Red Bone Marrow Density, Resistance Exercise, and Circulating Endothelial Progenitor Cells

This article is interesting [Payne et al., 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17408874)], and the authors suggest that physical exercise may increase the bone marrow density and that very sedentary lifestyles may contribute to the development of unexplained anemia, particularly the state of so-called "anemia of chronic disease." In anemia of chronic disease, pro-inflammatory cytokines, such as tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6), among others, are thought to be released from monocytes and endothelial cells (and adipocytes, even), etc., and suppress hematopoiesis. There are ways of helping to diagnose anemia of chronic disease by examining the ratio of soluble transferrin receptor (sTfR), which is not the same as serum transferrin (sTf), to serum ferritin. There are these algorithms for interpreting the ratio and drawing guarded conclusions from it.

I think the most interesting factor, from a mechanistic standpoint, that the authors discuss is the regulation of bone marrow mesenchymal stromal cell differentiation by growth hormone (GH) and other factors that are released in response to resistance exercise. Evidently GH (and presumably IGF-1) may be able to promote the osteogenic (osteoblastic) differentiation of stromal cells [Moerman et al., 2004: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1850101&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/15569355)], to some extent, and may be able to increase the numbers of osteoblasts. Gevers et al. (2002) [Gevers et al., 2002: (http://endo.endojournals.org/cgi/reprint/143/10/4065)(http://www.ncbi.nlm.nih.gov/pubmed/12239118?dopt=Abstract)] also suggested that lipids, mobilized from bone marrow adipocytes, could serve as an energy source for hematopoietic cells. Gimble et al. (1996) discussed the same concept, and those authors are basically saying the yellow marrow isn't necessarily completely undesirable. I think that GH release and the whole set of responses of growth factors to resistance exercise would be the type of thing that could help to lend credence to the article by Payne et al. (2007). Obviously, the mechanical stress from the resistance exercise would be necessary also, and mechanical stress activates localized growth-factor responses in osteoclasts and osteoblasts. But the proportion of bone marrow that's "yellow marrow" increases substantially between the ages of 17 and 21 or something, and it sort of corresponds to the decline in GH output and to the end of physical growth, etc. Gevers et al. (2002) also discuss the fact that decreases in bone mass (i.e. decreases in bone density, due to aging or immobility) are associated with increases in the numbers of bone marrow adipocytes (yellow marrow), and that's the type of thing that Payne et al. (2007) discuss. Gevers et al. (2002) note that there's generally an inverse relationship between the numbers (i.e. density) of osteoblasts, in parts of the bones, and the numbers of bone marrow adipocytes.

It's strange the way there's not much research on factors influencing the density of the red marrow, and the bone marrow is presented and viewed as if it's this invoilable and immutable organ with an almost endless capacity to adapt to any physiological challenge. It's obviously more or less true that the bone marrow has this "vast regenerative capacity" to recover from damage, etc. But there hasn't been much of an attempt to address the factors, apart from pharmacological interventions, that would allow that regeneration to occur. It's interesting that the hematocrit (i.e. percentage of the blood volume that is red blood cells) is known to be positively associated with the lean body mass (i.e. muscle and bone) [I think this is one of the many articles that shows that: Muldowney, 1957: (http://www.ncbi.nlm.nih.gov/pubmed/13414148)]. I think the assumption is that it's a result of a higher rate of oxygen consumption into cells (higher resting metabolic rate, etc.), but maybe the effect is partially due to changes in the red marrow in response to resistance exercise. There's probably a lot of research that discusses this in more depth, but I haven't really looked into it. This type of effect of resistance exercise could be important in the context of aging or atherosclerosis, given that one might expect a higher hematocrit to be accompanied by higher numbers of circulating potentially-antiatherogenic, CD34+ endothelial progenitor cells [this isn't all that good an example, but the point is that hematocrit seems to correlate positively with the numbers of circulating endothelial progenitor cells: Bahlmann et al., 2003: (http://www.nature.com/ki/journal/v64/n5/pdf/4494070a.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/14531796)].

Protection Against Postischemic Damage by Uric Acid and Evidence Arguing Against a Strongly Destructive Role for Xanthine Oxidase Activity

This article [Mink and Johnston, 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17335786)] shows that the intravenous infusion of 600 umol/kg/hr of hypoxanthine (HPX) reduced damage to the brain following experimental hypoxia and cerebral ischemia (8 minutes of each). The authors compared the degree of damage in the group that had received HPX, given during the interval of 30-minutes before and 30 minutes after the induction of ischemia (the 30 minutes "after" had included the 16 minutes during which the animals had been subjected to hypoxia and then ischemia), to the degrees of damage in the groups that had received either an infusion of xanthine (XAN) or an inactive vehicle. The relative degrees of damage were not significantly different in the groups that had received either xanthine or the vehicle, and the authors noted that these data, along with other data, argue against the notion that free-radical production from xanthine oxidase activity contributes strongly to damage following cerebral ischemia. The authors also found that the cerebral HPX levels, measured after the four hours of reperfusion had elapsed, were actually much lower in the group that had received the HPX infusion than in the other groups. Given that these decreases in the post-reperfusion HPX levels had been accompanied by reductions in postischemic brain damage, the authors noted that the HPX had probably been salvaged heavily or converted into xanthine, etc. Some of the XAN in the XAN group may have also been used for salvage (XAN can be salvaged to xanthosine and guanosine).

