Low Molecular Weight (poly)Phenol Metabolites Across the Blood-Brain Barrier: The Underexplored Journey

In silico predictions of passive permeation

Although the presence of the LMW (poly)phenols metabolites has been confirmed in circulation as mentioned, the mechanisms by which these molecules cross the cell membranes into circulation and reach target organs are still quite underexplored, especially in the case of the brain. Several types of membrane transport have been considered for (poly)phenols. For flavonoids, several mechanistic studies on their transport have been proposed, due to their molecular weight and other chemical properties like polar surface and number and position of the oxygen atoms among others discussed below. The main idea of a transport, based on passive transport, has been substituted by the concept of flavonoids transport be mediated by ATP-binding cassette transporters (ABC) and organic anion transporters (OAT) [3] and also organic anion transporting polypeptides (OATP) [28, 29]. However, in the case of LMW (poly)phenol metabolites, the facts about their transport is still largely unknown.

Considering that the transport of (poly)phenol metabolites across cell membranes is known to be either through passive permeation, or active transport [23, 30, 31], this could potentially be extended to LMW (poly)phenol metabolites as well. Never the less, in the case of small molecules, defined as <500 g/mol, and their potential transport through BBB, one of the most studied possibilities is passive permeation [25, 32–34]. Passive permeation depends on the physical-chemical properties of a molecule and could be predicted by in silico methods. Yet, a limited number of studies have been conducted on LMW (poly)phenol metabolites molecules. These methods rely on physical-chemical parameters denominated by molecular descriptors, that are intrinsic to a molecule, for example molecular weight, number of hydrogen acceptor and donor bounds, volume, dipole moment, and many others [32]. All these parameters can then be conjugated into a single property designated commonly as log BB that represents the logarithmic ratio between the concentration of a compound in the brain and blood. This log BB property can then be used to compare between a library of known molecules that cross the BBB and the potential molecules to predict. Many of these models are present in the literature and are usually associated with drug discovery [35]. Due to the different molecular descriptors used for each model, a direct comparison between log BB cannot be made and are usually model- dependent. The ability of several LMW (poly)phenol metabolites to passively cross the BBB according to different in silico models are reported in Table 2. The number of studies found on LMW (poly)phenol metabolites passive permeation was quite limited and data was only found for 23 of the described molecules in Table 1. Overall, most of the molecules evaluated demonstrate some probability of crossing the BBB, although this probability cannot be quantitative, it is indicating only that if the presence is expected in the brain it is due to passive permeation. One of the exceptions is salicylic acid where in silico predictions projected as not probable to reach the brain. However, this molecule has been detected in the rat brain and its presence was associated with transport-mediated mechanisms [36]. For some of these compounds like protocatechuic and, vanillic acid different estimates were given by different predictive models, demonstrating that the model used can influence the conclusions taken about passive permeability. All molecules analyzed in QikProp software seemed to indicate, through log BB, that the molecules could reach the brain. Nevertheless, different conclusions were drawn with SWISS ADME where di and tri-substituted phenolic acid metabolites (vanillic acid, protocatechuic acid and, gallic acid) were stated as not probable of crossing the BBB. Furthermore, SWISS ADME predicts ferulic acid has a molecule able to cross the BBB, together with p-coumaric and cinnamic acid but not sinapic acid. Information regarding other classes of LMW (poly)phenol metabolites like hippuric acids and propionic acids would greatly increase our knowledge of passive permeation for these types of molecules. Nevertheless, this data is still missing. Overall, in silico predictions since are less labor-intensive, inexpensive and high throughput, are quite useful for initial information about the potential of a metabolite to cross the BBB. Initial in silico screening in a set of compounds allows refining further analysis in the more technically challenging process of obtaining detailed in vivo BBB permeability data for the most promising compounds. A good example of this rationale was the comprehensive study that was done for the flavan-3-ols metabolites, phenyl-γ-valerolactones, phenylvaleric acids, and their conjugates. This study explored through an in silico analysis the most promising molecules which were tested in a cell-based assay. Finally, the presence of the above mentioned compounds was validated in the brain in three different in vivo studies [33]. It will be very interesting to perform a comparative study for all described LMW (poly)phenols metabolites in the same model of passive permeation - an evaluation that, to the best of our knowledge, to date still has not been done.

