Induction of glutamine synthetase in embryonic neural retina: its suppression by the gliatoxic agent α-aminoadipic acid

Developmental Brain Research, 1 (1981) 103-119 © Elsevier/North-Holland Biomedical Press

103

I N D U C T I O N OF G L U T A M I N E SYNTHETASE IN EMBRYONIC N E U R A L R E T I N A : ITS SUPPRESSION BY THE GLIATOXIC A G E N T a - A M I N O A D I P I C ACID

PAUL J. LINSER and A. A. MOSCONA Developmental Biology Laboratory, CummingsLife Science Center, University of Chicago, Chicago, Ill. (U.S.A.)

(Accepted October 7th, 1980) Key words: glutamine synthetase -- a-aminoadipic acid - - Mfiller cells -- embryonic retina -hormonal induction -- cell aggregates

SUMMARY Competence for cortisol-mediated induction of glutamine synthetase (GS) is a differentiation marker of embryonic neural retina. Earlier work has indicated that the induction and accumulation of GS is localized in the Mfiller glia cells. This localization was presently confirmed by the finding that the gliatoxin D,L-a-aminoadipic acid (AAA)reduces responsiveness to GS induction by 6 0 - 9 0 ~ due to preferential damage to Mfiller cells. The tests were performed on organ cultures of retina tissue from chick embryos, and on retina cell aggregates in which there is tissue reconstruction. The presence of GS-inducible Mfiller cells was monitored by immunostaining of tissue sections with anti-GS antiserum. Reduction of GS inducibility due to pretreatment with AAA resulted in virtual absence of cells that immunostained for GS. The preferential toxicity of AAA for Mfiller cells was also demonstrated by cell viability tests; it was further corroborated by the finding that treatment with AAA greatly reduced the level of carbonic anhydrase activity, another enzyme localized predominantly in Mfiller cells, but did not affect x-aminobutyric acid transaminase and choline acetyl transferase, neuronal enzymes. Susceptibility of Mfiller cells to AAA was found to increase with embryonic development of the retina. We suggest that acquisition of susceptibility for AAA represents another differentiation marker of embryonic Mfiller cells.

INTRODUCTION Glutamine synthetase (L-glutamate ammonia ligase; EC 6.3.1.2), referred to as GS, is a developmentally regulated biochemical marker of differentiation in the

104 vertebrate neural retinal7,20, 25. In the early chick embryo the level of GS in the retina is very low, but rises steeply on the sixteenth day of development, following elevation of systemic corticosteroid inducers, and increases 100-fold in the next few days. However, competence for GS induction is present in the retina long before it is normally expressed, since GS can be induced precociously several days ahead of its regular time by administering cortisol to premature embryos, or directly to early retina tissue isolated in organ culture24, ~9. In fact, the retina acquires induction competence between the sixth and eighth day of development, and thereafter its responsiveness to GS induction increases progressively 25. It has been shown that GS induction represents de novo synthesis and accumulation of the enzyme 19, and that it requires gene expression elicited by the steroidal inducer23, 3a. Especially significant for the present work has been the recent immunohistochemical demonstration that both in precocious and in normal induction, GS accumulates exclusively in the Mtiller cells 11,zl which are the glia of the avian retinal Also in other vertebrate retinas GS has been found to be confined to Mfiller cells 3~,34. These results indicated that inducibility for GS in the embryonic retina and the accumulation of this enzyme represent a characteristic feature of the differentiation of embryonic M/iller glia cells. In view of the importance of glia cells in retina development and function, and the significance of GS in a variety of metabolic processes, we considered it essential to look for additional biochemical markers of Mfiller cell differentiation, and to explore the relationship between Mfiller cells and GS induction by means of still other experimental approaches. Olney et al. 27 and Pedersen and Karlsen ''8 reported that D,L-a-aminoadipic acid (AAA), an aoalogue of glutamate 14,~7, can preferentially destroy Mfiller cells in adult rat and infant mouse retina. Although the mechanism of AAA's effects is unresolved, its reported selectivity led us to examine if embryonic Mfiller glia cells were sensitive to AAA, and whether this agent could be useful in studying the relationship between these cells and GS induction. We found that Mfiller cells in the retina of the chick embryo are differentially susceptible to AAA and that their susceptibility increases with development. The results corroborate previous evidence concerning the localization of GS induction in Mfiller cells; furthermore, the increased susceptibility to AAA represents still another differentiation marker of embryonic Mfiller glia cells. MATERIALS AND METHODS Animals White leghorn chick embryos, incubated at 37.2 °C and staged, were used throughout this study. D,L-a-aminoadipic acid AAA was purchased from Sigma Chemical. Tissue and cell cultures For organ culture, neural retinas were aseptically isolated from eyes of embryos

