Effect of Concanavalin A on Tyrosine Aminotransferase in Rat Hepatoma Tissue Culture Cells RAPID REVERSIBLE INACTIVATION OF SOLUBLE ENZYME

Concanavalin A added to intact cells at 37 degrees caused rapid and reversible inactivation of a soluble enzyme, tyrosine aminotransferase, in two lines of rat hepatoma tissue culture cells grown in monolayer culture. This temperature-dependent process was independent of de novo protein and RNA synthesis and independent of increased uptake of Ca2+ and Mg2+ or glucose. The inactivation could be reversed by adding alpha-methyl-D-mannopyranoside a competing sugar for concanavalin A binding. Other lectins known to bind to different sugars did not bring about the inactivation of tyrosine aminotransferase. Addition of concanavalin A did not result in the inactivation of another soluble enzyme, lactic dehydrogenase. The maintenance of tyrosine aminotransferase in an inactive form after the binding of concanavalin A to the cells required the continued presence of concanavalin A. This effect of concanavalin A could not be mimicked either by dibutyryl cyclic adenosine or guanosine monophosphoric acid. Incubation of cell extracts with concanavalin A did not result in inactivation nor did mixing of extracts from concanavalin A-treated cells with extracts from untreated cells. On the basis of these results we conclude that the following are the essential requirements for concanavalin A to bring about the inactivation of tyrosine aminotransferase: (a) the binding of native concanavalin A to the cells; (b) integrity of certain structural elements of the cells.

Concanavalin A added to intact cells at 3'7" caused rapid and reversible inactivation of a soluble enzyme, tyrosine aminotransferase, in two lines of rat hepatoma tissue cdture cells grown in monolayer culture. This temperaturedependent process was independent of de nouo protein and RNA synthesis and independent of increased uptake of Ca"+ and Mg2+ or glucose. The inactivation could be reversed by adding cu-methyl-o-mannopyranoside a competing sugar for concanavalin A binding. Other lectins known to bind to different sugars did not bring about the inactivation of tyrosine aminotransferase.
Addition of concanavalin A did not result in the inactivation of another soluble enzyme, lactic dehydrogenase.
The maintenance of tyrosine aminotransferase in an inactive form after the binding of concanavalin A to the cells required the continued presence of concanavalin A. This effect of concanavalin A could not be mimicked either by dibutyryl cyclic adenosine or guanosine monophosphoric acid. Incubation of cell extracts with concanavalin A did not result in inactivation nor did mixing of extracts from concanavalin A-treated cells with extracts from untreated cells. On the basis of these results we conclude that the following are the essential requirements for concanavalin A to bring about the inactivation of tyrosine aminotransferase: (a) the binding of native concanavalin A to the cells; (b) integrity of certain structural elements of the cells.  (3,4). Con A also evokes increased Ca2+ uptake in mouse T lymphocytes (5), an effect mediated by cyclic nucleotides. However, the relation between Con A-induced Ca"+ ion uptake and DNA synthesis is not understood.
Other cellular effects of Con A binding include agglutination of growing cells or protease-treated densityinhibited cells (6, 7) and, at sufficiently high doses, cell killing (8,9). No general model is available to explain the mechanisms by which surface binding of the l&ins leads to these effects. In this report we present data describing a new cellular effect of Con A. Binding of the lectin to either of two lines of rat hepatoma cells in tissue culture resulted in a rapid and reversible inactivation of the intracellular enzyme, tyrosine aminotransferase (EC 2.6.1.5). This enzyme, found in the soluble fraction of cell extracts, is induced in both lines by glucocorticosteroid hormones (10, 11) and insulin (12,13), and in one of them  by analogs of cyclic adenosine monophosphate (14,15 by 95%) was added to the cells 10 min before the addition of Con A. As can be seen from Fig. 2, the inactivation was unaffected.
