Roles of the tyrosine isomers meta-tyrosine and ortho-tyrosine in oxidative stress

Abstract

The damage to cellular components by reactive oxygen species, termed oxidative stress, both increases with age and likely contributes to age-related diseases including Alzheimer’s disease, atherosclerosis, diabetes, and cataract formation. In the setting of oxidative stress, hydroxyl radicals can oxidize the benzyl ring of the amino acid phenylalanine, which then produces the abnormal tyrosine isomers meta-tyrosine or ortho-tyrosine. While elevations in m-tyrosine and o-tyrosine concentrations have been used as a biological marker of oxidative stress, there is emerging evidence from bacterial, plant, and mammalian studies demonstrating that these isomers, particularly m-tyrosine, directly produce adverse effects to cells and tissues. These new findings suggest that the abnormal tyrosine isomers could in fact represent mediators of the effects of oxidative stress. Consequently the accumulation of m- and o-tyrosine may disrupt cellular homeostasis and contribute to disease pathogenesis, and as result, effective defenses against oxidative stress can encompass not only the elimination of reactive oxygen species but also the metabolism and ultimately the removal of the abnormal tyrosine isomers from the cellular amino acid pool. Future research in this area is needed to clarify the biologic mechanisms by which the tyrosine isomers damage cells and disrupt the function of tissues and organs, and to identify the metabolic pathways involved in removing the accumulated isomers after exposure to oxidative stress.

1. Introduction

Since 1956 when Denham Harman first postulated that aging and its associated diseases could be attributed to cellular damage caused by free radicals (Harman, 1956), the Free Radical Theory of Aging has been a central hypothesis in aging research. Extensive research has been conducted to determine the role of reactive oxygen species in the aging process, and arguments both for and against this hypothesis have been put forth (Buffenstein et al., 2008; Hekimi et al., 2011; Perez et al., 2009; Schaar et al., 2015; Vina et al., 2013). In seemingly direct conflict to this theory, multiple reports have shown the knockdown of key anti-oxidative genes and/or the increase of reactive oxygen species within an organism, not only fail to shorten lifespan, but in some cases increase it altogether (Doonan et al., 2008; Schulz et al., 2007; Van Raamsdonk and Hekimi, 2012). Yet the damage to cellular components by reactive oxygen and nitrogen species, termed oxidative stress, has also been reported to increase with age (Stadtman, 1992) and is likely to contribute to age-related diseases including Alzheimer’s disease (Shen et al., 2008; Tamagno et al., 2002; Tamagno et al., 2008; Wang et al., 2014), atherosclerosis (Li et al., 2014; Pennathur et al., 2001), diabetes (Brownlee, 2001; Giacco and Brownlee, 2010), and cataract formation (Fu et al., 1998b; Kruk et al., 2015). These discrepancies highlight the limited understanding we currently have regarding oxidative stress and the processes by which it mediates cell damage and contributes to aging and disease. With abnormal tyrosine isomers as an example, it is likely that many of the pathological mechanisms of oxidative stress occur insidiously and extend far beyond our current knowledge.

Under conditions of oxidative stress, the production of the abnormal tyrosine isomers meta- and ortho-tyrosine primarily occurs when hydroxyl radicals oxidize the benzyl ring of phenylalanine (Mager and Berends, 1974; Maskos et al., 1992). While elevations in m- and o-tyrosine concentrations were previously perceived to simply be a biological marker of reactive oxygen species, there is emerging evidence from bacterial, plant, and mammalian studies suggesting that these atypical isomers of tyrosine are actually mediators of oxidative stress. This represents a novel mechanism by which oxidative stress may disrupt cell structure and function. It is therefore conceivable that the cellular defenses against the accumulation and adverse effects of m- and o-tyrosine are two-fold: (1) the reduction of reactive oxygen species and other free radicals via classical anti-oxidative stress mechanisms (e.g. superoxide dismutase, glutathione, reactive oxygen species scavengers, etc.), and (2) the elimination of the abnormal tyrosine isomer pool via the activation of metabolic and degradation pathways. Only when both defenses are impaired would pathogenic phenotypes be observed. Thus previously unheralded metabolic pathways may prove to be a missing link in specifying the effects of free radicals in aging and disease progression.

Here we review the formation of these abnormal tyrosine isomers as well as their elevation in disease states. Additionally we will summarize the literature revealing their toxicity to cells and tissues and the mechanism(s) by which this may occur. Finally, we will propose future directions that may help elucidate their role in aging and disease as well as the physiologic and cellular defenses against their accumulation by excretion or metabolic degradation.

2. Formation of abnormal tyrosine isomers, meta- and ortho-tyrosine by oxidative stress

Apart from the optical isomers D- and L-tyrosine, there exist three structural isomers of tyrosine— para-, meta-, and ortho-tyrosine—that differ according to the position of their hydroxyl group on the benzyl side chain (Figure 1). The para isoform is the one depicted in biochemistry textbooks and the major isoform involved in metabolism and protein synthesis in the cell. While enzymatic pathways have been identified for the synthesis of m-tyrosine in certain species of bacteria (Zhang et al., 2011) and plants (Huang et al., 2012; Muller and Schutte, 1967), in animals the enzymatic oxidation of phenylalanine by phenylalanine hydroxylase appears to occur exclusively on carbon 4, which produces p-tyrosine (Halliwell and Whiteman, 2004; Kaur and Halliwell, 1994). However, under conditions of oxidative stress when levels of free radicals are elevated or following exposure to ionizing radiation, non-enzymatic hydroxylation of phenylalanine may occur resulting in the formation of the abnormal tyrosine isomers m- and o-tyrosine in addition to p-tyrosine (Figure 2) (Davies et al., 1999; Mager and Berends, 1974; Maskos et al., 1992).

Figure 1.

The enzymatic and radical hydroxylation of phenylalanine

Figure 2.

Mechanisms of abnormal tyrosine isomer formation. Utilizing m-tyrosine as a general example, there are multiple mechanisms by which non-enzymatic hydroxylation of phenylalanine may occur. A hydroxyl radical may oxidize the benzyl ring of phenylalanine, and the resultant radical intermediate may rapidly undergo abstraction, oxygenation, or disproportionation reactions to form the stable tyrosine isomer. Peroxynitrite may also react with phenylalanine to generate tyrosine isomers. Photoionization of phenylalanine or phenylalanine’s reaction with strong one-electron oxidants may result in the formation of a phenylalanine radical cation, and subsequent reactions with water will also form tyrosine isomers.

As depicted by Figure 2, the formation of these tyrosine isomers by hydroxyl radical oxidation of phenylalanine involves a two-step process in which the hydroxyl radical first attacks the phenyl ring as an addition reaction to produce the highly reactive hydroxyphenylalanine radical intermediate. Next, this intermediate product quickly undergoes a secondary reaction to generate the stable tyrosine isomer by one of three mechanisms: (1) abstraction, in which a second hydroxyl radical or other free radical can steal the hydrogen atom bonded to the hydroxylated carbon of the phenyl ring and, in the reaction with another hydroxyl radical, release water (Mujika et al., 2013; Wang et al., 1993); (2) oxygenation, in which the radical intermediate reacts with oxygen with the subsequent release of a hydroperoxyl radical (Maleknia and Downard, 2001); or (3) disproportionation, in which two radical intermediates react with one another producing one phenylalanine molecule, one tyrosine isomer, and water (Kaur and Halliwell, 1994; Solar, 1985). In addition to reactions with hydroxyl radical, the non-enzymatic hydroxylation of phenylalanine may also occur by alternative mechanisms so that the formation of m- and o-tyrosine are not unique to hydroxyl radicals. Powerful oxidants (e.g. sulfate radical anion) or photoionization via UV light may generate a phenylalanine radical cation by removal of an electron from the aromatic ring, which may subsequently react with water to produce the tyrosine isomers (Davies et al., 1999; Hawkins and Davies, 2001). Furthermore, the addition of peroxynitrite to an in vitro system also resulted in the formation of m- and o-tyrosine; however, it should be noted that hydroxyl radical scavengers reduced the generation of these isomers, suggesting the ability of peroxynitrite to hydroxylate phenylalanine occurs at least in part through the intermediate formation of hydroxyl radicals (Kaur et al., 1997). The radical oxidation process can affect either the free phenylalanine pool or phenylalanine that is incorporated into proteins (Kaur et al., 1996; Kelly and Lubec, 1995; Nair et al., 1995; Pennathur et al., 2001; Stadtman and Levine, 2003).