105 and each was cut into 8 sectors in sterile Tyrode's solution. Four sectors were then placed in a 25 ml Erlenmeyer flask with 3 ml Medium 199 (with Hanks' salts, Microbiological Associates), supplemented with 10 ~ fetal bovine serum (FBS), 1 penstrep mixture and 1% glutamine. The flasks were gassed with 5 ~ COz-air mixture, sealed, and placed on a rotary shaker (72 rpm) at 37 °C. Medium was changed daily. Monolayer cell cultures were prepared from cell suspensions obtained by trypsinization of retina tissue, or of cell aggregates (see below), according to standard procedure 16,2z. Briefly, the tissue (or cell aggregates) was treated with 0.3 ~ crystallipe trypsin (ICN) in calcium- and magnesium-free Tyrode's solution for 20 min at 37 °C, rinsed with Medium 199 that contained 150 #g/ml soybean trypsin inhibitor (Worthington), 15/~g/ml DNAase I (Worthington), and 1 ~ penstrep solution. Dispersion into single cells was by repeated gentle pipetting in the above medium. The cells were plated in Falcon tissue culture plastic dishes and placed in a 5 ~ CO2-air incubator at 37 °C. After 2 h, 10~ FBS was added to the medium. Cell aggregates were prepared as beforOl,16; suspensions of cells dissociated from 6-day embryonic retinas were dispensed into Erlenmeyer flasks with 3 ml culture medium, 1.5 × 107 cells per flask. The flasks were gassed with 5 ~ C O , a i r mixture, sealed and rotated on a gyratory shaker (72 rpm) at 37 °C. Medium was changed daily.

Induction of glutamine synthetase Induction of glutamine synthetase (GS) in cultures of retinal tissue or of cell aggregates was by addition of cortisol (hydrocortisone-free base; Sigma) to the medium to a final concentration of 0.33 #g/ml z3. After times specified in Results, the cultures were harvested and assayed for GS activity.

Enzyme assays The specific activity of GS was determined in sonicates of retinal tissue or of aggregates by the colorimetric assay previously described 23 modified for increased sensitivity (to be reported elsewhere). The assay follows the reaction: glutamine + ATP NHzOH As -+ ~,-glutamylhydroxamate(GHA) 6- NH3. Results are expressed as GSspecific activity, i.e., #M GHA formed/rag protein/h. Carbonic anhydrase (CAH) activity was measured by the micromethod of Maren ~3. y-Aminobutyric acid transaminase (GABA-T) activity was measured by the method of Hall and Kravitz 6 using y-[2,3-3H(N)]aminobutyric acid (New England Nuclear, NET-191) as substrate. Choline acetyl transferase (CAT) was assayed as reported by Schwarz and Coylezn. Protein was determined by the method of Lowry et al. 12.

Anti-GS lgG GS was purified from adult chicken retinas as described a2. Rabbits were immunized with GS and serum was collected as in previous work 11. Each serum batch was assayed for immunoprecipitation of GS 4. The monospecificity of the antiserum was established by Ouchterlony double-diffusion and immunoelectrophoresislL The

106 IgG fraction was purified from the antiserum and from preimmune serum by a m m o n i u m sulfate fractionation followed by DEAE-sephacell (LKB) chromatography 3.

Immunostaining with anti-GS IgG Detection of GS by immunostaining and indirect immunofluorescence was previously described 11. Tissue sections mounted on glass slides were exposed to a solution (0.2 mg/ml) of anti-GS I g G (that contained 1:50 normal goat serum) for 30 rain at 37 °C in a COz incubator. The sections were then rinsed extensively and exposed to 0.02 mg/ml fluorescein-conjugated goat-anti-rabbit IgG ( F I T C - G A R ; Miles-Yeda) for 30 min as above. After rinsing, the specimens were examined microscopically by illumination from an ultraviolet source for localization of immunofluorescence. Tissue sections were also stained with hematoxylin and eosin (H and E).

Measurements of DNA synthesis Cell cultures were exposed for 1 h to 1 #Ci/mt tritiated [methyl-aH]thymidine (New England Nuclear). The cultures were then rinsed in ice-cold Tyrode's solution, the cells were stripped into cold H20, disrupted by sonication, and the D N A content was determined by the diphenylamine reaction following precipitation and washing in cold trichloro-acetic acid 1. Radioactivity was measured in a Liquid Scintillation Counter. The results are expressed as dpm/#g DNA.

lO0-

ro

~ 60-

20I

I

0

.2

i

i

i

I

.4 .6 .8 1.0 r r ~ m l AAA

/I

I

J.O

Fig. 1. Dose-dependent effect of pretreatment of embryonic retina with AAA on GS induction. Retinas from 11-day chick embryos were cultured for 48 h in medium with AAA (medium changed daily), then were washed and transferred for 48 h to medium with cortisol (0.33/~g/ml) to induce GS. GS spec. act. was then determined. Each point represents the mean of 3 determinations; vertical bars indicate the range.

