Transepithelial Effect of Probiotics in a Novel Model of Gut Lumen to Nerve Signaling.

Recent studies have shown that the gut microbiome changes brain function, behavior, and psychiatric and neurological disorders. The Gut-Brain Axis (GBA) provides a neuronal pathway to explain this. But exactly how do commensal bacteria signal through the epithelial layer of the large intestine to activate GBA nerve afferents? An in vitro model is described. We differentiated two human cell lines: Caco2Bbe1 into mature epithelium on 0.4-micron filters and then SH-SY5Y into mature neurons in 24-well plates. These were co-cultured by placing the epithelium-laden filters 1 mm above the neurons. Twenty-four hours later they were tri-cultured by apical addition of 107Lactobacillus rhamnosus or Lactobacillus fermentum which settled on the epithelium. Alone, the Caco2bbe1 cells stimulated neurite outgrowth in underlying SH-SY5Y. Beyond this, the lactobacilli were well tolerated and stimulated further neurite outgrowth by 24 h post-treatment, though not passing through the filters. The results provide face validity for a first-of-kind model of transepithelial intestinal lumen-to nerve signaling. The model displays the tight junctional barrier characteristics found in the large intestine while at the same time translating stimulatory signals from the bacteria through epithelial cells to attracted neurons. The model is easy to set-up with components widely available.

Transepithelial Anti-Neuroblastoma Response to Kale among Four Vegetable Juices Using In Vitro Model Co-Culture System.

Juicing vegetables is thought to be an anticancer treatment. Support exists for a rank order of anticancer greens (kale > dandelion > lettuce > spinach) based on degrees of bioavailability of different phytochemicals, also offset by some noxious molecules (i.e., calcium-oxalate). We developed a new in vitro transepithelial anti-neuroblastoma model system. The juices were diluted as predicted once in the small intestine. They were applied to apical Caco-2Bbe1 cells atop dividing SH-SY5Y neuroblastoma cells, and changes in transepithelial electrical resistance (TEER) and cell growth were considered with juice spectroscopies. Studied first in monoculture, kale and dandelion were the most cytostatic juices on SH-SY5Ys, lettuce showed no effect, and high (4.2%) spinach was cytotoxic. In co-culture, high (4.2%) kale was quickest (three days) to inhibit neuroblastoma growth. By five days, dandelion and kale were equally robust. Lettuce showed small anti-proliferative effects at five days and spinach remained cytotoxic. Spinach's cytotoxicity corresponded with major infrared bands indicative of oxalate. Kale juice uniquely induced reactive oxygen species and S-phase cell cycle arrest in SH-SY5Y. The superiority of kale and dandelion was also apparent on the epithelium, because raising TEER levels is considered healthy. Kale's unique features corresponded with a major fluorescent peak that co-eluted with kaempferol during high performance liquid chromatography. Because the anticancer rank order was upheld, the model appears validated for screening anticancer juices.

Human Cells Grown With or Without Substitutes for Fetal Bovine Serum.

Safety concerns over cell-derived pharmaceutical products being manufactured in supplements of fetal bovine serum (FBS) have ignited pleas to replace FBS. Herein, four newly marketed alternatives to FBS were compared: a xeno-free product called Cell-Ess®, a human platelet lysate marketed as GroPro®, and two mixtures of adult bovine serum varying in their proportions of neonatal growth factors, called Liporo® and FetalGro®. An endothelial cell line (C2BBe1) and a neuronal cell line (SHSY5Y) near confluency in media with 10% FBS were selectively scraped and taken through a 25-day step-wise algorithm to replace FBS, and another human endothelial cell line (HRA-19) was studied to replicate C2BBe1. Cells were stained, counted, and compared for viability, migration, and spheroids. The C2BBe1 and HRA-19 cell lines failed to proliferate in 10% Cell-Ess® but grew in 10% GroPro® or 10% FetalGro® reasonably well compared to reference 10% FBS. With SH-SY5Y, only FetalGro® approached FBS's efficacy. These were all inferior to 11 different branded lots of FBS (positive controls), but five days into switching just amongst the FBS brands, 4 of 11 supported less proliferation than reference FBS in endothelial HRA-19 (p < 0.004). Moreover, neurospheres were enriched in two branded lots of FBS and FetalGro® (each p < 0.004), neurospheres being an unwanted phenotype for any neuronal cell application. Because platelet-derived GroPro® stood out amongst the non-FBS growth supplements to allow proliferation without inducing spheroids, it seems the best (mindful that the cells still grew slower in it compared to FBS). While no perfect replacement was found amongst the alternatives to FBS, the algorithm for switching should be useful in future testing of new alternatives to FBS as the need arises to switch from FBS and expand pharmaceutical products with safety for human use.

