Patent Application
20250110135
The proteome is usually described as the entire complement of proteins found in a biological system, such as, e.g., a cell, tissue, organ or organism. Proteomics is the study of the proteome expressed at particular times and/or under internal or external conditions of interest. Proteomics approaches frequently aim at global analysis of the proteome, and require that large numbers of proteins, e.g., hundreds or thousands, can be routinely resolved, identified and quantified from a single sample.
Analyses of proteins from large cohort samples are typically restricted to high-throughput automated assays using antibody-based or aptamer-based technologies. These assays are restricted to panel targets and are therefore not unbiased. Untargeted mass spectrometry proteomics offers a solution to this problem. Mass spectrometry (MS) represents one of the main technologies for quantitative proteomics with advantages and disadvantages. Quantitative MS has higher sensitivity but can provide only limited information about the intact protein. Quantitative MS has been used for both discovery and targeted proteomic analysis to understand global proteomic dynamics in populations of cells (bulk analysis) or in individual cells (single-cell analysis).
However, mass spectrometric methods introduce a number of other challenges, such as low sample preparation throughput and low protein identification rates without substantial instrument time. Proteolysis of complex biological samples can produce thousands of peptides, which may overwhelm the resolution capacity of known chromatographic and mass spectrometric systems, causing incomplete separation and impaired identification of the constituent peptides. Furthermore, performing large-scale plasma proteome profiling is challenging due to limitations imposed by lengthy preparation and instrument time. Throughput, reproducibility, time, and cost remain longstanding barriers to the necessary large-scale MS sample processing. There is a need to develop an accurate, reproducible, cost-effective, and time-effective method, which allows for identification and quantification of the various different peptides, polypeptides, and proteins.
Disclosed herein are methods of quantifying or peptides or polypeptides, comprising the steps of: (a) combining the peptides or polypeptides into one plex; (b) splitting the plex of peptides or polypeptides into two or more portions; (c) re-combining the portions into one plex; and (d) analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each peptide or polypeptide.
Also disclosed herein are methods of quantifying or identifying two or more chemically isotopic labeled peptides or polypeptides, comprising the steps of: (a) combining the two or more peptides or polypeptides into one plex; (b) splitting the plex of labeled peptides or polypeptides into two or more portions; (c) re-combining the portions into one plex; and (d) analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each peptide or polypeptide.
Also disclosed herein are methods of quantifying or identifying peptides or polypeptides, comprising the steps of: (a) combining the peptides or polypeptides into one plex; (b) mixing the peptides or polypeptides in the plex; (c) splitting the plex of peptides or polypeptides into two or more portions; (d) re-combining the portions into one plex; and (e) analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each peptide or polypeptide.
Also disclosed herein are methods of quantifying or identifying two or more chemically isotopic labeled peptides or polypeptides, comprising the steps of: (a) combining the two or more labeled peptides or polypeptides into one plex; (b) mixing the two or more labeled peptides or polypeptides in the plex; (c) splitting the plex of labeled peptides or polypeptides into two or more portions; (d) re-combining the portions into one plex; and (e) analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each peptide or polypeptide.
In some embodiments, the peptides or polypeptides are labeled. In some embodiments, the peptides or polypeptides are chemically isotopic labeled.
In some embodiments, the method further comprises mixing two or more peptides or polypeptides in a single plex prior to splitting a plex of peptides or polypeptides into two or more portions.
In some embodiments, the method further comprises clean-up of each portion. In some embodiments, the method further comprises clean-up of each portion prior to splitting a plex of peptides or polypeptides into two or more portions. In some embodiments, the clean-up method is selected from the group consisting of peptide precipitation, in-solution (IS), in-StageTip (IST), Single-Pot Solid-phase enhanced Sample Preparation (SP3), filter-aided sample preparation (FASP), S-Trap, or SepPak.
In some embodiments, the method further comprises normalizing the concentration values of the peptides or polypeptide to a reference sample. In some embodiments, the normalizing is prior to labeling the peptides or polypeptides. In some embodiments, the method further comprises normalizing the concentration values of the peptides or polypeptides to a reference sample prior combining the peptides or polypeptides into one plex.
In some embodiments, the method further comprises generating a bridge.
In some embodiments, the analyzing comprises protein identification using high-field asymmetric waveform ion mobility (FAIMS Pro) and Real Time Search (RTS).
In some embodiments, the chemically isotopic labeled peptides or polypeptides are isobarically labeled or otherwise isotope incorporated.
In some embodiments, the method is performed on an automated liquid-handling robot.
Also disclosed herein are methods for quantifying or identifying two or more chemically isotopic labeled peptides or polypeptides, comprising the steps of: (a) Combining the two or more labeled peptides or polypeptides into one plex; (b) Mixing the two or more labeled peptides or polypeptides in the plex; (c) Splitting the plex of labeled peptides or polypeptides into two or more portions; (d) Combining the portions into one plex; (e) Analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each chemically isotopic labeled peptide or polypeptide.
In some embodiments, the method further comprises clean-up of each portion prior to splitting the plex of labeled peptides or polypeptides into two or more portions, wherein the clean-up method is selected from the group consisting of peptide precipitation, in-solution (IS), in-StageTip (IST), Single-Pot Solid-phase enhanced Sample Preparation (SP3), filter-aided sample prepatation (FASP), S-Trap, or SepPak.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptide to a reference sample prior to labeling the peptides or polypeptides.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptides to a reference sample prior to combining the two or more labeled peptides or polypeptides into one plex.
In some embodiments, the method further comprises generating a bridge.
In some embodiments, the method further comprises correlating the mass spectrometry data to each peptide or polypeptide.
In some embodiments, the chemically isotopic labeled peptides or polypeptides are isobarically labeled or otherwise isotope-incorporated.
In some embodiments, the method is performed on an automated liquid-handling robot.
Also disclosed herein are methods for obtaining a plurality of enriched peptides or polypeptides in a sample, comprising the steps of: (a) Contacting the sample with beads comprised of a single type of bead; (b) Separating a portion of the sample comprising protein-bound beads from a portion of the sample comprising unbound proteins; and (c) Resuspending the portion of the sample comprising protein-bound beads, thereby obtaining a plurality of enriched peptides or polypeptides in the sample.
In some embodiments, the sample comprises whole blood, plasma, serum, urine, saliva, tears, spinal fluid, synovial fluid, cell lysate, tissue lysate, exosomes, individual cell organelles, or any combination thereof. In some embodiments, the sample comprises plasma.
In some embodiments, the beads are magnetic. In some embodiments, the size of the beads is about 0.1 μm to 10 μm in diameter or 1 nm to 1000 μm in diameter.
In some embodiments, the method further comprises incubating the beads with the sample at physiological conditions. In some embodiments, the method is performed under physiological conditions. In some embodiments, the method is performed at a pH of about 7.4. In some embodiments, the method is performed at a temperature of about 37 degrees Celsius.
In some embodiments, the plurality of enriched peptides or polypeptides comprises one or more protein complexes.
In some embodiments, the method comprises magnetic immobilization to separate a portion of protein-bound beads from a portion of unbound plasma proteins and unbound beads in the sample.
In some embodiments, the method further comprises reducing the plurality of enriched peptides or polypeptides. In some embodiments, the method further comprises alkylating the plurality of enriched peptides or polypeptides. In some embodiments, the method further comprises derivatization of cysteines in the plurality of enriched peptides or polypeptides. In some embodiments, the method further comprises clean-up of the plurality of enriched peptides or polypeptides.
In some embodiments, the method comprises resuspending which includes digesting.
In some embodiments, the method further comprises digesting the plurality of enriched peptides or polypeptides. In some embodiments, the digesting comprises adding protease to the plurality of enriched peptides or polypeptides.
In some embodiments, the method further comprises labeling the plurality of enriched peptides or polypeptides. In some embodiments, the method further comprises isobarically labeling the plurality of enriched peptides or polypeptides. In some embodiments, the labeling comprises incorporating chemically isotopic labels onto each peptide or polypeptide.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptide to a reference sample prior to labeling the peptides or polypeptides.
Also disclosed herein are methods for quantifying or identifying peptides or polypeptides, comprising the steps of: (a) Obtaining a plurality of labeled peptides or polypeptides according to a method disclosed herein; (b) Aliquoting the labeled peptides or polypeptides into two or more portions; (c) Combining the two or more labeled portions into one plex; (d) Splitting the plex of labeled peptides or polypeptides into two or more portions for clean-up; (e) Re-combining the portions post clean-up; and (f) Analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying peptide or polypeptide.
In some embodiments, the method further comprises mixing the two or more portions in the single plex prior to the splitting. In some embodiments, the method further comprises clean-up of each portion prior to re-combining.
In some embodiments, the method further comprises generating a bridge.
In some embodiments, the method is performed on an automated liquid-handling robot.
Also disclosed herein are methods for quantifying or identifying peptides or polypeptides in a sample, comprising the steps of: (a) Contacting the sample with beads comprising a single type of bead; (b) Separating a portion of the sample comprising protein-bound beads from a portion of the sample comprising unbound proteins; (c) Resuspending the portion of the sample comprising protein-bound beads; (d) Incorporating isotopic labels onto each peptide or polypeptide in the portion of the sample comprising protein-bound beads; (e) Aliquoting the labeled peptides or polypeptides into two or more portions; (f) Combining the two or more isotopically labeled samples into one plex; (g) Splitting the plex of labeled peptides or polypeptides into two or more portions for clean-up; (h) Re-combining the portions into a single portion; and (i) Analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying peptides or polypeptides in a sample.
In some embodiments, the method further comprises alkylating the portion of the sample comprising protein-bound beads. In some embodiments, the method further comprises clean-up of the portion of the sample comprising protein-bound beads. In some embodiments, the method further comprising the clean-up is prior to re-combining the portions into a single portion.
In some embodiments, the resuspending includes digesting. In some embodiments, the method further comprises digesting the peptides or polypeptides. In some embodiments, the digesting comprises adding protease to the plurality of enriched peptides or polypeptides.
In some embodiments, the method further comprises generating a bridge.
In some embodiments, the method is performed on an automated liquid-handling robot.
Also disclosed herein is a method for quantifying or identifying peptides or polypeptides in a sample, comprising the steps of: (a) Contacting the sample with a single type of bead; (b) Separating a portion of the sample comprising protein-bound beads from a portion of the sample comprising unbound proteins; (c) Resuspending the portion of the sample comprising protein-bound beads; and (d) Analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying peptides or polypeptides in a sample.
In some embodiments, the method further comprises alkylating the portion of the sample comprising protein-bound beads. In some embodiments, the method further comprises clean-up of the portion of the sample comprising protein-bound beads.
In some embodiments, the clean-up is prior to re-combining the portions into a single portion. In some embodiments, the resuspending includes digesting. In some embodiments, the method further comprises digesting the peptides or polypeptides. In some embodiments, the digesting comprises adding protease to the plurality of enriched peptides or polypeptides.
In some embodiments, the method further comprises generating a bridge.
In some embodiment, the method is performed on an automated liquid-handling robot.
The features and other details of the disclosure will now be more particularly described. Before further description of the present invention, certain terms employed in the specification, examples and appended claims are collected here. These definitions should be read in light of the remainder of the disclosure and understood as by a person of skill in the art. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by a person of ordinary skill in the art.
This description is not intended to be a detailed catalog of all the different ways in which the invention may be implemented, or all the features that may be added to the instant invention. For example, features illustrated with respect to one embodiment may be incorporated into other embodiments, and features illustrated with respect to a particular embodiment may be deleted from that embodiment. In addition, numerous variations and additions to the various embodiments suggested herein will be apparent to those skilled in the art in light of the instant disclosure which do not depart from the instant invention. Hence, the following specification is intended to illustrate particular embodiments of the invention, and not to exhaustively specify all permutations, combination and variations thereof.
Unless the context indicates otherwise, it is specifically intended that the various features described herein can be used in any combination. Moreover in some embodiments, any feature or combination of features set forth herein can be excluded or omitted. To illustrate, if the specification states that a complex comprises components A, B, and C, it is specifically intended that any of A, B, or C, or a combination thereof, can be omitted and disclaimed singularly or in any combination.
As used in the present specification, the following words and phrases are generally intended to have the meanings as set forth below, except to the extent that the context in which they are used indicates otherwise.
It is to be noted that as used herein and in the claims, the singular forms “a,” “an” and “the” include plural referents unless the context clearly dictates otherwise.
As used herein, “and/or” refer to and encompasses any and all possible combinations of one or more of the associated listed items, as well as the lack of combinations when interpreted in the alternative (“or”). Moreover, any feature or combination of features set forth herein can be excluded or omitted.
As used herein, the terms “comprises,” “comprising,” “includes,” “including,” “has,” “having,” “is,” “are” or any other variation thereof, are intended to cover a non-exclusive inclusion. They are to be interpreted synonymously with the phrases “having at least” or “including at least”. The term “consisting of” refers to including, and being limited to, whatever follows the phrase “consisting of.”
As used herein, the term “comprising” also specifically includes embodiments “consisting of” and “consisting essentially of” the recited elements, unless specifically indicated otherwise. Similarly, the term “consisting essentially of” is intended to include embodiments encompassed by the term “consisting of”.
The term “about” as used herein when referring to a measurable value such as an amount of a compound or agent, dose, time, temperature, and the like, is meant to encompass variations of ±10%, ±5%, ±1%, ±0.5%, or even ±0.1% of the specified amount.
Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description herein is for the purpose of describing particular embodiments only and is not intended to be limiting.
As used herein, the term “sample” refers to mean any fluid, cell, tissue, organ or a portion thereof. It also includes, but is not limited to, whole blood, plasma, serum, urine, saliva, tears, spinal fluid, synovial fluid, cell lysate, tissue lysate, exosomes, individual cell organelles, or any combination thereof. The sample can be from any organism. It includes, but is not limited to human, mouse, yeast, worm, fish, bacteria, etc.
The term “peptide” refers to a short polymer of amino acids linked by peptide bonds. It has the same chemical (peptide) bonds as proteins but is commonly shorter in length. The shortest peptide is a “dipetide” consisting of two amino acids joined by a peptide bond. There can also be tripeptides, tetrapeptides, pentapeptides, etc. A peptide has an amino end and a carboxyl end, unless it is a cyclic peptide.
The term “polypeptide” refers to a single linear chain of amino acids bonded together by peptide bonds and preferably comprises at least five amino acids. A polypeptide can be one chain or may be composed of more than one chains, which are held together by covalent bonds, e.g. disulphide bonds and/or non-covalent bonds. Polypeptides that can be purified with the method of the invention preferably have a length of at least five amino acids, more preferably a length of at least 10 amino acids, 15 amino acids, 20 amino acids, 25 amino acids, 30 amino acids, 35 amino acids, 40 amino acids, 45 amino acids or 50 amino acids or longer. In some embodiments, the peptides or polypeptides are suspended in a liquid solution. Examples of liquid solution include water, aqueous buffer mixtures, acidic or basic solutions, organic solvents such as alcohol or acetonitrile, or any combination thereof.