These results are important and suggest, as noted by the authors, that the protective effects of allopurinol, a xanthine oxidase inhibitor used to treat hyperuricemia, against cerebral ischemia are mediated by mechanisms other than XO inhibition per se. In other words, the reduction in uric acid is probably not protective in and of itself (especially since uric acid has been shown in many studies to protect against experimental, ischemic brain injuries) and, more importantly, the xanthine oxidase activity per se seems unlikely to be a major source of oxidative damage to the brain. The authors provided more XO substrates, which are HPX and XAN, and did not find increases in ischemic damage. The authors noted that the high Km value for XAN binding to XO (88 uM) as a substrate means that XO was not saturated with XAN during the experiments and wasn't in untreated animals either (the brain XAN content in the cerebral cortex of the rabbits was only 18 uM at 30 minutes after the initiation of reperfusion). This means that exogenous XAN is likely to have been utilized as a substrate for XO, as noted by the authors, but also means that the XO activity in the control group might not have necessarily been lower (the XO enzymes in the control rabbits would presumably have had access to plenty of endogenously-produced XAN, etc.). There's an article showing that allopurinol can exert antinociceptive effects that are evidently due to elevations in the cerebrospinal fluid guanosine and adenosine levels, and some of the antinociceptive effects can be blocked by adenosine receptor antagonists [Schmidt et al., 2009: (http://www.ncbi.nlm.nih.gov/pubmed/19133997)]. Those types of elevations in CSF guanosine and adenosine could also explain the neuroprotective effects of allopurinol. Even though the paper by Mick and Johnston (2007) doesn't necessarily preclude a damaging effect of XO-derived reactive oxygen species, the evidence, in my opinion, that XO activity is "all-bad" or that uric acid is a mediator of ischemic damage is not compelling at all, to say the least. Betz et al. (1991) [Betz et al., 1991: (http://www.ncbi.nlm.nih.gov/pubmed/1996699)] also noted that XO activity is unlikely to make a large contribution to oxidative damage following ischemia. To the contrary, uric acid has been shown to protect against postischemic damage, in many articles, in both humans and animals [here are a couple of the articles that show or discuss this: Yu et al., 1998: (http://www.ncbi.nlm.nih.gov/pubmed/9726432); Amaro et al., 2008: (http://www.ncbi.nlm.nih.gov/pubmed/18271711); Chamorro et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/14962621); Amaro et al., 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17525395)(http://stroke.ahajournals.org/cgi/reprint/38/7/2173.pdfhttp://stroke.ahajournals.org/cgi/reprint/38/7/2173.pdf); Romanos et al., 2007: (http://www.ncbi.nlm.nih.gov/pubmed/16596120); Teng et al., 2002: (http://www.ncbi.nlm.nih.gov/pubmed/12398932); Keller et al., 1998: (http://www.ncbi.nlm.nih.gov/pubmed/9425011)]. Uric acid and xanthine have both been shown to exert fairly strong feedback inhibition of XO activity [Rubbo et al., 1991: (http://www.ncbi.nlm.nih.gov/pubmed/1653611); Radi et al., 1992: (http://www.ncbi.nlm.nih.gov/pubmed/1322703)], and this feedback inhibition appears to be significant at normal, physiological concentrations in humans [Tan et al., 1993: (http://www.ncbi.nlm.nih.gov/pubmed/8134172)]. But the feedback inhibition has actually been shown to increase superoxide production by XO in vitro (by the isolated enzyme) (Rubbo et al., 1991; Radi et al., 1992). At the same time, Tan et al. (1993) found that 150 or 300 uM uric acid reduced the production of superoxide in human plasma overall by 23.2 and 32.0 percent, respectively. This indicates, in my opinion, that the overall effect of uric acid, despite its potential to increase superoxide production by XO, is to decrease superoxide formation. So elevations in uric acid levels, produced by exogenous purines or exogenous uric acid, during ischemia protect against ischemic damage, may increase free radical production by XO, but appear to decrease superoxide formation in the blood overall. This is consistent with the results of many other articles. The articles I've discussed provide more evidence, in my opinion, that uric acid per se and XO activity per se should not be viewed as being major factors, in the absence of exogenous uric acid or purine precursors of uric acid, contributing to oxidative damage following ischemia.

When one considers the nearly countless articles showing protection by uric acid against experimental autoimmune encephalomyelitis (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=encephalomyelitis+uric+OR+urate) and mitochondrial damage (http://scholar.google.com/scholar?q=mitochondrial+peroxynitrite+uric+OR+urate&hl=en&lr=), etc., these association studies, implying that uric acid is "independently" damaging, become bizarre to see, in my opinion. Hozawa et al. (2006) noted that elevations in serum urate are a risk factor for stroke but that uric acid itself "may" not cause strokes [Hozawa et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16239005)]. It's important to remember that ischemia itself (meaning ongoing, intermittent, low-level cerebral ischemia or ischemia in blood vessels outside the brain) increases purine nucleotide export and XO activity and uric acid production in a very reliable manner. It's worthwhile to remember that statistic-driven association studies are not a substitute for reasoning.

Interactions of Caffeine With Purine Metabolism, Ribose, and Uric Acid

The authors of this article [Herrick et al., 2009: (http://linkinghub.elsevier.com/retrieve/pii/S0306987709000061)] suggest that people could combine D-ribose with caffeine to augment the effect of caffeine and conceivably decrease the adverse effects associated with caffeine intake. The authors are essentially saying that ribose could augment ATP production and either decrease or increase the export of adenosine and its nucleotides from cells, meaning neurons, that have been stimulated by caffeine, etc. There's some validity to this suggestion, but, in my opinion, using purines or uridine as a source of small amounts of ribose would be a safer and more effective approach in the long term. Inosine monophosphate is ~43 percent ribose, and some similar percent of adenosine and guanosine are ribose. I don't feel like looking up the molar masses. In my opinion, high-dose ribose is not really a good idea, but I suppose one approach would be to combine small doses of ribose with purines and uridine or cytidine (uridine has been shown to elevated the cytidine and uridine nucleotide pools to significant extents, and so one doesn't need to take cytidine, really), etc. Barsotti and Ipata (2002) [Barsotti and Ipata, 2002: (http://www.ncbi.nlm.nih.gov/pubmed/11841784)] note that ribose has been shown to more effectively augment ATP repletion, following ischemia, when purines or purine bases are given along with the ribose (references 2 and 4, p. 130). Other articles have also shown that to be the case [Smolenski, 2000: (http://www.actabp.pl/pdf/4_2000/1171-1178s.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/11996106)].