107 RESULTS A A A effects on Miiller cells In preliminary dose-response tests retinas isolated from 11-12 day chick embryos were treated in vitro with AAA for 48 h; they were then rinsed, transferred for 48 h to medium with cortisol to induce GS, and assayed for GS. Tissue sections were also examined histologically and by indirect immunofluorescence for the presence of GS in Mtiller cells11. The results (Fig. 1) showed that pretreatment of the retina with AAA caused a dose-dependent reduction in its responsiveness to GS induction. Pretreatment with 200 /~g/ml AAA reduced GS inducibility by 60%; histological examination showed that this dose affected predominantly Mfiller cells, but caused little or no discernible pathological changes in neurons (see below). Higher doses had more generalized adverse effects and therefore were not used in this study. It should be noted that addition of AAA directly to GS-containing preparations during the enzyme assay had no effect on the enzyme activity. Histological examination of embryonic retinas treated for 24 and 48 h with AAA showed progressive degenerative changes in Miiller cells (Fig. 2) quite similar to those reported for AAA-treated adult rat retina zS. These included edema, 'rarification' of cytoplasm and nucleoplasm, and cell death. Since M/iller cell nuclei are located in the bipolar cell layer in close proximity with neurons, we cannot state with certainty that the latter were completely unaffected; however, our observations strongly point to Mtiller cells as, at least, the preferential target of AAA. As an expected consequence of Miiller cell degeneration, there was disruption of the outer limiting membrane, i.e. of the junctions that link M/iller and photoreceptor cells at the outer boundary of the retina. This resulted in progressive structural distortion and disorganization of the photoreceptor cell layer and, to some extent, also of the bipolar cell layer; however, there was no detectable loss of photoreceptor cells. After transfer of the retina to medium without AAA for 48 h only a few abnormal nuclei and edematous cells could be noted. However, the outer layer of the retina remained grossly distorted, while the other cell layers appeared nearly normal. Presumably, by now most of the originally lethally damaged M/iller cells had been lost, while those remaining were either unaffected, or had recovered. Immunostaining for GS of retinas not treated with AAA and exposed to cortisol to induce GS revealed, as expected, numerous M/iller cells that strongly reacted with the anti-GS lgG (Fig. 2), consistent with previous evidence of GS induction and localization in these cells11,21. In sharp contrast, when retinas were pretreated with AAA for 48 h before exposure to cortisol, immunostaining revealed very few Mfiller cells that reacted with the anti-GS IgG, consistent with the low level of GS activity detectable by enzyme assay. The above results showed that pretreatment of 11-day embryonic retina with AAA markedly reduced inducibility for GS, and that this coincided with a drastic reduction in the number of M~ller cells that immunostained for GS. While these results corroborate the localization of GS in M/iller cells, they did not conclusively resolve the question of whether AAA lethally damaged these cells or merely caused them to become non-responsive to GS induction.

Fig. 2. Histological effects of treatment of embryonic retina with AAA. Retinas from I 1-day embryos were cultured as described in Fig. 1, then sectioned and stained with H and E, or immunostained with anti-GS lgG. A: section of freshly dissected I l-day retina. B: section similar to A, but immunostained for GS with anti-GS IgG and FITC-GAR and examined in ultraviolet light; no immunofluorescence for GS. C: section of ll-day retina cultured for 2 days in normal medium, then for 2 days in cortisolmedium to induce GS. D: section similar to C, but immunostained for GS (as in B) showing intense fluorescence in Mi~ller cells. E: 1l-day retina cultured for 24 h in medium with AAA (200 itg/ml) ; note partial disorganization of the outer (photoreceptor) cell layer and perturbations in the other layers. F: I 1-day retina after 48 h culture in medium with AAA; persistence of cellular disorganization- G: I l-day retina cultured for 48 h in AAA, then for 48 h in cortisol-medium; note partial reorganization in the bipolar cell layer and continued distortion of the outer cell layer, l-I: section similar to G, but immunostained for GS (compare with D); faint staining of occasional elongated strands and possibly the ganglion layer. Abbreviations: PP, photoreceptor cell processes ; of residual M ~ller cell 'end feet' in olin, outer limiting membrane; pc, photoreceptor cell layer; up, outer plexiform layer; bc bipolar cell layer; ip, inner plexiform layer; gc, ganglion cell layer; t~f, nerve fiber layer; tim, inner limiting membrane. × 290.

109 TABLE I Reinitiation of D NA synthesis in monolayer cultures of cells dissociated from embryonic retina tissue: effect of pretreatment with AAA Retinas from 12-day embryos were cultured for 48 h in medium with or without 200/~g/ml AAA. They were then washed, dissociated into single cells and the cells were plated at equal viable cell densities. After 48 h the monolayer cultures were labeled with [aH]thymidine for 1 h and incorporation of the isotope into DNA was determined (see Methods).

Controls +AAA

I~g DNA/culture

dpm

dpm/l~g DNA

22.25 22.5 20.5 22.0 25.0 22.2

8109 5768 5765 625 666 631

364.4 256.3 281.2 28.4 26.6 28.4

This question was further examined as follows. Retinas from 11-12 day embryos were cultured for 48 h in m e d i u m With or without 200/~g/ml A A A ; they were then dissociated into suspensions of single cells a n d these were plated as monolayers at identical viable cell densities. F r o m previous work it is k n o w n that the retina in 11-12 day embryos has already reached the end of its growth phasO7,20,25; however, when

h

~

f

Fig. 3. Monolayer cultures of cells dissociated from AAA-pretreated (A) and control (B) retinas. Retinas from 1l-day embryos were maintained for 48 h in medium with 200 #g/ml AAA, or without it; they were then dissociated and the cells were plated at equal viable cell densities in culture dishes. After 5 days, the monolayer cultures were fixed and stained with H and E. A: mostly neuronal cells. B : neuronal and LER cells. × 250.

