Bidirectionality of mixed lymphocyte stimulation (Mls) response. Effects of Mlsb stimulator cells on Mlsa helper cells

The activation of BALB/c lymphocytes in the mixed lymphocyte reaction to Mls-disparate APC has been shown to encompass up to 20% of the mature resting helper T lymphocyte population. In addition to these overtly Mls-responsive cells, our studies have revealed a second population that respond to the Mls difference of DBA/2 spleen cells in conjunction with the mitogen Con A. This part of the Mls response is therefore latent. As mitogen and Mls-stimulating effect act in synergy, it is likely that both stimuli act on the same cell, and hence the Mls effect can be regarded as a regulatory interaction between APC and Th cell. By use of congenic BALB.Mlsa mice, the regulatory effect has been mapped to the Mls locus. The regulatory influence has also been demonstrated in DBA/2 Th cells (Mlsa) stimulated simultaneously with mitogen and Mls-disparate (Mlsb) APC, consistently causing inhibition of mitogen-induced proliferation in this reverse Mls direction. This antagonistic effect has also been linked to the Mls locus. We conclude that the Mls reaction governed by the a and b alleles is bidirectional, producing synergy with class II-dependent activation signals in the direction of Mlsa----Mlsb, and antagonism in the direction Mlsb----Mlsa. Both the classical Mls and the reverse Mls effects have been demonstrated at the clonal level. These results are in accord with the previously proposed hypothesis that the Mls molecule serves as a down-regulatory stimulus in the activation of Th cells. Mls responses of Mlsb T cells are explained as the consequence of a diminished down-regulation by Mlsa APC. Conversely, the reverse Mls response described here can be considered a consequence of inordinately high down-regulation of the Mlsa T cell responses by Mlsb APC.     

A new mouse cell-surface antigen (Ly-m20) controlled by a gene linked to Mls locus and defined by monoclonal antibodies

Five monoclonal antibodies were established by the fusion of mouse myeloma cells (NS.1) with spleen cells from A and (A x C3H/An)F₁ mice hyperimmunized with 70Z/3 tumor cells. These antibodies recognized a new antigenic specificity provisionally called Ly-m20.2. In direct cytotoxicity assays, 60 percent of cells in spleen, 40 percent in lymph node, 50 percent in bone marrow and less than 5 percent in thymus were found to react with three of the five antibodies, whereas the two others yielded somewhat lower cytotoxicity indices. The Ly-m20.2 antigen was also expressed on cells derived from liver and kidney but not on cells derived from brain. As judged from cytotoxicity assays with separated T and B cells, Ly-m20.2 antigen is carried preferentially on B lymphocytes. Direct plaque-forming cells (PFC) were completely eliminated by Ly-m20.2-specific antibody and complement. Linkage tests by analysis in 20 (CBA/J x C3H/An) x C3H/An backcross mice and by segregation analysis of BXH and SWXL recombinant inbred strains indicate close association of the loci controlling Ly-m20.2 and Mls antigens on chromosome 1.Abbreviations used in this paper MLR mixed lymphocyte reaction - MHC major histocompatibility complex - Mls minor lymphocyte-stimulating antigen - Ia I-region associated - PFC plaque-forming cell - SRBC sheep red blood cell - Con A concanavalin A - LPS lipopolysaccharide - GVH graft-versus-host - HVG host-versus-graft - CML cell-mediated lympholysis - LD lymph ocyte-defined - SD serologically defined The prefix m (monoclonal) is used following a suggestion by Klein and co-workers (1979).     

Isolation of cockroach Phe-Gly-Leu-amide allatostatins from the termite Reticulitermes flavipes and their effect on juvenile hormone synthesis

Immunoreactivity to cockroach Diploptera punctata allatostatin-7 (Dippu AST-7) has been demonstrated previously in axons innervating the corpora allata of the termite Reticulitermes flavipes. This peptide and Dippu AST-11 inhibited juvenile hormone (JH) synthesis by corpora allata (CA) of brachypterous neotenic reproductives (secondary reproductives) of termites. The present study shows that R. flavipes CA are also inhibited by Dippu AST-2, AST-5, AST-8, and AST-9 at approximately the same rank order of potency as demonstrated in D. punctata. Another allatostatin from Periplaneta americana (Peram AST-12) also inhibits JH synthesis by R. flavipes CA. Sensitivity to the allatostatins is higher in glands with low rates of JH synthesis than in those with relatively high JH synthetic rates as has been demonstrated in CA from male and female secondary reproductives as well as in those from non-egg-laying and egg-laying females. The identical inhibitory effects of R. flavipes brain extract on CA from both D. punctata and R. flavipes and the isolation and identification of five cockroach allatostatins (Dippu AST-1, AST-2, AST-5, AST-8, and Peram AST-12) from termite brain extract reflect the close relationship between cockroaches and termites.     

