Small phenolic and indolic gut-dependent molecules in the primate central nervous system: levels vs. bioactivity.

INTRODUCTION: A rapidly growing body of data documents associations between disease of the brain and small molecules generated by gut-microbiota (GMB). While such metabolites can affect brain function through a variety of mechanisms, the most direct action would be on the central nervous system (CNS) itself. OBJECTIVE: Identify indolic and phenolic GMB-dependent small molecules that reach bioactive concentrations in primate CNS. METHODS: We conducted a PubMed search for metabolomic studies of the primate CNS [brain tissue or cerebrospinal fluid (CSF)] and then selected for phenolic or indolic metabolites that (i) had been quantified, (ii) were GMB-dependent. For each chemical we then conducted a search for studies of bioactivity conducted in vitro in human cells of any kind or in CNS cells from the mouse or rat. RESULTS: 36 metabolites of interests were identified in primate CNS through targeted metabolomics. Quantification was available for 31/36 and in vitro bioactivity for 23/36. The reported CNS range for 8 metabolites 2-(3-hydroxyphenyl)acetic acid, 2-(4-hydroxyphenyl)acetic acid, 3-(3-hydroxyphenyl)propanoic acid, (E)-3-(3,4-dihydroxyphenyl)prop-2-enoic acid [caffeic acid], 3-hydroxybenzoic acid, 4-hydroxybenzoic acid, 2-acetamido-3-(1H-indol-3-yl)propanoic acid [N-acetyltryptophan], 1H-indol-3-yl hydrogen sulfate [indoxyl-3-sulfate] overlapped with a bioactive concentration. However, the number and quality of relevant studies of CNS neurochemistry as well as of bioactivity were highly limited. Structural isomers, multiple metabolites and potential confounders were inadequately considered. CONCLUSION: The potential direct bioactivity of GMB-derived indolic and phenolic molecules on primate CNS remains largely unknown. The field requires additional strategies to identify and prioritize screening of the most promising small molecules that enter the CNS.

Changes in the Serum Metabolome of Patients Treated With Broad-Spectrum Antibiotics.

BACKGROUND: The gut microbiome (GMB) generates numerous small chemicals that can be absorbed by the host and variously biotransformed, incorporated, or excreted. The resulting metabolome can provide information about the state of the GMB, of the host, and of their relationship. Exploiting this information in the service of biomarker development is contingent on knowing the GMB-sensitivity of the individual chemicals comprising the metabolome. In this regard, human studies have lagged far behind animal studies. Accordingly, we tested the hypothesis that serum levels of chemicals unequivocally demonstrated to be GMB-sensitive in rodent models would also be affected in a clinical patient sample treated with broad spectrum antibiotics. METHODS: We collected serum samples from 20 hospitalized patients before, during, and after treatment with broad-spectrum antibiotics. We also collected samples from 5 control patients admitted to the hospital but not prescribed antibiotics. We submitted the samples for a non-targeted metabolomic analysis and then focused on chemicals known to be affected both by germ-free status and by antibiotic treatment in the mouse and/or rat. RESULTS: Putative identification was obtained for 499 chemicals in human serum. An aggregate analysis did not show any time x treatment interactions. However, our literature search identified 10 serum chemicals affected both by germ-free status and antibiotic treatment in the mouse or rat. Six of those chemicals were measured in our patient samples and additionally met criteria for inclusion in a focused analysis. Serum levels of 5 chemicals (p-cresol sulfate, phenol sulfate, hippurate, indole propionate, and indoxyl sulfate) declined significantly in our group of antibiotic-treated patients but did not change in our patient control group. CONCLUSIONS: Broad-spectrum antibiotic treatment in patients lowered serum levels of selected chemicals previously demonstrated to be GMB-sensitive in rodent models. Interestingly, all those chemicals are known to be uremic solutes that can be derived from aromatic amino acids (L-phenylalanine, L-tyrosine, or L-tryptophan) by anaerobic bacteria, particularly Clostridial species. We conclude that judiciously selected serum chemicals can reliably detect antibiotic-induced suppression of the GMB in man and thus facilitate further metabolome-based biomarker development.

