In the adult brain, the neurons appear to undergo mitosis only to a very limited .... The protein content of glial cells of the Deiters' nucleus is only one fifth of that .... radioactive protein formed in a cell-free system from rabbit brain can be pre- ... 13 Hyden, H., "The neuron," in The Cell: Biochemistry, Physiology, Morphology, ed.
A GLIAL PROTEIN SPECIFIC FOR THE NERVOUS SYSTEM* BY H. HYD*N AND B. MCEWENt INSTITUTE OF NEUROBIOLOGY, FACULTY OF MEDICINE, UNIVERSITY OF G6TEBORG, SWEDEN
Communicated by A. E. Mirsky, December 9, 1965
In many organs, part of the over-all protein synthesis serves the growth of cells by completing the new members of the cell population which arise after mitosis. In the adult brain, the neurons appear to undergo mitosis only to a very limited extent, while glial cells do divide under special circumstances.' Thus, cell renewal and growth cannot account for the main part of the extensive protein synthesis occurring in the brain. In an effort to explain the role of protein turnover in brain function, extensive efforts are under way in a number of laboratories to isolate proteins specific to the brain and to determine their cellular localization, metabolic half life, and function. Moore and co-workers2 3 have recently isolated and characterized an acidic protein (named S100 protein) specific to nervous tissue which makes up 0.5 per cent of the brain's soluble proteins. Antiserum to this protein, isolated from beef brain, showed cross reaction to protein from all vertebrate species examined.4 The suggestion has been made that the S100 protein is located in glial cells,5 while another proposal has been made that it is a neuronal protein.3 In this paper, evidence will be presented that the S100 protein is mainly localized in the glial system of the Deiters' nucleus of the rabbit brain stem. The S100 protein is also found in the cell nuclei of the large neurons, while it is distinctly absent from the nuclei of the glial cells. Material and Methods.-Microanalysis: The large neurons and surrounding glia from the lateral vestibular (Deiters') nucleus of the rabbit brain stem were used. After rapid removal of the brain stem, a section was placed at the rostral boundary of the tubercula acustica. The sections of both sides, 3-4 mm thick, were placed in cold 0.25 M sucrose solution. The neurons or the glia were separated and cleaned by freehand dissection as previously described.6 The reason for using these types of neurons is that many data on their biochemical characteristics have been collected during the last 10 years. Also, the relationship between the Deiters' neuron and its surrounding glia has been elucidated in measurements of enzymatic activity and RNA studies.7 -10 Each Deiters' nerve cell has a dry weight of 2 X 10-8 gm. The dry weight per .s3 for both the nerve cell body and the surrounding glia is 0.20 juug. Therefore, biochemical data obtained on the same volume of neurons or glia can be compared. Twenty to thirty nerve cells or glial samples are transferred by freehand manipulation to a drop of homogenizing solution on top of a Teflon homogenizer pestle, 2 mm in diameter. The cells are homogenized in 20 Al of distilled water containing 0.5% w/v Triton X-100. After homogenization, centrifugation took place in 0.5-mm capillary tubes at 12,000 rpm. Single diffusion agar precipitation using antiserum against the S100 protein was performed in glass capillaries, 300 IA or 500 A in diameter (Microcaps: Drummond Scientific Co., Broomall, Pennsylvania), according to a micromodification of Oudin's technique."1 Agar in a 0.85% warm (450C) solution was introduced by capillary force in the middle third of a 40-mm-long capillary. By a micropipette, the top part of the capillary was filled with the antigen in 0.25 M sucrose, and the bottom part with a mixture of antiserum in 0.85% agar. The capillaries were placed vertically in a humidity chamber. Fluorescent antibody technique: The multiple-layer method (sandwich technique) of Coons12 for the localization of antigens by fluorescence was applied to identify the S100 protein in cryostat sections through the lateral vestibular nucleus. The sections were first treated with acetone for 30 see and then exposed to the specific antiserum for 30 see or to the following central procedures: (1) exposure to normal rabbit serum for 30 see and (2) no treatment at all. After 354
VOL. 55, 1966
355
ZOOLOGY: HYDAN AND MCEWEN
thorough washing of the sections in a pH 7.3 phosphate buffer, goat gamma globulin against rabbit gamma globulins, conjugated with fluorescin isothiocyanate (Microbiological Associates, Bethesda, Md.), was applied to the section for 30 see and the excess removed by repeated washing in the buffer. Sections were viewed in a Zeiss fluorescence microscope.
Results.-Figure 1 gives a schematic representation of the result of the single diffusion precipitation technique. The antigen consisted of soluble proteins extracted from 50 Deiters' nerve cells, or the corresponding amount of glia, or brain stem homogenate. On the fourth day, immunoprecipitation was obtained with proteins from both brain homogenates and glial cells. No precipitation was seen, using neuronal proteins. Figure 2 shows a photograph of the precipitation zone obtained from the glial sample. brain
neurons
q/(a
foinoqeuatcB
U
atiqqen
aqr
FIG. 2.-Photograph of
FIG. 1.-Localization of the protein S100 to the glia by single diffusion precipitation in agar capillaries.
the precipitation band of the protein S100 in a glass capillary 0.5 mm in diame-
ter.
