Cryostat Sections Of Rat Brain Research Articles

Mass spectrometry imaging of rat brain lipid profile changes over time following traumatic brain injury

BackgroundMild traumatic brain injury (TBI) is a common public health issue that may contribute to chronic degenerative disorders. Membrane lipids play a key role in tissue responses to injury, both as cell signals and as components of membrane structure and cell signaling. This study demonstrates the ability of high resolution mass spectrometry imaging (MSI) to assess sequences of responses of lipid species in a rat controlled cortical impact model for concussion. New methodA matrix of implanted silver nanoparticles was implanted superficially in brain sections for matrix-assisted laser desorption (MALDI) imaging of 50μm diameter microdomains across unfixed cryostat sections of rat brain. Ion-mobility time-of-flight MS was used to analyze and map changes over time in brain lipid composition in a rats after Controlled Cortical Impact (CCI) TBI. ResultsBrain MS images showed changes in sphingolipids near the CCI site, including increased ceramides and decreased sphingomyelins, accompanied by changes in glycerophospholipids and cholesterol derivatives. The kinetics differed for each lipid class; for example ceramides increased as early as 1 day after the injury whereas other lipids changes occurred between 3 and 7 days post injury. Comparison with existing method(s)Silver nanoparticles MALDI matrix is a sensitive new tool for revealing previously undetectable cellular injury response and remodeling in neural, glial and vascular structure of the brain. ConclusionsLipid biochemical and structural changes after TBI could help highlighting molecules that can be used to determine the severity of such injuries as well as to evaluate the efficacy of potential treatments.

Specific binding of [(18)F]fluoroethyl-harmol to monoamine oxidase A in rat brain cryostat sections, and compartmental analysis of binding in living brain.

We investigated [(18)F]fluoroethyl-harmol ([(18)F]FEH) as a reversible and selective ligand for positron emission tomography (PET) studies of monoamine oxidase A (MAO-A). Binding of [(18)F]FEH in rat brain cryostat sections indicated high affinity (KD = 3 nM), and density (Bmax; 600 pmol/g). The plasma free fraction was 45%, and untransformed parent constituted only 13% of plasma radioactivity at 10 min after injection. Compartmental analysis of PET recordings in pargyline-treated rats showed high permeability to brain (K1; 0.32 mL/g/min) and slow washout (k2; 0.024/min), resulting in a uniformly high equilibrium distribution volume (VD; 20 mL/g). Using this VD to estimate unbound ligand in brain of untreated rats, the binding potential ranged from 4.2 in cerebellum to 7.2 in thalamus. We also calculated maps of rats receiving [(18)F]FEH at a range of specific activities, and then estimated saturation binding parameters in the living brain. In thalamus, striatum and frontal cortex KD was globally close to 300 nM and Bmax was close to 1600 pmol/g; the 100-fold discrepancy in affinity suggests a very low free fraction for [(18)F]FEH in the living brain. Based on a synthesis of findings, we calculate the endogenous dopamine concentration to be 0.4 μM in the striatal compartment containing MAO-A, thus unlikely to exert competition against [(18)F]FEH binding in vivo. In summary, [(18)F]FEH has good properties for the detection of MAO-A in the rat brain by PET, and may present logistic advantages for clinical research at centers lacking a medical cyclotron. We made a compartmental analysis of [(18)F]fluoroethylharmol ([(18)F]FEH) binding to monoamine oxidase A (MAO-A) in living rat brain and estimated the saturation binding parameters from the binding potential (BPND). The Bmax was of comparable magnitude to that in vitro, but with apparent affinity (300 nM), it was 100-fold lower in vivo. PET imaging with [(18) F]FEH is well suited for quantitation of MAO-A in living brain.

Dendritic mRNA Targeting of Jacob and N-Methyl-d-aspartate-induced Nuclear Translocation after Calpain-mediated Proteolysis