Agmatine: clinical applications after 100 years in translation.

Agmatine (decarboxylated arginine) has been known as a natural product for over 100 years, but its biosynthesis in humans was left unexplored owing to long-standing controversy. Only recently has the demonstration of agmatine biosynthesis in mammals revived research, indicating its exceptional modulatory action at multiple molecular targets, including neurotransmitter systems, nitric oxide (NO) synthesis and polyamine metabolism, thus providing bases for broad therapeutic applications. This timely review, a concerted effort by 16 independent research groups, draws attention to the substantial preclinical and initial clinical evidence, and highlights challenges and opportunities, for the use of agmatine in treating a spectrum of complex diseases with unmet therapeutic needs, including diabetes mellitus, neurotrauma and neurodegenerative diseases, opioid addiction, mood disorders, cognitive disorders and cancer.

Putative agmatinase inhibitor for hypoxic-ischemic new born brain damage.

Agmatine is an endogenous brain metabolite, decarboxylated arginine, which has neuroprotective properties when injected intraperitoneally (i.p.) into rat pups following hypoxic-ischemia. A previous screen for compounds based on rat brain lysates containing agmatinase with assistance from computational chemistry, led to piperazine-1-carboxamidine as a putative agmatinase inhibitor. Herein, the neuroprotective properties of piperazine-1-carboxamidine are described both in vitro and in vivo. Organotypic entorhinal-hippocampal slices were firstly prepared from 7-day-old rat pups and exposed in vitro to atmospheric oxygen depletion for 3 h. Upon reoxygenation, the slices were treated with piperazine-1-carboxamidine or agmatine (50 µg/ml agents), or saline, and 15 h later propidium iodine was used to stain. Piperazine-1-carboxamidine or agmatine produced substantial in vitro protection compared to post-reoxygenated saline-treated controls. An in vivo model involved surgical right carotid ligation followed by exposure to hypoxic-ischemia (8 % oxygen) for 2.5 h. Piperazine-1-carboxamidine at 50 mg/kg i.p. was given 15 min post-reoxygenation and continued twice daily for 3 days. Cortical agmatine levels were elevated (+28.5 %) following piperazine-1-carboxamidine treatment with no change in arginine or its other major metabolites. Histologic staining with anti-Neun monoclonal antibody also revealed neuroprotection of CA1-3 layers of the hippocampus. Until endpoint at 22 days of age, no adverse events were observed in treated pups' body weights, rectal temperatures, or prompted ambulation. Piperazine-1-carboxamidine therefore appears to be a neuroprotective agent of a new category, agmatinase inhibitor.

Database of genetic studies of bipolar disorder.

This study describes the construction and preliminary analysis of a database of summary level genetic findings for bipolar disorder from the literature. The database is available for noncommercial use at http://bioprogramming.bsd.uchicago.edu/BDStudies/. This may be the first complete collection of published gene-specific linkage and association findings on bipolar disorder, including genome-wide association studies. Both the positive and negative findings have been incorporated so that the statistical and contextual significance of each finding may be compared semi-quantitatively and qualitatively across studies of mixed technologies. The database is appropriate for searching a literature populated by mainly underpowered studies, and if 'hits' are viewed as tentative knowledge for future hypothesis generation. It can serve as the basis for a mega-analysis of candidate genes. Herein, we discuss the most robust and best replicated gene findings to date in a contextual manner.

