<<

Therapeutic intervention in experimental allergic encephalomyelitis by administration of uric precursors

Gwen S. Scott, Sergei V. Spitsin, Rhonda B. Kean, Tatiana Mikheeva, Hilary Koprowski, and D. Craig Hooper*

Department of Microbiology and Immunology, Thomas Jefferson University, Philadelphia, PA 19107

Contributed by Hilary Koprowski, October 23, 2002 (UA) is a metabolite that selectively inhibits an ongoing neuroinflammatory condition with UA has benefi- peroxynitrite-mediated reactions implicated in the pathogenesis of cial effects (10, 11). (MS) and other neurodegenerative diseases. There are also indications that the elevated UA levels nor- Serum UA levels are inversely associated with the incidence of MS mally seen in may be beneficial. For example, MS in humans because MS patients have low serum UA levels and patients generally have serum UA levels that are lower than individuals with () rarely develop the disease. those of controls, and gout (hyperuricemia) patients very rarely Moreover, the administration of UA is therapeutic in experimental develop MS (10). Moreover, it has been demonstrated recently allergic encephalomyelitis (EAE), an animal model of MS. Thus, that serum UA levels are significantly lower in MS patients with raising serum UA levels in MS patients, by oral administration of a a clinical remission and that this correlates with blood–brain UA precursor such as , may have therapeutic value. We have barrier (BBB) dysfunction (17). We therefore have begun to assessed the effects of inosine, as well as , on param- examine the therapeutic potential of elevating UA levels in MS eters relevant to the chemical reactivity of peroxynitrite and the patients, first assessing whether oral administration of UA could pathogenesis of EAE. Both had no effect on chemical reactions be used to raise serum levels (23, 24). This proved ineffective associated with peroxynitrite, such as tyrosine nitration, or on the likely because of the production of uricase by enteric bacteria activation of inflammatory cells in vitro. Moreover, when mice (23). Our alternative approach was to administer a precursor of NEUROSCIENCE treated with the inhibitor potassium oxonate were UA that is not susceptible to breakdown by uricase because fed inosine or inosinic acid, serum UA levels were elevated mark- of dietary is a major source of UA. Inosine edly for a period of hours, whereas only a minor, transient increase is a purine that has been used extensively in humans in serum inosine was detected. Administration of inosinic acid and marketed for a number of years as an energy supplement and suppressed the appearance of clinical signs of EAE and promoted performance enhancer, although there is little scientific evi- recovery from ongoing disease. The therapeutic effect on animals dence that it has these properties (25, 26). A well known side with active EAE was associated with increased UA, but not inosine, effect of inosine use is raised serum UA levels. We have found levels in CNS tissue. We, therefore, conclude that the mode of that the low serum UA levels of MS patients can be raised action of inosine and inosinic acid in EAE is via their metabolism reliably and maintained at desired levels for extended periods of to UA. time by oral administration of inosine (24). Preliminary results suggest that there may be a therapeutic benefit to such treat- ment, and no signs of any deleterious effects have been noted he prospect that the manipulation of serum urate (UA) levels (24). Inosinic acid, the 5Ј-monophosphate of inosine, is used as Tmay provide a natural approach to therapeutic intervention a enhancer and, therefore, also is consumed extensively by in multiple sclerosis (MS) and perhaps other neurodegenerative humans, particularly in Japan (27). processes is supported by several lines of evidence, including (i) Although little is known about the effects of inosinic acid on the specificity of UA for the inhibition of certain peroxynitrite- physiological processes, there is increasing evidence that inosine mediated chemical reactions, such as tyrosine nitration (1–5), (ii) can modulate inflammatory processes (28–31) and neuronal evidence that these reactions may contribute to the pathogenesis growth properties (32, 33) that possibly are relevant to neuro- of MS and other neurodegenerative diseases (6–9), (iii) the degenerative disease. For example, administration of inosine in demonstration that UA is an effective treatment in several a mouse model of acute, lipopolysaccharide (LPS)-induced, lung animal models of CNS inflammation (8, 10–13), and (iv) obser- inflammation