Functional and structural properties of stannin: roles in cellular growth, selective toxicity, and mitochondrial responses to injury

Stannin (Snn) was discovered using subtractive hybridization methodology designed to find gene products related to selective organotin toxicity and apoptosis. The cDNAs for Snn were first isolated from brain tissues sensitive to trimethyltin, and were subsequently used to localize, characterize, and identify genomic DNA, and other gene products of Snn. Snn is a highly conserved, 88 amino acid protein found primarily in vertebrates. There is a minor divergence in the C-terminal sequence between amphibians and primates, but a nearly complete conservation of the first 60 residues in all vertebrates sequenced to date. Snn is a membrane-bound protein and is localized, in part, to the mitochondria and other vesicular organelles, suggesting that both localization and conservation are significant for the overall function of the protein. The structure of Snn in a micellar environment and its architecture in lipid bilayers have been determined using a combination of solution and solid-state NMR, respectively. Snn structure comprised a single transmembrane domain (residues 10-33), a 28-residue linker region from residues 34-60 that contains a conserved CXC metal binding motif and a putative 14-3-3xi binding region, and a cytoplasmic helix (residues 61-79), which is partially embedded into the membrane. Of primary interest is understanding how this highly-conserved peptide with an interesting structure and cellular localization transmits both normal and potentially toxic signals within the cell. Evidence to date suggests that organotins such as trimethyltin interact with the CXC region of Snn, which is vicinal to the putative 14-3-3 binding site. In vitro transfection analyses and microarray experiments have inferred a possible role of Snn in several key signaling systems, including activation of the p38-ERK cascade, p53-dependent pathways, and 14-3-3xi protein-mediated processes. TNFalpha can induce Snn mRNA expression in endothelial cells in a PKC-epsilon dependent manner. Studies with Snn siRNA suggest that this protein may be involved in growth regulation, since inhibition of Snn expression alone leads to reduced endothelial cells growth and induction of COP-1, a negative regulator of p53 function. A key piece of the puzzle, however, is how and why such a highly-conserved protein, localized to mitochondria, interacts with other regulatory proteins to alter growth and apoptosis. By knowing the structure, location, and possible signaling pathways involved, we propose that Snn constitutes an important sensor of mitochondrial damage, and plays a key role in the mediation of cross-talk between mitochondrial and nuclear compartments in specific cell types.

Calcium-dependent interaction of calcineurin with Bcl-2 in neuronal tissue

Calcineurin, a calmodulin-dependent protein phosphatase, regulates transcription and possibly apoptosis. Previous studies demonstrated that in baby hamster kidney-21 cells after co-transfection calcineurin interacts with Bcl-2, thereby altering transcription and apoptosis. Using co-immunoprecipitation and subcellular fractionation techniques, we observed that calcineurin occurred as a complex with Bcl-2 in various regions of rat and mouse brain. The calcineurin-Bcl-2 complex was identified in mitochondrial, nuclear, microsomal and cytosol fractions. In vitro induction of hypoxia and aglycia or N-methyl-D-aspartate treatment markedly altered both extent of complex formation and its subcellular localization. These observations suggest that Bcl-2 either sequesters calcineurin, that calcineurin dephosphorylates Bcl-2, or that Bcl-2 shuttles calcineurin to specific substrates. Calcineurin also co-immunoprecipitated with the inositol-tris-phosphate receptor. This interaction increased after in vitro hypoxia/aglycia. In Bcl-2 (-/-) mice, interactions between calcineurin- and inositol-tris-phosphate receptor occurred less frequently than in wild-type mice under both control and hypoxic conditions. Experiments involving cell-free systems, as well as brain slices treated with thapsigargin or with N-methyl-D-aspartate suggested that calcium and calmodulin activation of calcineurin leads to interactions between calcineurin and Bcl-2. These data indicate that during times of cellular stress and damage, Bcl-2 targets activated calcineurin to specific compartments and substrates.

In vitro hypoxia and excitotoxicity in human brain induce calcineurin-Bcl-2 interactions

Although pathogenesis of neuronal ischemia is incompletely understood, evidence indicates apoptotic neuronal death after ischemia. Bcl-2, an anti-apoptotic and neuroprotective protein, interacts with calcineurin in non-neuronal tissues. Activation of calcineurin, which is abundant in the brain, may play a role in apoptosis. Using co-immunoprecipitation experiments in biopsy-derived, fresh human cortical and hippocampal slices, we examined possible interactions between calcineurin and Bcl-2. Calcineuin-Bcl-2 interactions increased after exposure in vitro to excitotoxic agents and conditions of hypoxia/aglycia. This interaction may shuttle calcineurin to substrates such as the inositol-1,4,5-tris-phosphate receptor because under these experimental conditions interactions between calcineurin and inositol-1,4,5-tris-phosphate receptor also increased. A specific calcineurin inhibitor, FK-520, attenuated insult-induced increases in calcineurin-Bcl-2 interactions and augmented caspase-3 like activity. These data suggest that Bcl-2 modulates neuroprotective effects of calcineurin and that calcineurin inhibitors increase ischemic neuronal damage.

Phosphorylation of Syk activation loop tyrosines is essential for Syk function. An in vivo study using a specific anti-Syk activation loop phosphotyrosine antibody

Syk is an important protein-tyrosine kinase in immunoreceptor signaling. FcepsilonRI aggregation in mast cells induces tyrosine phosphorylation and increased enzymatic activity of Syk. The two adjacent tyrosines in the Syk activation loop are thought to be important for the propagation of FcepsilonRI signaling. To evaluate the phosphorylation of these tyrosines in vivo and further understand the relationship of Syk tyrosine phosphorylation with its function, an antibody was developed specific for phosphorylated tyrosines in the activation loop of Syk. FcepsilonRI aggregation on mast cells induced the phosphorylation of both tyrosine residues of the activation loop. The kinase activity of Syk played the major role in phosphorylating its activation loop tyrosines both in vivo and in vitro. In FcepsilonRI-stimulated mast cells, the total Syk tyrosine phosphorylation paralleled the phosphorylation of its activation loop tyrosines and downstream propagation of signals for histamine release. In contrast, the cell surface binding of anti-ganglioside monoclonal antibody AA4 induced only strong general tyrosine phosphorylation of Syk and minimal histamine release and weak phosphorylation of activation loop tyrosines. These results demonstrate that phosphorylation of the activation loop tyrosines is important for mediating receptor signaling and is a better marker of Syk function than is total Syk tyrosine phosphorylation.

Mechanisms of injury in the central nervous system

Neurotoxicants with similar structural features or common mechanisms of chemical action frequently produce widely divergent neuropathologic outcomes. Methylmercury (MeHg) produces marked cerebellar dysmorphogenesis during critical periods of development. The pathologic picture is characterized by complete architectural disruption of neuronal elements within the cerebellum. MeHg binds strongly to protein and soluble sulphydryl groups. Binding to microtubular -SH groups results in catastrophic depolymerization of immature tyrosinated microtubules. However, more mature acetylated microtubules are resistant to MeHg-induced depolymerization. In contrast to MeHg, the structurally similar organotin trimethyltin (TMT) elicits specific apoptotic destruction of pyramidal neurons in the CA3 region of the hippocampus and in other limbic structures. Expression of the phylogenetically conserved protein stannin is required for development of TMT-induced lesions. Inhibition of expression using antisense oligonucleotides against stannin protects neurons from the effects of TMT, suggesting that this protein is required for expression of neurotoxicity. However, expression of stannin alone is insufficient for induction of apoptotic pathways in neuronal populations. The aromatic nitrocompound 1,3-dinitrobenzene (DNB) has 2 independent nitro groups that can redox cycle in the presence of molecular oxygen. Despite its ability to deplete neural glutathione stores, DNB produces edematous gliovascular lesions in the brain stem of rats. Glial cells are susceptible despite high concentrations of reduced glutathione compared with neuronal somata in the central nervous system (CNS). The severity of lesions produced by DNB is modulated by the activity of neurons in the affected pathways. The inherent discrepancy between susceptibility of neuronal and glial cell populations is likely mediated by differential control of the mitochondrial permeability transition in astrocytes and neurons. Lessons learned in the mechanistic investigation of neurotoxicants suggest caution in the evaluation and interpretation of structure-activity relationships, eg, TMT, MeHg, and DNB all induce oxidative stress, whereas TMT and triethyltin produce neuronal damage and myelin edema, respectively. The precise CNS molecular targets of cell-specific lipophilic neurotoxicants remain to be determined.

