A Rapid, Scalable Method for the Isolation, Functional Study, and Analysis of Cell-derived Extracellular Matrix

The extracellular matrix (ECM) is recognized as a diverse, dynamic, and complex environment that is involved in multiple cell-physiological and pathological processes. However, the isolation of ECM, from tissues or cell culture, is complicated by the insoluble and cross-linked nature of the assembled ECM and by the potential contamination of ECM extracts with cell surface and intracellular proteins. Here, we describe a method for use with cultured cells that is rapid and reliably removes cells to isolate a cell-derived ECM for downstream experimentation. Through use of this method, the isolated ECM and its components can be visualized by in situ immunofluorescence microscopy. The dynamics of specific ECM proteins can be tracked by tracing the deposition of a tagged protein using fluorescence microscopy, both before and after the removal of cells. Alternatively, the isolated ECM can be extracted for biochemical analysis, such as sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting. At larger scales, a full proteomics analysis of the isolated ECM by mass spectrometry can be conducted. By conducting ECM isolation under sterile conditions, sterile ECM layers can be obtained for functional or phenotypic studies with any cell of interest. The method can be applied to any adherent cell type, is relatively easy to perform, and can be linked to a wide repertoire of experimental designs.


Introduction
Over the last few decades, the extracellular matrix (ECM) has become recognized as a diverse, dynamic, and complex environment that is involved in multiple cell-physiological and pathological processes. At the tissue level, the ECM influences cell signaling, motility, differentiation, angiogenesis, stem cell biology, tumorigenesis, fibrosis, etc. 1,2 . The study of ECM organization and ECM-dependent processes thus has wide implications for cell biology and tissue physiology. To reach a mechanistic understanding of ECM composition, organization, and functional properties, methods for the accurate isolation of ECM are required. Whereas the earliest identification of ECM proteins relied upon isolation from tissues 3 , the preparation of ECM from cultured cells has now become more prevalent.
The isolation and analysis of cell-derived ECM is complicated for two main reasons. Firstly, the presence of cells and their abundant intracellular proteins can make it difficult to isolate the ECM as a discrete extracellular structure. Indeed, some ECM proteins have roles inside the cell as well as in the ECM 4 ; therefore, the efficient removal of intracellular proteins from cell-derived ECM is vital if the study of proteins within the ECM is not to be confused with their roles inside the cell. Secondly, cell-derived ECM is composed of many large, oligomeric proteins, which are often covalently cross-linked upon ECM assembly and are therefore insoluble in standard detergents. These properties can complicate extraction and the further analysis of ECM. To address these issues, a method is required that enables the efficient separation of ECM proteins from cellular components.
Several methods have been described in the literature for the isolation of ECM from either cell culture or tissue extracts. Many of these methods are aimed at the extraction of the abundant ECM protein, collagen, from tissues and include the use of neutral salts 3 , acidic conditions 5,6 , or pepsin 7 . Isolation of total ECM from tissue extracts often involves decellularization of the tissue prior to the isolation of the ECM. For example, a sequential extraction method has been described that isolates ECM from human cardiac tissue 8 . First, loosely-bound ECM proteins were extracted with 0.5 M NaCl before sodium dodecyl sulphate (SDS) was used to remove the cells. Finally, the remaining ECM proteins were extracted using 4 M guanidine 8 . ECM can also be solubilized from decellularized samples using 8 M urea 9 . Other techniques use detergents, such as deoxycholic acid 10 , to extract both cells and ECM before separating insoluble ECM proteins from the soluble cell lysate by centrifugation.
The method for isolation and analysis of cell-derived ECM described in this report provides a reproducible method for removing cell material, leaving cell-derived ECM that can be analyzed by in situ immunofluorescence or extracted for further biochemical analysis. This method can be adapted for any adherent cell type and can be scaled up for downstream procedures, such as immunoblotting or mass spectrometry, or for utilization of the isolated ECM in functional studies. The method can also be used in conjunction with confocal microscopy of live cells to track ECM deposition of a tagged protein of interest in real time. This is achieved through the use of a gridded, glass-bottomed dish. Overall, the

