Epigenetic regulation of drug metabolism in aging: utilizing epigenetics to optimize geriatric pharmacotherapy

Abstract

We explore the relationship between epigenetic aging and drug metabolism. We review current evidence for changes in drug metabolism in normal aging, followed by a description of how epigenetic modifications associated with age can regulate the expression and functionality of genes. In particular, we focus on the role of epigenome-wide studies of human and mouse liver in understanding these age-related processes with respect to xenobiotic processing. We highlight genes encoding drug metabolizing enzymes and transporters revealed to be affected by epigenetic aging in these studies. We conclude that substantial evidence exists for epigenetic aging impacting drug metabolism and transport genes, but more work is needed. We further highlight the promise of pharmacoepigenetics applied to enhancing drug safety in older adults.

Tweetable abstract

Evidence is emerging that epigenetic aging affects the regulation of genes involved in drug metabolism and transport. This may stimulate development of new biomarkers for drug safety in older adults.

Normal aging includes a progressive physiological and functional loss of tissue and organ function in a time-dependent manner [1]. The global population of older adults (people aged 60 years or older) was 962 million in 2017, which was more than double the number of older adults in 1980 when there were only 382 million [2]. Moreover, it is anticipated that by 2050 the number of older adults will more than double again and reach 2.1 billion [2]. By that time, it is projected that there will be more adults aged 60 or older than adolescents and youth at ages 10–24 (2.1 vs 2.0 billion) [2]. This demographic shift has profound implications for healthcare because older individuals require distinct treatment regimens compared with their younger counterparts.

With increased age comes an increased burden of disease; aging is considered the major risk factor for most human diseases [3]. Multimorbidity is very common in older adults, and it often leads to the use of multiple medications. According to the Slone Epidemiology Center at Boston University, 18% of individuals aged 65 and above take between five and nine medications while around 58% rely on ten or more [4]. What is more concerning is that the occurrence of adverse drug reactions (ADRs) leading to hospitalization is seven-times higher in the geriatric population compared with younger patients [5]. Therefore, the development of biomarkers that could guide dosing decisions and prevent ADRs in older adults would have a major impact on healthcare.

The current standard of care for geriatric patients typically involves initiating treatment with a new medication to address a newly diagnosed medical condition, with subsequent regular clinical visits and follow-ups to monitor the patient's response and progress [6,7]. However, it is often necessary to make medication adjustments or dosage modifications over time to optimize therapeutic outcomes [6,7]. These adjustments require diligent monitoring and ongoing follow-up to mitigate the risk of ADRs [6,7]. Despite these efforts, ADRs continue to occur in a significant number of cases [5]. Finally, these approaches are not entirely personalized to each patient's unique needs.

Pharmacogenetics (PGx) stands out as a robust and effective approach for preventing ADRs [8]. Several tests for drug metabolizing genes have already gained approval for clinical implementation, and their utility extends to the older adult population as well [8]. Classical PGx focuses on associating heritable DNA sequence variants with drug response and has indeed allowed for more personalization of treatments by providing medications that ‘fit’ a patient's genetic data. However, in the aging field it has many limitations [8]. For example, a patient's genotype does not change over their life course, except for random somatic mutations [9]. Therefore, PGx cannot explain age-associated changes in drug metabolism in the same individual over time. Consequently, there is a compelling need to identify specific biomarkers that exhibit age-related changes, thereby providing valuable insights into the aging-associated transformation of drug metabolism. Such biomarkers would serve as crucial tools for optimizing medication regimens tailored to the unique needs of the geriatric population.

Age-related changes in pharmacokinetics: clinical practice & molecular evidence

Geriatric clinical pharmacology arose as a unique field of study that focuses on the pharmacokinetic and pharmacodynamic changes occurring with aging and how they impact drug treatment in older adults. Aging significantly affects drug absorption, distribution, metabolism and excretion (ADME) processes with notable implications for drug treatment as we now outline (and summarize in Table 1).

Table 1. . Summary of absorption, distribution, metabolism and excretion changes with age.