110 its cells are dissociated and plated in monolayer cultures, DNA synthesis is reinitiated in non-neuronal cells°, resulting in multiplication of large epithelioid retinocytes (LER cells) derived from Mfiller glia cells. We examined if cells dissociated from retinas pretreated with AAA would produce monolayer cultures of multiplying LER cells, or if the pretreatment destroyed the precursors of these cells, i.e. the Mfiller cells. To monitor DNA synthesis the cultures were pulsed for 1 h with [3H]thymidine 48 h after cell plating. The results showed 10 times more DNA synthesis in cell cultures established from control retinas than in those established from AAA-pretreated retinas (Table I); even after 5 days the latter contained very few Mfiller gila-derived LER cells and consisted mostly of non-replicating neurons, whereas the control cultures contained a dense population of LER cells in addition to neurons (Fig. 3). It might be suggested that AAA treatment of retina tissue affects the proliferative capacity of cells without actual cell killing. To test this possibility, we treated early retinas which are mitotically very active (i.e. day 6) with AAA for 48 h and then measured the rate of thymidine incorporation. We found no significant differences between the rates of incorporation in AAA-treated retinas and untreated controls, and concluded that AAA exposure did not affect DNA synthesis per se. The possibility that pre-treatment with AAA blocked proliferation of Mfiller cells without killing these cells cannot be completely dis-

-- NC

|

+HC

"5

II

12 Ernbrtjonic

15

14

15

age(de~$)

Fig. 4. GS induction in aggregates of embryonic retinal cells. Suspensions of trypsin-dissociated cells from 6-day embryo retinas were aggregated by rotation in 25 m! Erlenmeyer flasks (!.5 × 107 cells/3 m! flask); the cell aggregates were cultured for 3 through 7 days, i.e. to stages equivalent to embryonic ages of 9 through 13 days. They were then transferred for 48 h to medium with or without cortisol and assayed for GS. The equivalent embryonic age at the time of assay is given below each bar. + H C , cortiso] treated ; - - H C , control. Each bar represents the mean of 3 determinations; vertical lines indicate the range.

111 TABLE II Effect of pretreatment with AAA on GS induction in retina cell aggregates

Cultures of cell aggregateswere prepared from 6-day embryo retina cells. After 5 days, 200 ktg/mlAAA was added to some of the cultures for 48 h. All the cultures werethen transferred to media with or without cortisol for 48 h and assayed for GS activity. HC, cortisol.

GS spec. act.

--HC

+ HC

+ AAA + HC

54.1 55.9 37.3

162.3 146.1 115.0

13.2 11.8 10.8

counted on the basis of these findings. These results support the hypothesis that treatment.of 11-12 day retina with 200 #g/ml AAA damaged or killed Mfiller cells and, therefore, that the reduction of GS inducibility in the AAA-treated retina was due to adverse effects of AAA on this type of cells. Experiments with aggregates o f retina cells

When retinal tissue is isolated in organ culture it tends to roll up because of its sheet-like structure, even whert it is cut into fragments. Histological examination indicated that this may hinder exchange of nutrients and prevent uniform access of AAA to all parts of the tissue, thereby causing variability in results. To overcome this problem, and to increase precision and consistency, we turned to the use of cell aggregates. Since this is a well-established experimental system 18 it needs only a brief explanation. Numerous studies have shown that aggregates of various neural cells can be effectively used as in vitro model systems for analyzing a variety of developmental and pharmacological problemsT,S,30,35,3L When retinal tissue from 6-day chick embryos is dissociated by trypsinization into single cells and the cell suspension is swirled in flasks by rotation on a gyratory shaker (72 rpm; a/4 in. diameter of rotation), the cells associate into numerous small multicellular aggregates (0.3-0.5 mm in diameter). Within these aggregates the cells progressively reconstruct retinotypic tissue architecture and continue to develop15, aS. The aggregates contain numerous rosettes, each with an internal lumen bordered by a photoreceptor cell layer surrounded by bipolar and ganglion cells. Previous work has shown that these cell aggregates display developmental features characteristic of embryonic retinal tissue; they acquire competence for GS induction at the expected embryonic age; GS inducibility by cortisol increases with their differentiation similarly to intact retinal tissue15,25; and the enzyme is localized in Mfiller cellstl,~L Fig. 4 illustrates the increase in GS inducibility in cultures of retina cell aggregates with advancing embryonic age of the cells. Using cultures of retina cell aggregates, we examined the following questions: (1) does their pretreatment with AAA reduce responsiveness to GS induction ?; (2) can the identification of Mfiller cells as the preferential target of AAA be corroborated by the

112 use of other cell markers ?; (3) are there age-dependent differences in cell susceptibility to this effect of A A A ? Table II shows the effect of pretreatment with A A A on GS induction in cell aggregates. The aggregates were prepared from retina cells of 6-day embryos; they were cultured for 5 days (i.e. to a stage equivalent to I 1-day embryonic age), treated with 200/~g/ml A A A for 48 h, then transferred to medium with cortisol for 48 h and assayed for GS. Controls were not exposed to AAA. The results showed that pretreatment with A A A consistently reduced responsiveness to GS induction by more than 90 ~ compared with untreated controls.