Jacob is a recently identified plasticity-related protein that couples N-methyl-d-aspartate receptor activity to nuclear gene expression. An expression analysis by Northern blot and in situ hybridization shows that Jacob is almost exclusively present in brain, in particular in the cortex and the limbic system. Alternative splicing gives rise to multiple mRNA variants, all of which exhibit a prominent dendritic localization in the hippocampus. Functional analysis in primary hippocampal neurons revealed that a predominant cis-acting dendritic targeting element in the 3'-untranslated region of Jacob mRNAs is responsible for dendritic mRNA localization. In the mouse brain, Jacob transcripts are associated with both the fragile X mental retardation protein, a well described trans-acting factor regulating dendritic mRNA targeting and translation, and the kinesin family member 5C motor complex, which is known to mediate dendritic mRNA transport. Jacob is susceptible to rapid protein degradation in a Ca(2+)- and Calpain-dependent manner, and Calpain-mediated clipping of the myristoylated N terminus of Jacob is required for its nuclear translocation after N-methyl-d-aspartate receptor activation. Our data suggest that local synthesis in dendrites may be necessary to replenish dendritic Jacob pools after truncation of the N-terminal membrane anchor and concomitant translocation of Jacob to the nucleus.

The ‘allosteric modulator’ SCH‐202676 disrupts G protein‐coupled receptor function via sulphydryl‐sensitive mechanisms

1. Previous studies suggest that the thiadiazole compound SCH-202676 (N-(2,3-diphenyl-1,2,4-thiadiazol-5-(2H)-ylidene)methanamine) acts as an allosteric modulator of a variety of structurally distinct G protein-coupled receptors (GPCRs). It was postulated that SCH-202676 would directly bind a structural motif in the receptor molecule common to divergent members of the GPCR family. The molecular mechanisms of such a promiscuous action, however, remain obscure. 2. To clarify the mechanism of SCH-202676 action, we used the functional approach of [35S]GTPgammaS autoradiography with rat brain cryostat sections together with classical membrane [35S]GTPgammaS binding assays to evaluate how the thiadiazole affects G protein activity mediated by various receptors linked to the Gi-family of G proteins. 3. We found that in the absence of dithiotreitol (DTT), SCH-202676 (10(-7)-10(-5) M) elicits nonspecific effects in the [35S]GTPgammaS-based G protein activation assays, thereby severely compromising interpretations on the compounds ability to allosterically inhibit receptor-mediated G protein activity. Such a nonspecific behaviour was fully reversed upon addition of DTT (1 mM), revealing thiol-based mechanism of action. 4. In routine incubations containing DTT, SCH-202676 had no effect on receptor-driven G protein activity, as assessed for adenosine A1, alpha2-adrenergic, cannabinoid CB1, lysophosphatidic acid LPA1, muscarinic M2/M4, purinergic P2Y12 or sphingosine 1-phosphate receptors, suggesting that the thiadiazole does not act as an allosteric modulator of GPCR function. 5. 1H NMR analysis indicated that SCH-202676 underwent structural changes after incubation with the reducing agent DTT or with brain tissue. 6. We conclude that SCH-202676 modulates GPCRs via thiol modification rather than via true allosteric mechanisms.

Myelin contributes to the parallel orientation of axonal growth on white matter in vitro

BackgroundBrain and spinal cord white matter can support extensive axonal growth. This growth is generally constrained to an orientation that is parallel to the longitudinal axis of the fiber tract. This constraint is presumably due to permissive and non-permissive substrates that are interleaved with each other and oriented in parallel within the tract.ResultsEmbryonic chick sympathetic neurons were cultured on cryostat sections of rat brain and the orientation of neurite growth on white matter was assessed. To determine if haptotaxis is sufficient to guide parallel neurite growth, neurons were cultured under conditions designed to interfere with interactions between growing neurites and factors that act as biochemical contact guidance cues but not interactions with haptotactic cues. Under these conditions, neurites extending on white matter were not exclusively oriented in parallel to the fiber tract, suggesting that biochemical cues are involved. To assess the role of myelin in guiding parallel neurite growth, neurons were cultured on myelin-deficient corpus callosum. These neurons also extended neurites that were not constrained to a parallel orientation. Moreover, preincubation with NGF and treatment with cAMP analogs, manipulations that attenuate overall myelin-mediated inhibition of neurite growth, also led to a reduced parallel orientation of neurite growth.ConclusionsThe present studies suggest that some of the relevant factors that constrain axonal growth on white matter are not haptotactic in nature and appear to be partly mediated by factors that are associated with myelin and may involve myelin-associated "inhibitors".

