Induction of hepatic tyrosine aminotransferase in vivo by derivatives of cyclic adenosine 3':5'-monophosphate.

A number of 8- and N6-SUBSTITUTED DERIVATIVES OF CYCLIC ADENOSINE 3':5'-MONOPHOSPHATE-DEPENDENT PROTEIN KINASE, AND AS SUBSTRATES FOR RAT LIVER CYCLIC NUCLEOTIDE PHOSPHODIESTERASE. All of the analogs tested were able to induce the transaminase. The induction by the analogs was shown to be the result of an actual increase in the amount of enzyme, and the mechanism of induction was an increase in the rate of synthesis of the transaminase. The induced enzyme appeared to be immunologically similar to the non-induced enzyme. A good correlation was found to exist between the dose that produced 50% of maximal induction and a combination of the activation constant for cyclic adenosine 3':5'-monophosphate-dependent protein kinase by the analog and its susceptibility to hydrolysis by cyclic nucleotide phosphodiesterase. These data suggest that the phosphorylation of some site is involved in the mechanism by which cyclic adenosine 3':5'-monophosphate affects the rate of synthesis of tyrosine aminotransferase.

in rat liver in vivo, as activators of rat liver cyclic adenosine 3':5'-monophosphate-dependent protein kinase, and as substrates for rat liver cyclic nucleotide phosphodiesterase.
All of the analogs tested were able to induce the transaminase.
The induction by the analogs was shown to be the result of an actual increase in the amount of enzyme, and the mechanism of induction was an increase in the rate of synthesis of the transaminase.
The induced enzyme appeared to be immunologically similar to the noninduced enzyme.
A good correlation was found to exist between the dose that produced 50% of maximal induction and a combination of the activation constant for cyclic adenosine 3': 5'-monophosphate-dependent protein kinase by the analog and its susceptibility to hydrolysis by cyclic nucleotide phosphodiesterase. These data suggest that the phosphorylation of some site is involved in the mechanism by which cyclic adenosine 3': 5'-monophosphate affects the rate of synthesis of tyrosine aminotransferase.
Many of them are stable to enzymatic hydrolysis and at the same time are significantly more active than cAMI" as activators of cAMI'-dependent protein kinase. In addition, they do not require metabolic transformation for activity, as does Ilt,cA1lP (7)(8)(9). These new analogs of cAMP contain a wide range of chemical modifications and vary over a 2-order of magnitude range in their ability to activate cA1IPdependent protein kinase (l-6). This presents a situation in which a possible correlation between two different biological parameters can be tested.
One attractive hypothesis is that the phosphorylation of some site results in an increased rate of tyrosine transaminase synthesis.
It is conceivable that a cAMPstimulated protein kinase could mediate this phosphorylation.
In this paper we compare the ability of a number of 6-and W-substituted derivatives of cAMP to induce tyrosine transaminase in adrenalectomized rats and to activate rat liver CAMPdependent protein kinase.
EXPERIMENTAL PROCEDURES CAMP Derivatives-B&CAMP, N"Bt-CAMP, and 02'Bt-CAMP were from Sigma Chemical Co. The 6-and %substituted derivatives of CAMP were synthesized as previously described (l-6).
Enzyme Induction-Male Sprague-Dawley rats (100 to 150 g) were maintained in a 12-hour light-dark cycle on laboratory chow and 0.9% NaCl solution ad libitum for 72 hours after bilateral adrenalectomy under ether anesthesia.
All experiments were begun at the end of a dark cycle. The animals were fasted for 18 hours prior to the administration of the CAMP analogs intraperitoneally in sterile 0.9% NaCl solution.
The animals were killed by decapitation at appropriate times after dosing. Control animals received 0.9% NaCl solution only. The livers were excised, rinsed in ice-cold Buffer A (20 mM Tris-HCl (pH 7.5)-150 mM KCl-10 mM MgClz-5 mM 2-mercaptoethanol), homogenized in 3 volumes of Buffer A by using three strokes of a motor-driven glass-Teflon homogenizer, and centrifuged at 27,000 X 9 for 1 hour; the resulting supernatant fraction was assayed for tyrosine transaminase activity as described by Lin et al. (13). Protein was determined by the method of Lowry et al. (14). Protein concentrations that afforded linear initial rates were determined from pilot experiments. Basal transaminase activity ranged from 8 to 11 milliunits per mg, where one unit is the amount of enzyme that produces 1 pmol of product per min at 37". Immunochemical, Analyses-Antiserum prepared in rabbits to hydrocortisone-induced rat liver tyrosine transaminase was generously provided by Dr. Wesley D. Wicks. In the experiments with unlabeled transaminase, increasing amounts of the supernatant fraction were added to a constant amount of antiserum; the mixtures were incubated for 1 hour at 35" and then for 24 hours at 4", and then assayed for remaining transaminase activity.