Characterization of a xenogeneic antiserum raised against the fetal germ cells of the mouse: cross-reactivity with embryonal carcinoma cells.

A xenogeneic antiserum (PG-1) raised against the germ cells of the 13.5d p.c. mouse conceptus reacts with the fetal germ cells of both sexes and adult male germ cells, but not with any nongerminal tissue that has been tested. PG-1 can also react with the EC cells of four different teratocarcinoma cell lines. There are, however, marked differences in the absorptive capacity of the different EC cell lines. The antigen(s) recognized by the antiserum are confined to only a proportion of cells in some EC lines. This antiserum has a number of potential applications in studies of the origin and development of the mouse germ cell lineage.     

The Mlsd-defined primary mixed lymphocyte reaction: a composite response to Mlsa and Mlsc determinants

Considerable disagreement exists among immunologists regarding the polymorphic nature of the murine Mls system. An estimate of the capacity of a given putative Mls allelic gene product expressed on a stimulator population to elicit proliferation of H-2-compatible Mls-disparate unprimed T cells may vary widely among different groups of investigators. This laboratory has shown previously that preactivation of B lymphocytes in a splenocyte stimulator population by exposure to goat anti-mouse IgD (GaMD) before irradiation dramatically enhanced the in vitro presentation not only of the strongly stimulatory (and highly cross-reactive) Mlsa and Mlsd, but also the more poorly stimulatory Mlsc specificity. Therefore, by the use of GaMD-treated splenocytes that optimally present the various Mls non-H-2 stimulatory epitopes, we attempted in this study to obtain a clearer understanding of Mls polymorphism by re-examining the conflicting claims associated with the mixed lymphocyte reaction (MLR) stimulatory capacity of different Mls specificities. Among H-2k responder cells of the Mls null, Mlsa, Mlsb, or Mlsd genotypes, only T cells from Mlsd-bearing CBA/J mice did not respond to Mlsc determinants present on GaMD-treated C3H/HeJ stimulator cells. Crossing CBA/J with an Mlsc-responsive mouse strain yielded an F1 animal in which nonresponsiveness to Mlsc was dominant. Although Mlsa (AKR/J) and Mlsc (C3H/HeJ) parental T cells both proliferated vigorously to Mlsd (CBA/J) stimulator cells, the Mlsa/c (AKR X C3H)F1 T cells responded poorly to GaMD-treated Mlsd stimulator cells. In addition, Mlsd (CBA/J) T cells were nonresponsive to Mlsa (AKR/J), Mlsc (C3H/HeJ), and Mlsa/c (AKR X C3H)F1 GaMD-treated stimulator cells. Because Mlsa (AKR/J) and Mlsc (C3H/HeJ) specificities are mutually stimulatory, at least limited polymorphism must exist in the Mls system. However, because Mlsa/c (AKR X C3H) and Mlsd (CBA/J) specificities are mutually nonstimulatory, T cell proliferation in an Mlsd-defined primary MLR is most likely due to a composite response to Mlsa and Mlsc epitopes present on CBA/J stimulator cells.     

Characterization of a new minor lymphocyte stimulatory system. I. Cluster of self antigens recognized by "I-E-reactive" V beta s, V beta 5, V beta 11, and V beta 12 T cell receptors for antigen.