Coupling between agonist and chloride ionophore sites of the GABA A receptor: agonist/antagonist efficacy of 4-PIOL

Eight γ-aminobutyric acid (GABA) mimetics were tested on their ability to differentiate native GABA A receptor subtypes present in various rat brain regions. In rat brain cryostat sections, little regional variations by the agonistic actions of muscimol, thiomuscimol, 4,5,6,7-tetrahydroisoazolo(5,4-c)pyridin-3-ol, piperidine-4-sulphonic acid, taurine and β-alanine on [ 35S] t-butylbicyclophosphorothionate ([ 35S]TBPS) binding to GABA A receptor channels were found. They were very similar to those found for GABA itself and indicated no direct correlation with single subunit distributions for any of these compounds. Only the low-efficacy GABA mimetic 5-(4-piperidyl)isoxazol-3-ol (4-PIOL) acted like a weak partial agonist or antagonist depending on the brain area. As the cerebellar granule cell layer was relatively insensitive to both modes of action, we tested 4-PIOL in recombinant α1β2γ2 (widespread major subtype) and α6β2γ2 (cerebellar granule cell restricted) receptors where it had different effects on GABA-modulated [ 35S]TBPS binding and on electrophysiological responses. 4-PIOL may thus serve as a potential lead for receptor subtype selective compounds.

Multiple calmodulin genes exhibit systematically differential responses to chronic ethanol treatment and withdrawal in several regions of the rat brain

Ethanol induces profound alterations in the neuronal signaling systems, including the calcium (Ca 2+) signaling. Prolonged exposure to ethanol evokes adaptive changes in the affected systems as they strive to restore the normal neuronal function. We investigated the involvement of calmodulin (CaM) genes, coding for the major mediator protein of intracellular Ca 2+ signals, in these adaptive processes at the mRNA level. The changes induced in the regional abundances of the CaM I, II, and III mRNA classes by chronic ethanol treatment and withdrawal were examined by means of quantitative in situ hybridization, employing gene-specific [ 35S]cRNA probes on rat brain cryostat sections. Regional analysis of the resulting changes in mRNA levels highlighted brain areas that belong in neuronal systems known to be especially sensitive to the action of ethanol. The results revealed systematically differential regulation for the three mRNA classes: the CaM I and CaM III mRNA levels displayed increases, and CaM II levels decreases in the affected brain regions, in both chronic ethanol- and withdrawal-treated animals. As regards the numbers of brain regions undergoing significant alterations in mRNA content, the CaM I mRNA levels exhibited changes in most brain areas, the CaM II levels did so in a lower number of brain regions, and the CaM III levels changed in only a few brain areas. These results suggest a differential regulation for the CaM genes in the rat brain and may help towards elucidation of the functional significance of the multiple CaM genes in the mammalian genome.

Monoclonal antibodies to keratan sulfate immunolocalize ramified microglia in paraffin and cryostat sections of rat brain.

We used six monoclonal antibodies (MAb) recognizing epitopes within keratan sulfate (KS) chains for an immunocytochemical study of adult rat brain. One of the MAb selectively stained microglia and their ramified processes. KS-positive cells were found throughout the CNS in both paraffin-embedded and cryostat sections; the greatest number were present in hippocampus and brainstem. In the cortex the positive processes of some cells surrounded neuronal somata. In the white matter the processes were both parallel and perpendicular to the axon bundles. Double staining showed that KS-positive cells did not express astrocytic or oligodendroglial markers. By immunoelectron microscopy, the positivity was localized around the perikarya and cell processes of small cells with peripheral chromatin clumps and dark cytoplasm, which often contained secondary lysosomes. The KS-positive cells did not contribute to myelin sheaths and were not surrounded by a basal membrane. In addition to the cellular staining, three other MAb stained the white matter diffusely. Anti-KS MAb are therefore proposed as immunohistochemical markers for ramified microglia in both paraffin and cryostat sections of adult rat brain.

In vitro binding of [3H]bilirubin to neurons in rat brain sections.

Slide-mounted 12-microns cryostat sections of rat brain were incubated with [3H]bilirubin, washed, dried, and apposed to emulsion-coated coverslips for autoradiography. The binding appeared to be nonsaturable. Within the gray matter, binding was concentrated over neuronal cell bodies (particularly evident over hippocampal pyramidal and granular cells, and over cerebellar Purkinje cells), suggesting that the preferential neuronal toxicity of bilirubin may be related to neuronal binding. Regional differences in bilirubin binding were not demonstrated.

