Personalized prediction of optimal water intake in adult population by blended use of machine learning and clinical data

Growing evidence suggests that sustained concentrated urine contributes to chronic metabolic and kidney diseases. Recent results indicate that a daily urinary concentration of 500 mOsm/kg reflects optimal hydration. This study aims at providing personalized advice for daily water intake considering personal intrinsic (age, sex, height, weight) and extrinsic (food and fluid intakes) characteristics to achieve a target urine osmolality (U_Osm) of 500 mOsm/kg using machine learning and optimization algorithms. Data from clinical trials on hydration (four randomized and three non-randomized trials) were analyzed. Several machine learning methods were tested to predict U_Osm. The predictive performance of the developed algorithm was evaluated against current dietary guidelines. Features linked to urine production and fluid consumption were listed among the most important features with relative importance values ranging from 0.10 to 0.95. XGBoost appeared the most performing approach (Mean Absolute Error (MAE) = 124.99) to predict U_Osm. The developed algorithm exhibited the highest overall correct classification rate (85.5%) versus that of dietary guidelines (77.8%). This machine learning application provides personalized advice for daily water intake to achieve optimal hydration and may be considered as a primary prevention tool to counteract the increased incidence of chronic metabolic and kidney diseases.

Drinking Water Source and Intake Are Associated with Distinct Gut Microbiota Signatures in US and UK Populations

BACKGROUND

The microbiome of the digestive tract exerts fundamental roles in host physiology. Extrinsic factors including lifestyle and diet are widely recognized as key drivers of gut and oral microbiome compositions. Although drinking water is among the food items consumed in the largest amount, little is known about its potential impact on the microbiome.

OBJECTIVES

We explored the associations of plain drinking water source and intake with gut and oral microbiota compositions in a population-based cohort.

METHODS

Microbiota, health, lifestyle, and food intake data were extracted from the American Gut Project public database. Associations of drinking water source (bottled, tap, filtered, or well water) and intake with global microbiota composition were evaluated using linear and logistic models adjusted for anthropometric, diet, and lifestyle factors in 3413 and 3794 individuals, respectively (fecal samples; 56% female, median [IQR] age: 48 [36-59] y; median [IQR] BMI: 23.3 [20.9-26.3] kg/m2), and in 283 and 309 individuals, respectively (oral samples).

RESULTS

Drinking water source ranked among the key contributing factors explaining the gut microbiota variation, accounting for 13% [Faith's phylogenetic diversity (Faith's PD)] and 47% (Bray-Curtis dissimilarity) of the age effect size. Drinking water source was associated with differences in gut microbiota signatures, as revealed by β diversity analyses (P < 0.05; Bray-Curtis dissimilarity, weighted UniFrac distance). Subjects drinking mostly well water had higher fecal α diversity (P < 0.05; Faith's PD, observed amplicon sequence variants), higher Dorea, and lower Bacteroides, Odoribacter, and Streptococcus than the other groups. Low water drinkers also exhibited gut microbiota differences compared with high water drinkers (P < 0.05; Bray-Curtis dissimilarity, unweighted UniFrac distance) and a higher abundance of Campylobacter. No associations were found between oral microbiota composition and drinking water consumption.

CONCLUSIONS

Our results indicate that drinking water may be an important factor in shaping the human gut microbiome and that integrating drinking water source and intake as covariates in future microbiome analyses is warranted.

A Journey through the Early Evidence Linking Hydration to Metabolic Health

The idea that water intake or hydration may play an intrinsic, independent role in modulating metabolic disease risk is relatively recent. Here, we outline the journey from early experimental works to more recent evidence linking water and hydration to metabolic health. It has been known for decades that individuals with existing metabolic dysfunction experience challenges to body water balance and have elevated arginine vasopressin (AVP), <underline>a key</underline> hormone regulating body fluid homeostasis. Later, intervention studies demonstrated that altering fluid balance in these individuals could worsen their condition, suggesting that hydration played a role in modulating glycemic control. More recently, observational and interventional studies in healthy individuals have implicated the hydration-vasopressin axis in the pathophysiology of metabolic diseases. Individuals with higher AVP (or its surrogate, copeptin) are at higher risk for developing type 2 diabetes and components of the metabolic syndrome, an association that remains even when controlling for known risk factors. Supporting preclinical work also suggests a causal role for AVP in metabolic dysfunction. It is known that individuals who habitually drink less fluids tend to have higher circulating AVP, which may be lowered by increasing water intake. In the short term, water supplementation in habitual low drinkers with high copeptin may reduce fasting glucose or glucagon, generating a proof of concept for the role of water supplementation in reducing incident metabolic disease. A large randomized trial is ongoing to determine whether water supplementation for 1 year in subjects with low water intake can meaningfully reduce fasting glucose, risk of new-onset diabetes, and other cardiometabolic risk factors.

Associations between Drinking Water Source and Gut Microbiota Composition in the American Gut Project Database

INTRODUCTION

The gut microbiome exerts a fundamental role in host physiology. Extrinsic factors such as lifestyle and diet are widely recognized as the main drivers of gut microbiota composition [<xref ref-type="bibr" rid="ref1">1</xref>, <xref ref-type="bibr" rid="ref2">2</xref>]. While drinking water is among the food items consumed in the largest amount, little is known about its potential impact on gut microbiota structures [<xref ref-type="bibr" rid="ref3">3</xref>-<xref ref-type="bibr" rid="ref5">5</xref>].