The term “labeling” refers to a method of attaching a detectable moiety to polypeptides or introducing a detectably moiety into the polypeptides. Preferred are labels, markers or tags which provide quantification of labeled polypeptides in mass spectrometry analysis. The method of stable isotope labeling entails replacing specific atoms of the polypeptides, e.g. C, O, N, or S, with their isotopes. This includes but is not limited to methods like stable isotope labeling by amino acids in cell culture (SILAC), trypsin-catalyzed 180 labeling, isotope coded affinity tagging (ICAT), isobaric tags for relative and absolute quantitation (iTRAQ). One common method in the field of stable isotope labeling is dimethyl labeling comprising a reagent for the generation of a Schiff base with a primary amine, a reducing agent and a suitable buffer. Another widespread method is called isobaric labeling using tandem mass tag (TMT) labeling comprising a label reagent, a suitable buffer, an organic solvent, a denaturating reagent, a reducing reagent, an alkylating reagent, a quenching reagent and a protease.
The terms “splitting,” “separating,” or “portioning” and the like are used interchangeably herein to generally mean splitting one sample into a plurality of samples. The plurality of samples may have substantially equal volumes, the same volumes, or unequal volumes.
The term “plex” comprises isobaric labelsed peptides or polypeptides, isotopic labeled peptides or polypeptides, standard peptides or polypeptides, such as standard spike-in peptides or polypeptides, AQUA, PTM modified, non-PTM modified, noncanonical amino acids, unmodified peptides or polypeptides, unlabeled peptides or polypeptides, or any combination thereof.
The terms “enriched proteins,” “enriched polypeptides,” and “enriched peptides” refer to proteins, polypeptides, or peptides, respectively, that preferentially bind to and/or interact with a bead surface or a surface.
The term “physiological conditions” or “native” refers to conditions of the external or internal milieu that may occur in nature for an organism, cell, or cell system. Non-limiting examples of physiological conditions include a temperature range of 20-40 degrees Celsius, pH of 6-8, glucose concentration of 1-60 mM, or calcium concentration of 0.5-2 mM.
The term “non-physiological conditions” or “non-native” refers to conditions of the external or internal milieu that are in a higher or lower range than the conditions that may occur in nature for an organism, cell, or cell system. Non-limiting examples of non-physiological conditions include a temperature lower than 20 degrees Celsius, temperature higher than 40 degrees Celsius, pH less than 6, pH greater than 8, glucose concentration of less than 1 mM, glucose concentration greater than 60 mM, calcium concentration less than 0.5 mM, or calcium concentration greater than 2 mM.
Large-scale plasma proteome profiling typically requires a large time investment both in terms of sample preparation time and instrumental analysis. Non-limiting examples of samples includes whole blood, plasma, blood serum, red blood cells, white blood cells/PBMCs, cell lysate, tissue lysate, exosomes, samples enriched or sorted for specific cell types such as B cells, samples enriched for cell components such as organelles, samples enriched for PTMs (post-translational modifications), samples depleted or certain components such as high abundant proteins, samples enriched for certain proteins for example using immunoprecipitation.
The concentration of plasma proteins spans a dynamic range of at least 12 orders of magnitude, which results in poor protein identification depth. To combat the low number of proteins identified in plasma, the field has resorted to either extensive fractionation, or depletion of the most abundant proteins. Mass spectrometric analysis times with extensive fractions are expensive and are associated with a low rate of protein identification. While plasma depletion greatly increases the number of identifiable proteins, it is not amenable to automation, can lead to biases due to antibody cross-reactivity, and immunoprecipitation of antigen-bound proteins takes at least four hours to process. Furthermore, plasma depletion is costly and leads to increased variability.
Recent studies using multiplexed mass spectrometry-based proteomics analysis of human plasma quantified 652 proteins in depleted plasma in five hours of instrument time (10-plex quantification of approximately 131 proteins per hour) (Keshishian H, et al. 2017 Nat Protoc 12 (8): 1683-1701). From the same lab, depleted plasma combined with fractionation achieved 3,390 proteins with 4-plex quantitation (32 proteins per hour) (Keshishian H, et al. 2015 Mol Cell Proteomics 14 (9): 2375-2393). A recent study using undepleted plasma with 11-plex quantification, coupled to deep fractionation (180 fractions), achieved 4,826 proteins in approximately one month of instrument time (approximately 7 proteins per hour) (Dey K K, et al. 2019 Clinical Proteomics: 1-12). An alternative method is to use non-multiplexed quantitative proteomics (e.g. label-free or data-independent acquisition) with the largest undepleted dataset identifying 1,725 proteins at the time of this writing (Albrechtsen N J W, et al. 2018 Cell Systems 7 (6): 601-612.e3). However, non-multiplexed analyses require extensive instrument time and are sensitive to variations in liquid chromatography-mass spectrometry (LC-MS) performance (O'Connell J D, Paulo J A, O'Brien J J, Gygi S P 2018. J Proteome Res 17 (5): 1934-1942).
Nonlimiting examples of samples include any fluid, cell, tissue, organ, a portion thereof, or any combination thereof. Examples also include, but are not limited to, whole blood, plasma, serum, urine, saliva, tears, spinal fluid, synovial fluid, cell lysate, tissue lysate, exosomes, individual cell organelles, or any combination thereof. The sample can be from any organism. Examples include, but are not limited to, human, mouse, yeast, worm, or fish. In some embodiments, the methods disclosed herein utilize a sample volume of about 0.5 μL-5000 μL. In some embodiments, the methods disclosed herein utilize a sample volume of about 50-250 μL. In some embodiments, the methods disclosed herein utilize a sample volume of about 250 μL.
Disclosed herein is an automated platform (also referred to as “AutoMP3”) for high throughput proteome profiling. The platform was designed considering and optimizing sample preparation, ease of automation, 96-well plate compatibility, level of multiplexing, LC column robustness and selection, LC-MS method acquisition parameters, fractionation, and how resulting outputs would feed into a downstream data analysis pipeline supporting statistical analysis and visualization of a large number of samples.
Most studies investigating the plasma proteome either use extensive fractionation for multiplexed samples, consequently requiring long instrument run times (1,025 proteins×1-plex/0.75 hr run/60 min=˜22 proteins/min/sample), or single shot runs using label-free quantification (652 proteins×10-plex/5 hr run/60 min=˜23 proteins/min/sample). A higher rate of proteins quantified in a short amount of time is needed for large scale biomarker studies that require the analysis of thousands of plasma samples. The AutoMP3 platform is able to quantify ˜44 proteins/min/sample (489 proteins×16-plex/3 hr run/60 min), an increase of approximately 2× over other known methods. The use of TMTpro combined with FAIMS Pro and RTS on an Eclipse mass spectrometer increased the rate of protein identification.
An increase in the number of quantified proteins in plasma by around 50% was achieved using the AutoMP3 Platform. This improvement was made possible largely by the use of FAIMS Pro and RTS. The nature of DDA methods means that different peptides are selected for sequencing between plexes, which inevitably decreases overlap and leads to missing values. To maximize overlap between plexes, an inclusion list method approach was investigated. For large numbers of plexes (>15), using an inclusion list method on its own is likely the preferred method to achieve the largest peptide overlap compared to DDA. However, for smaller numbers of plexes, the regular DDA method, in our setup, resulted in the largest coverage and peptide overlap between plexes. Another advantage of the inclusion list approach is the ability to prioritize peptides of interest. This is particularly attractive for low abundance PTM peptides that would not otherwise be selected for analysis.
The utility of automated scale fractionation was also evaluated. A potentially worthwhile increase in peptide and protein identifications was observed, which needs to be weighed against longer instrument time, especially for large scale experiments. Fractionation was easily implemented with the AutoMP3 platform using a modified version of the peptide cleanup protocol, where instead of eluting peptides with a single buffer, three (or more) elution buffers could be applied (for example, a gradient of high to low concentration of acetonitrile) and each transferred to separate 96-well plates. Fractionating using AutoMP3 in the 96-well format allowed for higher throughput than our standard high-pressure reverse phase (HPRP) fractionation setup. As with other steps, fractionation was designed in a modular fashion. The effect of fractionating was tested on label-free mouse plasma samples into three parts, using various acetonitrile-based elution gradients. Fractionated sets were compared against three single shot injections of an unfractionated sample (
To minimize instrument acquisition time and the associated costs, sample multiplexing and throughput was maximized using TMTpro and single shot analyses. The main barrier to performing large-scale studies using TMT is sample preparation, since the labeling chemistry is very time consuming to perform manually. This challenge may be addressed by creating a magnetic bead-based, TMT-compatible/isobaric labeling sample preparation workflow.
In some embodiments, the present disclosure provides methods comprising quantitating the glycosylated peptide fragments by using a mass spectrometer (MS). In some embodiments, the methods employ liquid chromatography (LC).
In some embodiments, the method comprises further fractionation prior to analyzing. In some embodiments, the method comprises further splitting of the peptides or polypeptides prior to analyzing. In some embodiments, the method comprises further splitting of the combination of peptides or polypeptides prior to analyzing.
Disclosed herein, in certain embodiments are methods for quantifying or identifying two or more chemically isotopic labeled peptides or polypeptides, comprising the steps of: (a) combining the two or more labeled peptides or polypeptides into one plex; (b) splitting the plex of labeled peptides or polypeptides into two or more portions; (c) combining the portions into one plex; and (d) analyzing the peptide or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each chemically isotopic labeled peptide or polypeptide.
Disclosed herein, in certain embodiments are methods for quantifying or identifying two or more chemically isotopic labeled peptides or polypeptides, comprising the steps of: (a) combining the two or more labeled peptides or polypeptides into one plex; (b) mixing the two or more labeled peptides or polypeptides in the plex; (c) splitting the plex of labeled peptides or polypeptides into two or more portions; (d) combining the portions into one plex; and (e) analyzing the peptides or polypeptides so as to obtain mass spectrometry data, thereby quantifying or identifying each chemically isotopic labeled peptide or polypeptide.
In some embodiments, the method further comprises mixing the two or more labeled peptides or polypeptides in the single plex prior to splitting.
In some embodiments, the method further comprises clean-up of each portion of sample or plex of samples prior to combining. In some embodiment, the clean-up is selected from the group consisting of peptide precipitation, in-solution (IS), in-StageTip (iST), Single-Pot Solid-phase enhanced Sample Preparation (SP3), filter-aided sample prepatation (FASP), STRAP, or SEP PAK.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptide to a reference sample. In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptide to a reference sample prior to labeling the peptides or polypeptides.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptides to a reference sample prior to combining the two or more labeled peptides or polypeptide. In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptides to a reference sample prior to combining the two or more peptides or polypeptide.
In some embodiments, the method further comprises generating a bridge.
In some embodiments, the analyzing comprises protein identification. In some embodiments, the protein identification is using high-field asymmetric waveform ion mobility (FAIMS Pro) and Real Time Search (RTS).
In some embodiments, the chemically isotopic labeled peptides or polypeptides are isobarically labeled or otherwise isotope-incorporated.
In some embodiments, the method is performed on an automated device. In some embodiments, the method is performed on a liquid-handling robot.
Minimizing instrument time, while maximizing the number of proteins quantified per sample, is critical to allow large cohort analysis at a reasonable cost. When using label-free mas spectrometry type approaches, the instrument time to sample ratio increases at a much greater rate than for multiplexed isobaric tag methods, such as TMT 10plex and TMTpro 16-plex (
For example, analyzing 1,000 samples would require three months of continuous LC-MS time using label-free, whereas TMT multiplexing would require 19 days and would require 12 days (
Disclosed herein are methods of automated sample preparation solution for TMTpro. As depicted in
Labeling samples with stable isotope labels allows a mass spectrometer to distinguish between identical proteins in separate samples. Protein quantitation is accomplished by comparing the intensities of reporter ions in the MS/MS spectra. A key benefit of isobaric labeling over other quantification techniques (e.g. SILAC, ICAT, Label-free) is the increased multiplex capabilities and thus increased throughput potential. The ability to combine and analyze several samples simultaneously in one LC-MS run eliminates the need to analyze multiple data sets and eliminates run-to-run variation. Multiplexing reduces sample processing variability, improves specificity by quantifying the proteins from each condition simultaneously, and reduces turnaround time for multiple samples. Isobaric chemical tags can facilitate the simultaneous analysis of multiple samples.
In some embodiments, simultaneous analysis can be conducted for up to 100 samples, including but not limited to 2 samples, 6 sampes, 10 samples, 11 samples, 16 samples, and 18 samples.
In some embodiments, the peptides or polypeptides are isotopically labeled. In some embodiments, protein or peptide samples prepared from cells, tissues or biological fluids are labeled in parallel with the isobaric mass tags and combined for analysis.
Non-limiting examples of labels for relative quantification methods include isobaric labeling (tandem mass tags or TMT), isobaric tags for relative and absolute quantification (iTRAQ), label-free quantification metal-coded tags (MeCAT), N-terminal labeling, stable isotope labeling with amino acids in cell culture (SILAC), terminal amine isotopic labeling of substrates (TAILS), and Isotope-coded affinity tag (ICAT).
Metal-coded tags (MeCAT) method is based on chemical labeling, but rather than using stable isotopes, different lanthanide ions in macrocyclic complexes are used. Stable isotope labeling with amino acids in cell culture (SILAC) is a method that involves metabolic incorporation of “heavy” C- or N-labeled amino acids into proteins followed by MS analysis. Traditionally the level of multiplexing in SILAC was limited due to the number of SILAC isotopes available.
Isobaric mass tags (tandem mass tags or TMT) are tags that have identical mass and chemical properties that allow heavy and light isotopologues to co-elute together. These tags are designed to cleave at a specific linker region upon high-energy CID, yielding different-sized tags that are then quantitated by LC-MS/MS.
In some embodiments, the two or more peptides or polypeptides, comprise one or more chemically labeled peptides or polypeptides. In some embodiments, the two or more peptides or polypeptides, comprise one or more unlabeled peptides or polypeptides. In some embodiments, the two or more peptides or polypeptides, comprise one or more labeled peptides or polypeptides and/or one or more unlabeled peptides or polypeptides. In some embodiments, the two or more peptides or polypeptides, comprise one or more chemically labeled peptides or polypeptides and/or one or more unlabeled peptides or polypeptides. In some embodiments, the two or more peptides or polypeptides, comprise one or more chemically labeled peptides or polypeptides and/or one or more unlabeled PTM peptides or polypeptides. In some embodiments, the methods comprise two or more labeled peptides or polypeptides, wherein the labels are not the same. In some embodiments, the methods comprise two or more labeled peptides or polypeptides, wherein the labels are the same.
Magnetic bead type workflows are ideal for automation because they are magnetic and bead quantity is easily adjusted based on the starting amount. This is in contrast to solid phase systems, which are limited to cartridges with a set capacity.
In some embodiments, the beads are comprised of a single type of bead. In some embodiments, the beads are magnetic. In some embodiments, the beads are not magnetic.
