Development of a Microflow System for In-Cell Footprinting Coupled

Sep 28, 2016 - Department of Chemistry and Chemical Biology, Indiana University−Purdue University Indianapolis, Indianapolis, Indiana 46202, United ...
1 downloads 9 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Development of a Microflow System for In-Cell Footprinting Coupled with Mass Spectrometry Aimee Rinas,† Vishaal S. Mali,‡ Jessica A. Espino,†,§ and Lisa M. Jones*,†,§ †

Department of Chemistry and Chemical Biology, Indiana University−Purdue University Indianapolis, Indianapolis, Indiana 46202, United States ‡ Avon High School, Avon, Indiana 46123, United States S Supporting Information *

ABSTRACT: Fast photochemical oxidation of proteins (FPOP) has become a valuable tool for protein structural characterization. The method has recently been demonstrated to oxidatively modify solvent-accessible sites of proteins inside live cells (IC-FPOP). However, the flow system used for in vitro analysis is not well-suited for IC-FPOP as a number of factors can lead to cell aggregation, causing inconsistent labeling and clogging. Here, we present an IC-FPOP flow system that centrally focuses the cells, ensuring consistent radiation exposure. Fluorescence imaging was used to analyze the effectiveness of the system in focusing the cells. Analysis shows the cells flowing individually through the center of the capillary with the buffer visible along the walls and with no aggregation or clogging observed. To ensure the flow system does not disturb oxidative modification, Vero cells were labeled using the flow system and analyzed by liquid chromatography−mass spectrometry (LC−MS). The results demonstrate a 13-fold increase in the number of oxidized proteins and a 2 orders of magnitude increase in the dynamic range of the method.

H

for and for the analysis of proteins that are difficult to purify and study in vitro. While this initial report demonstrates a new approach for probing the solvent accessibility of proteins in their native cellular environment, only 105 endogenous proteins were identified as being oxidatively modified. This is only a small fraction of the thousands of human proteins found in the Swiss-Prot database7 that was used for data analysis. Surprisingly, because these 105 proteins had varying expression profiles, the low number of modified proteins cannot be attributed to only the most abundant proteins being modified. Three factors that could have led to the low number of modified proteins include excessive decomposition of H2O2 by cellular catalase, cell clumping, and cell settling. For the reaction of amino acids with hydroxyl radicals to reach completion, oxygen is required. Owing to the low levels of oxygen present in mammalian cells, the decomposition of H2O2

ydroxyl radical protein footprinting (HRPF) coupled with mass spectrometry is increasingly gaining popularity as a part of structural proteomics. The method, which utilizes hydroxyl (OH) radicals to oxidatively modify solvent-accessible sites in proteins, can probe protein binding and conformational changes. The efficacy of the method lies in the similarity of OH radicals to water which allows for higher structural resolution compared to larger-sized probes.1 The fact that the reactivities and reaction mechanisms of amino acid chains with OH radicals are well-characterized and that there are various methods for generating radicals add to the utility of the method.2−4 Fast photochemical oxidation of proteins (FPOP), the method of OH radical generation used for this work, generates radicals through laser-induced photolysis of hydrogen peroxide (H2O2).5 Recently, a novel in-cell FPOP (IC-FPOP) application was reported in which FPOP was used to label proteins in live cells.6 This is in contrast to previous FPOP studies that were performed in vitro on purified proteins. In-cell labeling allows for the complexity of the cellular environment to be accounted © XXXX American Chemical Society

Received: June 18, 2016 Accepted: September 28, 2016

A

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

low-cost options to make these chips are not UV transparent at 248 nm and the materials that are would be costly. To overcome this, we endeavored to design a microfluidics system that is compatible with IC-FPOP. The work described here focuses on a flow system prototype that was designed and built specifically for IC-FPOP to increase the labeling yield of the method. The flow system’s low cost and modular design makes it easy to change out parts if necessary. The efficacy of the flow system was evaluated for both its ability to centrally focus the cells as well as its performance when used for IC-FPOP.

