Two-Electron Reduction of Heme and Initiation of Heme Degradation

This is one possible psychedelic "loser" mechanism for the degradation of heme in association with the formation of a glutathionyl adduct. The "mechanism" is an adaptation, for better or worse, of the mechanisms proposed by these authors [Nakamura et al., 1984: (http://www.jbc.org/content/259/11/7080.full.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/6725282); Atamna and Ginsburg, 1995: (http://www.jbc.org/cgi/content/full/270/42/24876); Schaefer et al., 1985: (http://www.ncbi.nlm.nih.gov/pubmed/3927975)], and the main interest I have is in keeping track of the two-electron reduction of heme. It's difficult to keep track of things. It seems more likely that a thioperoxyl derivative of glutathione, as discussed by Atamna and Ginsburg (1995), would initiate the degradation reactions, but thioperoxyl species mimic the actions of peroxides, etc., and doing those again could get repetitive. Schaefer et al. (1985) proposed that the reaction would repeat more than once and produce dipyrroles and maleimides (or, here, related sulfhydryl compounds or disulfide cross-links with amino acid residues on proteins, etc.):

Note on the Reduction of Ferryl Heme by Hydrogen Peroxide

I was meaning to change this diagram to show the product that would be expected (the superoxide anion):

The point is that there's thought to be an unknown mechanism by which that superoxide can donate another electron to iron(IV), in ferryl heme in some peroxidase enzymes, to produce molecular oxygen and water, which I've drawn as one of the coordinating ligands. I should draw the reduction mechanism in the case of the hydrogen abstraction from an unsaturated fatty acid [RH ---> R(e-) + H(e-) (a hydrogen atom and an alkyl radical, R(e-), that can go on to form a lipid hydroperoxide and then a lipid alkoxyl radical, etc.)]. That's the overall reaction in the reduction of ferryl heme by hydrogen peroxide--a hydrogen abstraction (shown below). It makes more sense in view of that article by Traylor et al. (1993), cited in the previous posting. The reaction is very similar to an epoxidation, and Traylor et al. (1993) used different hemes to form epoxides in high yields. But ferryl heme doesn't normally form epoxides in vivo. They had to use special conditions. Incidentally, a hydrogen abstraction isn't the same as a hydride transfer:

Reductions of Ferryl Heme by Hydroperoxides; Reduction of Perferryl Heme via the Oxidation of a Tyrosine Residue; Heme-to-Protein Cross-Linking

It helps me to go through these reactions and to thereby avoid "having" to make use of these types of diagrams or equations in other contexts. If I can understand these things, I won't have to catalog equations, as so many articles do, or even discuss most of these things. One reason that many of the articles are so confusing is that hydrogen peroxide or an organic hydroperoxide (or a protein "substrate") can reduce ferryl hemes to ferric hemes or perferryl hemes to ferryl hemes. Those same compounds usually act as oxidants, however, and they also oxidize ferric heme to perferryl heme or ferrous to ferryl hemes. But anyway, the reduction of ferryl heme by HOOH (hydrogen peroxide) is more or less the same reaction as the reduction by an organic hydroperoxide, such as linoleate hydroperoxide or another lipid hydroperoxide. I'm not showing the reduction of perferryl to ferric heme, but that's the reaction that allows for the true "peroxidase" activity of non-protein-bound heme (or at least that's one way of looking at it). The products are O2, water, and ferric heme, even though there's a superoxo-Fe(III)-heme "intermediate." That's supposedly one reason some peroxidase enzymes can exert antioxidant effects at some low-to-intermediate concentrations. There's an intermediate range at which the peroxide-to-protein molar ratio allows for the consumption of hydrogen peroxide, by allowing the perferryl-to-ferric reduction pathway to predominate over the more damaging ferryl-to-ferric reduction pathway, but that doesn't usually work out too well in vivo, seemingly. The reaction I'm showing for the one-electron reduction of ferryl heme (one proposed by Traylor and Xu, 1987, cited below) by HOOH is similar, however, to that perferryl-to-ferric pathway. The reduction of ferryl to ferric heme is thought to produce a perhydroxyl species [Nagababu and Rifkind, 2004: (http://www.ncbi.nlm.nih.gov/pubmed/15548894)] or a peroxyl radical, etc. (that's the reason it's more damaging, under some conditions, than the other reduction pathway). The mechanisms aren't known (Traylor et al., 1993, cited below). Anyway, some antioxidants are likely to act by very similar one-electron reduction mechanisms. These two are adapted from the mechanisms proposed by Traylor et al. (1993) [Traylor et al., 1993: (http://pubs.acs.org/doi/abs/10.1021/ja00060a027)] and other authors [Traylor and Xu, 1987: (http://pubs.acs.org/doi/abs/10.1021/ja00254a059); Schaefer et al., 1985: (http://www.ncbi.nlm.nih.gov/pubmed/3927975)] and show the one-electron reduction of ferryl heme by an organic hydroperoxide, yielding ferric heme and an alkoxyl radical [the ROO radical shown, where R = a lipid or protein or other substrate, such as some antioxidants (ROO could also be RNO in some cases, etc.)] and, below that, a "hack" mechanism for the one-electron reduction of ferryl heme (to ferric heme) by hydrogen peroxide.