Tracking Deposition of an ECM Protein by Confocal Microscopy Imaging of Live Cells
1. Count the cells under a hemocytometer. For COS-7 cells, plate 250,000 cells onto a 60 mm cell culture dish for transfection. NOTE: For live cell imaging, the cells must be transfected with a vector encoding the ECM protein of interest, fused in frame with a genetically-encoded fluorescent tag. This is to enable the identification of the ECM protein of interest during live-cell imaging. 2. After a suitable time for protein expression (e.g., 16 h) remove the cells from the dish with trypsin-EDTA solution, count them in a hemocytometer, and plate ~150,000 cells onto a 35-mm gridded, glass-bottomed dish for imaging. Allow 2 h for the cells to reattach and to begin ECM deposition (the time period will depend on the cell type and must be established empirically). NOTE: Use a cell medium that does not contain phenol red, as this can give a red tinge that affects image quality. 3. During the above 2 h period, set up a confocal microscope with a 37 °C incubator chamber and a CO 2 supply. Allow the temperature in the chamber and in the dish to stabilize to 37 °C before imaging; this can take up to 1 h. 4. Incubate the cells plated onto the 35 mm gridded dish with a fluorescent cell marker for 30 min. Then, wash the cells twice in phenol red-free culture medium prior to imaging. NOTE: A cytoplasmic fluorescent marker can be used to visualize cell volumes, or a membrane-incorporating marker can be used to visualize cell edges. 5. Using the 20X objective and phase contrast on a confocal microscope, identify an appropriate grid square that contains a number (>3) of adherent, healthy cells. Confirm the expression of the target protein of choice in the cells within the selected grid square using the 63X or 100X objectives and the appropriate wavelengths under fluorescence microscopy. 6. Set up fluorescence and phase-contrast time-lapse imaging using a z-stack software. Set the "Begin" stack to be just below the ECM layer and the "End" stack to be just above the cell body. Set the z-step size to 0.3 µm and the z-volume to between a 5 and 10 µm depth in total, depending on the cell type. This will give between 15-30 z-sections per time point. 3. Maintaining sterile technique, gently wash the "matrix producer" cells on coverslips three times with sterile PBS (see step 1.2). 4. Remove the PBS by aspiration with a sterile pipette and dispose it in a suitable liquid waste container. Add 20 mM sterile ammonium hydroxide (3 mL/100 mm dish) and incubate them for 5 min with rocking at intervals. 5. Add copious amounts of sterile de-ionized H 2 O to the "matrix producer" dish, at least 20 mL per 100 mm dish, with rocking, and pour away the cell lysate into a suitable waste container. 6. Repeat step 3.5 four more times to ensure the complete removal of the ammonium hydroxide-solubilized material.
NOTE: The ECM deposited by the "matrix producer" cells on the coverslips then provides a sterile ECM for functional assays with any test cells of interest.  1. To analyze ECM by mass spectrometry, achieve a more concentrated sample by collecting the SDS-PAGE sample buffer from the dish (as described in steps 4.5 and 4.6, below), re-heating it to 95 °C, and using the same aliquot across 2-3 additional plates.
5. Use a cell scraper to thoroughly scrape the ECM off the dish, ensuring that all areas of the dish have been scraped. 6. With the cell scraper, gather the sample to one side of the dish and remove it with a 200 µL pipette tip into a tube. 7. Transfer any residual SDS-PAGE sample buffer extract from the cell scraper tip into the tube to minimize the loss of ECM material. For a concentrated sample, re-heat the same SDS-PAGE sample buffer aliquot to 95 °C and re-use it across 2-3 additional plates. 8. Resolve the ECM proteins by SDS-PAGE 12 , using either 7% or 4-7% gradient polyacrylamide gels. NOTE: It is convenient to use pre-stained protein markers. 9. To increase the resolution of the high-molecular-weight ECM proteins, extend the gel running time such that the 37 kDa molecular weight marker runs close to the bottom of the gel and molecular weight markers <37 kDa have run off the gel. 10. For proteomic analysis of ECM proteins, stain the gel with a protein stain and capture an image. Cut bands of interest from the gel, digest them with trypsin, and analyze them by matrix-assisted laser desorption/ionization (MALDI)-mass spectrometry 13 . 11. Alternatively, for a more comprehensive analysis of the ECM extract and to analyze the entire sample, slice the gel lane into sections, digest them with trypsin, and perform liquid chromatography-mass spectrometry (LC-MS) 14 . 12. Alternatively, for immunoblotting 15 , transfer proteins from the SDS-PAGE gel onto polyvinylidene fluoride (PVDF) membrane at 15 V for at least 1 h 10 min (semi-dry method) to ensure the transfer of the high-molecular-weight ECM proteins. Visualize the total protein transferred to the membrane with a Ponceau S stain. 13. Following step 4.11, use antibodies to establish the presence and/or make a semi-quantitative analysis of individual ECM proteins 15 .
1. Block the membrane with 2% (w/v) milk in TBS-Tween 20 (blocking buffer) overnight at 4 °C. Dilute specific antibodies to ECM proteins in blocking buffer and incubate with the membrane for 1.5 h at room temperature (antibody to fibronectin diluted 1/600 and antibody to thrombospondin-1 diluted 1/150; Supplementary Table 1).
Two hours post-plating, a suitable grid square that contained multiple healthy cells expressing mRFPovTSP1C was visualized under a confocal microscope. A reference phase contrast image was taken of this area, and then fluorescence time-lapse imaging was carried out every 1 min for 2 h (Figure 2A). Cells were then removed using ammonium hydroxide, and the same grid square on the dish was relocated under phase contrast and then re-analyzed by fluorescence microscopy. The cells were no longer present in the grid square, but mRFPovTSP1C remained within the ECM, detected by fluorescence microscopy as characteristic puncta (Figure 2A). Any mRFPovTSP1C deposited in the ECM prior to cell removal was identified by comparing the fluorescence pattern pre-and post-removal of cells. The fluorescent puncta identified in both images (Figure 2A, example arrowed) are inferred to be associated with the ECM.
Ammonium hydroxide treatment was then used to isolate native ECM from cultured, adherent cells. Human dermal fibroblasts (HDF) were grown on glass coverslips for 96 h. Cells were removed using ammonium hydroxide, and the ECM was probed with an anti-collagen I antibody, which demonstrated a characteristic fibrillar and meshwork staining pattern 16 ( Figure 2B). Fibronectin patterning was examined around fixed, nonpermeabilized rat chondrosarcoma cells (RCS) and within the ECM after the removal of the RCS cells ( Figure 2C). The ammonium hydroxide isolation of ECM was also carried out under sterile conditions so that "test" cells could be re-plated onto the ECM for phenotypic or functional studies. For example, "test" COS-7 cells were plated for 2 h onto sterile ECM produced by RCS cells, and then they were fixed, permeabilized, and analyzed for F-actin organization by FITC-phalloidin staining, in comparison to COS-7 cells producing their own ECM (steps 3.1-3.9) (Figure  3).
At a larger scale, the method was used to isolate ECM for biochemical procedures. For example, RCS cells were grown on two 100 mm cell culture dishes for 7 days, and the procedures in steps 4.1-4.7 were followed. Proteins in the cell-derived ECM were collected by scraping them into hot SDS-PAGE sample buffer and were separated by SDS-PAGE on a 7% polyacrylamide gel under reducing conditions. The gel was stained for protein, and the four major bands were each isolated and analyzed by mass spectrometry (Figure 4A). A number of ECM proteins, including fibronectin, thrombospondin 1 (TSP1), thrombospondin 5/cartilage oligomeric matrix protein (TSP5 / COMP), and matrilin-1 were identified ( Figure 4B).
Alternatively, smaller-scale preparations of ECM can be evaluated by immunoblots. For example, cell-derived ECM from COS-7 cells ectopically expressing mRFPovTSP1C was separated by SDS-PAGE, transferred to a PVDF membrane, and probed for RFP with an anti-RFP antibody ( Figure 4C). This technique is sensitive enough to detect differences in deposition between a native trimer of TSP1 (mRFPovTSP1C) and an engineered pentamer of the TSP1 C-terminal region (mRFP-TSP-5-1C) 17 ( Figure 4C). To investigate the endogenous ECM of HDF, the presence of specific proteins in cell lysate or the isolated ECM was examined by immunoblotting. The characteristic ECM proteins fibronectin and thrombospondin-1 were detected in the ECM, whereas the abundant intracellular proteins α-tubulin and β-actin were absent from the isolated ECM ( Figure 4D).