The protein content of glial cells of the Deiters' nucleus is only one fifth of that of an equal volume of the large Deiters' neurons."3 Therefore, it seemed likely from this immunoprecipitation experiment that the S100 protein is present in much higher concentrations in glia than in neurons. In order to see if any S100 protein could be detected in Deiters' neurons, a dilution experiment was performed to test the sensitivity of the immunoprecipitation technique. A soluble protein extract was prepared in 0.25 M1 sucrose from rabbit Deiters' nucleus. After determining the protein content of this extract by the method of Lowry,'4 varying amounts of protein, from 0.1 to 10 Aug, were placed in M\Jicrocaps under the same conditions used in the experiments with neurons and glia. Immunoprecipitates were observed with amounts of protein above 0.4 14g (Table 1). Following this experiment, another immunoprecipitation experiment was performed with 150 Deiters' neurons, corresponding to 2.0 /ig of protein. Again no immunoprecipitate was observed, leading to the conclusion that the concentration of S100 protein in the neurons is at least less than one fifth that of the proteins of the whole Deiters' nucleus. It should be noted that this technique is probably not sensitive enough to detect the S100 protein in neuronal nuclei, since the nucleus comprises only 3.5 per cent on the average of the mass of the nerve cell.13 Thus, the absence of precipitate in these experiments with neurons is not inconsistent with the finding of the S100 protein in neuronal nuclei by the fluorescent antibody technique to be described below.
PROC. N. A. S.
ZOOLOGY: HYDtJN AND MICEWEN
356
TABLE 1 DILUTION EXPERIMENT WITH ANTIGEN* Precipitation
Antigen in 10-6 gm
19.80 12.20 2.24 1.22 1.110 0.9 0.8 0.7 0.6 0.5 0.4 0.4- 0.04
+ +
+ + +
+ + +
Comments
2 sharp bands
1 sharp band + 2 faint bands 2 bands Iband 1 band 1 band 1 very thin band 1 very thin band
Negative
* Brain stem homogenate, rabbit, in sucrose at 21'C and constant humidity in amounts from 19.8 to 0.04 X 10-6 gM. Antiserum against protein S100 diluted 1:8. Reaction observed on the 3rd day.
A pertinent question is where in the glial cells and in which types of glial cells the S1i0 protein is found. Using the fluorescent antibody sandwich technique, it became apparent that the specific fluorescence occurred around the nuclei of oligodendrocytes and scattered as fluoresceint spots within the memlbranous system
a
c
b
d
FIG. 3.-(a, b) Specific fluorescence in cryostat sections throulgh the Deiters' nucleus treated with antisertum against the 8100 protein. Note the fluorescence in t~he glial cells aroulnd the nonfluorescent nusclei and the scattered fluorescent particles scattered throughout the glia. The ntuclei of the big Deiters' nerve cells are brightly fluorescent. (c) Control section treated with normal rabbit serum. (d) No rabbit serutm. Magnification 150 X.
VOL. 55, 1966
ZOOLOGY: HYDJlN AND MCEWEN
357
of the glia. Figures 3a and b demonstrate the fluorescence in cryostat sections through the Deiters' nucleus treated with the antiserum against the S100 protein. Figures 3c and d show control sections treated either with normal rabbit serum (Fig. 3c) or with no rabbit serum at all (Fig. 3d). The markedly reduced fluorescence in the control sections demonstrates that it is most likely the S100 protein which is being stained by the S100 antiserum. Neuronal cytoplasm shows some fluorescence in both experimental and control sections; this staining cannot be considered specific to the S100 protein. A striking observation was that the nuclei of the large neurons showed a bright fluorescence as an indication of the presence of the S100 protein (Figs. 3a, b). The nucleolus, however, did not show this phenomenon (although this does not clearly appear from the reproduced photographs). By contrast, the nuclei in the control sections showed some fluorescence in the form of small fluorescent specks (Figs. 3c, d). Furthermore, the nuclei in glial cells are not fluorescent at all. This differential staining of neuronal nuclei is of great interest. Comparing the results of the Oudin and the fluorescent tests, there may be the possibility that, with the extraction procedure used, the protein was not extracted from the nuclei of the neurons. Discussion.-The finding of a brain-specific protein similar among many vertebrate species widely separated on the evolutionary scale has given a new impetus for studies of brain proteins. The fact that the present study shows that the S100 protein is localized in the glial system and neuronal nuclei of the Deiters' nucleus raises tantalizing questions: Where is this protein made? Does it move from the glia to the neuronal nuclei? Above all, what is its function? Previous studies in this laboratory have shown that inverse changes occur between neurons and glia in respiratory enzyme activities and in the amount and base composition of RNA.7-10 RNA fractions with the same base composition have been found concurrently to increase in the neurons and to decrease in the glia in 1 hr.8 With respect to the metabolic turnover of the S100 protein, there are suggestions that it may be very high. Rubin and Stenzel"5 reported that 15 per cent of the radioactive protein formed in a cell-free system from rabbit brain can be precipitated with the specific anti-S100 serum. Studies in this laboratory16 have shown that the S100 protein reaches its highest specific activity in vivo within 30 min of an injection of radioactive amino acid and that the specific activity declines markedly by 24 hr. The relationship between this rapid turnover and the cellular localization, particularly the finding of the S100 protein in neuronal nuclei, presents an interesting problem for future investigation. There may be the possibility, considering the acidic nature of the protein, that it might react with histones in neuronal nuclei and thereby control the genetic activity. Summary.-The acidic protein called protein S100 and specific only for the nervous tissue has been localized mainly to the neuroglia by single diffusion agar precipitation in glass capillaries using antiserum. Application of the fluorescent antibody sandwich technique showed the specific fluorescence to be localized around the nuclei of oligodendrocytes and scattered throughout the glial membrane system. The neurons only showed specific fluorescence in the nuclei. The significance of these findings is discussed in relation to the rapid turnover of the protein S100.