had a broad range of antiinflammatory effects vations of a reciprocal relationship between the incidence of MS including the down-regulation of tumor necrosis factor-␣ ex- and serum UA levels (10, 14–17). A common feature of the pression (31). On the other hand, infusion of inosine directly into neuroinflammatory conditions that are influenced by UA is the the cisterna magna or lateral ventricle of the rat brain after presence in CNS tissues of cells expressing inducible nitric oxide experimental induction stimulated axonal outgrowth and (iNOS), an that produces high levels of NO (11). When rewiring in the intact side of the brain (33). In this study, we NO and superoxide are produced in close proximity, for exam- therefore have performed experiments to establish the nature of ple, by activated monocytes (18), they combine to form per- any inosine or inosinic acid effects on peroxynitrite-mediated oxynitrite, which rapidly decomposes in a biological milieu to chemical reactivities and the pathogenesis of EAE. • •Ϫ generate highly reactive intermediates, such as NO2 and CO3 , some of which are inactivated by UA (4, 5). UA is an interme- Experimental Procedures diate of in most but, because of the Assessment of Peroxynitrite-Mediated Oxidation. Peroxynitrite- unusual penetrance of a single point mutation in the urate mediated oxidation was measured in vitro by the conversion of oxidase gene, is the end product in higher-order primates and humans (19–21). When the normally low serum levels of UA in animals used to model neuroinflammatory disease are raised by Abbreviations: UA, uric acid͞urate; BBB, blood–brain barrier; EAE, experimental allergic encephalomyelitis; iNOS, inducible NO synthase͞NOS2; K-Ox, potassium oxonate; LPS, repeated administration of UA, they become resistant to the lipopolysaccharide; MS, multiple sclerosis; MBP, myelin basic protein; DHR, dihydrorho- induction of such diseases, which include experimental allergic damine; SIN-1, 3-morpholinosydnonimine⅐HCl. encephalomyelitis (EAE) and Borna disease (8, 10–13, 22). *To whom correspondence should be addressed at: 1020 Locust Street, JAH Room 454, Moreover, studies in EAE have demonstrated that treatment of Philadelphia, PA 19107-6799. E-mail: [email protected].

www.pnas.org͞cgi͞doi͞10.1073͞pnas.212645999 PNAS ͉ December 10, 2002 ͉ vol. 99 ͉ no. 25 ͉ 16303–16308 Downloaded by guest on September 25, 2021 Fig. 1. DHR123 oxidation by the peroxynitrite donor SIN-1 and activated monocytes is inhibited by UA but not by inosine or inosinic acid. UA, inosine, and inosinic acid were titrated into cultures containing 100 ␮M SIN-1 as a source of peroxynitrite (A) or added at a final concentration of 200 ␮M to cultures containing RAW 264.7 cells activated with LPS (B) as described in Experimental Procedures. The conversion of 5 ␮M DHR123 to rhodamine 123 was measured with a fluorometer (excitation, 485 nm; emission, 530 nm; gain 70). The conversion of DHR123 to rhodamine 123 was used as an indicator of oxidation. Data are expressed as DHR123 oxidation and are presented as means Ϯ SD.

dihydrorhodamine 123 (DHR123, Molecular Probes) to fluores- Induction of EAE. Female 8- to 10-week-old PLSJL mice (The ␮ cent rhodamine 123 (10, 12, 13). DHR123 (5 M) and different Jackson Laboratory) each were immunized s.c. at three sites concentrations of UA, inosine, and inosinic acid were added to along the back with 200 ␮l of an emulsion of 100 ␮g myelin basic cultures containing either 100 ␮M 3-morpholinosydnoni- protein (MBP) in complete Freund’s adjuvant (1:1) containing ͞ mine⅐HCl (SIN-1, Alexis Biochemicals, San Diego) or LPS- 0.05% Mycobacterium butyricum plus an additional 4 mg ml stimulated RAW 264.7 cells, as sources of peroxynitrite, and Mycobacterium tuberculosis H37 RA. Pertussis toxin (List Bio- incubated for1hat37°C. Rhodamine 123 fluorescence then was logical Laboratories, Campbell, CA), 400 ng, was given i.p. twice, measured in a microplate fluorometer (Cytofluor II, PerSeptive on days 0 and 2. Mice were scored for clinical signs of EAE twice Biosystems, Framingham, MA), with an excitation wavelength of daily on the basis of the presence of the following symptoms: 0, 485 nm and an emission wavelength of 530 nm, gain 70. normal mouse; 1, piloerection, tail weakness; 2, tail paralysis; 3, tail paralysis plus hindlimb weakness; 4, tail paralysis plus partial ؊ hindlimb paralysis; 5, total hindlimb paralysis; 6, hind- and Western Analysis of ONOO -Mediated Tyrosine Nitration. BSA (Sig- forelimb paralysis; 7, moribund͞dead. ma) at 1 mg͞ml in PBS was incubated with 1 mM SIN-1 for 2 h at 37°C in