Pro-inflammatory biomakers in depression: treatment with venlafaxine.

High levels of pro-inflammatory biomarkers have been reported in depression. In the present study, five pro-inflammatory biomarkers were measured in the blood of patients with major depressive disorder (MDD). Biomarker levels were compared to age- and sex-matched healthy subjects. Patients with MDD had significantly higher baseline levels of tumour necrosis factor-alpha (TNFalpha, P=0.04), interleukin-1beta (IL1beta, P=0.03), and monocyte chemotactic protein-1 (MCP-1; P=0.02) compared to controls. There were no differences between groups in levels of cell determinant-40 ligand (CD40L) and C-reactive protein (CRP). A subset of the MDD patients consented to undergo treatment with venlafaxine (an SNRI: at lower doses a selective serotonin reuptake inhibitor; at higher doses also a norepinephrine reuptake inhibitor) for 8 weeks. By week 8, all treatment completers had responded therapeutically. However, levels of TNFalpha, IL1beta, and MCP-1 remained elevated. A concave quadratic equation described the associations between plasma venlafaxine concentrations and IL1beta (P=0.03), TNFalpha (P=0.09), and MCP-1 (P=0.02), suggesting that these biomarkers may have become selectively lowered in the serotonergic dose range of venlafaxine. This is the first report of venlafaxine's possible effect on pro-inflammatory biomarkers.

IRAS, a candidate for I1-imidazoline receptor, mediates inhibitory effect of agmatine on cellular morphine dependence.

Agmatine, an endogenous ligand for the I1-imidazoline receptor, has previously been shown to prevent morphine dependence in rats and mice. To investigate the role of imidazoline receptor antisera-selected protein (IRAS), a strong candidate for I1R, in morphine dependence, two CHO cell lines were created, in which mu opioid receptor (MOR) was stably expressed alone (CHO-mu) or MOR and IRAS were stably co-expressed (CHO-mu/IRAS). After 48 h administration of morphine (10 microM), naloxone induced a cAMP overshoot in both cell lines, suggesting cellular morphine dependence had been produced. Agmatine (0.1-2.5 microM) concentration-dependently inhibited the naloxone-precipitated cAMP overshoot when co-pretreated with morphine in CHO-mu/IRAS, but not in CHO-mu. Agmatine at 5-100 microM also inhibited the cAMP overshoot in CHO/mu and CHO-mu/IRAS. Efaroxan, an I1R-preferential antagonist, completely blocked the effect of agmatine on the cAMP overshoot at 0.1-2.5 microM in CHO-mu/IRAS, while partially reversing the effects of agmatine at 5-100 microM. L-type calcium channel blocker nifedipine entirely mimicked the effects of agmatine at high concentrations on forskolin-stimulated cAMP formation in CHO-mu and naloxone-precipitated cAMP overshoot in morphine-pretreated CHO-mu. Therefore, IRAS, in the co-transfected CHO-mu/IRAS cell line, appears necessary for low concentrations of agmatine to cause attenuation of cellular morphine dependence. An additional effect of agmatine at higher concentrations seems to relate to both transfected IRAS and some naive elements in CHO cells, and L-type voltage-gated calcium channels are not ruled out. This study suggests that IRAS mediates agmatine's high affinity effects on cellular morphine dependence and may play a role in opioid dependence.

Expression of human arginine decarboxylase, the biosynthetic enzyme for agmatine.