In the experiments in which tyrosine transaminase was labeled, the animals were given 0.1 mCi of ['%]leucine (240 Ci per mole) intraperitoneally 20 min before they were killed.
Carrier hydrocortisone-Do 12345 0 427 for measuring the rate of transaminase synthesis were the same as those used bv Wicks et al. (10).
Preparation aid Assay of E&&-Rat liver CAMP-dependent protein kinase, assayed as described previously for the calf brain enzyme (9) were designed to minimize differences in absorption and metabolism. The general approach was as follows: by using an arbitrary dose, the time course of transaminase induction was followed to determine the time at which the largest-fold increase in enzyme activity occurred.
A dose-  N%t-CAMP was also slightly more potent than Ilt,cAMl' (Fig. lB), whereas 02'Bt-CAMP was much less active. These results indicate that N6Bt-CAMI' is the active metabolite of BtscAMl', a finding which is consistent with previous data (7)(8)(9).
&Substituted Derivatives of cALlIP-Similar studies performed on &substituted cA;\Il' derivat,ives arc summarized in Fig. 1, C to I'. With the exception of SH&cARIl', all of these derivatives were able to cause as great, an increase in transaminase activity as did l'&cAIJII' (Fig. 1, C and E). The lower inducing activity of 8H&cAMl' may be related to its rapid rate of hydrolysis by phosphodiesterase (Table II). Anot,her enzyme, tyrosine hydroxylase, has been shown to be similarly induced by 13tzcAMl-' and 8MeScAMl' in cultured neuroblastoma cells (17). The three &substituted cAh!Il' derivatives containing aromatic  8pCll'hS-, and 8PhCHzHN-cARlI') elicited a more rapid response than did the other 8-substituted CAMP derivatives.
With all of these 8-substituted derivatives and many of the NQubstituted cAhI1' derivatives and the 6-substituted cRhI1' derivatives, the transaminase activity returned very rapidly to basal levels, in some cases in as little as 1.5 to 2 hours. Although the data presented here did not allow an accurate determination of the half-life of the transaminase after induction by the analogs, previous studies in tivo suggested that induction of the enzyme by l%cAhIl' had no significant effect on the rate of enzyme degradation (10). The dose-response curves (Fig. 1, D and F) show clearly that many of the 8-substituted derivatives were significantly more potent than 13tzcA111'. 8pCll'hS-cAMI', for example, was 40 times more active than litzcAi\Il' in terms of Indso values.* In general, the arylthio-substituted derivatives (8pClI'hS-and 8l'hCH&cARll') were the most active, followed by 813r-, 8HS-, 8&S-, and 8HO-rAMI with intermediate activity; the 8RHN derivatives and 8hleO-cARlI' were the least active of the group. It was found that 8MeO-, 8MeHN-, and 81'hCHzHN-cAh21' were toxic to the animals at doses greater than 200 mg per kg. Therefore, the time course studies with these analogs were done at a dose of 200 mg per kg.
NQYubstiluted Derivatives of cAA4P-In this class of compounds, only N6EtzcAMP and N%tOCO-cAMI' were able to produce the magnitude of response produced by the 8-substituted derivatives and lltzcAhIl' (Fig. 1G). Although all of the NQubstituted derivatives were hJ.drolyzed by phosphodiesterase, the latter two compounds were hydrolyzed at a relatively slower rate than the others (Table II).
N"Et-cAhI1' and NGI'hH2C-cAhlI' were intermediate in both their ability to induce the transaminasc and to be hydrolyzed by phosphodiesterasc. N%tO-cAxI1' and NGI'hCH20-CAMP were the poorest inducers and the most rapidly hydrolyzed compounds of this group.
The rapid onset of induction (Fig. 1H) demonstrated by N6PhH2C-and N61'hCHzO-cAMI' sustains the suggestion that cyclic nucleotide derivatives containing aromatic substitutions arc more rapidly absorbed.
Other derivatives in this group required somewhat longer to elicit maximal induction of the transaminase.
The dose-response studies on these derivatives (Fig. lli) show that they all demonstrated similar IndSO values, with only approximately a 4.fold difference between the most and least active.
NVhH,C-cAhI1' proved toxic at doses greater than 200 mg per kg; therefore, the time course study was performed at this dose.