by endogenous catalase to generate oxygen aids modification by IC-FPOP. However, the mixing of H2O2 with the cells for several minutes prior to laser photolysis may lead to excessive decomposition limiting radical production. For IC-FPOP, cells and H2O2 are mixed in a syringe and flowed through a flow tube toward the excimer laser pulse. At a flow rate of 57 μL/ min it takes 10 min for all of the cells to reach the laser. By this time, much of the H2O2 may be decomposed by catalase leaving a limited concentration for laser photolysis leading to a low abundance of OH radicals. The second factor, cell clumping, is a phenomenon displayed by several cell types in suspension. In the case of IC-FPOP, this could lead to nonuniform labeling with some cells never being exposed to the laser light. It could also lead to flow tube clogging. The method is susceptible to cell settling in the syringe owing to the minutes-long time frame of the flow. This settling would lead to a lower number of cells being exposed to the laser and collected at the end of the experiment resulting in lowered sensitivity of the method. To address all of these factors and increase the number of oxidatively modified proteins, we wanted to employ a microfluidics system for IC-FPOP. For this system, we wanted to create a single stream of cells that flow toward the laser while mixing the cells with the H2O2 just prior to laser photolysis. When considering the requirements for IC-FPOP, we looked to flow cytometers, which have a key aspect in common: cells in solution are ordered into a centrally focused stream. In flow cytometry, single-cell flow is achieved via hydrodynamic focusing.8 This effect of fluid dynamics arises when two fluids in contact in a straight microfluidic channel are introduced at different velocities (Figure 1). The fluid moving at a higher

■

EXPERIMENTAL SECTION Chemicals and Reagents. Dimethylthiolurea (DMTU), 30% hydrogen peroxide, dimethyl sulfoxide (DMSO), N-tertbutyl-phenylnitrone (PBN), fluorescein isothiocyanate (FITC), and triethylammonium bicarbonate buffer (TEABC) were purchased from Sigma-Aldrich (St. Louis, MO). Vero cells were purchased from ATCC (Manassas, VA). Dulbecco’s modified Eagle’s medium (DMEM), sterile phosphate-buffered saline (PBS), trypsin−EDTA, penicillin−streptomycin, and 6diamidino-2-phenylindole (DAPI) were purchased from Life Technologies (Grand Island, NY). LC−MS grade solvents, acetone, fetal bovine serum (FBS), iodoacetamide (IAA), dithiothreitol (DTT), tetramethylrhodamine (TMRM), nuclease, and trypsin were purchased from Fisher Scientific (Thermo Fisher Scientific, Waltham, MA). Flow cell fittings were obtained from IDEX Health and Science (Oak Harbor, WA). Polyimide-coated fused silica was obtained from Polymicro Technologies (Phoenix, AZ). In-Cell FPOP. IC-FPOP was completed as previously described.6 Briefly, Vero cells were grown to 70% confluence in a T-175 flask in DMEM supplemented with FBS and streptomycin−penicillin. The cells were trypsinized, centrifuged, and resuspended in 6 mL of sterile PBS. Each sample was prepared with 450 μL of cells and 50 μL of PBS. For ICFPOP without the flow system, the cells and PBS were passed through a flow cell at 57 μL/min. For the mixing studies and IC-FPOP with the flow system, the cells were mixed in the syringe using VP710 tumble stirrer (V&P Scientific, San Diego, CA) with six VP724F stir discs to prevent cell settling. For ICFPOP with the flow system, hydrogen peroxide, 20 mM, was mixed with the cells at the mixing tee (Figure 2). Samples were

Figure 1. Hydrodynamic focusing of particles.

velocity, referred to as the sheath, fills a larger portion of the channel.9,10 The radius of the sample stream is a function of flow rate and varies when the flow rates of the sample and sheath streams are changed.11 Using hydrodynamic focusing, flow cytometers have been successful in single-cell analysis of several different cell types for immunophenotyping,10 cell sorting,12 cell cycle analysis,13 cell proliferation assays,14 and other applications. Previous work has been done designing miniature microfluidic flow cells using hydrodynamic focusing. However, the

Figure 2. IC-FPOP flow cell schematic. Blue arrows indicate flow, and colored lines represent tubing (not to scale). Optimal conditions were observed with a 10:1 sheath buffer to cellular analyte ratio: 360 μm o.d., 75 μm i.d. capillary in orange, 360 μm o.d., 150 μm i.d. capillary in red, 673 μm o.d., 450 μm i.d. capillary in dark blue, window for laser light in light blue. B