The identical result was obtained when Binding of IsHIAcetyl-Con A to FU-5-5 at 5" and 37"-FU-5-5 cells grown in 35-mm dishes were kept in PBS with (0.9 mM) Ca2+ and (0.99 mM) Mg*+ either at 5" (over ice) or at 37" (0.9 ml of PBS to each dish). To each dish 0.1 ml of 0.5 M cu-methyl-n-mannopyranoside solution was added and the dishes were maintained at their respective temperatures for 10 min. Another set of dishes which served as the uncompeted control received 0.1 ml of PBS. At the end of 10 min preincubation with the competing sugar or PBS, 50 ~1 of ["Hlacetyl-Con A (0.6 mg/ml, 3700 cpm/pg) were added to each dish. Cells were harvested at different times after exposure to Con A by aspirating the PBS, washing the monolayer of cells three times with cold PBS, scraping, and washing once more with cold PBS by centrifugation. The cell pellet was dissolved in 0.3 ml of 1 N NaOH and an aliquot of the solution was counted in 10 ml of Aquasol (New England Nuclear). The amount of Con A bound/mg of total cellular protein was converted into amount bound per lo6 cells by using the conversion factor determined empirically of lo6 cells = 260 fig of protein. The difference in the amount of Con A bound with and without a-methyln-mannopyranoside was taken as the specific binding of Con A. The binding of [3Hlacetyl-Con A to FU-5-5 cells in suspension was carried out in an identical manner, except that Con A was added to the cell suspension and cells were washed four times in PBS by centrifugation

Effect
of Con A on Tyrosine Amirwtransferase-When FU-5-5 cells with tyrosine aminotransferase preinduced by overnight exposure to 1 PM dexamethasone were added to a Con Atreated coverslip (24), we noted an immediate drop in the level of the enzyme.
This observation prompted us to study the effect of Con A on tyrosine aminotransferase. The effect of increasing concentrations of Con A on the induced level is shown in Fig. 1. In growth medium, above 25 pg/ml of Con A, the level of the enzyme decreased with increasing concentration of Con A and the effect was maximum at 200 pg/ml of Con A, no further effect being seen up to 500 pg/ml (data not included).
However, when Con A was added after replacing the growth medium with PBS containing Ca2+ and Mg2+, the maximum effect of Con A was seen at a much lower concentration ( Fig.  1). This could be due to the competition by serum glycoproteins and glucose for the binding of Con A to the cell FIG. 1. Effect of increasing concentrations of Con A on the induced level of tyrosine aminotransferase specific activity in FU-5-5 cells. FU-5-5 cells grown in 35-mm dishes were preinduced overnight (about 16 h) with 1 PM dexamethasone. Different concentrations of Con A were added to the induced cells incubated at 37". The cells were harvested after 30 min of treatment with Con A and enzyme assay was carried out as described under "Experimental Procedure." Con A was added to cells either in growth medium (O--O) or PBS with Ca'+ and 2. Kinetics of inactivation of induced level of tyrosine aminotransferase specific activity by Con A at 37" with and without cycloheximide.
FU-5-5 cells preinduced overnight with 1 PM dexamethasone in 35-mm dishes were treated with 200 pg/ml of Con A. At different times after the addition of Con A, the cells were harvested after washing the cells once with 2 ml of cold PBS. Tyrosine aminotransferase assay was done as described (0-O).
