Identification and sequencing of cDNA clones for the rodent negative ...

7 downloads 56694 Views 765KB Size Report
puter using the Apple I1 software for DNA and protein se- quence data ..... Northemann, W., Andus, T., Gross, V. & Heinrich, P. C. (1983). (1982) Biochem. J. 206 ...
Eur. J. Biochem. 164,375-381 (1987) 0 FEBS 1987

Identification and sequencing of cDNA clones for the rodent negative acute-phase protein al-inhibitor 3 Michael SCHWEIZER ’, Kenji TAKABAYASHI Thomas GEIGER’, Thomas LAUX ’, Gereon BIERMANN ’, Jean-Marie BUHLER’, Francis GAUTHIER4, Lilian M. ROBERTS and Peter C. HEINRICH’

*

Institut fur Mikrobiologie und Biochemie der Universitat Erlangen-Nurnberg Biochemisches Institut der Universitat Freiburg Service Biochimie, Centre dEtudes NuclCaires de Saclay, Gif-sur-Yvette Laboratoire de Biochimie, Universite Franqois-Rabelais, Tours

(Received September l/December 10, 1986) - EJB 86 0935

Rat al-inhibitor 3 clones were isolated by immunological screening of a 1 g t l l cDNA library prepared from rat liver poly(A)-rich RNA. The recombinant cDNA clones were identified by the absence of their immunoprecipitable products following hybrid-arrested in vitro translation. The size of the cognate poly(A)-rich RNA was estimated to be roughly SO00 residues. Approximately 16 h after induction of inflammation the amount of al-inhibitor 3 poly(A)-rich RNA decreases as shown by dot-blot hybridization and Northern analyses. The response of this negative acute-phase plasma protein to inflammation may therefore be considered to be at the pretranslational level. The characterized DNA constitutes an open reading frame of 225 amino acids followed by a canonical eucaryotic polyadenylation signal and a poly(A) tail. Sequence microheterogeneity, particularly in the 3’-flanking region was observed. An amino acid homology of 70% for al-inhibitor 3 with human and rodent a2-macroglobulin emphasizes the evolutionary relationship of the macroglobulins.

The globin and histone genes provide two of the many examples of multigene families in the eucaryotes [ l , 21. Even though the various genes of a family are differentially expressed, their products generally perform similiar functions PI. The rat a-macroglobulin gene family, currently consisting of three members, illustrates this point particularly well. The a-macroglobulins are glycoproteins containing unusual cyclic thiol esters essential for their proteinase-inhibiting activity [4]. Specifically,az-macroglobulinand al-inhibitor 3 are so-called acute-phase proteins, i.e. their plasma levels change, albeit in opposite directions, under varying physiological and pathological conditions [5 - 71. The serum concentration of a2macroglobulin may increase up to 1000-fold [8, 91, whereas that of al-inhibitor 3 undergoes a 70% reduction [lo]. In contrast, the serum level of al-macroglobulin remains virtually unchanged during acute inflammation [l11. cDNA clones for a,-macroglobulin have led to the elucidation of the molecular basis of its physiological reaction response [9, 121. Apparently, the regulation of a2-macroglobulin is at the level of mRNA synthesis. To date, al-inhibitor 3 has only been investigated biochemically [lo, 131. The native protein consists of a single polypeptide chain with a relative molecular mass of about 200000 [14]. For studying the regulation of this negative acutephase protein, al-inhibitor 3 cDNA clones are essential. Such

Correspondence to M. Schweizer, Institut fur Mikrobiologie und Biochemie der Universitat Erlangen-Nurnberg, StaudtstraSe 5 , D-8520 Erlangen, Federal Republic of Germany

clones have been isolated from a 1 gtl 1 recombinant cDNA library. Here we describe the characterization of these clones and show that as a result of inflammation there is a reduction in the amount of al-inhibitor 3 mRNA. The amino acid sequences derived from our clones and rat a2-macroglobulin cDNA are compared with each other and with the corresponding data for human a,-macroglobulin [lS, 161 and the significance of their similarities is discussed. MATERIALS AND METHODS Reagents

Reagents were obtained as follows: Boehringer Mannheim, Biolabs, Schwalbach and Gibco BRL, Eggenstein : restriction endonucleases, T4 DNA ligase, DNA polymerase I, Klenow fragment of DNA polymerase I, RNase T I from Aspergillus oryzae, S-bromo-4-chloro-3-indolylgalactopyranoside (X-Gal), isopropyl-P-D-thiogalactopyranoside (IPTG). Sigma, Miinchen: Sepharose CL6B-200, Dowex MR-3 mixed-bed resin, DNase (grade I), RNase (type 11-A) from bovine pancreas, agarose low EEO, acrylamide, N,W-methylene-bisacrylamide.Serva, Heidelberg: ampicillin, chloramphenicol. Amersham Braunschweig: [3sS]methionine (400 Ci/mmol), [u-~’P]~ATP (400 and 3000 Ci/mmol). Fluka, Neu-Ulm: N,N-dimethylformamide, formaldehyde, guanidinium hydrochloride. Wacker Chemie, Munchen: Silan GF31 (y-methacryloxypropyl-trimethoxysilan). BDH Poole, England: Repelcote (2% dimethyldichlorosilane in trichloromethane). Paesel, Frankfurt a. M. : oligo(dT)-cellulose type 2. Other reagents were of analytical grade.

