Mass-Spectrometry-Based Quantification of Protein-Bound Fatty Acid

Subscriber access provided by Flinders University Library

Article

Mass spectrometry-based quantification of protein-bound fatty acid synthesis intermediates from Escherichia coli Marek J. Noga, Mattia Cerri, Nicole Imholz, Pawel Tulinski, Enes #ahin, and Gregory Bokinsky J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00405 • Publication Date (Web): 05 Sep 2016 Downloaded from http://pubs.acs.org on September 5, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Mass spectrometry-based quantification of proteinbound fatty acid synthesis intermediates from Escherichia coli Marek J. Noga*, Mattia Cerri, Nicole Imholz, Pawel Tulinski, Enes Şahin, Gregory Bokinsky* Department of Bionanoscience, Delft University of Technology, Kavli Institute of Nanoscience Delft, Lorentzweg 1, 2628CJ Delft, The Netherlands KEYWORDS: post-translational modification, fatty acids, phospholipids, acyl carrier protein, phosphopantetheine, mass spectrometry, metabolomics, biofuels, quantitative proteomics, thioesters

ABSTRACT: The production of fatty acids from simple nutrients occurs via a complex biosynthetic pathway with dozens of intermediate compounds and multiple branch points. Despite its importance for microbial physiology and biotechnology, critical aspects of fatty acid biosynthesis, especially dynamics of in vivo regulation, remain poorly characterized. We have developed a liquid chromatography / mass spectroscopy (LC/MS) method for relative quantification of fatty acid synthesis intermediates in Escherichia coli, a model organism for studies of fatty acid metabolism. The acyl carrier protein, a vehicle for the substrates and intermediates of fatty acid synthesis, is extracted from E. coli, proteolytically digested, resolved using reverse-phase LC, and detected using electrospray ionization coupled with a tandem MS.

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 26

Our method reliably resolves 21 intermediates of fatty acid synthesis, with an average relative standard deviation in ratios of individual acyl-ACP species to total ACP concentrations of 20%. We demonstrate that fast sampling and quenching of cells is essential to accurately characterize intracellular concentrations of ACP species. We apply our method to examine the rapid response of fatty acid metabolism to the antibiotic cerulenin. We anticipate that our method will enable the characterization of in vivo regulation and kinetics of microbial fatty acid synthesis at unprecedented detail, and will improve integration of fatty acid synthesis into models of microbial metabolism.

INTRODUCTION Microbial fatty acid synthesis has attracted interest as an antibiotic target,1 and as a source of renewable fuels.2 Fatty acids are synthesized by iterative condensation of acetyl-CoA units with short-chain intermediates (Figure 1A). Intermediates of fatty acid synthesis are covalently attached to a small 8.6 kDa acyl carrier protein (AcpP in Escherichia coli, hereafter referred to as ACP) via a phosphopantetheine group (Figure 1B). ACP-bound fatty acid intermediates are shuttled between enzymes that catalyze elongation, dehydration, and reduction reactions. Finally, the acyl group is transferred from ACP to a phosphorylated head group, forming a phospholipid. In E. coli, the complete synthesis of a single 16-carbon saturated fatty acid chain involves 31 intermediates and 9 enzymes, placing fatty acid synthesis among the longest anabolic pathways. Acyl-ACP species are also used as precursors of other essential compounds, such as biotin.3 As membrane synthesis and integrity is essential for viability, fatty acid synthesis is subject to both genetic and enzyme-level regulation.4 However, despite decades of study,5-7 important aspects of fatty acid synthesis still remain poorly characterized. Investigations of in vivo fatty


2

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


acid synthesis regulation, which have led to the discovery of several important regulatory interactions, often rely upon quantification of ACP-bound fatty acid intermediates.8-10 These methods use either radioisotopes or ACP-specific antibodies for detection of ACP-bound intermediates, followed by resolution via liquid chromatography or polyacrylamide gel electrophoresis (PAGE).11,