B-Substituted Derivatives of cRJfP-Kane of the compounds 2 IndsO is the dose that produces 50% of the maximally obtainable induction.
in this class elicited more than a 1.5. to a-fold increase in tyrosine transaminase activity, and GHS-cRM1' produced less than a 0.5.fold increase (Fig. 11). In part, the reason for this low activity may be the rapid rate at which this class of compounds was hydrolyzed by phosphodiesterase (Table II), GHS-cRM1' being hydrolyzed even more rapidly than was CAMP.
As with the above groups of derivatives, the compound containing an aromatic 6-substituent, 61'hCH&cRM1~, demonstrated a more rapid induction of the transaminase than did the other R-substituted cRMP derivatives tested.
Studies on the dose-response by these derivatives (Fig. 1J) showed that 6MeS-and GPhCH& cRMP were significantly more potent than the other 6-substituted cRM1' derivatives.
In addition, 6HS-, 6HO-, and 6MeO-cRM1' demonstrated IndsO values similar to those of the N6substituted cAM1' derivatives discussed above. GHO-cRMP has been shown to be an inducer of tyrosine transaminase in fetal rat liver in utero (18).

Analysis of Induced Enzyme
A series of experiments \vas next performed to determine whether the analogs were actually causing an increase in the amount of enzyme and, if so, whether this was the result of an increased rate of synthesis of tyrosine transaminase.
Characteriaztion of Induced Enzyme by Immunotitration-In the experiments summarized in Fig. 2, increasing amounts of supernatant fraction from induced or noninduced livers were added to a constant amount of antiserum and the remaining transaminase activity was determined after immunoprecipitation. In this way an equivalence point was determined for the induced and non-induced enzymes.
In each case the relationship Adrenalectomieed rats received various CAMP analogs at their respective optimal doses and were killed at the time of optimal response for e&h analog. These conditions were as follows: 8Br-CAMP. 5 mn ner ke and 3 hours: 8PhCH2S-CAMP, 3 mg per kg and 2 hburs;Yi6EtZzMP, 100 mg'per kg aid 2.5 hours; and GMeS-cRMP, 50 mg per kg and 2 hours. For each analog the crude liver extracts from five identically treated animals were pooled. The increases in the specific activity of the transaminaie in the pooled extracts from thk treated animals over those of the controls were: 8Br-CAMP. 3.7-fold: 8PhCH&CAMP.  between the catalytic activity and the antigenicity of tyrosine transaminase was not changed as a result of the induction of the enzyme by these CAMP derivatives.
The results, therefore, show that the increase in catalytic activity of the transaminase actually represents an increase in the amount of enzyme, and not just activation of preexisting enzyme.
Isotopic-Zmmunochemical Analysis of Rate of Tyrosine Transaminase Synthesis in Response to CAMP Derivatives-In these experiments the labeling of tyrosine transaminase was compared with that of the other soluble proteins after a pulse of [%]leutine given 20 min before killing.
With all four CAMP derivatives tested , the incorporation of [Wlleucine into both soluble protein and the transaminase was increased, but the increase for the transaminase was approximately twice that for the soluble protein (Table I). In addition, there was a comparable increase in the specific activity of tyrosine transaminase (2.1. to 3-fold).
For two of the analogs (8pCll'hS-and 8MeS CAMP), the pulse labeling and immunoprecipitation were performed at a time after dosing when the specific activity of the transaminasc had not yet shown a significant increase (0.5 hour for 8pClPhS and 1 hour for 8MeScAMP). In these cases, the preferential incorporation of [r4C]leucine into tyrosine transaminase over the other soluble proteins was also observed. The results show that these analogs were able to effect an increase in the rate of synthesis of tyrosine transaminase. This 429 is consistent with the data of Wicks et al. (10) on &CAMP. Although this mechanism of action of CAMP derivatives in the induction of the transaminase has been rigorously demonstrated for only the few derivatives listed above, it is probably a valid assumption that all of these analogs are inducing the enzyme by a similar mechanism.
Wicks' group has found that some of the CAMP derivatives studied here were able to induce tyrosine transaminase by increasing its rate of synthesis in cultured Reuber H35 hepatoma cells (19).

Studies with Rat Liver Enzymes in Vitro
Two enzymes in liver with which these CAMP derivatives might interact are CAMP-dependent protein kinase and CAMP phosphodiesterase.
The first is important as a possible intermediary in the mechanism by which CAMP derivatives induce tyrosine transaminase, and the second is important because it can inactivate the analogs.