376 Biological ma ter ials

RR1 (F-hsdS2O recA' aral4 proA2 lacy1 galK2 rpsL20 xy15 mtll supE44 L)was used as recipient for Escherichia coli transformation [17]. 71 - 18 [F'lacIqZ A M15 pro A (lac pro)] was the recipient for M13 phage transfection [IS]. BNN 97 [= BNN93 (1gtll) hsdR- hsdM' supE thr leu thi lacy1 tonA211 was used for antibody purification and Y1090 [ A lacU169 proA+ d Ion araD139 strA supF (trpC22::TnlO) (pMC9 = pBR322-ladQ)][19] was used as the bacterial lawn when screening the rat cDNA library. The rat cDNA g t l l library was kindly provided by M. Beato (Universitat Marburg). Inflammation was induced by intramuscular injection of 0.5 ml turpentine/100 g body weight in male Wistar or female Sprague-Dawley rats weighing 200 - 300 g.

For Northern blotting the appropriate poly(A)-rich RNAs were separated on 1YO agarose/formaldehyde (6.6%) gels prior to transfer to nitrocellulose [42,43]. The air-dried filters were autoradiographed on Kodak XAR-5 at - 70°C. Immunological screening of a 1 g t l l cDNA library

Fatty acid synthase was isolated from rat liver using DEAE-cellulose and sucrose gradient centrifugation according to Nepokroeff et al. [44]. The purity of the enzyme was determined in SDS/polyacrylamide gel electrophoresis using a 5% lower gel and a 3% upper gel [45]. A rabbit was immunized against native rat fatty acid synthase and boostered 4 and 6 weeks later. Following purification [46], the antibody was titered by Ouchterlony immunodouble-diffusion analysis [47] against rat fatty acid synthase. Finally, possible bacterial contaminants were reD N A manipulations moved by incubation with BNN 97 lysate bound to CNBrPlasmid DNA was isolated by scaling-up the mini-prep activated Sepharose 4B (Pharmacia, Freiburg) (cf. Biological method of Birnboim and Doly (201 followed by Sepharose 4B materials above) according to the manufacturer's instructions. The cleaned-up antibody, diluted 1 : 100, was used for in (Pharmacia, Freiburg) column chromatography. 1gtl 1 DNA plaque hybridization (10 80-mm plates each with lo4 situ was prepared as described [21]. Single-stranded M13 phage DNA for sequencing was isolated according to Messing et al. plaques). After the incubation with the antibody and sub[18, 221. DNA restriction was carried out in TAM buffer sequent washes, the filters were exposed to either (a) 1251[23] and the products were electrophoresed in agarose gels labelled Staphylococcus aureus protein A [48] or (b) according to Maniatis [24]. When necessary, DNA fragments horseradish-peroxidase-conjugatedsecond antibody as dewere eluted from agarose by the freeze-squeeze method [25] scribed in the Bio-Rad instruction manual (Bio-Rad Laboraor by the NaI method [26]. For Southern hybridization cDNA tories, Munchen). Positive signals were visualized (a) by exinserts of al-inhibitor 3 and c/2-macroglobulin were 32P- posure to Kodak XAR-5 at -70 "C or (b) histochemically labelled either by nick translation [27] or via a random primer with the appropriate colour development substrates as described by DeBlas and Chenvinski [49]. 1281. Electroblots of proteins (Western blotting) were carried DNA sequencing was performed by the dideoxy-chaintermination method of Sanger and Coulson [29]by subcloning out as described by Towbin et al. [50]. Rat al-macroglobulin, a,-macroglobulin, a1-inhibitor 3 restriction fragments of the cDNA clones into the appropriate polylinker restriction sites of M13 vectors, so-called forced and the equivalent monospecific polyvalent antisera were precloning [22,30 - 321. Experimental details for sequencing have pared as described [8, 511. been given by Schweizer et al. [33]. Hybrid-arrested translation Computing Hybrid-arrested translation was performed essentially as DNA sequence data were collated in an Apple IIe com- described [52]. 5 pg poly(A)-rich RNA were hybridized to puter using the Apple I1 software for DNA and protein se- 0.1 pg al-inhibitor 3 cDNA insert in 80% deionized quence data (University of Minnesota, Apple I1 USCD formamide, 10 mM Pipes pH 6.4, 0.45 M NaCI, 0.5 mM Pascal, Version 2.1,1982) [34]. Comparison of the cDNA with EDTA at 48 "C for 2.5 h. After ethanol precipitation the pellet other sequenced genes/proteins available from the EMBL was washed thrice with 70% ethanol and dissolved in 5 pl nucleotide sequence library (Heidelberg) was done on a VAX H 2 0 . Cell-free translation with rabbit reticulocytes took place terminal in the Institut fur Klinische Virologie, Universitat in a total volume of 50 pl in the presence of 1 pCi Erlangen-Niirnberg, with the sequence analysis software [35S]methionine/plat 30 "C for 2 h [53]. Immunoprecipitation, package from the University of Wisconsin and EMBL SDS/polyacrylamide electrophoresis and autoradiography were according to previous protocols [54, 551 (and references Heidelberg (Version 3, June 1985). therein). R N A manipulations

