Adenosine and Guanosine in Animal Models of Depression

Several articles show that adenosine [Kaster et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/14729225); Kaster et al., 2005a: (http://www.ncbi.nlm.nih.gov/pubmed/16140163); Kaster et al., 2005b: (http://www.ncbi.nlm.nih.gov/pubmed/16202183); Kaster et al., 2007a: (http://www.ncbi.nlm.nih.gov/pubmed/17868670); Kaster et al., 2007b: (http://www.ncbi.nlm.nih.gov/pubmed/17296254)] or guanosine [Eckeli et al., 2000: (http://www.ncbi.nlm.nih.gov/pubmed/10884029)] can produce antidepressant ("antidepressant-like") effects in animal models of depression. It's important to note that guanosine and adenosine have not been proven to be effective or safe in the treatment of depression or any other disease, and one should always talk to his or her doctor before taking any of these supplements. It's also important to take the results from these animal studies with a grain of salt, but the results of a lot of these experiments are actually generally-consistent with the results of other research in humans. My sense is that guanosine would display more of a mild anticonvulsant effect than adenosine would, mainly because of the large amount of research showing anticonvulsant effects of oral or intraperitoneal guanosine in animals [(http://scholar.google.com/scholar?num=100&hl=en&lr=&safe=off&cites=13842524336757775794); (http://hardcorephysiologyfun.blogspot.com/2009/01/anticonvulsant-effects-of-oral.html)]. The apparent effectiveness of SAM-e in the treatment of depression (http://scholar.google.com/scholar?num=100&hl=en&lr=&safe=off&q=antidepressant+%22S-adenosylmethionine%22+OR+%22S-adenosyl-L-methionine%22), in spite of the relatively poor biovailability that I see SAM-e as having, is, in my opinion, an important line of evidence suggesting that adenosine could, in fact, display antidepressant effects and serve some kind(s) of adjunctive role(s) in that regard. SAM-e is rapidly metabolized into adenosine, and, as discussed below, some of the effects of SAM-e on neurons and other cell types can be blocked by adenosine receptor antagonists or mimicked by adenosine itself. Although adenosine is generally viewed as exerting primarily an inhibitory influence on catecholaminergic and glutamatergic transmission, S-adenosylmethionine (SAM-e), which is generally known to increase the pool of adenosine and its nucleotides in the cells it reaches [Smolenski, 2000: (http://www.actabp.pl/pdf/4_2000/1171-1178s.pdf)(http://www.ncbi.nlm.nih.gov/pubmed/11996106)], can nonetheless induce mania or hypomania in some people (an effect that is not consistent with an anticonvulsant effect) (http://scholar.google.com/scholar?q=S-adenosylmethionine+mania+OR+hypomania&hl=en&lr=). Thus, SAM-e can exert excitatory effects under some conditions, such as in the presence of medications that affect adenosine release or adenosine receptor sensitivities in the brain (many drugs used in psychopharmacology). Consistent with this, Saletu et al. (2002) [Saletu et al., 2002: (http://www.ncbi.nlm.nih.gov/pubmed/12486491)] found that the changes in the electroencephalograms of people who had been receiving parenteral SAM-e for 7 days indicated that SAM-e had been producing a mixture of both excitatory and inhibitory effects on overall brain activity. This article is a bit confusing at times but discusses some of the complexities of "adenosinergic" signalling in the context of dopaminergic transmission and psychiatric conditions [Cunha et al., 2008: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=2423946&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/18537674)]. What they're saying is that adenosine itself can exert excitatory effects on striatal dopamine release by activating (apparently-predominantly-presynaptic) "extrastriatal" A2A receptors on glutamatergic neurons, even in the face of the inhibition of dopamine release by the adenosine-induced, tonic activation of A1 receptors on neurons in the striatum (on dopaminergic terminals or interneurons, etc.). Adenosine could, for example, modulate the release of glutamate from glutamatergic, pyramidal neurons that project from the dorsolateral prefrontal cortex and provide excitatory inputs to dopaminergic neurons in the ventral tegmental area (VTA), thereby providing an excitatory influence on dopamine release in the striatum (given that many dopaminergic neurons in the VTA project to the ventral striatum, etc.) [Carr and Sesack, 2000: (http://www.jneurosci.org/cgi/reprint/20/10/3864)(http://www.ncbi.nlm.nih.gov/pubmed/10804226?dopt=Abstract)].