Overexpression of epidermal growth factor receptor in Peutz-Jeghers syndrome

Peutz-Jeghers syndrome is characterized by gastrointestinal hamartomatous polyposis, mucocutaneous pigmentation, and a predisposition to cancer. The etiology of this syndrome is unknown. We investigated the expression of epidermal growth factor receptor (EGFr), transforming growth factor-alpha (TGF-alpha), transforming growth factor-beta1 (TGF-beta1) and transforming growth factor-beta receptor (TGF-beta RII) between normal and Peutz-Jeghers small bowel tissues. In addition, immunoprecipitation by phosphotyrosine antibodies followed by EGFr western blotting was measured and compared between a Peutz-Jeghers hamartoma and normal duodenal tissue. EGFr expression was increased 2.5-fold in normal and hamartomatous tissue of Peutz-Jeghers patients compared to normal small bowel tissue. In Peutz-Jeghers tissues, the major EGFr immunoreactive band was increased size from 170 to approximately 200 kDa. Using an antibody specific for activated EGFr, this larger size band was predominant in Peutz-Jeghers tissue. Immunoprecipitation of a hamartoma by a phosphotyrosine specific antibody followed by western blotting for EGFr demonstrated this 200-kDa band. Expression of TGF-alpha, TGF-beta1, TGF-beta1 RII was not significantly different between normal and Peutz-Jeghers tissues. In conclusion, EGFr was overexpressed in normal and hamartomatous small bowel tissue of Peutz-Jeghers patients, which suggests that EGFr in Peutz-Jeghers tissue is persistently activated or highly stimulated by endogenous ligands and also suggests a possible role for EGFr in the pathogenesis of Peutz-Jeghers syndrome.

Reduction of calcineurin activity in brain by antisense oligonucleotides leads to persistent phosphorylation of tau protein at Thr181 and Thr231

Phosphorylation of tau protein promotes stability of the axonal cytoskeleton; aberrant tau phosphorylation is implicated in the biogenesis of paired helical filaments (PHF) seen in Alzheimer's disease. Protein kinases and phosphatases that modulate tau phosphorylation have been identified using in vitro techniques; however, the role of these enzymes in vivo has not been determined. We used intraventricular infusions of antisense oligodeoxynucleotides (ODNs) directed against the major brain isoforms of the Ca2+/calmodulin-dependent phosphatase calcineurin to determine how reduced activity of this enzyme would affect tau dephosphorylation. Five-day infusions of antisense ODNs (5 and 10 nmol/day) in rats decreased immunoreactive levels and activity of calcineurin throughout the brain; sense ODNs, scrambled ODNs, and infusion vehicle alone had no effect. When neocortical slices were prepared from antisense ODN-treated rats and incubated for 1 to 2 h in vitro, tau protein remained phosphorylated as determined by using the phosphorylation-sensitive monoclonal antibodies AT-180 (Thr231) and AT-270 (Thr181). In contrast, AT-180 and AT-270 sites were completely dephosphorylated during incubation of neocortical slices from vehicle-infused controls and sense ODN-treated rats. Neocortical slices from antisense-treated rats were incubated with the phosphatase inhibitors okadaic acid (100 nM; 10 microM) and FK-520 (5 microM); these preparations showed enhanced tau phosphorylation, consistent with a significant loss of calcineurin activity. Thus, we conclude that phosphorylation of at least two sites on tau protein, namely, Thr181 and Thr231, is regulated by calcineurin.

Alterations in tau phosphorylation in rat and human neocortical brain slices following hypoxia and glucose deprivation

Tau is a microtubule-associated protein which is regulated by phosphorylation. Highly phosphorylated tau does not bind microtubules and is the main component of the paired helical filaments seen in Alzheimer's and related neurodegenerative diseases. Recent reports suggested that patterns of tau phosphorylation changed following ischemia and/or reperfusion in vivo. We used an in vitro model employing rat and human neocortical slices to investigate changes in tau phosphorylation which accompany oxygen and glucose deprivation. Western blotting with polyclonal and phosphorylation-sensitive Tau-1 monoclonal antisera was used to monitor changes in tau which accompanied conditions of oxygen and glucose deprivation and reestablishment of these nutrients. In vitro hypoglycemia/hypoxia caused tau to undergo significant dephosphorylation in both rat and human neocortical slices after 30 and 60 min of deprivation. This dephosphorylation was confirmed using immunoprecipitation experiments after radiolabeling tau and other proteins with 32Pi. Okadaic acid, a phosphatase inhibitor, was able to prevent tau dephosphorylation in both control and ischemic slices. Lubeluzole, a benzothiazole derivative with in vivo neuroprotective activity, did not significantly alter patterns of tau phosphorylation. Restoration of oxygen and glucose following varied periods of in vitro hypoxia/hypoglycemia (15-60 min) led to an apparent recovery in phosphorylated tau. These data suggest that tau undergoes a rapid, but reversible dephosphorylation following brief periods of in vitro hypoxia/hypoglycemia in brain slices and that changes in tau phosphorylation help determine the extent of recovery following oxygen and glucose deprivation.

Chromosomal localization and characterization of the stannin (Snn) gene

Stannin is a protein that has been localized to trimethyltin-sensitive cell populations, and evidence suggests it plays a role in the toxic effects of organotins. In this study, we have isolated a mouse stannin genomic clone and have characterized the gene's intron-exon organization, promoter region, and chromosomal location. We have also isolated a partial human stannin cDNA clone and analyzed the open reading frame. The mouse genomic clone spans approximately 19 kb and consists of one intron and two exons. The splice site consensus sequence was maintained at all intron-exon junctions. Promoter analysis suggests that two putative promoter sites exist, each containing multiple regulatory elements and transcription factor-binding sites. Fluorescence in situ hybridization analysis localized stannin to mouse Chromosome (Chr) 16 at band A2. This region is homologous to the proximal region of human Chr 16 (16p13) to which stannin has been previously mapped. Sequence analysis revealed that the 264-bp open reading frame was identical between rat and mouse. The human sequence was 98% identical, with two amino acid substitutions near the c-terminal end of the peptide. These data suggest that stannin is highly conserved between species, and its unusual pattern of cellular expression may, in part, be explained via cell-specific promoters.

Localization and characterization of stannin: relationship to cellular sensitivity to organotin compounds

The cDNA encoding the protein stannin was isolated previously via subtractive hybridization, using differential expression after trimethyltin (TMT) intoxication, as a basis for isolating mRNA which may be expressed in TMT-sensitive cells. Initial characterization revealed a novel gene product which was differentially expressed in several tissues sensitive to TMT. In the current study, biochemical and molecular techniques were used to quantitate stannin expression at the cellular and subcellular levels. Northern blot analysis showed that the stannin 3.0 kb mRNA transcript was present, in decreasing amounts, in: spleen, hippocampus, neocortex, cerebellum, striatum, midbrain, kidney and lung. Liver, heart, skeletal muscle and testis showed no detectable expression of stannin mRNA. Immunoblot analysis using antipeptide antisera raised against stannin indicated a high level of expression in spleen, followed by brain and kidney. Stannin mRNA was present during early brain development and consolidated by post-natal day (PND) 20 to patterns and levels seen in adults. In situ hybridization studies showed widespread neuronal expression of stannin mRNA at PND 1, which shifted to a restricted pattern of expression in specific regions by PND 20. Stannin was partially purified from rodent brain and spleen using cation exchange, sizing and hydrophobic interaction chromatography. It behaved as a monomer throughout purification. Stannin was also expressed in a baculovirus system, using a series of constructs containing the entire cDNA, 1.0 kb of DNA flanking the open reading frame, and a 400 bp open reading frame minimal construct. While all constructs expressed stannin, the best expression was seen with the entire cDNA. Based on current findings, we suggest that stannin expression is necessary but not sufficient for TMT toxicity.

Regulated phosphorylation and dephosphorylation of tau protein: effects on microtubule interaction, intracellular trafficking and neurodegeneration

This review attempts to summarize what is known about tau phosphorylation in the context of both normal cellular function and dysfunction. However, conceptions of tau function continue to evolve, and it is likely that the regulation of tau distribution and metabolism is complex. The roles of microtubule-associated kinases and phosphatases have yet to be fully described, but may afford insight into how tau phosphorylation at the distal end of the axon regulates cytoskeletal-membrane interactions. Finally, lipid and glycosaminoglycan modification of tau structure affords yet more complexity for regulation and aggregation. Continued work will help to determine what is causal and what is coincidental in Alzheimer's disease, and may lead to identification of therapeutic targets for halting the progression of paired helical filament formation.