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry passed through a flow cell system at a total flow rate of 484 μL/ min. The excimer laser was set to a pulse frequency of 20 Hz, with a laser energy of 124 mJ and a pulse width of 2.55 mm. Cells were collected in a tube containing 4 mL of a quench solution composed of 100 mM DMTU and 100 mM PBN. DMSO (1%) was added to the quench solution to inhibit methionine sulfoxide reductase. Cells were labeled in biological duplicate, each in technical triplicate with an equal number of controls (no laser irradiation). Cell Lysis and Proteolysis. Post FPOP, cells were centrifuged, rinsed with PBS, resuspended in 100 μL of radioimmunoprecipitation assay (RIPA) buffer, frozen in liquid nitrogen, and stored at −80 °C overnight. Cells were thawed, incubated at 95 °C for 5 min, and cooled on ice for 5 min. The samples were treated with nuclease at room temperature for 15 min, centrifuged, and the supernatant was transferred to a new tube. The cell lysate was reduced with DTT, alkylated with IAA, and acetone-precipitated overnight. The precipitate was resuspended in 25 mM TEABC and digested with trypsin overnight at 37 °C. Formic acid (FA) was added to a final concentration of 5% to quench digestion. Liquid Chromatography−Tandem Mass Spectrometry (LC−MS/MS). Analysis was completed using an UltiMate 3000 RSLC and a Q Exactive mass spectrometer (Thermo Fisher Scientific, Waltham, MA) as previously described.6,15 For each experiment, the digest was loaded onto a 2 cm Acclaim Pepmap 100 C18 trap column (Thermo Fisher Scientific, Waltham, MA) and washed for 40 min with loading buffer (2% acetonitrile with 0.1% formic acid) at a flow rate of 2.5 μL/ min. The samples were separated on a 75 μm inner diameter analytical column packed in-house with a 30 cm bed of Magic 5 μm particles (Michrom Bioresources Inc., Auburn, CA). Peptides were eluted with a 141 min stepped gradient at a flow rate of 300 nL/min from 4% to 10% acetonitrile (ACN)/ 0.1% FA over 10 min and 10−45% ACN/0.1% FA over 140 min. The total run time was 207 min including loading, washing, and equilibration. MS1 spectra were acquired over an m/z range of 300−2000 at a resolving power of 70 000 for 400 m/z ions, with a dynamic exclusion of 20 s. The 25 most abundant ions were selected for MS2 at a resolving power of 17 500 for 400 m/z ions. Ions with a charge state of +7, +8, and >+8 were rejected. AGC targets were set to 3 × 106 for MS1 and 1 × 105 for MS2 with an under fill ratio of 5%, giving an intensity threshold of 5 × 103. Data Analysis. All data files were searched as previously described,16 using Proteome Discoverer version 1.4 (Thermo Fisher Scientific, Waltham, MA) with Sequest HT and Mascot version 2.4 (Matrix Sciences Ltd., London, U.K.) against the Swiss-Prot reviewed human FASTA database that contains 20 193 proteins. Extracted ion chromatogram (EIC) areas for each peptide spectrum match (PSM) were calculated using a custom multilevel workflow.16 Briefly, five search algorithm levels were constructed in Proteome Discoverer and the commonly observed FPOP modifications were dispersed across the individual search levels. The fragment ion tolerance was set at 0.02 Da and the parent ion tolerance at 10 ppm, with the enzyme specificity set to trypsin with one missed cleavage. Peptides were ungrouped and filtered to a 5% false discovery rate (FDR). The data were exported to Excel and summarized using the PowerPivot add-in. Proteins were accepted if at least two distinct peptides were identified with the 5% FDR filter. The fractional oxidation per peptide or residue was determined according to the following equation:

∑ EIC area modified ∑ EIC area

(1)

where, for peptide-level analysis, EIC area modified is the EIC area of a PSM containing an FPOP modification, and EIC area is the EIC area of any PSM with a sequence identical to that containing the modification. For residue-level analysis, EIC area modified is the EIC area of a PSM for a specific modified residue and EIC area is the EIC area of a PSM with sequences identical to those containing the modification. Fluorescence Imaging. Vero cells were grown and harvested as described above. Cells were incubated in 300 nM DAPI for 5 min and resuspended in PBS. TMRM and FITC were added at concentrations of 5 μg/mL to the cells/ H2O2 and sheath buffer, respectively. Flow system imaging was completed on a Leica SP8 confocal laser scanning microscope (Leica Microsystems Inc., Buffalo Groves, IL) with a Nikon Fluor 40×/0.80 W DLL objective (Nikon Instruments Inc., Melville, NY), and with an excitation/emission of 405/410− 483 nm for DAPI, 488/505−547 nm for FITC, and 552/652− 752 nm for TMRM.