Fig. 5. Histological effects of AAA on aggregates of embryonic retina cells. Cell aggregates were prepared from 6-day embryo retina cells. After 5 days AAA (200 pg/ml) was added to half the cultures for 48 h. Treated and control cultures were then transferred to normal medium with cortisol for 48 h to induce GS. A: section of control aggregate after 7 days in culture; note numerous typical rosettes (arrow). B: cell aggregate after 48 h exposure to AAA; note edematous cells and nuclei (arrows). C: an aggregate cultured in AAA, then in cortisol-medium ; partial recovery, but note pycnotic cells and absence of rosettes. D: control aggregate (as in A), immunostained with anti-GS lgG for cortisoMnduced GS ; intense staining of cells identifiable as MOilercells. E : aggregate pretreated with AAA, then exposed to cortisoJ and immunostained for GS (compare with D); faint staining in occasional elongated fibers. F: control aggregate induced for GS (as in D), stained with non-immune IgG to demonstrate specificity of staining with anti-GS lgG. x 240.

113

GS

CAH

GABA-T ]

Fig. 6. Effects of pretreatment with AAA of retina cell aggregates on levels of GS, CAH, GABA-T

and CAT. Cell aggregates were prepared from 6-day embryo retina cells. After 4 days, AAA (200 /~g/ml)was added to half the cultures for 48 h. All the cultures were then transferred to cortisol-medium for 48 h and assayed for the enzymes. The spec. act. for each of the controls (open bars) represents 100~ and the experimental values (black bars) are expressed accordingly. Each bar represents the mean of 2 or 3 determinations.

Histological examination of cell aggregates treated for 48 h with AAA revealed changes essentially similar to those described above for retinal tissue (Fig. 5): edema and 'rarification' of cytoplasm and nucleoplasm in cells identifiable as Mfiller glia, and distortion and disorganization of the photoreceptor cell layer in most of the rosettes. Following transfer and culture in medium without AAA, few edematous cells were seen, but the number of pycnotic cells increased, and most of the rosettes were disorganized. When AAA-pretreated cell aggregates were immunostained for GS, they showed only very few faintly staining areas (Fig. 5), consistent with the low level of GS induction detected by biochemical assay. Further evidence that Mfiller cells are the preferential target for AAA was provided by measurements of carbonic anhydrase (CAH), gamma aminobutyric acidtransaminase (GABA-T), and choline acetyl transferase (CAT). According to previous studies, C A H in the retina is confined to the glial compartment 26, while GABA-T and CAT are localized predominantly in the neuronal compartment 34. If AAA preferentially damages Mfiller cells, it should reduce the level of C A H in addition to reducing GS inducibility; on the other hand, the levels of GABA-T and CAT should not be reduced, considering their predominantly neuronal localization. The results were consistent with these expectations. Cell aggregates prepared from 6-day embryo retina cells were cultured for 4 days and then were exposed to AAA (200/~g/ml) for 48 h; they were then transferred to

114 cortisol-medium for 48 h and assayed for GS, CAH, GABA-T and CAT. The results (Fig. 6) showed that pretreatment with A A A resulted in markedly lower C A H activity as well as in the expected reduction of GS inducibility; in contrast, the levels of GABA-T and CAT were not reduced. In addition, immunostaining of AAA-treated retina cell aggregate or tissue sections with antiserum to chicken C A H - C showed no staining whereas untreated controls had very intense staining limited exclusively to the Mfiller cells (manuscript in preparation). While these results do not fully exclude possible effects of AAA on neurons, they are consistent with the conclusion that Mfiller glia cells are the preferential target for AAA in the retina. This conclusion was further supported by evidence that also in cell aggregates, treatment with AAA greatly reduced the number of viable Mfiller cells. Cell aggregates prepared from 6-day embryonic retina cells were cultured for 4 days and were then treated with AAA (200/~g/ml) for 48 h. They were then dissociated into single cells and the cells were plated as monolayer cultures. Similar cell cultures were prepared also from untreated aggregates. The cultures were examined for outgrowth of Mfiller gliaderived LER cells; and they were labeled with [ZH]thymidine to monitor D N A synthesis. The results (Table III) showed that in cell cultures derived from AAA-pretreated cell aggregates D N A synthesis was markedly lower than in controls. Such cultures contained very few LER cells, although numerous neurons were present, whereas cultures derived from control aggregates contained also abundant LER cells. These results parallel those seen in cell cultures from intact retina tissue in showing that treatment with AAA preferentially damages or destroys Mfiller glia cells. Age-dependent differences in susceptibility to A A A

In the above experiments the cell aggregates were exposed to AAA after they had reached a stage equivalent to 10-11 days embryonic age. To examine if there are agedependent differences in susceptibility to AAA, cell aggregates were treated for 48 h with AAA after 2 days or after 6 days in culture, i.e. at stages equivalent to 8 and 12 days of embryonic age. The aggregates were then dissociated into cells, and these were TABLE [II Reinitiation of DNA synthesis in monlayer cultures of cells dissociatedfrom retina cell aggregates : effect of pretreatment with AAA

Cultures of cell aggregates were prepared from 6-day embryo retina cells. After 4 days, 200 #g/ml AAA was added to half of the cultures for 48 h. They were then washed, dissociated into single cells and the cells were plated at equal viable cell densities. After 48 h, the cultures were labeled with [3H]thymidine for 1 h and incorporation of the isotope into DNA was determined (see Methods).