This is an adaptation of the mechanism proposed by Chu et al. (2000) for the formation of a tyrosyl phenoxyl radical via the reduction of a copper(II)-amine complex [Chu et al., 2000: (http://pubs.acs.org/cgi-bin/jtext?jpcbfk/104/i15/abs/jp994487d)(http://www.chem.yorku.ca/profs/hopkin/87.pdf)].
This shows the oxidation of a tyrosine residue by perferryl heme, and the reaction transfers the porphyrin radical and the "cation" of perferryl heme, in effect, to two different amino acid residues of a protein. This is known to occur to some extent, and the radical is delocalized throughout many amino acid residues of the protein, etc.

This is a hypothetical, proposed "bookkeeping" mechanism and shows a two-electron reduction
of perferryl heme to ferric heme that forms an intramolecular glycol and a heme-protein cross-link on one of the bridgehead carbons, I guess one would say:

Two-Electron Oxidation of Ferric Heme to "Perferryl" Heme: Cleavage of the Dioxygen Bond is Normally Heterolytic in Protic Solvents, for God's Sake

I'm basically putting these up for my own benefit and to see if it's possible to see them on Google's "2-inch-wide" blog space. It helps me to keep track of the "bookkeeping" of the electron distributions and formal charges to do this, even though perferryl heme has a net charge of 0, as far as I can tell, in the physiological pH range (and even though this is unlikely to be of interest to anyone). This type of thing is more like "meditation" than anything else, and doing it lets me use the obnoxious drawing program. The wireless is down, and, as a result, I'm "grounded" in nerd tasks.

The mechanism for the formation of perferryl heme {a.k.a. "Compound I," an intermediate that forms during the enzymatic cycling of ferryl and ferric heme, in peroxidase and catalase enzymes and in cytochrome P450-containing enzymes, and that also forms during the ferric-ferryl cycling of heme in aqueous solutions of non-protein-bound, monomeric or dimeric heme [initially as hemin (chloro-coordinated heme) & hematin (hydroxo-coordinated Fe(III), etc.)]} is my adaptation (meaning a "loser" electron-bookkeeping venture that makes sense) of the intramolecular mechanism proposed by Schaefer et al. (1985) [Schaefer et al., 1985: (http://www.ncbi.nlm.nih.gov/pubmed/3927975)] for the formation of a glycol, as an intermediate species in the degradation of heme to "oligopyrroles" and maleimides, and a cation radical (such that the electron density is "localized," in the porphyrin radical that's delocalized across the entire ring system, primarily on the meso carbons and pyrollic nitrogens, as it's supposed to be [La Mar et al., 1981: (http://www.jbc.org/cgi/reprint/256/1/237.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/7451437)]) and the mechanisms proposed by Ozaki et al. (2001) [Ozaki et al., 2001: (http://pubs.acs.org/doi/abs/10.1021/ar9502590)]. Perferryl heme, or Fe(V)-heme, is actually Fe(IV)-heme with a porphyrin radical, and researchers don't use consistent forms of notation in their descriptions of it as being either a pentavalent species or a tetravalent species with a pi-cation radical on the porphyrin ring. It's tetravalent, though, and it's almost impossible for me to make sense of the reactions without doing some electron bookkeeping and getting a crude understanding of the electron distribution in the different species.

