The protein pool in the human body was previously believed to be immobile until the innovative work of Schoenheimer et al. [1] in 1930s, the pioneer of metabolic isotope tracer technology, who first investigated in vivo protein kinetics using ¹⁵N labeled amino acids (AAs) and demonstrated the dynamic nature of body proteins (i.e., protein synthesis and protein breakdown). Entitled “The Dynamic State of Body Constituents”, the dynamic states of living systems were eloquently described in his posthumously assembled book in 1946 as “all constituents of living matter are in a steady state of rapid flux” [2]. Every molecule in the body turns over at various rates to accomplish “dynamic” homeostasis in the body. The protein pool size can change when an imbalance between rates of protein synthesis and breakdown occurs in normal or pathological processes over time. For example, loss of muscle mass can occur if the rate of protein breakdown is greater than protein synthesis, even if an increased rate of protein synthesis exists compared with the normal condition as long as the rate of protein breakdown is accelerated to a greater extent, as observed in catabolic muscle wasting conditions [3]. This underscores the importance of simultaneous assessment and understanding the rates of protein synthesis and breakdown. Although elucidation of molecular signaling implicated in protein kinetics is invaluable but limited to “snap-shot” information, the stable isotope tracer methodology can provide information on protein dynamics (kinetics), which are often not reflected in static information such as mRNA abundance, signaling, and protein abundance [4,5]. Therefore, assessing protein kinetics on both sides of the protein balance equation is important to better understand the dynamic nature of protein metabolism in conjunction with assessing the molecular signaling. In this review, the dynamic nature of protein metabolism, with an emphasis on skeletal muscle protein dynamics, which is important to understand normal and pathological muscle wasting conditions such as sarcopenia and cachexia, is presented. Then, various stable isotope tracer methodologies that enable assessment of in vivo kinetics of protein dynamics at the muscle and whole-body levels are reviewed using representative AA tracer infusion or deuterium oxide (D₂O), each with unique advantages and disadvantages.

The protein pool size or mass in the body is controlled by the balance between protein synthesis and breakdown rates in the different levels of organisms: specific individual proteins (e.g., titin, myosin light chain, creatine kinase) [6-9], subcellular organelles (e.g., contractile and mitochondria proteins) [10,11], tissue/organ (e.g., muscle, adipose tissue, liver) [12-14], or whole systems [15-18]. For example, in healthy adults, muscle mass is relatively well maintained due to the direct result of the close relationship between protein synthesis and breakdown rates. In anabolic states such as growing children or body builders, muscle mass increases because the rate of protein synthesis is greater than protein breakdown. Conversely, in catabolic states such as cancer cachexia, heart failure, and sarcopenia, the breakdown of protein is greater than synthesis. Furthermore, achieving similar protein pool sizes (or muscle mass) in states of different turnover is possible but can lead to changes in the quality of the muscle because muscle quality (i.e., muscle strength normalized to mass [19]) is positively associated with muscle protein turnover rates [20]. This indicates that rates of protein turnover within a normal physiological range ensures a constant provision of new “functional” proteins (i.e., protein synthesis) to replace old, nonfunctional, or damaged proteins that must be cleared via the process of protein breakdown. Despite the dynamic nature of muscle proteins in vivo, researchers have extensively relied on assessments of static information, termed “statomics”, such as abundance of metabolites, mRNA, and proteins, or the (in)activity of molecular signaling cascades [5], which can result in erroneous view of actual metabolic states. Significant mismatches between “statomics” and dynamics (i.e., flux) have been observed in human [4,21] and animal [5,22] studies. In the following sections, basic principles and applications (to research) of stable isotope tracer methodology to assess in vivo metabolic flux rates are reviewed, with a focus on human muscle dynamics, after briefly discussing how this understanding can be used for drug development in muscle wasting conditions.

