Abstract
Cytosolic sulfotransferases are believed to play a role in the neuromodulation of certain neurotransmitters and drugs. To date, four cytosolic sulfotransferases have been shown to be expressed in human brain. Recently, a novel human brain sulfotransferase has been identified and characterized, although its role and localization in the brain are unknown. Here we present the first immunohistochemical (IHC) localization of SULT4A1 in human brain using an affinity-purified polyclonal antibody raised against recombinant human SULT4A1. These results are supported and supplemented by the IHC localization of SULT4A1 in rat brain. In both human and rat brains, strong reactivity was found in several brain regions, including cerebral cortex, cerebellum, pituitary, and brainstem. Specific signal was entirely absent on sections for which preimmune serum from the corresponding animal, processed in the same way as the postimmune serum, was used in the primary screen. The findings from this study may assist in determining the physiological role of this SULT isoform.
R
Cytosolic sulfotransferases expressed in the human brain include SULTs 1A1, 1A3, 2A1, and 1E1. However, cloning, enzymological, and immunohistochemical (IHC) studies have provided little information about the distinct localization and level of expression of each isoform in the brain (Rein et al. 1982; Rivett et al. 1982; Young et al. 1984; Viani et al. 1990; Zou et al. 1990; Duffel et al. 1991; Homma et al. 1994; and Beaujean et al. 1999). Recently, cytosolic sulfotransferase consensus sequences have been used to identify a novel protein from the expressed sequence tag databases (Falany et al. 2000; Liyou et al. 2000b). Falany and co-workers (2000) named this protein BR-STL and reported on its expression in the human and rat brain. No expression was found in other tissues investigated with Northern blotting analysis. This work revealed that the mRNA of this SULT isoform was highly expressed in the cortex of rat and human brain (Falany et al. 2000). However, the authors were unable to demonstrate sulfotransferase activity.
After cloning the corresponding coding region from human brain cDNA, we expressed the recombinant protein in E. coli. We named this protein, according to the current nomenclature of sulfotransferases, SULT4A1 (Falany 1996). The recombinant protein, containing an N-terminal (His)6 tag, was purified using Ni-NTA affinity chromatography and used to raise polyclonal antibodies in the goat suitable for IHC studies. This study provides an IHC examination of the distribution and cellular localization of SULT4A1 in the human and rat brain.
Materials and Methods
Cloning the Human Brain SULT4A1 cDNA
The SULT4A1 coding sequence was identified by routine screening of the expressed sequence tag databases using the sulfotransferase-specific consensus sequences PKSGT and RKGxxGxWKxxFT. Primers were designed to amplify the coding region using human brain cDNA as template (5′-CATATGGCGGAGAGCGAGGCCG-3′ and 5′-CTCGAGTTATAAATAAAAGTCAAACGTGAGCTC-3′, forward and reverse, respectively). These primers incorporated NdeI (forward primer) and XhoI (reverse primer) sites at the 5′ ends to enable directional cloning. The resultant fragment was subcloned into PCRBlunt (Invitrogen; Carlsbad, CA) and the DNA sequence confirmed (ABI BigDye Terminator Cycle Sequencing Ready Reaction Kit, PE Applied Biosystems, Foster City, CA. Sequence analysis was performed by the Australian Genome Research Facility, University of Queensland, Australia). The fragment was excised from the shuttle vector using the above-mentioned restriction enzymes and cloned into pET28a(+) (Novagen).
Bacterial Expression
A single E. coli BL21 (DE3) plysS cell colony, transformed with pET28a(+) containing the SULT4A1 insert, was used to inoculate 10 ml of LB Broth, supplemented with 30 μg/ml kanamycin and 34 μg/ml chloramphenicol, and incubated at 37C overnight. This overnight culture was then used to inoculate 250 ml LB Broth, containing 30 μg/ml kanamycin and 34 μg/ml chloramphenicol, to an OD600 of 0.1 and incubated at 37C with shaking until OD600 reached 0.6–0.8. Expression of the recombinant protein was induced by the addition of isopropyl-β-
Protein Purification
All steps were carried out at 4C in the presence of protease inhibitors (Cømplete EDTA-free protease inhibitor cocktail; Boehringer Mannheim, Mannheim, Germany) unless otherwise stated. Protein analysis was conducted using SDS-PAGE and silver staining as described by Liyou et al. (1999). Protein concentrations were estimated either from the absorbance at 280 nm or using the Pierce BCA reagent as described by Liyou et al. (2000a).
