Introduction

Maple syrup urine disease (MSUD) (MIM 248600) is an autosomal recessive inherited disorder, with a worldwide incidence of 1:185,000 live births. It is caused by defects of the branched-chain α-ketoacid dehydrogenase (BCKDH, EC 1.2.4.4) complex, leading to accumulation in tissues and biological fluids of the branched-chain α-keto acids (BCKA), α-ketoisocaproic acid (KIC), α-keto-β-methylvaleric acid, and α-ketoisovaleric acid, and the equivalent amino acids leucine, isoleucine, and valine [1, 2]. MSUD is classified into four genetic variants, the classic, the intermediate, the intermittent, and the thiamine-responsive forms [1]. The classical form is the most severe, presenting during the neonatal period, sometimes called neonatal MSUD [3, 4].

Patients with classical neonatal MSUD usually manifest with nonspecific signs of metabolic intoxication, such as irritability, somnolence, anorexia, vomiting, dehydration, and ketoacidosis, which is followed by neurological symptoms, such as lethargy, coma, apnea, opisthotonos, and cytotoxic cerebral edema. Later in infancy or childhood, they develop physical and psychomotor delay with intellectual disability and movements disorders such as dystonia and tremor [1, 2]. Neuropathological findings comprise hypomyelination and cytotoxic edema in various brain areas, including the basal ganglia and cerebral cortex [5,6,7], and brain atrophy [8, 9]. Noteworthy, MSUD is considered an intoxicating inherited disorder in which KIC and leucine are the main neurotoxic metabolites, once augmented plasma concentrations of these compounds are associated with the appearance or worsening of the neurological symptoms that take place during crises of metabolic decompensation accompanied by excessive proteolysis [1, 2]. Although various mechanisms have been proposed to explain the cerebral damage in MSUD, the exact underlying pathophysiological mechanisms are still poorly known.

Regarding the BCKA, intracerebroventricular (icv) administration of KIC to adult (60-day-old) or young (30-day-old) rats was shown to cause learning deficits and behavioral changes [10]. KIC icv injection was also shown to induce oxidative damage to biomolecules [11], decrease of brain-derived neurotrophic factor and nerve growth factor levels [12], as well as of the activities of the respiratory chain complexes I and II-III [13] in different cerebral regions. KIC was also shown to alter the levels of interferon-γ and tumor necrosis factor-α [14].

Furthermore, in vitro exposition of brain to the BCKA, particularly to KIC, caused lipid oxidative damage and impairment of the antioxidant defenses [15, 16], as well as inhibition of α-ketoglutarate dehydrogenase and pyruvate dehydrogenase activities [17, 18]. KIC also reduced oxygen consumption in state 3 respiration and disrupted mitochondrial membrane potential [19]. Another work showed that the BCKA accumulating in MSUD inhibit CO2 formation and increase lactate release in the brain of young rats [20].

On the other hand, several studies revealed that in vivo and in vitro exposition of rat brain to the BCAA induce oxidative stress through lipid peroxidation, DNA, and protein damage, and altered activities of antioxidant enzymes [21,22,23,24]. BCAA also decreased CO2 production and the activities of creatine kinase, and respiratory chain complexes II–III, III, and IV [18, 25, 26]. Of note, lipid and protein oxidative damage was found in the plasma of MSUD patients [27,28,29,30], corroborating the investigations with the chemical animal models of this disease. Noteworthy, all these previous studies designed to clarify MSUD pathogenesis were performed in the brain of young or adult rats, whereas the onset of the neurological findings in MSUD patients is usually manifested in the first few days of life. Therefore, further studies using neonatal brain are necessary to clarify the pathophysiology of the brain damage in this disease.

Therefore, since to the best of our knowledge, there were no reports on the deleterious effects of KIC on brain functioning in early days of development, the present work evaluated the ex vivo effects of intracerebral administration of KIC to neonate rats on redox homeostasis, as well as on neuropathological landmarks in the cerebral cortex and striatum. The antioxidants N-acetylcysteine (NAC) and melatonin (MEL) were used in some experiments to evaluate their neuroprotective effects on redox homeostasis.

Material and Methods

Animals and Reagents

A total of 84 1-day-old (neonate) Wistar rats from Centro de Reprodução e Experimentação de Animais de Laboratório (CREAL) at the Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, Brazil, were used in this work. The newborn rats were kept with their dam on a 12:12-h light/dark cycle at temperature of 22 ± 1 °C. Free water and commercial chow with 20% (w/w) protein were available to dams.