Isolation of total and poly(A)-rich RNA from rat liver with or without turpentine pretreatment was carried out as described [35 - 381. The quality of the RNA preparations was checked by electrophoresis in 1.5% agarose/5 M urea gels

WI.

For dot blots [40], RNA was spotted onto nitrocellulose (BA 85) using a 96-well Manifold (Schleicher & Schiill, Dassel) apparatus. The filter-bound RNA was hybridized with 32P-labelledcDNA for al-inhibitor 3 or a2-macroglobulin to a sequence homology of around 90% [41] for 20 h. The filters were washed thrice in 0.1 x standard saline/phosphate/EDTA, 0.1% SDS at 55°C for 30 min.

RESULTS Physical characterization of clones isolated by immunological screening of a A g t l l rat liver cDNA library

Here we present evidence that the recombinant A gtll clones which we isolated with rat fatty acid synthase antiserum are in fact specific for rat al-inhibitor 3. This finding suggests that the rat fatty acid synthase preparation may have been contaminated with al-inhibitor 3. In the first screening 28 weak signals were obtained from l o 5 plaques. After two rescreens six positive candidates remained. The cDNA inserts were released from the ;1 gtll

377 Sgl If

PSI1

I

5'

I

I

Hind 111

PSI1

Psfl

I

Hind I l l

1 C G A C A C A G G A C A t 1 G T G G I G G C C C I C G A T G C T C l G T C C 444 1 A C GGA G C A G C l A C T I T 1 s i n I H R G L N A S P r m n V A L V A L A L A L r u As? A L A L i u s r R Lrs I I R GLr A L A A L A r n n rnr

13'

Fig. 1. Restriction map and sequencing strategy of al-inhibitor ~DIVA. The black bar is the insert in pI3cDNA6 containing pI3cDNA1 and pI3cDNA4. The open bar represents pI3cDNA3 or pI3cDNA5. The 5' + 3' orientation was deduced from the open reading frame of pI3cDNA6. The extent and orientation of the M13 subclones of pI3cDNA6 and pI3cDNA3 are shown in the lower part of figure

vector by EcoRI restriction and their sizes determined by agarose gel electrophoresis. Thus the six clones could be assigned to four groups according to the lengths of their inserts: two contain about 600 bp, another two contain a 720-bp insert, the one member of the third group has an insert of 830 bp whereas the sixth clone has the largest insert, 1780 bp [56, 571. It seems likely that the clones having inserts of the same size are identical even though each clone was an independent isolate. For ease of manipulation the cDNA fragments were recloned into a derivative of pBR322 and designated pI3cDNAl -p13cDNA6. Southern hybridization of the six clones with 32P-labelled cDNA from pI3cDNAl revealed the presence of only two groups: five of the six clones cross-hybridize with each other (data not shown). Confirmation of this classification was obtained by Northern blotting. pI3cDNA6, as the representative of the five clones in the larger group, hybridizes with a 5.0 - 5.3-kb mRNA, whereas the remaining clone, pI3cDNA2, hybridizes with a mRNA less than half this size (about 2.3 kb) (data not shown). Further investigations were carried out solely with the five related clones since the multiple isolation of a certain cDNA fragment may support the validity of it being specific. Fig. 1 summarizes the restriction data for the overlapping cDNAs. The black bar depicts pI3cDNA6, the largest clone in the group; it contains pI3cDNA1 and pI3cDNA4. The open bar underneath represents the two remaining members of the group, pI3cDNA3 and pI3cDNA5. Regarding the restriction map only the two 3'-proximal Pstl restriction sites are common to all five clones. The unique BglII and the 3'proximal Hind111 restriction sites are missing in p13cDNA3 and pI3cDNAS (open bar). The second HindIII site is present in all clones; it is, however, displaced marginally towards the 3'-end of the insert in p13cDNA3 and pI3cDNAS. In spite of the differences in their restriction maps, Fig. 1 clearly shows the common 3' ends and the colinearity of the five clones, as exemplified by p13cDNA6 (black bar) and pI3cDNA3 (open bar). cDNA sequencing After subcloning into appropriate M13 vectors, the cDNA clones were sequenced by the dideoxy-chain-termination method [29, 321. The sequencing strategy is shown below the restriction map in Fig. 1. A total length of 802 nucleotides after subtraction of the poly(A) tail was obtained from the combined sequencing data of the overlapping clones