I've shown a way of viewing the formation of the porphyrin radical that helps me understand the electron donation. One electron of the iron-oxygen "double bond," which is really a mess of complex d(pi)-p(pi) interactions and d(pi)-p*(pi), or whatever, backbonding interactions (it's basically intramolecular hyperconjugation that they're talking about, as far as I can tell), comes from iron, and one comes from the porphyrin ring, from a bookkeeping standpoint. The bond order is larger than 2 for the ferryl (it's analogous to a "carbonyl," C=O, group) oxygen-iron bond. In any case, it's a two-electron oxidation of iron and of the porphyrin ring, as a whole, from a formal-charge-sort-of oxidation state of Fe(3+) to Fe(5+), even though the heme iron is hexacoordinated or maybe pentacoordinated in the physiological pH range and is likely to be neutral, except in the "crypto-" protonated ferryl and perferryl heme species, in which the net charge is +1 [Carlsen et al., 2005: (http://cat.inist.fr/?aModele=afficheN&cpsidt=16551176)]. The site(s) of protonation are unknown but are near the heme iron in an unidentified location (Carlsen et al., 2005) [maybe on the nitrogens and the oxygen, rapidly exchanged by proton transfers in perferryl heme (a.k.a. compound I)? (http://scholar.google.com/scholar?hl=en&q=%22proton+transfer%22+nitrogen+pyrrolic+heme+cation+radical&as_ylo=&as_vis=0)]. I say the oxygen because there's evidence that the ferryl oxygen can be protonated at physiological pH values in some chloroperoxidase enzymes [Newcomb et al., 2008: (http://www.ncbi.nlm.nih.gov/pubmed/18174331)]. Damn, that's exciting--"basic ferryl groups" (Newcomb et al., 2008, p. 8179) on some heme-containing enzymes. But one way of looking at the internal electron transfer that forms perferryl heme, as shown in Figure 11 of Schaefer et al. (1985) and Figure 1 of Pan et al. (2006) [Pan et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16500709)], is to view pentavalent heme as being an "intermediate" or resonance form that doesn't really exist as a distinct species or that involves an abstraction, in another step that doesn't really work that way, of a second electron from the porphyrin ring.

I've shown the heterolytic cleavage of the dioxygen bond of an organic hydroperoxide (RO-OH of ROOH), and the upshot of a lot of articles is that the cleavage of that bond is heterolytic and not homolytic, under most conditions (except in the presence of micellar aggregates, to some extent). But lipid hydroperoxides (LOOH) and other ROOH compounds can still be oxidized by ferryl heme species to form radical products (Traylor et al., 1993: (http://pubs.acs.org/doi/abs/10.1021/ja00060a027)]. I'm showing the oxidation of ferric heme (to perferryl heme) by an ROOH species (this includes hydrogen peroxide, or HOOH). I'm too tired to upload my proposed mechanism for the transfer of a cation radical from the porphyrin ring to a tyrosine residue, to form a tyrosyl phenoxyl radical. Why not save some fun for another day.

Here's the initial step, showing the oxidation of ferric heme by an organic hydroperoxide or hydrogen peroxide, and the participation of or "catalysis" by a generic base, B(-):

___________________________

Trick for Overriding "Stemming" on Google Scholar

This is a not-very-advanced trick for overriding the "stemming" feature that can sometimes become problematic in searches on google scholar. In some cases, google scholar automatically stems words, such as creatine, to include the word creatinine, and that can give excessive numbers of results. I've seen other examples of search terms that produce that kind of "garbage" in the search results, but I can't think of those terms now. The big secret is to put the single word in quotations. This search gives a lot of junk on creatinine (http://scholar.google.com/scholar?hl=en&q=turnover+export+creatine) and provides 8,640 results, but this search, in which I meant to find articles discussing the mechanisms governing the export of creatine from cells and, hence, the total, intracellular creatine+phosphocreatine pools in cells, gave 3,860 results (http://scholar.google.com/scholar?hl=en&q=turnover+export+%22creatine%22) and less junk.

Magnesium as a CCK-Releasing Stimulus: Potential Relevance to the Use of Magnesium as an Adjunctive Treatment in Liver Disease

In this article [Ko, 2008: (http://www.ncbi.nlm.nih.gov/pubmed/18289201)], Ko (2008) discussed this article (the article the editorial accompanies) [Tsai et al., 2008: (http://socgastro.com/biblioteca/rev-articulos/pdf/magnesio200802.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/18076730)]. Tsai et al. (2008) found that higher intakes of dietary magnesium (Mg2+) intakes were associated with a decrease in the risk of developing symptomatic gallstones. I guess they looked at men only (?). It's not just relevant to men, of course. But Ko et al. (2008) cited some of that old research showing that intraduodenal Mg2+, in the form of Mg2+ salts and not chelates, induces gallbladder contractions by inducing cholecystokinin CCK release (http://scholar.google.com/scholar?hl=en&q=cholecystokinin+magnesium+gallbladder) from, I guess, enterocytes. I haven't read enough of all that research to get a sense of the mechanism, but the site of action is apparently localized. They've used large doses of Mg2+ in some of those articles, but I did some quick searching and reading and found a dose-response study in dogs that would tend to argue against the need for massive dosages. Ko (2008) was basically saying that normal amounts of Mg2+ could exert the same effect on a smaller scale.

