Here, we present a protocol for live cell imaging of TGF-β/Smad3 signaling activity using an adenovirus reporter system. This system tracks transcriptional activity in real-time and can be applied to both single cells in vitro and in live animalmodels.
Transforming Growth Factor β (TGF-β) signaling regulates many important functions required for cellular homeostasis and is commonly found overexpressed in many diseases, including cancer. TGF-β is strongly implicated in metastasis during late stage cancer progression, activating a subset of migratory and invasive tumor cells. Current methods for signaling pathway analysis focus on endpoint models, which often attempt to measure signaling post-hoc of the biological event and do not reflect the progressive nature of the disease. Here, we demonstrate a novel adenovirus reporter system specific for the TGF-β/Smad3 signaling pathway that can detect transcriptional activation in live cells. Utilizing an Ad-CAGA12-Td-Tom reporter, we can achieve a 100% infection rate of MDA-MB-231 cells within 24 h in vitro. The use of a fluorescent reporter allows for imaging of live single cells in real-time with direct identification of transcriptionally active cells. Stimulation of infected cells with TGF-β displays only a subset of cells that are transcriptionally active and involved in specific biological functions. This approach allows for high specificity and sensitivity at a single cell level to enhance understanding of biological functions related to TGF-β signaling in vitro. Smad3 transcriptional activity can also be reported in vivo in real-time through the application of an Ad-CAGA12-Luc reporter. Ad-CAGA12-Luc can be measured in the same manner as traditional stably transfected luciferase cell lines. Smad3 transcriptional activity of cells implanted in vivo can be analyzed through conventional IVIS imaging and monitored live during tumor progression, providing unique insight into the dynamics of the TGF-β signaling pathway. Our protocol describes an advantageous reporter delivery system allowing for quick high-throughput imaging of live cell signaling pathways both in vitro and in vivo. This method can be expanded to a range of image based assays and presents as a sensitive and reproducible approach for both basic biology and therapeutic development.
Transforming Growth Factor β (TGF-β) is an essential cytokine implicated in human development that signals through a heterodimeric complex consisting of type II and type I receptors1. Binding to type II receptor results in the recruitment and phosphorylation of type I receptor, which in turn phosphorylates downstream Smad2/3 proteins2,3. These activated Smad2/3 proteins bind to Smad4, forming a complex that translocates into the nucleus and regulates gene transcription4. Under homeostatic conditions TGF-β/Smad signaling is tightly regulated; however, in many diseases, the signaling pathway is deregulated and often overexpressed leading to progression of disease5,6,7. Recent studies have demonstrated that cellular response to TGF-β is heterogenous and subpopulations of TGF-β/Smad active cells are responsible for biological function in a time-dependent manner8,9. Common cellular analysis of TGF-β/Smad signaling involves the use of fixed endpoint assays that provide only a snapshot of cellular activity and often quantitate the average TGF-β/Smad effect10. These methods, however, may not accurately represent the molecular behaviors of TGF-β/Smad signaling in the physiological state during disease progression. Image-based analysis of live cells capture the dynamics of cellular and biological processes with both a spatial and temporal understanding.
Our goal was to develop a sensitive high-throughput method for live cell imaging of TGF-β/Smad signaling using adenovirus-based reagents. Here, we infected the human breast cancer cell line MDA-MB-231 with an adenovirus expressing the Smad3 CAGA motif binding sequence and a luciferase (Luc) or Td-Tomato (Td-Tom) reporter gene. Adenoviral reporter systems provide a quick and cheap method for plasmid introduction that can result in a 100% infection rate in cancer cell lines. Adenoviral reporter systems have also been successfully applied to cell lines that are difficult to transfect with conventional plasmid11. In this protocol we will describe a reproducible and noninvasive process to achieve live cell imaging of the TGFβ/Smad signaling pathway both in vivo and in vitro.
All animal experiments were approved by the University of Melbourne Animal Ethics Committee.
Note: The sequence, construction and generation protocol for the adenoviral vectors pAd-CMV-Td-Tom, pAd-CMV-GFP and pAd-CAGA12-Luc/Td-Tom have been previously described11,12,13. All vectors are commercially available.