Alterations in hippocampal expression of SNAP-25, GAP-43, stannin and glial fibrillary acidic protein following mechanical and trimethyltin-induced injury in the rat

A set of well-defined antisera against neuronal and glial proteins were used to characterize patterns of protein expression in rat hippocampus following transection of the fimbira-fornix and perforant pathways or after administration of the selective neurotoxicant trimethyltin (8 mg/kg, i.p.). SNAP-25 (synaptosomal protein, mol. wt 25,000) is a neuron-specific, developmentally regulated presynaptic protein, stannin is a protein enriched in cells sensitive to trimethyltin, and GAP-43 (growth-associated protein, mol. wt 43,000) is associated with axonal growth and regeneration. Glial fibrillary acidic protein is an astrocyte-specific intermediate filament protein and a marker for reactive gliosis. SNAP-25 immunoreactivity was altered following both neurotoxicant and mechanical injury. Three days after fimbria-fornix/perforant path lesions, there was a loss of SNAP-25 immunoreactivity in hippocampal efferent pathways and in the lesioned entorhinal cortex. By day 12, there was evidence of reinnervation of hippocampal subfields by SNAP-25-immunopositive commissural afferent fibers. On day 3, immunoblots showed the appearance of SNAP-25a, a developmental isoform produced by alternative splicing of nine amino acids in exon 5, in lesioned tissues. This isoform declined by day 12 and was not found in contralateral control hippocampus or non-lesioned brain regions. Stannin immunoreactivity was unchanged, while GAP-43 was prominent on day 12 post-lesion. Glial fibrillary acidic protein immunoreactivity indicated gliosis near the site of pathway transection. In contrast, trimethyltin induced a marked loss of stannin immunoreactivity in hippocampal neurons seven days after injection. Trimethyltin increased glial fibrillary acidic protein staining in the hippocampus and other damaged regions. SNAP-25 immunoreactivity was markedly increased in mossy fibers and other hippocampal fields seven days following trimethyltin. Immunoblot analysis showed that only the adult SNAP-25b isoform was expressed after trimethyltin intoxication. These data suggest that SNAP-25 is a useful marker for presynaptic damage. Furthermore, reexpression of developmental isoforms of SNAP-25a may precede functional reinnervation when the postsynaptic target remains intact.

Microtubule assembly competence analysis of freshly-biopsied human tau, dephosphorylated tau, and Alzheimer tau

Phosphorylation of the microtubule-associated protein tau regulates its binding to microtubules; highly phosphorylated tau is also a prime component of paired helical filaments (PHFs) of Alzheimer's disease (AD). Tau from freshly biopsied human, monkey, and rat brain share similar electrophoretic mobility patterns and overlapping phosphorylated epitopes when compared to AD tau isolated from AD brain. We compared the microtubule reassembly competence of fresh isolates of phosphorylated tau to that of maximally dephosphorylated tau and tau from AD brain. A rapid procedure was developed which permitted the enrichment of phosphorylated and dephosphorylated tau from human biopsies in the absence of protein kinase and phosphatase activity. Microtubule assembly assays, using a spectrophotometric measure and purified bovine brain tubulin, were used to correlate assembly competence with states of tau electrophoretic mobility. Maximally dephosphorylated human biopsy-derived tau and monkey tau were assembly competent; tau from AD brain was virtually unable to direct microtubule assembly. Unmodified, biopsy-derived tau from non-AD brain was intermediate in assembly competence relative to AD tau and dephosphorylated tau. Several lines of evidence were used to correlate phosphorylation states of tau with microtubule assembly. First, in vitro dephosphorylation of human biopsy-derived tau with either PP2A or PP2B alone or in combination led to increasing assembly competence as the electrophoretic mobility of tau increased. Second, addition of the protein phosphatase inhibitor okadaic acid (10 microM) to brain-slice preparations slowed electrophoretic mobility of tau and decreased binding competence. We suggest that tau derived from freshly-biopsied brain exists in a range of phosphorylated states, and that dephosphorylation by PP2A and/or PP2B increases microtubule assembly competence.

Induction of apoptosis by organotin compounds in vitro: neuronal protection with antisense oligonucleotides directed against stannin

Immortalized cell lines and primary neuronal cultures were used to characterize the selective toxicity of trimethyltin (TMT),triethyltin (TET) and tributyltin (TBT). TBT and TET were cytotoxic at similar concentrations in the immortalized cell lines tested; the 50% toxic concentration (TC50) was 1 to 11 microM. In contrast, immortalized cell lines varied considerably in their sensitivity to TMT, with sensitive cell lines (neuroblastomas, T-, B-cell lines) showing TC50 values of 2 to 8 microM, whereas insensitive cells (NIH-3T3 fibroblast, HTB-14 glioma, TC-7 kidney cells) had TC 50 values > 100 microM. Primary neuronal cell cultures were very sensitive to organotins (TC50 values, 1-10nM), and showed patterns of selective toxicity with respect to neuronal and glial cells. Because organotin toxicity evolves over 24 to 48 hr. we determined whether these compounds induced apoptosis in primary cultures. TMT increased (P < .05) the fraction of apoptotic cells 6 and 12 hr after treatment with TMT at TC50 concentrations. Prior studies suggested that a protein, stannin, was localized in cells sensitive to organotins. Stannin was expressed in several TMT-sensitive cell lines (PC12, T, B cells) and in primary neurons in culture. Stannin was absent in the resistant HTB-14 glioma cell line. The role of stannin in mediating TMT toxicity in primary cultures was investigated by blocking stannin expression with specific antisense oligonucleotides. Treatment of primary cultures with antisense oligonucleotides for 48 hr before and during TMT treatment significantly protected neurons from the neurotoxic and apoptotic effects of TMT. This effect was not observed with scrambled oligonucleotide controls. Thus, TMT may induce apoptosis in sensitive cells, which is partly mediated by stannin. Based on the available data we conclude that stannin expression is necessary, but not sufficient for TMT toxicity.

Induction and variants of neuronal nitric oxide synthase type I during synaptogenesis

In the adult central nervous system, nitric oxide (NO) is formed from L-arginine by the so-called constitutive or type I NO synthase (NOS-I155). However, expression of NOS-I155 immunoreactivity and activity was low or not detectable in developing mouse and rat brain. NOS-I155 was sharply induced coincident with the onset of synaptogenesis in specific brain regions. This was followed by a second phase in which total NOS-I155 expression decreased both in specific cell populations and in the total synaptosomal subcellular fraction.Furthermore, two putative variants of NOS-I were transiently observed: an NOS-I-immunoreactive protein with increased electrophoretic mobility (NOS-I144) and a transient hypersensitivity of NOS-I155 to the competitive substrate inhibitor N omega-nitro-L-arginine. It is concluded that NOS-I expression is not constitutive but locally induced. In the central nervous system, this regionally specific, biphasic pattern of postnatal NOS-I induction is consistent with a role for NO in synaptogenesis and synaptic plasticity.

Reversal of toxicity using avidin-based hemoperfusion: a model system in rats using biotinylated melittin

The high-affinity interaction between avidin and biotin (Kd = 10(-15)M) can be exploited to develop specific protocols for retrieval of biotinylated drugs and toxicants from biological fluids. Melittin, the main toxic component of bee venom, was biotinylated and used as a model toxicant to determine whether avidin-based extracorporeal hemoperfusion could remove biotinylated melittin and thus alter the severity of the toxic response in rats. Melittin was biotinylated using N-hydroxysuccinimide-long-chain biotin. Biotinylated melittin produced 100% lethality in rats by 120 min following four sequential intravenous injections of 1.7 mg/kg biotinylated melittin (0, 5, 20, and 35 min). An avidin hemoperfusion column was constructed (10 mg avidin/1 ml gel) and connected via the femoral vasculature to rats intoxicated with biotinylated melittin. Controls rats were hemoperfused using avidin columns blocked with d-biotin. None of the 6 rats hemoperfused using the biotin-blocked avidin column control survived, whereas 5 of 9 of the experimental rats survived to 120 min. The difference between the two survival rates was statistically significant (p < 0.0048). Thus, avidin-based hemoperfusion improved survival following biotinylated melittin toxicity and strengthens the concept that avidin-based hemoperfusion can reverse the toxicity of biotinylated toxicants.

Tau phosphorylation in brain slices: pharmacological evidence for convergent effects of protein phosphatases on tau and mitogen-activated protein kinase

Tau is a neuron-specific, microtubule-associated protein that forms paired helical filaments (PHFs) of Alzheimer's disease when aberrantly phosphorylated. We have attempted to elucidate the protein kinases and phosphatases that regulate tau phosphorylation. Incubation of rat, human, and rhesus monkey temporal neocortex slices with the phosphatase inhibitor okadaic acid induced epitopes of tau similar to those found in PHFs. Okadaic acid (1-20 microM) induced variant forms of tau at 60-68 kDa, which were recognized by the monoclonal antibodies Alz-50 (in humans only) and 5E2 and two polyclonal antipeptide antisera, OK-1 and OK-2. The phosphorylation-sensitive monoclonal antibody Tau-1 failed to recognize the slowest mobility forms of tau after okadaic acid treatment. FK-520 (1-10 microM), a potent inhibitor of calcineurin activity, was tested in brain slices and found not to alter tau mobility. However, combinations of FK-520 (5 microM) and okadaic acid (100 nM) caused tau mobility shifts similar to those seen after 10 microM okadaic acid treatment; similar results were seen using the calcineurin-selective inhibitor cypermethrin. Treatment of human slices with 10 microM okadaic acid decreased both protein phosphatase 2A and calcineurin activity; FK-520 inhibited only protein phosphatase 2B activity. A proposed tau-directed kinase, 42-kDa mitogen-activated protein kinase (p42mapk), was activated by okadaic acid (> 100 nM) but not FK-520 (5 microM). Nerve growth factor (100 ng/ml) activated p42mapk, particularly when used in combination with 100 nM okadaic acid; changes in tau mobility were seen when this kinase was activated. Forskolin (2 microM) antagonized the effects of nerve growth factor on both p42mapk activity and tau phosphorylation; forskolin alone had little effect on PHF-like tau formation induced by phosphatase inhibitors. These results outline complex interactions between tau-directed protein kinases and protein phosphatases and suggest potential sites for therapeutic intervention.