The frozen cell pellet was resuspended by direct addition of extraction buffer (50 mM Na2PO4, 300 mM NaCl, 5 mM imidazole, pH 7.0) and sonication (three times for 20 sec) on ice. DNase I (2.5 U/μl) was added to the sonicate and incubated for 20 min at 37C, followed by further sonication as above. The soluble fraction was collected via ultracentrifugation at 100,000 × g for 1 hr.
The supernatant obtained above was bound to Nickel Superflow IMAC resin (5 ml; Progen, Brisbane, Australia) with rolling for 2 hr at 4C. After washing with five column volumes of wash buffer (50 mM Na2PO4, 300 mM NaCl, 20 mM imidazole, pH 7.0) in batch form, the resin with bound protein was packed into a column for elution of specifically bound protein using FPLC (Pharmacia; Uppsala, Sweden) with a flow rate of 1.0 ml/min. The resin was washed with a further 10 column volumes of wash buffer. Elution was effected by the application of a 20–500 mM linear imidazole gradient over 40 column volumes. The fractions (1.2 ml) were analyzed using A280 measurement and SDS-PAGE as described by Liyou et al. (1999). Based on the absorbance intensity, the recombinant SULT4A1 protein was eluted at around 220 mM imidazole and then concentrated using a Centriprep10 (10-kD cut-off) concentrating cell. SULT4A1 was further purified using anion exchange chromatography (MonoQ; Amersham, Castle Hill, Australia). Concentrated recombinant SULT4A1 was added to an equal volume of wash buffer 1 (50 mM Tris-HCl, pH 8.3) to give a binding NaCl concentration of 150 mM. The sample was applied to the resin at a flow rate of 1.0 ml/min, then washed with 10 column volumes of wash buffer 2 (50 mM Tris, 150 mM NaCl, pH 8.3). Elution of protein was effected using a linear 150–300 mM NaCl gradient over 40 column volumes, collecting 1.5 ml fractions. The selected fractions were pooled, concentrated, and stored in 50% glycerol at −80C until further use.
Generation of Polyclonal Antibodies
Purified protein was emulsified with Freund's complete adjuvant for injection into goats. Before immunization, goats were bled to collect control serum (40 ml/animal). Two goats were immunized with approximately 500 μg protein each in multiple sites on the hind legs. Follow-up booster injections with 500 μg protein in incomplete Freund's adjuvant were performed 28 days after the initial injection. The goats were bled 14 days later (40 ml/animal) and serum collected. All animals used in this study were handled according to NHMRC Animal Ethics Committee guidelines. Ethical approval was obtained through the Animal Ethics Committee of the University of Queensland.
Antibodies were purified from the immune serum by affinity chromatography using E. coli recombinant SULT4A1 immobilized on cyanogen bromide-activated Sepharose according to the manufacturer's instructions (Pharmacia). Purified antibodies (in TBS) were concentrated and stored at −20C in 50% glycerol (to a final concentration of 0.46 mg/ml). Corresponding preimmune goat serum was also applied to the affinity resin and the eluate concentrated and stored as above for use as the negative control in the IHC investigation.
Immunoblotting of Purified Recombinant Human SULTs
The specificity of the goat antihuman SULT4A1 antibody was tested using a panel of expressed human sulfotransferases (SULTs 1A1, 1A2, 1A3, 1E1, 2A1, and 1C2). Proteins were separated on gradient SDS-PAGE under reducing conditions and transferred to nitrocellulose as described in Liyou et al. (2000a).
Very slight crossreactivity occurred with SULT1A1 but not with the other SULTs. To correct for this, the purified antibody was applied to an affinity resin of immobilized recombinant SULT1A1 as above. Antibody that did not bind was collected, concentrated, and stored in glycerol. The antibody purified in this manner showed specificity only towards SULT4A1.