	Physiological changes	Medication effect
Absorption	Decreased acid secretion and thinning of epidermis	Drug absorption remains generally unaffected by age; epidermal thinning might increase dermal drug absorption.
Distribution	Reduction in total body water and increased body fat	Reduced volume of distribution, elevated plasma concentrations and decreased half-lives of hydrophilic drugs. Increased volume of distribution, tissue accumulation and longer half-lives of lipophilic drugs
Metabolism
First pass	After reaching the age of 40, the decrease occurs at a rate of approximately 1% per year	Older individuals may experience increased systemic exposure of some drugs and decreased systemic exposure of some prodrugs
Phase I	Reduction in functional capability of the monooxygenase system (decreased activity of certain cytochrome P450 enzymes)	The activation of specific medications may be prolonged, leading to an increased risk of adverse drug reactions. Other medications may experience diminished efficac
Phase II	Preserved Phase II metabolic pathways in healthy older adults, possible decrease in frail older adults	May reduce conjugative metabolism for some medications
Phase III	Altered expression of major transporters in kidney tissue with age; age-related declines in transporter activity	Altered drug concentrations, drug interactions, variability in drug response and an increased risk of drug-related problems
Excretion	Progressive decline in kidney function and volume, reduced renal plasma flow and decreased ability to improve baseline GFR	Decreased clearance of drugs and metabolites excreted by the kidneys

Changes in drug absorption & bioavailability

Most medications are administered orally [10]. Gastric pH, peristalsis, intestinal permeability, mucosal integrity, drug transporters and gastrointestinal blood flow are among the factors affecting oral medication absorption. For the most part, drug absorption appears to be unaffected by age unless there are underlying comorbidities, dietary factors or concurrent medications that contribute to altered gastrointestinal tract function. However, first pass hepatic or gut wall metabolism may decrease with age, resulting in increased bioavailability of some drugs and reduced bioavailability of some prodrugs [11]. Moreover, in older individuals, diminished tissue perfusion following intramuscular or subcutaneous injections may also have implications for drug absorption. Consequently, understanding these age-related considerations may be important in optimizing drug therapy for specific older patients.

Changes in drug distribution

Age-related factors also play a crucial role in the distribution process, as they determine the concentration of active substances at specific targets within the body. Several factors, including tissue and plasma protein binding, changes in body composition and protein synthesis, can impact drug distribution [12]. Modifications in body composition including decreased body water and a relative increase in body fat may impact the pharmacokinetic profile of drugs based on their lipid solubility [12]. Moreover, the distribution volume is subject to the impact of the proportion of lean body mass to adipose tissue mass. The geriatric population experiences a reduction in the distribution volume, leading to elevated plasma concentrations and decreased half-lives of hydrophilic drugs. Conversely, lipophilic drugs may accumulate and exhibit longer half-lives in older individuals which may lead to toxicity [12].

Changes in liver & kidney metabolism & excretion

Age related changes in liver function have a significant impact on the metabolism and elimination of drugs, which encompasses three essential phases: phase I, which involves functionalization reactions; phase II, which entails conjugation reactions; and phase III, which encompasses the elimination of drugs and metabolites through transporters found in the liver and other tissues [13].

Phase I enzymes

The cytochrome P450 (CYP450) enzymes are involved in catalyzing the oxidative biotransformation of most drugs [14]. A large body of work in laboratory animals has demonstrated substantial age-related declines in CYP450 content, activity and inducibility [15,16]. In humans, there is a general consensus that physiological changes such as liver volume, liver blood flow and hepatocyte mass change with age, which can influence metabolism of drugs [17–20]. CYP-mediated phase I reactions are more likely to be impaired in older adults than phase II reactions [18]. However, the degree to which this impairment is the result of broader physiological changes with age, versus reductions in the expression and function of human CYPs, has been debated. For example, Sotaniemi et al. analyzed 226 subjects with equal histopathologic conditions and reported that cytochrome P450 activity, in general, was shown to be 32% lower in patients 70 years and older when compared with patients aged 20–29 years, or declines at a rate of 0.07 nmol/g of liver after approximately 40 years of age [21]. George et al. examined human liver tissues from a group of 71 subjects, including those with histologically normal livers and several with chronic liver disease [22]. Their study showed that overall CYP450 levels decrease by -3.5% for each decade of life, with the expression of CYP2E1 and CYP3A4 enzymes declining by 5 and 8% per decade, respectively. However, other studies have found increases in the levels of some CYP450s with age [23].

A more recent study by Dücker and Brockmöller [24] found that the pharmacokinetic properties of certain medications such as digoxin, venlafaxine, ezetimibe, fluvoxamine and clomipramine are impacted by the age of the individual and to a lesser extent by their genotype. Conversely, other medications like risperidone, clomipramine, thioridazine, desmethyldiazepam and paroxetine are affected by both factors (age and genotype). In terms of drug-metabolizing enzymes, the study suggested that drugs which are primarily metabolized by CYP2C19 and CYP2D6 enzymes may be affected by age and genotype, leading to a possible increase of twofold or more of systemic exposure [24].