All reagents were purchased from Merck (St. Louis, MO, USA). On the day of the experiments, KIC, MEL, and NAC solutions were made, and the pH adjusted to 7.4.

icv Injection of KIC

A schematic representation of the experimental design is shown in Fig. 1. Neonate rats were submitted to an icv injection of KIC (1 μmol/g body weight, dissolved in 0.01 M phosphate-buffered saline (PBS), pH 7.4, in a volume of 0.5 μL/g body weight), or an equal volume of 0.01 M PBS, pH 7.4 (control), into the cisterna magna, as previously described [31, 32]. After recovering in a heating pad at 37 °C, neonate animals returned to their dam until euthanasia that was immediately followed by the various analyses. In some experiments, neonatal rats were pre-treated with two intraperitoneal injections (volume of 2 μL/g body weight) of NAC (300 mg/kg and 400 mg/kg, respectively) or MEL (40 mg/kg, each) 24 and 1 h prior to KIC icv administration [32, 33]. The animals were then euthanized at 6 h after KIC icv infusion to evaluate redox homeostasis parameters and heme oxygenase-1 (HO-1) protein immunocontent, at postnatal day (PND) 7 to determine oxidative stress measures, or at PND 15 for the immunofluorescence analyses. It is stressed that the neonates were randomly distributed into the various experimental groups, the researchers that conducted the experimental procedures did not know which solution was administered to the animals (PBS, KIC, NAC, or MEL), and the groups were not identified during the analyses. The sample size and power calculations for the biochemical assays, Western blotting, and immunofluorescence analysis utilized the software Minitab 16, and considered the following parameters: standard deviation of 10%, power of 0.9, difference of 25%, and 2 or 3 levels (experimental groups). These parameters were based on previous reports from our and other research groups [32, 34, 35].

Fig. 1
figure 1

Schematic representation of the experimental design showing the procedures performed in Wistar rats from postnatal day 1 (PND 1) to postnatal day 15 (PND 15). HO-1, heme oxygenase-1; icv, intracerebroventricular; ip, intraperitoneal; KIC, α-ketoisocaproic acid; MEL, melatonin; NAC, N-acetylcysteine; PND, postnatal day; PBS, phosphate-buffered saline

Full size image

Parameters of Redox Homeostasis

Preparation of Tissue Homogenates

The animals were euthanized by decapitation, the brain removed, and the cerebral cortex and striatum dissected. These cerebral structures were then homogenized (1:10) in 20 mM sodium phosphate buffer, pH 7.4, with 140 mM KCl, centrifuged at 800 g and the supernatants used for the determination of redox homeostasis parameters, excepting for the nitrate and nitrite levels that were measured with the forebrain supernatants.

2',7'-Dichlorofluorescein (DCFH) Oxidation

Reactive oxygen species (ROS) production was measured by DCFH oxidation, as described by LeBel et al. [36]. Tissue homogenates were incubated for 30 min at 37 °C with the cell-permeant diacetate form (DCF-DA), which is cleaved to generate DCFH that is oxidized by ROS to form the fluorescent product DCF. Fluorescence was measured in a Spectramax M5 at 480 nm (excitation) and 535 nm (emission). A standard curve was carried out with DCF (0–10 μM) and the data expressed as μmol DCF per milligram protein.

Nitrate and Nitrite Levels

Reactive nitrogen species (RNS) levels were evaluated by the determination of the nitric oxide derivatives nitrate (NO3−) and nitrite (NO2−), according to the method described by Navarro-Gonzálvez et al. [37]. Tissue homogenates were deproteinized by a solution of 75 mM ZnSO4, centrifuged, and the supernatant neutralized with a solution of 55 mM NaOH. Then, all nitrates were converted to nitrite by the addition of copper-coated cadmium granules. Next, an aliquot of the samples was treated with the Griess reagent and incubated for 10 min at room temperature in the dark. The absorbance was determined in a Spectramax M5 microplate reader at 505 nm. A standard curve was performed with sodium nitrite (2.5–100 μM). Results were expressed as μmol nitrates and nitrites per milligram protein.

Malondialdehyde (MDA) Levels

MDA levels were determined by measuring thiobarbituric acid reactive substances, according to the method of Yagi [38]. After pre-incubation, tissue homogenates were incubated for 1 h at 100 °C in the presence of 10% trichloroacetic acid (TCA) and 0.67% thiobarbituric acid. The resultant pink color was extracted with butanol and separated by centrifugation. The fluorescence of the organic phase was measured at 515 nm (excitation) and 553 nm (emission). A standard curve was made by using 1,1,3,3-tetramethoxypropane. Results were expressed as nmol MDA per milligram protein.