To study the effect of cycloheximide on the inactivation brought about by Con A, cycloheximide at a concentration of 100 pg/ml was added to the induced cells 10 min before the addition of Con A. The rest of the procedure was the same as described before (O---O). some mechanism related to the prior treatment of the cells with steroid, Con A was added to uninduced cells. There was a rapid-decrease in the basal activity of the enzyme (Fig. 3). Con A is effective only in its native tetravalent form in bringing about the inactivation of tyrosine aminotransferase in FU-5-5 cells, for the divalent succinylated Con A did not do so in these cells (Table  I)  The addition of Con A to FU-5-5 cells did not result in any decrease in the level of lactic dehydrogenase and not more than a 30% decrease in alanine aminotransferase as shown in Table II. Thus the effect seems lectin-specific, and probably enzyme-specific as well. Many of the effects caused by Con A seem to be temperature-dependent (26-28). For example, it has been shown that cells bind Con A at 0" without agglutination (7). However, when the temperature was raised to 22" after preincubating the cells with Con A at 0", the cells did agglutinate (29). We therefore undertook to study the temperature dependence of the Con A effect on tyrosine aminotransferase. When Con A was added at 5" to preinduced FU-5-5 cells, the process of inactivation was not blocked but was slowed. Thus, complete inactivation which took only 10 min at 37" took nearly 3 h at 5" (Fig. 4). To see whether the rapid decay of the enzyme would return at higher temperatures, the preinduced FU-5-5 cells were incubated with Con A at 5", washed, and then shifted to 37" in the absence of Con A. The result of such an experiment culture. These experiments suggest that for Con A to bring is shown in Fig. 5. As expected there was a slow fall of tyrosine about its full effect on the inactivation, the cells have to aminotransferase with Con A treatment at the low tempera-remain attached to the growth surface. However, the specific ture, then a rapid fall after the temperature increase. The binding of labeled Con A to either scraped or trypsinized FU-5enzyme activity decreased to about 50% of the level present at 5 cells was not significantly different from the binding obthe time of the temperature shift but did not reach as low a served in FUG-5 cells that were still attached (Table IV); thus level as when Con A was present at 37" continuously. The the binding of Con A to the cells alone is not sufficient for decreased enzyme level persisted no more than 15 min after bringing about inactivation in these cells, The addition of which it returned to the preinduced level within 1 h. This colchicine (100 PM) or cytochalasin B (10 pg/ml) simply did not experiment suggests that the continued presence of Con A is mimic the effect of Con A, nor did these compounds block the required for the maintenance of decreased levels of enzyme. effect of Con A when added together with it (data not shown), This happened even in the case of the cells treated with Con A suggesting that the effect of Con A does not involve microtuat 37" and then replaced in growth medium without Con A as bules and microfilaments. shown in Fig. 6. The temperature shift experiment also shows that binding of Con A and inactivation of tyrosine aminotransferase can be dissociated to some extent, and that the overall process is temperature-dependent.
The delayed effect of Con A on the inactivation of tyrosine aminotransferase at 5" could be due to a difference in the 280 1 I I I 1 binding of Con A to FU-5-5 cells at 5" and 37". This possibility 240 was investigated by studying the cell binding of radiolabeled Con A. The specific binding of [3H]acetyl-Con A to FU-5-5 cells 5 xa at the two temperatures is shown in Fig. 7. The binding 5 150 increases with time both at 5" and 37" for up to 60 min at least. Y However, the specific binding is considerably higher (3-to 4fold) at 37" than at 5". This difference in binding at these temperatures offers one possible explanation for the delayed effect of Con A at 5".
When Con A was added to a suspension of preinduced FU-5-:)~ I 2 '": m; do1 * I * 5 cells prepared either by scraping or trypsinization, rather o ' 1 ? 3 1 5 HOURS _ _ ^__. _ than to the undisturbed monolayer, there was little la11 in the activity of tyrosine aminotransferase (Table III). The data FIG. 6 (lefi). Reactivation of tyrosine aminotransferase specific activity in the absence of Con A in FU-5-5 cells treated with Con A at there also show that Con A did not nroduce anv inactivation of 37". Preinduced FU-5-5 cells were treated with 200 tie/ml of Con A at "Experimental Procedures." The difference in the amount of Con A bound with and without cu-methyl-n-mannopyranoside treatment subsequent to the binding with labeled Con A was taken as the _ specific binding of Con A. of induced FU-5-5 cells with Con A transferase. Con A (200 @g/ml) was also added to preinduced HTCat 5" followed by shift to 37" in the absence of Con A on the inactiva-Hl cells grown in suspension culture and the cells were collected 30 tion of tyrosine aminotransferase specific activity. FU-5-5 cells were preinduced overnight with 1 pM dexamethasone in 35-mm dishes.
min later and processed for enzyme assay. Control cells were treated with PBS alone. The medium was replaced by tricine-buffered cold Improved Minimal Essential Medium containing 1 pM dexamethasone and 200 @g/ ml of Con A. The dishes were maintained at 5" over ice for 30 min, then the medium was removed and the cells washed once with cold PBS. Regular growth medium at 37", containing 1 pM dexamethasone. was added to the dishes and the dishes were maintained at 37". Cells were harvested at different times during the course of various treatments and assayed for tyrosine aminotransferase as described previously.
