Rat alpha 1-acid glycoprotein mRNA. Cloning of double-stranded cDNA and kinetics of induction of mRNA levels following acute inflammation.

Messenger RNA preparations from the livers of normal and acutely inflamed rats were translated in a mRNA-dependent cell-free protein-synthesizing system. Immunoprecipitation of the translation products with a specific antibody prepared against purified rat plasma al-acid glycoprotein (al-AGP) detected an abundant 23,000 molecular weight peptide induced by inflammation. In order to examine the regulation of this acute phase protein, a double-stranded cDNA to partially purified rat al-AGP mRNA was synthesized, inserted into the Pst I site of the plasmid pBR322 by a GC-tailing technique, and used to transform Escherichia coli RR1. A recombinant plasmid containing a 740-base pair insert with a contiguous poly(dA) segment was identified as containing al-AGP cDNA sequences by partial DNA sequence analysis, and by specific hybrid selection of mRNA followed by in vitro translation and immunoprecipitation. The cloned double-stranded cDNA isolated  from  the  recombinant pBR322 vector was recloned in the single-stranded DNA bacteriophage M13mp7 in order to develop a specific hybridization probe for mRNA quantitation. The al-AGP-specific cDNA probe was used to examine a1AGP mRNA levels in total liver RNA during a period of 48 h following the induction of acute inflammation by the administration of turpentine. In normal rat liver, aI-AGP mRNA comprised about 0.0006% of total cellular RNA. An increase in the amount of al-AGP mRNA was first detected 4 hr after the onset of inflammation, and it reached a maximum level of induction at 36 h following the  administration of the  inflammatory agent. At this point, al-AGP mRNA comprised about 0.053% of total cellular RNA, representing a 90-fold increase over its normal level. This induction was associated with a substantial increase in the circulating

Messenger RNA preparations from the livers of normal and acutely inflamed rats were translated in a mRNA-dependent cell-free protein-synthesizing system. Immunoprecipitation of the translation products with a specific antibody prepared against purified rat plasma al-acid glycoprotein (al-AGP) detected an abundant 23,000 molecular weight peptide induced by inflammation. In order to examine the regulation of this acute phase protein, a double-stranded cDNA to partially purified rat al-AGP mRNA was synthesized, inserted into the Pst I site of the plasmid pBR322 by a GC-tailing technique, and used to transform Escherichia coli RR1. A recombinant plasmid containing a 740-base pair insert with a contiguous poly(dA) segment was identified as containing al-AGP cDNA sequences by partial DNA sequence analysis, and by specific hybrid selection of mRNA followed by in vitro translation and immunoprecipitation. The cloned double-stranded cDNA isolated from the recombinant pBR322 vector was recloned in the single-stranded DNA bacteriophage M13mp7 in order to develop a specific hybridization probe for mRNA quantitation. The al-AGP-specific cDNA probe was used to examine a1-AGP mRNA levels in total liver RNA during a period of 48 h following the induction of acute inflammation by the administration of turpentine. In normal rat liver, aI-AGP mRNA comprised about 0.0006% of total cellular RNA. An increase in the amount of al-AGP mRNA was first detected 4 hr after the onset of inflammation, and it reached a maximum level of induction at 36 h following the administration of the inflammatory agent. At this point, al-AGP mRNA comprised about 0.053% of total cellular RNA, representing a 90-fold increase over its normal level. This induction was associated with a substantial increase in the circulating * This research was supported by Grant CA 18450 to the Specialized Cancer Center from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. plasma concentration of al-AGP, as measured by quantitative rocket immunoelectrophoresis. These results demonstrate that acute inflammation causes an increased accumulation of a specific mRNA which is responsible for the increased plasma concentration of al-AGP.
a1-Acid glycoprotein is an M, = 41,000 plasma protein that is produced by the liver. It consists of a single polypeptide chain and contains five carbohydrate side chains that account for approximately 45% of its mass (1). In normal animals, al-AGP' has a relatively low plasma concentration. However, in response to various acute inflammatory agents such as turpentine or bacterial endotoxin, the plasma level of al-AGP increases rapidly. This response is typical of the class of plasma proteins known as the acute phase reactants (2).