On the surface, that research doesn't look like much, and the same effect can supposedly be shown for other electrolytes. I don't know, though. This article [Colombato et al., 1977: (http://www.ncbi.nlm.nih.gov/pubmed/835560)] shows a 17-fold increase in the rate of bilirubin output from the gallbladder/common bile duct in normal people, in response to Mg2+. That's significant, but the main thing I was thinking, in view of that editorial by Ko (2008), is that changes in the relative abundances of bile acids or increases in the biliary and intestinal-luminal-fluid concentrations of unconjugated bilirubin occur in people with liver disease and can bind massive amounts of magnesium [see here for some articles: (http://hardcorephysiologyfun.blogspot.com/2009/09/heterogeneous-precipitationnucleation.html)]. In one of those articles in the past posting [Heubi et al., 1997: (http://www.ncbi.nlm.nih.gov/pubmed/9285381)], the authors found that the equivalent of an adult's intake of 1100-2300 mg of Mg2+ per day was necessary to just prevent overt Mg2+ deficiency in children with liver disease, and that was attributed mainly to the sequestration of Mg2+ in the GI tract. In any case, in theory, that Mg2+ would exist mainly as mixed precipitates, bound to free fatty acids or bile salts or bilirubin, etc., and wouldn't be capable of participating in osmoregulatory disturbances. So the people might be able to tolerate much higher dosages of Mg2+ than people who do not have liver disease would. Some authors have remarked on the drastic differences in the amounts of Mg2+ that people can tolerate. That type of precipitation could be one reason for the large amount of variation. I think Ko (2008) also mentioned the anti-nucleating effect that Mg2+ could have. Magnesium binds to bilirubin or otherwise interacts with it in complex ways, also (http://scholar.google.com/scholar?hl=en&q=magnesium+bilirubin+binding), and can apparently prevent bilirubin from binding to precipitated calcium phosphate [Van der Veere et al., 1995: (http://www.jlr.org/cgi/reprint/36/8/1697.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/7595091)]. In general, Mg2+ has an "anti-nucleating" effect, in different contexts, even though I've seen articles discussing the capacity of Mg2+ to promote amorphous calcium phosphate formation but to also help to prevent, to a significant degree, soft-tissue hydroxyapatite/apatite crystallization. Amorphous calcium phosphate is basically innocuous, as far as I know, particularly in comparison to hydroxyapatite (in soft tissues). Anyway, the point is that people who have liver disease could require large amounts of Mg2+ to just to prevent deficiency (there's a lot of research showing that Mg2+ depletion contributes to exercise intolerance in liver disease, etc.), and even larger dosages could be required to produce the CCK-releasing effects that would occur in normal people at lower dosages. They used to say that CCK suppresses the appetite and used to suggest that people use L-phenylalanine to release CCK, etc. I don't think that's a good idea, but, in view of the problematic quality of much of the research on Mg2+ (research that is plagued by the use of miniscule and arbitrary dosages of oral Mg2+ and massive dosages of intravenous Mg2+ for everything under the sun), I wouldn't be surprised if the regulation of gallbladder contraction by Mg2+ is another thing that people stopped paying attention to in the 1980's, after the Bayh-Dole Act was passed or something. Bayh and Dole told them to buy influence in Universities via special interest groups and dole out advertising money for endless tv ads, and they "Bayh-ed it." No, that's nonsense.