Carbonyl Content

In order to determine the carbonyl content, tissue homogenates were treated with 2,4-dinitrophenylhydrazine (10 mM) and incubated for 1 h in the dark at room temperature. Proteins were precipitated with 20% TCA, the sediment obtained after centrifugation was washed with a mixture of ethanol:ethyl acetate (1:1, v/v) and resuspended in 6 M guanidine. The carbonyl content was spectrophotometric determined at 365 nm in a Spectramax M5 [39]. Results were expressed as nmol carbonyl content per milligram protein.

Reduced Glutathione (GSH) Levels

Tissue homogenates reacted with o-phthaldialdehyde (1 mg/mL) and were incubated for 15 min at room temperature in the dark. Next, the fluorescence was determined in Spectramax M5 microplates at 350 nm (excitation) and 420 nm (emission) [40]. GSH levels were calculated from a standard curve of GSH (0.001–1 mM) as nmol GSH per milligram protein.

Glutathione Peroxidase (GPx) Activity

GPx activity was determined in tissue homogenates by the method of Wendel [41], using tert-butyl hydroperoxide as substrate. The decrease of NADPH absorbance at 340 nm was monitored in a Spectramax M5. GPx activity was expressed as unit per milligram protein.

Glutathione Reductase (GR) Activity

GR activity was measured in tissue homogenates according to the method described by Carlberg and Mannervik [42], using oxidized glutathione (GSSG) and NADPH as substrates. This enzymatic activity was determined by monitoring the decrease of NADPH absorbance at 340 nm in a Spectramax M5. GR activity was expressed as unit per milligram protein.

Superoxide Dismutase (SOD) Activity

SOD activity was measured in tissue homogenates according to the method of Marklund [43], which is based on the autoxidation of pyrogallol that can be measured by determining the absorbance at 420 nm in a Spectramax M5. A standard curve was made with commercially purified SOD (S8408, Merck) to calculate the SOD activity. The results were expressed as unit per milligram protein.

Western Blotting

For the determination of HO-1 protein immunocontent, animals were euthanized by decapitation; the cerebral cortex was dissected, homogenized, and centrifuged; and the supernatants were prepared for the analyses, as described by Seminotti et al. [33]. The primary antibody anti-HO-1 (1:500, Abcam ab13248) was used and the results expressed as a ratio comparative to β-actin (1:20000, Merck A1978).

Neurochemical Markers

The neurochemical markers were measured by immunofluorescence. A transcardiac perfusion with 0.9% saline containing 0.16% sodium citrate was first carried out in the anesthetized rats and followed by a second perfusion with 4% paraformaldehyde, after which the brain was removed and coronal slices (35 μm) prepared using a Vibratome (VT1000S; Leica, Nussloch, Germany). Three transverse brain slices containing the cerebral cortex and striatum were obtained from each animal and the immunostaining procedures were performed according to Seminotti et al. [33]. An overnight incubation at 4 °C with the following primary antibodies were then performed: mouse anti-neuronal nuclear protein (NeuN) (1:400, Millipore #MAB377), rabbit anti-glial fibrillary acidic protein (GFAP) (1:500, Thermo Fisher # MA5-12,023), mouse anti-myelin basic protein (MBP) (1:500, Merck #AMAb91064), and rabbit anti-2′,3′-cyclic-nucleotide 3′-phosphodiesterase (CNPase) (1:500, Cell Signaling #D83E10). Next, a 90-min incubation at room temperature with a secondary mouse or rabbit antibody was carried out (1:500, Thermo Fisher Scientific). Slices were then assembled with fluoroshield (Merck) and the images were captured by an Olympus FV300 confocal microscope using 488 and 546 nm wavelengths. NeuN-positive cells and the fluorescence intensity of GFAP, MPB, and CNPase were assessed in three randomly chosen fields per slice.

Determination of Protein Concentrations

Protein concentrations were measured by the method of Lowry and collaborators [44] with bovine serum albumin as standard.

Statistical Analysis

The normality of data distribution was initially assessed using the Shapiro-Wilk test, which showed a normal distribution pattern. Thus, the statistical analyses were one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test when more than two groups were compared, whereas Student’s t test for unpaired samples was used to compare two groups. Results were expressed as mean ± standard deviation and differences rated significant when P < 0.05. All analyses were made in the GraphPad Prism 9.0 software.