OBJECTIVE

We explored the associations between plain drinking water source and gut microbiota compositions in a large microbiota-based cohort.

METHODS

Participants in the American Gut Project database provided fecal samples and completed health, lifestyle, and food records which included plain drinking water source (bottled, tap, filtered, or well water). Associations between drinking water source and gut microbiota were evaluated using models adjusted for anthropometric, diet, and lifestyle factors in 3,413 individuals [<xref ref-type="bibr" rid="ref6">6</xref>]. Index of intra-individual fecal microbial diversity, inter-individual differences in composition, and taxa abundance were estimated by 16S rRNA sequencing.

RESULTS

The type of drinking water was associated with fecal microbiota composition. Drinking water source ranked among the key contributing factor explaining the gut microbiota variation both in alpha and beta diversity analyses, with effect sizes comparable to that of alcohol or diet type [<xref ref-type="bibr" rid="ref6">6</xref>] (Fig. <xref ref-type="fig" rid="f01">1</xref>). Subjects drinking different sources of water had differences in gut microbiota signatures, as revealed by beta diversity analyses (p < 0.05; Bray-Curtis dissimilarity, Weighted UniFrac distance) [<xref ref-type="bibr" rid="ref6">6</xref>]. Subjects drinking mostly well water also had higher fecal alpha diversity than the other groups (p < 0.05; Faith's PD, Observed OTUs) [<xref ref-type="bibr" rid="ref6">6</xref>]. Taxonomic differences were found in well water drinkers, with clinically important taxa, such as Bacteroides, Odoribacter, and Streptococcus being depleted and Dorea being increased as compared to the other groups [6].

CONCLUSIONS

Our results reveal that drinking water may be an important factor in shaping the gut microbiome. Future research investigating the gut microbiota in relation to environmental factors may benefit from integrating drinking water source as a covariate in the analyses.

Hydration for health hypothesis: a narrative review of supporting evidence

PURPOSE

An increasing body of evidence suggests that excreting a generous volume of diluted urine is associated with short- and long-term beneficial health effects, especially for kidney and metabolic function. However, water intake and hydration remain under-investigated and optimal hydration is poorly and inconsistently defined. This review tests the hypothesis that optimal chronic water intake positively impacts various aspects of health and proposes an evidence-based definition of optimal hydration.

METHODS

Search strategy included PubMed and Google Scholar using relevant keywords for each health outcome, complemented by manual search of article reference lists and the expertise of relevant practitioners for each area studied.

RESULTS

The available literature suggest the effects of increased water intake on health may be direct, due to increased urine flow or urine dilution, or indirect, mediated by a reduction in osmotically -stimulated vasopressin (AVP). Urine flow affects the formation of kidney stones and recurrence of urinary tract infection, while increased circulating AVP is implicated in metabolic disease, chronic kidney disease, and autosomal dominant polycystic kidney disease.

CONCLUSION

In order to ensure optimal hydration, it is proposed that optimal total water intake should approach 2.5 to 3.5 L day^-1 to allow for the daily excretion of 2 to 3 L of dilute (< 500 mOsm kg^-1) urine. Simple urinary markers of hydration such as urine color or void frequency may be used to monitor and adjust intake.

Osmotic stimulation of vasopressin acutely impairs glucose regulation: a counterbalanced, crossover trial

BACKGROUND

Epidemiological studies in humans show increased concentrations of copeptin, a surrogate marker of arginine vasopressin (AVP), to be associated with increased risk for type 2 diabetes.

OBJECTIVES

To examine the acute and independent effect of osmotically stimulated AVP, measured via the surrogate marker copeptin, on glucose regulation in healthy adults.

METHODS

Sixty subjects (30 females) participated in this crossover design study. On 2 trial days, separated by ≥7 d (males) or 1 menstrual cycle (females), subjects were infused for 120 min with either 0.9% NaCl [isotonic (ISO)] or 3.0% NaCl [hypertonic (HYPER)]. Postinfusion, a 240-min oral-glucose-tolerance test (OGTT; 75 g) was administered.

RESULTS

During HYPER, plasma osmolality and copeptin increased (P < 0.05) and remained elevated during the entire 6-h protocol, whereas renin-angiotensin-aldosterone system hormones were within the lower normal physiological range at the beginning of the protocol and declined following infusion. Fasting plasma glucose did not differ between trials (P > 0.05) at baseline and during the 120 min of infusion. During the OGTT the incremental AUC for glucose from postinfusion baseline (positive integer) was greater during HYPER (401.5 ± 190.5 mmol/L·min) compared with the ISO trial (354.0 ± 205.8 mmol/L·min; P < 0.05). The positive integer of the AUC for insulin during OGTT did not differ between trials (HYPER 55,850 ± 36,488 pmol/L·min compared with ISO 57,205 ± 31,119 pmol/L·min). Baseline values of serum glucagon were not different between the 2 trials; however, the AUC of glucagon during the OGTT was also significantly greater in HYPER (19,303 ± 3939 ng/L·min) compared with the ISO trial (18,600 ± 3755 ng/L·min; P < 0.05).

CONCLUSIONS

The present data indicate that acute osmotic stimulation of copeptin induced greater hyperglycemic responses during the oral glucose challenge, possibly due to greater glucagon concentrations.This study was registered at clinicaltrials.gov as NCT02761434.