Tau phosphorylation in human, primate, and rat brain: evidence that a pool of tau is highly phosphorylated in vivo and is rapidly dephosphorylated in vitro

The extent of tau phosphorylation is thought to regulate the binding of tau to microtubules: Highly phosphorylated tau does not bind to tubules, whereas dephosphorylated tau can bind to microtubules. It is interesting that the extent of tau phosphorylation in vivo has not been accurately determined. Tau was rapidly isolated from human temporal neocortex and hippocampus, rhesus monkey temporal neocortex, and rat temporal neocortex and hippocampus under conditions that minimized dephosphorylation. In brain slices, we observed that tau isolated under such conditions largely existed in several phosphorylated states, including a pool that was highly phosphorylated; this was determined using epitope-specific monoclonal and polyclonal antibodies. This highly phosphorylated tau was dephosphorylated during a 120-min time course in vitro, presumably as a result of neuronal phosphatase activity. The slow-mobility forms of tau were shifted to faster-mobility forms following in vitro incubation with alkaline phosphatase. Laser densitometry was used to estimate the percent of tau in slow-mobility, highly phosphorylated forms. Approximately 25% of immunoreactive tau was present as slow-mobility (66- and 68-kDa) forms of tau. The percentage of immunoreactive tau in faster-mobility pools (42-54 kDa) increased in proportion to the decrease in content of 66-68-kDa tau as a function of neuronal phosphatases or alkaline phosphatase treatment. These data suggest that the turnover of phosphorylated sites on tau is rapid and depends on neuronal phosphatases. Furthermore, tau is highly phosphorylated in normal-appearing human, primate, and rodent brain.(ABSTRACT TRUNCATED AT 250 WORDS)

Cardiac noradrenergic mechanisms mediate GABA-enhanced ouabain cardiotoxicity

Peripherally administered gamma-aminobutyric acid (GABA) alters cardiovascular function and has been reported to enhance ouabain-induced cardiotoxicity in vivo. Control and reserpinized rat hearts were perfused in vitro to determine if GABA directly evokes bradycardia by GABAA receptors, interacts with ouabain, and if noradrenergic mechanisms are required. Also, double-staining immunohistochemistry was employed to determine whether GABA-ergic and noradrenergic synthetic enzymes were juxtaposed within atrial tissue. The main results were as follows. GABA produced a dose-dependent bradycardia (p < 0.05) by stimulating GABAA receptors in Langendorff-perfused hearts. Reserpinized hearts were unresponsive (p < 0.05) to GABA, except at the highest dose (20 mg/ml). Ouabain-induced cardiotoxicity was enhanced (p < 0.05) by GABA in isolated control, but not reserpinized hearts. Lastly, glutamic acid decarboxylase and tyrosine hydroxylase immunoreactivities were in close proximity in atrial slices. Collectively, the results document that GABA causes bradycardia and enhances ouabain cardiotoxicity by modulating noradrenergic mechanisms in the isolated rat heart. Since the synthetic enzymes for GABA and norepinephrine were in close proximity in atrial tissue, these transmitters may interact under physiological conditions.

Biopsy-derived adult human brain tau is phosphorylated at many of the same sites as Alzheimer's disease paired helical filament tau

Tau from Alzheimer's disease (AD) paired helical filaments (PHF-tau) is phosphorylated at sites not found in autopsy-derived adult tau from normal human brains, and this suggested that PHF-tau is abnormally phosphorylated. To explore this hypothesis, we examined human adult tau from brain biopsies and demonstrated that biopsy-derived tau is phosphorylated at most sites thought to be abnormally phosphorylated in PHF-tau. These sites also were phosphorylated in autopsy-derived human fetal tau and rapidly processed rat tau. The hypophosphorylation of autopsy-derived adult human tau is due to rapid dephosphorylation postmortem, and protein phosphatases 2A (PP2A) and 2B (PP2B) in human brain biopsies dephosphorylate tau in a site-specific manner. The down-regulation of phosphatases (i.e., PP2A and PP2B) in the AD brain could lead to the generation of maximally phosphorylated PHF-tau that does not bind microtubules and aggregates as PHFs in neurofibrillary tangles and dystrophic neurites.

Enhanced elimination of biotinylated antibodies by avidin-based hemoperfusion in rats

We designed a hemoperfusion system using immobilized avidin for selective removal of biotinylated therapeutic antibodies from rats. Two prototype therapeutic antibodies, ovine antidigoxin Fab fragment and murine monoclonal 8A1 against Escherichia coli J5 lipopolysaccharide endotoxin, were biotinylated using biotin-long-chain-N-hydroxysuccinimide ester. Biotinylated antibodies were administered i.v. to anesthetized rats which were instrumented for measurement of cardiovascular parameters and connected to avidin-hemoperfusion devices. By using several different protocols of antibody administration and hemoperfusion, we found that the half-lives and areas under the time vs. concentration curves of biotinylated antibodies were reduced significantly after passage over immobilized avidin. Avidin hemoperfusion was not associated with adverse changes in blood pressure, heart rate or excess hemolysis. These data suggest that avidin hemoperfusion affords a means for selective removal of biotinylated ligands from serum, and that other therapeutic and potentially toxic compounds could be removed in this manner.

Calcineurin immunoreactivity in Alzheimer's disease

Aberrant phosphorylation of tau is linked to formation of the paired helical filaments (PHF) seen in Alzheimer's disease. Protein kinases such as mitogen-activated protein kinase, and calcium-regulated protein kinases may, in part, be responsible for addition of phosphate groups to serine residues of PHFtau; however, less is known concerning the phosphatases which regulate tau. In this report, we used several well-characterized antibodies to document calcineurin immunoreactivity in brain tissue from patients with Alzheimer's disease. We now report that levels of immunoreactive calcineurin are not significantly altered in neocortex and cerebellum of Alzheimer's patients relative to similar regions of age-matched controls. Immunocytochemical studies indicated that calcineurin immunoreactivity was present in dendrites and perikarya of many different neuronal populations in both control and Alzheimer brain. When specific antibodies against PHFtau were used in double-labeling experiments with anti-calcineurin antibodies, calcineurin immunoreactivity was seen in association with neurofibrillary tangles. However, calcineurin was not seen in all tangle bearing neurons. These data suggest that calcineurin levels per se are not significantly altered in Alzheimer's disease, but that calcineurin is distributed around some neurofibrillary tangles and may play a role in regulation of tau phosphorylation.

Adrenalectomy-induced alterations of calmodulin-dependent hippocampal adenylate cyclase activity: role of guanine nucleotide-binding proteins

Ca2+/calmodulin-dependent processes are altered by manipulations of the hypothalamic-pituitary-adrenal axis. In particular, adrenalectomy (ADX) attenuates hippocampal, but not cortical, calmodulin-dependent adenylate cyclase activity measured during the active (waking) phase of rats. The involvement of calmodulin- and guanine nucleotide (G)-binding proteins in the effects of ADX on the activity of calmodulin-dependent adenylate cyclase were investigated. In hippocampal membranes, inclusion of the GTP antagonist guanosine 5'-O-(2-thiodiphosphate) (250 microM) caused pronounced inhibition of calmodulin-stimulated adenylate cyclase activity. Guanosine 5(1)-O-(2-thiodiphosphate) had much smaller effects on calmodulin-independent (basal and forskolin-stimulated) enzyme activity. Substitution of Mn2+ for Mg2+ in the assay medium increased basal and forskolin-stimulated adenylate cyclase activity, but abolished calmodulin-dependent activation of this enzyme in both hippocampal and cortical membranes. These treatments blunted ADX-induced attenuation of hippocampal adenylate cyclase. ADX, with or without corticosterone administration (40 mg/kg, sc, once daily), failed to alter either Gi alpha or Gs alpha membrane protein content in either hippocampus or cortex. The levels of major membrane calmodulin-binding proteins in hippocampus and cortex also were not significantly altered by ADX. These results confirm that hormonal and biochemical regulation of calmodulin-dependent adenylate cyclase is distinct from that of other adenylate cyclase family members. Changes in Gs alpha and Gi alpha protein content alone cannot account for the effects of ADX on this enzyme. Overall, our studies suggest that the effects of ADX on calmodulin-dependent adenylate cyclase may occur through a reduction in the absolute amount of the catalytic subunit or an alteration(s) in the efficiency of coupling between adenylate cyclase and its modulators.