■

RESULTS AND DISCUSSION Flow Cell Construction. The flow system (Figure 2) was constructed entirely from standard IDEX parts and Polymicro fused-silica capillaries with a modular design that can be adapted for different experimental requirements. The cellular analyte solution and H2O2 enter the system through separate capillaries (360 μm o.d., 75 μm i.d.) where they are mixed by a mixing tee to keep the exposure time to H2O2 constant throughout the experiment and to prevent endogenous catalase from completely decomposing the H2O2. The capillary (360 μm o.d., 75 μm i.d.) carrying the mixed cellular analyte and H2O2 is mounted so that it passes through the cross and is concentric with the larger diameter flow cell capillary (673 μm o.d., 450 μm i.d.). As shown in Figure 2, the smaller outer diameter capillary carrying the mixed cellular analyte (orange) fits entirely within the larger capillary leading to the laser (blue) leading them to share a common axis. The two capillaries (360 μm o.d., 150 μm i.d.) carrying the sheath buffer are flushmounted with the screw which enables the sheath buffer to completely surround the cellular analyte/H2O2 capillary before any cells enter the capillary, centrally focusing the cells and reducing the diameter of the solution carrying the cells/H2O2. Cells are irradiated through a window in the flow cell capillary created by burning off a portion of the polyimide coating. Flow Dynamics. Similar to flow cytometry, the flow system has a central capillary through which the cells are injected, surrounded by a faster flowing sheath buffer. Laminar flow prevents the cells and sheath from mixing. However, there is one significant design difference between this flow system and those used in flow cytometers. Typically, the central capillary is tapered, and this, along with the faster flowing sheath, hydrodynamically focuses the cells. The IC-FPOP experimental parameters as well as the backpressure limitations of the syringe pump and fittings did not allow for the central capillary to be tapered. As such, the blunt end of the central capillary in the IC-FPOP flow system creates an area that may potentially cause turbulence and disrupt laminar flow. Fluorescence imaging was performed to evaluate the ICFPOP flow systems effectiveness in hydrodynamically focusing the cells. Fluorophores were added to each solution to ensure laminar flow was maintained and to establish the location of C

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 3. (A) Average intensity projection of sheath buffer (green) and cellular analyte (red). (B) Maximum intensity heat map showing the locations of detected cells. Three-dimensional average intensity heat map of the sheath buffer (C) and cellular analyte (D). Lower intensities are blue and highest are red (B−D). Orthogonal YZ stack illustrating 3D focusing (E) and location of cells (F, red) with the sheath and cellular analyte in buffer in gray.

each solution within the flow cell; FITC (green) was added to the sheath PBS, and TMRM (red) was added to the cellular solution with PBS (used in place of H2O2 for imaging) and, collectively, will be referred to as cellular analyte. Additionally, the cell nuclei were stained with DAPI (blue). To establish the average position of each of the flowing solutions, an average intensity projection was generated using the Icy bioimaging platform.17,18 This shows the average intensity of each pixel location over all of the 2666 frames (recorded at 26.6 fps) in the run (Figure 3A). To generate the projection, the RGB channels were separated from the multichannel image stack, an average intensity projection for the red (TMRM) and green (FITC) channels was performed, and the resulting projections were then merged. Analysis shows the cellular analyte was centrally focused into a single stream approximately 50 μm in diameter which is 25 μm smaller than the capillary inner diameter. The blurring between the sheath and cellular analyte is expected as it is an average over all of the frames. Each individual frame is grainy owing to the flow velocity and image capture speed. Other contributing factors may be vibrations causing slight movements of the capillary and crosstalk between the FITC and TMRM. To address this issue, a threedimensional (3D) average heat map was generated for the individual TMRM and FITC projections (Figure 3, parts B and C). The map clarifies the blur between the channels and clearly shows the highest intensity of the sheath toward the sides of the

