Monitoring the synthesis of biomolecules using mass spectrometry

Abstract

The controlled and selective synthesis/clearance of biomolecules is critical for most cellular processes. In most high-throughput ‘omics’ studies, we measure the static quantities of only one class of biomolecules (e.g. DNA, mRNA, proteins or metabolites). It is, however, important to recognize that biological systems are highly dynamic in which biomolecules are continuously renewed and different classes of biomolecules interact and affect each other's production/clearance. Therefore, it is necessary to measure the turnover of diverse classes of biomolecules to understand the dynamic nature of biological systems. Herein, we explain why the kinetic analysis of a diverse range of biomolecules is important and how such an analysis can be done. We argue that heavy water (²H₂O) could be a universal tracer for monitoring the synthesis of biomolecules on a global scale.

This article is part of the themed issue ‘Quantitative mass spectrometry’.

1. Introduction

Biological systems are highly complex, dynamic and to some degree unpredictable. The human body is made up of 37 trillion cells [1]. Each cell contains many organic (e.g. DNA, proteins and lipids) and inorganic (e.g. water, oxygen and sodium chloride) molecules. The numbers of these molecules in a cell are incredibly large, for example 6 billion protein molecules per human cell [2]. The status of a cell never becomes truly a static state nor a steady state, as the cellular status continually changes in response to external factors such as the circadian cycle, temperature and nutrient abundance/depletion. Furthermore, these biological systems behave in a nonlinear fashion; a small insult or change may result in a large later effect, for example a single mutation in a gene often causes a life-threatening disease in the organism. Every molecule in a cell is connected to everything else directly or indirectly, and emergent properties arise out of the interactions of these individual molecules. This connectedness necessitates studying the cell as a whole, rather than individual components within the cell in isolation. While it would be wonderful if we could quantitatively monitor all the biomolecules in a living organism, it is not currently feasible to do so due to technological limitations. Instead, what we have are high-throughput holistic approaches for genomics, transcriptomics, proteomics and metabolomics. These techniques disrupt cells and isolate and analyse a single class of biomolecules. In their current incarnation, these techniques collectively cover a diverse range of biomolecules (DNA, RNA, proteins and metabolites) in a cell and provide data on the static quantities of these molecules.

Although the static levels of individual biological molecules are useful, information on the synthesis and clearance rates of these biomolecules is needed to understand the underlying dynamic nature of physiological processes. This is because the static levels of biomolecules are the net result of their production and clearance regulated by the activities of proteins involved in their synthesis, degradation and transportation. The synthesis/clearance of biomolecules is, however, not directly observable. Such an analysis is commonly done by administering a stable isotope-labelled precursor molecule and monitoring the kinetics of isotope transfer into the product molecules. Mass spectrometry (MS) has been the most widely used technique for this purpose because it is capable of tracking individual molecules in a complex mixture and offers high detection sensitivity.

Different classes of biomolecules interact and affect each other's production and clearance. For instance, there are proteins that interact with DNA and mRNA molecules to modulate the production of proteins. Many enzymes are involved in the synthesis and clearance of various metabolites, and some metabolites are in turn coupled to the synthesis of DNA, mRNA and protein molecules as the building blocks of these molecules. It is, therefore, desirable to measure the turnover rates of all classes of biomolecules to obtain an understanding of the interplay between different classes of molecules. Despite the informational content of such a comprehensive measurement, it is rare to measure the turnover rates of more than one class of molecules in an ‘omics’-scale study. This is probably because the analytical techniques used for different types of molecules are quite distinct and require unique expertise for each class of molecules. We expect, however, that soon comprehensive kinetic analysis on multiple classes of biomolecules will become much more routine as the key technologies for the analysis are already available.

2. Measuring the synthesis of biomolecules

In contrast with static measurements of concentrations of biomolecules, determining the rate of synthesis requires the measurement of temporal changes of newly synthesized molecules. This can be done by administering a stable isotope-labelled precursor and monitoring the time-dependent changes of isotope labelling of the products, whereas the rate of clearance can be determined by monitoring the decay of labelled products following the elimination of the labelled precursor. The stable isotopes commonly used are the isotopes of hydrogen (²H), carbon (¹³C) and nitrogen (¹⁵N). This technique is readily applicable to both in vitro and in vivo kinetics studies and has become a powerful approach with the advancement of high-resolution MS as a detection tool for labelled products. Note that nuclear magnetic resonance (NMR) spectroscopy can also be used as the detection tool, but the throughput and sample requirements are inferior to MS, which offer a femtomole detection limit.