Water Supplementation Reduces Copeptin and Plasma Glucose in Adults With High Copeptin: The H2O Metabolism Pilot Study

OBJECTIVE

Because elevated copeptin, a marker of vasopressin, is linked to low water intake and high diabetes risk, we tested the effect of water supplementation on copeptin and fasting glucose.

DESIGN, SETTING, AND PARTICIPANTS

Thirty-one healthy adults with high copeptin (>10.7 pmol · L-1 in men and >6.1 pmol·L-1 in women) identified in a population-based survey from 2013 to 2015 and with a current 24-hour urine osmolality of >600 mOsm · kg-1 were included.

INTERVENTION

Addition of 1.5 L water daily on top of habitual fluid intake for 6 weeks.

MAIN OUTCOME MEASURE

Pre- and postintervention fasting plasma copeptin concentrations.

RESULTS

Reported mean water intake increased from 0.43 to 1.35 L · d-1 (P < 0.001), with no other observed changes in diet. Median (interquartile range) urine osmolality was reduced from 879 (705, 996) to 384 (319, 502) mOsm · kg-1 (P < 0.001); urine volume increased from 1.06 (0.90, 1.20) to 2.27 (1.52, 2.67) L · d-1 (P < 0.001); and baseline copeptin decreased from 12.9 (7.4, 21.9) pmol · L-1 to 7.8 (4.6;11.3) pmol · L-1 (P < 0.001). Water supplementation reduced fasting plasma glucose from a mean (SD) of 5.94 (0.44) to 5.74 (0.51) (P = 0.04). The water-associated reduction of both fasting copeptin and glucose concentration in plasma was most pronounced in participants in the top tertile of baseline copeptin.

CONCLUSIONS

Water supplementation in persons with habitually low water consumption and high copeptin levels is effective in lowering copeptin. It appears a safe and promising intervention with the potential of lowering fasting plasma glucose and thus reducing diabetes risk. Further investigations are warranted to support these findings.

Osmotic Stimulation of Thirst in Men and Women (P15-017-19)

Objectives

Examine potential sex differences in osmotically stimulated thirst.

Methods

60 healthy adults (30 female - measured in early follicular phase, age: 39 ± 8 y, weight: 78.2 ± 15.2 kg, body mass index: 26.9 ± 4.0 kg·m−2) were infused intravenously with 3% (HYPER) or 0.9% (ISO) NaCl for 120 min (0.1 ml·kg−1·min−1) in a counterbalanced, cross-over design. The same volume of saline was infused in both trials. Blood samples and thirst responses were collected prior to infusion and every 30 minutes. Plasma osmolality (POsm) and sodium (PNa) were measured. Thirst was assessed via a 180-mm visual analog scale consisting of a line with an anchor on the left side (0 mm, “not at all”) and a second anchor on the 125-mm mark with the label “extremely”.

Results

From 0 min to 120 min in the HYPER trial, PNa increased from 136 ± 2 mmol ·L−1 to 146.5 ± 3 mmol ·l−1 (P < 0.001). Similarly, POsm and thirst increased significantly from 286 ± 3 mmol·kg−1 to 305 ± 4 mmol·kg−1 (P < 0.001) and from 38 ± 29 mm to 88 ± 35 mm (P < 0.001), respectively. During the ISO trial, PNa (0 min: 135.9 ± 1.9; 120 min: 137.6 ± 1.6 mmol·L−1), POsm (0 min: 285 ± 4 mmol·kg−1; 120 min: 287 ± 3 mmol·kg−1), and thirst (0 min: 30 ± 27 mm; 120 min: 44 ± 32 mm) remained unchanged (P > 0.05). POsm at the end of the infusion did not differ between men and women for the HYPER (men: 306 ± 5 mmol·kg−1; women: 305 ± 3 mmol·kg−1) or ISO (men: 288 ± 3 mmol·kg−1; women: 288 ± 3 mmol·kg−1) trials. When thirst responses were split by sex women exhibited greater thirst (98 ± 27 mm) than men (77 ± 38 mm) at the end of the HYPER trial (P < 0.0001), but not at the end of the ISO trial (women: 43 ± 32 mm, men 44 ± 31 mm; P > 0.05).

Conclusions

These data suggest that similar hypertonic stimulus may lead to greater thirst stimulation in women.

Funding Sources

This study was funded by Danone Research.

Hypertonic Saline Infusion Acutely Degrades Mood in Healthy Volunteers (P23-014-19)

Objectives

Mild and moderate dehydration adversely affect mood and cognitive function. During dehydration, hypertonic hypovolemia activates both osmo- and baro-receptors but it is not known which physiological pathway is associated with degraded mood state. This study examined the acute effect of osmoreceptor stimulation on mood.

Methods

Sixty healthy adults (50% females, 30 ± 1 y; BMI: 26.9 ± 4.0 kg·m−2) were infused intravenously with 3.0% (HYPER) or 0.9% (ISO) NaCl for 2 h (0.1 ml·kg−1·min−1) using a counterbalanced, crossover design. Blood samples were collected every 30 minutes to measure plasma osmolality (POsm), copeptin (a surrogate marker of vasopressin), and renin-angiotensin-aldosterone system (RAAS) hormones. Mood was assessed with the short version of Profile of Mood State (POMS) questionnaire before and after the infusion.

