Hydroxyl radical (·OH) is a reactive oxygen species produced during severe hypoxia, asphyxia, or ischemia that can cause cell death resulting in brain damage. Generation of ·OH may occur in the fetal brain during asphyxia in utero. The very short half-life of ·OH requires use of trapping agents such as salicylic acid or phenylalanine for detection, but their hydroxylated derivatives are either unstable, produced endogenously, or difficult to measure in the small volume of microdialysis samples. In the present study, we used terephthalic acid (TA), hydroxylation of which yields a stable and highly fluorometric isomer (excitation, 326 nm; emission, 432 nm). In vitro studies using ·OH generated by the Fenton reaction showed that hydroxylated TA formed quickly (<10 s), was resistant to bleaching (<5% change in fluorescence), and permitted detection of <0.5 pmol ·OH. In vivo studies were performed in fetal sheep using microdialysis probes implanted into the parasagittal cortex. The probe was perfused at 2 μl/min with artificial cerebrospinal fluid containing 5 mM TA, and samples were collected every 30 min. Fluorescence measured in 10 μl of dialysate was significantly greater than in the efflux from probes perfused without TA. High-performance liquid chromotography analysis showed that the fluorescence in dialysis samples was entirely due to hydroxylation of TA. Thus this study shows that it is possible to use TA as a trapping agent for detecting low concentrations of ·OH both in vitro and in vivo and that low concentrations of ·OH are present in fetal brain tissue and fluctuate with time.
- terephthalic acid
- Fenton reaction
reactive oxygen species (ROS) are free radicals produced by most cells under normal and pathological conditions but particularly by activated neutrophils (4, 8, 20), microglia (15, 19, 34) and macrophages (9). High concentrations of ROS, particularly the highly reactive hydroxyl radical (·OH), induce lipid peroxidation (16), DNA and RNA fragmentation (38), and mitochondrial compromise that can lead to cell death (7, 31). Because ·OH has a very short half-life (e.g., nanoseconds), it is usually measured indirectly by using trapping agents in both in vitro and in vivo experiments (36). Salicylic acid and phenylalanine are the most commonly used trapping agents. Salicylic acid has a relatively high sensitivity for ·OH and is readily measured by electrochemistry (6). Hydroxylation of salicylic acid produces two isomers, viz., 2,3- and 2,5-dihydroxbenzonic acid (DHBA). Because 2,5-DHBA can be produced by in vivo enzymic reaction, the only reliable measurement to determine the change of ·OH is from 2,3-DHBA derived from exogenously administered salicylic acid. Further to this complication in hydroxylation products, salicylic acid also inhibits cyclooxygenase activity (2, 17, 18, 24), which is itself an important source of ROS (29) and the means by which prostaglandins are produced. Phenylalanine has been suggested as an alternative to salicylic acid (37), but its low sensitivity and rate of hydroxylation have restricted its use. Furthermore, phenylalanine reacts with ·OH to produce meta-, para-, and orthotyrosine, and paratyrosine is produced endogenously and, therefore, cannot be used to report changes of ·OH production. Thus both methods are somewhat inaccurate and unreliable, particularly when the endogenous ·OH concentration is low.
Real-time measurement of ·OH in vivo has used microdialysis to deliver the trapping agent into the tissue and then to collect the hydroxylated product in the efflux from the probe. In this study, we have advanced the use of this technique to an in utero study of the fetal sheep brain, but it has the disadvantage that relatively long (>80 cm) lengths of inlet and outlet tubing must be used to traverse the uterine wall and maternal abdomen. The transit time of the sample from the tip of the probe to the collection vial is therefore quite long (∼1 h), necessitating the use of hydroxylation products that are very stable. Salicylic acid proved to be unsuitable for this purpose, leading us to search for an alternative compound that was both a sensitive and stable marker of endogenously produced ·OH.