Phase II enzymes

In laboratory rodents, some studies have reported that UGT enzyme activity, which mediates glucuronidation reactions, declines with age [25]. Decreases in glutathione S-transferase have also been detected in some tissues as rats get older [26]. Fu et al. [27] studied over 100 xenobiotic metabolism and transport genes in male and female mice, and found several phase II genes that were downregulated in aging, such as the glutathione S-transferases mu 4 and pi 1 (Gstm4, Gstp1) and sulfotransferase 3 a1 (Sult3a1). They also found some up-regulated phase II genes in older mice such as Gstm2 and Sult2a1. Substantial sex differences in the age-related effects were noted [27]. Meanwhile, some human studies have suggested decreased hepatic glucuronidation capabilities among older individuals [28]. However, the number of studies reporting age-related change in phase II metabolism in humans is relatively modest and current consensus is that human phase II metabolism is mostly preserved in older adults [29,30], with some exceptions.

Frailty is a state characterized by diminished physiological reserves and heightened susceptibility to negative consequences when faced with stressors [31]. The concept of frailty is underpinned by studies documenting a decline in drug metabolism and changes in disposition in frail older people compared with either healthy older adults or the young. Frail older adults may experience a reduction in phase II metabolism, as described by Wynne et al. [32]. They reported that aging adults exhibiting frailty may exhibit a decreased ability to clear acetaminophen/paracetamol via glucuronidation, and metoclopramide via glucuronidation and sulfation. Meanwhile, non-frail individuals do not show similar issues. Moreover, alterations in gut microbiota composition can have an impact on drug metabolism, particularly among frail older adults. With aging comes bodily changes resulting in biochemical imbalances within an individuals' microbiota, a process referred to as dysbiosis. This dysbiosis can impact the metabolism and bioavailability of drugs, potentially altering their therapeutic efficacy and safety profiles. To improve understanding regarding these potential alterations researchers must be mindful of connected age-related factors such as nutritional status; inflammatory conditions; and gut microbiota composition impacting their internal balance [33,34]. Understanding the complex interplay between age, frailty, nutrition, inflammatory conditions and gut microbiota composition is essential while considering these age-related factors [10].

Phase III transporters

Considering transporters that mediate hepatic phase III reactions, several rodent studies have examined how transporter expression changes from early development into adulthood [35], but relatively fewer have looked at old age. One such study of older rats indicated that expression of several organic anion-transporting polypeptides (Oatps) decreased with age [36]. Another study by Fu et al. [27] found that Oatp1a1 was downregulated in the livers of both male and female mice, whereas organic anion transporter Oat2 (Slc22a7) was upregulated in males. Zhu et al. [37] found that ATP-binding cassette sub-family G member 2 (Abcg2) was reduced in livers of older rats. In the rat kidney, expression levels of several transporters are low in early development, but increase gradually after birth and into adulthood; however, levels decrease in old age [38]. Human studies of hepatic and renal transporter function in normal aging are lacking, but age-related diseases and inflammation can impair transporter function [20,39].

Changes in renal excretion

With normal healthy aging, the kidney undergoes a progressive increase of nephrosclerosis (thickening and hardening of the kidney's blood vessels) with loss of functional glomeruli, which results in an overall decrease in kidney function as determined by the patient's glomerular filtration rate (GFR). This decline in kidney function and volume begins in patients older than 50 years and can affect the clearance of medications that are excreted by the kidneys [40]. Aymanns et al. suggest that the decline in renal function is the most prominent change in organ function that impacts drug pharmacokinetics in older adults. Furthermore, they note that 15 to 30% of older individuals have a GFR of 30–60 mL/min, which is indicative of stage 3 kidney disease [41,42]. Ultimately, medication doses should be adjusted to account for these changes. Prescribing medications excreted by the kidneys requires special attention in older patients, especially for medications with a narrow therapeutic index.

Aging also reduces renal plasma flow by 50% and significantly reduces the kidney's capacity to improve baseline GFR. In addition, medications that promote the excretion of salt and water may have harmful effects on older adults because of their reduced ability for sodium reabsorption. Renal failure, low blood volume and hyponatremia are possible outcomes of using such medications. Furthermore, hyperkalemia may arise when certain drugs are given to geriatric patients with decreased renal potassium excretion (ACE inhibitors, ARBs, aliskiren, digoxin, potassium-sparing medicines, beta-blockers and NSAIDs) [10]. In general, frail older adults often experience a decline in renal function when compared with younger adults, leading to a general decrease in the clearance of medications that are primarily eliminated through the kidneys. For older adults facing frailty, keeping their bodily systems in equilibrium becomes increasingly difficult and poses more risks for them. These issues may be related to past complications following kidney injury from medications or other sources, ultimately leading to lowered renal performance [43,44].