Because of the resistance of C–C covalent bonds to non-enzymatic transformations, ¹³C-labelled tracers (e.g. ¹³C₆-glucose) are extensively used in metabolic flux studies. In this approach, ¹³C-enrichment and the positions of ¹³C-labelling in metabolites, knowledge of the metabolite concentrations, along with the knowledge of the metabolic network are combined to calculate the intracellular flux of the metabolites [3,4]. Similar stable isotope-based tracer experiments have been carried out to study the metabolism of the tricarboxylic acid cycle [5], fatty acid synthesis/oxidation [6,7], glucose production/utilization [8,9], protein synthesis/breakdown [10,11] and DNA synthesis [12]. These studies demonstrate that ¹³C-labelled tracers are useful in measuring the synthesis of biomolecules, especially the flux of metabolites in well-characterized metabolic pathways. However, ¹³C-labelled tracers are relatively expensive, and administration of ¹³C-tracers (e.g. ¹³C₆-glucose) may alter metabolic pathways. ²H₂O is an alternative to ¹³C-tracers and has a great potential to be a universal tracer in monitoring the biosynthesis of various classes of molecules for the reasons described below.

3. Can ²H₂O be a universal tracer?

The ubiquitous presence of hydrogen atoms in biomolecules allows utilization of ²H₂O as a unique tracer for monitoring the synthesis of virtually all small and large biomolecules. ²H₂O has been used to study the synthesis, interconversion and degradation of biomolecules since the mid-1930s [13]. Deuterium atoms incorporated metabolically in biomolecules and bonded to carbon atoms are chemically stable, and no non-enzymatic exchange occurs under physiological conditions. While hydrogens in biomolecules that are bonded to heteroatoms such as oxygen, nitrogen and sulfur are non-enzymatically exchanged with deuterons, these deuterons are exchanged back to protons almost instantaneously when they are exposed to a standard medium (H₂O) [14]; therefore, these deuterons are not measured by MS analysis. A low dose of ²H₂O (less than 2%) is safe for consumption [15] and is relatively inexpensive, thus making it a useful tracer for human applications [16]. A label-chase experiment using ²H₂O is simple. ²H₂O can be given to free-living organisms by multiple oral doses over the course of a study in their drinking water. ²H₂O freely and rapidly equilibrates with total body water in all organs, organelles and cells within an organism, and labels the metabolic precursors of all biopolymers, i.e. amino acids for proteins, acetyl-CoA/NADPH for fatty acids and cholesterol, deoxyribose for DNA, and phosphoenolpyruvate for glucose (figure 1) [17,18]. Because it is easy to maintain the constant labelling of ‘body water’ over an extended period, it is possible to study molecules that have slow turnover rates. Furthermore, the easy access to ‘body water’ for ²H₂O enrichment assays eliminates the difficulty of measuring isotope enrichment in the precursor molecules, which is critical for data interpretation. Thus, ²H₂O offers many advantageous properties to be used as a ‘universal tracer’. The biosynthesis of various molecules, including proteins [17,19–21], carbohydrates [22], lipids [23,24] and DNA [25], has been monitored using ²H₂O as a tracer. These studies underscore the applicability of ²H₂O to measure the synthesis of various classes of biomolecules.

Figure 1.

²H₂O as a unique tracer for different classes of biomolecules. The figure illustrates how the deuterium atom(s) from ²H₂O is rapidly incorporated into metabolic precursors of different classes of biomolecules and then slowly incorporated into their products. Note that the use of ²H₂O is not limited to these molecules.

There are also limitations of ²H₂O as a tracer that should be noted. Firstly, it is known that enzymatic reactions that involve deuterium transfer in their rate-determining steps are several times slower than the corresponding reactions in which hydrogens are transferred [26]. The deuterium atom that is transferred could be of solvent (²H₂O) or the one already incorporated in one of the catalytic residues of the enzyme or its substrate (note that as mentioned above deuterium atoms can be incorporated into various places of biomolecules enzymatically as well as non-enzymatically). This phenomenon is known as the solvent kinetic isotope effect. Therefore, a high concentration of ²H₂O impacts the metabolism of biomolecules significantly. To minimize the kinetic isotope effect of deuterons, ²H₂O content in the water administered needs to be low (less than 2% in humans). For this reason and also because high doses of ²H₂O are toxic in mammals [27], ²H₂O can only be used in a small amount. However, this is not a serious limitation that prevents the use of ²H₂O because current MS technology allows accurate quantification of the deuterium enrichment in biomolecules with body water enrichment of less than 0.5% [28]. Secondly, assignment of hydrogens in the MS/MS spectra of biomolecules is more challenging than that of carbons due to the complicated gas-phase fragmentation [29]. Therefore, it is difficult to determine the positional information of deuterons incorporated in biomolecules. The positional information is required to determine the isotope enrichment patterns [30]. Thirdly, one of the key assumptions in the ²H₂O tracer experiment is that the labelling of the immediate precursors of the biomolecules being measured is much faster than the synthesis of their products. Therefore, ²H₂O tracer cannot be used for molecules whose synthesis rates are faster than their precursor labelling.

4. Bottlenecks of ²H₂O-based global kinetic analysis

As in all other stable isotope-based turnover experiments, ²H₂O-based approaches require accurate and reproducible measurements of isotope enrichment in the products for which the baseline separation between adjacent isotopic peaks is required. This is particularly critical when one aims to quantify subtle changes in the synthesis rate of molecules due to a disease or therapeutic intervention. Fortunately, modern high-resolution mass spectrometers such as time-of-flight (TOF) and Fourier transform ion cyclotron (FT-ICR) instruments have the necessary resolution and provide sufficient accuracy to determine only slight percentage differences in isotope enrichment among samples [31]. Because of its utility and ubiquitous adoption across biology and chemistry, MS will continue to be the key technique used in kinetic analysis unless more powerful technology emerges.