Previous studies suggest that the increased plasma concentrations of the acute phase reactants are due to an increased synthesis of these proteins (3-5). However, little is known about the mechanisms which underly this increased synthesis. In this report we describe a preliminary characterization of the induction of al-AGP following the onset of acute inflammation. A cloned cDNA hybridization probe was utilized to demonstrate that the increased plasma concentration of al-AGP appears to be the result of a dramatic increase in the level of liver al-AGP mRNA.

EXPERIMENTAL PROCEDURES
Animals-Acute inflammation was induced in Sprague-Dawley rats (230-280 g) by a single subcutaneous injection of turpentine (0.5 ml/lOo g of body weight) in the dorsal lumbar region. At different times following injection, the rats were anesthetized with ether, plasma was obtained from the inferior vena cava, and the livers were excised and rapidly frozen in liquid nitrogen.
Protein Purification and Antibody Production-cr-AGP was purified from rat plasma obtained 48 h after turpentine injection by a combination of ammonium sulfate precipitation, DEAE-Sepharose (Pharmacia) and CM-Sepharose (Pharmacia) ion exchange chromatography, affinity chromatography on wheat germ lectin Sepharose (Pharmacia), and preparative polyacrylamide gel electrophoresis. Antiserum was prepared in rabbits with complete Freund's adjuvant (Difco) by the method of Crowle (6). Antibodies to purified al-AGP showed a single precipitin line in crossed immunoelectrophoresis (7) with rat plasma.
Preparation of Total RNA-Total RNA was extracted from frozen liver essentially as described by Chirgwin et al. (8). Livers were ' The abbreviations used are: a,-AGP, al-acid glycoprotein; SDS, sodium dodecyl sulfate; Hepes, 4-(2-hydroxyethyl)-l-piperazine/ethanesulfonic acid; ds-cDNA, double-stranded cDNA; Rot, the product of initial RNA concentration in moles of nucleotides per liter and the time in seconds; Rotl/S, the Rot value at 50% hybridization. homogenized in 15 volumes of 4 M guanidine thiocyanate, 25 mM sodium citrate at pH 7.0, 0.5% N-laurylsarcosine, 0.1 M 2-mercaptoethanol, and 0.2% Antifoam A (Sigma). The homogenate was centrifuged at 6000 rpm in a Sorvall GSA rotor for 15 min a t 10 "C. The supernatant fluid was adjusted to pH 5.0 by addition of acetic acid and the RNA was precipitated by 0.75 volume of ethanol at -20 "C for 2 h. RNA was collected by centrifugation and redissolved in 7.5 M guanidine hydrochloride containing 25 mM sodium citrate and 5 mM dithiothreitol. Following two additional precipitations using 0.5 volume of ethanol, the residual guanidine hydrochloride was extracted from the precipitate with absolute ethanol. RNA was dissolved in sterile water, insoluble material was removed by centrifugation, and the pellets were re-extracted with water. The RNA was adjusted to 0.2 M potassium acetate and precipitated by addition of 2.5 volumes of ethanol at -20 "C overnight.
Preparation of Poly(A)-containing RNA-Total RNA was dissolved in 20 m~ Hepes buffer at pH 7.2 containing 10 mM EDTA and I% SDS, heated at 65 "C for 10 min, then quickly cooled to 25 "C. The RNA solution was then diluted with an equal volume of water and adjusted to 300 m~ in NaCl. Samples containing up to 2400 A m units' of RNA were mixed with 2.5 g of poly(U)-Sepharose (Pharmacia), equilibrated in the above final buffer, and mixed gently at room temperature for 30 min. The affinity matrix was then collected by filtration and washed with 70 ml of the above buffer. Poly(A)containing RNA was eluted with 70% formamide containing 1 mM Hepes buffer (pH 7.2) and 2 mM EDTA. The eluate was adjusted to 0.24 M NaCl and the RNA was precipitated by 2.5 volumes of ethanol at -20 "C. Eluted RNA was further purified by CsCl centrifugation as described by Glisin et al. (9), and then subjected to a second cycle of chromatography on poly(U)-Sepharose.