Human Tissue Samples
The samples examined in this study were collected according to NHMRC Human Ethics Committee guidelines. Ethical approval was obtained through the Human Ethics Committee of the University of Queensland. Relevant patient information is provided in Table 1.
Immunoblotting of Human Brain Cytosolic Fractions
Frozen human brain tissue (midtemporal, parietal, occipital and midfrontal cortex, cerebellum, caudate nucleus, and putamen) was collected from four individuals (Table 1: case numbers 1–4). Approximately 100 mg of each tissue sample was thawed in tissue extraction buffer (50 mM phosphate, 10 mM EDTA, pH 7.4, 0.5% Triton X-100). Each specimen was homogenized using a mortar and pestle (on ice) and then centrifuged (12,000 × g) to remove cell debris. Further centrifugation at 100,000 × g for 1 hr was performed to remove the membrane protein fraction. Supernatant was collected and protein estimates were performed for each sample. For SDS-PAGE electrophoresis (conducted under non-reducing conditions), approximately 150 μg of total protein was loaded per sample. Proteins were transferred to nitrocellulose and probed with affinity-purified humSULT4A1 antibody (1:1000 concentration). Purified recombinant SULT4A1 was used as control.
Immunohistochemistry
Cases 3 and 4 were examined for this study (Table 1). Immunoperoxidase staining for humSULT4A1 protein was performed on 6-μm sections using the DAKO (Glostrup, Denmark) LSAB+ (labeled streptavidin-biotin) peroxidase system. Sections were deparaffinized in xylene and rehydrated.
Patient details
aPM, post mortem.
Target retrieval was performed using DAKO Target Retrieval Solution (25 min, 97C) and then allowed to cool at room temperature for a further 20 min. Sections were then blocked using 1% H2O2 in buffer A (50 mM Tris-HCl, 0.5 M NaCl, 0.5% Tween-20, pH 7.6), washed in buffer A and incubated overnight (4C) with the goat anti-hum-SULT4A1 IgG (1:200) in buffer A. Sections were washed in buffer A and DAKO biotinylated linking antibody solution was applied (150 μl/section) for 25 min. Sections were then washed in buffer A before application of DAKO enzyme conjugated streptavidin (150 μl/slide, 20 min). Slides were washed in buffer A and Substrate Chromagen (DAKO) added for 5 min. For negative controls, the primary antibody was replaced with the corresponding affinity-purified preimmune serum.
Rat brain sections were obtained from adult male Wistar rats that were deeply anesthetized with sodium pentobarbitol (80 mg/kg IP) and perfused transcardially with 2% sodium nitrite solution (in 0.1 M PBS, pH 7.4) followed by 4% formaldehyde (in 0.1 M PBS, pH 7.4). Brains were quickly removed, postfixed for 2 hr in 4% formaldehyde, then cryoprotected overnight in 10% sucrose (in 0.1 M PBS, pH 7.4, 4C; Buller and Day 2002). Serial forebrain (40-μm) and brainstem (50-μm) sections were collected using a microtome. Two 1-in-5 brainstem series (250-μm interval) and two 1-in-4 forebrain series (160-μm interval) from each rat were immunolabeled, one brain series for the preimmune serum and one brain series for the SULT14A antibody. Sections were stained according to the human procedures described above except that dewaxing and rehydration were omitted. The sections were prepared by rinsing in TBS and blocking in TBS containing 1% H2O2. Labeling then proceeded as described for the human sections. After development the sections were washed in water and placed directly on glass slides to be air-dried overnight, then coverslipped.
Results
Cloning and Expression of humSULT4A1
The novel sulfotransferase-like sequence identified from the expressed sequence tag database was amplified using PCR, cloned, sequenced, and expressed in E. coli. The sequence (genomic clone Z97005, PAC 388M5) corresponds to that described by Falany et al. (2000). The corresponding amino acid sequence showed less than 35% identity with the other human cytosolic SULTs and therfore, based on the current classification system employed (Falany 1996), represents a new SULT family.