1. Virus Titer Determination Using 50% Tissue-culture Infectious Dose (TCID50)
2. Adenovirus MOI Determination
3. Dual-adenovirus Transduction in Live Single Cells
4. Live Single Cell Signaling in Wound Healing Assay
5. In Vitro Luciferase Assay
6. Cell Preparation for Live Signaling Imaging In Vivo
7. Orthotopic Implantation
8. IVIS Imaging
9. Data Analysis
Live single cell imaging in vitro
To accurately assess the activation of TGF-β/Smad signaling in single cells using adenovirus based reagents, it is important to first determine the optimal Multiplicity of Infection (MOI) for each cell line. The optimal MOI is determined when 100% of cells are positive for adenoviral infection with no present cytopathic or cytotoxic effects. To determine this, we used a constitutively active Ad-CMV-Td-Tom reporter that acted as a marker of positive infection. The percentage of adenovirus infected cells was increased in a MOI dependent manner from 0 to 5,000 MOI. At 2,500 MOI, we observed 100% Td-Tom expression in MDA-MB-231 cells (Figure 1A, 1B). Pixel intensity/cell also increased with MOI, providing enhanced sensitivity and detection of transcriptional output (Figure 1C). Therefore, a working MOI of 2,500 was used to perform a dual infection where cells were infected with Ad-CMV-GFP (served as positive control for cell infection) and Ad-CAGA12-Td-Tom simultaneously. Single cells presented varying levels of Td-Tom expression indicating heterogenous activation of TGF-β/Smad3 transcription across cells in both live and fixed cells. (Figure 2A, 2B). 100% of MDA-MB-231 cells presented with GFP expression while only around 26% of cells displayed detectable TGF-β/Smad3 transcriptional activity 24 h after TGF-β stimulation, compared to 0% positivity without TGF-β treatment (Figure 2C). To use the adenovirus based reagent to investigate the TGF-β/Smad3 transcriptional activity during biological processes, MDA-MB-231 cells were infected with Ad-CAGA12-Td-Tom at 2500 MOI and subjected to a wound healing assay. Infected MDA-MB-231 cells exhibited normal migration behavior and cells treated with 5 ng/mL TGF-β showed enhanced wound closure compared to non-TGF-β treated cells (Figure 3A). Notably, about 62% cells in the wound area presented positive TGF-β/Smad3 transcriptional activity, which is significantly higher than observed in non-wound area (32%) after TGF-β treatment (Figure3B). As such, the adenovirus-based reagents validated their application for imaging of the TGF-β signaling pathway in live single cell in vitro.
Live TGF-β signaling imaging in vivo
The sensitivity of the Ad-CAGA12– Luc reporter was first tested in vitro in response to TGF-β stimulation and the presence of a TGF-β inhibitor. When infected MDA-MB-231 cells were stimulated with 1 ng/mL TGF-β, luciferase reporter activity increased 1,200-fold compared to basal untreated MDA-MB-231 levels. This response was remarkably blocked by 12 µg/mL of TGF-β inhibitor (Figure 4A). The Ad-CAGA12-Luc reagent was subsequently tested in a breast tumor orthotopic implantation animal model. Both control and the TGF-β inhibitor treated groups demonstrated similar luciferase reporter activity before treatment. However, for the control group, the luciferase signal did not change after PBS treatment, whereas, mice treated with the TGF-β inhibitor showed around 3-fold signal reduction (Figure 4B, 4C). Collectively, these results demonstrate the potential for live and real-time monitoring of TGF-β signaling activity in vivo without the need for animal sacrifice.
Figure 1. Determination of the MOI of adenovirus-based reagent. MDA-MB-231 cells were infected with Ad-CMV-Td-Tom at different MOI and seeded into a 12 wells plate. 24 h after infection, cells were fixed and nuclear stained by Hoechst. (A) Images were taken to visualize Td-Tom positive cells (red) and nucleus (blue) and quantified by ImageJ. (B) Percentage of Td-Tom positive cells. (C) Td-Tom pixel intensity/cell normalized to background. These figures are representative of at least 3 independent experiments. Data shown are means of triplicates ± SD. Scale bar = 50 µm (200X). Please click here to view a larger version of this figure.
Figure 2. Single cell imaging for dual infection. MDA-MB-231 cells were dual infected with Ad-CMV-GFP and Ad-CAGA12-Td-Tom at MOI (2,500) and seeded into a 12 wells plate. 24 h after infection, cells were treated with or without 5 ng/mL TGF-β. (A) 24 h after treatment, cells were live imaged to visualize GFP positive (green) and Td-Tom positive (red) cells. Cells were then fixed and nuclear stained by Hoechst for quantification. (B) Images were taken to visualize GFP positive (green), Td-Tom positive (red) cells and nucleus (blue). (C) The GFP or Td-Tom positive cells were quantified by ImageJ and the percentage of positive cells was calculated. These figures are representative of at least 3 independent experiments. Data shown are means of triplicates ± SD. Statistical significance was determined by Student's t-test (**** p ≤0.0001). Scale bar = 50 µm (200X). Please click here to view a larger version of this figure.