Cell based data on BBB permeability

In vitro cell-based studies, mostly taking advantage of brain endothelial cellular models, have been a very important tool to address the ability of metabolites permeation across the BBB endothelium in a more realistic way. For some flavonoids and their known circulating metabolites, the permeability was tested on in vitro models of the BBB [21]. The uptake of hesperetin, naringenin, and their respective in vivo glucuronides, as well as the anthocyanins cyanidin-3-rutinoside and pelargonidin-3-glucoside has been reported in two brain endothelial cell lines from mouse (b.END5) and rat (RBE-4). In this study flavonoids permeation across in vitro BBB mono layers was compared to the potential for permeation for their more polar glucuronidated conjugates and to specific phenolic acids derived from colonic biotransformation of flavonoids. The authors concluded that hesperetin and naringenin present a high apparent permeability [21]. Moreover, in another study, both catechin and epicatechin were able to cross two BBB cell lines, RBE-4 cells and hCMEC/D3 (immortalized human cerebral micro vessel endothelial cell line) in a time-dependent manner [43]. In vitro transmembrane transport of flavonols, flavan-3-ols and anthocyanins and some of their methylated and glucuronidated metabolites were also observed in the hCMEC/D3 cells [44]. Interestingly, the authors concluded that in most cases, the metabolites exhibited higher transport efficiency than their unconjugated parent compounds [44]. Overall, data strongly suggest the effective uptake of dietary (poly)phenols and their metabolites through the BBB endothelium. However, there are very few studies focusing on the LMW (poly)phenols metabolites and their ability to reach the brain. In fact, a huge potential in this field has been underestimated, and only for a few classes such as benzene diols and triols, benzoic acid, and cinnamic acid data was collected (Table 3).

In a recent study, the endothelium transport of some known LMW (poly)phenol metabolites, products of sulfate conjugations was addressed by using a human brain micro vascular endothelial cell (HBMEC) line [25]. The same model was used to confirm flavan-3-ols metabolites BBB permeability followed by in vivo validation [33]. The BBB cellular models consist of the ability to cultivate cells on a transwell that divides two separate chambers mimicking the luminal (blood) and the abluminal (brain) sides separated by brain micro vascular endothelial cells. Catechol-O-sulfate, pyrogallol-2-O-sulfate, 1-O-methylpyrogallol-O-sulfate, 2-O-methylpyrogallol-1-O-sulfate, vanillic acid-4-O-sulfate, 4-O-methylgallic acid, and 4-O-methylgallic acid-3-O-sulfatehave been tested for BBB permeability and all of them were detected in the abluminal side [25]. However limited amounts of these LMW metabolites were observed to be transported, ranging from 1.3 to 12.6% of transport [25]. One of the possibilities to originate the low passive diffusion could be the involvement of efflux transporters pumping back the metabolites to the luminal side. In the same study the efflux transporters P-gp, BRCP, and MRP1 activity were tested in the cells for the transport of some LMW (poly)phenols metabolites and the percentage of transport was not significantly affected [25]. Interestingly, phase II derivatives, sulfated and methylated, presented a higher percentage of transport comparing to their counterparts, suggesting being more prone to be brain permeable. In any case, passive permeation may not be the only method of transport across the BBB, since it could also be mediated by carriers [25].

Some ready-to-use BBB model kits were already tested for LMW (poly)phenol metabolites BBB permeability. In that sense, resorting to the BBB kit (RBT-24), made of primary cultures of rat brain endothelial cells, pericytes and astrocytes, the transport for gallic acid, pyrogallol and pyrogallol-1-O-glucuronoide was demonstrated [45,46]. Despite of the pyrogallol is less polar than gallic acid, this last one presented a higher permeation since it is known that it can be incorporated into cells through an organic anion carrier (OAT3) [48]. Furthermore, another commercially available in vitro model of BBB (RTU), consisting of bovine endothelial cells, was used to test the potential of ferulic acid, dihydroferulic acid, and dihydrocaffeic acid to cross the BBB [47]. In fact, none of these metabolites presented a high capacity to cross these endothelial cells, presenting permeability coefficient (Pe) values slightly above the adopted cutoff value of 1×10^–3 cm.min^–1 (Table 3). This means that their permeation to the brain could be considered as negligible. This medium capacity to cross the BBB, could result, as mentioned before, from a low passive diffusion or even can be due to the involvement of efflux transporters pumping back the metabolites to the luminal side. In opposition, in Madin-Darby Canine Kidney (MDCK) cells, which is considered a useful model to evaluate cell permeability, ferulic acid rapidly crossed this barrier, increasing the transport rate overtime, reaching approximately 25% [49]. In a different BBB model consisting of ECV-304 bladder carcinoma cells and C6 glioma cells in co-culture, also described as a good indicator of BBB permeability, the authors were unable to detect the transport of (3,4-dihydroxyphenyl) acetic acid, 3-(4-hydroxyphenyl) propionic acid and 2-(4-hydroxyphenyl) acetic acid [21].