In the mouse, two sets of V beta gene products have been shown to be associated with T cell recognition of endogenous self Ag. One of these is the set of V beta associated with T cell reactivities to stimulatory Mls gene products, Mlsa (V beta 6, V beta 8.1, V beta 9) or Mlsc (V beta 3); another is the set of V beta, such as V beta 5, V beta 11, V beta 12, or V beta 17a, which were originally found to be related to I-E recognition. Although the Mls system has been well characterized, little is known about the nature of the ligands for the second set of V beta. In this work, we describe the evidence that the natural ligand or ligands of V beta 5, V beta 11, and V beta 12 may be novel Mls determinants that are recognized by naive T cells at a high precursor frequency and function as the ligand for clonal deletion of self-reactive T cells by negative selection. However, surprisingly, unlike the conventional Mls system, in which all V beta associated with Mlsa recognition or Mlsc recognition are uniformly deleted in those animals expressing the relevant Mls type, expression of these three V beta segregates independently among strains. Based on these observations, the nature of T cell recognition for this new Mls gene product(s) is discussed.     

Specific neonatal induction of functional tolerance to allogeneic Mls determinants occurs intrathymically and the tolerant state is Mls haplotype-specific

The studies described show that functional Mls specific tolerance, which we previously reported in peripheral spleen cells of mice injected within 24 h of birth with Mls incompatible spleen cells, is observable in the thymus on day 6. At this time a significant positive response is not detectable in spleen cells of normal mice. In the limiting dilution assay, we are able to detect a more profound depletion than others have found with anti-TCR antibodies. The tolerance in the thymus is not due to active suppression or simple dilution of responders by nonresponsive cells of the neonatal inoculum. By tolerizing BALB/c (Mls(b,c] mice with spleen cells from Mls(a) congenic mice, we show that Mls(a) incompatibility alone is sufficient for tolerance induction. Data from these experiments also show that the T cells seen responding at high frequency to stimulators from mice expressing Mls(a) determinants, as well as many other non-H-2 encoded incompatibilities, are indeed responding to Mls(a) determinants. In addition, experiments involving neonatal injection of Mls(b) mice with Mls(a) and Mls(c) spleen cells show no cross-reactivity of tolerance between Mls(a) and Mls(c) haplotypes. Our findings also show coexpression of determinants common to both Mls(a) and Mls(c) haplotypes by the Mls(d) haplotype. In all, the described experiments elucidate a pattern of Mls determinant specific hyporesponsiveness, in mice neonatally injected with appropriate allogeneic spleen cells, which bears all the hallmarks of functional, alloantigen specific, clonal deletion type tolerance.     

Soluble suppressor of T cell proliferation in primary in vitro response to murine minor histocompatibility antigens

Normal mouse lymphocytes are not capable of mounting a primary cytotoxic T cell response to Mls encoded, non H-2, allodeterminants, although a strong lymphoproliferative response is observed in primary MLR between Mls incompatible cells. In this study it is reported that in the supernatant of primary cultures between AKR macrophages and CBA/H lymphocytes (H-2 identical, incompatible for Mls and other minor antigens) a suppressor of T cell proliferation in MLR is detected. By contrast, a suppressor is not detected in supernatants from primary cultures between BALB/C macrophages and CBA/H lymphocytes (H-2 incompatible, Mls identical), B10.BR macrophages and CBA/H macrophages and CBA/H lymphocytes (syngeneic) suggesting that the production of the suppressor factor occurs only when an Mls incompatibility exists. The suppressive activity of the Mls incompatible culture supernatant upon MLR between incompatible macrophages and lymphocytes is neither antigen specific nor Mls or H-2 restricted, nor is it due to an irreversible toxic effect on T lymphocytes or macrophages. The inhibition of T cell proliferation could be explained by inhibition of IL 2 production, by blocking its union to T cells or by a combination of both effects. Our findings could help explain previous observations that lymphocytes from mice preimmunized with Mls incompatible cells have a depressed proliferative response as well as depressed cytotoxicity against alloantigens.     

T cell recognition of Mls. T cell clones demonstrate polymorphism between Mlsa, Mlsc, and Mlsd