Poly(A)-containing RNA recovered from the second poly(U)-Sepharose chromatography step was heated at 65 "C for 10 min, quickly cooled to 25 "C, and sedimented through 5-29.9% isokinetic sucrose gradients (10) containing 1% SDS, 25 m~ Hepes (pH 7.4). and 5 mM EDTA in a Beckman SW 41 rotor at 27,000 rpm for 17 h at 20 "C. The gradients were collected in 0.4-ml fractions and the RNA was precipitated twice with ethanol. RNA samples were redissolved in water and used for translation analyses.
Cell-free Translation a n d Immunoprecipitation-Cell-free translations in the mRNA-dependent protein-synthesizing system derived from rabbit reticulocyte lysates were carried out as described (11). Immunoprecipitation was performed by an initial binding reaction with monospecific antibodies to rat a,-AGP, followed by adsorption to a staphylococcal protein A (IgGSORB, The Enzyme Center Inc., Boston) antibody adsorbent (12). Electrophoresis of translation products and immunoprecipitates was performed on 10% or 10-17.5% gradient polyacrylamide gels containing 0.1% SDS, essentially as described by Laemmli (13). To prepare fluorograms, the gels were impregnated with EN3HANCE (New England Nuclear), dried under vacuum, and exposed to Kodak XRP film a t -70 "C for several days.
Construction of Recombinant Plasmids-Single-stranded cDNA was synthesized with avian myeloblastosis virus reverse transcriptase (kindly provided by Dr. Joseph Beard, Life Sciences, Inc., St. Petersburg, FL), with a 4-fold weight excess of oligo(dT) primer to mRNA, essentially as described previously (Il), but with the omission of actinomycin D and the inclusion of 2 m~ sodium pyrophosphate. Double-stranded cDNA was synthesized from a single-stranded cDNA template with reverse transcriptase as described by Norgard et al. (14). The ds-cDNA was made blunt-ended by incubation with SI nuclease (14). The reaction mixture consisted of 0.2 M sodium acetate (pH 4.5), 0.4 M sodium chloride, 2.5 mM zinc acetate, and 1-5 units of SI nuclease/ng of ds-cDNA in a final reaction volume of 100 p1. The ds-cDNA was incubated at 37 "C for 1 h, supplemented with a IO-fold excess of E . coli tRNA, extracted with phenol:chloroform, ethanol-precipitated, and then sedimented through a 527.8% isokinetic sucrose gradient (10) in an SW 41 rotor at 41,000 rpm at 5 "C for 24 h. Reaction products with a length of 200 base pairs or longer were pooled, dialyzed against H20, and lyophilized. Terminal transferase was employed to add approximately 30 dCMP residues to the ds-cDNA 3'-termini with reaction condtions as described by Wahl et al. (15). About 15 residues of dGMP were added in a similar reaction to the plasmid pBR322 that had been previously cleaved by Pst I restriction endonuclease. Equimolar amounts of dC-tailed ds-cDNA and dG-tailed plasmid were annealed in 150 m~ NaC1, 10 mM Tris-HC1 (pH 7.6), and 1 mM EDTA at a final DNA concentration of 10 One A m unit corresponds to the optical absorbance a t 260 nm of -50 pg of RNA in 1 ml of water. pg/ml. The annealing mixture was placed at 75 "C and the temperature was lowered to 0 "C over a 5-h time period. Transformation of E .
coli RRl was carried out essentially as described by Norgard et al. (14). Transformants were selected by growth on tetracycline. All recombinant DNA work was carried out according to "The National Institutes of Health Revised Guidelines for Research Involving Recombinant DNA Molecules" issued in January 1980. The experiments were covered by Section 111-0 of the revised guidelines according to an approved "Memorandum of Understanding and Agreement." Identification Of Transformants Containing a, -AGP Sequences-Transformants were grown on replicate nitrocellulose filters, and prepared for hybridization according to Grunstein and Hogness (16). Filters were hybridized with "'P-labeled single-stranded cDNAs synthesized from normal and acute phase poly(A)-containing RNA, then they were washed, dried, and autoradiographed. Colonies that showed a strong hybridization signal with acute phase cDNA and a weak hybridization signal with normal cDNA were selected for further analysis. Plasmid DNA was isolated (17), cleaved with restriction endonuclease Pst I, and the cleavage fragments were separated by electrophoresis in horizontal 1.5% agarose gels (18). One of these recombinant plasmids (pAGP 663) had a ds-cDNA insert size of approximately 740 base pairs, and was selected for further analysis as described below.