Dependence of Antiapoptotic Effects of Glutamine on de novo Uridine Biosynthesis

Evans et al. (2005) [Evans et al., 2005: (http://ajpgi.physiology.org/cgi/reprint/289/3/G388)(http://www.ncbi.nlm.nih.gov/pubmed/15878985?dopt=Abstract)] found that the antiapoptotic effects of physiological concentrations of glutamine (GLN) (500 uM) in cultured, colonic epithelial cells were mediated, at least in part, by the capacity of GLN to serve as a precursor for pyrimidine (uridine and the other pyrimidine nucleosides and nucleotides derived from it) biosynthesis. One thing that's interesting is that the GLN-induced maintenance of pyrimidine nucleotide concentrations did not depend on transcription or on DNA replication, and it's possible that the maintenance of the pyrimidine pool maintained glucose uptake by maintaining UDP-hexosamine-dependent glycosylation reactions, etc. These results don't mean that GLN didn't partially prevent apoptosis by serving as a precursor of 2-oxoglutarate and the TCA cycle intermediates that are formed from the entry of 2-oxoglutarate into the TCA cycle. The results just mean that the maintenance of a pyrimidine pool can be an obligatory condition for the prevention of apoptosis in response to certain apoptotic stimuli. The fact that inosine, at millimolar concentrations, could prevent apoptosis in those experiments shows that ribose-5-phosphate and hypoxanthine, derived from inosine, can help maintain the ATP and PRPP, etc., that are required for pyrimidine salvage and de novo pyrimidine formation and for other antiapoptotic effects.

One interpretation of that article is that glutamine, taken at excessive dosages in people who have chronic liver disease or other inflammatory conditions, could exacerbate liver disease or other disease processes that are characterized by impairments in mitochondrial functioning. Orotate, an intermediate in uridine biosynthesis, can be used to induce fatty liver disease in animals, in part by depleting ATP and purine nucleotide levels [see here: (http://hardcorephysiologyfun.blogspot.com/2009/08/some-more-old-papers-of-mine.html)]. GLN is a substrate of type II (cytosolic) carbamoyl phosphate synthase ["glutamine-utilizing carbamoyl phosphate synthase (CPS-II)], and there would mainly be an accumulation of cytosolic carbamoyl phosphate (CP) in people exhibiting impairments in the overall activity of the urea cycle, such as in chronic liver disease. In the context of a trauma or other acute injury that can increase glutamine turnover, there probably wouldn't be the kind of damage to the mitochondria that would cause mitochondrial CP to be exported to the cytosol and cause orotate formation to increase pathologically. But mitochondrial dysfunction could be expected to cause orotate and dihydroorotate to accumulate, given that dihydroorotate dehydrogenase is a mitochondrial enzyme whose activity can be impaired in mitochondrial disorders. Szondy and Newsholme (1989) [Szondy and Newsholme, 1989: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1138925&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/2803258)] noted, however, that only something like 0.4 percent of the glutamine that's utililzed (the Km of GLN for glutamine utilization overall, by all pathways, is 2.5 mM) in proliferating lymphocytes goes toward pyrimidine biosynthesis, and the Km of glutamine for binding to CPS-II is 16 uM. That implies that the rate of pyrimidine biosynthesis is unlikely to be sensitive to increases in GLN availability. Also, Evans et al. (2005) used 500 uM GLN to prevent apoptosis, and the plasma GLN concentration and, hence, extracellular fluid GLN concentrations are around 400-600 uM in humans. But the regulation of pyrimidine biosynthesis may not always be normal or comparable in other tissues or individuals, and that's another reason not to use excessive dosages. Ammonia derived from GLN deamidation could also contribute to adverse effects in some cases, but one would mainly expect to see that in some disease states.

The Alanine-Glucose Cycle: Relevance to the Effects of Exercise on the Brain

The authors of this article [Roe and Mochel, 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16763896)] discussed the fact that skeletal-muscle myocytes normally export some alanine (ALN) and glutamine (GLN) and that the liver may utilize significant amounts of that ALN and GLN. Roe and Mochel (2006) referred to the process as being the "alanine cycle," and the process is also known as the "alanine-glucose cycle." Roe and Mochel (2006) noted that the ALN cycle is a "one-way street" that allows for alanine to be exported from the skeletal muscles and utilized by the liver as a precursor of pyruvate, particularly during strenuous exercise. Exercise is known to increase serum ALN, but that's obviously only one effect that occurs during some types of exercise. Evans et al. (2004) found that the intravenous infusion of L-ALN (that's not the same as beta-alanine, and beta-alanine supplementation can produce symptoms of neuropathy at relatively low dosages, in some cases [Harris et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16554972)]) improved some measures of cognitive functioning in nondiabetic people who were being made artificially hypoglycemic, but it's not clear if the effects were solely due to the entry of ALN into the brain or if the ALN-induced elevations in plasma lactate contributed to the effects [Evans et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/15089788)]. Incidentally, oral GLN reliably elevates plasma ALN, and, as a representative article, Déchelotte et al. (1991) found that jejunally-administered GLN elevated plasma ALN in humans [Déchelotte et al., 1991: (http://www.ncbi.nlm.nih.gov/pubmed/1903599)].