field of view (Figure 3B) and down the center for the cellular analyte (Figure 3C). This illustrates that the flow system is operating under laminar flow conditions as mixing due to turbulence would not display a clear delineation in intensity.19 Fluorescence imaging was also used to determine the location of the cells as they pass through the flow cell capillary. In addition to the DAPI staining of the nucleus, the TMRM dye also stains the mitochondria of live cells as the accumulation of the dye is driven by the mitochondria membrane potential.20,21 Since not every pixel location contains cells, a maximum intensity projection was generated, rather than an average, to show the location of the cells. To generate this projection, a median filter was applied to remove many of the individual pixels. These pixels are likely from background noise or the TMRM that was in the cellular analyte but not taken up by the cells. The maximum intensity projections for the filtered TMRM and DAPI channels were merged, and a two-dimensional (2D) heat map was generated (Figure 3D). Nearly all of the pixels from this map are located within the same dimensions of the cellular analyte from Figure 3A. This demonstrates that the cells are kept within the focused cellular analyte channel at the center of the flow cell capillary. As a confocal microscope was used for imaging, the images can only demonstrate that the cellular analyte is centrally focused on one XY plane of the flow cell capillary and cannot provide any evidence of 3D hydrodynamic focusing. To assess the 3D focusing, a cross-sectional YZ stack was collected D

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (Figure 3E). Although the focusing in the YZ plane is not as condensed as the XY plane, the cellular analyte is unquestionably centered through the flow cell capillary, confirming the flow system is able to centrally focus the cells. Additionally, the cross-sectional stack captured cells as they flowed through the capillary (Figure 3F), further showing the cells stay centrally 3D focused. The discrepancy in XY and YZ focusing is likely due to the cross since the sheath buffer enters the flow system horizontally perpendicular to the cellular analyte capillary. While the focusing is adequate for IC-FPOP, it could be improved by using a six-port flow-through manifold with the sheath buffer entering the flow system both horizontally and vertically perpendicular to the cellular analyte capillary. IC-FPOP Using the Flow System. IC-FPOP of Vero cells using the flow system was completed as previously described so that the data could be compared to the previously published results.6 However, there were three experimental design parameters that were adjusted from those published. First, a flow rate of 484 μL/min was used. This was the fastest that could be used at the 20 Hz pulse frequency with no exclusion volume, and is about half the rate used without the flow system when scaled for the flow system. Second, syringe stir bars were used to prevent the cells from settling in the syringe. Lastly, there were minor changes in the digestion protocol. The Thermo Scientific Pierce mass spectrometry prep kit for cultured cells protocol was used, but with reagent substitutions; RIPA and TEABC buffers were used, and only trypsin was used for digestion. All other parameters including the preparation of cells, sample volume, quench, H2O2 final concentrations, and three technical replicates were consistent. The sheath buffer used for IC-FPOP with the single-cell flow system was sterile PBS. Increased Oxidized Protein Identifications. The data from IC-FPOP coupled with the flow system were searched against human FASTA sequences as the Vero cell proteome is incomplete. The search identified a total of 1391 oxidatively modified endogenous proteins over the two biological replicates (Table S1). This corresponds to a 13.2-fold increase in oxidized proteins identified in the previously published work without the flow system (Figure 4A). Individually, the biological replicates also had sizable gains in the number of