In this study, we evaluate the use of terephthalic acid (TA) as a trapping agent for ·OH. TA has a configuration of two carboxylate anion (COO−) side groups attached to a six-carbon ring at positions 1 and 4 to form a structurally symmetrical compound (Fig. 1A). Reaction of ·OH at any one of the four unsubstituted carbons will form only one hydroxylated product, 2-hydroxy-TA (2-OH-TA). TA is nonfluorescent, whereas 2-OH-TA is highly fluorescent. Neither TA nor 2-OH-TA is present in tissues or is known to be involved in normal cellular functions. No toxicity has been reported regarding these compounds. Ortho-OH-TA fluorescence can be determined directly by using a microfluorometer, thus obviating the need to use high-pressure liquid chromatographic (HPLC) separation and electrochemical detection, which is necessary when salicylic acid and phenylalanine are used as the trapping agent. In this paper, we have demonstrated the advantage and potential use of TA as trapping agent for ·OH in both in vivo and in vitro experiments.
MATERIALS AND METHODS
All chemicals were of the highest grade available from commercial suppliers. TA, salicylic acid, and 2-bromoterephthalic acid were obtained from Sigma-Aldrich. Methanol, NaH2PO4·H2O, Na2HPO4, and H2O2 (30% vol/vol) were obtained from Merck, and FeSO4·7H2O, NaCl, KCl, HCl, NaOH, KOH, KH2PO4, H3PO4, sodium acetate, copper powder, and phenolphthalein were from ICN.
Because authentic 2-OH-TA is not a commercially available compound, it was synthesized according to Miura et al. (25) and Field and Englehardt (14) after minor variation. Ortho-bromoterephthalic acid (3.00 g, 12.2 mmol) and NaOH (0.98 g, 24.5 mmol) were dissolved in water (57 ml). To the solution were added sodium acetate (2.21 g, 26.9 mmol), copper powder (15.6 mg, 0.24 mmol), and a few drops of phenolphthalein. The mixture was heated to reflux and stirred for 24 h. Aqueous KOH (10% wt/vol) was added occasionally to keep the reaction mixture alkaline. On completion, the solution was cooled to room temperature and filtered, and the filtrate was acidified with 2 M HCl. The resulting precipitate was collected by filtration and washed with water. The product was dried to give 2-OH-TA (2.06 g, 92%) as a cream-colored solid and was characterized by proton nuclei magnetic resonance spectroscopy and mass spectrometry [melting point: 319.8–323.1°C (literature melting point: 320–322°C); 1H NMR (300 MHz, d6-dimethyl sulfoxide): δ 7.86 (doublet, 1H; J = 8.0 Hz, Ar-H), 7.46, 7.42 (doublet of doublets, 1H; J = 8.0, 1.6 Hz, Ar-H), and 7.42 (multiplet, 1H, Ar-H); 13C NMR (75 MHz, d6-dimethyl sulfoxide): δ = 171.0, 166.3, 160.5, 136.8, 130.7, 119.5, 117.7, 116.8; electrospray ionization mass spectrometry, mass-to-charge ratio 181 mass of parent ion (molecule) minus a proton].
In vitro Fenton reaction.
Free ·OH was generated by the well-known Fenton reaction. Equal molar concentrations of H2O2 and FeSO4·7H2O (1–30 μM) were mixed, with or without the addition of 1 ml of 5 mM TA in 0.05 M phosphate buffer (in mM: 12.5 NaH2PO4·H2O, 37.5 Na2HPO4, pH 7.4). The reaction mixture was incubated at 37°C for 15 min. At the end of incubation, a 10-μl aliquot was immediately injected directly into a microfluorometer equipped with an 8-μl quartz cell or onto the C18 column of an HPLC system equipped with the same fluorescent detector, as described below.
In vivo microdialysis study.