Critical Steps within the Protocol
The method has some critical steps that must be followed to ensure the successful isolation and analysis of ECM. For example, ammonium hydroxide is used to remove the cells and to isolate ECM for analysis. Although ammonium hydroxide is stable in aqueous solutions in the dark and can be stored at room temperature, bottles must be tightly re-sealed between each use. Because the gas may evaporate from the solution, thus reducing the concentration, we strongly recommend starting a fresh bottle every 6-8 weeks. In addition, the working solution should be made freshly for each experiment. It is also essential that the cells are fully removed with copious washes in de-ionized water. If these steps are not performed adequately, cellular material may remain on the dish and contaminate downstream analyses. If cells have been grown for 10 days Copyright © 2017 Creative Commons Attribution 3.0 License January 2017 | 119 | e55051 | Page 8 of 9 or longer, additional water washes may be needed. It is important to note that the isolated ECM layer is typically very thin in comparison to the cells 11 , so care is needed when identifying the ECM during imaging. For effective extraction of the isolated ECM into SDS-PAGE sample buffer, it is essential that the sample buffer contains a reducing agent. For the most effective removal of the ECM, boiling-hot sample buffer must be used. For experiments in which ECM samples will be resolved by SDS-PAGE electrophoresis for protein staining or proteomics, it is critical to achieve a concentrated sample by scraping several plates-worth of ECM into the same aliquot of sample buffer.