the presence and absence of UA, inosine, and inosinic Ј ␮ Ϫ Treatment of Mice. Inosine and inosinic acid (5 -monophosphate, acid (100–800 M) and 10 mM HCO3 added as NaHCO3 ␮ disodium , from yeast, Sigma) were administered twice daily (Sigma). Immediately after incubation, 5 l of each sample was either as i.p. doses of 500 mg͞kg in 100 ␮l of saline or by gastric separated on a 12.5% SDS-polyacrylamide gel and transferred intubation of 1,500 mg͞kg in 100 ␮l of saline with and without onto a poly(vinylidene difluoride) membrane (NEN; ref. 13). two daily i.p. injections of 250 mg͞kg potassium oxonate (K-Ox) Nitrotyrosine-containing proteins were detected with rabbit in 100 ␮l of saline to inhibit breakdown of UA by urate oxidase. polyclonal anti-nitrotyrosine antibody (Upstate Biotechnology, To assess the effects on existing disease, treatment of mice began Lake Placid, NY) and developed with a diaminobenzidine sub- when clinical signs of EAE reached a score of at least 3. strate by using the Vectastain detection kit according to man- ufacturer recommendations (PK-6101, Vector Laboratories). HPLC Analysis of UA, Inosine, and Inosinic Acid Levels in Sera and CNS Tissue. For sera, heparinized blood was collected and sera was Activation of Monocytes and Assessment of iNOS Activity in Vitro. isolated by centrifugation and deproteinized by using HClO4 and Cells of the mouse monocyte–macrophage cell line RAW 264.7 K2HPO4 as described (11). Spinal cord tissue was homogenized (American Type Culture Collection) were grown to 80% con- in 0.1 M perchloric acid, and the supernatant was deproteinized fluence and activated with 1 ␮g͞ml LPS (Escherichia coli with K2HPO4 as detailed (11). HPLC analysis was performed by serotype 055:B5, Sigma) in RPMI medium 1640 supplemented using a C18 reverse-phase column and a 30-min convex gradient with 10% heat-inactivated FBS, 50 units of penicillin, 50 ␮g͞ml of 100% buffer A (0.06 M K2HPO4 and 0.04 M KH2PO4 in H2O, pH 6.0) to 100% buffer B (0.06 M K2HPO4 and 0.04 M KH2PO4 streptomycin, and 5 mM L-glutamine in the presence or absence ͞ of UA, inosine, and inosinic acid (200 ␮M) overnight. Nitrite in 25% methanol 75% H2O, pH 6.0) on a Beckman Coulter 125 solvent module and 168 diode array detector. UA was detected accumulation in the medium was assessed by using the Griess at 292 nm at Ϸ5 min, and inosinic acid and inosine were detected reaction as detailed (10). The activation of iNOS genes was at 254 nm at Ϸ7 and 22 min, respectively. Concentrations were assessed by real-time quantitative RT-PCR analysis of the determined by comparison of the integral of peak areas with expression of the specific mRNAs by using a Bio-Rad (Hercules, those of known controls. CA) iCycler iQ real-time detection system as detailed (22). Data were calculated based on a threshold cycle (Ct), determined as Results the cycle with a signal higher than that of the background (signal Chemical Reactivities Mediated Through Peroxynitrite Decomposition detected in cycles 2–10) plus 10 times its SD. Data are expressed Are Inhibited in Vitro by UA but Not Inosine or Inosinic Acid. Per- as a fold increase in mRNA expression calculated by exp[Ct oxynitrite decomposition mediates several chemical reactions lowest expresser (e.g., unstimulated cells) Ϫ Ct test value] that are highly sensitive to inhibition by UA, including oxidation divided by the same value determined for the housekeeping gene of DHR123 to rhodamine 123 and tyrosine nitration (13). Neither GAPDH. inosine nor inosinic acid had any effect on the oxidation of

16304 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.212645999 Scott et al. Downloaded by guest on September 25, 2021 Fig. 2. UA but not inosine or inosinic acid inhibits tyrosine nitration by SIN-1. BSA (1 mg͞ml) in PBS was incubated with 1 mM SIN-1 for2hat37°Cinthe Ϫ presence of 10 mM HCO3 , with and without decreasing concentrations of UA, inosine, and inosinic acid. Immediately after incubation, 5 ␮l of each sample was separated on a 12.5% SDS͞PAGE gel and transferred onto a poly(vinyli- dene difluoride) membrane. Nitrotyrosine-containing proteins were detected by using a rabbit polyclonal anti-nitrotyrosine antibody as detailed in Exper- Ϫ imental Procedures. 