Five pregnant ewes (Merino-Leicester cross breed) at ∼125 days gestation (0.85 of term) were used in this study. The ewes were housed in individual cages with a constant environment of 22°C and a 12-h light-dark cycle. They were fed daily and given free access to water. All the experimental procedures had received prior approval from the Standing Committee on Ethics and Animal Experimentation of Monash University. Surgery was performed using halothane (Merial) general anesthesia. A microdialysis probe with a 4-mm polycarbonate membrane (molecular weight cut-off of 20,000; CMA/20, CMA Microdialysis, Sweden) was chronically implanted into gray matter (8-mm depth) of the parasagittal cortex of fetal sheep brain. The probe was inserted through a 1-mm hole drilled in the skull, with coordinates of 5 mm forward of the coronal suture and 10 mm lateral of the midline, and secured in place with tissue adhesive (Vetbond no. 1469, 3M) and covered with dental acrylic (Orthodontic Resin, Caulk Dentsply). After a recovery period of 3 days, the probe was perfused at 2 μl/min with modified artificial cerebrospinal fluid (aCSF; in mM: 12.5 NaH2PO4·H2O, 37.5 Na2HPO4, 90 NaCl, 4 KCl, pH 7.4) containing 5 mM TA. The perfusate was made freshly each time just before the experiment was commenced and was passed through a 0.22-μm filter before use. Osmolality was ∼308 mosM/l. The microdialysis probe was perfused for at least 6 h to allow equilibration before the collection of dialysate samples was commenced. Samples were then collected into ice-chilled Eppendorf tubes over 30-min intervals for the next 24 h. On the completion of each collection period, the dialysate was frozen on dry ice and kept at −70°C until analysis.
Determination of 2-OH-TA.
Fluorescence due to the presence of 2-OH-TA was determined in two ways. First, 10 μl of sample were injected directly into the microfluorometer cell. In this case, distilled deionized water (Milli-Q Water Purification System, Millipore) was pumped through the microfluorometer at 1 ml/min, and the sample was injected into the flow line from a manual injector (model 7125, Rheodyne) placed immediately before the cell. The total fluorescent intensity of the sample was measured at an excitation of 326 nm and emission of 432 nm. Second, samples were injected into a HPLC system consisting of a solvent delivery system (model LC1110, GBC), a manual injector (model 7125, Rheodyne), online degasser (model 3425, ERC), a C18 reverse-phase column (NOVA PAK C18, Waters), a fluorescent detector (model LC1250, GBC) and data collection software WinChrom (GBC). The mobile phase contained 50 mM KH2PO4 and 30% methanol, with the final pH adjusted to 3.2 with 1 M H3PO4, and was delivered at a flow rate of 0.8 ml/min. Ten microliters of sample were injected into a HPLC system with a total elution time of 20 min. Excitation and emission wavelengths of fluorescent detector were set at 326 and 432 nm, respectively. The concentration of 2-OH-TA in each sample was determined against a standard curve of 2-OH-TA generated from the authentic standard.
In vitro microdialysis probe recovery rate.
The recovery rate of the microdialysis probe was determined in vitro in both new probes and probes recovered from the fetal brain at postmortem, after implantation for 7–9 days. Each probe was perfused at 2 μl/min with modified aCSF (in mM: 12.5 NaH2PO4·H2O, 37.5 Na2HPO4, 90 NaCl, 4 KCl, pH 7.4) while immersed in the medium, which consisted of 50 mM PB and 1 μM 2-OH-TA standard. The probe was perfused for at least 30 min to equilibrate the system before the sample collection. Three samples were collected from the microdialysis probe efflux in 10-min fractions. The recovery rate was tested in the condition of medium, which was at both room temperature and 39°C. The sample and medium were analyzed for 2-OH-TA using HPLC, and recovery in the dialysate was expressed as a percentage of concentration in the medium.
Histology and immunohistochemistry.
Both ewe and fetus were killed using an overdose of pentobarbitone sodium (Lethabarb, Virbac, Australia) 24 h after the collection of microdialysis samples was started. The fetal brain was perfused in situ with ∼1 liter of cold saline and then with 1 liter of 4% paraformaldehyde in 0.1 M phosphate buffer (0.025 M NaH2PO4·H2O, 0.075 M Na2HPO4, pH 7.4). The brain was then removed from the skull and postfixed in 4% paraformaldehyde for 24 h before a block of cortex that contained the probe insertion site was removed, embedded in paraffin, and then cut into 10-μm-thick sections using a microtome. The sections were stained with cresyl violet (Sigma-Aldrich) and acid-fuchsin (Sigma-Aldrich) for general histology.