Fig. 2. Nucleotide and the predicted amino acidsequences of pI3cDNA6 andpI3cDNA3. The numbering of the nucleotide sequence starts at the first nucleotide of pI3cDNA6. The sequence of pI3cDNA3 is given below that of pI3cDNA6 and starts at nucleotide 460. For pI3cDNA3 only nucleotides or amino acids differing from those in pI3cDNA6 are shown. -; indicates no mismatch. The overlined HindIII restriction site (AAGCTT) in the 3'-untranslated region of pI3cDNA6 is mutated in pI3cDNA3 to AACCTT. The bracketed sequence of eight nucleotides is inserted before the poly(A) tail of pI3cDNA3

pI3cDNAl and pI3cDNA6 (Fig. 2). There is an uninterrupted reading frame corresponding to 225 amino acids. The untranslated sequence following the stop codon contains the characteristic eucaryotic polyadenylation signal AATAAA, 11 bp before the poly(A) tail. This location of the polyadenylation signal relative to the poly(A) sequence has been described previously [%I. pI3cDNA3 was sequenced to about 50% of its length (cf. Figs 1 and 2). Its nucleotide sequence and the amino acids encoded therein are indicated below pI3cDNA6 in Fig. 2 (starting at nucleotide 460 of pI3cDNA6); this aligns the sequences of pI3cDNA3 and pI3cDNA6 with respect to their stop codons. A comparison of the two nucleotide sequences shows them to be virtually identical. There are, however, within the common 343 nucleotides, including the 3'untranslated regions, 19 nucleotide substitutions. Of the four nucleotide substitutions in the coding region two are silent whereas the other two cause an amino acid exchange; in one

378

86

GLU

I335

I04 1353

I?? 1371

Lrs

*

G L H G I N P R O A L A P n f A L A i f u Lrs V A L G L H G L U G L U PHI P R O GlY

A S H A S H P R O l Y S 6 1 1 G i n ASH ASP GLU I A l A H I S THR

t i r s f f i ARC PRO A L A S I R

Y R L PRO LEU I H R CYS 1fU GlH

*

SIR PHf G L N ILI S L R L f U O l U I L E S E R T Y R H € T I t f I t t THR 564 V A L X

fisn n u

V A L i i c A L A ASP V A L " 1 1 A . Y A l "

* S I R "

t

inn

Lrs nu II

L E U sffi GLY . V A L * t

I40 1309

P H E 111 P R O L I U L Y S P R O T H R V A L L Y S L Y S L f U G L U A R C L E U GLY H I S V A L 5 E R t I t I r n r i r * S I R A S N ~ t t

1?4 1443

G L n P R O A L A 111 V R L

ZI? 1661

[rs

t

ALA GLU x

*

t

*

irfi sifi SIR t rsn f i l l

Lrs t

PRO

.

.

V A L i r R A S P irR T r R G L U T u n A S P G L U V A L A L A t (I PnE

crs t

,

*

stfi S f f i A S P R S P G L N A S H V A L I

Lrs * L f u GLr

I:

,

Fig. 4. RNA dot-blot hybridization with al-inhibitor 3 and ct2macroglobulin. Rats were injected intramuscularly with turpentine and total RNA was isolated from the livers of animals sacrificed at the times indicated. For each time point 10,30 and 60 pg of total RNA were spotted onto nitrocellulose and hybridized with 32P-labelledaZmacroglobulin cDNA (upper panel) and El-inhibitor 3 pI3cDNA1 (lower panel)

pnt

rno

Aia f11u

Fig. 3. Amino acid sequence homology between pI3cDNA6, human and rat a2-macroglobulin. The deduced amino acid sequence of pI3cDNA6 is numbered from the first amino acid encoded by its cDNA. (*) Identical amino acids; ( 0 )Cys residues (positions 1321,1352,1467); (9) glycosylated residue (position 1424) in human az-macroglobulin

instance Gly is converted to Glu and in the other Asp to Asn in pI3cDNA6 and pI3cDNA3, respectively (nucleotide position 460 and 680, Fig. 2). The remaining 15 substitutions are clustered in the 3'-untranslated moiety of the cDNAs. One of these nucleotide substitutions leads to the loss of the Hind111 restriction site in pI3cDNA3 (see overlining of p13cDNA6 in Fig. 2). The loss of the BglII restriction site in pI3cDNA3, referred to above, may also be due to a nucleotide substitution. In addition to these differences there is an insertion of eight nucleotides (bracketed in Fig. 2 ) in pI3cDNA3 immediately before the poly(A) tail.