Agmatine, an amine formed by decarboxylation of L-arginine by arginine decarboxylase (ADC), has been recently discovered in mammalian brain and other tissues. While the cloning and sequencing of ADC from plant and bacteria have been reported extensively, the structure of mammalian enzyme is not known. Using homology screening approach, we have identified a human cDNA clone that exhibits ADC activity when expressed in COS-7 cells. The cDNA and deduced amino acid sequence of this human ADC clone is distinct from ADC of other forms. Human ADC is a 460-amino acid protein that shows about 48% identity to mammalian ornithine decarboxylase (ODC) but has no ODC activity. While naive COS-7 cells do not make agmatine, these cells are able to produce agmatine, as measured by HPLC, when transfected with ADC cDNA. Northern blot analysis using the cDNA probe indicated the expression of ADC message in selective human brain regions and other human tissues.

Effect of agmatine against cell death induced by NMDA and glutamate in neurons and PC12 cells.

AIMS: Agmatine is an endogenous guanido amine and has been shown to be neuroprotective in vitro and in vivo. The aims of this study are to investigate whether agmatine is protective against cell death induced by different agents in cultured neurons and PC12 cells. METHODS: Cell death in neurons, cultured from neonatal rat cortex, was induced by incubating with (a) NMDA (100 microM) for 10 min, (b) staurosporine (protein kinase inhibitor, 100 nM) for 24 h, and (c) calcimycin (calcium ionophore, 100 nM) for 24 h in the presence and absence of agmatine (1 micro M to 1 mM). Cell death in PC12 cells was induced by exposure to glutamate (10 mM), staurosporine (100 nM), and calcimycin (100 nM). The activity of lactate dehydrogenase (LDH) in the medium was measured as the marker of cell death and normalized to cellular LDH activity. RESULTS: Agmatine significantly reduced the medium LDH in NMDA-treated neurons but failed to reduce the release of LDH induced by staurosporin or calcimycin. In PC12 cells, agmatine significantly reduced LDH release induced by glutamate exposure, but not by staurosporine or calcimycin. Agmatine itself neither increased LDH release nor directly inhibited the enzyme activity. CONCLUSION: We conclude that agmatine protects against NMDA excitotoxicity in neurons and PC12 cells but not the cell death induced by protein kinase blockade or increase in cellular calcium.

Differential expression of alpha2-adrenoceptor vs. imidazoline binding sites in postmortem orbitofrontal cortex and amygdala of depressed subjects.

Clonidine is a well established antihypertensive agent that is also used effectively to treat a variety of psychiatric disorders. Clonidine is a prototypic imidazoline compound that acts as an alpha(2)-adrenergic agonist but possesses nearly equivalent affinity for non-adrenergic imidazoline binding sites (I-sites). Receptor autoradiography of [(3)H]-clonidine binding presented herein compares densities of alpha(2)-adrenoceptors and I-sites (under a noradrenergic-mask) in Brodmann's area 47 of the left orbitofrontal cortex (OFC) and in six amygdaloid nuclei of subjects with major depression (n=12) vs. controls with no psychiatric history (n=11). Postmortem diagnoses were made from psychiatric interviews with next-of-kin. [(3)H]-Clonidine binding to alpha(2)-adrenoceptors in each of six OFC layers was lower, although not reaching statistical significance in any one layer by multivariate analysis, in depressives vs. control subjects. Binding to I-sites was conversely higher in depressives compared to control OFC layers, but did not reach statistical significance alone. However, the ratios of alpha(2)-adrenoceptor : I-sites in all six layers of OFC of depressed subjects were nearly half that of control subjects (P<0.008). In amygdalas from a different group of depressed patients there were no changes in alpha(2)-adrenoceptors or I-sites, or their ratios, compared with controls. The results support previous western blot data indicating a cortex-selective shift away from alpha(2)AR towards I-site preponderance in depressed patients.

Non-adrenergic exploratory behavior induced by moxonidine at mildly hypotensive doses.