The sections adjacent to those used for cresyl violet and acid-fuchsin staining were processed for glial fibrillary acidic protein (GFAP) and lectin immunoreactivities to illustrate the astrocytic and other responses to tissue injury, respectively. The paraffinized sections were dewaxed in xylene and rehydrated in H2O. The sections were treated with 0.3% H2O2 to reduce endogenous peroxidase and were then incubated in mouse anti-GFAP antibody (Sigma-Aldrich; 1:500 dilution) or peroxidase-labeled lectin (Isolectin B4 from Bandeiraea Simplicifolia, Sigma-Aldrich; 1:200 dilution) overnight at 4°C. For GFAP staining, the sections were subsequently incubated with rabbit anti-mouse secondary antibody (DAKO; 1:500 dilution) and streptavidin horseradish peroxidase (Amersham Pharmacia Biotech; 1:200 dilution). For GFAP immunoreactivity, color was visualized by incubating the sections with diaminobenzidine (Pierce Chemical). For lectin staining, diaminobenzidine was directly added after overnight incubation for visualization. Between each incubation step, the sections were washed 3 × 5 min in phosphate-buffered saline (0.1 M, pH 7.4).
In vitro data.
Fluorescent excitation and emission wavelengths of 2-OH-TA have been identified in a previous study (35), where peak emission occurred at 432 nm for an excitation wavelength of 326 nm (Fig. 1B). In this study, scanning fluorometry of authentic 2-OH-TA standard confirmed that maximum emission occurred at excitation 326 nm but that the peak emission was closer to 440 nm for 2-OH-TA in phosphate buffer measured by direct fluorometry (Fig. 2B) or 420 nm when 2-OH-TA was resolved using HPLC (Fig. 2A). This difference was due to the presence of methanol in the HPLC mobile phase. In both systems, the emission scan for 2-OH-TA in microdialysis samples was virtually identical for that of authentic 2-OH-TA (Fig. 2).
Chromatographic analysis of microdialysis samples showed that all fluorescence in the sample could be accounted for by one peak, eluting at the same time (9 min) as authentic 2-OH-TA (Fig. 3A). There was a linear relationship between fluorescence intensity and the amount of 2-OH-TA (Fig. 3B), and the standard curve shows that at least 0.2 pmol ·OH can be detected when TA is used as the trapping agent. The in vitro recovery rate of 2-OH-TA using a new CMA/20 microdialysis probe perfused at 2 μl/min was 11.7% at room temperature (20–22°C) and 13.2% at 39°C (the body temperature of fetal sheep). Recovery was not different for probes retrieved from the fetal brain after 5–6 days of implantation (11.0% at 20–22°C).
Ortho-OH-TA was found to be very stable (<5% changes in fluorescence intensity) when stored at −70°C over a 5-mo testing period (Fig. 4A). Ortho-OH-TA was also very stable when left at room temperature for up to 6 h (Fig. 4B). In contrast, the hydroxylation product of salicylic acid (2,3-DHBA) was broken down to <40% of the initial concentration after storage at −70°C for up to 20 days (Fig. 4A) and increased significantly within 1.5 h and with considerable variation between samples when left at room temperature (Fig. 4B).
In vivo experiment.
The presence of 5 mM TA in the perfusate of microdialysis probes was sufficient to detect the basal ·OH concentration in the fetal brain (Fig. 5). This was confirmed by showing that the fluorescence of dialysate samples without TA in the perfusate was considerably less than when TA was present. Furthermore, as shown above, fluorescence in microdialysis samples had the same emission spectrum (Fig. 2), and with HPLC eluted at the time (Fig. 3A), as authentic 2-OH-TA. Because 2-OH-TA was the only source of fluorescence in the sample, it was possible to measure 2-OH-TA using the microfluorometer directly without a prior HPLC setup. When dialysate samples were first measured by direct microfluorometry and subsequently by the HPLC method, the total fluorescence and estimated 2-OH-TA concentration, respectively, and their variation over 20 h of continuous microdialysis were essentially the same (Fig. 5).