In some embodiments, the size of a bead is less than 1 micron in diameter. In some embodiments, the size of a bead is about 1-10,000 nanometers in diameter. In some embodiments, the size of a bead is 0.1-1,000 micrometers in diameter. Suitable examples of ranges of a bead diameter include, but are not limited to, for example, a bead from about 5 nm to about 1000 nm in diameter, about 5 nm to about 500 nm, about 10 nm to about 500 nm, about 50 nm to about 500 nm, alternatively about 50 nm to about 250 nm, alternatively about 50 nm to about 200 nm, alternatively about 50 nm to about 100 nm, and any combinations, ranges or amount in between (e.g., 10 nm, 20 nm, 30 nm, 40 nm, 50 nm, 60 nm, 70 nm, 80 nm, 90 nm, 100 nm, 150 nm, 200 nm, 250 nm, 300 nm, 350 nm, 400 nm, 450 nm, 500 nm, 550 nm, 600 nm, 650 nm, 700 nm, 750 nm, 800 nm, 850 nm, 900 nm, 1000 nm, etc.).
In some embodiments, the beads are, for example, nanoparticles, microparticles, protein-based particles, or any combinations thereof. In some embodiments, the beads have a solid core. In some embodiments, the beads have a porous core. In some embodiments, the beads surface is functionalized. Examples of functional group include but are not limited to hydrophobic surfaces (e.g. C18) with carboxylate groups to provide hydrophilicity.
Non-limiting examples of beads include 1 μm average diameter hydrophobic surface carboxylate SpeedBead SeraMag beads (Cytiva, #44152105050350), 0.5 μm, carboxylate SpeedBead SeraMag beads, solid core, 5 μm, amine, porous core, 5 μm, carboxylated, porous core, 5 um, HILIC, porous core, 5 μm, hydroxyl, porous core, 1 μm, carboxylated, solid core, or a combination thereof.
In some embodiments, a method disclosed herein comprises hydrophobic SeraMag SpeedBeads (Cytiva, #44152105050350). In some embodiments, a method disclosed herein comprises hydrophilic SeraMag SpeedBeads (Cytiva, #45152105050350). In some embodiments, a method disclosed herein comprises a mixture of hydrophobic and hydrophilic SeraMag SpeedBeads (Cytiva, #44152105050350 & #45152105050350). In some embodiments, a mixture of hydrophobic and hydrophilic SeraMag SpeedBeads is in a 1:1 ratio.
Disclosed herein are methods of automated sample preparation. Automated sample preparation allows a 96-well plate automation solution, as well as a high degree of flexibility for future improvements (e.g., multiplexing reagents).
Multiple automated sample preparation platforms for label-free proteomics are known in the art. A non-limiting example of a commercially available robot that can handle chemically isotopic labeling includes the PreON workflow (Preomics, Planegg-Martinsried, Germany). However, the PreON only allows multiplexing of up to 11 samples simultaneously. Additionally, the PreON is not compatible with TMTpro, and it cannot multiple plexes because it can only process 12 samples in one round. The PreON system is capable of processing up to 36 samples per day, which then allows for three 11-plexes (10 samples+bridge). However, this solution is not compatible with 16-plex automation, does not include automated TMT label ratio check sample generation, does not include adjustments of samples with different protein concentrations (normalization), does not include bridge generation and it is not compatible with a 96-well plate format design.
In some embodiments, the method relates to an automated system with centralized control for performing proteomics research and sample preparation. In some embodiments, the automated system is a liquid-handling. In some embodiments, the automated system is a pipetting robot. In some embodiments, the automated system is a microfluidic device. In some embodiments, a method disclosed herein is performed on an automated device.
In some embodiments, a method of quantifying or identifying is performed on an automated device. In some embodiments, a method of quantifying or identifying is performed on a liquid-handling robot. Examples of liquid handler platforms include but are not limited to Hamilton Vantage™, Tecan, Agilent Bravo, Opentrons OT-2, or epMotion 5073m. An example of a microfluidic device includes but is not limited to a digital microfluidic device (Sci-bots, DropBot DB3-120). In some embodiments, MP3 sample preparation may be automated on a Hamilton Vantage™ liquid handler.
While the most difficult and important steps in automating the manual protocol were the protein cleanup and TMT labeling, there were additional steps that were also critical to automate when scaling up from a single 16-plex to dozens and even hundreds of plexes (
In addition to automating a pooled bridge design for combining data between plexes, an in-house software platform was developed for analyzing time courses based on previously described protein quantification models. This allows for processing large numbers of plexes together and fit linear, quadratic, or cubic time effects, while using the within-sample variability to weight each observation. This weighting prevents highly variable measurements from unduly influencing estimated trajectories.
In some embodiments, a method disclosed herein further comprises normalizing concentration values of the two or more peptides or polypeptides to a reference sample prior to labeling the peptides or polypeptides. In some embodiments, the normalizing is by BCA.
In some embodiments, the method further comprises normalizing the concentration values of the two or more peptides or polypeptides to a reference sample prior to step (a). In some embodiments, the normalizing is by ratio check. In some embodiments, the ratio check normalizes peptide content in each channel to ensure equal total amounts of total peptide/protein in each sample using mass spectrometry readout of peptide concentration.
A challenge in a TMT workflow arises when the number of experimental samples exceeds the number of TMT multiplex channels. Typically, this challenge is mitigated with a bridge channel that contains a mixture of all samples. Each individual experimental channel can be normalized to the bridge to enable combining multiple plexes together. To construct the bridge, a small portion of each sample at the peptide level may be combined prior to labeling.
In some embodiments, a method disclosed herein further comprises generating a bridge. In some embodiments, a bridge comprises a small portion of each sample. In some embodiments, a bridge is constructed by taking a portion of each sample and combining each portion into one. In some embodiments, a bridge comprises taking a portion of select samples and combining each portion into one. In some embodiments, a bridge comprises a mixture of synthetic peptides, a select set of samples, or a portion of each sample. In some embodiments, a bridge is constructed prior to labeling. In some embodiments, a bridge is generated prior to labeling the peptides or polypeptides.
A ratio check sample is generated by taking a portion of each samples (e.g. 0.1-20%), combined, and then performing clean-up prior to mass spectrometry analysis to obtain an estimate of peptide/protein abundance of each of the samples. A ratio check step allows for correction to ensure equal peptide/protein abundance of each of the samples. Non-limiting examples of methods of protein/peptide quantification includes colorimetric, fluorescent based assays such as the BCA or Bradford assay, and use of a ratio check sample to ensure substantially equal protein concentrations between samples.
In some embodiments, a method disclosed herein further comprises a ratio check. In some embodiments, a ratio check comprises normalizing peptide content in each channel to ensure equal peptide/protein concentrations in each sample prior to combining the samples for analysis by mass spectrometry.
In some embodiments, a method disclosed herein further comprises clean-up. In some embodiments, the clean-up method is selected from the group consisting of peptide precipitation, in-solution (IS), in-StageTip (IST), Single-Pot Solid-phase enhanced Sample Preparation (SP3), filter-aided sample preparation (FASP), S-TRAP, or C18 column based clean-up such as SepPak (Waters). In some embodiments, clean-up is by de-salting. In some embodiments, clean-up is by detergent removal. In some embodiments, clean-up is by lipid or metabolite removal. In some embodiments, clean-up is by small molecule removal.
In some embodiments, a method disclosed herein further comprises a peptide cleanup protocol. In some embodiments, a peptide cleanup protocol comprises eluting peptides with a single buffer. In some embodiments, a peptide cleanup protocol comprises more than one elution buffer. In some embodiments, a peptide cleanup protocol comprises three or more elution buffers. In some embodiments, in a peptide cleanup protocol comprising more than one elution buffer, a gradient of high to low concentration can be created. In some embodiments, a gradient of high to low concentration is of acetonitrile or other oganic solvent. In some embodiments, a gradient of low to high acetonitrile or other organic solvent. In some embodiments, a peptide cleanup protocol comprises transferring each elution to separate 96-well plates or single vessel consumables, or 384-well plates, or any number in between/higher etc.
In some embodiments, a peptide cleanup protocol is prior to combining the two or more labeled peptides or polypeptides. In some embodiments, a peptide cleanup protocol is prior to combining two or more portions into one plex. In some embodiments, a peptide cleanup protocol is prior to MS data acquisition analysis.
In some embodiments, a peptide cleanup protocol comprises a gradient of high to low organic solvent. In some embodiments, a peptide cleanup protocol comprises a gradient of low to high organic solvent. Non-limiting examples of organic solvents suitable for use in a peptide cleanup protocol include acetonitrile (ACN), methanol (MeOH), ethanol (EtOH), isopropyl alcohol (IPA), and combinations thereof.
In some embodiments, a peptide cleanup protocol comprises an organic solvent for about 2.5 hrs, with increasing time up to about 7.5 hrs. In some embodiments, a peptide cleanup protocol comprises an organic solvent for about 1.0 hrs, for about 1.5 hrs, for about 2.0 hrs, for about 2.5 hrs, for about 3.0 hrs, for about 3.5 hrs, for about 4.0 hrs, for about 4.5 hrs, for about 5.0 hrs, for about 5.5 hrs, for about 6.0 hrs, for about 6.5 hrs, for about 7.0 hrs, or for about 7.5 hrs.
In some embodiments, a peptide cleanup protocol disclosed herein alters pH.
In some embodiments, a peptide cleanup protocol comprises a gradient of high to low ethanol (EtOH). In some embodiments, a peptide cleanup protocol comprises a gradient of low to high EtOH. In some embodiments, a peptide cleanup protocol comprises a 95% final concentration EtOH. In some embodiments, a peptide cleanup protocol comprises a 80% final concentration EtOH. In some embodiments, a peptide cleanup protocol comprises the addition of an amount of 100% EtOH to reach a final concentration of 95% EtOH. In some embodiments, a peptide cleanup protocol comprises the addition of an amount of 100% EtOH to reach a final concentration of about 80% EtOH, about 85% EtOH, about 90% EtOH, or about 95% EtOH.
In some embodiments, a peptide cleanup protocol comprises a 95% final concentration EtOH for about 2.5 hrs, with increasing time up to about 7.5 hrs. In some embodiments, a peptide cleanup protocol comprises a 95% final concentration EtOH for about 1.0 hrs, for about 1.5 hrs, for about 2.0 hrs, for about 2.5 hrs, for about 3.0 hrs, for about 3.5 hrs, for about 4.0 hrs, for about 4.5 hrs, for about 5.0 hrs, for about 5.5 hrs, for about 6.0 hrs, for about 6.5 hrs, for about 7.0 hrs, or for about 7.5 hrs. In some embodiment, once all plates were started with an incubation step, wash steps began.
In some embodiments, a peptide cleanup protocol comprises a gradient of high to low methanol (MeOH). In some embodiments, a peptide cleanup protocol comprises a gradient of low to high MeOH. In some embodiments, a peptide cleanup protocol comprises a 95% final concentration MeOH. In some embodiments, a peptide cleanup protocol comprises a 80% final concentration MeOH. In some embodiments, a peptide cleanup protocol comprises the addition of an amount of 100% MeOH to reach a final concentration of about 80% MeOH, about 85% MeOH, about 90% MeOH, or about 95% MeOH.
In some embodiments, a peptide cleanup protocol comprises a gradient of high to low isopropyl alcohol (IPA). In some embodiments, a peptide cleanup protocol comprises a gradient of low to high IPA. In some embodiments, a peptide cleanup protocol comprises a 95% final concentration IPA. In some embodiments, a peptide cleanup protocol comprises a 80% final concentration IPA. In some embodiments, a peptide cleanup protocol comprises the addition of an amount of 100% IPA to reach a final concentration of 95% IPA. In some embodiments, a peptide cleanup protocol comprises the addition of an amount of 100% IPA to reach a final concentration of about 80% IPA, about 85% IPA, about 90% IPA, or about 95% IPA.
In some embodiments, a peptide cleanup protocol comprises a gradient of high to low MeOH and ACN. In some embodiments, a peptide cleanup protocol comprises a gradient of low to high MeOH and ACN. In some embodiments, a peptide cleanup protocol comprises a 40% final concentration MeOH, a 45% final concentration MeOH, a 50% final concentration MeOH, or a 55% final concentration MeOH. In some embodiments, a peptide cleanup protocol comprises a 40% final concentration ACN, a 45% final concentration ACN, a 50% final concentration ACN, or a 55% final concentration ACN.
In some embodiments, a method disclosed herein further comprises one or more wash steps after peptide binding. In some embodiments, a method disclosed herein further comprises one or more wash steps comprising one or more organic solvents. In some embodiments, a method disclosed herein further comprises a wash protocol. In some embodiments, a wash protocol comprises one or more organic solvents. Non-limiting examples of organic solvents suitable for use in one or more wash steps or a wash protocol include acetonitrile (ACN), methanol (MeOH), ethanol (EtOH), and isopropyl alcohol (IPA).
In some embodiments, a wash protocol disclosed herein alters pH.
In some embodiments, a sample is washed with one more organic solvents. In some embodiments, a method disclosed herein further comprises a wash comprising one or more organic solvents comprises 80% MeOH, 95% EtOH, 100% EtOH, or 100% ACN. In some embodiments, a method disclosed herein further comprises a wash comprising one or more organic solvents comprises 50% ACN and 45% MeOH, or 45% ACN and 50% MeOH.
Disclosed herein are profiling methods that preserve the native form of the protein and allow the global identification and/or quantification of proteins and/or protein complexes in their native states.
As known to one of skill in the art, proteins rarely function alone and their activity is often determined by their binding partners and interactions. However, the majority of high throughput proteome profiling techniques first denature protein complexes, resulting in a loss of information about protein complexes and protein-protein interactions. Non-limiting examples of advantages of native profiling of protein complexes disclosed herein include: preserving a native form of the protein and allowing a global quantification of proteins in their native states.
A number of native profiling methods have been developed to profile protein-protein interactions and protein complexes such as yeast two-hybrid, affinity purification, APEX (Rhee et al. Science 2013, 339:1328-1331) and BioID (Roux et al. J. Cell. Biol. 2012, 196:801-810) and there have been efforts to try to characterize the human interactome in 293T cells using the affinity purification of 10,128 proteins (Huttlin et al. Cell 2021, 184:3022-3040.e28). However, such native profiling methods are extremely resource intensive and not easily scalable to profile multiple samples to compare changes in protein-protein interactions and protein complexes.
In some embodiments, a method disclosed herein comprises co-fractionation mass spectrometry (CF-MS) for native protein profiling. In some embodiments, CF-MS is advantageous since it uses endogenous lysate/samples and does not require genetic manipulation or heterologous expression. In some embodiments, during CF-MS, proteins are extracted in their native form, separated using native chromatography and fractions analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). In some embodiments, proteins in a complex co-fractionate and have similar profiles when separated with native chromatography. In some embodiments, protein complexes are inferred using one or more bioinformatic methods to determine which proteins have high correlations and likely are part of complexes. In some embodiments, a correlation method to predict complexes is by Pearson, Spearman, Kendall, Euclidean distance or a mixture of scores.