Closer inspection of the sequences suggests that the similarities are not restricted to the primary structure of the polypeptides but may be extended further. For instance, a functional similarity is the existence of cysteine residues capable of forming disulphide bridges at identical positions within human a2-macroglobulin and rat pI3cDNA6 (Fig. 3 : black circles below the human a,-macroglobulin amino acid sequence, Cys-I 321/Cys-1352/Cys-1467). Furthermore one of the eight glycosylated amino acids of the human a2macroglobulin is to be found in the sequence shown in Fig. 3 (5; 1424) [15, 161. At the same position in pI3cDNA6 there is an identical amino acid which could also be glycosylated. Since the homology of pI3cDNA6 with rat a2-macroglobulin is no better than with human a,-macroglobulin we do not feel justified in assuming pI3cDNA6 to be a cDNA clone for rat a,-macroglobulin. The question remains: do our clones (pI3cDNA6 and p13cDNA3) represent an allelic variation of rat a,-macroglobulin or another gene of the macroglobulin family? Functional analysis of pI3cDNAl

Ident$cation of the clones To ascertain the identity of our cDNA clones to other known sequences, we searched the EMBL (Heidelberg) gene and protein data bank and found a 69% amino acid homology ofpI3cDNA6 with the carboxy-terminal region of the primary structure of human a,-macroglobulin [15, 591. In addition there is a 69% homology with the amino acid sequence of rat a,-macroglobulin as derived from the corresponding cDNA clones [12]. The similarities between pI3cDNA6 and the two macroglobulins points to our cDNA clones representing a rat a,-macroglobulin gene and not the gene for rat fatty acid synthase. The deduced amino acid sequences of pI3cDNA6, together with the corresponding data for human and rat a,macroglobulins, can be seen in Fig. 3.

Experimental verification of the theoretically deduced identity of the cDNA clones was initially obtained with monospecific antisera against rodent a,-macroglobulin, cL1-macroglobulin and a,-inhibitor 3. Plate lysates of the original A gtll clones were transferred to nitrocellulose filters and challenged with each of the three antisera. A strong immunoresponse was obtained only with al-inhibitor 3 (data not shown). A diagnostic feature of al-inhibitor 3 is the reduction of its serum concentration following induced inflammation [13]. It could be that this change in the protein concentration reflects an alteration at the RNA level. Therefore dot-blot hybridization and Northern analysis with poly(A)-rich RNA and pI3cDNA1 as a representative probe of the cDNA clones were performed. The results are depicted in Fig. 4: 16 h after turpentine administration, the amount of al-inhibitor 3

379

Fig. 5. Northern blot analysis of El-inhibitor 3 and a2-macroglobulin poly(A)-rich R N A . 8 pg poly(A)-rich RNA from livers of control and turpentine-treated (18 h) rats were electrophoretically separated on a 1 % agarose gel under denaturing conditions. The nitrocellulosebound RNA was hybridized with nick-translated El-inhibitor 3 pl3cDNAl or cc2-macroglobulincDNA. 4.8 kb and 2.12 kb represent the positions of 28s and 18s rodent ribosomal RNA as determined by ethidium bromide staining

Fig. 6. Hybrid-arrested translation of El-inhibitor 3 mRNA. 5 pg poly(A)-rich RNA were incubated in the absence (lane 1) or presence (lane 2) of 100 ng ul-inhibitor 3 pI3cDNA1 at 48°C for 2.5 h and subsequently translated in a cell-free system from rabbit reticulocytes as described in Materials and Methods. al-Inhibitor 3 and the heavy chain of al-macroglobulin were simultaneously immunoprecipitated from both translation mixtures by their respective monospecific antisera. The radioactively labelled and immunoprecipitated polypeptides were separated by SDS/polyacrylamide gel electrophoresis and detected by autotadiography. The molecular mass standards are shown on the left