Summary

There is substantial evidence for age-related decline in the body's ability to process drugs, although current evidence suggests that some systems are more likely to be impaired than others. A key aspect of these age-related changes, however, is that they do not occur uniformly at specific ages in all individuals and there are substantial individual differences in the extent of age-related change. We now turn to understanding molecular mechanisms of aging that can potentially give insight into these phenomena.

The complex mechanisms of aging: understanding the effect of epigenetics

Aging is a natural and complex process which is manifested by a gradual decline of physiological function in a time-dependent manner [45]. As the world's older adult population continues to increase, there is a need to understand the underlying mechanisms of aging and develop biomarkers associated with healthy aging. Aging can be characterized by complex, intra-individual processes associated with 12 major cellular and molecular hallmarks [46,47]. In their 2013 paper López-Otín et al. proposed the first nine hallmarks of aging – which they updated later in 2023 to become 12. They refer to the molecular and cellular changes that ‘are generally considered to contribute to the aging process and together determine the aging phenotype'. Epigenetic changes, one of the primary hallmarks of aging, are modifications to DNA that can affect gene expression without altering the underlying genetic sequence. These changes can include DNA methylation, histone modifications and non-coding RNA. We will demonstrate in this review how epigenetic changes in normal aging can influence the function and expression of drug-metabolizing genes. Current evidence already exists for age-related changes in the expression of genes related to inflammation, stress response and metabolic genes while many other tissue-specific expression changes have been observed [48]. Considering the liver, previous reports have indicated an association between expression of xenobiotic metabolism genes and longevity in mice [49–52] and other organisms including humans [53,54]. However, the extent to which these expression changes are driven by epigenetics have yet to be fully elucidated.

Our investigation of age-associated epigenetic changes in this paper resonates with the ‘aging as a software design flaw' theory [55]. According to De Magalhaes, aging occurs not as a result of physical damage accumulated by cellular components, but due to inherent flaws in our cells' genetic code [55]. This software design leads to a change in epigenetic states with age, which is one of the driving forces for gene expression. It is therefore conceivable that epigenetically-driven changes to gene expression in aging could cause changes in drug metabolism. This loss (or change) of information with age, as described by the informational theory of aging, applies to one of the most broadly studied epigenetic modifications, DNA methylation (DNAm). Across the genome, both hypermethylation and hypomethylation can occur at different genomic loci, but a global hypomethylation trend is typically observed (Figure 1) [56]. These changes in DNAm levels can be due to random chance (DNAm drift) or can be consistent changes at specific positions in the genome that occur in a time-dependent manner i.e., age-associated differentially methylated regions (a-DMRs) [57]. These epigenetic changes lead to the phenomena of chromatin decondensation with age or the heterochromatin loss model of aging [58]. Therefore, this general loss of information results in the misreading of genes and consequently age-associated physiological changes, possibly including changes in drug metabolism.

Figure 1. . Model of epigenetic modifications with aging.

Some epigenetic marks reduce gene expression like DNA methylation, others increase the gene expression by changing the configuration of histones that provide structural support to the DNA sequence, like the addition of acetyl group on the 27th lysine residue of Histone 3 (H3K27ac).

The concept of biological age of an individual, as defined by epigenetics, as opposed to their chronological age, has gained significant traction in recent years. ‘Epigenetic clocks’ are mathematical models that estimate an individual's biological age by analyzing specific variations in DNA methylation patterns [59]. These normal age-related variations in DNA methylation can indicate health and illness risk. In comparison to earlier clocks, the second-generation epigenetic clocks feature more accuracy, precision, and generalizability. These clocks were created using deep learning and are pan-tissue DNA-methylation epigenetic clocks. The biological age estimates that they yield more accurately reflect disease morbidity and mortality risk than first generation clocks [59,60].

The tight association between DNA methylation-based biomarkers and the aging process suggests that this epigenetic mark could be excellent biomarker candidate for aging pharmacoepigenetics [61,62]. More specifically, if DNA methylation changes occur at genes associated with drug metabolism, this could lead to misreading of the genes and consequently result in age-associated ADRs. However, if these changes can be predicted and preventative measures taken, this could allow for safer use of medications in the geriatric population.

Epigenetic changes in the aging liver: effect on drug metabolizing genes

In this section, we present current evidence for age-related epigenetic changes at drug metabolizing genes in the liver. We focus on liver studies for two reasons: 1) the liver is the main site of xenobiotic metabolism; and 2) multiple studies have shown that DNA methylation changes with age are tissue-specific [63]. A summary of the studies used for this section is shown in Table 2.

Table 2. . Genome-wide DNA methylation datasets summary.