2.1 Ratio of m- and o-tyrosine formation

Multiple studies have sought to determine whether phenylalanine hydroxylation in the setting of oxidative stress occurs indiscriminately along the aromatic ring or if there are positional preferences. Molecular modeling studies have suggested that the attack of the hydroxyl radical is energetically favorable at all sites on the aromatic ring over attack of the carbons in the amino acid backbone, and that there should be no significant preference for the para, meta, or ortho positions based on thermodynamic, kinetic, and steric data (Mujika et al., 2013). However, other work provides contradictory data. For example, using in vitro reactions, multiple research groups reported o-tyrosine to be the preferential product of the reaction of hydroxyl radical and phenylalanine followed by the formation p-tyrosine and then m-tyrosine. However, the degree of these preferences varied due likely to differences in experimental design. The o-:p-:m-tyrosine isomer ratio in aqueous solutions following irradiation has been found to vary between 1.0:0.60:1.0 (o-:p-:m-tyrosine) (Simic et al., 1985) and 1.0:0.60:0.28 (Solar, 1985)—the later data is consistent with mathematical modeling predicting a ratio of 1.0:0.37:0.11 (Galano and Cruz-Torres, 2008). The ratio of the tyrosine isomers formed also varies depending on the chemical oxidizer utilized or the addition or absence of secondary oxidants following solution irradiation. For example, the o-:p-:m-tyrosine ratio in irradiated aqueous solution was 1.0:0.5:0.6 with the addition of iron monocyanide; 1.0:1.0:0.9 with the addition of oxygen; and 1.0:1.5:1.4 in the absence of a secondary oxidant (Wang et al., 1993). Likewise, in vitro metal-catalyzed oxidation of human blood plasma resulted in a greater production of free m-tyrosine than o-tyrosine (m-:o-tyrosine = 1.0:0.48) (Blount and Duncan, 1997). Based on the above studies, there seems to be a slight general preference for o-tyrosine formation in vitro; although, p- and m-tyrosine are also produced in appreciable quantities.

3. Abnormal tyrosine isomers as biochemical markers of oxidative stress

Because of their suspected role in aging and disease, multiple techniques, including the goldstandard approach of electron spin resonance (ESR), have been developed over the years to detect and measure reactive oxygen species (Hawkins and Davies, 2014). However, by nature of its high reactivity, the hydroxyl radical is extremely short-lived, and therefore, it cannot accumulate to levels high enough for direct detection by ESR (Halliwell and Whiteman, 2004). This renders direct quantification of the hydroxyl radical challenging. Because hydroxyl radicals readily react with aromatic compounds (Stein and Weiss, 1950), the hydroxylation of aromatic traps was developed as a surrogate measurement for the presence of hydroxyl radicals in a system (Freinbichler et al., 2011); and phenylalanine—representing an aromatic compound that may be non-enzymatically hydroxylated to form stable hydroxyl adducts (see Section 2)—began to be extensively used for this purpose in the mid-1980s as in vitro data established the utility of this method.

Ishmitsu et al. showed not only that m- and o-tyrosine were formed when phenylalanine was added to a solution containing hypoxanthine and xanthine oxidase under aerobic conditions but also that the addition of antioxidants (superoxide dismutase, catalase, or hydroxyl radical scavengers (e.g. potassium iodide, potassium bromide, mannose, etc.)) or pro-oxidants (lactoferrin or Fe³⁺) could inhibit or augment phenylalanine hydroxylation, respectively (Ishimitsu et al., 1984). Likewise, incubation of phenylalanine with a mixture of Fe²⁺, ethylenediaminetetraacetic acid (EDTA), and hydrogen peroxide (H₂O₂), which are known to generate hydroxyl radicals via the Fenton reaction (Haber and Weiss, 1934; Winterbourn, 1995), resulted in the formation of all three tyrosine isomers (Kaur et al., 1988). In addition, m- and o-tyrosine were detected following gamma-radiolysis of an aqueous solution containing phenylalanine (Maskos et al., 1992). Thus, the production of abnormal tyrosine isomers from free phenylalanine is a sensitive measurement of hydroxyl radical formation in vitro. It was further demonstrated that similar hydroxylation by hydroxyl radicals occurs to phenylalanine incorporated into protein (Huggins et al., 1993) and that neither tissue homogenization nor the hydrolysis of proteins necessary for the measurement of specific amino acids induces radical damage or skews quantification (Aronson and Wermus, 1965; Fu et al., 1998a; Leeuwenburgh et al., 1998). Following these proof of concept studies, the measurement of m- and o-tyrosine as markers of reactive oxygen species was first employed in the food industry. As the use of ionizing radiation for purposes of food preservation became more popular, there were concerns regarding the effect of irradiation on food products. Consequently, m- and o-tyrosine were proposed as biomarkers for the quantification of hydroxyl radicals within food, and their formation in cubed chicken breast subjected to irradiation was demonstrated (Karam and Simic, 1988).

Later the use of aromatic traps was expanded to detect hydroxyl radicals in vivo, and this methodology was applied more directly to biomedical research. Although salicylate was initially suggested for this purpose (Grootveld and Halliwell, 1986; Maskos et al., 1990), its known pharmacologic and toxic effects (e.g. inhibition of prostaglandin production, ototoxicity, etc.) raised concerns for its usage within organisms (Brien, 1993; Kaur and Halliwell, 1994). By consequence, phenylalanine was assessed for its effectiveness in detecting hydroxyl radicals in vivo and also compared to other known aromatic traps (Maskos et al., 1992; Themann et al., 2001). Although the rate constant of the reaction between phenylalanine and hydroxyl radical was found to be lower than it is with salicylate (Maskos et al., 1992), the assumption that phenylalanine—as an amino acid endogenous to cells—would have fewer toxic side effects made it an favorable alternative (Kaur and Halliwell, 1994). As a result, since the 1990s the formation of m- and o-tyrosine has been measured extensively in physiological and diseased states an indicator of hydroxyl radicals and oxidative stress, and various methods (e.g. highperformance liquid chromatography, gas chromatography, mass spectroscopy, etc.) have been optimized for their specific detection (Halliwell and Kaur, 1997; Heinecke et al., 1999; Kaur et al., 1988; Kaur and Halliwell, 1994; Li et al., 2003; Reddy et al., 1999).

4. Association of m- and o-tyrosine with disease states

When measured as a biomarker for hydroxyl radicals, elevations in the concentrations of both free and protein-bound m- and o-tyrosine have been observed in many diseases in which oxidative stress is thought to play a pathological role. Table 1 summarizes the literature that has utilized the identification of tyrosine isomers in this manner. Of note, increases in m- and/or o-tyrosine have been detected in aging cardiac muscle (Leeuwenburgh et al., 1997b) and in multiple diseases including, among others, diabetes (Brasnyo et al., 2011; Molnar et al., 2005b; Pennathur et al., 2005; Pennathur et al., 2001), cataract formation (Fu et al., 1998b; Wells-Knecht et al., 1993), atherosclerosis (Fu et al., 1998a; Leeuwenburgh et al., 1997a), and hyperoxia-induced chronic lung disease (Kelly and Lubec, 1995; Lubec et al., 1997).