One major reason I say that ribose from purines would, in my opinion, be safer is that ribose has been shown to increase the export of purine nucleotides and nucleosides from skeletal muscle cells and other cell types (including cells in the liver, etc.). The "purine-wasting" effect is not nearly as large as the effect of xylitol (I've cited much of this research in postings in December and early January) or fructose, but there is the potential for that to occur, in my opinion. That essentially means, in my view, that there would be short-term export of purines that would appear to be beneficial and long-term sequalae that would not be desirable. A similar effect occurs with something like zinc, which shows all these apparent "antidepressant" effects in short-term animal studies. Zinc increases the export of adenosine and other purines, partly by serving as a cofactor for a number of nucleotidase enzymes that degrade intracellular and extracellular nucleotides. That can produce short-term benefits, but the articles showing neurotoxicity from excessive zinc supplementation or from derangements in zinc homeostasis, in the absence of supplementation, are almost endless in numbers and are just absolutely appalling to see. I collected about a hundred of them in a list, and I'll try to link to a large number of them one of these days. They're not pleasant to look for or to read. That's an extreme example, in any case. Also, adenosine is exported from neurons in a generalized manner in response to neuronal activity, and that would be part of the rationale for providing actual exogenous purines instead of ribose. This is basically the same thing I've been saying over and over again, and there are almost innumerable articles showing that purine export is a generalized response to the excitation of neurons, either by electrical stimulation or pharmacological manipulations that increase excitatory neurotransmission. This effect, again, to the extent that it would be therapeutic, might not be long-lived in the absence of some attempt to address the resulting deficit in the intracellular purine pools. I came across a reference to an old article suggesting that antidepressants may exert some of their effects by increasing adenosine availability (this was discussed in the context of the capacity of low levels of adenosine to produce activation, rather than inhibition, of adenylate cyclase) [cited as reference 7 on p. 598 in Cooper et al., 1980: (http://www.ncbi.nlm.nih.gov/pubmed/6162091)]. I'm not sure if that cited article is talking about a reduction in the export of adenosine or about the export of adenosine from astrocytes leading to the import of adenosine into neurons. But I have multiple articles showing that either exogenous guanosine, adenosine, or inosine can elevate cAMP in various cell types, and I don't feel like linking to them right now. cAMP signaling is really complex, though, and an increase in the activities of cAMP-dependent protein kinases can be "pathological" or undesirable under some circumstances, and adenosine exerts a very complex set of effects on cAMP signaling.

Ribose mainly would contribute to the pool of intermediates in the nonoxidative pentose cycle, and this would mainly assist, to some extent, in the salvage of purines and in the provision of ATP by glycolysis, etc. (the activities of the enzymes of the de novo purine biosynthetic pathway are very low in the brain, and any increase in de novo inosine monophosphate formation from exogenous ribose alone would, in my opinion, be fairly minimal).

This suggestion about ribose is actually sort of minimally interesting in the context of the bizarre formulation of some of these well-known energy drinks advertised heavily on tv. I'm not going to say the brand, but I looked up the ingredients to see what was supposed to be so special about one of them. I see nothing very special about it and don't see the appeal of it. But an ingredient that stands out, in combination with caffeine, as being unusual is glucuronolactone. This is metabolized into glucuronic acid, and some of labeled glucuronolactone is converted into L-xylulose and then ribose [Hiatt, 1958: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1062823&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/13575548)]. So, in my opinion, glucuronolactone is like a third-rate substitute for ribose (the conversion of L-xylulose into xylulose-5-phosphate is ATP-consuming, and this ATP consumption and phosphate sequestration is basically the mechanism whereby xylulose produces more purine depletion than ribose, etc.). These types of differences in the point of entry into the pentose cycle have been shown, fairly clearly, in my opinion, to produce surprisingly significant effects on ATP levels and phosphate sequestration, and it's partly because of the large amounts of the sugar(s) entering cells at one time.

Another advantage of providing the actual purines, as a source of ribose, would be, in my opinion (apart from the peroxynitrite scavenging effect of uric acid), the capacity of uric acid (urate) to influence purine metabolism indirectly (also, oral inosine has been shown to elevate plasma hypoxanthine and xanthine in humans, and the contribution of those elevations in hypoxanthine to the effects associated with inosine should not be underestimated). Hunter et al. (1990) [Hunter et al., 1990: (http://www.ncbi.nlm.nih.gov/pubmed/2345757)] found that elevating uric acid in the blood of rats decreased the density of A1 adenosine receptors in the striatum, a site of action of caffeine. In contrast, caffeine upregulated the A1 adenosine receptor density. This basically shows that uric acid elevations decreased the tolerance to caffeine by downregulating the chronic caffeine-induced upregulation of A1 adenosine receptor density [caffeine is a nonselective adenosine receptor antagonist, but its acute blocking effect on (antagonism of) A1 adenosine receptors figures prominently into its stimulant effects]. Thus, in a person who has never taken caffeine, the antagonism is robust, but the receptor density becomes gradually upregulated in response to the presence of the antagonism by caffeine. Increases in extracellular or intracellular uric acid may oppose that, but Hunter et al. (1990) found that uric acid was not a direct A1 adenosine receptor antagonist at physiologically-relevant concentrations. One mechanism could be feedback inhibition of xanthine oxidase activity [cited as reference 16 in Kroll et al., 1992: (http://www.ncbi.nlm.nih.gov/pubmed/1539702)]. I've never heard anyone talk about that mechanism, and it could be really important for understanding purine metabolism. There's research showing that the Ki for the feedback inhibition of xanthine oxidase by uric acid (urate) may be as low as 200 uM, a physiologically relevant concentration. This could spare ATP, given that xanthine oxidase is an ATP-consuming reaction and consumes reducing equivalents. I saw that the articles showing feedback inhibition of xanthine oxidase by endogenousl purines and purine-based drugs haven't been cited very many times, but, if that effect occurs in humans, it would be a really important mechanism, in my opinion. An excess of intracellular urate can obviously be detrimental, and one would want to discuss any of this with one's doctor and have one's uric acid checked. There's evidence that urate can inhibit glycogen phosphorylase, and caffeine can also inhibit glycogen phosphorylase. That would be counterproductive to any supposed therapeutic effects, and the levels of plasma urate at which those undesirable effects might begin to occur are not well-known. But there's research showing, for example, that hyperuricemia induced by excessive inosine supplementation (in a study in athletes whose urate levels were already high-normal) can worsen exercise performance, and inhibition of glycogen phosphorylase could account for that.

Arginine, Uric Acid, and Peroxynitrite in Neurodegenerative and Psychiatric Conditions

This article [Xia et al., 1996: (http://www.pnas.org/cgi/reprint/93/13/6770.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/8692893)] is one of many showing that depletion of intracellular arginine tends to increase peroxynitrite formation by multiple mechanisms. When nitric oxide synthase (NOS) enzymes are not occupied by arginine or are competitively inhibited by asymmetric N(G),N(G)-dimethylarginine (ADMA) or by other methylarginines, which are inhibitors of NOS enzymes that are produced normally during the breakdown of proteins, the NOS enzymes produce superoxide and can also produce NO and superoxide at the same time. The NOS-derived NO and superoxide tend to react to form peroxynitrite. The depletion of cytosolic arginine by roughly half produced a fivefold increase in the sensitivities of the cells to a cytotoxic stimulus that increased nNOS activity. Similarly, Xia and Zweier (1997) [Xia and Zweier, 1997: (http://www.pnas.org/cgi/content/full/94/13/6954)] found that arginine depletion from activated macrophages produced large increases in peroxynitrite levels, and these increases were almost entirely blocked by either 1 mM extracellular arginine or 1 mM extracellular urate (the form that uric acid is in at physiological pH values).