Results

POsm and copeptin increased from 286 ± 3 mmol·kg−1 to 305 ± 4 mmol·kg−1 and from 4.5 ± 3.7 pmol·L−1 to 20.4 ± 12.8 pmol·L−1, respectively in HYPER (P < 0.05), and were unchanged in ISO (P > 0.05). No hormonal differences were observed between trials for RAAS hormones (P > 0.05). During HYPER copeptin, following the 2-h infusion, was greater in females than in males (female: 23.4 ± 13.9 pmol·L−1, male: 17.4 ± 10.9 pmol·L−1; P < 0.05). The POMS total mood disturbance (TMD) score increased from 10.5 ± 0.9 to 16.5 ± 1.6 in HYPER (P < 0.05), but not in ISO (P > 0.05). Among POMS subscales, depression-dejection and fatigue-inertia increased in HYPER compared to ISO (P < 0.05). When TMD responses in the HYPER trial were analyzed with sex as a between-subjects factor, the increase was significant in females (pre: 10.2 ± 1.0, post: 18.6 ± 2.3; P < 0.001) but not in males (pre: 10.8 ± 1.4, post: 14.0 ± 2.0; P > 0.05). The confusion-bewilderment subscales and fatigue-inertia of the POMS were also elevated post HYPER in females (P < 0.05), but not in ISO (P > 0.05) in either sex.

Conclusions

Hypertonic saline infusion acutely degrades mood state, and women appear to have a more pronounced response. The underlying mechanisms remain to be determined but may be related to higher copeptin levels in women.

The study was registered at ClinicalTrials.gov as NCT02761434.

Funding Sources

Danone Research.

Supporting Tables, Images and/or Graphs

L. fermentum CECT 5716 prevents stress-induced intestinal barrier dysfunction in newborn rats

BACKGROUND

Intestinal epithelial barrier (IEB) dysfunction plays a critical role in various intestinal disorders affecting infants and children, including the development of food allergies and colitis. Recent studies highlighted the role of probiotics in regulating IEB functions and behavior in adults, but their effects in the newborn remain largely unknown. We therefore characterized in rat pups, the impact of Lactobacillus fermentum CECT 5716 (L. fermentum) on stress-induced IEB dysfunction, systemic immune response and exploratory behavior.

METHODS

Newborn rats received daily by gavage either L. fermentum or water. Intestinal permeability to fluorescein sulfonic acid (FSA) and horseradish peroxidase (HRP) was measured following maternal separation (MS) and water avoidance stress (WAS). Immunohistochemical, transcriptomic, and Western blot analysis of zonula occludens-1 (ZO-1) distribution and expression were performed. Anxiety-like and exploratory behavior was assessed using the elevated plus maze test. Cytokine secretion of activated splenocytes was also evaluated.

KEY RESULTS

L. fermentum prevented MS and WAS-induced IEB dysfunction in vivo. L. fermentum reduced permeability to both FSA and HRP in the small intestine but not in the colon. L. fermentum increased expression of ZO-1 and prevented WAS-induced ZO-1 disorganization in ileal epithelial cells. L. fermentum also significantly reduced stress-induced increase in plasma corticosteronemia. In activated splenocytes, L. fermentum enhanced IFNγ secretion while it prevented IL-4 secretion. Finally, L. fermentum increased exploratory behavior.

CONCLUSIONS & INFERENCES

These results suggest that L. fermentum could provide a novel tool for the prevention and/or treatment of gastrointestinal disorders associated with altered IEB functions in the newborn.

Coatings in food contact materials: Potential source of genotoxic contaminants?

Up till now, no harmonized EU regulation exists on chemicals used in coatings for food contact materials (FCM). Therefore, these substances need to comply with the general provisions of EU Regulation 1935/2004 and, if present, with national legislation. Different 'inventory lists' of compounds that might be present in coatings are available, but for hundreds of these substances, the potential human health impact of their use in FCM coatings has not (recently) been evaluated. Since detailed evaluation of all compounds is not feasible, a pragmatic approach was developed to identify substances with a potential concern for human health. First, an inventory was assembled containing all substances potentially used in coatings. Afterwards, the genotoxic potential of the non-evaluated substances was predicted in silico using two structure-activity relationship (SAR) software programs. For substances yielding structural alerts in both models, genotoxicity data were collected from previous European evaluations in a non-FCM context and from the European CHemicals Agency (ECHA) website. In total, 53 substances were identified as genotoxic in both in silico models, of which ten were considered to be of high concern. For most of the substances, additional toxicological information is needed.

Drug-induced liver injury: mechanisms, types and biomarkers

Drug-induced liver injury is a ubiquitous issue in clinical settings and pharmaceutical industry. Hepatotoxicity elicited by drugs may be intrinsic or idiosyncratic, both which are driven by different molecular mechanisms. Recently, a unifying mechanistic model of drug-induced liver injury has been introduced. According to this model, drug-induced hepatotoxicity relies on 3 consecutive steps, namely an initial cellular insult that leads to the occurrence of mitochondrial permeability transition, which in turn ultimately burgeons into the onset of cell death. Clinically, drug-induced liver injury can be manifested in a number of acute and chronic conditions, including hepatitis, cholestasis, steatosis and fibrosis. These pathologies can be diagnosed and monitored by addressing well-established physical, clinical chemistry and histopathological biomarkers. In the last few years, several novel read-outs of drug-induced liver injury have been proposed, involving genetic, epigenetic, transcriptomic, proteomic and metabolomic parameters. These new concepts and recent developments in the field of drug-induced liver injury are revised in the current paper.