1, Molecular mass markers; 2, SIN-1 and HCO3 ; 3, SIN-1, Ϫ ␮ Ϫ ␮ Ϫ HCO3 , and 800 M UA; 4, SIN-1, HCO3 , and 400 M UA; 5, SIN-1, HCO3 , and ␮ Ϫ ␮ Ϫ ␮ 200 M UA; 6, SIN-1, HCO3 , and 100 M UA; 7, SIN-1, HCO3 , and 800 M Ϫ ␮ Ϫ ␮ inosine; 8, SIN-1, HCO3 , and 400 M inosine; 9, SIN-1, HCO3 , and 200 M ino- Ϫ ␮ Ϫ ␮ sine; 10, SIN-1, HCO3 , and 100 M inosine; 11, SIN-1, HCO3 , and 800 M Ϫ ␮ Ϫ inosinic acid; 12, SIN-1, HCO3 , and 400 M inosinic acid; 13, SIN-1, HCO3 , and ␮ Ϫ ␮ 200 M inosinic acid; 14, SIN-1, HCO3 , and 100 M inosinic acid; and 15, Fig. 3. LPS-induced iNOS expression and nitrite formation by monocytes in molecular mass markers. vitro are unaffected by the addition of UA, inosine, and inosinic acid. RAW 264.7 cells were activated by LPS as described in Experimental Procedures in the presence and absence of 200 ␮M UA, inosine, or inosinic acid. iNOS DHR123, whereas UA at physiological levels completely inhib- expression was determined by real-time quantitative RT-PCR (A), and nitrite ited the reaction, whether mediated by a peroxynitrite donor or production in culture medium was measured by the Griess reaction (B)as Ϯ activated monocytes (Fig. 1). A similar result was observed for detailed in Experimental Procedures. Results are expressed as mean SD of a NEUROSCIENCE the nitration of serum proteins by peroxynitrite, where UA minimum of two independent samples assayed in duplicate. strongly prevented the reaction, whereas comparable concen- trations of inosine and inosinic acid had no noticeable effect (Fig. 2). As is evident from Fig. 4, serum UA levels can be raised only transiently in mice likely because of the rapid metabolism of UA iNOS Gene Expression by Monocytes in Vitro Is Unaffected by UA, by urate oxidase. To maintain elevated serum UA levels over a Inosine, or Inosinic Acid. Our previous studies have not detected more extended period, we have examined the utility of several any changes in monocyte or lymphocyte activation associated urate oxidase inhibitors including K-Ox, 8-azaxanthine, and with UA treatment (12, 13). However, there is evidence that 9-methyl-UA. K-Ox has been used more extensively in vivo (e.g., inosine may have inhibitory effects on pathways related to ref. 34) and therefore has become our urate oxidase inhibitor of immune function (28–31). iNOS is an enzyme that catalyzes the choice. By titrating K-Ox in mice treated with UA, we have production of NO by monocytes during an inflammatory re- determined that serum UA levels can be raised more effectively sponse and has been implicated in the pathogenesis of both MS and maintained at higher levels for over 8 h when using a single and EAE (6–9, 11–13). Therefore, we assessed whether, unlike i.p. dose of 10 mg of K-Ox (Fig. 4). UA, inosine or inosinic acid could inhibit the activation of iNOS Using K-Ox to inhibit the metabolism of UA, we compared the in cultures of mouse RAW monocytes stimulated by LPS. Fig. 3 efficacy of inosinic acid vs. inosine in raising serum UA levels shows that neither the levels of iNOS mRNA expression nor the after oral administration. After a single i.p. dose of K-Ox, serum accumulation of the NO metabolite, nitrite, in the cultures is UA levels were elevated strongly for periods of up to8hbythe altered by the addition of UA, inosine, or inosinic acid. This administration of inosinic acid, whereas inosine, initially, was indicates that none of these compounds has a direct inhibitory effect on the induction of iNOS.

Oral Administration of Inosine and Inosinic Acid Primarily Raises Serum UA Levels in Mice. The oral administration of inosine to MS patients has proven to effectively raise their normally low serum UA levels (23, 24). Because, unlike humans, mice metabolize UA to , we first determined whether their serum UA levels could be elevated similarly. In addition to inosine, we also assessed the capacity of its derivative, inosinic acid, to raise serum UA levels in mice. Inosinic acid is particularly interesting in this regard because it has been consumed regularly by a large number of individuals as a flavor enhancer, without reported side effects (27). The i.p. administration of 500 mg͞kg of either inosine or inosinic acid in 100 ␮l of saline to adult female PLSJL mice caused a transient, minor increase in serum inosine levels Ϸ ␮ Fig. 4. Inhibition of urate oxidase with potassium oxonate delays UA (peak, 30 M at 15 min postinjection) as well as a strong metabolism in mice. Groups of mice (n ϭ 3–4) were immunized with MBP as elevation in serum UA (data not shown). Serum UA levels were Ϸ ␮ detailed in Experimental Procedures and injected i.p with UA (1, 5, or 10 mg) maximal at 15 min postinjection ( 400 M) but declined rapidly and K-Ox (1, 5, or 10 mg) as indicated. Serum samples were collected 30 min, to between 100 and 150 ␮M, where it was maintained for 2–4h 1 h, 4 h, and 8 h postinjection. Serum UA levels were assessed by HPLC as (data not shown). described in Experimental Procedures. Data are presented as mean Ϯ SD.