Rat Liver CAMP-dependent Protein KirLase-The activation constants (K, values) were determined for each of the CAMP derivatives with CAMP-dependent protein kinase from rat liver. The results (Table II) indicate that the 8-substit.uted CAMP derivatives were widely divergent in their ability to activate this enzyme. The most active of this group was 8pClPhScAMP, which was nearly 100 times more effective than CAMP.
The K, values for each of the CAMP derivatives with rat liver cAMPdependent protein kinase were of the same relative magnitude as the K, values previously reported for these analogs with other CAMP-dependent kinases (l-6, 20-23).3 Rat Liver Phosphodiesterase-A comparison of the K, values with the corresponding Indso values (Table II) indicated that, although some correlation appeared to exist between these two groups of data, many of the compounds demonstrated higher IndsO values than would be predicted on the basis of their K, values. One possible reason for this could be that some of the compounds were being hydrolyzed by phosphodiesterase.
This was indeed the case, as seen by the results in Table II. Of the 8-substituted derivatives, only 8H&cAMP was hydrolyzed at a very rapid rate, compared with that of CAMP.
All of the NQubstituted CAMP derivatives and g-substituted cRMP derivatives were hydrolyzed at significant rates. GHS-cRM1' was an even better substrate than was CAMP for the rat liver phosphodiesterase preparation used. In general, these results indicate that many of the analogs which demonstrated higher Indso values than expected were hydrolyzed by the phosphodiesterase.
The relative rates of hydrolysis of the analogs were in the same relative ranges as those reported previously for enzyme preparations from other sources (l-6, 20-23).

Analysis of Correlation of Zndso with K, and Phosphodiesterase Hydrolysis Values
The lndsO values of the cyclic nucleotides were examined for possible correlation with their K, values and with both K, and phosphodiesterase hydrolysis values (expressed as S, the ratio of the rate of hydrolysis of the analog to the rate of hydrolysis of CAMP) by regression analysis.
Although many other factors would be expected to influence the lndso of the cyclic nucleot,ides in tivo, S and K, were chosen to construct a mathematical model of induction of the transaminase because of their accessibility and their reflection of two fundamental events in which CAMP participates. Indso refers to the dose of compound that produces 50% of maximal tyrosine transaminase induction.
The observed values were determined by inspection of Fig. 1 The cyclic nucleotides were examined in three groups, the &substituted compounds and the 6-substituted compounds, and as a whole.
Data in Table 11 were used in the regression analysis.
The 8MeHN-, 81'hCHzS-, and NGEt-CAMP values were eliminated from the fit because their values were anomalous. Equations 1 to 6 give the results of the regression analysis. The values given in parenthesis are t,he 95% confidence intervals; n is the number of points, sd is the standard deviation, and r is the correlation coefficient. Table II gives the values of Indso calculated from the regression Equation 6. Indso was expressed in milligrams per kilogram and K, in nanomoles per liter.  Examination of the equations generated by multiple regression analysis of the data in Table II indicated some obvious trends.
Considering either the S-or B-substituted derivatives separately or all analogs together, a better fit was obtained in each case by incorporation of the phosphodiesterase hydrolysis term S. The regression coefficient of S was not as significant in the two smaller groups of compounds (Equations 2 and 4) as it was in the analysis of all the compounds (Equation 6). This was perhaps because the values of S are fairly similar within each group, but the average S value of the 6-substituted compounds was quite a bit greater than that of the 8-substituted compounds.
A qualitative overview of the information derived from regression analysis of the data w-ould strongly suggest, but by no means prove, a mechanism of tyrosine transaminase induction by the cyclic nucleotide analogs which involves activation of a protein kinase (hence subsequent phosphorylation of a protein).
The significance of the coefficient of log(S) in Equation 6 suggests that the better substrates for phosphodiesterase (with a great,er negative log(S)) require higher concentrations for transaminase induction than would be estimated from examination of the K, values alone. A fair rank correlation (r = 0.762) has been noted between tyrosine transaminase induction in H35 cells and K, values for bovine brain CAMPdependent protein kinase for a series of 8-substituted CAMP derivatives (1,3,24).
Acknowledgments-We wish to thank Dr. Wesley Wicks for supplying us with the antityrosine transaminase antibody and the results of his group on the induction of the transaminase in H35 cells in response to CAMP derivatives, Dr. Paul Greengard for his unpublished results on activation of bovine heart cAMPdependent protein kinase by some of the cAMI' derivatives studied here, and Mary K. Dimmitt for her able technical assist-8-Substituted-CAMP derivatives (Compounds 3, 5 to 9, 11, ante in the early stages of this investigation.