That's a fairly well-established neuronal pathway that's important for cognitive functioning and mood, but the research on the effects of adenosine (and guanosine) in the brain is very vast and extremely complicated. That's one reason I tend to focus more on in vivo studies in animals and to not focus too heavily on articles that discuss, for example, the intracellular signalling cascades that guanosine and adenosine activate, via the binding of GTP to G-proteins or the binding of adenosine to adenosine receptors or intracellular binding sites, etc. But given the importance of dopaminergic transmission to cognitive functioning and mood disorders, it is noteworthy that, as discussed by Carr and Sesack (2000), glutamatergic inputs to striatal dopaminergic neurons are crucially important for the burst firing patterns of those dopaminergic neurons to be maintained normally. Too much or too little dopamine release in the prefrontal cortex, from dopaminergic neurons that project to the prefrontal cortex from different sites in the basal ganglia (including the striatum), is detrimental to working memory and therefore to cognitive functioning, and the firing patterns of glutamatergic neurons in the prefrontal cortex and other "extrastriatal" sites, as would be regulated by, for example, the degree of activation of A2A receptors outside the striatum by adenosine, are a major factor that regulates dopamine release in the prefrontal cortex. In some articles, a discussion of "a receptor in the striatum" can mean that the receptor is on an axon terminal, in the striatum, of a dopaminergic neuron whose cell body is not in the striatum but is in, for example, the midbrain, in the VTA. This type of discussion can just become absurdly complicated and can just get crazy, especially given the concentration-dependence of the effects of adenosine, etc. Cunha et al. (2008) discussed the fact that A2A receptors seem to be activated more by adenosine that is derived from extracellular ATP hydrolysis than adenosine that is derived from other sources (such as by the export from the cytosol, via equilibrative adenosine transporters, etc.). That's just one example of an obscure "mechanistic" explanation for the complexity of the adenosinergic or purinergic modulation of neuronal activity. In any case, those excitatory effects that adenosine, such as can be derived from SAM-e or oral adenosine monophosphate or triphosphate, can sometimes have on some glutamatergic neurons are one mechanism that could account for the acute increases in dopamine and noradrenaline release that have been shown to occur in animals in response to SAM-e administration [cited in Benelli et al., 1999: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1566059&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/10401554)]. In a related vein, Eckeli et al. (2000), cited above, found that the apparent antidepressant effects of guanosine monophosphate in animals was partially dependent on serotonin biosynthesis or release. The antidepressant-like effect of guanosine monophosphate was mimicked by an NMDA receptor antagonist or fluoxetine, a serotonin reuptake inhibitor and prescription antidepressant. In that case, the general idea is essentially that a stressor is producing excessive glutamatergic activation and that the glutamatergic activity is indirectly disturbing the feedback inhibition of the firing rates, which increase in response to acute stress, of noradrenergic neurons in the locus ceruleus. Guanosine or NMDA receptor antagonism may attenuate this excessive glutamatergic activity, as implied by the results of Eckeli et al. (2000). There are more than a dozen articles showing that oral or intraperitoneal guanosine can exert anxiolytic (anti-anxiety) or anticonvulsant or neuroprotective effects that are at least partially a result of the attenuation of glutamate release or augmentation of synaptic glutamate reuptake by guanosine (http://scholar.google.com/scholar?num=100&hl=en&lr=&safe=off&cites=13842524336757775794). This seems confusing, but under conditions of chronic stress or neuronal injury, for example, decreasing glutamatergic transmission (such as by NMDA receptor antagonism) can actually increase dopamine release or, more precisely, increase the responsiveness of (postsynaptic) neurons receiving dopaminergic inputs to, for example, D1 dopamine receptor activation in response to dopamine release [see, for example, Peeters et al., 2002: (http://www.ncbi.nlm.nih.gov/pubmed/12213297); Deep et al., 1999: (http://www.ncbi.nlm.nih.gov/pubmed/10529725); Arai et al., 2003: (http://www.ncbi.nlm.nih.gov/pubmed/12711097); Konradi et al., 1996: (http://www.jneurosci.org/cgi/reprint/16/13/4231)(http://www.ncbi.nlm.nih.gov/pubmed/8753884?dopt=Abstract); Tokuyama et al., 2001: (http://www.ncbi.nlm.nih.gov/pubmed/11408088); Boyce-Rustay et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16482087)]. In the absence of stress (and even in the "presence of stress," up to a point), however, decreasing glutamatergic transmission (let's say to the same "degree" as in the presence of stress and at the same dosages of the same drug, for example) will tend, with some degree of predictability, to decrease dopamine release (many more references on this). In the former case, mild NMDA receptor antagonism, such as by amantadine (which also exerts numerous other effects that confound the discussion, but so be it), produces an excitatory effect on normal, burst-firing-associated dopaminergic transmission, but NMDA receptor antagonism in the latter case--antagonism that is stronger or that occurs in the absence of some factor that is producing an "abberant" or excessive increase in the firing rates of the neurons in question--will tend to inhibit dopaminergic (or noradrenergic) transmission. Here's an article that highlights some of the complexity with which adenosine can modulate the release of dopamine in response to NMDA receptor (a glutamate receptor) activation in the striatum [Quarta et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/15525341)].