Scott et al. PNAS ͉ December 10, 2002 ͉ vol. 99 ͉ no. 25 ͉ 16305 Downloaded by guest on September 25, 2021 Fig. 6. Effects of inosine and inosinic acid administration on the levels of inosine, inosinic acid, and UA in the CNS tissues of PLSJL mice with active EAE. Fig. 5. Effect of oral administration of inosinic acid and inosine on serum UA, Groups of three to five PLSJL mice with active EAE (mean clinical score, 3.8 Ϯ inosine, and inosinic acid levels. Groups of MBP-immunized mice (n ϭ 5) were 0.2) were treated once with 5 mg of K-Ox i.p. and 30 mg of either inosinic acid treated once with 5 mg of K-Ox i.p. and 30 mg of either inosinic acid or inosine or inosine in 100 ␮l saline by gastric intubation. Untreated, healthy PLSJL mice in 100 ␮l of saline by gastric intubation. Serum samples were collected at 15 served as donors of normal tissue. At 90 min, 4 h, and 8 h posttreatment, min, 30 min, 1 h, 2 h, and 4 h posttreatment. Serum levels of UA, inosine, and animals were killed and spinal cord tissues were collected for analysis. The inosinic acid were measured by HPLC as described in Experimental Procedures. levels of inosinic acid, inosine, and UA in the spinal cord tissues were deter- Results are expressed as mean Ϯ SD. mined by HPLC as described in Experimental Procedures. Data are expressed as the fold increase in tissue levels, determined by dividing the levels detected in test animals by the average normal level. Results are presented as mean Ϯ less effective (Fig. 5). As shown in Fig. 5, oral administration of SD. *, P Ͻ 0.05; **, P Ͻ 0.01; and ***, P Ͻ 0.001 compared with normal levels both inosine and inosinic acid caused a substantial elevation in by one-way ANOVA with post hoc Dunnett’s test. serum UA levels but only a minor increase in serum inosine. Inosinic acid ingestion resulted in peak serum UA levels at 15 min that were roughly twice those caused by inosine, but these our prior experiments with UA (10–13). With two daily doses of declined within 2 h such that the elevated UA levels caused by inosinic acid, the onset of clinical signs of EAE was delayed and both reagents were approximately the same (Fig. 5). In both the severity of the disease was reduced (Fig. 7). Moreover, the cases, serum inosine levels dropped to background between 30 30% mortality seen in vehicle-treated controls also was pre- and 60 min after ingestion, whereas UA levels remained elevated vented by inosinic acid treatment (Fig. 7). Our previous studies for more than 4 h (Fig. 5). Inosinic acid was not detected established that mice with clinical signs of EAE recover rapidly consistently in serum at any time point (Fig. 5). when given UA (10, 11). To assess whether raising serum UA levels by oral administration of a precursor has similar effects, we Oral Administration of Inosine and Inosinic Acid Raises UA Levels in the fed inosinic acid to mice with existing signs of active EAE while CNS Tissues of Mice with EAE. Normally, the BBB is relatively prolonging raised UA levels by i.p. administration of K-Ox. As impervious to UA and UA levels in CNS tissues are low (11). can be seen readily from Fig. 8, such treatment promoted early When the BBB becomes compromised during the course of EAE recovery from clinical signs of EAE, whereas disease severity in or MS, UA accumulates in CNS tissue (11, 35, 36). Presumably, mice that received K-Ox alone remained relatively constant. ingested inosine or inosinic acid also could accumulate in CNS Like K-Ox treatment alone, sham treatment with saline also had tissue under similar circumstances, provided that they can cross no effect on the clinical course of the disease (data not shown). the BBB and are metabolized slowly in CNS tissues. To deter- Discussion mine whether this may be the case, mice with clinical signs of EAE were fed inosine or inosinic acid and their CNS tissues were Raising their normally low serum UA levels protects mice from assessed for levels of these and UA. As shown in Fig. 6, UA levels the development of EAE (10). Moreover, UA treatment of mice in the spinal cord tissues of mice with EAE became elevated but with EAE promotes their recovery (10, 11). The therapeutic there was no significant effect on