12

However, radiolabeling ACP requires auxotrophic strains, and

antibodies raised against ACP may exhibit different affinities to the various acyl-ACP intermediates. Furthermore, these techniques cannot easily resolve all fatty acid intermediates, nor can they be used for stable isotope tracing experiments, which are ideally suited for determining intracellular fluxes. A previous report used matrix-assisted laser desorption ionization coupled with a time-of-flight mass spectrometer (MALDI-TOF) to quantify several ACP species from an in vitro mixture of purified fatty acid synthesis enzymes. However, quantifying most of the ACP-bound intermediates was beyond the resolution of the technique.13 Furthermore, MALDI-TOF may not be optimal for quantifying ACP species within a cell extract. A method for quantifying ACP-bound fatty acid intermediates would greatly advance our understanding of fatty acid synthesis and better guide efforts to engineer fatty acid synthesis for production of useful compounds. Here we describe a semi-quantitative liquid chromatography/mass spectroscopy (LC/MS) method for measuring ACP-bound fatty acid intermediates within E. coli. Because of the hybrid nature of fatty acid intermediates (proteins covalently modified with fatty acid metabolites) our method combines aspects of proteome and metabolome analysis protocols. As turnover of ACP species is rapid, we use a fast-quenching approach typical of metabolite extraction methods to minimize perturbation of the ACP pools during sampling.10 Next, ACP species are proteolytically digested and peptide fragments bearing the acylated phosphopantetheine group


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

are separated using reverse-phase LC. Finally, we use electrospray ionization (ESI) and collisional fragmentation within a tandem mass spectrometer to eject the acyl-bearing pantetheine groups from the ACP peptide.14, 15 We demonstrate that tandem MS is well-suited for selectively detecting ACP species within a complex cell extract. We demonstrate the applicability of our method by examining rapid changes in acyl-ACP concentrations after exposure to an antibiotic.

Figure 1. Overview of the experimental method and the role of ACP in fatty acid synthesis. A. The E. coli fatty acid synthesis pathway. Fatty acid initiation begins with the carboxylation of acetyl-CoA by acetyl-CoA carboxylase (ACC), forming malonyl-CoA. Next, the malonyl group is transferred to the terminal thiol of a phosphopantetheine linker (represented by a dashed line) covalently linked to acyl carrier protein (ACP). Malonyl-ACP provides activated 2-carbon units used for both the initiation reaction (condensation with acetyl-CoA, catalyzed by FabH) and the elongation reactions (condensation with an acyl-ACP, catalyzed by FabF or FabB). Both


4

Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


elongation and initiation produce a β-ketoacyl-ACP, which is subsequently reduced, dehydrated, and reduced again by FabG, FabZ/FabA, and FabI respectively, generating a saturated acyl-ACP species. When the acyl group has reached 16 carbon units, the acyl chain is converted by an acyltransferase enzyme (PlsB, PlsC, or PlsX) into a phospholipid for subsequent processing and incorporation into the membrane, releasing holo-ACP. Cerulenin inhibits the elongation enzymes FabF and FabB. B. Chemical structure of the phosphopantetheine linker attached to ACP. C. Amino acid sequence of E. coli ACP, indicating relevant phosphopantetheinylated peptides generated by trypsin and GluC proteases. The signal from GluC peptide LVMALE is used to normalize phosphopantetheinylated peptide signals, as described in Experimental Methods. EXPERIMENTAL SECTION All chemicals used were purchased by Sigma Aldrich unless noted otherwise. Solvents and buffer additives used for LC/MS (acetonitrile (ACN), methanol, acetic acid, formic acid, ammonium formate, all of UPLC-MS quality) were obtained from Biosolve. Cerulenin was obtained from Cayman Chemical Company and was dissolved in dimethyl sulfoxide prior to use. All aqueous solutions were prepared in Merck Millipore Milli-Q® water. Preparation of ACP standards. acpP was amplified from E. coli MG1655 genomic DNA with primers 5ʹ-GCG CTC CAT ATG ATG AGC ACT ATC GAA GAA CGC GTT AAG AAA ATT ATC GGC G-3ʹ and 5ʹ-GCG CTC GGA TCC TTA CGC CTG GTG GCC GTT GAT GTA ATC AAT G-3ʹ (IDT DNA), digested with restriction enzymes NdeI and BamHI, and ligated into pET28a plasmid (Novagen), which appended a thrombin-cleavable N-terminal His6-tag to expressed ACP. The fusion construct was overexpressed in E. coli BL21(DE3) strain (Novagen) and pre-purified using Ni-IMAC (Biorad) using a modified procedure described by Nguyen and colleagues.16 ACP species were further isolated by pH-dependent precipitation,17 followed by