Although powerful analytical technologies for the detection of biomolecules are already in place, there is one major technical hindrance that must be solved before this technique can be widely used for monitoring the biosynthesis of a diverse range of molecules. The problem is the lack of robust computational tools that can calculate the synthesis rates of biomolecules from the MS data obtained over a time-course experiment. This requires the following steps: (i) identification of biomolecules from the acquired MS or MS/MS spectra, (ii) calculation of the extent of deuterium enrichment for the identified molecules from their MS spectra, and (iii) plotting the deuterium enrichment for each molecule as a function of time and calculating the synthesis rates. In the first step, experimentally obtained MS or MS/MS spectra are compared with theoretical or actual spectra of molecules stored in a database to identify the molecules that produced the experimental spectra. For this, various database search engines are already available for different classes of biomolecules, e.g. Mascot database search software (Matrix Science, London, UK) for proteins and METLIN (http://metlin.scripps.edu) for metabolites. In the second step, the extent of deuterium enrichment is calculated from the acquired MS spectra. Labelling with heavy water with low concentration primarily affects the relative isotopomer distribution without a measurable mass shift. This partial labelling of biomolecules with overlapping isotope profiles of labelled and unlabelled species complicates the isotope enrichment calculation, requiring knowledge of the elemental compositions of identified molecules to correct for the basal heavy isotope enrichment levels. There are software tools that carry out semi-automatic quantification of peptide isotope enrichment for global proteome dynamics studies [32,33]. However, these software tools are still in the early stages of development. The third step is straightforward as long as the biological systems being studied can be assumed to be in a steady state. Measurements of the changes in quantities of biomolecules and complex modelling of the data are required to calculate the synthesis rates in a non-steady-state experiment. Software capable of performing automated analysis of MS data that carries out all these steps successively without significant human intervention is not currently available.

The other complication preventing the carrying out of global kinetic analysis may be the scientists ourselves; various fields of science have been separated depending on the molecules studied, with each field employing different underlying technology to derive the data. Scientists in the field of proteomics and metabolomics are trained differently and attend different scientific conferences. Although MS is the common technique used regardless of what class of biomolecules is analysed, sample preparation, chromatographic separation techniques and MS data analysis techniques are different depending on the classes of molecules analysed. Nevertheless, all the biomolecules in a sample subjected to a ²H₂O-metabolic labelling experiment incorporate deuterium. If one wishes, the kinetic information of many of these molecules can be obtained. It seems a waste of an opportunity when only one class of molecules is analysed. To realize this, scientists with various ‘omics’ expertise need to work together. The most realistic way would be by dividing deuterium-labelled samples into several aliquots and each aliquot being subjected to the analysis optimum for one of the classes of biomolecules unless efficient serial extraction of different classes of molecules becomes available.

5. Future applications of ²H₂O-based global kinetic analysis

A major challenge remaining in current ‘omics’ research is the integration of the quantitative information obtained by the different ‘omics’ approaches to arrive at a holistic understanding of cellular function. Integrating the kinetic information on the synthesis of various molecules using ²H₂O tracer should be more reliable because all the kinetic measurements are based on the incorporation of deuterium atoms from the single tracer. For this reason, ²H₂O-based kinetic analysis has a potential to be adopted for routine in vivo kinetic analysis on a global scale to measure the synthesis rates of biomolecules. ¹³C-labelled tracers can then be used for detailed flux analysis of individual biomolecules in a more focused and targeted manner.

Static quantitative metabolomics, lipidomics and proteomics are emerging as diagnostic tools in clinical research and patient care. It is reasonable to assume that ²H₂O-based kinetic analysis will complement the traditional static ‘omics’ measures and will be applied to clinical diagnostics in the near future, paving the way for kinetic-based descriptors for decision making in personalized medicine. The most realistic clinical application of the ²H₂O-labelling method would be its utilization for the assessment of the dynamics of metabolites and proteins in biofluids. Although the high sensitivity of existing mass spectrometers makes it possible to monitor the biosynthesis of various molecules in small tissue biopsy samples, the ability to perform non-invasive, more sophisticated ‘virtual biopsies’ based on the kinetic measurement of tissue-derived molecules in circulation may become a method for evaluating tissue damage and other disease pathologies [34]. However, the application of ²H₂O-based kinetic measures to clinical applications is still in its infancy and will require the continued development of the underlying technologies including automated sample processing, MS detection and data analysis before it can be widely adopted.

Authors' contributions

M.M. initiated the plan for the article and its overall structure. T.K. led the review of the methods used for monitoring the turnover of biomolecules. Both the authors contributed equally to writing the article.

Competing interests

We declare we have no competing interests.

Funding

T. K. is supported by NIH grant no. RO1GM112044.