Results

Markers of Redox Homeostasis

Cerebral cortex and striatum of neonate rats were assayed to determine a large spectrum of redox homeostasis parameters 6 h after KIC icv injection (1 μmol/g) or at PND 7.

icv Administration of KIC to Neonate Rats Disrupts Brain Redox Homeostasis

We initially observed that icv KIC administration caused significant increases of DCFH oxidation (t(10) = 4.329, P < 0.01), MDA levels (t(10) = 4.769, P < 0.001), and carbonyl content (t(10) = 3.458, P < 0.01), besides a decrease of GSH levels (t(10) = 11.72, P < 0.001), and GR activity (t(10) = 3.649, P < 0.01) in the cerebral cortex of the neonate rats 6 h after icv injection. In contrast, SOD and GPx activities were not changed by this treatment (Fig. 2). Furthermore, increased DCFH oxidation (t(9) = 3.449, P < 0.01) and MDA levels (t(10) = 13.72, P < 0.001), and decreased GSH concentrations (t(10) = 2.4, P < 0.05) were observed in the striatum of neonate rats following KIC treatment (Fig. 3). Increased activities of GPx (t(10) = 7.641, P < 0.001), GR (t(10) = 6.389, P < 0.001), and SOD (t(10) = 3.235, P < 0.01) were also found in the striatum 6 h after KIC injection (Fig. 3). Finally, we found that nitrate and nitrite concentrations were not changed in the forebrain of neonate rats 6 h after the administration of the metabolite (Fig. 4). Taken together, these data clearly reveal that a single icv in vivo KIC injection causes short-lived oxidative stress in the brain of neonate rats probably secondary to increased ROS formation.

Fig. 2
figure 2

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on redox homeostasis in the cerebral cortex 6 hours after injection. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), carbonyl formation (C), reduced glutathione (GSH) levels (D), and glutathione peroxidase (GPx) (E), glutathione reductase (GR) (F), and superoxide dismutase (SOD) (G) activities were determined. Values are mean ± standard deviation of six independent experiments (N) expressed as μmol/mg protein, nmol/mg protein or U/mg protein. **P < 0.01, ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 3
figure 3

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on redox homeostasis in the striatum 6 hours after injection. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), reduced glutathione (GSH) levels (C), and glutathione peroxidase (GPx) (D), glutathione reductase (GR) (E), and superoxide dismutase (SOD) (F) activities were determined. Values are mean ± standard deviation of five to six independent experiments (N) expressed as μmol/mg protein, nmol/mg protein, or U/mg protein. *P < 0.05, **P < 0.01, ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 4
figure 4

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on nitrate and nitrite concentrations in the forebrain 6 h after injection. Values are mean ± standard deviation of six independent experiments (N) expressed as μmol/mg protein. No significant differences were detected (Student’s t test for unpaired samples)

Full size image

We also measured the effects of icv KIC injection to the neonate rats on redox homeostasis parameters in cerebral cortex and striatum at PND 7. We observed that KIC only significantly elevated SOD activity (t(8) = 6.392, P < 0.001) in cerebral cortex, with no alterations of DCFH oxidation, MDA and GSH concentrations, as well as of GPx and GR activities (Fig. 5). Regarding the striatum, we found that all antioxidant enzymes had their activities increased by this treatment (GPx: t(8) = 9.987, P < 0.001; GR: t(8) = 2.483, P < 0.05; SOD: t(8) = 4.249, P < 0.01), without any change in DCFH oxidation, MDA, and GSH levels (Fig. 6).

Fig. 5
figure 5

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on redox homeostasis in the cerebral cortex at postnatal day (PND) 7. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), reduced glutathione (GSH) levels (C), and glutathione peroxidase (GPx) (D), glutathione reductase (GR) (E), and superoxide dismutase (SOD) (F) activities were evaluated. Values are mean ± standard deviation of five to six independent experiments (N) expressed as μmol/mg protein, nmol/mg protein, and U/mg protein. ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 6
figure 6

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on redox homeostasis in the striatum at postnatal day (PND) 7. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), reduced glutathione (GSH) levels (C), and glutathione peroxidase (GPx) (D), glutathione reductase (GR) (E), and superoxide dismutase (SOD) (F) activities were evaluated. Values are mean ± standard deviation of five independent experiments (N) expressed as μmol/mg protein, nmol/mg protein, and U/mg protein. *P < 0.05, **P < 0.01, ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image

NAC and MEL Prevent the Alterations Induced by an icv Administration of KIC on Redox Homeostasis in Cerebral Cortex and Striatum of Neonate Rats