This is relevant to the effects of arginine and purines in the brain. I think that maintaining an adequate urate level in the cerebrospinal fluid and also intracellularly, in neurons and astrocytes, is likely to be really important for maintaining cellular energy metabolism and also for maintaining the normal nitrergic regulation of noradrenergic and dopaminergic transmission, such as through the effects of nitric oxide on NMDA receptor activation. Roitman et al. (2007) [cited here: (http://hardcorephysiologyfun.blogspot.com/2009/03/arginine-agmatine-and-nitric-oxide-in.html)] noted some of the evidence that impairments in cellular energy metabolism can be found in people with depression and other psychiatric symptoms. Phosphocreatine (PCr) levels have been shown, in research using magnetic resonance spectroscopy techniques, to be drastically depleted in the brains of people with depression, for example (cited in Roitman et al., 2007). Mitochondrial dysfunction and ATP depletion would reasonably be expected to produce PCr depletion, even in the absence of a deficit in the formation of new creatine from arginine, etc. There's actually a large amount of research showing protection by uric acid/urate against mitochondrial damage due to peroxynitrite (inactivation of complex I and mitochondrial dysfunction) (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=mitochondrial+peroxynitrite+uric+OR+urate).

The reason I didn't see those articles in the past is that I think the authors of many articles showing protective effects of urate tend to not mention multiple sclerosis or Parkinson's disease, two neurodegenerative diseases in which inosine, a precursor of urate, is being tested (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=Parkinson%27s+uric+OR+urate); (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=%22multiple+sclerosis%22+uric+OR+urate)]. They would be much better off (and would, in my opinion, get substantially more robust therapeutic effects) using guanosine and adenosine monophosphates or triphosphates as precursors of urate instead of inosine, in my opinion, but that's beside the point. The research tends to be very focused in on one little area, and I've never seen those articles on mitochondrial protection, by urate, cited in the context of Parkinson's disease or MS. But the potential for the protection, by urate (or arginine), against the compromising of cellular energy metabolism, by peroxynitrite, would be very important, in my opinion, in the contexts of those and other disorders. West et al. (2002) [cited here: (http://hardcorephysiologyfun.blogspot.com/2009/03/arginine-agmatine-and-nitric-oxide-in.html)] cited research showing that peroxynitrite tends to decrease dopamine release (as in tonic, excitatory, nitrergically-mediated dopamine release), and that's obviously relevant to cognition and psychiatric conditions. I think the peroxynitrite-reducing potential of arginine would be more likely to be effective in combination with normalization of CSF and intracellular urate levels in the brain. If a person's plasma urate (blood uric acid) is already high, the person wouldn't need to do this. But the notion that any old urate level is as "good" as any other, within the normal range, is not defensible, given the overwhelming evidence, in my opinion, showing major effects across small increments in extracellular and, by extension, intracellular urate concentrations. I don't have time to go into the articles showing high intracellular urate levels, but the main idea is, in my opinion, that the use of urate as a peroxynitrite scavenger makes the use of most other antioxidants look like child's play. The concentrations of urate that scavenge peroxynitrite meaningfully, in vitro, are comparable to achievable and normal in vivo concentrations (this is not the case at all for many antioxidants). The concentrations of urate, both intracellularly and extracellularly, are much, much higher than the concentrations one is going to achieve with most antioxidants, in my opinion. Additionally, urate is less like an antioxidant scavenger of peroxynitrite than it is like a peroxynitrite "sink" and is, for the most part, excreted. It is not regnerated (doesn't need to be regenerated) by oxidoreductase enzymes but can actually be degraded, in a series of intramolecular degradative reactions (to allantoin or other molecules), upon its nitrosylation/nitration/etc. That's a really unique property that sets it apart from most other so-called "antioxidants" that consume reducing equivalents in their regeneration. I do think there's a lot of validity to the peroxynitrite-reducing effects of arginine in the context of brain disorders [Wiesinger, 2001: (http://www.ncbi.nlm.nih.gov/pubmed/11275358)], but I think that approach would work better in combination with normalization of urate levels. Obviously, one should discuss this type of thing with one's doctor.

Antagonism of the Spermidine-Induced Potentiation of NMDA Receptor Activation by Agmatine: Implications for the Actions of Agmatine and Arginine

This article is really good, and the authors discuss the discrepancies in past research on the antagonism (or negative allosteric modulation) of NMDA receptor (NMDA-R) activation by agmatine [Gibson et al., 2002: (http://www.ncbi.nlm.nih.gov/pubmed/12363406)]. I've seen authors of some articles make statements that agmatine doesn't bind to "the polyamine binding site" on NMDA receptors, but Gibson et al. (2002) point out the fact that a minimum of three polyamine binding sites are known to exist on different NMDA-R subunits. Gibson et al. (2002) found that agmatine displaced spermidine from binding to NMDA-Rs and antagonized the spermidine-mediated positive allosteric activation of MK-801-induced NMDA-R activation. I'm going to refer to this effect of spermidine as positive allosteric modulation, even though Gibson et al. (2002) discuss the inchoate understanding of the mechanism by which spermidine potentiates ligand-induced NMDA-R activation. The research Gibson et al. (2002) discuss, in relation to the precise mechanisms, is just crazy, and so I have to use *some* term to refer to the effect. But the research of Gibson et al. (2002) just shows that spermidine, a polyamine that can be produced from putrescine (and, by extension, agmatine and its precursor, arginine), potentiated the excitatory effects of NMDA-R activation (which would occur in vivo in response to the binding of glutamate, rather than MK-801). Agmatine produced essentially "mild" NMDA-R antagonism (by acting as an antagonist of the binding of spermidine to one of the polyamine binding sites on NMDA-Rs or to one of the "crazy-mystery-non-polyamine-binding-sites" that binds polyamines) at remarkably low concentrations (5 uM and above, with a Ki value for inhibition of spermidine-induced MK-801 potentiation of 14.8 uM). The Ki value is lower than the Ki value for irreversible inhibition of nNOS activity by agmatine aldehyde (29 uM). Gibson et al. (2002) note that researchers may have been looking for direct antagonism of NMDA-R ligands in past assays, and those different ligand-receptor binding assays produced a bizarre range of Ki values for the antagonism or negative modulation of NMDA-Rs by agmatine (12 uM to 1000 uM, or 1 mM).