Again, my opinion is that many of the effects of SAM-e on the brain are mediated by adenosine, which SAM-e is metabolized into, and by the nucleotides derived from adenosine and its metabolites (including hypoxanthine and inosine and also some guanosine formed from inosine monophosphate, derived from adenosine nucleotides, etc.). I've discussed some of the evidence for this in past postings, but, for example, Renshaw et al. (2001) suggested that SAM-e may exert antidepressant effects, in part, via its metabolism into adenosine [cited here: (http://hardcorephysiologyfun.blogspot.com/2009/01/details-on-nucleotides-bioavailability.html)]. Additionally, the effects of exogenous adenosine or S-adenosylmethionine or inosine monophosphate, in combination with guanosine and other nucleotides, can produce increases in the phosphocreatine (PCr) to creatine (Cr) ratio (PCr/Cr ratio) that are reminiscent of the effect of exogenous SAM-e on the PCr/NTP ratio [see Silveri et al. (2003) and other articles cited and discussed here: (http://hardcorephysiologyfun.blogspot.com/2009/03/creatine-cr-phosphocreatine-pcr-and.html)]. In reference to the article by Silveri et al. (2003), it is noteworthy that NTP refers to the pool of beta-nucleotide triphosphates, and this "quantity" is thought to primarily reflect the pool of ATP that is measured, in a given part of the brain, by magnetic resonance spectroscopy (MRS). Also, some of the antiinflammatory effects of SAM-e can be antagonized by adenosine receptors and mimicked by the administration of adenosine, as a substitute for SAM-e [Song et al., 2005: (http://www.ncbi.nlm.nih.gov/pubmed/15843034)]. Song et al. (2005) also found that SAM-e elevated the pool of intracellular adenosine in cultured monocytes. Travagli et al. (1994) [Travagli et al., 1994: (http://www.ncbi.nlm.nih.gov/pubmed/7698179)] found that SAM-e reduced the firing rates of neurons in the vagal motor nucleus, and these effects were not only abolished by adenosine receptor antagonists but were shown to be "reversed" in the presence of those antagonists. SAM-e increased the firing rates of those neurons in the presence of the adenosine receptor antagonists, and Travagli et al. (1994) noted that increases in the availability of adenosine, produced from SAM-e via the hydrolysis of S-adenosylhomocysteine (SAH), had probably caused the changes in the neuronal firing rates. The attenuation of the SAM-e-induced increases in the firing rates of the neurons by SAH, in the presence of the adenosine receptor antagonists, might be explained by the inhibition of methyltransferases by SAH. Inhibiting methyltransferase enzymes in the presence of an excess of SAM-e would prevent much of the conversion of the excess SAM-e into SAH and then into adenosine. Alternatively, the excess SAH may itself have been converted into adenosine and thereby produced effects on adenosine receptors or purinergic receptors that were, as a result of the higher concentration of adenosine, in opposition to the effects that had been produced by the exogenous-SAM-e-derived adenosine (i.e. by that lower concentration). In either case, the results could conceivably be explained in terms of the effects of nucleotides or nucleosides derived from adenosine (i.e. inosine, hypoxanthine, xanthine, uric acid, or guanosine, etc.) or the effects of different concentrations of extracellular adenosine, derived from exogenous SAM-e or SAH, on different adenosine receptor subtypes or on other intracellular or extracellular adenosine binding sites on other proteins. There are lots of other articles that show similarly-comparable effects of adenosine and either SAM-e or 5'-methylthioadenosine (MTA), which is converted into adenosine [these are two that I found without any effort, but there are many others: Song et al., 2004: (http://www.ncbi.nlm.nih.gov/pubmed/15566950); Song et al., 2003: (http://www.ncbi.nlm.nih.gov/pubmed/12736147)]. I view the adenosine-receptor dependence of the anti-inflammatory effects of SAM-e, along with the elevation of intracellular adenosine by SAM-e (Song et al., 2005, cited above; Smolenski, 2000, cited above), as evidence that the adenosine nucleotide pool is being elevated and equilibrating with the extracellular pool, thereby causing increases in the activation of extracellular adenosine receptors. There are other interpretations, but that's just my opinion.