	Type	N	Age	Sex	Ref.
Human studies
Bysani et al.	Illumina 450K	95	28–65 years	Males and females	[64]
Bacalini et al.	Infinium HumanMethylation450 BeadChip	45	13–90 years	27 males, 18 females	[65]
Mouse studies
Sandoval-Sierra et al.	MBD-seq (methyl-binding domain protein capture and sequencing)	70 livers	6–25 months	Female BXD murine family	[66]
Sleiman et al.	RRBS (reduced representation bisulfite sequencing)	3 vs. 3	6 vs 24	Male C57BL/6J mice	[67]

MBD-seq: Methyl-binding domain protein capture and sequencing; RRBS: Reduced representation bisulfite sequencing.

Human studies

Bysani et al. aimed to investigate the effect of aging on DNA methylation and mRNA expression in the human liver [64]. Illumina 450K and HumanHT12 expression BeadChip arrays were used to analyze genome-wide DNA methylation and gene expression in liver samples. Their sample included 95 individuals aged 28–65 years old with an average of 49.5 ± 7.6 years old. The investigators studied liver biopsies collected during surgery as a component of the Kuopio Obesity Surgery Study from patients who underwent Roux-en-Y gastric bypass surgery [68]. Their results showed that age was significantly associated with altered DNA methylation at 20,396 CpG sites and they identified 151 genes whose liver expression also correlated with age. Interestingly, when the liver methylation data were compared with published methylation data in other tissues, it was found that the age-associated CpG sites overlapped primarily between liver and blood. Thus, blood DNA methylation levels may reflect aging-related changes in DNA methylation patterns and gene expression in the liver. Of the significant CpG sites that change with age, several are located in genes encoding drug metabolizing enzymes and transporters including: CYPs (CYP1A1, CYP1B1), alcohol dehydrogenases (ADH1B), aldehyde dehydrogenases (ALDH1A2, ALDH1A3), ATP-binding cassette transporters (ABCB4, ABCC1, ABCC4, ABCC8, ABCG1) and UGTs (UGT2B7, UGT3A2). However, it is important to note that the subjects were part of an obesity study and previous studies have shown that obesity accelerates epigenetic aging of the human liver [69]. Moreover, obesity can affect rates of drug metabolism across the life course [70]. This may limit the generalizability of the findings to the broader population, but it is worth considering that this population might represent individuals experiencing premature or accelerated aging effects. Furthermore, another limitation of this study is its age cutoff of 65, and so did not encompass the full spectrum of older adult population.

A study by Bacalini et al. investigated the epigenomic and transcriptional changes that occur during aging in the human liver [65]. The liver biopsies were collected from 45 liver donors (27 males and 18 females) ranging from 13 to 90 years old and epigenome-wide DNA methylation levels were assessed using the Infinium HumanMethylation450 BeadChip. The results indicated that the epigenetic aging rate in the liver decreases after 60 years of age. The study also identified a liver-specific epigenetic signature of aging, with 75 a-DMRs validated in hepatocytes. One of their targeted analyses included CYP1B1 which showed a high DNA methylation correlation with age. Interestingly, there was a significant overlap observed between the list of differentially methylated regions in this study and those previously identified by Bysani et al. The overlap was substantial for regions showing a positive association with age (3475 out of 5572), while it was relatively lower for regions exhibiting a decrease in methylation with age (43 out of 3051). However, the study had limitations, such as the liver biopsies being collected from brain-dead, heart-beating donors, which may have impacted liver transcription and epigenetic states relative to the healthy population.

Animal studies

Mice serve as valuable model organisms for DNA methylation studies due to their genetic similarity to humans, shorter lifespan and faster reproduction [71]. Moreover, employing an animal model allows for better control over confounding factors, such as biological and environmental influences, which all can affect epigenetic signatures. This control is crucial in minimizing the impact of these factors and maintaining consistent experimental conditions [72].

Sleiman et al. analyzed epigenetic modifications in the livers of male C57BL/6J mice in adults 6 vs 24 months, including the DNA methylome, histone modifications as well as their transcriptome [67]. In their study, they compared the epigenetic profiles of aging livers against muscle and heart, concluding that epigenetic changes are tissue specific. For example, the enrichment analyses showed that the liver had the most changes in gene expression, while the quadriceps and heart had more changes in histone modifications. Interestingly in their study, they reported that their multi-layer and multi-tissue analysis reveals that aging affects different facets of gene regulation in a tissue-dependent manner. Whereas the liver has the strongest gene expression and DNA methylation effect, the heart and muscle show strong epigenetic alterations, mainly toward a gain in activating and loss of repressive marks. Of note, their a-DMR list included CpG sites located in genes involved in drug metabolism including: Cyp1a1, Cyp1a2, Cyp1b1, Cyp2d9, Aldh1a2 and Abcd2.