That capacity of the skeletal muscles to export GLN and ALN is likely to be relevant to research on the effects of exercise on the brain. These articles don't show the effect as unequivocally as some other articles do [Dalsgaard et al., 2004: (http://jp.physoc.org/content/554/2/571.full)(http://www.ncbi.nlm.nih.gov/pubmed/14608005?dopt=Abstract); Kemppainen et al., 2005: (http://jp.physoc.org/content/568/1/323.full)(http://www.ncbi.nlm.nih.gov/pubmed/16037089?dopt=Abstract)], but the utilization of serum lactate during exercise allows the brain to decrease glucose utilization, and that means that the oxidation of lactate in astrocytes and neurons can exert a "glucose-sparing" effect. It's fairly clear to me that the utilization of lactate becomes more significant during high-intensity exercise than during low-intensity exercise, but the release of lactate from the muscles is also higher. Resistance exercise is likely to ultimately, in the long term, allow the muscles to export substantially more ALN and lactate and GLN than endurance exercise is, and I'm talking about the postexercise period and in the fasted state. Endurance exercise tends to not increase muscle mass (as in the percent lean body mass) very much, and an increase in the efficiency of something like the ALN-glucose cycle requires, in my opinion, some sort of actual increase in muscle tissue. I've discussed some of the details about resistance exercise in past postings.

More Details on Glutamine Metabolism in the Liver, With Reference to the Brain, and in the Brain, Also With Reference to the Brain

These are some articles [Iglesias et al., 2001: (http://cat.inist.fr/?aModele=afficheN&cpsidt=13399682); Jia et al., 2006: (http://wjg.wjgnet.com/downpdf.asp?url=/1007-9327/12/1373)(http://www.ncbi.nlm.nih.gov/pubmed/16552804); Schuster et al., 2009: (http://www.ncbi.nlm.nih.gov/pubmed/19324476); Hong et al., 1992: (http://www.ncbi.nlm.nih.gov/pubmed/1546897)] that show the capacity of glutamine (GLN) to protect against various types of liver damage in animals. GLN reduced the amount of damage produced by acetaminophen-induced liver failure in rats (Hong et al., 1992), for example. The mortality rate in the group that had received GLN was 15% (4/26 rats), but the mortality rate in the control group was 46% (13/28). Hong et al. (1992) focused, to some extent, on the role that the preservation of hepatic glutathione (GSH) concentrations are thought to play in the resistance to acetaminophen-induced liver failure, and they discussed some of the research in which GLN availability, in some cell types, had been shown to be rate-limiting, for all practical purposes, for GSH biosynthesis. There is, actually, more of a rationale for the use of GLN as a precursor of GSH in something like acetaminophen-induced liver failure, as noted by the authors, than there is for the use of GLN as a GSH precursor in other contexts. As discussed by Hong et al. (1992), acetaminophen can reduce the overall GSH pool ([GSH] + [GSSG]) and doesn't just disturb the intracellular redox state, as reflected in the [GSSG]/[GSH] ratio. The authors cited research that had shown a [GSSG]/[GSH] ratio of less than .01 in the livers of animals that had been subjected to acetaminophen-induced liver failure, and, even in the presence of that ratio, the absolute levels of reduced GSH (GSH is the reduced state) levels were drastically depleted. But, in many cases, the assumption has been that it's enough to simply replenish the overall GSH+GSSG levels, and that's not necessarily going to be beneficial in many other contexts. But my point was that the hepatoprotective effects were not necessarily a result of the GLN-induced increases in hepatic GSH availability. But I've seen other research that supports the authors' assertion that GLN availability can be limiting for GSH biosynthesis under some conditions. I'd add that there's a lot of research showing that orally or parenterally-administered nucleotides can protect against experimental liver injuries in animals, and I tend to think that nucleotide monophosphates or diphosphates or triphosphates (or triacetyluridine), along with reduced folates, such as L-methylfolate or levoleucovorin, and some amount of methylcobalamin, would be more effective than trimethylglycine (betaine) or phosphatidylcholine or many of the other approaches to liver damage, but that's only my opinion. I've discussed that in past postings. Experimental liver damage is almost used as a kind of generic method, in animals, for evaluating mitochondrial toxicity or any number of other processes, and the research tends to be relevant to many other disease states that involve organs other than the liver.