A method for the direct measurement of mRNA in discrete regions of mammalian brain

A rapid and nearly quantitative method for the direct analysis of steady-state mRNA levels in microgram quantities of frozen mammalian brain is described. Briefly, tissue punches 0.5–1.0 mm in diameter were sampled from 250-μm-thick cryostat sections of rat brain (approximately 50–200 μg tissue). The samples were homogenized in 50 μl of a denaturing gel loading buffer and applied directly to a 2.2 m formaldehydeagarose gel for electrophoresis and subsequent RNA blot analysis. The method is extremely rapid, results in excellent recovery of intact RNA, and allows the direct assay of mRNA levels in discrete subregions of the mammalian brain.

Selective growth of hippocampal neurites on cryostat sections of rat brain

Dissociated rat hippocampal neurons were cultured on horizontal cryostat sections from neonatal and adult rat brain and their growth patterns visualized by brightfield and interference contrast microscopy. Cells adhered to the sections as individuals or in small clusters and grew extensive neurites. Neurites grew over all areas of neonatal sections without apparent selectivity. For adult sections, however, neurites grew almost exclusively on areas of grey matter: there was no neurite growth on areas of white matter, irrespective of the location of that white matter within the brain. The transition from the neonatal to the adult pattern of growth occurred for sections from animals aged between 14 and 21 days.

Corticosterone decreases 3H-glutamate binding in rat hippocampal formation.

The effect of in vivo corticosterone manipulation on binding sites for the excitatory neurotransmitter glutamate was determined using radioligand binding to rat brain cryostat sections and crude synaptic membranes. 3H-glutamate binding to regions of the dorsal hippocampal formation was significantly decreased in adrenalectomized animals following 5-10 days of corticosterone treatment. Corticosterone did not alter 3H-glutamate binding in several other brain regions. The loss in 3H-glutamate binding appeared to be due to a decrease in the maximal number of binding sites, with little change in binding affinity. Both chloride-dependent and chloride-independent 3H-glutamate binding sites were decreased by corticosterone treatment. results indicate that corticosterone can selectively alter binding sites for excitatory amino acids in hippocampal tissue.

In vitro autoradiographic localization of guanine nucleotide binding sites in sections of rat brain labeled with [ 3H]guanylyl-5′-imidodiphosphate

In vitro labeling of guanine nucleotide binding proteins has been accomplished for the microscopic localization and quantitation of these sites in individual brain regions. Cryostat sections of rat brain were labeled with [ 3H]guanylyl-5′-imidodiphosphate ([ 3H]Gpp(NH)p) using parameters which resulted in a high specific to non-specific (signal to noise) binding ratio. Gpp(NH)p binding sites were identified to some extent in all brain regions examined. The highest densities existed in the olfactory bulb, cerebral cortex, caudate-putamen, interpeduncular nucleus, substantia nigra and cerebellum. Thus, GTP binding proteins can be labeled and localized using a novel application of an autoradiographic technique developed for microscopic detection of receptors.

Timm's method modified for human tissue and compatible with adjacent section histofluorescence in the rat

A simple dip method of Timm staining was developed. Because perfusion with sulphides and/or aldehydes is not required for adequate staining, the method can be employed with more control on human tissue taken at autopsy. Timm-stained human hippocampus reveals silver deposition in the mossy fiber region with a distribution similar to that in rats, but with some variation in the extent of penetration into stratum pyramidale. Adjacent cryostat sections of rat brain can be reacted for glyoxylic acid induced fluorescence by using the SPG dip method.

Autoradiographic identification and topographical analyses of high affinity serotonin receptor subtypes as a target for the novel putative anxiolytic TVX Q 7821

TVX Q 7821 is a putative non-benzodiazepine anxiolytic which has a high affinity to 5-HT 1 receptors. In this study some of the binding characteristics of the radiolabeled drug using rat brain cryostat sections and the autoradiographic localization of binding sites using the LKB-Ultrofilm technique have been investigated. Parallel experiments have been performed using [ 3H]serotonin ([ 3H]5-HT). Both [ 3H]TVX Q 7821 and [ 3H]5-HT bound specifically and in a saturable manner to tissue sections, the K d values being 6.8 and 3.7 nmol/l, respectively. Quantitative autoradiography using computer-assisted image analysis revealed a mean inhibition by TVX Q 7821 of [ 3H]5-HT binding of 56% in many brain areas. The inhibition ranged from 80% in the hippocampus and entorhinal area to practically none in the substantia nigra and the dorsal subiulum. Color coded autoradiograms obtained either with [ 3H]5-HT or [ 3H]TVX Q 7821 showed a nearly identical pattern of labeling with high receptor densities in the hippocampus, the entorhinal area, the septum, the interpeduncular nucleus and the dorsal raphe. However, in some brain areas striking differences in the intensity of labeling were found. [ 3H]5-HT but not [ 3H]TVX Q 7821 bound strongly in the substantia nigra, the dorsal subiculum and the globus pallidus. It is proposed that TVX Q 7821 binds to a subtype of 5-HT 1 receptor (the so-called 5-HT 1A sites as recently proposed). Thus, the putative anxiolytic TVX Q 7821 may provide a means for the study of the functional role of 5-HT 1 receptors.