Characterization of Polyclonal SULT4A1 Antibodies
Isoform-specific polyclonal antibodies were generated against recombinant humSULT4A1 as described in Materials and Methods. Specificity was determined by probing a range of purified recombinant human cytosolic SULTs via immunoblotting. Figure 1 shows that only recombinant SULT4A1 is recognized by the antibodies that were further purified with an antibody affinity column specific for the proteins of the SULT1A family.
The polyclonal antibodies generated against E. coli recombinant SULT4A1 are specific for this SULT isoform. Immunoblotting analysis probing purified recombinant sulfotransferases (50 ng) with affinity-purified goat anti-SULT4A1 antibodies (1:1000 dilution) showed that only SULT4A1 was recognized by these antibodies. Molecular weight markers (Lane S); humSULT1A1 (Lane 1); humSULT1A2 (Lane 2); humSULT1A2v (Lane 3); humSULT1A3 (Lane 4); humSULT1E1 (Lane 5); humSULT2A1 (Lane 6); humSULT1C2 (Lane 7); humSULT4A1 (Lane 8). The apparent molecular mass (kD) of the markers is indicated at left.
The purified recombinant His-tagged SULT4A1 protein appears as a doublet in this non-reducing electrophoresis system. These bands resolve to a single band under reducing conditions (result not shown). This phenomenon has been observed for several of the N-terminally His-tagged recombinant SULTs.
Localization of SULT4A1 in Human Brain Regions
To identify the specific regions of the brain in which SULT4A1 was expressed, homogenized human brain sections from four individuals were analyzed by immunoblotting. Figure 2 shows the result for one subject, demonstrating that SULT4A1 is expressed strongly, but to various degrees, in all regions examined. The other three subjects studied showed a similar pattern of SULT4A1 expression (results not shown).
Immunohistochemistry
A more specific study of the localization of SULT4A1 was conducted by performing IHC analyses of human and rat brain sections. Distribution, cellular localization, and intensity of staining have been determined by observation using light microscopy. The distribution and relative abundance of SULT4A1 in regions of the human and rat brain examined in this study are presented in Tables 2 and 3 and representative brain sections are shown in Figures 3 (human) and 4 (rat). Control human and rat brain sections immunolabeled with the preimmune serum were devoid of immunolabeling for SULT4A1 (Figures 3 and 4). We found that SULT4A1 is extensively but specifically expressed in the human and rat brain. In addition, we found very little variation in staining among subjects. In one human subject no labeling was observed in the corticospinal tracts but weak staining was observed in other subjects (Table 3).
SULT4A1 is expressed extensively in the human brain. Immunoblotting analysis of human brain tissue (150 μg of total protein per lane) probing with SULT4A1 specific polyclonal antibodies (1:1000 dilution). Molecular weight markers (Lane S); midtemporal cortex (Lane 1); parietal cortex (Lane 2); occipital cortex (Lane 3); midfrontal cortex (Lane 4); cerebellum (Lane 5); caudate nucleus (Lane 6); putamen (Lane 7); pure His-tagged recombinant humSULT4A1 (Lane 8). The apparent molecular mass (kD) of the markers is indicated at the left.
Staining of human cortical sections reveal localization of SULT4A1 to neuronal cell bodies and dendrites in all six layers of the cerebral cortex but predominantly in layers 2, 3, 5, and 6. Large pyramidal and fusiform neurons were prominently labeled. A very similar distribution and cellular localization of SULT4A1 was observed in the rat cortex, with strong immunolabeling of neurons throughout the cortex including the motor, somatosensory, and cingulate cortices (Figure 4B). In the prefrontal cortex, SULT4A1-positive neurons were located mostly in the infralimbic and dorsal peduncular regions and were also observed in the tenia tecta and prelimbic regions. Positive neurons were concentrated in the islands of Calleja of the rat and human. However, in both the rat and human, the shell and core of the nucleus accumbens were devoid of staining. In addtion, few positive elements were found in the caudate nucleus and putamen. Large numbers of SULT4A1-positive neurons were observed in the lateral septum and the diagonal band of Broca in the rat. Many neurons were also labeled in the dorsal and ventral regions of the bed nucleus of the stria terminalis.