Understanding protein dynamics is important for the development of effective drugs targeting muscle wasting conditions. As discussed above, loss of muscle mass can occur as the result of various permutations of protein turnover in which the rate of protein breakdown is greater than protein synthesis, independent of their absolute rates. Depending on the primary side of the balance equation that has been altered to cause loss of muscle mass, a molecular targeting approach for drug development must be considered accordingly. For example, if the primary cause of a catabolic response is the attenuation of protein synthesis, three major targets exist for drug development: 1) implicated signaling pathways [23], 2) translational processes (i.e., translational capacity and efficiency) [24], and the availability of precursor AAs [25]. Although signaling pathways and translational processes are important, protein synthesis is dependent on adequate precursor AAs in the intramuscular fluid compartment, particularly essential AAs (EAAs), which are not endogenously produced. The response to anabolic stimuli, such as resistance exercise and anabolic hormones [15,26], may be limited by precursor availability, thus ensuring that an adequate precursor availability is necessary to obtain the desired results. Therefore, we feel that efforts to discover drugs that target muscle wasting will not likely result in clinically meaningful results without understanding the nature of protein turnover. Consistent with this concept, clinical studies have not shown a significant improvement in muscle mass and/or strength by targeting only the molecular pathways of protein synthesis or breakdown (e.g., myostatin inhibitors, leucine supplementation, etc.) without increasing EAA precursor availability [27-29]. Conversely, if a catabolic response is primarily caused by accelerated protein breakdown, the loss of muscle mass may theoretically be effectively counteracted by inactivating the responsible molecular pathways because “exogenous precursors” are not required in contrast to protein synthesis (i.e., EAAs). Thus, determining the primary problem regarding protein turnover is critical for development of “effective” drugs to treat muscle wasting diseases. Therefore, the basic principles of stable isotope tracer methodology are reviewed to determine the characteristics of the proteome in vivo, which are applicable to humans, animals, and in vitro models, with emphasis on human skeletal muscle protein dynamics.

In vivo information on protein dynamics can be assessed using various tracer methods; Fig. 1 shows analysis using stable isotope-based tracing. Numerous tracer methods exist; however, all are fundamentally based on two tracer models, i.e., tracer dilution and tracer incorporation. The models can be used either independently or in combination to assess in vivo kinetics [30-32]. Various aspects of quantitative metabolic fluxes of nearly every biological compound kinetics including glucose, fatty acids, glycerol, and various AAs can be dissected at cellular, organ, and systemic levels using these state-of-the-art methodologies. Several representative methods use either AA tracers (EAAs such as leucine [Leu] or phenylalanine [Phe] tracer are popular because they are not produced endogenously) or D₂O, both of which trace naturally occurring tracee AAs. Both tracer methods employ a stable isotopically labeled compound (either AA tracer or heavy water) in which one or more heavier isotopes (¹³C, ²H, or ¹⁵N isotopes in AA tracers and ²H isotope in D₂O method) are incorporated somewhere in the tracer compound [31,33,34]. In the following section, the basics of tracer methodology are reviewed and representative tracer methods that enable determination of protein dynamics in vivo evaluated. For more comprehensive information, please refer to the following references [30,31,33,35,36] or to the annual workshop on isotope tracer methodology sponsored by the National Institutes of Health – Mouse Metabolic Phenotyping Center (https://www.mmpc.org/shared/tracers.aspx).

Fig. 1. Schematic diagram of stable isotope tracer-based muscle metabolic research. (A) One or more of tracers are administered into the circulation typically via intravenous infusion or oral bolus to assess muscle protein dynamics. (B, C) Samples are processed for mass spectrometry analysis of tracer enrichment of water, free amino acids (AAs) or AAs bound to proteins. (D) Finally, tracee kinetics are mathematically calculated based on the dose information of the rate of tracer administered and measured enrichment. This figure was created with BioRender.com.
D₂O = Deuterium oxide.

Determination of protein dynamics in vivo includes either 1) administration of a stable isotope AA tracer into the system, typically intravenously (or bolus injection, which is not considered in the present review) or 2) administration (i.p. injection into animal models or oral consumption in humans) of heavy water which labels precursor AAs with deuterium. In both cases, “labeled” AAs are incorporated into product proteins used in conjunction with precursor enrichment determined by mass spectrometry (MS) over the defined experimental period for determination of protein kinetics in muscle and other tissues.