Ouabain cardiotoxicity is enhanced by GABA in anesthetized rats

Potential alteration of ouabain-induced cardiotoxicity by gamma-aminobutyric acid (GABA) in rats was tested by infusing ouabain for 10 min (0.7 mg/kg/min, i.v.) before or after continuous infusion of Ringer's solution with or without GABA (1 mg/min, i.v.). GABA evoked hypotension and bradycardia of similar magnitude under both conditions. The incidence of ouabain-induced ventricular fibrillation (VF) or cardiac arrest (CA) was similar in both groups. However, the time intervals to onset of VF and CA, in rats given ouabain before, but not after, GABA were shorter than in rats treated with Ringer's solution (p < 0.05). In experiments where baclofen (0.034 mg/min, i.v.) was infused after ouabain, hypotension and bradycardia occurred, but the incidence and times of ouabain-induced VF and CA were similar to control values. These results suggest that the enhancement in ouabain cardiotoxicity was mediated by GABAA receptors.

Okadaic acid induces hyperphosphorylated forms of tau protein in human brain slices

Hyperphosphorylated forms of the microtubule-associated protein tau are components of the paired helical filaments (PHFs) seen in patients with Alzheimer's disease. Slices of human lateral temporal cortex were obtained from tissues removed incidental to resections for intractable hippocampal epilepsy. Tau phosphorylation in temporal lobe slices was determined using mobility shifts after sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunodetection with the monoclonal antibodies Alz-50, 5E2, and Tau-1. The results indicate that tau phosphorylation was altered in a dose-dependent manner by the phosphatase inhibitor okadaic acid, but not by N-methyl-D-aspartate, quisqualate, or kainate. The slowest mobility forms of tau, termed "PHF-like tau," produced by okadaic acid treatment were dephosphorylated by purified protein phosphatase 2B (calcineurin). Formation of PHF-like tau peptides was blocked by KN-62, 1[N,O-bis(1,5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazi ne, an inhibitor of Ca2+/calmodulin-dependent protein kinase II. The protein kinase inhibitor staurosporine also prevented formation of PHF-like tau. These data suggest that phosphorylation of tau is regulated by Ca(2+)-dependent protein kinases and okadaic acid-sensitive protein phosphatases, alterations of which may be implicated in the pathogenesis of Alzheimer's disease.

Molecular neurotoxicology of trimethyltin: identification of stannin, a novel protein expressed in trimethyltin-sensitive cells

The molecular basis of selective vulnerability of specific neuronal populations to neurotoxicants remains a key focus in neurotoxicology. Trimethyltin (TMT) selectively damages neurons in rodent and human central nervous system after a single exposure. By coupling subtractive hybridization with molecular cloning techniques, we isolated a cDNA specifically localized in TMT-sensitive cells. This 2.9-kilobase cDNA encodes a putative 10-kDa peptide of 88 amino acids, termed "stannin." In immunocytochemical experiments, antisera raised against the amino terminus of stannin exhibited strong immunoreactivity in TMT-sensitive neurons in the hippocampus and entorhinal cortex, areas previously identified by in situ hybridization. Northern blot and in situ hybridization experiments detected a 3.0-kilobase stannin mRNA in brain, spleen, and kidney; expression occurred as early as embryonic day 15 in rat brain and thymus. In situ hybridization in human hippocampus demonstrated a stannin mRNA in pyramidal and dentate gyrus neurons. High stringency Southern blot analysis of genomic DNA identified stannin homologs in rabbit, Drosophila, and human. These findings indicate that stannin is present in TMT-sensitive cells and may play a role in the selective toxicity of organotin compounds.

Pyruvate carboxylase in genetic obesity

Immunoblotting and protein microsequencing were used to identify several adipocyte proteins expressed in an obesity-related fashion in the Zucker rat. One of these was a 116-kDa particulate protein (p116). The p116 levels in adipocytes from 5- to 7-wk-old obese Zucker rats were two- to fivefold higher on a per milligram of protein basis than levels in lean animals and decreased after the induction of streptozotocin-induced diabetes mellitus. This suggests the change may be related to the actions of insulin. Hepatic levels of p116 did not change. The p116 was purified to homogeneity from obese Zucker rat adipocytes, and polyclonal antisera were prepared against the purified protein in rabbits. Microanalysis of electroblotted p116 proteolytic fragments suggested that p116 was pyruvate carboxylase (PC). Other evidence that p116 was PC included the following: 1) p116 contained biotin, 2) p116 in particulate subcellular fractions was soluble after freeze-lysis, 3) antibodies to p116 reacted with purified hepatic PC, 4) p116 and purified hepatic PC had identical pI and relative molecular weight values, and 5) similar changes were detected in adipocyte p116 and PC enzyme activity during obesity and after the induction of streptozotocin-induced diabetes mellitus. Increased adipose tissue PC probably contributes to the increased lipogenic capacity of young obese Zucker rat adipocytes.

Distribution and expression of SNAP-25 immunoreactivity in rat brain, rat PC-12 cells and human SMS-KCNR neuroblastoma cells

Immunocytochemical, immunoblotting and in situ hybridization studies were used to map the distribution of SNAP-25 protein and mRNA in the rodent nervous system. These experiments demonstrated that subsets of neurons expressed SNAP-25, and that several patterns of expression emerged: SNAP-25 expression in caudate nucleus was initially concentrated in axons, which subsequently was localized in presynaptic regions of these axons. Other regions, typified by neocortex, showed developmental increases and persistent adult neuronal immunoreactivity for SNAP-25. Finally, olfactory bulb contained neurons which initially expressed SNAP-25, but lost expression during maturation. Additional studies in cultured human and rat cell lines derived from neural crest suggested that SNAP-25 is expressed in such lines, but not in glial or fibroblast lines. Differentiation of rat PC-12 cells with nerve growth factor failed to alter steady-state levels of SNAP-25 protein; similar responses were seen in human SMS-KCNR neuroblastoma cells differentiated using retinoic acid. The presence of SNAP-25 in presynaptic regions of numerous neuronal subsets and in neural crest cell lines suggests that this protein subserves an important function in neuronal tissues.

Expression of the calmodulin-dependent protein phosphatase, calcineurin, in rat brain: developmental patterns and the role of nigrostriatal innervation

The distribution of neurons expressing the calmodulin-dependent protein phosphatase, calcineurin (CN) was characterized in developing and adult rat brain using a combination of immunocytochemical, immunoblot and in situ hybridization approaches. Immunoblot analysis revealed a strong increase postnatally in CN protein expression. Four differently-charged isoforms of CN were observed in adult brain with apparent regional differences in isoform expression. Immunocytochemistry showed highest levels of CN in hippocampus, striatum, substantia nigra, amygdala and septal nuclei with immunoreactivity first appearing in striatum and septal nuclei, followed by hippocampus, neocortex and limbic structures. In situ hybridization demonstrated that mRNA for the catalytic subunit of CN was seen as early as postnatal day (PND) 1 in striatum, cortex and hippocampus. Since immunoreactivity was not detectable until day 4, this suggests that mRNA expression may precede that of protein by several days in these regions. Lesioning of developing and adult nigrostriatal dopamine neurons either with 6-hydroxydopamine or by surgical hemitransection had little effect on expression of CN, suggesting that CN expression is not influenced transsynaptically by dopamine. Collectively, these findings demonstrate that CN protein and mRNA expression are subject to regional and temporal control during brain development suggesting that specific synaptic connections may influence CN gene expression. However, in striatum, dopaminergic innervation does not appear to affect CN levels.

Interspecies comparison of calmodulin binding proteins throughout the gastrointestinal tract: comparison with human colon adenomas and adenocarcinomas

Calmodulin is an ubiquitous cytoplasmic protein which mediates many of the actions of calcium on intestinal tissue including regulation of growth and differentiation of normal and neoplastic cells. Using a biotinylated calmodulin overlay system, we compared the pattern of calmodulin binding proteins throughout the gastrointestinal tract of mice, rats, rabbits, and humans, and in human colonic adenomas and adenocarcinomas. A common calmodulin binding protein of 67 kDa was found in membrane and cytosolic fractions of oesophagus, stomach, proximal and distal small intestine, and colon from all four species. In human tissue this 67 kDa protein was present in greatest concentration in stomach tissue. Furthermore, a 67 kDa binding protein was the major calmodulin binding protein from human stomach and ileum as determined by ion exchange and calmodulin affinity chromatography. A similar pattern of binding proteins was noted between rabbit and human cytosolic fractions; proteins of 60/67 kDa and 105 kDa were present in stomach tissue. A 94 kDa protein was present in samples of rabbit and human ileum but not of mouse or rat. A similar pattern of calmodulin binding proteins was seen in normal and neoplastic large bowel tissue, apart from one of nine adenocarcinomas, where a distinct 54 kDa band was noted in both cytosolic and membrane fractions. The results of this study show interspecies and organ differences between calmodulin binding proteins, but suggest that a 67 kDa protein is the major binding protein present throughout normal gastro-intestinal tract and neoplastic human tissue.