Cloning ofpAGP 663 in Bacteriophage M13mp7-Isolated recombinant plasmid DNA (1 pg) was digested with 3 units of Pst I in a final volume of 10 pl. The enzyme was inactivated by heating to 65 "C for 10 min, and the DNA cleavage fragments were diluted to a final concentration of 10 ng/pl. The replicative form of the vector DNA, M13mp7 described by Messing et al. (19), was cleaved with Pst I and treated in an identical manner. Vector DNA (20 ng) and insert DNA (60 ng) were ligated in a 10-pl reaction mixture containing 70 mM Tris-HC1 (pH 7.5), 7 mM MgC12, 7 mM dithiothreitol, 0.07 mM ATP, and 0.2 unit of Tq DNA ligase. The reaction mixture was incubated at 15 "C for 2 h and then diluted to 80 pl with 25 mM EDTA.
Preparation of M13mp7 Recombinant DNA-Preparation of M13mp7 recombinant DNA was carried out essentially as described by Sanger et al. (22). A fresh overnight culture of E. coli strain JM103 was diluted 1:100 into 2 ml of 2 X YT medium. The cultures were incubated at 37 "C for 1.5 h and the recombinant plaques were transferred by pipettes to the medium. Incubation of the culture was continued for 8 h and then the bacteria were pelleted by centrifugation for 10 min. A portion of the supernatant fluid (40 pl) was treated with 1 pl of 20% SDS and the insert size was determined by electrophoresis on 0.7% agarose gels. The DNA was transferred to nitrocellulose fdters (23) and then hybridized to a "P-labeled cDNA prepared from acute phase liver mRNA. Recombinants containing the coding strand were identified by a positive hybridization signal, whereas recombinants containing the noncoding complementary DNA strand were identified by a negative hybridization signal and an insert size identical with the coding strand clones. Purified single strand DNA, suitable for use in cDNA synthesis or hybrid selection of mRNA, was prepared from culture media by 2 cycles of polyethylene glycol precipitation and phenol extraction as described by Sanger et al. (22).
Hybrid Selection of mRNA a n d Cell-free Translation"M13mp7 phage DNA, carrying the inserted cDNA strand, was immobilized on nitrocellulose filters essentially as described by Ricciardi et al. (24). DNA (5 pg) was adjusted to 0.2 N NH,OH and 2 M NaCl in a final volume of 20 p1, heated to 100 "C for 1 min, and immediately spotted on a 1 cm2 nitrocellulose filter. The filter was air dried for 5 min and then baked under vacuum a t 80 "C for 2 h. The filter was cut into approximately 20 small pieces and incubated in 1 ml of 70% (v/v) formamide, 0.3 M NaCl, 10 mM Hepes (pH 7.5), 0.2% SDS, and 5 mM EDTA for 5 min at 47 "C. Hybridization was carried out subsequently in 150 pl of the above solution containing 20 pg of total poly(A)containing RNA derived from 36-h acute phase liver. Following a,-Acid Glycoprotein mRNA hybridization for 3 h a t 47 "C, the filters were washed five times for 15 min each a t 47 "C with 1 ml of 75 mM NaCI, 7.5 m~ sodium citrate (pH 7.4), and 0.5% SDS. Hybridized RNA was eluted by two successive 10-min incubations at 47 "C in 150 p1 of 9 0 % (v/v) formamide, 0.1% SDS, 10 m~ Hepes (pH 7.5), 1 mM EDTA, and 33 pg/ml of tRNA. The RNA was precipitated twice with ethanol and then translated in the reticulocyte lysate protein-synthesizing system.