Muscle protein synthesis (MPS) and muscle protein breakdown (MPB) can be assessed using AA tracers and typically determined as the fractional synthetic rate (FSR) and fractional breakdown rate (FBR). The assessment of muscle protein FSR is predicated on the incorporation of precursor (AA) tracer into product (e.g., protein) tracer incorporation approach, which is commonly used to calculate synthesis of any polymers such as DNA [37,38], protein [39,40], glucose (i.e., gluconeogenesis and oxidation) [41-43], fatty acids (i.e., de novo lipogenesis and oxidation) [14,44], and triglycerides [45]. Since AAs are used as building blocks of proteins, the administration of any AA tracer (e.g., ²H₅-phenylalanine, or 1-¹³C-leucine) as a bolus or constant infusion measures the incorporation of the AAs into muscle protein over time (Fig. 2). When assessing FBR, proteins are considered precursors that release AAs from the process of protein breakdown. Based on this principle, the FBR in muscle can be assessed using the 3-pool A-V tracer model [46], administration of L-[3-methyl-²H₃] histidine [47], or decay of the bound enrichment method [48]. When comparing FSR and FBR, the pool size (i.e., muscle protein mass) should be considered because FSR and FBR represent the fraction of the pool size that has been newly produced in a defined experimental period. In general, FSR can be directly compared in acute study as the muscle pool size is virtually unchanged, but when the pool size is significantly altered (e.g., muscle hypertrophy, or wasting conditions), the absolute MPS and MPB should be determined to compare effects between groups (absolute MPS or MPB=FSR or FBR×pool size÷100). Detailed information regarding the MPS and MPB using substrate tracers is available elsewhere [31,33,40]. Fig. 2 shows the basic principles of muscle protein FSR and FBR assessment.

Fig. 2. Assessments of muscle protein fractional synthetic rate using amino acid (AA) tracer method. Both labeled tracer and unlabeled AAs cross the muscle cell membrane in proportion to their relative concentrations (i.e., tracer-to-tracee ratio) to enter the free AA precursor pool from where they are either oxidized or incorporated into proteins. For a given precursor enrichment, a greater labeling of product (i.e., protein enrichment) is achieved in the faster synthesis condition (A) compared with the slower synthesis condition (B). This figure was created with BioRender.com.
MPS = muscle protein synthesis; MPB = muscle protein breakdown.

Heavy water labeling is another tracer method used to determine muscle protein FSR with several advantages. Cumulative muscle protein FSR can be assessed over prolonged periods of time under free-living conditions with simultaneous FSR assessment of different polymers with no apparent true precursor enrichment issue over the traditionally used AA tracer infusion method (e.g., ¹³C, ¹⁵N, or ²H labeled AA) [34,49,50].

The basic principle of the heavy water labeling method is depicted in Fig. 3. FSR can be assessed via oral administration in humans [34] or in small animals such as mice [51] and rats [7]. A rapid equilibrium occurs at a certain level of body water enrichment (animals: ~5% MPE [52], humans: ~0.5% MPE [34]) which generates monodeuterated water (²H₁HO) via exchange of hydrogen (¹H or ²H) between water and precursor AAs through trans-amination reactions [34] and subsequently incorporated into proteins. Finally, muscle protein FSR can be calculated as the change (increase) in enrichment of specific AAs (e.g., alanine with four labeling sites) bound to proteins divided by the precursor (AA) enrichment (directly reflected by water enrichment) because a fixed enrichment ratio exists between water and specific AAs. The enrichment of water or muscle tissue is measured using gas or liquid chromatography-MS [34]. Regarding FBR measurements, since the changes in muscle mass (i.e., protein pool size) are the cumulated difference between MPS (FSR×pool size) and MPB (FBR×pool size), the integrated MPB over time can be deduced by subtracting the cumulated MPS from the changes in muscle mass.

Fig. 3. Assessments of muscle protein fractional synthetic rate (FSR) using heavy water labeling method. (A) A small dose of heavy water (deuterium oxide, D₂O) is orally administered to achieve target enrichment (~1% for humans and ~5% for animals) and assess FSR of muscle protein. (B) D₂O administered into the skeletal muscle will form mono-deuterated water (i.e., ²H¹HO). The ²H will be incorporated into precursors, such as amino acids (AAs). The ²H-labeled AAs are incorporated into new muscle proteins. (C, D) Muscle protein FSR (individual, subgroup, or mixed proteins) is determined based on the information on precursor AA enrichment and changes in product enrichment over a defined time. (E) In addition, enrichment is accessed using gas or liquid chromatography-mass spectrometry in samples obtained via needle biopsy, blood, or urine. This figure was created with BioRender.com.
mCK = muscle creatine kinase.

The rate of protein turnover measured from muscle samples can serve as a key diagnostic marker for skeletal muscle health [32]. However, collection of muscle samples via needle biopsy is quite invasive and hinders human participation in research, particularly for children, women, or older adults [53,54]. For example, obtaining sufficient muscle tissue in older adults is inherently difficult due to greatly reduced muscle mass (i.e., sarcopenia) [40]. Therefore, a “virtual biopsy” method provides a solution to this issue [7,55] and can be used to quantify muscle protein FSR from blood [7] or urine [9] samples (Fig. 3). The method consists of deuterium labeling of muscle proteins through ingestion of a small amount (100 mL) of heavy water daily for a few days to several months, and measuring muscle-derived proteins from blood (e.g., muscle creatine kinase, carbonic anhydrase 3 isoform) [7] or urine (e.g., muscle-specific contractile proteins such as titin and myosin light chain 1/3) [9], which can accurately reflect FSR of muscle proteins in normal heathy individuals (adolescents or adults who do or do not exercise) [7] and patients such as individuals with Duchenne muscular dystrophy [9] or chronic liver disease [56].