proteins oxidatively modified, with an 8.2-fold increased for the first replicate (n = 864) and a 9.8-fold increase for the second (n = 1025), with 498 proteins in common between the replicates (Table S1, Figure 4B). Each biological replicate is a sum of the oxidized proteins across three technical replicates. The variation between the numbers of proteins modified in the two biological replicates is not well-understood but may be attributed to the cell-to-cell variability in protein expression.22,23 Although there is significant overlap in the technical replicates, variation was also observed (Figure S1). This variation in the technical replicates was not expected, and further optimization of method parameters will be carried out to increase reproducibility. Since the magnetic stirrer was not used for the previously published results, the effect of the stirring system was evaluated by running three technical replicates with controls without the flow system (Figure 4C). While using the magnetic stirrers did double the number of oxidized protein identified (n = 211 vs n = 105), the majority of the gains made can be attributed to the flow system, as there are 1183 more proteins identified. Fifty-eight of the 105 proteins modified in the absence of the single-cell flow system were also modified with the flow system (Table S2, Figure 4A). Of these, 33 were modified in both biological replicates, 3 were modified in biological replicate one only, and 22 were modified in biological replicate two only. This overlap in modified proteins allows for a secondary examination of the efficacy of the flow system for IC-FPOP. The number of oxidatively modified residues in each of the 58 proteins modified both with and without the flow system was analyzed (Table S2). This data demonstrates that, along with an increase in the number of modified proteins, the flow system also led to an increase in the number of modifications per protein. Of the 36 proteins that were oxidatively modified both with the flow system in biological replicate one and without the flow system, 11 proteins demonstrated a decrease in the number of oxidations and 2 proteins had an equal number of oxidations (Table S2). The remaining 23 proteins displayed an increased number of oxidatively modified residues. Although some gains were modest, as is the case for aconitate hydratase and galectin-1 that only show one additional modification, several proteins had a large increase in the number of modified residues. Heat shock protein HSP-90α and filamin-C had an additional eight and six residues modified, respectively (Table S2). The differences were even more pronounced in biological replicate two. Of the 55 proteins in biological replicate two that overlapped with the data without the flow system, 37 had increased oxidations, 13 had decreased oxidations, and 5 had the same number of oxidations. Myosin-9 had an additional 54 residues modified in comparison to without the flow system. Heat shock protein HSP 90-α, heat shock protein HSP 90-β, and filamin had an additional 20, 11, and 13 oxidatively modified residues (Table S2). The single-cell flow system does have a higher efficacy with IC-FPOP than when the flow system is not used both in the number of proteins modified and in the number of residues modified in each protein. The discernible advantage of using the flow system for ICFPOP is likely attributed to several factors. One can postulate that some of the cells exposed to the radiation will lyse, exposing the DNA which can lead to cell clumping.24 These clumps may then stick to the wall of the capillary as the DNA would be attracted to the fused silica.25 The buffer sheath keeps the cells away from the walls of the flow cell capillary, and any cell aggregates that may form will be pushed through the flow

Figure 4. Comparison of oxidized proteins identified with and without the flow system (A), the biological replicates (B), and (C) with and without mixing. E

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

for 137 different proteins that were oxidatively modified (Table S1), yielding a fold change of 1.25 × 104. The proteins used to determine the dynamic range in the previously published report were also evaluated with this database for an equivalent comparison and resulted in a fold change of 350, with 278 FPKM for heat shock cognate protein and 0.8 FPKM < for shroom2. On the basis of these expression profiles, using the flow system increased the dynamic range of IC-FPOP by approximately 2 orders of magnitude.

system before they accumulate on the walls and grow. Another factor is the method of adding the H2O2 to the sample. With the flow system, the H2O2 is mixed with the cells just prior to entering the flow cell capillary (Figure 2), thereby reducing the time the peroxide is exposed to endogenous catalase and reducing peroxide decomposition. Finally, the smaller inner diameter of the capillary used to deliver the cells in the flow system along with the reduced width of the stream from the hydrodynamic focusing may have an overall smaller dead volume depending on the length of the capillary used. As illustrated in Figure 4C, the use of stir bars to keep the cells in suspension played a minimal role in the increased identifications. Previous studies with in vitro FPOP determined the efficacy of separating the peroxide from the protein samples and using hydrodynamic focusing with FPOP. Zhang et al.26 demonstrated a modified FPOP setup with peroxide and protein mixing just prior to laser photolysis that provides better reproducibility and less peroxide-induced oxidation. Vahidi et al.27 utilized hydrodynamic focusing with FPOP for sample mixing to investigate protein unfolding. However, this study utilizes hydrodynamic focusing differently than our flow system. First, we perform mixing prior to hydrodynamic focusing, whereas Vahidi et al.27 uses focusing for mixing to induce a pH change in the protein buffer. Second, our system uses the hydrodynamic focusing to form a single line flow of cells. The results of our flow system utilizing peroxide mixing and hydrodynamic focusing leading to increased modifications aligns with these previous results that mixing and hydrodynamic focusing are compatible with FPOP. Properties of Oxidized Proteins. IC-FPOP without the flow system showed that proteins in various cellular compartments were oxidatively modified. This indicates the H2O2 was able to permeate the various organelles of the cell prior to laser photolysis. It is possible that the decreased time where the cells are in contact with H2O2 limits the permeation of the peroxide to the organelles. To investigate this, the oxidized proteins identified using the flow system were analyzed to establish the location of the protein within the cellular organelles. Oxidatively modified proteins were found in 27 different compartments within the cell (Table S3). There were a large number of proteins located in the cytoplasm and membrane which is expected. However, there were also a significant number of proteins identified that are located within protein complexes, such as the cytoskeleton, and organelles, such as the nucleus. This provides evidence that using the mixing tee does indeed mix the cells with the H2O2, and does not affect the ability of H2O2 to fully permeate the cells. The dynamic range of the IC-FPOP method was evaluated in the previous work by comparing the oxidized proteins identified with their expression level (transcripts per million, TPM) in human kidney cells, estimated at 3 orders of magnitude. This was based on the fold change between the protein with the highest TPM’s heat shock cognate protein at 4635, and the lowest, protein shroom2 at 4 TPMs. Both of these proteins were oxidized using the flow system, indicating that the dynamic range did not decrease. The full dynamic range of IC-FPOP using the flow system was assessed using data from the Human Protein Atlas,28 which contains expression profiles on 83 normal human cell lines from 44 tissues, including the kidney which is used here. Maximum and minimum fragments per kilobase million (FPKM) for oxidized proteins were 1227 for elongation factor 1-α (EEF1A1) and 0.1

