci R, Huang TH-M, Elizalde PV, Downregulation of the tumor-suppressor miR-16 via progestin-mediated oncogenic signaling contributes to breast cancer development, Breast Cancer Research. 14 (2012) R77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Sanchez-Beato M, Sanchez-Aguilera A, Piris MA, Cell cycle deregulation in B-cell lymphomas, Blood. 101 (2003) 1220–1235. [DOI] [PubMed] [Google Scholar]
  • [50].Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, Shimizu M, Wojcik SE, Aqeilan RI, Zupo S, Dono M, Rassenti L, Alder H, Volinia S, Liu CG, Kipps TJ, Negrini M, Croce CM, miR-15 and miR-16 induce apoptosis by targeting BCL2, Proc Natl Acad Sci U S A. 102 (2005) 13944–13949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [51].Howell PM Jr., Li X, Riker AI, Xi Y, MicroRNA in Melanoma, Ochsner J. 10 (2010) 83–92. [PMC free article] [PubMed] [Google Scholar]
  • [52].Kim IS, Ganesan P, Choi DK, Cx43 Mediates Resistance against MPP(+)-Induced Apoptosis in SH-SY5Y Neuroblastoma Cells via Modulating the Mitochondrial Apoptosis Pathway, Int J Mol Sci. 17 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [53].Denuc A, Nunez E, Calvo E, Loureiro M, Miro-Casas E, Guaras A, Vazquez J, Garcia-Dorado D, New protein-protein interactions of mitochondrial connexin 43 in mouse heart, J Cell Mol Med. 20 (2016) 794–803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [54].Vinken M, Decrock E, Vanhaecke T, Leybaert L, Rogiers V, Connexin43 signaling contributes to spontaneous apoptosis in cultures of primary hepatocytes, Toxicol Sci. 125 (2012) 175–186. [DOI] [PubMed] [Google Scholar]
  • [55].Kameritsch P, Khandoga N, Pohl U, Pogoda K, Gap junctional communication promotes apoptosis in a connexin-type-dependent manner, Cell Death Dis. 4 (2013) e584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [56].Pan H, Li W, Chen Z, Luo Y, He W, Wang M, Tang X, He H, Liu L, Zheng M, Jiang X, Yin T, Liang R, Ma Y, Cai L, Click CAR-T cell engineering for robustly boosting cell immunotherapy in blood and subcutaneous xenograft tumor, Bioact Mater. 6 (2021) 951–962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [57].Zhao W, Jia L, Zhang M, Huang X, Qian P, Tang Q, Zhu J, Feng Z, The killing effect of novel bi-specific Trop2/PD-L1 CAR-T cell targeted gastric cancer, Am J Cancer Res. 9 (2019) 1846–1856. [PMC free article] [PubMed] [Google Scholar]

RESOURCES