Primary hepatocyte cultures for pharmaco-toxicological studies: at the busy crossroad of various anti-dedifferentiation strategies

Continuously increasing understanding of the molecular triggers responsible for the onset of diseases, paralleled by an equally dynamic evolution of chemical synthesis and screening methods, offers an abundance of pharmacological agents with a potential to become new successful drugs. However, before patients can benefit of newly developed pharmaceuticals, stringent safety filters need to be applied to weed out unfavourable drug candidates. Cost effectiveness and the need to identify compound liabilities, without exposing humans to unnecessary risks, has stimulated the shift of the safety studies to the earliest stages of drug discovery and development. In this regard, in vivo relevant organotypic in vitro models have high potential to revolutionize the preclinical safety testing. They can enable automation of the process, to match the requirements of high-throughput screening approaches, while satisfying ethical considerations. Cultures of primary hepatocytes became already an inherent part of the preclinical pharmaco-toxicological testing battery, yet their routine use, particularly for long-term assays, is limited by the progressive deterioration of liver-specific features. The availability of suitable hepatic and other organ-specific in vitro models is, however, of paramount importance in the light of changing European legal regulations in the field of chemical compounds of different origin, which gradually restrict the use of animal studies for safety assessment, as currently witnessed in cosmetic industry. Fortunately, research groups worldwide spare no effort to establish hepatic in vitro systems. In the present review, both classical and innovative methodologies to stabilize the in vivo-like hepatocyte phenotype in culture of primary hepatocytes are presented and discussed.

Inhibition of NF-kappaB activation by the histone deacetylase inhibitor 4-Me2N-BAVAH induces an early G1 cell cycle arrest in primary hepatocytes

OBJECTIVE

Benzoylaminoalkanohydroxamic acids, including 5-(4-dimethylaminobenzoyl)aminovaleric acid hydroxamide (4-Me(2)N-BAVAH), are structural analogues of Trichostatin A, a naturally occurring histone deacetylase inhibitor (HDACi). 4-Me(2)N-BAVAH has been shown to induce histone hyperacetylation and to inhibit proliferation in Friend erythroleukaemia cells in vitro. However, the molecular mechanisms have remained unidentified.

MATERIALS AND METHODS

In this study, we evaluated the effects of 4-Me(2)N-BAVAH on proliferation in non-malignant cells, namely epidermal growth factor-stimulated primary rat hepatocytes.

RESULTS AND CONCLUSION

We have found that 4-Me(2)N-BAVAH inhibits HDAC activity at non-cytotoxic concentrations and prevents cells from responding to the mitogenic stimuli of epidermal growth factor. This results in an early G(1) cell cycle arrest that is independent of p21 activity, but instead can be attributed to inhibition of cyclin D1 transcription through a mechanism involving inhibition of nuclear factor-kappaB activation. In addition, 4-Me(2)N-BAVAH delays the onset of spontaneous apoptosis in primary rat hepatocyte cultures as evidenced by down-regulation of the pro-apoptotic proteins Bid and Bax, and inhibition of caspase-3 activation.

Kinetic Characteristics of Acidic and Alkaline Ceramidase in Human Epidermis

It has recently become evident that at least five ceramidase (CDase) isoforms are present in human epidermis, and that specifically acidic CDase (aCDase) and alkaline CDase (alkCDase) activities increase during keratinocyte differentiation, and thus might play a pivotal role(s) in permeability barrier function. Prior to investigating their possible roles in the epidermal barrier function, it is necessary to characterize basic kinetic parameters for these enzymes, as well as to determine the effects of the established CDase inhibitors and their activities. In this study, assays for both aCDase and alkCDase activities in fully differentiated human epidermis were optimized using a radiolabeled substrate. These studies revealed that aCDase activity is substantially higher than alkCDase activity, and that both isoenzymes are inhibited by a CDase inhibitor N-oleylethanolamine. These findings were also confirmed using an in situ enzyme assay.

A Metabolic Screening Study of Trichostatin A (TSA) and TSA-Like Histone Deacetylase Inhibitors in Rat and Human Primary Hepatocyte Cultures

Hydroxamic acid (HA)-based histone deacetylase (HDAC) inhibitors, with trichostatin A (TSA) as the reference compound, are potential antitumoral drugs and show promise in the creation of long-term primary cell cultures. However, their metabolic properties have barely been investigated. TSA is rapidly inactivated in rodents both in vitro and in vivo. We previously found that 5-(4-dimethylaminobenzoyl)aminovaleric acid hydroxyamide or 4-Me2N-BAVAH (compound 1) is metabolically more stable upon incubation with rat hepatocyte suspensions. In this study, we show that human hepatocytes also metabolize TSA more rapidly than compound 1 and that similar pathways are involved. Furthermore, structural analogs of compound 1 (compounds 2-9) are reported to have the same favorable metabolic properties. Removal of the dimethylamino substituent of compound 1 creates a very stable but 50% less potent inhibitor. Chain lengthening (4 to 5 carbon spacer) slightly improves both potency and metabolic stability, favoring HA reduction to hydrolysis. On the other hand, Calpha-unsaturation and spacer methylation not only reduce HDAC inhibition but also increase the rate of metabolic inactivation approximately 2-fold, mainly through HA reduction. However, in rat hepatocyte monolayer cultures, compound 1 is shown to be extensively metabolized by phase II conjugation. In conclusion, this study suggests that simple structural modifications of amide-linked TSA analogs can improve their phase I metabolic stability in both rat and human hepatocyte suspensions. Phase II glucuronidation, however, can compensate for their lower phase I metabolism in rat hepatocyte monolayers and could play a yet unidentified role in the determination of their in vivo clearance.