Those articles, along with other articles, are relevant to an understanding of the way GLN behaves in cells that are largely nonmitotic or postmitotic, such as the brain. GLN is known to be utilized as a precursor of tricarboxylic acid (TCA) cycle intermediates and for other purposes by Kupffer cells and hepatic stellate cells in the liver, especially following a liver injury, and those cells are mitotic (again, especially after an injury). But my point is that it's possible to make crude comparisons between the effects of GLN on the liver and the effects of GLN on the brain. In contrast, the effects of GLN on proliferating lymphocytes can vary throughout the cell cycle, etc.

More specifically, the traditional model of the intercellular compartmentation of glutamine and glutamate metabolism in periportal and perivenous hepatocytes [Souba, 1991: (http://www.ncbi.nlm.nih.gov/pubmed/1892702)] is reminiscent of the traditional model of the glutamine-glutamate cycle in the brain, but neither one of those traditional models is especially helpful in allowing one to understand the effects of exogenous GLN on the brain (or liver). In each case, the traditional model is an oversimplification of the reality, in vivo, and can even become problematic, particularly if one is trying to understand the effects of exogenous (supplemental) GLN in any kind of disease state, as noted by Souba (1991). According to the traditional model, GLN is transported into periportal hepatocytes and deamidated into glutamate by glutaminase, a mitochondrial enzyme that is abundantly-expressed in periportal hepatocytes and that is presented as being expressed to a negligible extent in perivenous hepatocytes (Souba, 1991). Periportal hepatocytes also display more abundant (or, supposedly, exclusive) expression of urea cycle enzymes and utilize glutaminase-derived ammonia in the urea cycle. Perivenous hepatocytes are much less numerous than periportal hepatocytes [Haussinger, 1990: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1131284&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/2185740)] and supposedly are the primary or exclusive subtype of hepatocytes that express glutamine synthetase (Souba, 1991). That type of model would seem to suggest that exogenous GLN would be utilized only or primarily by periportal hepatocytes, but that's unlikely to really be the case. For example, Watford and Smith (1990) [Watford and Smith, 1990: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1131276&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/1970242)] found that perivenous hepatocytes did, in fact, display glutaminase activity, even though the glutaminase activity in periportal hepatocytes was 2.33 times the activity in perivenous hepatocytes. Those authors cited research that had shown that perivenous hepatocytes evidently comprise the only subtype of hepatocytes that expresses GLN synthetase. But the perivenous hepatocytes wouldn't even necessarily have to express glutaminase in order to be influenced by exogenous GLN, however. Many articles have shown that increases in the intracellular GLN concentration, up to 650 uM or so [Smith et al., 1984: (http://www.ncbi.nlm.nih.gov/pubmed/6146632)], at least, are accompanied by decreases in GLN synthetase activity, and Sandrasagra et al. (1988) [Sandrasagra et al., 1988: (http://www.ncbi.nlm.nih.gov/pubmed/2903721)] cited a lot of those articles, including some that have shown an inverse relationship between intracellular GLN concentrations and GLN synthetase activity. But the point is that some of the research has shown that GLN can fairly directly and rapidly, such as within 1.5 hours (Sandrasagra et al., 1988), decrease GLN synthetase activity. The mechanism still isn't known, but it could be that it's an allosteric effect or that the GLN-mediated increase in the proteasomal degradation of GLN synthetase requires the formation of glutaminyl-tRNA, etc. But some of the articles have shown changes in the Vmax in response to GLN, and it's unlikely that the ubiquitination of the GLN synthetase protein, by proteasomal enzymes directed to GLN synthetase by exogenous GLN (via unknown mechanisms), would be responsible for a GLN-induced decrease in the Vmax (Smith et al., 1984) of GLN synthetase, although anything's possible. So exogenous GLN could decrease GLN synthetase activity in perivenous hepatocytes, and that could reduce ATP consumption. Also, 2-oxoglutarate or alanine or aspartate or other substrates, derived from the metabolism of GLN in periportal hepatocytes, could enter perivenous hepatocytes, etc. The point is that the model of the "cycle" in the liver has the potential to be misleading and make it seem as if cells in the liver are limited, in a "strict" way, in their capacities to utilize or respond to exogenous GLN. One can look at research on the GLN-GLT-GABA cycle in the brain and get the same type of misleading sense that GLN can only be used by neurons and that it's going to build up in astrocytes and cause encephalopathy or something. It's ammonia that causes most of those effects, in my opinion, in the context of hepatic encephalopathy. Ammonia causes excessive calcium influx and inhibits multiple mitochondrial enzymes, and the resulting ATP depletion basically cripples the volume regulatory functions of astrocytes and causes astrocytic swelling that is partially a consequence of a cell-energy-metabolism-failure-induced impairment in the export of GLN from astrocytes. But the fundamental problem is not that GLN per se is causing osmoregulatory disturbances. The problem is mitochondrial dysfunction in astrocytes, in my opinion. The beneficial effects of GLN synthetase inhibitors in hepatic encephalopathy could, incidentally, be partly a result of decreases in ATP consumption and not just a result of decreases in GLN formation and in GLN-dependent astrocytic swelling.