In some embodiments, a CF-MS method disclosed herein comprises: i) native separation of the complexes, ii) proteome profiling including sample preparation and LC-MS quantification and iii) data analysis of protein complex inference and comparison.
Non-limiting examples of native fractionation for native plasma profiling and chromatographic fractionation methods include: size-exclusion chromatography (SEC), ion exchange chromatography (IEX), native-PAGE, and differential centrifugation.
In some embodiments, a method disclosed herein comprises size-exclusion chromatography (SEC). In some embodiments, native fractionation for protein profiling is by SEC. In some embodiments, native fractionation for plasma profiling is by SEC. In some embodiments, fractionation of plasma by SEC comprises label-free AutoP3 and/or AutoMP3.
In some embodiments, a method disclosed herein comprising SEC has one or more improved properties. Nonlimiting examples of one or more improved properties of SEC includes: samples for plasma profiling can be injected automatically with autosampler and an automated fraction collector, and ease of downstream sample preparation compared to, for example, extraction of proteins from native gels.
Although CF-MS is much less resource intensive than a global IP-type approach involving the pull-down of thousands of proteins it is inherently still very time consuming both from a sample preparation angle as well as the LC-MS instrument time angle since one sample inevitably gets fractionated into tens of fractions each requiring analysis.
To enable CF-MS to become high-throughput (HT), an automated sample preparation is advantageous due to the large number of fractions that need to be analyzed. In some embodiments, LC-MS acquisition time for label free experiments is decreased by minimizing the gradient length and minimizing sample loading time.
Non-limiting examples of LC-MS methods for native plasma profiling include: label-free with input not adjusted for protein concentration, label-free with input adjusted for protein concentration, isobaric labeling with input not adjusted for protein concentration, isobaric labeling with input adjusted for protein concentration, isobaric labeling where each SEC fraction is a distinct plex, isobaric labeling where a single plex is selected and one or more samples with no bridge, and isobaric labeling where each SEC run is split across multiple plexes and performed with one or more samples having a bridge.
Non-limiting examples of advantages of a LC-MS method for native plasma profiling disclosed herein include ability to estimate stoichiometry, improved detection of complexes with high molecular weight (MW), decreased instrument time, and ability to further fractionate plexes.
In some embodiments, throughput of a method disclosed herein is increased with multiplexing analysis using isobaric tags, such as Tandem Mass Tags (TMTpro). In some embodiments, a method disclosed herein comprises comparing proteins/complexes in specific SEC fractions or across the whole SEC gradient.
In some embodiments, a sample is label-free with input not adjusted. In some embodiments, a method disclosed herein comprising label-free sample with input not adjusted has improved protein correlation profiling (PCP). In some embodiments, a sample is label-free with input adjusted. In some embodiments, a method disclosed herein comprising label-free sample with input adjusted has improved detection of low abundant complexes. In some embodiments, a sample is isobaric labeled. In some embodiments, a method disclosed herein comprising isobaric labeled sample has improved protein quantification between samples.
In some embodiments, a method disclosed herein is used to identify antigens in immune complexes in a sample. In some embodiments, a method disclosed herein is used to identify immunoglobulin binding partners in immune complexes in a sample. In some embodiments, a sample is plasma.
In some embodiments, a method disclosed herein has one or more improved properties. In some embodiments, a method disclosed herein has one or more improved properties compared to standard methods known in the art. Nonlimiting examples of improved properties includes: increased throughput for native separation and reduced experimental variation. In some embodiments, protein-level TMT labeling is performed in plasma. In some embodiments, protein-level TMT labeling is performed in a body fluid. In some embodiments, protein-level TMT labeling is performed in cell lysate and tissues. In some embodiments, cell lysate and tissues are lysed using non-denaturing conditions.
Disclosed herein is an automated platform with label-free batch processing (also referred to as “AutoP3”) for high throughput proteome profiling.
Non-limiting examples of samples for processing by AutoP3 include plasma, cell or tissue lysate, other body fluids, and any combination thereof. As understood by one of skill in the art, the number of samples can be increased or decreased based on the experimental design.
Plasma is typically a difficult sample for deep proteome profiling as a result of the large dynamic range with just a few proteins dominating the majority of the protein content. In some embodiments, a method disclosed herein comprises native fractionation. In some embodiments, native fractionation comprises a tandem column set-up of Sepharose columns. In some embodiments, native fractionation comprises a tandem column set-up of Sepharose columns provides higher resolution compared to a single column. In some embodiments, native fractionation for plasma profiling is by SEC.
In some embodiments, fractionation of plasma by SEC comprises label-free AutoP3. It is estimated that the sample preparation time that would be needed to prepare the 954 samples manually would be around two years using standard approaches prepared sequentially. In comparison, HT label-free AutoP3 requires just 4 days to go from fractionated SEC in 96-well plates to the samples in ready-to-shoot LCMS vials.
In some embodiments, fractionation of plasma by SEC comprises isobarically labeled AutoMP3. It is estimated that application of the AutoMP3 workflow to SEC samples where the sample preparation of 954 samples requires an additional 2 days in addition to the 4 days required for label-free sample preparation. However, the increased sample preparation time comes with the advantage of reduced LC-MS acquisition time on the instrument.
Using a 3 h method per sample requires under a week of MS time (6.625 days) with no washes. Using a 2 h method requires just 4.4 days and a 1 h method just 2.2 days for the multiplexed analysis of 954 samples. Greater depth and accuracy is achieved with the longer gradients, especially when using FAIMS, real time search (RTS) and MS3 with synchronous precursor selection (SPS) methods on a tribrid mass spectrometer but depending on the throughput required, greater numbers of samples may be advantageous at the expense of depth of coverage, especially if the specific complexes of interest are known to be fairly abundant.
In some embodiments, each sample in a SEC fraction is labeled individually. In some embodiments, samples are normalized across different SEC fractions prior to TMT labeling.
Nonlimiting examples of comparisons of performance include comparisons of: average number of peptide and protein identifications, missed cleavage rates, variability in method, labeling efficiency (%) MC vs iST vs MP3, quantified total peptides, unique peptides, and total proteins, accuracy and precision, based on the yeast spike-in peptides.
The disclosure is further illustrated by the following examples. The examples are provided for illustrative purposes only, and are not to be construed as limiting the scope or content of the disclosure in any way.
Four sample preparation methods were evaluated as alternatives to MC: in-solution (IS), in-StageTip (IST), Single-Pot Solid-phase enhanced Sample Preparation (SP3) and S-Trap (ST).
In some embodiments, clean-up is prior to combining the two or more label peptides or polypeptides. In some embodiments, clean-up is prior to combining the two or more portions into one plex. In some embodiments, the clean-up method is selected from the group consisting of peptide or protein precipitation, in-solution (IS), in-StageTip (IST), Single-Pot Solid-phase enhanced Sample Preparation (SP3), filter-aided sample preparation (FASP), S-TRAP, or C18 column based clean-up such as SepPak (Waters). In some embodiments, clean-up is by de-salting. In some embodiments, clean-up is by detergent removal. In some embodiments, clean-up is by lipid or metabolite removal. In some embodiments, clean-up is by small molecule removal.
Method Comparison and Optimization Samples. For all label-free method comparison studies as well as SP3 digestion optimization either cell lysate of A-375 melanoma cells or mouse plasma from naive C57/B6 mice were used. Mouse plasma was collected in-house from C57/B6 mice. For the yeast spike-in experiment to assess precision and accuracy of TMT labeled 11-plex, DBY12000 yeast strain and A-375 melanoma cells were used.
Lysate Generation. Unless otherwise indicated, lysis buffer of 75 mM NaCl, 50 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) pH 8.5, 3% SDS (sodium dodecyl sulfate), with Roche Protease Inhibitor cocktail was added to A375 cell pellet or mouse plasma. Cell lysates were additionally sonicated with EpiShear sonicator probe (ActivMotif, Carlsbad, CA) (cycle settings: amplitude of 40%, pulse 10 seconds ON, 5 seconds OFF, for a total time of 50 seconds). Protein quantification was performed using the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific, Waltham, MA). Proteins were reduced with dithiothreitol (DTT, 5 mM, 30 min, 25° C.), alkylated with iodoacetamide (IAA, 15 mM, 30 min, 25° C., in the dark), and excess IAA was quenched with equivalent DTT (5 mM, 25° C., 5 min in the dark). An aliquot of 100 μg of protein was then purified using either methanol-chloroform protein precipitation (MC), In-solution (IS), in-StageTip (IST), Single-pot solid-phase enhanced sample preparation (SP3) or S-Trap (ST).
Vacuum Assisted Sample Desalting using SepPack columns. Column was first equilibrated with 1×1 mL 100% methanol. Followed by equilibration with 1×1 mL 80% acetonitrile (ACN). Followed by 1×1 mL wash of 0.5% acetic acid. Followed by 3×1 mL washes of 0.1% trifluoroacetic acid (TFA). Acidified sample was then bound to column resin. This was followed by another 3×1 mL wash of 0.1% TFA. Followed by 1×1 mL wash of 0.5% acetic acid. Cleaned up peptides were then eluted with 1×0.75 mL 40% ACN, 0.5% acetic acid, and 1×0.75 mL 80% ACN, 0.5% acetic acid. Elutions were combined and dried down using a speedvac.
Methanol/Chloroform Sample Preparation. Samples were methanol-chloroform precipitated as previously described (McAlister G C, et al. (2014) MultiNotch MS3 Enables Accurate, Sensitive, and Multiplexed Detection of Differential Expression across Cancer Cell Line Proteomes. Anal Chem 86 (14): 7150-7158). Briefly, to one volume of sample, four volumes of 100% methanol, one volume of 100% chloroform, and three volumes of 100% water were added with a vortex mix step after each addition, followed by a centrifugation at 15,000 g for 30 minutes at 4° C. to clarify the precipitated protein pellet, which was followed by removing both of the liquid phases, a 1000 μL 100% methanol wash and centrifugation at 15,000 g for 10 minutes at 4° C., removal of wash and air drying. The precipitated protein was resuspended in 100 μL 8M urea, 50 mM HEPES pH 8.5. The sample was then diluted to 4M urea and digested first with LysC at a ratio of 1:100 protease: protein (16 h, 25° C.). The samples were then diluted to 1 M urea and digested with trypsin at 1:25 protease: protein (6 h, 300 rpm, 37° C.). Samples were then acidified (final concentration 0.5% TFA). Samples were centrifuged to remove cellular debris, and supernatant saved (15,000 g, 10 min, 4° C.). Samples were then desalted using C18 solid-phase extraction (SPE) columns (SepPak) as described above and dried down in a speedvac overnight. Peptide BCA was then performed to quantify peptide content. (Thermo Fisher Scientific, Waltham, MA). For a label-free mass spectrometry experiment, samples were resuspended in 5% formic acid & 5% ACN (MS Loading Buffer), and 1 μg was injected on the instrument. For TMT experiments, 50 μg peptide material was then aliquoted per sample to prepare for labeling. For an 11-plex/16-plex TMT experiment a ratio of 4:1/8:1 of TMT reagent: peptide was used per sample for cell lysate and mouse plasma respectively. Labeling was performed in 50 μL 200 mM HEPES, 30% anhydrous ACN for 1 hour shaking at 300 rpm, and room temperature. Reaction was quenched with 5% hydroxylamine, 200 mM HEPES pH 8.5 to final concentration of 0.3% v/v for 15 minutes at room temperature. For label ratio check, 4 μL was taken out each labeled sample combined, desalted using a StageTip, and analyzed on the instrument. Using the label ratio check, samples were normalized, combined, and desalted using C18 SepPak, as described above. High pH Reversed-Phase (HPRP) fractionation, was performed as described below. For mass spectrometry analysis TMT-based fractionated experiments, 12 fractions of set A were each resuspended in 20 μL MS Loading Buffer and 2 μL per fraction were analyzed by Orbitrap Lumos Fusion mass spectrometer. For label-free unfractionated experiments, 10 μg per sample was resuspended in 20 μL MS Loading Buffer and 1 μg analyzed on the instrument.
In-Solution (IS). Both mouse plasma and cells were lysed as described above except with the following lysis buffer: 8 M Urea 50 mM HEPES pH 8.5, with Roche Protease Inhibitor cocktail. No cleanup post lysis and prior to digestion was performed. Digestion and peptide desalting were performed the same way as described in methanol/chloroform method. Samples were resuspended in MS Loading Buffer and a BCA assay was then performed to assess peptide material recovered. For label-free instrument analysis, 10 μg were resuspended in 20 μL MS Loading Buffer and 2 μL (1 μg) were analyzed by LC-MS/MS.
Label-free in-StageTip (IST) method. For initial method comparison the label-free iST method is used as per manufacturer's instructions (Preomics, Martinsried, Germany). In brief, approximately 100 μg mouse plasma or cell lysate samples were reduced and alkylated at 95° C. for 10 minutes, shaking at 1,000 rpm. BCA assay was performed to adjust sample amounts to be exactly 100 μg for the proceeding digestion step. Samples were digested over the filter cartridge for 3 hours at 37° C., 500 rpm. Digestion was quenched and samples were bound to filter by spinning at 3,800 g for 2 minutes. Samples were washed once with two sequential wash solutions, and then eluted twice with 100 μL of elution buffer. Samples were dried down in a speed-vac overnight, resuspended in 200 μL MS Loading Buffer at an approximate concentration of 0.5 μg/L, and 2 μL (1 μg) was analyzed by LC-MS/MS.
TMT-compatible in-StageTip (iST-NHS) method. For yeast spike-in experiment the TMT-compatible iST method was used as described by the vendor in an earlier version where it was advised to perform TMT labeling on top of the filter. A more recent version advises to perform TMT labeling in-solution in a tube, and then transferring the labeled sample to the filter for wash—this produces better labeling efficiency. In brief, approximately 100 μg mouse plasma or cell lysate samples were lysed, reduced and alkylated using iST specific alkylating reagent (C6H11NO composition, 113.084 Da) in a TMT-compatible lysis buffer at 95° C. for 10 minutes, shaking at 1,000 rpm. BCA assay was performed to adjust sample amounts to be exactly 100 μg for the proceeding digestion step. Samples were digested over the filter cartridge for 3 hours at 37° C., 500 rpm. TMT reagent was added at a ratio of 8:1 for mouse plasma, and 4:1 for cell lysate and labeling was performed at room temperature, 500 rpm, for 2 hours. Labeling reaction was quenched with 12 μL 5% hydroxylamine solution for final concentration of 0.3% v/v. Digestion was quenched and samples were bound to filter by spinning at 3,800 g for 2 minutes. Samples were washed once with two sequential wash solutions, and then eluted twice with 100 μL of elution buffer. Samples were dried down in a speed-vac overnight, resuspended in 200 μL MS Loading Buffer (˜0.5 μg/μL) and 2 μL (1 μg) analyzed on the mass spectrometer for a ‘ratio check’. Samples were then combined based on ratio corrected sum signal to noise values per TMT channel and dried down. HPRP fractionation was performed as described below. For mass spectrometry analysis, 12 fractions of set A were each resuspended in 20 μL MS Loading Buffer per fraction and 2 μL were analyzed by Orbitrap Lumos Fusion mass spectrometer.