[The role of intercellular communication via "gap junctions" in disease]

Gap junctional intercellular communication plays an important role in the maintenance of cellular homeostasis. The flow of chemical messengers through gap junctions, gap junctional intercellular communication, is essential in processes such as electrical coupling, embryonic development and adaptive tissue response. Gap junctions are formed by connexin proteins. Mutational alterations in the connexin genes are associated with the occurrence of multiple diseases, such as peripheral neuropathy, cardiovascular disease, dermatological disease, hereditary deafness and cataract. Consequently, modulation of gap junctional intercellular communication is a potential pharmacological target. Future research, based, for example, on the recent developments in genetics, may clarify gap junction physiology. This will in turn provide promising perspectives for the development of targeted drugs.

Effect of oxygen concentration on the expression of glutathione S-transferase activity in periportal and perivenous rat hepatocyte cultures

Cultures of perivenous (PV) and periportal (PP) hepatocytes could provide suitable in vitro models for studying the zone-specific hepatotoxic potential of xenobiotics. However, it is not known whether cultured PP and PV hepatocytes keep their phenotypes when the microcirculation of the liver changes. This question has been studied by culturing rat hepatocytes at 13 and 4% (v/v) O(2), respectively, mimicking the acinar oxygen gradient. PP and PV adult rat hepatocytes were isolated by digitonin-collagenase in situ perfusion and cultured on plastic Falcon and gas-permeable Petriperm dishes in Williams' E medium and kept at 13 and 4% (v/v) O(2), respectively. Cultures at 20% (v/v) O(2) on plastic dishes served as a control. Two types of cultures were studied, namely conventional cultures either unsupplemented or supplemented with 30 mM pyruvate. The activities of glutamine synthetase (GS) and glutathione S-transferase (GST) were measured in freshly isolated PP and PV hepatocytes and all cultures. The heterogeneous expression of GS (PV>PP), observed in freshly isolated hepatocytes, was kept for at least 4 days in culture. Total, Mu and Alpha class GST activities were predominantly expressed in PV freshly isolated cells. However, no beneficial effect could be observed in culture by exposing the cells to their specific in vivo oxygen concentration. The best maintenance of GST PV predominance in culture was observed in Petriperm dishes at 20% (v/v) O(2), as well in pyruvate-supplemented as unsupplemented cultures. PV GST predominance was thus kept in particular when the highest oxygen concentration was used and made available to the cells through the gas-permeable membranes. The results on GS PV predominance support these findings.

Pyruvate-induced long-term maintenance of glutathione s-transferase in rat hepatocyte cultures

The addition of pyruvate to the culture medium has been reported to improve the maintenance of P450-dependent enzyme expression in primary rat hepatocyte cultures. In this study, the effects of 30mM pyruvate on cell morphology, albumin secretion and glutathione S-transferase (GST) expression were investigated as a function of the time in culture. The effect of triiodothyronine (T3) exposure on GST expression was also measured in pyruvate-treated cultures. Transmission electron microscopy showed that untreated hepatocytes deteriorated after culture for 7 days, whereas the morphology of the pyruvate-treated cells was similar to that observed in intact liver tissue. The albumin secretion rate was significantly higher in rat hepatocytes exposed to pyruvate than in control cells. In the presence of pyruvate, mu and alpha class GST activities were well maintained, whereas GST pi activity was increased over the entire culture period. HPLC analysis revealed that the complement of GST subunits present in hepatocytes is altered during culture with pyruvate: mu,class proteins remained relatively constant, whereas a decrease in the alpha class content was accompanied by a strong increase in GST subunit P1 (GSTP1). The induction of GSTP1 was confirmed at the mRNA level. In control cultures, pi class GST activity was increased, but total, mu, and alpha class GST activities continuously declined as a function of culture time and became undetectable beyond 7 days in culture. At the protein and mRNA levels, a much smaller increase in GSTP1 was observed than in the pyruvate cultures. When the pyruvate-treated cell cultures were exposed to T3, an inhibitory effect on GST activities and proteins was found. These results indicate that this simple culture model could be useful for studying the expression and regulation of GST.

Hydroxypropyl-beta-cyclodextrin as delivery system for thyroid hormones, regulating glutathione S-transferase expression in rat hepatocyte co-cultures

Thyroid hormones play a role in the regulation of glutathione S-transferase (GST) expression. Here, co-cultures of rat hepatocytes with bile duct epithelial cells have been used to study the direct effects of both triiodothyronine (T3) and thyroxine (T4) on GST activities and proteins. Because T3 and T4 are poorly water soluble and organic solvents used to dissolve them often interfere with biotransformation pathways, an alternative delivery system namely hydroxypropyl-beta-cyclodextrin (HPBC) has been applied. Appropriate control cultures contained either 0.02 or 0.10% (w/v) HPBC, the concentrations necessary to supply T3 and T4 (10(-9) to 10(-5) M) to the cells, respectively. No effect of the vehicle HPBC on the different GST isoenzyme activities and proteins could be observed. On the contrary, after 10 days of co-culture, T3 and T4 decreased GST protein concentrations as well as GST activities measured by 1-chloro-2,4-dinitrobenzene (broad spectrum), 1,2-dichloro-4-nitrobenzene (Mu class M1/M2-specific) and 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (Alpha class A1/2-specific) in a concentration-dependent manner. The Alpha class subunits A1/2 and A3, and the Mu class subunit M2 were mostly affected. No effect was observed on the Pi class enzyme. These findings indicate that a combination of co-cultured hepatocytes with an HPBC-based delivery system for hydrophobic compounds represents a powerful in vitro tool in drug development.