Anyway, this posting is getting too long. I also found multiple articles that show that even intravenous GLN infusions tend to not elevate intracellular GLN concentrations in the liver or skeletal muscles, and the articles show the extremely high rate of turnover of GLN. Souba (1991), cited above, cited reference 6, on p. 291, as an example of research that had shown no increase in intracellular GLN in the liver in response to an infusion of GLN under pathological conditions, and that's relevant to the research that supposedly shows no entry of GLN into the brains of humans with traumatic brain injuries (see past postings). It's very likely that GLN did enter the brain in those people. Although Souba (1991) argued that it had been the slow rate of glutamine uptake that had been a limiting factor in the utilization of GLN, that's basically like saying that the rate of increase of intracellular GLN, in response to uptake + synthesis from glutamate or, by transamination, from aspartate, etc., had been less than the rate of decrease, either by glutaminase activity or export or deamidation by that family of non-glutaminase enzymes in the cytosol, etc. That's very similar to the situation at the blood-brain and blood-CSF barriers. The rate of efflux of GLN from the brain is 3-20 times (or something like that) higher than the rate of influx of GLN, but that doesn't mean that no GLN is being transported into the brain (and into astrocytes and neurons) (!). It says absolutely nothing about the amount of GLN that's passing through the interstitial fluid (ISF) in the CNS (to look at the ISF GLN concentration, in view of the research as a whole), especially given the drastic increases in the oxidation of GLN carbons in the TCA cycle, following ischemia, and the countless articles showing no elevations of intracellular GLN or even plasma GLN in response to parenteral or oral GLN. Watford and Smith (1990) discussed the concept that the intercellular cycling of GLN and ammonia in the liver is essentially a futile cycle (or, similarly, in the brain, although the discussions of the GLN-GLT cycle in the brain generally focus on GLT as a substrate of GLN synthetase and only focus on ammonia, as a substrate, in the context of hepatic encephalopathy and other pathological states), although the cycling is clearly not only an ATP-consuming system. But a lot of ATP is consumed. Anyway, the point is that there's a lot of research showing that GLN decreases GLN synthetase activity across small changes in extracellular GLN, and there's also research showing that GLN lessens the glucocorticoid-induced increases in GLN synthetase activity [Hickson et al., 1996: (http://www.ncbi.nlm.nih.gov/pubmed/8945950)]. Another relevant point is that Hickson et al. (1996) cited articles, on p. R1165, in which researchers had found that the glucocorticoid-induced increases in GLN synthetase expression had correlated positively with the degree of muscle atrophy or denervation of the skeletal muscles, in animal experiments. That's important and is another line of indirect evidence that the pathological effects of elevations in GLN synthetase activity are, at least partially, a consequence of increases in ATP consumption, as a result of increases in GLN synthetase activity. But a lot of articles discuss GLN formation as if it's favorable to energy metabolism. In fact, astrocytes oxidize GLN carbons very readily, just as neurons do, and the GLN-GLT cycle can appear to be "robust" or "harmonious" and can actually be consuming large amounts of ATP with very little to show for it, at least in terms of the amelioration of disease states. The articles that have shown the resistance of intracellular and extracellular GLN concentrations to infusions of large amounts of GLN also suggest, in view of the other research that has shown GLN-mediated decreases in GLN synthetase activity, that a lot of the research showing no apparent effect of lower-intensity exercise on GLN metabolism might be sort of missing the point. If one only looks at the plasma GLN level and finds no change, one could erroneously conclude that exercise produced no significant effect on GLN metabolism. One could even do an MRS study and show no change in the intracellular GLN levels, but the point is that there could be a major change in ATP consumption by GLN synthetase, even though lower-intensity, endurance exercise has sometimes been shown to decrease GLN synthetase activity or expression in skeletal muscle myocytes. Hickson et al. (1996) cited some of the other paradoxes. The main point I'd make is that there needs to be more research that looks at the GLN-GLT cycle in the brain in the context of GLN supplementation, given the vast and increasing amounts of clinical research on the effects of GLN supplementation.