Next, we evaluated whether the antioxidants NAC and MEL (intraperitoneal injections 24 and 1 h before KIC administration) could prevent the alterations of the various redox homeostasis parameters caused by KIC administration 6 h after the injection. We found that MEL normalized DCFH oxidation, MDA levels, and GSH concentrations, whereas NAC totally prevented GSH decrease and the increase of MDA levels in the cerebral cortex (Fig. 7) (DCFH: F(3,16) = 7.359, P < 0.01; MDA: F(3,12) = 5.555, P < 0.05; GSH: F(3,16) = 7.861, P < 0.01). Furthermore, MEL fully prevented DCFH oxidation, MDA, and GR activity increase, but did not change KIC-elicited increase of GPx, and exacerbated KIC-induced increase of SOD activity in the striatum. In turn, NAC normalized GPx activity, exacerbated KIC-induced SOD activity increase and did not change KIC-induced alterations on DCFH oxidation, MDA levels and GR activity (Fig. 8) (DCFH: F(3,12) = 6.313, P < 0.01; MDA: F(3,10) = 18.67, P < 0.001; GPx: F(3,17) = 94.67, P < 0.001; GR: F(3,16) = 23.55, P < 0.001; SOD: F(3,15) = 57.63, P < 0.001).

Fig. 7
figure 7

Effects of the antioxidants N-acetylcysteine (NAC) and melatonin (MEL) on the alterations of redox homeostasis parameters caused by an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats in the cerebral cortex 6 h after the injection. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), and reduced glutathione (GSH) levels (C) were determined. Animals received two intraperitoneal injections of NAC (300 mg/kg and 400 mg/kg) or MEL (40 mg/kg) 24 and 1 h prior to KIC icv administration. Values are mean ± standard deviation of four to six independent experiments (N) expressed as μmol/mg protein and nmol/mg protein. *P < 0.05, **P < 0.01 compared to control. #P < 0.05, ##P < 0.01 compared to KIC (one-way ANOVA, followed by post hoc Tukey’s range test)

Full size image
Fig. 8
figure 8

Effects of the antioxidants N-acetylcysteine (NAC) and melatonin (MEL) on the alterations of redox homeostasis parameters caused by an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats in the striatum 6 h after the injection. 2′,7′-Dichlorofluorescein (DCFH) oxidation (A), malondialdehyde (MDA) levels (B), reduced glutathione (GSH) levels (C), and the activities of glutathione peroxidase (GPx) (D), glutathione reductase (GR) (E), and superoxide dismutase (SOD) (F) were determined. Animals received two intraperitoneal injections of NAC (300 mg/kg and 400 mg/kg) or MEL (40 mg/kg) 24 and 1 h prior to KIC icv administration. Values are mean ± standard deviation of three to six independent experiments (N) expressed as μmol/mg protein, nmol/mg protein, or U/mg protein. *P < 0.05, **P < 0.01, ***P < 0.001 compared to control. ##P < 0.01, ###P < 0.001 compared to KIC (one-way ANOVA, followed by post hoc Tukey’s range test)

Full size image

icv Administration of KIC to Neonate Rats Induces HO-1 Production in Cerebral Cortex

To find out a possible signaling mechanism involved in the KIC-induced oxidative stress, we measured HO-1 protein levels, an enzyme with cytoprotective effects on oxidative stress by degrading free heme [45], in cerebral cortex of the KIC-treated neonate rats 6 h after injection. Figure 9 shows that the levels of HO-1, which are controlled by the nuclear factor erythroid 2-related factor 2 (Nrf2), were markedly increased after KIC administration (t(10) = 9.064, P < 0.001).

Fig. 9
figure 9

Effect of an intracerebroventricular (icv) administration of α-ketoisocaproic acid (KIC, 1 μmol/g) to neonate rats on heme oxygenase-1 (HO-1) immunocontent in cerebral cortex 6 h after injection. Values are mean ± standard deviation of three to six independent experiments (N) expressed as arbitrary units. ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image

Neurochemical Markers Measured by Immunofluorescence

We also measured the effects of icv administration of KIC (1 μmol/g) to neonate rats on NeuN, GFAP, MBP, and CNPase immunofluorescence that indicate neuronal number, reactive astrogliosis, and myelination, respectively, in cerebral cortex and striatum at PND 15.

icv Administration of KIC to Neonate Rats Provokes Long-Term Neuronal Loss, Astrocytic Reactivity, and Myelination Impairment at PND 15

Next, we investigated whether an icv administration of KIC to neonate rats could alter the immunofluorescence staining of neuronal (NeuN), astrocytic (GFAP), and myelin (MPB and CNPase) proteins at PND 15. It was found that KIC significantly decreased NeuN (cerebral cortex: t(4) = 3.301, P < 0.05; striatum: t(4) = 12.04, P < 0.001) (Fig. 10) and increased GFAP (cerebral cortex: t(5) = 2.829, P < 0.05; striatum: t(5) = 2.087, P < 0.05) (Fig. 11) immunofluorescence staining. In addition, MBP (striatum: t(6) = 2.026, P < 0.05) (Fig. 12) and CNPase (striatum: t(6) = 2.870, P < 0.05) (Fig. 13) staining were significantly reduced in the striatum of these animals. These results suggest long-term neuronal loss, astrocyte reactivity, and myelination injury in the brain following a single administration of KIC in the neonatal period.