Another study in mice by Sandoval-Sierra et al. investigated epigenetic marks in livers of aged females [66]. They investigated the association between body weight, high-fat diet and epigenetic aging in female members of the BXD mouse line. The study found that both higher body weight and high-fat diet were associated with accelerated epigenetic aging, as measured by an epigenetic clock defined from a-DMRs. Their a-DMR list includes genes involved in drug metabolism: Cyp2a4, Cyp3a41, Abcc4, Cyp46a1, Cyp2e1, Cyp2d9, Slc6a6.

Targeted analysis of specific genes in the mouse liver has also shown that the function of specific drug-metabolizing enzymes can be affected by epigenetic aging. Kronfol et al. demonstrated how epigenetic aging can affect the function of two drug-metabolizing genes: Cyp2e1 (phase I) and Sult1a1 (phase II). They used liver tissues from male C57BL/6 mice aged 4–32 months and found that DNA methylation significantly increased with age at Cyp2e1 and that this increase was associated with a decrease in gene and protein expression. Moreover, CYP2E1-mediated intrinsic clearance of the probe drug chlorzoxazone, which is almost exclusively metabolized by CYP2E1, was associated with DNA methylation levels and histone 3 lysine 9 acetylation (H3K9ac) at the Cyp2e1 locus but not with chronological age [73]. This analysis also showed that this epigenetic biomarker was more closely associated with rates of CYP2E1-mediated clearance across ages than chronological age itself [73]. In a separate study, they found a significant decrease of DNA methylation with age at Sult1a1 in addition to an increase in H3K9ac. Moreover, H3K9ac levels accounted for 23% of the variance in Sult1a1 gene expression across all ages studied [74]. These results are two targeted examples of how epigenetic changes in normal aging correlate with the expression and function of genes encoding both phase I and phase II drug metabolizing enzymes.

Discussion

Pharmacoepigenetics is an emerging field and can be defined as the study of the epigenetic basis for individual variation in drug response [75]. A search of PubMed using the keywords ‘drug-metabolism’ and ‘epigenetics’ resulted in 178 articles. However, the same search adding the any of the keywords ‘aging’, ‘elderly’, ‘geriatric’, ‘older adults’ results in only 14 articles. Moreover, a search of ‘pharmacoepigenetics’ yields 136 entries while adding the same keywords reduces the results to only 17 articles. The scarcity of research on how epigenetic modifications regulate drug metabolism during the process of aging is an oversight because of these markers change so much during aging. Furthermore, it has been shown that many epigenetics modifications can indeed affect phase I, II, and III gene with age, as revealed in epigenome-wide studies and summarized in Table 3. These studies should not be viewed as exhaustive because there are limitations to current epigenome-wide studies of aging in the human liver, as we outlined above, and many important human drug metabolizing genes such as CYP3A4 and CYP2D6 do not have clear rodent orthologs, reducing the utility of rodent studies. Moreover, the human liver epigenome-wide studies described above used microarrays that assay DNA methylation at several hundred thousand CpG sites but there are approximately 28 million CpG sites in the human genome, so these studies only cover a portion of the epigenome. Bisulfite-based methods for detecting DNA methylation, such as used for the Illumina arrays, cannot distinguish between DNA methylation and hydroxymethylation, which can have very different effects on expression. Other important marks such as histone acetylation have not been investigated to date in this context. Therefore, large-scale, epigenome-wide studies of human samples that employ next-generation sequencing technology and account for a broad range of potential confounders such as sex, obesity, smoking status etc, will be necessary to fully map epigenetic aging effects at relevant drug metabolizing genes.

Table 3. . Genes of drug-metabolizing enzymes demonstrating epigenetic regulation with age in the liver.