Controls ÷AAA

l+g DNA/culture

dpm

dpm/l~g DNA

5.5 4.9 4.6 5.2

3309 4079 680 825

601.6 832.4 147.8 158.6

115 plated as monolayer cultures to determine survival of Mfiller cells. Examination of these cell cultures revealed striking age-related differences. Those established from aggregates treated with AAA at the older age contained neuroos, but almost no Mfiller cell-derived LER cells; in contrast, those established from aggregates treated at the younger age contained also abundant outgrowth of LER cells. These results suggest that Mfiller cell susceptibility to AAA increases with embryonic age, i.e. as a function of their progressive differentiation. To further examine this relationship, cell aggregates were treated with AAA at different embryonic ages, and the effect on GS inducibility was determined. Cultures of cell aggregates were prepared from 6-day embryo retinal cells. Some were treated with AAA (200 #g/ml) between 0 and 48 h, i.e. while the cells were between the sixth and eighth day of embryonic age; others were treated between 48 and 96 h; still others between 72 and 120 h. All the aggregates, including untreated controls, were exposed to cortisol-medium between 120 and 168 h and then were assayed for GS. In parallel tests, cell aggregates were set up as above and were treated with AAA for periods of 48 h at staggered times; however, in this series, the aggregates were exposed to cortisol immediately following treatment with AAA, rather than after 120 h. In both series, responsiveness to GS induction served as an indicator of Mfiller cell susceptibility to AAA at different embryonic ages. The results are presented in Figs. 7 and 8. They clearly show that treatment with AAA at an older embryonic age resulted in much greater reduction of GS inducibility than treatment at a younger age. Therefore, Mfiller cells become increasingly

¢J

eO-

It) U

"¢ a..

so-

4020-AAA

conf

+ AAA

I

~

÷ AAA

÷ AAA

0-4813 4 8 - 9 6 h ?2-1,~Oh

Fig. 7. Developmental increase in susceptibility to AAA. Cell aggregates were prepared from 6-day embryo retina cells. Some were treated with A A A (200 #g/ml) from 0 to 48 h, others from 48 to 96 h, and still others from 72 to 120 h. Controls were not treated (--AAA). All were transferred into cortisolmedium from 120 to 168 h, then were assayed for GS. Each bar represents the mean of 3 determinations.

116

+ AAA

"~ ~t

| -AAA

~.

~

-

II

12

13

14

15

Embr~fonic acje (do~/s)

Fig. 8. Similar experimental protocol as in Fig. 7, except that in each case the cell aggregates were transferred to cortisol-medium immediately following pretreatment with A A A and were assayed for GS 48 h later. Comparison is between the level of GS induced in aggregates pretreated with AAA (open bars) and in untreated aggregates (black bars). Points connected by line show the percentage reduction in GS inducibility due to pretreatment with AAA, as a function of embryonic age of the cells at assay time. All data represent means of 3 determinations.

susceptible to the effect of AAA as their differentiation progresses; at early developmental stages, Mfiller cells (or their precursor stem cells) are either unaffected or can recover from the exposure to AAA. Accordingly we suggest that acquisition of susceptibility to AAA is a developmental attribute of M/filler cells and that this represents an additional 'marker' characteristic for these cells. It is noteworthy that development of Mfiller cell susceptibility to AAA parallels temporally the increase in their responsiveness to GS induction by cortisol (Fig. 8). While this relationship may be merely coincidental, it deserves future examination considering that AAA is an analogue of glutamate and GS is involved in glutamate-glutamine metabolism. DISCUSSION

M/filler cells are the characteristic glial elements of the vertebrate retina. Their analysis, particularly during embryonic differentiation, would be greatly aided by availability of biochemical markers and cytotoxic agents selective for these cells. A characteristic developmental feature of the embryonic retina of the chick is its inducibility for glutamine synthetase (GS) by cortiso125. Previous work has indicated