The results of Gibson et al. (2002) suggest, in my opinion, that agmatine, derived from exogenous agmatine or arginine, could produce direct antagonistic effects (i.e. not mediated by inhibition of nNOS activity) on NMDA-R activation by glutamate, and those effects could help to account for the antidepressant (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=agmatine+antidepressant) and antihyperalgesic (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=agmatine+pain)/cognition-enhancing (http://scholar.google.com/scholar?q=agmatine+cognitive&hl=en&lr=) effects of agmatine in animal models. The newer NMDA-R antagonists, such as memantine, have been tested for the antidepressant effects that they can produce, but my sense is that they're more useful as strategies to augment or restore the effectiveness of other treatments. The authors of many articles discuss the fact that mild NMDA-R antagonism can sometimes enhance cognitive functioning and produce antidepressant effects, but more potent inhibition of NMDA-R activation can easily impair cognition and worsen depression, etc. This article by Gibson et al. (2002) helps explain the complex and conflicting dose-response relationships that researchers have found for arginine in animal models of depression or chronic stress or pain (http://hardcorephysiologyfun.blogspot.com/2009/03/arginine-agmatine-and-nitric-oxide-in.html), given that arginine is a precursor of glutamate (and, by extension, GABA), agmatine, putrescine, spermidine, and spermine. Each of those compounds can produce different effects on NMDA-R activation. I tend to think that lower doses of arginine would produce lower levels of agmatine and produce mainly inhibitory effects at NMDA-Rs, but, under conditions of inflammation or chronic stress, the excitatory effects of spermidine or spermine (or the supposed capacity of agmatine to induce glutamate release at high concentrations) might begin to predominate. The articles on polyamines tend to be really confusing for everyone, including the researchers writing them, seemingly. It's really a strange area of research, and polyamine metabolism is very dynamic and context-specific. There's a whole area of research on the effects of MAO inhibitors on polyamine metabolism, given that MAO inhibitors can inhibit the recycling of polyamines by producing inhibition of diamine oxidase, evidently (and the N-acetylpolyamines are substrates for MAO-B). But there's the initial effect and then there's the long-term response to the accumulation of N-acetylspermine and N-acetylspermidine, etc., etc. I don't need to say that MAO-B inhibitors have been used for cognitive enhancement and Parkinson's disease and depression and everything else under the sun, and some of those effects are thought to be due to the changes, produced by MAO-B inhibition, in the polyamine-dependent modulation of glutamatergic transmission [Youdim et al., 1993: (http://www.ncbi.nlm.nih.gov/pubmed/8302308); (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=%22MAO-B%22+polyamine)].

Conversion of Free-Form L-Tryptophan Into Norharmane, Tetrahydroharmane, Harmane, and Other Beta-Carboline Compounds in Vivo: Doesn't Look Good to Me

If I don't put this up, I'll forget about it. This article [Robinson et al., 2003: (http://www.ncbi.nlm.nih.gov/pubmed/15028582)] discusses the fact that L-tryptophan can be converted enzymatically into various beta-carboline compounds, including tetrahydroharmane and harmane, following the conversion of tryptophan into tryptamine. Fekkes et al. (2001) [Fekkes et al., 2001: (http://www.ncbi.nlm.nih.gov/pubmed/11323106)] found that as little as 1.2 grams of free-form tryptophan elevated plasma levels of norharmane in humans. When the articles say these beta-carbolines are formed enzymatically, this means that the compounds are formed in tiny amounts from dietary tryptophan but could be expected to be formed from supplemental tryptophan in larger amounts, in the absence of any contamination of the supplemental tryptophan. Half of that issue of the Annals of the New York Academy of Sciences (http://www3.interscience.wiley.com/journal/118876207/issue) is devoted to harmane and norharmane as endogenous "clonidine-displacing substances," (as imidazoline receptor agonists, etc.) and they can have anxiolytic effects. But they're obviously really potent compounds, in my opinion, because they're formed in tiny amounts and are thought to meaningfully regulate imidazoline receptor and alpha2-adrenoreceptor activation. Beta-carboline compounds aren't associated with a lot of desirable effects in the literature, and they tend to just be really potent (pM-range Kd values for binding to benzodiazepine binding sites on GABA receptors, etc.).

In my opinion, these are disturbing articles and help to explain the ongoing association of free-form tryptophan ingestion with eosinophilic disorders, etc. [some occurred before or after the contamination of a batch of tryptophan with a beta-carboline compound that caused the eosinophilia myalgia disease (http://scholar.google.com/scholar?num=100&hl=en&lr=&q=carboline+tryptophan+eosinophilia) that led to the banning of tryptophan from OTC sales] [Blauvelt et al., 1991: (http://www.ncbi.nlm.nih.gov/pubmed/1863073)]. In my opinion, it's pretty obvious that there's some problem with administering tryptophan in free form, because it's converted into beta-carboline compounds (which incidentally, along with kynurenine-pathway metabolites, can be really potent and might account for the growth hormone releasing effects of tryptophan). Also, kynurenine metabolites strongly regulate eosinophil function under normal circumstances [Odemuyiwa et al., 2004: (http://www.jimmunol.org/cgi/content/full/173/10/5909)(http://www.ncbi.nlm.nih.gov/pubmed/15528322?dopt=Abstract)] (in the absence of supplementation with tryptophan or any contaminant), and free form tryptophan ingestion has been associated, albeit in poorly-defined ways in any one individual, with a panoply of rheumatic disorders and other Th2-driven disorders (if a person's eosinophils are elevated in a blood panel, this suggests, most basically, some allergic disorder, but the abnormalities linked to tryptophan supplementation are much more complex). Researchers are arguing about which diseases are linked to it (diffuse fasciitis vs. scleroderma, etc.) (http://scholar.google.com/scholar?q=scleroderma+tryptophan&hl=en&lr=), but the point is that the endogenously-produced beta-carbolines and kynurenine metabolites, in my opinion, seem to be able to tip the scales of T-cell differentiation toward some of these pathological, Th2-cytokine-driven disorders, especially in people who might already have some susceptibility to those disorders. One can connect the dots, in my opinion. It's confusing and almost mind-bending to think about, but I don't see how one can explain away all of these separate facts in terms of one, isolated contamination.