S-Trap (micro): range of <=50-100 μg & S-Trap (mini): range of 100-300 μg. Sample preparation was performed as indicated by the kit (Protifi, Huntington, NY). Epishear probe sonication method was performed as described above for cell lysis. Protein- and peptide-level quantification was performed using the BCA assay. A ratio of 1:25 of trypsin: protein was used to digest 100 μg of either mouse plasma or cell lysate material. For label-free instrument analysis, 10 μg were resuspended in 20 μL MS Loading Buffer and 2 μL (1 μg) were analyzed by LC-MS/MS.
Optimization of peptide recovery for SP3. For each trial 100 μg of starting protein material (human cell lysate and mouse plasma) was processed starting with lysis through digestion stage with and without peptide cleanup. Peptide level quantitation was performed using BCA assay in the digestion buffer or diluted elution buffer (to maintain compatibility with BCA assay reagents) utilized for the trial. Variations of digestion volumes, Sera-Mag magnetic beads: protein ratios, wash, elution buffers as well as conditions and switching to processing samples in PCR tubes up to peptide cleanup stage were performed to improve peptide recovery.
Optimization of digestion efficiency of plasma samples utilizing SP3. For each condition, in triplicate, 100 μg of starting mouse plasma protein was processed starting with lysis through digestion stage with peptide cleanup and followed by analysis on LC-MS2. Various digestion buffer reagent combinations were utilized as indicated in
Single-pot solid-phase-enhanced sample preparation (SP3) is based on the unbiased immobilization of proteins and peptides on carboxylate modified hydrophilic and hydrophobic coated surfaces mix of beads. Immobilization is promoted through trapping proteins and peptides in an aqueous solvation layer on the bead surface through increasing the concentration of organic additive in solution. Once on the bead surface, proteins and peptides can be processed and rinsed to remove contaminants, such as detergents. Allows for binding of proteins in a non-selective mode. The purified proteins can be eluted from the beads through the addition of an aqueous solution. Resulting peptides can be used in downstream HPLC-based fractionation methods. Or, the peptides can be re-immobilized on the paramagnetic beads for sample clean-up prior to MS-analysis, thus eliminating common sample handling steps.
Multiplexed Single-Pot Solid-Phase enhanced Sample-Preparation (MP3). Lysis, reduction and alkylation were performed as described in methanol/chloroform sample preparation described above. Protein cleanup, digestion, and peptide cleanup sections were performed using SP3 as described previously (McAlister G C, et al. (2014) MultiNotch MS3 Enables Accurate, Sensitive, and Multiplexed Detection of Differential Expression across Cancer Cell Line Proteomes. Anal Chem 86 (14): 7150-7158) except with the following modifications and the addition of the TMT labeling stage. Initial binding of protein to beads was performed at 75% ACN final concentration. Post-binding, two 70% ethanol washes followed by 100% ACN wash were performed. For protein digestion, the protein-bead mixture was resuspended in 90 μL 50 mM HEPES pH 8.5, 10 mM CaCl2. Digestion was performed with Lys-C overnight at room temperature (1:100 protease-to-protein), followed by trypsin (1:25 for trypsin) for 6 h at 37° C. on a thermal mixer. For biological experiments requiring multiple plexes, at this stage a small portion from each sample, or as indicated, was taken to generate a ‘bridge’ sample of 50 μg peptide material for each plex (i.e. 15 samples+bridge per plex). TMT labeling was performed using 50 μg in 80 μL digestion buffer following removal from the magnet immobilized beads. TMT reagent was used at a ratio of 4:1 for cell lysate samples, and 8:1 for mouse & naked mole-rat plasma samples. Excess TMT was quenched with 5% hydroxylamine to a final concentration of 0.3% v/v. Samples per plex were then combined, pipette mixed, and split back out into portions containing no more than 100 μL aqueous volume to bind and cleanup with magnetic beads in 2 mL tubes or 96-well 2 mL plate, to maintain >=95% final ACN concentration to bind peptides to beads. Bead-bound samples were washed 1× with 1000 μL 100% ACN. The portions were eluted 3× with 100 μL 5% ACN, re-combined and dried down in a speedvac overnight. The combined samples were then fractionated offline as described by the HPRP method below. Twelve fractions of set A, resuspended manually in 20 μL MS Loading Buffer and 2 μL per injection, were analyzed on either Orbitrap Lumos Fusion, or Orbitrap Eclipse with FAIMS Pro and RTS, mass spectrometers as indicated.
Automated Multiplexed Protein Profiling Platform (AutoMP3). MP3 method described above was automated using Hamilton Vantage™ liquid-handling platform. With the exception of manual lysis step, all of the steps are automated with short hand-on preparation time prior to each step: i) reduction & alkylation, ii) protein cleanup, iii) LysC & Trypsin digestion, iv) bridge generation, v) peptide material normalization prior to TMT labeling, vi) TMT labeling, vii) labeling efficiency check, viii) TMT quenching, ix) Combine-Mix-Split, x) Peptide Cleanup. Eluted cleaned up TMT labeled peptides are then dried down over night and resuspended either for offline HPRP fractionation as described below, or with direct resuspension in MS loading buffer for analysis on either Orbitrap Lumos Fusion, or Orbitrap Eclipse with FAIMS Pro and RTS, mass spectrometers as indicated. Minimal operator intervention was necessary prior to run for each stage to loading the necessary tips, plates and reagents onto the deck.
HPRP Fractionation. The pooled TMT-labeled peptides were fractionated using HPRP. The samples were resuspended in 5% ACN, 10 mM Ammonium Bicarbonate pH 8.0, 95% H2O and fractionated over a ZORBAX extended C18 column (Agilent, 5 μm particles, 4.6 mm inner diameter and 250 mm in length). Peptides were separated on a 75 min linear gradient from 5% to 35% ACN in 10 mM ammonium bicarbonate pH 8.0 at a flow rate of 0.5 mL/min on an Agilent 1260 Infinity pump equipped with a degasser and a diode array detector (set at 214-, 220-, and 254-nm wavelength) from Agilent Technologies (Waldbronn, Germany). The samples were fractionated into a total of 96 fractions and then consolidated in a checkerboard pattern sets of twelve, set A and set B. Samples were dried down under vacuum and reconstituted in MS Loading Buffer for LC-MS/MS processing.
AutoMP3 on Hamilton Vantage™. Liquid classes were optimized, especially for high organic solutions handling. Necessary reagents (lysis buffer, 100 mM DTT, 250 mM IAA) are loaded manually onto respective 96-well plates prior to start. The AutoMP3 protocol begins with 100 μg protein plasma sample with 41.6 μL with a lysis buffer in 200 μL 96-well plates. Followed by the addition of 2 μL 100 mM DTT using the to reduce disulfide bonds and incubated at room temperature for 30 minutes. Then the samples were cysteine alkylated with addition of 2.4 μL 250 mM IAA, incubated in the dark with a plate lid at room temperature for 30 minutes. This was followed up by an additional 2 μL 100 mM DTT to quench the reaction.
While quenching is occurring, the deck is prepared for protein binding stage with manual aliquoting of Sera-Mag magnetic beads (GE Healthsciences) freshly resuspended in H2O as well as stock of 100% ACN, 70% EtOH, and digestion buffer in separate 96-well plates (200 μL plate for digestion buffer and 2 mL plate for rest of reagents). The rest of the stage was completed by the liquid-handling platform. Stock beads are aliquoted into a custom 3D printed 200 μL plate and supernatant removed using the 96-well Alpaqua magnet. Samples were then thoroughly mixed using the 96-channel head to ensure fast and consistent mixing across the entire plate. Followed by addition of 150 μL 100% ACN to reach final organic concentration of 75% to bind protein to beads, and incubated for 18 minutes, then immobilized for 2 minutes on the Alpaqua magnet. Supernatant was removed while the sample plate is on the magnet, followed by two washes with 100 μL 70% EtOH, and one wash with 100 μL 100% ACN and an additional wash removal step to ensure no ACN is left prior to adding the digestion buffer. The digestion buffer (87 μL 50 mM HEPES, pH 8.5, 10 mM CaCl2) was added once the sample plate was been moved off of the magnet.
This was followed by manual setup for digestion with addition of a 200 μL 96-well plate containing LysC protease. Then, 5 μL of LysC (10 AU resuspended in 5 mL ice-cold water) are added per sample. At each column addition, samples were thoroughly mixed to ensure beads were completely resuspended prior to manually sealing the plate with 8-well strip caps and incubating for 16 hours, 1000 rpm, room temperature using the heater/shaker module. Trypsin was added similarly in the morning at a 1:25 ratio of enzyme to protein (4 μg in 8 μL) and thoroughly mixed to completely resuspend the magnetic beads. And again, manually sealing the plate with 8-well strip caps, followed by an incubation for 6 hours, 1000 rpm, 37° C. using the CPAC module. Peptide BCA was performed automatically on the platform with manual setup of necessary reagents. A worksheet was automatically generated using an in-house python script to direct the 8-channel head to normalize sample volume amounts to be at 50 μg peptide material. If multiple plexes were necessary, at this stage an additional worksheet was generated to create a pooled ‘bridge’ sample of 50 μg peptide material per plex.
In preparation for TMT labeling stage, worksheets were automatically generated to direct the 8-channel head aspiration/dispense steps of the TMTpro reagent as well as combine-mix-split strategy. Necessary reagents were manually set up on the platform including each TMTpro reagent that was aliquoted into a separate 1.7 mL tube and placed in the 32-tube rack. TMT labeling was then started on the platform by first aliquoting portions of TMTpro reagent into a 200 μL 96-well plate to minimize variability of ACN (that the reagent was resuspended in) evaporating during addition to samples. All of the portioned TMTpro reagent is then added at once (24 μL/50 μg plasma sample) to the sample plate using the 96-channel head for rapid, consistent and thorough mixing prior to sealing of the plate using a flexible 96-well mat (Corning) and moved to heater/shaker for 1 hour incubation, room temperature, 300 rpm shaking. Samples are then quenched with 5% hydroxylamine from a reservoir, mixing upon dispensing, followed by 15-minute incubation on the plate rack.
Labeling ratio check evaluations were then performed by taking 2 μL, using the 8-channel head and a specifically built worksheet, from each sample, then combining, followed by peptide cleanup and analyzing on the instrument. Combine-mix-split strategy is employed: samples, per 16-plex, are combined (with optional input from ratio check results) using the 8-channel head into separate wells in a 2 mL 96-well plate. Each plex is then thoroughly mixed prior to splitting out portions for peptide cleanup.
A new 2 mL plate was prepared with another aliquot of 50 μL of magnetic beads per well in preparation for TMT labeled peptide cleanup, and supernatant is removed using the Alpaqua magnet. Then each plex was split back out into the 2 mL 96-well plate containing the magnetic beads using two columns per plex, which completed the combine-mix-split strategy. Peptides were bound to the beads by reaching >95% organic concentration through addition of 1900 μL 100% ACN, and thorough mixing. Samples were incubated for 18 minutes, followed by incubation on the Alpaqua magnet for 2 minutes to immobilize the beads. A single 1 mL 100% ACN wash was performed. Peptides were then eluted into 1 mL 96-well plate (Eppendorf) three times through addition of 100 μL 5% ACN, thorough mixing and 1000 rpm shaking for 18 minutes on the heater/shaker. For elution, the magnetic beads were then immobilized one more time and eluate is transferred to the 1 mL 96-well plate. For fractionation testing, elution was instead performed using different solutions as indicated in
Evaluation of methods compatible for automated proteomics sample preparation. For the first round of comparisons the peptide recovery was evaluated for each of the methods for both plasma and cell lysate (
Next, the number of peptide and protein identifications and the missed cleavage rate was evaluated using each sample preparation method (1 μg). The number of proteins identified were similar across all the methods for both cell lysate and mouse plasma sample types, with the exception of SP3, due to the low peptide recovery amount (
Although SP3 performed the worst in this initial comparison, it was kept in the evaluation for a number of reasons including high score for ease of automation (
IS was not tested further because it is difficult to automate and limited reagent compatibility due to lack of robust and fast cleanup mechanism. The iST method both has large and optimized method coverage (lysis through TMT labeling steps), and 96-well format, aside from its cost, making it an attractive option. ST did not perform as well as iST and had less method optimization and coverage. No further testing was performed.
Comparison of the methods most amenable to automation (iST vs SP3). Following through some optimization of SP3 steps, much improved peptide recovery and digestion efficiency was achieved (
Optimization of SP3: peptide recovery, digestion conditions, TMT labeling. Post SP3 optimization 70 μg recovery for 100 μg starting protein amount, (
The optimizations included varying digestion volume and buffer composition, changing from 2% DMSO to 5% ACN elution solution, varying bead to protein ratios. To address the low digestion efficiency in plasma for SP3 a matrix scheme was devised with 24 conditions to evaluate three fundamental parts of the digestion buffer based on the literature: buffer composition, use of CaCl2, and use of surfactants. For buffer composition, 50 mM HEPES pH 8.5, +/−1% Deoxycholate (DOC), +/−5% Trifluoroethanol (TFE) was evaluated. The effect of addition of 10 mM CaCl2, as well as the addition of Rapigest, PPS, and Invitrosol surfactants was investigated.
The presence of small molarity calcium chloride improved digestion efficiency across all conditions (
The large number of steps involved in a multiplexed TMT experiment makes creating a fully automated method challenging (
βTargets for future improvements.
The two most attractive methods to automate in a 96-well plate format were SP3 and IST (
Initial experiments with SP3 resulted in both poor recovery and digestion efficiency (
To optimize the digestion efficiency, a matrix of 24 different conditions was prepared, investigating the effect of adding 1% Deoxycholate (DOC), +/−5% Trifluoroethanol (TFE), Rapigest, PPS, and Invitrosol surfactants as well as CaCl2. The greatest improvement was observed by the addition of CaCl2 and 5% TFE. With these changes the missed cleavage rate decreased from 30% to <5% (
To decide whether to automate SP3 or iST, the performance of these two methods were compared using both label-free and TMT workflows. In the label-free experiment, SP3 and iST methods had similar performance for mouse plasma samples, both in terms of the average number of peptide and protein identifications as well as missed cleavage rates. However, there was higher variability in iST method (
A TMT-based two organism (human/yeast) spike-in experiment was performed (Table 3) to compare the precision and accuracy of the MP3 and iST against MC.