Table 1.

The use of m- and o-tyrosine as indicators of oxidative stress in physiological and pathological states

Publication	Disease/Condition	Species	Measurement	Findings
Sun et al. (1993)	Myocardial Ischemia	Dog	Plasma m- and o-tyrosine	By intravenously infusing phenylalanine (either solely or combined with antioxidants) into dogs subjected to induced myocardial ischemia and then measuring resultant increases in plasma m- and o-tyrosine, Sun and colleagues provided evidence for hydroxyl radical production following myocardial ischemia in vivo and a causal role of increased reactive oxygen species in myocardial stunning.
Wells-Knecht et al. (1993)	Noncataractous lens	Human	Protein-bound o-tyrosine within healthy cataracts	No statistically significant differences in the measurement of o-tyrosine incorporated into lens proteins was found with age.
Kelly and Lubec (1995)	Hyperoxia-induced chronic lung disease	Guinea pig	Protein-bound o-tyrosine within pulmonary tissue	Whereas lung tissue o-tyrosine concentration remained constant over the course of 28 days in preterm pups maintained in normoxic conditions (21% oxygen), o-tyrosine concentration significantly increased in pups maintained in hyperoxic conditions (85% oxygen), rising from 0.51% of the total tyrosine pool on day 7 to 1.45% on day 28.
Nair et al. (1995)	Betel quid chewing/oral cancer	Human	Free m- and o-tyrosine in saliva	Saliva collected following betel quid chewing to which phenylalanine had been supplemented contained significantly higher concentrations of m- and o-tyrosine compared to controls (2025 ± 821 nM vs. 20 ± 11 nM for m-tyrosine and 1764 ± 773 nM vs. 39 ± 21 nM for o-tyrosine) supporting the hypothesis that betel quid chewing increases the risk for oral cancers by generating hydroxyl radicals.
Kaur et al. (1996)	Rheumatoid arthritis	Human	Free m- and o-tyrosine in synovial fluid and blood	Synovial fluid was aspirated from the knee of 53 patients with active rheumatoid arthritis, and when added to saline containing phenylalanine, tyrosine isomers could be detected in 36 of the patients (whereas none could be detected when added to saline alone); similar results were observed with blood withdrawn from these same patients suggesting elevated reactive oxygen species may contribute to rheumatoid arthritis progression.
Lubec et al. (1996)	Hydrogen peroxide poisoning	Human	Protein-bound o-tyrosine within brain tissue	By error 100 mL of 3% hydrogen peroxide was fatally administered instead of saline to a 7 month-old boy; post-mortem analysis of brain tissue revealed twice the concentration of o-tyrosine in the child administered H2O2 compared to 5 controls (0.328 nmol/g brain tissue vs. 0.163±0.013 nmol/g).
O’Neill et al. (1996)	Myocardial ischemia/reperfusion	Cat	Free m- and o-tyrosine in blood	Cats were administered phenylalanine in saline immediately before they were subjected to varying lengths of ischemia (2, 5, and 10 min.) by reversible occlusion of the left anterior descending artery, and the formation of tyrosine isomers was measured during reperfusion; elevations in m- and o-tyrosine concentrations corresponded to lengthened time of ischemia suggesting hydroxyl radical production correlates to extended periods of ischemia.
Frischer et al. (1997)	Ozone exposed airways	Human	o-Tyrosine in nasal lavage	The ratio of o-tyrosine to p-tyrosine detected nasal lavages from children was increased following days of high ozone exposure (>180 μg/m3) compared to the ratio following days of low exposure (<140 μg/m3) with mean ratios being 0.18 vs. 0.02, respectively.
Leeuwenburgh et al. (1997a)	Atherosclerosis	Human	m- and o-tyrosine incorporated in low-density lipoproteins (LDL) and vascular proteins	LDL was extracted from atherosclerotic lesions or from the circulating plasma pool; following protein hydrolysis, concentrations of m- and o-tyrosine were measured but not found to differ across samples; similarly, the concentration of tyrosine isomers did not statistically change within aortic vascular tissue at varying stages of atherosclerosis.
Leeuwenburgh et al. (1997b)	Aging tissues	Mouse	Protein-bound o-tyrosine within skeletal muscle, heart, liver and brain tissue	Cardiac tissue o-tyrosine concentrations increased significantly between young mice (4 months) and aged mice (14 months); and while there was a trend towards increasing o-tyrosine concentrations in skeletal muscle and brain tissue, these values did not achieve significance—perhaps due to the small sample size. No differences were seen within the liver with age nor did caloric restriction greatly affect measurements.
Lubec et al. (1997)	Oxygen-treated infants	Human	Urinary o-tyrosine	Urinary o-tyrosine concentration expressed as a percentage of phenylalanine was found to be elevated in infants who received supplemental oxygen compared to infants on room air (0.40 ± 0.028 percent vs. 0.18 ± 0.012 percent, respectively).
Fu et al. (1998a)	Atherosclerosis	Human	Protein-bound m- and o-tyrosine within atherosclerotic plaques	Atherosclerotic plaques contained increased concentrations of protein-bound tyrosine isomers compared to the concentrations measured in intimal tissue of the iliac arteries obtained from non-atherosclerotic patients (3.35 ± 1.88 nmol/g weight of tissue vs. 0.71 ± 0.17 nmol/g for m-tyrosine and 13.53 ± 4.19 nmol/g vs. 5.90 ± 1.77 nmol/g for o-tyrosine).
Fu et al. (1998b)	Cataractogenesis	Human	Protein-bound m- and o-tyrosine within cataractous lens	The severity of the cataract lens (as indicated by classification) directly correlated with increased concentrations of m- and o-tyrosine detected in samples. Very low levels were measured in normal lens tissue, but a near 10-fold increase was observed when measured in nigrescent Type IV cataractous lenses.
Fu et al. (1998c)	Advanced glycation end products in collagen	Rat	Protein-bound m-tyrosine within collagen	Exposure of collagen extracted from rats’ tails was incubated in a buffer containing glucose under aerobic conditions for three weeks, and this led to weekly increases in the concentration of m-tyrosine. Incubation under anaerobic, anti-oxidative conditions did not produce any increases in m-tyrosine during this same period
Leeuwenburgh et al. (1998)	Aging tissues	Rat	Protein-bound o-tyrosine within cardiac, skeletal muscle and liver tissues	The concentrations of o-tyrosine were not found to change in cardiac, skeletal muscle, and liver tissue between young (9 months) and old (24 months) rats, nor did antioxidant supplements alter the concentrations in any of these tissues.
Lamb et al. (1999)	Acute respiratory distress syndrome	Human	Protein-bound o-tyrosine within bronchoalveolar lavage fluid	Compared to healthy controls, bronchoalveolar lavage fluid from patients with acute respiratory distress syndrome (ARDS) had significantly higher concentrations of o-tyrosine (10.43 ± 5.79 μmol/g total protein in ARDS vs. 0.58 ± 0.22 μmol/g total protein in controls).
Davies et al. (2001)	Aging mitochondria	Rat	Protein-bound m- and o-tyrosine within cardiac, liver, and brain tissues	No statistical differences were observed between tyrosine isomer concentrations within the cell homogenates, cytosol, or mitochondria of any tissue between young (2–3 months) and old (24 months) rats; nor were there differences found between o- and m-tyrosine concentrations between the mitochondrial matrix and membrane in young vs. old rats.
Pennathur et al. (2001)	Diabetic vascular disease	Cynomolgus monkey	Protein-bound m- and o-tyrosine within vascular wall of the thoracic aorta	Proteins from aortas dissected from diabetic monkeys contained 60% and 40% more m and o-tyrosine, respectively, compared to controls (0.18 ± 0.01 mmol/mol phenylalanine vs. 0.11 ± 0.01 mmol/mol for m-tyrosine and 0.36 ± 0.01 mmol/mol vs. 0.26 ± 0.01 mmol/mol for o-tyrosine). Furthermore, the concentrations of the tyrosine isomers correlated closely with serum levels of glycated hemoglobin (r2 = 0.80 for m-tyrosine and r2 = 0.91 for o-tyrosine).
Ogihara et al. (2003)	Hypoxic ischemic encephalopathy	Human	Free m- and o-tyrosine within cerebral spinal fluid	Cerebral spinal fluid from infants with hypoxic ischemic encephalopathy had significantly higher concentrations of m- and o-tyrosine than healthy controls (20.5 ± 24.9 nM vs. 8.0 ± 3.1 nM for m-tyrosine and 20.5 ± 18.6 nM vs. 8.7 ± 2.6 nM for o-tyrosine).
Molnar et al. (2005a)	Cataractogeneis	Human	Protein-bound m- and o-tyrosine in soluble and insoluble protein fractions within the lens	Within the total protein homogenate, m- and o-tyrosine were significantly elevated in cataractous lenses compared to controls (median: 20.28 nmol per gram of protein in cataractous lenses vs. 3.41 nmol/g in controls for m-tyrosine; 211.78 nmol/g in cataractous lenses vs. 38.61 nmol/g in controls for o-tyrosine
Molnar et al. (2005b)	Diabetes mellitus and chronic kidney disease	Human	Urinary and plasma o-tyrosine	The urinary concentration of o-tyrosine was significantly higher in both diabetic and chronic kidney diseased (CKD) patients compared to healthy controls (interquartile ranges: 2.72–4.99 μmol/day for diabetics, 0.94–1.83 μmol/day for CKD, and 0.00–0.35 μmol/day for controls). While the plasma concentration of o-tyrosine followed a similar trend, the values failed to reach significance.
Pennathur et al. (2005)	Diabetes	Rat	Protein-bound m- and o-tyrosine within retinal tissue	Hyperglycemic rats had 85% and 67% higher levels of m- and o-tyrosine, respectively, within retinal tissues than controls (0.19 ± 0.02 mmol/mol phenylalanine vs. 0.11 ± 0.01 mmol/mol for m-tyrosine and 0.35 ± 0.02 mmol/mol vs. 0.21 ± 0.01 mmol/mol for o-tyrosine). No differences in tyrosine isomers were observed in hyperlipidemia rat models alone.
Brasnyo et al. (2011)	Diabetes	Human	Urinary o-tyrosine	Caucasian males with type II diabetes were treated with either resveratrol (n=10) or placebo (n=9) for four weeks. Compared to controls, resveratrol significantly improved insulin resistance (as measured by HOMA-IR, p = 0.044) and reduced urinary o-tyrosine concentration (p = 0.043).
Escobar et al. (2013)	Chronic fetal hypoxia	Human	Free m-tyrosine within amniotic fluid	m-Tyrosine concentration measured within amniotic fluid from 19 pregnant women with either type I or gestational diabetes strongly and significantly correlates with erythropoietin levels (r = 0.90), which is a clinical marker of fetal hypoxia.
Szelig et al. (2015)	Sepsis	Human	Plasma and urinary m and o-tyrosine	Compared to healthy controls, septic patients had higher m-tyrosine serum concentrations on days 2 and 3 of hospital admission (16 nM (day 2) & 21 nM (day 3) vs. 4 nM) with no significant differences in urine concentration. The reverse was true for o-tyrosine with elevated urinary o-tyrosine concentration but not serum concentration in septic patients versus controls. Furthermore serum procalcitonin (PCT) concentration, which is used as a clinical marker of sepsis severity, paralleled m-tyrosine levels in septic patients.
Torres-Cuevas et al. (2016)	Hypoxia	Pig	Plasma m- and o-tyrosine and protein-bound m- and o-tyrosine within liver and brain tissue	Compared with normoxic controls, an elevated concentration of m-tyrosine was found to be significantly elevated within the liver of hypoxia-treated piglets (253 ± 35 mmol m-tyrosine/mol phenylalanine in treated vs. 132 ± 19 mmmol m-tyrosine/mol phenylalanine in controls). No significant differences were seen in o-tyrosine concentration within the liver nor were there any differences observed within brain tissue or plasma.