The location of the microdialysis probe tip in the brain is very important due to differences in cell type and cellular biochemistry of different brain regions. Only animals with the microdialysis probe tip located in gray matter were included in data analysis (n = 4). Figure 6 shows a section stained with cresyl violet and acid-fuchsin, GFAP, and lectin in an area immediately adjacent to where the tip of the microdialysis probe had been located. There was no evidence of cell death, as shown by the absence of pyknotic cells (Fig. 6, B and C). The appearance of the GFAP-positive cells was similar to that observed in the normal fetal brain, with no indication of astrogliosis seen in this or other animals (Fig. 6E). The density of GFAP-positive cells did not change and was similar in staining patterns to the adjacent areas. Lectin-positive cells, presumed to be microglia from their appearance, were also observed, but their number was not greatly different from that seen in the adjacent area and elsewhere in the brain (Fig. 6D).
·OH is a very reactive ROS produced during normal (27) and especially pathological conditions of cell respiration (1). The mechanism of ·OH-induced damage is mainly associated with oxidative damage to cellular components such as protein, lipid membrane, and nucleic acids (26). Three methods have been commonly used to study the cellular generation of ·OH: direct measurement with electron spin resonance (13), measurement of oxidized products such as lipid peroxidation and DNA fragmentation, and measurement of the free radical by a trapping agent.
The salicylate hydroxylation method has been widely used in the past two decades in both in vitro and in vivo settings and, among many applications, to investigate the effects on the brain of cerebral ischemia (21, 40), head injury (10), and neurological disease (11). However, the use of this trapping agent has recently been questioned (3, 12). Because salicylate inhibits cyclooxygenase activity, an important enzyme for prostaglandin production and a pathway with many physiological functions in both normal and disease conditions (33, 39), it has a significant limitation when prostaglandins are involved in responses to stress and trauma in the brain. This is almost certainly the case in the fetal and newborn brain, where prostaglandins are involved in the regulation of cerebral blood flow and neural activity that determine sleep and breathing activity (22). Furthermore, salicylate inhibits phospholipase C and interacts with several transition metal ions (5).
Another benzoic acid compound, phenylalanine, has been suggested as the alternative to salicylate because it has fewer identified side effects (37). However, chemokinetic studies show that hydroxylation of phenylalanine is very slow (37), and the amount of ·OH required to hydroxylate phenylalanine is rarely present in tissues, even in abnormal conditions. Furthermore, a cautionary note suggested that unknown compound(s) coeluted with m-tyrosine in ischemia-reperfusion experiments, thereby casting doubt on the accuracy of using this adduct for the measurement of ·OH in vivo (30).
The present study shows that TA has properties that make it a superior trapping agent for ·OH compared with either salicylic acid or phenylalanine. TA is not normally present in tissues, nor is it a reactant in biochemical processes. It is very reactive with ·OH with a faster hydroxylation rate compared with other trapping agents (35). Only one hydroxylated stereoisomer is formed when TA reacts with ·OH. Whereas TA is a nonfluorescent compound, 2-OH-TA is highly fluorescent, making it easy to distinguish between the parent molecule and the hydroxylated derivative. Fluorescence detection is technically much easier than the electrochemical methods required for detecting the hydroxylated derivatives of salicylic acid and phenylalanine. Furthermore, no other fluorescent moiety appears to be present in dialysates (from fetal brain, at least), obviating the need to use HPLC methods to measure 2-OH-TA. Ortho-OH-TA is free of bleaching for many hours when kept at room condition and stable for many months when stored at −70°C. The detection limit of 2-OH-TA is well above fluorescent levels present in microdialysis samples; 0.02–0.05 μM of 2-OH-TA is easily measured by fluorescence detection, which is above the level of ∼0.1 μM present in the late-gestation fetal sheep brain under resting conditions.