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 26

purification of holo-ACP and apo-ACP by preparative HPLC (Agilent). G1311B quaternary pump was set to deliver a linear gradient composed of 0.1% trifluoroacetic acid (TFA) in water (A) and 0.1% TFA in ACN at a flow rate of 5 mL/minute. 100 µL protein extract was injected using a G1329B autosampler cooled at 4°C onto a Zorbax 300SB column (9.4 mm ID, 150 mm length, 5 µm particle size). Elution was monitored using a G1365C UV detector and detection of peaks at a wavelength of 215 nm triggered collection of fractions by a G1364C analytical-scale fraction collector, cooled at 4°C. Fractions corresponding to holo-ACP and apo-ACP (as verified by LC/MS) were frozen and lyophilized. Protein quantity was estimated by absorbance at 280 nm, using the extinction coefficient previously determined for ACP.18 Acetyl-, malonyl-, butyrl-, octanoyl-,

dodecanoyl-,

myristoyl-

and

palmitoyl-ACP

were

each

produced

by

phosphopantetheinylation of purified ACP with appropriate CoA thioester catalyzed by Sfp synthase (NEB). Apo-ACP (0.5 µg/µL final) was dissolved in phosphate buffer (pH 7.2, 50 mM final) supplemented with MgSO4 (10 mM final) CoA thioester (2 mM final) and SFP synthase (1 µM final). Reactions were incubated for 16-18h at 4°C and full conversion of apo-ACP into acyl-ACP was verified by LC/MS analysis of GluC-digested standards (Figure S-1, Supporting Information). Acyl-ACP standards used for identification purposes were used directly from reaction mixes without further purification. Culture conditions. E. coli K-12 strain MG1655 (DSMZ 18039) was grown in MOPSbuffered minimal medium with 0.2% glucose,19 with or without 0.2% Cas-amino acids (Sigma) in a Grant Instruments Sub Aqua Pro dual water bath at 37°C. 50 mL cultures were grown in 250 mL Erlenmeyer flask agitated via stirring with a 12 mm magnetic stir bar (VWR), coupled to a magnetic stir plate (2mag MIXdrive 1 Eco\MIXcontrol 20) set at 350 rpm.


6

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Sample harvesting. 1 mL of culture at ~0.5 OD was pipetted into a 1.5 mL Eppendorf tube containing 250 µL ice-cold 10% w/vol trichloroacetic acid (TCA). The tube was inverted immediately and placed on ice. Quenched cells were pelleted by centrifugation at 4°C and the supernatant discarded by decantation. After washing once with ice-cold acetone, the pellet was dried via vacuum centrifugation. Dried samples were stored at -80°C. Preparation of

15

N-labeled internal standards. Cells were grown in MOPS-buffered

minimal medium (0.2% glucose) in which unlabeled NH4Cl was replaced with an equal concentration of 15NH4Cl (Cortecnet). At OD600 0.5, 10 mL of culture was pipetted into a 50 mL Falcon tube containing 2.5 mL ice-cold 10% w/vol TCA, inverted, and chilled on ice. Quenched cells were then pelleted by centrifugation at 4°C and the supernatant discarded by decantation. One mL of cold 1% w/vol TCA was used to wash the pellet. Aliquots of quenched 15N-labeled cells were stored at -80°C. When samples were ready for analysis, one