Interestingly, in several studies the concentrations of m- and/or o-tyrosine directly correlated with other indicators of disease severity. For example, in the context of diabetic vascular disease, cynomolgus monkeys with streptozotocin-induced diabetes exhibited elevated concentrations of m- and o-tyrosine in aortic tissue, and these measurements correlated with serum glycated hemoglobin (HbA_1C) with the coefficient of determination for both m-tyrosine (r² = 0.80) and o-tyrosine (r² = 0.91) achieving statistical significance (p < 0.001) (Pennathur et al., 2001). Additionally, in the human lens, m- and o-tyrosine levels paralleled the severity of cataract with virtually no detectable tyrosine isomers in the normal lens but increasing amounts measured in every classification up to Type IV (Fu et al., 1998b). The correlation between tyrosine isomer concentrations and disease states may be interpreted three ways: (1) as the underlying disease progresses, more reactive oxygen species are produced due to reductions in cellular functions, and the tyrosine isomers are generated as downstream byproducts and are simply a lagging marker of the disease state; (2) elevated reactive oxygen species play a causative role in disease pathogenesis, and the coincident production of tyrosine isomers becomes a sensitive marker of the underlying levels of oxidative stress promoting disease pathogenesis; or (3) the accumulation of tyrosine isomers has a direct role in disrupting normal cellular homeostasis and provoking disease, and the changes in tyrosine isomer levels represent a driving force contributing to the onset or the rate of progression of a specific disease. While initial consideration may discount the third option under the claim that correlation does not signify causation, this hypothesis becomes more intriguing as studies elucidate the effects of m- and o-tyrosine in cells and tissues.

5. meta- and ortho-tyrosine are directly toxic to cells and organisms

Far from simply being a surrogate measurement for the presence of hydroxyl radicals, emerging evidence suggests m- and o-tyrosine may directly contribute to the toxic effects of oxidative stress in cells and tissues. Notably, as will be discussed in the subsequent subsections, the adverse effects of these abnormal tyrosine isomers have been observed across species from bacteria and plants to mammalian cells and vertebrate animals.

5.1. Effects on bacteria

The first report of the toxic effects of these isomers was published over 50 years ago when Smith et al. noted that the addition of either 0.06 mM m-tyrosine or 0.02 mM o-tyrosine to culture media inhibited the growth of E. coli by 50 percent (Smith et al., 1964). A year later, Aronson and Wermus published data further showing the growth inhibitory effect of m-tyrosine as 1 mM supplementation to culture caused significant reduction in the growth rate of E. coli and multiple species of Bacillus with mass doubling times being increased one- to two-fold—although it should be noted that they did not observe any reduction to the maximal growth of these microorganisms. Sporulation of three of the four Bacillus species tested was also reduced (Aronson and Wermus, 1965). However, a recent study reported no effects on the growth rate of wild-type E. coli when m-tyrosine (3 mM) was supplemented to minimal media (Bullwinkle et al., 2014).

5.2. Effects on yeast

The growth inhibitory effects of the tyrosine isomers also extend to eukaryotic organisms. Bullwinkle et al. demonstrated that the addition of 0.3 mM m-tyrosine to minimal media reduced the growth rate of wild type S. cerevisiae strains by approximately 50 percent. In contrast, the addition of p-tyrosine had no effect on growth rate even at concentrations four-fold higher than what was observed for m-tyrosine (Bullwinkle et al., 2014).