The arrow indicates the time when cells were shifted to 37" without Con A. To be sure that the Con A-provoked inactivation of tyrosine aminotransferase was occurring intracellularly, and not during sonication of Con A-treated cells, a histochemical assay for enzyme activity by the method of Thompson and Tomkins (30) was applied to cells in situ. Fig. 8 shows that the Con Atreated cells possessed very little stainable enzyme compared to untreated cells. Thus, the effect of Con A in bringing about inactivation seems to be an in vivo effect and not an artifact caused during the handling of the cells subsequent to the treatment with Con A.

Condition of cells
Externally added cAMP or cGMP, their analogs, or theophylline did not mimic the effect of Con A in bringing about the inactivation of tyrosine aminotransferase in FU-5-5 cells, nor did adding these compounds with the lectin alter the usual inactivation (Table V). This suggests that the interaction of Con A with the cell does not inactivate the enzyme via changes in the intracellular level of either CAMP or cGMP. To test the possibility that the inactivation could be due to the release of some unknown diffusible inhibitor, an extract prepared from Con A-treated FU-5-5 cells was mixed with an extract of induced FU-5-5 cells. As can be seen from Table VI, this did not result in any inactivation.
A similar result was obtained when the cell extract from Con A-treated cells was prepared 2 min after the addition of Con A. Further, when cells treated with Con A for 2 min were mixed with untreated induced cells and then sonicated together immediately after mixing, enzyme activity in the mixed cells' extract was simply additive (data not shown). The effect of the addition of Con A to the cell extract prepared from induced FU-5-5 cells is shown in Table   TABLE IV Specific binding of L3Hlacetyl-Con A to FU-5-5 cells either in monolayer or in suspension at 5" and 37", respectively FU-5-5 cells either in monolayer or in suspension (obtained either by scraping or by trypsinization) were treated with labeled Con A at 5" and 37" for 10 min and the specific binding of Con A was estimated as described under "Exuerimental Procedures."   VII; the inactivation of tyrosine aminotransferase brought about by Con A is not due to the direct interaction of Con A with the enzyme.
Reversibility of Tyrosine Aminotransferase Inactivation Caused by Con A-It is known that Con A interacts with polysaccharides that contain a-n-glucopyranosyl, cY-n-mannopyranosyl, or cY-n-glucosaminyl residues (2). Ifthe inactivation of tyrosine aminotransferase brought about by adding Con A to preinduced FU-5-5 cells is also due to its binding to LX-Dglucopyranosyl or cY-n-mannopyranosyl residues on the cell surface, then this effect should be reversed by adding the competing sugar. In fact, this was found to be the case. When   after suspending the cells in 0.9 ml of sonication buffer. The cell extract was divided; to one portion increasing concentrations of Con A in NaCl-saturated PBS were added. After thorough mixing and incubation at 0" for 2 h the extracts were assayed for tyrosine aminotransferase.
The tyrosine aminotransferase in a second portion of the extract was partially purified by affinity chromatography (31) and then incubated with Con A and assaved. cY-methyl-n-marmopyranoside was added to cells treated with Con A, the enzyme returned to induced level as rapidly as it had been inactivated by the addition of Con A, as shown in Fig. 9. This rapid reactivation was independent of de nouo protein or RNA synthesis.
The rapid reactivation brought about by this sugar was a specific effect, since N-acetylglucosamine and N-acetylgalactosamine, sugars known to compete for different lectins, could not reverse the effect of Con A (Table VIII).
The inactivation of tyrosine aminotransferase by Con A and the reactivation of the inactivated enzyme by LYmethyl-D-mannopyranoside could be repeated a number of times, without any great loss of the induced level, as shown in Fig. 10. This result suggests that when this enzyme is inactivated by Con A, it is converted to a relatively stable inactive form and returns to its active form immediately upon removing the Con A bound to the cell. so as to bring the final concentration of cycloheximide to 100 pgiml and of actinomycin D to 1 pgiml, respectively, was added to three different sets of Con A-treated cells. Then 15 min after the addition of PBS or cycloheximide or actinomycin D, 0.2 ml of 01methyl-n-mannopyranoside solution in PBS was added to all the three sets of dishes, bringing the final concentration to 50 mM. Samples were removed at different times after the addition of 01methyl-n-mannopyranoside and assayed for tyrosine aminotransferase. by Con A and its reactivation by cu-methyl-n-mannopyranoside in induced FU-5-5 cells, FU-5-5 ceils preinduced overnight with 1 pM dexamethasone in 35mm dishes were treated with 200 pg/ ml of Con A for 15 min at 37". Then, 0.2 ml of an ol-methyl-nmannopyranoside solution in PBS was added to all the dishes, to a final concentration of 50 mM. The medium was replaced 15 min later by regular growth medium containing 200 pgiml of Con A and kept at 37" for 15 min. They were all then treated with ol-methyl-nmannopyranoside as before for 15 min. This cycle was repeated two more times. Samples were harvested for enzyme assay at the end of each treatment of either Con A or a-methyl-n-mannopyranoside. The solid arrows and dashed arrows represent the time of addition of Con A and ol-methyl-n-mannopyranoside, respectively.