Determining protein kinetics at the whole-body level is important for several reasons. First, determination of protein kinetics in muscle may underestimate total protein dynamics, particularly when assessing the overall maximal total anabolic response to dietary protein because a significant portion of protein turnover occurs in other tissues such as rapidly turning-over tissue like the gut, which can later affect muscle protein kinetics [57]. Second, determination of muscle protein FBR has technical difficulties such as complicated mathematical equations and experimental protocols [33,46,58]. Therefore, determining the whole-body rate of protein breakdown can be a surrogate measure for muscle protein FBR. Furthermore, simultaneous determination of MPS and MPB is impossible within the same time period [33]. Finally, whole-body protein kinetics can be determined relatively noninvasively from blood samples without sampling muscle tissues [31,33]. In the following section, representative tracer methods to determine whole-body protein kinetics using Leu tracer method or dual phenylalanine-tyrosine (Phe-Tyr) are reviewed. Furthermore, we will discuss two approaches to assess postprandial whole-body protein kinetics: digestibility method and intrinsically labeled protein method.

The basic principle for determining whole-body protein kinetics is depicted in Fig. 4. The general approach to assess in vivo whole-body protein kinetics is based on determining the rate of appearance (R_a) into and rate of disappearance (R_d) from plasma of a representative EAA, particularly in the Leu and Phe-Tyr methods. The R_a (=R_d in a physiologically steady state) of the unlabeled EAA (tracee) estimated from the EAA tracer based on the dilution model directly reflects the rate of protein breakdown in the fasted state since there is no endogenous production of EAAs (e.g., Leu). The determination of whole-body protein kinetics is predicated on a most complex tracer model which both tracer dilution and incorporation models are integrated with multiple pools and precursors [30,59]. For example, in physiological steady states, R_d of Leu (e.g., of EAA) is equal to R_a of Leu, the latter of which reflects protein breakdown [36] and has two metabolic fates: rate of 1) protein synthesis (R_d²) and 2) oxidation (R_d³). Protein breakdown cannot be directly determined but deduced from protein synthesis and oxidation (i.e., protein breakdown=protein synthesis–oxidation), the latter of which (i.e., Leu oxidation) can be quantified based on the information of labeled CO₂ enrichment in expired air, R_d Leu, and infusion rate (F) of ¹³C labeled Leu. Actual protein synthesis and protein breakdown, thus, net protein balance (protein synthesis–protein breakdown) can be calculated by dividing EAA kinetics by its fractional contribution to protein (e.g., 4% in the case of Phe) [60,61]. Compared with the Leu method, the Phe-Tyr method determines hydroxylation of Phe to Tyr, typically from blood, but not oxidation, consequently eliminating the requirement of isotope ratio mass spectrometry for measuring expired CO₂ enrichment [62]. For the postprandial assessment of whole-body protein kinetics, the Phe-Tyr method is further discussed due to our extensive experience with this method [63,64].

Fig. 4. Assessment of whole-body protein kinetics. Both labeled (from outside of the body via plasma) and unlabeled amino acids (AAs) cross the cell membrane in proportion to their relative concentrations (i.e., tracer-to-tracee ratio) to enter the free intracellular AA precursor pool (i.e., R_d¹). Once inside cells, the tracer can be oxidized (or hydroxylated in the case of Phe, R_d³) or incorporated into protein (R_d²). In a physiological steady state, R_a¹ reflects the rate of protein breakdown and is equal to R_d¹. This figure was created with BioRender.com.
R_d = rate of disappearance; R_a = rate of appearance; Leu = leucine; Phe = phenylalanine.