5.3. Effects on plants

Although m-tyrosine synthesis by the myrtle spurge (Euphorbia mysinites) had been discovered previously (Mothes et al., 1964), the phytotoxic property of m-tyrosine was first described in 2007 when Bertin et al. identified certain species of fescue grass that secrete m-tyrosine within their root exudates to render allelopathic advantage (i.e. the process of producing biochemicals by one species that suppress the growth and development of neighboring organisms of differing species) (Bertin et al., 2007; Movellan et al., 2014). Treatment with m-tyrosine was shown to inhibit root growth of a variety of plants (e.g. Lactuca sativa (lettuce), Arabidopsis thaliana, blue fescue, crabgrass, dandelion, tomato) with the half maximal inhibitory concentrations (IC₅₀) ranging between 10–160 μM. Surprisingly, o-tyrosine, by contrast, stimulated root growth of lettuce plants when tested (Bertin et al., 2007).

5.4. Effects on insects

Through a series feeding assays, Gautam and Henderson demonstrated the detrimental effects of m-tyrosine on termite survival. They administered m-tyrosine by pipetting various concentrations of m-tyrosine solution onto dry filter paper, which the termites then ingested as food. After 12 days, survival of the termites was assessed, and whereas 100 percent survival was observed in the control group, the highest concentration of m-tyrosine (50 mM) was found to reduce termite survival by approximately 60 percent. Additionally when multiple filter papers each prepared with different concentrations of m-tyrosine were simultaneously offered as food, the termites showed definitive preference to filter paper prepared with control and the lowest concentration of m-tyrosine (5 mM) compared to filter paper prepared with higher concentrations as determined by measurement of the amount of filter paper consumed (Gautam and Henderson, 2008).

5.5. Effects on mammalian cells

Utilizing Chinese-hamster ovary (CHO) cells incubated with various concentrations of m-tyrosine, Gurer-Orhan, et al. first provided evidence for the toxic effects of the abnormal tyrosine isomers to mammalian cell cultures. Following a 24-hour incubation period with m-tyrosine, CHO cells showed evidence of decreased cell vitality as indicated by the inhibition of colony formation, the diminished ability to reduce the tetrazolium dye MTS, and the increased release of lactate dehydrogenase (LDH). Colony formation was inhibited in a concentration-dependent manner with 60 percent reduction observed when the cells were incubated with 0.25 mM m-tyrosine compared to control incubation with 1 mM p-tyrosine. The reduction of MTS by NAD(P)H-dependent dehydrogenase enzymes produces a formazan product that can be detected by colorimetric means and can serve as a correlative measurement to the number of metabolically active cells in culture. A concentration-dependent inhibition of MTS reduction was observed when CHO cells were treated with m-tyrosine with nearly a 50 percent decrease when the cells were incubated with 0.5 mM m-tyrosine. Finally, LDH release was reported to increase when cells were incubated with high concentrations of m-tyrosine, which suggests at these doses m-tyrosine is sufficient to cause cell damage (Gurer-Orhan et al., 2006).

Work by Mikolás et al. further supported the inhibitory effects of m- and o-tyrosine on the proliferation of cultured cells. The addition of 110 μM of either m- or o-tyrosine to the culture media of TF-1 erythroblasts significantly reduced erythropoietin-induced cell proliferation. This effect is likely mediated by the uptake of these isomers into cells because supplementing the culture with additional p-tyrosine counteracted this proliferative decline (Mikolas et al., 2013).

It should also be noted for cell culture experiments that all of the commonly used media contain elevated levels of tyrosine and phenylalanine especially compared to fasting plasma levels. Consequently, there is likely significant competition between the tyrosine isomers and phenylalanine or tyrosine for cellular proteins including membrane transporters for entry into the cell.

5.6. Effects on cancer cells

Select primary tumors are able to inhibit the growth of secondary tumors and metastasis within the host through a phenomenon called concomitant tumor resistance (Chiarella et al., 2012). Although this phenomenon was initially believed to be mediated by T-cell-dependent processes, Ruggiero et al. produced data suggesting these tumors induce the production of reactive oxygen species by myeloid-derived suppressor cells, which in turn generate m- and o-tyrosine that are capable of inhibiting secondary tumor growth. Using a non-immunogenic lymphoma that confers concomitant tumor resistance, m- and o-tyrosine were identified within the serum of tumor-bearing mice and found to cause secondary tumor growth inhibition. These results were confirmed by additional in vitro and in vivo experiments. Both isomers inhibited lymphoma cell proliferation in vitro with m-tyrosine producing the greater effect (50 percent reduction in proliferation by a concentration of 24.8 ± 5.0 μM for m-tyrosine versus 257.7 ± 28.7 μM for o-tyrosine). The suppression of cell proliferation was reversible because the replacement of tyrosine isomer-supplemented culture media with fresh media without isomers permitted cells to proliferate in a manner similar to controls. Furthermore, excess phenylalanine neutralized the effects of m- and o-tyrosine, which suggests that phenylalanine and m- and o-tyrosine may compete for cell transporters or perhaps, as will be discussed in 6.1, for cellular targets, such as charging to phenylalanine tRNA (tRNA^Phe). Mice injected with m-tyrosine became resistant to lymphoma, fibrosarcoma, and epidermoid carcinoma implants as well as lung metastases of mammary adenocarcinoma (Ruggiero et al., 2011). More recent work has shown m-tyrosine injections given to mice with metastatic mammary carcinoma (C7HI or LMM3 strain) significantly reduced the number of lung metastases; and m-tyrosine treatment following surgical removal of LMM3 primary tumors drastically improved the survival percentage of the treated animals 140 days post-surgery (Machuca et al., 2015).

5.7. Effects on mammals

Rats supplemented with o-tyrosine (1.76 mg/day) for four weeks had significantly elevated levels of o-tyrosine within proteins of the vascular wall of the femoral artery (mean ± SEM: 0.04 ± 0.011 mmol/mol phenylalanine for controls vs. 0.08 ± 0.004 mmol/mol phenylalanine for o-tyrosine-supplemented). This correlated with a significant reduction in insulin-induced relaxation of femoral arteries isolated from o-tyrosine supplemented rats compared to controls (Szijarto et al., 2014). Further experiments observed an inverse correlation between o-tyrosine concentration within arterial walls and acetylcholine-induced vasorelaxation (Molnar et al., 2015).

6. Potential mechanisms for the adverse effects of m- and o-tyrosine

Despite the significant number of studies showing the adverse effects of m- and o-tyrosine to organisms, the mechanisms accounting for these effects are poorly understood. However, ongoing research in the field has suggested several possibilities. These mechanisms are also graphically depicted in Figure 3.

Figure 3.

Suggested mechanisms by which m- and o-tyrosine mediate oxidative stress are shown with blue arrows. Known and proposed protective antioxidant enzymes and pathways are shown with red arrows. Abbreviations: ROS (reactive oxygen species); SOD (superoxide dismutase), UPR (unfolded protein response), HSR (heat shock response).