Expression of calmodulin-dependent enzymes in developing rat striatum is not affected by perturbation of dopaminergic systems

Transsynaptic regulation is one mechanism that controls expression of several calmodulin (CaM)-dependent enzymes. This observation and the demonstration that expression of several CaM-dependent enzymes in developing striatum occurred with a spatial and temporal pattern similar to that seen for dopamine and tyrosine hydroxylase suggested that the nigrostriatal pathway may influence the expression of CaM-binding proteins (CaM-BPs) during striatal development. Therefore, the possible role of nigrostriatal dopamine systems regulating the expression of CaM-dependent enzymes was studied in Sprague-Dawley rats by using surgical hemitransections of brain, 6-hydroxydopamine lesions, and chronic haloperidol treatments. Alterations in CaM-BP expression following perturbation of the developing nigrostriatal tract were analyzed by using immunoblots, biotinylated CaM overlays, and enzyme assays. The extent of nigrostriatal lesions was assessed by using depletion of immunoreactive tyrosine hydroxylase levels in striatum. All three experimental paradigms failed to alter the normal developmental expression of CaM-dependent enzymes. From these results we conclude that the increased expression of CaM-dependent enzymes during striatal development is not directly dependent on synaptic input from the nigrostriatal dopamine system.

Developmental expression of the 25-kDa synaptosomal-associated protein (SNAP-25) in rat brain

The developmental expression and subcellular distribution of the neuron-specific 25-kDa synaptosomal protein (SNAP-25) were investigated by using Northern (RNA) blots, immunoblots, and immunocytochemistry. Both SNAP-25 protein and mRNA were present at low levels in embryonic day 15 rat brain, and levels of both increased during early postnatal maturation. Developmental immunoblots with antipeptide antisera demonstrated that a 25-kDa peptide was the major isoform in brain, and this form increased steadily from embryonic day 15 through adulthood. A second 27-kDa immunoreactive isoform was present in brain only during early development. Immunoblots of two-dimensional SDS/polyacrylamide gels revealed the presence of a predominant 25-kDa isoform of SNAP-25 in adult brain. Immunocytochemical studies indicated that as immunoreactivity for SNAP-25 increased during development, the cellular localization of SNAP-25 immunoreactivity concomitantly shifted from axons and cell bodies to presynaptic terminals. These data suggest that the SNAP-25 protein shifts in subcellular localization during development and may play a role in the establishment and stabilization of specific presynaptic terminals in brain.

Preparation, characterization and biological properties of biotinylated derivatives of calmodulin

Biotinylated derivatives of calmodulin (CaM) were prepared and their biological properties characterized by using enzyme assays, affinity and hydrophobic-interaction chromatography. Several N-hydroxysuccinimidobiotin derivatives [sulphosuccinimidobiotin (sulpho-NHS) and sulphosuccinimido-6-(biotinamido)hexanoate (BNHS-LC)] differing in spacer arm length were used to modify CaM. The shorter-spacer-arm CaM derivative (sulpho-CaM) activated CaM-dependent cyclic nucleotide phosphodiesterase and CaM-dependent protein kinase II; preincubation with avidin blocked its ability to activate these enzymes. The extended-spacer-arm derivative (BNHS-LC-CaM) activated CaM-dependent enzymes both in the presence and in the absence of avidin, suggesting that the longer spacer arm diminished steric effects from avidin preincubation. Other biotinylated CaM derivatives were prepared with biotinylated tyrosine and/or histidine residues (diazobenzoylbiocytin; DBB-CaM) or nucleophilic sites (photobiotin acetate; photo-CaM). These derivatives activated CaM-dependent enzymes in the presence and in the absence of avidin. Oriented affinity columns were constructed with covalently immobilized avidin complexed to each biotinylated CaM derivative. The chromatographic profiles obtained revealed that each column interacted with a specific subset of CaM-binding proteins. Elution profiles of biotinyl CaM derivatives on phenyl-Sepharose hydrophobic-interaction chromatography suggested that several derivatives displayed diminished binding to the matrix in the presence of Ca2+. Development and characterization of a series of biotinylated CaM molecules can be used to identify domains of CaM that interact with specific CaM-dependent enzymes.

Developmental expression of calmodulin-dependent cyclic nucleotide phosphodiesterase in rat brain

The patterns of expression of calmodulin-dependent cyclic nucleotide phosphodiesterase (CaM-PDE) have been studied in developing and adult rat brain using affinity-purified polyclonal antibodies against CaM-PDE. An immunocytochemical map of adult brain regions expressing CaM-PDE, constructed from serial coronal brain sections, illustrated that CaM-PDE was expressed in specific neuronal subpopulations throughout the adult rat brain. Immunoblot analysis coupled with subcellular fractionation indicated that CaM-PDE was primarily localized to cytoplasmic fractions, with a small amount associated with synaptosomal membranes. Immunoblots from developing brain indicated that CaM-PDE expression increased dramatically during postnatal days 7-20 (PND 7-20); parallel increases in CaM-PDE enzyme activity occurred during this same time. Immunocytochemical studies indicated that several distinct patterns of CaM-PDE expression occurred during development. Neocortex showed low levels of CaM-PDE immunoreactivity in neuronal somata of layers III, V and VI on PND 4 that increased by PND 11; the adult somatodendritic pattern of immunoreactivity was observed by PND 60. Similar patterns were observed in cerebellar Purkinje cells, with somatodendritic staining observed by PND 12. By contrast, caudate-putamen, the inferior olive and the hypoglossal nuclei expressed high levels of CaM-PDE on PND 4, with levels considerably lower in the adult animal. The different patterns of expression suggest that in neocortex and cerebellum, CaM-PDE increases during the period of neuronal differentiation and active synaptogenesis, while in the caudate-putamen, inferior olive and hypoglossal nucleus, high levels may be required early in development.

Use of avidin-biotin subtractive hybridization to characterize mRNA common to neurons destroyed by the selective neurotoxicant trimethyltin

Trimethyltin (TMT), a selective neurotoxicant, destroys a distinct subpopulation of neurons which possess no known biochemical or anatomic linkage. However, TMT-sensitive neurons may share common gene products related to susceptibility. In an effort to isolate mRNAs common to TMT-sensitive neurons, avidin/biotin based-subtractive hybridization was used to generate a cDNA library specifically related to TMT-toxicity. Out of 50 cDNAs, two clones hybridized only to poly(A+) mRNA isolated from the brains of saline-treated rats. Two of these cDNAs, p9T10 and p9T19, were used for in situ hybridization; both hybridized to hippocampus, limbic cortex, amygdala and other regions destroyed by TMT, suggesting that these probes identified mRNA enriched in TMT-sensitive neurons. The patterns of in situ hybridization coupled with the loss of p9T10 and p9T19 hybridization to mRNA isolated from the brains of TMT-treated rats suggests that one or both of these two clones may represent mRNA found in neurons damaged by TMT. The combination of selective neurotoxic lesions followed by cDNA subtractive hybridization should prove to be a useful strategy for the isolation of gene products from specific neuronal populations.

Developmental expression of neuronal calmodulin-binding proteins in rat brain

The developmental patterns of calmodulin-binding proteins (CaM-BPS) in rat brain were examined using biotinylated calmodulin overlays of one- and two-dimensional gels. Hippocampus showed the earliest onset of CaM-BP expression (postnatal day 5; PND5), followed by cerebral cortex and striatum, both of which had detectable levels of CaM-BPs by PND7. Cerebellum had the latest onset of CaM-BP expression; CaM-BPs were not detectable until PND9. Very few CaM-BPs were present in brain before PND5 and all regions reached near adult levels by PND20. However, several unique CaM-BPs were seen in embryonic brain and these proteins may have an important role in developing neurons. These data suggest an orderly, complex expression of CaM-BPs which increases during times of synaptogenesis and synaptic maturation.

Evidence for transsynaptic regulation of calmodulin-dependent cyclic nucleotide phosphodiesterase in cerebellar Purkinje cells

Calmodulin-dependent phosphodiesterase (CaM-PDE) is selectively expressed in specific neuronal populations in adult rat brain. In cerebellar cortex, it is expressed at high levels in Purkinje cells (soma and dendrites). Climbing fiber ablation by intraperitoneal injections of 3-acetylpyridine resulted in a selective depression of cerebellar CaM-PDE expression using Western immunoblot procedures; neither calcineurin (calmodulin-dependent protein phosphatase) nor other calmodulin binding proteins, detected by biotinylated calmodulin overlays, were affected. Immunocytochemical staining of cerebellum revealed a loss of detectable CaM-PDE immunoreactivity in Purkinje cells, with no appreciable change in calcineurin immunoreactivity. Cerebral cortex was examined as a control for a direct effect of 3-acetylpyridine on CaM-PDE expression, independent of climbing fiber deafferentation. There were no detectable changes in CaM-PDE or calcineurin immunoreactivity in cortical pyramidal cells, and no changes were detected, either in Western blot analyses for CaM-PDE or calcineurin or in biotinylated calmodulin overlays. These data suggest that CaM-PDE expression in Purkinje cells is regulated transsynaptically by climbing fiber inputs.