Interestingly, in silico predictions about the permeability of the specific molecules with available information show corroborating data between in vitro and in silico predictions. However, the number of in vitro studies is still limited to generalize this validation of the in silico data.

Still, although the data available regarding LMW (poly)phenols BBB permeability is scarce, it has gained a lot of attention since the BBB represents the entrance door to reach the brain. Noteworthy, sometimes the goal of the research is to find methods to avoid brain penetration of toxins and potentially dangerous molecules but sometimes it is convenient to promote the entrance of drugs. An interesting branch of research has been focusing on increasing the efficiency of drug delivery with the purpose of overcoming the tight regulation characteristic of the BBB. In that sense, sinapic acid, conjugated with a zwitterionic polymer in an encapsulated bovine serum albumin-based nanoparticles, proved to be a novel bioinspired BBB-permeable ligand for delivery of cargo into the brain [50], which is seen as a key factor for the therapeutic efficacy in the treatments of CNS-related diseases.

In vivo experiments confirm the presence of LMW (poly)phenol metabolites in the brain

For a full comprehension regarding the brain uptake of phenolic metabolites, in vivo experiments are crucial since they are the only way to fully recapitulate the complexity of the BBB. Being this a very emergent issue, several studies have been interested in confirming the brain bioavailability of (poly)phenols, having already been identified some of them in different brain regions of rats [51, 52] and pigs [53, 54] which can accumulate in a non regional-specific manner [55]. Micro dialysis sampling in rats also showed the presence of (+)-catechin and (–)-epicatechin in the brain, demonstrating their ability to cross the BBB [56]. The exact localization and target sites of (poly)phenols in the brain represent an important aspect for understanding its molecular mechanism but this issue remains unlooked. The main reason for this is the fact that the major analytical approach for this evaluation in biological samples relies on chromatographic techniques. Some attempts to address this was done by using monoclonal antibodies for the compounds [57, 58]. Also, the brain presence of (poly)phenols seems to be independent of their route of administration, since different compounds were detected in the brain after different types of administration. For instance, epigallocatechin gallate [59, 60], epicatechin [61–64], anthocyanins [52, 65], quercetin [66, 67] and naringenin [68], were detected in the brain after oral administration while naringenin and its glucuronide were also detected after intravenous injection [69] as well as hesperetin that was detected in the brain striatum [70]. However, there are some contradictory results as is the case of anthocyanins that could not be detected in brains of rats after supplementation with raspberry juice in a dose equivalent to 700 mL to a 70 kg human [71] but were detected in other experimental designs that involved either higher doses or chronic administration [52, 72, 73]. For analysis of a compound penetration in the brain it is the gold standard to use exsanguinated or perfused animals or a correction for residual blood in the brain [23]. This was one critical aspect that has raise some controversy since these appropriated control procedures were lacking in some studies.

By contrast to the dietary (poly)phenols, only a very limited number of studies have focused on LMW (poly)phenol metabolites, although some evidence can be found of their potential to cross the BBB and reach the brain. Among predictive studies, whose outcomes provide an idea about the ability of metabolites to cross the BBB endothelium by passive diffusion and in vitro studies that can give us a wide-ranging clue about their permeation, both approaches are still limited. On the other hand, in vivo models, despite being based on rodent and pig models, are those that get closest to recapitulate the human physiology, undergoing the influence of natural processes of a living organism. This means that in animal studies the brain bioavailability of phenolic metabolites will be influenced by different cell types, enzymes, transporters, endogenous molecules, etc. mimicking in a better way the organism environment.

