Identifying and measuring endogenous peptides through peptidomics

Abstract

Endogenous peptides, such as neuropeptides and peptide hormones, play important roles in intercellular communication, can provide information on physiology, and are potential sources of biomarkers. Mass spectrometry-based peptidomics methods are underutilized tools to identify and measure endogenous peptides in a relatively non-targeted manner. The purpose of this Viewpoint is to serve as a brief introduction to the field of peptidomics so that researchers interested in studying endogenous peptides are aware of this powerful approach and can consider its application.

Introduction

When introduced to neuropeptides and peptide hormone research, some people are not used to thinking about small endogenous peptides as bioactive molecules. A number of researchers are often surprised to hear that peptide sequences as short as three amino acid residues (e.g., thyrotropin-releasing hormone) are synthesized in a regulated manner and engage in specific peptide-receptor interactions to enact physiological functions. Neuroscientists, of course, are familiar with peptide neurotransmitters and hormones, but often are not aware of how mass spectrometry can be used to gain insight into these molecules in living systems. The goal of this Viewpoint is to give a brief overview of mass spectrometry-based “peptidomics”, a collection of non-targeted approaches to identify and measure endogenous peptides isolated from biological sources. Several such studies have recently been published in ACS Chemical Neuroscience,^1–4 and this Viewpoint may serve as a useful introduction to peptidomics to this community. As mass spectrometry becomes more accessible (especially through core facilities), so too do capabilities to perform peptidomics experiments. This overview is written especially for researchers who are not familiar with mass spectrometry, with the hopes that these scientists may consider pursuing such experiments to answer their research questions.

Why study endogenous peptides?

In this Viewpoint, the term “endogenous peptide” refers to a peptide sequence shorter than most traditional proteins (e.g., <100 amino acid residues) that is naturally generated by cells or tissues (Figure 1). Endogenous peptides in animals are typically synthesized by the ribosome first as larger proteins, and then processed into smaller peptides through peptidase-mediated amide bond cleavages. However, mature peptides can also be directly translated from small open-reading frames (<100 codons). For neuropeptides and peptide hormones, processing occurs as the protein precursors (termed “prohormones/propeptides”) travel through the secretory pathway in large dense-core vesicles. These peptides are processed by highly specific processing enzymes to generate the final mature peptide molecules which are released from the cell in a regulated, stimulation-dependent manner. In other cases, general (and often non-specific) protease-mediated degradation of larger non-prohormone proteins can give rise to peptides (“protein fragments”).

Figure 1.

Peptidomics approaches identify and measure mature endogenous peptides in a biological sample. Cartoon representation of the typical biosynthetic mechanism for endogenous peptides, which are genetically encoded, transcribed, and translated into precursor proteins. These precursor proteins then undergo extensive post-translational modifications to give rise to the mature endogenous peptides. The sequences and relative abundances of these endogenous peptides cannot be determined from genomics, transcriptomics, or proteomics methods. Antibody-based measurement methods require preselection of peptides of interest, and often fail to differentiate between peptides with very similar sequences (examples shown in red text).

Identifying and quantifying endogenous peptides are important goals because these molecules play important roles in a variety of biological contexts. Neuropeptides and peptide hormones are secreted from cells in a controlled manner, travel through extracellular space, and engage specific receptor proteins to facilitate cell-to-cell communication. These prohormone-derived peptides play critical roles in diverse biological responses, including feeding behavior, metabolism, stress response, pain perception, thermal regulation, and many more. In some cases, “protein fragments” (peptides derived from non-prohormone proteins) can also act as direct modulators of cell signaling, enzyme inhibitors, and more. Finally, even the study of endogenous peptides without known functions can be valuable since the abundance of specific endogenous peptides can provide insight into changes in physiology (e.g., progression of disease or identification of biomarkers).

Typical methods to measure peptides

Two common methods used to measure peptides are transcriptomic methods and antibody-based methods (e.g., immunoassay, ELISA, etc.). Although transcript measurements can be informative, these approaches cannot give information on the translation of mRNA into protein (transcript levels do not necessarily correlate with protein levels), nor can they give information on the post-translational processing of the protein precursors to generate the final peptide products. As a result, transcript measurements cannot reliably be used to measure the mature peptides in most cases. Antibody-based assays can quantitatively measure the final peptide products, but require both the preselection of specific peptides of interest and obtaining reliable and highly selective antibodies to those peptides. Thus, these approaches must be targeted to only a small number of peptides. Highly similar peptides, such as acetylated and non-acetylated forms of the same sequence or two sequences differing by one or more residue at one terminus (Figure 1), are likely to cross react with antibodies, making it nearly impossible to gain insight into the mature forms of the peptides measured.

What is peptidomics and how does it work?