The determinants encoded by the minor lymphocyte stimulating locus (Mls) are defined as determinants that induce strong T cell proliferative responses in primary mixed lymphocyte reactions. Although the Mls locus was originally described as having four alleles, a, b, c, and d, a number of recent observations have led several investigators to challenge the idea that Mls is truly a polymorphic system. To better define this system of determinants recognized at high frequency by T cells, the present studies were undertaken to evaluate the polymorphism of Mls products. In the present study, the in vitro proliferative responses of Mlsa- and Mlsc-specific T cell clones were employed to analyze Mls products. The identification of determinants recognized by Mlsa- and Mlsc-reactive clones was established by the pattern of responses to stimulators derived from congenic strains, recombinant inbred strains, and backcross mice. T cell clones and unprimed T cells gave concordant responses that confirmed the Mlsa or Mlsc specificity of the cloned populations. With the use of these two sets of Mls-specific T cell clones, the existence or absence of polymorphism of Mls-encoded gene products was examined. It was found that Mlsa-specific cloned T cells responded to Mlsa but not Mlsc stimulators, whereas Mlsc-specific clones responded to Mlsc but not Mlsa. This reciprocal pattern of specificity indicates that the Mls system as currently defined is therefore truly polymorphic. In addition, it was observed that both Mlsa- and Mlsc-specific clones were stimulated by Mlsd stimulators. In particular, the possibility that Mlsa and Mlsc are not alleles but products of different loci and that Mlsd strains are those that express both Mlsa and Mlsc is considered.     

Gene cloning of Porphyromonas gingivalis specific antigens recognized by serum of adult periodontitis patient.

1. Porphyromonas gingivalis is believed an important pathogen of adult periodontitis. A gene library of P. gingivalis 381 was constructed in lambda phage vector L47.1. The library was probed with serum obtained from patients of severe adult periodontitis. Two clones, lambda MDBG101 and lambda MDBG103 which were expressed, 200 and 160 kDa respectively, were selected and further studied. 2. The expressed antigens in these two clones were also reacted with rabbit antiserum against whole cells, capsular fraction and cell surface fraction of P. gingivalis. 3. Genes coding protein antigens in lambda MDBG101 and lambda MDBG103 were subcloned into high-copy-number plasmid vector pACYC184 and subclones obtained were designated as MD101 and MD103. Recombinant plasmids, pMD101 and pMD103, differed in their restriction endonuclease digestion. 4. Immunodiffusion analysis showed that cloned proteins from MD101 and MD103 reacted with antiserum against P. gingivalis but did not react with antiserum against Prevotella intermedia, Prevotella loescheii and Prevotella asaccharolyticus. 5. These data suggest that P. gingivalis species-specific antigens has been successfully cloned and expressed in Escherichia coli. Since these cloned specific antigens were recognized by adult periodontitis patient sera, the recombinant antigen will be useful material for the development of serodiagnosis of P. gingivalis infection in adults periodontitis.     

Activation requirements of cloned inducer T cells. III. Need for two stimulator cells in the response of a cloned line to Mls determinants

To gain insight into the nature of Mls determinants, we examined the stimulator cells responsible for the activation of inducer T cell clones by Mls determinants. Two types of clones responding to Mls determinants were identified. One type responded to purified B cells, but not to splenic adherent cells (SAC), from mice bearing Mls stimulatory determinants. The other type of Mls-reactive T cell clone, including the representative clone Ly1-N5, demonstrated a vigorous response to unfractionated spleen cells, but showed little or no response to B cells alone or to SAC alone from mice bearing the Mlsa or Mlsd stimulatory determinant. The response of these clones to Mls determinants required stimulation by two cell types. The failure of clone Ly1-N5 to respond to Mlsa-bearing B cells was reversed by the addition of SAC taken from mice bearing the Mlsa allele. In addition, SAC from mice bearing the nonstimulatory Mlsb allele could synergize with B cells from Mlsa-bearing animals. B cells were required to provide the Mlsa determinant, because the combination of Mlsa-bearing SAC and Mlsb-bearing B cells did not activate the clone. The response of clone Ly1-N5 to Mls is restricted by Ia determinants (shared by H-2b, H-2d, and H-2k haplotypes but not by the H-2q haplotype). The permissive H-2 alleles can be present either on the stimulator B cell or on the SAC. The optimal response of the clone was obtained by using B cells bearing Mlsa and the permissive Ia epitopes. However, a significant response of the clone to B cells bearing Mlsa but an inappropriate Ia (Iaq) was also seen in the presence of SAC bearing the nonstimulatory Mlsb allele but the permissive Ia epitopes.     

Expression of a type 2 pneumocyte-specific antigen by a cell strain from normal adult mouse lung

The mouse lung epithelial cell strain NAL 1A has been confirmed as type 2 pneumocyte-related by immunostaining with a type 2 pneumocyte-specific antiserum. Cytoplasmic immunoreactivity with the antiserum paralleled the appearance of phospholipid-containing osmiophilic cytoplasmic inclusions, which were much more abundant in confluent cell cultures at low passage number than in exponentially growing cultures. However, phospholipid analysis indicated that NAL 1A cells were impoverished in phosphatidylglycerol as compared to lung type 2 pneumocytes. Confluent cultures of the neoplastic cell lines NAL 1AM and NUL 1 did not reveal any reactivity with the specific antiserum.     