■

CONCLUSIONS The work presented here has demonstrated a process improvement for IC-FPOP using a custom flow system prototype. This low-cost, modular design hydrodynamically focuses the cells in a small stream down the center of the flow cell capillary preventing cell aggregation and system clogging. Additionally, the focusing of the cells ensures that radiation exposure remains consistent throughout the experiment. The IC-FPOP data presented here has validated that using the flow system for IC-FPOP results in a dramatic increase in the identification of oxidized proteins and the number of residues modified within those proteins, without compromising the dynamic range of the method or the ability to modify proteins in various cellular compartments. Although this manuscript discusses the use of this flow system for the protein footprinting method IC-FPOP coupled with mass spectrometry, the system can be used for other applications. There is a need for inexpensive, portable flow cytometers where small optical components can be incorporated.29 The flow system described here could be used in this manner as it is small and portable and can be used with a programmable syringe pump. Here, the flow system was paired with an excimer laser, but it is feasible that another optical component can be paired with the system.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b02357. Comparison of technical replicates, list of oxidatively modified proteins and residues, list of residues modified both with and without the flow system, and abbreviations of cellular components (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 410-706-3380. Present Address §

J.A.E. and L.M.J.: Department of Pharmaceutical Sciences, University of Maryland, Baltimore, MD 21201, United States. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare the following competing financial interest(s): The authors declare that a patent (U.S. Patent No. PCT/US2016/025188) has been granted for this flow system. F

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

■

G.; Freed, D.; Zahari, M. S.; Mukherjee, K. K.; Shankar, S.; Mahadevan, A.; Lam, H.; Mitchell, C. J.; Shankar, S. K.; Satishchandra, P.; Schroeder, J. T.; Sirdeshmukh, R.; Maitra, A.; Leach, S. D.; Drake, C. G.; Halushka, M. K.; Prasad, T. S.; Hruban, R. H.; Kerr, C. L.; Bader, G. D.; Iacobuzio-Donahue, C. A.; Gowda, H.; Pandey, A. Nature 2014, 509, 575−581. (29) Ateya, D. A.; Erickson, J. S.; Howell, P. B., Jr.; Hilliard, L. R.; Golden, J. P.; Ligler, F. S. Anal. Bioanal. Chem. 2008, 391, 1485−1498.

ACKNOWLEDGMENTS This work was supported by a Grant from the National Science Foundation (1552509). The authors would like to thank Seth Winfree for help with imaging.