Fig. 10
figure 10

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on NeuN immunofluorescence in the cerebral cortex and striatum at postnatal day (PND) 15. Images are representative of three independent experiments (N). Quantitation of NeuN-positive cells was performed on 400× magnification images using the mean of three randomly selected fields of each brain structure per slice. Data were obtained using three slices (containing cerebral cortex and striatum) per rat brain from three animals in each experimental group and are expressed as mean ± standard deviation of NeuN-positive cells. *P < 0.05, ***P < 0.001 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 11
figure 11

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on GFAP immunofluorescence in the cerebral cortex and striatum at postnatal day (PND) 15. Images are representatives of three independent experiments (N). Quantitation of GFAP ratio intensity was performed on 400× magnification images using the mean of three randomly selected fields of each brain structure per slice. Data were obtained using three slices (containing cerebral cortex and striatum) per rat brain from three to four animals in each experimental group and are expressed as mean ± standard deviation of GFAP ratio intensity (% of control). *P < 0.05 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 12
figure 12

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on MBP immunofluorescence in the cerebral cortex (A) and striatum (B) at postnatal day (PND) 15. Images are representatives of four independent experiments (N). Quantitation of MBP ratio intensity was performed on 200× magnification images using the mean of three randomly selected fields of each brain structure per slice. Data were obtained using three slices (containing cerebral cortex and striatum) per rat brain from four animals in each experimental group and are expressed as mean ± standard deviation of MBP ratio intensity (% of control). *P < 0.05 compared to control (Student’s t test for unpaired samples)

Full size image
Fig. 13
figure 13

Effects of an intracerebroventricular (icv) administration of KIC (1 μmol/g) to neonate rats on CNPase immunofluorescence in the cerebral cortex (A) and striatum (B) at postnatal day (PND) 15. Images are representatives of four independent experiments (N). Quantitation of CNPase ratio intensity was performed on 200× magnification images using the mean of three randomly selected fields of each brain structure per slice. Data were obtained using three slices (containing cerebral cortex and striatum) per rat brain from four animals in each experimental group and are expressed as mean ± standard deviation of CNPase ratio intensity (% of control). *P < 0.05 compared to control (Student’s t test for unpaired samples)

Full size image

Discussion

MSUD is a severe life-threatening inherited metabolic disease clinically characterized by neurological symptoms that worsen during crises of metabolic decompensation usually triggered by infections and other stressful conditions associated with fasting. These episodes are accompanied by a huge elevation in the levels of the BCKA and BCAA, mainly KIC and leucine, which are thought to be the major neurotoxic metabolites in MSUD [1, 2, 46]. MSUD children that survive the initial episodes of metabolic decompensation during the first weeks of life usually present long-term neurological sequelae, such as cortical atrophy and basal ganglia alterations, whose pathogenesis is still poorly established [8, 47]. In this particular, there is evidence suggesting that oxidative stress and neuroinflammation are associated with MSUD neurodegeneration [27,28,29,30, 48,49,50]. In line with this, various studies found that the metabolites accumulating in this disease disturb redox homeostasis and cause mitochondrial dysfunction in vivo in brain of young and adult rats [11, 13, 18, 22,23,24, 26]. Furthermore, in vivo icv injection of KIC to 30-day-old rats showed to cause oxidative stress [11], decrease of neurotrophic factors [12], and of respiratory chain complexes activities [13], as well as alterations of pro-inflammatory cytokines levels [14] in various cerebral regions. It is emphasized that most of these experimental works testing the effects of the BCAA and BCKA were carried out using 30-day-old or older rats. Investigating the role of these metabolites on neonatal brain seems therefore justified since MSUD patients manifest with neurological dysfunction in the first days of life.

Thus, the present investigation assessed biochemical markers of oxidative stress and neurochemical landmarks by immunofluorescence staining aiming to determine the number of neurons, astroglial reactivity, and myelination in the cerebral cortex and striatum following an acute cerebral overload of KIC to neonate rats in an effort to clarify the mechanisms responsible for brain injury in MSUD at this age. It is stressed that KIC can easily penetrate into the brain through the monocarboxylate transporter (MCT/SLC16A1) localized in the blood-brain barrier [51,52,53], being therefore able to accumulate in the central nervous system. We found that KIC caused short-term oxidative stress, as well as long-standing neurochemical alterations indicative of neuronal loss, astrogliosis, and hypo/demyelination in the brain of the treated animals. We also explored whether the antioxidants MEL and NAC could be protective against KIC-induced oxidative stress in these cerebral structures.