15

N-labeled cell pellet

was resuspended in 2 mL freshly-prepared lysis buffer (50 mM sodium phosphate buffer (pH 7.2), 10 mM N-ethylmaleimide (NEM), 1 mM ascorbic acid, 2 mM EDTA, 6 M urea). Cell lysis and protein precipitation. Lysis of the cell samples and protein precipitation was performed by a modified methanol/chloroform procedure.20, 21 Dried pellets were resuspended in 100 µL lysis buffer containing internal standards and mixed with 400 µL methanol. Following addition of 100 µL chloroform, cell pellets were resuspended by sonication within an ultrasonic bath (Branson CPX3800H) for 10 minutes at room temperature. Next, phase separation was triggered by the addition of 300 µL 200 mM formate buffer (pH 3.9) to increase precipitation of ACP.17 Samples were centrifuged, the upper layer discarded, and protein extracts pelleted by adding 300 µL of methanol and centrifuging again. The resulting pellet was washed again using 300 µL of methanol and dried via vacuum centrifugation.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 26

Proteolysis of protein extract. Dried protein extract was resuspended in 10 µL resuspension buffer (50 mM sodium phosphate (pH 6.8) and 0.4% ProteaseMAX™ (Promega), vigorously mixed and sonicated for 10 min at room temperature in an ultrasonic bath. Samples were centrifuged to remove insoluble debris, and 5 µL of the supernatant was transferred to a 0.5 mL Eppendorf tube. Proteolytic digestion was performed by addition of 10 µL 0.1 mg/mL GluC (Promega) and incubation at 37°C for at least 12 hours. In order to maintain constant protease to protein ratio, the same amount of biomass (1 mL of OD 0.5) was used for analysis, an approximate protease to protein ratio of 1:50. Proteolysis was quenched by addition of 15 µL methanol. Quantification of acyl-ACP species with LC/MS. The LC/MS system (Agilent) consisted of a binary pump (G1312B), autosampler (G7167A), temperature-controlled column compartment (G1316A) and triple quadrupole mass spectrometer (G6460C) equipped with a standard ESI source, all operated using MassHunter data acquisition software (version 7.0). 10 µL of sample was injected for analysis on an Acquity UPLC CSH130 C18 column (Waters) with 1.7 µm particle size, 2.1 mm ID and 50 mm length. A binary gradient was formed by mixing phase A (25 mM formic acid in water) and phase B (25 mM formic acid in ACN) at 0.6 mL/minute. Initial %B was set to 15%, followed by a 3 minute ramp to 25% and a 9 minute increase to 95%, supplemented by a 1 minute hold at 95% B and a 3 minute re-equilibration at initial conditions before next injection. Mass spectrometer operated in dynamic MRM mode set to unit resolution using transitions defined in Table S-1 (Supporting Information). Fragmentor setting of 130 and collision energy of 30 were used for all the acyl-ACP MRM traces. A time filter of 0.035 min was applied to the date during acquisition. Masses of phosphopantetheinylated peptides and


8

Page 9 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


corresponding acyl-pantetheine ejection ions were generated in silico with aid of Marvin Beans suite (version 16.4.18.0, ChemAxon). Analysis of LC/MS data. Acquired data was processed using Mass Hunter Quantitative Analysis software, version 7.0 (Agilent), by integration of MRM peaks. All peaks were inspected by operator and automatic integration adjusted manually where necessary. Peaks with low intensity (S/N peak height < 3) or showing heavy distortion were rejected from further analysis. Relative response (RR) of each species i was calculated by dividing the peak area by the area of the corresponding 15N-labeled species. Follow-up processing and subsequent normalization was performed using in-house developed Python scripts, using Pandas,22 Matplotlib,23 and Seaborn.24 15

N-labeled internal standard (IS) corrected signals were calculated by multiplying each relative

response by average peak area of the relevant internal standard across the whole measurement batch. Finally, to calculate relative counts of each species i, total-ACP corrected peak areas were calculated by dividing each IS-corrected signal by the peak area of LVMALE peptide, thus correcting for variations in total unlabeled ACP abundance in the sample: Relative counts =