endogenous inosine levels and effects of UA in EAE and other animal models of CNS inosinic acid remained essentially undetectable. inflammation (10–13, 22), together with evidence of an inverse correlation between serum UA levels and the incidence of MS, Oral Administration of Inosinic Acid Inhibits the Development of EAE have renewed interest in the specificity of this purine metabolite. and Promotes Recovery from Clinical Signs of the Disease. Based on Although UA has long been recognized as an (2), its efficacy in raising serum UA levels in mice, and apparent recent observations suggest that it is not a general antioxidant safety in humans, we have examined the effects of inosinic acid but a specific inhibitor of radicals generated by the decomposi- on the development of EAE. As an initial test, we administered tion of peroxynitrite, the product of NO and superoxide, in a inosinic acid i.p. to PLSJL mice immunized with MBP, based on biological milieu (4, 5). These UA-sensitive, peroxynitrite-

16306 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.212645999 Scott et al. Downloaded by guest on September 25, 2021 Fig. 7. Effects of inosinic acid administration on the clinical course of EAE in PLSJL mice. Groups of PLSJL mice (n ϭ 10) were immunized with MBP as detailed in Experimental Procedures. From day 7 postimmunization (arrow in A), animals received two daily i.p. doses of either 10 mg of inosinic acid in saline (INA) or saline alone (Vehicle). Mice were scored twice daily for clinical signs of EAE by independent investigators as detailed (10). Results are presented as percent incidence of disease (A), mean disease score Ϯ SE (B), and percent mortality (C). The clinical scores between inosinic acid and vehicle-treated mice are significantly different (P Ͻ 0.001) by the Wilcoxon signed-rank test.

• •Ϫ decomposition radicals, possibly NO2 and CO3 , therefore are only transiently, attests to the fact that inosine is metabolized likely to be responsible for the elements of a CNS inflammatory rapidly to UA. As is the case with UA (10–13), inosinic acid response that have been associated with peroxynitrite (4, 5). treatment not only inhibits the development of EAE but also Using UA administration to probe the effects of peroxynitrite in promotes recovery from existing disease. CNS inflammation, we have discovered that it contributes not Is it possible that increased inosine levels are directly respon- only to CNS tissue damage but also to enhanced BBB perme- sible for the therapeutic effects seen in EAE? To significantly ability and immune cell invasion into the CNS (11–13, 22). reduce the production of several cytokines by LPS-stimulated Peripheral aspects of the immune response that leads to CNS peritoneal macrophages in vitro, 30-min pretreatment with 10– 100 ␮M inosine followed by a 24-h incubation was necessary (28). inflammation, such as the up-regulation of iNOS expression by NEUROSCIENCE ͞ monocytes, are unchanged by UA treatment (12, 13). We, Pretreatment for 30 min with 100 mg kg inosine applied directly therefore, have concluded that the peroxynitrite-based radicals into the peritoneum was required to transiently inhibit cytokine inactivated by UA not only cause tissue damage by tyrosine production in mice in response to LPS that also was administered nitration and other chemical reactions but also trigger changes i.p. (28). Moreover, the continued presence of inosine evidently in BBB function that promote cell infiltration into CNS tissues was necessary to stimulate axonal growth in vivo (33). In our (11–13, 22). Ultimately, these processes result in clinical signs of studies, the activation of monocytes to express iNOS or produce disease that can be prevented by the administration of UA or a NO in vitro was unaffected by the presence of inosine, inosinic acid, or UA. In addition, unlike UA, neither inosinic acid nor precursor such as inosinic acid. inosine inhibits the peroxynitrite-mediated nitration of tyrosine Inosine administration to humans clearly can be used to raise or oxidation of DHR, reactivities that have been associated with serum UA levels in MS patients (23, 24). The current findings the pathogenesis of EAE (10–13). Because