Whole-body protein kinetics can be determined in the non-steady state (e.g., following consumption of a meal containing proteins or AAs). The appearance of ingested tracee into the peripheral blood must also be quantified to calculate the rate of protein breakdown in the fed state. The postprandial assessment can be accomplished by adding the chemically identical but differently labeled tracer AA from the tracer infused intravenously to the protein meal. Two methods have been widely used: 1) intrinsically labeled protein method [65] and 2) bioavailability method [66,67]. In the intrinsically labeled protein method, research participants consume labeled protein (labeled AAs in the proteins) [68,69]. Although in this method, dilution of tracer coming from the labeled proteins assumedly does not occur between the point of the ingestion and the point of blood sampling, this is not the case. This dilution underestimates contribution of exogenous AAs to peripheral blood (R_a EXO) and consequently overestimates protein breakdown [36,61] resulting in underestimation of net protein balance: Total R_a (R_a TOT)=R_a endogenous (R_a ENDO, reflecting protein breakdown)+R_a EXO (AAs from food). When solving for R_a ENDO, protein breakdown=R_a TOT–R_a EXO. In the bioavailability method, R_a EXO is estimated by correcting the total protein consumption for true ileal digestibility [61,70]. The method assumes no gain or loss of intestinal protein in the process of digestion and absorption, which may in turn introduce errors, which would be random but not accompanied with any systematic bias in the calculated protein kinetics [61,70].

Knowledge regarding the dynamic nature of muscle proteins (“kinetics”) is important as well as determining direct muscle mass (changes) to examine the efficacy of new therapeutics (nutrition, exercise, or drug) for muscle, and can be accomplished with the recently developed non-invasive method, called the D₃-creatine dilution method (Fig. 5). The method is predicated based on several assumptions: orally consumed D₃-creatine is 1) completely absorbed; 2) remains in the body, more specifically in skeletal muscle; and 3) skeletal muscle creatine concentration is consistent [71-73]. These assumptions are arguably valid [40] although some controversy has recently been suggested [74], which should be examined more carefully in the future. Briefly, an orally consumed small dose of D₃-creatine will be diluted as the direct function of the abundance of existing creatine pool, which reflects muscle mass since creatine concentration is constant and resides mostly in skeletal muscle [55,71]. In addition, a constant irreversible conversion from creatine (both labeled and unlabeled) to creatinine (preservation of the labeled to unlabeled ratio) exists and excreted into urine in accordance with their respective abundance [55]. Based on the ratio between creatine and creatinine, the total creatine pool size can be obtained by dividing the dose of D₃-creatine by the urine D₃-creatinine enrichment. The skeletal muscle mass can be estimated by dividing the known creatine pool size (4.3 g/kg) by the creatine concentration in the muscle [9].

Fig. 5. Direct assessments of whole-body functional muscle mass using D₃-Cr dilution method. A small dose of absorbed D₃-Cr and endogenous Cr in the skeletal muscle will irreversibly be converted to Crn. The Crn enrichment from urine reflects the dilution of D₃-Cr as a function of the abundance of existing creatine pool. Knowledge of 1) whole-body Cr pool size (determined) and 2) muscle Cr concentrations (known to be constant) enables calculations of skeletal muscle mass. For a given D₃-Cr oral consumption, a greater dilution of D₃-Cr will occur in cases of large muscle mass (a) compared with smaller muscle mass (b). This figure was created with BioRender.com.
Cr = creatine; Crn = creatinine.

Skeletal muscle protein is in a constant state of turnover (i.e., rates of protein synthesis and breakdown). Although protein turnover is in a balanced state (synthesis=breakdown) in normal healthy individuals over time, a prolonged imbalance of muscle protein turnover (breakdown>synthesis) can lead to muscle wasting conditions such as sarcopenia and cachexia. Loss of muscle mass has been a target of drug development; however, this effort has not had clinical success to date, possibly due to a lack of understanding the dynamic nature of the muscle proteome. Stable isotope tracer methodology enables quantification of protein dynamics (i.e., kinetics) in humans and animals as well as in in vitro models. In conjunction with molecular and cellular biology tools, stable isotope tracer methodology provides an in-depth assessment of dynamic changes in the proteome as well as simultaneous understanding of the molecular mechanisms for the observed kinetics of proteome, allowing researchers to develop effective therapeutic strategies (e.g., nutrition, exercise, and drugs) to treat muscle wasting diseases.

Conceptualization: IYK. Funding acquisition: IYK, SP, JJ. Visualization: HJK. Writing – original draft: IYK. Writing – review & editing: IYK, SP, JJ, YK, HJK.

IYK and SP are shareholders in Myocare, Inc. and Myocro, Inc. The other authors declare no competing interests.

This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (2021R1A2C3005801), the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A3A13071529), the Brain Pool Program funded by the Ministry of Science and ICT through the National Research Foundation of Korea (2019H1D3A1A01071043), and the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1I1A1A01074380).