Modifications and Troubleshooting
The production and secretion of ECM proteins can vary significantly between cell types, so time periods for secretion and deposition of ECM should be determined de novo for each cell type. It is also important to be aware that ECM composition will change dynamically in relation to cell density and time in culture 18 . This factor should be taken into account when designing experiments. Clearly, the selection of a cell type for this method is important, particularly when extracting ECM for analysis by mass spectrometry, which may require a substantial amount of total ECM protein.

Limitations of the Technique
Cells that secrete very low levels of ECM proteins will be more challenging for use with this method. Non-adherent cells, 3D cell cultures, or cell invasion assays are unsuitable at present for ECM isolation by this method. The method is most suitable for use with standard "2-dimensional" cell cultures. Because the ammonium hydroxide extraction removes the cells completely, ECM associated with the upper sides of cells and secreted, but non-cross-linked, ECM proteins are also removed, leaving on the dish a thin layer of assembled ECM from below and around the bases of the cells 11,17 . It is complex to quantify how much ECM is released by the ammonium hydroxide extraction, because the released material also includes ECM proteins that are intracellular either prior to secretion or after uptake from the cell surface.

Significance of the Technique with Respect to Existing/Alternative Methods
The ability to conduct a wide range of experiments with isolated ECM is important in the context of cell biology and tissue physiology, and it also has implications for the fields of tissue regeneration and tissue engineering. The method has advantages over gels formed from purified collagens or other purified ECM proteins in that the cells assemble complex ECM structures at their native size, with native cross-linking mechanisms and with individual ECM proteins present in ratios appropriate to the cell type of origin. For example, the proteomic study of RCS ECM demonstrated abundant matrilins and COMP/TSP5, as expected for a cartilaginous ECM 19,20 (Figure 4). In general, the analysis of cellderived ECM is hampered by its nature as an extensive, multi-protein network that results in covalent crosslinking and the insolubility of many ECM proteins, and also by the potential contamination of ECM extracts with intracellular proteins. A multi-step procedure designed to isolate the ECM fraction from tissues for proteomic analysis yielded 8% of ECM proteins in the total set of proteins identified 21 . However, peptides from ECM proteins were 73% of the total peptides identified. This method for isolation and analysis of cell-derived ECM rapidly and reliably removes cellular material whilst retaining ECM proteins. This method can be used for a range of cell types and downstream applications and facilitates the analysis of ECM biology and cell-ECM interactions in cell culture. Our protocols detail how the method can be applied at different scales to address a wide range of experimental questions with regard to ECM organization, dynamics, or composition, and to the functional effects of isolated ECM on cells.

Future Applications or Directions after Mastering This Technique
Looking to the future, the method could be adapted for use with 3D cell cultures; this would expand its physiological relevance. Decellularized native tissue has great potential as a scaffold for the regeneration of lost or damaged tissue and as an alternative to transplantation, such as after heart failure 22 . It has the potential to overcome the problems of donor availability and immunorejection. Future applications of the method described in this report could investigate its use with 3D cell cultures or tissue decellularization, which would further increase the scope of this approach.

Disclosures
The authors declare that they have no competing financial interests.