Cyclic AMP regulation of lactate dehydrogenase. Quantitation of lactate dehydrogenase M-subunit messenger RNA in isoproterenol-and N6,O2'-dibutyryl cyclic AMP-stimulated rat C6 glioma cells by hybridization analysis using a cloned cDNA probe.

We have cloned DNA complementary to mRNA coding for rat C6 glioma cell lactate dehydrogenase M-subunit. Double-stranded DNA complementary to a portion of lactate dehydrogenase mRNA was inserted into the Pst I site of plasmid pBR322 by the dC.dG tailing technique and amplified in Escherichia coli HB101. A recombinant plasmid containing lactate dehydrogenase cDNA was identified by colony hybridization to a cDNA prepared from partially purified lactate dehydrogenase mRNA and by hybridization-selected translation. The recombinant plasmid (pRLD42) contains a 680 nucleotide insert of lactate dehydrogenase mRNA. Hybridization of nick-translation pRLD42 to glioma cell poly(A)+RNA separated on agarose gel and transferred to nitrocellulose exhibited Mr = 5.9 X 10(5) for lactate dehydrogenase mRNA. Furthermore, Northern blot analysis of RNA from unstimulated and isoproterenol-stimulated glioma cells indicated a 2-fold increase of lactate dehydrogenase mRNA molecules in stimulated cells. The 2-fold increase of lactate dehydrogenase mRNA was confirmed by RNA-excess kinetic hybridization using pRLD42 DNA and poly(A)+RNA from unstimulated, isoproterenol-, and dibutyryl cAMP-stimulated glioma cells. These data demonstrate that isoproterenol and dibutyryl cAMP cause an increase of the number of lactate dehydrogenase M-subunit mRNA molecules in glioma cells which, in part, determines the extent of synthesis of the lactate dehydrogenase M-subunit.

1 TO whom correspondence should be addressed.
The abbreviations used are: CAMP, adenosine 3':5"monophosphate; dibutyryl CAMP, N6,O2'-dibutyryl adenosine 3':5'-monophosphate; SDS, sodium dodecyl sulfate; NaCl/P,, phosphate-buffered saline; Poly(A)'RNA, RNA containing a polyadenylic acid sequence; dscDNA, double-stranded complementary DNA, kbp, kilobase pair; bp, base pair. complexity of eukaryotic systems has, to date, prevented the elucidation of the mechanism underlying cAMP control of eukaryotic protein synthesis. There is evidence that cAMP modulates the de novo synthesis of specific proteins rather than the regulation of the abundance of proteins by modification of protein degradation (1,2,(4)(5)(6)(7)(8)(9)(10)(11). Recent results have suggested that cAMP modulation of the synthesis of several eukaryotic proteins may be due to an accumulation of specific mRNA (9-13). Indirect quantitation through in vitro translation has shown CAMP-mediated increases in the functional levels of mRNA for tyrosine aminotransferase (12, 13), phosphoenolpyruvate carboxykinase (lo), and lactate dehydrogenase "subunit (11). Furthermore, Brown and Papaconstantinou (9), using a cDNA probe for mouse albumin, demonstrated that dibutyryl cAMP is capable of increasing the number of molecules of albumin mRNA in the Hepa-2 mouse hepatoma cell line. While these findings indicate a regulation by cAMP of the levels of mRNA of the proteins studied, there is no evidence which would allow a determination whether cAMP acts through regulation of transcriptional and/or posttranscriptional events. Several investigators (1, 14-16), using inhibitors of protein and RNA synthesis, have come to the conclusion that cAMP acts at a post-transcriptional level, conceivably through stimulation of protein chain elongation (17). While it is possible that cAMP may regulate different enzymes in a specific way and may have diverse points of action which include transcriptional and/or post-transcriptional regulatory events, additional studies are required to clarify these questions.