Chromatographic analysis showed that the fluorescence in microdialysis samples was present as a single peak, eluting at the same time as authentic 2-OH-TA. This raises the possibility that 2-OH-TA can be measured directly in dialysis samples without prior HPLC separation. Figure 5 shows the sample measured by both HPLC separation and direct fluorometry, which have shown an identical fluctuation of ·OH changes over the sampling period. Furthermore, samples collected from probes perfused with aCSF alone had a level of fluorescence lower than samples where TA was present in the perfusate. This shows that there is an in vivo reaction that increases fluorescence when TA is present and, for reasons given elsewhere, is likely to be due to the reaction of TA with ·OH. The fluorescence present in aCSF alone is autofluorescent, because no peak can be resolved by HPLC (Fig. 5).
Using TA as a trapping agent, we have shown that a period of 10 min of fetal asphyxia, induced by complete occlusion of the umbilical cord, significantly increased the efflux of 2-OH-TA from a microdialysis probe implanted in the cerebral cortex (23). This study also demonstrated the induction of lipid peroxidation (4-hydroxynonenal) and DNA oxidation (8-hydroxy-2′-deoxyguanosine) after the asphyxial insult (23). The increase in 2-OH-TA during or soon after the asphyxial insult demonstrates the sensitivity of the TA method to detect changes of ·OH production using microdialysis, and the increase in lipid peroxidation is consistent with the presence of oxidative stress. ·OH arises from hydrogen peroxide in the presence of iron as a Fenton reaction product, and because iron is abundant in the developing brain (32), this is likely to be a major reason for the in vivo production of ·OH. Although we were not able to measure iron in the microdialysis samples, this might simply mean that sufficient free iron does not enter the extracellular fluid or is under the detection limit of our assay.
Limitation and advantage of the methodology.
Microdialysis is a powerful method for measuring biochemical changes in tissues in real time, in situ. However, it has had limited application to the study of fetal brain metabolism, partly because of the long transit time along the outlet tubing. Although the dialysed sample is largely free of proteins and enzymes that degrade cellular metabolites, other compounds are inherently prone to oxidation (such as catecholamines, metabolites, glutamate, etc.) making them difficult to measure when the efflux time is prolonged. Measurement of ·OH using salicylic acid is difficult under these conditions, but the stability of 2-OH-TA makes it suitable as a trapping agent to measure ·OH in the fetal sheep brain in utero, where the transit time through a microdialysis probe and its tubing is necessarily prolonged (∼1 h). TA does not appear to have toxic effects on the fetal brain from the observations of both physiological parameters and brain histology. However, the toxicity of TA and its hydroxylated product have not yet been established, and further elucidation of possible side effects (e.g., on neural transmission) of TA and 2-OH-TA is required.
Although microdialysis provides real-time measurement of ·OH production, limitations of the microdialysis technique remain; viz., insertion of a probe into the brain is invasive and necessarily produces some tissue damage; reactive tissue processes occur that could decrease efficiency of the dialysis membrane; and only a relatively small volume of tissue is sampled by the probe. Cerebral perfusion is another factor that may affect the recovery rate of the probe since the concentration of substances in extracellular fluid is dependent on local perfusion. The present study was not designed to address this question, and further study is required. Nevertheless, in this study, we show that microdialysis of the fetal brain is possible for an extended time, and in another study we have used it to sample the neurosteroid level in gray and white matter of the cortex continuously over 96 h (28).
This project was supported by a National Health and Medical Research Council of Australia project and fellowship grants to M. Castillo-Melendez, S. L. Miller, and D. W. Walker, and a Monash University Graduate Scholarship to E. B. Yan.
The authors thank Alex Satragno and Jan Loose for excellent assistance with animal surgical procedures and experiments.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society