inhibitor 3 have similar antigenic determinants or the native fatty acid synthase preparation used to raise polyvalent antiserum was contaminated with @,-inhibitor 3. The first possibility may be discarded since in Western blotting there is no cross-reaction between antisera against rat a1-inhibitor 3, a,-, a2-macroglobulin and purified rat fatty acid synthase (T. Laux, unpublished observation). However, our antiserum against rat fatty acid synthase cross-reacted with the purified al-inhibitor 3 and al-macroglobulin. There was no detectable reaction with a2-macroglobulin since this, as a positive acutephase protein, is present in the serum only in very small amounts under normal physiological conditions [7]. According to various biochemical criteria, e.g. the presence of an internal cyclic thiol ester bond directly involved in the proteinase inactivation of all macroglobulins studied, a l inhibitor 3 belongs to the a-macroglobulin family 1141. Evidence that our clones represent a member of this family comes from amino acid homology. Approximately 70% homology at the amino acid level is revealed by a comparison of pI3cDNA6 sequence with the partially known rat a2-macroglobulin sequence [4, 121. A similar degree of homology was found to the human a2-macroglobulin for which both the complete amino acid [15] and cDNA sequences [16] have been published. Furthermore it was observed that the serum concentration DISCUSSION of al-inhibitor 3 decreases to 30% of the normal after The rodent al-inhibitor 3 cDNA clones were isolated with turpentine injection [lo]. This decrease has been shown by us polyvalent antiserum against fatty acid synthase. The fact that to be caused by a reduction of similar magnitude in the they do not represent fatty acid synthase cDNA could be due amount of poly(A)-rich RNA, suggesting a pretranslational to one of two reasons: either rat fatty acid synthase and al- control of al-inhibitor 3. Whether this regulation is due to a mRNA decreases by about 80%. As a reference the amount of a2-macroglobulin mRNA was determined. a2-Macroglobulin reacts to turpentine administration not only 8 h earlier but also in the opposite fashion from al-inhibitor 3. Enhanced transcription of a2-macroglobulin mRNA has been described previously [9, 121. The Northern analyses corroborate these observations (Fig. 5). Following 18-h exposure to turpentine, al-inhibitor 3 poly(A)-rich RNA was found at a much lower concentration. For control purposes, a Northern blot of a2macroglobulin mRNA is shown. A signal is only visible after turpentine treatment. The final proof that pI3cDNA1 is in fact al-inhibitor 3 cDNA was obtained by hybrid-arrested in vitro translation (Fig. 6). Only when pI3cDNA1 was omitted from the in vitro translation reaction mix, was it possible to immunoprecipitate al-inhibitor 3 with the corresponding antiserum (Fig. 6, see lanes 1 and 2). In both the intact and the hybrid-arrested in vitro translation reaction mixes the heavy chain of a l macroglobulin could be immunoprecipitated by the cognate antiserum, showing the integrity of the in vitro translation experiment for such large molecules.

380 change in the rate of mRNA synthesis or, less likely, to a change in its stability has still to be resolved. Final confirmation that this mRNA codes for al-inhibitor 3 was obtained by hybrid-arrested in vitro translation. Prior hybridization to al-inhibitor 3 cDNA prevented the immunoprecipitation of the in vitro-synthesized product with monospecific al-inhibitor 3 antiserum (Fig. 6). In addition, the experiment also shows the mRNA hybridizing with the al-inhibitor 3 cDNA has a coding potential for a protein of approximately 200 kDa, the molecular mass of al-inhibitor 3 (Figs 5 and 6) [14]. Not to be ignored are the nucleotide substitutions between pI3cDNA6 and pI3cDNA3: 15 of 19 substitutions occur in the untranslated region of the cDNA which represents only a very small part of the reverse-transcribed mRNA. We feel justified in ascribing these differences to the 3’-flanking region of the corresponding al-inhibitor 3 gene rather than to mistakes made during reverse transcription [60]. Microheterogeneity in the nucleotide sequences in other cDNA clones, e.g. rat al-acid glycoprotein [61], another acute-phase protein, and rat aldolase [62] has been reported. In addition to the nucleotide substitutions the 3’-untranslated region of pI3cDNA3 is eight base pairs longer than that of pI3cDNA6 (Fig. 2). The translational stop codons and the polyadenylation signals are, however, at the same relative positions. Probably pI3cDNA3 is not an isoform of al-inhibitor 3 since isozymes are generally tissue-specific [63] and the cDNA library from which we isolated pI3cDNA3 and pI3cDNA6 was constructed using only rat liver mRNA. cDNA synthesis primed by an oligonucleotide close to the 5‘ end of pI3cDNA6 has provided us with further evidence for localized microheterogeneity in the al-inhibitor 3 coding region (unpublished results). This microheterogeneity may reflect an intraspecies polymorphism at the DNA level or the presence of at least two non-allelic genes. In Southern blotting of rat genomic DNA with our pI3cDNA6 probe there are apparently two hybridization signals, suggesting a minimum of two genes for al-inhibitor 3 (M. Mittag, unpublished observation). Questions regarding the organization of the rat amacroglobulin genes may be answered by the isolation of the corresponding genomic clones. Furthermore one might be able to demonstrate how a family of functionally different macroglobulin genes has arisen from a common ancestral gene. M. S. is very grateful to Prof. Dr E. Schweizer for his continued and stimulating interest in this project. This work was supported by the Deutsche Forschungsgemeinschaft and the Deutsche Akademische Austauschdienst (K.T.).