It was found that MP3 had the smallest coefficient of variation when considering human only peptides (
The MP3 method was automated end-to-end on a Hamilton Vantage™ robot. There were a number of requirements for the automated workflow. One, sufficient capacity for consumables such that end-to-end automation of each of the steps (
For the Vantage™ configuration, two heads were selected: a 96-well fixed-head capable of picking up volumes between 1-1,000 μL, and an individually operated multichannel head consisting of eight 1-1,000 μL channels. A two-meter deck-length option was selected for its large capacity. The internal plate gripper (IPG), which is capable of transferring items around most of the deck space, was a crucial element of the platform. Additionally, two heater-shakers, a cold plate air cooled (CPAC) shaker, and a magnetic rack for magnetic bead immobilization were added.
The large deck was capable of fitting all of the tips, plates, liquid and tip waste compartments, and modular devices necessary for the method. The deck is spacious enough for future additions. Aside from bulk processing of 96 samples at a time, the presence of an 8-channel head allowed for the freedom to finely control key stages of the sample processing workflow. First, various configurations of TMTpro, or alternate reagents could be patterned across the 96-well plate. Second, it was possible to normalize sample content at the protein, peptide, and TMT-labeling-stages. Third, it was possible to generate bridge samples per plex at any stage.
All of the liquid-handling steps were tested and optimized, especially for small volume handling steps such as the addition of LysC and trypsin for the digestion. The liquids used in the protocol vary considerably in physical properties, from the viscous bead slurry to the volatile 100% acetonitrile. To meet this challenge the Vantage™ pipetting channels and software allow for detailed control of all the parameters of the pipetting process. Thorough mixing is critical for a number of the protocol steps, in particular the resuspension of the magnetic beads with protein lysate and the addition of the TMT reagent for peptide labeling. Care was taken to mix reagents as quickly as possible, but without causing bubbles to form or residual liquid to be retained in the tips.
The main challenges with automating MP3 were variability, TMT labeling and cleanup post TMT. The sequence timing for the pipetting steps and plate movements were carefully considered to keep well variation low. The TMT labeling stage was difficult due to the high vapor pressure of acetonitrile which the TMT reagents were dissolved in. To address this, liquid classes for the steps involving acetonitrile were optimized. Cleanup post TMT labeling was a major challenge since the total volume for a 16-plex experiment results in very large volumes (for example, 1.6 mL aqueous volume requires 30.4 mL 100% acetonitrile for the peptide binding stage). Large volumes are difficult to deal with in a 96-well plate format. To solve this problem, a combine-mix-split strategy was devised to keep the volumes low (
In some embodiments, the method further comprises mixing the labeled peptides or polypeptides. In some embodiments, the mixing is by aspiration and dispensing. In some embodiments, the mixing is by vibration. In some embodiments, the mixing is by passive diffusion.
In some embodiments, one plex is split into two or more portions. In some embodiments, the plex-portions are substantially homologous. In some embodiments, each plex-portion is substantially homogeneous. In some embodiment, one plex is split into a plurality of plex-portions. In some embodiments, the plurality of plex-portions have substantially equal volumes. In some embodiments, the plurality of plex-portions have identical volumes. In some embodiments, the plurality of plex-portions have unequal volumes.
In some embodiments, increasing the number of samples increases the number of plex-portions due to an increased volume of samples. For examples, 16 samples may be split into about 16 plex-portions and the volume of each portion is about 91 μL.
In some embodiments, the volume of each plex-portion is less than about 1 mL. In some embodiments, the volume of each portion is about 1 μL to about 100 μL. In some embodiments, the volume of each portion is more than 6 μL. In some embodiments, the volume of each portion is less than 100 μL. In some embodiments, the volume of each portion is about 5 μL. In some embodiments, the volume of each portion is about 10 μL. In some embodiments, the volume of each portion is about 20 μL. In some embodiments, the volume of each portion is about 30 μL. In some embodiments, the volume of each portion is about 40 μL. In some embodiments, the volume of each portion is about 50 μL. In some embodiments, the volume of each portion is about 60 μL. In some embodiments, the volume of each portion is about 70 μL. In some embodiments, the volume of each portion is about 80 μL. In some embodiments, the volume of each portion is about 90 μL. In some embodiments, the volume of each portion is about 100 μL.
To validate the combine-mix-split strategy for TMT peptide cleanup, manual and automated preparations of mouse plasma labeled with TMTpro were compared using 16 technical replicates all mixed at 1:1 ratios. Comparable low variability was observed between the standard and combine-mix-split approaches (
The Hamilton Vantage™ liquid-handler system performed all of the liquid-handling for the sample preparation, with limited interactions from the end user. Plasma samples were aliquoted into plates, then reagents for reduction and alkylation were added and mixed automatically. Incubation steps were timed and controlled by the system. The protein purification steps were accomplished utilizing Sera-Mag magnetic beads. Throughout purification, the Hamilton Vantage™ was able to move the sample plate on and off of the magnetic rack to immobilize the beads, as well as to perform all of the washing and elution steps. The proteins were digested with LysC and trypsin, and the peptide concentration is measured via a bicinchoninic acid (BCA) assay (Thermo Fisher Scientific). The system used the individual concentration values to normalize each sample so that a consistent mass of 50 μg of peptides were labeled. The TMTpro labeling was automated, with the ability to adjust for the number of samples and degree of sample multiplexing. The post-labeling cleanup and combine-mix-split strategy took advantage of the Vantage™'s ability to handle deep-well plates and pipette larger volumes of 100% acetonitrile. TMT labeling, and cleanup were followed by a dry down step using a speedvac overnight. Plexes can then either be fractionated or resuspended in 5% acetonitrile/5% formic acid and directly analyzed on the mass spectrometer. A detailed description of all the steps is included herein.
The robustness of the automated MP3 method was evaluated on the Vantage™ across an entire 96-well plate. Samples prepared in a 96-well plate using a checkerboard pattern (
Transferring the MP3 method to an automated liquid-handling platform resulted in significant time savings for sample preparation (
In order to maximize throughput, performing single shot analysis was optimized while still maximizing the number of protein identifications (IDs) from blood plasma. To that end, the effect of including Real Time Search (RTS) and high-field asymmetric waveform ion mobility (FAIMS Pro) was evaluated. Previously both of these MS tools have been shown to increase the number of peptide and protein identifications, but neither of those studies tested the performance of these tools in the context of blood plasma. Extensive optimization was needed of both ion times and AGC targets for the MS3 scan in combination with testing out different compensation voltage (CV) settings (Table 6). The greatest number of quantified plasma peptides was achieved using CVs of −40, −50, −60, and −70. A pre-release version of the instrument control software was utilized for the Orbitrap Eclipse that allowed RTS to be used across multiple CVs in the same run (Tune 3.4). The addition of FAIMS Pro and RTS increased the number of peptide IDs over 64% and quantified peptides 48%, while the number of proteins IDs increased 63% and quantified proteins 49% (
The addition of FAIMS Pro alone resulted in a 13% increase in peptide IDs, while the addition of RTS alone resulted in a 37% increase (
One of the main issues when combining data from multiple TMT-plexes is that the overlap in identified and quantified peptides decreases as the number of plexes increases. While data dependent acquisition (DDA) is the most efficient way to quickly produce a large number of peptide IDs, it often lacks adequate peptide identification reproducibility due to the way it stochastically samples precursors. This becomes even more challenging in a sample matrix like plasma due to the large dynamic range of plasma proteins. In order to maximize chances of comparing protein and peptide abundances based on the measurement of the same peptides across multiple TMT-plexes, the potential effect of the usage of inclusion lists was evaluated across series of both a smaller (n=5) and larger (n=15) set of replicate injections. For both of the test series, the target masses on the inclusion list were derived by selecting the best scoring peptides from five initial DDA runs of the sample series. For the smaller sample set, the largest coverage and greatest overlap was achieved using a regular DDA method without an inclusion list: consistent measurement of 2,226 peptides (amounting to 44% overlap out of total unique peptides identified) across five single plasma injections (
Fractionation yields additional proteome depth for complex samples such as plasma. Fractionation was easily implemented with the AutoMP3 platform using a modified version of the peptide cleanup protocol, where instead of eluting peptides with a single buffer, three (or more) elution buffers could be applied (for example, a gradient of high to low concentration of acetonitrile) and each transferred to separate 96-well plates. Fractionating using AutoMP3 in the 96-well format allows for higher throughput than our standard high-pressure reverse phase (HPRP) fractionation setup. As with other steps, fractionation was designed in a modular fashion. The effect of fractionating label-free mouse plasma samples into three parts was evaluated, using various acetonitrile-based elution gradients. Fractionated sets were compared against three single shot injections of an unfractionated sample (
When considering the analysis of thousands of samples, it is critical to have robust chromatography. This is of particular importance when using an inclusion list method with retention time scheduling. Three LC columns from different suppliers were evaluated: New Objective (NO, 100 μm ID, C18 1.7 μm 130 Å, 25 cm bed), IonOpticks Aurora (IO, 75 μm ID, C18 1.6 μm 120 Å, 25 cm bed), and Pharmafluidics uPAC (uPAC, 40 μm ID, C18 pillar surface, 50 cm bed) (
Each column has its advantages and disadvantages (lifetime, pressure during runtime, capacity, and resolution) and each requires different optimizations. Initial comparison shows that, on average, the IonOpticks Aurora 25 cm column performed the best in our hands for both label-free and TMTpro labeled samples (
Sample Collection for Naked Mole-rat Circadian Rhythm Study. Naked mole-rats were raised under Association for Assessment and Accreditation of Laboratory Animal Care International [AAALAC] standards. Animal protocol A10199 was approved by the Buck Institute For Research on Aging Institutional Animal Care and Use Committee. Two-year-old naked mole-rats, one male and one female, were sacrificed using isoflurane inhalation followed by cardiac exsanguination every 2 hours over a 48 hr period. Whole blood was collected using EDTA as an anticoagulant and samples were centrifuged at 5000× rpm for 5 minutes. Plasma was separated from whole blood, aliquoted, flash frozen and stored at −80° C. until use. In total 48 naked mole-rats were euthanized in this study. Corresponding male and female timepoints were treated as duplicates in the analysis.
Sample Collection for UV Treated Mice and Naked mole-rats. Naked mole-rats and Skh1 hairless mice were raised under Association for Assessment and Accreditation of Laboratory Animal Care International [AAALAC] standards. Animal protocol A10139 was approved by the Buck Institute For Research on Aging Institutional Animal Care and Use Committee. Naked mole-rats were provided ad libitum access to food and mice were provided ad libitum access to food and water. Physiologically age-matched male naked mole-rats and Skh1 hairless mice were irradiated with the same dose of UV light (180 mJ/cm2). Two-year-old male naked mole-rats and 2-month-old male mice were placed on a rotating platform under 8 UV lamps emitting 72.6% UVB and 27.4% UVA. UV emission was measured using an ultraviolet sensor. Control animals were sham treated on the UV exposure platform. Animals were sacrificed via isoflurane and cardiac exsanguination 48-(2 days) and 168-hours (1 week) after UV exposure. Whole blood was collected using EDTA as an anticoagulant and samples were centrifuged at 5000× rpm for 5 minutes. Plasma was separated from whole blood, aliquoted, flash frozen and stored at −80° C. until use.
Molecular circadian rhythm changes were investigated in naked mole-rats using plasma samples, collected every 2 hours over a 48 hr period from young male and female naked mole-rats (
A power analysis was performed to determine the minimum number of replicates needed to detect circadian rhythm changes at several magnitudes over a 48-hour period (
Two-month-old, male Skh1 hairless mice and two-year-old, male naked mole-rats were treated with UV radiation (180 mJ/cm2) and sacrificed at both 48 hours and 7 days after a single UV exposure (
Comparison of protein fold changes between control and 48 hours post UV exposure, as well as control and 1-week post UV exposure, shows a larger number of >4-fold changes in abundance for the mouse plasma proteome. Changes in the naked mole-rat proteome were much more modest (
UV radiation induced signs of local low-grade inflammation, are clearly evident in the mouse plasma levels of inflammation-associated proteins, apolipoproteins, and metabolic proteins commonly associated with lipid metabolism and protein stability. High-density lipoprotein (HDL) proteins showed a tendency to change. These included a decline in Apoa1 and Apoa4 suggesting systemic inflammation and dysregulated lipid homeostasis. In addition to these observed changes in the mouse, most of the UV exposure induced changes in plasma proteins over the week-long monitoring period were in keeping with an Acute Phase Response (APR)—a pronounced systemic reaction elicited in response to local trauma, infection, and inflammation. These APR responsive proteins, commonly known as Acute Phase Proteins (APPs), showed the most pronounced changes at the 48 hr time point. Both negative increases and decreases in APP protein levels were evident. A marked decline in abundance of albumin (Alb), as well as a transthyretin (Ttr), Apoa1, Interleukin-1 receptor accessory protein (Il1rap), and retinol binding protein (Rbp4), were evident with the abundance in these negative APPs declining by 20-53% (
Although it was speculated that mole-rats would be more sensitive to UV, having lived in an environment devoid of UV light for more than 18 million years, surprisingly acute UV exposure has little effect on their plasma proteins. None of the negative APPs significantly changed over the time course post UV exposure (Table 7). Similarly, abundance of Apoa1, Itih4, and fibrinogen proteins were unaffected by UV. Hp, complement factors (C3 and C9) and Crp abundance changed after UV exposure, but their pattern of response was markedly different to that of mice (Table 7). Generally, reactions to UV were modest, and unlike mice, changes in the abundance levels of these proteins seven days after exposure were similar to that at 48 hr after exposure. For example, compared to the mouse response, Crp showed a significant but small change 48 hr after UV treatment and continued to show a modest increase at the one-week time point (
Two experimental methods were evaluated as alternatives to standard mass spectrometer data acquisition: (1) Native-MP3 binding+AutoMP3 with TMT labeling, and (2) Native-MP3 binding+label-free AutoMP3.
As depicted in
Proteins obtained from the Native-MP3 workflow were reduced by the addition of DTT (5 mM final concentration, 300 rpm, 30 mins, RT). This was followed by alkylation using iodoacetamide (15 mM final concentration, 300 rpm, RT, 30 mins in the dark). The alkylation reaction was subsequently quenched by the addition of DTT (final concentration 5 mM, RT). Protein cleanup was performed by the addition of 100% acetonitrile to reach 75% concentration to re-bind the protein sample back to the beads, with an 18-minute incubation period (RT). Particles were then immobilized on the magnet for 2 minutes, supernatant removed, followed by two 100 μL 70% ethanol washes, with a final 100 μL 100% acetonitrile wash. The sample was then resuspended in 90 μL 100 mM EPPS 10 mM CaCl2 pH 8.5 buffer, followed by protein digestion by adding 3.33 μg of LysC/Trypsin protease mixture in 10 μL 100 mM EPPS, 10 mM CaCl2 pH 8.5 buffer. Digestion was performed overnight at 37 degrees Celsius, 600 rpm. Evaluation of recovered peptide material was then performed using Pierce's BCA Assay kit as per manufacturer's instructions.