cDNA Synthesis from Ml3AGP 663m Recombinant DNA-One microgram of recombinant M13mp7 DNA was mixed with 2 p g of oligo(dT) primer (Collaborative Research) in a buffer containing 125 mM NaCI, 17.5 m~ Tris (pH 7.5), 17.5 mM MgC12, and 0.1 m M EDTA in a final volume of 11 pl. The DNA was denatured by heating a t 100 "C for 3 min. then plunged immediately into an ice water bath for 10 min to allow annealing of the oligo(dT) primer. The reaction mixture was then supplemented to give final concentrations of 30 pM each of unlabeled deoxyribonucleoside triphosphates (dATP, dGTP, dTTP), 5 p~ of [a-"2P]dCTP (410 Ci/mmol; Amersham), 1 mM dithiothreitol, and 2 units of the Klenow fragment of E. coli DNA polymerase 1 (New England Nuclear) in a final reaction volume of 25 pl. The reaction mixture was incubated a t 25 "C for 1 h, followed by phenol extraction and Sephadex (3-100 column chromatography. The DNA was denatured in 0.2 M NaOH by heating a t 75 "C for 15 min, then sedimented through 5-27.8% isokinetic sucrose gradients (10) containing 0.1 M NaOH, 0.9 M NaCI, and 1 mM EDTA for 24 h in a Beckman SW 41 rotor a t 41,000 rpm a t 4 "C. Fractions containing radioactive DNA (2 X 10" cpm pg") were pooled and the DNA was collected by ethanol precipitation.

RESULTS
Identification of the mRNA for al-AGP-Total poly(A)containing RNA isolated from normal and acute phase livers was translated in the mRNA-dependent protein-synthesizing system derived from rabbit reticulocyte lysates. Analysis of the total translation products by gel electrophoresis and fluorography indicated that an abundant M, = 23,000 peptide had been induced by the action of the inflammatory agent (Fig.  1). Immunoprecipitation of the total translation products with a monospecific antibody to aI-AGP (Fig. 1) demonstrated a single radioactive band of M, = 23,000, indicating that this peptide corresponded to the translation product of aI-AGP mRNA.
An RNA preparation enriched for al-AGP mRNA was obtained by sedimenting total poly(A)-containing RNA through isokinetic sucrose gradients. The RNA fractions were translated and analyzed by gel electrophoresis and fluorography. A comparison of normal and acute phase translation products indicated that density gradient fraction 10 ( Fig. 2) was substantially enriched for al-AGP mRNA. This poly(A)containing RNA fraction was then used for ds-cDNA synthesis and subsequent cloning in pBR322.
Construction a n d Identification of Recombinant Plasmids Containing aI-AGP cDNA-A ds-cDNA was synthesized from the gradient fraction enriched for aI-AGP mRNA and was examined by agarose gel electrophoresis. Most of the radioactive material migrated at a size of 400-800 base pairs (data not shown). The ds-cDNA was digested with S1 nuclease, separated from low molecular weight products by sedimentation through sucrose gradients, and subsequently inserted into the Pst I site of the ampicillin resistance gene of the plasmid pBR322 by the oligo(dG) oligo(dC)-joining technique. The DNA was used to transform E. coli RR1 and approximately 1000 tetracycline-resistant transformants were selected. Transformants were screened as described under "Experimental Procedures," and a recombinant DNA plasmid (pAGP 663) containing a 740-base pair insert was tentatively identified as being a candidate for containing al-AGP mRNA sequences.
In order to establish the identity of the al-AGP candidate plasmid, the ds-cDNA insert was employed to select its corresponding mRNA by hybridization, followed by cell-free translation of the hybrid-selected mRNA and immunoprecipitation of the translation products. Since this procedure requires denatured or single-stranded DNA that is complementary to the mRNA, the ds-cDNA was recloned into the single strand phage vector M13mp7. Single-stranded DNA from one of the cDNA strand clones (M13AGP 663c) was immobilized on nitrocellulose and used to specifically select its corresponding mRNA. Fig. 3 (Lane B) demonstrates that M13AGP 663c specifically hybridizes to an mRNA which then directs the cell-free synthesis of a single peptide with an apparent M , = 23,000. The translation product is immunoprecipitated with monospecific antibodies to rat al-AGP (Fig. 3, Lane 0 , thus establishing the identity of the recombinant plasmid.