REFERENCES 1. Collins, F. S. & Weissman, S. M. (1984) Progr. Nucleic Acid Res. Mol. Biol. 31, 351 -458. 2. Lewin, B. (1980) in Gene expression, vol. 2, Eucaryotic chromosomes (2nd edn) pp. 907-961, A. Wiley Interscience Publication, New York. 3. Doolittle, R. F. (1985) Trends Biochem. Sci. 10, 233-237. 4. Gordon, A. H. (1976) Biochem. J. 159,643 - 650. 5. Koj, A,, Dubin, A,, Kasperczyk, H., Bereta, J. & Gordon, A. H. (1982) Biochem. J . 206, 545-5553, 6. Kushner, I. (1982) Ann. N. Y. Acad. Sci. 389,39-48. 7. Harpel, P. C. & Brower, M. S. (1983) Ann. N.Y. Acad. Sci. 421, 1-9. 8. Northemann, W., Andus, T., Gross, V. & Heinrich, P. C. (1983) Eur. J. Biochem. 137,251-262.

9. Hayashida, K., Okubo, H., Noguchi, M., Yoshida, H., Kangewd, H., Matsuo, H. & Sakaki, Y. (1985) J. Biol. Chem. 260,1422414229. 10. Gauthier, F. & Ohlsson, K. (1978) Hoppe-Seyler’s Z . Physiol. Chem. 359,987 - 992. 11. Schaeufele, J. T. & Koo, P. H. (1982) Biochem. Biophys. Res. Commun. 108,l-7. 12. Northemann, W., Heisig, M., Kunz, D. & Heinrich, P. C. (1985) J . Biol. Chem. 260,6200 - 6205. 13. Esnard, F. & Gauthier, F. (1980) Biochim. Biophys. Acta 614, 553 - 563. 14. Esnard, F., Gutman, N., EIMoujahed, A. E. & Gauthier, F. (1985) FEBS Lett. 182, 125- 129. 15. Sottrup-Jensen, L., Stepanik, T. M., Kristensen, T., Wierzbicki, D. M., Jones, C. M., Lonblad, P. B., Magnusson, S. & Petersen, T. E. (1984) J. Biol. Chem. 259,8318-8327. 16. Kan, C. C., Solomon, E., Belt, K. T., Chain, A. C., Hiorns, L. R. & Fey, G. (1985) Proc. Natl Acad. Sci. USA 82,2282 - 2286. 17. Bolivar, F., Rodriguez, R. L., Betlach, M. & Boyer, H. W. (1977) Gene 2,75 - 93. 18. Messing, J., Gronenborn, B., Miiller-Hill, B. & Hofschneider, P. H. (1977) Proc. Natl Acad. Sci. USA 74, 3642-3646. 19. Young, R. A. & Davis, R. W. (1983) Proc. Nut1 Acad. Sci. USA 80,1194-1198. 20. Birnboim, H. C. & Doly, J. (1979) Nucleic Acid Res. 7, 15131523. 21. Huynh, T. V., Young, R. A. & Davis, R. W. (1985) in DNA cloning, vol. 1 (Glover, D. M., ed.) pp. 49-78, IRL Press, Oxford, Washington DC. 22. Messing, J. &Vieira, J. (1982) Gene 19,269-276. 23. O’Farrell, P. H., Kutter, E. & Nakanishi, M. (1980) Mol. Gen. Genet. 179,421-435. 24. Maniatis, T., Fritsch, E. & Sambrook, J. (1982) Molecular cloning. A laboratory manual, pp. 156- 172, Cold Spring Harbor Laboratory Press, New York. 25. Tautz, D. & Renz, M. (1983) Anal. Biochem. 109, 123-129. 26. Vogelstein, B. & Gillespie, D. (1979) Proc. Nut1 Acad. Sci. USA 76,615-619. 27. Schweizer, M., Case, M. E., Dykstra, C. C., Giles, N. H. & Kushner, S. R. (1981) Proc. Natl Acad. Sci. USA 78, 50865090. 28. Feinberg, H. P. & Vogelstein, B. (1983) Anal. Biochem. 132, 613. 29. Sanger, F. & Coulson, A. R. (1977) Proc. Natl Acad. Sci. USA 74, 5463 - 5467. 30. Messing, J., Crea, R. & Seeburg, P. H. (1981) Nucleic Acid Res. 9,309 - 321. 31. Schreier, P. H. & Cortese, R. (1979) J. Mol. Biol. 129, 169-172. 32. Messing, J. (1983) Methods Enzymol. 101, 20-89. 33. Schweizer, M., Roberts, L. M., Holtke, H. J., Takabayashi, K., Hollerer, E., Hoffmann, B., Miiller, G., Kottig, H. & Schweizer, E. (1986) Mol. Gen. Genet. 203,479-486. 34. Larson, R. & Messing, J. (1983) DNA 2, 31 -35. 35. Gregori, C., Besmond, C., Kahn, A. & Dreyfus, J.-C. (1982) Biochem. Biophys. Res. Commun. 104,369-375. 36. Chirgwin, J. M., Przybyla, A. E., MacDonald, J. & Rutter, W. J. (1979) Biochemistry 18, 5294- 5299. 37. Clemens, M. J. (1984) in Transcription and translation (Hames, B. D. & Higgins, S. J., eds) pp. 211 -230, IRL Press, Oxford, Washington DC. 38. Aviv, H. & Leder, P. (1972) Proc. Natl Acad. Sci. USA 69,1408 1412. 39. Locker, J. (1979) Anal. Biochem. 98, 358 - 367. 40. Spandidos, D. A. & Wilkie, N. M. (1984) in Transcription and translation (Hames, B. D. & Higgins, S. J. eds) pp. 1-48, IRL Press, Oxford, Washington DC. 41. Hutton, J. A. (1977) Nucleic Acids Res. 4, 3531-3555. 42. Patel, V., Schweizer, M., Dykstra, C. C., Kushner, S. R. & Giles, N. H. (1981) Proc. Natl Acad. Sci. U S A 78, 5783 - 5787. 43. Raue, N., Crkvenjakoo, R. & Boedtkes, H. (1979) Nucleic Acids Res. 6, 3559 - 3567. 44. Nepokroeff, C. M., Lakshmanan, M. R. & Porter, J. W. (1975) Methods Enzymol. 35, 37-44.