Phase I	Epigenetic change with age	Source of evidence	Gene expression change with age	Source of evidence	Ref.
CYP1A1/Cyp1a1	DNA methylation ↑	Human and animal studies	Gene expression ↓	Animal study	[65–67]
Cyp1a2	DNA methylation ↓	Animal study	Gene expression ↓	Animal study	[66,67,76]
CYP1B1/Cyp1b1	DNA methylation ↑	Human and animal studies	Gene expression ↓	Animal study	[65,67,76]
Cyp2d9	DNA methylation ↑	Animal studies	Gene expression ↓	Animal study	[66,67]
Cyp2a4	DNA methylation ↑	Animal study	–	–	[66]
Cyp2e1	DNA methylation ↑	Animal studies	Gene expression and protein expression ↓	Animal study	[66,73]
CYP3A7	DNA methylation ↑	Human study	–	–	[65]
Cyp46a1	DNA methylation ↑	Animal study	Gene expression ↑	Animal study	[66,76]
CYP26C1	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
CYP26B1	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]
ADH1B	DNA methylation ↓	Human study	–	–	[65]
Aldh1a2	DNA methylation ↓	Animal study	Gene expression ↑	Animal study	[67,76]
ALDH1A3	DNA methylation ↑	Human study	–	–	[65]
Sult1a1	DNA methylation ↓	Animal study	Gene expression ↑	Animal study	[66,74]
Phase II
Ugt1a2	DNA methylation ↑	Animal study	Gene expression ↑	Animal study	[67,76]
UGT2B7	DNA methylation ↑	Human study	–	–	[65]
UGT3A2	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
Phase III
ABCB4	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,66,76]
ABCC4/Abcc4	DNA methylation ↑	Human and animal studies	Gene expression ↓	Animal study	[65,66,76]
ABCC8	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]
ABCG1	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]
Abcd2	DNA methylation ↑	Animal study	Gene expression ↑↓	Animal study	[66,67,76]
Slc6a6	DNA methylation ↓	Animal study	Gene expression ↑	Animal study	[66,76]
SLC3A1	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,66,76]
SLC4A11	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]
SLC5A2	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
SLC6A1	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
SLC6A3	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
SLC6A5	DNA methylation ↑	Human study	Gene expression ↑	Animal study	[65,76]
SLC6A7	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]
SLC7A4	DNA methylation ↑	Human study	Gene expression ↓	Animal study	[65,76]

Nevertheless, from Table 3, it is evident that for those epigenetic modifications detected so far, many are closely linked to changes in gene expression. Notably, genes like CYP1A1, Cyp1a2, CYP1B1, Cyp2e1, CYP26B1, Aldh1a2, Sult1a1, ABCC4, ABCC8, ABCG1, Abcd2, Slc6a6, SLC4A11, SLC6A7 and SLC7A4 show a consistent pattern: increased DNA methylation, a suppressive epigenetic mark, aligns with decreased gene expression [77]. On the other hand, there are cases where DNA methylation changes do not correlate with expression shifts, such as for Cyp1a2, warranting further investigation into these complexities. PGx testing as currently implemented does not eliminate all ADRs [8]. Moreover, it is doubtful that even a complete knowledge of all common genetic variation would fully explain ADRs. Genome-wide association studies have consistently revealed a gap, known as the ‘missing heritability’, between the trait variance explained by single nucleotide polymorphism associations and the heritability of the trait [78]. The work by Preskorn et al. has shed light on the complex relationship between an individual's functional drug metabolism phenotype and their DNA genetic sequence genotype [79]. Their research showed that the expression of an individual's genetic code does not always align with their observed drug metabolizer status. This underscores the possibility that alternative factors like epigenetics could exert considerable influence. To gain a comprehensive understanding of drug response patterns, it is imperative to explore the intricate interplay between genetics, epigenetics, and personalized medicine, especially in vulnerable populations like older adults.

The Beers Criteria from the American Geriatrics Society aim to alert healthcare professionals to potentially inappropriate medications where the risks generally outweigh the benefits of treatment in the older adult population [80]. It classifies five categories of medicines requiring specific consideration during prescription, including: 1) potentially inappropriate medications (PIMs); 2) medications considered generally inappropriate in patients having syndromes or particular diseases; 3) drugs that should be used with caution; 4) potentially inappropriate drug-drug interactions; and 5) medications with recommended dosage adjustments with reduced renal function [80]. The Beers Criteria are a valuable tool for healthcare providers in making informed decisions when prescribing medications for older adults. These guidelines focus on drugs that may have an increased risk of adverse events or have minimal therapeutic benefit in older adults due to physiological changes related to aging and increased susceptibility to side effects [80]. The Beers Criteria take into consideration the impact of age-related physiological changes on drug metabolism and response, but do not directly mention or relate to any epigenetics modifications.

We speculated that drugs in the list of PIMs may be metabolized by enzymes affected by epigenetic aging. While many of the genes involved in the metabolism of PIMs are epigenetically regulated, studies are lacking when it comes to investigating epigenetic changes with age. Therefore, we need to begin by acknowledging that evidence regarding epigenetic changes associated with aging is quite limited. The available studies, largely conducted on humans with certain limitations such as obese patients or organ donors, or on mice, are constrained in their generalizability to the normal aging populations. It's important to note that many of the orthologs of drug metabolizing genes are absent in mice, preventing a direct translation of findings to humans [81]. This underscores the challenge of extrapolating findings from mouse studies to human systems. Nevertheless, CYP1A2 metabolizes many of the PIMs (diphenhydramine, doxylamine, clonidine, olanzapine, zolpidem, nabumetone and metaxalone) and the Cyp1a2 gene shows evidence of epigenetic aging in mice.