Changes in expression of tyrosine hydroxylase immunoreactivity in human SMS-KCNR neuroblastoma following retinoic acid or phorbol ester-induced differentiation

Human SMS-KCNR cells differentiated in response to either retinoic acid or phorbol esters; differentiated cells extended numerous, complex neurites and showed reduced proliferation. Tyrosine hydroxylase (TH) immunoreactivity was measured in this cell line following treatment with retinoic acid (1-10 microM), 12-O-tetradecanoyl-phorbol-13-acetate (TPA; 16-160 nM), or combinations of these agents. After 21 days of treatment with either TPA or retinoic acid (RA), TH immunoreactivity was measured in this using densitometric scans of Western blots, was doubled relative to untreated or serum-deprived SMS-KCNR cultures. Increases in TH immunoreactivity could be detected after 6 days of treatment. Treatment with RA for 3 days followed by phorbol esters for an additional 3 days resulted in a 3-fold increase in TH immunoreactivity at day 6; reversing the order of drug treatment did not have this effect. Treatment of cultures with the divalent cationophore A23187 caused treated cells to retract neurites; expression of TH immunoreactivity was decreased relative to drug-treated and control cultures. These results suggest that retinoic acid treatment may 'prime' SMS-KCNR cells for the subsequent effects of phorbol esters, and indicate that the patterns of biochemical differentiation induced by TPA or RA are different.

Expression of calmodulin-dependent phosphodiesterase, calmodulin-dependent protein phosphatase, and other calmodulin-binding proteins in human SMS-KCNR neuroblastoma cells

Calmodulin (CaM)-dependent enzymes, such as CaM-dependent phosphodiesterase (CaM-PDE), CaM-dependent protein phosphatase (CN), and CaM-dependent protein kinase II (CaM kinase II), are found in high concentrations in differentiated mammalian neurons. In order to determine whether neuroblastoma cells express these CaM-dependent enzymes as a consequence of cellular differentiation, a series of experiments was performed on human SMS-KCNR neuroblastoma cells; these cells morphologically differentiate in response to retinoic acid and phorbol esters [12-O-tetradecanoylphorbol 13-acetate (TPA)]. Using biotinylated CaM overlay procedures, immunoblotting, and protein phosphorylation assays, we found that SMS-KCNR cells expressed CN and CaM-PDE, but did not appear to have other neuronal CaM-binding proteins. Exposure to retinoic acid, TPA, or conditioned media from human HTB-14 glioma cells did not markedly alter the expression of CaM-binding proteins; 21-day treatment with retinoic acid, however, did induce expression of novel CaM-binding proteins of 74 and 76 kilodaltons. Using affinity-purified polyclonal antibodies, CaM-PDE immunoreactivity was detected as a 75-kilodalton peptide in undifferentiated cells, but as a 61-kilodalton peptide in differentiated cells. CaM kinase II activity and subunit autophosphorylation was not evident in either undifferentiated or neurite-bearing cells; however, CaM-dependent phosphatase activity was seen. Immunoblot analysis with affinity-purified antibodies against CN indicated that this enzyme was present in SMS-KCNR cells regardless of their state of differentiation. Although SMS-KCNR cells did not show a complete pattern of neuronal CaM-binding proteins, particularly because CaM kinase II activity was lacking, they may be useful models for examination of CaM-PDE and CN expression. It is possible that CaM-dependent enzymes can be used as sensitive markers for terminal neuronal differentiation.

Trimethyltin-induced neuronal damage in the rat brain: comparative studies using silver degeneration stains, immunocytochemistry and immunoassay for neuronotypic and gliotypic proteins

Trimethyltin is a neurotoxicant which produces a distinct pattern of neuronal cell death following peripheral administration of a single dose (8 mg/kg, i.p.) in rats. The cupric-silver degeneration stain was used to produce an atlas documenting the distribution and time course of trimethyltin-induced neuronal damage in adult, male Long-Evans rats. Animals were examined at survival times of 1, 2, 3, 4, 5, 7, 10 and 18 days after intoxication. The earliest degeneration was observed at day 1 in the intermediate and ventral divisions of the lateral septal nucleus, followed by development of degeneration on days 2-4 in neuron populations including the septohippocampal nucleus, septohypothalamic nucleus, anterior olfactory nucleus, bed nucleus of the stria terminalis, endopiriform nucleus, parafascicular nucleus, superior colliculus, interstitial nucleus of the posterior commissure, inferior colliculus, pontine nuclei, raphe nuclei, pars caudalis of the spinal trigeminal nucleus, the caudal aspect of nucleus tractus solitarius, dorsal vagal motor nucleus, granule cells in the dentate gyrus, pyramidal cells in CA fields of the hippocampus, and of neurons in the subiculum, pyriform cortex, entorhinal cortex and neocortex (mainly layer Vb and VI). This was followed by degenerative changes on days 5-7 in other structures, including the amygdaloid nuclei, the ventral posterolateral and ventral posteromedial thalamic nuclei and the periaqueductal gray. The distribution of terminal degeneration from these neurons indicate that specific pools of cells are affected in each structure, and the time course suggests somatofugal degeneration. The trimethyltin damage was also assessed with immunocytochemical visualization of a neuronotypic protein, protein-O-carboxyl methyltransferase and a radioimmunoassay for glial fibrillary acidic protein. Protein-O-carboxyl methyltransferase immunoreactivity was altered in neuronal populations damaged by trimethyltin, but did not appear to be either as sensitive or selective an assay of neuronal damage as the silver stain, especially at short survival times. Glial fibrillary acidic proteins were dramatically elevated 21 days after trimethyltin intoxication, particularly in areas of extensive damage. These studies revealed advantages and problems encountered in the use of each technique in assessing neurotoxic effects, forming a basis for discussion of the relative merits of using a battery of specific molecular probes for neurotoxicity evaluations.

Calmodulin-binding proteins in human Y-79 retinoblastoma and HTB-14 glioma cell lines

The Y-79 human retinoblastoma cell line has been used as a model system for studying differentiation of primitive neuroectodermal cells into either glial-like (glial fibrillary acidic protein positive) or neuron-like (neuron-specific enolase-positive) cells. To determine whether Y-79 retinoblastoma cells express neuronotypic calmodulin-binding proteins, Y-79 cells were either treated with butyrate or dibutyryl cyclic AMP (dbcAMP) in serum-containing medium or were maintained in serum-free media. Using a biotinylated calmodulin blot overlay technique, we found that Y-79 cells treated with dbcAMP or butyrate expressed low levels of membrane-bound calmodulin-binding proteins of 150, 147, 127, and 126 kilodaltons (kDa); butyrate-treated cells also expressed a calmodulin-binding peptide of 135 kDa. Since butyrate treatment of Y-79 cells induces the expression and the secretion of interphotoreceptor retinoid-binding protein (IRBP, 140 kDa), we tested the hypothesis that the calmodulin-binding protein of 135 kDa induced by butyrate treatment was IRBP. Purified bovine IRBP did not bind calmodulin; further, the 135-kDa calmodulin binding protein was not immunoreactive with antisera directed against IRBP. Since dbcAMP and butyrate induce some glial-like characteristics in Y-79 cells, we compared the calmodulin-binding protein pattern in these cells with that seen in human HTB-14 glioma cells. The HTB-14 line did not express calmodulin-binding proteins, even after treatments with agents that induce morphologic change in these cells. Thus, we conclude that Y-79 cells express membrane-bound calmodulin-binding proteins, but in a pattern different from that seen with adult, differentiated neurons or from human HTB-14 glioma cells.

Differential expression of calmodulin-binding proteins in B, T lymphocytes and thymocytes

Changes in intracellular free Ca2+ are involved in the transmembrane signalling of different cells, including lymphocytes. Since calmodulin (CaM) is a primary receptor for Ca2+ (ref. 4), it may mediate the activation of crucial enzymes after antigen-induced increases in cytosolic Ca2+. Using a biotinylated-CaM (Bio-CaM) detection procedure to identify such proteins, we found that a peptide of relative molecular mass 59,000 (59K) was the predominant soluble CaM-binding protein (CaM-BP) in T cells and B lymphocytes from murine spleen; immunoblotting experiments identified it as a subunit of the CaM-dependent phosphatase, 'calcineurin' (CN). Smaller amounts of larger CaM-BPs, thought to be cytoskeletal-binding proteins, were also detected. CaM-BPs were expressed differentially, with B lymphocytes having four times more of the CN-like protein than T lymphocytes, while in thymocytes, a 65K polypeptide was the major CaM-BP. However, limited proteolysis analysis suggested that this thymus-specific peptide may be a precursor of CN. These data suggest that Ca2+-stimulated protein dephosphorylation may be an important and highly regulated function in lymphoid cells.

