Metabolomics Captures the Biological Signatures of Aging and Health Span and Identifies Pathway Targets for Intervention

Graphical Abstract

The complexity of the biology that underpins physical, cognitive, and psychosocial decline with aging and the need for robust diagnostic and treatment specificity warrants the use of novel, high-throughput measurements to identify the metabolic pathways that drive aging and functional decline. The development and implementation of metabolomic measurements over the past decade have enabled deeper dissection of interrelated pathways on the molecular level in an efficient and high-throughput manner. Advancing beyond studies in the past that have examined genes and proteins, metabolomics captures the state of the molecular building blocks of biological systems. Metabolomics-based measurements of circulating pools of metabolites can identify changes stimulated by aging and disease as well as by pharmaceutical interventions. Several recent articles in Journals of Gerontology, Biological Sciences (JGBS) have used this approach to gain new insight into the complex pathophysiological changes that occur with aging and functional decline.

The Promise and Peril of Metabolomics

Metabolomic studies employ multiple analytical methods (NMR, LC/GC-MS, etc.) to measure small molecules known as metabolites (ie, sugars, amino acids, or lipids). There are multiple approaches to metabolomics experiments, but these mainly fall in 2 categories: untargeted metabolomics and targeted metabolomics. They have a number of key differences. Untargeted metabolomic experiments are used primarily to “discover” or to identify changes between groups (1,2). This approach measures a large number of metabolites but lacks quantification of metabolites and specificity in identification of metabolites. On the other hand, targeted metabolomics measures fewer metabolites but adds improved identification and quantification to the measurements (3,4). These approaches have advantages and disadvantages and can be suitable for different experiments depending on the objectives of the study being discovery or hypothesis testing. For instance, an untargeted analysis can give you an idea of broad categories of metabolites that are altered in a cohort but cannot reliably pinpoint a single molecule of interest or quantify its concentration. On the other hand, a targeted analysis can identify and quantify metabolites but may miss changes that are in molecules that fall outside of its predetermined targets.

Altered Lipid Metabolism as a Biological Signature of Aging Which May Be Etiologically Linked With Health Span

Epidemiological studies and cross-sectional studies are advancing the search for biological signatures that are characteristic of age-related diseases. These insights will guide mechanistic studies and enable the revision and fine tuning of clinical diagnostic criteria. Numerous studies have now shown age, sex and disease related differences in the metabolome. These differences include changes in major categories of metabolites including lipids and energy pathway intermediates.

Alterations in numerous lipids are a consistent finding associated with the biological signatures associated with aging (5). Di Cesare et al. in a relatively large sample size showed age-dependent remodeling of lipid metabolism (1). Using an untargeted approach, they showed correlations between age and concentrations of blood metabolites, predominately lipids and that these changes were distinct in men and women. Lipids represent a large part of the metabolome and consist of hundreds of unique species that can be structurally dissimilar (ie, phosphotidylcholine compared to stearic acid) or very closely related and differ only by the position or number of carbon double bonds (ie, hexadecenoylcarnitine compared to hexadecadienoylcarnitine). This compounded with the uncertainty of where, and in what pathway the lipids measured originate from, makes drawing conclusions from lipid based metabolomic data somewhat difficult. Future studies will untangle this for more clear insights as to the meaning of lipidomic data.

Metabolomic data can be more impactful when paired with other “omic” data. Liu et al. (2) combined untargeted metabolomic data with transcriptomic bioinformatics analysis. They showed that specific metabolomic profiles associated with decreased skeletal muscle mass in postmenopausal women. This included numerous changes in lipid metabolism with aging and a negative association between stearic acid and skeletal muscle mass. By following up on their cross-sectional metabolomic findings with in vitro studies they showed that stearic acid can negatively affect muscle mass. This approach of combining metabolomic data with direct testing of the effects of molecular interventions in vitro can yield mechanistic insights into biological processes such as aging. Given that metabolomics is species agnostic, comparative metabolomics between humans and animals with frailty and physical decline is another approach that can reveal specific pathways and can buttress the insight/evidence of the involvement of a specific pathway in the age-related disease process (5).

Altered Energy Metabolism as a Biological Signature of Aging That Is Etiologically Linked With Health Span and Life Span and Pharmacologically Targetable

Metabolomic changes connected to muscle mass and muscle strength are not limited to lipids but can also be linked to changes in metabolites involved in energy metabolism. Using untargeted and targeted metabolomic approaches, changes in energy metabolism-related metabolites from glycolysis and the TCA cycle were quantified in a subset of young, robust older and frail vulnerable older adults (4). This study showed elevated levels of numerous TCA cycle and glycolytic intermediates in frail subjects. Furthermore, many of these metabolites were linked to frailty status, decreases in grip strength, and slower walking speed.

In addition to cross-sectional and observational studies, metabolomics can be used in intervention studies to monitor responses to interventions and can be linked to drug levels as a tool to reveal novel drug targets and mechanisms of action. Lee et al. showed that 6-month treatment with the antihypertensive drug Losartan, led to improvements in mitochondrial health biomarkers, such as higher serum concentrations of arginine and spermidine, and lower concentrations of markers of oxidative stress such as nitrotyrosine; and lower extracellular concentrations of several intermediates of energy metabolism. These findings suggest that Losartan treatment alters the circulating metabolome and improves molecular changes associated with frailty and may show a role for Losartan’s nonangiotensin PPARγ pathway-related activity.

Metabolomic profiling has highlighted changes in energy metabolism as a central part of the biological signature of aging. This data support the need for interventional studies that target aging by altering these central metabolic pathways. Kumari et al. have outlined a randomized, double-blind clinical trial in adults free of chronic disease in which they are stratified as either insulin sensitive or insulin resistant based on homeostatic model assessment of insulin resistance and take 1 500 mg/d of metformin or placebo for 12 weeks (6). Given the role of metformin in the control of glucose levels, energy metabolism, and through the action of insulin, lipid metabolism, such study may yield a wealth of information about how targeting these metabolic pathways can improve life span and health span in older adults.

Future Directions

The next generation of metabolomic studies will give more in-depth information about aging and disease and targets for intervention. Although many studies are currently examining the metabolome with cross-sectional study designs, these studies provide only static views of the very dynamic metabolic system in living organisms. Future studies examining the rates of flow of metabolites through the intermediate steps of metabolic pathways, or their “metabolic fluxes,” will yield more detailed information about where metabolic changes happen in aging and diseases. The field of “fluxomics” is gaining ground as an informative tool that can improve our understanding of both complex and poorly understood processes and basic biological processes that were thought to be well understood. Connecting the changes in the metabolome that occur between different tissues and determining how these changes correlate with and contribute to circulating levels of metabolites is another goal that will push the field forward. This emphasizes the need for translational studies combining findings from animal models and human subjects to support these research goals. Also, studies examining sex-related differences and those that examine patients longitudinally are warranted to increase the understanding of the biology of aging and disease within individuals and within sexes.

Funding

This study was supported by the Johns Hopkins Older Americans Independence Center National Institute on Aging (grants P30 AG021334), Bright Focus Foundation Research Award (PMA), the Glenn Foundation for Medical Research and American Federation for Aging Research Grant for Junior Faculty (RW), and National Institute on Aging 1K01AG076873-01 (RW).

Conflict of Interest

None declared.