6.1. Aberrant charging to phenylalanine tRNA

The incorporation of free m- and o-tyrosine into proteins during protein synthesis has been observed when bacteria, plants, cells, and rats are treated with exogenous m- or o-tyrosine ((Aronson and Wermus, 1965; Bertin et al., 2007; Gurer-Orhan et al., 2006; Rodgers et al., 2002; Szijarto et al., 2014). Their incorporation is dependent on protein synthesis because cycloheximide treatment inhibits m-tyrosine from being incorporated into proteins by 88 percent (Gurer-Orhan et al., 2006). An important effect following the treatment of cells with tyrosine isomers is the substitution of m- and o-tyrosine for phenylalanine (Gurer-Orhan et al., 2006; Klipcan et al., 2009; Popp et al., 2015). This is due to the ability of phenylalanyl-tRNA synthetase to erroneously charge phenylalanine tRNA (tRNA^Phe) with the atypical tyrosine isomers (Bullwinkle et al., 2014; Klipcan et al., 2009; Popp et al., 2015; Smith and Fowden, 1968). This mischarging was calculated to occur at 27 percent of the efficiency of the correct charging of phenylalanine in the mung bean plant (Smith and Fowden, 1968). More recent kinetic data has indicated that the catalytic efficiency of human mitochondrial phenylalanyl-tRNA synthetase to attach m-tyrosine to tRNA^Phe is 20 percent that of phenylalanine; and while m-tyrosine charging onto tRNA^Phe does occur in the cytosol, the efficiency of cytosolic phenylalanyl-tRNA synthetase to charge tRNA^Phe with the tyrosine isomers has not been clearly determined (Klipcan et al., 2009). While post-transfer editing normally deacylates mischarged tRNA and significantly reduces the rate of amino acid substitution based on the rate of mischarged tRNA alone, only minimal deacylation of tRNA^Phe charged with m-tyrosine occurs in humans (Klipcan et al., 2009). This poor editing can perhaps lead to the disproportionate accumulation of tRNA^Phe charged with m-tyrosine over time. Furthermore, even low rates of substitution are enough to cause adverse consequences. In E. coli, a one-percent misincorporation rate of m-tyrosine for phenylalanine in proteins was sufficient to significantly restrict growth and impair cellular viability (Bullwinkle et al., 2014).

The cellular consequences of substituting a hydrophilic amino acid for a hydrophobic acid could be potentially extensive, particularly with a high level of substitution. Paredes et al. studied the global effects of substituting serine for leucine—which have comparable amino acid hydrophobicity ranks to tyrosine and phenylalanine, respectively—and showed that even a low level of amino acid mistranslation that does not affect growth rate in yeast (~1%) is sufficient to up-regulate genes associated with oxidative stress, ER stress, and heat shock while down-regulating genes related to protein synthesis. At the same time the enzymatic activity of specific proteins was reduced, and protein aggregation increased (Paredes et al., 2012). It is therefore possible that the resulting substitution of a polar tyrosine isomer for non-polar phenylalanine during protein synthesis produces adverse changes to protein structure and activity and elicits various downstream cellular response pathways, including activation of the unfolded protein response, the heat shock response, augmented proteasome activity, and/or promotion of apoptosis.

While less studied, it is possible that the tyrosine isomers could also be charged to tRNA^Tyr and substituted for p-tyrosine in proteins (Bullwinkle et al., 2014). While the chemical differences between p-tyrosine and m- or o-tyrosine are less marked, the location of the hydroxyl group of p-tyrosine is often critical for the folding of specific protein domains, the activity of enzyme catalytic sites, and/or phosphorylation by tyrosine kinases (Hunter, 2014). As a result, even the substitution of m- or o-tyrosine for p-tyrosine could produce important changes in the spatial location of the hydroxyl group and thereby protein function.

6.2. Enhanced protein degradation

When the J774 mouse macrophage cell line was incubated with 2 mM m-tyrosine, the degradation rates of leucine radiolabeled proteins increased compared to controls; but it was unclear whether this was a result of enhanced proteasome activity or autophagy (Rodgers et al., 2002). These results were confirmed by a follow-up study, which reported similar effects utilizing the same J774 cell line. Treatment with either 1 mM m-tyrosine or 1 mM o-tyrosine significantly increased the rate of protein degradation in these cells with m-tyrosine having the greater effect (Dunlop et al., 2008). Furthermore, while gene expression of the lysosomal enzyme cathespin B was enhanced in cells treated with either tyrosine isomer, cystatin B, which is an endogenous inhibitor of cathepsin proteases, was also significantly expressed in treated cells; and no changes in the enzymatic activity of cathespin B, cathespin L, or arylsulfatase could be detected when compared to controls. The impact of m- and o-tyrosine on the expression and activity of the 20S and 26S proteasomes was also measured, and only m-tyrosine was shown to have any effect, which was a diminished expression of the 19S proteasomal subunit; however, no changes to proteasomal activity were observed. Lastly, no detectable aggregation, as screened for by the detection of autofluorescence under fluorescence microscopy, occurred in cells treated with m- or o-tyrosine (Dunlop et al., 2008). As increased protein turnover can represent either a beneficial response or a sign of disrupted proteostasis, further work is needed to elucidate how and why these changes in overall protein degradation occur.

6.3. Induction of apoptosis

The potential for tyrosine isomers to induce apoptosis has been experimentally evaluated in a human monocytic cell line (THP1 cells) treated with o-tyrosine. Dunlop et al. incubated THP1 cells for 24 hours in phenylalanine-depleted media supplemented with 500 μM o-tyrosine and then assessed apoptosis via Annexin V staining and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL). By both measurements, apoptosis significantly increased approximately one- to two-fold compared to controls. The percentage of cells with lysosome membrane permeabilizaiton, as determined by the leakage of acridine orange dye from the lysosome following staining, and depolarized mitochondria were significantly greater in cells incubated with o-tyrosine 20 hours after the start of treatment. Caspase-3 activity was also elevated in o-tyrosine treated cells compared to controls. However, incubation of cells in complete media (i.e. containing abundant phenylalanine) neutralized the effects of o-tyrosine and eliminated the observed differences in all parameters measured suggesting that phenylalanine can compete with o-tyrosine either for transport into the cell, charging to tRNA, or effects on another cellular target (Dunlop et al., 2011).

6.4. Inhibition of cell proliferation

There is also evidence suggesting that m- and o-tyrosine can inhibit cell proliferation by disrupting mitogenic signaling pathways. Two reports have independently observed a reduction in the levels of phosphorylated ERK and STAT proteins following the treatment of cells with the tyrosine isomers. Ruggiero et al. showed lymphoma cells cultured in media with m-tyrosine (1.38 mM) had significantly reduced levels of phosphorylated ERK1/2 and STAT3 compared to controls, and the addition of phenylalanine counteracted these results (Ruggiero et al., 2012). Mikolás et al. reported similar results when TF-1 erythroblasts were incubated with either m- or o-tyrosine (110 μM). Treatment with tyrosine isomers inhibited the usual phosphorylation of the ERK1/2 and STAT5 proteins after stimulation with erythropoietin (Mikolas et al., 2013). It is worthwhile to note that the cell lines used for both studies have inherently high proliferative potential, and the reduction of mitogenic signaling may be characteristic only of the effects of m- and o-tyrosine on these types of cells.

6.5. Alternative mechanisms

While the above subsections represent all of the data to date regarding the mechanisms of m- and o-tyrosine toxicity, there remain other potential processes by which tyrosine isomers may elicit their adverse effects. The substitution of an abnormal tyrosine isomer for either phenylalanine or p-tyrosine in a protein may disrupt the tertiary structure and/or the catalytic site of the protein and therefore reduce its biologic activity. Additionally, the misfolded proteins may stimulate one or more cellular stress response pathways, such as activating the unfolded protein response due to the presence of unfolded proteins in the endoplasmic reticulum or mitochondria or the heat-shock response due to similar processes occurring in the cytoplasm. Lastly, due to the aberrant charging of m- or o-tyrosine to tRNA^Phe, there is potential to deplete the cellular pool of tRNA^Phe should tRNA^Phe charged with m- or o-tyrosine be utilized at a diminished rate by the ribosome. This could result in ribosomal stalling and perhaps the premature termination of protein translation.