Influence of culture system and medium enrichment on sulfotransferase and sulfatase expression in male rat hepatocyte cultures

The expression of sulfotransferase and steroid sulfatase was studied in rat liver using the most promising culture models of hepatocytes, including monolayer culture with a pyruvate (30 mM) enriched medium, co-culture with rat epithelial cells from primitive biliary origin and collagengel sandwich culture. In the latter, addition of dexamethasone (1 microM) to the medium was examined. Phenol sulfotransferase enzymes (SULT1) were studied by measuring activities towards 4-methylphenol and estradiol, hydroxysteroid sulfotransferase (SULT2A) activity was determined towards dehydroepiandrosterone (DHEA). Microsomal steroid sulfatase activity was measured towards estrone sulfate. Western blot analysis was carried out using polyclonal antibodies raised against rat phenol sulfotransferase SULT1A1 (ASTIV), estrogen sulfotransferase SULT1E1 (EST) and hydroxysteroid sulfotransferase (HST). SULT2A activity towards DHEA was maintained at a high level during the whole culture time. In the co-culture it even reached the level of freshly isolated cells. Addition of pyruvate had no positive effect on the activity measured in monolayer cultures. High SULT1A1 activity towards 4-methylphenol was found in the co-culture system. In the monolayer culture, the activity initially decreased with 35% but was then kept at a constant level, while in the sandwich culture low activities were measured. For dexamethasone, an inducing effect on the various SULT activities could not be detected. Independently of the culture model used, the SULT1E1 activity towards estradiol decreased to 20% and 5% of the initial activity after four and seven days of culture, respectively. Microsomal steroid sulfatase activity was best maintained in collagengel sandwich cultures. During the first four days in culture it retained 73% of the initial activity, afterwards it decreased to 40% of the activity found in freshly isolated hepatocytes, irrespective of the culture conditions. High expectations exist for collagengel sandwich cultures, however, in our study the results were rather disappointing. Monolayer is a suitable culture model for short-term purposes. For long-term in vitro biotransformation studies, co-culture is preferred but is rather complex.

Triiodothyronine downregulates the periportal expression of alpha class glutathione S-transferase in rat liver

Most drug-metabolizing phase I and phase II enzymes, including the glutathione S-transferases (GST), exhibit a zonated expression in the liver, with lower expression in the upstream, periportal region. To elucidate the involvement of pituitary-dependent hormones in this zonation, the effect of hypophysectomy and 3,3',5-triiodo-L-thyronine (T3) on the distribution of GST was studied in rats. Hypophysectomy increased total GST activity both in the periportal and perivenous liver region. Subsequent T3 treatment counteracted this effect in the perivenous zone. However, analysis for either mu class M1/M2-specific (1,2-dichloro-4-nitrobenzene) or alpha class A1/A2-specific (7-chloro-4-nitrobenzo-2-oxa-1,3-diazole) GST activity revealed that T3 treatment did not significantly affect the perivenous activity of these GST classes. In contrast, T3 was found to significantly counteract the increase of alpha class GST activity caused by hypophysectomy in the periportal zone. To establish whether this effect was T3-specific, hepatocytes were isolated from either the periportal and perivenous zone by digitonin/collagenase perfusion and cultured either as pyruvate-supplemented monolayer or as co-culture with rat liver epithelial cells. Only in the latter it was found that T3 suppressed the A1/A2-specific GST activity and alpha class proteins predominantly in periportal cells. The data demonstrate that T3 is an important factor responsible for the low expression of alpha GST in the periportal region. T3 may be involved in the periportal downregulation of other phase I and II enzymes as well.

Effect of long-term ethanol exposure on the acinar distribution of hepatic glutathione S-transferase

The ability of ethanol to affect the regional distribution of individual glutathione S-transferase (GST) isoenzymes in rat liver was investigated by analyzing the expression levels in cell lysates obtained from the periportal or perivenous liver region after in situ digitonin perfusion. In control rats, a significant perivenous dominance of GST proteins and activities measured by the substrates 1-chloro-2,4-dinitrobenzene (broad spectrum), 1, 2-dichloro-4-nitrobenzene (M1/M2-specific), and 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (A1/A2-specific) was found. In pair-fed rats exposed to ethanol (36% of total calories) for 2 weeks, all GST activities measured were significantly increased in both acinar zones. However, the relative increase was greater in the perivenous region. The induction of the A1/A2-specific activity was the most pronounced. HPLC analysis revealed for both regions that this increase was largely confined to the A2 subunit, with only minor effects observed on the A1 subunit. At the mRNA level, the constitutive perivenous dominance of both GST A1 and GST A2 expression became more pronounced after ethanol administration. The results demonstrate that long-term ethanol exposure induces individual GST isoenzymes differently and might have a profound effect on xenobiotic-induced regional liver damage.