Autoradiographic localization of kappa opiate receptors in CNS taste and feeding areas

Recent evidence suggests that kappa opiate receptors may play a key role in the regulation of appetite [21]. Such evidence implies that kappa receptors might be localized within specific brain areas known to regulate ingestive behaviors. On the basis of this implication we employed an in vitro film autoradiographic technique using 3H-ethylketocyclazozine as ligand to identify putative kappa receptors [7] within CNS “taste” nuclei [23] and surrounding areas. Coronal cryostat sections of rat brain were incubated with ligand in the presence of D-Ala 2, D-Leu 5-enkephalin (DADLE) and morphine, apposed to LKB Ultrofilm for 60 days, processed and kappa receptor densities evaluated with the aid of a hand held photometer and video image analyzer. Highest kappa receptor densities were found within various gustatory and feeding sites including the rostral pole of the nucleus of the solitary tract, parabrachial nuclei, ventral posterior and medial portions of the thalamus, medial hypothalamus, medial nuclei of the amygdala and bed nucleus of the stria terminalis. Various other midline and medial limbic areas also showed significant kappa densities.

Pathways of hydrogen utilization from NADPH generated by glucose-6-phosphate dehydrogenase in circumventricular organs and the hypothalamo-neurohypophysial system: A cytochemical study

Cytochemistry was used to examine the distribution of two pathways of utilization of hydrogen (Type I and Type II H) generated by glucose-6-phosphate dehydrogenase (G6PD) in circumventricular organs (CVOs) and the hypothalamo-neurohypophysial system in cryostat sections of rat brain. Type I H is defined as that portion of the total reducing equivalents (Total H) that is passed, in the intact cells, along the cytochrome chain (NADPH-diaphorase system). In the liver, energy from Type I H is used for cytochrome P-450-dependent oxidation of steroids, as well as xenobiotics. We proposed that mixed function oxidation, and therefore Type I H, would be preferentially localized in brain regions lacking a blood-brain barrier, such as CVOs and magnocellular cells with terminals in such brain regions. Type I H was identified in tissue sections using neotetrazolium. This reagent, when reduced, precipitates as formazan granules that can be quantified. The large difference in redox potential between NADPH and neotetrazolium ensures that only hydrogen (Type I H) passed in the intact cell along the cytochrome chain, can reduce the tetrazole. Total NADPH generation (Total H) from glucose-6-phosphate, was identified using medium containing phenazine methosulphate, a hydrogen acceptor that transfers all reducing equivalents from NADPH to the tetrazole. Type II H, the difference between Total and Type I H, is presumed to be used for NADPH-dependent biosynthetic functions such as lipid synthesis, or reduction of glutathione. In CVOs formazan granules indicative of Type I H were selectively concentrated and localized within cells throughout the SFO, organum vasculosum of the lamina terminals, pineal gland and in the apical cytoplasm of columnar ependymocytes in the subcommissural organ. Formazan granules attributable to Type I H were also prominent throughout the hypothalamo-neurohypophysial system. Reaction product was present in the cytoplasm of some magnocellular neurons in both the supraoptic and paraventricular nuclei, in the median eminence, including the zona interna, and in and between cells in the neurohypophysis. The distribution of NADPH-diaphorase in sections incubated with NADPH instead of glucose-6-phosphate was similar to that of type I H. These findings are consistent with the hypothesis that mixed function oxidation involving NADPH and the cytochrome chain occur in these brain regions. Formazan granules attributable to Total H were both more ubiquitously distributed than Type I H throughout the brain, and heavily concentrated in CVOs, particularly in the SFO, and in or between cells in the supraoptic and paraventricular nuclei. Reaction product attributable to Total and Type I H in magnocellular neurons was quantified by microspectrophotometry. Absorbance (mean integrated extinction × 100) for Total H was approximately 5–6× greater than for Type I H in magnocellular neurons in both the supraoptic and paraventricular nuclei. Relatively more Type II than Type I H was generated in magnocellular neurons and ependymal cells from glucose-6-phosphate (ratio Type I/Type II= 0.13 ± 0.01to0.29 ± 0.08). No reaction product was formed in the absence of substrate (glucose-6-phosphate) or NADPH.