As far as the dosage considerations go, a number of articles have shown that adenosine and guanosine can exert effects in animal models at remarkably low dosages. Guanosine has been shown to exert neuroprotective and anticonvulsant effects at 7.5 mg/kg bw per day, given orally or intraperitoneally, in rats. That works out to something like ~111 mg per day, scaled to a dosage for a 70-kg human, and that's an extremely low dosage. I would think there might be some kind of issue with that dosage or the scaling factor (4.71) that I used, but essentially the same dosage (8 mg/kg bw/day, given intraperitoneally), was shown to promote restoration of functioning in animals following experimental spinal cord injuries [Jiang et al., 2008: (http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=2072916&blobtype=pdf)(http://www.ncbi.nlm.nih.gov/pubmed/18404454)]. If one were to not scale the dosage and use 8 mg/kg bw per day in a 70-kg human, the dosage would be 560 mg/d. Even under those circumstances, that's a very low dosage, as purines go. That suggests that guanosine is fairly potent, particularly in comparison to inosine. The dosages of inosine used to promote recovery after spinal cord injuries, in animal models, are much higher. For example, Liu et al. (2006) [Liu et al., 2006: (http://www.ncbi.nlm.nih.gov/pubmed/16317421)] used inosine to prevent the ongoing degneration that occurs after a spinal cord injury in adult rats, and the dosage used was 75 mg/kg bw, given intraperitoneally, every eight hours. That's 225 mg/kg bw per day in rats. The fact that Jiang et al. (2008) found therapeutic effects of guanosine at 8 mg/kg bw/day is remarkable and shows, along with the many articles showing anticonvulsant and neuroprotective effects of oral or i.p. guanosine or guanosine monophosphate at 7.5 mg/kg bw/day, that, in my opinion, guanosine would be expected to produce meaningful effects on the brain at doses that would be compatible with human physiology. I say that because the dosage of inosine of 225 mg/kg bw/day (or the 100 mg/kg bw/day used in many other animal experiments showing anti-inflammatory effects of inosine, for example) scales to 3344 mg/day of inosine for a 70-kg human. If one looks at the trials using inosine to elevate uric acid in people with multiple sclerosis, one sees that that dose of inosine could produce hyperuricemia in many humans. Using that dose in the short term, such as after a spinal cord injury, would not be expected to produce hyperuricemia, and researchers would get larger effects on the brain by administering inosine intravenously or intraperitoneally than they would by administering it orally. But guanosine appears to exert many of its effects on the brain at remarkably low dosages.

