Preparation and Use of Reverse Protein Microarrays

UNIT 27.7

Elisa Pin,1,2 Giulia Federici,1,3 and Emanuel F. Petricoin, III1 1

Center for Applied Proteomics and Molecular Medicine, George Mason University, Manassas, Virginia 2 Division of Experimental Oncology 2, Centro di Riferimento Oncologico-IRCCS, National Cancer Institute, Aviano, Italy 3 Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanit`a (ISS), Rome, Italy

ABSTRACT Reverse-phase protein array (RPPA) is a multiplex, high-throughput proteomic technique for profiling the activation status of signal transduction pathways involved in cancer survival and progression, potentially allowing for identification of new biomarkers and drug targets. On RPPA, the entire patient proteome is immobilized on a spot and single proteins can be quantified across a set of samples, spotted on the same array, with high specificity and sensitivity. Array immunostaining and signal amplification systems are used to generate a signal proportional to the concentration of the analyte. Dedicated scanners and software are used to detect spots, measure intensity, subtract background, normalize signal, and generate a numeric value as output. The generated output file is then analyzed using several different bioinformatic and biostatistical tools. In this unit, the RPPA procedure is described in depth, from sample handling and preparation to data analysis, with particular emphasis on tissue sample analysis. Curr. Protoc. Protein Sci. C 2014 by John Wiley & Sons, Inc. 75:27.07.1-27.07.29.  Keywords: tissues r cells r protein r microarray r antibody r analysis

INTRODUCTION One of the main goals in cancer research is to optimize treatment through improvement of personalized therapy. Cancer derives from variations and dysregulations of cellular signaling pathways due to genetic and/or proteic alterations. Therefore, mapping patients’ cellular network activation is a necessary starting point to identify new tailored therapies. Genomics and transcriptomics give, respectively, a picture of genome status (e.g., presence of gene mutations or variations) and gene activation/expression, while proteomics analysis can map protein activation status. Since proteins are the functional elements of cellular signaling pathways, proteomics represents the optimal method to describe the functional status of networks controlling processes like growth, differentiation, proliferation, migration, apoptosis, and metabolic activity; in this context, reverse-phase protein array (RPPA) is a powerful tool developed to study post-translational modifications that can alter these fundamental cellular processes. RPPA technology is a robust, sensitive, and quantitative technology that allows analysis of protein modifications such as phosphorylation, glycosylation, cleavage, and ubiquitination, all of which are regulated by the activation of signaling cascades involving protein families such as kinases, phosphatases, glycosyltransferases, and ubiquitin-conjugating proteins. Clarifying the entire cascade that leads to protein activation is the first step to identify alterations in cell homeostasis that will drive cells to carcinogenesis. Moreover, RPPA is a versatile technique useful not only for the identification of cancer-targeted Protein Arrays Current Protocols in Protein Science 27.7.1-27.7.29, February 2014 Published online February 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/0471140864.ps2707s75 C 2014 John Wiley & Sons, Inc. Copyright 

27.7.1 Supplement 75

therapy but also for treatment of many other diseases, as well as for diagnostic and prognostic biomarker identification. RPPA descends directly from miniaturized immunoassays (Ekins et al., 1989) and it is called “reverse phase” to be distinguished from forward phase microarrays, since the molecule immobilized on nitrocellulose is the analyte (antigen) and not the antibody. RPPA was first described in 2001 by Paweletz et al. (2001). Each array can comprise hundreds to thousands of samples, controls, and calibrators in duplicate or triplicate, depending on the dimension of the spots. The number of samples that can be printed depends on the number of replicates, dilutions, controls, and calibrators to be included and the pin size. For example, 185-μm pins allow printing of >2000 spots, while 350μm pins allows printing of 600 spots. Consequently, the more replicates, dilutions, and controls printed, the fewer samples fit in the array. The use of three technical replicates with at least two dilutions for each sample, such as a neat and a 1:4 dilution, is recommended to provide appropriate experimental conditions (i.e., good spot averaging, analysis of dilutions in case artifacts occur on the neat spots, and good antibody reactions). Samples to be tested by RPPA can be whole tissue lysates, enriched tissue lysates (e.g., microdissected tissues) (Petricoin et al., 2007; Silvestri et al., 2010), cell cultures (Mazzone et al., 2010), body fluids such as serum/plasma (Mueller et al., 2010b), and CSF (Gyorgy et al., 2010). A few picograms are sufficient to test the expression/activation of a single analyte (Paweletz et al., 2001) and this allows printing of hundreds of arrays with just a few microliters of sample. Each array is then probed with a single antibody allowing comparison of the expression/activation pattern of the analyte across samples in the same condition. Considering the variability of proteins, especially their phosphorylated forms, reproducible and consistent sample collection and storage methods are critical and necessary components of cell signaling analysis. One of the most important pre-analytical variables to control is the experimental duration, since enzymes such as phosphatases and proteases continue to function at room temperature after the tissue/cell sample has been collected. To block protein degradation, it is advisable to inhibit proteases at all phases of the analysis, from tissue collection to lysis. New fixatives are now available and could substitute for more classical fixatives (e.g., OCT) due to their innovative formulations that improve protein preservation. On the other hand, it is always necessary to add protease and phosphatase inhibitors to the lysis buffer used to extract proteins from the sample and to all aqueous solutions used to treat samples prior to lysing. Unlike DNA arrays, which are generally printed on amine- or lysine-coated glass slides because of the molecule’s negative charge (Miller et al., 2002), the optimal substrate for RPPA protein arrays is nitrocellulose, which has low background but high affinity and binding capacity for proteins without altering their structure (Tonkinson and Stillman, 2002). RPPA samples are printed on nitrocellulose mostly using solid pin-based contact printing. Pins have a flat end that is dipped in the sample (generally loaded in a multiwell microplate), allowing the transferring of a small amount of it to the pin-head. Subsequently, contact between the pins and nitrocellulose surface transfers the sample to the substratum (printing). Several types of pins with different diameters are available and can be chosen on the basis of required sample volume and spot size (Rose, 2000).