T o further c o n f m t h e identity of the recombinant plasmid, a partial nucleotide sequence of the ds-cDNA insert was determined. The restriction endonuclease cleavage map used in the nucleotide sequence determination is shown in Fig. 4.
The 271-base pair Ava I1 fragment was labeled at the 3' end, then subjected to DNA sequence analysis by the method of Maxam and Gilbert (25). Fig. 5A shows the sequence of 78 nucleotides of this fragment and the amino acid sequence predicted from one of the six possible reading frames. Multiple termination codons were observed in each of the five other reading frames. Fig. 5B indicates the alignment of the inferred amino acid sequence with the known protein sequence (1) of human al-AGP. Of the 26 amino acid residues shown in the rat sequence, 19 are identical with the corresponding residues in the human protein, providing further support for the identity of the recombinant plasmid.
Size Estimation ofal-AGPmRNA-Total poly(A)-containing RNA samples from normal and acute phase livers were denatured by reaction with glyoxal followed by electrophoresis in agarose gels (27). Following electrophoresis the RNA was transferred to diazotized paper (28) and hybridized to a 'lPlabeled cloned DNA probe. As shown in Fig. 6, the cloned cDNA probe hybridized to an RNA species of about 850 nucleotides in length. Furthermore, the relative amount of aI-AGP mRNA in the acute phase liver is increased greatly compared to normal liver. The broad width of the mRNA gradient fraction from which the RNA was collected. N indicates RNA obtained from normal liver, and A indicates RNA obtained from acute phase liver. The molecular weight markers as indicated are m-macroglobulin (180,000), phosphorylase b (92,500), bovine serum albumin (68,000), ovalbumin (45,000). carbonic anhydrase (30,000), and cytochrome c (12,300).

Human U1-ACP
FIG. 5. Nucleotide sequence of a portion of the coding strand of the a,-AGP ds-cDNA insert. A, DNA from plasmid pAGP 663 was cleaved with restriction endonuclease Aua I1 and labeled at the 3' ends with E. coli DNA polymerase I (Klenow fragment) and [a- '"PIdCTP (25). The 5'-terminal Aua I1 fragment was isolated by polyacrylamide gel electrophoresis and its sequence was determined as described by Maxam and Gilbert (25). The amino acid sequence predicted from the only open reading frame is shown. The numbers above the line indicate the corresponding position in human al-AGP as described by Schmid (1). B, the boxes enclose amino acid residues that are identical in both rat and human al-AGP. The one-letter notation for amino acids is according to the recommended convention of Dayhoff (26). The double letter code for three of the amino acid positions represents normal variation in the human protein as described by Schmid (1). 36-h acute phase liver (Lane I ) or normal liver (Lane 2) was denatured with glyoxal and examined by electrophoresis in 1.5% agarose gels (27). The RNA was transferred to diazotized paper (28) and hybridized to a '"P-labeled DNA probe prepared by 3' endlabeling of an Aua I1 digest of cloned al-AGP DNA insert using the Klenow fragment of E. coli DNA polymerase I (25). The molecular weight markers used are Hind111 restriction endonuclease fragments of SV40 DNA that were labeled with [y""P]ATI' at the 5' ends (25).
band detected is probably a function of heterogeneity in the 3"terminal poly(A) segment reported for most eucaryotic mRNAs (29).