Minor lymphocyte stimulatory antigen-bearing stimulator cells require lipopolysaccharide activation to induce programmed cell death in T cell hybridomas.

T cell hybridomas respond to conventional peptide Ag associated with self major histocompatibility restriction elements, as well as to alloantigens, activating lectins, and stimulatory forms of mAb by producing lymphokines and undergoing programmed cell death (PCD). We show here that the level of PCD and IL-2 production correlate well in responses to CD3 or allostimulation. The response to minor lymphocyte-stimulatory (Mls) Ag, members of the family of endogenous superantigens, however, are marked by divergence in the levels of the PCD and lymphokine responses. Specifically, PCD in response to Mls activation is achieved poorly despite vigorous IL-2 production. B lymphoma cell stimulators induced PCD in alloreactive T cell hybridomas but not in Mls-reactive T cell hybridomas. This suggests that the absence of PCD in the Mls response is a function of superantigen recognition rather than the stimulator cell type. LPS-preactivated Mls+ stimulators, either splenic B or B lymphoma cells, are shown to trigger PCD in the T cell hybridomas. These results imply that T cell interaction with Mls presented by untreated stimulator cells is not sufficient for induction of PCD and thus is distinct from interactions with conventional Ag.     

Targeting of malate synthase 1 to the peroxisomes of Saccharomyces cerevisiae cells depends on growth on oleic acid medium.

The eukaryotic glyoxylate cycle has been previously hypothesized to occur in the peroxisomal compartment, which in the yeast Saccharomyces cerevisiae additionally represents the sole site for fatty acid beta-oxidation. The subcellular location of the key glyoxylate-cycle enzyme malate synthase 1 (Mls1p), an SKL-terminated protein, was examined in yeast cells grown on different carbon sources. Immunoelectron microscopy in combination with cell fractionation showed that Mls1p was abundant in the peroxisomes of cells grown on oleic acid, whereas in ethanol-grown cells Mls1p was primarily cytosolic. This was reinforced using a green fluorescent protein (GFP)-Mls1p reporter, which entered peroxisomes solely in cells grown under oleic acid-medium conditions. Although growth of cells devoid of Mls1p on ethanol or acetate could be fully restored using a cytosolic Mls1p devoid of SKL, this construct could only partially alleviate the requirement for native Mls1p in cells grown on oleic acid. The combined results indicated that Mls1p remained in the cytosol of cells grown on ethanol, and that targeting of Mls1p to the peroxisomes was advantageous to cells grown on oleic acid as a sole carbon source.     

A mouse plasma cell surface antigenic determinant (PLKB) recognized by xenoantiserum. Its relationship to the PC.1 antigenic determinant.

The specificities of the xenoantisera made against mouse myeloma cells have been compared to those recognized by alloantiserum by studying patterns of cytotoxicity on both normal and malignant plasma cells. Goat antiserum obtained by immunization with Balb/c mouse myeloma ADJ-PC-22A cells and purified by in vivo absorption could detect cell surface antigenic determinants present on plasma cells and on cells of liver, kidney, and brain (PLKB antigen), as we had previously reported for a similarly prepared rabbit antiserum. In spite of an apparent similarity between the tissue representation of the PLKB determinant and that of PC.1 antigenic determinants which were detected by DBA/2 anti-ADJ-PC-22A cell alloantiserum, the PLKB antigenic determinant is not identical with the PC.1 antigenic determinant, since the former is found on the tissues of PC.1-negative as well as PC.1-positive strains of mice. However, it was deduced that the PLKB antigenic determinant and the PC.1 antigenic determinant reside in close proximity on the cell surface or maybe even on the same molecule, since Fab fragments of antiserum against either PLKB or PC.1 blocked the cytotoxicity against both antigens. On the other hand, these Fab fragments did not inhibit the cytotoxicity of anti-H-2 antiserum, indicating that neither PLKB nor PC.1 antigenic determinants are in close proximity to H-2 antigens. Association of PLKB and PC.1 determinants was further supported by the finding that the loss of the PLKB determinant in a variant of myeloma MOPC-70A corresponds to the loss of PC.1 determinant on the same cells.     