Reverse Protein Microarrays

Each RPPA spot is representative of the entire cellular proteome; therefore, primary antibody selection and validation are significant to guarantee high specificity. High sensitivity is ensured by using commercially available signal amplification systems based on catalyzed reporter deposition (e.g., CSA kit, Dako) (Bobrow et al., 1989). The detection system can be based on colorimetric or fluorimetric methods.

27.7.2 Supplement 75

Current Protocols in Protein Science

tissue (frozen and OCT-embedded formalin-fixed and paraffin-embedded)

cell culture

laser-capture microdissection

serum, plasma, CSF, vitreous, synovial fluid

protein extraction

protein quantitation (Bradford and/or test print)

printing

immunostaining and signal amplication

signal detection

Figure 27.7.1

spot analysis

RPPA workflow. The schematic shows the main RPPA phases.

Spot intensity is revealed by slide scanning and image analysis using one of several dedicated software tools such as Microvigene (VigeneTech). The analyte concentration is proportional to signal intensity after subtraction of background and negative control signal and normalization to total protein amount. The normalized data can be used to generate “heat maps” and network profiles across patient samples (Liotta et al., 2003). This unit describes the phases of the RPPA process: (1) sample collection, preservation, storage, and protein extraction (see Basic Protocols 1 through 6), (2) array construction (see Basic Protocol 7), (3) immunostaining and total protein determination (see Basic Protocols 8 through 10), and (4) image analysis (see Basic Protocol 11). Figure 27.7.1 presents the workflow of the whole process.

PROTEIN SAMPLE PREPARATION FOR REVERSE-PHASE PROTEIN MICROARRAY Using RPPA technology, a large number of samples (hundreds to thousands) can be analyzed simultaneously. It is possible to investigate the protein content from cell cultures, tissues, and biological fluids, such as serum and urine.

Cell Cultures There are many different types of cell cultures and they can differ according to cell origin (tumor or normal cells, from blood or from tissues) and morphology (adherent or suspension cells). Protein lysates can be obtained following different protocols depending on the type of sample. It is important to note that proteins must be maintained in their posttranslationally modified forms (e.g., phosphorylated) for analysis of protein activation states; therefore, it will be necessary to add protease and phosphatase inhibitors to the lysis solution to prevent protein degradation and dephosphorylation, respectively. In the case of a cell culture system, the operator will utilize the cell culture itself, or directly the cell pellets derived from it. Protein Arrays

27.7.3 Current Protocols in Protein Science

Supplement 75

BASIC PROTOCOL 1

Lysing Cells in Suspension This protocol describes a procedure for lysing cells in suspension culture. It is important to maintain proteins in their post-translational forms, so it is necessary to include protease and phosphatase inhibitors in the lysis buffer solutions to prevent protein degradation and dephosphorylation, respectively. In this case, collection of cells is simplified since physical detachment from the plate is not needed.