Moxonidine is a centrally-active imidazoline compound with preferential affinity for imidazoline receptors (IR) over alpha(2)-adrenoceptors (alpha(2)AR). Clinically, moxonidine has proven advantageous for treating hypertension over pure alpha(2)-adrenergic agonists (i.e., guanabenz) due to its lowered incidence of sedative side effects. The present experiments reveal divergent behavioral effects of low doses of moxonidine and guanabenz in C57Bl/6 mice in an exploratory arena. Low-dose moxonidine (0.05 mg kg(-1) i.p.) elicited an increase in novel object contacts (+36%) and more movement into central space (+56%; P<0.01) compared to saline-injected controls; whereas guanabenz induced only dose-responsive sedative-like behaviors in the same paradigm. Yet, the two agonists were indistinguishable in terms of blood pressure changes over a similar dose range (0.025-0.1 mg kg(-1) i.p.) in consciously free-moving mice (Delta mean+/-S.E.M.=-12.3+/-3.2 mm Hg for moxonidine versus -13.5+/-1.9 mm Hg for guanabenz). As expected of alpha(2)AR involvement, the sedative-like effects of guanabenz were completely blocked by pretreatment with the non-imidazoline alpha(2)AR-antagonist, SKF86466 (0.5 or 1.0 mg kg(-1) i.p.). However, the pro-exploratory effects of low doses of moxonidine (0.05 or 0.1 mg kg(-1)) were not antagonized by SKF86466. These results suggest that moxonidine acts preferentially through a non-adrenergic mechanism, possibly IR-mediated, to elicit pro-exploratory behavior.

Regulation of inducible nitric oxide synthase and agmatine synthesis in macrophages and astrocytes.

Agmatine is a novel endogenous guanido amine synthesized from arginine by arginine decarboxylase. Among several biologic effects, the ability of agmatine to protect against ischemic injury and chronic neuropathic pain is particularly interesting. Because inflammation is a common contributor to these conditions, we sought to determine if agmatine acts by decreasing the production of proinflammatory molecules such as nitric oxide and if agmatine synthesis is regulated by inflammatory stimuli. We tested whether agmatine affects astroglial and macrophage (RAW 264.7 cell line) nitric oxide synthase-2 (NOS-2) expression. NOS-2 was induced in these cells by incubation with lipopolysaccharide (LPS) plus three cytokines for astrocytes and LPS alone for RAW 264.7 cells in the presence and absence of varying concentrations of agmatine. NOS-2 activity was assessed after 24 hours by nitrite accumulation in the culture media. Agmatine dose-dependently inhibited nitrite accumulation, and shorter incubation with agmatine (1 and 4 hours) also caused significant reduction. Agmatine decreased the expression of NOS-2 activity and NOS-2 protein as determined by immunoblot analysis. Incubation of astrocytes and RAW 264.7 cells with LPS/cytokines for 2 hours resulted in an increase in arginine decarboxylase (ADC) activity, whereas longer-term incubation (12-17 hours) lowered ADC activity. Agmatine levels in these cells are increased after 6-hour incubation with LPS/cytokines. These results show that agmatine inhibits the production of nitric oxide by decreasing the activity of NOS-2 in macrophages and astroglial cells by decreasing the levels of NOS-2 protein. These findings provide a molecular basis for the neuroprotective and anti-inflammatory actions of agmatine.

Is agmatine an endogenous factor against stress?