Urinary Metabolites of Green Tea as Potential Markers of Colonization Resistance to Pathogenic Gut Bacteria in Mice.

BACKGROUND: The gut microbiome (GMB) generates numerous chemicals that are absorbed systemically and excreted in urine. Antibiotics can disrupt the GMB ecosystem and weaken its resistance to colonization by enteric pathogens such as Clostridium difficile. If the changes in GMB composition and metabolism are sufficiently large, they can be reflected in the urinary metabo-lome. Characterizing these changes could provide a potentially valuable biomarker of the status of the GMB. While preliminary studies suggest such a possibility, the high level of data variance presents a challenge to translational applications. Since many GMB-generated chemicals are derived from the biotransformation of plant-derived dietary polyphenols, administering an oral precursor challenge should amplify GMB-dependent changes in urinary metabolites. METHODS: A course of antibiotics (clindamycin, piperacillin/tazobactam, or aztreonam) was administered SC daily (days 1 and 2) to mice receiving polyphenol-rich green tea in drinking water. Urine was collected at baseline as well as days 3, 7, and 11. Levels of pyrogallol and pyrocatechol, two phenolic molecules unequivocally GMB-dependent in humans but that had not been similarly examined in mice, were quantified. RESULTS: In confirmation of our hypothesis, differential changes in murine urinary pyrogallol levels identified the treatments (clindamycin, piperacillin/tazobactam) previously associated with a weakening of colonization resistance to Clostridium difficile. The changes in pyrocatechol levels did not withstand corrections for multiple comparisons. CONCLUSIONS: In the mouse, urinary pyrogallol and, in all likelihood, pyrocatechol levels, are GMB-dependent and, in combination with precursor challenge, deserve further consideration as potential metabolomic biomarkers for the health and dysbiotic vulnerability of the GMB.

The phenolic interactome and gut microbiota: opportunities and challenges in developing applications for schizophrenia and autism.

Schizophrenia and autism spectrum disorder have long been associated with elevated levels of various small phenolic molecules (SPMs). In turn, the gut microbiota (GMB) has been implicated in the kinetics of many of these analytes. Unfortunately, research into the possible relevance of GMB-mediated SPMs to neuropsychiatry continues to be limited by heterogeneous study design, numerous sources of variance and technical challenges. Some SPMs have multiple structural isomers and most have conjugates. Without specialized approaches, SPMs can be incorrectly assigned or inaccurately quantified. In addition, SPM levels can be affected by dietary polyphenol or protein consumption and by various medications and diseases. Nonetheless, heterotypical excretion of various SPMs in association with schizophrenia or autism continues to be reported in independent samples. Recent studies in human cerebrospinal fluid demonstrate the presence of many SPMs A large number of these are bioactive in experimental models. Whether such mechanisms are relevant to the human brain in health or disease is not known. Systematic metabolomic and microbiome studies of well-characterized populations, an appreciation of multiple confounds, and implementation of standardized approaches across platforms and sites are needed to delineate the potential utility of the phenolic interactome in neuropsychiatry.

Quantification of phenolic acid metabolites in humans by LC-MS: a structural and targeted metabolomics approach.

AIM: Co-metabolism between a human host and the gastrointestinal microbiota generates many small phenolic molecules such as 3-hydroxy-3-(3-hydroxyphenyl)propanoic acid (3,3-HPHPA), which are reported to be elevated in schizophrenia and autism. Characterization of these chemicals, however, has been limited by analytic challenges. METHODOLOGY/RESULTS: We applied HPLC to separate and quantify over 50 analytes, including multiple structural isomers of 3,3-HPHPA in human cerebrospinal fluid, serum and urine. Confirmation of identity was provided by NMR, by MS and other detection methods. The highly selective methods support rapid quantification of multiple metabolites and exhibit superior chromatographic behavior. CONCLUSION: An improved ultra-HPLC-MS/MS and structural approaches can accurately quantify 3,3-HPHPA and related analytes in human biological matrices.