7. How could tyrosine isomers contribute to the adverse effects of oxidative stress?

Resulting from the collection of data compiled over the course of the past half-century, the previously unknown roles of m- and o-tyrosine in mediating aspects of the harm produced by oxidative stress and contributing to the development of disease have emerged and are outlined in Figure 3. While it has long been known that amino acids and proteins can undergo oxidation (Garrison, 1972), it is now clear that abnormal tyrosine isomers are directly generated by hydroxyl radical attack of phenylalanine when levels of reactive oxygen species are elevated, and they can be formed within either intact proteins or in the free amino acid pool. Free m- and o-tyrosine may then be charged to tRNA^Phe and/or tRNA^Tyr and subsequently synthesized into proteins (Gurer-Orhan et al., 2006; Klipcan et al., 2009; Popp et al., 2015). In either case, the incorporation of abnormal tyrosine isomers can alter protein function and structure, which may in turn trigger cell stress responses and either directly or indirectly lead to a global increase in protein degradation. The degradation of proteins containing the tyrosine isomers could actually be detrimental by augmenting the cellular pool of free m- and o-tyrosine, and the cycle leading to the incorporation of the isomers into newly synthesized proteins would then continue. Thus the accumulation of abnormal tyrosine isomers within cells could have devastating consequences and contribute to disease pathogenesis. Indeed a role for tyrosine isomers has already been hypothesized in the development of insulin resistance (Kun et al., 2015; Molnar et al., 2015; Szijarto et al., 2014). With elevations of these isomers being noted in many other diseases (see Table 1), they may likely mediate some aspects of the disruption of cellular homeostasis produced in each of these pathologic processes as well. Despite their comparatively low physiological levels relative to the concentrations of other amino acids, the elevated ratio of m and o-tyrosine measured in disease states does not deviate far from their cellular concentration in feeding studies in which they were shown to induce adverse effects (Table 2). Thus the tight control of oxidative stress relies not solitarily on the antioxidant capacity of the cell (i.e. removal of free radicals) but also on its ability to eliminate stable, but toxic radical adducts such as m and o-tyrosine. Only when both processes fail would the diseased state become evident.

Table 2.

Physiological and pathological levels of m- and o-tyrosine

Organ/tissue	Physiological controls a		Pathological levels a			References
	m-Tyrosine	o-Tyrosine	m-Tyrosine	o-Tyrosine	Disease
Lens	0.003 (0.003–0.0042)	0.035 (0.011–0.046)	0.021* (0.007–0.059)	0.291* (0.054–1.167)	Cataract formation	Molnar et al. (2005a)
	0.13 ± 0.051	0.15 ± 0.041	3.53 ± 0.531,*	5.25 ± 0.081,*		Fu et al. (1998b)
		0.69 ± 0.14				Wells-Knecht et al. (1993)
Arterial tissue	0.020 ± 0.0031	0.182 ± 0.0171	0.122 ± 0.0201,*	0.468 ± 0.0451,*		Fu et al. (1998a)
	0.12 ± 0.041	0.28 ± 0.101	0.25 ± 0.111	0.75 ± 0.381	Atherosclerosis	Leeuwenburg et al. (1997a)
	0.11 ± 0.01	0.26 ± 0.01	0.36 ± 0.01*	0.18 ± 0.01*	Diabetes	Pennathur et al. (2001)
Plasma		0.54 (0.32–1.33)		1.82 (0.87–5.27)	CKD	Molnar et al. (2005b)
		0.54 (0.32–1.33)		0.74 (0.48–0.93)	Diabetes	Molnar et al. (2005b)
	0.351,2 (0.32–0.58)	0.791,2 (0.22–1.14)	0.681,2,* (0.55–1.15)	0.461,2 (0.24–0.77)	Sepsis	Szelig et al. (2015)
CSF (infant)	0.68 ± 0.24	0.75 ± 0.26	1.01 ± 0.55	1.1 ± 0.35	HIE	Ogihara et al. (2003)
CHO cells	1.011		7.061,3,*			Gurer-Orhan et al. (2006)
Femoral arterial tissue		0.04 ± 0.0111		0.08 ± 0.0041,4,*		Szijarto et al. (2014)

Values are expressed as mmol tyrosine isomer per mol phenylalanine (except where otherwise indicated) and represent either mean ± SD/SEM or median (interquartile range). CHO cells and femoral arterial tissue are presented for comparison against typical levels of tyrosine isomers present in feeding experiments.

^*p < 0.05 compared to physiological controls

¹Values are estimated by WebPlotDigitizer, a web based tool to extract numerical data from plots and graphs.

²Values are expressed as mmol m- or o-tyrosine per mol p-tyrosine and are representative of concentrations of isomers within plasma on the 2^nd day of hospitalization

³m-Tyrosine incorporated into CHO cells incubated with 0.2 mM m-tyrosine, which causes ~40% reduction in colony formation

⁴Following o-tyrosine supplementation (1.76mg/day) for 4 weeks, which significantly reduced insulin-induced arterial relaxation

7.1. Clearance of tyrosine isomers as an anti-oxidative response

In conjunction with our above hypothesis, the clearance of m- and o-tyrosine would therefore serve as a protective mechanism during conditions of oxidative stress.

On the organismal level, there is some evidence to suggest that m- and o-tyrosine can be cleared at least in part via the renal system. Molnar et al. compared the renal excretion of p-tyrosine versus o-tyrosine in patients with chronic kidney disease (CKD) and/or diabetes and healthy controls. Utilizing the fractional excretion (FE) for each tyrosine isomer, which is the percentage of the isomer filtered by the kidney and then excreted in the urine and is calculated by dividing the clearance of the measured isomer by the clearance of creatinine (as an approximation of glomerular filtration rate), they were able to show in healthy individuals that renal excretion of o-tyrosine (median FE = 7.86%) was greater than that for p-tyrosine (median FE = 0.67%), suggesting a preference for the reabsorption of p-tyrosine within the proximal tubule. The difference in the fractional excretion of p-tyrosine versus o-tyrosine is exacerbated in patients with chronic kidney disease and/or diabetes. While the renal excretion of p-tyrosine does not significantly differ in these disease states compared to healthy controls (median FE in CKD = 1.36%; in diabetes = 0.98%; in CKD and diabetes = 1.06%), there is a marked increase in the renal excretion of o-tyrosine in patients with these diseases (median FE in CKD = 27.28%; in diabetes = 125.29%; in CKD and diabetes = 111.89%); although it should be noted that in patients with chronic kidney disease alone these values failed to reach significance (Molnar et al., 2005b). Recently similar results were also reported in septic patients with the fractional excretion of m-tyrosine (mean FE = 61.0%) and o-tyrosine (mean FE = 243%) being significantly greater than that of p-tyrosine (mean FE = 1.6%) (Szelig et al., 2015). It is, however, unclear whether these results represent an adaptation of the kidneys to preferentially clear m- and o-tyrosine under conditions that are potentially favorable for the production of the abnormal tyrosine isomers, a reflection of some aspect of kidney damage in these diseases, or whether this could represent in loco production of the abnormal tyrosine isomers within the kidney during diseased states. If these data do represent an increase in the removal of the tyrosine isomers via the kidney in the setting of these diseases, which importantly show elevated blood levels of these same isomers, then these measurements may underestimate the actual production rate (as well as tissue concentrations) of m- and o-tyrosine.