Our laboratory has recently begun a systematic investigation of the CAMP-mediated induction of lactate dehydrogenase in the rat C6 glioma cell line. This model offers the advantages that one can study the regulation of synthesis of a well characterized protein under conditions that can be rigorously controlled. In addition, lactate dehydrogenase represents a substantial portion of the total protein and mRNA activity in rat C6 glioma cells, thus facilitating quantitative studies and purification of lactate dehydrogenase and its mRNA. We have previously demonstrated that the isoproterenol-or dibutyryl CAMP-mediated induction of lactate dehydrogenase activity in the rat C6 glioma cells is preceded by specific activation of cytoplasmic and nuclear CAMP-dependent protein kinase (18), phosphorylative modification of several histones (19), and involves an increased de nouo synthesis of the lactate dehydrogenase M-type subunit and increased functional levels of lactate dehydrogenase mRNA (11). In order to more clearly define the mechanism of regulation of lactate dehydrogenase gene expression, we have developed a hybridization probe for lactate dehydrogenase mRNA. We report here our use of a partially purified lactate dehydrogen-ase mRNA to construct and select a cloned recombinant plasmid DNA containing a portion of the lactate dehydrogenase mRNA sequence. Furthermore, we have utilized this plasmid for the quantitation of lactate dehydrogenase mRNA molecules from control and isoproterenol-or dibutyryl CAMPstimulated glioma cells.

RESULTS
Partial Purification of Lactate Dehydrogenase mRNA-In our previous publication, we have indicated that lactate dehydrogenase mRNA could be enriched 8-fold by isoproterenol stimulation of confluent C6 glioma cells (11). The stimulation produced a mRNA population in which 0.8% of the sequences coded for the lactate dehydrogenase M-type subunit as estimated by in vitro translation. We attempted to enrich the lactate dehydrogenase mRNA by size fractionation of poly(A)'RNA on sucrose gradients. Total mRNA activity and lactate dehydrogenase mRNA activity were estimated either after trichloroacetic acid precipitation or after immunoprecipitation of in vitro translation products using RNA from selected gradient fractions as templates. Fig. 1 shows the results of a representative sucrose gradient fractionation. Whereas total mRNA activity increases as the size range increases, lactate dehydrogenase mRNA activity displays a sharp peak migrating slightly slower than the 18 S ribosomal RNA peak. Judging by the position of molecular weight markers run on a parallel gradient (rabbit globin mRNA, 18 S and 28 S ribosomal RNA), the size of lactate dehydrogenase mRNA is estimated a t 6.0 X 10" daltons. In the peak fractions (fractions (12-14), lactate dehydrogenase accounted for 4% of the total mRNA activity, representing an enrichment of only Since size distribution of lactate dehydrogenase mRNA relative to total poly(A)'RNA was not sufficiently distinct to allow for a meaningful purification by size fractionation, we utilized a modification of the polyribosome immunoprecipitation procedure (44) for the isolation of lactate dehydrogenase mRNA. We observed that it was very important to saturate the polysomes with primary antibody and to separate the polysome-antibody complex from unbound antibody prior to reaction with the GAR-cellulose in order to maximize yield and decrease nonspecific precipitation (results not shown). The enrichment procedure was monitored by using the reticulocyte in vitro translation system. Starting with 300 A260 units of polyribosomes, we recovered 0.22 mg of polysomal RNA (Table I). This procedure resulted in an approximately 10-fold enrichment of the lactate dehydrogenase mRNA activity. As shown by translation of this partially purified mRNA in the reticulocyte lysate system and analysis of translated protein by SDS-polyacrylamide gel electrophoresis before and after immunoprecipitation of lactate dehydrogenase (Fig. 2), lactate dehydrogenase appears to be the predominant mRNA species, although there is a considerable number of contaminating mRNA species.