〈Peak area,N 〉 Peak area · Peak areaLVMALE Peak area,N


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 26

Figure 2. Chromatographic traces of acyl-ACP peptides extracted from an E. coli culture during exponential growth in 0.2% glucose and 0.2% Cas-amino acids (OD = 0.5) and mixed with 15Nlabeled cells serving as internal standards. Each panel depicts a separate class of acyl-ACP (top: saturated acyl-ACP; middle: hydroxyacyl-ACP; bottom: unsaturated acyl-ACP). All unlabeled and

15

N-labeled acyl-ACP peptides were detected in a single run using the method described

(aside from in vitro standards, which were detected in a separate run). Signals corresponding to ketoacyl ACP species were detected at very low levels, precluding reliable assignments and thus are not shown here. Abbreviations: M = malonyl-ACP; H = holo ACP. Chromatograms showing alignment of

15

N-labeled acyl-ACP peptides with unlabeled acyl-ACP peptides, and individual

species with ACP standards, are shown for clarity in Figure S-4 and Figure S-5, respectively (Supporting Information).


10

Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


RESULTS AND DISCUSSION Development and validation of sample preparation and ACP detection using synthetic standards. Conventional methods for protein quantification by LC/MS rely upon proteolysis of protein samples, usually by trypsin, into peptides of sizes suitable for ionization and fragmentation. Full sequence coverage or detection of a specific sequence fragment is rarely needed for comparing protein quantities. However, identifying and quantifying acyl-ACP species requires the detection of peptides bearing the phosphopantetheine group, to which acyl intermediates are covalently linked via the side chain of serine 36. Unfortunately, digestion of ACP using trypsin generates a phosphopantetheine-bearing peptide too large to be readily analyzed by LC/MS. Therefore, we used an alternative protease to generate an appropriate peptide. GluC, which cleaves C-terminally to glutamate residues,25 produces a peptide within optimal range for LC/MS analysis (Figure 1C). To develop an ACP digestion and detection protocol, eight acyl-ACP standards of varying hydrophobicity (holo-, malonyl-, acetyl- (C2), butyryl- (C4), octanoyl- (C8), dodecanoyl- (C12), myristoyl- (C14), and palmitoyl-ACP (C16)) were prepared and digested with GluC in conditions that preserve the thioester linkage between the acyl chain and the phosphopantetheine arm (thioesters experience increased rate of hydrolysis with increased pH26) while preserving GluC activity (pH optimum 7.2). In order to prevent oxidation of the free thiol, holo-ACP was treated prior to digestion with N-ethylmaleimide (NEM).12 Digestion was performed overnight and quenched with addition of 50% methanol to maintain the solubility of peptides bearing hydrophobic acyl chains. The digested peptides were resolved using a C18 reverse-phase column, and detected with ESI-MS/MS. Modified peptides were selected based on calculated m/z of doubly charged


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 26

pseudomolecular ions using the first quadrupole. Borrowing an approach previously used to analyze polyketide synthesis intermediates tethered to a phosphopantetheine-modified peptide,15 we used collisional fragmentation within the ionization chamber to separate the acyl-pantetheine from the peptide, and set the third quadrupole to detect the expected m/z of the acyl-pantetheine arm (Figures S-2, and S-3, Supporting Information). Our method successfully detected acylpantetheine groups ejected from all eight acyl-ACP standards (Figure 2). Proteolysis of E. coli cell extract yields modified ACP peptides quantifiable by LC/MS. We developed a method for extracting ACP from growing E. coli prior to proteolysis. An aliquot of culture was withdrawn and quickly quenched into TCA to interrupt fatty acid metabolism.12 Cells were solubilized and mixed with