Differential localization of calmodulin-dependent enzymes in rat brain: evidence for selective expression of cyclic nucleotide phosphodiesterase in specific neurons

High-affinity antibodies against calmodulin (CaM)-dependent cyclic nucleotide phosphodiesterase and protein phosphatase (calcineurin) were purified and characterized. Rabbit anti-phosphodiesterase antibody did not react with other phosphodiesterases or with the regulatory subunits of cAMP-dependent protein kinase. Affinity-purified goat anti-calcineurin antibody recognized both the 61-kDa catalytic subunit and the 18-kDa Ca2+-binding subunit of the phosphatase. Neither antibody reacted with CaM, several CaM-binding proteins (calmodulin-dependent protein kinase, myosin light chain kinase, fodrin), or other cytosolic proteins from brain. The antibodies were used to compare the cellular localization of these two CaM-dependent enzymes in rat brain. Both calcineurin and phosphodiesterase were found predominantly in nerve cells; however, phosphodiesterase was restricted to very specific neuronal populations. Phosphodiesterase was prominent in the somatic cytoplasm and dendrites of regional output neurons--e.g., cerebellar Purkinje cells and hippocampal and cortical pyramidal cells. The extensive and uniform staining in the dendrites was consistent with postsynaptic localization and suggested an important function for this enzyme in neurons that integrate multiple convergent inputs. Calcineurin was present in virtually all classes of neurons, with immunoreactivity confined primarily to cell bodies. Both diffuse cytoplasmic staining and characteristic punctate staining of cell bodies were observed; the latter suggested compartmentalization of calcineurin at or near the plasma membrane. The results of this study demonstrate that calcineurin and phosphodiesterase are differentially localized in the central nervous system. Thus, the expression and compartmentalization of CaM-binding proteins may be highly regulated and specific for particular differentiated nerve cell types.

Protein-O-carboxylmethyltransferase in the rat brain: high regional levels in the substantia nigra, locus coeruleus and paraventricular nucleus

Immunocytochemical techniques were used to localize protein-O-carboxylmethyltransferase in the rat brain. Particularly high levels of immunoreactive protein-O-carboxylmethyltransferase were found in the paraventricular and supraoptic nucleus, the substantia nigra and the locus coeruleus. The enhanced expression of the methyltransferase in these brain regions suggests that protein carboxylmethylation is of particular importance in these areas. These findings are consistent with previous biochemical studies which suggest that protein methylation plays a role in presynaptic monoaminergic neurons and in the release and/or processing of neurohypophyseal peptides.

A rapid and sensitive method for detection and quantification of calcineurin and calmodulin-binding proteins using biotinylated calmodulin

Purified bovine brain calmodulin was biotinylated with biotinyl-epsilon-aminocaproic acid N-hydroxysuccinimide. Biotinylated calmodulin was used to detect and quantify calmodulin-binding proteins following both protein blotting and slot-blot procedures by using alkaline phosphatase or peroxidase coupled to avidin. When purified bovine brain calcineurin, a calmodulin-dependent protein phosphatase, was immobilized on nitrocellulose slot blots, biotinylated calmodulin bound in a calcium-dependent saturable manner; these blots were then quantified by densitometry. Biotinylated calmodulin was able to detect as little as 10 ng of calcineurin, and the binding was competitively inhibited by addition of either native calmodulin or trifluoperazine. When biotinylated calmodulin was used to probe protein blots of crude brain cytosol and membrane preparations after gel electrophoresis, only protein bands characteristic of known calmodulin-binding proteins (i.e., calmodulin-dependent protein kinase, calcineurin, spectrin) were detected with avidin-peroxidase or avidin-alkaline phosphatase procedures. Purified calcineurin was subjected to one- and two-dimensional gel electrophoresis and protein blotting; as expected, only the 61-kDa calmodulin-binding subunit was detected. When the two-dimensional protein blot was incubated with biotinylated calmodulin and detected with avidin-alkaline phosphatase, several apparent forms of the 61-kDa catalytic subunit were detected, consistent with isozymic species of the enzyme. The results of these studies suggest that biotinylated calmodulin can be used as a simple, sensitive, and quantifiable probe for the study of calmodulin-binding proteins.

Is Ca2+-calmodulin-dependent protein phosphorylation in rat brain modulated by carboxylmethylation?

Calmodulin stimulation of protein kinase activity in calmodulin-depleted preparations of rat brain cytosol or synaptosomal membranes was attenuated by prior carboxylmethylation of the enzyme source with purified protein-O-carboxylmethyltransferase. Similarly, calmodulin stimulation of highly purified Ca2+-calmodulin-dependent protein kinase was reduced if the kinase was exposed to methylating conditions prior to addition of calmodulin. Biochemical and acidic sodium dodecyl sulfate-gel electrophoretic analyses indicated that all sources of protein kinase activity were substrates for methylation. The specific activity of methyl group incorporation into protein kinase increased with increasing purity of the preparation, reaching values of 1.72 pmol CH3/micrograms protein or 0.15-1.12 mol CH3/mol of holoenzyme. Analysis of ATP binding in cytosol with the use of the photoaffinity probe [32P]8-azido-ATP indicated that carboxylmethylation reduced ATP binding. These results suggest that carboxylmethylation of Ca2+-calmodulin protein kinase may modulate the activity of this enzyme in rat brain.

Immunohistochemical localization of protein-O-carboxylmethyltransferase in rat brain neurons

The distribution of the enzyme protein-O-carboxylmethyltransferase (EC 2.1.1.24) has been investigated in the rat brain using both immunohistochemical and biochemical techniques. The enzyme, which carboxylmethylates free aspartic and glutamic acid residues of protein substrates, was localized in neurons, but not other cell types throughout the brain. The highest immunoreactivity was detected throughout the cortex, followed by the hippocampus, the corpus striatum, the thalamus and the amygdala. Immunoreactive cells were detected in other brain regions but were not as prominent as those regions listed above. The distribution of immunoreactivity in the hippocampus was most striking, with considerable labelling of the pyramidal and granule cells in all regions. Numerous pyramidal cells were labelled in the cerebral cortex, with some ascending processes exhibiting immunoreactivity. The corpus striatum was uniformly labelled, suggesting that the enzyme was not localized to any specific neurotransmitter system. The antisera employed in this study was generated against purified bovine brain protein-O-carboxylmethyltransferase and Western immunoblot analysis showed cross reactivity against both rat brain and human erythrocyte forms of the enzyme. Enzyme activity and methyl acceptor protein capacity were examined in 1.5 mm coronal sections of rat brain. The regions with highest enzyme activities were found in cross-sections containing cortex and corpus striatum or cortex and hippocampus. The lowest enzyme activities were noted in slices of brainstem and cerebellum, areas exhibiting low amounts of immunoreactive protein-O-carboxylmethyltransferase. Methyl acceptor protein capacity was highest in slices of cortex and corpus striatum, cortex and hippocampus and was lowest in slices of brainstem and cerebellum. These results demonstrate that protein-O-carboxylmethyltransferase has an unique neuronal pattern of distribution in the rodent central nervous system, and suggest that the carboxylmethylation of proteins may be of functional significance in these neurons.

Thermal denaturation of rat pulmonary and testicular angiotensin-converting enzyme isozymes. Effects of chelators and CoCl2

In an attempt to assess the biochemical consequences resulting from structural differences between rat pulmonary and testicular angiotensin-converting enzyme, the thermal stability of crude and purified preparations of each enzyme was compared. Structural heterology was verified by molecular weight determinations and by peptide mapping after limited proteolysis with Staphylococcus V8 proteinase. Thermal stability was monitored by changes in catalytic activity following incubations at 55 degrees C in the presence of chelators and CoCl2. Purified pulmonary angiotensin-converting enzyme was more sensitive to inhibition by the chelators EDTA and 1,10-phenanthroline and by the site-directed inhibitor captopril than was the testicular isozyme. Although the pulmonary holoenzyme was unaffected by cobalt, the testicular holoenzyme was inhibited by cobalt in a concentration-dependent manner. Crude pulmonary angiotensin-converting enzyme was significantly more resistant to thermal denaturation than its crude testicular counterpart. The differences in the thermal lability of each isozyme were still present in purified preparations, although the purified enzymes appeared to be more thermally stable than their crude counterparts. Both chelators and cobalt markedly potentiated the thermal denaturation of each isozyme. These data suggest that the structural heterology of the pulmonary and testicular isozymes may affect the interaction of zinc with the respective enzymes and that zinc may contribute to the structural integrity and thermal stability of angiotensin-converting enzyme in each tissue.