Although limited, evidence of the BBB capacity to uptake LMW (poly)phenol metabolites using in vivo models have been increasing in the literature. However the main findings attained in these studies, have different experimental designs, particularly regarding the animal model, the gender, the type of administration, the dose, and the compounds sourced, hampers an exact comparison. Within this section, we review some of these studies, highlighting the compounds that were tested to cross the BBB, the dosage, and the brain uptake, using in vivo models. An important difference in these studies, that lead to different conclusions, is the type of administration of the (poly)phenols. Compounds were sourced either by extract form or pills being administered intragastrically or orally mixed with the diet (Table 4), or they can be administered themselves directly either by intravenous or intraperitoneal injection (Table 5). The difference between these two approaches is the metabolic route that compounds will follow. If they enter the organism orally, the pharmacokinetics of the metabolites increase the difficulty for the exact quantification of the amount that reaches the brain and the calculation of the proportion that is brain-bioavailable. The extensive metabolism over the (poly)phenols, present in the extracts, by the microbiota and the entero-hepatic enzymes, will lead to difficulties in estimating the concentration of a specific metabolite in blood and the kinetics (T_max, T_1/2) associated with their presence. On the other hand, if the administration is directly in the blood stream it will bypass most hepatic metabolic pathways, reaching the tissues faster and making it easier to know with more accuracy the percentage that reached the brain.

Margalef et al. have administered orally different acute doses (125, 250, 375, and 1000 mg/Kg BW) of grape seed (poly)phenol extract (GSPE) to the male rats [74]. The total of 16 different phenolic acids that generate a set of phase II metabolites found at different concentrations in different tissues were identified from this extract [74]. As expected, given that the brain is a peripheral organ difficult to cross, metabolites that were quantified in this tissue presented much lower concentrations than those found in the liver and kidney. Notably, 4 of the human LMW (poly)phenol metabolites (listed on Table 1) were detected in brain plus phenylpropionic acid, an LMW (poly)phenol metabolite detected in rodents, being that a proportional increase with the increased dose of GSPE was not observed. Benzoic acid, 3-hydroxybenzoic acid, homovanillic acid, and hippuric acid were detected after 2 hours of GSPE ingestion. In turn, gallic acid, 3-O-methylgallicacid, and 4-hydroxyhydrocinnamic acid were not detected in the brain of these rats for any dose tested, despite being detected in other organs in the same animals. Moreover, gallic acid was present in the extract and 4-hydroxyhydrocinnamic acid could be generated by the metabolism of quercetin-3-O-galactoside, which was also present in GSPE [74]. The same authors evaluated if the GSPE extract could originate a brain accumulation of the metabolites after a long-term administration in the rats [75]. The authors concluded that the concentrations detected in the brain, 21 hours after daily GSPE administration of 100 mg/Kg BW for 12 weeks, do not reflect an accumulation but instead originated from the last acute administration dosage [75]. This conclusion was based on the fact that flavanol metabolites in plasma were not detected 21 hours after GSPE administration [75]. Moreover, very few metabolites were detected in the studied tissues and at levels much lower than the total amounts that were quantified 2 hours after a single GSPE ingestion of a similar dose in the previous study [74]. In the study, conducted by Wang et al., also using GSPE administered for 11 days, the content for 12 LMW (poly)phenol metabolites in the brain were analyzed [76]. Among the 12 screened compounds 8 were found in the brain but only 3-hydroxyhydrocinnamic acid and 3-hydroxybenzoic acid were found significantly increased compared with the control animals (4.5 and 2.5-fold change respectively). Hippuric acid, 4-hydroxybenzoic acid, 3-hydroxyphenylacetic acid, homoprotocatechuic acid, dihydrocaffeic acid and phenylacetic acid were also detected in the brain, but no differences were found compared with rats without GSPE [76]. This differential profile of the LMW (poly)phenol metabolites detected in the brain using the same source the GSPE extract could be due to several reasons. One possibility is the time of analysis, in the latter study, the analysis of LMW (poly)phenol metabolites present in the brain was done 6 hours post-GSPE administration [76] the opposite of 2 and 21 hours post-administration in the previous studies [74, 75]. LMW (poly)phenol metabolites appearance in circulation in vivo peaks around 6 hours post-consumption and have longer elimination half-life [85, 86]. Accordingly, it is not clear if a chronic consumption vs. acute ingestion could impact the levels detected since there is no a direct comparison for the same time point of tissue collection after the last intake. Another aspect that may account for different results is the fact that these studies have used different rat strains (Sprague-Dawley vs. Wistar rats) whose endocrine system was described as different and this causes differences in food conversion and food intake [87]. Altogether, these differences in the experimental design may justify the qualitative and quantitative differences found in the rats’ brains of LMW (poly)phenol metabolites in both studies.