38 1 45. Laemmli, U. K. (1970) Nature (Lond.) 227, 680-685. 46. Campbell, D. H., Garvey, J. S., Cremer, N. E. & Sussdorf, D. H. (1970) in Methods in immunology (3rd edn) pp. 189-191, W. A. Benjamin? New York. 47. Ouchterlony, 0. (1962) in Progress in allergy, vol. VI, (Kallos, P. & Waksman, B. H., eds) pp. 30 - 154, Karger, Basel. 48. Greenwood, F. C., Hunter, W. M. &Glover, J . S. (1963) Biochem. J . 89, 114-123. 49. DeBlas, A. L. & Cherwinski, H. L. (1983) Anal. Biochem. 133, 214 - 21 9. 50. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl Acad. S C ~USA . 76,4350 -4354. 51. Westrom, B. R., Karlsson, B. W. & Ohlsson, K. (1983) HoppeSeyler’s 2. Physiol. Chem. 364, 315-381. 52. Patterson, B. M., Roberts, B. E. & Kuff, E. L. (1977) Proc. Natl Acad. Sci. USA 74, 4370 - 4314. 53. Pelham, R. B. & Jackson, R. J. (1976) Eur. J. Biochem. 67,241 256. 54. Andus, T., Gross, V., Tran-Thi, T. A., Schreiber, G., Nagashima, M. & Heinrich, P. C. (1983) Eur. J. Biochem. 133, 561 -571.

55. Gross, V., Tran-Thi, T. A., Vosbeck, K. & Heinrich, P. C. (1983) J . Biol. Chem. 258,4032 -4036. 56. Takabayashi, K. (1986) Doctoral thesis, Universitat ErlangenNiirnberg. 57. Biermann, G. (1986) Diploma thesis, Universitat ErlangenNiirnberg. 58. Birnstiel, M. L., Busslinger, M. & Strub, K . (1985) Cell41, 349359. 59. Sottrup-Jensen, L., Stepanik, T. M., Wierzbicki, D. M., Jones, C. M., Lonblad, P. B., Kristensen, T., Petersen, T. E. & Magnusson, S. (1 982) EMBL (Heidelberg) protein data bank. 60. Chen, M. J., Shimadd, T., Moulton, A. D., Harrison, M. & Nienhuis, A. W. (1982) Proc. Natl Acad. Sci. USA 79, 14357439. 61. Liao, Y.-C. J., Taylor, J. M., Vannice, J. L., Chauson, G. A. & Smuckler, E. A. (1985) Mol. Cell Biol. 5, 3634-3639. 62. Joh, K., Mukai, T., Yatsuki, H.& Hori, K . (1985) Gene 39, 1724. 63. Schibler, U., Hagenbiichle, O., Wellauer, P. K. & Pittet, A. C. (1983) Cell 33, 501 -508.