117 that this induction and the subsequent accumulation of GS takes place in MOiler cellslX,21. The present study substantiates this by exploring the'effects of the gliatoxic agent AAA (a-aminoadipic acid). This analogue of glutamate has been shown by otherslO,27,2a to be cytotoxic for MOiler cells in infant mice and adult rats without noticeable effects on retinal neurons. We found that also in the retina of the chick embryo AAA preferentially damages MOiler cells and consequently severely reduces inducibility for GS. Of particular interest is the finding that the susceptibility of Mfiller cells to AAA increases with their differentiation, i.e. with the embryonic age of the retina. Therefore, acquisition of susceptibility to AAA can be considered a 'differentiation marker' characteristic of embryonic MOiler cells. The above results and conclusions are based on tests performed in part on cultures of isolated embryonic retinal tissue, but mostly on cultures of cell aggregates prepared from suspensions of freshly dissociated retinal tissue. The analytical advantages of this experimental system were explained in the Results. The preferential susceptibility of MOiler cells to AAA was determined by several criteria: (1) histological examination of AAA-treated retina and retina cell aggregates revealed cytopathic changes confined predominantly to M filler cells. Immunostaining for GS showed that pretreatment with AAA resulted in marked reduction or virtual absence of GS-containing MOiler cells, while in the untreated controls such cells were abundant; (2) in addition to severely reducing inducibility for GS, pretreatment with AAA markedly lowered the level of CAH, another enzyme marker reportedly confined to MOiler cellse6 and glia5, whereas it did not reduce the levels of GABA-T and CAT, predominantly neuronal enzymes34; (3) viability of MOller cells was tested by plating cells dissociated from embryonic retina tissue or cell aggregates. Normally, in monolayer cultures established from such cells, MOiler cells begin to multiply and form an abundant population of LER cells; this was almost completely prevented by pretreatment with AAA, evidently because of lethal damage to Mfiller cells. These findings agree with electron microscopic studies by others ~8, and biochemical analysis of various neuronal markersaO,27 and justify the conclusion that AAA preferentially damages MOiler glial cells in the embryonic neural retina. Our results do not address themselves to, or exclude the possibility of effects of this agent on some types of retinal neurons. AAA is a structural analogue of glutamatO4, 27, but its mechanism of action on cells is unknown. Electron microscopic examination of Miiller cells in adult rat retina led Pedersen and Karlsen2s to suggest that A A A acts directly rather than through a secondary metabolite. Olney et al. raised the interesting possibility that the effect of A A A on glial cells may be due to reaction with the cell membrane 27. If a membrane receptor is involved (as in the case of kainic acid's cytotoxic interaction with neuronsa6), its nature deserves investigation, especially in view of the age or differentiation-dependent increase in susceptibility of embryonic MOiler cells to AAA. Our finding that embryonic MOller cells acquire increasing sensitivity to AAA with differentiation strongly suggests that the specific changes which cause them to become a preferential target for AAA represent a developmentally regulated 'marker' of these cells. Analysis of these changes and their relationship to the inducibility of GS

118 may help in clarifying aspects of differentiation that distinguish Mfiller cells from other types of glia a n d from neurons. Thus, A A A emerges as a promising probe for studying cell differentiation in the retina and other neural tissues. ACKNOWLEDGEMENTS This work is part of a research p r o g r a m supported by G r a n t HD01253 from the N a t i o n a l Institute of Child Health a n d H u m a n D e v e l o p m e n t (to A.A.M.), by Basic Research G r a n t 1-733 from the March of Dimes-Birth Defects F o u n d a t i o n , and by a postdoctoral fellowship (to P.J.L.) from T r a i n i n g G r a n t GM7542 from the N a t i o n a l Institute of Health to the University of Chicago.

REFERENCES 1 Burton, K., A study of the conditionsand mechanism of the diphenylamine reaction for colorimetric estimation of deoxyribonucleic acid, Bioehem. J., 62 (1956) 315-323. 2 Cajal, S. R. In R. W. Rodieck (Ed.), The Vertebrate Retina, Freeman, San Francisco, 1973, pp. 838-852. 3 Deutsch, H. F., Preparation ofimmunoglobulinconcentrates. In C. A. Williams and M. W. Chase (Eds.), Methods in Immunolagy and Immunochemistry, Vol. 1, Academic Press, New York, 1967, pp. 315 328. 4 Garfield, S. and Moscona, A. A., Glutamine synthetase in the embryonic chick neural retina: the effect of cycloheximide on the conservation of labile templates for enzyme synthesis, Mech. Age. Develop., 3 (1974) 253-269. 5 Ghandour, M. S., Langley, O. K., Vincendon, G. and Gombos, G., Double labelling immunohistochemical technique provides evidence of the specificity of glial cell markers, J. Histochem. Cytochem., 27 (1979) 1634-1637. 6 Hall, Z. W. and Kravitz, E. A., The metabolism of 7-aminobutyric acid (GABA) in the lobster nervous system - - I, J. Neurochem., 14 (1967) 45-54. 7 Hatten, M. and Sidman, R., Cell reassociation behavior and lectin induced agglutination of embryonic mouse cells from different brain regions, Exp. Cell Res., 113 (1978) 111-120. 8 Honegger, P. and Richelson, E., Neurotransmitter synthesis, storage and release by aggregating cell cultures of rat brain, Brain Research, 162 (1979) 89-99. 9 Kaplowitz, P. B. and Moscona, A. A., Stimulation of DNA synthesis by ouabain and concanavalin A in cultures of embryonic neural retina cells, Cell Differentiation, 5 (1976) 109-119. 10 Karlsen, R. L., The toxic effect of sodium glutamate and D,L-a-aminoadipic acid on rat retina: changes in high affinity uptake of putative transmitters, J. Neurochem., 31 (1978) 1055-1061. 11 Linser, P. and Moscona, A. A., Induction of glutamine synthetase in embryonic neural retina: localization in Mhller fibers and dependence on cell interactions, Proc. nat. Acad. Sci. (Wash.), 76 (I 979) 6476-6480. 12 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J., Protein measurement with the Folin phenol reagent, J. biol. Chem., 193 (1951) 265. 13 Maren, T. H., A simplified micromethod for the determination of carbonic anhydrase and its inhibitors, J. Pharmacol. exp. Ther., 130 (1960) 26-29. 14 Meister, A., Glutamine synthetase of mammals. In P. Boyer (Ed.), The Enzymes, 3rd Edn., Academic Press, New York, 1974, pp. 699-754. 15 Morris, J. E. and Moscona, A. A., The induction of glutamine synthetase in cell aggregates of embryonic neural retina : correlations with differentiation and multicellular organization, Develop. Biol., 25 (1971) 420-444. 16 Moscona, A. A., Rotation-mediated histogenic aggregation of dissociated cells: a quantifiable approach to cell interactions in vitro, Exp. Cell Res., 22 (1961) 455-475. 17 Moscona, A. A., Induction of glutamine synthetase in embryonic neural retina: a model for the regulation of specific gene expression in embryonic cells. In A. Monroy and R. Tsanev (Eds.), Biochemistry of Cell Differentiation, FEBS Symp., Vol. 24, Academic Press, London, 1972, pp. 1 23.