We initially found that KIC increased ROS generation (DCFH oxidation), induced lipid (MDA levels) and protein (carbonyl content) oxidative damage and decreased the GSH levels and GR activity in the cerebral cortex of the neonate rats. Similar findings were obtained in the striatum of these animals, except that all antioxidant enzymes evaluated (GPx, GR, and SOD) had their activities increased. It was also seen that nitrate and nitrite values were not changed by this treatment, ruling out a possible action of RNS in these effects. Noteworthy, since MDA is a lipoperoxidation-derived aldehyde, increases of its levels are indicative of lipid peroxidation, which may lead to severe cell damage, particularly to the brain that is rich in polyunsaturated fatty acids highly vulnerable to oxidation [54, 55]. With respect to the carbonyl groups, they result from the oxidation of amino acid side chain residues, indicating protein oxidation that is potentially harmful to cellular physiology [56, 57].

Regarding the decreased levels of GSH, the primary non-enzymatic antioxidant defense of the central nervous system that is critical to the clearance of free radicals [58], it is feasible that elevated levels of ROS could oxidize GSH, forming GSSG, but this should be confirmed since the oxidized form of GSH was not measured here. We might also suggest that the low capacity to prevent oxidative damage (GSH decrease) caused by KIC in the main structures injured in MSUD, may have contributed to lipid and protein oxidation. In turn, the decreased activity of GR in the cerebral cortex, which is a crucial enzyme to restore GSH concentrations, may also have a role in the diminution of GSH levels. Otherwise, we do not know yet the exact mechanisms for the increased antioxidant enzyme activities in the striatum following KIC injection, but our data showing that this organic acid also increased HO-1 protein content suggest that KIC leads to the upregulation of antioxidant defense genes possibly in an attempt to overcome the redox imbalance generated in this cerebral structure, reflecting therefore a cytoprotective strategy [59, 60].

With respect to SOD, since this enzyme converts the superoxide radical to hydrogen peroxide mainly to preserve GSH [61, 62], it is possible that SOD increased activity could result from its transcriptional overproduction to combat the increase in the anion superoxide radical [63]. The same reasoning could explain GPx activity augment in the striatum, to counterattack hydrogen peroxide species.

Since oxidative stress represents an unbalance between reactive species generation and antioxidant defenses [63], our present findings indicate that KIC induces oxidative stress in the brain of neonate rats, which to our mind was not yet described at this early development stage.

Our results also revealed that pre-treatment of neonate rats with the antioxidants NAC and MEL before KIC injection prevented or attenuated most of the altered biochemical markers of redox homeostasis. Therefore, we presume that particularly NAC, which is used in other pathologies to restore cell redox homeostasis [64,65,66], could serve as an adjuvant therapy to avoid, or ameliorate acute or chronic oxidative damage to critical cell macromolecules in MSUD.

In this regard, NAC is an effective antioxidant, neutralizing ROS and RNS [67]. This compound has a direct effect due to its free thiol group and it is an essential substrate for the enzyme γ-glutamylcysteine synthetase in the synthesis of GSH [68]. NAC has also anti-inflammatory properties by suppressing the nuclear factor κ B (NF-κB), diminishing the production of tumor necrosis factor-α, interleukin-1, and interleukin-6 (pro-inflammatory cytokines) [69].

MEL, a neurohormone secreted by the pineal gland, acts primarily on circadian rhythms [70], but has been shown to have neuroprotective effects [71, 72] mainly attributed to its antioxidant activity and improvement of mitochondrial function [70, 73,74,75,76,77,78,79,80]. In this context, MEL was demonstrated to mitigate oxidative stress induction by regulating the Nrf2 signaling pathway [81,82,83], which is mainly responsible for modulating the expression of many antioxidant genes [84].

Another interesting observation of this work was that oxidative stress induction following KIC injection at neonatal period was much less evident at PND 7, implying a short-lived effect of this metabolite on brain redox status.

The present data obtained very early during rat development, at an age where severe neurological symptoms appear in humans, are in agreement with previous results that reported increased MDA levels, protein carbonylation, DCFH oxidation, DNA oxidative damage, and SOD activity induced by KIC icv administration in the brain of young rats [11, 13], as well as with in vitro studies demonstrating that KIC disturbs brain redox homeostasis in young rats [15], C6 astroglial cells [16], human leukocytes [85, 86], and hippocampal neuronal cells [13]. Oxidative damage to lipids and proteins has been also shown in plasma of MSUD patients [27,28,29,30], corroborating with the in vivo and in vitro animal studies.