L-Tyrosine availability affects basal and stimulated catecholamine indices in prefrontal cortex and striatum of the rat.

We previously found that L-tyrosine (L-TYR) but not D-TYR administered by reverse dialysis elevated catecholamine synthesis in vivo in medial prefrontal cortex (MPFC) and striatum of the rat (Brodnik et al., 2012). We now report L-TYR effects on extracellular levels of catecholamines and their metabolites. In MPFC, reverse dialysis of L-TYR elevated in vivo levels of dihydroxyphenylacetic acid (DOPAC) (L-TYR 250-1000 µM), homovanillic acid (HVA) (L-TYR 1000 µM) and 3-methoxy-4-hydroxyphenylglycol (MHPG) (L-TYR 500-1000 µM). In striatum L-TYR 250 µM elevated DOPAC. We also examined L-TYR effects on extracellular dopamine (DA) and norepinephrine (NE) levels during two 30 min pulses (P2 and P1) of K+ (37.5 mM) separated by t = 2.0 h. L-TYR significantly elevated the ratio P2/P1 for DA (L-TYR 125 µM) and NE (L-TYR 125-250 µM) in MPFC but lowered P2/P1 for DA (L-TYR 250 µM) in striatum. Finally, we measured DA levels in brain slices using ex-vivo voltammetry. Perfusion with L-TYR (12.5-50 µM) dose-dependently elevated stimulated DA levels in striatum. In all the above studies, D-TYR had no effect. We conclude that acute increases within the physiological range of L-TYR levels can increase catecholamine metabolism and efflux in MPFC and striatum. Chronically, such repeated increases in L-TYR availability could induce adaptive changes in catecholamine transmission while amplifying the metabolic cost of catecholamine synthesis and degradation. This has implications for neuropsychiatric conditions in which neurotoxicity and/or disordered L-TYR transport have been implicated.

Large neutral amino acids levels in primate cerebrospinal fluid do not confirm competitive transport under baseline conditions.

In rodents, transport of large neutral amino acids (LNAAs) across the blood brain barrier (BBB) and blood-cerebrospinal fluid (CSF) barrier is mediated by high affinity carriers. Net brain LNAA levels are thought to be determined mainly by this competitive transport from plasma. Since the affinity for LNAA transport at the BBB in primates is considerably higher than in rodents, brain influx and by extension LNAA brain levels, should be even more dependent on competitive transport. Given that LNAA levels in CSF and brain interstitial fluid are usually similar, we analyzed serum and CSF of fasted subjects (n=24) undergoing spinal anesthesia and calculated brain influx and transporter occupancy using a conventional model of transport. Despite predicted near-full transporter saturation (99.7%), correlations between CSF levels and brain influx were modest, limited to tyrosine (r=0.60, p<0.002) and tryptophan (r=0.50, p<0.01) and comparable to correlations between CSF and serum levels. We also analyzed serum and CSF in (n=5) fasted vervet monkeys. Tyrosine and phenylalanine levels in CSF were positively correlated with those in serum, but correlations with calculated brain influx, which takes competition into account, were weaker or absent. We conclude that in primates i) baseline CSF LNAA levels do not confirm competitive transport, ii) brain LNAA levels should not be estimated on the basis of serum indices alone. This has implications for amino acid challenge studies and for neuropsychiatric disorders associated with dysregulated LNAA transport in which quantitative information about brain LNAA levels is needed.

A simplified method to quantify dysregulated tyrosine transport in schizophrenia.