Preparation of an al-AGP-specific Hybridization Probe-Single-stranded DNA from one of the M13 mRNA strand clones (M13AGP 663m) was employed to develop a highly specific, easily prepared hybridization probe for use in mRNA quantitation in an RNA-driven reaction in solution. This method is a modification of a technique previously described by this laboratory (30). A detailed characterization of the procedure and the evidence supporting the model shown in Fig. 7 are described elsewhere." Briefly, the technique takes " G . A. Ricca, J. M. Taylor, and J. E. Kalinyak, manuscript in preparation.
advantage of the availability of a 2-fold rotational axis of symmetry in the cloning site of the M13mp7 phage vector DNA (Fig. 7). Thus, when DNA is inserted into the Pst I site of bacteriophage M13mp7, the single-stranded DNA can fold back, allowing the palindromic sequence to anneal and thus form a stable hybrid with a single-stranded loop the size of the inserted DNA sequence. In the case of M13AGP 663m, stretches of homopolymeric dG and dC tracts at the boundaries of the inserted DNA sequence further stabilize the hybrid structure (Fig. 7). Direct DNA sequence analysis (data not shown) at the 3' end of pAGP 663 demonstrated the presence of a contiguous poly(dA) tract that is 21 nucleotides in length (Fig. 4). Thus, oligo(dT) was annealed to the poly(dA) segment of the M13 coding strand insert and was employed as a primer for the Klenow fragment of DNA polymerase I to synthesize the functional equivalent of a complementary DNA copy of the mRNA coding strand. Following the synthesis reaction, the radioactively labeled DNA material was isolated by sedimentation in an alkaline isokinetic sucrose gradient that readily dissociated the noncovalently attached probe from its template. Radioactive DNA having a very sharp sedimentation profile was obtained, suggesting DNA of uniform length. Gel electrophoresis of the labeled denatured (27) cDNA indicated a single band with an electrophoretic mobility corresponding to the length of the inserted sequence (data not shown), strongly suggesting that the transcription with DNA polymerase I appeared to terminate near the oligo(dG) segment at the 5' end of the template al-AGP coding strand (Fig. 7).

Measurement of al-AGP mRNA Levels-The MIS-derived al-AGP DNA was employed in
RNA-driven hybridization reactions to determine the relative concentrations of aI-AGP mRNA in total cellular RNA extracted from rat liver before, and at various times after the animals received a single subcutaneous injection of turpentine. Representative hybridization curves are shown in Fig. 8 and the R"t1/2 values for all time points are indicated in Fig. 8. The hybridization reaction went to greater than 90% completion with pseudo fmt order kinetics, occurring over a log Rot range of approximately 2. The S1 nuclease-resistant background of the cDNA in the absence of RNA was about 4% of the input probe.
The relative increase in the concentration of al-AGP mRNA sequences following the induction of acute inflammation was measured by comparing the R0tl,2 values of the hybridization reactions; 2 h after the onset of inflammation no change could be detected in the mRNA levels for al-AGP (Figs. 8 and 9). At 4 h after injection there was a 3.4-fold increase in these levels, followed by a progressive increase which reached a maximum level 36 h after injection (Fig. 9), at which point the level was 90-fold higher than that of the normal liver. This induction was then followed by a decline which could be detected by 48 h after the onset of inflammation.
aI-AGP mRNA levels were correlated with the plasma levels of its corresponding protein. Rocket immunoelectrophoresis (35) of rat plasma samples indicated that the al-AGP concentration in the plasma of untreated rats was 0.064 mg/ ml. Fig. 9 shows that an increase in the plasma level of a ] -AGP was first detected between 4 and 6 h after the onset of inflammation. This induction was followed by a progressive increase to a concentration of 2.5 mg/ml, a value that is in excellent agreement with the serum concentration of 2.7 mg/ ml previously reported by Nagashima et al. (36), under conditions of inflammation identical with those in this report. However, we observe 2.7-fold lower levels of aI-AGP in normal plasma than that which was reported by Nagashima et al. These results suggest that basal levels of plasma aI-AGP may differ substantially, depending on a variety of hormonal and environmental factors.

DISCUSSION
By employing a monospecific antibody to rat plasma al-AGP, we have identified a peptide with a M , = 23,000 as being the primary translation product of rat a,-AGP mRNA. This size is similar to the molecular weight of the peptide portion of human al-AGP determined from the amino acid sequence by Schmid (1). When the same antibody is employed to immunoprecipitate a [3H]leucine-labeled rat liver perfusate, a protein with an apparent M, = 44,000 is ob~erved,~ in agreement with a previous report (36) on the molecular weight of rat a,-AGP.