I have no idea why free form tryptophan would cause more potent effects on eosinophils than tryptophan from food would, but foods don't contain all that much tryptophan. Additionally, free form tryptophan, like any free form amino acid, would be expected to provide some cell types, such as eosinophils, with much larger amounts of tryptophan than those cells would ever have access to in the face of increases in dietary tryptophan (i.e. tryptophan from foods). Food-derived tryptophan would be released very slowly, in comparison to free form tryptophan. Also, there's some unique aspect of the albumin-binding of tryptophan. I forget the details, but high-carbohydrate meals elevate plasma tryptophan more than other aromatic or other large neutral amino acids and supposedly also elevate free (unbound) tryptophan significantly (?). I forget the details, and this is just my opinion and sense of it (that an increase in free, unbound tryptophan in the blood could occur in response to supplementation and help to explain some of these supposed effects on the functions or Th2-differentiation-inducing capacities of eosinophils). I think it's just one of those things, one of those phenomena that's strange but that, in my opinion, could be associated with worrisome effects. I posted that information about free form tyrosine and phenylalanine and tryptophan, but there's, in my opinion, the most evidence suggesting that D-phenylalanine might have some therapeutic effects, but, even with that, there get to be all these issues with competition for uptake into the brain, etc.

[Incidentally, in a matter off-topic, D-phenylalanine is apparently not incorporated into proteins and not incorporated into aminoacyl-tRNAs (doesn't cause growth suppression in animals, like D-tyrosine). I couldn't find any article showing that it absolutely is never a substrate for aminoacyl-tRNA synthetases, though, so that would be the main concern I'd have with it. It looks like it probably isn't, but that's the type of thing one would want to talk with one's doctor about. I think that would have shown up, as it did with D-tyrosine administration, in animal experiments. D-phenylalanine substituted for L-phenylalanine dose-dependently, up to very high doses, and didn't produce the growth suppression that D-tyrosine did. That suggests that it probably isn't incorporated into proteins (and isn't normally a precursor of a D-phenylalanyl-containing aminoacyl-tRNA), and I found a couple of articles addressing the question directly. But apparently beta-phenylethylamine, derived in tiny amounts from D-phe, can be incorporated into proteins in tiny amounts (?). I forget the details, but that's the type of thing I would be concerned about and that one would want to keep in mind.]

Arginine, Agmatine, and Nitric Oxide in Psychiatric Conditions and Neuroprotection: Abbreviated Posting

This article shows that exogenous L-arginine exerted a biphasic effect on adult mice in the forced swim test, an animal model of depression [Ergun and Ergun, 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17125765)], such that low but not high doses of arginine produced antidepressant effects. The higher doses increased immobility (suggestive of a depressogenic or depressant effect), and the authors cite two other articles showing similar biphasic effects of arginine [Da Silva et al., 2000: (http://www.ncbi.nlm.nih.gov/pubmed/11117475); and Inan et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/15006455)]. The authors also found evidence that this biphasic response can be explained in terms of "depressant" effects of high levels of nitric oxide (NO) and antidepressant effects of smaller increases in NO formation. Arginine supplementation tends to increase the production of NO, a short-lived signalling molecule that plays a major role in maintaining blood vessel dilation and is released from nitrergic neurons that provide synaptic inputs to neurons throughout the brain, including glutamatergic and dopaminergic neurons in the striatum and noradrenergic neurons in the locus ceruleus [French et al., 2005: (http://www.med.upenn.edu/taylor/pubs/french%20neurosci%202005.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/16084659)]. Most of the nitrergic neurons in the striatum are nonspiny interneurons. Ergun and Ergun (2007) cited research [Ergun et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16940926/); Harkin et al., 1999: (http://www.ncbi.nlm.nih.gov/pubmed/10395013)] that the administration of different dosages of NOS inhibitors (which, under these conditions, are thought to be exerting their effects through the inhibition of neuronal NOS (nNOS) activity and not through the inhibition of inducible NOS (iNOS), given the absence of inflammation and short-term time frame, etc.) [West et al., 2002: (http://www.med.wayne.edu/neuroscience/labs/bird/pdfs/Tony-NO-rev.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/11984858)] also produces biphasic antidepressant effects in animal models.

Some of these effects of arginine and NOS inhibitors in animal models of depression are likely to be occurring in parts of the brain other than the striatum, such as in the locus ceruleus, etc. This is a really complicated area, and the effects of arginine in animal models are mainly acute or short-term effects and would not be expected to be the same in the context of chronic stress or depression or in other situations. And the same compound or physiological process that is beneficial to one person may be detrimental to another, and that's the reason it's always necessary for people to discuss these things with one's doctor. But I think there's enough research on arginine and agmatine and nitrergic transmission to get a general picture of the dose-response relationships and some of the predominant pharmacological effects that would be expected to occur in humans (under various conditions).

The general idea is that, under conditions that are not producing acute or chronic activation of the stress response pathways in the brain (i.e. increases in the firing rates of noradrenergic neurons in the locus ceruleus, etc.), exogenous arginine at "low doses" will, in my opinion, exert its short term effects by producing NO-dependent inhibition of the glutamatergically-mediated release of noradrenaline or dopamine or both and simultaneously decrease the firing rates of noradrenergic neurons. More specifically, I think the NO-mediated decrease in calcium influx in response to NMDA receptor activation will be the type of effect that will, at lower doses of arginine, predominate over the excitatory effects of NO. Some of the effects of the acute administration of "low" dosages of arginine may also be mediated by agmatine, which is produced in the brain in large amounts in response to arginine supplementation in animals and probably also would be, in my opinion, in humans. At higher dosages, I think arginine would tend to produce enhancements in glutamatergic transmission, mediated by both increases in nNOS-derived NO and increases in agmatine levels (agmatine may release glutamate and thereby produce excitatory effects at higher dosages [Halaris and Pleitz, 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17927294)]), that could well be counterproductive to the treatment of depression or chronic stress. Researchers have found that the dosages of intraperitoneal (i.p.) L-arginine, in mice, that produce "antidepressant-like" effects in animal models of depression scale to acute, single human dosages, for a 70-kg human (http://hardcorephysiologyfun.blogspot.com/2008/12/equations-for-animal-food-intake-and.html), of 3022-6045 mg (Da Silva et al., 2000), 6045-12090 mg (Inan et al., 2004), or 1209 mg (Ergun and Ergun, 2007). Inan et al. (2004) found that low dosages of i.p. arginine (dosages that scale to i.p. dosages of 302 or 1209 mg for a 70-kg human) produced "depressant" effects and blocked the antidepressant effects of potassium channel blockers or nNOS inhibitors, but Ergun and Ergun (2007) found that only low but not high doses of arginine were consistent with "antidepressant-like" effects. It's possible to make sense of these discrepancies, to some extent, but one can't look at data from those models in excessively-rigid terms.