15

N-labeled cells, which serve as internal acyl-ACP

standards, prior to lysis and protein precipitation. Extending the approach used for detection of standard acyl-ACP, corresponding MRM transitions (Tables S-1 and S-2, Supporting Information) were tested using GluC digests of mixtures of

15

N-labeled and unlabeled E. coli. Chromatograms reveal that the majority of acyl-

ACP species can be identified (Figure 2), by simultaneous detection of co-eluting unlabeled and 15

N-labeled species. We routinely detect and quantify 21 species corresponding to predicted m/z

of holo-, malonyl-, acetyl-, and saturated, unsaturated and hydroxyl acyl intermediates with chain length between 4 and 16 carbons. In addition to the unsaturated intermediates generated during elongation of saturated fatty acids, E. coli produces unsaturated fatty acids for incorporation into the membrane by elongating trans-2-decenoyl-ACP (C10:1).27 Thus we expect peaks corresponding to two potential isomers for chain lengths of 10, 12, 14 and 16 carbon atoms. Indeed two peaks are observed for C10:1 and C12:1 (Figure S-4, Supporting Information), however full identification of cis and trans isomers is not possible without synthesis of additional


12

Page 13 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


standards or biological controls. We are unable to unambiguously detect β-keto intermediates, or hydroxy and di-unsaturated intermediates of the unsaturated branch of the pathway, likely due to their low abundance. However, we do often detect signals matching predicted m/z of these species (Figure S-4 and Table S-2, Supporting Information). The identifications of holo-, malonyl-, acetyl- (C2), butyryl- (C4), octanoyl- (C8), dodecanoyl(C12), myristoyl- (C14), and palmitoyl-ACP (C16) species were confirmed by spiking a mixture of in vitro standards into a sample of

15

N-labeled culture at the cell lysis step (Figure S-5,

Supporting Information). This distinguishes acyl-peptides originating from malonyl-ACP and 3hydroxy-butyryl-ACP (C4-OH), which appear identical within our mass resolution. We confirmed identification of acyl-ACP signature peptides without available in vitro standards using a triggered MRM method to acquire signal for 4 additional fragments: a2 and y1 ions of the peptide backbone, and three variants of phosphopantetheine ejection (Figure S-6, Supporting Information). By spiking acyl-ACP standards at different levels we obtained linear response curves (R2 > 0.94) for each acyl-ACP standard (Figure S-7, Supporting Information). Normalization using signals from isotopically-labeled internal standards proved to be essential for correcting for variations in protein precipitation, digestion, and quantification. We found that while variations in the ratio of protease to extracted protein affected quantification, normalization with internal standards corrected for this variation (Figure S-8, Supporting Information). Counts from unlabeled species are further normalized for variations in loading of unlabeled cells based on counts from the ACP peptide LVMALE, yielding relative counts for each ACP species. The relative standard deviations (RSD) for all major species after these normalizations are listed in Table S-1, with an overall average of 20%, well below that needed


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 26

for detecting significant changes in fatty acid synthesis intermediates. RSD is generally improved by ensuring complete solubilization of 15N-labeled cell pellets in the lysis buffer.

Figure 3. Slow quenching of fatty acid metabolism severely perturbs concentrations of several important ACP species. Aliquots of cells were withdrawn from a well-stirred culture and were either immediately quenched into TCA (“rapid quenching” executed within < 5 seconds), or placed on ice for 10 minutes before centrifugation, followed by either quenching with TCA (“slow quenching”) or freezing (“no quenching”). Depicted here are ratios from each quenching technique compared to in-flask quenching, in which TCA is added directly to a well-stirred culture. Error bars depict propagated error, calculated as the geometric mean of standard deviations in each measurement. ACP species were quantified in each condition with a minimum of 4 independent replicates. Rapid quenching of metabolism is essential for accurate characterization of intracellular acyl-ACP concentrations. Unlike protein concentrations, many intracellular metabolite


14

Page 15 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


concentrations respond quickly to perturbations due to fast turnover by enzymes (

Mass-Spectrometry-Based Quantification of Protein-Bound Fatty Acid

Recommend Documents