Most “peptidomics” approaches utilize mass spectrometry to identify peptides present in a complex mixture (e.g., tissue extracts) in a non-targeted manner (Figure 2).^5–7 Broadly, peptidomics methods measure the molecular masses of ionized molecules present in a mixture (these are called the “parent ions”, at the “MS1” level). Measured molecules are then subjected to fragmentation, often by collision with an inert gas, which fractures peptides by cleaving along the amide backbone. The masses of these fragments can then also be measured by the mass spectrometer (“fragment ions”, produced from “tandem MS”, also referred to as the “MS2” or “MS/MS” level). The fragmentation data can then be interpreted with the help of specialized software (e.g., PEAKS, Mascot, etc.) to determine the sequences of the peptides analyzed. While peptide sequences can be determined de novo using certain software with high-quality data, peptidomics experiments often search against a protein database to narrow potential sequence space. This protein database is often a file containing the sequences of predicted proteins derived from transcriptome data. An important note is that this database need only contain predicted protein sequences – it does not need to include all possible peptide sequences or post-translational processing. For commonly studied animals such as humans, rats, and mice, well-curated protein databases are readily available from public sources such as UniProt. For less commonly studied species, predicted protein databases derived from transcriptome data are often available, although these databases are usually not as well curated as the more common species and some effort may be needed to build a usable database. To aid in measurements, peptides are often separated prior to mass spectrometry analysis, usually using liquid chromatography connected directly to the mass spectrometer (LC-MS).

Figure 2.

Simplified representation of peptide identification by tandem mass spectrometry. At the MS1 level, the mass-to-charge ratio of the full-length ionized peptide is measured by the mass spectrometer. The peptide is then fragmented in the instrument, and the resulting fragment ions are measured at the MS2 level. Analysis of fragment masses (aided by software) allow for the determination of peptide sequence. Because of the high mass resolution and mass accuracy of the instrument (often to several digits past the decimal), most modifications (e.g., represented by the green circle) can be confidently identified and localized.

In contrast to transcriptomics and antibody-based methods, peptidomics experiments measure the mature peptide products that are present and can easily differentiate between highly similar sequences. For example, modern mass spectrometers can reliably differentiate between a C-terminal amidated and non-amidated forms of a peptide, which differ by only 0.98 Da! Although there are certain modifications that are difficult to detect via mass spectrometry, such as isomerization or relatively labile sulfation, differences in amino acid sequence (e.g., truncation of residues from either terminus) and most post-translational modifications are readily identifiable. While identification of peptides is the most direct result obtainable from a proteomics experiment, relative quantification of peptides between different groups can be achieved either using covalent labeling with isotopic mass tags or using “label-free” approaches.⁷

Peptidomics experiments are typically most straightforward from tissues that generate relatively large amounts of endogenous peptides, including peptide-rich CNS regions such as the hypothalamus and the pituitary. However, experiments can also be performed from diverse matrices such as cerebral spinal fluid, urine, milk, and more.^{6, 7} Measuring peptides released from cells into media or in serum can be challenging due to low relative peptide concentrations, although can be feasible with advanced methods.⁷

How does one perform a peptidomics experiment?

In practice, peptidomics experiments are similar to mass spectrometry-based proteomics experiments that are routinely carried out in proteomics facilities at many institutions. Peptidomics approaches make use of the same LC-MS instrumentation used in proteomics workflows, so most proteomics-capable facilities can perform peptidomics experiments, although some important adjustments must be made. One important factor that must be considered is the rapid rate of post-mortem peptide degradation that occurs in biological tissues by proteases and peptidases, which must be halted through rapid tissue processing or tissue “stabilization” (often in the form of heating to denature proteases). Other major differences between peptidomics and proteomics arise because the major goal of peptidomics is to measure small, endogenous peptides and not larger molecular-weight proteins. Typical proteomics procedures utilize some steps that specifically remove small endogenous peptides, including analyte extraction conditions and the inclusion of precipitation or filtration steps to isolate larger proteins. Importantly, common “bottom-up” proteomics workflows utilize enzymatic digestion (typically trypsin) to cleave larger proteins into smaller peptides within the mass range most readily analyzed by LC-MS instrumentation. Such enzymatic digestion steps are not included in peptidomics workflows since they will destroy any peptides that contain enzymatic cleavage sites, leading to the loss of information about the endogenous peptides originally present in the sample.

The major deviations from a typical proteomics workflow that should be followed when adapting for peptidomics include:

1. Tissues must be “stabilized” (e.g., by heat treatment) or processed as quickly as possible (within minutes of death) to minimize post-mortem peptide degradation.

2. Peptide extraction conditions from tissue often differ from typical protein extraction conditions.

3. Some steps typically used to isolate proteins, such as protein precipitation or molecular weight-based filtration, should be adapted, modified, or removed to ensure isolation of smaller peptides rather than larger proteins.

4. No enzymatic digestion is used in peptidomics to conserve the endogneous sequences present in the sample.

5. Database search parameters using proteomics software should be adjusted to ensure detection and quantification (if appropriate) of individual peptides, not larger proteins. Most software is designed for proteomics, so result outputs are often designed with only the protein-level in mind.

Those wishing to perform a peptidomics experiment should discuss these aspects with their local proteomics facility, who may be able to adapt their workflows to enable identification of endogenous peptides by examining prior published peptidomics protocols.

Conclusions

The goal of this Viewpoint was to give a “big picture” overview to introduce this topic to new researchers who have never before considered peptidomics experiments. To learn more, recent technical protocols and reviews are available.^5–7 Those interested in endogenous peptides are encouraged to reach out to local proteomics facilities, or contact a researcher in the field to propose a collaboration.