BACKGROUND: Schizophrenia is associated with altered tyrosine transport across plasma membranes. This is typically demonstrated by measuring the uptake of radiolabeled tyrosine in cultured human fibroblasts. Our primary goal was to determine whether tyrosine uptake could be characterized using unlabeled tyrosine. A secondary goal was to assess the effect of antipsychotic drugs added during the incubation. METHOD: Epithelium-derived fibroblast cultures were generated from patients with schizophrenia (n=6) and age-matched controls (n=6). Cells between cycles 8-12 were exposed to an amino acid free medium for 60min and then for 1min to media containing tyrosine (0.008-1.0mM). Amino acid levels were measured and Michaelis-Menten parameters determined. Uptake of tyrosine (0.5mM) was also measured in control cells after antipsychotic drugs were introduced during the depletion or uptake phases. RESULTS: Tyrosine uptake was sodium-independent. The maximal transport velocity (Vmax) was significantly lower in patients with schizophrenia than in controls (p<0.01). The transporter affinity (Km) did not differ between the groups. Tyrosine uptake was differentially affected (p<0.001) by inclusion of 10(-4)M haloperidol, chlorpromazine or clozapine during different periods of incubation. CONCLUSION: Dysregulated tyrosine kinetics in schizophrenia can be readily studied without the use of radiolabeled tracers. The data also indicate that tyrosine uptake may be subject to complex pharmacological effects.

Presynaptic regulation of extracellular dopamine levels in the medial prefrontal cortex and striatum during tyrosine depletion.

RATIONALE: Available neurochemical probes that lower brain dopamine (DA) levels in man are limited by their tolerability and efficacy. For instance, the acute lowering of brain tyrosine is well tolerated, but only modestly lowers brain DA levels. Modification of tyrosine depletion to robustly lower DA levels would provide a superior research probe. OBJECTIVES: The objective of this study was to determine whether the subthreshold stimulation of presynaptic DA receptors would potentiate tyrosine depletion-induced effects on extracellular DA levels in the medial prefrontal cortex (MPFC) and striatum of the rat. METHODS: We administered quinpirole, a predominantly DA type 2 (D2R) receptor agonist, into the MPFC and striatum by reverse dialysis. A tyrosine- and phenylalanine-free neutral amino acid mixture [NAA(-)] IP was used to lower brain tyrosine levels. DA levels in the microdialysate were measured by HPLC with electrochemical detection. RESULTS: Quinpirole dose-dependently lowered DA levels in MPFC as well as in the striatum. NAA(-) alone transiently lowered DA levels (80 % baseline) in the striatum, but had no effect in MPFC. The co-administration of NAA(-) and a subthreshold concentration of quinpirole (6.25 nM) lowered DA levels (50 % baseline) in both the MPFC and striatum. This effect was blocked by the mixed D2R/D3R antagonist haloperidol at IP doses that on their own did not affect DA levels (10.0 nmol/kg in the MPFC and 0.10 nmol/kg in the striatum). CONCLUSIONS: Pharmacological stimulation of inhibitory D2R receptors during tyrosine depletion markedly lowers the extracellular DA levels in the MPFC and striatum. The data suggest that combining tyrosine depletion with a low dose of a DA agonist should robustly lower brain regional DA levels in man.

Tyrosine depletion lowers in vivo DOPA synthesis in ventral hippocampus.

In vivo dopamine synthesis in the medial prefrontal cortex of the rat is sensitive to the availability of tyrosine. Whether other limbic cortical dopamine terminal regions are similarly tyrosine-dependent is not known. In this study we examined the effects of tyrosine depletion on dopamine synthesis and catecholamine levels in the ventral hippocampus. A tyrosine- and phenylalanine-free neutral amino acid mixture was used to lower brain tyrosine levels in rats undergoing in vivo microdialysis. In one group, NSD-1015 was included in perfusate to permit measurement of DOPA levels. In a second group, NSD-1015 was not included in perfusate so that catecholamine levels could be assayed. Tyrosine depletion significantly lowered DOPA levels in the NSD-1015 treated group and lowered DOPAC but not dopamine or noradrenaline levels in the group not exposed to NSD-1015. We conclude that while catecholamine synthesis in the ventral hippocampus declines when tyrosine availability is lowered, under basal conditions, compensatory mechanisms are able to maintain stable extracellular catecholamine levels.