Figure 3. Live cell TGF-β signaling promotes cell migration. MDA-MB-231 cells were infected with Ad-CAGA12-Td-Tom at MOI (2,500) and seeded onto a 6 wells plate. 24 h after infection, cells were scratched using a P200 pipette tip to create a wound and media replaced with or without 5 ng/mL TGF-β. After another 24 h, cells were fixed and nuclear stained by Hoechst for visual representation. (A) Phase contrast of wound closure, Scale bar = 100 µm (100X). (B) Td-Tom positive (red) cells and nucleus (blue). Scale bar = 50 µm (200X) (C) Td-Tom positive cells were quantified by ImageJ and the percentage of positive cells was calculated. These figures are representative of at least 3 independent experiments. Data shown are means of triplicates ± SD. Statistical significance was determined by Student's t-test (**** p ≤0.0001). Please click here to view a larger version of this figure.
Figure 4. In vivo live TGF-β signaling imaging. (A) MDA-MB-231 cells were infected with Ad-CAGA12-Luc and then seeded into 96 wells plate for 24 h. Thereafter, cells were treated with or without 1 ng/mL TGF-β in the presence or absence of 12 µg/mL TGF-β inhibitor overnight. The luciferase activity was measured based on the whole cell lysate. Luciferase activities were presented as the fold of induction normalized to no treatment control. Data shown are means of triplicates ± SD. (B) MDA-MB-231 cells were infected with Ad-CAGA12-Luc 24 h before implantation. Cells were implanted into mammary fat pad on both sides of SCID mice. 72 h later, IVIS imaging was performed before and after 24 h treatment (PBS for control group and 50 µg/tumor TGF-β inhibitor for treatment group). (C) Photon flux was quantified by Living Image software. Data shown are means of each treatment group (n = 12) ± SD. Statistical significance was determined by Student's t-test (** p ≤0.01, **** p ≤0.0001). p/s refers to photons/s. Please click here to view a larger version of this figure.
We have developed a technique to allow for real-time imaging of TGF-β/Smad3 signaling in single live cells. Using this novel method, we have previously identified a sub-population of cells with dynamic TGF-β/Smad3 transcriptional activity that was associated with enhanced invasion and migration8. This method improves on traditional assays for TGF-β signaling, such as western blots of Smad3 phosphorylation and TGF-β targeted gene expression, by capturing the heterogeneity of TGF-β/Smad3 signaling within the tumor population and highlighting the importance of single cell biology. Additionally, alternate methods of single cell imaging such as nuclear translocation can be time consuming and expensive if performed in live cells and may not always translate to gene transcription16,17. While these methods are important to examine the upstream pathways of signal transduction, the adenovirus reporter system provides a unique perspective on TGF-β/Smad3 signaling in individual cells in relation to biological function.
Significantly, our method provides a sensitive and reproducible method for in vivo detection of the TGF-β/Smad3 signaling pathway in real-time during tumor progression and treatment. Similar to traditional stable luciferase in vivo imaging, detection of TGF-β/Smad3 signaling within the mouse is dependent on the number of cells and level of promoter activity18. While subcutaneous tissue provides little interference for analysis, increased depth of cells will reduce the sensitivity of signal detection. It is possible to expand the method for use in other organs, such as lung and bone, and other cancers; however, signal optimization or ex vivo imaging is recommended to achieve the sensitivity required for the cancer model of interest.
An additional benefit of using the adenovirus reporter system is that its application is likely to expand to analysis of multiple signaling pathways simultaneously in live cells. In Figure 2A, we demonstrated that MDA-MB-231 cells could be transduced with 2 separate adenovirus vectors, CMV-GFP and CAGA-TOM. This dual-infection system can be further expanded to report two different signaling pathways via both fluorescence reporter proteins and luciferase reporters (using Firefly and Guassia luciferase reporter proteins). Our lab has achieved simultaneous signaling of BMP/Smad1 and TGF-β/Smad3 signaling pathways in a number of live cancer cell lines (data not shown).
Adenovirus reporter systems provide an easy and convenient method for live cell imaging of the TGF-β/Smad signaling pathway both in vitro and in vivo. This system has distinct advantages over other commonly used reporter systems. Firstly, adenovirus vectors are reported to have among the highest viral transduction efficacy, making them an optimal system for cell lines that are difficult to transduce19,20,21,22. Additionally with current advancements in viral technologies, 'gutless' adenovirus vectors can be generated by removing most of the viral genome allowing for increased exogenous DNA capacity23. In comparison to non-viral delivery systems, adenovirus vectors represent a cost-effective, quick and easy application for cell transduction that results in far greater delivery efficiency19,20,21,22.