serum inosine levels suggest that inosinic acid is adsorbed in the mouse gut more Ϸ ␮ Ϸ ͞ were elevated only to a peak of 100 M for 30 min and no rapidly efficiently than inosine but that both can effectively raise increase in CNS levels were seen in our experiments, we consider serum UA levels. That serum UA levels become elevated for it unlikely that inosine is directly therapeutic in EAE. As extended periods of time after inosine or inosinic acid admin- expected if UA is the functional molecule after the administra- istration when K-Ox is present, but inosine levels are increased tion of inosinic acid to mice with EAE, serum UA levels were maintained for extended periods of time and elevated levels of UA, but not inosine or inosinic acid, were found in diseased spinal cord tissue. We therefore conclude that the effects of inosinic acid in EAE are likely to be mediated through its metabolism to UA, culminating in the inactivation of peroxyni- trite-decomposition radicals at the BBB and in diseased spinal cord tissue. Perhaps the best direct evidence of peroxynitrite’s involve- ment in the pathogenesis of MS is that active plaques often contain accumulations of iNOS-positive cells and nitrotyrosine (6–9). However, the inverse relationship between serum UA levels and the incidence of MS also attests to the possibility of an association between peroxynitrite and the disease process. Several studies have confirmed our original observation that serum UA levels tend to be lower in MS patients than controls Fig. 8. Administration of inosinic acid plus potassium oxonate promoted (14–17). We recently have found that this difference extends to recovery from clinical signs of active EAE. PLSJL mice were immunized with identical twins, where individuals with MS have lower serum UA MBP as detailed in Experimental Procedures. When mice developed a clinical levels than their healthy siblings (16). Analysis of the 1995 EAE score between 3 and 4, they were randomized into two groups of six mice Medicare͞Medicaid database revealed that the incidence of MS each, one group receiving 5 mg of K-Ox in 100 ␮l of saline i.p. three times daily plus 100 ␮l of saline by gastric intubation twice a day (K-Ox ϩ Saline), and the among gout patients was exceptionally low (10). Even if most of second receiving the same regimen of K-Ox plus 30 mg of inosinic acid in 100 the gout patients were male, which is not the case for the ␮l of saline by gastric intubation twice a day (K-Ox ϩ INA). Mice were scored predominant age group in the database, and two-thirds of the twice daily for clinical signs of EAE by independent investigators. *, P Ͻ 0.05; MS patients were female, the combined incidence of MS and **, P Ͻ 0.01 compared with K-OX ϩ saline group by the Mann–Whitney U test. gout should have been Ϸ10-fold higher than observed. The

Scott et al. PNAS ͉ December 10, 2002 ͉ vol. 99 ͉ no. 25 ͉ 16307 Downloaded by guest on September 25, 2021 implication from these observations is that UA may protect through inactivation of peroxynitrite decomposition products by against the development of MS, a possibility made more distinct UA. We expect that this also will be the case if inosine by the demonstration in animal models that UA administration, administration to MS patients proves therapeutic. in addition to preventing a particular element of CNS tissue damage, can inhibit CNS inflammation through protecting We thank C. M. Brimer, W. K. Hooper, and K. M. Hooper for technical against BBB-permeability changes (11, 13, 22). That the thera- assistance. This work was supported by Award Number RG 2896A8͞5to peutic effects of inosinic acid treatment of EAE are associated D.C.H. from the National Multiple Sclerosis Society and by a grant to with increased serum and tissue levels of UA, rather than the Biotechnology Foundation Laboratories from the Commonwealth of inosine, leads us to conclude that the mechanism of action is Pennsylvania.