The translation assay was utilized to monitor the partial purification of the mRNA and to c o n f i i the identity of recombinant bacterial plasmids containing ds-cDNA sequences specific for al-AGP. Recloning the ds-cDNA insert 3.16, 12 h; 1.19, 18 h; 1.05,24 h; 0.75, 36 h; and 1.54, 48 h. The fraction of al-AGP mRNA present in total RNA can be calculated as follows: R0t,,2 of pure aI-AGP mRNA = (Rd1/2 of pure rat albumin mRNA) (complexity of al-AGP mRNA)/(complexity of albumin mRNA). The complexity of pure albumin mRNA in 2265 nucleotides (34) and the complexity of aI-AGP rnRNA is 850 nucleotides. Under identical conditions purified albumin mRNA hybridizes to albumin cDNA with a Rotl/* of 0.001.4 Thus, the RUtI,2 value for hybridization of purified a,-AGP mRNA to a,-AGP cDNA would be 0.0004. Comparison of this value to the f 1 0 t 1~~ value obtained upon hybridization of al-AGP cDNA to total RNA from normal rats (66.9) indicates that al-AGP mRNA comprises approximately 0.0006% of total normal liver RNA. The fraction of al-AGP mRNA present in total mRNA can be calculated based on the observation4 that 2% of the total liver RNA is mRNA.  The relative change in al-AGP mRNA levels was determined from the hybridization data shown in Fig. 8 by comparison of the ROtIr2 value of normal liver RNA to the Rotll2 values obtained for RNA isolated from liver a t various times after the onset of inflammation. Plasma levels of al-AGP were determined by rocket immunoelectrophoresis as described by Weeke (35). Each time point represents the average of individual analysis of plasma from 4 separate rats and is expressed as the relative increase over normal plasma levels of a,-AGP. Plasma concentrations were determined from a standard curve employing purified al-AGP, with the concentration being estimated from the absorption at 280 nm, assuming an E& I cn, of 6.7 (36).
into the single-stranded bacteriophage M13mp7 was found to be particularly advantageous with regard to identifying recombinant clones by hybridization selection of mRNA and subsequent translation. Since the M13 bacteriophage DNA is obtained in a single strand form, potential problems associated with renaturation of the DNA during binding to the nitrocellulose are completely avoided. We have found that hybridization selection of mRNA using M13 recombinant DNA is far more reproducible than the same procedure with denatured double-stranded DNA inserts derived from bacterial plasmids such as pBR322. The usefulness of recloning a cDNA sequence into M13mp7 is underscored when one considers the ease of obtaining a hybridization probe from this vector, using the method described in this report.
Our results demonstrate directly that administration of an inflammatory agent can result in a dramatic accumulation of the mRNA coding for an individual component of the seromucoid fraction (2) of plasma, al-acid glycoprotein. It is noteworthy that no change in mRNA levels occurs within 2 h after the onset of inflammation, but a significant increase has occurred by 4 h. This lag time may be required for the host response to inflammation through the production of various leukocyte factors or other intermediate humoral factors that may stimulate the liver to increase the production of the acute phase proteins (2). Our observations are in agreement with the results of Neuhaus et al. (41, who found that the increased synthesis of the seromucoid fraction, following inflammation, could be blocked by the administration of actinomycin D not later than 4 h after injury. The 90-fold increase in the relative amount of aI-AGP mRNA a t 36 h after the onset of inflammation results in this mRNA species becoming a major component of the liver mRNA. Assuming a sequence complexity of 850 nucleotides for al-AGP mRNA, it can be calculated (see Fig. 8) that this mRNA is induced to become about 2.7% of the total liver mRNA at the height of the acute phase response. In this regard, albumin mRNA levels show a transient 5-fold decrease, to about 2.2% of total liver mRNA during the acute inflammatory r e a~t i o n .~ Thus, al-AGP mRNA may be induced to become a major liver mRNA species during the maximum response of the liver to acute inflammatory agents. The mechanisms responsible for this apparent induction of al-AGP mRNA are currently under investigation.