Immunohistochemical differences between neurofilaments in perikarya, dendrites and axons: Immunofluorescence study with antisera raised to neurofilament polypeptides (200K, 150K, 70K) isolated by anion exchange chromatography

Neurofilament (NF) proteins (70K, 150K and 200K D) were isolated from 2 M urea extracts of bovine spinal cord by anion exchange chromatography. Antisera to the individual NF polypeptides were produced in rabbits and affinity-purified on Sepharose columns prepared with their own antigen. The NF antisera were completely absorbed by their own antigen at protein concentrations that did not decrease the staining when the absorption was conducted with the heterologous NF antigens. Partial absorption (decrease in immunofluorescence titer) occurred at higher concentrations of the heterologous antigens. Cross-reactivity between the polypeptides of the NF triplet could not be detected by double immunodiffusion. The antisera formed immunoprecipitin lines only when reacted with their own antigen. Conversely, cross-reactivity was demonstrated by the immune blotting procedure. Anti-70K stained all three NF polypeptides. Anti-200K and anti-150K stained both 200K and 150K but not 70K, the main reaction being with their own antigen. The antisera were rendered monospecific by adsorption of the common antigenic determinants on Sepharose columns prepared with the heterologous NF antigens. The localization of the NF proteins was studied by immunofluorescence on cryostat sections of rat brain, cerebellum, spinal cord and posterior root ganglia. All NF antisera (anti-70K, anti-150K and anti-200K) stained axons including Purkinje cell baskets with identical pattern. Spinal cord motor neurons, posterior root ganglia neurons and pyramidal neurons in the cerebral cortex stained with anti-70K and anti-200K. No staining of neuronal perikarya and dendrites was observed with anti-150K. Aluminium-induced neurofibrillary tangles in rabbit spinal cord stained with anti-70K and anti-200K. The tangles were not decorated by anti-150K. It is concluded that a marked difference exists in the concentration of 150K depending on the location, i.e., cell body or axon; or, alternatively, that 150K undergoes modification of antigenic sites within the axon so that it may not be recognized immunologically as a component of the neurofilament within perikarya and dendrites.

Adenosine receptor density and the depression of evoked neuronal activity in the rat hippocampus in vitro

The relationship between the heterogenous distribution of A-1 adenosine receptors and the capacity of adenosine to depress neuronal activity was examined in the rat hippocampus. Utilizing autoradiographic techniques, the distribution of A-1 adenosine receptors was assessed by the binding of [ 3H]cyclohexyladenosine ([ 3H]CHA) to cryostat sections of rat brain. The apical dendritic region of CA1 showed a differential distribution of adenosine receptors between the stratum radiatum and stratum lacunosum/moleculare. The physiological relevance of this binding differences was studied in the hippocampal slice by examining the capacity of adenosine to depress evoked potentials in these two strata. It was observed that the receptor differences correlated with differential sensitivities to adenosine modulation of the evoked potentials. These data suggest that receptor density, as shown by binding techniques, may provide not only a qualitative but also a quantitative map of the sites of adenosine action.

Immunohistological localization of desmin, the muscle-type 100 A filament protein, in rat astrocytes and Müller glia.

The distribution of glial fibrillary acidic (GFA) protein and desmin was compared in cryostat sections of rat brain, spinal cord, and eye by immunofluorescence and peroxidase-antiperoxidase (PAP) staining. Desmin antisera were raised to antigen purified from chicken gizzard. In rat brain and spinal cord, GFA protein and desmin were selectively localized in astrocytes. Neurons and axons were not stained. The only difference between GFA and desmin antisera was the staining of smooth muscle in small arteries with anti-desmin. It was only in retinal glia that a difference in the localization of the two proteins was apparent. As previously reported, only the glia limitans on the inner surface of the retina was demonstrated with GFA antisera in the normal eye. With anti-desmin Müller fibers spanning the whole thickness of the retina were stained. It is concluded that GFA and desmin form two distinct systems of 100 A filaments in astroglia, as previously reported for GFA and vimentin.