Concerning the presence of gallic acid in the brain, another study also using Sprague-Dawley rats, likewise has not detected this metabolite in the brain when rats were orally gavaged with acute doses of GSPE (50, 100 or 150 mg/Kg BW) [79]. Only upon repetitive dosing with GSPE for 10 days, trace levels of gallic acid were detected [79]. In another study, gallic acid was also not found in the brain of control animals that were orally treated with Gualou Guizhi granules, a standard prescribed drug from Traditional Chinese Medicine. In the same trial, however, the rats that were subjected to cerebral ischemia/reperfusion injury and, treated also with the Gualou Guizhi granules, presented higher permeability of BBB, favouring the passage of several metabolites, including the gallic acid [88]. The BBB integrity is an important factor, even though BBB is highly regulated, towards an insult or a disease it may be disrupted, and therefore it increases its permeability, favouring the entrance of molecules that otherwise would not enter.

Protocatechuic acid was detected in the brain of rats that received Danshen extract intragastrically (containing danshensu 40 mg/Kg BW, protocatechuic aldehyde 149 mg/Kg BW and salvianolic acid B 50 mg/Kg BW) [78]. Additionally, male Balb/cA mice with a diet supplemented with protocatechuic acid have also shown increased levels of this metabolite in the brain, which were not detectable in brains from mice not supplemented [77]. Ferulic acid was also found in the brain of rats that took Shunaoxin pills, a traditional Chinese medicine product enriched in ferulic acid. It was showed that ferulic acid is rapidly absorbed and distributed in the brain, being detectable 5 minutes after the administration and no longer at 4 hours [49].

Although not derived from colonic metabolism, metabolites derived from intake of hydroxytyrosol food sources, like virgin olive oil, are also LMW (poly)phenol metabolites. Homovanillic acid and some related metabolites detected in rodents like homovanillic acid sulfate and homovanillic alcohol sulfate, were detected in the brains of the rats 5 hours after an acute high dose of hydroxytyrosol (100 mg/Kg BW) [81]. All these metabolites appear in the brain of control animals derived from dopamine metabolism which masks the appearance of the metabolites for the intake of lower doses of hydroxytyrosol. Homovanillic alcohol sulfate was also reported to increase in brains of rats that consumed daily standard diet supplemented with 5 mg/Kg BW of either hydroxytyrosol or hydroxytyrosol derivatives for 21 days [89]. Moreover, sulfates of vanillin, vanillic acid and homovanilic were also detected in the brain of rats after the ingestion of phenolic extract from olive cake [80].

Interestingly a significant gender effect in the plasma, liver, and kidney concentrations of hydroxytyrosol metabolites was observed [81]. However, no gender effect was observed in the uptake of hydroxytyrosol metabolites in the brain and heart [81]. On the other hand, gender effects were considered to influence brain uptake of (poly)phenols in the case of GSPE derived flavanol metabolites where differences between males and females were observed [74, 90, 91]. Being (poly)phenols recognized by the human organism as xenobiotics, it is not surprising to observe differences between genders in its metabolism. Such differences can be potentiated by age, estrous/menstrual cycle and pregnancy. Generally, it is described that male rodents tend to have higher levels of cytochrome P450, while phase II metabolism seems to be quite similar between genders. Some reports, however, show higher glucuronidation in males in the liver, compensated with higher glucuronidation in the female kidney. Moreover, some CYP450 expression seems to be sex-specific, and their expression levels quite different between genders meaning the rate of metabolism and excretion can be highly dependent on the substrate [92]. Gender affects the balance between the sulfation, glucuronidation, and methylation of (poly)phenols and the way how xenobiotics affect metabolic enzymes in the liver of rats is also different between sexes [90]. It has been reported for in vitro models of BBB that estrogens may play a role in modulating free flavanol uptake, and it has been further suggested that progesterone can act as an endogenous factor that modulates P-glycoprotein ability to serve as transporters of flavan-3-ol’s across the BBB [43].