Increased tyrosine availability increases brain regional DOPA levels in vivo.

Tyrosine hydroxylation is considered to be the rate-limiting step in catecholamine synthesis. It is also assumed that under usual conditions, tyrosine 3-monooxygenase (EC 1.14.16.2) (tyrosine hydroxylase - TH) is close to full saturation with its l-tyrosine substrate and hence that raising the availability of l-tyrosine does not substantially increase 3,4-dihydroxyphenylalanine (DOPA) synthesis. We evaluated this in vivo by reverse dialysis of the aromatic-l-amino-acid decarboxylase (EC 4.1.1.28) inhibitor NSD-1015 (20µM) and selected concentrations of l- or d-tyrosine. In striatum, extracellular DOPA levels increased linearly (maximum 250% control) as l-tyrosine concentrations were raised from 0-1000µM. In medial prefrontal cortex, DOPA levels reached a maximum (300% control) at l-tyrosine 62.5-125µM but still remained significantly elevated (200% control) at higher l-tyrosine concentrations (250-500µM). At the l-tyrosine concentrations tested, DOPA levels were never below those of controls. d-tyrosine (62.5µM) did not affect DOPA levels. The degree to which the elevation of DOPA levels represents a net increase in tyrosine hydroxylation as opposed to heteroexchange of l-TYR for intracellular DOPA remains to be determined. However, one interpretation of the data is that under usual in vivo conditions brain TH may not be near full saturation with l-tyrosine and that mechanisms other than tyrosine hydroxylation may be more important in the acute regulation of brain catecholamine synthesis than previously appreciated. This would have implications for the pathophysiology and treatment of neuropsychiatric conditions in which dysregulation of DA transmission and l-tyrosine transport have been implicated.

Acute lithium administration selectively lowers tyrosine levels in serum and brain.

Lithium exerts anti-dopaminergic behavioral effects. We examined whether some of these might be mediated by changes in brain levels of tyrosine (TYR), the precursor to dopamine. Lithium chloride (LiCl(2)) 3.0mEq/kg IP acutely lowered serum TYR and the ratio of serum TYR to other large neutral amino acids (LNAAs); it also selectively lowered striatum TYR levels as measured in tissue or in vivo. While LiCl(2) 3.0mEq/kg IP also augmented haloperidol (0.19mg/kg SC)-induced catalepsy, this lithium effect was not attenuated by administration of TYR 100mg/kg IP. We conclude that lithium acutely and selectively lowers brain TYR by lowering serum levels of tyrosine relative to the LNAAs that compete with it for transport across the blood-brain barrier. However, the lowering of TYR does not appear to significantly contribute to the ability of lithium to potentiate haloperidol-mediated catalepsy.

Relationships between large neutral amino acid levels in plasma, cerebrospinal fluid, brain microdialysate and brain tissue in the rat.

Experimental limitations may preclude direct measurement of large neutral amino acid (LNAA) levels in brain tissue. Some data suggest that serum or cerebrospinal fluid (CSF) may provide an index of LNAA brain levels. We examined this in a series of experiments in rats, administering tyrosine, phenylalanine or valine IP 60min prior to harvesting of blood, CSF or brain tissue or during in vivo microdialysis of the brain. Serum indices of the administered LNAA generally showed a significant (r>0.8) correlation with brain tissue levels but the linear relationships varied significantly across brain regions, the LNAA and its dose. Increases in levels of an administered LNAA were consistently greater in CSF than in brain tissue. In contrast, changes in LNAA levels in brain tissue and in vivo microdialysate were generally comparable. We confirm that changes in serum and CSF LNAA levels can support limited, qualitative inferences about changes in brain tissue LNAA levels; quantitative inferences should not be drawn without prior validation under relevant experimental conditions.