Post digestion, samples were immobilized for 2 minutes and the peptides (supernatant) moved to a fresh batch of 1:1 mixed hydrophobic: hydrophilic, 1000 μg SeraMag SpeedBeads (Cytiva, #44152105050350 & #45152105050350) at no more than 100 μL aqueous volume to bind and cleanup with magnetic beads in 2 mL tubes or 96-well 2 mL plate. Binding was achieved through the addition of 1900 μL 100% acetonitrile (>=95% final ACN concentration), and incubated for 18 minutes, no shaking. Bead-bound samples were washed 1× with 1000 μL 100% ACN. The portions were eluted 3× with 20 μL 5% ACN, 2 minutes 600 rpm, followed by magnet immobilization, elutions re-combined and dried down in a speedvac overnight.
Post digestion, following removal from the magnet immobilized beads, TMT labeling was performed by aliquoting out 5 μg peptides per sample (based on peptide BCA measurements) into 80 μL digestion buffer. TMT reagent was used at a ratio of 8:1. Excess TMT was quenched with 5% hydroxylamine to a final concentration of 0.3% v/v. Samples per plex were then combined, pipette mixed, and split back out into portions containing no more than 100 μL aqueous volume to bind and cleanup with magnetic beads in 2 mL tubes or 96-well 2 mL plate. Binding was achieved through addition of 1900 μL 100% acetonitrile (>=95% final ACN concentration), and incubated for 18 minutes, no shaking. Bead-bound peptides were washed 1× with 1000 μL 100% ACN. The portions were eluted 3× with 20 μL 5% ACN, 2 minutes 600 rpm, followed by magnet immobilization. Elutions and plex ‘portions’ were re-combined and dried down in a speedvac overnight.
Samples were analyzed either on an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) coupled to an Easy-nLC1200 (Thermo Fisher Scientific) or on an Orbitrap Eclipse mass spectrometer with Field Asymmetric Ion Mobility Spectrometry unit (FAIMSpro) with Real Time Search (RTS), (Thermo Fisher Scientific) coupled to an UltiMate 3000 (Thermo Fisher Scientific) as indicated. For both label-free and TMT multiplexed peptides, online separation was performed on an IonOpticks Aurora microcapillary column (75 μm inner diameter, 25 cm long, C18 resin, 1.6 μm, 120 Å). The total LC-MS run length for each sample was 180 min including a 165 min gradient from 8 to 30% ACN in 0.1% formic acid for TMT 16-plex multiplexed samples. For label-free experiments the total run length was 85 minutes, with 75-minute gradient from 1 to 25% ACN in 0.1% formic acid. The flow rate was 300 nL/min, and the column was heated at 60° C. Data was collected using the data-dependent acquisition (DDA) mode.
A high resolution MS1 scan in the Orbitrap (m/z range 375-1,500, 120 k resolution, Automatic Gain Control (AGC) 1×10{circumflex over ( )}6, max injection time 100 ms, RF for ion funnel 30%) was collected, from which the top 10 precursors were selected for MS2 followed by synchronous precursor selection (SPS) MS3 analysis. For MS2 spectra, ions were isolated with the quadrupole mass filter using a 0.5 m/z isolation window. The MS2 scan was performed in the quadrupole ion trap (Collision Induced Dissociation (CID), AGC 2×10{circumflex over ( )}4, normalized collision energy 30%, max injection time 35 ms) or with Higher-energy collisional dissociation (HCD) in the ion trap, AGC 2×10{circumflex over ( )}4, normalized collision energy 20%, max injection time 35 ms.
RTS and FAIMS Pro were combined with DDA. RTS utilized a UniProt mouse database. Four FAIMS Pro compensation voltages were used: −40, −50, −60, and −70 Volts. Each of the four experiments had a 1.25 seconds cycle time. A high resolution MS1 scan in the Orbitrap (m/z range 400-1,600, 120 k resolution, “Standard” AGC, “Auto” maximum injection time, ion funnel RF of 30%) was collected, from which the top 10 precursors were selected for MS2 followed by SPS MS3 analysis. For MS2 spectra, ions were isolated with the quadrupole mass filter using a 0.7 m/z isolation window. The MS2 product ion population was analyzed in the quadrupole ion trap (CID, “Standard” AGC, normalized collision energy 35%, “Auto” max injection time) and the MS3 scan was analyzed in the Orbitrap (HCD, 50 k resolution, 200% “Normalized AGC Target”, max injection time 200 ms, normalized collision energy 45%). Up to ten fragment ions from each MS2 spectrum were selected for MS3 analysis using SPS. The mass tolerance for the target masses were 10 ppm (low and high).
V. Comparison of Data Acquired from Native MP3 and Regular MP3
Method Comparison and Optimization Samples. For all methods mouse plasma from naive C57/B6 mice are used. Mouse plasma was collected in-house from C57/B6 mice.
Samples are prepared and analyzed as described above in Examples 1-6.
Using a Native-MP3 coupled with transition to label-free protocol (see Example 6), 250 μL plasma was processed using either 0.5 μm or 1 μm carboxylated, solid core magnetic particles for the 1 hour incubation. One μg of recovered peptide material was injected over a 85 minute gradient on EASY-nLC1200 with 25 cm IonOpticks Aurora column on a Lumos Fusion (no FAIMS: no RTS). Data was searched with Comet using a conservative SwissProt FASTA database with reverse hits and common contaminants, and filtered to 1% FDR at the protein level.
As shown in Table 8, using the Native-MP3 workflow, a greater number of proteins was identified with 1 μm carboxylated beads than 0.5 μm carboxylated beads.
Additionally, there was a large overlap of identified proteins was detected between Native-MP3 performed on different sized beads (
Using Native-MP3 coupled with transition to label-free protocol (see Example 6), 250 μL plasma was processed using 1 μm carboxylated, solid core magnetic particles for the 1 hour incubation, and the natively bound portion was then prepared as previously described. Unbound material (supernatant) was removed. Additionally, 100 μg samples were prepared using label-free Regular-MP3 (see Examples 1-5). For both samples (label-free MP3 and label-free Regular-MP3), 1 μg of recovered label-free peptide material was injected over a 85 minute gradient on an EASY-nLC1200 with 25 cm IonOpticks Aurora column on a Lumos Fusion (no FAIMS: no RTS). Data was searched with Comet using a conservative SwissProt only FASTA database with reverse hits and common contaminants, and filtered to 1% FDR at the protein level.
As shown in Table 9, using the single-bead Native-MP3 workflow, a greater number of proteins was identified than with the label-free Regular-MP3 workflow.
Additionally, there was a large overlap of identified proteins detected between Native-MP3 and Regular-MP3, with approximately a 2-fold larger identification rate for MP3-Native (
An experimental design was created to compare workflow methodologies for Native-MP3 and Regular-MP3 (
Using Native-MP3 coupled with transition to TMT 16-plex multiplexed AutoMP3 protocol, 250 μL plasma, in 4 replicates per Control or FAS mutant mouse plasma were processed using 1 μm carboxylated, solid core magnetic particles. Additionally, 2 μL of plasma (or approximately 120-100 μg) was processed with the same experimental setup using Regular-MP3 workflow. For both methods, 5 μg of peptide material per sample were TMT labeled.
For mass spectrometer analysis, an 8-plex was created by pooling 1 μg TMT labeled peptides for the samples from the Native-MP3 method, and similarly a separate 8-plex with the Regular-MP3 method. For each plex, 1 μg injection was injected over a 180 minute gradient on UltiMate 3000 LC with 25 cm IonOpticks Aurora column on a Orbitrap Eclipse (with FAIMS & RTS). Data was searched with Comet using a SwissProt mouse protein FASTA database with reverse hits and common contaminants, and filtered to 1% FDR at the protein level.
As shown in Table 10, the number of identified proteins using the Native-MP3 method was approximately 3× higher compared to the Regular-MP3 method. There was a large overlap of quantified proteins that were detected between Native-MP3 and Regular-MP3 (
A greater number of differentially expressed protein changes were detected for Native-MP3 (
As shown in Table 11, the number of differentially expressed proteins using Native-MP3 is approximately 2.5-fold higher compared to Regular-MP3. For differential expression changes greater than 40%, there are approximately 3-fold more proteins identified using Native-MP3 compared to Regular-MP3. At greater than 4-fold changes the number of proteins quantified are approximately the same using Native-MP3 compared to Regular-MP3. Native-MP3 allows us to quantify more differentially expressed proteins than Regular-MP3, especially those with smaller fold changes.
Similar patterns and fold changes were observed for differentially expressed proteins detected in both Native-MP3 and Regular-MP3. Examples of proteins which exhibited similar patterns and fold changes include CD5L (
Comparing quantified protein expression of all overlapping quantified proteins for WT versus mutant (FAS) mice, the majority of overlapping quantified proteins trended to have similar magnitude, between Native-MP3 and Regular-MP3, especially when the change was small (
For the label-free SEC processing, AutoMP3 was adapted for label-free batch processing, referred to as automated proteome profiling platform (AutoP3).
An exemplary AutoP3 method for the processing of 18 plasma samples, using liquid chromatography-tandem mass spectrometry (LC-MS/MS) and an exemplary method for analysis of the 18 human plasma samples that are separated into 53 SEC fractions, generating a total of 954 samples for LCMS analysis.
Exemplary HT CF-MS sample preparation processing procedure that was adapted to different numbers of samples and sample types. Label-free methods, isobaric labeling methods, and a combination of both of these methods allows for comprehensive analysis of quantifying protein complexes and native plasma proteome profiling.
It is estimated that the sample preparation time that would be needed to prepare the 954 samples manually would be around two years using standard approaches prepared sequentially. In comparison, HT label-free AutoP3 requires just 4 days to go from fractionated SEC in 96-well plates to the samples in ready-to-shoot LCMS vials.
Application of the AutoMP3 workflow to SEC samples where the sample preparation of 954 samples requires an additional 2 days in addition to the 4 days required for label-free sample preparation. However, the increased sample preparation time comes with the advantage of reduced LCMS acquisition time on the instrument.
Using a 3 h method per sample requires under a week of MS time (6.625 days) with no washes. Using a 2 h method requires just 4.4 days and a 1 h method just 2.2 days for the multiplexed analysis of 954 samples.
Plasma samples were prepared by thawing 50 μL of frozen plasma at room temperature until ice crystals were no longer visible. Each tube of sample was flicked to mix. The plasma samples were centrifuged at 20,000 RCF at 4° C. for 10 minutes to collect all residual plasma from sides of the tube and to pellet large debris. The plasma was diluted by pipetting 45 μL of the centrifuged plasma sample into 450 μL of ice-cold PBS (Corning, Product 21-040-CV). The diluted plasma sample was transferred into a 0.45 μM PVDF spin filter tube (Millipore, Product UFC30HV00) and centrifuged at 12,000 RCF at 4° C. for 4 minutes. The filtered plasma sample was collected and drawn into a 500 μL gastight syringe (Hamilton, Product 81216).
The instrument was initiated and the plasma sample was injected into the 500 μL sample loop of an ÄKTA Pure liquid chromatography system (Cytiva, Product 29018228) with two serially linked in-tandem Superose6 Increase 10/300 columns pre-equilibrated with PBS (Corning, Product 21-040-CV) with a flow of 0.3 mL/min at 4° C. After the first 14 mL of PBS passed through the column, 0.35 mL aliquot fractions were collected in a 96-deepwell plate until a total of 60 mL passed through the columns.
Subsequently, 5 mL of 1M sodium hydroxide (ThermoFisher, Product A16037-36) was run through the sample loop, into the columns and through the fractionation outlet at 0.3 mL/min to sanitize the system and prevent carryover. The cleaning was finished in place and re-equilibrated by washing the sodium hydroxide through the previous flowpath with PBS at 0.3 mL/min until a total of 60 mL have passed through the system.
To validate the integrity of the columns, 500 μL of 1.7 mg/mL Cytiva Gel Filtration Protein Standards (Cytiva, Product 28403842) was injected into the 500 μL sample loop of an ÄKTA Pure liquid chromatography system (Cytiva, Product 29018228) with two serially linked in-tandem Superose6 Increase 10/300 columns (Cytiva, Product 29091596) pre-equilibrated with PBS at 4° C.
Using 10 μL per fraction, protein concentration was determined for each SEC fraction for a representative plasma sample using BCA quantitation to assess protein abundance range across the SEC fractions and to determine starting material amounts as well as amounts for downstream steps such as protein cleanup and protease addition. The set of SEC fractions was limited for analysis to within the relevant separation molecular weight (MW) of the column (e.g., approximately 53 fractions for wells B8 through F12 for SEC column). For sample processing, 75 μL (21.4% of total volume) was used which equates to at most 75 μg of protein material. The SEC fractions were arranged from 1-53 for each sample into nine 96-well DNA LoBind Eppendorf plates such that each contains 2 plasma samples worth of fractions (e.g., rows C,D,E, and F), and a final 10th plate containing wells B8-B12 for all 18 samples.
Each of the 53 fractions was reduced by adding 6 μL 64 mM DTT and incubated at room temperature for 30 minutes with lids covered. The samples were alkylated by adding 7.125 μL 160 mM IAA per sample and incubating at room temperature for 30 minutes, lids covered, in the dark. The alkylation reaction was quenched by adding an additional 6 μL 64 mM DTT, and allowed to stand at room temperature for 5 minutes, lids covered.
Protein samples were purified using SeraMag beads 1:1 mix of hydrophobic: hydrophilic surface at a ratio of 10:1 beads: protein to ensure adequate binding. To initiate binding samples, beads were premixed for 1 minute at 1500 rpm, add 282.4 μL 100% Acetonitrile (ACN), and mixed at 1300 rpm for 1 minute, and then incubated with sample for 18 minutes to facilitate binding. Samples were immobilized with a magnet and supernatant was removed. Subsequently, samples were washed twice with 200 μL 70% ethanol. Followed by one wash with 200 μL 100% acetonitrile.
Each sample was resuspended with 90 μL digestion buffer (10 mM CaCl2, 100 mM EPPS pH 8.5). A minimum of 0.5 μg LysC/trypsin mix (Promega) was added to the wells where BCA quantitation was below the limit of detection (with the assumption of these wells containing approximately 5 μg of protein). For wells above the limit of detection, add LysC/trypsin to reach a 20:1 protein: protease ratio. All wells were normalized with an additional digestion buffer to ensure the final volume was 100 μL. Samples were mixed at 1500 rpm for 1 minute, and digested overnight at 37 C, 300 rpm, lidded to limit evaporation.
Samples were immobilized with the magnet and transfer the supernatant into a new 1.3 mL NUNC 96-well plate (ThermoFisher). Optionally, samples were subjected to peptide quantification.