Materials Fresh cell cultures Phosphate-buffered saline, calcium- and magnesium-free (CMF-PBS), ice cold Cell lysis buffer (see recipe), prepared fresh and store on ice 15-ml and 1.5-ml collection tubes Refrigerated centrifuge Vortexer 1. Collect the medium containing the cells in a collection tube and centrifuge 10 min at 600 × g, 4°C. 2. Discard supernatant, and wash two times with 2 to 4 ml of ice-cold CMF-PBS. Centrifuge 5 min at 600 × g after each wash. 3. After the last wash, carefully discard all supernatant so there is no residual PBS and place the cell pellet on ice. The cell pellet can be stored up to 3 months at −80°C, upon pre-freezing in liquid N2 , for later lysis and processing.

4. Add an appropriate amount of cell lysis buffer to the pellet (100 μl lysis buffer/1 × 106 cells). If unsure about the number of cells, estimate the volume of lysis buffer to be used as a function of the size of the pellet. For example, for invisible pellets, do not use >15 μl of lysis buffer. It is preferable to keep the volume low, rather than add too much buffer and dilute the sample. The number of cells/volume ratio may change by cell type.

5. Vortex the cell pellets for 15 sec and incubate 20 min on ice. If solution is too viscous after vortexing, more buffer may be added or sample can be sonicated on ice for three cycles of 10 sec each at 15-μm amplitude.

6. Centrifuge sample 5 min at 14,000 × g, 4°C. 7. Collect supernatants containing the proteins in a fresh labeled tube and place on ice for immediate use in RPPA technology or store up to 1 year at −80°C. BASIC PROTOCOL 2

Lysing Adherent Cells This protocol describes the procedure for lysing adherent cells. It is important to maintain proteins in their post-translational forms, so it is necessary to include protease and phosphatase inhibitors in the lysis buffer solutions to prevent protein degradation and dephosphorylation, respectively. In this case, cell scrapers are used for the collection of cells. Enzymatic detachment procedures are avoided so proteins on the cell membrane are not damaged or modified.

Materials Reverse Protein Microarrays

27.7.4 Supplement 75

Fresh cell cultures Phosphate-buffered saline, calcium- and magnesium-free (CMF-PBS), ice cold Current Protocols in Protein Science

Cell lysis buffer (see recipe), prepare fresh Flask scrapers 15-ml and 1.5-ml collection tubes Refrigerated centrifuge 1. Discard medium from cell culture flask. 2. Wash cells in flask two times with 5 ml of ice-cold CMF-PBS by gently shaking. 3. After the final wash, add 2 ml of ice-cold CMF-PBS, scrape the cells away from the bottom of the flask using a scraper and collect the solution containing the cells in a fresh labeled tube. 4. Centrifuge 5 min at 250 × g, 4°C. 5. Add 300 to 500 μl of cell lysis buffer (amount of buffer for cells 80% confluent collected from a 75-cm2 flask) and allow it to react for 15 min on ice or at 4°C. 6. Centrifuge 10 min at 14,000 × g, 4°C. 7. Collect the supernatant containing the proteins in a new tube and place on ice for immediate use for RPPA. Alternatively, the supernatant can be stored for at least 1 year at −80°C.

TISSUES It is possible to print on nitrocellulose slides material derived from patients’ biopsies or mice xenografts. Tissues can be handled differently—they can, for example, be freshly frozen (embedded or not in OCT) at −80°C or fixed in formalin and embedded in paraffin followed by cutting thin (8-μm) sections to be mounted on glass slides. Freshly frozen tissues can either be lysed as whole chunks, or as whole tissue sections, or after cell isolation from tissue sections using laser-capture microdissection (LCM). LCM permits isolation of highly pure cell populations from a heterogeneous tissue section through the reaction of the laser with a specific area of the tissue (i.e., the single cell of interest) and transfer onto a thermolabile polymer placed on a cap (Espina et al., 2009). In any of these cases, the lysis buffer is different from the cell lysis buffer mentioned before, since a Laemmli-based buffer containing a reducing agent must be used. Whole-Tissue Chunk Lysis

BASIC PROTOCOL 3

This protocol describes a procedure for lysing whole-tissue chunks. When it is not possible to laser-capture microdissect (LCM) the tissue due to technical or time limitations, patient biopsies can be directly dissolved in extraction buffer. This procedure can be applied if an LCM station is not available or when the number of samples is large. However, this approach does not guarantee tumor tissue purity, since biopsies can also contain stroma and immunological infiltrates. It is desirable to check tissue composition by H&E staining. This lysis procedure requires the use of manual or electronic pestles to break and dissolve the biopsy tissues.