for tyrosine aminotransferase can take place normally in the presence of Con A, uninduced FU-5-5 cells were treated with Con A in the presence of dexamethasone for 6 h, washed, and then a-methyl-n-mannopyranoside was added with and without dexamethasone. After the addition of the sugar there was a burst of enzyme activity, which peaked in 1 h, then declined and reached a steady level which was independent of the presence of dexamethasone (Fig. 11). However, if either actinomycin D (1 pg/ml) or cycloheximide (50 pg/ml) was present during the initial treatment of the uninduced cells with Con A in the presence of dexamethasone, and then the pyranoside was added after once washing the cells, there was no burst of tyrosine aminotransferase activity (data not shown). This experiment suggests that transcription and translation of the mRNA occurred during treatment of the cells with Con A, even though the enzyme synthesized was inactive. Addition of competing sugar allowed the preformed enzyme activity to be expressed. The maximum induced level attained after the addition of a-methyl-n-mannopyranoside was only about 50% of the usual level of tyrosine aminotransferase seen after 6-h induction with dexamethasone. This could be due to the impairment of protein synthesis by the exposure of the cells to Con A. Protein synthesis in cells treated with Con A for 4 h occurred with only 48% efficiency of that in normal cells (data not shown). Also possible is an effect of Con A on RNA synthesis, but this possibility has not as yet been checked.

DISCUSSION
Lectins and their binding to cell surfaces are known to influence many intracellular events. Con A, for instance, may have either a mitogenic or a cytotoxic effect, depending on cell type and concentration used (8,9,32 11. Induction of tyrosine aminotransferase specific activity in FU-5-5 cells after treatment with Con A. Fresh medium was added to uninduced FU-5-5 cells grown in 35-mm dishes containing 200 /*g/ ml of Con A and 1 +V dexamethasone. Samples were harvested for tyrosine aminotransferase assay as described previously at different times (up to 6 h). The medium was replaced in the rest of the dishes with fresh medium containing 50 rnM a-methyl-n-mannopyranoside along with (O-0) or without (O-0) 1 ELM dexamethasone. Samples were harvested for enzyme assay at different times after the addition of cu.methyl-n-mannopyranoside.
which these cell surface active compounds influence intracellular events, however, remains unknown. The experiments presented here show for the first time the effect on a soluble intracellular enzyme of the binding of Con A to the cell. In both FU-5-5 and HTC cells, practically all tyrosine aminotransferase activity except the small amount associated with mitochondria is found in the 100,000 x g supernatant fraction after cell lysis and differential centrifugation (33). Consequently, the enzyme is thought to exist in the soluble fraction. Recently, a similar inactivation of an enzyme by Con A was reported, but the enzyme affected was the membrane-associated ecto-5'-nucleotidase of glioma cells (341, and not one found in the soluble fraction. The rapidity (Fig. 2) of the inactivation of tyrosine aminotransferase by Con A suggests that this effect could be due either to the direct interaction of the enzyme with Con A, resulting in its inactivation or to the release of some signal subsequent to the binding of Con A, an event which in turn brings about the inactivation.