119 18 Moscona, A. A., Surface specification of embryonic cells: lectin receptors, cell recognition, and specific cell ligands. In A. A. Moscona (Ed.), The Cell Surface in Development, John Wiley, New York, 1974, pp. 67-99. 19 Moscona, M., Frenkel, N. and Moscona, A. A., Regulatory mechanisms in the induction of glutamine synthetase in the embryonic retina: immunochemical studies, Develop. Biol., 28 (1972) 229241. 20 Moscona, A. A., Linser, P., Mayerson, P. and Moscona, M., Regulatory aspects of the induction of glutamine synthetase in embryonic neural retina. In Glutamine Synthetase Symposium, University of Mexico, 1979, in press. 21 Moscona, A. A., Mayerson, P., Linser, P. and Moscona, M., Induction of glutamine synthetase in the neural retina of the chick embryo: localization of the enzyme in Mfiller fibers and effects of BrdU and cell separation. In E. Giacobini, A.. Vernadakis and A. Shahar (Eds.), Tissue Culture in Neurobiology (Proc. Bat-Sheva Symposium, August, 1979), Raven Press, New York, 1980, pp. 111-127. 22 Moscona, A. A. and Moscona, M. H., Aggregation of embryonic cells in a serum-free medium and its inhibition at suboptimal temperatures, Exp. Cell Res., 41 (1966) 697-702. 23 Moscona, A. A., Moscona, M. and Saenz, N., Enzyme induction in embryonic retina: the role of transcription and translation, Proc. nat. Acad. Sci. (Wash.), 61 (1968) 160-167. 24 Moscona, A. A. and Piddington, R., Stimulation by hydrocortisone of premature changes in the developmental pattern of glutamine synthetase in embryonic retina, Biochim. biophys. Acta (Amst.), 121 (1966) 409-411. 25 Moscona, M. and Moscona, A. A., The development of inducibility for glutamine synthetase in embryonic neural retina: inhibition by BrdU, Differentiation, 13 (1979) 165-172. 26 Musser, G. L. and Rosen, S., Localization of carbonic anhydrase activity in the vertebrate retina, Exp. Eye Res., 15 (1973) 105-109. 27 Olney, J. W., Ho, O. L. and Rhee, V., Cytotoxic effects of acidic and sulphur containing amino acids on the infant mouse central nervous system, Exp. Brain Res., 14 (1971) 61-76. 28 Pedersen, O. O. and Karlsen, R. L., Destruction of Miiller cells in the adult rat by intravitreal injection of D,L-a-aminoadipic acid. An electron microscopic study, Exp. Eye Res., 28 (1979) 569-575. 29 Piddington, R. and Moscona, A. A., Precocious induction of retinal glutamine synthetase by hydrocortisone in the embryo and in culture: age-dependent differences in tissue response, Biochim. biophys. Acta ( Amst.) , 141 (1967) 429-432. 30 Ramirez, G., Manrique, E., Barat, A. and Villa, S., Analysis of developmental interactions in reaggregated embryonic cell cultures. In E. Giacobini, A. Vernadakis and A. Shahar (Eds.), Tissue Culture in Neurobiology, Raven Press, New York, 1980, pp. 129-145. 31 Riepe, R. E. and Norenburg, M., MiJller cell localization of glutamine synthetase in rat retina, Nature (Lond.), 268 (1977) 654-655. 32 Sarkar, P. K., Fischman, D. A., Goldwasser, E. and Moscona, A. A., Isolation and characterization of glutamine synthetase from chicken neural retina, J. biol. Chem., 247 (1972) 7743-7749. 33 Sarkar, P. K. and Moscona, A. A., Glutamine synthetase in embryonic neural retina: immunochemical identification of polysomes involved in enzyme synthesis, Proc. nat. Acad. Sci. (Wash.), 70 (1973) 1667-1671. 34 Sarthy, P. V. and Lain, D. K., Biochemical studies of isolated glial (MOiler) cells from the turtle retina, J. Cell Biol., 78 (1978) 675-684. 35 Seeds, N. W., Haffke, S. C. and Krystosek, A., Cell migration and recognition in cerebellar reaggregate cultures. In E. Giacobini, A. Vernadakis and A. Shahar (Eds.), Tissue Culture in Neurobiology, Raven Press, New York, 1980, pp. 129-145. 36 Schwarz, R. and Coyle, J. T., Kainic acid: neurotoxic effects after intraocular injection, Invest. Ophthal. Vis. Sci., 16 (1977) 141-148. 37 Trapp, B. D., Honegger, P., Richelson, E. and Webster, H., Morphological differentiation of mechanically dissociated fetal rat brain in aggregating cell cultures, Brain Research, 160 (1979) 117-126.