The NMDA antagonistic effects of arginine-derived nitric oxide (and, more indirectly, of arginine-derived agmatine) would, in my opinion, have relevance to the augmentation of conventional antidepressants or to restoring the effectiveness of some psychopharmacological strategies, but there are a lot of details that are really complicated to get into. It's fairly clear, from the literature, that nitric oxide normally produces "tonic" (meaning under baseline conditions, in conditions other than animal models of chronic stress or depression, in which the firing rates of locus ceruleus neurons are going to be increased) excitatory influence on both noradrenaline release, by noradrenergic neurons in the locus ceruleus, and on dopamine release in the striatum. For example, arginine can increase dopamine release in the striatum by enhancing nNOS-derived NO [Liang and Kaufman, 1998: (http://www.ncbi.nlm.nih.gov/pubmed/9685635)] and can also increase the firing rates of noradrenergic neurons in the locus ceruleus, evidently by nNOS-independent glutamatergic effects (this might be due to agmatine) [Torrecilla et al., 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17473915)]. Liang and Kaufman cite four other articles [Hirsch et al., 1993: (http://www.ncbi.nlm.nih.gov/pubmed/15335838); Lonart et al., 1992: (http://www.ncbi.nlm.nih.gov/pubmed/1425999); Strasser et al., 1994: (http://www.ncbi.nlm.nih.gov/pubmed/7533554) Zhu and Luo, 1992: (http://www.ncbi.nlm.nih.gov/pubmed/1494918)] showing the same nitrergically-mediated dopamine release in response to L-arginine, and I'm sure there are other articles showing that. That's probably partly a glutamatergic effect that's mediated by the nitric oxide-induced activation of glutamate release or inhibition of chloride influx in response to GABA-A-receptor activation, etc. But under conditions of chronic increases in the firing rates of noradrenergic neurons or pathological increases in glutamatergic transmission, more agmatine is formed from arginine and may help to limit the firing rates of noradrenergic neurons and thereby produce an "anti-stress" effect (this stress-induced increase in agmatine formation has been shown in animal experiments, in which acute stress roughly triples the tissue agmatine contents in multiple parts of the brain). In the striatum, NMDA receptor antagonism, such as by nitric oxide, can actually sensitize striatal neurons to D1 dopamine receptor activation via dopaminergic inputs from the ventral tegmental area. This can produce beneficial effects on working memory, to a point, but can then, at higher degrees of NMDA receptor antagonism, impair working memory by causing a breakdown of the organized, burst firing patterns of glutamatergic pyramidal neurons in the prefrontal cortex (in association with working memory impairment). So basal dopaminergic activity and noradrenergic activities appear to be dependent on an adequate level of NO-mediated glutamatergic activity, but the stress-induced activation of noradrenergic neurons, among other effects, could, in my opinion, narrow or even "abolish" the supposed therapeutic dosage range for something like arginine, in this context.

Incidentally, the acute oral dosages that would produce comparable effects in the brain (comparable to the i.p. dosages) would be expected to be higher than the i.p. dosages, and the increases in plasma arginine [Bode-Boger et al., 1998: (http://www.ncbi.nlm.nih.gov/pubmed/9833603)(http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1873701&blobtype=pdf)], following daily supplementation, can take 4-8 weeks to reach a steady state in humans [Campbell et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16928472)]. Bode-Boger et al. (1998) found that the absolute bioavailability of oral arginine in humans, in relation to i.v. arginine, was ~68 percent, and the ratio of the AUC(i.v.)/AUC(oral) was ~1.542 for dosages of 6 grams of arginine (by i.v. and oral routes). I think the ratio of the Cmax(i.v.)/Cmax(oral) (= 2.652 for arginine at dosages of 6 grams, given by i.v. and oral routes) is likely to be a more important determinant of the effects of arginine on the brain than the ratio of the AUC values, because the concentration of agmatine in the CSF (and, by extension, in the CNS intraparenchymal interstitial fluid) increases and decreases rapidly. This suggests to me that the rate of entry of arginine into the brain may importantly determine the metabolic fate of arginine. More specifically, the arginine-induced production of agmatine in the brain would be expected to be higher in response to arginine taken in the fasted state, meaning before breakfast in the morning, than in response to arginine taken between meals. The amounts of agmatine formed in the brain from exogenous arginine appear to be quite significant in primates and mice [Piletz et al., 2003: (http://www.ncbi.nlm.nih.gov/pubmed/15028571)], and I'll include my analysis of that type of data, together with my estimate of the short-term extracellular and intracellular concentrations of agmatine in response to scaled doses of arginine, in another posting. Suffice it to say that a single dose of intraperitoneally-administered arginine in a monkey (this scales to a human dose of 6176 mg arginine, given intraperitoneally) produced a peak level of 2200 nM (2.2 uM) agmatine in the CSF (up from baseline values of 46.9-181 nM). The Kd values for the binding of agmatine to I1 and I2 imidazoline receptors (IRs), producing competitive inhibition, are 700 nM (0.7 uM) and 1000 nM (1 uM), and the Kd for the binding of agmatine to alpha2-adrenoreceptors (alpha2-ARs), as an agonist, is ~4000 nM (4 uM). Clonidine produces activation of both I1-IRs and alpha2-ARs, and agmatine, to a meaningful extent, essentially produces clonidine-like effects. But the point is that the more robust activation of I1-IRs by agmatine, at extracellular agmatine levels less than the range of the 4 uM Kd value for strong alpha2-AR activation by agmatine, could reasonably be expected to augment the effect of low-level alpha2-AR activation by the arginine-induced increases in agmatine. I haven't seen much compelling evidence that meaningful antagonism of NMDA receptors occurs at extracellular concentrations of agmatine that occur under normal or therapeutic circumstances, in response to either arginine or agmatine administration. The Ki value is really high for NMDA receptor antagonism by agmatine, and the authors of one article, showing evidence of NMDA receptor antagonism in response to intrathecal agmatine (or i.c.v.--I can't remember right now and don't want to look it up), found that the effect only lasted between 10 and 30 minutes. The extracellular concentration was very high. One key point, though, is that i.v. arginine increased the CSF agmatine to a peak level that was 22 times as high as the increase in plasma agmatine. Thus, these articles that measure plasma agmatine in response to arginine are not going to be detecting the agmatine formed en masse in the CSF. I think arginine-induced agmatine participates in the arginine-induced growth hormone release, also. Additionally, I think the agmatine levels would accumulate over time intracellularly. It's rapidly transported into neurons by a polyamine transporter that transports spermine and putrescine (and presumably spermidine), and the results of the article by Piletz et al. (2003) (cited above) suggest that most of the tissue agmatine levels, at three hours post-i.p. injection, will be intracellular and not extracellular. I can't extrapolate the intracellular concentration that accompanied the large peak in CSF agmatine that occurred in the primates, but, under steady-state conditions, it's pretty clear that about 95.6 percent of a measurement of the tissue agmatine levels, in ng/g wet weight, is intracellular (with the rest being extracellular). I'll put the simple calculations up on another posting. (You can estimate this if you know, roughly, both the steady-state CSF concentration and the tissue concentration, in ng/g ww.) Finally, in the longer term, there's evidence to suggest that the inhibitory effects of arginine-derived agmatine on nNOS activity may become more significant and produce meaningful effects, by glutamatergic or other mechanisms, on noradrenergic or dopaminergic transmission. The Ki value for irreversible (noncompetitive) inhibition of nNOS by agmatine (as agmatine aldehyde) is 29 uM (Piletz et al., 2003, cited above) (the Ki value for competitive inhibition by agmatine is much higher), and the supposed accumulation of stored, intracellular agmatine could conceivably produce some gradual inhibition of nNOS activity over time. That could produce NMDA receptor antagonism, by reducing nNOS-derived NO, but the direct antagonism of NMDA receptors by agmatine seems to be an effect that, in my opinion, is unlikely to become meaningful.