Materials Frozen, whole-tissue chunks Dry ice Extraction buffer (see recipe), prepare fresh Protein Arrays

27.7.5 Current Protocols in Protein Science

Supplement 75

Pestles and appropriate 1.5-ml vials 100°C dry heat block 1. Place frozen, whole-tissue chunks in a closed box with dry ice prior to use. 2. Place the whole frozen sample into a 1.5-ml vial specific for pestle fitting. Keep the tissue frozen. 3. Add freshly prepared extraction buffer in a volume depending on the size (or weight) of the tissue sample. Start energetically disrupting the specimen at room temperature with the pestle until the chunk is dissolved. 4. Heat the lysate 10 min on a 100°C dry heat block. It is possible that the chunk will not dissolve completely; this is not an issue.

5. Allow the tubes to cool for 10 min, vortex 10 sec, and briefly spin the lysates down. 6. Recover the lysate in a new labeled 1.5-ml tube and store it at −20°C prior to printing. Store lysates for at least 6 months at −20°C. BASIC PROTOCOL 4

Whole or Laser-Capture Microdissected Tissue Section Lysis This protocol describes a procedure for lysing whole or laser-capture microdissected (LCM) tissue sections. Tissues to be used for genomic and proteomic analysis should be frozen as soon as possible after procurement. When using tissue sections, it is desirable to freeze specimens embedded in OCT, by putting the tissue-containing cryomold on dry ice and then storing it for at least 24 hr at −80°C prior to cutting sections. Cut sections should be stored in a slide storage box at −80°C for an additional 24 hr before lysing or microdissecting. For formalin-fixed, paraffin-embedded (FFPE) tissues, the paraffin must be dissolved and removed before lysis by incubating the slide-mounted sections in xylene and allowing them to rock on an agitator for 15 min at room temperature two times. The extraction buffer for tissue lysis can be used for performing both western blots and RPPAs and consists of a detergent, a denaturing agent, and a buffer. This buffer is considered a mild denaturing extraction buffer for the solubilization of cellular proteins. NOTE: If microdissected tissue specimens are desired, an LCM station is needed.

Materials Frozen, whole-tissue sections (8-μm thick) on glass slides Dry ice Hematoxylin and eosin (H&E) staining solution (see Support Protocol 1) Extraction buffer (see recipe) LCM station for tissue dissection 0.5-ml screw-cap tubes 100°C dry heat block 1. Place frozen tissue slides in a closed box with dry ice prior to use. 2. Fix and stain slides following an H&E staining procedure (see Support Protocol 1).

Reverse Protein Microarrays

Avoid using eosin that binds membrane proteins, which possibly interferes with primary antibody binding during immunostaining. Add protease inhibitor cocktail tablets to the solutions up to the second 70% ethanol. Protease inhibitors tablets are water-soluble: dissolve them in water, vortexing the tubes before adding the alcoholic component.

27.7.6 Supplement 75

Current Protocols in Protein Science

3. Proceed with LCM if desired. Microdissect tissues no longer than 40 min to remain within the active window of the protease and phosphatase inhibitors. The selected cells will bind to the cap polymer. A minimum of 15,000 cells is needed for RPPA printing. Label and store the cap in dry ice, placing it into a 0.5-ml tube. Store caps at −80°C if the lysis will not be performed immediately after the dissection.

4. Cover the section or microdissected tissue with a drop (3 to 5 μl) of freshly prepared extraction buffer. Incubate 1 min at room temperature. For microdissected tissue, adjust the extraction buffer volume based on the number of cells—generally 1 μl/1000 cells, but this may vary with tissue type. The lysates from multiple caps per single sample can be pooled together to increase the total volume of the lysate and to ensure a sufficient number of cells for adequate total protein concentration (0.25 μg/μl). Considering the high neurotoxicity of 2-mercaptoethanol, it is always better to work under a chemical hood, even if the air flux can cause partial evaporation of samples. When working with small volumes of buffer (e.g., LCM material lysis), it is possible to substitute 2-ME with a non-toxic TCEP-based solution.

5. Pipet the extraction buffer up and down five to ten times, then allow lysis to continue for an additional 1 min. 6. Collect tissue lysate in a 0.5-ml screw-cap tube using a micropipette. 7. Heat the lysates for 7 min in a 100°C dry heat block. 8. Allow the tubes to cool for 10 min, vortex 10 sec, and briefly centrifuge the lysates 10 sec at maximum speed. 9. Remeasure cell lysate volume and, if evaporation occurred, bring sample to starting total volume adding more extraction buffer. 10. Recover the lysate in a new labeled tube and store up to 1 year at –20°C prior to printing. Extraction of Proteins from Biological Fluids and Low-Molecular-Weight Serum Fraction Samples