Gamma-butyrolactone-induced dopamine accumulation in prefrontal cortex is affected by tyrosine availability.

Gamma-butyrolactone (GBL) elevates striatal and prefrontal cortex dopamine levels; only the striatal dopamine levels are elevated by increased dopamine synthesis. If increased dopamine synthesis is necessary in order for dopamine levels to be affected by tyrosine availability, then GBL-induced prefrontal cortex dopamine levels should be tyrosine insensitive. Rats received either vehicle, tyrosine (50 or 200 mg/kg i.p.) or a tyrosine-depleting mixture prior to GBL 750 mg/kg i.p.. GBL-induced dopamine levels in prefrontal cortex were lowered by tyrosine depletion. GBL-induced striatal dopamine levels were not affected. Hence, increased dopamine synthesis may not be necessary in order for tyrosine availability to affect pharmacologically elevated prefrontal cortex dopamine levels.

Tyrosine availability modulates potassium-induced striatal catecholamine efflux in vivo.

The relationship between tyrosine availability and high potassium (K+) induced dopamine (DA) and norepinephrine (NE) efflux was examined in striatum using in vivo microdialysis. High K+ (80 mM) was included in perfusate for two 30 min periods, 2.5 h apart. After the first high-K+ perfusion, a tyrosine- and phenylalanine-free mixture of large neutral amino acids (LNAA(-)) was administered (IP) to lower brain tyrosine. Tyrosine (0, 25, 50 or 100 mg/kg IP) was administered 30 min prior to the second high-K+ perfusion. The ratio of catecholamine efflux during the two perfusions (P2/P1) was compared between groups. LNAA(-) significantly lowered P2/P1 for both DA and NE. Tyrosine 25-50 mg/kg blocked the LNAA(-) effect. We conclude that catecholamine efflux during prolonged depolarization is tyrosine dependent. Analyses of LNAA levels suggest that availability of tyrosine for tyrosine hydroxylation may be modulated by competition between LNAAs within brain extracellular fluid.

Tyrosine depletion lowers dopamine synthesis and desipramine-induced prefrontal cortex catecholamine levels.

The relationship between limited tyrosine availability, DA (dopamine) synthesis and DA levels in the medial prefrontal cortex (MPFC) of the rat was examined by in vivo microdialysis. We administered a tyrosine- and phenylalanine-free mixture of large neutral amino acids (LNAA-) IP to lower brain tyrosine, and the norepinephrine transporter inhibitor desipramine (DMI) 10 mg/kg IP to raise MPFC DA levels without affecting DA synthesis. For examination of DOPA levels, NSD-1015 20 microM was included in perfusate. Neither NSD-1015 nor DMI affected tyrosine levels. LNAA- lowered tyrosine levels by 45%, and lowered DOPA levels as well; this was not additionally affected by concurrent DMI 10 mg/kg IP. In parallel studies DMI markedly increased extracellular levels of DA (420% baseline) and norepinephrine (NE) (864% baseline). LNAA- had no effect on baseline levels of DA or NE but robustly lowered DMI-induced DA (176% baseline) as well as NE (237% baseline) levels. Even when DMI (20 microM) was administered in perfusate, LNAA- still lowered DMI-induced DA and NE levels. We conclude that while baseline mesocortical DA synthesis is indeed dependent on tyrosine availability, the MPFC maintains normal extracellular DA and NA levels in the face of moderately lower DA synthesis. During other than baseline conditions, however, tyrosine depletion can lower ECF DA and NE levels in MPFC. These data offer a potential mechanism linking dysregulation of tyrosine transport and cognitive deficits in schizophrenia.