1. Aruoma, O. I. & Halliwell, B. (1989) FEBS Lett. 244, 6–80. 19. Varela-Eschavarria, A., Montes de Oca-Luna, R. & Barrera-Saldana, H. A. 2. Whiteman, M. & Halliwell, B. (1996) Free Res. 25, 275–283. (1988) FASEB J. 2, 3092–3096. 3. Santos, C. X. C., Anjos, E. I. & Augusto, O. (1999) Arch. Biochem. Biophys. 372, 20. Wu, X. W., Muzny, D. M., Lee, C. C. & Catskey, C. T. (1992) J. Mol. Evol. 34, 285–294. 78–84. 4. Squadrito, S. L., Cueto, R., Splenser, A. E., Valavanidis, A., Zhang, H., Uppu, 21. Yeldandi, A. V., Yeldandi, V., Kumar, S., Murthy, C. V., Wang, X. D., Alvares, R. M. & Pryor, W. A. (2000) Arch. Biochem. Biophys. 376, 333–337. K., Rao, M. S. & Reddy, J. K. (1991) Gene 109, 281–284. 5. Whiteman, M., Ketsawatsakul, U. & Halliwell, B. (2002) Ann. N.Y. Acad. Sci. 22. Hooper, D. C., Kean, R. B., Scott, G. S., Spitsin, S. V., Morimoto, K., Bette, 962, 242–259. M., Ro¨hrenbeck, A. M., Dietzschold, B. & Weihe, E. (2001) J. Immunol. 167, 6. Bo, L., Dawson, T. M., Wesselingh, S., Mork, S., Choi, S., Kong, P. A., Hanley, 3470–3477. D. & Trapp, B. D. (1994) Ann. Neurol. 36, 778–786. 23. Koprowski, H., Spitsin, S. V. & Hooper, D. C. (2001) Ann. Neurol. 49, 139. 7. Bagasra, O., Michaels, F. H., Zheng, Y. M., Bobroski, L. E., Spitsin, S. V., Fu, 24. Spitsin, S., Hooper, D. C., Leist, T., Streletz, L. J., Mikheeva, T. & Koprowski, Z. F., Tawadros, R. & Koprowski, H. (1995) Proc. Natl. Acad. Sci. USA 92, 12041–12045. H. (2001) Multiple Sclerosis 7, 313–319. 8. Hooper, D. C., Bagasra, O., Marini, J. C., Zborek, A., Ohnishi, S. T., Kean, R., 25. Williams, M. H., Kreider, R. B., Hunter, D. W., Somma, C. T., Shall, L. M., Champion, J. M., Sarker, A. B., Bobroski, L., Farber, J. L., et al. (1997) Proc. Woodhouse, M. L. & Rokitski, L. (1990) Med. Sci. Sports Exercise 22, 517–522. Natl. Acad. Sci. USA 94, 2528–2533. 26. Starling, R. D., Trappe, T. A., Short, K. R., Sheffield-Moore, M., Jozsi, A. C., 9. Cross, A. H., Maning, P. T., Keeling, R. M., Schmidt, R. E. & Misko, T. P. Fink, W. J. & Costill, D. L. (1996) Med. Sci. Sports Exercise 28, 1193–1198. (1998) J. Neuroimmunol. 88, 45–56. 27. Fuke, S. & Konosu, S. (1991) Physiol. Behav. 49, 863–868. 10. Hooper, D. C., Spitsin, S., Kean, R. B., Champion, J. M., Dickson, G. M., 28. Hasko´, G., Kuhel, D. G., Ne´meth, Z. H., Mabley, J. G., Stachlewitz, R. F., Chaudhry, I. & Koprowski, H. (1998) Proc. Natl. Acad. Sci. USA 95, 675–680. Vira´g, L., Lohinai, Z., Southan, G. J., Salzman, A. L. & Szabo´, C. (2000) 11. Hooper, D. C., Scott, G. S., Zborek, A., Mikheeva, T., Kean, R. B., Koprowski, J. Immunol. 164, 1013–1019. H. & Spitsin, S. V. (2000) FASEB J. 14, 691–698. 29. Liaudet, L., Mabley, J. G., Soriano, F. G., Pacher, P., Marton, A., Hasko´,G. 12. Spitsin, S. V., Scott, G. S., Kean, R. B., Mikheeva, T. & Hooper, D. C. (2000) & Szabo´, C. (2001) Am. J. Respir. Crit. Care Med. 164, 1213–1220. Neurosci. Lett. 292, 137–141. 30. Marton, A., Pacher, P., Murthy, K. G., Nemeth, Z. H., Hasko´, G. & Szabo´,C. 13. Kean, R. B., Spitsin, S. V., Mikheeva, T., Scott, G. S. & Hooper, D. C. (2000) (2001) Int. J. Mol. Med. 8, 617–621. J. Immunol. 165, 6511–6518. 31. Liaudet, L., Mabley, J. G., Pacher, P., Virag, L., Soriano, F. G., Marton, A., 14. Sotgiu, S., Vanoli, M., Pugliatti, G., Fresu, V., Agnetti, I. & Rostati, A. G. Hasko´, G., Deitch, E. A. & Szabo´, C. (2002) Ann. Surg. 235, 568–578. (1999) J. Neuroimmunol. 101, A53. 32. Benowitz, L. I., Goldberg, D. E., Madsen, J. R., Soni, D. & Irwin, N. (1999) 15. Drulovic, J., Dujmovic, I., Stojsavljevic, N., Mesaros, S., Miljkovic, D., Peric, Proc. Natl. Acad. Sci. USA 96, 113486–113490. V., Dragutinovic, G., Marinkoviv, J., Levic, Z. & Stojkovic, M. M. (2001) J. Neurol. 248, 121–126. 33. Chen, P., Goldberg, D. E., Kolb, B., Lanser, M. & Benowitz, L. I. (2002) Proc. 16. Spitsin, S. V., Hooper, D. C., Mikheeva, T. & Koprowski, H. (2001) Multiple Natl. Acad. Sci. USA 99, 9031–9036. Sclerosis 7, 165–166. 34. Mangoff, S. C. & Milner, J. A. (1978) Proc. Soc. Exp. Biol. Med. 157, 110–115. 17. Toncev, G., Milicic, B., Toncev, S. & Samardzic, G. (2002) Eur. J. Neurol. 9, 35. Honegger, C. G., Krenger, W. & Langemann, H. (1986) Neurosci. Lett. 69, 221–226. 109–114. 18. Ischiropoulos, H., Zhu, L. & Beckman, J. S. (1992) Arch. Biochem. Biophys. 298, 36. Langemann, H., Kabiersch, A. & Newcombe, J. (1992) Eur. Neurol. 32, 446–451. 248–252.

16308 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.212645999 Scott et al. Downloaded by guest on September 25, 2021