Synthesis a n d Molecular Cloning of Double-stranded cDNA-The lactate dehydrogenase-enriched polysomal RNA was selected for poly (A) Electrophoretic analysis of lactate dehydrogenase synthesized in a reticulocyte lysate system from various RNA templates. Rat C6 glioma cells were stimulated with M isoproterenol for 8 h. RNA was isolated from glioma cell polysomes either after immunoprecipitation of polysomes or directly without prior immunoprecipitation. RNA was subsequently translated in the reticulocyte lysate system. The in uitro-synthesized [:"'S]methionine-labeled proteins were isolated either by trichloroacetic acid precipitation (A) or by immunoprecipitation ( B ) and subjected to SDS-polyacrylamide gel electrophoresis and autoradiography. Lane 1, translation products of the endogenous reticulocyte lysate R N A lane 2, translation products from RNA isolated from total cellular polysomes; lane 3, translation products from RNA isolated from immunoprecipitated polysomes. LDH, lactate dehydrogenase. tog-raphy on oligo(dT)-cellulose prior to oligo(dT)-primed cDNA synthesis. Both yield and size of cDNA were maximized by using saturating levels of reverse transcriptase. Generally, we obtained efficiencies of 15-20s (cDNA synthesized/input RNA) and a modal size of the DNA product of approximately 1500 nucleotides (data not shown). Second strand synthesis by Escherichia coli DNA polymerase I utilized the self-priming capability of the f i t strand (26) and proceeded to 50-75% completion (percentage of fwst strand). Following S, nuclease trimming, the double-stranded cDNA was fractionated on agarose gels after which two size ranges of molecules were recovered. Starting with 2.2 pg of poly(A)'RNA, we recovered 4.4 ng of a 1-to 2.5-kbp dscDNA and 10.6 ng of a 0.5-to 1.0kbp dscDNA. The dscDNA of both sue ranges was then dCtailed and annealed with pBR322 that had been dG-tailed at the Pst I site. Transformation of E. coli HBlOl resulted in 119 Tc' colonies from the 1-to 2.5-kbp material and 331 colonies from the 0.5-to 1.0-kbp cDNA. Since we were interested in generating a cDNA probe of maximal size, only clones with the 1-to 2.5-kpb insert were screened further.
Screening of Clones for Lactate Dehydrogenase cDNA Sequences-Replicate plating of Tc' 1-to 2.5-kbp clones on ampicillin-containing agar showed that 104/119 colonies were Tc'/Amp". These clones were analyzed by colony hybridization using [""PIcDNA synthesized from a mRNA population partially enriched for lactate dehydrogenase mRNA by immunoprecipitation of polyribosomes. The mRNA preparation used to generate this screening cDNA probe was, however, isolated from batches of cells different from those used to synthesize the cloned cDNA in an effort to decrease the crossreactivity with non-lactate dehydrogenase-containing colonies. This preliminary screening identified 11 Tc'/Amp" colonies as potentially containing lactate dehydrogenase DNA sequences.
In order to additionally increase the specificity of the filter hybridization analysis, the ['"PIcDNA probe used for subsequent colony hybridization was enriched for a predominant cDNA species (45). Since size analysis of the cDNA itself did not yield any distinct species, the cDNA was digested with Hae 111 restriction endonuclease and separated by alkaline agarose gel electrophoresis (46). Autoradiography of the gel identified a distinct band a t 500 nucleotides, superimposed on the heterogeneous background of the cDNA (data not shown). The 500-nucleotide cDNA band was recovered from the gel and used in a filter hybridization analysis of the 11 Tc'/Amp" colonies identified by preliminary screening. Using the Hae 111-restricted ['"PIcDNA probe, four clones (Nos. 2, 42, 55, and 97) were identified as containing a predominant cDNA species, presumably lactate dehydrogenase which was the most abundant mRNA present in the RNA preparation used for cDNA synthesis.
The four clones (Nos. 2, 42, 55, and 97) were subsequently analyzed for the size of the inserted DNA fragment. Plasmid DNA was partially purified, digested with Pst I, and analyzed on 1.5% neutral agarose gels. Analysis of the ethidium bromide staining pattern revealed that clones 2, 42, 55, and 97 contained a DNA insert consisting of a 480-bp fragment (see Fig.  3) and a 200-bp fragment (not visible on this gel), thus indicating the presence of an internal Pst I site. It appears that the four clones contain plasmids with identical inserts, resulting from a single initial transformation event.