■

REFERENCES

(1) Xu, G.; Chance, M. R. Chem. Rev. 2007, 107, 3514−3543. (2) Wang, L.; Chance, M. R. Anal. Chem. 2011, 83, 7234−7241. (3) Watson, C.; Janik, I.; Zhuang, T.; Charvatova, O.; Woods, R. J.; Sharp, J. S. Anal. Chem. 2009, 81, 2496−2505. (4) Hayes, J. J.; Kam, L.; Tullius, T. D. Methods Enzymol. 1990, 186, 545−549. (5) Hambly, D. M.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2005, 16, 2057−2063. (6) Espino, J. A.; Mali, V. S.; Jones, L. M. Anal. Chem. 2015, 87, 7971−7978. (7) Bairoch, A.; Apweiler, R. Nucleic Acids Res. 2000, 28, 45−48. (8) Watson, J. V. Introduction to Flow Cytometry; Cambridge University Press: Cambridge, U.K., 1991. (9) Golden, J. P.; Justin, G. A.; Nasir, M.; Ligler, F. S. Anal. Bioanal. Chem. 2012, 402, 325−335. (10) Brown, M.; Wittwer, C. Clin Chem. 2000, 46, 1221−1229. (11) Zarrin, F.; Dovichi, N. J. Anal. Chem. 1985, 57, 2690−2692. (12) Shapiro, H. M. Practical Flow Cytometry, 4th ed.; Wiley-Liss: New York, 2003; pp l−681. (13) Galbraith, D. W.; Harkins, K. R.; Maddox, J. M.; Ayres, N. M.; Sharma, D. P.; Firoozabady, E. Science 1983, 220, 1049−1051. (14) Vermes, I.; Haanen, C.; Steffens-Nakken, H.; Reutellingsperger, C. J. Immunol. Methods 1995, 184, 39−51. (15) Rinas, A.; Jones, L. M. J. Am. Soc. Mass Spectrom. 2015, 26, 540− 546. (16) Rinas, A.; Espino, J. A.; Jones, L. M. Anal. Bioanal. Chem. 2016, 408, 3021−3031. (17) de Chaumont, F.; Dallongeville, S.; Chenouard, N.; Herve, N.; Pop, S.; Provoost, T.; Meas-Yedid, V.; Pankajakshan, P.; Lecomte, T.; Le Montagner, Y.; Lagache, T.; Dufour, A.; Olivo-Marin, J. C. Nat. Methods 2012, 9, 690−696. (18) de Chaumont, F.; Dallongeville, S.; Olivo-Marin, J. C. IEEE International Symposium on Biomedical Imaging 2011, 8, 234−237. (19) Park, H. Y.; Qiu, X.; Rhoades, E.; Korlach, J.; Kwok, L. W.; Zipfel, W. R.; Webb, W. W.; Pollack, L. Anal. Chem. 2006, 78, 4465− 4473. (20) Ehrenberg, B.; Montana, V.; Wei, M. D.; Wuskell, J. P.; Loew, L. M. Biophys. J. 1988, 53, 785−794. (21) Farkas, D. L.; Wei, M. D.; Febbroriello, P.; Carson, J. H.; Loew, L. M. Biophys. J. 1989, 56, 1053−1069. (22) Albert, F. W.; Treusch, S.; Shockley, A. H.; Bloom, J. S.; Kruglyak, L. Nature 2014, 506, 494−497. (23) Cookson, N. A.; Cookson, S. W.; Tsimring, L. S.; Hasty, J. Nucleic Acids Res. 2010, 38, 2676−2681. (24) Crissman, H. A.; Mullaney, P. F.; Steinkamp, J. A. Methods Cell Biol. 1975, 9, 179−246. (25) Vandeventer, P. E.; Mejia, J.; Nadim, A.; Johal, M. S.; Niemz, A. J. Phys. Chem. B 2013, 117, 10742−10749. (26) Zhang, Y.; Rempel, D. L.; Zhang, H.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2015, 26, 526−529. (27) Vahidi, S.; Stocks, B. B.; Liaghati-Mobarhan, Y.; Konermann, L. Anal. Chem. 2012, 84, 9124−9130. (28) Kim, M. S.; Pinto, S. M.; Getnet, D.; Nirujogi, R. S.; Manda, S. S.; Chaerkady, R.; Madugundu, A. K.; Kelkar, D. S.; Isserlin, R.; Jain, S.; Thomas, J. K.; Muthusamy, B.; Leal-Rojas, P.; Kumar, P.; Sahasrabuddhe, N. A.; Balakrishnan, L.; Advani, J.; George, B.; Renuse, S.; Selvan, L. D.; Patil, A. H.; Nanjappa, V.; Radhakrishnan, A.; Prasad, S.; Subbannayya, T.; Raju, R.; Kumar, M.; Sreenivasamurthy, S. K.; Marimuthu, A.; Sathe, G. J.; Chavan, S.; Datta, K. K.; Subbannayya, Y.; Sahu, A.; Yelamanchi, S. D.; Jayaram, S.; Rajagopalan, P.; Sharma, J.; Murthy, K. R.; Syed, N.; Goel, R.; Khan, A. A.; Ahmad, S.; Dey, G.; Mudgal, K.; Chatterjee, A.; Huang, T. C.; Zhong, J.; Wu, X.; Shaw, P. G

DOI: 10.1021/acs.analchem.6b02357 Anal. Chem. XXXX, XXX, XXX−XXX