Agmatine is an endogenous amine synthesized from the decarboxylation of arginine. A proposed intracellular role of agmatine is to balance the production of polyamines (a promitotic process) and nitric oxide (an inflammatory process). Agmatine is also released from neurons upon depolarization. We previously reported that agmatine concentrations are increased in rat pups' brains shortly after hypoxic-ischemia and in the plasma of depressed patients. Herein, male rats (270-290 g) were divided into four groups receiving different degrees of known stress: 2-hour restraint at 21 degrees C, 4-hour restraint at 21 degrees C, 4-hour restraint at 4 degrees C, and control rats only handled at 21 degrees C. Cortex, cerebellum, medulla, hippocampus, hypothalamus, and blood plasma samples were collected for determination of endogenous agmatine levels. No changes in agmatine levels were detected after 2-hour and 4-hour restraint at room temperature, but concentrations of agmatine were increased in all brain regions except cerebellum after 4-hour restraint in the cold. Plasma agmatine levels (ng/mL) were 6.8 +/- 0.6 in controls versus 58.1 +/- 12.8 in the 4-hour restraint-plus-cold group. Cortical agmatine levels (ng/g wet tissue) were 15.3 +/- 2.4 in controls versus 57.4 +/- 19.6 in the 4-hour restraint-plus-cold group. Therefore, endogenous agmatine was increased in response to cold-restraint stress, possibly as a neuroprotective agent.

Effect of postmortem delay on imidazoline receptor-binding proteins in human and mouse brain.

Immunoreactive proteins of 45-kD and 29/30-kD doublet bands are candidate imidazoline receptor binding proteins (IRBP) based on associations with I(1) or I(2) binding sites, respectively. It was reported that the density of cortical membrane 29/30-kD I(2) protein is diminished whereas a 45-kD I(1) protein is increased in depressed suicide victims versus controls. IRBP immunoreactive bands of similar size have been suggested to be breakdown products of the 170-kD protein known as IRAS (putative full-length I(1) receptor). This study compares nonpathologic human brains collected and frozen after postmortem delays of 13.4 hours +/- 1.7 (SEM) with brains of longer postmortem delays (26.1 hours +/- 1.2). The fresher human brains possessed more full-length IRAS (P = 0.05). In another study, the postmortem decay of IRBP bands in mouse brain was shown to be linear over time. The results are relevant to previous studies of IRBP bands in postmortem brains of depressed suicide victims.

IRAS is an anti-apoptotic protein.

Active cell death, also known as apoptosis, has been implicated in the pathophysiology of diseases such as cancer, heart failure and neurodegenerative disorders. We report the anti-apoptotic function of IRAS, which was previously shown to bind imidazoline ligands. The amino acid sequence of human IRAS (hIRAS) is unrelated to known proteins, except for rat IRAS and a mouse homologue named nischarin, which binds the alpha5 integrin subunit of the fibronectin receptor. When stably transfected into PC12 cells, hIRAS localizes to the cytosol as a 167 kDa immunoreactive protein. Clonal PC12 cell lines expressing hIRAS displayed normal serum growth responses. However, hIRAS expression led to prolonged cell survival against known apoptotic stimuli: serum starvation or thapsigargin or staurosporine treatments. The apoptotic population of hIRAS-expressing cells was significantly reduced, and this protection was achieved by a decrease in caspase-3 activity, phosphatidylserine translocation, and nuclear fragmentation. Similar protective effect was obtained in COS7 cells transiently transfected with hIRAS. A partial activation of the PI3 kinase pathway is possibly implicated in the anti-apoptotic effect of IRAS. Thus, IRAS appears to represent a previously unknown anti-apoptotic protein involved in the regulation of cell survival.

Intracellular effect of imidazoline receptor on alpha(2A)-noradrenergic receptor.

Imidazoline-1 receptors (I(1)R) and alpha(2)-noradrenergic receptors (alpha(2)AR) are known to coexist in many cell types and bind many of the same imidazoline ligands. Herein, the possibility of an interaction between these receptors was explored using a cloned cDNA that encodes a protein with I(1)R-like binding properties, designated imidazoline receptor antisera-selected (IRAS). Chinese hamster ovary (CHO) sublines permanently expressing the human subtype alpha(2A)AR cDNA were transiently cotransfected with the human IRAS cDNA (pIRAS). Saturation radioligand binding experiments on membranes isolated from the various sublines allowed distinction between I(1)R and alpha(2A)AR. Transfection of pIRAS into either subline led to a rise in membrane I(1)R-binding sites. Immunoblotting revealed that IRAS was enriched in membranes more than in cytosolic fractions. Transfection of pIRAS in CHO cells harboring the alpha(2A)AR cDNA resulted in a twofold increase in alpha(2A)AR binding sites with no change in alpha(2A)AR binding affinity, compared with controls. Immunoblotting also revealed increased expression of membranous alpha(2A)AR by IRAS. Thus, pIRAS transfection led to I(1) binding sites and to an increase in alpha(2A)AR binding sites in CHO cells expressing the human alpha(2A)AR. Although the mechanism is unclear, this increase in binding sites may explain previous imidazoline drug effects suggestive of interactions between these two receptors.