A study of histocompatibility-2 antigens in wild mice from Santiago,Chile

LyM-1 is the provisional designation given to a system of murine cell-surface alloantigens which are controlled by genes closely linked to those of theMls system. Formal genetic analysis has failed to disclose separation of genes determiningMls and LyM-1 antigens, but studies of the distribution of these antigens among inbred strains shows that the LyM-1 polymorphism is not primarily responsible for the MLR activity associated with Mls differences, and suggests that LyM-1 and Mls substances are products of genes at closely linked, but probably separate loci. Absorption analysis shows that strains whose cells react with anti-LyM-1.2 can be divided into at least two classes on the basis of the efficiency with which their cells remove anti-LyM-1.2 antibodies. This provides evidence for the existence of two LyM-1 alleles in addition to the one(s) possessed by nonreactive mouse strains.     

Induction of tolerance against the Mls antigen in mice no need for Mls-bearing cells

Infusion of CBA mice with spleen cells from the H-2-compatible, but Mis antigen-incompatible, strain C3H leads to a specific reduction of the MLC reactivity of the host's lymphocytes. One explanation of the reduced reactivity could be that the specifically Mls antigen-responsive CBA T cells become exhausted by intense antigen stimulation or that the infused cells actively neutralize specifically responsive cells. In this investigation, we have shown that depletion of membrane Ig-positive cells from C3H × CBA spleen cell preparations strongly reduces their capacity to stimulate CBA lymphocytes in the MLC, indicating that the Mls antigen is expressed on B but not on T cells. However, such B-cell depleted cell preparations were fully capable of reducing the MLC response of CBA hosts. Cell preparations of spleen and lymph nodes exhibited high stimulatory and inhibitory effects. Thymic cells lacked both these characteristics, whereas bone marrow cells were weak stimulators but relatively strong inhibitors. The results support the proposal that the observed reduction of MLC reactivity is due to an active process of the injected cells. The cell type which is working as an inhibitor has not been clearly defined yet.     

Genetics of a non-H-2 antigen that stimulates "unrestricted" helper T cells and MLR

The inheritance of antigens expressed by C3H/Tif B cells that stimulate MHC-unrestricted helper T cells from C3H/HeJ was investigated. F1 hybrids between C3H/HeJ and C3H/Tif and 39 C3H/HeJ X F1 backcross mice were characterized as to the ability of their spleen cells to stimulate a proliferative C3H/HeJ T helper cell response and to respond to helper cell activity by the development of polyclonal plaque-forming cell responses. Backcross progeny wee also typed for the following markers segregating in this cross: 1) Responsiveness to the B cell mitogen lipopolysaccharide (LPS); 2) LyM-1 allotype; 3) antigen(s) stimulating a primary non-H-2 MLR between these strains, previously ascribed to Mls locus differences, 4) expression of target antigens for cytotoxic T cells raised in the same strain combination. The antigen(s) recognized by helper cells and those stimulating primary MLR are controlled by autosomal gene(s) and segregate as a single trait. These antigens, however, are not encoded in genes linked to either the Lps or the Mls loci, and are not recognized by cytotoxic T cells raised in the same strain combination.     

Sulforhodamine 101 as a specific marker of astroglia in the neocortex in vivo

Glial cells have been identified as key signaling components in the brain; however, methods to investigate their structure and function in vivo have been lacking. Here, we describe a new, highly selective approach for labeling astrocytes in intact rodent neocortex that allows in vivo imaging using two-photon microscopy. The red fluorescent dye sulforhodamine 101 (SR101) was specifically taken up by protoplasmic astrocytes after brief exposure to the brain surface. Specificity was confirmed by immunohistochemistry. In addition, SR101 labeled enhanced green fluorescent protein (EGFP)-expressing astrocytes but not microglial cells in transgenic mice. We used SR101 labeling to quantify morphological characteristics of astrocytes and to visualize their close association with the cortical microvasculature. Furthermore, by combining this method with calcium indicator loading of cell populations, we demonstrated distinct calcium dynamics in astroglial and neuronal networks. We expect SR101 staining to become a principal tool for investigating astroglia in vivo.