The former alternative has been ruled out because of the lack of inactivation by the direct addition of Con A (up to 1 mg/ml) to crude or partially purified extract prepared from preinduced cells (Table VII). The inactivation could be mimicked neither by CAMP nor by cGMP (Table Vl. These are the two most widely studied mediators for the intracellular actions of membrane active ligands. Hence, the effect reported here seems to be due to an alternate mechanism. It is known that Con A stimulates the uptake of Ca2+ in T lymphocytes (5); it is possible that the inactivation of tyro-Inactivation of Tyrosine Aminotransferase by Concanavalin A sine aminotransferase brought about by Con A binding could also be mediated by changes in the permeability for different ions. The experiments, in which Con A added to cells in Ca*+, W+, and glucose-free medium still caused rapid inactivation (data not shown), suggest that the inactivation is not due to altered uptake of these cations (or of glucose). So far all our attempts to demonstrate the inactivation by mixing experiments (Table VI) have been without success. Either the putative inhibitor is very highly labile, or some other mechanism is involved. Our prejudice is that some such inactivator exists, but that conditions necessary to demonstrate it in broken cell preparations have not yet been found. The rapid activation of tyrosine aminotransferase brought about by the addition of cl-methyl-n-mannopyranoside to the Con A-treated cells (Fig. 9) suggests that for the maintenance of the enzyme in an inactive form, it is necessary to have the Con A in a bound form. Moreover, the rapid reactivation of the enzyme further suggests that the putative inhibitor must be very labile. The inactivation of tyrosine aminotransferase brought about by Con A occurred much more slowly at 5" than at 37" (Figs. 4 and 2) a fact which could partially be accounted for by the lower specific binding of Con A to FU-5-5 cells at 5" compared to 37" (Fig. 7). The membrane is rigid at low temperatures (35); this prevents the agglutination of cells by Con A an event which is dependent upon the fluidity of the cell membrane. In contrast, the inactivation brought about by the binding of Con A to cell membrane was not an all or none phenomenon, for the addition of Con A to cells at 5" only prolonged the inactivation period. This may be due to the lack of a stringent requirement for membrane fluidity to bring about the release of the putative inhibitor for tyrosine aminotransferase once Con A binds to the cell.
The lack of inactivation of tyrosine aminotransferase when Con A was added to cells that were in suspension (Table III) demands the assumption of a correlation between the extended state of the cells when they are attached to a substratum and the release of the putative inhibitor for the enzyme since the binding of Con A was not impaired significantly when the lectin was added to a cell suspension rather than to a monolayer culture. The lack of inactivation in suspension cells (HTC-Hl) could possibly be due to a difference in the arrangement of microtubules and microfilaments in these cells as compared to cells remaining attached to a substratum. However, a mere perturbation of either the microtubules or microfilaments by the addition of colchicine or cytochalasin B was not sufficient to bring about inactivation in these cells. The observation that the reactivation of the enzyme in cells treated with Con A either at 5" or 37", after simply washing off excess Con A, was very slow compared to the reactivation brought about by the addition of a-methyl-n-mannopyranoside may be due to the slow or incomplete removal of membrane-bound Con A by simple washing. These results also imply that once the excess Con A is washed off from the medium, the Con A that was responsible for the inactivation is released slowly from its original site, resulting in the slow reactivation of the enzyme.
The addition of Con A to uninduced FU-5-5 cells in the presence of dexamethasone did not interfere with the induction of tyrosine aminotransferase (Fig. 11). However, the newly made enzyme was inactivated immediately after being synthesized since the addition of a-methyl-n-mannopyranoside to cells treated with Con A in the presence of dexamethasone resulted in bringing the newly made enzyme to an active form. This suggests that Con A did not simply act by interfer-ing with overall macromolecular synthesis, but brought about a relatively specific effect on the enzyme.
Although the results presented in this report did not enable us to arrive at the exact mechanism of inactivation of tyrosine aminotransferase by Con A, the following steps seem essential for the effect to be seen in FU-5-5 and HTC-Hl cells: (a) binding of native Con A to the cells; (b) cells growing in undisturbed monolayers.
In conclusion, we feel that the results reported here describe a novel phenomenon in the interactions of lectins with cells, the rapid, reversible inactivation of a soluble enzyme resulting from lectin binding. The rapidity, specificity, and ease of assay of this inactivation make it attractive as a model for studying cell surface-intracellular communication systems. We are currently studying this effect of Con A in hepatoma cells using both biochemical and genetic approaches.