Relationship between platelet imidazoline receptor-binding peptides and candidate imidazoline-1 receptor, IRAS.

A candidate human imidazoline-1 receptor, designated imidazoline receptor antisera-selected (IRAS) protein, was cloned based on immunoreactivity with antiserum against a purified imidazoline receptor binding peptide (IRBP antiserum). Human IRAS is 167 kD in size, different from 33- to 85-kD IRBP bands previously linked to the human platelet I(1) receptor. To explore the possible relationship between IRAS and these smaller proteins, seven different epitope-specific antisera against IRAS were raised in rabbits for comparison with IRBP antiserum. Focus was on antiserum(227-241), corresponding to amino acids No. 227 to 241 in IRAS, because this antiserum was found uniquely able to immunoprecipitate non-denatured 85-kD and 170-kD forms of IRAS from a human megakaryoblastoma cell line (MEG01), a model of platelet-producing cells. Human platelets lacked the 170-kD form of IRAS, but 33-kD and 85-kD bands were detectable and seemed to be possible fragments of full-length IRAS. The intensity of the 85-kD band detected by antiserum(227-241) was significantly correlated (r = 0.62, P = 0.04) with the intensity of the 33-kD band across 11 human platelet samples. A positive correlation between the intensities of the 33-kD and 85-kD bands is consistent with both being fragments of IRAS.

Agmatine suppresses nitric oxide production and attenuates hypoxic-ischemic brain injury in neonatal rats.

Nitric oxide and excitatory amino acids contribute to hypoxic-ischemic brain injury. Agmatine, an endogenous neurotransmitter or neuromodulator, is an inhibitor of nitric oxide synthase and an antagonist of N-methyl-D-aspartate receptors. Does agmatine reduce brain injury in the rat pup hypoxic-ischemic model? Seven-day old rat pups had right carotid arteries ligated followed by 2.5 h of hypoxia (8% oxygen). Agmatine or vehicle was administered by i.p. injection at 5 min after reoxygenation and once daily thereafter for 3 d. Brain damage was evaluated by weight deficit of the right hemisphere at 22 d after hypoxia by a blinded observer. Agmatine treatments significantly reduced weight loss in the right hemisphere from -30.5 +/- 3.6% in vehicle-treated pups (n = 22) to -15.6 +/- 4.4% in the group treated with 50 mg/kg (n = 18, p < 0.05) and to -15.0 +/- 3.7% in the group treated with 100 mg/kg (n = 18, p < 0.05), but the group treated with 150 mg/kg showed no reduction. Other pups received agmatine or vehicle at 5 min after reoxygenation, and brain biochemistry was assessed. Levels of endogenous brain agmatine rose 2- to 3-fold owing to hypoxic-ischemic (3 h), whereas pups treated with agmatine (100 mg/kg) showed 50-fold higher brain agmatine levels (3 h). Agmatine (100 mg/kg) blocked a hypoxia-induced increase in brain nitric oxide metabolites at 6 h (vehicle-treated, +60.2 +/- 15.2%; agmatine-treated, +4.2 +/- 8.4%; p < 0.05). Agmatine thus reduces brain injury in the neonatal rat hypoxic-ischemic model, probably by blunting the rise in nitric oxide metabolites normally seen after hypoxia.