Another approach to evaluate brain permeability to (poly)phenol metabolites in vivo is its direct administration to the circulation bypassing the gastrointestinal tract (Table 5). In that sense, a pharmacokinetic study tried to uncover the fate of microbial metabolites of dietary (poly)phenols upon intravenous injection in the dorsal penis vein of rats [51]. The injection site allows the minimal manipulation of anesthetized animals avoiding the release of inflammatory mediators that might affect BBB permeability, the normal distribution, and excretion of metabolites. From the 23 metabolites present in the mixture injected, 18 are LMW metabolites described in Table 1. From these, 13 were found in control brains as endogenous but only 9 were found increased compared to the control group (Table 5). Among them, caffeic acid has reached the brain at 15 minutes at approximately 34 times the basal concentration (0.11 pmol/g of tissue) and has presented the biggest difference between controls and injected rats, but it was the one detected in lower concentration (3.81 pmol/g of tissue). On the other hand, vanillic acid presented the highest percentage of transport to the brain considering both the dose injected and the mean concentration found in control brains [51]. Interestingly within the 9 LMW (poly)phenol metabolites, the top 4 that presents the higher percentage of transport are all phenolic acids (gallic acid < 4-hydroxybenzoic acid < homoprotocatechuic acid < homovanillic acid < vanilic acid). In fact, an interesting consideration could be made about these molecules, since homovanillic acid, one of such top 4 molecules was already described to be transported in rats brain by the activity of the Organic Anion Transporter 3 (rOAT3) [93]. Based on this, we may expect that this type of transport could also be occurring for other phenolic acids.

In an attempt to identify circulating hydroxytyrosol metabolites and their deposition in tissues, D'Angelo, and colleagues injected intravenously radiolabelled ¹⁴C hydroxytyrosol in Sprague-Dawley rats [94]. Although the radioactivity up taken by the brain was lower than in other organs, besides the hydroxytyrosol, homovanillic alcohol, 3,4-dihydroxyphenylacetaldehyde, 3,4-dihydroxyphenylacetic acid, homovanillic acid and sulfate conjugates were also detected in the brain. The authors consider that some of these metabolites were presumably derived from enzymatic activities operating in the brain, like methylation that reflects COMT activity [94]. In fact, in order to evaluate the effects of hydroxytyrosol and its nitro derivatives on COMT activity, compounds were administered into the rat striatum through a microdialysis probe [95]. All compounds increased extracellular levels of 3,4-dihydroxyphenylacetic acid during the perfusion and did not produce a decrease in the extracellular homovanilic acid suggesting their effect as COMT inhibitors [95]. The same results were observed in rats that received either an acute (single dose; 20 mg/Kg, i.p.) or chronic (one daily dose for 5 days; 20 mg/Kg, i.p.) treatment with the compounds [96].

Another study has addressed the penetration of catechol and pyrogallol into male mouse brain after intraperitoneal injection of 60 mg/Kg BW and 120 mg/Kg BW respectively, being both compounds detected in the brain [83]. Similar observations were made in female mice after the administration of both compounds intraperitoneally [97]. In both studies, catechol and pyrogallol reached the brain at very high concentrations, reflecting the supraphysiological concentrations administered to mice. Although it was not the goal of these studies, the results do not indicate noticeable sex differences in the ability of the compounds to reach the brain. However, this could be an important factor to be considered in futures studies, since the sex dimorphisms in many aspects of COMT’s function are already described and therefore relevant for benzene diols/triols and also dihydroxyphenylacetic acids. For instance, it was detected a prominent sex differences in the impact of tolcapone, COMT inhibitor, on dopamine metabolite levels. Specifically, females showed a greater tolcapone-related change in 3,4-dihydroxyphenylacetic acid in prefrontal cortex and cerebellum, and in homovanilic acid in the prefrontal cortex and striatum [98]. In fact, gender effects were already considered to influence (poly)phenol brain uptake as previously discussed.

Concerning these data, different conclusions can be drawn if the proper interpretation is not made. There are many variables between studies regarding their experimental designs that influence the amount and type of LMW (poly)phenol metabolites detected in the brains that should be taken into consideration. From the studies herein presented, the source of (poly)phenols is quite wide and variable and even the regular diet given to the rats can be different. The time of fasting and the time after the last (poly)phenol administration and the quantification are critical parameters that generate different results. As we already highlighted, the rat strain and gender, the concentration of the compound, as well as the route and the frequency of the administration have also a strong influence on the amount and type of LMW metabolites able to reach the brain.