One caveat to the renal excretion of m- and o-tyrosine as a protective mechanism against their accumulation and adverse effects is that in order for these isomers to be cleared by the kidney, they would first need to be transported from the tissues and cells in which they were formed into the blood. It is therefore likely that, in addition to their clearance via the renal urinary system, elimination of m- and o-tyrosine via metabolism or degradation also occurs at the cellular level. It is most likely that these abnormal tyrosine isomers are metabolized by one of the established pathways for tyrosine, either through tyrosine decarboxylase to form m- and o-tyramine or through tyrosine aminotransferase to form 2- and 3-hydroxyphenylpyruvate. In support of the metabolism through tyramine, mammalian DOPA-decarboxylase and bacterial tyrosine decarboxylase were shown to maintain specificity for both m- and o-tyrosine (Blashcko, 1950). Additionally, when 5 mg/kg m-tyrosine was administered to humans, the major metabolite was m-hydroxyphenylacetic acid, which was believed to follow decarboxylation of m-tyrosine to m-tyramine (Fell et al., 1979). The detection of tyramine and its metabolites were also detected after either tyrosine isomer was administered to various animal models including cats and rats ((Edwards, 1982; Edwards and Rizk, 1981; Gibson and Wurtman, 1978; McQuade and Juorio, 1983; Mitoma et al., 1957; Pogrund et al., 1961). Whether decarboxylation of m- and o-tyrosine to their respective isomers of tyramine is sufficient to prevent toxicity has yet to be tested. While evidence to support m- and o-tyrosine metabolism through the tyrosine degradation pathway is lacking to date, the gene encoding tyrosine aminotransferase has been shown to be upregulated in response to oxidative stress in C. elegans (Powolny et al., 2011). Furthermore, deamination via tyrosine aminotransferase to the isomers of hydroxyphenylpyruvate would likely be sufficient to reduce the adverse effects of the radical adduct, as it has previously been shown that the IC₅₀ of 3-hydroxyphenylalanine is five-fold higher that of m-tyrosine when tested for the inhibition of plant root growth (Bertin et al., 2007). Biochemical assays will be necessary to assess the enzymatic activity towards both m- and o-tyrosine in each of these pathways and whether these are protective against the adverse effects of the tyrosine isomers.

As additional phenylalanine was reported by multiple studies to counteract the effect of the tyrosine isomers (Bertin et al., 2007; Dunlop et al., 2011; Huang et al., 2010), a temporary reduction in phenylalanine hydroxylase activity could also prove protective during oxidative stress in two important ways: (1) this would increase the pool of free phenylalanine that could compete with the tyrosine isomers for charging to tRNA^Phe, and (2) it would reduce the pool of free p-tyrosine that could compete with its m- and o-isomers for degradation by the tyrosine metabolizing enzymes. These reductions in phenylalanine hydroxylase could occur at the transcriptional or post-transcriptional level after exposure to oxidative stress (Miranda et al., 2002; Ying et al., 2010). Alternatively, heterozygous genetic mutations affecting phenylalanine hydroxylase could also serve to produce chronic low-level elevations of phenylalanine that could be protective against oxidative stress and other adverse effects of the abnormal tyrosine isomers. Perhaps this could be a possible explanation for the wide-range of mutations affecting phenylalanine hydroxylase seen in human populations (Scriver et al., 1996). Experiments testing whether phenylalanine hydroxylase is repressed in response to oxidative stress have not been conducted. It should, however, be noted that in a few studies supplementation of p-tyrosine limited the adverse effects of the tyrosine isomers (Mikolas et al., 2013; Molnar et al., 2016; Selley et al., 2015). While an explanation for this is unclear, it could contradict the rationale that reducing phenylalanine hydroxylase activity may also be protective. Although, another explanation could be that in these feeding studies p-tyrosine simply blocked or reduced m- and o-tyrosine transport into cells, or the elevated p-tyrosine levels lead to a feedback reduction in phenylalanine hydroxylase activity and increased phenylalanine levels.

7.2. Differences between the effects of m- and o-tyrosine

Finally, there were observed differences in the effects of m- and o-tyrosine in cells with m-tyrosine in some settings being much more toxic of the two (Bertin et al., 2007; Dunlop et al., 2008; Rodgers et al., 2002; Ruggiero et al., 2011). The investigation into the mechanisms that account for these differences could be meaningful and help identify critical biological pathways involved in producing the harmful effects of tyrosine isomers. As suggested by Dunlop et al., one potential explanation for this difference in biologic activity could be due to how the position of the hydroxyl group effects protein structure (Dunlop et al., 2008). In the ortho-conformation, the hydroxyl group would be positioned in closest proximity to the peptide backbone and could potentially act chemically more similarly to phenylalanine; whereas, in the meta-conformation, the hydroxyl group would protrude further out and could have a more significant effect on protein structure due to a greater chemical similarity to tyrosine. Of course, if o-tyrosine is substituted for p-tyrosine then o-tyrosine could be more harmful by more closely resembling phenylalanine chemically. Alternatively, varying secondary modifications to the tyrosine isomers within protein could be another cause. Lastly, Popp et al. have suggested that phenylalanyl-tRNA synthetase could be less stringent in the exclusion of m-tyrosine versus o-tyrosine due to differences in available hydrogen bounds (Popp et al., 2015), and data in bacterial experiments have confirmed this (Bullwinkle et al., 2014). This results in phenylalanine-tRNA being more readily charged with m-tyrosine than o-tyrosine and would lead to greater incorporation of m-tyrosine into protein. Although, this difference would only account for cells undergoing active protein translation, and it should be noted that measured levels of m-tyrosine versus o-tyrosine incorporated into protein vary dependent upon the cell or tissue type. Indeed, o-tyrosine concentration within protein often is found to be higher than that of m-tyrosine, particularly in tissues in which there is minimal protein synthesis (e.g. lens) (Fu et al., 1998b; Molnar et al., 2005a). What accounts for this variability is currently unknown, but it could be that in addition to the preference for the non-enzymatic hydroxylation of free phenylalanine to generate o-tyrosine (see section 2.1), o-tyrosine formation may be favored when phenylalanine in intact protein is hydroxylated by nature of phenylalanine’s orientation within peptides. The hydrophobic nature of phenylalanine’s benzyl ring might position the amino acid in such a way that the ortho positions are more readily accessible for oxidation. However, further studies are necessary to examine this line of reasoning.

8. Conclusion

In conclusion, the abnormal tyrosine isomers m- and o-tyrosine can be generated under conditions of oxidative stress by the attack of the hydroxyl radical on phenylalanine. Though it was previously believed these isomers had no downstream physiological effects and could be used simply as markers of elevated hydroxyl radical levels both in vitro and in vivo, it is becoming apparent that they actually have a range of harmful effects in organisms including eliciting inhibitory effects on cell growth and survival. Hence the accumulation of tyrosine isomers after exposure to oxidative stress could represent a novel and critical mechanism by which oxidative stress may mediate adverse effects in cells and contribute to aging and the development of age-related diseases. The current evidence suggests that the aberrant incorporation of these isomers into proteins is at least one of the mechanisms contributing to the toxic effects of m- and o-tyrosine. Either the effects of this misincorporation event on the cellular proteome or perhaps other undiscovered cellular effects produced by the tyrosine isomers may lead to the observed alterations in cellular function and activity, including enhanced protein degradation, stimulation of apoptosis, and inhibition of mitogenic signaling pathways. However, our understanding of the downstream effects of exposure to the tyrosine isomers remains incomplete, and further characterization of these cellular responses is necessary. More data is needed to determine if the formation or adverse effects of the tyrosine isomers differ between tissues, particularly in animal models, and to identify their role in disease pathogenesis in vivo. Lastly, the cellular metabolism and elimination of m- and o-tyrosine is a poorly understood process, which could represent a novel oxidative stress defense mechanism.

Highlights.

• meta- and ortho-tyrosine isomers are formed under conditions of oxidative stress.

• Measurements of these isomers were first used to quantify hydroxyl radicals.

• Mounting evidence suggests m- and o-tyrosine adversely affect cells and tissues.

• This represents a novel mechanism by which oxidative stress causes cellular damage.

• Further directions are suggested to understand their roles in aging and disease.

Abbreviations

CKD

chronic kidney disease

CSF

cerebral spinal fluid

HIE

hypoxic ischemic encephalopathy

CHO

Chinese-hamster ovary