The final screening procedure used for positive identification of lactate dehydrogenase cDNA clones was hybridizationselected translation. Plasmid DNA was purified from clones 2 and 42 and immobilized on DBM paper. After hybridization of poly(A)'RNA from isoproterenol-stimulated C6 glioma cells to the immobilized DNA, the bound RNA was eluted and translated in the reticulocyte lysate cell-free translation system. Translation products were analyzed by SDS-polyacrylamide gel electrophoresis. As shown by the autoradiograph in Fig. 4, RNA eluted from filters containing pBR322 DNA (lane 3) did not direct the synthesis of any proteins other than the endogenous products of the reticulocyte lysate (lane 4). Filters containing plasmid DNA from clones 2 and 42 (lanes 1 and 2), however, retained a RNA species which coded for a 35,000-dalton protein (arrow) co-migrating with authentic rat liver lactate dehydrogenase (not shown). The identity of this protein was further established by immunoprecipitation with rabbit anti-rat lactate dehydrogenase-5 isozyme antiserum. The results (Table 11) indicate that clones 2 and 42 hybridized to a mRNA coding for a protein having the antigenic characteristics of rat lactate dehydrogenase M-type subunit. We conclude, therefore, that these recombinant plasmids contain a rat lactate dehydrogenase "type subunit DNA sequence.
Restriction Endonuclease Analysis of Cloned Lactate Dehydrogenase cDNA-Purified plasmid DNA from clone 42 (pRLD42) was further analyzed by digestion with restriction endonucleases as summarized in Fig. 5. The cloned lactate dehydrogenase cDNA appears to be 690 bp in length (assuming 30 bp of G-C tails a t each end). Endonucleases that did not cut the lactate dehydrogenase cDNA include Eco RI, HindIII, Bam HI, Sam I, Sal I, and Hha I.
Northern Blot Analysis of Lactate Dehydrogenase mRNA-Poly(A)'RNA isolated from either unstimulated C6 glioma cells or cells stimulated for various periods with isoproterenol ( M) was denatured with glyoxal and separated by electrophoresis on a 1.5% agarose gel. Following transfer of the RNA to nitrocellulose, lactate dehydrogenase mRNA was detected by hybridization to nick-translated [:"P]pRLD42. The autoradiograph of the Northern blot (Fig. 6)  Analysis of putative lactate dehydrogenase recombinant clones after Pst I digestion and agarose gel electrophoresis. Plasmid DNA was isolated from clones 2. 42, 55, and 97 and digested with Pst I, and the fragments were analyzed by electrophoretic separation on a neutral 1.5% agarose gel. DNA bands were visualized by ethidium bromide staining. For further experimental details see "Experimental Procedures." All four clones display a "pBH322 DNA" band at 4.3 kbp and "inserted DNA" bands at 480 and 200 bp (indicated by arrows). Molecular weight markers (on each side of the gel) are X DNA digested with HindIII and @X 174 DNA digested with Hue 111.