We believe since aging can significantly alter these epigenetic markers, thereby they might influence drug metabolism potentially leading to altered drug response in older adults. Gaining a deep understanding of the relationship between genetics, epigenetics and aging is crucial for optimizing pharmacotherapeutic outcomes in older adults. This knowledge will help in tailoring treatments according to an individual's needs.

Conclusion

In conclusion, the integration of pharmacoepigenetics into current drug treatment approaches is a promising path in advancing precision medicine, especially for the aging population. The physiological and epigenetic changes that occur with age can impact the metabolism and effectiveness of medications. Therefore, a comprehensive understanding of these changes, their interactions with genetic factors and their influence on drug response are imperative for optimizing pharmacotherapy among older adults.

Future perspective

To develop pharmacoepigenetics as a practical tool to adjust dosing and prevent ADRs in older adults, the most urgently needed next steps are large-scale studies correlating epigenetic biomarkers with pharmacokinetic outcomes in humans. Unlike DNA sequence, epigenetic states are tissue-specific so epigenetic changes in peripheral tissues may not reflect what is going on in internal organs. However, as described above, there is evidence from human studies that epigenetic states in blood and liver overlap to an extent [64]. The key development now, will be to determine the precise epigenetic markers to assay in blood for maximum predictive power with respect to age-related hepatic or renal metabolic changes. While studies of epigenetic aging and gene regulation the human post-mortem liver are highly desirable from a mechanistic perspective, practical biomarkers for clinical use will have to be derived from peripheral tissues.

Future research might unravel possibilities of modulating epigenetic age-related changes in drug metabolism through interventions. Numerous studies have elucidated the interconnectedness of nutrition, epigenetics, aging and chronic inflammatory diseases [82,83]. Various dietary components and metabolites have been identified to influence the activity of epigenetic enzymes and factors that regulate gene expression and chromatin structure, thereby suggesting their potential role in regulating drug metabolism changes with age [84]. Methods like caloric restriction have been shown to influence DNA methylation in mouse liver [85,86]. Studies of anti-aging interventions including caloric restriction have been conducted on mice and have been shown to deaccelerate epigenetic clocks and reverse or prevent 20 to 40% of the age-related changes in DNA methylation, indicating that they might be helpful in modulating DNA methylation changes in humans [87]. Furthermore, evidence suggests the potential of pharmaceutical interventions, such as metformin and rapamycin in mitigating age-associated epigenetic changes [88–92]. As such, these methods may serve as potential approaches to address epigenetic changes influencing drug metabolism among older individuals. However, it seems more promising in practice that we could tailor the dosage according to specific epigenetic characteristics than attempting to modify the epigenetic profile to accommodate a standard adult dose.

In this review we focused on epigenetic modifications as it is one of the hallmarks of aging, however, nutrient sensing, which is another hallmark of aging, is an important factor to consider when examining the effects of aging on drug metabolism. It involves cellular mechanisms that monitor and respond to the availability of nutrients. These pathways, such as mTOR and AMPK, play a role in regulating metabolism, energy balance, and the aging process [93,94]. Research suggests that changes in nutrient sensing pathways can impact aging and its influence on drug metabolism. For example, disruptions in these pathways, particularly mTOR signaling, have been associated with impaired drug metabolism and increased risk of drug-related issues [95]. Moreover, nutrient sensing pathways can interact with epigenetic modifications and gene expression, further influencing drug-metabolizing enzymes and transporters [96]. Therefore, we believe that understanding how nutrient sensing and epigenetics collectively influence drug metabolism and transport could have significant implications.

Executive summary.

• The application of classical pharmacogenomics, based on DNA sequence variant, in personalizing treatments in older adults has limitations. However, new tools like epigenetics hold promise for this population.

• Aging leads to a decline in overall CYP450 metabolism, which may influence the pharmacokinetics of several medications.

• Phase II metabolism is generally preserved in older adults, but some studies suggest possible age-related decreases in specific enzymes, especially in frail older adults.

• Age-related changes in the activity of transporters involved in phase III reactions can affect drug processing and function.

• Several studies on humans and mice have considered epigenetic modifications in the aging liver, and showed evidence that epigenetics can influence drug metabolism with age.

• Some medications labeled by the Beers Criteria as potentially inappropriate for older adults are metabolized by enzymes susceptible to and regulated by epigenetic aging. However, further studies are needed in this area.

• The integration of pharmacoepigenetics into the current pharmacological approaches stands as a promising path in advancing precision medicine for the aging population. Future research should focus on mapping those epigenetic changes with the greatest power to predict drug response in older adults, in addition to exploring innovative strategies to optimize drug therapies tailored to the unique needs of the aging population.