Distribution and cellular localization at SULT4A1 in human brain regions
aThe relative intensity of labeling was ranked by two independent observers. Ratings reflect mainly the density of SULT4A1-labeled cells. Negative, –; weak, (+); weak/moderate, +; moderate, ++; moderate/strong, + + +; strong, + + + +.
bGranular staining suggests synapses labeled.
Distribution and cellular localization at SULT4A1 in rat brain regions
aThe relative intensity of labeling was ranked by two independent observers. Ratings reflect mainly the density of SULT4A1-labeled cells. Negative, –; weak, (+); weak/moderate, +; moderate, ++; moderate/strong, + + +; strong, + + + +.
bGranular staining suggests synapses labeled. BNST, bed nucleus of the stria terminalis; PAG, periaqueductal gray.
IHC staining of human cerebellum sections reveal localization of SULT4A1 to granular neuronal cell bodies within the folial granular layer of the cerebellum (Figure 3B). Staining was also observed in the Purkinje cells in the lateral lobe of the folia; however, staining was not observed in the vermis. A large number of SULT4A1-positive neurons were also seen in the human pituitary sections (Figure 3F). However, owing to the procedure of brain removal in the rat, the pituitary was not collected for analysis.
SULT4A1-positive neurons were visible in both the human and rat sections in layers of the hippocampus, including Ammon's horn and Sommer's sector. In both the rat and human, many labeled neurons were localized in all divisions of the thalamic nuclei (Figure 3D). In particular, the anterior and subthalamic nuclei were labeled intensely. SULT4A1 was localized to neurons (with and without neuromelanin, which is naturally stained a deep brown) and dendrites in the substantia nigra in human and rat midbrain sections (Figure 3H). Staining was apparent in the pars compacta and pars reticularis. In the hypothalamus, only scattered labeled neurons were found in the paraventricular nucleus. A group of labeled neurons was concentrated dorsal to the paraventricular nucleus in the zona incerta in the rat brain.
SULT4A1-positive neurons were found in the peri-aqueductal gray and were particularly localized in the dorsal, lateral, and ventrolateral divisions. Analysis of sections of the red nucleus of the midbrain also showed SULT4A1 localized to neurons of both human and rat sections (Figure 4F). The magnocellular neurons of the rat were especially strongly labeled compared to those observed in the human red nucleus.
In both human and rat brainstem sections, SULT4A1 was expressed strongly in neurons of the III cranial nerve (oculomotor nucleus; Figure 4D). Neurons containing SULT4A1 were localized to neuronal cell bodies of the inferior olivary nucleus in the human. In the brainstem, neurons were also strongly labeled in the XII (hypoglossal nucleus, Figure 4H), VII (facial nucleus), and V (trigeminal nucleus) cranial nerves of both human and rat. Many SULT4A1-positive neurons were localized in the lateral reticular nucleus in the brainstem. Labeled cells were also found in the gracile and cuneate nuclei. Very few labeled neurons were observed in the nucleus tractus solitarius. Neurons were labeled in the dorsal motor nucleus of the vagus in both the rat and human. The locus ceruleus was virtually devoid of positive SULT4A1 elements in the human, but in the rat the locus ceruleus was labeled strongly. The parabrachial nucleus of the rat also contained prominent SULT4A1 immunolabeling.
Photomicrographs of cross-sections through human brain tissue immunolabeled with pre-immune goat serum (
Photomicrographs of cross-sections through rat brain tissue immunolabeled with preimmune goat serum (
Discussion
Cytosolic sulfotransferases are a superfamily of enzymes important in phase II metabolism of both xenobiotics and endogenous substrates such as catecholamines and steroids. The present study identifies the distribution and cellular localization of a novel brain cytosolic sulfotransferase isoform in the human and rat brain: SULT4A1. Knowing the pattern of expression of this enzyme in the brain, along with structural and functional information, will assist with determining the functional significance of SULT4A1 localization in the brain.