strates that pRLD42 hybridizes to a RNA species of 5.9 X IO5 daltons present in all of the poly(A)'RNA fractions. This value corresponds to approximately 1800 nucleotides, a value which agrees well with our estimate of lactate dehydrogenase mRNA molecular weight obtained by sucrose density gradient centrifugation (Fig. 1). Furthermore, the Northern blot allows a relative quantitation of the amount of lactate dehydrogenase mRNA in the various poly(A)'RNA fractions (49). The autoradiograph (Fig. 6) shows an increased intensity of hybridization of the band corresponding to lactate dehydrogenase mRNA over the time course after isoproterenol-stimulation, although visual inspection appears to indicate that RNA extracted from cells incubated with isoproterenol for 2 h hybridizes more strongly than RNA from cells stimulated for 4 h. However, excision of the bands and quantitation of :v2P radioactivity by liquid scintillation counting (Fig. 7) shows the increase of hybridizable counts in RNA samples from isoproterenol-stimulated C6 glioma cells as a function of stimulation time. The discrepancy between visual inspection of the autoradiograph (Fig. 6) and the actual "' P counts in the hybrids is not clear but the problem may be due to the geometry of the Confluent C6 glioma cells were stimulated with lo-" M isoproterenol for the time periods indicated and poly(A)' RNA was isolated. The poly(A)'RNA was denatured with glyoxal and subjected to electrophoresis on a neutral 1.5% agarose gel. Following electrophoresis, RNA was transferred to nitrocellulose. Lactate dehydrogenase mRNA was detected by hybridization with nick-translated '"P-labeled pRLD42 (clone 42) followed by autoradiography. The numbers shown above each fane refer to the hours of glioma cell stimulation with isoproterenol; C, unstimulated glioma cell RNA. The molecular weight of the LDH mRNA was estimated using HindIII-digested X DNA and rat ribosomal RNA as standards. The position of 18 S and 28 S RNA is indicated.
= I E n HOURS AFTER ISOPROTERENOL   FIG. 7. Quantitation of lactate dehydrogenase mRNA after Northern blot analysis. Glioma cell poly(A)'RNA was separated by agarose gel electrophoresis and subjected to Northern blot analysis as described in Fig. 6. The bands corresponding to lactate dehydrogenase mRNA were excised from the nitrocellulose sheet, and the hybridized '"P radioactivity was quantitated by liquid scintillation counting. The amount of hybridized radioactivity is displayed as percentage of unstimulated control.
in the plateau hybridization values (see Fig. 8). This suggests a partial difference in the nucleotide sequence of the lactate dehydrogenase mRNA present in control uersus isoproterenol-or dibutyryl CAMP-stimulated C6 glioma cells.  I digestion of pRLD42). Hybridizations were carried out for various time periods as described under "Experimental Procedures." The data are presented after adjustment of the plateau hybridization level to 100% and subtraction of background hybridization (12%). The actual plateau hybridization value was 32% for isoproterenol-and dibutyryl CAMP-stimulated RNA, and 26% for RNA from unstimulated cells. Computer-derived ROt1,2 values are: RNA from isoproterenol-stimulated cells, 1.63 mol X min X liter"; RNA from dibutyryl CAMP-stimulated cells, 1.54 mol X min X liter"; RNA from unstimulated cells, 3.06 mol X min X liter".

DISCUSSION
Although cAMP or effector agents that generate cAMP are known to induce a number of proteins in eukaryotes (1, 2,4ll), the molecular mechanisms of induction are virtually undisclosed. Recently, our laboratory has investigated the catecholamine-mediated induction of lactate dehydrogenase in the rat C6 glioma cell line as a model system to elucidate the role of cAMP in eukaryotic enzyme induction. In the present study, we have generated a specific cDNA probe complementary to lactate dehydrogenase mRNA through the technique of molecular cloning and have utilized this probe to quantitate lactate dehydrogenase mRNA levels in rat C6 glioma cells stimulated with either isoproterenol or dibutyryl CAMP.
Our strategy consisted of purification of lactate dehydrogenase mRNA, generation of dscDNA, and amplification of this cDNA through molecular cloning. Partial purification of lactate dehydrogenase mRNA was achieved through immunoprecipitation of polyribosomes isolated from isoproterenolstimulated C6 glioma cells, resulting in a mRNA preparation in which lactate dehydrogenase mRNA represented 20% of the total mRNA activity as evaluated by in vitro translation analysis. The relative efficiency of this purification procedure is demonstrated by the facts that it resulted in a 10-fold enrichment of lactate dehydrogenase mRNA as compared to mRNA which was not isolated by polysome immunoprecipitation and in a 200-fold enrichment as compared to the lactate dehydrogenase mRNA activity in poly(A)'RNA preparations from unstimulated glioma cells.