358
ZOOLOGY: BACHVAROVA ETl AL.
PROC. N. A. S.
The authors would like to express their gratitude to Dr. L. Lindholm, Department of Bacteriology, for his instruction in and assistance with the fluorescent antibody technique. They would also like to thank Mrs. K. Bjurstam for excellent technical assistance. The antiserum prepared against rabbit protein S100 was kindly given by Dr. L. Levine, Department of Bacteriology, Brandeis University, Waltham, Massachusetts, and by Dr. A. Rubin, Cornell Medical School, New York. Economic support was given by the Swedish Medical Research Council and by the U.S. Air Force European Office of Aerospace Research grant no. EOAR 63-28 and by the Wallenberg Foundation, Stockholm. t Recipient of a U.S. Public Health Service postdoctoral fellowship, 2-F2-SM-22, 097-02. This support is gratefully acknowledged. 1 Altman, J., Science, 135, 1127 (1962). 2 Moore, B. W., and D. McGregor, J. Biol. Chem., 240, 1647 (1965). 3 Moore, B. W., Biochem. Biophys. Res. Commun., 19, 739 (1965). 4Levine, L., and B. W. Moore, Neurosci. Res. Progr. Bull., 3, 18 (1964). 5 Levine, L., personal communication. 6 Hyden, H., Nature, 184, 433 (1959). 7 Hyden, H., and A. Pigon, J. Neurochem., 6, 57 (1960). 8 Hyde'n, H., and P. W. Lange, J. Cell Biol., 13, 233 (1962). 9 Hamberger, A., and H. Hyden, J. Cell Biol., 16, 521 (1963). 10 Hyden, H., and P. W. Lange, Naturwissenschaften, in press. 11 Oudin, J., in Methods in Medical Research, ed. M. Cohn (Chicago: The Year Book Publishers, 1951), vol. 5, p. 335. 12 Weller, T. H., and A. H. Coons, Proc. Soc. Exptl. Biol. Med., 86, 789 (1954). 13 Hyden, H., "The neuron," in The Cell: Biochemistry, Physiology, Morphology, ed. J. Brachet and A. E. Mirsky (New York: Academic Press, 1960), vol. 4, chap. 5. 14 Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem., 193, 265 (1951). 16 Rubin, A. L., and K. H. Stenzel, these PROCEEDINGS, 53, 963 (1965). 6 McEwen, B. S., and H. Hyden, in preparation. *
ACTIVATION OF RNA SYNTHESIS ASSOCIATED WITH GASTR ULA TION* BY R. BACHVAROVAI E. H. DAVIDSON,t V. G. ALLFREY, AND A. E. MIRSKY THE ROCKEFELLER UNIVERSITY
Communicated December 6, 1965
(Gastrulation in the amphibian embryo marks the onset of widespread, de novo cell differentiation. In this communication we present the results of a radioautographic and biochemical study of gene activity as gastrulation takes place in Xenopus laevis. We have found that gene activation occurs abruptly within a 1-hr period preceding the onset of gastrulation and that most of the newly synthesized RNA resulting from this activation appears to be messenger RNA. Pregastrular and Postgastrular Gene Activity.-From- gastrulation onward, development is organized through the controlled activity of the cell-borne genetic apparatus. Although the zygote genome and its blastomeric descendants manifest a slowly rising level of activity all through cleavage and blastulation,'- it is not until gastrulation that the indispensability of embryonic gene action can be demonstrated. This position dates back to the classical sea u'rchin hybridization studies