The novel sequence identified, cloned, and expressed in the present study establishes criteria for a new SULT family and therefore was named SULT4A1. The generation of isoform-specific polyclonal antibodies towards SULT4A1 in this study represents a powerful tool for examining the role of this enzyme. Immunoblotting analyses showed that SULT4A1 is expressed extensively, and apparently exclusively, in the brain. This is in contrast to other SULT isoforms, which are expressed in brain tissue as well as other tissues (Rein et al. 1982; Young et al. 1984; Zou et al. 1990; Duffel et al. 1991; Homma et al. 1994; Beaujean et al. 1999). At present, the physiological substrate of SULT4A1 is unknown. However, the cDNA cloned in this study is 100% identical to that identified by Falany et al. (2000) and has all the signatory sequences of a sulfotransferase cDNA (Kakuta et al. 1997; Bidwell et al. 1999).
The present study collectively demonstrates that SULT4A1 is expressed in the human and rat brain in a restricted manner. Furthermore, the region-specific expression pattern of SULT4A1 in the brain implies a function in the central nervous system, although the nature of this function is not yet clear. In the forebrain we observed the greatest degree of immunolabeling for SULT4A1 in the cortex (motor, cingulate, frontal, somatosensory), globus pallidus, islands of Calleja, septum, thalamus (lateral, medial dorsal, anterior, subthalamic), red nucleus, substantia nigra and pituitary. Therefore, in support of the only other study to have examined brain expression levels, a Northern blotting analysis conducted by Falany and co-workers (2000), we found a high degree of SULT4A1 expression in the cortex of both the human and rat. Interestingly, SULT4A1 expression was observed in neuronal cell bodies and dendrites in all six layers of the cerebral cortex. In the midbrain and brainstem, staining was particularly prominent in the cerebellum, nuclei of the III, V, VII, and XII cranial nerves, red nucleus, and lateral reticular nucleus.
Importantly, we report that many motor nuclei in the brain were found to express SULT4A1. SULT4A1 was localized to granular neuronal cell bodies in the folial granular layer of the cerebellum. The granular cells are the only excitatory neurons in the cerebellum and receive the main input into this region. The neurons of the red nucleus of the midbrain, which express SULT4A1, have cerebral, thalamic cerebellar, and brainstem connections. This pathway is concerned with reflexes involved in motor coordination and maintaining posture. SULT4A1 was also expressed in the III cranial nerve nucleus (oculomotor), which produces certain intrinsic and extrinsic movements of the eyeball. The neuronal cell bodies of the inferior olivary nucleus, which acts on cerebellar circuits to integrate sensory and motor information about movements in real time, also demonstrated the presence of SULT4A1. SULT4A1 was also present in neurons and dendrites in the substantia nigra in human midbrain sections. Many neurons in the substantia nigra send fibers to the basal ganglia, which are involved in the coordination of skeletal muscle movements.
It is notable that there was a high degree of similarity in the localization and level of intensity of SULT4A1 immunolabeling in the human and rat brain. Only relatively minor variations were observed. For example, only a paucity of staining was found in the human locus ceruleus but in the rat the degree of labeling was quite marked. On the other hand, in the human a strong signal was apparent in the amygdala but in the rat very little SULT4A1 immunolabeling was observed. Nevertheless, the high degree of conformity between the rat and the human is favorable if one were to subsequently use the rat as an animal model to investigate human brain SULT4A1 neurochemistry.
The present study greatly extends our knowledge of the distribution and cellular localization of cytosolic sulfotransferases in the brain and provides a foundation for understanding how SULT4A1 might regulate the functions and neurochemistry of different regions of the brain. Our laboratory is now undertaking structural and functional studies to determine the specific role played by SULT4A1.
Footnotes
Acknowledgements
Supported by a grant from the National Health and Medical Research Council (#9936607). PRD is an NHMRC Principal Research Fellow.
We would like to express our gratitude to Mr Roger Pearson of CSIRO Livestock Industries and to Dr Conrad Sernia and Paul Addison of the School of Biomedical Sciences.