For peptide cleanup, samples were mixed with beads, and bound at a 95% final concentration EtOH for 2.5 hrs (plate 1) with increasing time up to 7.5 hrs (plate 10). Once all plates were started with the incubation step, wash steps began. Samples were immobilized on a magnet for 2.5 minutes and supernatant was removed. Samples were washed 2× with 100 μL 95% EtOH, and then 1× with 100 μL 100% EtOH, then dried under nitrogen to ensure no carryover of EtOH (nitrogen flow was adjusted so as not to disturb the dried peptide-bead material), then eluted with the addition of 72.5 μL 5% ACN, 1500 rpm shake for 1 minute 2× sequentially. The eluent was saved in separate Protein LoBind Eppendorf 1 mL 96-well plates. Elution was performed twice total. The eluent plates were combined such that each sample contained 145 μL. To minimize carryover of spurious magnetic beads, an additional immobilization on the magnet was performed and the eluent was slowly aspirated and dispensed to a final 1 mL 96-well Protein LoBind Eppendorf plate, and frozen at −80° C.
The plates were aliquoted for i) non-adjusted label-free, ii) adjusted proportion label-free and iii) isobaric TMT labeling as appropriate.
For label free AutoP3 where the input amounts were not adjusted, a portion (e.g. 6% the samples) was acidified with 100% formic acid to a final concentration of 5% formic acid/5% acetonitrile and followed with LC-MS/MS analysis described herein below.
For label free adjusted input for LC-MS/MS, ˜1 μg digest per fraction was removed based on previous BCA assay and transferred to a new 96-well plate. The calculation to generate the portions was performed either through a sensitive enough quantification assay for the range of abundances concerned or dead-reckoned based on standard peptide level recovery observed (at minimum 50%) from starting protein material. For wells under limit of detection of BCA an assumption was made to use half of the available stock volume (minus some backup dead volume), i.e. 63 μL. The adjusted amount of peptide digest was dried in a speedvac and resuspended in 5% formic acid/5% acetonitrile such that the concentration was ˜1 μg/μL.
AutoMP3 was applied to SEC fractions to provide an alternative where manual preparation would be extremely time consuming.
For TMT labeling, ˜1 μg digest per fraction was removed and transferred to a 96-well plate and the samples were arranged such that the layout is the simplest for TMT labeling. For example, each 18-plex comprising all 18 samples from a specified SEC fraction were arranged together. For wells under limit of detection of BCA an assumption was made to use half of the available stock volume (minus some backup dead volume), i.e. 63 μL. The samples were dried down in 96-well plates using a speed-vac.
To prepare for TMT labeling, the TMTpro kit was taken out of −80° C. and allowed to reach room temperature. A quick spin down was performed to push reagent to bottom of the tubes and 250 μL anhydrous 100% acetonitrile was added per TMTpro channel, resulting in 20 μg/μL concentration. Each was vortexed for 10 seconds per channel, and a quick spin down was performed to collect resuspended reagent at bottom of tube. Each well was resuspended in 4.5 μL 5% Acetonitrile, 120 mM EPPS pH 8.5 and mix at 2500 rpm for 1 minute on a thermomixer to resuspend peptides.
TMT was aliquoted into ‘stock’ eighteen 1.5 mL Protein LoBind Eppendorf tubes and out of those, aliquoted into a ‘stock reagent’ 150 μL Eppendorf DNA LoBind plate in a pattern matching the 18-plex layout (up to five plexes per 96-well plate). Using 96-well pipettor stamp, the TMTpro reagent was added across the whole sample plate at a ratio of 9.6:1 w:w reagent:peptide in 5 μL 100% acetonitrile per sample, and mixed at 2350 rpm for 10 seconds to ensure mixing. The plate was sealed with aluminum foil film and incubated at room temperature for 1 hour to complete labeling. The TMT reaction was quenched with 1 μL of 5% hydroxylamine in water per sample, and the plate was mixed at 2350 rpm for 10 seconds to ensure mixing. This was followed by incubation for 15 minutes at room temperature to complete quenching.
To prepare for the peptide cleanup, 1000 μg of 1:1 hydrophobic: hydrophilic mixture (at 20 μg/μL) of SeraMag beads was pipetted into 53 wells an empty 2 mL Corning Square 96-well plate and water was removed through magnet immobilization.
Once TMT quenching was completed, pool each plex directly into a separate well containing pre-aliquoted 1000 μg of SeraMag beads (prepared in the previous step).
Once plex pooling was completed, the plate was placed under nitrogen drier for 1 minute to slightly evaporate organic solvent (acetonitrile) content in the wells samples to ensure the targeted peptide binding solvent ratio was minimally impacted. The samples were thoroughly mixed together at 1500 rpm for 1 minute, and the binding process was started through the addition of 1900 μL 100% EtOH to reach a final concentration of 95% EtOH. Samples were incubated for 2.5 hours at RT to ensure binding of peptides to beads and allow beads to settle. Samples were immobilized using the magnet and the supernatant was removed. Washed were performed 2× using 100 μL 95% EtOH. A 100 μL 100% EtOH rinse was performed 1× and then gently dried under nitrogen (approximately 3 minutes). This was followed by elution twice with the addition of 72.5 μL 5% ACN and shaken at 1500 rpm for 1 minute 2× sequentially. Eluent was saved in a separate Protein LoBind Eppendorf 1 mL 96-well plate. Total elution equaled a volume of 145 μL.
To minimize carryover of spurious magnetic beads, the combined eluent plate was magnet immobilized and slowly and carefully transferred to a final 1 mL 96-well Protein LoBind Eppendorf plate, and frozen at −80° C. until ready for LC-MS/MS analysis.
When ready for LC-MS/MS analysis, samples were resuspended in 5% acetonitrile/5% formic acid at a concentration targeting injecting ˜1 μg of material in 2 μL of volume, and load into the autosampler.
Peptides were analyzed using an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) equipped with a nano-electrospray ion source and coupled to an Easy-nLC1200 (Thermo Fisher Scientific). For label free LC-MS/MS analysis, one or more mass spectrometers was selected (e.g., Orbitrap, QTOF, ion trap, FTICR). Chromatographic separation of peptides was performed using an IonOpticks Aurora microcapillary column (75 μm inner diameter, 15 cm long, C18 resin, 1.6 μm, 120 Å). Total run length was 30 minutes, with 27-minute gradient from 5 to 30% ACN in 0.125% formic acid. The flow rate is 400 nL/min, and the column was heated at 60° C.
Data was acquired using data-dependent acquisition (DDA) mode. A high resolution MS1 scan in the Orbitrap (m/z range 375-1,500, 120 k resolution, Automatic Gain Control (AGC) 1×10{circumflex over ( )}6, max injection time 100 ms, RF for ion funnel 30%) was collected. For MS2 spectra, ions were isolated with the quadrupole mass filter using a 0.5 m/z isolation window. The MS2 scan was performed in the quadrupole ion trap (CID, AGC 2×10{circumflex over ( )}4, normalized collision energy 30%, max injection time 35 ms or HCD, AGC 2×10{circumflex over ( )}4, normalized collision energy 20%, max injection time 35 ms).
Peptides were analyzed on an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) coupled to an Easy-nLC1200 (Thermo Fisher Scientific). For label free LC-MS/MS analysis additional mass spectrometers include Orbitrap, QTOF, ion trap, and FTICR. Chromatographic separation of peptides was performed using an IonOpticks Aurora microcapillary column (75 μm inner diameter, 25 cm long, C18 resin, 1.6 μm, 120 Å). The total LC-MS run length for each sample was 90 min including a 75 min gradient from 8 to 30% ACN in 0.125% formic acid. The flow rate was 300 nL/min, and the column was heated at 60° C.
Data was acquired on an Orbitrap Fusion Lumos using the data-dependent acquisition (DDA) mode. A high resolution MS1 scan in the Orbitrap (m/z range 500-1,200, 60 k resolution, Automatic Gain Control (AGC) 5×10{circumflex over ( )}5, max injection time 100 ms, RF for ion funnel 30%) is obtained, from which the top 10 precursors were selected for MS2 followed by synchronous precursor selection (SPS) MS3 analysis. For MS2 spectra, ions were isolated with the quadrupole mass filter using a 0.5 m/z isolation window. The MS2 scan was performed in the quadrupole ion trap (Collision Induced Dissociation (CID), AGC 1×10{circumflex over ( )}4, normalized collision energy 34%, max injection time 35 ms) and the MS3 scan was analyzed in the Orbitrap (Higher-energy collisional dissociation (HCD), 60 k resolution, max AGC 5×10{circumflex over ( )}4, max injection time 250 ms, normalized collision energy 45). Up to six fragment ions from each MS2 spectrum were selected for MS3 analysis using SPS. If available, the use of an Eclipse mass spectrometer with FAIMS-RTS (High Field Asymmetric Waveform Ion Mobility Spectrometry with real-time search) to increase depth of proteins identified was preferred.
Mass spectra were interpreted with Proteome Discoverer v2.4 (Thermo Fisher Scientific, San Jose, CA). In brief, the parent mass error tolerance was set to 10 ppm and the fragment mass error tolerance to 0.6 Da. Strict trypsin specificity was required allowing for up to two missed cleavages. Carbamidomethylation of cysteine (+57.021242 Da) was set as a static modification while methionine oxidation (+15.995 Da) was set to a variable modification. In addition, N-terminal protein acetylation (+42.011 Da), was set as a variable modification. The minimum required peptide length was set to seven amino acids. Spectra are queried against a “target-decoy” protein sequence database consisting of human proteins, common contaminants, and reversed decoys with the SEQUEST algorithm. The Percolator algorithm (Käll et al. Nat Methods 2007, 4, 923-925) was used to estimate and remove false positive identifications to achieve a strict false discovery rate of 1% at both peptide and protein levels. For the peak and feature detection the minimum number of non-zero points that must exist in a chromatographic trace (trace length) was set to 5, the max ΔRT of isotope pattern multiplets was set to 0.2 (min) and feature to id linking peptide-spectrum match (PSM) confidence is set to high. In order to detect changes in the proportions of protein complexes, the intensity vectors were first normalized by sum to a total signal of 1. To minimize the impact of missing data and noisy measurements a filtered dataset was created by removing proteins with total signal less than 9×107 (approximately the median total signal). Additionally proteins were removed that show little overlap between the samples by requiring a correlation between the vectors of at least 0.2. For each protein, vectors of intensities from each sample were converted to cumulative sums and euclidean distances between the samples were calculated
Mass spectrometry data are processed using an in-house software pipeline [19]. Raw files are converted to mzXML files and searched against a human UniProt containing sequences in forward and reverse orientations using the Comet algorithm. Database searching matched MS/MS spectra with fully tryptic peptides from this composite dataset with a precursor ion tolerance of 20 p.p.m. and a product ion tolerance of 0.6 Da. Carbamidomethylation of cysteine residues (+57.02 Da) and TMTpro tags on peptide N-termini and lysines (+304.20 Da) are set as static modifications. Oxidation of methionine (+15.99 Da) is set as a variable modification. Linear discriminant analysis is used to filter peptide spectral matches to a 1% FDR (false discovery rate) (see eg., Huttlin et al. Cell 2010, 143, 1174-1189). Non-unique peptides that are matched to multiple proteins are assigned to proteins that contain the largest number of matched redundant peptide sequences using the principle of Occam's razor. Quantification of TMTpro reporter ion intensities is performed by extracting the most intense ion within a 0.003 m/z window at the predicted m/z value for each reporter ion. TMT spectra were used for quantification when the sum of the signal-to-noise for all the reporter ions was greater than 200 (Ting et al. Nat Methods 2011, 8, 937-940).
Size exclusion chromatography tandem Superose 6 Increase 10/300 GL setup is calibrated using a standards kit containing Blue Dextran (2,000 kDa, void volume estimation only), Thyroglobulin (669 kDa), Ferritin (440 kDa), Aldolase (158 kDa), Conalbumin (75 kDa), Ovalbumin (43 kDa). Molecular weight estimation is then calculated per well fraction using the formula: Kav=(Ve−Vo)/(Vc−Vo) where Ve=elution volume of target peak, Vo=column void volume (˜16.8 mL based on Blue Dextran elution volume peak), Vc=geometric column volume (48 mL). Thyroglobulin is excluded from calculation as indicated due to known deviation from Log (MW) to Kav linearity in this column. The estimated standard curve fit had an R{circumflex over ( )}2=0.993. At Kav=0.1 and Kav=0.9 are upper and lower MW bounds, respectively, within which Log (MW) to Kav linear trend is expected to approximately hold true, and outside of which (Kav>0.9 and Kav<0.1) is not expected to hold true. Linear fit equation generated from the four standards is used for molecular weight estimation: y=−0.06132181*x+0.9769997, where y is Kav of given fraction at the heaviest MW volume elution range of the fraction (fraction 1 began elution at 14 mL, therefore Kav at 14 mL was used), and x is log 2 (kDa)
Algorithms for correlation profiling to infer protein complexes that can be used include Pearson correlation, Spearman correlation, Kendall correlation, Euclidean distance, Co-Apex (Heusel et al. Mol Syst Biol 2019 15, e8438), Bray-Curtis similarity or a mix of the above. A number of open source software toolkits/workflows are available for CF-MS inference of protein complexes. These include EPIC (elution profile-based inference of complexes) (Hu et al. Nat Methods 2019, 16, 737-742) which enables the automated scoring of CF-MS data using supervised machine learning to create probabilistic protein-protein interaction networks followed by clustering to define protein complexes that have high confidence. Other software includes PrinCE, CCprofiler, EGAD, or CEDAR.
Total peptides and unique peptides were quantified for exemplary peptide clean-up methods for the processing of human cell lysate and mouse plasma samples.
Total peptides and unique peptides were quantified for exemplary peptide clean-up methods of 10 μg peptide samples mixed with beads, and bound with various solvents, including final concentrations of 95% IPA, 80% EtOH, 95% EtOH, 95% MeOH, a mixture of 45% ACN and 50% MeOH, a mixture of 50% ACN and 45% MeOH, and a mixture of 40% ACN and 40% MeOH (
Total peptides and unique peptides were quantified for 1 μg human cell lysate samples, 1 μg mouse plasma samples, 0.1 μg mouse plasma samples, and/or 0.2 μg mouse plasma samples subjected to various exemplary peptide clean-up methods comprising various solvents or solvent combinations for binding, with or without SeraMag beads, followed by an optional wash comprising various solvents or solvent combinations, followed by an optional rinse comprising various solvents (
The entire disclosure of each of the patent documents and scientific articles cited herein is incorporated by reference for all purposes.
The disclosure can be embodied in other specific forms with departing from the essential characteristics thereof. The foregoing embodiments therefore are to be considered illustrative rather than limiting on the disclosure described herein. The scope of the disclosure is indicated by the appended claims rather than by the foregoing description, and all changes that come within the meaning and range of equivalency of the claims are intended to be embraced therein.
This application is a continuation of International Application No. PCT/US2023/025485, filed on Jun. 15, 2023, which claims priority to and the benefit of U.S. Provisional Patent Application No. 63/366,448, filed on Jun. 15, 2022, the entire contents of each of which are incorporated by reference herein for all purposes.
| Number | Date | Country | |
|---|---|---|---|
| 63366448 | Jun 2022 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/US2023/025485 | Jun 2023 | WO |
| Child | 18981314 | US |