We also evaluated whether the injection of KIC to neonate rats could provoke neuronal damage, astrocyte reactivity, and myelination injury in striatum and cerebral cortex at PND 15 since there is no report on whether an acute intracerebral KIC overload during this period of life could cause histopathological alterations along development.

KIC administration led to a long-standing (PND 15) neuronal damage or/and loss (decreased NeuN positive cells) and astroglial reactivity (elevation of the astrocytic marker GFAP) in the cerebral cortex and striatum, whereas significant decreases of the immunoreactivity of the myelin proteins MBP and CNPase, indicating myelination injury, were found in the striatum.

It is emphasized that a moderate activation of astrocytes usually represents a mechanism of protection to surrounding neurons, whereas exacerbated glial reactivity may be deleterious to neurons, as evidenced in various neurodegenerative pathologies [87,88,89]. Noteworthy, GFAP, which is the main protein constituting astrocyte intermediate filaments, has been considered a classic marker of reactive astrocytes [90, 91]. In this scenario, the present investigation also showed a decrease of NeuN immunoreactivity in cerebral cortex and striatum, suggesting neuronal damage associated with glial reactivity in these cerebral structures that are mostly injured in MSUD.

In what concerns to myelination, we found a significant decrease of the content of sensitive protein markers of the myelination status in the striatum. A significant decrease in MBP immunoreactivity was observed in the striatum at PND 15, indicating myelin vulnerability secondary to neonatal intracerebral accumulation of KIC. Of note, the MBP is a sensitive marker of oligodendrocyte function and myelination status once it maintains myelin stability and integrity [92, 93]. In addition, MBP is involved in the maintenance of calcium and cell homeostasis [93]. The immunoreactivity of the CNPase protein, which is important in the regulation of oligodendrocyte process outgrowth [94, 95], was also significantly reduced in the striatum of these animals at PND 15, further indicating myelin injury in the rats neonatally injected with KIC. No significant effect of this treatment towards myelination was observed in the cerebral cortex, indicating that striatum is more vulnerable to myelin injury caused by an acute overload of KIC during the neonatal period that possibly mimics a metabolic episode of decompensation in MSUD affected neonates.

The present neurochemical findings corroborate with previous post-mortem studies carried out in MSUD patients showing defective myelination in various brain regions, including subcortical areas and basal ganglia [96, 97], as well as decrease of oligodendrocyte number in brain white matter [97,98,99], neuronal abnormalities in the cerebral cortex, and astrogliosis in the hippocampus and basal ganglia [100].

It should be also emphasized that the development and formation of a normal adult brain may be affected by injuries in the early postnatal period that lead to future neurologic symptoms and brain abnormalities. In this particular, it is interesting to note that all KIC-injected animals manifested with seizures immediately after its administration and recovered after approximately 5 min, in contrast to the PBS-injected rats (controls) that had no seizures at all. These observations revealed that our model of MSUD was toxic to the animals. Based on the data of the present investigation, we postulate that intracerebral toxic levels of KIC at the neonatal period may imbalance cellular redox status, disturbing cellular functions, and ultimately leading to neuronal loss, astrocyte reactivity, and myelination impairment.

We cannot, however, rule out that other pathomechanisms of brain damage are caused by increased concentrations of KIC or of the other metabolites accumulated in MSUD, including mitochondrial dysfunction and increased inflammatory response [46], that may also play an important role in MSUD neuropathology. In this context, we highlight that oxidative stress and neuroinflammation are inter-connected pathogenic processes [101], so that it is feasible that increased pro-inflammatory cytokines in the neural cells might also contribute to the deleterious effects caused by KIC. Indeed, elevation of these inflammatory biomarkers was demonstrated in MSUD patients [48, 50] and in animal models of this disease [102, 103]. However, further studies are required to corroborate this hypothesis.

In conclusion, the findings obtained in this study indicate that KIC-induced oxidative stress early during development may represent a central deleterious mechanism involved in neurological injury in MSUD and that antioxidants may represent a potential new therapeutic strategy for patients affected by this disease. We emphasize that the main purpose of our experimental model using a single icv administration of KIC was to mimic the acute episodes of metabolic decompensation that take place during stressful situations in MSUD patients and are accompanied by high elevation of the brain levels of the accumulating metabolites, particularly KIC. It is also of note that in the affected patients KIC produced peripherally may enter the brain from the systemic circulation, being transported through the blood-brain barrier where it accumulates [53]. Furthermore, this organic acid can be also produced from leucine and accumulate inside the brain cells of MSUD patients, because of the deficient activity of the BCKDH complex in this tissue [53].