Effect of ethanol on the expression of hepatic glutathione S-transferase: an in vivo/in vitro study

Ethanol, a human toxicant and a solvent in pharmacological research, is known to interfere with biotransformation of xenobiotics. We compared the in vivo and in vitro long-term effects of ethanol exposure on the expression of glutathione S-transferases (GST, EC 2. 5.1.18) in rat liver. Long-term in vivo ethanol treatment to achieve blood ethanol levels ranging between 10-50 mM was by liquid diet feeding. For in vitro experiments, rat hepatocytes co-cultured with rat liver epithelial cells were exposed to 17 and 68 mM ethanol for up to 10 days. Two weeks of liquid diet ethanol treatment increased total GST activity. Both Mu and Alpha classes and in particular the A1 and A2 subunits and the amount of their corresponding mRNAs were increased. Total GST activity was also increased in co-cultures after exposure to 68 mM ethanol for 10 days. However, the Mu class subunits M1 and M2 and the corresponding mRNAs were increased, rather than the Alpha class subunits. Thus, long-term exposure to ethanol induces hepatic GST both in vivo and in vitro, but different isoenzymes are affected. Consequently, extrapolation of in vitro data on GST expression and regulation to the in vivo situation must be judicious. During xenobiotic metabolism in cell culture, a shift in relative expression and induction of different GST forms may occur, resulting in either an under- or overestimation of effects.

Hormonal regulation of glutathione S-transferase expression in co-cultured adult rat hepatocytes

Glutathione S-transferases (GSTs) are subject to regulation by thyroid and sex hormones and by GH. We have used an in vitro experimental system comprising adult rat hepatocytes co-cultured with rat liver epithelial cells of primitive biliary origin, to distinguish between direct and indirect effects of various hormones on GSTs; to identify the GST subunits affected by individual hormones; and to investigate the level at which the hormones act. Tri-iodothyronine (T3), thyroxine (T4) and 17beta-oestradiol (OE2) reduced GST activities, whereas testosterone, dihydrotestosterone, and human growth hormone (hGH) had little effect on total GST activity. HPLC separation of the various GST subunits revealed that T3 and T4 reduced total GST content, in particular the abundance of subunits M1 and M2. The amount of the Pi-class subunit P1 was reduced by OE2. Treatment of the co-cultured cells with this hormone altered the GST subunit profile to one that is more similar to that observed in freshly isolated hepatocytes. Analysis of mRNAs demonstrated that some of the hormones act at a pre-translational level, whereas others act at a translational or post-translational level to regulate the expression of various GST subunits.

Observation of hepatotoxic effects of 2-n-pentylaminoacetamide (Milacemide) in rat liver by a combined in vivo/in vitro approach

Milacemide or 2-n-pentylaminoacetamide hydrochloride, a new glycine derivative, was found to cause elevations of plasma transaminases in patients suffering from severe depression and Alzheimer's disease. However, no signs of liver toxicity were observed during the course of earlier conducted subchronic and chronic in vivo studies in rodents and cynomolgus monkeys. In this study an in vivo/in vitro approach has been proposed to detect early alterations in key metabolic and functional liver capacities. Milacemide was administered by continuous i.v. infusion for 7 days to male Sprague-Dawley rats using subcutaneously implanted osmotic pumps. Doses were given of 0, 250 and 500 mg/kg per day. Body weight and food intake were recorded and at day 7 of exposure, Milacemide concentration, glucose, urea, triglycerides and cholesterol levels and alanine (ALT) and aspartate aminotransferase (AST) activities were measured in plasma. Non-esterified fatty acids were determined in serum. On day 8, after overnight fasting, hepatocytes were isolated. A portion of the cells derived from untreated animals (no osmotic pumps) were cultured in a primary monolayer and exposed in vitro to different Milacemide concentrations. The xenobiotic biotransformation capacity of the isolated hepatocytes was studied by measuring the cytochrome P450 content, ethoxycoumarin-O-deethylase (ECOD), pentoxyresorufin-O-deethylase (PROD), ethoxyresorufin-O-deethylase (EROD), aldrin epoxidase (AE), epoxide hydrolase (EH) and glutathione S-transferase (GST) enzyme activities. Triglycerides, cholesterol and phospholipid contents were measured on the isolated cells. At plasma concentrations of 43 and 130 microM Milacemide, the ALT activity was unchanged or significantly decreased, whereas the AST activity was increased in both cases. Other clinical chemistry parameters remained unchanged. Weight gain was significantly lower in rats treated with the high Milacemide dose. In addition, decreased food consumption was observed in all treated animals leading to significantly lower food efficiency factors for the rats treated with the high dose. Milacemide had a specific inhibitory effect on xenobiotic biotransformation: ECOD activity decreased to 60% of the control value for both Milacemide doses, PROD activity remained unaffected whereas EROD activity decreased to 65% of the control value. A decrease was also observed at the highest drug concentration for AE (to 41%), EH (to 65%), cytochrome P450 content (to 80%) and GST (to 85%). At 500 mg Milacemide kg/day, hepatocyte triglycerides levels increased 3.1-fold while cholesterol and phospholipid levels remained unaffected. Electron and light microscopy on total liver and isolated hepatocytes indicated a concentration-dependent accumulation of lipid droplets, the occurrence of numerous vacuoles in the cytoplasm and other structural abnormalities. When the cultured hepatocytes of control animals (without osmotic pumps) were exposed to Milacemide, the appearance of vacuoles and myeloid bodies could be confirmed in vitro. The results of this study using an in vivo/in vitro approach clearly show potential hepatotoxic properties of Milacemide, an effect not observed in conventional toxicity studies.