In any case, this is too complicated a topic to discuss all at once. Arginine has also been shown to augment creatine formation in the brain in humans, in two articles, and in animals, and low-dose creatine (3-5 grams/d) was shown to be beneficial, in a small study, in augmenting conventional antidepressant medications in people with treatment-resistant depression [Roitman et al., 2007: (http://www.ncbi.nlm.nih.gov/pubmed/17988366)]. I think arginine would be safer for that effect [discussed here: (http://hardcorephysiologyfun.blogspot.com/2009/02/glutamate-glutamine-cycle-de-novo.html)], but one approach could be to use extremely small doses of creatine, such as 1-2 grams/d. Supposedly, diets high in red meat provide 1-2 grams of creatine/d, but I think some of those estimates of dietary creatine overestimate the amounts that people would be getting. It's interesting that Roitman et al. (2007) found that a couple of the people lost the mood-elevating effect from creatine as they increased the dosage from 3 to 5 grams per day. When they went back to the 3 g/d dosage, they benefited from it again. I think that's plausible, that there would be a dose-response relationship across very small dosage increments. The main problem, in my opinion, is not with the short-term effects of creatine but with the fact that the long-term effects can be very complicated and can become counterproductive, even in the extreme, in my opinion. The therapeutic window for creatine dosing, for psychiatric or cognitive effects [Roitman et al. (2007) cite research showing cognitive enhancement from something like 5 grams/d of creatine in normal people, and there's a vast amount of research on its effects on the brain], is probably very narrow, and the therapeutic dosages could be, in my opinion, as low as 1-3 grams/d. Also, there's the issue of the conversion of ornithine or agmatine, both derived from arginine, into putrescine and the other polyamines, and putrescine has been shown to produce antidepressant effects in animal models. The antidepressant effects that S-adenosylmethionine (SAM-e) has been shown to exert in animal models may partly be due to the SAM-e-induced elevations in the levels of putrescine and other polyamines in the brain, but, in my opinion, the effects of SAM-e have more to do with increases in the adenosine nucleotide pools in neurons and astrocytes and could be mimicked by exogenous adenosine (and guanosine). Decarboxylated SAM-e is a cofactor for spermine synthase, which converts spermidine into spermine, and spermidine synthase, an enzyme that converts putrescine into spermidine. Putrescine can be formed by the catabolism of agmatine by agmatinase or by the pyridoxal 5'-phosphate (PLP)-dependent enzyme ornithine decarboxylase. The authors of one article I discussed [Geng et al., 1995: (http://hardcorephysiologyfun.blogspot.com/2009/01/mechanisms-of-neuroprotection-by.html)] found evidence that the neuroprotective effects of vitamin B6 (pyridoxine), in cultured cells, were partially mediated by the PLP-dependent increase in ornithine decarboxylase activity (via the formation of putrescine and the other polyamines, which then can either antagonize or produce positive allosteric activation of NMDA receptors), given that ifenprodil partially blocked the PLP-mediated neuroprotection. The same modulation of NMDA receptor activation could account for the effects of changes in polyamine levels on cognitive functioning or psychiatric conditions, but research on polyamines seems to be rather chaotic and confusing for just about everyone who writes about it.

Arginine could be expected to increase both agmatine and ornithine and to increase putrescine formation from either agmatine or ornithine, whereas ornithine may not increase agmatine nearly as much and would tend to inhibit creatine formation by arginine:glycine amidinotransferase (AGAT) (the first enzyme in creatine biosynthesis that is expressed throughout the brain). There's also the issue of peroxynitrite formation by the reaction of excessive NO with superoxide or by the uncoupling of nNOS by asymmetric N(guanidino),N(guanidino)-dimethylarginine (ADMA) or N(omega)-monomethylarginine (endogenously-produced inhibitors of NOS enzymes), etc., and maintaining a high-normal CSF uric acid level, with low-dose purines, could, in my opinion, help to limit peroxynitrite formation and downregulate iNOS expression. I can't link to 100 articles in one posting, though.