This protocol uses a second-generation adenovirus expression system, which is replication-incompetent in mammalian cells that do not express the E1a and E1b proteins, minimizing biosafety risk24. However, despite these enhanced biosafety features, adenovirus particles can still transduce primary human cells with high efficacy25,26. US Biosafety Level 2 and containment of adenoviral vectors is recommended when handling and disposing of adenovirus contaminated equipment. Large-scale amplification of adenovirus in HEK293A cells can result in a rare generation of replication-competent "wild-type" adenovirus27. PCR screening should be performed to identify wild-type contamination27.
Replication-incompetent adenovirus vectors are transient in expression and do not integrate with host DNA20,21. It is recommended that the experimenter identify the stability in expression of their reporter protein when using the adenovirus reporter system to confirm the validity of long-term experiments. The expression time of the adenovirus vector within a cell is proportional to the division time of the cell and viral MOI. To ensure reproducible results, it is critical that the viral titer is accurately determined for each viral batch. Additionally, excessively high volumes of the adenovirus can result in cytotoxic or cytopathic effects and it is important that an optimal MOI is empirically determined for each cell lined used to ensure best results. Further, in vivo use of adenovirus vectors can induce immune responses; where immune response is not the primary observation of the experiment, immune-compromised animals are recommended20,21. Additional quality control is required when using the adenoviral vectors to target immune responses, though the use of the adenovirus has been validated as an adjuvant to vaccine therapies28,29.
The adenovirus reporter system can be used on a broad range of primary and cultured cell lines of both dividing and non-dividing nature20,22. Using the adenovirus reporter system, we have achieved 100% infection rates in a number of cancer cell lines including brain, breast, colorectal as well Primary Mouse Embryological Fibroblast (MEF) and Canine Kidney Epithelial (MDCK) cells. Additionally, adenoviral vectors can be further modified to increase the affinity for cell-type specific receptors, providing improved organ targeting and efficacy for transduction30,31.
Applications of this method can be expanded to other signaling pathways and cell hosts. The principal advantages of the adenoviral reporter system include the cheap,high transduction efficacy, quick experimental setup and lack of integration with host DNA21,32. This method can be applied to many current assays and in vivo models, including in the evaluation of new targeted therapeutics33. We hope that the use of this technique will enhance our understanding of the relationship between signaling pathways and biological processes leading to new biological discoveries and the development of new therapies.
The authors have nothing to disclose.
This work was supported by grants from the National Health and Medical Research Council (NHMRC) to H-JZ. TMBW is a recipient of an Australia Postgraduate Award from the Australian Government and the Ann Henderson Top-Up Scholarship from Australian Rotary Health in partner with Rotary of Templestowe and Dine for a Cure.
DMEM | ThermoFisher | 1881024 | Warm in 37 °C waterbath before use |
Foetal Bovine Serum | Scientifix Life | FBS500-S | Heat inactivated before use |
Recombinant Human TGF-β1 | PEPROTECH | 100-21 | Aliquot in DDW to make final concentration at 10 mg/mL |
Hoechst-33258 | Tocris Bioscience | 5117 | Dilute in to PBS to make final concentration at 1 μL/mL |
Luciferase Reporter Assay Kit | Promega | 197897 | Dilute 5x in PBS before use |
Luminometer | Promega | 9100-002 | |
Phase contrast fluorescence microscopy | OLYMPUS | IX50 | |
Centrifuge | eppendorf | 5810 R | |
VivoGl Luciferin | Promega | P1041 | |
IVIS Lumina III In Vivo Imaging System | PerkinElmer | CLS136334 | |
0.5% Trypsin-EDTA (10x) | ThermoFisher | 15400-054 | Diltue to 0.05% (1x) in PBS |
Cell Culture Lysis 5x Reagent | Promega | E153A | Dilute to 1x in DDW |
10% Formalin | Sigma-Aldrich | F5554-4L | |
HEK 293A | ThermoFisher | R70507 | |
MDA-MB-231 | ATCC | CRM-HTB-26 | |
PRKDC-SCID | Animal Resources Centre | SCIDF6 | |
Matrigel | Corning | 354234 | |
Isoflurane | Zoetis | 26675-46-7 | |
Ethanol | Chem-supply | EA043-10L-P | |
Refresh Night Time | Allergan | 1750D | Lubricating Eye Ointment |
Solution | Composition | ||
Phosphate-Buffered Saline (PBS) | NaH2PO4.2H2O (4 mM); NaHPO4 (16 mM); NaCl (0.12M) | ||
FBS-DMEM | 5% heat inactivated FBS; 10 μg/mL penicillin; 100 μg/mL streptomycin |