The technique of molecular cloning was utilized to amplify and purify cDNA prepared from this partially purified lactate dehydrogenase mRNA. A crude ["PIcDNA probe generated from the partially enriched lactate dehydrogenase mRNA was used for the initial screening of recombinant clones by filter hybridization. This initial screening identified a subset of bacterial colonies which potentially contained lactate dehydrogenase cDNA sequences. Further screening of these colonies was accomplished after enrichment of the cDNA probe for an abundant sequence (45) prepared by Hue 111 digestion of the cDNA probe and isolation of a 500-nucleotide fragment. Use of this restriction fragment as a hybridization probe allowed us to identify 4 bacterial colonies as putative lactate dehydrogenase cDNA clones. Size fractionation of Pst I digests of plasmid DNA isolated from the 4 colonies identified them as containing identical plasmids. The presence of lactate dehydrogenase DNA sequences in these plasmids was conf i e d by hybridization-selected in vitro translation resulting in the synthesis of a protein which was identified as lactate dehydrogenase on the basis of its molecular sue and its specific reaction with rabbit anti-rat lactate dehydrogenase-5 isozyme antiserum. Restriction endonuclease analysis indicated the cloned lactate dehydrogenase cDNA to be approximately 700 nucleotides in length.
The use of the cloned cDNA probe in a Northern blot analysis of electrophoretically separated denatured poly(A)' RNA indicated lactate dehydrogenase mRNA to be about 1800 nucleotides in length. Excluding a poly(A) sequence of about 150 nucleotides, the composite of the translated and untranslated regions of lactate dehydrogenase mRNA is thus formed by about 1650 nucleotides. Based on M , = 35,000 for the in vitro translation product of lactate dehydrogenase mRNA, approximately 600 bases form the untranslated region of the mRNA. The relative size of this untranslated region is similar to that observed for a number of other proteins (50-55).
Using the discriminating ability of the cDNA probe to select specifically for lactate dehydrogenase mRNA, we have demonstrated by Northern blot as well as by kinetic hybridization analysis that isoproterenol (or dibutyryl CAMP) stimulation of C6 glioma cells causes a 2-fold increase in the number of lactate dehydrogenase mRNA molecules 8 h after addition of the inducing agents. Although the rate of accumulation of lactate dehydrogenase mRNA after isoproterenol stimulation followed a time course similar to the one previously determined by us using in vitro translation of poly(A)+RNA isolated from isoproterenol-stimulated glioma cells (exhibiting maximal activity 8 h after stimulation) (ll), we previously found a n 8-to 10-fold maximal increase of the functional level of lactate dehydrogenase. Since we have ruled out technical problems in either the hybridization or translation assays (all analyses were repeatedly carried out under different conditions using the same poly(A)'RNA preparations), this discrepancy in our finding suggests that isoproterenol or dibutyryl cAMP not only cause an increase of the number of lactate dehydrogenase mRNA molecules but that the newly induced mRNA exhibits a higher translational efficiency per molecule of mRNA than lactate dehydrogenase mRNA from unstimulated glioma cells. The increased translation efficiency could be the result of factors present in the induced poly(A)'RNA fraction allowing for more effective utilization of lactate dehydrogenase mRNA or could be the result of inherent structural modifications of the mRNA.
The reproducible significant difference observed in the plateau hybridization values after kinetic hybridization of poly(A)+RNA (Fig. 8) suggests that lactate dehydrogenase mRNA from isoproterenol-or dibutyryl CAMP-stimulated cells exhibits a higher degree of homology with the cloned DNA template (which was prepared from mRNA of stimulated cells) than mRNA from unstimulated cells. Since the products of in vitro translation of lactate dehydrogenase mRNA from unstimulated and stimulated cells are identical (ll), the difference in homology may reflect significant base sequence changes in the untranslated region of lactate dehydrogenase mRNA from stimulated cells as compared to mRNA from unstimulated cells. While our data do not prove a transcriptional role for CAMP, the model we consider pre-14.

Cyclic AMP and
Lactate Dehydrogenase mRNA  Sucrose aradlent. LOH mRNA 1 C t l v l t Y and t o t a l mRNA