BASIC PROTOCOL 5

It is possible to print material obtained from biological fluids, such as serum, CSF, synovial fluid, and vitreous. The extraction buffer for serum is a modified Laemmlibased buffer, which contains a higher percentage of glycerol than cell extraction buffer. The higher glycerol content provides a higher viscosity, which limits spot diffusion on the printed microarray. Analysis of low-molecular-weight (LMW) proteins that were previously recovered by continuous elution electrophoresis allows the identification of proteins or protein fragments 2000-fold in some instances) and linearity without saturation. This characteristic affords the flexibility to print a sample without a sample dilution curve, which can reduce sample loss. Chromogenic detection must utilize dilution curve printing since color detection has an extremely shallow dynamic range. The lack of a dilution curve permits use of less sample and, therefore, printing more arrays and analyzing a larger number of endpoints. This is important, in particular, when a small quantity of samples is available (e.g., microdissected tissues). Another important advantage of the fluorescence protocol is the non-toxicity of the reagents. Not only is this important for the health of the

27.7.26 Supplement 75

Current Protocols in Protein Science

A

B

C

D

E

F

Figure 27.7.5 RPPA results and spot artifacts. (A) Good spots. Good intensity and low background. (B) Two examples of “dust”. While in the left image, just one of the three spots is affected, the image on the right is more difficult to analyze because all three spots contain dust. (C) Sample concentration problem. This effect generally happens when the sample in the plate starts to concentrate, creating shadows on the nitrocellulose. (D) Low sample volume. When the volume is too low, some spots are missed during printing. (E) High background. This effect is due to a primary antibody affinity problem or insufficient blocking. (F) Shadows on nitrocellulose. This effect is generated by bad washing.

operator (DAB in carcinogenic), but it also reduces costs for waste disposal.

Image adjustments All adjustments of array images obtained by slide scanning must be performed prior to spot detection and analysis and must be consistent for all of the tested slides. Adobe Photoshop is suggested for image manipulation because it does not change the actual pixels. Avoid using other programs that change the pixel intensities in a non-linear manner.

Analysis of spot quality After spot detection by MicroVigene analysis, it is important to exclude all spots that have unconventional shapes or non-homogeneous signal intensity (Fig. 27.7.5). Some of these effects are due to pin-specific deposition issues. Other artifacts are due to regional differences in nitrocellulose or primary antibody staining that give a high background signal. The coefficient of variation (CV) is a good parameter that helps to evaluate the homogeneity of spots inside a single sample. In general, a good MicroVigene CV is considered to be 8. MicroVigene permits elimination of dust from spots but this tool sometimes does not work—in these cases, it is possible to manually exclude the altered spot. For these reasons, it is necessary to print the sample in triplicate or quadruplicate on the slide, permitting to have a signal mean and a standard deviation even when it is necessary to exclude one of the spots.

Inter-array calibrators To ensure comparisons of different printing runs, it is important to include in the arrays some shared calibrators in a dilution curve that allow the normalization of different array sets. This is important, in particular, if the arrays are used for clinical aims, like for clinical trials.

Troubleshooting Problems, possible causes, and solutions for each step of the RPPA process are discussed in Table 27.7.3.

Anticipated Results Figure 27.7.5 represents a typical result for RPPA analysis and some of the most frequent spot effects that can affect the analysis. An ideal spot appearance is represented in Figure 27.7.5A—full spots with high signal on a low background. Signals should also be very similar among triplicate, assuring low CV values during MicroVigene analysis.

Time Considerations Sample printing Sample plating and arrayer priming take 1 hr, depending on the quantity of samples to be plated. The whole printing run can take from 10 min to days depending on how many samples and how many slides need to be printed. Staining run Slide preparation (stripping plus blocking) takes 1.5 hr. The time necessary for a staining run may vary depending on the protocol used. An automated protocol saves a lot of time and

Protein Arrays

27.7.27 Current Protocols in Protein Science

Supplement 75

Table 27.7.3 Troubleshooting RPPA Problems

Phase

Problem

Possible cause(s)

Cell culture lysis

Lysate is too viscous Small volume of lysis buffer

Solution(s) Increase the volume of lysis buffer used

High DNA concentration

Sonicate sample on ice for three cycles of 10 sec each at 15-μm amplitude

Tissue does not detach from the slide

Too long first fixation in 70% ethanol

Decrease the dipping time to

Preparation and use of reverse protein microarrays.

Reverse-phase protein array (RPPA) is a multiplex, high-throughput proteomic technique for profiling the activation status of signal transduction path...
863KB Sizes 0 Downloads 0 Views