One can, in my opinion, get a sense of the possible "dosage ranges" for guanosine monophosphate and adenosine monophosphate by looking at the dosages of S-adenosylmethionine and inosine that have been used in trials in humans. Guanosine monophosphate is available, evidently only in combination with other nucleotides, from a limited number of manufacturers [for example: (http://www.google.com/products?q=bluebonnet+nucleotide+complex+&hl=en)], but I'm not sure how much guanosine monophosphate is in that product. Again, I don't have any financial interest in any of these products or in anything I've discussed on this blog. The combinations of nucleotides that are used in research (http://scholar.google.com/scholar?q=dietary+nucleotide&hl=en&lr=) typically contain guanosine monophosphate and the other nucleoside monophosphates (a nucleotide, for the purposes of this discussion, is a nucleoside that's been phosphorylated in the 5'-position) in a either roughly 1:1:1:1 mass ratio or molar ratio. If the nucleotides and nucleosides were in a 1:1:1:1 mass ratio, each of those capsules would contain ~50 mg guanosine monophosphate and 10 mg guanosine and 15 mg guanine (and those same amounts for adenosine monophosphate, adenosine, and adenine). But I don't know if that's correct. One could call the manufacturer, and I'm sure they'd probably tell you what the composition is. But, from the standpoint of uric acid production, the elevation in uric acid is going to be different for every person. That's one reason it's necessary to have one's uric acid monitored by one's doctor, if one's going to take the full range of dosages for oral adenosine or guanosine monophosphates. One way to determine a dosage, with one's doctor, in the context of antidepressant medication, would be to look at the "percent adenosine" in S-adenosylmethionine and the dosage range of inosine used to treat multiple sclerosis. Less uric acid is made from adenosine than from inosine, in general, because adenosine is more efficiently salvaged. The ratio is something like 1.3/2.5, meaning the ratio of the AUC or Cmax (I forget which) for the serum uric acid response to oral adenosine to the uric acid response to oral inosine is about 0.52 or 0.55 or something. SAM-e is ~67.1 percent adenosine (the ratio of the molar mass of adenosine to the molar mass of SAM-e is 267.241/398.44), and the maximum dosage of SAM-e that I've seen used to treat depression is 3600 mg/day [Di Rocco et al., 2000: (http://www.ncbi.nlm.nih.gov/pubmed/11104210)]. If one takes 67.1 percent of that dosage, one gets 2415 mg/day of adenosine (or adenosine + guanosine). That works out to about 16 of those nucleotide capsules per day, assuming about half of the 300 mg is adenosine + guanosine (or purine bases, which are nucleic acids that can would probably be converted into uric acid). I'm not suggesting anyone would want to take any of these sorts of full dosage ranges without talking to one's doctor, because elevations in uric acid are not the only potential side effect. It's conceivable that adenosine or guanosine could produce vasodilation or reduce blood pressure and interact with sedatives or blood pressure medications or any number of medications and produce serious side effects. It's actually the case that adenosine monophosphate and guanosine monophosphate are 70-some percent adenosine and guanosine (the rest is phosphate), and that would affect the "calculation," crude as it is (that would mean the dose would be a little bit higher to be equivalent to the 2415 mg of adenosine that can be derived from 3,600 mg of SAM-e). Adenosine triphosphate disodium is more widely available than guanosine [(http://hardcorephysiologyfun.blogspot.com/2009/02/brief-note-on-purines.html); (http://hardcorephysiologyfun.blogspot.com/2009/01/details-on-nucleotides-bioavailability.html)].

One way to minimize the production of uric acid from adenosine monophosphate and guanosine monophosphate would be to dissolve them in water and take them on an empty stomach (which means before breakfast, not between meals), but I'm not sure if that would meaningfully increase the bioavailabilities of those nucleotides. There's increasingly been a recognition that the bioavailability of physiological substrates can depend on seemingly-trivial factors, such as the rate of dissolution, and this has been discussed in the context of the bioavailability of creatine. When researchers [Deldicque et al., 2008: (http://www.ncbi.nlm.nih.gov/pubmed/17851680)] gave 2 grams of creatine monohydrate dissolved in aqueous solution (water) or in a protein bar or in some kind of complex with "beta-glucan" (I have no idea what the rationale for testing that is), the Cmax for creatine in water was 299 uM (up from a baseline of 10 uM), and that was higher than the Cmax for creatine from the protein bar (237 uM) or the beta-glucan source (174 uM). I've seen multiple articles in which the researchers have dissolved creatine monohydrate in water, ostensibly with the aim of improving the availability of creatine for entry into the brain. I don't know if that type of factor would really matter all that much, and it might be cumbersome to do. But the same concept has been discussed in the context of orally-administered nucleotides. The activity of xanthine oxidase ("xanthine oxidoreductase") is thought to be lower in the fasting state, and this means that less of a dose of adenosine or guanosine may be converted, following the conversion into hypoxanthine or xanthine, into uric acid, in the fasting state, in the short term [see references 75-77, cited on p. 63 of Carver and Walker, 1995: (http://cat.inist.fr/?aModele=afficheN&cpsidt=3477199)]. Note that that paper by Carver and Walker (1995) is not especially up-to-date, as far as information on nucleotide bioavailability is concerned.

I'll put the discussion of mechanisms in a separate posting. This is too long to put in one posting.